Distribution and Metabolism of 14C-Tryptophan

Transcription

1 INFECTION AND IMMuNITY., Oct. 197:3, p Copyright ( 197:3 American Society tor Microbiology Vol. 8. No. 4 Printed in UT.S.A. Distribution and Metabolism of 14C-Tryptophan in Normal and Endotoxin-Poisoned Mice ROBERT J. MOON, ELAINE S. TREMBLAY, AND KATHERINE M. MORRIS Department of Microbiology and Public Health, Michigan State University, East Lansing, Michigan Received tor publication 16 May 1973 DL-Tryptophan (benzene ring-'4c) and its metabolites persist longer and in greater quantity in tissues of endotoxin-poisoned mice than in tissues of' normal mice. Correspondingly less label is excreted in urine and feces and expired as "4CO2 in the poisoned animals. The distribution of label (1.1 x 16 dpm per microgram of tryptophan) was relatively constant whether it was administered alone or in combination with 2 mg of unlabeled L-tryptophan. Tryptophan must be metabolized through the tryptophan oxygenase pathway to be converted to carbon dioxide, but attempts to quantitatively correlate depressed tryptophan oxygenase activity with depressed carbon dioxide production were unsuccessf'ul. It appears that neither tryptophan oxygenase nor substrate availability exclusively determine the quantity of tryptophan converted to,4c2 except under highly selected conditions. The validity of' an earlier suggestion that expired "4CO2 could be used to monitor in vivo tryptophan oxygenase activity is not supported by our data. The disruptive effects of endotoxin on host metabolism (2) must be considered as an important basis for explaining the primary toxicity of the bacterial poison. In recent years numerous changes in tryptophan metabolism in infected and in endotoxin-poisoned animals have been described (1, 9, 1, 12-15) and yet, for the most part, the precise mechanisms and signif'icance of such changes remain unclear. It is well established that a tryptophan load (1 mg per g body weight) is lethal to endotoxin-poisoned mice. Such animals die, frequently in convulsions, within 4 to 8 h and it has been suggested that altered tryptophan metabolism forms the basis of the hyperreactivity (9, 1). More quantitative data must be obtained before it will be possible to describe the specif'ic metabolic alterations in tryptophan metabolism that account for the hyperreactivity. The present paper represents an initial effort in that direction. We have chosen DL-tryptophan (benzene ring-l4c [U]) to investigate the distribution of' tryptophan and its metabolites. Although it is possible that the D and L isomers are metabolized differently (4), Madras and Sourkes have suggested that the D isomer is rapidly converted to the L form in vivo (8). In our investigation we have assumed this to be the case. Our initial objective was to determine whether endotoxin poisoning significantly altered the quantitative distribution and excre- tion of DL-tryptophan- 14C in mice. Experiments were performed in animals with and without tryptophan load. It was observed that the quantity of labeled carbon dioxide expired by endotoxin-poisoned mice was significantly less than that by normal mice. Madras suggested that the amount of 14CO2 expired from benzene ring labeled tryptophan was a direct function of tryptophan oxygenase activity in vivo (8). Since it is well established that tryptophan oxygenase is depressed in endotoxin-poisoned animals (3, 9, 1) a second objective of our study was to determine whether a direct relationship exists between the level of' tryptophan oxygenase activity and the quantity of '4CO2 produced in normal and endotoxin-poisoned mice. MATERIALS AND METHODS Mice. Female CF-I mice (18 to 22 g, Carworth Farms, Portage, Mich.) were used throughout the study. Mice were housed six per cage and wood chips served as bedding. Food (Purina Laboratory Chow, Ralston Purina Co., St. Louis, Mo.) and water were available ad libitum. Endotoxin. Heat-killed Salmonella typhimurium, strain SR-11, served as endotoxin in all experiments. Preparation of the cells has been described previously (1). The mean lethal dose (LD5), determined by the method of Reed and Muench (16), was approximately 4 x 19 dead organisms given intraperitoneally. Chemicals, radioisotopes, and injection schedules. L-Tryptophan was purchased from Nu- 64 Downloaded from on July 26, 218 by guest

2 VOL. 8, 1973 TRYPTOPHAN METABOLISM IN ENDOTOXEMIA 65 tritional Biochemical Co., St. Louis, Mo. DL-tryptophan (benzene ring-14c) was purchased from Amersham Searle, Chicago, Ill., in 5-uCi lots with a specific activity of 95 mci/mmol. Radioisotopes were diluted in nonpyrogenic isotonic saline (Travenol Laboratories, Morton Grove, Ill.). For studies with tryptophan load, L-tryptophan was dissolved in the solution containing the isotope. For the distribution studies, normal or endotoxinpoisoned mice were injected intraperitoneally with 1.1 x 16 dpm of benzene-ring-labeled tryptophan (approximately 1.7,ug of tryptophan) either alone or with 2 mg of unlabeled L-tryptophan (tryptophan load). In studies correlating 14CO2 expired with liver tryptophan oxygenase activity, mice were given either 1, 5, 1, or 2 mg of L-tryptophan containing 5.5 x 14 dpm benzene-ring-labeled tryptophan per milligram of unlabeled amino acid. All tryptophan injections were made between 8 and 9 a.m. Endotoxin was given 1 h prior to tryptophan. Collection and quantitation of isotope in urine and carbon dioxide. To determine the amount of 14C excreted in urine, mice were placed in a 5-ml Erlenmeyer flask containing Whatman filter paper stapled to the underside of a stainless steel grid. At the end of the experiment, the filter paper was removed from the grid, dried, cut into several pieces, and placed in scintillation vials containing 2 ml of counting fluid (5 mg of 1, 4-bis-2- [5-phenyloxazolyl] benzene [POPOP] and 4 g of 2,5-diphenyloxazole [PPO] in 1, ml of toluene). Counts per minute were determined with a Packard 224 liquid scintillation spectrometer system. Preliminary studies showed that more than 95% of the label excreted in urine was trapped on the filter paper. A counting efficiency of 62% was used in calculating disintegrations per minute Ċarbon dioxide was collected according to the method of Madras and Sourkes (7) with slight modifications in our laboratory. In the distribution studies the flask containing the experimental animals was fitted with a 2-hole stopper, open on one side to the air. The other side was connected by glass tubing to a 125-mI Erlenmeyer flask containing 2 ml of ethylene glycol monomethyl ether: ethanolamine (2:1). Respired air was drawn through the CO2 trapping mixture by vacuum. At the end of the experiment the quantity of trapping fluid was measured and 1 ml was transferred to a counting vial containing 1 ml of toluene-ppo-popop and 5 additional milliliters of the trapping fluid. Carbon dioxide collection was modified slightly in the experiments correlating the quantity of '4CO2 expired with tryptophan oxygenase activity to assure maximum quantitation. Mice were placed in a 25-ml plastic bottle which had air holes drilled in the bottom. The opposite end was fitted with a 1-hole rubber stopper and glass tubing linked the plastic bottle sequentially to three scintillation vials. Each vial contained 5 ml of the CO2 trapping fluid described above. Air was pulled through the apparatus by vacuum. Preliminary experiments showed that over 99% of the 14CO2 was trapped in the first two vials. The third vial usually contained negligible counts. A 1-ml amount of toluene-ppo-popop was added to the vials along with sufficient trapping fluid to make the solution miscible. In both instances a counting efficiency of 63% was used in calculating the disintegrations per minute of 14CO2 expired. Collection and quantitation of isotope in tissues. Mice were exsanguinated through the retro-orbital plexus. The stomach and intestine, liver, spleen, kidneys, heart, lung, and brain were removed and wrapped in Saran Wrap. The carcass minus feet, skin, and tail was homogenized in 5 ml of water for 1 to 3 min in a Waring blender. A 5-ml sample of this preparation was removed and processed as described below. The remainder was dried in a tared flask to determine the dry weight of the carcass. All tissues, including the blood, were frozen overnight at C. The frozen tissues were lyophilized, weighed, ground with a mortar and pestle, and 1 mg of the ground tissues was placed in a scintillation vial containing.5 ml of Soluene TM 1 (Packard Instrument Co., Downers Grove, Ill.). The tissues were left at room temperature for 24 h, after which time 1 ml of toluene-ppo- POPOP was added and the disintegrations per minute were determined. Samples were read several times over a period of 7 to 1 days to assure that all phosphorescence was eliminated. The disintegrations per minute were determined through the use of external standard determinations. Data are expressed either as disintegrations per minute of 14C per organ or tissue or as disintegrations per minute per milligram (dry weight) of organ or tissue. All points represent pooled values from at least seven mice. Tryptophan oxygenase assay. Tryptophan oxygenase was assayed according to the method of Knox and Auerbach (7) as modified in our laboratory (1) to insure maximum activation of all latent enzyme. RESULTS Percentage of isotope accounted for in mice without and with tryptophan load. Table 1 shows that 8 to 9% of the 1.1 x 16 dpm of benzene-ring-labeled tryptophan was consistently accounted for at all observation periods in both normal and endotoxin-poisoned mice without tryptophan load. In mice receiving trypto- TABLE 1. metabolites accounted for from all sources' in normal and endotoxin-poisoned mice both with and without tryptophan load Percentage of total DL-tryptophan-'4C and its Normal Endotoxin Hours No trypto- Tryptophan No trypto- Tryptophan phan load load phan load load k a Sources include urine, carbon dioxide, feces, carcass, intestine, liver, kidney, spleen, lung, heart, blood, and brain. h Percentage of isotope recovered. Downloaded from on July 26, 218 by guest

3 - -_ LI~~~~~~~VER 66 MOON, TREMBLAY, AND MORRIS INFECT. IMMUNITY phan load, recoveries ranged between 6 and milligram (dry weight) of tissue or disintegrations per minute in the whole organ. 75%. All calculated percent recoveries expressed in subsequent figures and tables are based on Figure 3 shows the distribution of DL-tryptophan-54C and its metabolites over a 6-h observa- the total number of counts injected. Distribution of DL-tryptophan- 4C and its tion period among the organs and tissues of metabolites among the organs and tissues of normal and endotoxin-poisoned animals given a normal and endotoxin-poisoned mice. One 2-mg L-tryptophan load. The data are expressed as disintegrations per minute per milli- hour after injection of' trace quantities of' DLtryptophan-54C into normal mice 22% of the gram (dry weight) of' tissue. The kidneys, liver, total label injected was found in the carcass intestine, and spleen had the greatest specif'ic plus intestine (Fig. 1). Smaller quantities of activity, whereas the carcass, brain, and blood label were found in the liver and kidneys. Blood, had significantly lower amounts. In all instances endotoxin-poisoned mice had more spleen, heart, lung, and brain cumulatively contained less than 4% of' the isotope. The label in their tissues than normal mice. Compared with Fig. 2, the quantities of label ap- percentage of counts in a given organ remained relatively constant throughout the subsequent peared considerably less in animals with load 5-h observation period. The organs and tissues than without load. It must be remembered, of endotoxin-poisoned mice consistently contained more label than normal mice (Fig. 1). ceived 2, times more tryptophan and hence however. that mice with tryptophan load re- Figure 2 shows the data of Fig. 1 plotted as much greater quantities of the amino acid and disintegrations per minute per milligram (dry its metabolites remain in vivo. weight) of tissue. The spleen which contained Recovery of DL-tryptophan- I4C and its metabolites in urine, carbon dioxide and feces of less than 1% of the total counts (cf. Fig. 1) had the greatest specific activity of' isotope. The normal and endotoxin-poisoned mice. Within carcass, which had 11 to 12% of the total counts, 1 h after injection of' labeled tryptophan without contained only a small concentration of' label load slightly more than 2% of the isotope per milligram of tissue. Analysis of the remaining organs and tissues shows little in the way of and by 6 h nearly 3% of the isotope had been appeared in the urine of normal mice (Table 2), significant differences whether the data were excreted. After 1 h 3.4% of the label appeared as expressed as disintegrations per minute per carbon dioxide and this increased to 7.4% after 6 NORMAL E NDOTOXIN WITHOUT LOAD WITHOUT LOAD. I I -. INTESTINE c INTESTINE 1.:' LLVER b- KIDNEY BLOOD 1. -,SPLEEN SPLEEN LUNG "RAIN BLOOD, H~~~~~~EART LUNG 13RAIN HEART 24II I I FIG. 1. Distribution of DL-tryptophan- 4C and its metabolites among the organs and tissues of normal and endotoxin-poisoned mice over a 6-h observation period. Data are expressed as the percent per organ or tissue of the total disintegrations per minute injected. Downloaded from on July 26, 218 by guest

4 o-j I FIG. 2. Distribution of DL-tryptophan-'4C and its metabolites among organs and tissues of normal and endotoxin-poisoned mice over a 6-h observation period. Data are expressed as disintegrations per minute per milligram (dry weight) of organ or tissue * 3- O 25- \ - ~~~~~~~SPLEEN INTESTINE NORMAL WITHOUT LOAD ~~~~~~~~~LUNG ENDOTOXI N WITHOUT LOAD IT E STINE CAK SEY / ~~~~~~~~~SPLE EN -LIVER \ o 2- K IDNEY \ < - - ~~~~~~~~LIVER 4 - g ~~~~~~~HEART B LOOD B LOOO "EART = 8RAi N C AR CASS _ - 2- CP :11 I5 - E a. 5 - O-J r E ' 15- * NORMAL WITH LOAD INTESTINE LIVER KIDNEY SPLEEN LUNG HEART BLOOD BRAIN ENDOTOX IN WITH LOAD KIDNEY LIVER INTESTINE SPLEEN LUNG BLOOD HEART BRAIN I hou rs FIG. 3. Distribution of DL-tryptophan-'4C and its metabolites over a 6-h observation period among organs and tissues of normal and endotoxin-poisoned mice given 2 mg of L-tryptophan concurrently with labeled tryptophan. Data are expressed as disintegrations per minute per milligram (dry weight) of organ or tissue. 67 Downloaded from on July 26, 218 by guest

5 68 MOON, TREMBLAY, AND MORRIS INFECT. IMMUNITY h. Trace amounts of label were found in the feces but this figure was quite variable and may reflect contamination by urinary metabolites. No attempt was made to wash the intestine free of feces. Cumulatively, over 25% of the label was excreted by 1 h and slightly more than 4% by 6 h in normal mice. In endotoxin-poisoned mice given labeled tryptophan without load only 6% of the isotope was found in urine after 1 h. Seventeen percent had been excreted by 6 h but this was still significantly less than in normal animals (Table 2). Likewise, 14CO2 and isotope in feces was less in endotoxin-poisoned mice. The quantities of label excreted by normal and endotoxin-poisoned mice given a 2-mg load was similar to mice not receiving a tryptophan load (Table 3). The only difference was an increase of approximately 1% in the total label excreted after 6 h by endotoxin-poisoned animals given the tryptophan load (Table 3 versus Table 2). Labeled carbon dioxide expired as a function of dose of tryptophan in normal and endotoxin-poisoned mice. Normal and endotoxin-poisoned mice were injected with either 1, 5, 1, or 2 mg of L-tryptophan containing 5.5 x 14 dpm per mg of DL-tryptophan- 4C. Normal mice given 1 mg of tryptophan converted only small amounts of label to 14CO2 (Fig. 4). In mice given 5, 1, or 2 mg of tryptophan the quantity of 14CO2 was constant for the first hour. In subsequent the quantity of' '4CO2 was a function of the dose of tryptophan injected. By 6 h only negligible quantities of label were detected as 14CO2, even at the 2-mg dose level. Endotoxin-poisoned mice expired less 14CO than normal mice at all time periods. While there was some suggestion that the quantity of '4CO2 expired by poisoned animals was dose dependent, mice given 2 mg of tryptophan produced less 14CO2 than mice given 1 mg of the amino acid. The production of '4CO2 was actually greater in endotoxin-poisoned compared to normal mice in the latter of the observation period, presumably due to the longer persistance of' the substrate. Tryptophan oxygenase activity in normal and endotoxin-poisoned mice given graded doses of tryptophan. In normal mice tryptophan oxygenase increased in proportion to the dose of tryptophan (Fig. 5). By 4 and 6 h after substrate, the enzyme activity was elevated TABLE 2. Percentage of recovery of DL-tryptophan-'4C and its metabolites from urine, carbon dioxide, and feces of normal and endotoxin-poisoned mice Normal Endotoxin Experimental Hours Hours Urine 21.6a Carbon dioxide Feces Total a Percentage of isotope recovered. Downloaded from on July 26, 218 by guest TABLE 3. Percentage of recovery of DL-tryptophan- 4C and its metabolites from urine, carbon dioxide, and feces of normal and endotoxin-poisoned mice given a 2-mg L-tryptophan load concurrently with the labeled amino acid Normal Endotoxin Experimental Hours Hours Urine 22.7a Carbon dioxide Feces Total a Percentage of isotope recovered.

6 NORMAL ENDOTOX IN io 6 - x * 4- a,1 a) 8-2- TRYPTOPHAN: 1 I MG * 1 MG E 5 MG 2 2 MG FIG. 4. "CO2 expired per hour from normal and endotoxin-poisoned mice given either 1, 5, 1, or 2 mg of L-tryptophan containing 5.5 x 14 dpm of DL-tryptophan- 14C per milligram. Endotoxin was injected 1 h prior to tryptophan. O._._ a ) (o CP ; a c C a NORMAL TRY PTOP HAN: D I MG ENDOTOXIN S MG * 1 MG E 2 MG Downloaded from on July 26, 218 by guest J r FIG. 5. Adaptive changes in liver tryptophan oxygenase in normal and endotoxin-poisoned mice given 1, 5, 1, or 2 mg of tryptophan. Endotoxin was injected 1 h prior to tryptophan. Data are expressed as micromoles of kynurenine per gram (dry weight) of liver per hour. Each point represents the average value of six mice. 69

7 61 MOON, TREMBLAY, AND MORRIS INFECT. IMMUNITY from 1 to 3 U above control at all doses tested except 1 mg. In endotoxin-poisoned mice tryptophan oxygenase was also inducible, but to a lesser degree. By 4 to 6 h after tryptophan the enzyme was consistently lower in endotoxin-poisoned mice than in normal mice. DISCUSSION The most striking observation in Fig. 1 to 3 is that tryptophan and its metabolites persist longer and in greater quantities in tissues of endotoxin-poisoned than normal mice. It appears that by 1 to 2 h after tryptophan administration an equilibrium is reached in which the isotope is more concentrated in organs and tissues which actively metabolize tryptophan (such as the intestine and liver) than in organs and tissues which do not actively metabolize the amino acid (such as the carcass and heart). Very little isotope was detected in the blood and brain. The increased quantities of label in tissues of poisoned mice suggest that toxic tryptophan metabolites such as serotonin and quinolinic acid may persist longer in vivo in poisoned animals. Such information has particular significance in determining the basis of the hyperreactivity of endotoxin-poisoned mice to tryptophan. The greatest single quantitative difference in the distribution data was seen in the amount of label excreted in urine. Significantly less tryptophan and its metabolites were excreted by endotoxin-poisoned mice than by normal mice (cf. Tables 1 and 2). Since similar quantities of label were excreted whether the animal was given labeled tryptophan with or without load, a passive rather than an active mechanism appears to be involved in the decreased excretion of label in the poisoned animals. The most likely explanation for such a change is the decreased blood flow to the kidneys due to vascular shock (5, 11). It is not possible to consistently correlate tryptophan oxygenase activity with the quantity of carbon dioxide produced from tryptophan in either normal or endotoxin-poisoned animals. Although there was a suggestion that the 14CO output may correlate with tryptophan oxygenase activity during the first 2 h after a 1 or 2 mg load in normal mice, there was considerably more enzyme potential available than there was "CO2 produced after 2 h. At lower doses (1 and 5 mg) the amount of '4CO2 expired was exclusively a function of substrate availability. Cumulatively, our data in normal mice favors Kim and Miller's suggestion (6) that substrate availability plays a more significant role than tryptophan oxygenase activity in regulating the conversion of' tryptophan to carbon dioxide. The carbon dioxide data from endotoxin-poisoned mice suggests that factors other than the enzyme activity and substrate availability may be influencing the quantity of 14CO2 produced in endotoxin-poisoned animals. Substrate induction of tryptophan oxygenase, while frequently lower in endotoxin-poisoned mice than in normal mice (cf. Fig. 5), did not show constant significant differences. By contrast, the magnitude of the depression in 14CO2 production observed between normal and endotoxin-poisoned mice (cf. Fig. 4) was readily apparent, frequently different statistically, and could not be correlated whatsoever with changes in tryptophan oxygenase activity. Such evidence indicates that decreased tryptophan oxygenase per se is probably not a significant variable accounting for the impaired 14CO2 production. Since the data in Fig. 1 to 3 clearly show that the quantity of isotope from tryptophan persists in greater quantity in livers of endotoxin-poisoned mice, it seems improbable that a lack of substrate in the liver is responsible for the decreased carbon dioxide expired. Conceivably, hemodynamic changes which result in altered blood flow through the organs may be responsible for the observed metabolic alterations. Such a consideration emphasizes the importance of cautious interpretation of data from whole animals so as not to assume that alterations in metabolism reflect exclusively disruption of enzyme regulation in vivo. The use of perfused organs and isolated cell systems should be very useful tools in resolving such conflicts. It is not understood why we were able to account for greater quantities of label in animals without tryptophan load than with tryptophan load (Table 1). Care had been taken to inject the same quantity of isotope into all mice. Extensive studies on our collection procedure for carbon dioxide and urine as well as our preparation of tissues have been performed without providing clues for this discrepency. After 1 h, peritoneal fluid contained less than.1% of the label both with and without tryptophan load (unpublished observation). ACKNOWLEDGMENTS This investigation was supported by Public Health Service grant AI-9394 from the National Institute of Allergy and Infectious Diseases and is Journal Article no from the Michigan Agricultural Experiment Station. LITERATURE CITED 1. Beisel, W. R., and M. I. Rapoport Inter-relations between adrenocortical functions and infectious illness. N. EngI. J. Med. 28: Downloaded from on July 26, 218 by guest

8 VOL. 8, 1973 TRYPTOPHAN METABOLISM IN ENDOTOXEMIA Berry, L. J Metabolic eff'ects of' bacterial endotoxins, p In S. Kadis, G. Weinbaum, and S. J. Ajl (ed.), Microbial toxins, vol. 5. Academic Press Inc., New York. 3. Berry, L. J., and D. Smythe Eff'ects of' bacterial endotoxins on metabolism. VI. The role of tryptophan pyrrolase in response of' mice to endotoxin. J. Exp. Med. 118: Hankes, L. V., and R. R. Brown Metabolism of D- and L-kynurenine-keto-'4C in rats and the effects of unlabeled enantiomers. Proc. Soc. Exp. Biol. Med. 129: Hinshaw, L. B Release of vasoactive agents and the vascular effects of endotoxin, p In S. Kadis, G. Weinbaum and S. J. AjI (ed.), Microbial toxins, vol. 5. Academic Press Inc., New York. 6. Kim, J. H., and L. L. Miller The functional signifticance of changes in activity of' the enzymes tryptophan pyrrolase and tyrosine transaminase after induction in intact rats and the isolated, perfused rat liver. J. Biol. Chem. 244: Knox, W. E., and V. H. Auerbach The hormonal control of tryptophan peroxidase in the rat. J. Biol. Chem. 214: Madras, B. K., and T. L. Sourkes Formation of respiratory "4CO, from variously labeled forms of tryptophan-c'4 in intact and adrenalectomized rats. Arch. Biochem. Biophys. 125: Moon, R. J., and L. J. Berry Role of' tryptophan pyrrolase in endotoxin poisoning. J. Bacteriol. 95: Moon, R. J Tryptophan oxygenase and tryptophan metabolism in endotoxin-poisoned and allopurinoltreated mice. Biochim. Biophys. Acta 23: Nagler, A. L., and S. M. Levenson Experimental hemorrhagic and endotoxic shock, p In S. Kadis, G. Weinbaum, and S. J. Ajl (ed.). Microbial toxins, vol. 5. Academic Press Inc., New York. 12. Powanda, M. C Excretion of tryptophan metabolites in three patients with sandfly f'ever. Amer. J. Clin. Nutr. 24: Powanda, M. C., R. W. Wannemacher, Jr., and G. L. Cockerell Nitrogen metabolism and protein synthesis during pneumococcal sepsis in rats. Infect. Immunity 6: Rapoport, M. I., W. R. Beisel, and R. B. Hornick Tryptophan metabolism during infectious illness in men. J. Infect. Dis. 122: Rapoport, M. I., G. Lust, and W. R. Beisel Host enzyme induction of bacterial inf'ection. Arch. Int. Med. 121: Reed, L. J., and H. A. Muench A simple method for estimating fifty percent endpoints. Amer. J. Hyg. 27: Downloaded from on July 26, 218 by guest