NEFROLOGIA. Vol. XIV. Suplemento 2, 1994 Membrane selection and muscle protein catabolism J. Bergström, A. Alvestrand and A. Gutiérrez, Department of Renal Medicine, Karolinska Institute, Huddinge University Hospital, Stockholm, Sweden. S e v e r a l reports have documented that protein malnutrition is frequently present in patients undergoing maintenance hemodialysis (HD) therapy 1-4. It is generally accepted that suboptimal nutritional status is associated with increased morbidity and may contribute to poor rehabilitation and quality of life 4-6, Low intake of protein and energy, because of anorexia, is a major contributing factor to malnutrition in many HD patients. In addition, the requirements of protein appear to be higher in HD patients than in healthy individuals and nondialyzed patients with chronic uremia. In normal adults the average minimum requirements for protein are about 0.6 g/kg body wt/day which, after correction for 25 % variability to include 97.5 % of the population of young adults, raises the safe level of intake to 0.75 g/kg body wt 7. Results of nitrogen balance studies in patients on maintenance dialysis twice a week suggested that approximately 0.75 g/kg/day of high biological value protein is necessary to maintain nitrogen equilibrium 8 or a slightly positive nitrogen balance 9. According to more recent long-term studies, this amount of protein may not be adequate. Signs of malnutrition have been observed in a substantial proportion of apparently well-rehabilitated patients on maintenance HD, who had a daily protein intake of about 1 g/kg/day. On the basis of clinical results with protein and energy supplements to the diet, it was suggested that 1.2 g of protein, primarily of high biological value and an energy intake of 35 kcal/kg body wt/day should be prescribed for HD patients 1. There is evidence that the HD procedure per se is a strong catabolic stimulus 10. A contributing factor is the loss of free and conjugated amino acids into the dialysis fluid, amounting to about 9 to 13 g per dialysis 11, 12. However, the loss of amino acids is not sufficient to account fully for the increased protein requirements in HD patients compared to non-uremic i n d i v i d u a l s and non-dialyzed uremic patients. Consequently, the possibility must be considered that additional factors not related to the dialytic removal of amino acids are involved. Interaction between blood and artificial membranes in a dialyzer induces biological effects, such as stimulation of the complement system with release of anaphylotoxins (C3a, C5a), aggregation of granulocytes, transient leukopenia and release of granulocytic enzymes, clotting and thrombocyte activation 13. Numerous studies document that membranes of different materials differ markedly in their complement activating actively, which is generally accepted as an index of biocompatibility 14. Monocyte activation with increased release of monokines (IL-1, TNF) may result from direct membrane contact 15, activated complements (C5a) 16,, endotoxin fragments passing through the membrane 17,18, or acetate in the dialysis fluid 19. IL-I and TNF may act in concert and induce inter alia lysosomal catabolism of muscle protein 20, an effect which is mediated by the release of prostaglandin E2 21, 22. This mechanism has been proposed for the increase in muscle protein breakdown observed in sepsis and trauma 23. More recently it has been observed that IL-l. TNF and endotoxin may induce net catabolism of muscle protein by stimulating branched-chain keto acid dehydrogenase, which leads to on enhanced oxidation of branched-chain amino acids 24. Correspondencia: Jonas Bergström, M.D. Professor Department of Renal Medicine. Karolinska Institute. Huddinge University Hospital. Stockholm. S-14186 Huddinge. Sweden. Efects of blood membrane contact in vivo on muscle protein catabolism To investigate whether the blood-membrane interaction might induce increased muscle protein catabolism, we studied the exchange of free amino acids across the leg in fasting healthy subjects before and 89 A. GUTIERREZ y cols Table I . Sham dialysis studies with different membranes and dialyzers Croup Number of studies Membrane Dialyzer Surface area m2 cu CU-IND CA AN PS 10 6 6 8 8 8 Cuprophane @ Cuprophane @ Cellulose acetate Acrylonitrile-Na methalylsulphonate (AN -69 @) Polysulphone Gambro Lundia Plate 11 p Gambro Fiber GF 120 H Cambro Lundia Plate l l 1 C-DAKTM3500 FiltraI 12 Hemoflow F60 1 .o 1.3 1 .0 0.9 1.15 1.2 after a sham HD procedure, that is, in vivo passage of blood, but with no circulating dialysate, through a dialyzer. The release of amino acids was investigated using dialysis membranes recognized as strong (cuprophane), moderate (celulose acetate), and weak (AN 69, polysulphone) complement activations. The results of these studies, which have recently been published 25, 26, are summarized and discussed in the present communication. Materials and metbods Five groups of healthy subjects (mean age 21, range 20 to 56 years) were studied. They adhered to their ordinary diet until the evening before the study, when they began an overnight fast. The number of subjects, dialysis membranes and dialyrers used are presented in table I . Cuprophane membranes were used in two groups; in one of them the subjects took indomethacin 100 mg before and 50 mg at the end of the sham dialysis procedure 1.4 1.2 1 .o 0.8 0.6 0.2 0.0 90 3 Polyethylene catheters were inserted into a femoral artery and an ipsilateral femoral vein, for sampling of blood for amino acid determination. Access for blood circulation through the dialyzer was obtained with a single lumen catheter and in the contralateral femoral vein with a dialysis needle in a cubital vein. After a careful rinsing procedure using saline 25 blood was circulated through the dialyzer at a rate of 100 ml/min for 150 minutes, with the dialysate ports clamped. To prevent cooling of blood, the venous blood line was heated to 37 ºC. Before start of sham dialysis a single bolus of heparin (100 IU/kg body wt) was administered intravenously. Blood samples from the catheters in the femoral artery and vein were taken in quadriplicate for determination of plasma amino acids, before (basal values), at 150 minutes, and at 345 minutes after start of sham dialysis. Leg blood flow was measured by venous occlusion pletysmography immediately after blood sampling at each time point. Plasma flow (Qp) was derived from the blood flow (Qb) and hematocrit (Hct), using the formula Qp = Qb X (100 Hct)/l00. Leg amino acid balance was calculated by multiplb T l F i g 1 Leg tissue p l a s m a flow before, a f t e r 150 min (at the end of sham-HD), and after 3 4 5 min in five groups of healthy subjects undergoing sham-HD with cuprophane (CU, n ), cuprophane with oral supply of indomethacin (CU-IN, ffl) cellulose acetate (CA,=), AN 69 (AN Bj and polysulphone ( P S , 0 ). Significant increases in leg plasma flow were observe in groups CU, CU-IND a n d C A , t h a t is, the groups in w h i c h cellulosic membranes were used. Data are mean v a l u e s f sem. a = P < 0,05; b = P < 150 Min ' M E M B R A N E S AND PROTEIN CATABOLISM A L AA release P < 0,Ol Sham HD Basa¡ 150 Min 345 Min B 1 Phe release peo,05 - Basal 150 Min 345 Min Fig. 2.-Release of the sum of all measured amino acids (Z AA) ana p h e n y l a l a n i n e (PHE) before, after 7.50 min (at the end of shamH D ) and after 345 min in groups CU (filled bars) and CU-/ND (striped bars). There was a significant increase in (A) C AA release a n d (B) PHE release at 345 min in group CU, but no increase in group CU-IND, demonstrating that blood-membrane contact with c u p r o p h a n e elicits net protein catabolism in leg tissue (mainly muscle) and that this effect can be abolished by cyclooxygenase inhibition. Data are mean values f SEM. cal HD treatment, and to avoid the loss of amino acids from the blood by dialysis. Thus, it can be assumed that any significant effect on leg blood flow and amino acid exchange was the result of blood-membrane interaction and not due to loss of amino acids or to uptake of substances from the dialysis fluid. Leg plasma flow. Figure 1 demonstrates the effect of sham HD on the leg plasma flow in the five groups. Following sham HD the plasma flow increased in groups CU, CU-IND and CA but no significant change in plasma flow was recorded in groups AN and PS, that is, with the membranes known to be more biocompatible than cellulosic membranes as reflected in their potency to activate complement. Conceivably, the increase in leg blood flow using ceIlulosic membranes might be caused by release of inflammatory mediators 28, 29. The mobilization of amino . acids from muscle may also result in increased muscle blood flow 30. Amino acid release after sham HD with cuprophane without and with indomethacin. The net release of the sum of measured amino acids and of phenylalanine from leg tissue is shown in figure 2. In group CU there was a significantly increased net release from the leg of the sum of the measured amino acids, and also of some individual amino acids (alanine, histidine, isoleucine, leucine, lysine, phenylalanine, threonine and tyrosine) 25,26. The peak effect of blood-membrane contact on leg amino acid balance in group CU was observed 345 minutes after the start of the 150 minute sham-HD 25, an observation of interest when discussing the potential role of IL-I as'a mediator of the protein breakdown associated with HD, since in human studies increased release of IL-I has been observed three to four hours after activation ying the arterio-femoral venous concentration differente for individual amino acids, multiplied by the leg plasma flow. Plasma concentrations of individual amino acids were measured by ion exchange chromatography using an automatic amino acid analyzer (LKB 4460). Asparagine and glutamate were not well separated by the chromatography and are not reported. Plasma concentration of 3-methylhistidine was measured by HPLC-technique based on precolumn derivatization of orthophthaldehyde 27. 400 1 Sham H D P < 0,Ol 3 s .c, b.0 0 2 .E 300 - 200 - . 100' 0 E S Basal 150 Min 345 Min Results and discussion The sham HD procedure was designed to establish a continuous in vivo contact between blood and dialysis membranes, resembling that occurring during a clini- Fig. 3.-Net release of the sum of all measured amino acids (c AA) before, after 150 min (at the end of sham-HD) and after 345 min in the four groups of subjects using different membranes. For explanation of symbols see table I and fig. 1. The efflux of amino acids increased significantly only in group CU. 91 A. GUTIERREZ y cols A Art concentration P < 0.01 0.8 6 V-A diif. P < 0.02 1 0.6 2 5 2 0.4 I 0.2 0.0 0.8 0.6 ,' 0 1 C Leg eiilux P < 0.01 1 0.4 0.0 345 Min Fig. 4.-Venous-arterial concentration difference [V- A diff) and leg e f f l u x of 3-methylhistidine in subjects undersgoing sham-HD with cuprophane membranes (group CU). T h e increase in 3methylhistidine arterial level. V-A diff and leg efflux indicates that t h e sham-HD procedure stimulated proteolysis in muscle Data are mean values + SEM. of monocytes 31 In five subjects in group CU who were followed for more than 345 minutes after start of sham HD, the release of amino acids tended to decrease, but was still higher at 540 minutes than before sham HD 25. In the subjects who received indomethacin before and at the end of sham HD (group C U - I N D ) the release of amino a c i d s w a s n o t increased 25. Based on measurement of the balance across the leg of amino acids that are not metabolired in muscle (tyrosine and phenylalanine), and the knowledge of the amino acid composition of skeletal muscle , protein 32, a rough quantitative estimation of the net proteolytic effect of blood-membrane interaction can be made. Applying this method to the present experimental situation of sham HD the increased release of amino acids in group CU can be calculated to correspond to a loss of 13 to 16 g of muscle protein. Since it has been demonstrated that monokines induce increased muscle proteolysis through activation of PGE2 21, our observation that sham HD did not increase amino acid release in group CU-IND to whom indomethacin was administered, supports the suggestion that the catabolic effect of blood membrane interaction may be mediated through stimulation of prostaglandins production by some monokine(s). An effect on leg release of amino acids by indomethacin through any other mechanism than prostaglandin inh i b i t i o n is not likely, since results from in vitro studies indicate that indomethacin does not influence the release of amino acids from skeletal muscle in the absence of a catabolic stimulus 23. Leg amino acid release wifh different membranes m e a s u r e d . The effect of sham HD on leg release of the sum of measured amino acids, using different membranes is demonstrated in figure 3. In groups CA and PS the release of the sum of amino acids tended to be higher after 345 minutes compared with the basal release, but only in group CU did the increase r each statistical significance. In group AN there was no change in amino acid release after sham HD. The results strongly suggest that not only the leg blood flow, but also the efflux of amino acids from leg tissue is dependent on the properties of the dialyrer membrane. Although the changes in amino acid metabolism during a single sham HD in healthy subjects do not necessarily reflect those occurring in uremic patients undergoing maintenance HD, it is not inconceivable that the stimulation of protein catabolism could be even higher during clinical dialysis than in our experiments, since most HD treatments are performed using higher blood flow, longer duration, and sometimes larger surface area than in the present experimental study (blood flow 100 ml/min/m2 during 150 min). 92 MEMBRANES AND PROTEIN CATABOLISM . L e g muscle release of 3-methylhistidine. Amino acids that are Iiberated during the course of muscle protein breakdown are extensively reutilized for protein synthesis. The increased net efflux of amino acids from the leg may therefore result from increased protein degradation, decreased protein synthesis or a combination of these alterations. Some inferences on the contribution of protein breakdown may be d r a w n from the meaurement of leg release of 3m e t h y l h i s t i d i n e . This amino acid is formed posttranslationally by the irreversible methylation of histidine in a c t i n o m y o s i n e p r o t e i n a n d c a n n o t b e reutilized after being Iiberated during the process of protein degration 33. We observed that the arterial 3m e t h y l h i s t i d i n e concentration, the V-A difference and the efflux of 3-methylhistidine from leg tissue increased significantly in group CU (fig. 4), but were u n c h a n g e d in group CA and PS. The increase in leg e f f l u x and elevated arterial concentration of 3methylhistidine following sham HD using cuprophane dialyzers implies that increased muscle proteolysis plays an important part in the net catabolic process induced by blood-membrane contact. References 1. Kluthe R, Luttgen FM, Capetianu T, Heinze V, Katz N, Südhoff A : Protein requirements in maintenance hemodialysis. Am J Clin Nutr 31:1812-1820, 1978. 2 . Y o u n g GA, Swanepoel CR, Croft MR, Hobson SM, Parsons FM: Anthropometry and plasma valine, amino acids, and proteins in the nutritional assessment of hemodialysis patients. Kidney Int 21:492-499, 1982 3 . Guarnieri G, Toigo G, Situlin R, Faccini L, Coli U, Lanini S, Bazzato G, Dardi F, Campanacci L: Muscle biopsy studies in c h r o n i c a l l y uremic patients: Evidente for malnutrition Kidney Int 24 (Suppl 16):187-193, 1983. 4. M a r c k m a n n P: Nutritional status and mortality of patients in regular dialysis therapy. J Intern Med 226:429-432, 1989. 5. Lowrie EG, Lew NL: Death risk in hemodialysis patients: T h e predictive value of commonly measured variables and an eval u a t i o n of death rate differences between facilities. Am J Kidney Dis 15:458-482, 1990. 6. Acchiardo SR, Moore LW, Latour PA: Malnutrition as the main f a c t o r in morbidity and mortality of hemodialysís patients. Kidney Int 74 (Suppl 16):199-203, 1983. 7. F A O / W H O : Energy and protein requirements. Report of a joint FAO/WHO ad hoc Expert Commíttee, in Tech Rep Ser, No 522, Geneva, World Health Organization, 1973. 8 . Ginn HE, Frost A, Lacy WW: Nitrogen balance in hemodialysis patients. Am JClin Nutr 21 :385-393, 1968. 9 . K o p p l e JD, Shínaberger JH, Coburn JW, Sorensen MK, R u b i n i ME: Optimal dietary protein treatment during chron i c hemodialysis. Trans Am Soc Artif Organs 15:302-308, 1969. 10. B o r a h M F , S c h o e n f e l d P Y , G o t c h F A , S a r g e n t JA, W olfson M, Humphreys MH: Nitrogen balance during int e r m i t t e n t dialysis therapy of uremia. Kidney Int 14:491500,1978. 11. Kopple JD, Swendseid ME, Shinaberger JH, Umezawa CY: The free and bound amino acids removed by hemodialysis. Trans , Am Soc Artif Intern Organs 19:309-313, 1973. 12. Wolfson M, Jones MR, Kopple JD: Amino acid losses during h e m o d i a l y s i s with infusion of amino acids and glucose. Kidney Int 21:500-506,1981. C r a d d o c k PR, Hammerschmidt DE: Complement-mediated g r a n u l o c y t e activation and down-regulation during hemodialysis. ASAIO J 7:50-56, 1984 . C h e n o w e t h DE, Cheung AK, Henderson LW: Anaphylatoxin f o r m a t i o n during hemodialysis: Effects of different dialyzer membranes. Kidney Int 24:764-769, 1983. B e t z M, Haensch GM, Rauterberg EW, Bommer J, Ritz E: C u p r a m m o n i u m membranes stimulates interleukin -I release and arachidonic acid metabolísm in monocytes in the absence of complement. Kidney Int 34:67-73, 1988. H a e f f n e r - C a v a i l l o n N, Cavaillon MJ, Laude M, Kazatchekine MD: C3a/C3ades Arg induces production and release of interl e u k i n - l (IL-l) by cultured human m o n o c y t e s . J Immunol 139:794-799, 1987. L o n n e m a n n G, Bíngel M, Floege J, Koch KM, Shaldon S, D i n a r e l l o CA: Detection of endotoxin-Iike interleukin-l-inducíng actívíty duríng in vítro díalysís. Kidney Int 33:29-35, 1988 Evans RC, Holmes CJ: In vitro study of the transfer of cytokineí n d u c i n g substances across selected high-flux hemodialysis membranes. Blood Purif 9:92-101, 1991. Bingel M, Lonnemann G, Koch KM, Dinarello CA, Shaldon S: E n h a n c e m e n t of in-vitro human interleukin-l production by sodium acetate. Lancet 1:14-16, 1987. Flores EA, Bistrian BR, Pomposelli JJ, Dinarello CA, Blackburn GL, Istfan NW: Infusion of tumor necrosis factor. Cachectin promotes muscle catabolism in the rat. J Clin Invest 83:16141622,1989. Conclusions O u r sham HD studies demonstrate that in vivo contact between blood and complement activating dialysis membranes may induce loss of muscle protein without the interaction with other factors associated with clinical dialysis. The results suggest that this effect is caused primarily by enhanced protein breakdown and that the effect is mediated by prostaglandins. Sham HD using membranes characterized by low complement activation did not elicit enhanced protein breakdown. However, it should be emphasized that in addition to the membrane material, the catabolic effect of clinical dialysis will also dep e n d on the presence of other factors such as acetate or endotoxin in the dialysis fluid which may activate release of cytokines from monocytes. 13. 14. 15. 16. 17. Acknowledgments 18. These studies were supported by grants from the Swedish Medical Research Council T Project No. 1002. AB Gambro and from Fresenius AG. This work was presented by lonas Bergström at the meeting New Frontiers in Renal Disease, Monte-Carlo, March 1 9 - 2 2 , 1992. The authours wish to thank A n n Hellstrom for secretaria1 assistance. 19. 20. 93 A. GUTIERREZ y cols. 2 1 . Baracos V, Rodeman HP, D i n a r e l l o CA, Goldberg AL: Stimulation of muscle protein degradation and prostaglandin El release by leukocytic pyrogen (interleukin-I). N Engl J M e d 308:553-558, 1983. 22. Goldberg AL, Kettelhut IC, Furuni K, Fagan JM, Baracos V: Activation of protein breakdown and prostaglandin E2 product i o n in rat skeletal muscle in fever is signaled by a macrophage product distinct from interleukin-l or other known monokines. J Clin Invest 81:1378-1383, 1988. 23. Hasselgren PO, Talamini M, Lafrance R, James H, Peters J, Fisher JE: Effect of indomethacin on proteolysis in septic muscle. Ann Surg 202:557-562, 1985. 2 4 . N a w a b i M D , Block K P , C h a k r a b a r t i M C , Buse M G : Administration of endotoxin, tumor necrosis, or interleukin-l to rats activates skeletal muscle branched-chain alpha-keto acid dehydrogenase. J Clin Invest 85:256-263, 1990. 2 5 . Gutiérrez A, Alvestrand A, Wahren J, Bergström J: Effect of in vivo contact between blood and dialysis membranes on protein catabolism in humans. Kidney Int 38:487-494, 1990. 26. Gutierrez A, Bergstrom J, Alvestrand A: Protein catabolism in sham hemodialysis. The effect of different membranes. Clin Nephrol 38: 20.29. 1992. 27. Halawa J, Baig S, Qureshi GA: Use of high performance liquid chromatography in defining the abnormalities in the free amino acid patterns in the cerebrospinal fluid of patients with aseptic meningitis. Biomed Chromatogr 5:8-13, 1991. 28. Beasley D, Cohen RA, Levinsky NC: Interleukin I inhibits contraction of vascular smooth muscle. J Clin Invest 83:331-335, 1989. 29. Vicaut E, Hou X, Payen D, Bousseau A, Tedgui A: Acute effects of tumor necrosis factor on the microcirculation in rat cremaster muscle. J Clin Invest 87:1537-1540, 1991. 3 0 . F i n l e y Rj, D u f f IH, H o l l i d a y RL, Jones D, Marchuk JB: Capillary muscle blood flow in human sepsis. Surgery 78:8794,1975. 31. Dinarello CA: The biology of interleukin-1 and its relevance of hemodialysis. Blood Purif 1 :197-224. 1983. 3 2 . Cilowes GHA, Randall HT, Cha CJ: Amino acid and energy metabolism in septic and traumatized patients. J Parent Ent Nutr 4:195-205, 1980. 33. Young VR, Munro HN: 3-methylhistidine and muscle protein turnover: An overview. Fred Proc 37:2291-2300, 1978. 94