BIOTROPIA No. 8, 1995: 1 - 10 

NUTRIENT TRANSFER IN VESICULAR-ARBUSCULAR MYCORRHIZAS: 

A NEW MODEL BASED ON THE DISTRIBUTION OF ATPases ON FUNGAL 

AND PLANT MEMBRANES*) 

F.A. SMITH 
Department of Botany, The University of Adelaide, South Australia, 5005, 

Australia 

S.E. SMITH 
Department of Soil Science, The University of Adelaide, South Australia, 5005, 

Australia 

ABSTRACT 

In this paper we review the membrane transport processes that are involved in the transfer of mineral 

nutrients and organic carbon between the symbiotic partners in mycorrhizas. In particular, we reassess the 

prevailing hypothesis that transfer in vesicular-arbuscular (VA) mycorrhizas occurs simultaneously and 

bidirectionally across the same interface and that arbuscules are the main sites of transfer. 
Using cytochemical techniques, we and our collaborators have reexamined the distribution of ATPases 

in the arbuscular and intercellular hyphal interfaces in VA mycorrhizas formed between roots ofAllium cepa 

(onion) and the fungus Glomus intraradices. The results showed that H 
+
-ATPases have different localisation on 

plant and fungal membranes in arbuscular and hyphal interfaces (Gianinazzi-Pearson et al. 1991). While some 

arbuscular interfaces had H
+
-ATPase activity on both fungal and plant membranes, in most cases the fungal 

membrane lacked this activity. In contrast, the plasma membranes of intercellular hyphae always had H 
+
 

-ATPase and the adjacent root cells did not. This suggests that the different interfaces in a VA mycorrhiza may 

have different functions. 
We propose that passive loss of P from the arbuscules is associated with active uptake by the energised 

(ATPase-bearing) plant membrane and that passive loss of carbohydrate from the root cells is followed by 

active uptake by the intercellular hyphae. If this model is correct, then variations in "mycorrhizal efficiency" 

(i.e. the extent to which mycorrhizal plants grow better than non-mycorrhizal controls) might be determined by 

differences in the numbers of active arbuscules as a proportion of the total fungal biomass within the root. 
As a first step towards investigating this possibility, we have developed methods for measuring the 

surface areas of arbuscular and hyphal interfaces in different fungus-host combinations, Glomus spp./ Allium 

porrum (leek). We have also measured fluxes of P from fungus to plant and have been able to partition these 

between the arbuscular and total (arbuscular plus hyphal) interfaces. The implications of this work, and 

suggestions for future investigations of the molecular mechanisms involved in nutrient transfer in mycorrhizas, 

are discussed. 

Key words: Mycorrhizas/Glomus intraradices/ATPases/Allium cepa. 

*)Paper presented at the Second Symposium on Biology and Biotechnology of Mycorrhizae and Third Asian 

Conference on Mycorrhizae (ACOM III), 19-21 April, Yogyakarta, Indonesia. 

 

1 



BIOTROPIA No. 8, 1995 

SMITH, S.E., S. DICKSON, C. MORRIS and F.A. SMITH. 1994b. Transfer of phosphate from fungus to plant in VA 

mycorrhizas: calculation of the area of symbiotic interface and fluxes of P from two different fungi to 

Allium porrum. New Phytologist. In press. 

SMITH, S.E. and F.A. SMITH. 1990. Structure and function of the interfaces in biotrophic symbioses. New 

Phytologist, 114, 1-38.  

SMITH, S.E. and F.A. SMITH. 1994. Membranes in mycorrhizal interfaces: specialised functions in symbiosis. 

In: Membranes - Specialised Functions in Plant Cells. Ed. by M. Smallwood and D. Bowles, JAI Press 

Inc. (in press). 

 TESTER, M. 1990. Plant ion channels: whole cell and single channel studies. New Phytologist, 114, 305-340. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 



BIOTROPIA No. 8, 1995 

  

Transfer of nutrients is of central importance in the functioning of the very 

widespread symbioses between mycorrhizal fungi and their host plants. The host gains 

mineral nutrients (e.g. phosphate and zinc), and in all mycorrhizas except orchid 

mycorrhizas the fungus gains carbohydrate. This two-way transfer occurs across 

interfaces which comprise the membranes of both organisms and an apoplastic region 

between them. The interfacial apoplast may contain considerable amounts of wall or 

wall-derived material, which varies according to the type of mycorrhiza and the particular 
interface (Smith and Smith 1994). In ectomycorrhizas, the interface is entirely intercellular 

(extracellular), whereas in some endomycorrhizas (e.g. ericoid mycorrhizas), the interface 

is virtually entirely intracellular. Other endomycorrhizas (e.g. vesicular-arbuscular 

mycorrhizas) have both an intercellular interface — where the fungus penetrates 

between cortical cells, and an intracellular interface - where the fungus penetrates the 

cell walls of its host and develops arbuscules. 

In all cases, the interface has a restricted volume. It is isolated from the external 

environment, and membrane transport processes in both organisms will control the 

physico-chemical conditions in the apoplast and hence the overall transport of solutes from 

one organism to the other. Fig. 1 summarises the uptake and transfer of nutrients in all 

mycorrhizas except orchids. As well as emphasising phosphate (P) and 

 

Figure 1. Transfer of mineral nutrients and carbohydrates (sugars) in mycorrhizas (excluding orchids). With the 

exception of H 
+
 -ATPases (see text), the mechanisms of membrane transport are not specified, fpm: 

fungal plasma membrane; rpm: root plasma membrane; Pi: inorganic phosphate. 

2 

INTRODUCTION 



Nutrient transfer in vesicular-arbuscular mycorrhizas - F.A. Smith and S.E. Smith 

zinc (Zn), it shows that other mineral nutrients can be transferred from soil to the host via 

the fungus. In the case of NH4
+
, there is conversion to organic N compounds and 

subsequent transfer of some of these across the interface. The transfer of sugars to the 

fungus is shown in Fig. 1, and also the possible transfer of "growth factors" i.e. 

phytohormones, vitamins etc. Finally, Fig. 1 shows the presence of H
+
-transporting 

ATPases that are present in the functional plasma membranes of all fungal and plant cells. 

These "energise" the membranes both electrically, via the generation of electric potential 

differences (PDs), and chemically via the development of pH differences, and so allow 

selective transport of other solutes. For example, phosphate and sugars are transported 

into fungal and plant cells by diferrent transport proteins which catalyse co-transport of 

phosphate or sugar with H
+
 (not shown in Fig. 1, for simplicity). Other types of transport 

proteins are "channels" which can open and close in response to various signals at the 

plasma membrane, including changes in PDs and pH. For a general discussion of interfaces 

and transport in mycorrhizas, see Smith and Smith (1990, 1994). Tester (1990) describes 

factors that control the operation of channels. 

MEMBRANE TRANSPORT AND ATPases IN VA MYCORRHIZAS 

For VA mycorrhizas, the prevailing hypothesis has been that transfer of 

nutrients and carbohydrates occurs bidirectionally and simultaneously across the same 

interface and that the arbuscules, having large surface-to-volume ratio, are important as 

the main site of transfer. It has been suggested in the past - and sometimes even now — 

that- P transfer can be explained by breakdown (digestion) of arbuscules, releasing P to the 

host. However, calculations by Cox and Tinker (1976) showed that this would be far too 

slow the account for the fluxes of P across the interface. To make these calculations, 

Cox and Tinker (1976) measured surface areas of arbuscules in the Glomus 

mosseae/Allium cepa (onion) symbiosis using electron microscopy and image analysis, 

the amount of infection and the total numbers of arbuscules per unit length of root. 

They also estimated the life- span of the arbuscules (about 4 days), their P content and 

the inflow of P - i.e. the uptake of P to the host per unit length of root. In this way, they 

showed that the P that could be released by digestion was about 100 times too small for 

the actual P flux; hence continuing transfer from living arbuscules is required. This is an 

important conclusion, and the calculated value for the P flux - 13 nmol m (arbuscular 

surface area)
-2s-1

 -is also important since it is similar to values for uptake of P into cells 

by active transport (i.e. by H
+
-P co-transport), but much larger than ty- 

 

3 



BIOTROPIA No. 8, 1995 

pical values for efflux of P. An important question thus becomes: what is the mechanism 

for release of P from the fungus? This issue is discussed by Smith et al. (1994a). 

The hypothesis of simultaneous two-way transport of P and sugar in VA my 

corrhizas is summarised in Fig. 2. Details of co-transport and passive transport are shown. 

The hypothesis has been supported by the cytochemical demonstration of ATPases on 

both fungal and plant plasma membranes in the arbuscular interface (Marx et al. 1982), 

but does not take into account the potential role of the intercellular fungal hyphae in 

transport processes (Smith and Dickson 1991). Likewise, the possible role of the 

intercellular interface was not considered by Cox and Tinker (1976). 

    

 

Figure 2. Diagrammatic representation of the distribution of H 
+
-ATPases and associated transport 

processes in VA mycorrhizas that would result in bidirectional transfer of sugar and phosphate 

across the same arbuscular interface. Based on Figure 7 of Gianinazzi-Pearson et al. (1991). pam: 

periarbuscular (plasma) membrane of the root cortical cells. 

 

 

 

4 



Nutrient transfer in vesicular-arbuscular mycorrhizas - F.A. Smith and S.E. Smith 

SPATIAL SEPARATION OF TRANSPORT? 

Gianinazzi-Pearson et al. (1991) re-examined the distribution of ATPases in the 

arbuscular and intercellular (hyphal) interfaces in VA mycorrhizas formed between 

roots of Allium cepa (onion) and Glomus intraradices. The cytochemical methods were 

similar to those used by others: they involved the deposition of lead phosphate at sites of 

ATPase activity and the use of inhibitors (e.g. vanadate and molybdate) to distinguish 

H
+
-ATPases likely to be involved in active membrane transport from other enzymes 

with ATP-hydrolysing ability. The results showed that H
+
-ATPases have different 

localisation on plant and fungal membranes in arbuscular and hyphal interfaces 

(Gianinazzi-Pearson et al., 1991). While some arbuscular interfaces had H
+
-ATPase 

activity on both fungal and plant membranes, in most cases the fungal membrane 

lacked this activity. In contrast, the plasma membranes of intercellular hyphae always 

had H
+
-ATPase and the adjacent root cells did not. 

These results suggest that the different interfaces in a VA mycorrhiza may have 

different functions (Fig. 3). The arbuscules may indeed play a major role in P transfer. 

Passive P loss from the fungus to the inter facial apoplast would be associated with 

active uptake (H
+
-P co-transport) by the energised (ATPase-bearing) plant 

membrane. Reabsorption of P from the interfacial apoplast by the fungus may be 

reduced in the arbuscules, because the fungal membrane frequently appears to lack 

ATPase activity and hence the capacity for active uptake. These 

Figure 3. Spatial separation of transfer in VA mycorrhizas. A: passive loss of P from the fungus and 

active uptake (H 
+
 -phosphate co-transport) across the energised periarbuscular membrane. B: 

passive loss of sugar to the intercellular apoplast and active uptake (H 
+
-hexose 

co-transport) across the energised hyphal plasma membrane. Based on Figure 7 of 

Gianinazzi-Pearson et al. (1991). 

5 



BIOTROPIA No. 8, 1995 

processes would result in net transfer of P from fungus to plant. We have shown that 

transfer of P between the symbionts occurs at rates consistent with high passive efflux from 

the fungus and active uptake by the plants (Smith et al. 1994a & b). Gianinazzi-Pearson 

et al. (1991) propose that the intercellular hyphae (but not the arbuscules) may be 

responsible for carbohydrate transfer from plant to fungus. In this case, passive loss of 

carbohydrate from the root cortical cells to the intercellular apoplast would be followed by 

active uptake by the fungal hyphae. As far as we are aware, the proposal that transfers of 

P and carbohydrate are spatially separated had not been made previously. 

ARBUSCULES AND VA MYCORRZHIZAL "EFFICIENCY" 

An attraction of this new model is that variations in mycorrhizal "efficiency" (i.e. the 

extent to which mycorrhizal plants grow better than non-mycorrhizal controls) might be 

explained by differences in the numbers or size of metabolically active arbuscules as a 

proportion of the total fungal biomass within the root. Without arbuscules, the fungus 

would be essentially a parasite, draining carbohydrate from its host but giving no 

phosphate in exchange. It is well known that mycorrhizal "efficiency" does vary in 

different fungus/host combinations, or under different environmental conditions, but 

there has been little attempt to correlate these variations with numbers of arbuscules or 

their activity. One reason is the difficulty of measuring numbers and surface areas of 

arbuscules, and length and surface area of intercellular hyphae, rather than just taking 

the length or percent of the root that contains fungus as a measure of infection. 

In order to remedy this deficiency, we have developed an updated version of the 

methods used by Cox and Tinker (1976), including computer-aided image analysis techniques 

and the use of nitroblue tetrazolium to distinguish metabolically active fungal tissues from 

the total fungal biomass (Smith and Dickson 1991). By means of a size facility in the 

program, it is possible to distinguish between intercellular hyphae and arbuscules. Table 

1 shows data for Allium porrum infected by Glomus sp. 'City Beach' (WUM16). 

Percentage infection was obtained conventionally and the areas of interface obtained by 

image-analysis. Details for the experiment and methods are given by Smith and Dickson 

(1991) and Smith et al. (1994b). Table 1 shows that the surface area of metabolically 

active arbuscules per metre of root declined with time, while that of hyphae increased. 

After 63 days, the hyphae had the larger surface area, illustrating the danger of ignoring 

them as a site of nutrient transfer, irrespective of what model is adopted. 

Table 2 shows the inflows of P (i.e. amounts of P taken up per metre of root length) 

into non-mycorrhizal and mycorrhizal plants, with the amount taken into 

 

 

 

 

 

6 



Nutrient transfer in vesicular-arbuscular mycorrhizas - F.A. Smith and S.E. Smith 

the latter via the fungus calculated by subtraction. We are aware that this calculation, 

which assumes that fungal infection does not affect P uptake by the host, may produce 

errors, but this possibility has been ignored in our calculations. Table 2 shows that 

inflow of P decreased with time, particularly in non-mycorrhizal plants. Table 3 shows the 

fluxes of P across the interfaces, as obtained by dividing the inflows (Table 2) by the 

average surface areas obtained from Table 1. The increasing divergence with time between 

the values that were obtained assuming that the flux occurred 1) only across the arbuscular 

interface and 2) across both arbuscules and intercellular hyphae reflects the increasing 

contribution of the latter to the total surface area. The range of values in Table 3 (3.7-12.8 

nmol m
-2
 s-

1
) agrees quite well with the single value of nmoll m

-2
 s

- 1
 that was calculated 

by Cox and Tinker (1976). 

 

 

7 



BIOTROPIA No. 8, 1995 

Although these results show that it is possible to make separate calculations for 

influx of P across the arbuscular interface and the total interface, they do not by 

themselves answer the question of whether the arbuscular interface alone is responsible 

for P transfer. To do this, it will be necessary to attempt to correlate P transfer with 

arbuscular surface area, e.g. in comparison with the total fungal biomass within the root. 

Results that we have obtained so far suggest that where different VA mycorrhizal fungi 

infect the same host species, there is no simple correlation. Table 4 shows again the 
results for Allium porrum infected by Glomus sp. 'City Beach' over 21-42 days of growth 

and compares them with results for Allium porrum/Glomus mosseae obtained at the 

same time and under the same conditions (Smith and Dickson 1991; Smith et al. 1994b). 

G. mosseae produced a slightly larger % infection but a slightly smaller inflow of P (19.0 

compared to 22.3 pmol m
- 1

s
- 1

)- The areas of interfaces with G. mosseae were more than 

twice that with G. sp. 'City Beach', and the flux of P via mosseae was accordingly much 

smaller than with G. sp. ' City Beach', whether calculated using the area of arbuscular 

interface or the total interface. These results show that G. sp. 'City Beach' loses P faster 

on a surface area basis. There are several possible explanations including not only 

differences in numbers of membrane transport proteins at the interface but also different 

rates of translocation of P through the fungi or different rates of uptake from the soil which 

could in turn be associated with different lengths of external hyphae. To pursue further the 
question of the role of the arbuscule as the sole location of P transfer in comparison with 

that of the intercellular hyphae (or total interface) as the location of exchange of 

carbohydrate, it will be necessary to do more sophisticated experiments with more 

fungus/host combinations. The use of VA mycorrhizas in which development of 

arbuscules is suppressed will be particulary valuable. 
 

 

 

8 



Nutrient transfer in vesicular-arbuscular mycorrhizas - F.A. Smith and S.E. Smith 

CONCLUSIONS 

In this paper we have reviewed briefly the mechanisms for transfer of nutrients in 

mycorrhizas - mainly VA mycorrhizas - and have then proceeded to re-evaluate the sites 

of nutrient transfer, particularly the question of whether simultaneous bidirectional 

transfer occurs in arbuscules. We suggest that this may not be the case, and this in turn has 

led us to quantify the development of the arbuscular interface vis-a-vis the total interface 

and to attempt to calculate P fluxes across both interfaces. As well as pursuing this 

approach further, we are starting to attempt to identify the molecular mechanisms for 

transport at the interface and also to measure the physico-chemical conditions there, 

including the ionic conditions in the apoplast. In this way we hope that the control of 

nutrient transfer and its role in determining "mycorrhizal efficiency" will be better 

understood. 

ACKNOWLEDGEMENTS 

We thank our collaborators, particularly Vivienne Gianinazzi- Pearson, Silvio 

Gianinazzi, Sandy Dickson, Christina Morris, Mark Tester and Rob Reid for all their 

help in discussions and experiments. Funding came from the Australian Research Council, 

an INRA international collaborative research program and The University of Adelaide. 

REFERENCES 

Cox, G. and P.B. TINKER. 1976. Translocation and transfer of nutrients in vesicular-arbuscular mycorrhizas. 

I. The arbuscule and phosphorus transfer: a quantitative ultrastructural study. New Phytologist, 77, 

371-378. 

GIANINAZZI-PEARSON, V., S.E., SMITH, S. GIANINAZZI and F.A. SMITH. 1991. Enzymatic studies on the metabolism of 

vesicular-arbuscular mycorrhizas V. Is H 
+
 -ATPase a component of ATP hydro-lysing enzyme activities 

in plant/fungus interfaces? New Phytologist, 117, 61-74. 

MARX, C., J., DEXHEIMER, V. GIANINAZZI-PEARSON and S. GIANINAZZI. 1982. Enzymatic studies on the metabolism of 

vesicular-arbuscular mycorrhizas IV. Ultracytoenzymological evidence (AT-Pase) for active transfer 

processes in the host-arbuscular interface. New Phytologist, 90, 37-43. 

SMITH, F.A., S., DICKSON, C., MORRIS, R.J., REID, M., TESTER and S.E., SMITH. 1994a. Phosphate transfer in VA 
mycorrhizas: special mechanisms or not? Proceedings of the 4th International Symposium on the 
Structure and Function of Roots. In press.  

SMITH, S.E. and S. DICKSON. 1991. Quantification of active vesicular-arbuscular mycorrhizal infection using image 
analysis and other techniques.  Australian Journal of Plant Physiology,  18, 637-648. 

 

 

 

 

 

9 


	1.pdf
	10.pdf
	2.pdf
	3.pdf
	4.pdf
	5.pdf
	6.pdf
	7.pdf
	8.pdf
	9.pdf