Upsala J Med Sci 78: 181-188, 1973 Effects of lndomethacin on the Transcapillary Leakage of Macromolecules and the Efflux of Prostaglandins in the Paw Lymph Following Experimental Scalding Injury GOSTA ARTURSON and CARL-EVERT JONSSON From the Burn Center, Department of Plastic Surgery, University Hospital, (lppsala, Sweden ABSTRACT Transport of macromolecules (dextrans and proteins) from blood to lymph and efflux of prostaglandins into lymph were studied in dogs following scalding injury of the paw and treatment with indomethacin. Indomethacin inhibited the efflux of prostaglandins following scalding injury, in- dicating an inhibition of the biosynthesis of prostaglandins. A pronounced suppression of the increased lymph flow and transcapillary transoort of macromolecules following scalding was found after treatment with indomethacin. The increased microvascular permeability in scalded tissue was not significantly altered by indomethacin. These results indicate that the major effect of indomethacin on the microcirculation in the scalded tissue is a reduction of the capjllary surfaee area available for exchange due t o a re- duced number of capillaries perfused with blood. The re- sults d s o support the hypothesis that some of the vascular reactions following thermal injury may be mediated by prostaglandins. INTRODUCTION Prostaglandins, a group of biologically active lipids (for reviews see 9, 20, 22, 41) have been demon- strated in tissue fluids following chemical in- flammation (44, 4 3 , scalding injury (2, 24) and anaphylaxis (16). Recently it was shown that scald- ing injury in guinea pig is followed by an increased biosynthesis of prostaglandins (1 8). Prostaglandin E, (PGE,) is the major prostaglandin found in inflammatory fluids. The vascular effects of this compound fulfil many of the criteria as demanded by a chemical mediator of the inflammatory re- sponse (37). Evidence has accumulated for structural changes of the blood-lymph-barrier, i.e. increased micro- vascular permeability (4, 29), increased capillary surface area due to dilatation of resistance vessels as well as increased tissue osm2tic forces in the burn wound (8). Recently it was shown that drugs like aspirin and indomethacin inhibit the biosynthesis of pros- taglandins (33, 39), suggesting that these drugs owe their anti-inflammatory effects to this mech- anism. In the present communication we report effects of indomethacin (47) on prostaglandin efflux in paw lymph and on microvascular reactions follow- ing scalding injury. MATERIAL AND METHODS Six healthy dogs of the Vorsteh-type, weighing between 17 and 28 kg and of the same breed, were used. All experi- ments were done using sodium pentobarbital anaesthesia (NembutalO, Abbott). F o r induction of anaesthesia a dose of 30 mg per kg of body weight was given. Small additional doses were given when required. T h z dog was placed on an operating table equipped with electric heat- ing to maintain a normal body temperature; this was con- trolled throughout the experiment. Free airway was ascertained by endothracheal intubation. Ringer solution was given as a slow intravenous infusion during the ex- periments in a dose of 5 ml per kg of body weight per hour. Peripheral lymph was drained from the dog's hind paw via a short PE 50 Intramedic polyethylene cannula (Clay- Adams, Inc., New York). Th: operative procedures and methodological details have been published elsewhere (4). Lymph flow was facilitated by regular, passive move- ments of the paw. Lymph was collected in ice-chilled poly- ethylene tubes for assay of dextran, protein and prosta- glandins. The thermal trauma was inflicted by immersing the animal's paw for 10 sec in water at 100°C OJ for 20 sec in water at 70°C. This was done about 180 min affer clamping of the renal pedicles. Four dogs were treated with indomethicin (10 o r 2 0 m g / kg body weight) intravenously immediately after scalding and on- hour later except one dog which only received a single dose of indomethacin (20 mg/kg body weight). lndomethacin (Merck, Scharp and Dohme, Rahway, N.J., USA) was dissolved in 0.15 M p3tassium phosphate buf- fer, p H 7.4, immediately prior to administration. Two dogs served as controls and received n o treatment af:er scalding other than Ringer solution. U p ~ ! u J M e d Sci 78 182 G . Arturson and C.-E. Jonsson Permeability studies The local microvascular permeability in the scalded tissue was studied before and at different times after the trauma. D:xtran was used as test substance. RheomacrodexO,, 10% in 0.9 saline, average molecular weight, M , = 40 000 (range 5 000-90 000) was given slowly in a dose of 0.5 g / kg bady weight. This low dose produces only very small changes in the plasma volum-, (15). Before t h s dextran in- fusion, th: renal pedicles were clamped to prevent th? rapid elimination of dextran molecules of low molecular weight via the kidneys. After an equilibration period of 120 min the molecular weights of dextran molecules in lymph and plasma were determined and the lymph/plasma concentration ratio ( C , / C p ) of different molecular sizes was calculated. The total concentration of dextran in plasma and lymph was determined by the anthrone method (23, 40). This methJd has an error of 1.8% S.D. (40). The total concentration of protein in lymph was deter- mined according to a modified Folin method (3). The molecular weight distribution of dextran was estimated by gel chromatography adapted t o automated routine and computer analysis ( 6 ) . The sensitivity, repro- ducibility and resolving power of the gel chromatography method has been thoroughly investigated (13, 27). The accuracy of the determination of dextran molecular weight distribution is in the range of 3% S.D. (Granath, personal communication). In all experiments the regional transport of dextran across thz blood-lymph barrier was calculated according to: C , j C , . v j t = T where C,, and C , (in mg/ml) is the concentration of dextran i n lymph and plasma respectively, and v the volume (in ml) of the lymph collected during the sampling period t (in min). T will then be equal to the net transport of th: substance from blood to the collected lymph can- nula in mg/min at unit plasma concentration ( = 1 mg/ml). The physical dimension of T is actually ml/min. The Cp value used in this ca!culation corresponded t o the plasma concentration at the middle of the lymph sampling period. It was not possible to correct for the time delay bstween th- actual transport through the capillary wall and the collection from the lymph cannula. Determination of prostaglandins Lymph was precipitated in absolute ethanol containing 0.1 % d,l a-tocopherol as antioxidant and stored at - 20°C before processing. Lipid extracts were prepared as de- scribed by Unger et al. (38) and subjected to silicic acid chromatography (30) prior to bioassay. The silicic acid columns (Unisil, 0.5 g, 100-200 mesh, Clarkson Chemical Corp., Williamsport, Penns.) were made up in ethyl acetate-benzene (1 : 9, v/v) and eluted with 20 ml of ethyl acetate-benzene (1 : 9) followed by 20 ml of ethyl acetate. Thes-, fractions were evaporated under reduced pressure and assayed for smooth muscle stimulating activity on colon of the gerbil (42). The recovery was determintd in each sample by addition of *H-PGE, (0.1 pC, specific activity 87.3 C/mmole, New England Nuclear Corp., Boston, Mass.), to the lymph-ethand mixture and deter- Upsala J M e d Sci 78 mination of radioactivity in aliquots of the ethyl acetate fractions and was about 50% ( x = 49.9, S.D. = 9.1,. n = 74). When 50 ng PGE, was added to 5 ml human plasma and processed as described the recovery of smooth muscle stimulating activity was similar (n = 4 5 % , S.D. = & 1 8 , n = 17). Indomethacin (100 p g / m l plasma) did not affect the recovery of smooth muscle stimulating ac- tivity of PGE, (35, 39). In all assays t h t colon of the gerbil responded to 0.5 ng PGE, added to the organ bath. Levels below 2 ng per sample could not be detected and were thus regarded as zero values. Smooth muscle stimu- lating activity was expressed in terms of PGE, after cor- rection for recovery of radioactivity. Radioactivity was determined in a Packard liquid spectrometer (model 3320). 10 ml of Instagel (Packard Instr.) was used as scintillator. Corrections for counting efficiency were made by use of external standardization. RESULTS Transcapillary leakage of macromolecules No significant differences were found in lymph flow, in protein o r in dextran concentration of the lymph drained from the different paws before in- fliction of the thermal trauma. The changes in sieving ratio ( C L / C p ) in the scalded areas were found to be related to the severity of the burn trauma, being somewhat less pronounced after scalding for 20 sec at 70°C than for 10 sec a t 100°C (Figs. 1 and 2). The micro- vascular permeability and the lymph flow were maximal about 1-4 hours after scalding. After immersion of a dog’s paw for 20 sec in water of 70°C the lymph flow increased about 10 times, the protein concentration of the lymph in- creased from 0.7 to about 3.0 g/100 ml and the relative amount of dextran transported transcapil- lary (T-values) increased 20 times (Fig. 1). The lymph plasma dextran concentration ratio ( C L / C p ) for molecular weights of 10 000-80 000 increased immediately after the trauma and reached max- imal values 1-3 hours post burn (Fig. 1). After scalding of another dog’s paw for 20 sec at 70°C and treatment with indomethacin (10 mg/ kg body weight), repeated twice, immediately and 60 min after the burn, a pronounced decrease of the lymph flow and a slight reduction of the total protein concentration in the lymph compared with non-treated dog was observed (Fig. 1). The relative amount of dextran transported transcapillary from the blood to the extravascular space after scald- ing was reduced to between 25 and 35% of the amount transported transcapillary in the non- treated dog (Fig. 1). The C , / C p ratios for dif- ferent molecular weights of dextran were about Effects of indomethacin on the transcapillary leakage o f macromolecules 183 [Controd [Treatment with indomcthacinl min ( e a t r a n ) 1.2- 1.0- 0.8- 0.4 0.2- 0- I h o u r p o s t burn M o I weight x l o 3 I 300 min O 2 1 O J & ' ' " T " r m ~ 1 0 200 300 min P r P h U r n 0 20 60 60 80 M o I w e i g h t x 10.' min O 5 1 1 ° j bl h o v r p o s t burn 0 5 MoI weigh1 x l o ' [friatment with indomethacin 1 mllrnin O b1 Paw burn U ( n . 2 0 ~ ) k m p ii mplkp b.w. 3 hours posl bvrn 0 5 "1 \\I hour posl burn o j , L , P r e b u r n 0 20 40 60 80 MoI w e i g h t I 16.' Fig. 1. Lymph flow, total protein concentration of lymph, T-values, efflux of prostaglandins and C , / C , ratios for dextran in lymph from paws scalded for 20 sec at 70°C in two dogs, one without treatment and one treated with indomethacin 10 mg/kg body weight immediately and one hour after scalding. _ _ _ _ _ , lymph flow 0- - - -0, protein conc. Fig. 2. Lymph flow, total protein concentration of lymph, T-values, efflux of prostaglandins and C , / C , ratios for dextran in lymph from paws scalded for 10 sec at 100°C in two dogs, one without treatment and ons treated with indomethacin 20 mg/kg body weight immediately and one hour after scalding. _ _ - - - , lymph flow; 0- - - -0, protein conc. Upsala J M e d Sci 78 184 G . Arturson and C.-E. Jonsson \Treatment with indomethacin I 0 100 200 300 10mg 10mg/kg b.w. 0 0 min cone. j l j f , , , . , , , , . , , , , , , , , , , , . . , , , . . . . I b h mg/min at unit Cp 0.2 (dextran) 01 u 0 100 2 00 300 mi n n g PGE2 O 2 ' k h j 200 300 0 100 mi n n g PGE2 O 2 ' k h j 200 300 0 100 rnin 3 hours post burn 1 hour post burn 0 0 20 40 60 80 Mol weight x Fig. 3. Lymph flow, total protein concentration of lymph, T-values, efflux of prostaglandins and C,/C, ratios for dextran in lymph from a paw scalded for 10 sec a t 100°C in one dog, treated with indomethacin 10 mg/kg body weight immediately and one hour after scalding. - - _ _ - , lymph flow; 0----0, protein conc. equally increased in both the scalded non-treated and treated paws (Fig. 1). Figs. 2 and 3 show results from a comparative study in three dogs of the microcirculation in scalded tissue and with treatment with indo- methacin a t two dose levels. The scaldings were made in exactly the same way in all three dogs, i.e. immersion of the paws for 10 sec in water of 100°C. The lymph flow, the total protein con- centration of lymph and the 7'-values increased immediately after scalding and reached the same level in the non-treated dog and the dog treated with indomethacin in the low dose (10 mg/kg body weight). Following scalding and treatment with indomethacin in a dose of 20 mg/kg body weight the increase in lymph flow and the T- values were reduced to 30-40% of those found in the scalded non-treated dog (Figs. 2 and 3). The sieving ratios for dextran molecules of Upsala J M e d Sci 78 various sizes in lymph/plasma (C,/C,) increased following the thermal trauma in a similar,way in all three dogs irrespective of treatment or not (Figs. 2 and 3). A study of the microcirculation in scalded tissue with and without treatment with indomethacin made on one and the same dog is shown in Fig. 4. A burn trauma on the right hind paw resulted in increases in lymph flow, transcapillary transport of dextran (T-values) and in the C L / C p ratio for all dextran molecules in the preparation. Two hours after the first trauma a second one was made in exactly the same way on the left hind paw and the dog was treated with indomethacin (20 mg/kg). This resulted in less pronounced in- creases in lymph flow and T-values compared t o the scalded non-treated paw (Fig. 4). No signifi- cant differences were found in the C,/C, ratios obtained from the two paws (Fig. 4). Efflux of prostaglandins in lymph After scalding of the dog's paw for 20 sec a t 70°C and for 10 sec at 100°C and without treat- ment there was a rapid increase in the concen- tration of prostaglandins in the lymph drained from scalded tissue with a decline after about 150 min. This was followed by a second increase of the efflux of prostaglandins in the lymph 3-5 hours after the trauma (Figs. 1 and 2). After treatment with indomethacin in a dose of 10 o r 20 mg/kg body weight, administered immediately after scalding and one hour later, prostaglandins were not recovered in the paw lymph (Figs. 1 , 2 and 3). In the study with and without treatment with indomethacin made on one and the same dog total inhibition of the efflux of prostaglandins was found following treatment (Fig. 4). Also on the previously scalded non-treated paw the amount of prostaglandins in the lymph decreased to zero values concomitant with the intravenous adminis- tration of indomethacin two hours after the first thermal trauma. DISCUSSION I n the acute phase of a burn injury there is a rapid loss of intravascular fluid into the burned area. The fluid loss has been ascribed to func- tional disturbances in the blood lymph barrier, increased effective filtration area and increased tissue osmotic forces (8). By selecting appropriate Effects of indomethacin on the transcapillary leakage of macromolecules 185 (dextran) 0 50 100 150 min m i n [ Treat m e n t with indomet hact n 1 mllmin Paw burn 0.2\;] 0 0 0'23 0 0 1.01 rnin 120 rnglkg b.w. 0 50 100 150 rnin 0 50 100 1% min Fig. 4. Lymph flow, total protein concen- tration of lymph, T-values, efflux of prosta- glandins and C L / C , ratios for dextran in lymph following scalding for 20 sec at 70°C of two paws of a dog, one paw without 2 hours post burn ted pre burn and one and two hours after 0 20 4 0 60 80 0 20 4 0 60 80 the trauma, respectively. - - - - -, lymph flow: 0- - - -0, protein conc. I M o l . weight x l O J Mol. weight x l o 3 treatment and one paw treated with 20 mg indomethacin per kg body weight immedi- ately after scalding. T h e C , / C , ratios for dextran molecules of various sizes are plot- 2 hours post burn thermal stimuli to experimental animals two phases of vascular reactions have been discerned, an immediate and a delayed phase (32). Vaso- active-compounds have been suggested t o mediate some of these vascular reactions following injury (37, 46). Attempts have been made to modify the vas- cular reactions following thermal injury by drugs. Spector & Willoughby (36) demonstrated that by treating rats with mepyramine maleate, a power- ful antihistamine drug, prior t o moderate thermal injury, the immediate vascular reactions, expressed as leakage of trypan blue and oedema formation, were suppressed. Administration of salicylate sup- pressed both phases. These observations indicate that histamine is involved only in the immediate response to thermal injury suggesting that his- tamine may initiate vascular reactions which are maintained by other mechanisms. The occurrence of PGE-compounds in inflam- matory fluids (2, 16, 24, 44, 45) and the vascular effects of these compounds suggest that they participate in the inflammatory response. Paren- terally administered PGE-compounds induce hypo- tension due to vasodilatation and reduced periph- eral resistance (11, 14, 28). Intradermally ad- ministered PGE-compounds elicit erythema and oedema (10, 25, 34) due to vasodilatation. In rats and guinea pigs, preinjected with protein bound dyes, PGE-compounds induce extravasation of color on intradermal administration (10, 21, 2 6 ) . This effect has been claimed to be due to in- creased capillary permeability. In canine adipose tissue PGE, is a powerful vasodilatating agent with comparatively small effects on capillary per- meability (1 1). Recently it was demonstrated that anti-inflam- matory drugs like aspirin, sodium salicylate and indomethacin inhibit the synthesis of prostaglan- dins (12, 17, 33, 39). This was found both in cell- free homogenates and in the whole animal. In the present investigation the effect of indo- methacin on the efflux of prostaglandins in lymph drained from scalded tissue and the effect on the transcapillary leakage of macromolecules into the burn oedema was studied simultaneously. The efflux of prostaglandins in the lymph was assumed to reflect the biosynthesis of prostaglandins in the tissue. The sieving ratio for dextran molecules of various sizes in lymph/plasma (C,/C,) was meas- Upsala J M e d Sci 78 186 G . Arturson and C.-E. Jonsson ured and used as a n index of the microvascular permeability. Recently Arturson ( 5 ) demonstrated a moderate suppression of the increased micro- vascular permeability after scalding by 0-(P-hy- droxyethy1)-rutosides (HR) a compound with anti- oedematous effects (48). In the present investigation the scalding injury resulted in efflux of prostaglandins (cf. 2 , 24) con- comitant with a n increased lymph flow and an increased concentration of both protein and dex- tran in the lymph drained from scalded tissue (cf. 4). The sieving ratio ( C L / C p ) and the regional transport of dextran across the blood-lymph bar- rier also increased. This indicates an increase of the microvascular permeability after scalding (cf. 4, 7 ) . In all dogs except one treatment with indo- methacin resulted in a suppression of the increased lymph flow, a slight reduction of the total con- centration of protein in lymph and a pronounced reduction of dextran transported from the blood to the extravascular space as compared to scalded non-trcated dogs. The sieving ratios for dextran molecules of various sizes in lymph/plasma were unaffected by the treatment with indomethacin. No smooth muscle stimulating activity was found in lymph fluid after scalding and subsequent treat- ment with indomethacin in any dogs. The inhibition of prostaglandin efflux indicate an inhibition of prostaglandin biosynthesis by indomethacin. I t seems improbable that in the tissue with less rapid lymph drainage there was an increased metabolism of prostaglandin E-com- pounds into metabolites without biological activity ( 1 ) . This is supported by the finding that no smooth muscle stimulating activity was recovered after scalding and subsequent treatment with indo- methacin in a low dose (10 mg/kg twice) which did not affect the lymph flow (cf. Fig. 3). The mechanism underlying the effects of indo- methacin on the microcirculation in scalded tissue is not clear. I t could be either a decreased capil- lary surface area available for exchange due to a reduced number of capillaries perfused with blood, or t o a suppression of the increased micro- vascular permeability caused by the thermal trauma, or both. The very small changes of the sieving ratios for dextran molecules of various sizes in lymph/plasma ( C L / C p ) and the pro- nounced suppression of the lymph flow as well as the amount of dextran transported transcapillary Upsala J M e d Scr 78 following treatment with indomethacin indicate that a reduction of the effective capillary surface area is the most important mechanism for the ef- fect of indomethacin. Lately a modulatory role of prostaglandins on the sympathetic nervous transmission has been implied (19). Evidence exists to suggest that pros- taglandins inhibit the release of norepinephrine (43). Blocking of the local formation of prosta- glandins leads to a considerably increased release of norepinephrine on sympathetic stimulation (31). In the same way it may be suggested that after scalding injury and treatment with indomethacin prostaglandins are not formed in sufficient amounts to oppose the increased secretion of catecholamines. However, removal of the vaso- dilatory effect of prostaglandin E-compounds may also be one explanation. As it has not been demonstrated that indomethacin exclusively in- hibits prostaglandin biosynthesis, other effects of the drug itself cannot be excluded. No suppressive effect on the lymph flow or the transcapillary macromolecular transport was ob- served in the dog with a paw immersed for 10 sec in water of 100°C and treated with indomethacin in the low dose (Fig. 3) although no PGE, was found in the lymph. This might indicate either that the efflux of prostaglandins in lymph is not a satisfactory estimate of the prostaglandin bio- synthesis in tissues or that the changes of the microvasculature are not mediated via PGE,. The fact that clear suppressive effects on both lymph flow and transcapillary transport of dextrans were found following scalding for 10 sec at 100°C and treatment with indomethacin in high dose (Fig. 2) favours the hypothesis that som: of the vascular reactions after scalding injury are mediated via the prostaglandin system. ACKNOWLEDGEMENT This work was supported by Swedish Medical Research Council (Projects No. B73-40Y-2370-06 and B73-40X-676- 08) and Pharmacia AB, Uppsala, Sweden. REFERENCES AnggBrd, E.: The biological activities of three meta- bolites of prostaglandin El. Acta Physiol Scand 66: 509, 1966. Anggird, E. & Jonsson, C.-E.: Efflux of prostaglan- dins in lymph from scalded tissue. Acta Physiol Scand 81: 440, 1971. Aronsson, T., Arturson, G . & Wallenius, G . : Dcter- E f f e c t s of indomethacin on the transcapillary leakage o f macromolecules 187 mination of serum protein in the presence of dextran. Scand J Clin Lab Invest 18:458, 1966. 4. Arturson, G.: Pathophysiological aspects of the burn syndrome. Acta Chir Scand, Suppl. 274, 1961. 5 . - Effects of 0-(P-hydroxyethy1)-rutosides (HR) on the increased microvascular permeability in skin burns. Acta Chir Scand 138: 1 1 1 , 1972. 6. Arturson, G. & Granath, K.: Dextrans as test mole- cules in studies of the functional ultrastructure of biological membranes. Clin Chim Acta 37: 309, 1972. 7. Arturson, G., Groth, T. & Grotte, G . : The functional ultrastructure of the blood-lymph barrier. Computor analysis of data from dog heart-lymph experiments using theoretical models. Acta Physiol Scand, Suppl. 345, 1972. 8. Arturson, G. & Mellander, S.: Acute changes in capil- lary filtration and diffusion in experimental burn in- jury. 62:457, 1964. 9. Bergstrom, S., Carlson, L. A. & Weeks, J. R.: The prostaglandins: a family of biologically active lipids. Pharmacol Rev 20: 1, 1968. 10. Crunkhorn, P. & Willis, A. L.: Cutaneous reactions to intradermal prostaglandins. Brit J Pharmacol 41: 49, 1971. 1 1 . Fredholm, B. B., Oberg, B. & Rosell, S . : Effects of vasoactive drugs o n circulation in canine subcutaneous adipose tissue. Acta Physiol Scand 79: 564, 1970. 12. Ferreira, S . H., Moncada, S. & Vane, J. R.: Indo- methacin and aspirin abolish prostaglandin release from the spleen. Nature New Biol 231:237, 1971. 1 3 . Granath, K. A. & Kvist, B. E.: Molecular weight distribution analysis by gel chromatography on Sepha- dex. J Chromatogr 28169, 1967. 14. Greenberg, R. A. & Sparks, H. V.: Prostaglandins and consecutive vascular segments of the canine hindlimb. Am J Physiol 216:567, 1969. 15. Grotte, G.: Passage of dextran molecules across the blood-lymph barrier. Acta Chir Scand, Suppl. 211, 1956. 16. Greaves, M. W., Sondergird, J. & McDonald-Gibson, W.: Recovery of prostaglandins in human cutaneous inflammation. Brit Med J 2:258, 1971. 17. Hamberg, M . & Samuelsson, B.: On the metabolism of prostaglandins E, and E, in the guinea pig. J Biol Chem 247: 345, 1972. 18. Hamberg, M. & Jonsson, C.-E.: Increased synthesis of prostaglandins in the guinea pig following scalding injury. Acta Physiol Scand 87: 240, 1973. 19. Hedqvist, P.: Studies o n the effect of prostaglandins EL and Ez o n the sympathetic neuromuscular trans- mission in some animal tissues. Acta Physiol Scand, Suppl. 345, 1970. 20. Hinman, J. W.: Prostaglandins. Ann Rev Biochem 41: 161, 1972. 21. Horton, E. W.: Action of prostaglandin El o n tissues which respond to bradykinin. Nature 200: 892, 1963. 22. - Prostaglandins. Springer-Verlag, Berlin-Heidelberg- New York 1972. 23. Jenner, H.: Automated determination of th? molecular weight distribution of dextran. Automation in analyti- cal chemistry. Technicon Symposia 2: 203, 1967. 24. Jonsson, C.-E.: Smooth muscle stimulating lipids in peripheral lymph after experimental burn injury. Scand J Plast Reconstr Surg 5: 1, 1971. 25. Juhlin, L. & Michaelsson, G.: Cutaneous vascular re- actions to prostaglandins in healthy subjects and in patients with urticaria and atopic dermatitis. Acta derm-venerol (Stockh.) 49: 251, 1969. 26. Kaley, G. & Weinsr, R.: Prostaglandin E,: a potential mediator of the inflammatory response. Ann N Y Sci 180: 3 3 8 , 1971. 27. Laurent, T. C. & Granath, K. A,: Fractionation of dextran and ficoll by chromatography on Szphadex (3-200. Biochim Biophys Acta 136: 191, 1967. 28. Nakano, J. & McCurdy, J. R.: Cardiovascular effects of prostaglandin 6. J Pharmacol Exp Ther 156:538, 1967. 29. Netsky, M. G. & Leiter, S . S . : Capillary permeabiilty to h x s e proteins in burn-shock. Am J Physiol 140: 1, 1943. 30. Samuelsson, B.: Isolation and identification of prosta- glandins from human seminal plasma. J Biol Chem 238: 3229, 1963. 3 1 . Samuelsson, B. & Wennmalm, A,: Increased nerve stimulation induced release of noradrenaline from the rabbit heart after inhibition of prostaglandin synrhesis. Acta Physiol Scand 83: 163, 1971. 32. Sevitt, S.: Early and delayed oedema and increase in capillary permeability after burns of the skin. J Pathol Bacteriol 75: 27, 1958. 3 3 . Smith, J. B. & Willis, A. L.: Aspirin selectively in- hibits prostaglandin production in human platelets. Naturs New Biology 231:235, 1971. 34. Solomon, L. M., Juhlin, L. & Kirschcnbaum, M. B.: Prostaglandins o n cutaneous vasculature. J Invest Derm 51: 280, 1968. 3 5 . Sorrentino, L., Capasso, F. & Di Rosa, M.: Indometha- cin and prostaglandins. Europ J Pharmacol 17: 306, 1972. 36. Spector, W . G. & Willoughby, D. A.: Experimental suppression of the acute inflammatory changes of t h m n a l injury. J Pathol Bacteriol 78: 121, 1959. 37. - The pharmacology of inflammation. The English Universities Press Ltd., London 1968. 38. Unger, W. G., Stamford, I . F . & Bennett, A.: Extrac- tion of prcstaglandins. Nature 233: 336, 1971. 39. Vane, J. R.: Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol 231:232, 1971. 40. Wallenius, G.: Renal clearance of dextran as a meas- ure of glomerular permeability. Acta SOC Med Up- saliens 59: Suppl. 4, 1954. 41. Weeks, J . R.: Prostaglandins. Ann Rev Pharmacol 12: 317, 1972. 42. Weeks, J. R., Schultz, J. R. & Brown, W. E.: Evalua- tion of smooth muscle bioassays for prostaglandins E, and FIc. J Appl Physiol 2 5 ~ 7 8 3 , 1968. 43. Wennmalm, A. & Stjarne, L.: Inhibition of the release of adrenergic transmittor by a fatty acid in the per- fusate from sympathetically stimulated rabbit heart. Life Sci 10r471, 1971. 44. Willis, A. L.: Parallel assay of prostaglandin-like activity i n rat inflammatory exudate by means of cascade superfusion. J Pharm Pharmac 21: 126, 1969. Upsala J M e d Sci 78 188 G . Arturson and C.-E. Jonsson 45. - Identification of prostaglandin E, in rat inflam- matory exudate. Pharmacol Res Com 2:297, 1970. 46. Winkelmann, R. K.: Molecular inflammation of the skin. J Investig Dermatol 57: 197, 1971. 47. Winter, C. A., Risky, E. A. & Nuss, G . W.: Anti- inflammatory and antipyretic activities of indometha- cin, l-(gchlorobemoyl)-5-methoxy-2-methylindole-3- acetic acid. J Pharmac Exp ThIherap 141: 369, 1963. 48. Wismer, R.: The action of tri-hydroxyethyl-rutoside on the permsability of the capillaries in man. Praxis 52: 1412, 1963. Received December 19, 1972 Address for reprints: Carl-Evert Jonsson Departmmt of Plastic Surgery University Hospital S-750 14 Uppsala 14 Sweden Upsala J M e d Sci 7 8