Supplementation of pig starter diets with carbohydrate-degrading
enzymes - stability, activity and mode of action

Johan Inborr

University of Helsinki
Department of Animal Science

FIN-00014 University of Helsinki, Finland

Academic dissertation
To he presented, with the permission of the Faculty of
Agriculture and Forestry of the University ofHelsinki,

for public criticism in the Small Hall, Fabianinkatu 33, Helsinki,
on April 14th, 1994, at 10 am.

https://www.c-info.fi/en/info/?token=esU0A6tLTOwmnolA.h1ld-EZcaFXorfX-JGNAug.ouNV-clf8r-Fj1rz7iV8T__B1kg7zxJdGiLXmeU7rFxVIpBK_OACX8EYZ81Zk4VTZZOS9-Vv39NssMG9H3bvVBh9y7pErPFGc8Yi1nFQULFF4bdidM3w8NLL9KSBHFA2r4BNGCgHZKbknXUUe3vEAo8ujHMQ5E0_2h2y5K94FidQtAIsizpKaPKOdbCZ7XDKv_LUkcN0WRwx9r82P9oHqwcXgU-Fmus2qxiwG4d3dC9pPjHZzNdsZ8l1ZYbbTGHC4y2Z




3

PREFACE

This work is a result ofan idea born within the CultorLtd. Group almost ten years ago. This
idea developed into a concept, which was later introduced to the international scene, where
it found a wide acceptance and gained momentum. Over the years, a large number animal
nutritionists and biochemists have contributed to the development of this concept, simply
called ’pig feed enzymes’, to give it the credibility it enjoys today. To mention all these
persons would be an impossible task, but in the following I have tried to list the ones that
significantly contributed to this thesis.

First ofall I want to thankProfessorLiisa Syijälä-Qvist, Department ofAnimal Science,
University ofHelsinki, for her positive attitude towards this work and Associate Professor
Matti Näsi, at the same department for valuable critisism of the manuscripts and his
encouragement.

I also wish to express my sincere gratitude to:
Dr Martin Schmitz and Dr Frank Ahrens at the IS Forschungsgesellschaft GmbH&Co,

Wahlstedt, Germany for invaluable assistance in conducting the slaughter experiment and
with the statistical analysis of the data.

Dr Jan van der Meulen at the DLO Institute for Livestock Feeding and Nutrition,
Lelystad, The Netherlands for supervising the cannulation experiment and analysing the
data, and for valuable critisism of the manuscripts.

Dr Kirsten Jakobsen, Dr Knud Erik Bach Knudsen, Dr Bent Borg Jensen and Ms Mette
Skou Jensen at the National Institute of Animal Science (NIAS), Foulum, Denmark for
supervising the work at the institute and for the practical arrangements and the day-to-day
care of the pigs, data collection and sample analysis.

Mr Walter Michie, Scottish Agricultural College in Aberdeen, Craibstone Poultry
Research Unit, for arranging and supervising the broiler experiment.

Dr Mike Bedford at Finnfeeds International Ltd., Marlborough, United Kingdom, for
his professional assistance with the statistical analysis of the data and valuable critisism of
the manuscripts, and all my friends at Finnfeeds InternationalLtd. for support and encour-
agement during the course of the work.

Mr Jari Puhakka and Mrs Anne Grönlund at the Cultor Ltd. Technology Centre for
arranging and supervising the enzyme activity analyses and for valuable assistance in
preparing the manuscripts.

Mr Tom von Weymam, Cultor Ltd., Helsinki, Finland, and Mr Richard Cooper,
Finnfeeds International Ltd., for allowing the entire project to be carried out within the
scope of CultorLtd. Animal Nutrition Group’s and Finnfeeds InternationalLtd.’s Research
and Development programmes.

And last but certainly not least I am forever greatful to the four most amiable ladies in
the world, my wife Maggi and daughters Sara, Jennie and Ina for their never failing support
and encouragement, and for being a source of inspiration during the course of this work.

The experiments included in this thesis were conducted during 1988-1993 as part of the
research programme ofFinnfeeds International Ltd., Marlborough, England.

Hamburg, 22nd December 1993



4

LIST OF PUBLICATIONS

This thesis is based on the following five papers which will be referred to by their Roman
numerals:

I Inborr, J. & Bedford, M.R. 1994. Stability offeed enzymes to steam pelleting during
feed processing. Animal Feed Science and Technology (in press).

II Inborr, J. & Grönlund, A. 1992. Stability of feed enzymes in physiological condi-
tions assayed by in vitro methods. Agricultural Science in Finland 2: 125-132.

11l Inborr, J., VAN der Meulen, J. & Puhakka, J. 1993. Nutritional implications of
feeding enzyme-treated wheat bran to pigs. 1. Recovery of added enzyme activities in
the stomach and terminal ileum. Submitted to British Journal of Nutrition (Publisher:
Cambridge University Press).

IV Inborr, J., Schmitz, M. & Ahrens, F. 1993. Effect ofadding fibre and starch-degrad-
ing enzymes to a barley/wheat based diet on performance and nutrient digestibility in
different segments of the small intestine of early weaned pigs. Animal Feed Science
and Technology 44: 113-127.

V INBORR, J. 1994. Gastrointestinal parameters influencing performance of pigs fed
enzyme-supplemented barley-based feeds. Submitted to British Journal of Nutrition
(Publisher: Cambridge University Press).



5

CONTENTS

PREFACE 3

LIST OF PUBLICATIONS 4

ABSTRACT 7

INTRODUCTION 8
I Dietary fibre in pig diets 8
2 Opportunities for supplementary enzymes in pig feeds 9

OBJECTIVES 10

MATERIALS AND METHODS 10
I Enzyme stability and recovery 10
2 Pig performance and nutrient digestibility II
3 Enzyme mode of action in pigs II

RESULTS AND DISCUSSION 12
I Stability and activity ofsupplementary enzymes 12
2 Site and mode of action of carbohydrate-degradingenzymes 13

CONCLUSIONS 17

REFERENCES 18

SELOSTUS 21

SAMMANDRAG 22

ZUSAMMENFASSUNG 23

PUBLICATIONS I-V





Supplementation of pig starter diets with carbohydrate-degrading
enzymes - stability, activity and mode of action

Johan Inborr

Inborr, J. 1994. Supplementation of pig starter diets with carbohydrate-degrading
enzymes - stability, activity and mode of action. Agricultural Science in Finland 3:
Supplement No. 2. 23 p. Academic dissertation. (Department of Animal Science, P.O.
Box 28, FIN-00014 University of Helsinki, Finland.)

A total of five experiments were conducted to investigate the stability of feed enzymes
to steam pelleting and the proteolytic conditions in the gastrointestinal (GI) tract of pigs
and poultry, and to try and elucidate the mechanisms behind the improved performance
of pigs fed enzyme-supplemented barley/wheat-based diets.

The results of the pelleting stability experiment showed that the commercial feed
enzyme employed maintained most of its activity in conditioning temperatures up to
85°C. Furthermore, it became evident that measuring enzyme recovery in pelleted feeds
by in-vitro assay methods underestimated the actual activity. For this purpose in-vivo
models such as that based on gut viscosity measurements in broiler chickens gives a
more accurate estimate. Gut viscosity also correlated highly with live weight gain
(r2 =0.624) and feed utilisation (r2 =0.616) of broiler chickens. The in-vitro incubations
using conditions similar to those of the GI tract showed that enzymes are not readily
denatured and inactivated in such conditions and indicated that wheat and wheat gluten,
and possibly similar feed ingredients, may help to maintain the activity longer either due
to their buffering capacity or by providing substrates for the enzymes. This was
supported by the results of the in-vivo measurements. In the stomach of pigs, 10-20per
cent of the xy lanase and (i-glucanase activities added to the diets could still be recovered
4 hours after feeding. In the ileum, proportionally more added enzyme activities were
recovered between 4 and 6 than 0 and 2 hours after feeding. In broiler chickens fed an
enzyme-supplemented barley-based diet, [Tglucanasc was fully recovered in the proxi-
mal part of the small intestine, giving further proof of the stability of the enzymes
employed to the conditions of the GI tract.

When a mixture of fibre- and starch-degrading enzymes were added to a diet based
on wheat and barley, (3-glucan, starch and dry matter digestibilities were significantly
(P<0.05) improved in the last quarter of the small intestine of early-weaned pigs but did
not translate into improved growth or feed utilisation. However, the results showed that
enzyme supplementation increased the rate of digestion and more nutrients were
absorbed higher up in the small intestine. In a similarly designed experiment, adding a
single (3-glucanase preparation todiets based on either a low- or a high-(5-glucan barley,
improved live weight gain (P=0.074) and feed utilisation (P=0.058) of early-weaned
pigs over a three-week experimental period. Although there was no significant improve-
ment in nutrient digestibility, enzyme supplementation reduced digesta viscosity
(P<0.03) and the concentration of digestive enzymes (P<0.08) in the three proximal
quarters of the small intestine. Thus it appears that conditions for a more efficient
digestion were brought about due to the reduction of digesta viscosity, enabling less
production of pancreatic enzymes without affecting digestibility. Further analysis of
the digesta samples showed a significant (P=0.044) reduction in the concentration of
VFAs in the distal small intestine and hind gut of thepigs fed the enzyme-supplemented
diets. This further indicates that enzyme supplementation leads to fewer nutrients
escaping digestion and absorption in the small intestine such that less readily fer-
mentable material was available for microbial growth.

7

Agric. Sei. Fint. Suppl. No. 2 (1994)



It is concluded that appropriately selected fibre- and starch-degrading enzymes added
to pig starter diets based on wheat and barley exert their activity in the GI tract by
reducing digesta viscosity and thereby increasing the efficiency and rate of digestion.
Less digestive enzymes are needed and a greaterportion of the nutrients are absorbed in
the upper part of the small intestine. This leads to reduced microbial activity in the distal
parts of the digestive tract, resulting in less fermentation losses and digestive upsets. All
these effects contribute to overall improved performance of the pigs.

Key words: barley, wheat, (3-glucan, viscosity, (l-glucanase, xylanase, amylase, diges-
tion, microbial fermentation

INTRODUCTION

1 Dietary fibre in pig diets

By definition, pigs like other monogastric animals
are not capable of digesting dietary fibre (DF) by
means of their own digestive enzymes (Trowell et
al. 1976). Since the concentration of DF in barley
and wheat can be as high as 14 and 25% respec-
tively (ÅMAN 1987), a significant proportion of
diets based on these cereal grains is usually poorly
utilised by the pig.

Dietary fibre is composed of a very heterogenous
mixture of substances, mainly associated with plant
cell walls, which may be defined as non-starch
polysaccharides (NSP) and lignin. Due to its com-
plex physical and chemical characteristics, DF can
influence many processes and reactions in the di-
gestive system of the pig, thus influencing the util-
isation of feed. Some of these effects were exten-
sively reviewed by Low (1985).

Apart from lignin, which is an aromatic polymer
of phenolic alcohols, DF consists of a wide range of
polymers such as cellulose, hemicellulose, (3-glu-
cans, pentosans (arabinoxylans), pectic substances
and oligosaccharides (raffinose and stachyose). It
has also been proposed that certain types of proc-
essed starch that escape digestion by the host amy-
lases, so called ’resistant starch’, should be in-
cluded in DF (Englyst and Cummings 1985).
Nine monosaccharides dominate quantitatively as
the building blocks of fibre polysaccharides. These

are the pentoses (xylose and arabinose), the hexoses
(glucose, mannose and galactose), the 6-deoxy-
hexoses (rhamnose and fucose), and the hexauronic
acids (galacturonic and glucuronic acid) (Graham
1988).

The physical properties of DF are very much
dependent upon their source and composition. Nev-
ertheless, the ability to hold water seem to be a
feature common to most DF sources. Flowever, the
extent of the water holding capacity varies signifi-
cantly between sources and depends on the source
of fibre, maturity of the plant, processing, particle
size, pH and electrolyte composition. Many of
these substances also tend to increase the viscosity
of feeds when soaked (Cherbut et al. 1990) and
gut contents (Meyer et al. 1986, Potkins et al.
1991, Roberts et al. 1990), which can influence
the rate of gastric emptying (Meyer et al. 1986,
Rainbird and Low 1986), feed transit time (Cher-
but et al. 1990) and digestive secretions (IKEGAMI
et al. 1990). In particular, nitrogen secretion seems
to increase with increased amounts ofdietary fibre,
an observation made with both insoluble and sol-
uble types of dietary fibre (Zebrowska and Low
1987, Langlois et al. 1987). The apparent digest-
ibility of nitrogen may thus be a function not only
of the inherent digestibility of the dietary protein,
but also of the type and amount ofdietary fibre with
which it is incorporated in the diet.

Nutrient absorption in the small intestine tend to

8

Agric. Sei. Fin!. Suppl. No. 2 (1994)



be impaired at higher levels of DF. Just et al.
(1983) measured decreased apparent ileal digest-
ibilities of nitrogen and fat as the crude fibre con-
tent increased from 33 to 161 g/kg in diets fed to
growing pigs. The most striking effect, however,
was seen in the digestibility of energy.

Once DF leaves the small intestine its role as an
agent which influence the digestion and absorption
of other nutrients changes to that of being a nutrient
in its own right, as a result of microbial fermenta-
tion. Again the type and amount ofdietary fibre will
influence the extent of this fermentation, which
appears to be most intensive in the caecum and
colon (Bach Knudsen et al. 1991). Microbial fer-
mentationof fibre also occurs in the small intestine
and the magnitude of this ’digestion’ increases with
age. Graham et al. (1988a) reported the digest-
ibility of fibre polysaccharides in 20, 40 and 80 kg
pigs to be 12, 35 and 57%, respectively, at the end
of ileum. The end products of the fermentation of
dietary fibre are the volatile fatty acids (VFA).
These are absorbed through the gut wall into the
portal blood and contribute to the total energy yield
from the feed (Argenzio 1982). However, the
magnitude of this contribution in relation to the
total energy supply is a matter ofdebate(Goodlad
and Mathers 1991) but has been estimated to be
approximately 30 per cent of the energy require-
ment for maintenence in growing pigs (Rerat et al.
1987, Yen et al. 1991). Some chemical and physi-

cal properties of DF in relation to nutrition are
presented in Table I.

2 Opportunities for supplementary
enzymes in pig feeds

Based on the above it would seem that adding
fibre-degrading enzymes to diets with high concen-
trationof DF could alleviate some of the anti-nutri-
tive effects. Such enzymes are today available at
relatively low cost and their use would, therefore,
appear more attractive than in the past. In fact,
enzyme supplementation of pig feeds has attracted
increasing interest in recent years although the re-
sults have been fairly inconsistent compared to
those obtained with broiler chickens (CHESSON

Table 1. The fibre properties implicated, and some of the
possible mechanisms involved in the modification at varoius
stages of nutrient assimilation in pigs (Graham 1988).

Stage of Fiber properties Possible
assimilation implicated mechanisms

involved

Intake Bulk Mechanical
Water-binding capacity Viscosity
Energy dilution Taste

Hormonal

Passage rate Bulk Mechanical
Water-binding capacity Viscosity

Hormonal
Enzyme hydrolysis Water-binding capacity Viscosity
and absorption Cation-exchange capacity Adsorption

Architecture Hormonal
Hydrophobicity

Bacterial Cell wall composition Potential
fermentation and structure degradation

Transit time

1987, Dierick 1989). Hence in many experiments
supplementation of barley-based diets with p-glu-
canases (cellulases) has improved pig performance
(Thomke et al. 1980, Newman et al. 1983, Mark-
STRÖM et al. 1985), increased ileal (Graham et al.
1988b, Bedford et al. 1992) and faecal (Graham

et al. 1988b, Inborr and Graham 1991) nutrient
digestibility and reduced the incidence of digestive
upsets (Inborr and Ogle 1988, Böhme 1990).
However, there are also reports of non-significant
responses to p-glucanase supplementation of bar-
ley-based diets (Newman et al. 1980, Graham et
al. 1986). Enzyme treatment of other types of diets
or raw materials such as soybean and rapeseed meal
(Näsi 1991), rye (Buraczewska 1988, Thacker
et al. 1992) and wheat (McClean et al. 1993) has
sometimes met with success. However, Tangend-
JAJA et al. (1988) failed to improve the nutritive
value of rice bran to pigs by cellulase treatment. It
should be noted that more consistent positive re-
ponses to enzyme supplementation have been ob-
tained with young pigs in the post-weaning period
up to 25 kg liveweight than with growing/finishing
pigs.

9

Agric. Sei. Finl. Suppl. No. 2 (1994)



One explanation to the inconsistent responses to
enzyme supplementation of pig feeds may be found
in the stability and activity of the enzymes em-
ployed. If for any reason the enzymes employed
cannot exert their activity at the expected site of
action, no physiological response can be expected.
Being proteins, possessing very specific features
both in terms of their structure and mode ofaction,
enzymes are susceptible to, for example, changes in
pH, temperature and metal ion concentration,
which can result in partial or total inactivation
(Godfrey and Reichelt 1983). Heat-treatment of
feeds and employing enzymes with inappropriate
pH and temperature optima may, therefore, have
contributed to the failure of many experiments in
the past. Consequently, enzymes with proven sta-
bility and suitable pH and temperature optima need
to be identified before initiating elaborate experi-
ments.

In summary, based on the results cited above,
there seems to be an opportunity for improving the

nutritive value ofraw materials with relatively high
DF content by fibre-degrading enzymes. However,
based on the information generated to date it is not
possible to predict the magnitude of the response or
explain the mode of action of supplementary en-
zymes in pig feeds. In broiler chickens, a reduction
of digesta viscosity due to added enzymes has been
shown to highly correlate with improved bird per-
formance regardless of DF source (Bedford and
Classen 1992, Inborr et al. 1993). This was not
the case in early-weaned pigs fed hulless barley
(Inborr et al. 1991). Digesta viscosity, therefore,
does not seem to interfere with digestion to the
same extent in the pig as it does in the chick. The
mechanisms behind the improved performance and
nutrient digestibility that have been reported in re-
sponse to enzyme supplemenation of diets for
early-weaned and growing pigs are likely to relate
to factors such as feed passage rate, pancreatic se-
cretion, plant cell wall degradation and changes in
the activity of the microflora.

OBJECTIVES

The objectives of the investigations reported in this
thesis were:

1. To investigate the stability of feed enzymes to
steam pelleting;

2. To investigate the stability and activity of feed
enzymes in physiological conditions by in-vitro
and in-vivo methods;

3. To study the effect of adding fibre- and starch-
degrading enzymes to diets based on barley and

wheat on nutrient digestibility and growth per-
formance of early-weaned pigs;

4. To try and explain the mechanisms involved in
the enzyme-induced growth response and im-
proved nutrient digestibility of early-weaned
pigs fed diets based on barley.

Reference will be made to similar studies with
broiler chickens for interspecies comparison.

MATERIALS AND METHODS

1 Enzyme stability and recovery leting both in-vitro (I, II) and in-vivo (I and III)
methods were employed.

For the purpose of estimating the stability of en
zymes in various conditions and through steam pel

In experiment I, a barley-based broiler starter
feed with and without added enzyme (Avizyme

10

Agric. Sei. Fin!. Suppt. No. 2 (1994)



(R) . .SX ) was pelleted after conditioning at 75, 85 and
95°C for either 30 seconds or 15 minutes. (3-gluca-
nase activity of the feeds was determined by a
viscosimetric method to assess the degree ofinacti-
vation due to pelleting. The feeds were fed to day-
old broiler chickens for 19 days to study the relative
activity of the enzyme based on bird performance.
On day 19, four birds per treatment were sacrificed
and the viscosity and (3-glucanase activity of the
digesta of the upper halfof the small intestine meas-
ured. The results from the in-vitro ((3-glucanase in
feed) and in-vivo (digesta viscosity and (3-glucanase
activity) enzyme activity measurements were then
used for assessing the extent of enzyme survival
and compared with bird performance.

In experiment 11, a crude xylanase preparation
and a commercially available feed enzyme product
were subjected to conditions similar to those in the
stomach and small intestine of pigs. The treatments
comprised of incubation in low pH (2.5) with pep-
sin for 30 minutes with and without wheat gluten
followed by incubation in neutral pH with pan-
creatin for 30 minutes. Recoveries of xylanase and
(3-xylosidase of the crude xylanase preparation and
xylanase and [3-glucanase of the commercial pro-
duct were measured after each incubation by en-
zyme activity analysis methods based on the release
of reducing sugars and di-nitro salicylic acid (DNS).

Experiment 111 was designed to measure enzyme
recovery and rate and magnitude of enzyme inacti-
vation in the stomach and ileum of growing pigs.
Five pigs of approximately 31 kg live weight were
fitted with cannulas in the stomach and terminal
ileum and fed one of five diets over five 14-day
periods in a 5 x 5 Latin Square design. The experi-
mental diets were based on wheat bran (40%) that
had been either incubated (3.5 hours at 39°C) or
supplemented with one oftwo crude enzyme prepa-
rations (one cellulase and one xylanase) prior to
feeding. Wheat bran incubated without enzyme
served as a control. Other main components of the
experimental diets were maize starch (49.3%) and
casein (8.2%). Samples of the stomach contents
were taken immediately after feeding and then 2
and 4 hour after feeding, whereas ileal chyme was
collected between 0 and 2, 2 and 4, and 4 and 6
hours after feeding. Xylanase and (3-glucanase ac-

tivities of the samples were measured by DNS
methods using standard conditions (pH, tempera-
ture, incubation time) or modified to actual sample
pH and physiological temperature (39°C) with a
prolonged incubation lime.

2 Pig performance and nutrient
digestibility

A total of 96 pigs, weanedbetween 21 and 24 days
ofage, were divided into twelve groups ofeight and
housed in flat-deck pens for three weeks (experi-
ment IV). A basal diet based on barley (35%),
wheat (35%) and soybean meal (22%) was used as
a control or supplemented with either of two en-
zyme premixes containing (3-glucanase, xylanase
and a-amylase. Titanium oxide was added as an
indigestible marker. The diets were fed ad libitum
for 21 days and liveweight gain and feed consump-
tion recorded on a weekly basis. At the end of the
experiment two pigs per pen were sacrificed, the
small intestine removed and divided into four seg-
ments of equal length. Dry matter, protein, starch
and fibre digestibilities were estimated in the three
distal segments.

3 Enzyme mode of action in pigs

Forty pigs weaned at three weeks of age were
housed individually in metabolism cages in blocks
of eight animals (experiment V). There were two
pigs per dietary treatment in each block with all
pigs of the same replicate selected from the same
litter. There were two basal diets, one based on a
hulled (Arra) and one based on a hulless (Condor)
barley cultivar. Each diet was supplemented with
and without a [3-glucanase preparation (Multifect
CS , Genencor International Ltd., Helsinki, Fin-
land) and fed ad libitum for 21 days. Liveweights
and feed consumption were recorded weekly. On
day 21 the pigs were slaughtered and the digestive
tract removed and divided into the following seg-
ments; stomach (Sto), four sections ofequal length
of the small intestine (SI 1-4), caecum (Cae), as-
cending (Col) and descending (Co 2) colon and

Agric. Sei. Finl. Suppl. No. 2 (1994)



rectum (Rec). The contents of each segment was
collected quantitatively. Viscosity of the superna-
tant was measured directly after centrifugation of
the fresh digesta samples. After freeze drying, sam-
ples were analysed for p-glucan and nutrient digest-

ibility based on chromic oxide, which was used as
an indigestible marker. In addition, the concentra-
tion of volatile fatty acids (VFA) was measured in
the distal small intestine, caecum, colon and rec-
tum.

RESULTS AND DISCUSSION

1 Stability and activity of supplementary
enzymes

Based on the p-glucanase assay of the pelleted feed
samples, enzyme activity decreased with increasing
conditioning times and temperatures (Figure 1, I).
A similar trend was observed in the digesta viscos-
ity values, although to a lesser degree. These results
would indicate that the assay method employed to
measure the P-glucanase activity in feed underesti-
mated the actual activity. This is probably due to
the enzyme binding so strongly to its substrate that
it cannot be fully extracted during the assay. Many
cell wall degrading enzymes (cellulases, endohy-
drolases and cellobiohydrolases) have been found
to possess a cellulose binding domain (CBD) with
which they bind to cellulose (Nieves et al. 1991,
Poole et al. 1991). Despite being immobilised,

they are still able to hydrolyse fibre polysaccha-
rides, since they possess an arm-like movable struc-
ture which allows the active site to continue in the
formation of enzyme-substrate complexes (Din et
al. 1991). The attraction between the CBD and
crystalline cellulose seems to be mediated through
a hydrophobic interaction. The rapidity of this in-
teraction will, therefore, be dependent upon the
presence ofwater and temperature. This may partly
explain the apparently low enzyme activity recov-
eries in steam heat-treated feeds assayed by meth-
ods including an extraction step. Estimating en-
zyme activity in pelleted feeds using methods with-
out an extraction step or in-vivo would, therefore,
seem more appropriate. Based on the results of the
digesta viscosity measurements (I) it would appear
that the enzyme employed had retained most of its
activity during conditioning in temperatures up to
85°C. At the longer conditioning times enzyme
recoveries were reduced.

Susceptibility to inactivation (denaturation) of
enzymes in low pH and proteolytic conditions ap-
peared to differ between enzyme source and
method of stabilisation (II). In general, the crude
enzymes seemed to possess an inherent stability to
conditions resembling those in the GI tract of the
pig. With the addition of wheat gluten to the incu-
bations the magnitude ofinactivation was generally
reduced. When a commercial feed enzyme product
was employed, which had been stabilised by spray-
ing liquid enzymes onto a wheat-based carrier ma-
terial and then dried in low temperature, neither
P-glucanase nor xylanase activities were signifi-
cantly reduced due to the treatments. This is prob-
ably due to the buffering capacity of the carrier
material, which also provided substrates to the en-

Fig. 1. p-glucanase recovery (% of corresponding control)
in feed samples after pelleting at different conditioning
temperatures and times estimated either by in-feed enzyme
analysis (□) or digesta viscosity reduction in broiler chick-
ens (■).

12

Agric. Sd. Fin!. Suppl. No. 2 (1994)



zymes. De Cordt et al. (1992) found that both
polyols and carbohydrates are powerful stabilisers
for both dissolved and immobilised B. licheni-
formis a-amylase. When the stabilisedfeedenzyme
product was fed to broiler chickens, P-glucanase
was fully recovered in digesta samples obtained in
the proximal small intestine (Figure 2,1). In sam-
ples obtainedfrom the stomach ofpigs fed enzyme-
supplemented diets, recovery rates decreased with
timeand were between 10and 20% four hours after
feeding (Figure 3, III). This was probably due to
inactivation as pH of the gastric contents decreased
and would suggest that the physiological conditions

of the upper gastrointestinal tract may cause a cer-
tain degree of inactivation of supplementary en-
zymes. Interspecies differences in this respect may
be due to different feed passage rates and pH in the
gastric regions. In samples obtained at the terminal
ileum ofpigs the proportion of added xylanase and
p-glucanase activities increased during the period
between 2 and 6 hours after feeding compared to
that between 0 and 2 hours after feeding (Figure 3,
III). These results give evidence of partial survival
of the added enzymes through the GI tract to the
end of the small intestine. Consequently, provided
the enzymes to be employed are selected according
to the target substrates, are added in sufficient
amounts and are active in the conditions of the GI
tract (appropriate pH and temperature optima) a
physiological response can be expected.

2 Site and mode of action of carbohydrate-
degrading enzymes

Employing techniques based on sampling of ileal
digesta by cannulation or total collection of faeces
has usually failed to demonstrate any significant
effects (Graham et al. 1986, Thacker et al. 1992)
or has only given a small response (GRAHAM et al.
1988a, b, Inborr and Graham 1991) in nutrient
digestibility ofpigs, although the effects on growth
and feed conversion in some experiments have

Fig. 2. Correlation between p-glucanase activity in feed and
digesta samples of broiler chickens (r 2=0.948).

Fig. 3. Endogenous (■) and added (E3) xylanase and (f-glu-
canase activities of feed and digesta samples of pigs fed
enzyme-supplemented diets. Digesta samples were collected

from the stomach 0 (StmO), 2 (Stm2) and 4 (Stm4) hours and
terminal ileum between 0-2 (IleO), 2-4 (Ile2) and 4-6 (Ile4)
hours after feeding.

13

Agric. Sei. Finl. Suppl. No. 2 (1994)



been quite dramatic following enzyme supplemen-
tation(Böhme 1990, Newman etal. 1992).Conse-
quently, it appears that these techniques are not
sufficiently sensitive to describe the action of the
added enzymes and that the point of digesta sam-
pling need to be more anterior to aviod treatment
effects being masked by e.g. microbial fermenta-
tion. Hence, collecting samples from the entire
small intestine or even from the whole gastrointes-
tinal tract by slaughter technique may prove a more
successful way of investigating the effects of sup-
plementary enzymes. This technique was success-
fully employed by INBORR et. al. (1991) and, there-
fore, selected for experiments IV and V. The results
of the two latter experiments will be used to try and
explain how and where the added enzymes exert
their activity in the pig. The results of these two
experiments will be discussed in the following.

In both experiments, supplementation of the
diets with a mixture of enzymes containing ()-glu-
canase (IV) or a single P-glucanase preparation (V)
significantly increased the digestibility of dietary
P-glucans in the small intestine. In experiment V,
this was accompanied by a significant reduction in
digesta viscosity in the stomach and the three ante-
rior sections of the small intestine (Figure 4). In

experiment IV, microbial amylase was included in
the enzyme mixture employed, whereas in experi-
ment V, only a single enzyme preparation contain-
ing p-glucanase was used. This may explain why in
experiment IV starch digestibility increased signifi-
cantly due to enzyme supplementation (Figure 5),
but was unaffected in experiment V. Interestingly,
the digesta samples obtained from the three proxi-
mal quarters of the small intestine contained on
average 20% less amylase activity when P-gluca-
nase was added, without reducing starch digest-
ibility (Table 2, V). This observation wouldsuggest
that the conditions for starch digestion were more
optimal in the presence of the P-glucanase, prob-
ably due to reduced viscosity in the lumen. These
observations are in contrast with those reported
from studies with poultry (Almirall et al. 1993).
These workers found that addition of a P-glucanase
to diets based on barley significantly increased
amylase activity in the small intestine of broiler
chickens and one-year-old cocks. It is possible that
there are species differences with regard to the
feed-back mechanisms regulating pancratic exo-
crine secretion. It has been shown that increased
viscosity caused by rye non-starch polysaccharides
reduce the rate ofnutrient diffusion and absorption

Fig. 4. Viscosity of digesta of pigs fed a hulled barley (var,
Arra) with (A-A) and without (O-O) and a hulless barley
(var. Condor) with (□ -□) and without (O -O) p-glucanase
supplementation. A significant effect of the enzyme was
observed in the stomach (P=0.029), first (SI1; P=0.021),
second (SI2; P=0.004) and third (SI3; P=0.007) quarter of the
small intestine.

Fig. 5. Effect of supplementation of a barley/wheat-based
diet without (■) or with enzyme C (E 3) orenzyme M (□) on
starch digestibility in the three last quarters of the small
intestine of early-weaned pigs. Enzyme mixtures C and M
contained amylase and xylanase from the same sources but
differed by containing p-glucanase from two separate T.
longibrachiatum sources.

14

Agric. Sei. Fin!. Suppl. No. 2 (1994)



Table 2. Mean digestive enzyme activities (mU/g digesta dry matter) of the three proximal quarters of the
small intestine of pigs fed either Arra or Condor barley without (-) or with ( + ) added p-glucanase.

Arra- Arra+ Condor- Condor+ P-value Pooled
(enzyme) variance

Trypsin 10.510.6 12.510.2 0.0346.38
Chymotrypsin 0.100.15 0.150.09 0.0590.074
Lipase 76.667.5 91.268.3 0.07776.6
Amylase 997 787 816 626 0.038 576

in-vitro (Fengler and Marquardt 1988) and that
reduction of intestinal viscosity improves the nutri-
ent digestibility and subsequent performance of
broiler chickens (Pettersson 1988,Bedford and
Classen 1992, Almirall et al. 1993). In pigs,
feeding viscous polysaccharides has been shown to
increase digesta viscosity (Roberts et al. 1990,
POTKINS et al. 1991) and pancreatic exocrine sec-
tretion (Cherbut et al. 1990). P-glucanase supple-
mentation of a pig starter diet based on a hulless
barley significantly improved the apparent protein
digestibility over the entire small intestine and
tended to reduce digesta viscosity (Bedford et al.
1992). These workers assumed that the enzyme
increased the rate of protein digestion and absorp-
tion due to the reduced viscosity and by degrading
cell wall polysaccharides interfering with the diges-
tion. However, based on the results of experiment
V, a reduced output ofdigestive enzymes may have
contributed to the increased apparent protein di-
gestibility. In pigs fitted with catheters in the pan-
creatic duct and fed the same diets as in experiment
V, it was found that in the presence of P-glucanase
the pancreatic exocrine secretion of protein was
reduced (Skou Jensen, unpublished). If reduction
of the luminal viscosity allows nutrients to be di-
gested and absorbed at an equal rate with signifi-
cantly less digestive enzyme production, this
should lead to reduced endogenous losses and re-
duced energy requirements for enzyme synthesis
and secretion and, consequently, an improvement
in dietary energy utilisation. Pierzynowski (1991)
estimated the daily pancreatic exocrine protein se-
cretion in pigs up to 20 kg liveweight between 2 and
6 g, whereas in pigs of varying age this secretion
has been estimated to between 9 and 36 g (SOUF-

FRANT 1991). If intestinal amylases are added to
this a 20% reduction in enzyme activity may repre-
sent approximately 2 g protein or 5 per cent of the
daily protein retention of pigs during the post-
weaning period. The increased liveweight gain and
improved feed utilisationobserved in experiment V
would indicate that dietary energy and protein
were more efficiently utilised in the presence of the
P-glucanase.

Increasing the rate of digestion and nutrient ab-
sorption in the small intestine effectively means
less available substrates for the intestinal mi-
croflora. In experiments withbroiler chickens it has
been shown that enzyme supplementation of bar-
ley-based diets reduces the microbial count in the
small intestine (Salih et al. 1991). This is thought
to be a consequence of an increased feed passage
rate and reduced concentration of nutrients in the
lumen of the posterior intestine. In studies with
early-weaned pigs, application ofmixtures of fibre-

and starch-degrading enzymes has lead to signifi-
cant reductions in the frequency and severity of
diarrhoea (Inborr and Ogle 1988,Böhme 1990),
indicating reduced microbial activity in the hind
gut. These results are supported by the ones ob-
tained in the two slaughter experiments (IV, V). In
experiment IV, the application of a mixture con-
taining p-glucanase, xylanase and amylase resulted
in a significantly improved starch digestibility in
the last quarter of the small intestine (Figure 5).
However, the most dramaticresponse was seen in
the second quarter (Figure 5, P<0.06), indicating a
clear shift of the digestion from the distal to the
proximal parts of the small intestine. Based on the
above one can assume that increasing the rate and
changing site ofdigestion by enzyme supplementa-

15

Agric. Sei. Finl. Suppl. No. 2 (1994)



tion in this fashion leads to a reduced microbial
activity in the pig caecum and colon, thus reducing
the incidence ofdigestive disorders such as fermen-
tative diarrhoea as described by Skadhauge
(1985). Kamphues (1987) observed increased
starch, lactic acid and VFA concentrations in the
hindgut of pigs suffering from diarrhoea after hav-
ing consumed large amounts of feed following a
24-hour period of feed deprivation. Hence increas-
ing nutrient digestibility and the rate of absorption
by enzyme addition should alleviate the effects of
such nutrient-induced digestive disorders. This as-
sumption is supported by the results from experi-
ment V, where enzyme supplementation of the bar-
ley-based diets significantly (P=0.044) decreased
the concentration of VFA and lactic acid in the
distal sections of the GI tract (Figure 6) and in-
creased the dry matter content of the chyme. Since
many microbes in the gut interfere with host diges-
tive enzymes and processes (Ratcliffe 1985), a
reduction of the microbial count will improve the
overall efficiency of digestion.

in broiler chickens fed diets based on wheatand
rye, it has been established that gut viscosity in-
creases exponentially in relation to the concentra-
tion of a water-solublehigh-molecular weight car-

bohydrate (HMC) complex (Bedford and Clas-
sen 1992). With increasing fibre solubility or
amount of soluble fibres the higher the viscosity
and the amount ofwater being bound in the digesta.
This can lead to severe deterioration oflitter quality
and a high incidence of vent pasting of broilers. In
experiment I, there was a high positive correlation
between digesta viscosity and the incidence of vent
pasting. McCracken et al. (1992) reported in-
creased dry matter content of excreta when a com-
mercial feed enzyme was added to a broiler diet
based on wheatand barley that was produced either
without or with conditioning at 85°C for 15 minutes
before pelleting. Interestingly, conditioning in-
creased digesta viscosity of the birds, indicating
solubilisation of the fibre components due to the
heat-treatment. Graham et al. (1989) also ob-
served higher concentrationof soluble (f-glucans in
the digesta of pigs fed pelleted feeds compared to
mash feeds. Adding a (3-glucanase increased the
digestibility of NSP more in the pelleted than the
mash feed. Consequently, enzymes added to cereal-
based diets appear to primarily hydrolyse the sol-
uble fibres, thus reducing digesta viscosity and the
ability ofsoluble fibres to bind water. In experiment
V, addition of the (i-glucanase significantly in-
creased the dry matter content of the intestinal
chyme along the entire GI tract. In another experi-
ment with early-weaned pigs, enzyme supplemen-
tation of a barley/wheat-based diet resulted in in-
creased faecal nutrient digestibility and dry matter
content and reduced manure output (INBORR 1992).
These are additional effects of feed enzymes, which
may be of varying economical and practical signifi-
cance but, nevertheless, give further evidence of
their activity.

In summary, it appears that reduction of the in-
testinal viscosity plays an important part of the
action of supplementary dietary enzymes also in
pigs (Figure 7a). However, it was not possible to
find any correlation between the viscosity reduc-
tion and any of the other GI or performance para-
meters. This is in contrast with observations made
with poultry. It is possible that the magnitude of the
viscosity reduction observed in the pigs was insuf-
ficient to make accurate correlation analysis and
that the number of observations were too low. Re-

Fig. 6. Concentration of the main VFAs (acetic, propionic
and butyric acids) and lactic acid in the third (Sl3) and fourth
(SI4) quarters of the small intestine, ceacum, ascending
(Colonl) and descending (Colon2) colon and rectum of pigs
fed either a hulless (var. Condor) without (■) orwith (E 3), or
a hulled (var. Arra) barley without (□) or with (^) added
(Cglucanase. A significant effect of enzyme (P=0.044) was
observed.

16

Agric. Sei. Finl. Suppl. No. 2 (1994)



ducing digesta viscosity by adding only fibre-de-
grading enzymes resulted in a decreased secretion
of digestive enzymes in the pigs without affecting
nutrient digestiblility. This means a more efficient
digestion resulting in less endogenous losses and
better utilisation of dietary energy for growth. In
this situation, application of starch degrading in
combination with fibre degrading enzymes appears

to be of additional benefit. Not only can starch be
more rapidly hydrolysed and absorbed but less eas-
ily fermentable material will reach the lower parts
of the small intestine, thus reducing microbial
growth and fermentation. These effects contribute
into an improved animal performance, less diges-
tive disorders and increased faeces dry matter con-
tent.

CONCLUSIONS

The stabilised feed enzyme product employed
maintained its activity during the steam-pelleting
process up to 85°C conditioning temperatures. In
assessing enzyme stability and predicting the re-
sponse to enzyme supplementation on animal per-
formance the in-vivo enzyme activity assay meth-

ods appeared to be superior to those carried out
in-vitro.

The fungal enzymes employed appeared to pos-
sess an inherent stability to the proteolytic activities
encountered in the digestive tract of pigs and poul-
try and it was found that wheat and wheat gluten

Fig. 7a. Mode of action of carbohydrate-degrading enzymes in pig starter diets based on barley and wheat
(References: I=Cherbut et al. 1990, 2=lnborr 1992, 3=Bedford et al. 1992, 4=Kamphues 1987,s=lnborr
and Ogle 1988, 6=Böhme 1990).

17

Agric. Sei. Finl. Suppl. No. 2 (1994)



can act as stabilisers to microbial enzymes probably
due to their buffering capacity and by providing
substrates to the enzymes.

In the stomach of pigs xylanase and (3-glucanase
activities were completely recovered 0.5 hours after
feeding but decreased with time and were less than
20% of the initial values four hours after feeding.
At the terminal ileum no added enzyme activities
could be found during the first two hours after
feeding, suggesting complete inactivation during
the 12-hour feeding intervals. During the two sub-
sequent two-hour periods the proportion of added
in relation to endogenous enzyme activities in-
creased, indicating that part of the added enzymes
maintained their activity through to the end of the
small intestine.

In pigs fed barley-based diets, (3-glucanase sup-
plementation reduced digesta viscosity and the con-
centration of digestive enzymes in the small intes-
tine without affecting nutrient digestibility. These
results indicate that conditions for a more efficient
digestion were created, which decreased the need
for digestive enzymes, thus reducing endogenous

losses and increasing the amount of dietary energy
and protein available for growth. This assumption
was supported by the improved performance of
pigs fed the enzyme supplemented barley-based
diets.

Adding fibre-degrading in combination with
starch-degrading enzymes to a dietbased on barley
and wheat significantly increased the digestibility
of starch and generally inceased therate ofnutrient
absorption in the proximal parts of the small intes-
tine of early-weaned pigs. This decreases the
amount of easily fermentable substrates in the hind
gut and reduce microbial fermentation. (3-glucanase
supplementation of barley-based diets reduced the
VFA and lactic acid concentration in the caecum,
colon and rectum indicating reduced microbial ac-
tivity in these segments.

Supplementation of pig starter diets based on
barley and wheat with a combination of fibre- and
starch-degrading enzymes improves liveweight
gain and feed utilisation by reducing gut viscosity,
rendering the digestion more efficient and reducing
the microbial activity in the hindgut.

REFERENCES

Almirall, M., Brufau, J. & EsteveGarcia, E. 1993. Effects
of intestinal viscosity on digestive enzyme activities of
intestinal content and ileal digestibilities of poultry fed
barley diets at different ages supplemented with (3-gluca-
nases. Proceedings of the Symposium on Enzymes in
Animal Nutrition, ETH Zurich, p. 69-72.

Åman, P. 1987. Analys och kemisk sammansättning av
svensk spannmål. Fakta husdjur, Nr 3. Sveriges Lant-
bruksuniversitet. p. 4.

Argenzio,R. A. 1982.VFA production and absorption from
the large intestine of thepig. In: Laplace, J.P. et al. (eds.).
Physiologic Digestive Chez le Pore. Colloques de
ITNRA, No. 12. p. 207-215.

Bach Knudsen, K.E., Borg Jensen, 8., Andersen, J.O. &

Hansen, I. 1991. Gastrointestinal implications in pigs of
wheat and oat fractions. 2. Microbial activity in the gas-
trointestinal tract. British Journal of Nutrition 65: 233-
248.

Bedford, M.R. & Classen, H.L. 1992. Reduction of intesti-
nal viscosity through manipulation of dietary rye and
pentosanase concentration is effected through changes in
the carbohydrate composition of the intestinal aqueous
phase and results in improved growth rate and feed con-
version. Journal of Nutrition 122: 560-569.

—, Patience, J.F., Classen, H.L. & Inborr, J. 1992. The
effect of dietary enzyme supplementation of rye- and
barley-based diets on digestion and subsequent perform-
ance in weanling pigs. Canadian Journal of Animal
Science 72: 97-105.

Böhme, H. 1990. Experiments on the effect of enzyme sup-
plements as a growth promotor for piglets. Landbaufor-
schung Völkenrode 40: 213-217.

Burazcewska, L . 1988. Ileal digestibility of rye fed to pigs
with ’Enzyme PCT from B. subtilis. Proceedings of the
IV International Symposium on Digestive Physiology in
the Pig. Polish Academy of Sciences, Institute of Animal
Physiology and Nutrition. Jablonna, Poland, p. 378-380.

Cherbut, C., Albina, E., Champ, M., Doublier, J.L. & Le-
cannu, G. 1990. Action of guar gums on the viscosity of
digestive contents and on the gastrointestinal motor func-
tion in pigs. Digestion 46: 205-213.

Chesson, A. 1987. Supplementary enzymes to improve the
utilisation of pig and poultry diets. In: Haresign, W. &

Cole, D.J.A. (eds.). Recent advances in animal nutrition.
Butterworths, London, p. 71-89.

Decordt, S., Sara via, J., Hendrickx, M., Maesmans, G. &

Tobback, P. 1992. Changing the termostability on Bacil-
lus licheniformis a-amylase. Proceedings of the Interna-

18

Agric. Sei. Fint. Suppl. No. 2 (1994)



tional Symposium on ’Stability and stabilisation of en-
zymes’. Maastricht, The Netherlands. (Poster.)

Dierick, N.A. 1989. Biotechnology aids to improve feed and
feed digestion: enzymes and fermentation. Archives of
Animal Nutrition, Berlin. 39: 241-261.

Din, N., Gilkes, N.R., Tekant, 8., Miller, Jr., R.C., War-
ren, R.A.J. & Kilburn, D.G. 1991. Non-hydrolytic dis-
ruption of cellulose fibres by the binding domain of
bacterial cellulase. Biotechnology 9: 1096-1099.

Englyst, H.N. & Cummings, J.H. 1985. Measurement of
starch and non-starch polysaccharides and their break-
down in the small intestine of man. In; Just, A. et al.
(eds.). Digestive physiology in the pig. Copenhagen, Na-
tional Institute of Animal Science, p. 188-191.

Fengler, A.S. & Marquardt,R.R. 1988. Watersoluble pen-
tosans from rye: 11. Effects on rate of dialysis and on the
retention of nutrients by the chick. Cereal Chemistry 65:
298-302.

Goodfrey, T. & Reichelt, J. 1983. Industrial enzymology.
582 p. Stockton Press, New York, NY.

Goodlad, J.S. & Mathers, J.C. 1991. Digestion by pigs of
non-starch polysaccharides in wheat and raw peas (Pisum
sativum) fed in mixed diets. British Journal ofNutrition
65: 259-270.

Graham, H. 1988. Dietary fibre concentration and as-
similation in swine. ISI Atlas Sci. Anim. Plant Sei. 1:
76-80.

—, Åman, P. & Löworen, W. 1988a. Enzyme supplementa-
tion of pig diets. In: Buraczewska, L. et al. (eds.). Pro-
ceedings of the 4th International seminar on Digestive
physiology in the pig. Jablonna, Polish Academy of
Sciences, p. 371-376.

—, Fadel, J.G., Newman, C.W. & Newman, R.K. 1989.
Effect of pelleting and (J-glucanase supplementation on
the ileal and fecal digestibility of a barley-based diet in
the pig. Journal of Animal Science 67: 1293-1298.

—, Hesselman, K. Jonsson, E, & Åman, P. 1986, Influence
of (J-glucanase supplementation on digestion of a barley-
based diet in the pig gastrointestinal tract. Nutrition Re-
ports International 34: 1089-1096.

—, Löworen, W., Pettersson, D. & Åman, P. 1988b. Effect
of enzyme supplementation on digestion of abarley/pol-
lard-based pig diet. Nutrition Reports International 38:
1073-1079.

Ikegami, S., Tsuchihashi, F., Harada, H., Tsuchihashi, N.,
Nishide, E. & Innami, S. 1990. Effect of viscous indigest-
ible polysaccharides on pancreatic-biliary secretion and
digestive organs in rats. Journal of Nutrition 120: 353-
360.

Inborr, J. 1992. The manure threat - a challenge for feed
enzymes. Pig International, January, p. 18-19.

—, Bedford, M.R. & Graham, H. 1993. Stability and mode
of action ofpoultry feed enzymes in diets based on wheat
and barley. Proceedings of the Australian Poultry Science
Symposium, Sydney. Vol. 5: 53-56.

—, Bedford, M.R., Patience, J.F. & Classen, H.L. 1991.
The influence of supplementary feed enzymes on nutrient
disappearance and digesta characteristics in the Gl-tract
of early-weaned pigs. Proceedings of the sth Intema-

tional Symposium on ’Digestive physiology in pigs’.
Wageningen, The Netherlands. EAAP publication No.
54: 405-410.
& Graham, H. 1991. The effect of enzyme supplementa-
tion ofa wheat/barley-based starter diet on nutrient faecal
digetibility in early-weaned piglets. Animal Production
52: 565. (Abstract from the BSAP winter meeting in
Scarborough, April 1991).
& Ogle, B. 1988. Effect of enzyme treatment of piglet
feeds on performance and post-weaning diarrhoea.
Swedish Journal of Agricultural Research 18: 129-133.

Just, A., Fernandez, J.A. & Jorgensen, H. 1983. The net
energy value of diets for growth in pigs in relation to the
fermentative processes in the digestive tract and the site
of absorption of the nutrients. Livestock Production
Science 10: 171-186.

Kamphues, J. 1987. Untersuchungen zu Verdauungsvor-
gängen bei Absetzferkeln in Abhängigkeit von Futter-
menge und -zubereitung sowie von Futterzusätzen. Ha-
bilitationsschrift zur Erlangung der VENIA LEGENDI
an der Tierärtzlichen Hochschule Hannover.

Langlois, A., Corring, T. & Fevrier, C. 1987. Effects on
wheat bran on exocrine pancreas secretion in the pig.
Reproduction Nutrition Development 27: 929-939.

Low, A.G. 1985. The role of dietary fibre in digestion,
absorption and metabolism. In: Just, A. et al. (eds.).
Digestive physiology in the pig. Copenhagen, National
Institute of Animal Science, p. 157-179.

McClean, D., McEvoy, J. & McCracken, K.J. 1993. Effects
of processing and feed enzymes on nutrient digestibility
in diets for weaned pigs. Proceedings of the Nutrition
Society. Vol. 52:211A.

McCracken, K.J., Urquhart, R. & Bedford, M.R. 1992.
Effect of heat treatment and enzyme supplementation of
barley-based diets on performance of broiler chickens.
Proceedings of the Nutrition Society (in press).

Markström, 8., Pettersson, D. & Hesselman, K. 1985.
p-glukanastillsats till grov- och finmalet korn - ett smält-
barhetsförsök med smågrisar. Swedish University of Ag-
ricultural Sciences, Department of Animal Nutrition and
Management. Report 149.

Meyer, J.H., Gu, Y., Elashoff, J., Reedy, T., Dressman, J.
& Amidon, G. 1986. Effects of viscosity and fluid out-
flow on postcibal gastric emptying of solids. American
Journal of Physiology 250; Gl6l-G164.

Näsi, M. 1991. Digestibility and protein utilisation responses
of soybean and rape seed meal tophysical and enzymatic
treatments in diets for growing pigs. Journal of Agricul-
tural Science in Finland 63: 465-474.

Newman, C.W., Eslick, R.F. & ElNegoumy, A.M. 1983.
Bacterial diastase effect on the feed value of two hulless
barleys for pigs. Nutrition Reports International 28: 139-
146.

—, Eslick, R.F., Pepper, J.W. & ElNegoumy, A.M. 1980.
Performance ofpigs fed hulless and civered barleys sup-
plemented with or without bacterial diastase. Nutrition
Reports International 22: 833-836.

—, Newman, R.K. & Danielson, A.D. 1992. Effect of sup-
plemental enzymes on the performance ofyoung pigs and

19

Agric. Sei. Finl. Suppl. No. 2 (1994)



broiler chicks. Proceedings of the 43rd Montana Live-
stock Nutrition Conference. Manuscript, p. 8.

Nieves, R.A., Ellis, R.P., Todd, R.J., Johnson, T.J.A.,
Grohmann, K & Himmel. 1991. Visualization of Tricho-
derma reesei cellobiohydrolase I and endoglucanase I on
aspen cellulose by using antibody-colloidal gold conju-
gates. Applied and Environmental Microbiology 57:
3163-3170.

Pettersson, D. 1988.Composition and productive value for
broiler chickens of wheat, triticale and rye. PhD thesis.
Swedish University of Agricultural Sciences, Depart-
ment of Animal Nutrition and Management. Rapport
177, Uppsala.

Pierzynowski, S. 1991. Development and regulation of por-
cine pancreatic function with special reference to the
exocrine pancreas. PhD thesis. Dept, of Animal Physiol-
ogy, University of Lund, Sweden.

Poole, D.M., Durrant, A.J., Hazelwood, G.P. & Gilbert,
H.J. 1991.Characterisation of hybrid proteins consisting
of the catalytic domains of Clostridium and Ruminococ-
cus endoglucanases, fused Pseudomonas noncatalytic
cellulose-binding domains. Biochemical Journal 279:
787-792.

Potkins, Z.V., Lawrence, T.L.J. & Thomlinson, J.R. 1991.
Effects ofstructural and nonstructural polysaccharides in
the diet of the growing pig on gastric emptying rate and
rate of passage of digesta to the terminal ileum and
through the total gastrointestinal tract. British Journal of
Nutrition 65: 391-413.

Rainbird, A.L. & Low, A.G. 1986.Effect of various types of
dietary fibreon gastric emptying ingrowing pigs. British
Journal of Nutrition 55: 111-121.

Ratcliffe, B. 1985. The influence of the gut microflora on
the digestive processes. In: Just, A. et al. (eds.). Digestive
physiology in the pig. Copenhagen, National Institute of
Animal Science, p. 245-267.

Rerat, A., Fiszlewicz, M., Giusi, A. & Vaugelade,P. 1987.
Influence of meal frequency onpostprandial variations in
the production and absorption of volatile fatty acids in the
digestive tract of concious pigs. Journal of Animal
Science 64; 448-456.

Roberts, F.G., Smith, H.A., Low, A.G. & Ellis, P.R. 1990,

Influence of wheat breads containing guar flour supple-
menation ofhigh and low molecular weights on viscosity
of jejunal digesta in the pig. In; Southgate, D.A.T. et al.
(eds.). Dietary fibre: Chemical and biological aspects.
Special publication No. 83, Royal Society of Chemistry,
p. 164168.

Salih, M.E., Classen, H.L. & Campbell, G.L. 1991. Re-
sponse of chickens fed hull-less barley to dietary P-glu-
canases at different ages. Animal Feed Science and Tech-
nology 33: 139-149.

Skadhauge, E. 1985. The secretory response of the digestive
tract to the diet. In: Just, A. et al. (eds.). Digestive physi-
ology in the pig. Copenhagen, National Institute ofAni-
mal Science, p. 81-100.

Souffrant, W.G. 1991. Endogenous nitrogen losses during
digestion in pigs. In: Verstegen, M.W.A. et al. (eds.).
Proceedings of the Vth International Symposium on Di-
gestive Physiology in Pigs. EAAP publication No. 54.
Pudoc, Wageningen. p. 147-167.

Tangendjaja, 8., Johnson, Z.B. & Noland, P.R. 1988. Ef-
fect of cooking and addition of enzymes on feeding value
of rice bran for swine. Nutrition Reports International 37:
449-459.

Thacker, P.A., Campbell, G.L. & Grootwassink, J.W.D.
1992. Effect of salinomycin and enzyme supplementa-
tion on nutrient digestibility and the performance ofpigs
fed barley- or rye-based diets. Canadian Journal ofAni-
mal Science 72: 117-125.

Thomke, S., Rundgren, M. & Hesselman, K. 1980. Effects
of feeding high-viscosity barley to pigs. 31st EAAP Con-
gress, Munich. Session N 6:2.

Trowell, H., Southgate, D.A.T., Wolever, T.M.S., Leeds,
A.R., Gassull, M.A. & Jenkins, D.J.A. 1976. Dietary
fibre redefined. Lancet 1: 97.

Yen, J.T., Nienaber, J.A., Hill, D.A. & Pond, W.G. 1991.
Potential contribution of absorbed volatile fatty acids to
whole-animal energy requirement in concious swine.
Journal of Animal Science 69: 2001-2012.

Zebrowska, T. & Low, A.G. 1987. The influence of diets
based on whole wheat, wheat flour and wheat bran on
exocrine pancreatic secretion in pigs. Journalof Nutrition
117: 1212-1216.

20

Agric. Sei. Finl. Suppl. No. 2 (1994)



SELOSTUS

Hiilihydraatteja pilkkovien entsyymien lisäys porsasrehuihin
- stabiilisuus, aktiivisuus ja toimintatavat

Johan Inborr

Helsingin yliopisto

Rehuentsyymien stabiilisuutta rakeistuksessa sekä sian ja
kanan ruoansulatuskanavaa muistuttavissa olosuhteissa tut-
kittiin in-vitro- ja in-vivo -menetelmiä käyttäen kaikkiaan
kolmessa eri kokeessa. Entsyymilisäysten aikaansaamia por-
saiden tuotantotuloksiin vaikuttavia tekijöitä tutkittiin lisäksi
ohra- ja vehnäpohjaisilla rehuilla kahdessa eri kokeessa.

Käytetyt rehuentsyymit säilyttivät rakeistuskokeiden tulos-
ten mukaan aktiivisuutensa 85 °C:een asti. Käytetyn in-vitro
-menetelmän todettiin aliarvioivan rakeistetussa rehussa säi-
lyvää entsyymiaktiivisuutta. Broilereiden ohutsuolen sisällön
viskositeetin mittaamiseen perustuva in-vivo -menetelmä so-
veltui paremmin tähän tarkoitukseen. Ohutsuolen sisällön vis-
kositeetin, broilereiden kasvun ja rehun hyväksi-
käytön (r=0.616) välillä todettiin voimakas negatiivinen kor-
relaatio. Ruoansulatuskanavaa jäljittelevissä olosuhteissa teh-
dyt in-vitro -inkuboinnit osoittivat entsyymien inaktivoitumi-
sen olevan vähäistä. Lisäksi vehnä javehnägluteeni paransivat
entsyymien stabiilisuutta niiden puskoroivien vaikutusten an-
siosta tai toimimalla entsyymien kiinnityskohteina. Saadut
tulokset olivat yhdenmukaisia broilereilla ja sioilla saatujen
tulosten kanssa. Rehuun lisätty koko ksylanaasi- ja (3-gluka-
naasimäärä säilyi sian mahalaukussa puolen tunnin ajan ruo-
kinnasta ja 10-20% lisätystä entsyymimäärästä todettiin vielä
4 tuntiaruokinnan jälkeen. Lisätyistä entsyymeistä säilyi suu-
rempi osuus 2-6 tuntia ruokinnan jälkeen kuin kahden ensim-
mäisen tunnin aikana. Ohrapohjaiseen rehuun lisätty (1-gluka-
naasi säilyi kokonaisuudessaan broilerin ohutsuolessa. Nämä
tulokset osoittavat, että käytetyt entsyymit kestävät ruoansu-
latuskanavan proteolyyttisiä olosuhteita suhteellisen hyvin.

Porsaat sulattivat ohra- ja vehnäpohjaisten rehujen (l-glu-
kaania, tärkkelystä ja kuiva-ainetta paremmin (P<0.05), kun
rehuun oli lisätty kuitua ja tärkkelystä hajottavaa entsyymise-
osta. Tästä huolimatta porsaiden kasvuja rehun hyväksikäyttö
eivät parantuneet. Lisättäessä rehuun entsyymejä todettiin,
että ruoansulatus nopeutui ja suurempi osuus ravintoaineista
imeytyi ohutsuolen alkupuolella. Toisen kokeen tulosten mu-
kaan (3-glukanaasilisäys paransi porsaiden kasvua (P=0.074)
ja rehun hyväksikäyttöä (P=0.058), kun rehussa oli joko
vähän tai runsaasti (3-glukaania sisältävää ohraa. Ravintoai-
neiden sulavuuteen ei käsittelyllä ollut vaikutusta, mutta sekä
ruokasulan viskositeetti (P<0.03) että ruoansulatusentsyy-

mien määrä (P<0.08) vähenivät. Viskositeetin vähentyminen
sai aikaan optimaalisemmat olosuhteet ruoansulatuksen te-
hostumiselle, Tällöin tarvittiin vähemmän ruoansulatusent-
syymejä vertailuryhmään nähden sulavuuden kuitenkaan hei-
kentymättä. Tämä on ilmeisesti yhteydessä alentuneeseen en-
dogeeniseen eritykseen ja tehostuneeseen energian hyväksi-
käyttöön. Tämän lisäksi entsyymilisäys vähensi (P=0.044)
haihtuvien rasvahappojen määrää umpi- ja paksusuolessa,
mikä viittaa alentuneeseen mikrobiaktiivisuuteen.

Yhteenvetona voidaan todeta, että ohra- ja vehnäpohjaisiin
rehuihin lisätyt kuitua ja tärkkelystä hajoittavat entsyymit
säilyttivät aktiivisuutensa ja toimivat ohutsuolessa vähentäen
ruokasulan viskositeettia sekä luoden ruoansulatukselle otol-
lisemmat olosuhteet. Ruoansulatusentsyymejä tarvitaan siten
vähemmän jaravintoaineiden imeytyminen tehostuu ohutsuo-
len alkuosasta. Tästä on seurauksena mikrobitoiminnan vä-
hentyminen umpi- japaksusuolessa, millä on vaikutusta por-
saiden terveydentilaan. Nämätekijät yhdessä edistävät porsai-
den kasvua and rehun hyväksikäyttöä (Kuvio 7b).

Kuva 7b. Ohra-ja vehnäpohjaisiin rehuihin lisättyjen hiilihy-
draatteja pilkkovien entsyymien toimintatavat porsailla (Kir-
jallisuusviitteet: I=Cherbut et ai. 1990, 2=lnborr 1992,
3=Bedford et ai. 1992, 4=Kamphues 1987, s=lnborr ja Ogle
1988, 6=Böhme 1990).

21

Agric. Sei. Finl. Suppl. No. 2 (1994)



SAMMANDRAG

Tillsats av kolhydratspjälkande enzymer till smågrisfoder
- stabilitet, aktivitet och verkningssätt

JohanInborr

Helsingfors universitet

Inom detta projekt genomfördes sammanlagt fem försök för
att undersöka stabiliteten hos foderenzymer under pelletering
och de proteolytiska förhållanden som råder i matspjälk-
ningskanalen hos svin och höns. Dessutom gjordes ett försök
att förklara de mekanismer som står bakom den höjda
tillväxthastigheten och det förbättrade foderutnyttjandet hos
grisar som utfodrats med kom- och vetebaserade foder med
enzymtillsats.

Pelleteringsförsöket visade att den foderenzym som använ-
des tålde en förkonditionering i temperaturer upp till 85 °C
utan att förlorasin aktivitet. Vidare kunde konstateras att mäta
den kvarblivna aktiviteten efter pelleteringen i foder med
in-vitro analysmetoder underestimerar den verkliga aktivite-
ten. För detta ändamål är in-vivo modeller, som t.ex. den som
baserar sig på mätning av tarminnehällets viskositet hos
slaktkycklingar, bättre lämpade. Dessutom var korrelationen
hög mellan tarminnehällets viskositet och slaktkycklingamas
tillväxt (r2=0.624) och foderutnyttjande (r2 =0.616). Inkuba-
tionema in-vitro , motsvarande förhållandena i matspjälk-
ningskanalen, visade att enzymer inte inaktivieras under kor-
tare perioder i en sådan miljö och att vete och vetegluten
förbättrade stabiliteten antingen genom sin bufferverkan eller
genom att de innehåller substrat som enzymerna kan bindas
till. Dessa resultat stöds av de observationer som gjordes med
slaktkycklingar och svin. Tillsatt xylanas och (3-glucanas
kunde helt återfinnas i prov tagna ur magsäcken hos svin en
halv timme efter utfodring, medan 10-20% fortfarandekunde
återfinnas 4 timmar efter utfodring. Proportionellt sett kunde
mer tillsatt enzymeaktivitet återfinnas i prov tagna frän slutet
av tunntarmen (ileum) under perioden 2 till 6 timmar jämfört
med perioden 0 till 2 timmarefter utfodringen. Hos slaktkyck-
lingar som utfodrades med tillsatta enzymer i fodretkunde allt
(3-glukanas återfinnas i prov tagna från tunntarmens övre
halva. Dehär resultaten ger bevis pä att de enzymer som
användes ärrelativt stabila i matspjälkningskanalen.

Genom att tillsätta en blandning av fiber- och stärkelsened-
brytande enzymer till ett foder innehållande vete och kom,
förbättrades smältbarheten hos (3-glukan, stärkelse och torr-
substans (P<0.05) i den sista fjärdedelen av tunntarmen hos
smågrisar. Trots att denhär förbättringen inte ledde till en
bättre tillväxt eller ett bättre foderutnyttjande, visade resulta-
ten att matspjälkningen accelererades och en större mängd
näringsämnen absorberades högre upp i tunntarmen då enzy-
merna tillsattes. I ett annat likadant försök tillsattes ett (3 -

glukanas till foderbaserade pä antingen ett korn med låg eller
hög (3-glukanhalt. Detta resulterade i en högre daglig tillväxt

(P=0.074) och ett bättre foderutnyttjande (P=0.058). Fodrets
smältbarhet förblev oförändrat, men tarminnehällets viskosi-
tet (P<0.03) och mängden pankreasenzymer (P<0.08) i de tre
första ijärdedelarna av tunntarmen reducerades genom (3 -
glukanastillsatsen. Det verkar som om förutsättningar för en
mera effektiv matspjälkning skapades genom den sänkta
tarmviskositeten. Därigenom krävdes mindre mängd
matspjälkningsenzymer föratt uppnå en lika hög smältbarhet
som hos kontrollgruppen. Detta torde leda till mindre endo-
gena förluster och ett förbättrat utnyttjande av energin för
tillväxt. Dessutom uppmättes en lägre (P=0.044) koncentrati-
on av flyktiga fettsyror i blind- och tjocktarmen hos grisarna,
vilket tyder pä en minskad mikrobien fermentation.

Sammanfattningsvist kan konstateras att rätt valda fiber-
och stärkelsenedbrytande enzymer, som tillsatts till foder ba-
serade på korn och vete, bibehåller sin aktivitet och verkar i
tunntarmen genom att minska tarminnehällets viskositet och
skapa förutsättningar för en effektivare matspjälkning (Figur
7c). Mindre mängd matspjälkningsenzymer behövs och en
större mängd näringsämnen absorberas i de främredelarna av
tunntarmen. Detta leder vidare till en minskad mikrobien
aktivitet i blind- och tjocktarmen, vilket främjar grisarnas
hälsotillstånd. Alla dessa faktorer bidrar till en förbättrad
tillväxt och ett mera effektivt foderutnyttjande.

Fig. 7c. Verkningssätt hos kolhydratspjälkande enzymer i
smågrisfoder baserat pä korn och vete (Litteraturhän-
visningar: I=Cherbut et al. 1990, 2=lnborr 1992, 3=Bed-
ford et al. 1992, 4=Kamphues 1987, s=lnborr och Ogle
1988, 6=Böhme 1990),

22

Agric. Sd. Fin!. Suppl. No. 2 (1994)



ZUSAMMENFASSUNG

Zusatz von kohlenhydratspaltenden Enzymen zu Ferkelfutter
- Stabilität, Aktivität und Wirkungsweise

Johan Inborr

Universität Helsinki

In insgesamt fiinf Versuchen wurde die Stabilität von Futter-
enzymen unter Bedingungen der Dampfpelletierung und der
proteolytischen Gegebenheiten im Verdauungstrakt bei
Schweinen und Gefliigel untersucht. In einzelnen wurde zu
klären versucht, welche Mechanismen die Leistungssteige-
rung bei Schweinen bewirken, wenn Futterenzyme Getreide-
rationen zugemischt werden.

Die Versuchsergebnisse zur Pelletierstabilität zeigen, daß
die Enzymaktivitäten des untersuchten Enzympräparates bis
zu Temperaturen von 85 °C (bei der Konditionierung) un-
beeinflußt bleiben. Untersuchungen zur Ermittlung der Wie-
derfmdungsrate haben ergeben, daß m-vitro-Metoden die tat-
sächliche Enzymaktivität im pelletierten Futter deutlich unter-
schätzen. /n-vivo-Metoden, die auf Messungen der Viskosität
des Darminhaltes der Mastkiiken beruhen, spiegeln die Ver-
hältnisse exakter wieder, was sich auch in der hohen Korrela-
tion zum Lebendmassezuwachs (r2 =0.624) und zurFutterver-
wertung (r=0,616) zeigte. Die in v/tra-Untersuchungen, bei
denen im Verdauungstrakt vergleichbar Inkubationsver-
hältnisse simuliert werden, ergaben, daß die Enzyme nur
unvollständig abgeaut und inaktiviert werden. Weizen und
Weizenglutein begunstigten die Aufrechterhaltung der Akti-
vitäten durch ihre Pufferkapazität bzw. durch ihre Eigenschaft
als Substratdonator. Diese Befunde konnien durch in-vivo -

Untersuchungen gestutzt werden. So werden bei Schweinen 4
Stunden nach der Fiitterung 10%-20% der Aktivitäten im
Magen wiedergefunden; 4-6 Stunden nach dem Futtern sogar
deutlich höhere Anteile am Ileum. Auch bei Mastkiiken,
denen bei einer Gerstenration Enzymzusätze verfuttert wor-
den waren, wurde die (5-Glucanase im proximalen Teil des
Dunndarms vollständig wiedergefunden, was fur eine aus-
reichende Stabilität dieses Enzyms im Verdauungstrakt
spricht.

Bei Zusätzen von geriistsubstanz- und stärkespaltenden
Enzymen zu Weizen/Gerste-Rationen wurde bei friih ent-
wöhnten Ferkeln eine verbesserte Verdaulichkeil der (3-Glu-
kane, der Futtertrockenmasse und der Stärke (P<0.05) im
letzten Viertel des Dunndarms festgestellt. Auswirkungen auf
dieWachstumsleistung und auf die Futterverwertung wurden
jedoch nicht beobachtet. Es zeigte sich jedoch, daß ein größe-
rer Anteil der Nährstoffe im vorderen Diinndarm absorbiert
wurde.

In Rationen, deren Gerste einen hohen bzw. niedrigen (3
Glucangehalt aufwies, fiihrte der Zusatz von (3-Glucanase -

bei 3wöchiger Versuchsdauer - ebenfalls zu einem besseren
Lebendmassezuwachs (P=0.074) und zu einer besseren Futter-

ausnutzung (P=0.058). Bei den Ferkeln war die Verdaulich-
keit der Nährstoffe nicht gesteigert; die Viskosität des Dar-
minhaltes war jedoch herabgesetzt (P<0.03) und die Konzent-
ration der Verdauungsenzyme (P<0.08) in den drei proxima-
len Viertein des Dunndarms verringert. Offensichtlich bewir-
ken Enzymzusätze durch Verringerung der Viskosität und
durch verminderte Produktion von Pankreasenzymen effekti-
vere Verdauungsprozesse, die sich in geringeren endogenen
Verlusten und einer besseren Ausnutzung der fur das Wachs-
tum zur Verfiigung stehenden Energie äußert. Im distalen
Dimndarmabschnitt sowie im Dickdarm wurde zudem eine
niedrigere Mikrobentätigkeit (durch die Konzentration der fl.
Fettsäuren gemessen) festgestellt (P=0.044).

Geriistsubstanz- und stärkespaltende Enzyme - vorausge-
setzt, daß sie entsprechend selektiert sind - wirken beim Fer-
kel im gesamten Verdauungstrakt dadurch, daß die Viskosität
des Darminhaltes herabsetzen und den Verdauungsprozeß ef-
fektiver gestalten (Abbildung 7d): Die Bauchspeicheldriise
hat entsprechend wenigerEnzyme zu produzieren und gröSe-
re Nährstoffmengen werden im vorderen Teil des Dunndarms
absorbiert. Eine geringere Mikrobenaktivität in den hinteren
Abschnitten des Verdauungstraktes ist die Folge und die Fer-
mentationsverluste liegen niedriger. Insgesamt bewirken
dieseEffekte bei Schweinen eine Leistungssteigerung.

Abbildung 7d. Wirkungsweise von Enzymzusätzen bei Ver-
fiitterung von GersteAVeizen-Rationen an Ferkel (Literatur-
hinweise: I=Cherbut et ai. 1990, 2=lnborr 1992, 3=Bed-
ford et ai. 1992, 4=Kamphues 1987, s=lnborr und Ogle
1988, 6=Böhme 1990).

23

Agric. Sei. Fin!. Sappi. No. 2 (1994)



Agric. Sei. Fin!. Suppl. No. 2 (1994)

24



ö= LSETOS ANIEEXSS64

ANIMAL FEED
SCIENCE AND

TECHNOLOGY
ELSEVIER Animal Feed Science and Technology 46 (1994) 179-196

Stability of feed enzymes to steam pelleting during
feed processing

J. Inborr* 1
, M.R. Bedford

Finnfeeds InternationalLid., Market House, High Street, Marlborough SNB lAA, UK

(Received 3 February 1993; accepted 12 October 1993)

Abstract

A barley-based diet for broiler chickens was supplemented with a commercial feed en-
zyme product (Avizyme SX®, containing Trichoderma longibrachiatum /Lglucanase) at
0, 1 and 10 g kg -1 and pelleted after conditioning at 75, 85 or 95°C for either 30 sor 15
min in a 3 X 3 X 2 factorial design. The pelleted feeds were analysed for /(-glucanase activ-
ity, starch, total and soluble/J-glucan and non-starch polysaccharides (NSP) and fed to 1-
day-old male broiler chickens for a 19 day period, at the end ofwhich, weight gain, feed
intake and the incidence ofvent pasting were recorded. On Day 19, four birds per treat-
ment were killed and digesta viscosity and /(-glucanase activity in the proximal part of the
small intestine evaluated.

Conditioningat 75°C for 30 s reduced /?-glucanase activity compared with control mash
diets to 66% of initialactivity, whereas 15 min conditioning at 75 °C reduced recovery to
49%. At 85°C with 30 s and 15 min conditioning, the recoveries were 56% and 31%, and
at 95°C, these were 16% and 11%, respectively. There was a negative quadratic effect of
conditioning temperature and a positive linear effect of enzyme level on liveweight gain
and feed efficiency. Conditioningtime did not influence chick performance and there were
no treatment effects on feed intake. The incidence of vent pasting decreased linearly with
increasing enzyme inclusion rates and was significantly higher after conditioning for 15
min compared with 30 s.

Dietary /Lglucanase correlated linearly with digesta /i-glucanase activity and the slope of
the relationship indicatedfull recovery of the enzyme in the small intestine of the birds.
Digesta /J-glucanase activity correlated linearly with viscosity, which in turn correlated
highly with bird performance. Despite this, dietary /J-glucanase values required transfor-
mation to their logarithms before they could be used to accurately predict bird performance.

The results from this experiment suggest that partial enzyme inactivation occurs at pel-
leting. The magnitude of the inactivation depends on the pelleting conditions employed,
with higher temperatures and prolonged conditioning times increasing inactivation. How-

'Correspondingauthor.
'Present address:Fifttmigaf LD-20095 r Hamburg, Germany, f
0377-8401/94/507.00 © 1994Elsevier Science B.V. All rights reserved
55D1 0377-8401,/SSDI 0377-8401/ ,

/ /ajV.-r,,
_Z A . U‘H)VCbU,-C

cSv/e-rvslto, Feeler Åb , , Bex S-S3IIC- LicLlec-jsirvj, ,



Cf LSETOS AI9JFEXSS64

180 J. Inborr, M.R. Bedford /AnimalFeed Science and Technology 46 (1994) 179-196

ever, in this experiment, bird performance was only affected when feeds were pelleted at
temperatures over 85°C. The viscosimetric method used for measuring/Tglucanase activ-
ity in the feeds proved to give a good estimate of the in vivo activity based on digesta
viscosity. However, values obtained with this method needed a logarithmic transforma-
tion for accurate prediction of chick performance, thereby reducing sensitivity with values
normally encountered in feed. Digesta viscosity was a good indicator of bird performance.

1. Introduction

There is a considerable amount ofpublished work showing the response in an-
imal performance to enzyme supplementation of feeds (reviewed by Chesson
(1987) and Dierick (1989)). With increased commercial use of feed enzymes in
processed feeds the question of enzyme stability becomes more and more of an
issue. In particular, the ability of enzymes to withstand the various heat treat-
ments (pelleting, expansion, extrusion, etc.) now being employed in feed manu-
facturing is seriously questioned. Furthermore, once the enzymes enter into the
gastrointestinal tract, where the enzyme action occurs, they are subjected to a
range ofconditions that can cause denaturation, thus rendering them inactive.

Cellulases and hemicellulases are the most commonly used enzymes in feed
applications. They are mainly microbial enzymes produced by fungi and bacte-
ria. In general, the inherent heat stability ofmesophilic microbial enzymes is poor,
although enzymes ofdifferent origin can differconsiderably in this respect (God-
frey and Reichelt, 1983). A number of stabilisation methods have been devel-
oped to minimise these problems.

Results from chick performance experiments show significant responses to en-
zyme supplementation of the same feed fed either as a mash or as pellets (Inborr
and Graham, 1991), indicating sufficient enzyme activity after pelleting. Con-
sequently, enzyme stability to heat processing may not be estimable by current in
vitro methods, and so alternative methods need to be developed. Bedford and
Classen (1992) found a high correlation between reduced digesta viscosity and
improved performance of broiler chickens fed wheat- and rye-based diets. Simi-
lar relationships were also found in barley-fed broiler chickens (Inborr et al.,
1993). Monitoring intestinal viscosity would therefore appear suitable for esti-

mating the relative enzyme activity in feed after pelleting.
The objectives ofthis study, therefore, were to investigate the effect of condi-

tioning temperature and time during pelleting on the recovery of/Tglucanase from
the feed by traditional methods. These results were then to be compared with the
response of broilers fed such feeds. The parameters chosen were feed conversion
ratio, weight gain and intestinal viscosity.

2. Materials and methods

2.1. Diets and pelleting

Three target conditioning temperatures (75, 85 and 95°C) and two condition-
ing times (30 s and 15 min) were employed to investigate the influence ofsteam



a- LSETOS ANIFEXSS64

J. Inborr, M R. Bedford /Animal Feed Science and Technology 46 (1994) 179-196 181

pelleting on enzyme activity recovery in feeds. A commercial feed enzyme prod-
uct (Avizyme SX®, Finnfeeds International, Marlborough, UK) was used as the
)?-glucanase source and added at three levels ofinclusion (0, 1 and 10 gkg -1 ) to
the basal diet to give a 3 X 3 X2 factorial design.

A basal feed mix ofapproximately 1800kg, based on barley (hulless, var. Con-
dor) and soya-bean meal, was prepared by blending all the ingredients (Table I),
except soya oil, in a Nauta mixer before grinding in a hammer mill (sieve 2.0
mm). After grinding, soya oil was added during blending. Three batches of ap-
proximately 600 kg each were drawn from the basal feed mix to be used for the

Table 1
Composition and nutrient content of the basal diet (g kg 'as fed)

Ingredients
Barley 602.3

100Wheat
Soya-bean meal
Fish meal

200
20

Soya oil 35
Monocalcium phosphate
Limestone

12
13

L-Lysine-HCI
DL-Methionine

1.8
3.4

Vitamin/mineralpremix'
Salt

10
2.5

Analysed nutrients
Crude protein 203
Crude fat 65
Ash 56
Starch 393
Calcium 8.7
Phosphorus 7.3
Mixed-linked /f-glucans

Total 29
Insoluble
Soluble

15
14

Non-starch polysaccharides
Total 95
Insoluble
Soluble

66
29

Calculated nutrients
Lysine 12.5
Methionine + cystine 9.5
Metabolisable energy (MJ kg~') 12.6

'The premix supplied per kg diet: vitamin A, 9000 IU; vitamin D 3, 2000 IU; vitamin E, 12 IU; vita-
min 8,, 0.5 mg; vitamin 82,B 2, 4 mg; vitamin 86,B6, 1 mg; vitamin 812,B 12, 20/tg; vitamin K.3 , 2 mg; folic acid,
0.3 mg; niacin, 15 mg; pantothenic acid, 12.5 mg; choline, 173.5 g; Ca, 1.4 g; Na, 0.1 mg; Fe, 20 mg;
Mg, 0.37 g; Mn, 70 mg; Zn, 50 mg; Cu, 7 mg; I, 0.8 mg; Se, 0.2 mg; K, 0,24 mg; Co, 0.25 mg; Endox
anti-oxidant, 100 mg.



Q- LSETOS ANIFEXSS64

182 J. Inborr, M R. Bedford /Animal Feed Science and Technology 46 (1994) 179-196

different enzyme treatments. The enzyme was added at 0 gkg”l gkg” 1 and 10
g kg -1 to one of the three batches to form Treatments NE, AO and AT, respec-
tively. To ensure homogeneous mixing of the enzyme supplement, the required
amount was first mixed with 25 kg of the basal feed before being blended into the
rest of the batch.

The feeds were pelleted in batches of approximately 200 kg. When the target
conditioning temperature ofeach batch was reached, half of the batch (100 kg)
was collected from the conditioner into insulated and numbered plastic buckets
and stored for 15 min before being fed into the pelleter by emptying the buckets
in the same sequence as they had been filled. The rest of the feed was pelleted
directly after the 30 s ofconditioning. After pelleting, the feed was spread out on
a plastic sheet for cooling and drying.

Conditioning temperatures were monitored at four points of the conditioner
with fixed thermometers and the temperature of the pellets exiting the die with
an Infra Red thermometer (Infratherm IT 6, IMPAC Electronic GmbH, Ger-
many). Samples were taken for enzyme activity analysis of the mash feed after
mixing, and of all feeds after the cooling and drying stage following pelleting.

The pelleting machine used had a capacity of 500 kg h” 1 (Model V3-30, Ro-
binson Milling (Simon Heesen), Boxtel, Netherlands) with a power output of
18.5 kW. The pellet diameter was 3.5 mm, die thickness was 65 mm and die
diameter 300 mm. The amperage used during the pelleting varied between 16and
17 A.

2.2. Chemical analysis of the diets

The dry matter content of the feeds was determined by oven drying (103°C
overnight). Nitrogen was determined by the Kjeldahl procedure (Association of
Official Analytical Chemists (AOAC), 1990), crude fat by chloroform/methanol
extraction (Atkinson et al., 1972), crude fibre by the method ofAOAC (1990)
with Fibertec (Tecator) and starch enzymatically by the method of Åman and
Hesselman (1984). The ash content of the feeds was determined after the sam-
ples were ashed in an oven at 500°C for 4 h.

Total and soluble non-starch polysaccharide (NSP) content was determined
according to the method ofAOAC (1990). Insoluble NSP was calculated as the
difference between total and solubleNSP contents.

Total and insoluble /i-glucan was determined by the method ofÅman and Gra-
ham (1987) and soluble /?-glucan was calculated as the difference between these
two fractions.

Calcium and phosphorus were determined according to the methods described
by the Ministry ofAgriculture, Fisheries and Food (1985).

2.3. In vitroassays

Enzyme activity measurements in feed and digesla
A 10 g feed sample was accurately weighed into a 100 ml volumetric flask.

Distilled water at room temperature was added to bring the volume up to 100 ml



3s LSETOS ANIFEXSS64

J. Inborr, M R. Bedford / Animal Feed Science and Technology 46 (1994) 179-196 183

and the suspension stirred for 30 min and then filtered through a glass fibre filter
(Macherey Nagel 85/90, Diiren, Germany). The extract was diluted to the ap-
propriate concentration with buffer solutions according to the procedures of the
methods employed.

For the collection of digesta, birds were killed by cervical dislocation and the
abdominal cavity exposed. The small intestine was ligated at the duodenum and
at Meckel’s diverticulum. The total content of this proximal part of the small
intestine was collected into a 100 ml beaker and thoroughly mixed. Approxi-
mately 1.5 g (wet weight) ofthe fresh digesta were immediately placed in a mi-
crocentrifuge tube and centrifuged at 12 700Xg for 2 min. The supernatants were
instantly frozen in liquid nitrogen. Before enzyme analysis the samples were
thawed and filtered (Macherey Nagel 85/90).

Measurements of)?-glucanase (EC 3.2.1.6; endo-(l,3)(l,4)-glucanase) activ-
ity in feed were carried out viscosimetrically using 1.0% barley /i-glucan (Biocon
Biochemicals, Cork, Ireland) in Mcllvaine’s buffer as substrate by the method
described by Bathgate (1979) with minor modifications. This method measures
the reduction in substrate viscosity and enzyme activity is expressed as the in-
crease in reciprocal viscosity (IRV).

One percent (w/w) yS-glucan substrate was prepared by solubilising 1 g ofbar-
ley /J-glucan in 6 ml ofethanol in a tared beaker. To this was added 10 ml of 0.5
M sodium acetate buffer (pH 4.0), then distilled water to a final volume of 100
ml. The /?-glucan solution was standardised to initial reciprocal specific viscosity
of 0.13 by adjusting the /?-glucan:buffer ratio. In the measurements the volume
of the /?-glucan substrate varied between 5.3 and 5.5 ml and that of the enzyme
solution between 2.0 and 2.2 ml. The total volume of the substrate/enzyme so-
lution was adjusted to 7.5 ml before each measurement.

After enzyme addition, four to six flow times were determined over a 30 min
period in an Ostwald capillary viscometer (Model No. 11, 75-100 s, Rudolf Brand
GmbH, Wertheim, Germany), immersed in a water bath set at 30°C. The flow
times were used to calculate the reciprocal specific viscosities, using the equation

1/
dT°

/T,ip dTs -dT0

where 1 /f/sp is the reciprocal specific viscosity, dT0 is the flow time of the buffer
and dTs is the flow time of the substrate-enzyme solution.

/?-glucanase activity is expressed as the increase in reciprocal specific viscosity
(IRV) and calculated using the equation

kD/?-glucanase, IRV units g" 1
=—

where k is the slope of the curve plotted from the reciprocal specific viscosity
versus hydrolysis time (min), Dis the total dilution factor and V is sample vol-
ume. One IRV unit (IRVU) is defined as the change of one (l/f/sp ) per min.
Each measurement was carried out in duplicate.

/?-glucanase activity in the digesta supernatants of birds fed the control (NE)



Or LSETOS ANIFEXSS64

184 J. Inhon. MR. Bedford / Animal Feed Science and Technology 46 (1994) 179-196

and Avizyme SX, 10 g kg -1 (AT) diets were measured by the method suggested
by the MegaZyme (North Rocks, N.S.W., Australia) malt assay kit
with assay conditions modified to 30 min incubation time at 40 °C in 50 mM
acetate buffer (pH 4.6). This assay gives arelative measure of enzyme activity.
To obtain quantitative data, a standard assay based on release ofreducing sugars
using a non-dyed substrate was run parallel to the dye release method. By using
Avizyme SX for calibration of the assays, the unknown activity of the digesta
supernatants could be calculated. One yS-glucanase unit is defined as the amount
of enzyme needed to release 1 /rmol glucose min-1 in the described conditions.

2.4. In vivo assays

2.4.1. Broiler performance trial
This experiment was carried out in two subsequent blocks of 18 treatments

each. In the first block, 270 1-day-old male broiler chickens (hybrid Ross I) were
allocated to cages in groups of five birds and randomly assigned to one of the 18
treatments. In the second block the number ofbirds per cage was four. In all other
aspects, the design was similar to that of the first block.

Birds were fed ad libitum for 19 days and water was freely available at all times
through nipple drinkers. Each diet was fed as a crumble for the first week of the
experiment.

The birds were bulk weighed by cage on Days 1 and 19 and feed inputs and
residuals on Days 7, 14 and 19. Vent pasting was recorded in conjunction with
the weighing on Day 19. All birds were scored for vent pasting and any bird with
excreta adhering to theplumage around the cloaca was recorded.

Dead and culled birdswere weighed at the time ofremoval and abnormal mor-
talities investigated by post mortem examination.

2.4.2. Digesta viscosity
On Day 19, birds per treatment were randomly selected for digesta viscos-

ity measurements. Supernatant (500 /A) obtained as described above was care-
fully withdrawn in a 200-1000//I pipette and viscosity determined using a Brook-
field digital viscometer (Model LVTDVCP-11, Brookfield Engineering
Laboratories, Stoughton, MA) maintained at 20°C and a shear rate of 6-30 s-1.

Digesta viscosity measurements were carried out in both blocks. In addition,
during the second block, digesta viscosities ofthree birds fed a mash control diet
(NE-M) and three birds fed a pelleted (75°C) control diet (NE-P) from Days
17 to 19 were measured by the procedures described above to investigate the
effect ofpelleting on digesta viscosity.

2.5. Calculations and statistical analysis

The experiment was designed as a 3 X2 X2 factorial randomised complete block
design, with blocks representing time. The data were analysed by ANOVA ac-
cording to the general linear models procedure (Statistical Analysis Systems In-



Cr LSETOS ANIFEXSS64

J. Inborr, M R. Bedford / Animal Feed Science and Technology 46 (1994) 179-196 185

stitute, 1982). Means were separated where appropriate by contrasts. Main ef-
fects are reported only when interactions were found to be non-significant
(P> 0.05). Where interactions were significant, analysis of variance was per-
formed by each interactive parameter to determine the simple effects. Owing to
unequal replication, least square means were used where appropriate (gut
viscosity).

3. Results

3.1. Chemical composition ofthe diets

Starch and protein were the main components of the basal mash diet (Table
1). Conditioningtemperature and time did not influence the composition of the
diet (Table 2), whereas enzyme supplementation reduced the starch content
(P< 0.001). Solubility ofthe mixed-linked /?-glucans was on average 82% in the
pelleted feeds compared with 48% in the mash.

3.2. Enzyme activity recovery

Actual conditioning temperatures were equal to or slightly lower than the tar-
gets for the two lower temperatures, whereas 95°C was not achieved during any
of the batch runs (Table 3).

Endogenous /?-glucanase activity in the unsupplemented mash feed (NE-M)

Table 2
Effect of enzyme level, pelleting timeand temperature on starch, total and soluble non-starch polysac-
charides (NSP) and content of the feeds (g kg“' as fed)

Treatment Starch Total NSP Soluble Total/?- Soluble
NSP glucan

Enzyme (gkg-')
0 423 113 36 40 32

1 373 107 35 40 34
10 406 112 35 39 33
f-value 0.00080.3825 0.98360.9319 0.4560

Temperature (° C)
75 409 112 37 39 33
85 395 113 36 40 33
95 400 106 34 40 33
/'-value 0.11040.2569 0.80130.5802 0.9221

Time (min)
0.5 402 112 37 40 33

15 400 109 34 39 33
/'-value 0.30720.4728 0.41120.5762 0.7655



O’ LSETOS ANIFEXSS64

186 J. Inborr, M R. Bedford /Animal Feed Science and Technology46 (1994) 179-196

Table 3
Measured conditioning and pellet exit temperatures (°C), and /f-glucanase activities in mash and
pelleted feed samples

Enzyme Treatment In the conditioner Pellet exit /J-glucanase 1 CV2

dosage temperature (IRVUkg -1 ) (%)

(gkg -1 ) Time Target Measured
(min) temperature temperature

0 NEcontrol (mash) 4.0±0
NE1 3 0.5 75 74 78 2.510.7
NE2 0.5 85 84 84 1.5±0.7
NE3 0.5 95 90 94 1.510.7
NE4 15 75 70 82 2.510.7
NES 15 85 77 83 2.010
NE6 15 95 89 92 1.510.7 28.0

1 AOcontrol (mash) 23.513.5
AOl 3 0.5 75 74 81 15.011.4
AO2 0.5 85 83 82 16.510.7
AO3 0.5 95 91 82 4.512.1
AO4 15 75 71 86 13.513.5
AO5 15 85 83 88 8.510.7
AO6 15 95 91 88 3.510.7 18.4

10 AT control (mash) 27119
ATI 5 0.5 75 75 81 18517
AT2 0.5 85 85 82 113114
AT3 0.5 95 93 83 3617
AT4 15 75 76 84 11214
ATS 15 85 85 85 7115
AT6 15 95 92 91 2013 9.2

AvizymeSX 239001280 1.2

‘Values are means 0 = 2) ±SD.
Coefficient ofvariance percentage, average within each level of enzyme concentration (for NE, AO
and AT n = 7; for Avizyme SX n= I).
3 NE, no enzyme added; AO, enzyme added at 1 g kg~ AT, enzyme added at 10 g kg -1 .

was 4 IRVU kg -1 (Table 3). In the pelleted unsupplemented feeds, between 1.5
and 2.5 IRVU kg -1 was found, indicating significant inactivation even at the
lowest temperatures. /?-glucanase recovery in the supplemented mash feeds (AO-
M and AT-M) was complete. At the lower (1 gkg -1 ) inclusion rates, condition-
ing at 75, 85 and 95°C for 30 s resulted in enzyme recoveries of 64%, 70% and
19% of the mash, respectively. Conditioning for 15 min at the three target tem-
peratures (75°C, 85°C and 95°C) resulted in enzyme activity recoveries of 57%,
36% and 15%, respectively, of the mash feed. The enzyme activity recoveries with
Avizyme added at lOgkg-1 and after conditioning at 75°C, 85°C and 95°C were
respectively 68%, 42% and 13% of the mash when conditioning time was 30 s and



O’ LSETOS ANIFEXSS64

J. Inborr, M.R. Bedford /Animal Feed Science and Technology 46 (1994) 179-196 187

41 %, 26% and 7% when conditioning time was 15 min. The coefficient of vari-
ance (CV) of the recovery values decreased generally with increasing enzyme
concentration (Table 3). Average CV percentages ofthe NE, AO and AT samples
were 28.0, 18.4 and 9.2, respectively, and that of the enzyme premix, 1.2.

3.3. Broiler performance trial

There were no significant interactions between enzyme, temperature and time
for all parameters determined in the broiler experiment apart from an enzyme by
temperature interaction for the parameter digesta viscosity. Consequently, main
effects are given in Table 4 for all parameters except for viscosity.

Liveweight gain and feed efficiency improved linearly with increased levels of
enzyme addition (Table 4). Furthermore, there was a significant quadratic live-
weight gain and feed efficiency response to increased conditioning temperatures,
indicating increasing enzyme activity loss with increased temperature. There was

Table 4
Liveweight gain (g), feed intake (g), feed:gain ratios, gut viscosities (cPs) and vent pasting (% of
birds affected) at Day 19 ofbirds fed the experimental diets

Treatment Liveweight Feed Feed: gain Vent
gain intake ratio pasting

Enzyme (gkg~ ')
0 517 859 1.677.4

1 545 840 1.553.4
10 558 831 1.490.8

P-value 0.00010.6397 0.00010.0200
Contrast

Linear 0.00010.3557 0.00010.0055
Quadratic 0.09860.8493 0.11770.7688

Temperature (° C)
75 547 839 1.544.6
85 556 851 1.531.9
95 517 840 1.645.1

f-value 0.00010.9165 0.00010.2554
Contrast

Linear 0.00010.9530 0.00010.8068
Quadratic 0.00420.6806 0.00450.1027

Time (min)
0.5 544 851 1.571.7

15 539 836 1.576.0

lvalue 0.63210.5587 0.68900.0130
Contrast

Linear 0.63210.5587 0.68900.0130



Ö LSETOS ANIFEXSS64

188 J. Inborr, MR. Bedford /Animal Feed Science and Technology 46 (1994) 179-196

no difference in bird performance when feeds were conditioned at 75 and 85 °C.
Conditioningtime had no effect on bird performance and none of the treatments
had any effect on feed intake.

3.4. Vent pasting

There was a significant negative linear response to enzyme supplementation on
the vent pasting of birds (Table 4). There was a tendency towards a quadratic
effect of conditioning temperature (P=0.1027), positively correlated with the
measured gut viscosities. Conditioning for 15 min significantly (T=0.013) in-
creased the incidence of vent pasting at 19 days of age compared with 30 s
conditioning.

3.5. Digesta viscosity

There was no significant effect of conditioning time on digesta viscosity (data
not shown). Pelleting by itself markedly increased digesta viscosity but owing to
the high variation between the birds, this was not significant (P=0.3724, data
not shown). Increasing pelleting temperatures resulted in a significant
(P=0.0392) quadratic response in birds fed the control diet (Table 5), whereas
there was a linear response with enzyme supplemented feeds (P< 0.002), indi-
cating reduced enzyme activity at higher temperatures. Digesta viscosity de-
creased linearly with increased enzyme inclusion rates at all conditioning tem-
peratures employed (P<0.0005).

3.6. Enzyme activity in the small intestine

With the azo-glucan method employed, it was not possible to reliably measure
)?-glucanase activity in digesta samples obtained from birds fed the unsupple-
mented diets (NE treatments). /?-glucanase activity in samples obtained from

Table 5
Effect of temperature ( °C) and enzyme concentration (g kg~ ') on digesta viscosity (cPs)

Temperature P-value Contrasts

75 85 95 Linear Quadratic

Enzyme
0 31.316.7 29.10.2846 0.76890.0392

1 5.48.6 17.70.0001 0.00010.1059
10 2.93.8 6.00.0280 0.00120.3847

f-value 0.00010.0044 0.0035
Contrast

Linear 0.00010.0004 0.0003
Quadratic 0.00870.5024 0.9719



O- LSETOS ANIFEXSS64

J. Inborr, M.R. Bedford /Animal Feed Science and Technology 46 (1994) 179-196 189

Table 6
/Tglucanaseactivity (U per 1000 ml supernatant) in thedigesta ofbirds fed the AT diets

/J-glucanase 1Treatment

ATI 2

AT2
AT3
AT4
AT5
AT6

31.9113.0
21.817.7
4.913.1
15.612.2
7.017.8
2,611.0

‘Valuesare means (n =4)±SD.
2Enzyme (Avizyme SX) added at 10 g kg

birds fed Avizyme SX at 10 g kg" 1 (AT treatments) feed could be reliably mea-
sured and the results are presented in Table 6. Enzyme activity decreased with
increasing conditioning temperature and time.

3. 7. Enzyme activity in feed and digesta in relation to digesta viscosity and bird
performance

Plotting gut viscosity against dietary /?-glucanase activity on a logarithmic scale
revealed a high correlation between these parameters (r 2 =0.855, Fig. 1). How-

Fig. 1. Relationship between the logarithm of dietary /?-glucanase activity and digesta viscosity: +,

AT (r2
= 0.957); O. NEand AO (r2

= 0.746); for the whole data set r 2 =0.855.



a LSETOS ANIFEXSS64

190 J. Inborr, M R. Bedford /Animal FeedScience and Technology 46 (1994) 179-196

ever, at the low dietary yS-glucanase concentrations (NE and AO treatments), the
correlation was lower (r 2 =0.746) compared with high /J-glucanase concentra-
tion (r2 =0.957). Plotting weight gain and feed efficiency against dietary fi-glu-
canase activity on a linear scale gave r 2 values of0.188 and 0.372, respectively.
By changing the /J-glucanase values to their logarithms, r 2 values of 0.431 and
0.639 were obtained for weight gain and feed efficiency, respectively (Fig. 2).
Digesta yS-glucanase activity correlated well with digesta viscosity (r 2 =0.677, Fig.
3) and even better with the /?-glucanase activity in feed (r2 =0.948, slope 0.183,
Fig. 4).

3.8. Gut viscosity in relation to birdperformance

The slope of the relationship between digesta viscosity and liveweight gain was
2.13 (r 2

= 0.624) and between digesta viscosity and feed conversion efficiency
0.00267 (r2

= 0.616, Fig. 5). When correlations were calculated for the values
obtained with the lower enzyme concentrations (NE and AO treatments) r 2 val-
ues for weight gain was 0.566 and for feed efficiency 0.467. At the higher enzyme
concentrations (AO and AT treatments) the r 2 value for weight gain was 0.442
and for feed efficiency 0.352.

Fig. 2. Relationship between the logarithm of dietary activity and chick performance. O,
weight gain (r2 =0.431) X, feed efficiency (r 2 =0.639).



L'l- LSETOSAI9IFEXSS64

J. Inborr, M R. Bedford /Animal Feed Science and Technology 46 (1994) 179-196 191

Fig. 3. Relationship between yJ-glucanase activity in digesta and digesta viscosity: O, AT treatment
means (r 2 =0.677).

Fig. 4. Relationship between /3-glucanase activity in feed and digesta; +, AT treatment means
(r2

= 0.948).



3- LSETOS ANIFEXSS64

192 J. Inborr, M.R. Bedford /Animal Feed Science and Technology 46 (1994) 179-196

4. Discussion

Feed intake of the birds was not influenced by either enzyme supplementation
or conditioning temperature, which would suggest that digesta viscosity was not
high enough to influence the rate offeed passage (Salih et al., 1991; Bedford and
Classen, 1992). There was a linear effect of enzyme supplementation on feed
efficiency and liveweight gain. These responses are in accordance with those re-
ported by Bedford and Classen (1992) feeding rye- and wheat-based feeds to
broiler chickens. They found ahigh correlation between digesta viscosity and these
two performance parameters, indicating that intestinal viscosity has an impor-
tant role in determining the performance of broilers fed diets with high soluble
dietary fibre content. Increased viscosity is known to depress nutrient diffusion
in vitro (Fengler and Marquardt, 1988). This would be expected to reduce ab-
sorption in the small intestine, thus depressing chick performance. In the present
experiment, there was a relatively high correlation between digesta viscosity and
chick performance (Fig. 5), with slope values close to those reported by Bedford
and Classen (1992) feeding rye- and wheat-based diets to broiler chickens. The
results from the present experiment, wherebarley was used as the main source of
NSP, provide further evidence that it is not the chemical composition of the sol-

Fig. 5. Relationship between digesta viscosity and weight gain (x) and feed efficiency (O) in broiler
chickens fed a barley-based diet supplemented with Avizyme SX at 0, 1 and 10 g kg -1 and pelleted
after conditioning at 75, 85 and 95°C for either 0.5 or 15 min: X, weight gain (r 2=0.624); O, feed
efficiency (r 2 =0.616).



a- LSETOS ANIFEXSS64

J. Inborr, M R. Bedford /Animal Feed Science and Technology 46 (1994) 179-196 193

üble NSP fraction per se but the viscosity it creates in the lumen that causes the
deterioration in the performance of broiler chicks. The quadratic response in
weight gain and feed utilisation due to increasing conditioning temperatures in-
dicates that the negative effect of temperature had not been reached at 85 °C,
suggesting that the enzyme was stable up to this temperature as far as these two
parameters are concerned. Conditioning time did not influence any of the per-
formance parameters, suggesting that once the target temperature was achieved
it could be maintained for 15 min with no additional positive or negative effects.
This strongly indicates that conditioning temperature is ofgreat importance, with
little effect of time. Interestingly, conditioning at 85 °C resulted in the best per-
formance, lowest digesta viscosity values and lowest incidence of vent pasting of
birds fed the unsupplemented diets. It is possible that these processing conditions
were the most optimal, resulting in a combination of the highest degree of starch
gelatinisation and least degree of formation of Maillard products and resistant
starch.

Digesta viscosity of birds fed the unsupplemented diets showed a quadratic
response to increased conditioning temperatures, whereas there was a linear ef-
fect, and a tendency (/>=0.1059) towards a quadratic effect of temperature for
the diets with the lower concentration of the enzyme. These data would suggest
that the stability ofthe enzyme was better between 75 and 85°C thanbetween 85
and 95°C. At the highest level of enzyme inclusion there was a linear effect of
temperature, although values were low even at 95 °C. This would indicate partial
inactivation, but due to the overwhelming concentration of enzyme, there was
still sufficient activity even after conditioning at 95°C to maintain low digesta
viscosities. With conditioning at 75°C there was a quadratic response to enzyme
supplementation, indicating that increasing the enzyme concentration ten-fold
was of marginal benefit. At higher conditioning temperatures, enzyme supple-
mentation decreased digesta viscosity in a linear fashion, suggesting that the ben-
efits from increasing the enzyme concentration were more apparent.

Digesta viscosity and vent pasting values seemed to correspond well with each
other in terms of the response to enzyme and temperature. However, the inci-
dence of vent pasting was significantly increased when using a conditioning time
of 15 min compared with 30 s, whereas no such effect was observed for digesta
viscosity. It is possible that long-term conditioning causes components of the NSP
fraction to be solubilised further down in the GI tract of the bird, so that digesta
viscosity in the proximal part ofthe small intestine is not affected but water con-
tent and stickiness ofexcreta is increased. McCracken et al. (1992) reported both
decreased excreta dry matter content and increased digesta viscosity when feed-
ing broiler chickens with feeds pelleted after conditioning at 85°C for 15 min
compared with cold pelleted feeds. In this experiment, digesta from the entire
small intestine was collected for the viscosity measurements, which could explain
the discrepancy between this and the present experiment in terms ofdigesta vis-
cosity responses.

Apparent loss ofenzyme activity during pelleting has been reported for phytase
(Jongbloed and Kemme, 1990) and hemicellulases such as glucanase and xylan-



(j- ESETOS ANIFEXSS64

194 J. Inborr, MR. Bedford /AnimalFeed Science and Technology 46 (1994) 179-196

ase (Pickford, 1992). These results were based on in vitro measurements using
conventional enzyme activity assay methods. The results from the present exper-
iment also show significant loss ofyS-glucanase activity after pelleting when mea-
sured using the in vitro viscosimetric method. Despite this, enzyme supplemen-
tation significantly improved broiler performance, suggesting either that there
was still sufficient enzyme activity in the feed or that the in vitro assay method
underestimated the actual /i-glucanase activity. Conditioning the diets at 75 and
85 °C resulted in almost identical overall bird performance, but both liveweight
gain and feed utilisation were significantly worse after conditioning at 95°C be-
fore pelleting.

Dietary /?-glucanase correlated linearly with digesta /?-glucanase to a high de-
gree with a slope of 0.183 (Fig. 4). This means, for each unit increase ofdietary
/?-glucanase, the /?-glucanase activity in the upper small intestine increases by
0.183. By assuming the dry matter content of the digesta to be between 16 and
18% (Bedford et al., 1991), a five-fold dilution of the dietary /?-glucanase activity
is to be expected. This comes very close to the factor of0.183 and indicates more
or less full recovery of the measured activity of the feed in the small intestine.
Moreover, there was a high linear correlation between digesta /?-glucanase activ-
ity and digesta viscosity (Fig. 3), which in turn correlated linearly with both weight
gain and feed efficiency of the broiler chickens. Based on the above, one would
assume that dietary /i-glucanase wouldalso correlate with the chick performance
parameters. However, this was not the case until the dietary /?-glucanase values
had been transformed to their logarithms (Fig. 2). This decreases the discrimi-
natory power ofthis method compared with digesta viscosity.

It would appear, based on the relatively high correlation between /?-glucanase
activity in feed and digesta viscosity (Fig. 1), that the cellulose binding domain
(CBD) effect on interfering with in-feed analysis was not evident in this experi-
ment. It is possible that the recovery of/?-glucanase in the extraction was high and
the amount of enzyme extracted represents the amount of active /?-glucanase in
the feed. The binding power of the CBD is variable and that of some cellulases
of, for example Trichoderma ssp. can be quite low (Poole et al., 1991). Hence
this may allow a high recovery even after heat treatment.

There was a ten-fold higher activity measured in the AT than in the AO feeds,
indicating that the assay is linear over this range. The coefficient ofvariance (CV)
of the /?-glucanase recovery values decreased rapidly with increased levels of en-
zyme concentration of the diets. Thus, with the enzyme added at 1 g kg~‘, CV
decreased from 28.0 (unsupplemented diets) to 18.2% and further to 9.2% when
the enzyme was added at 10 g kg~'. This would indicate that the assay method
employed can be used for ‘in-feed’ analysis to measure /?-glucanase activity at
commercially used concentrations. It should be emphasised, however, thatabso-
lute /?-glucanase activity in feed is not necessarily a good predictor ofbroiler per-
formance, since plotting dietary /?-glucanase activity against liveweight gain and
feed efficiency coefficients gave very low r 2 values. By using the /?-glucanase val-
ues on a logarithmic scale, correlations increased considerably and the value for
feed efficiency was relatively high (0.651).



ar LSETOS AMIFEXSS64

J. Inborr, M R. Bedford /Animal Feed Scienceand Technology 46 (1994) 179-196 195

Pelleting diets for broiler chickens has usually resulted in improved perform-
ance (Inborr and Graham, 1991; Teigte et al., 1991), although it has been shown
that pelleting increases the proportion of soluble NSP in the ileal digesta of pigs
(Graham et al., 1988), which is likely to increase digesta viscosity and therefore
could affect chick performance (Choct and Annison, 1992). In the present ex-
periment, pelleting the basal diet at 75 °C numerically increased digesta viscosity
and there was an increased incidence of vent pasting with the long-term condi-
tioning. This may be explained by the increased solubility of/f-glucans observed
as a result of pelleting. Inborr and Graham (1991) observed a greater response
to enzyme supplementation ofpelleted than mash feeds when fed to broiler chick-
ens, suggesting that the enzymes had maintained their activity and a possible pos-
itive interaction between exogenous enzymes and pelleting. Similar results were
reported by Teigte et al. (1991) who found that both pelleting and micronisation
ofrye increased the chick performance response to pentosanase supplementation
compared with untreated rye, indicating that heat treatment renders the anti-
nutritive substances in rye more accessible to the enzyme.

Based on these results it can be concluded that the higher the conditioning tem-
perature and the longer the conditioning time, the lower the enzyme recovery and
apparent activity when measured by in vitro methods. Conditioning at 85°C be-
fore pelleting did not reduce enzyme activity compared with 75 °C based on bird
performance, whereas significant inactivation occurred at 95°C. The high corre-
lation between activity in feed and digesta viscosity on the one hand,
and gut viscosity and bird performance on the other hand, suggests that both the
in vitro and in vivo methods employed in this study give a good estimate of en-
zyme survivability through pelleting.

5. Acknowledgements

The authors would like to thank John Bouwman and Willem van de Veen,
Trouw International BV, Putten, Netherlands, for arranging and carrying out the
pelleting of the experimental feeds and Dr. Walter Michie at the Scottish Agri-
cultural College in Aberdeen (Craibstone Estate), Scotland, for supervising the
broiler performance experiment. The /J-glucanase activity measurements (feed
samples) carried out by Jari Puhakka at the Cultor Technology Centre, and by
Dr. Peter Hotten and Yvonne Turnbull at Rowett Research Services, Aberdeen,
Scotland (digesta samples) are highly appreciated. This experiment was funded
by Finnfeeds International, Wiltshire, UK.

6. References

Änian, P. and Graham, H., 1987. Analysis of total and insoluble raixed-linked (1-3), (1-4)- /?-D-glu-
cans in barley and oats. J. Agric. Food Chem., 35; 704-709.



LSETOS ANITEXSS64

196 J. Inborr, M.R. Bedford /AnimalFeed Science and Technology 46 (1994) 179-196

Äman, P. and Hesselman, K., 1984. Analysis ofstarch and other main constituents of cereal grains.
Swed. J. Agric. Res., 14: 135-142.

Association of Official Analytical Chemists, 1990. Official Methods ofAnalysis, 15th edn. AOAC,
Washington, DC.

Atkinson, T., Fowler, V.R., Garten, G.A. and Lough, A.K., 1972.A rapid method for accurate deter-
mination of lipid. Analyst, 97: 562-568.

Bathgate, G.N., 1979.The determination of endo-/Fglucanase activity in malt. J. Inst. Brew., 85: 92-
94.

Bedford, M.R. and Classen, H., 1992. Reduction of intestinal viscosity through manipulation of die-
tary rye and pentosanase concentration is effected through changes in the carbohydrate composi-
tion of the intestinal aqueous phase and results in improved growth rate and feed conversion. J.
Nutr., 122; 560-569.

Bedford, M.R., Classen, H.L. and Campbell, G.L., 1991.The effect ofpelleting, salt, and pentosanase
on the viscosity of intestinal contents and the performance of broilers fed rye. Poult. Sci., 70:
1571-1577.

Chesson, A., 1987. Supplementary enzymes to improve the utilisation ofpig and poultry diets. In: W.
Haresign and D.J.A. Cole (Editors), Recent Advances in Animal Nutrition. Butterworths, Lon-
don, pp. 71-89.

Choct, M. and Annison, G., 1992. Anti-nutritive effect ofwheat pentosans in broiler chickens: roles
of viscosity and gut microflora.Br. Poult. Sci., 33: 821-834.

Dierick, N.A., 1989. Biotechnology aids to improve feed and feed digestion: enzymes and fermenta-
tion. Arch. Anim. Nutr. Berlin, 39; 241-261.

Fengler, A.S. and Marquardt, R.R., 1988. Water-soluble pentosans from rye: 11. Effects on rate of
dialysis and on the retention ofnutrients by the chick. Cereal Chem., 65: 298-302.

Godfrey, T, and Reichelt, J. (Editors), 1983. Industrial Enzymology. The Application ofEnzymes in
Industry. Macmillan Publishers, Chippenham, UK, 582 pp.

Graham, H., Fadel, J.G., Newman, C.W. and Newman, R.K., 1988. Effect of pelleting and /i-glucan-
ase supplementation on the ileal and faecal digestibility of a barley-based diet in the pig. J. Anim.
Sci., 67:1293-1298.

Inborr, J., Bedford, M.R.and Graham, H., 1993.Stability and mode ofaction ofpoultry feed enzymes
in diets based on wheat and barley. Proc. Australian Poultry Science Symp., The WPSA (Austra-
lian Branch), University ofSydney, Sydney, N.S.W., Vol. 5, pp. 53-56.

Inborr, J. and Graham, H., 1991. Effect of enzyme supplementation ofwheat-based diets on the per-
formance of broiler chickens. Proc. Australian Poultry Science Symp., The WPSA (Australian
Branch), University of Sydney, Sydney, N.S.W., pp. 50-55.

Jongbloed, A.W. and Kemme, P.A., 1990. Effect ofpelleting mixed feeds on phytase activity and the
apparent absorbability ofphosphorusand calcium in pigs. Anim.Feed Sci. Technol., 28; 233-242.

Ministry ofAgriculture, Fisheries and Food, 1985.Analysis ofAgricultural Materials. Reference Book
427, MAFF, p. 36.

McCracken, K.J., Urquharl, R. and Bedford, M.R., 1993. Effect ofheat treatment and enzyme sup-
plementation ofbarley-based diets on performance ofbroiler chickens. Proc. Nutr. Soc., in press.

Pickford, J.R., 1992. Effects of processing on the stability ofheat labile nutrients in animal feeds. In:
P.C. Gamsworthy, W. Haresign and D.J.A. Cole (Editors), Recent Advances in Animal Nutri-
tion. Butterworth-Heinemann, Oxford, pp. 177-192.

Poole, D.M., Durrani, A.J., Hazlewood, G.P. and Gilbert, H.J., 1991.Characterisation ofhybrid pro-
teins consisting of the catalytic domains of Clostridium and Ruminococcus endoglucanases, fused
Pseudomonas non-catalytic cellulose-binding domains. Biochem. J., 279: 787-792.

Salih, M.E., Classen, H.L. and Campbell, G.L., 1991. Response of chickens fed hulless barley to die-
tary /?-glucanase at different ages. Anim.Feed Sci. Technol., 25: 193-200.

Statistical Analysis Systems Institute, 1982. SAS Users Guide: Statistics. SAS Institute, Cary, NC.
Teigte, D.A., Campbell, G.L., Classen, H.L. and Thacker, P.A., 1991. Heat pretreatmenl as a means

of improving the response to dietary pentosanase in chicks fed rye. Can. J. Anim. Sci., 71: 507-
513.



Agric. Sei. Fint. 2 (1993)

Stability of feed enzymes in physiological conditions assayed by
in vitro methods

Johan Inborr and Anne Grönlund

Inborr, J. & Grönlund, A. 1993. Stability of feed enzymes in physiological condi-
tions assayed by in vitro methods. Agric. Sci. Finl. 2: 125-132. (Finnfeeds Interna-
tional Ltd., Market House, High Street, Marlborough, Wiltshire SNB IAA, United
Kingdom and Cultor Ltd., Technology Centre, FIN-02640 Kantvik, Finland.)

A series of in vitro incubations were carried out to investigate the stability of two
enzyme preparations in conditions similar to those in the upper gastrointestinal tract of
monogastric animals. The two enzyme products, one crude xylanase from Trichoderrm
longibrachiatum (Multifekt K) and the other a specifically manufactured feed enzyme
(Avizyme SX®), were subjected to incubations at low and neutral pH with and without
proteolytic enzymes (pepsin and pancreatin). Wheat gluten was employed together with
the crude xylanase to investigate its potential as a stabilising agent. Due to the buffering
effect of Avizyme SX®, incubations were carried out with (pH 2.5) and without (pH 3.2)
addition of either citric or hydrochloric acid. Incubation of the crude xylanase at low pH
followed by incubation at neutral pH resulted in negligible loss of xylanase activity
whereas P-xylosidase recovery fell to 57 per centof the initial value (P<0.05). Addition
of wheat gluten resulted in full recovery of P-xylosidase. The recoveries of both
P-glucanase and xylanase were significantly (P<0.05) lower than the initial values after
incubation of Avizyme SX® in pH 2.5. However, with no pH adjustment (pH 3.2) the
recoveries were significantly higher (P<0.05 for P-glucanase and P<o.lo for xylanase).
The results from the pepsin and pancreatin incubations showed similar trends as the
ones of the pH stability experiments. Consequently, gluten addition and no pH adjust-
ment gave the highest enzyme activity recoveries.

The results suggest that partial enzyme inactivation may occur due to the low pH and
proteolytic activities and hence in the GI-tract of monogastrics. Feeds and feedstuffs can
due to their buffering capacity and possibly by providing substrates for the enzymes
markedly reduce the rate of inactivation. Results from a number of pig and poultry
experiments appear to support this assumption. In vivo recovery measurements using
animal models are needed to substantiate this.

Key words: xylanase, P-glucanase, P-xylosidase, wheat gluten, enzyme survivability,
pH, pepsin, pancreatin, Multifect K, Avizyme SX®.

Introduction

Since the early 1980’s the use ofenzymes in animal
feeds has received much attention both in the aca-
demic and commercial worlds. There is now a con-
siderable amount of published work showing the

response in animal performance to enzyme supple-
mentation of feeds, which has been extensively
reviewed by Chesson (1987) and Dierick (1990).
With increased commercial use of feed enzymes
stability has become a frequently debated issue. In
particular, enzyme stability during feed manufac-

125

Research Note



Agric. Sd. Finl. 2(1993)

hiring, due to the different heat treatments (pellet-
ing, expansion, extrusion etc.) now being employed
in the feed industry, has caused some concern. Fur-
thermore, once the enzymes enter into the gastroin-
testinal tract, they are subjected to a range ofcondi-
tions that can cause inactivation. Firstly, gastric pH
may be as low as 1, but usually around 2 in an
empty stomach, with rapid increase to between 4
and 5 as ingested food arrives (Kidder and Man-
ners 1978). Keeping in mind that enzymes are
proteins and, therefore, subject to proteolysis, low
pH in combination withpepsin would appear to be
a potential hazard for the enzyme molecules. Sec-
ondly, the pancreatic proteases e.g. tiypsin and chy-
motrypsin represent a major threat to the enzymes
due to their abundance and high activity in duode-
num.

Cellulases and hemicellulases are the most com-
monly used enzymes in feed applications. These are
mainly microbial enzymes produced by different
fungi and bacteria. In general, the inherent heat
stability of mesophilic microbial enzymes is poor,
although there are considerable differences be-
tween enzymes from different sources in this re-
spect (Godfrey and Reichelt 1983). Through
modem biotechnology a number of stabilisation
methods have been developed. For example, it is a
known fact that substrates and inhibitors that bind
specifically to the native conformation of an en-
zyme will increase the stability (Pace 1990). This
method of stabilisation is called immobilisation and
has been shown by Simons and Georgatsos

(1990) to decrease heat inactivation of barley (3-
glucosidase. Other promising stabilisation methods
involve adding crosslinks (disulfide bonds) to, sub-
stituting amino acids (Pace 1990) and crystal-
lisation of the enzyme proteins (Sx. Clair and
Navia 1992). Furthermore, De Cordt et al. (1992)
found that polyols and carbohydrates are powerful
stabilisers for both dissolved and immobilised B.
licheniformis a-amylase.

A series of in vitro experiments were designed to
investigate the survivability ofa few main activities
of two enzymes subjected to conditions similar to
those prevalent in the digestive tract of pigs and
poultry. Two enzyme preparations, one crude and
one specifically manufactured feed enzyme prod-

uct, were employed. Xylanase and B-xylosidase ac-
tivities of the crude enzyme preparation and 6-glu-
canase and xylanase of the feed enzyme product
were investigated.

Material and methods

A liquid crude xylanase preparation (Multifect K,
Genencor International Ltd., Helsinki, Finland)
from a Trichoderma longihrachiatum fermentation
and a stabilised(liquid enzymes adsorbed and dried
onto a wheat-based carrier; Patent FI 77359) com-
mercial feed enzyme product (Avizyme SX ,

Finnfeeds International Ltd., Marlborough, Wilt-
shire, United Kingdom) containing (3-glucanase
and xylanase from Trichoderma longihrachiatum
were subjected to treatments in conditions similar
to those in the stomach (proventriculus) and small
intestine of pigs and poultry. A schematic descrip-
tions of the incubation procedure employed is pre-
sented in Figure 1.

Incubation in low and neutral pH

The conditions employed in these incubations were
identical with those described below with the ex-
ception that no pepsin and pancreatin were used.

Incubation with pepsin and pancreatin

Crude xylanasepreparation

20 ml ofpepsin solution (Merck 7189), containing
0.25 protease units (haemoglobin, pH 1.2, 25°C)
per ml, in 0.1 M Mcllvaine’s buffer (pH 2.5) was
equilibrated at 40°C. 10 ml of xylanase solution in
Mcllvaine’s buffer (1 ml xylanase/50 ml buffer)
was added and the solution diluted with buffer to 50
ml followed by incubation at 40°C, pH 2.5 for 30
minutes. The reaction was stopped in an ice/water
bath immediately after incubation.pH was adjusted

126

Research Note



Agric. Sei. Fin!. 2 (1993)

with 0.5 M NaaHPOa to 5.0 and the final volume
was 100 ml. Xylanase and (3-xylosidase activities
were measured. The pepsin treated sample (10 ml)
was then adjusted to pH 7.0 with 0.5 M Na2HPC>4.
30 ml ofpancreatin solution (Merck7230), contain-
ing 30 protease units per ml, in Mcllvaine’s buffer
(pH 7.0), was added, and the solution diluted to 50
ml and adjusted to pH 7.0 with Mcllvaine’s buffer.

The solution was incubated at40°C for 30 minutes,
after which time the reaction was stopped in an
ice/water bath. The pH was adjusted to 5.0 by 0.5
M citric acid and xylanase and (3-xylosidase activ-
ities measured.

A parallel incubation was carried out with wheat
gluten (2% w/v) solubilised in a NaCl solution
(0.9% w/v) as a stabiliser. When wheat gluten was

Fig. I. Flow chart diagramme of the incubations.

127

Research Note



Agric. Sei. Finl. 2 (1993)

used, samples were filtered through fibre glass filter
paper before enzyme activity analysis.

Commercialfeed enzyme

Two proteolysis procedures were carried out. In the
first one, the pH was kept constant at 2.5 by addi-
tion of 6 N HCI, whereas in the second one no acid
was added during the pepsin treatment, and the pH
rose to 3.2 (buffering effect of the feed enzyme
product).

20 g feed enzyme premix was suspended in 0.1
M Mcllvaine’s buffer (pH 2.5) and the pH was
either maintained at 2.5 by acid (6 N HCI) addition
or allowed to rise due to the buffering capacity of
the feed enzyme product (parallel incubation). 10
ml pepsin solution was added to give a final volume
of 70 ml. The pH was adjusted after incubation to
5.0 using 2.5 M NaOH and samples were taken for
enzyme activity analysis (P-glucanase and xyla-
nase) after filtration through fibre glass filter paper.

The procedure continued with 70 ml unfiltered
pepsin-treated sample, employing the same condi-
tions as with the crude xylanase preparation but
without addition of wheat gluten. 15 ml pancreatin
solution in 0.1 ml Mcllvaine’s buffer (pH 7.0) was
added giving a volume of 85 ml during the incuba-
tion. After the pH adjustment (15 ml 2 M citric
acid) the final volume was 100ml. P-glucanase and
xylanase activities were measured after filtration.

of reducing sugars (expressed as glucose equival-
ents) in one minute under the conditions described.

Xylanase (EC 3.2.1.8; endo-p-(l,4)-xylanase)
activity (modified from Poutanen et al. 1986) was
determined using 1% oat spelt xylan (Sigma X-
-0376) in Mcllvaine’s buffer at pH 5.0 as substrate.
0.2 ml of suitable enzyme dilution was incubated
with 2.0 ml of substrate solution at 40°C for 30
minutes. The reducing sugars were assayed as de-
scribed above. One xylanase unit is the amount of
enzyme that liberates 1 pmol of reducing sugars
(expressed as xylose equivalents) in one minute
under the conditions described.

p-xylosidase (EC 3.2.1.37; 1,4-p-D-xylan xylo-
hydrolase) activity was determined (modified from
Deleyn et al. 1978) using 2 mmol/1 p-nitrophenyl-
(3-D-xylopyranoside in Mcllvaine’s buffer at pH
5.0 as substrate. 0.2 ml of suitable enzyme dilution
in deionized water was incubated with 2.0 ml sub-
strate solution in 40°C at pH 5.0 for 30 minutes.
After adding 1.0 ml of NazCOs, absorbance was
measured at 400 nm. One p-xylosidasc unit is the
amount of enzyme that liberates one pmol of p-ni-
trophenol in one minute under the conditions de-
scribed.

Each enzyme activity measurement was carried
out in duplicates or triplicates unless otherwise
stated. Values are expressed as means with standard
deviations of each sample. Means were separated
where appropriate by paired t-test.

Enzyme activity analysis

p-glucanase (EC 3.2.1.6; endo-P-(l,3)(l,4)-gluca-
nase) activity was determined
cally using 1.0% barley P-glucan (Biocon Bio-
chemicals Ltd., Ireland) in Mcllvaine’sbuffer at pH
5.0 as substrate. 0.2 ml of suitable enzyme dilution
in deionized water was incubated with 2.0 ml of
substrate solution at 40°C for 30 minutes. Reducing
sugars were assayed by additiön of 3.0 ml 3,5-dini-
trosalisylic acid (DNS) reagent (Sumner and
Somers 1949), boiling for 5 minutes, cooling and
measuring absorbance at 540 nm. One p-glucanase
unit is the amount of enzyme that liberates 1 pmol

Results and discussion

pH stability

Incubation in pH 2.5 did not reduce the xylanase
activity of Multifect K, whereas the p-xylosidase
activity was reduced to 57% (P<0.05) of the initial
value (Table 1). Subsequent incubation at pH 7
only marginally reduced xylanase activity and had
no effect on P-xylosidase. Addition ofwheat gluten
markedly increased the recovery of p-xylosidase
after incubation in pH 2.5 (P<0.05) and pH 7
(P<0.10), resulting in full recovery of this activity.

Maintaining pH at 2.5 by addition of citric acid

128

Research Note



Agric. Sei. Finl. 2 (1993)

Table I . Initial and residual xylanase and p-xylosidase activities of Multifect K, measured at pH 5, after incubation at acidic
and neutral pH with and without pepsin and pancreatin, and with (+ ) and without (-) gluten (means ± sd).

Xylanase (3-xylosidase

Gluten + - +

U/g % U/g % U/g % U/g %

Initial pH 5 4510 ± 80 100 4970 + 27 100 44 + 3 100 30 ± 2 100
Residual pH 2.5 4440* 98 4780' 96 25' +1 57 30» ± 1 100

pH 7 4070 ± 235 90 4720* 95 25» +1 57 30" ± 2 100
Pepsin pH 2.5 4270* 95 4520* 91 30> ± 2 68 18»+ 0 60
Pancreatin pH 7 2590» ±4 57 3400* 68 25 ± 5 57 29 ± 2 97

’ result of one measurement
x differ from initial value (P<0.05)
» differ from initial value (P<0.10)
■ differ from corresponding control ( + ) (P<0.05)
b differ from corresponding control ( + ) (P<0.10)

Table 2. Initial and residual P-glucanase and xylanase activities of Avizyme SX, measured at pH 5, after incubation at
acidic and neutral pH with and without pepsin and pancreatin, and with (!) and without (-) pH adjustment (means 1 sd).

P-glucanase Xylanase

pH adjustment l + +

U/g % U/g % U/g "o U/g %

Initial pH 5 870 ± 5 100 850 ± 15 100 460 + 16 100 467 ± 38 100
Residual pH 2.5/3.2 370" ± 6 43 830" ±2B 96 340? ± 6 75 510b ±2O 109

pH 7 390» ±8 45 790» ± 28 92 260" ±2 57 422» ±6 90
Pepsin pH 2.5/3.2 370" ± 17 42 850» ± 13 100 360* ± 3 78 510" ± 18 110
Pancreatin pH 7 310» ±9 36 760» ±3O 89 260" ± 12 56 470" ± 21 100

1 ( + ) means pH kept constant at 2.5 by addition of either 2 M citric acid (pH stability) or 6 N HCI (pepsin stability), (-)

means no acid addition during the incubation with pH rising to 3.2.
x differ from initial value (P<0.05)
» differ from initial value (PcO.lO)
' differ from pH 2.5 (P<0.05)
• differ from corresponding control ( +) (P<0.05)
b differ from corresponding control ( +) (P<0.10)

resulted in 43% (Pc0.05) recovery of p-glucanase
and 75% recovery of xylanase (PcO.10) in
Avizyme SX (Table 2). Subsequent incubation at
pH 7 had no effect on the P-glucanase activity,
whereas xylanase was reduced to 57% (Pc0.05) of
the initial activity. When pH was not adjusted, lead-
ing to a 0.7 unit increase in pH, p-glucanase recov-
ery after the first incubation (pepsin in pH 3.2) was
96% (P>0.05) and after the second incubation (pan-

creatin in pH 7) 92%.The corresponding recoveries
for xylanase were 109 and 90%. This difference
was not significant (P>0.10).

From these results it appears that low pH per se
did not cause any dramatic reductions in enzyme
activity and that addition of wheat gluten may re-
duce the degree of inactivation of more sensitive
enzymes i.e. in this case p-xylosidase. Whether this
effect is due to an increased dry matter content of

129

Research Note



Agric. Sei. Finl. 2(1993)

the system or e.g. binding between the gluten and
the enzyme is not known. Interestingly, with wheat
being the carrier material of Avizyme SX thus
providing the system with gluten, xylanase recover-
ies were lower with and higher without pH adjust-
ment than when gluten was added to MultifectK.

Pepsin and pancreatin stability

Xylanase and P-xylosidase recoveries after incuba-
tion of Multifect K with pepsin was 95 and 68%
withoutand 91 and 60% withaddedgluten, respect-
ively (Table 1). These effects were not significant
(P>0.10), with the exception being P-xylosidase
when incubated at pH 2.5 with added gluten
(P<0.10). After the subsequent incubation in pan-
creatin the corresponding recoveries were 57 and
57% and 68 and 97%, respectively (P>0.10). Thus
p-xylosidase seemed to be more sensitive to pepsin
than xylanase. Incubation with pancreatin tended to
further reduce enzyme activity, except the activity
of P-xylosidase when gluten was added.

P-glucanase activity ofAvizyme SX® decreased
to 42% (P<0.05) of the initial valueafter incubation
in pepsin with pH kept at 2.5 by addition of hydro-
chloric acid and was further reduced to 36% after
incubation in pancreatin (Table 2). With no pH
adjustment, the P-glucanase recoveries after pepsin
treatment was 100% and after incubation with pan-
creatin 89% (P>0.10). These recoveries were sig-
nificantly higher than when pH was kept at 2.5
(P<0.05). Xylanase recoveries after pepsin and
pancreatin incubations withpH adjustment were 78
and 56% of the initial activity (P<0.10), respect-
ively. Without pH adjustment xylanase was fully
recovered. These recoveries were significantly

higher than withpH adjustment (P<0.10).
Addition of wheat gluten to the enzyme/substrate

systems tended to increase the final enzyme recov-
eries but this effect was not observed at pH 2.5,
except for P-xylosidase after incubation at pH 2.5
withoutpepsin. De Cordt et al. (1992) found that
both polyols and carbohydrates increased the tem-
perature stability of bacterial a-amylase. The
mechanisms involved were not specifically eluci-
dated but "preferential protein hydration", changes
in the chemical potential of the proteins, changes in
the solvent dielectric constant, changes in the water
activity of the solvent system, degree of water or-
ganisation were suggested. It is possible that wheat
gluten used in this present experiment exerted one
of these effects leading to increased pH and prote-
olytic stability. Interestingly, the buffering capacity
of the feed enzyme product, probably due to the
cereal carrier material, had a clear stabilising effect
on the enzyme activities measured. Feed arriving in
the stomach of pigs exert similar pH buffering ef-
fects (Kidder and Manners 1978) and it can
therefore be assumed that the feed acts as a potent
stabilizer in the animal.

Results from a number of experiments withpigs
and poultry would suggest that sufficientactivity of
supplementary enzymes survive the potential haz-
ards of the Gl-tract to improve animal performance
(Chesson 1987,Dierick 1990). Enzyme recovery
measurements in different segments of the Gl-tract
would be needed to exactly establish the rate and
extent of feed enzyme inactivation in vivo.

Acknowledgements. The authors want to thank Ms Päivi
Rantalaiho and Mrs Jaana Oksanen at the Cultor Ltd. Tech-
nology Centre for technical assistance during these investi-
gations. This work was funded by Finnfeeds International
Ltd., Marlborough, Wiltshire, U.K.

References

Chesson, A. 1987. Supplementary enzymes to improve the
utilisation of pig and poultry diets. In: Haresign W. &

Cole D.J.A. (eds.). Recent Advances in Animal Nutrition
1987. Butterworths, London, p. 71-89.

Dierick, N.A. 1990. Biotechnology aids to improve feed and
feed digestion; enzymes and fermentation. Arch. Anim.

Nutr. Berlin 39: 241-261.
De Cordt, S., Sara via, J., Hendrickx, M., Maesmans, G. &

Tobback, P, 1992. Changing the thermostability ofBacil-
lus licheniformis ot-amylase. Proc. Int. Symp. on ’Stabil-
ity and stabilisation of enzymes’. Maastricht, The Neth-
erlands. Poster.

130

Research Note



Agric. Sei. Fint. 2 (1993)

Deleyn, F., Claeyssens M, van Beeumen, J. & De Bruyne,
C.K. 1978. Purification and properties of P-xylosidase
from Penicillium wortmanni. Can. J. Biochem. 56:4350.

Godfrey, T. & Reichelt, J. 1983. Industrial enzymology.
582 p. Stockton Press, New York, NY.

Kidder, D.E. & Manners, M.J. 1978. Digestion in the pig.
204 p. Scientechnica, Bristol, U.K.

Pace, C.N, 1990. Measuring and increasing protein stability.
Tibtech 8: 93-98.

Poutanen, K., Puls, J. & Linko, M. 1986. The hydrolysis of
steamed birchwood hemicellulose by enzymes produced
by Trichoderma reesei and Aspergillus awamori. Appi.
Microbiol. Biotechnol. 28: 419-424.

Simons, G. & Georoatsos, J.G. 1990. Immobilisation of
barley p-glucosidase on solid supports - yields and prop-
erties. Appi. Microbiol. Biotechnol. 33: 51-53.

St. Clair, N.L. & Navia, M.A. 1992. Crosslinked enzyme
crystals as robust biocatalysts. J. Amer. Chem. Soc. 114:
7314-7316.

Sumner, J.B. & Somers, G.F. 1949. Dinitrosalicylic acid for

glucose. In: Laboratory experiments in biological chem-
istry, 2nd ed„ Academic Press, New York. p. 3839.

Manuscriptreceived February 1993

Johan Inborr
Finnfeeds International Ltd.
Market House, High Street
Marlborough, Wiltshire SNS IAA
United Kingdom

Present address:
Finnsugar Gmbh
Raboisen 58
D-20095 Hamburg 1,Germany

Anne Grönlund
CultorLtd.,Technology Centre
FIN-02640 Kantvik, Finland

131

Research Note



Agric. Sd. Finl. 2(1993)

SELOSTUS

Rehuentsyymien stabiilisuus fysiologisissa olosuhteissa in vitro -menetelmin mitattuna

Johan Inborr ja Anne Grönlund

Finnfeeds International Ltd. ja Cultor Ltd.

Kahden entsyymipreparaatin, Trichoderma longihrachiatu-
min tuottaman ksylanaasin (Multifekt K) ja rehuentsyymi-
tuotteen (Avizyme SX®), stabiilisuutta tutkittiin in vitro -me-
netelmin sellaisissa olosuhteissa, jotkavallitsevat yksimahais-
ten ruoansulatuskanavan yläosissa. Ksylanaasipreparaatista
tarkasteltiin ksylanaasi- ja P-ksylosidaasiaktiivisuuksien-,
sekä Avizyme SX®:stä fi-glukanaasi- ja ksylanaasiaktiivi-
suuksien stabiilisuutta. Aluksi tutkituin entsyymiaktiivisuuk-
sien stabiilisuutta eri pH:ssa (2,5 ja7) ja sen jälkeen pepsiinin
ja pankreatiinin vaikutusta entsyymiaktiivisuuksiin. Tutki-
muksessa selvitettiin myös vehnägluteenin entsyymiaktiivi-
suuksia stabiloivaa vaikutusta.

Inkubointi pH 2,s:ssä ja sen jälkeen pH7:ssäei vaikuttanut
merkitsevästi Multifekt K:n ksylanaasiaktiivisuuteen. Sen si-
jaan p-ksylosidaasiaktiivisuus aleni 57 %:iin alkutasosta
(P<0.05) ilman vehnägluteenilisäystä. Vehnägluteenilisäyk-
sellä p-ksylosidaasiaktiivisuus pysyi aikatasolla. Avizyme
SX®:n P-glukanaasi-ja ksylanaasiaktiivisuus aleni merkitse-
västi kun sitruunahappoa lisättiin inkuboinneissa. Ilman hap-
polisäystä enstyymiaktiivsuudet pysyivät aikatasoilla.

Inkubointi pepsiinillä japankreatiinilla alensi merkitsevästi
(P<0.05) Multifekt K:n ksylanaasiaktiivisuutta. P-ksylosi-
daasi aleni 68 %:iin pepsiini-inkuboinnin jälkeen (P<0.10) ja
57 %:iin pankreatiini-inkuboinnin jälkeen (P>0.10). Pepsii-
niinkuboinnissa vehnägluteenilisäys ei parantanut stabiili-
suutta, mutta pankreatiini-inkuboinnissa vehnägluteenin lisä-
ys'nosti |3-ksylosidaasiaktiivisuuden 97 %:iin alkutasosta.
Avizyme SX*:n pepsiini- japankreatiini-inkuboinnit ja hap-
polisäys alensivat merkitsevästi (P<0.05) (3-glukanaasi- ja
ksylanaasiaktiivisuutta. Ilman happolisäystä aktiivisuudet py-
syivat alkuperäisillä tasoilla.

Tulokset osoittavat, että entsyymien osittainen inaktivoitu-
minen voi tapahtua matalassa pH:ssa ja proteolyyttisissä olo-
suhteissa. Avizyme SX®:llä javehnägluteenilisäyksellä saatu-
jen tulosten perusteella on kuitenkin odotettavaa, että rehun
mahalaukun sisältöä puskuroiva vaikutus suojaa rehuun lisät-
tyjä entsyymejä inaktivoitumiselta suhteellisen tehokkaasti.
Rehuun lisättyjen entsyymien todellisen inaktivoitumisasteen
määrittämiseksi joudutaan kuitenkin suorittamaan vastaavat
mittaukset kohde-eläimiä käyttämällä.

132

Research Note



Nutritional implications of feeding enzyme-treated wheat bran

to pigs

1. Recovery of added enzyme activities in the stomach and terminal

ileum

BY JOHAN INBORR 12’, JAN VAN DER MEULEN 1 AND JARI PUHAKKA 5

'DLO Research Institute for Livestock Feeding and Nutrition (IWO-DLO),

P.O. Box 160, NL-8200AD Lelystad, The Netherlands

2Finnfeeds International Ltd., Market House, High Street, Marlborough,

Wiltshire SNB IAA, United Kingdom

3Cultor Ltd. Technology Centre, SF-02460 Kantvik, Finland

Present address: Svenska Foder Ab, Kinnegatan 23, Box 673, S-53116 Lidköping, Sweden



2

Five crossbred barrows of approximately 31 kg live weight were fitted with a stomach

and a post valve T-caecum cannula and fed five experimental diets for five two-week

periods in a 5 x 5 Latin Square design. The diets contained wheat bran as the only

source of non-starch polysaccharides (NSP). The wheatbran of the control diet (C) was

incubated with a watenacetic acid mixture at 39°C and pH 5.0 for 3.5 hours. For

treatments Cel-I and Xyl-I the wheat bran was incubated in the same conditions with

either an added crude celluiase or a xylanase preparation. Immediately before feeding,

wheat bran treated in the same way as C was supplemented with either the celluiase or

xylanase preparation to give treatments Cel-A and Xyl-A, respectively. The pigs were

fed twice daily and gastric and ileal digesta collected on separate days between days 7
and 14in 2-hour intervals for 12 hours. Samples collected between 0 and 6 hours alter

feeding were analysed for xylanase and B-glucanase activities using either standard or

modified (according to sample pH and physiological temperature) assay methods to
estimate the degree of survivability. Immediately after feeding both the xylanase and /3-
glucanase activities in the stomach of pigs fed the Cel-diets were significantly (P<0.05)

higher than the controls. At 4 hours after feeding only Cel-A had higher ft-glucanase

activity (P<0.05) than the control pigs. In the gastric samples the modified method gave
significantly (P =0.001) lower xylanase values than the standard method. Between 2 and

6 hours after feeding the xylanase activity of ileal samples of pigs fed the xylanase-

treated wheat bran were significantly (P<0.05) higher than the controls, whereas /?-

glucanase activity in samples obtained from the pigs fed the cellulase-treated diets were

higher than the controls between 4 and 6 hours after feeding. In the ileal samples, the

modified method gave significantly higher xylanase and lower/?-glucanase activity values

than the standard method (P=0.001). It is concluded that xylanase and /I-glucanase

activities can be measured in digesta samples with the methods employed and that the

added enzymes were largely inactivated between the feedings. Measuring digesta enzyme
activities in conditions similar to the ones in the GI-tract may give a more reliable

picture of the actual activity of added enzymes.

INDEXING KEY WORDS: wheat bran, enzyme treatment, xylanase, celluiase, p-
glucanase, assay method, enzyme recovery, gastric and ileal cannula, dietary fibre, NSP.



3

Dietary fibre (DF) is composed of a very heterogenous mixture of substances, mainly

associated with plant cell walls, which may be defined as non-starch polysaccharides (NSP)

and lignin. Due to its complex physical and chemical characteristics, DF influences many

processes and reactions in the digestive system of the pig, thus influencing the utilisation of

feed (Low, 1985).

The main DF components of wheat and wheat by-products are the so called pentosans or

arabinoxylans (Annison et al., 1992). In wheat bran, the content of arabinoxylans (the sum

of the pentose sugars arabinose and xylose) can vary. Graham et al. (1986a) found 20.6 per

cent (48 per cent of DF) and Bach Knudsen & Hansen (1991) 26.2 per cent (52 per cent of

DF) arabinoxylans on a dry matter basis in wheat bran. Thus depending on the origin and

milling process, the DF content and chemical composition of wheat bran may vary

considerably.

Since the pig does not produce enzymes capable of degrading dietary fibre components it

has to rely on the microbial fermentation in the gastrointestinal tract to extract energy from

high fibre feedsluffs such as wheat bran. Treating wheat bran with cell wall degrading

enzymes before feeding or supplementation of the feeds with such enzymes may increase its

digestibility and hence utilisation. Neudoerffer and Smith (1969) reported increased protein
digestibility and performance of rats after cellulase and subtilisin treatment of wheat bran.

In an experiment with broiler chickens, supplementation of a diet containing 20 per cent

wheatbran with a Trichoderma viride cellulase significantly increased the digestibility of cell

wall components and tended to improve feed utilisation (Nahm and Carlson, 1985). In a

study by McCleanand McCracken (1992) enzyme supplementation of a wheat feed-based diet

did not influence ileal or faecal dry matter, crude protein and energy digestibility, whereas

Schulte el al. (1990) reported increased feacal digestibility of crude fibre, protein and fat due
to supplementation of wheat bran with a commercial cellulolytic enzyme preparation. Thus
there appear to be an opportunity for cell wall degrading enzymes to improve the nutritional

value of wheal bran also for pigs.

Enzyme activity assays are usually carried out according to standardised procedures. The

conditions of an assay (pH, temperature, substratre concentration etc.) are in general

reflecting the conditions of the application in which the enzyme is used or determined by the

optimum conditions for the enzyme itself. This can limit the applicability of the enzyme assay
or possibly give a false picture of how an enzyme will perform in other applications.



4

Modification of current standard assay methods may be necessary in order to estimate the

actual activity in different sections of the GI-tract and possibly explain the response in

nutrient digestibility and performance of the animals due to enzyme supplementation.

The aims of this study were to investigate the nutritional implications of treating wheat bran

with fibre degrading enzymes on the digestibility of nutrients along the digestive tract and

other physiological parameters including pH, osmolality and VFA concentration. In addition,
recoveries of two dietary enzyme activities in the stomach and ileum were measured. Two

enzyme preparations were employed either for pre-treatment of the wheat bran or to be added

to the diets immediately prior to feeding.

In this first paper the results of the enzyme recovery measurements, using two (either a

standard or a modified) enzyme assays, are presented.

MATERIALS AND METHODS

Animals and experimental design

Five crossbred (Dutch Landrace x Yorkshire x Finnish Landrace) barrows of 30.7±2.0 kg

body weight were fitted with a stomach cannula (Mroz & van der Mculen, in preparation)

and a post valve T-caecum (PVTC) cannula (van Leeuwen et ai, 1988) under inhalation

anaesthesia. After a recovery period of four weeks, the pigs were housed individually in

standard metabolism cages at an average room temperature of 18°C. Each pig was randomly

assigned to one of five dietary treatments consisting of semisynthetic diets fed in five

consecutive periods in a balanced 5x5 Latin Square design.

Diets and dietpreparation

All diets had the same composition with wheat bran as the only source of NSP (Table 1).
The wheal bran was incubated with water at a ratio of 6.25:1 (v/w) and acetic acid (96%)

was added (0.02:1 (v/w) to give pH 5.0. The mixture was incubated at 39°C for 3.5 hours

to give diet C (control). For the preparation of treatments Cel-I and Xyl-I, wheat bran was

incubated in the same conditions as the control with either an added cellulase (10 g/kg; Cel-I)

or a xylanase preparation (0.1 g/kg; Xyl-I). After the incubations, the wheat bran and the

incubation buffer were divided into portions of equal size and stored in plastic boxes at



5

-20°C. Incubations were carried out weekly during the course of the experiment. Before

feeding the frozen portions were thawed at room temperature and thoroughly mixed with the

balance of the experimental diets, resulting in a slurry with a water:feed ratio of 2.5:1 (v/w).

Immediately before feeding, 10 g of the cellulase and 0.1 g of the xylanase preparation per

kg wheat bran were added to diet Cto form treatments Cel-A and Xyl-A, respectively. All
diets contained 2 g/kg chromium-mordanted starch (Cr-starch) as an indigestible marker.

Table 1. Composition of the basal diet (g/kg)

Ingredients

Wheat bran 1 400.00

Maize starch 493.05

Casein 82.00

Limestone 14.00

Sail 2.50

DL-Methioninc 0.80

Mineral/vitamin premix2 1.40

Choline chloride 0.25

Cr-slarch 2.00

Co-EDTA 4.00

1 incubated in buffer with or without enzymes

2 provided the following ingredients (mg/kg diet): retinol 3.9, cholecalciferol 0.04, DL-a-

locopherol 8, riboflavin 4, cyanocobalamin 0.02, nicotinic acid 20, pantothenic acid 7.84,
antioxidant 125, FeS0 4*7H20 430, MnO 50, ZnS04*H 20 155, KI 2, Na2Se0 3 *5H20 7 and

carrier 587.2.



6

Enzymes

The enzymes used in the experiment were a Trichoderma longibrachialum cellulase and a

Trichoderma viride xylanase preparation (Genencor International Ltd., Helsinki, Finland).

Both products were crude enzyme preparations containing cellulolytic, hemicellulolytic and

xylanolytic activities (Table 2). The xylanase preparation contained approximately 100 times

more xylanase (oat spelt xylan, pH 5.3) and slightly less /1-glucanase (barley /J-glucan, pH

5.0) activity than the cellulase preparation. Since the arabinoxylans are the main NSP

fraction of wheat bran, and hence considered as the main target substrate, the two enzyme

preparations were added to provide equal amount of xylanase activity based on the activities

measured in the crude preparations. This resulted in a many-fold higher /1-glucanase activity
in the diets supplemented with the cellulase compared with those supplemented with the

xylanase preparation.

Table 2. Main activities of the crude enzyme preparations measured by standard methods 1

(units/g)

Substrate pH Cellulase Xylanase

T. virideT. longibrachialum

Barley /J-glucan

Oat spell xylan
5.0 10390 8440

5.3 1445 142050

Carboxymclhylcellulose
Filler paper

5.0 7370 5045

5.0 355 220

1 measured at the Cultor Ltd. Technology Centre, Kantvik, Finland



7

Feeding and collection procedures

The pigs were fed twice daily (at 08.30 and 20.30 hours) at 2.3 times maintenance

requirement (=4lB kJ ME/BW0 75 ) and had no access to water between the feedings. The

pigs were weighed weekly.
Each treatment period lasted 14 days and the animals were switched directly form one diet

to the next at the end of each period. Ileal digesta was quantitatively collected on the 4th and

11th day in two-hour intervals for 12 hours, beginning at 08.30 h. The digesta was collected

in polyethylene bags attached to the cannula and kept on dry ice. Each two-hourly sample

was weighed and immediately after pH had been measured, frozen and stored at -20°C. On

days 6 and 13 faeces and urine were collected quantitatively for 24 hours, then frozen and

stored at -20°C.

Gastric digesta was collected once daily on days 7,8, 9, 10 and 14 after the morning

feeding. Collection times were immediately after the pigs had finished their meals (0.5 h),

2,4, 6 and 8 hours after feeding. The stomach was evacuated in a polyethylene tube attached

to the cannula. The digesta sample was weighed, sampled and the remainder returned to the

pig. Thereafter, pH was measured and the sample was frozen and stored at -20°C.

Chemical and enzyme activity analysis

Dry matter of the stomach and ileal digesta samples was determined after freeze drying
according to the procedure of the Netherlands Normalization Institute (NEN 3332, 1974).

In addition to the wheat bran and diets, digesta samples taken from the stomach 0, 2 and

4 hours and ileum from collections between 0-2, 2-4 and 4-6 hours after feeding were

analysed for xylanase and /J-glucanase activities. Alter thawing, samples of the incubated

wheal brans, stomach and ileal digesta samples were centrifuged (g»13700) for 30 minutes

at room temperature. The supernatants were retained and kept frozen until used for the

enzyme activity measurements.

Xylanase (endo-/J-(l,4)-xylanase; EC 3.2.1.8) activity was determined by incubating 1 ml

substrate solution (1% (w/v) oat-spell xylan (Fluka 95590) in 0.05 M sodium acetate buffer

at pH 5.3) with 1 ml supernatant diluted to suitable concentration at 50°C for 30 minutes.



8

The reaction was terminated and the amount of reducing sugars released determined (Sumner
& Somers, 1949) by adding 3 ml 3,5-dinitrosalicylic acid (DNS) reagent, boiling for 5

minutes and measuring the absorbance at 540 nm after cooling. The measured absorbance

was subtracted with that of a blank sample (inactivated enzyme and DNS added to the

supernatant) to estimate the amount of reducing sugars liberated due to enzyme activity. One

xylanase unit is defined as the amount of enzyme needed to release 1 /rmol of reducing
sugars (expressed as xylose) in one minute under the assay conditions specified.

/3-glucanase (endo-/J-(l,3)(l,4)-glucanase; EC 3.2.1.6) activity was determined

spectrophotometrically. 1 ml of substrate solution (1% w/w) barley /?-glucan (Biocon

Biochemicals Ltd., Co Cork, Ireland) in distilled water) was incubated with 1 ml

supernatant (diluted with 0.1 M sodium acteate buffer at pH 5.0 to suitable concentration)

for 10 minutes at 40°C. The reaction was terminated and the amount of reducing sugars

released determined as described above in the xylanase assay. One /J-glucanase unit is defined

as the amount of enzyme required to release 1 /nnol of reducing sugars (expressed as

glucose) in one minute under the assay conditions specified.

Assay method modification

Since pH and temperature in the GI-lract of the pig are different from the ones used in the

above described standardised enzyme assays, it was decided to measure the enzyme activities

of digcsla samples also by ’modified’ methods to give a belter estimate of the actual enzyme

activity in different parts of the Gl-lract. Consequently, the assay temperature used in the

modifiedmethods was 39°C, incubation time 60 minutes and assay pH (and substrate solution

in the xylanase assay) was adjusted with Mcllvaine’s buffer to ±O.l pH units of the

measured sample pH.

Calculations and statistical analysis

The enzyme activities are expressed on a dry matter (DM) basis and were calculated using

the following equation;

Units/ml supernatant * (g sample - g sample DM)
U/g DM =

g sample DM



9

assuming that 1 ml supernatant weighs 1 g. The survivability (residual activity) of added

xylanase and /1-glucanasc activities were calculated based on the concentrations of the

markers using the equation

markerf„d
* activity djgcslJ

1 -( 1 ) * 100

markerdlgesla
* aclivity f„d

The data was subjected to analysis of variance according to the Latin square design, using

the GENSTAT 5 statistical package (Payne et al., 1987).

RESULTS

Digesla pH

pH of stomach contents decreased from an average value of 4.9 immediately after feeding

to less than 3 after 4 hours (Table 3), whereas pH of the ileal chyme was fairly stable and

decreased from 8.3 to 7.8 between 0 and 6 hours after feeding. There were no significant

treatment effects on pH over this time period.

Dietary and gastric enzyme activities

Diels containing wheat bran treated with the xylanase preparation (Xyl-I) contained

approximately 57% of the xylanase activity and 50% of the /1-glucanase found in the wheat

bran (Table 4). The corresponding values for the feeds containing the cellulase-lreated wheat

bran (Ccl-1) were 49 and 48%, respectively.



Table 3. Average pH of stomach and ileal digesta samples

(means and SD).

Stomach Ileum

Time, h 1 mean SD Time, h 1 mean SD

0 4.860.19 0-2 8.310.15

2 4.290.26 2-4 7.830.09

4 2.940.19 4-6 7.830.11

1 hours after feeding

Table 4. /?-glucanase and xylanase activities (U/g DM) of the enzyme-treated wheat brans

and corresponding experimental diets.

Treatment Xylanase /?-glucanase

Wheat bran Diet Wheat bran Diet

Control 5 3 10 5

Cel-I 41 20 146 70

Xyl-1 14 8 12 6

10



11

In pigs fed the cellulase-treated diets (Cel-I and Cel-A), xylanase activities of stomach

digesta were significantly (P <0.05) higher than in the control pigs immediately after feeding
when measured by the standard method (Table 5), whereas in pigs fed the xylanase-trealed
diets (Xyl-I and Xyl-A), xylanase activities were not significantly higher than in the controls.

Similar observations were made with regard to values obtained using the modified methods.
At 2 and 4 hours after feeding, stomach digesta samples of all pigs fed the enzyme-treated

diets tended to contain more enzyme activity than the controls, although none of them

significantly. Overall xylanase activity measured by the modified method was significantly

(P=0.001) lower than when the standard method was employed (Table 9).

Table 5. Xylanase activity measured by standard and modified methods in units/g DM of

gastric contents collected 0, 2 and 4 hours after feeding.

Time, h C Cel-I Cel-A Xyl-I Xyl-A sed

Standard

O 2.6" 13.8bc 16.4C 9.0"b 8.9"b 3.0

2 1.67.6 10.13.8 5.64.2
4 0.60.9 2.51.5 2.41.1

Modified
O 1.1" 7.2 b 7.4 b 3.7,b 3.2"b 1.9
2 0.84.1 5.31.7 2.72.0
4 0.30.4 0.80.6 0.70.3

Contrast P-values Standard vs Modified
0 0.5260.006 < 0.0010.023 0.016
2 0.7760.238 0.1180.500 0.326
4 0.7980.591 0.1310.259 0.044

ab values within a row with no common superscripts arc significantly different (P<0.05)



12

/J-glucanase activities of the stomach digesta samples taken 0 and 2 hours after feeding of

pigs fed the cellulase-treatedfeeds (Cel-I and Cel-A) were significantly (P <0.05) higher than

the control and the xylanase-treated feeds with both assay methods (Table 6). At 2 and 4
hours after feeding, /J-glucanase activity of Cel-A was significantly (P<0.05) higher than

the control, Xyl-I and Xyl-A when measured by the standard method, and higher than the

control, Xyl-I and Xyl-A 2 hours after feeding when measured by the modified method.

Overall, values obtained with the modified method tended to be lower (P =0.071) than the

ones obtained with the standard method (Table 9).

Table 6. /1-glucanasc activity measured by standard and modified methods in unils/g DM of

gastric contents collected 0, 2 and ,4 hours after feeding.

Time, h C Ccl-I Cel-A Xyl-I Xyl-A scd

Standard

O 2.9" 65. 8b 63. lb 3.3" 4.6" 11.8
2 3.5" 40.7" 47.3" 2.9" 3.6" 14.3

4 1.9" 6.5"b 16. lb 3.6' 1.9" 3.8

Modified
0 0.8" 52.6b 51. lb I.o* 1.2" 10.6

2 0.5" 34. Ob 38. 6b 0.5* 0.6" 12.6

4 0.12.2 10.40.8 0.13.6

Contrast P-values Standard vs Modified
0 0.8300.186 0.2280.812 0.735
2 0.8020.575 0.4660.837 0.804

4 0.3640.229 0.0930.588 0.568

"b values within a row with no common superscripts are significantly different (P<0.05)



13

In the stomach, xylanase activity recovery (survivability) based on Cr decreased rapidly and
was less than 10% of the initial activity in the diets supplemented with the cellulase and

between 10-20 % in the diets supplemented with the xylanase at 4 hours after feeding (Figure

1). The recovery of /f-glucanase activity decreased to less than 15% at 4 hours after feeding

(Figure 2).

Figure 1. Recovery (% of dietary concentration) of added xylanase activity in the stomach

based on chromium at 0.5 (H), 2 and 4 (CD) hours after feeding.

Figure 2. Recovery (% of dietary concentration) of added /?-glucanasc activity in the stomach

based on chromium at 0.5 (fl), 2 and 4 (CD) hours after feeding.



14

Ileal enzyme activities

At the terminal ileum, both xylanase and /1-glucanasc activity of the control samples were

relatively high and on the same level as that of the enzyme-treated diets in samples obtained

during the first two hours after feeding (Table 7). Xyl-I and Xyl-A samples obtained between

2 and 6 hours after feeding contained approximately twice as much (P<0.05) xylanase

activity as the control and the cellulase-treated diets when measured by the standard method.

With the modified method, the only significant (P <0.05) differences obtained were between

the xylanase-treated diets and Cel-I for samples taken between 2-4 hours, and between the

control and the xylanase-treated feeds for samples taken between 4-6 hours after feeding.

Table 7. Xylanase activity measured by standard and modified methods in units/g DM of

ileal contents collected during three two-hour periods after feeding.

Time, h C Cel-I Cel-A Xyl-I Xyl-A sed

Standard
0-2 20.520.1 15.323.2 24.65.2

2-4 13.5" 15.2" 11.9" 25. 3b 37. 6C 4.1
4-6 14.1" 14.9" 15.0" 26.5 b 28. lb 3.2

Modified
0-2 25.522.8 22.929.5 14.47.0

2-4 43.7" 1' 30.8" 36.5"b 58.1bc 67.3C 8.2

4-6 26.6" 30.0"bc 28.7"b 38. l c 34. 5b 2.9

Contrast P-values Standard vs Modified
0-2 0.3900.625 0.1850.271 0.080
2-4 0.0110.077 0.0280.001 0.002

4-6 0.0340.070 0.0560.052 0.259

,bc values within a row with no common superscripts are significantly different (P<0.05)



15

In contrast to the stomach samples, overall xylanase activity values obtained using the

modified method were higher than with the standard method (Table 9).

/?-glucanase activity of the ileal samples were low with no significant difference between
control and enzyme-treated samples except for the samples takenbetween 4 and 6 hours after

feeding (Table 8).

Table 8. /1-glucanase activity measured by standard and modified methods in units/g DM of

ileal contents collected collected during three two-hour periods after feeding.

Time, h C Cel-I Cel-A Xyl-I Xyl-A sed

Standard
0-2 17.210.0 12.016.1 14.73.4
2-4 17.126.3 18.214.1 13.14.7
4-6 10.3" 18.6b 17.5b 15.4"b 10.6» 2.6

Modified
0-2 4.11.3 2.33.8 1.71.1
2-4 4.54.7 4.17.1 3.62.2
4-6 2.9" 4.7b 5.1b 5.0b 3.2" 0.6

Contrasi P-valucs Standard vs Modified
0-2 <0.0010.006 0.003 < 0.001 < 0.001
2-4 0.002 < 0.001 <0.0010.080 0.008

4-6 0.001 < 0.001 < 0.0010.001 0.001

"b values within a row with no common superscripts are significantly different (P<0.05)



16

In these samples, regardless of the assay method employed, the cellulase-treated feeds
contained significantly (P <0.05) more /1-glucanase than the control feeds and Xyl-A. Values

obtained with the standard method were on average five times higher (P=0.001) than those

obtained with the modified method (Table 9). The high endogenous enzyme activities found

in the ileal samples made calculation of the recoveries meaningless.

Table 9. Average xylanase and /J-glucanase activities in stomach and

ileal samples measured by standard or modified methods.

Standard Modified sed P-value

Xylanase
Stomach 6.0 2.8 4.4 0.001

Ileum 20.0 32.5 2.2 0.001

P-glucanase

Stomach 19.0 14.0 2.7 0.071

Ileum 15.2 3.8 0.8 0.001

DISCUSSION

Enzyme application to animal feeds is a relatively new concept, and as a consequence,
specific feed enzyme assays have not yet been developed. Given the conditions in the

gastrointestinal (GI) tract of e.g. the pig, with pH ranging from approximately 2 to 8 (Kidder

and Manners, 1978), one can argue that an enzyme assay carried out in fixed conditions

cannot give the true picture of the actual activity in different parts of the gut. It, therefore,
seems necessary to try and measure activities of supplementary enzymes in conditions similar

to those of the digestive tract. Successful attempts have previously been made to measure

xylanase and /J-glucanase activity of ileal digesta of broiler chickens (Annison, 1992; Inborr
& Bedford, 1994). In these studies spectrophotometric assay methods based on dyed



17

substrates were employed. Similar studies with pigs have not been reported. In the present
study the absorbance measurements were carried out after the DNS reagent had reacted with

the reducing sugars liberated during the incubation, resulting in a coloured complex. Thus

these two methods are comparable.

Xylanase activity of the cellulase-treated wheat brans and diets was approximately three

limes higher than that of those treated with the xylanase preparation. This discrepancy may

be due to variation in the assay method and the high dilution rate required to measure

enyzme activities in the highly concentrated crude enzyme preparations. Hence even small

variations in the assay will multiply many thousand times, which in turn can lead to under-

or overdosing of the products. However, the measured activity values of the diets were used

as references when calculating the actual recoveries.
A change in assay temperature of 10°C means a change in enzyme activity by a factor of

approximately 2 (Dixon et al., 1979). Consequently, xylanase activity values obtained by the

modified method would be expected to be at a level of 50% of the ones obtained by the

standard method used in this experiment, provided that the other assay parameters were the

same. Deviations from this relationship should thus mainly reflect enzyme activity changes

due to changes in pH.

Xylanase activity values of the stomach samples measured by the modified method were

approximately 50 per cent lower than those measured with the standard method regardless

of time after feeding. Based on the above, this could be expected at pH close to the optimum

for the enzymes. This indicates that the conditions of the standard method were more

favourable to the enzymes and apparently lead to an overestimation the actual (true) activity

in the stomach of the pigs. Also /1-glucanasc activities were lower with the modifiedmethod.

ln-vilro studies with a crude xylanase preparation and a commercial feed enzyme have

shown that these enzymes are not readily denatured and possess an inherent stability to

proteolysis (pcpsin-HCI and pancreatin) and that the stability increased in the presence of

suitable substrates (Inborr & Grönlund, 1993). In their study the in-vitro incubations lasted

30 minutes, which was approximately the time elapsed between the consumption of the feeds

and the first sampling of the stomach in the present experiment. The results of these two

studies arc in good agreement, since the rates of enzyme inactivation were of the same

magnitude. However, both xylanase and /i-glucanasc activities decreased rapidly and were

on average 16 and 46 per cent of the initial value at 4 hours after feeding, respectively, when



18

measured by the standard method and 14 and 28 per cent of the initial value, respectively,

when measured by the modified method. Thus there was an apparent loss of enzyme activity

in these conditions. This is in contrast with the results by Inborr & Bedford (1994), who

reported full recovery of added /1-glucanase in the proximal part of the small intestine of

broiler chickens fed ad libitum. This discrepancy may be due to different pH and feed

retention times between the two animal species.

The relatively high endogenous enzyme activities (control diets) measured in the ileal

samples during the first two hours after feeding of the control diets is probably due to

microbial activity in the lumen (Tables 7 and 8). It is conceivable that the main part of the

first collection after feeding contained material from the previous feeding that had been

subjected to microbial degradation for several hours, leading to increased concentration of

fibre-degrading enzymes. Graham et al. (1986b) found that approximately 10 per cent of

lactobacilli of pig ileal digesta were capable of degrading pectin and mixed-linked /3-glucans
when pigs were fed a diet based on barley and soybean meal. In the present study the main

DF components were the arabinoxylans and /1-glucans, which may explain the relatively high

concentration of xylanase and /J-glucanase in the ileum of the control pigs. During the two

subsequent two-hour collections both xylanase and /1-glucanase concentration of the control

diet tended to decrease, whereas that of the enzyme-treated feeds generally increased. This

observation supports the assumption that the endogenous enzyme activities found in the ileal

samples immediately after feeding were a result of microbial fermentation, since ’fresh’

material arriving at the terminal ileum would not have been subjected to microbial

degradation to the same extent as ’old’ material and hence wouldcontain less background and

more added enzyme activity. Moreover, since there were no differences in enzyme

concentration between the control and the enzyme-treated feeds in the samples taken

immediatley after feeding, one can assume that most of the added enzymes had been

inactivated during the 12-hour time period between the feedings. With ’fresh’ material

arriving at the ileum (between 2 and 6 hours after feeding), the proportion of endogenous

activity of total decreased based on values using the modified assay method. This indicates

a certain degree of survivability of the added enzymes up to 6 hours after feeding (Tables

7 and 8). With the modified assay method, ileal xylanase activities were almost two times

higher than when the standard method was employed. This indicates a considerable

production of xylanase by the indigenous microflora, which was detectable by the modified



19

but not the standard assay method. One can assume that microbes living in conditions of pH

around 8 also produce enzymes for their own "digestion" with pH optima in this region. pH

optima of xylanase, ft- and endo-glucanases produced by rumen bacteria vary between 5.8

and 6.9, which correspond to pH of rumen contents (Cheng et al. , 1991). Cellulolytic and

peclinolytic microbes have also been found in the small intestine of pigs (Chesson et al.,
1985). Interestingly, with regard to /1-glucanase activity, this relationship between the assay

methods was reversed. It is possible that the /J-glucanase production by the small intestinal

microbes was very low on this diet and that differences between the two assay methods were

due to different assay conditions (pH and incubation time) or substrate affinity of the

enzymes.

When enzyme activity values obtained at the same site by the two assay methods were

compared, there was either a tendency to a significant (stomach /?-glucanase activity) or a

significant difference between the two methods employed. In three out of four cases, the

modified method gave lower values than the standard. However, in only one of these three

situations (xylanase in the stomach), the activity was approximately half of that of the

standard method as expected based on the changes in assay temperature. Thus other factors

than assay temperature appear to be involved such as pH, enzyme origin and substrate

affinity.
Based on these results it can be concluded that dietary xylanase and /1-glucanase activities

can be measured in the stomach and ileum of pigs with the assay methods employed.
Although a great portion of the added enzyme activities were recovered in the stomach

digesta samples immediately after feeding, enzyme activities decreased with time and were

almost nil at the end of the small intestine 12 hours after feeding as indicated by the high

endogenous activities found in the ileal samples immediately after feeding. When feed was

present, considerable amounts of xylanase appeared to be produced by the microflora of the

small intestine. The modified assay method gave generally significantly lower values than the

standard, indicating a need for measuring gut enzyme activities in conditions similar to those

found in the digestive tract in order to get a reliable picture regarding the actual enzyme
activity. The two enzyme preparations employed appeared to behave differently in the two

sections of the GI tract under study.



ACKNOWLEDGEMENTS

The authors would like to thank Messrs J.G.M. Bakker, R.A. Dekker and J. Veldhuis for

assisting with the sample collection and analysis. This investigation was funded by

Financieringsoverlag Mest- en Ammoniakonderzoek, Wageningen, the Netherlands and

Finnfeeds International Ltd., Marlborough, United Kingdom.



21

REFERENCES

Annison, G. (1992). Commercial enzyme supplementation of wheat-based diets raises ileal
glycanase activities and improves apparent metabolisable energy, starch and pentosan
digestibilities in broiler chickens. Animal Feed Science and Technology 38, 105-121.

Bach Knudscn, K.E. & Hansen, I. (1991). Gastrointestinal implications of wheat and oat
fractions. 1. Digestibility and bulking properties of polysaccharides and other major
constituents. British Journal of Nutrition 65, 217-232.

Cheng, K.-J., Forsberg, C.W., Minato, H. & Costerton, J.W. (1991). Microbial ecology
and physiology of feed degradation within the rumen. In Physiological Aspects of Digestion
and Metabolism in Ruminants, Proceedings of the 7th InternationalSymposium on Ruminant
Physiology, pp 595-624. [T. Tsuda, Y. Sasaki and R. Kawashima, editors]. Tokyo:
Academic Press.

Chcsson, A., Richardson, A.J. & Robertson, J.A. (1985). Fibre digestion and bacteriology
of the digestive tract of pigs fed cereal and vetetable fibre. In Proceedings of the 3rd
International Seminar on Digestive physiology in the Pig, pp. 272-275. [A. Just, J.
Fernandez and H. Jorgensen, editors]. Copenhagen: National Institute of Animal Science.

Dixon, M., Webb, E.C., Thorne, C.J.R. & Tipton, K.F. (1979). Enzymes, 3rd ed., New
York: Academic Press.

Graham, H., Hesselman, K. & Åman, P. (1986a). The influence of wheat bran and sugar-
beet pulp on the digestibility of dietary components in a cereal-based pig diet. Journal of
Nutrition 116, 242-251.

Graham, H., Hesselman, K., Jonsson, E. & Aman, P. (1986b). Influence of /1-glucanase
supplementation on digestion of a barley-based diet in the pig gastrointestinal tract. Nutrition
Reports International 34, 1089-1096.

Inborr, J. & Bedford, M.R. (1993). Stability of feed enzymes to steam pelleting during feed
processing. Animal Feed Science and Technology. (In press).

Inborr, J. & Grönlund, A. (1993). Stability of feed enzymes in physiological conditions
assayed by in-vitro methods. Agricultural Science in Finland 2, 125-132.

Kidder, D.E. & Manners, M.J. (1978). Digestion in the pig. pp. 201. Scientechnica, Bath,
U.K. ISBN 0 85608 022 5.

Lecuwen, van, P., Kuisman, J., Verstegen, M.W.A., Baak, M.J., van Kleef, D.J., van
Weerdcn, E.J. & den Hartog, L.A. (1988). A new technique for collection of ileal chyme
in pigs. In Proceedings of the 4th International Seminar on Digestive Physiology in the Pig,
pp. 289-296. [L. Buraczewska, S. Buraczewski, B. Pastuszewska and T. Zebrowska,
editors], Jablonna: Polish Academy of Science.



22

Low, A.G. (1985). The role of dietary fibre in digestion, absorption and metabolism. In
Proceedings of the 3rd International Seminar on Digestive physiology in the Pig, pp 157-179
[A. Just, J. Fernandez and H. Jorgensen, editors]. Copenhagen: National Institute of Animal
Science, pp. 157-179.

McClean, D. & McCracken, K.J. (1992). Effects of enzyme supplementation on digestibility
of wheat feed by weaned pigs. Proceedings of the Nutrition Society 51, 3.

Nahm, K.H. & Carlson, C.W. (1985). Effects of cellulase from Trichoderma viride on
nutrient utilisation by broilers. Poultry Science 64, 1536-1540.

Neudoerffer, T.S. & Smith, R.E. (1969). Enzymic degradation of wheat bran to improve its
nutritional value for monogastrics. Canadian Journal of Animal Science 49, 205-214.

Netherlands Normalisation Institute (1974). Test methods for feeding stuffs: determination
of the moisture content. NEN 3332. Delft, The Netherlands.

Payne, R.W., Lane, P.W., Ainsley, A.E., Dicknell, K.E., Digby, P.G.N., Harding, S.A.,
Leech, P.K., Simpson, H.R., Todd, A.D., Verner, P.J., White, P.W., Gower, J.C.,
Tunnicliffe Wilson, G & Paterson, L.J. (1987). Genslat 5 Reference manual. Oxford: Oxford
University Press.

Schulte, J.8., van Kempen, G.J.M. & Hamer, R.J. (1990). Possibilities to improve the
utilisation of feed ingredients rich in non-starch polysaccharides for poultry. In Proceedings
of the Bth European Poultry Conference, pp. 128-135. Barcelona, Spain.

Sumner, J.B. & Somers, G.F. (1949). Dinitrosalisylic acid for glucose. In Laboratory
experiments in biological chemistry, 2nd edition. New York; Academic Press.



Animal FeedScience and Technology , 44 (1993) 113-127 113
0377-8401/93/$06.00 © 1993 - Elsevier Science Publishers B.V. All rights reserved

Effect of adding fibre and starch degrading
enzymes to a barley/wheat based diet on

performance and nutrient digestibility in different
segments of the small intestine of early weaned

pigs

J. Inborr* 3
, M. Schmitzb

, F. Ahrens*5

‘Finnfeeds InternationalLtd., Market House, High Street, Marlborough, SNB IAA, UK
bIS Forschungsgesellschaft fur experimentelle Tierphysiologie und Tierernährung mbH und Co. KG,

Wiesenweg 10, D-2362 Wahlstedt, Germany

(Received 26 October 1992; accepted 11 May 1993)

Abstract

A total of 96 pigs were divided into 12 groups ofeight and housed in flat-deck pens for 21 days. A
basal diet based on barley (35%), wheat (35%) and soybean meal (22%) was used as control or
supplemented with either of two enzyme premixes (M and C), containing/1-glucanase, xylanase and
amylase. Both premixes contained the same sources ofxylanase and amylase, whereas the/J-glucanase
in premix M was of a different Trichoderma longibrachiatum strain and produced under different
conditions than the/9-glucanase in premix C. Diets were fed ad libitum during the 3-week period, with
weight gain and feed consumption recorded weekly. At the end of the experiment, two pigs from each
pen were killed, the small intestine removed and divided into four sections of equal length. Nutrient
and fibre component digestibilities were measured in the three lower sections.

Enzyme supplementation didnot significantly influence weight gain or feed utilisation of the pigs
during the experiment. Dry matter and starch digestibilities in the fourth quarter ofthe small intestine
were improved (P<0.05) by enzyme M. Both enzyme supplements increased /?-glucan digestibility
in the third and fourth quarter (P<0.05). The digestibility of soluble NSP increased significantly
(P<0.05) in the third and fourth quarter due to enzyme supplementation, whereas there was no
effect on the total NSP fraction. These results suggest that fibre and starch degrading enzymes are of
potential benefit to early weaned pigs fed dietsbased on barley and wheat.

Introduction

Weaning from the sow imposes a major stress on the pig, which has been
shown to cause a dramatic decrease in the production of, for example, pan-
creatic enzymes (Lindemann et al., 1986; Owsley et al., 1986). Furthermore,
weaning involves a change of diet, both in terms of composition and physical

•Correspondingauthor.



114 J. Inborr et ai /Animal Feed Scienceand Technology44 (1993) 1 13-127

form. This has been shown to provoke changes in the composition of the pan-
creatic digestive enzymes (Mourot and Corring, 1979; McCracken, 1984;
McCracken and Kelly, 1984). Also the change of environment usually asso-
ciated with weaning in terms of temperature may influence the production of
digestive enzymes (Szaboetal., 1976).

As the sow’s milk is withdrawn and dry feed is introduced, the digestive
system of the pig is challenged by a completely new set of chemical com-
pounds such as vegetable protein, starch and fibre. The reduced secretion of
digestive enzymes at this stage may lead to growth depression (Okai et al.,
1976) and digestive disorders due to malabsorption (Hampson and Kidder,
1986) as a result of the reduced digestive capacity. In particular, the indiges-

tible fibre components, such as the non-starch polysaccharides (NSP) may
impede the digestion of protein and energy (Low, 1985). Enzyme supple-
mentation of pig starter feeds could therefore serve two purposes. First, to
complement the production of the pig’s own digestive enzymes, and second,
to increase the digestibility ofthe fibre components of the diet.

In many European countries, wheat and barley are the most commonly used
cereal grains in pig and poultry feeds. In poultry, the soluble fibre fractions of
barley (mixed-linked/?-glucans) and wheat (pentosans) have been shown to
be responsible for the reduced nutrient digestibility in the small intestine
(White et al., 1981; Bedford et al., 1991). In a large number of experiments,
enzyme supplementation of barley and wheat based feeds has resulted in sig-
nificant improvements in bird performance and the energy value of the diet
(Chesson, 1987; Dierick, 1989). More recently, Bedford and Classen (1992)
have shown that the improved performance of broiler chickens fed enzyme-
supplemented feeds correlates to a high degree with reduced digesta viscosity
in the small intestine. There is also evidence that the mixed-linked /?-glucans
in barley may reduce the productive value of barley in young pigs, since sup-
plementation of barley based feeds with /?-glucanases has resulted in im-
proved performance (Thomke et al., 1980) and nutrient digestibility (Mark-
ström et al., 1985; Graham et al., 1986, 1989; Inborr et al., 1991).
Furthermore, enzyme treatment of barley reduced the incidence and severity
of diarrhoea in early weaned piglets (Inborr and Ogle, 1988). The mecha-
nisms involved are still unclear.

The objectives of this experiment were to investigate the effect of adding
two enzyme mixtures, containing /?-glucanase, xylanase and amylase, to a bar-
ley/wheat based pig starter feed on performance and digestibility of nutrients
and fibre components in the small intestine of early weaned pigs.

Materials and methods

Enzyme premixes

Two enzyme premixes were produced to contain equal concentration of /?-

glucanase, xylanase and a-amylase (Table 1). Both preparations contained



J. Inhorr etai. /AnimalFeed Scienceand Technology 44(1993) 113-127 115

Table 1
Measured activities of the enzyme premixes

Activity Premix M Premix C

/J-Glucanase (IRVU g 1 )
Xylanase (Ug-1 )

34.9
590

40.7
740

a-Amylase (U g“') 3300 3300

the same sources of xylanase ( Trichoderma viride) and a-amylase (Bacillus
subtilis), whereas two different y?-glucanase preparations were employed.

The two /?-glucanases (MF and CL), both from Genencor International
Ltd., Helsinki, Finland, were produced by different Trichoderma longibra-
chiatum strains under different conditions. This resulted in different activity
spectra, and pH and temperature optima of the enzymes. At 60°C, the pH
optimum of the MF cellulase activity was quite broad, between 3.5 and 5.5,
whereas that of the CL product was much narrower, between 3.8 and 4.2.

Diets

The basal diet was formulated to contain 18% crude protein, 1.2% lysine,
0.72% methionine + cystine, 0.78% threonine and a metabolisable energy
content of 13.2 MJ per kg feed (Table 2). The basal mash diet was ground
through a 4 mm sieve and then divided into three batches of approximately
equal size. Enzyme premix M (containing enzyme MF) was added to one of
the batches at 1.0 g kg -1 to form Diet M and enzyme premix C (enzyme CL)
at 0.95 g kg~ 1 to form Diet C. Based on the enzyme activity analysis, enzyme
premixes C thus provided slightly more /?-glucanase and xylanase and less a-
amylase per kg feed than premix M. The third batch was unsupplemented and
used as a control (Diet A). Titanium oxide (Tio2 ) was added as an undiges-
tible marker at 5 g kg~ 1 for the digestibility study.

Pigs

A total of 96 crossbred (German LandracexDuroc) barrows weaned at an
age of 21-24 days were divided into groups of eight and randomly assigned
to one ofthe three treatments by litter origin and weight. The pigs were housed
in flat-deck pens at an initial room temperature of26 °C, which was decreased
to 22 °C by the end of the experiment. Water was freely available from nipple
drinkers placed in the pens. Lights were on for 12 h a day between 05.30 and
17.30 h.



116 J. Inhonel ai /AnimalFeed Science and Technology 44 (1993) 113-127

Table 2
Composition and calculated content ofchemical constituents of the basal diet (gkg -1 as fed)

Ingredient

Composition
Wheat 348
Barley 348
Soybean meal ( CP 44%) 220
Soya oil 30
Vitamin/Mineralpremix 1 49
Titanium oxide 5

Calculated composition
Crude protein 180
Crude fat 45
Crude fibre 40
Lysine 12.0
Methionine and cystine 7.2
Threonine 7.8
Calcium 9.0
Phosphorous 7.0
Metabolisableenergy (MJkg -1 ) 2 13.2

‘The premix supplied per kilogram diet: retinyl acetate, 722 fig; cholecalciferol, 6 fig; 2-dl-a-tocopher-
ylacetate, 13.2 mg; vitamin 8,, 1.2 mg; vitamin 82,B2, 3.6 mg; vitamin 86,B6, 1.8 mg; vitamin 8,2> 18 fig;
vitamin K 3, 0.6 mg; folic acid, 0.36 mg; niacin, 15 mg; biotin, 0.06 mg; pantothenic acid, 10.8 mg;
choline, 0.48 g; Ca, 7,9 g; P, 3.3 g;Na, 1.6 g; Fe, 96 mg; Mg, 0.48 g; Mn, 3.8 mg; Zn, 96 mg; Cu, 6 mg;
I, 0.17 mg; Se, 0.3 mg; L-lysine-HCI, 4.1 g; DL-methionine, 1.5 g; L-threonine, 1.5 g; salinomycin 40
mg.
Calculated according to DLG (1991).

Performance study

The experimental diets were fed ad libitum for 20 days. The pigs were
weighed individually on Days 1,8, 15 and 21 and feed consumed per pen was
recorded on Days 8, 15 and 21, thus dividing the experiment into two periods
of 7 days and one period of 6 days. Live weight gains and feed consumption
were calculated for all three periods and feed conversion ratios for periods
two and three, since a few pigs lost weight during the first period. Health sta-
tus was monitored daily and any abnormalities recorded.

Digestibility study

On Days 21 and 22 of the experiment, pigs weighing between 10 and 13 kg
were killed and digesta was collected from the small intestine. Each day, four
pigs per treatment (one pig from each pen) wererandomly selected at least 3
h after the time the lights were switched on in the morning. Thus eight pigs
per treatment were used for digestiblity measurements.



117J. Inborret ai. /AnimalFeed Science and Technology 44(1993) 113-127

Pigs were killed and the small intestine was ligated at the pylorus and the
ileocaecal junction. Then the mesentry was cut and the small intestine di-
vided into four segments of equal length. Owing to the small amount of di-
gesta, the first segment was discarded. Digesta was gently squeezed out from
each single segment into separate plastic containers. The gut was then rinsed
with a saline solution to wash out any remaining digesta and the washing added
to the digesta sample in the plastic container. Digesta samples were immedi-
ately put into a freezer for storage at 20°C. Before analysis, the digesta sam-
ples were freeze-dried and ground through a 0.5 mm sieve.

Chemical analysis

The dry matter content of the feeds was determined after the samples had
been kept at 105°C for 4 h, and the dry matter content of the digesta samples
by freeze drying for 96 h.

Nitrogen was determined by the Kjeldahl procedure, crude fat by petro-
leum ether extraction after acid (HCI) hydrolysis, crude fibre by the method
of VDLUFA (1975) and starch enzymatically by the method of Brandt et al.
(1987). Ash content of the feeds was determined after the samples were ashed
at 550°C overnight.

Total NSP content was determined by employing the method of Englyst
and Hudson (1987) with the following modifications: the enzymes employed
were Termamyl 300 L and Promozyme (both Novo Nordisk a/s, Copen-
hagen, Denmark). Isolated fibre polysaccharides were hydrolysed for 0.5 h at
25 °C in 12 M H2S0 4, followed by 2 h at 100°C in 1 M H2S04 . Monomers
were then derivatised to their alditol acetates and quantified by gas chroma-
tography (SP-2340 column).

The soluble NSP content was determined by applying the same method as
above after the sample had been extracted in 0.2 M phosphate buffer (pH 7)
for 60 min, the resultant supernatant liquid destarched by thermophilic am-
ylase and pullulanase and the soluble NSP components precipitated with an
80% ethanolic solution. Beta-glucan was determined by the method of Mc-
Cleary and Codd (1991), using reagents supplied by Megazyme Pty Ltd.
(North Rocks, NSW, Australia). Titanium oxide was measured according to
the method of Brandt and Allam (1987).

Enzyme activity analysis

AlO g feed sample was weighed into a 100 ml volumetric flask. Distilled
water was added to 100 ml and the suspension stirred for 30 min in room
temperature and then filtered through a glass fibre filter (Macherey Nagel
85/90). The extract was diluted to the appropriate concentration with buffer
solutions according to the methods employed.



118 J. Inborr etai. /AnimalFeed Science and Technology44 (1993) 113-127

Beta-glucanase (EC 3.2.1.6; endo-/?-(l, 3)(1, 4)-glucanase) activity was
determined viscosimetrically by employing the method described by Bath-
gate (1979) with minor modifications (pH adjusted to 4.0 by 0.5 M sodium
acetate buffer and initial reciprocal viscosity of the substrate solution was
0.13). This method measures the reduction in substrate viscosity and enzyme
activity was expressed as the increase in reciprocal viscosity (IRV).

After enzyme addition, four to six flow times were determined over a 30
min time period in an Ostwald capillary viscometer (Model No. 11, 75-100
s), immersed in a water bath at 30°C. The flow times were used to calculate
the reciprocal specific viscosities, using the equation:

1/ dT°

/r,sp dTs -dT0

where \/rjsp is the reciprocal specific viscosity, dT0 is the flow time of the
buffer and dTs is the flow time of the substrate-enzyme solution.

Beta-glucanase activity was expressed as IRV and calculated using the
equation:

k'X.D/?-glucanase, IR V Units g~ 1
= ——

where k is the slope of the curve plotted from the reciprocal specific viscocity
versus hydrolysis time (min), Dis the total dilution factor and V is the sam-
ple volume. One IRV unit is defined as the change of one (1 /r] sp ) min~ 1

Xylanase (EC 3.2.1.8; endo-(l, 4)-/?-xylanase) activity was determined
using 1% oat spelt xylan (Fluka 95590) in 0.05 M sodium acetate buffer, pH
5.3, as substrate. One ml of enzyme dilution was incubated with 1 ml of sub-
strate solution at 50°C for 60 min. Reducing sugars formed were assayed by
adding 3 ml of 3, 5-dinitrosalicylic acid (DNS) reagent, boiling for 5 min and
measuring the absorbance at 540 nm. One xylanase unit is the amount of
enzyme that releases 1 /rniol of reducing sugars (expressed as xylose) in 1
min.

Alpha-amylase (EC 3.2.1.1) activity was determined using the Phadebas ®

Amylase Test kit (Pharmacia Diagnostics AB, Uppsala, Sweden), with buffer
(9.0 g NaCl, 2.0 g bovine serum albumin and 2.2 g CaCl 2 in 1000 ml distilled
water; pH 6.7) used instead of distilled water for dilution of the sample and
dissolution of the tablet. One «-amylase unit is the amount of enzyme needed
to hydrolyse one /rmol of glycosidic linkages in 1 min under the analytical
conditions.

Calculations and statistical analysis

Pig live weight gains are reported as treatment means based on individual



119J. Inborr etai. /Animal Feed Scienceand Technology 44 (1993) 113-127

pig weights, and feed intake is reported as treatment means based on individ-
ual pen (replicate) values. Feed conversion ratio is calculated as feed intake
divided by mean replicate live weight gain.

Digestibility coefficients of the analysed components are reported as treat-
ment mean values (eight pigs per treatment) and were calculated using the
formula;

I
nutrient concentration in digestaxTi02 concentration in feed
nutrient concentration in feedxTio2 concentration in digesta

Normal distribution of data and equality of variances were tested accord-
ing to Bartlett (Sachs, 1992). Data on weight gain, feed intake and feed con-
version ratio ofthe 3-week period was normally distributed and showed equal
variances. They were analysed by ANOVA with diet as factor. Digestibilities
were normally distributed and analysed separately for each small intestinal
section by ANOVA with diet as factor. This procedure was adopted because
variances in the single segments were different and could not be made equal
by transformation calculation. Differences between means were tested ac-
cording to Scheffe (Sachs, 1992).

Results

Diets

On an air dry basis, the total /J-glucan content of the diets ranged between
1.2 and 1.3% and the total NSP between 10.8 and 11.7% (Table 3). Glucose,
uronic acid, xylose and arabinose were the most abundant NSP sugars of the
diets, with rhamnose, fucose and mannose in concentrations less than 1%.
Solubility ofthe NSP sugars ranged from approximately 17% (xylose) to 68%
(glucose) with an average of 40%.

Measured enzyme activities of the supplemented diets were close to the
calculated values for /?-glucanase and xylanase, whereas that of «-amylase in
Diet M was lower (Table 3). Endogenous/?-glucanase and xylanase activities
of the unsupplemented diet were relatively high and represented approxi-
mately 15% of the total /?-glucanase and 57% of the total xylanase activity of
the supplemented diets. This high value for endogenous xylanase may have
resulted from high endogenous activity in, for example, wheat or from com-
ponents in the feed (e.g. metal ions) reacting with the DNS reagent causing a
colour reaction, which then increases the spectrophotometric absorbance of
the solution. The endogenous «-amylase activity of the unsupplemented diet
was low.



Table 3
Chemical composition (g kg -1 as fed) and measured enzyme activities of the experimental diets for
Pigs

Chemical constituent Experimental diet

A M C

Crude protein 182 181 181
Crude fat (HCI) 55 55 52
Ash 65 66 66
Starch 397 397 399
Mixed-linked /?-glucans 13 12 13

Non-starch polysaccharide residues
Total 117 108 114

Soluble
Arabinose 5 5 5
Xylose 5 6 5
Galactose 5 6 6
Glucose 23 26 25
Uronic acid 4 4 4
Mannose 2 2 1
Rhamnose -'

- -

Fucose - - -

Total 44 49 46

Insoluble
Arabinose 14 12 12
Xylose 25 21 25
Galactose 5 4 4
Glucose 18 12 16
Uronic acid 5 4 5
Mannose 4 4 4
Rhamnose 1 1 1
Fucose 1 1 1
Total 73 59 68

Enzyme activities
/?-glucanase(lßVUkg-') 6 34 36
xylanase(Ug-') 1.01.7 1.8
a-amylase (U g_l

) <O.l 2 3

'Concentration under detection limit.

Performance study

There were no mortalities or cases of illness among the treatment groups
during the experiment.

There were no significant differences in either weight gain or feed utilisa-



121J Inhorrelal. /Animal Feed Scienceand Technology44 (1993) 113-127

Table 4
Weight gain, feed intake and feed conversion ratio of pigs in the treatment groups during the experi-
mental period (values are means ±SEM) 1

Diet A Diet M Diet C

Weight gain (g day ~')

Week 1 56±10 64± 11 39±11
Week 2 203112 209113 210117
Week 3 379+16 423 + 26 392+18
Weeks 1-3 20419 223110 205113

Feed intake (g day~ ')

Week 1 10616 11419 10919
Week 2 254117 28018 231117
Week 3 517117 566117 541133
Weeks 1-3 281111 307+10 281 + 18

Feed: gain
Week 2 1.20+0.02 1.3610.09 1.1310.07
Week 3 1.37+0.02 1.3410.04 1.3810.02
Weeks 1-3 1.3810.02 1.3810.02 1.3910.06
'No significant treatment effects observed (/> >0.05).

tion of the pigs of the three treatment groups during the overall experimental
period (Table 4).

Digestibility study

Dry matter digestibility tended to be higher in the second and third quarter,
and was significantly higher (Pc 0.05) in the fourth quarter of the small in-
testine of pigs fed the enzyme supplemented diets (Table 5). Crude protein
digestibility was numerically but not significantly higher in the pigs fed the
enzyme-treated diets compared with the pigs fed the control diet.

Starch digestibility was significantly higher ( P< 0.05) in the fourth quarter
of the small intestine of pigs fed Diet M compared with the control pigs.

In pigs fed the basal diet (Diet A), the digestibility of/?-glucans was gener-
ally low and on the same level inall segments (Table 5). Supplementation of
the basal diet with the enzymes significantly (.Pc 0.05) increased the diges-
tibility of/?-glucans in the third and fourth section of the small intestine.

The digestibility of the soluble NSP fraction of the unsupplemented diet
was negative in all three sections of the small intestine investigated, with the
lowest value in the second and the highest in the fourth quarter (Table 6).
Enzyme supplementation significantly (P<0.05) increased the digestibility
of soluble NSP in the third and fourth quarters but did not have any signifi-
cant effect in the second quarter.



122 J. Inhonet al. /AnimalFeed Science and Technology 44(1993) US-127

Table 5
Digestibility of dry matter, crude protein, starch and mixed-linked/J-glueans in the last three quarters
of the small intestine of the pigs (treatment means ±SEM)

Section of the Nutrient Diet A Diet M Diet C
small intestine

Second quarter Dry matter 0.245 ±0.030 0.270 ±0.052 0.321 ±0.029
Protein 0.310±0.055 0.420±0.018 0.327±0.051
Starch 0.332 ±0.052 0.509 ±0.035 0.379 ±0.045
/Tglucan 0.340±0.021 0.447±0.069 0.559±0.076

Third quarter Dry matter 0.526±0.027 0.550±0.027 0.589±0.014
Protein 0.619±0.029 0.646±0.032 0.699±0.010
Starch 0.807 ±0.017 0.857 ±0.030 0.854 ±0.021
y?-glucan 0.304±0.055 0.578±0.029a 0.565±0.070a

Fourth quarter Dry matter 0.643 ±O.OOB 0.679 ±O.OO6a 0.671 ±0.009
Protein 0.768±0.018 0.784±0.013 0.779±0.011
Starch 0.949 ±O.OOB 0.977 ±O.OO6 a 0.968 ±0.004
/J-glucan 0.321 ±0.030 0.719±0.031 a 0.719±0.033a

treatment mean value is significantly different (P<0.05) from control (Diet A).

Table 6
Digestibility of total and soluble non-starch polysaccharides in the last three quarters of the small
intestine of the pigs (treatment means ± SEM)

Section of the Non-starch Diet A Diet M Diet C
small intestine polysaccharide

fraction

Second quarter Total 0.142 ±0.055 0.032 ±0.098 0.092 ±0.052
Soluble —0.385 ±O.OBO -0.304±0.122 -0’.328±0.086

Third quarter Total 0.045 ±0.055 0.005 ±0.049 0.032±0.056
Soluble -0.149 ±0.046 0.091 ±0.048“ 0.091 ±0.050“

Fourth quarter Total 0.013±0.025 —0.004±0.042 0.001 ±0.039
Soluble -0.054 ±0.030 0.219 ±0.016“ 0.160±0.030“

“Treatment mean value is significantly different (P<0.05) from control (Diet A).

Discussion

Enzyme supplementation of pig starter diets has given inconsistent re-
sponses in terms live weight gain and feed utilisation, whereas the response
to enzymes in broiler chickens has been much more consistent (Chesson, 1987;
Dierick, 1989). Recent studies with broilers have shown that reduced digesta
viscosity due to enzyme supplementation is the main factor responsible for
the performance response in diets based on barley (Bedford, 1993), wheat
and rye (Bedford and Classen, 1992). In contrast, digesta viscosity in the
small intestine of early weaned pigs fed hulless barley supplemented with /?-

glucanase did not correlate with improved performance (Inborr et al., 1991),
suggesting that viscosity is not a factor reducing the nutritive value of barley



123J. Inborr et al. /AnimalFeed Science and Technology44 (1993) 113-127

in young pigs. Interestingly, however, in the same study, protein digestibility
decreased as digesta viscosity increased in the first and third quarter of the
small intestine in pigs fed rye-based diets.

In the present study, enzyme supplementation increased the digestibility of
starch in the small intestine. This is in contrast with the results of a previous
study (Inborr et al., 1991) with pigs fed a hulless barley-based diet supple-
mented with a single /?-glucanase source. The use of a-amylase in combina-
tion with fibre-degrading enzymes in the present experiment may explain the
increased starch digestibility. Graham et al. (1986,1988, 1989) also reported
increased ileal starch digestibility in pigs fed enzyme-supplemented diets based
on barley or barley and pollard.

Since starch was the major source of energy in the diets employed, in-
creased weight gain and improved feed utilisation could perhaps have been
expected due to the increased digestibility. However, considering the magni-
tude of improvement, which was in the range of 2.5 percentage units at the
end of the small intestine, corresponding to approximately 4 g less feed con-
sumption per day (assuming equal weight gain) and a feed conversion ratio
reduction of0.02, the lack of performance response is not surprising.

Increased starch digestibility in the upper small intestine effectively means
that less easily fermentable substrate arrives in the lower digestive tract to be
fermented by the microbes. This would reduce the proliferation and activity
of the microflora in the caecum and large intestine (Bach Knudsen et al.,
1991), thus possibly reducing digestive upsets. Inborr and Ogle (1988) re-
ported reduced incidence and severity of diarrhoea in pigs fed enzyme-treated
barley, while Böhme (1990) found a reduced need for antibiotic treatments
due to less digestive disorders in pigs fed feeds supplemented with a /?-glucan-
ase and amylase containing enzyme preparation. In the present experiment,
none of the treatment groups showed any signs of digestive disorders.

Protein digestibility was not significantly increased by enzyme supplemen-
tation. This is in contrast with the results by Inborr et al. (1991), who re-
ported improved protein digestibility over the entire small intestine and Gra-
ham et al. (1988), who obtained an increased ileal protein digestibility in pigs
fed barley-based diets supplemented with /?-glucanase.

Addition of the enzyme supplements markedly increased the digestibility
of /?-glucans in the third and fourth quarters of the small intestine. This is in
agreement with the results by Graham et al. (1988, 1989). Ileal digestibility
of/?-glucan in the pig seems to increase with age (Graham et al., 1988) but is
also influenced by the source and solubility of the /J-glucans (Bach Knudsen
and Hansen, 1991). In a pig of about 20 kg live weight, /?-glucan digestibility
in the terminal ileum was reported to be about 40% (Graham et al., 1988),
which is higher than the results in this experiment with pigs weighing between
10 and 13 kg. This discrepancy may be due to lower microbial activity in the
smaller pigs, resulting in less degradation of the /?-glucans.



124 J. Inborret al. /AnimalFeed Scienceand Technology 44 (1993) 1 13-127

Although digesta viscosity in the small intestine has not been proved to be
a factor influencing the performance of barley fed pigs (Inborr et al., 1991),
it is possible that soluble /?-glucans increase the viscosity of the gastric con-
tent. This, in turn, may alter the rate of gastric emptying (Potkins et al., 1991)
and the size of particles passing through the pyloric sphincter into the small
intestine (Meyer et al., 1986), hereby affecting the rate end efficiency of
digestion. Consequently, the increased digestibility of /?-glucans in the pres-
ence of enzymes observed in the third quarter and a similar tendency in the
second quarter of the small intestine may indicate more rapid degradation
higher up in the gastrointestinal tract, for example in the stomach. This would
mean reduced viscosity and better conditions for the digestion of starch and
protein in the duodenum and jejunum due to better gastric function.

The digestibility of the soluble NSP fraction of the unsupplemented diet
was negative in all segments of the small intestine. This is in accordance with
the results obtained by Vervaeke et al. (1991) and is a consequence of fibres
being solubilised as they pass through the upper parts of the gastrointestinal
tract. The digestibility coefficients were approaching zero as the digesta passed
through the small intestine. This is likely to be a result of the microbial deg-
radation of these carbohydrates. Graham et al. (1986) showed that, for ex-
ample lactobacilli from the small intestine are capable of degrading pectins
and /7-glucans.

Since soluble fibres of wheat and barley tend to increase digesta viscosity,
thereby impairing digestion (White etal., 1981; Bedford and Classen, 1992),
this solubilisation of fibres may well reduce the efficiency of digestion in the
young pig. Enzyme supplementation did not have any effect on the digestibil-
ity of the soluble NSP in the second quarter of the small intestine and it may
be that the rate of solubilisation was much greater than the rate by which the
added enzymes were capable of degrading these fibres. In the third and fourth
quarter, enzyme supplementation significantly increased the digestibility of
the soluble NSP, which resulted in positive digestibility coefficients as op-
posed to the negative value of the control diet. This, together with the in-
creased starch and dry matter digestibilities, would strongly suggest that the
enzymes are still active in the upper jejunum and ileum and can improve the
efficiency of digestion.

Based on the digestibility of the total and soluble NSP, some degradation
of insoluble NSP appeared to take place in the small intestine. Vervaeke et al.
(1991) measured an average of 21 % degradation of insoluble NSP in the ileum
of pigs fed diets containing cereal grains (minimum 25%), sugar beet pulp,
wheatbran and alfalfa. It is possible that these diets contained more easily
fermentable components than those employed in the present experiment. Bach
Knudsen and Hansen (1991) and Bach Knudsen et al. (1991) found that the
recovery of insoluble NSP in the ileum of growing pigs was lower when wheat
flour was fed alone compared with wheat flour fed with oat bran but they



125J. Inborr etai. /Animal Feed Scienceand Technology 44(1993) 113-127

could not find any difference between these treatments in terms of microbial
activity in the last third of the small intestine. On the other hand, feeding
rolled oats resulted in almost full recovery of insoluble NSP in the ileum de-
spite the highest microbial activity, indicating that the source and structure
of carbohydrates influences the degradation ofthis fraction in the small intes-
tine of pigs.

It seems that fibre-degrading enzymes, whether added to the feed or pro-
duced by the gut microflora, have a preference for solubilised substrates over
insoluble, which is why enzyme supplementation in this experiment in-
creased the digestibility of soluble NSP but had no effect on the total NSP
fraction. This is a fortunate feature of exogenous enzymes, since the soluble
fibres are probably interfering more with food digestion than the insoluble
ones. This possibly means greater potential for increasing the efficacy of feed
utilisation by enzyme supplementation of feeds based on ingredients high in
soluble fibre.

There was no difference in efficacy between the two enzyme supplements
employed. This suggests that differences between the two fungal /?-glucanases
in terms ofpH optima and activity spectra were of insignificant importance.
One could possibly expect greater differences between fungal and bacterial
sources of /?-glucanase, since they would differ more in terms of pH optima
and substrate specificity than two sources of either fungal or bacterial origin.

The results from this experiment show that enzyme supplementation sig-
nificantly increased the digestibility of starch and fibre components, but that
the magnitude of improvement was too small to give a significant improve-
ment in growth performance or feed utilisation of the pigs. However, these
results show the potential benefits of fibre and starch degrading enzymes in
barley/wheat based pig starter feeds. They also give some indications regard-
ing the mode of action of the added enzymes in that the site of digestion and
absorption of nutrients seem to have been shifted upwards in the small intes-
tine when enzymes were added. Whether this apparent change in the rate of
digestion is due to reduced gut viscosity or simply to alleviation ofthe imped-
ing effects of fibre needs to be further investigated.

Acknowledgements

The authors would like to thank Jari Puhakka at the Cultor Ltd. Technol-
ogy Centre, Kantvik, Finland, for carrying out the enzyme activity analysis.
This work was funded by Finnfeeds International Ltd., Wiltshire, UK.

References
Bach Knudsen, K.E. and Hansen, 1., 1991. Gastrointestinal implications in pigs of wheat and

oat fractions. 1.Digestibility and bulking properties ofpolysaccharides and other major con-
stituents. Br. J. Nutr., 65: 217-232.



126 ,/, Inhonet ai. /AnimalFeed Scienceand Technology 44 (1993) 113-127

Bach Knudsen, K.E., Borg Jensen, 8., Andersen, J.O. and Hansen, 1., 1991. Gastrointestinal
implications in pigs of wheat and oat fractions. 2. Microbial activity in the gastrointestinal
tract. Br. J. Nutr., 65: 233-248.

Bathgate, G.N., 1979. The determination of endo-/?-glucanase activity in mall. J. Inst. Brew.,
85: 92-94.

Bedford, M.R., 1993. Mode of action of feed enzymes. J. Applied Poultry Res., Vol. 2: 85-92.
Bedford, M.R. and Classen, H.L., 1992. Reduction of intestinal viscosity through manipulation

of dietary rye and pentosanase concentration is effected through changes in the carbohydrate
composition of the intestinal aqueous phase and results in improved growth rate and food
conversion efficiency of broiler chicks. J. Nutr., 122: 560-569.

Bedford, M.R., Classen, H.L. and Campbell, G.L., 1991. The effect of pelleting, salt and pen-
tosanase on the viscosity of intestinal contents and the performance ofbroilers fed rye. Poult.
Sci.,7o: 1571-1577.

Böhme, H., 1990. Experiments on the effect of enzyme supplements as a growth promotor for
piglets. Landbauforsch. Voelkenrode, 40(3): 213-217.

Brandt, M. and Allam, S.M., 1987. Analytik von TiG2 im Darminhalt und Kot nach Kjeldah-
laufschluss. Arch. Anim. Nutr., 37: 453-454.

Brandt, M., Schuldt, A. Mannekorpi, P. and Vearasilp, T., 1987. Zur enzymatischen Stärkebes-
timmung im Darminhalt und Kot von Kiiken mit hitzestabiler Amylase. Arch. Anim. Nutr.,
37; 455.

Chesson, A., 1987. Supplementary enzymes to improve the utilisation ofpig and poultry diets.
In; W. Haresign and D.J.A. Cole (Editors), Recent Advances in Animal Nutrition. Butter-
worths, London, pp. 71-89.

Dierick, N.A., 1989. Biotechnology aids to improve feed and feed digestion: enzymes and fer-
mentation. Arch. Anim. Nutr., 39: 241-261.

DLG, 1991. DLG-Futterwerttabellen-Schweine. DLG-Verlag, Frankfurt am Main, pp. 13-15.
Englyst, H.N. and Hudson, G.J., 1987. Colorimetric method for routine measurement of die-

tary fibre as non-starch polysaccharides. A comparison with gas-liquid chromatography. Food
Chem., 24; 63-76.

Graham, H., Hesselman, K., Jonsson, E. and Åman, P., 1986. Influence of/(-glucanase supple-
mentation on digestion of a barley based diet in the pig gastro-intestinal tract. Nutr. Rep.
Int., 34: 1089-1096.

Graham, H., Löwgren, W., Pettersson, D. and Åman, P., 1988. Effect of enzyme supplementa-
tion on digestion ofa barley/pollard-based pig diet. Nutr. Rep. Int., 38: 1073-1079.

Graham, H., Fadel, J.G., Newman, C.W. and Newman, R.K., 1989. Effect of pelleting and /?-

glucanase supplementation on the ileal and whole-tract digestibility of a barley-based diet in
the pig. J. Anim. Sci., 67: 1293-1298.

Hampson, D. J. and Kidder, D.E., 1986. Influence of creep feeding and weaning on brush border
enzyme activities in the piglet small intestine. Res. Vet. Sci., 40: 24-31.

Inborr, J., Bedford, M.R., Patience, J.F. and Classen, H.L., 1991. The influence of supplemen-
tary feed enzymes on nutrient disappearance and digesta characteristics in the GI-tract of
early weaned pigs. Proc. Vth Int. Symp. on Digestive Physiology in Pigs, Wagcningcn, The
Netherlands, 24-26 April. Pudoc, Wageningen, EAAP Pubi., No. 54, 401-405.

Inborr, J. and Ogle, R.8., 1988. Effect of enzyme treatment ofpiglet feeds on performance and
post-weaning diarrhoea. Swedish J. Agric. Res., 18: 129-133.

Lindemann, M.D., Cornelius, S.G., El Kandelgy, S.M., Moser, R.L. and Pettigrew, J.E., 1986.
Effect of age, weaning and diet on digestive enzyme levels in the piglet. J. Anim. Sci., 62:
1298-1307.

Low, A.G., 1985. The role of dietary fibre in digestion, absorption and metabolism. In: Proc.
3rd Internatl. Seminar on Digestive Physiology in the Pig, National Institute of Animal Sci-



J. Inhon et ai /AnimalFeed Science and Technology 44 (1993) 113-127 127

ence, Copenhagen, Denmark. Landhusholdningsselskabets forlag, Copenhagen, Beretning
fra Statens Husdyrbrugsforsog, No. 580, 157-179.

Markström, 8., Pettersson, D. and Hesselman, K., 1985. /i-glukanastillsats till grov- och fin-
malet korn - ett smältbarhetsforsök med smågrisar. Swedish University of Agricultural Sci-
ences, Department ofAnimal Nutritionand Management, Report 149, 15 pp.

McCleary, B.V. and Codd, R., 1991. Measurement of (1-3), (l-4)-/i-D-glucan in barley and
oats: A streamlined enzymic procedure. J. Sei. Food Agric., 55: 303-312.

McCracken, K.J., 1984. Effect ofdiet composition on digestive development of early-weaned
pigs. Proc. Nutr. Soc., 43: 109 A (abstract).

McCracken, K.J. and Kelly, D., 1984. Effect ofdiet and post-weaning food intake on digestive
development ofearly-weaned pigs. Proc. Nutr. Soc., 43: 110 A (abstract).

Meyer, J.H., Gu, Y., Elashoff, J., Reedy, T., Dressman, J. and Amidon, G., 1986. Effects of
viscosity and fluid outflow on postcibal gastric emptying of solids. Am. J. Physiol., 250;
Gl6l-G164.

Mourot, J. and Corring, T., 1979. Adaption of the lipase-colipase system to dietary lipid con-
tent in pig pancreatic tissue. Ann. Biol. Anim. Biochem. Biophys., 19 (1A): 119-124.

Okai, D.8., Aherne, F.X. and Hardin, R.T., 1976. Effects of creep and starter composition on
feed intake and performance of young pigs. Can. J. Anim. Sci., 56: 573-586.

Owsley, W.F., Orr, Jr., D.E. and Tribble, L.F., 1986. Effects of age and diet on the development
of the pancreas and the synthesis and secretion of pancreatic enzymes in the young pig. J.
Anim. Sci., 63: 497-504.

Potkins, Z.V., Lawrence, T.L.J. andThomlinson, J.R., 1991. Effects of structural and non-struc-
tural polysaccharides in the diet of the growing pig on gastric emptying rate and rate ofpas-
sage ofdigesta to the terminal ileum and through the total gastrointestinal tract. Br. J. Nutr.,
65: 391-413.

Sachs, L., 1992. Angewandte Statistik. 7. Auflage, Springer Verlag, Berlin, 848 pp.
Szabo, J., Ribiczeyne-Szabo, P. and Rafai, P., 1976. A pancreas-hidrolazok aktivitasanak

vizsgalata kulonbozo kornyezeti homesekleten tartott sertesekben. Magyar Allatorvosok
Lapja, 31: 125-128.

Thomke, S., Rundgren, M. and Hesselman, K., 1980. The effect of feeding high-viscosity barley
to pigs. 31st Annual Meeting of EAAP, Munich. Commission on Animal Nutrition, p. 5
(Abstract).

VDLUFA, 1975. Methodenbuch VDLUFA, Band 111. Methode 6.1.1., Darmstadt, Germany,
pp. 1-4.

Vervaeke, 1.J., Graham, H., Dierick, N.A., Demeyer, D.I. and Decuypere, J.A., 1991. Chemical
analysis of cell walls and energy digestibility in growing pigs. Anim. Feed Sci. Technol., 32:
55-61.

White, W.8., Bird, H.R., Sunde, M.L., Prentice, N., Burger, W.C. and Marlett, J.A., 1981. The
viscosity interaction ofbarley beta-glucan with Trichoderma viride cellulase in the chick in-
testine. Poult. Sci., 60: 1043-1048.





Gastrointestinal parameters influencing performance of pigs

fed enzyme-supplemented barley-based feeds.

BY JOHAN INBORR 1

Finnfeeds International Ltd., Market House, Ailesbury Court, High Street, Marlborough,

Wiltshire SNB IAA, United Kingdom.

1 Present address: Svenska Foder Ab, Kinnegatan 23, Box 673, S-53116 Lidköping, Sweden



2

Abstract

Forty early-weaned pigs were housed individually in metabolism cages and fed diets

based on either a bulled (var. Arra) or bulless (var. Condor) barley with or without

added /J-glucanase for a 21-day period during which feed intake and live weight gain

were monitored. At the end of the experiment the pigs were sacrificed and the entire

gastrointestinal tract removed and divided into nine sections (stomach, four small

intestinal sections of equal length, caecum, ascending and descending colon and rectum).

Digesta from each of the GI segments was analysed for viscosity, dry matter and

nutrient digestibility using chromic oxide as an indigestible marker. Addition of enzyme

increased live weight gain (P = 0.074) and improved feed utilisation (P=0.058) of the

pigs. Enzyme supplementation significantly (P< 0.0001) increased the digestibility of /3-
glucans and reduced digesta viscosity in the stomach and three proximal quarters of the

small intestine (P<0.03). Furthermore, the activity of digestive enzymes was reduced

(P<0.08) due to addition of the enzyme without affecting dry matter, starch, nitrogen

or energy digestibility. It appears that through enzyme supplementation more optimal

conditions for digestion of the nutrients were created and endogenous losses reduced,

which, in turn, resulted in improved performance of the pigs.

INDEXING KEY WORDS: pig, enzyme supplementation, /?-glucanase, barley, nutrient

digestibility, digesta viscosity, digestive enzymes.



3

Pigs like other monogaslric animals do not produce enzymes capable of digesting fibre.

Indeed dietary fibre components can interfere with the digestion and production of digestive
enzymes (Graham, 1988; Ikegami et al., 1990). Dietary fibre is usually definedas non-starch

polysaccharides and lignin (Low, 1985) and consists of a large number of complex

carbohydrates with different chemical and physical properties. In barley, the main dietary

fibre component are cellulose and the mixcd-linked /1-glucans (Aman & Graham, 1987), the

concentration and solubility of which can change with both cultivar and conditions during the

growing and harvest seasons (Hesselman & Thomke, 1982). In poultry, soluble /J-glucans

form a viscous environment in the intestinal lumen (While et al., 1981), which impaires

optimal nutrient digestibility and performance of the birds (Hesselman, 1983).Similar effects

have not been observed in pigs. However, replacing maize or wheat with barley in rations

for pigs has sometimes resulted in poorer growth and feed utilisation (McConnell et al.,
1975; Lawrence, 1973) and reduced nutrient digestibility (Lawrence, 1972; Furuya and Kaji,

1991), indicating some anti-nutritive effects of the barleys employed. Furthermore,
supplementation of barley-based diets with fibre degrading enzymes has been reported to

improve pig performance and nutrient ileal (Graham et al., 1988; Bedford el al., 1992;
Inborr et al., 1993) and faecal digestibility (Thomke et al. , 1980; Markström et al., 1985),

although not consistently (Graham el al., 1986).
Viscous polysaccharides may exert their anti-nutritive activity by changing the rale of

gastric emptying (Meyer et al. , 1986; Rainbird & Low, 1986) and feed transit time (Cherbul

el al., 1990). They have also been shown to influence the amount of digestive secretions

(Ikegami et al., 1990). The fact that addition of /1-glucanase to diets based on barley
improves both nutrient digestibility and performance of pigs indicates that the enzyme is

capable of removing some of the anti-nutritive activities of the mixed-linked /1-glucans. In
broiler chickens, reduced intestinal viscosity following enzyme supplementation ofwheat and

barley-based diets has been found to be highly correlated with improved performance

(Bedford & Classen, 1992; Inborr el al., 1993). Similar relationships have not been found

in pigs fed enzyme-supplemented barley-based feeds (Bedford el al., 1992) and it is possible

that other mechanisms are of greater importance.

This investigation was undertaken to study the effect of /1-glucanase supplementation of

barley-based diets on pig performance and some gastrointestinal parameters relating to the



4

digestive processes. The aim of the study was to investigate the relationships between the

physiological parameters and pig performance in order to explain some of the mechanisms

behind the enzyme-induced performance responses.

EXPERIMENTAL

Animals and experimental design

A total of 40 pigs (alternating DL x LW) weaned at about four weeks of age with an

average live weight of 9.5 kg were divided into five blocks of eight and placed into
individual metabolism cages and assigned to one of four dietary treatments based on live

weight, sex and litter origin. The pigs of each block were fed the experimental diets for 21-
days. There were two treatment replicates per block with each replicate consisting of four

litlermale pigs. The experimental diets were based on one of two barley cultivars either with

or without added /5-glucanase to give a 2 x 2 factorial design.

Diets and feeding

The basal diets were based on barley, soybean and fish meal (Table 1). There were two

barley cultivars, one hulless (var. Condor, Canadian Wheat Board, Winnipeg, Canada) with

relatively high /1-glucan content and low crude fibre, and one hulled (var. Arra, Valtion

Viljavarasto, Ylivieska, Finland) with relatively low /J-glucan and high crude fibre content

(Table 1). The two basal diets were supplemented with a Trichoderma longibrachialum

cellulase preparation (Genencor International Ltd., Helsinki, Finland) with high /f-glucanase

activity. The diets were fed semi ad libitum in mash form during the entire experimental
period with water freely available.



5

Table 1. Composition of the experimental diets and the two barleys (g/kg)

Diet/barley... A C Arra Condor

Ingredients

Arra barley 755.4

Condor barley - 755.4

Soybean meal (48% CP) 150 150

Fish meal 40 40

Soya oil 20 20

Limestone 13 13

Dicalcium phosphate 10 10

Salt 3 3

HCI-Lysine 2.6 2.6

Vilamin/mincral mix 1 4 4
Chromic oxide (marker) 2 2
/1-glucanase 2 -/0.25 -/ 0.25

Nutrients (analysed)

Protein (Nx6.25) 225 224 159 171

HCI-fat 63 65 30 22
Ash 64 63 45 49

Starch 391 452 513 583
Total /?-glucan 20 34 26 48

Crude fibre 61 21
Total NSP 178 142 195 183

1 Provided the following (mg/kg diet); retinol 5.5, cholecalciferol 0.05, «-tocopherol 85,
menadione 2.4, thiamin 2.4, riboflavine 4.8, pyridoxine 3.6, pantothenic acid 12, nicotinic
acid 24, biotin 0.24, cyanocobalamin 0.024, Fe (FeSO 4 « 7H 20) 200, Zn (ZnO) 200, Mn
(MnO) 55.4, Cu (CuSO4

. 5H 20) 165, I (KI) 0.3, Se (Na2Seo 3) 0.3.
2 added at the expense of barley in diets A+ and C+ (see section Diets and feeding)



6

Digesta sample collection

On day 21 the pigs were sacrificed by means of an overdose of sodium pentobarbiturate and

bled. The digestive tract was rapidly removed and divided into the following sections by

ligatures; stomach (Sto), four small intestinal sections of equal length (SH, SI2 SI3 and SI4),

caecum (Cae), ascending (Col) and descending (Co2) colon and rectum (Rec). The total

contents of each section was carefully collected and weighed. Aliquots of suitable size were

taken immediately after collection for the viscosity measurements. The other samples were

frozen immediately and stored at -20 °C until needed for further analysis.

Chemical analyses and measurements

All analyses were made in duplicate. Cr 203, nitrogen and digesta viscosity determinations

were performed on wet material. All other analyses were carried out on freeze-dried

materials. Dry matter content of feed and digesta was determined by drying at 105 °C to

constant weight. Protein (N x 6.25) was determined by the Kjeldahl method using a Kjell-

Foss 16200 autoanalyser and energy by an IKA calorimeter C 400 (Janke & Kunthel KG

IKA-Werk, Germany) using benzoic acid (BCS-CRM No. 190 n BAS) for calibration. Ash

was analysed according to AOAC (Association of Official Analytical Chemists, 1975), while

fat (hydrochloric acid-fat) was extracted with diethyl ether after acid-hydrolysis (Stoldt,

1957). Cr 203 was determined using the method of Schiirch et al. (1950). Total /1-glucan was

determined by the enzymatic methods of McCleary & Glennie-Holmes (1985) and starch was

analysed by modification of the enzymatic method of Bach Knudsen et al. (1987). Starch was

gelatinised and quantitatively removed by incubation (100 °C, 60 min., 60 °C, 16 h) with

thermostable a-amylase (EC 3.2.1.1. Termamyl®, Novo Nordisk A/S, Denmark), which at

100" C was /l-glucanasc-free, and a /1-glucanase-free amyloglucosidase (EC 3.2.1.3., Cat No.

1060 074, Boehringcr Mannheim, GmbH, Mannheim, Germany). The resulting glucose

monomers were quantified with glucose oxidase reagent (EC 1.1.3.4., Cat No. 124001,
Boehringcr Mannheim GmbH).

Digesta viscosity was measured in the supernatant after centrifugation of the digesta
samples at 13,000 rpm for 3 minutes (Biofuge A; Heareus Sepatech GmbH, Germany) using



7

a Brookfield viscometer (model DV-II + ; Brookfield Engineering Laboratories Inc.,
Stoughton, MA, USA) operated at room temperature.

Pigs were weighed on days 1,7, 14 and 21 and the amount of feed consumed recorded at

the end of each week.

Enzyme activity analyses

/I-glucanase (EC 3.2.1.6; endo-/?-(l,3)(l,4)-glucanase) activity measurements of the feed

and stomach samples were carried out viscosimetrically using 1.0% barley /5-glucan (Biocon

Biochemicals Ltd., Cork, Ireland) in Mcllvaine’s buffer as substrate by the method described

by Bathgate (1979) with minor modifications. This method measures the reduction in

substrate viscosity and enzyme activity is expressed as the increase in reciprocal viscosity

(IRV).

1% (w/w) /1-glucan substrate was prepared by solubilising 1 g of barley /1-glucan (Biocon

Biochemicals Ltd., Cork, Ireland) in 6 ml of ethanol in a tared beaker. 10 ml of 0.5 M

sodium acetate buffer (pH 4.0) was added and then distilled water to a final volume of 100
ml. The /?-glucan solution was standardised to initial reciprocal specific viscosity of 0.13 by
adjusting the /1-glucan:buffer ratio. In the measurements the volume of the /1-glucan substrate

varied between 5.3 and 5.5 ml and that of the enzyme solution between 2.0 and 2.2 ml. The

total volume of the substrate/enzyme solution was adjusted to 7.5 ml before each

measurement.

After enzyme addition, four to six flow times were determined over a 30 minute lime

period in an Oslwald capillary viscometer (Model No. 11, 75-100 s), immersed in a water

bath set at 30°C. The flow times were used to calculate the reciprocal specific viscosities,
using the equation;

dT0

1/7sr
=

dTs - dT0

where l/j/sp is the reciprocal specific viscosity, dT0 the flow lime of the buffer and dTs the
flow lime of the substrate-enzyme solution.



8

/1-glucanasc activity is expressed as increase in reciprocal specific viscosity (IRV) and

calculated using the equation;

k * D
/?-glucanase, IRV Units/g =

V

where k is the slope of the curve plotted from the reciprocal specific viscosity versus

hydrolysis time (minutes), D is the total dilution factor and V sample volume. One IRV unit

is defined as the change of one iVrjsp ) min' 1 . Each measurement was carried out in

duplicates.

Enzyme activity analyses of the samples obtained from the three proximal quarters of the
small intestine were performed on lyophilized material, which was extracted with ImM HCI

(50 mg lyophilized digesta in 1 ml HCI) for one hour at 4°C followed by centrifugation
(3000Xg). The supernatants were then collected for activity analyses.

Trypsin (EC 3.4.21.4) activity was determined using Benzoyl-DL-arginine-p-nitroanilidc
(DL-BAPA) as substrate according to Erlanger et al., (1961) with minor modifications.

Aliquots of 450 fil of substrate were incubated at 37 °C and mixed with 50 //I extract before
transferred to a cuvette equlibraled at 37 °C in a Milton Roy (spectronic 1201)

spectrophotometer. Absorbance was measured every 20 seconds for 100 seconds and enzyme

activity was calculated from the initial velocity of the reaction. Chymotrypsin (EC 3.4.21.1)

was determined according to Erlanger et al. (1966) and Boisen et al. (1981) using glutaryl-1-

phenylalanin-p-nitroanilid (GPNA) as substrate.

Amylase (EC 3.2.1.1) activity was determined using the Phadebas® Amylase Test kit

(Pharmacia Diagnostics, Uppsala, Sweden) and lipase (EC 3.1.1.3) by a pH-slat titration

method using tributyrine as substrate according to Erlandsson-Albertsson et al., (1987). 15
ml of buffer containing ImM Tris-HCI at pH 7.0, ImM CaCl2, 150 mM NaCl and 4mM

sodium taurodeoxycholate were mixed with 500 /rl tribulyrin and 100-400 fi\ of extract.

Enzyme activities are expressed as units (micromoles substrate hydrolysed per minute per
ml) per gram sample dry matter and and per total segment dry matter (U/gXg segment dry
matter).



9

Calculations and statistical analyses

All data were subjected to ANOVA according to the GLM procedure (Statistical Analysis

System Institute, 1982). Where relevant, meanswere separated by orthogonal contrasts. Main

effects, or in the event of interactions, single effects are displayed. When analysing the

performance data the barley Xenzyme interaction was found non-significant and, therefore,
excluded from the model. Data presented graphically were analysed by linear regression

analysis using the linear and quadratic coefficients of gut segment. Nutrient digestibilities
were calculated relative to the indigestible marker (Cr2o3) content;

r2*-'3(diel) X X(digesU)
Digestibility of X = 1 - (1)

X

where X is the nutrient in question. X(die.t) and X(digesla) are concentrations of specific nutrients

in the diet and digesta taken from the different sections of the GI tract. Due to the small

samples obtained in the two most proximal sections of the small intestine these samples were

pooled when possible. When calculating starch digestibility it was assumed that free glucose
in the digesta was derived from starch.

RESULTS

Chemical composition of the barleys and experimental diets

Diels based on the Arra barley contained less starch and /J-glucans than those based on

Condor barley (Table 1). This reflected the concentrations of these two constituents in the

barleys. The content of protein, fat and ash was equal in all experimental diets.

Pig performance
Feed intake was not influenced, whereas live weight gain (P=0.074) and feed conversion

ratio (P=0.058) were improved by the enzyme treatment (Table 2). Pigs fed the diets based

on Arra barley tended to have lower body weight gain (P=0.16) than those fed Condor.



Table 2. Live weight gain (gAi), feed intake (g/d) and feed conversion ratio of the pigs during the three-week period.

Diet... A- A+ C- C+ P-values for main effects Pooled

barley enzyme variance

Live weight gain 200 208 223 229 0.1600.074 0.022

Feed intake 340 339 350 337 0.6750.340 0.028

Feed conversion 1.7081.646 1.5971.516 0.2270.058 0.146



10

Digestibility of nutrients

/J-glucan digestibility was significantly higher in the diets containing the added /1-glucanase

(Figure 1) but there was no significant effect of the enzyme on dry matter or starch

digestibility (Figures 2 and 3). Dry matter digestibility of the Condor-based diets was higher

(P <0.001) than of those based on Arra, whereas regarding starch digestibility the order was

reversed between the barleys (P=0.154). Protein digestibility tended to be higher for the

enzyme-supplemented diets (P=0.117) with the greatest improvement due to enzyme

supplementation observed in SI3 (Figure 4).

Figure 1. /?-glucan digestibility in the stomach (Sto), proximal two (Sll+2), third (SB) and
fourth (SI4) quarters of the small intestine, caecum (Cae), ascending colon (Col), descending
colon (Co2) and rectum (Rec) of pigs fed Arra barley without (O) or with (•), or Condor
barley without (□) or with (■) added /?-glucanase. No significant barleyXenzyme
interaction observed. Regression lines shown for barley with (—) and without (—) enzyme
supplementation.
Regression = 0.248(0.051) + 0.144(0.023) Xsegment - 0.004(0.003) X segment Xsegment
+ 0.1636(0.033)Xenzyme inclusion level. Model P-valuc <O.OOOl (r2 =0.643).



Figure 2. Dry matter digestibility in the stomach (Sto), proximal two (SII + 2), third (SI3)
and fourth (SI4) quarters of the small intestine, caecum (Cae), ascending colon (Col),
descending colon (Co2) and rectum (Rec) of pigs fed Arra barley without (O) or with (•),

or Condor barley without (□) or with (■) added /?-glucanase. No significant effect of
enzyme observed. Regression lines are shown for Arra (—) and Condor (—).

Regression =-0.070(0.025) + o.264(o.ol2)Xsegment - 0.019(0.001 )xsegmentXsegment
+ o.osB(o.ol2)Xbarley. Model P-value = 0.0001 (r2 =0.844).

Figure 3. Starch digestibility in the stomach (Sto), proximal two (SII + 2), third (SI3) and
fourth (Sl4) quarters of the small intestine, caecum (Cae), ascending colon (Col), descending
colon (Co2) and rectum (Rec) of pigs fed Arra barley without (O) or with (•), or Condor
barley without (□) or with (■) added /l-glucanase. No significant effect of enzyme
observed. Regression lines shown for Arra (—) and Condor (—).

Regression = -0.249(0.036) + o. 44s(o.ol7) Xsegment - 0.038(0.002) X segment Xsegment
+ 0.034(0.024)Xbar1ey. Model P-value = 0.154 (r2 =0.820).



12

Energy digestibility in the last quarter of the small intestine (segment SI4) and rectum was

not influenced by the enzyme treatments, whereas the values for Condor were significantly

higher (P =0.017 in Sl4 and P=o.ool in rectum) than those for Arra (Table 3).

Enzyme activities in feed and digesta

Endogenous /J-glucanase activity of the control diet based on Arra was higher than that of

the diet based on Condor (Table 4). Stomach digesta /Tglucanase activity of the enzyme-

supplemented diets was approximately 25% of that in the diets when expressed on a fresh
material basis, which reflected the dilution of the feed by water and gastric secretions.

Figure 4. Nitrogen digestibility in the stomach (Sto), proximal two (Sll+2), third (SB) and
fourth (Sl4) quarters of the small intestine, caecum (Cae), ascending colon (Col), descending
colon (Co2) and rectum (Rec) of pigs fed Arra barley without (O) or with (•), or Condor
barley without (□) or with (■) added /?-glucanase. No significant barleyXenzyme
interaction observed. Regression lines shown for barley with (—) and without (—) enzyme
supplementation. Regression = -0.059(0.035) + 0.189(0.014)X segment -

0.010(0.001) X segment Xsegment + o.o2Bs(o.o23)X enzyme inclusion level. Model P-value
= 0.117 (r2 =0.784).



Table 3. Digestibility of energy in segment SI4 and rectum of pigs fed either Arra or Condor barley with and without supplementary ft-glucanase.

Diet... A- A+ C- C+ P-values for main effects Pooled
barley enzyme bar*enz variance

SI4 0.6080.650 0.7290.700 0.0170.851 0.3060.096

Rectum 0.8130.814 0.8620.873 0.0010.580 0.6400.028



Table 4. /J-glucanase activity of the diets and stomach digesta samples, mU/g fresh weight

Did... A- A+ C- C +

Feed 11.5 (3.5)* 95 (2.0) 2 (0.0) 89 (0.5)

n 2 2 2 2

Digesla 1.5 (0.6) 24 (6.6) 1.2 (0.6) 23.7 (9.4)

n 5 8 8 5

1 Values are means (standard deviation)

When expressed on a sample dry matter basis digestive enzyme activities in the three

proximal quarters of the small intestine of pigs fed the enzyme-supplemented diets were

significantly lower (P<0.08) than in pigs fed the control diets (Table 5). There was a

significant barleyXenzyme interaction (P=0.002) for chymotrypsin. This showed that

enzyme supplementation of the Condor-based diet significantly reduced chymotrypsin activity

but tended to have the opposite effect on the diet based on Arra (Table 6), whereas feeding
the two barleys without enzyme resulted in significantly higher chymotrypsin levels in pigs
fed the Condor compared to Arra barley.

Table 6. P-valucs for the chymotrypsin (mU/g digesta DM) interactions.

Enzyme Barley Pooled variance

Arra 0.068 0.085

Condor 0.015 0.067

No enzyme

Enzyme
0.023 0.071

0.044 0.078

13



Table 5. Trypsin, chymotrypsin, lipase and amylase activity in segments SII + 2 and SI3 of the GI-tract expressed as mU/g digcsta DM.

Diet.. A- A+ C- C+ P-values for main effects Pooled

barley enzyme bar'enz seg variance

Trypsin

Sll+2 10.15.2 9.86.4 0.5920.034 0.396 0 0001 6.38

SI3 10.915.9 15.2 13.9

Chymotrypsin

SU+2 0.120.14 0.170.07 0.9960.059 0.002 0.641 0.074

SI3 0.070.16 0.130.11

Lipase

Sll+2 98.653.5 93.049.3 0.653 0.077 0.689 0.791 76.6

SI3 54.781.5 89.4 87.2

Amylase
Sll+2 1073 479 659 422 0.2050.038 0.9700.025 576

SI3 920 1094 973 829

Table 7. Trypsin, chymotrypsin, lipase and amylase activity in segments Sll+2 and SI3 of the GI-tract expressed as mU/segment DM.

Diet A- A- C- C+ P-values for main effects Pooled
barley enzyme bar'enz aeg variance

Trypsin

Sll+2 49 15 45.2 23 0.6770.107 0.1410.0001 52.0

SI3 106 148 154 116

Chymotrypsin

Sll+2 0.54 0.40 1.04 0.39 0.3740.087 0.0470.005 0.76
SI3 0.751.35 1.270.97

Lipase

Sll+2 501 154 386 240 0.606 0.197 0.9790.001 553

SI3 643 751 874 804

Amylase
Sll+2 5209 1479 3569 2016 0.4170.050 0.9780.0001 3912

SI3 8290 9366 8460 7388



14

When the enzyme activities were expressed on a total segment dry matter basis, enzyme

supplementation significantly (P =0.050) reduced the amylase activity and tended to reduce

the chymolrypsin (P=0.087) activity (Table 7). There was no effect of barley on any of the

enzyme activities. A significant (P=0.047) barleyXenzyme interaction was observed for

chymolrypsin. Enzyme supplementation tended to reduce (P=0.106) chymolrypsin activity

in pigs fed the Condor barley (Table 8), whereas there was no effect of the enzyme in the

Arra-based diets. With no added enzyme, Condor produced higher chymolrypsin (P=0.067)

activity than Arra. Segment SI3 contained significantly more digestive enzyme activity than

segment SI 1+2.

Table 8. P-values of the barley*enzyme interactions for chymolrypsin

when enzyme activities were expressed on a segment dry matter basis.

Barley Enzyme Pooled variance

Ärra 0.2900.688

Condor . 0.1060.852

No enzyme 0.067 . 0.843

Enzyme 0.3740.661

Digesta viscosity and dry matter content

Enzyme supplementation significantly (P<0.03) reduced viscosity in the stomach and the

three proximal quarters of the small intestine (Table 9). In the caecum and the descending

colon of pigs fed the Condor barley and in the colon of the Arra-fed pigs this relation was

reversed but the differences were not significant. Condor barley caused significantly

(P <0.02) higher digesta viscosities in caecum and Co 2 than Arra. Dry matter content of the

digesta of the pigs fed the enzyme supplemented diets was significantly higher (P=0.0003)
than that of those fed the unsupplemented diets (Figure 5).



15

DISCUSSION

/3-glucanase activity of the stomach digesta samples showed full recovery of the added

enzyme when expressed on a dry matter basis (data not shown). This is In accordance with

results from in-vitro experiments (Inborr and Grönlund, 1993) and those obtained by

measuring /5-glucanase activity in the proximal small intestine of broiler chickens fed

enzyme-supplemented diets (Inborr and Bedford, 1993).

Figure 5. Dry matter content of the stomach (Sto), proximal two (SI 1 + 2), third (SI3) and
fourth (SI4) quarters of the small intestine, caecum (Cae), ascending colon (Col), descending
colon (Co2) and rectum (Rec) of pigs fed Arra barley without (O) or with (•), or Condor
barley without (□) or with (■) added /?-glucanasc. No significant barleyXenzyme
interaction observed. Regression lines shown for barley with (—) and without (—) enzyme
supplementation.
Regression = 21.8271(0.833)-4.631(0.377)X5egment - o. 662(o.o4l)XsegmcntXsegmcnt
+ 1.992(0.542)Xenzyme inclusion level. Model P-value = 0.0003 (r2 =0.659).



Table 9. Digesta viscosity (cP) of the different segments.

Diet... A- A+ C- C+ P-values for main effects Pooled

barley enzyme bar*enz variance

Stomach 1.38 1.33 1.71 1.31 0.146 0.029 0.098 0.3 l
511 1.48 1.43 1.93 1.32 0.384 0.021 0.065 0.36
512 1.93 1.74 2.94 1.77 0.123 0.004 0.057 0.48

513 2.77 2.01 3.82 2.33 0.143 0.007 0.188 0.96
514 2.55 2.44 2.94 2.52 0.773 0.306 0.556 0.68
Caecum 1.87 1.68 1.96 2.70 0.007 0.451 0.167 0.67
Colon 1 6.15 7.93 9.85 11.11 0.143 0,337 0.685 4.99

Colon 2 7.50 10.42 14.50 13.00 0.016 0.451 0.878 3.78



16

Addition of /?-glucanase to the diets improved live weight gain and feed utilisation of the

pigs. This effect is consistent with the results of performance experiments, where /Tglucanase
(Thomke et al., 1980) or /Tglucanase-containing enzyme mixtures (Böhme, f990) were fed

to early-weaned pigs. In broilers, supplementary enzymes reduce the viscosity of the small

intestinal chyme (Bedford and Classen, 1992), thereby improving nutrient digestibility and

absorption (Almirall et al., 1993). A similar, but apparently weaker, relationship appears to

exist also in young pigs (Bedford et al. , (1992). In the present experiment, using the same

technique as in the last cited experiments, enzyme supplementation reduced digesta viscosity
in the stomach and small intestine. This is in agreement with the previously reported results

and suggests a similar enzyme mode of action of supplementary /?-glucanase in the young pig
as in broilers fed viscous polysaccharides. The reduction of the viscosity coincided with a

dramatic increase in /1-glucan digestibility, suggesting that this component was an important

contributor to the intestinal viscosity. This effect on the /1-glucans are in agreement with the

results reported by Graham et al. (1989) when a barley-based diet was fed to finishing pigs.

However, the level of viscosity and magnitude of the reduction due to enzyme
supplementation in the present experiment were considerably lower than that reported in
broiler chickens (Inborr and Bedford, 1993). This, in combination with a relatively low

number of observations, may explain why no significant correlation was found between

reduced gut viscosity and improved feed utilisation (data not shown). However, it has

recently been shown that in broiler chickens the relative reduction of gut viscosity due to

enzyme supplementation is more important than the absolute value (Graham and Bedford,
1993). Consequently, a 50% reduction at any level of viscosity should result in a

performance improvement of similar magnitude. If this relationship is valid also in the pig,

the reduction in small intestinal viscosity observed in the present experiment should translate

into an improved performance. The improved live weight gain and feed utilisation observed

in the present experiment would support this.

Increasing gut viscosity by feeding viscous polysaccharides such as guar gum has been

shown to increase the output of pancreatic juice and enzyme activities (Ikcgami et al., 1990).

In the present experiment, /1-glucanase supplementation generally reduced the amount of the

digestive enzymes in the first three quarters of the small intestine. With Condor barley

creating a more viscous environment, this effect was particularly evident. Among the

enzymes monitored, amylase was the one that seemed most affected by the /J-glucanase.



17

However, starch digestibility was not reduced, which suggests that there was sufficient starch

degrading activity despite the considerably lower amylase concentration. Corring (1982)

reported that a pig of 50 kg live weight produced enough amylase to hydrolyse 30 kg of

soluble starch daily. It, therefore, seems unlikely that even a considerable reduction of

amylase production would reduce starch digestibility, in particular, if more optimal

conditions for enzymatic hydrolysis are concomitantly being created. Leclerce et al. (1993)

reported reduced starch digestibilities in jejunum and ileum of pigs fed diets with increasing

amounts of viscous guar gum. Interestingly, these workers reported reduced amylase activity
in the small intestine when more viscous guar gums were used. This is in contrast with the

results of the present and other experiments (Ikegami et al., 1990; Cherbut et al., 1990).

Increased viscosity has been shown to reduce the rate of nutrient diffusion in-vitro (Fengler

and Marquardt, 1988) and to reduce nutrient digestibility in broiler chickens (Choct &

Annison, 1992). Consequently, one can assume that by reducing intestinal viscosity less

digestive enzymes are required to achieve the same rate of digestion as in a more viscous

environment. This assumption is supported by the results obtained in this experiment, with

no reduction in nutrient digestibility despite on average 20 per cent less digestive enzyme

activities. In fact, protein digestibility tended to be improved due to the addition of the

enzyme. Bedford et al. (1992) observed a significantly higher protein digestiblity over the

whole small intestine of early-weaned pigs when a barley-based diet was supplemented with

/1-glucanase. They assumed this effect being mainly related to improved utilisation of the

dietary protein. The results from the present experiment would suggest that a reduction in

the production of digestive enzymes is also contributing to the increased protein digestibility.

The energy saved by the reduced production of digestive enzymes is thus potentially available

to increase the efficiency of nutrient assimilation and improve growth performance and feed

utilisation. Again, the observed significant improvement in live weight gain and feed

utilisation due to enzyme addition would support this assumption. The daily exocrine

pancreatic protein output has been estimated to between 2 and 6 g in pigs during the weeks

following weaning (Pierzynowski, 1991), whereas in older pigs values between 9 and 25 g
have been reported (Souffrant, 1991). Assuming 10 g daily protein output in digestive
enzyme production (including intestinal amylase) in the pigs of the present experiment a 20

per cent reduction would mean 2 g less protein required for the digestive enzymes. This

represents approximately 5 per cent of the daily protein retention of pigs growing at a similar



18

rale as the pigs in the present study (Lassota, 1987), which would then be availvable for lean

growth. Conversely, this could lead to a more efficient utilisation of dietary energy and

protein, resulting in an improved feed utilisation.

In conclusion, enzyme supplementation significantly reduced digesta viscosity in the small

intestine and reduced the concentration of digestive enzymes without affecting nutrient

digestibility. It appears that conditions for a more energy efficient digestion were brought
about by the addition of the /3-glucanase, reducing endogenous losses and leading to increased

weight gain and improved feed utilisation.

ACKNOWLEDGEMENTS

This experiment was carried out at the National Institute of Animal Science (NIAS),

Department of Animal Physiology and Biochemistry, Research Centre Foulum, Denmark

under the supervision of Dr Kirsten Jakobsen and Ms Mette Skou Jensen. Their assistance

and the technical support of Dr K.E. Bach Knudsen and Dr B. Borg Jensen are highly

appreciated. The experiment was funded by NIAS and Finnfeeds International Ltd.,
Marlborough, United Kingdom.



19

REFERENCES

Åman, P & Graham, H. (1987). Analysis of total and insoluble mixed-linkcd (l-3),(l-4)-/J-D
glucans in barley and oats. Journal of Agricultural Food Chemistry 35, 704-709.

Association of Official Analytical Chemists (1975). Official Methods of Analysis. 11th ed.
Washington D.C.: Association of Official Analytical Chemists.

Bach Knudsen, K.E., Åman, P & Eggum, 8.0. (1987). Nutritive value of danish-grown
barley varilies, I, Carbohydrates and other major constituents. Journal of Cereal Science 6,
173-186.

Bathgate, G.N. (1987). The determination of endo-/?-glucanase activity in malt. Journal of
the Institute of Brewing 85, 92-94.

Bedford, M.R. & Classen, H.L. (1992). Reduction of intestinal viscosity through
manipulation of dietary rye and pentosanase concentration is effected through changes in the
carbohydrate composition of the intestinal aqueous phase and results in improved growth rate
and feed conversion. Journal of Nutrition 122, 560-569.

Bedford, M.R., Classen, H.L., Patience, J.F. & Inborr, J. (1992). Effect of dietary enzyme
supplementation of rye- and barley-based diets on digestion and subsequent performance in
pigs. Canadian Journal of Animal Science 72, 97-105.

Boisen, S., Andersen, C. Y. & Hejgaard, J. (1981). Inhibitors of chymotrypsin and
microbial serine proteases in barley grains. Physiology of Plants 52, 167-176.

Böhme, H. (1990). Experiments on the effect of enzyme supplements as a growth promotor
for piglets. Landbauforschung Völkenrode 40, 213-217.

Cherbul, C., Albina, E., Champ, M., Doublier, J.L. & Lecannu, G. (1990). Action of guar
gums on the viscosity of digestive contents and on the gastrointestinal motor function in pigs.
Digestion 46, 205-213.

Chocl, M. & Annison, G. (1992). Anti-nutritive effect of wheat pentosans in broiler
chickens: roles of viscosity and gut microflora. British Poultry Science 33, 821-834.

Corring, T. (1982). Enzyme digestion in the proximal digestive.tract of the pig: A review.
Livestock Production Science 9, 581-590.

Erlangcr, 8.F., Kokowsky, N. & Cohen, W. (1961). the preparation and properties of two
new chromogenic substrates of trypsin. Archive of Biochemical Biophysiology 95, 271-278.

Erlangcr, 8.F., Edel, F. & Cooper, A.G. (1966). the action of chymotrypsin on two
chromogenic substrates of trypsin. Archive of biochemical Boiphysiology 115, 206-210.



20

Erlanson-Albertsson, C., Larsson, A & Duan, R. (1987). Secretion of pancreatic lipase and
colipase from rat pancreas. Pancreas 2, 531-535.
Fengler, A.S. & Marquardt, R.R. (1988). Water-soluble pentosans from rye; 11. Effects on
rate of dialysis and on the retention of nutrients by the chick. Cereal Chemistry 65, 298-302.

Furuya, S. & Kaji, Y. (1991). Additivity of the apparent and true ileal digestible amino acid
supply in barley, maize and wheat or soya-bean meal based diets for growing pigs. Animal
Feed Science and Technology 32, 321-331.

Graham, H. (1988). Dietary fibre concentration and assimilation in swine. ISI Atlas of
Science: Animal and Plant Sciences 1, 76-80.

Graham, H. & Bedford, M.R. (1993). Feed enzymes: Mode of action and application to heat
processed poultry feeds. Amandus Kahl Seminar, EUROTIER, Hannover, Germany.
Manuscript, pp.l2.

Graham, H., Åman, P. and Löwgren, W. (1988). Enzyme supplementation of pig diets. In
Proceedings of the 4th International seminar on Digestive physiology in the pig, pp 371-376
[L. Buraczewska, S. Buraczewski, B. Pastuszewska and T. Zebrowska, editors]. Jablonna,
Polish Academy of Sciences.

Graham, H., Fadel, J.G., Newman, C.W. & Newman, R. K. (1989). Effect of pelleting and
/J-glucanase supplementation on the ileal and fecal digestibility of a barley-based diet in the
pig. Journal of Animal Science 67, 1293-1298.

Hesselman, K. (1983). PhD Thesis. Swedish University of Agricultural Sciences, Uppsala.

Hessclman, K. & Thomke, S. (1982). Influence of some factors on development of viscosity
in the water-extract of barley. Swedish Journal of Agricultural Research 12, 17-22.

Ikegami, S., Tsuchihashi, F., Harada, H., Tsuchihashi, N., Nishide, E. & Innami, S.
(1990). Effect of viscous indigestible polysaccharides on pancreatic-biliary secretion and
digestive organs in rats. Journal of Nutrition 120, 353-360.

Inborr, J. & Bedford, M.R. (1993). Stability of feed enzymes to steam pelleting during feed
processing. Animal Feed Science and Technology. (In press).

Inborr, J. & Grönlund, A. (1993). Stability of feed enzymes in physiological conditions
assayed by in-vitro methods. Agricultural Science in Finland 2, 125-132.

Inborr, J., Bedford, M.R. & Graham, H. (1993). Stability and mode of action of poultry
feed enzymes in diets based on wheat and barley. In Proceedings of the Australian Poultry
Science Symposium, Vol. 5, 53-56. University of Sydney, Sydney, February.

Inborr, J., Schmitz, M. & Ahrens, F. (1993). Effect of adding fibre and starch degrading
enzymes to a barley/wheat-based diet on performance and nutrient digestibility in different
segments of the small intestine of early weaned pigs. Animal Feed Science and Technology
44. 113-127.



21

Lassota, L. (1988). The efficiency of nitrogen utilisation in very early weaned pigs. In
Proceedings of the sth International Symposium on Protein Metabolism and Nutrition
[Wissenschaflliche Zeitschrift der Wilhelm-Pieck-Universität, Rostock], EAAP publication
No. 35, pp. 67.

Lawrence, T.L.J. (1972). High level cereal diets for the growing/finishing pig. VI. An
evaluation of flaked maize, wheat and barley when included at high levels in the diet of the
weaned pig grown to cutter weight (160 lb). Journal of Agricultural Science, Cambridge, 79,
155-160.

Lawrence, T.L.J. (1973). An evaluation of the micronization process for preparing cereals
for the growing pig. 2. Effects on growth rate, food conversion efficiency and carcass
characteristics. Animal Production 16, 109-116.

Leclerce, C., Champ, M., Cherbut, C. & Delorl-Laval, J. 1993. Starch digestion and
amylase activity in the small intestine in the presence of guar gums. Sciences des aliments,
13, 325-332.

Low, A.G. (1985). The role of dietary fibre in digestion, absorption and metabolism. In
Proceedings of the 3rd International Seminar on Digestive physiology in the Pig, pp 157-179
[A. Just, J. Fernandez and H. Jorgensen, editors], Copenhagen: National Institute of Animal
Science, pp. 157-179.

Markström, 8., Pettersson, D. & Hesselman, K. (1985). /J-glukanastillsats till grov- och
finmalet korn - ett smältbarhetsförsök med smågrisar. Sveriges Lantbruksuniversitet, Inst.
för husdjurens utfodring och vård. Rapport 149. Swedish University of Agricultural Sciences,
Uppsala.

McConnell, J.C., Skelley, G.C., Handiin, D.L. & Johnston, W.E. (1975). Corn, wheat,
milo and barley with soybean meal or roasted soybeans and their effect on feed lot
performance, carcass traits and pork acceptability. Journal of Animal Science 41, 1021-1030.

Meyer, J.H., Gu, Y., Elashoff, J., Reedy, T., Dressman, J. & Amidon, G. (1986). Effects
of viscosity and fluid outflow on postcibal gastric emptying of solids. American Journal of
Physiology 250, Gl6l-G164.

McCleary, B. V. & Glennie-Holmes, M. (1985). Enzymic quantification of (l-3),(l-4)-/?-D-
-glucan in barley and malt. Journal of the Institute of Brewing 91, 285-295.

Pierzynowski, S. (1991). Development and regulation of porcine pancreatic function with
special reference to the exocrine pancreas. PhD Thesis, University of Lund, Sweden.

Rainbird, A.L. & Low, A.G. (1986). Effect of various dietary fibres on gastric emptying of
growing pigs. British Journal of Nutrition 55, 111-121.

Statistical Analysis System Institute, Inc. (1982). SAS user’s guide: Statistics. SAS Inst., Inc.
Cary, NC.



22

Schiirch, A., Lloyd, L. & Crapton, E.W. (1950). The use of Chromic oxide as an index for
determining the digestibility of a diet. Journal of Nutrition 50, 629-636.

Souffrant, W.B. (1991). Endogenous nitrogen losses during digestion in pigs. In Proceedings
of the Vth International Symposium on Digestive Physiology in Pigs, pp 147-167 [M.W.A.
Verstegen, J. Huisman and L.A. den Hartog, editors], EAAP publication No. 54. Pudoc,
Wageningcn.

Stoldt, W. (1957). Vorschlag zur Vereinheitiichung der Fettbestimmung in Lebensmitteln.
Fetle, Seifen, Anstrichmittel 54. 206-207.

Thomke, S., Rundgren, M. & Hesselman, K (1980). The effect of feeding high-viscosity
barley to pigs. In Proceedings of the 31st annual meeting of the European Association of
Animal Production, Munich, Germany. Commission on Animal Nutrition, pp 5.

While, W.8., Bird, H.R., Sunde, M.L., Prentice, N., Burger, W.C. & Martlett, J.A.
(1981). The viscosity interaction of barley beta-glucan with Trichoderma viride cellulase in
the chick intestine. Poultry Science 60, 1043-1048.