Tryptophan Catabolism by Tryptophan Pyrrolase in Rat Liver THE EFFECT OF TRYPTOPHAK LOADS AKD CHAKGES IN TRYPTOPHAN PYRROLASE ACTIVITY*

We investigated how changes in tryptophan pyrrolase activity and tryptophan loads affect the breakdown of tryptophan was estimated by injecting rats with [ring-2-14-C]tryptophan and measuring respiratory 14-CO2. We concluded, contrary to previous reports, that induction of tryptophan pyrrolase definitely will increase the rate of tryptophan breakdown. Tryptophan loads also increase tryptophan breakdown even in circumstances where there is no increase in tryptophan pyrrolase activity, presumably by increasing the saturation of the enzyme. After a tryptophan load (50 mg per kg) the increase in liver tryptophan concentration lasts only 30 min. The rapid return of liver tryptophan to normal may be due partly to the high turnover rate of liver tryptophan. We estimate that tryptophan pyrrolase degrades tryptophan in vivo at a rate that is equivalent to the whole liver tryptophan concentration in 7.5 min or less.


SUMMARY
We investigated how changes in tryptophan pyrrolase activity and tryptophan loads affect the breakdown of tryptophan by tryptophan pyrrolase. Breakdown of tryptophan was estimated by injecting rats with [ring-ZJ%]tryptophan and measuring respiratory 14C02. We concluded, contrary to previous reports, that induction of tryptophan pyrrolase definitely will increase the rate of tryptophan breakdown.
Tryptophan loads also increase tryptophan breakdown even in circumstances where there is no increase in tryptophan pyrrolase activity, presumably by increasing the saturation of the enzyme. After a tryptophan load (50 mg per kg) the increase in liver tryptophan concentration lasts only 30 min. The rapid return of liver tryptophan to normal may be due partly to the high turnover rate of liver tryptophan.
We estimate that tryptophan pyrrolase degrades tryptophan in vivo at a rate that is equivalent to the whole liver tryptophan concentration in 7.5 min or less.
It is 23 years since Knox (I) showed that tryptophan pyrrolase (L-tryptophan: oxygen oxidoreductase, EC 1.13.11.11) is induced by adrenal cortical hormones. However, it is still unccrtain whether or not the increase in enzyme activity brought about by cortisol causes an increase in the rate of tryptophan degradation. The common assumption that a rise in tryptophan pyrrolase activity would increase the rate of tryptophan catabolism was first challenged by Kim and Miller (2). They showed that in the isolated perfused rat liver induction of tryptophan pyrrolase with hydrocortisone did not alter the rate of clearance of a load of tryptophan from the perfusate or increase the rate of accumulation of kynurenine. Also, using the in viva tryptophan pyrrolase assay of Madras and Sourkes (3), which involves injecting rats with labeled tryptophan and measuring respiratory 14C02, thev found that several-fold increases in tryptophan pyrrolase activity, induced by treatment of rats with cortisol or tryptophan, were associated with at most a 207, increase in the dose of labeled tryptophan that is converted to 14C02. However, under conditions in which the enzymatic activity was already *  maximal after a tryptophan load, a further increase in the dose of tryptophan caused a substantial increase in the conversion of labeled tryptophan to 14C02. They concluded from this that substrate concentration and not enzyme activity was the factor controlling tryptophan catabolism. Their conclusion is supported by work of Powanda and Wannemacher (4) who measured the concentration of NAD, a rnetabolite of tryptophan down the pathway initiated by tryptophan pyrrolase, in the livers of mice treated with either cortisol or tryptophan.
They found that a tryptophan load increased hepatic NAD whereas cortisol did not.
Other reports support the idea that cortisol can increase the rate of tryptophan catabolism. Joseph (5) found an increase in 14C02 output from labeled tryptophan after cortisol administration. Llamas and Vichido (6) reported that cortisol causes a small but significant increase in mouse liver pyridine nucleotides. Formate derived from tryptophan catabolism is used in purine biosynthesis in the mouse, and cortisol treatment increases the rate of utilization of tryptophan for this process (7). We have reinvestigated this problem, in order to determine whether changes in tryptophan pyrrolase activity do affect the rate at which the enzyme catabolizes tryptophan and, if possible, to reconcile the apparently contradictory results described above.  (13) and liver radioactivity as described previously (8).

MATERIALS
Chromatography of Liver Extracts-Chromatography of liver extracts was performed as described previously (8). In this method the liver extract was treated with ethanol to precipitate proteins. The protein-free extract was concentrated by evaporation under a stream of nitrogen and spotted on Whatman No. 3MM chromatography paper. Descending chromatograms were run using butan-1-al/acetic acid/water (4:1:5). After chromatography the paper strips were dried and cut into l-cm lengths which were put in counting vials with 5 ml of toluene containing 5 g of 2,5-diphenyloxazole per liter.

ESULTS
Effect of Cortisol on Formation of Respiratory 14C02 from Labeled Tryptophan-When rats were pretreated with cortisol the total production of r4C02 from [methylene-%]tryptophan over 5 hours was increased by 36%. However, when rates of i4C02 production were plotted as a function of time (Fig. I), it was evident that the effect of cortisol is not uniform over the whole 14C02 collection period. Its effect was greatest between 0.5 and 1 hour after labeled tryptophan administration, when the increase in r4C02 is production between 0.25 and 0.5 hour after labeled tryptophan administration increased by 170% (Fig. 2). While this 170% increase in tryptophan pyrrolase activity over control values measured in viva is not as great as the 260% increase in the conjugated enzyme measured in vitro after cortisol (Table I), it does indicate that an increase in tryptophan pyrrolase activity can increase the rate of tryptophan degradation.
Effect of crMeTrp and Tryptophan Loads on Tryptophan Pyrrolase Activity-To investigate further the effect of increases of tryptophan pyrrolase activity we used rats that had been pretreated with aMeTrp.
In this experiment LvMeTrp was used to induce tryptophan pyrrolase because, unlike cortisol, it not only 1  causes increased synthesis of the enzyme, but also causes enzyme protein already present to conjugate with heme (14). Tryptophan has the same effect (14), so a tryptophan load would not be expected to increase conjugated enzyme in cuMeTrp-pretreated rats, as it would in cortisol-pretreated rats. Nevertheless, some of these animals were also given tryptophan loads to see how this might affect their r4C02 production from labeled tryptophan. Thus, adrenalectomized rats were pretreated (65 hours) with cYMeTrp (100 mg per kg) or 0.9% NaCl solution. Half of each group were given a load of tryptophan (50 mg per kg) at zero time with the labeled tryptophan and i4C02 was collected for 1 hour. The results are shown in Fig. 3. This experiment was per-  When it was repeated with normal rats that had been starved for 15 hours before the experiment, similar results were obtained (Fig. 4) in Rat Liver-The rate of tryptophan degradation is not the only factor that might affect 14C02 production. The uptake of labeled tryptophan into the liver and the specific activity attained will also be of importance. To see how much the variations in '%02 output described above could be attributed to these different ceived a tryptophan load, including the labeled amino acid. The rats were killed 1 hour after receiving the labeled tryptophan, the collection of 14C02 being omitted. Tryptophan pyrrolase and radioactivity were measured in the livers. Tryptophan was also measured in the livers of rats pretreated with saline solution. It was not possible to measure tryptophan in the aMeTrp-pretreated rats because cYMeTrp, as well as tryptophan, is detected by the method of Denckla and Dewey (13). The results are shown in Table II.
In rats pretreated with either saline solution or cYMeTrp, a tryptophan load did not increase the total (apo + holoenzyme) tryptophan pyrrolase within an hour. However, in saline-pretreated rats there was an increase in the conjugated enzyme. This was not so in aMeTrp-pretreated rats, where the cvMeTrp had already increased the extent of conjugation maximally. A tryptophan load also increased the radioactivity found in the liver in both groups of rats.
When rats pretreated with cvMeTrp and saline solution, respectively, were compared, again crMeTrp increased both total and conjugated tryptophan pyrrolase. In these rats with high tryptophan pyrrolase there is also significantly more 14C in the livers. However, the increase is small, of the order of 23%.
It may be noted that a tryptophan load does not cause a significant increase in liver tryptophan at least at 1 hour (Table II)   Rats were injected with saline solution (---) or tryptophan (50 mg per kg) (---) and the liver tryptophan concentration was measured at various intervals as described under "Materials and Methods." Each point represents the mean for six rats f standard error.
implies that tryptophan in the liver must have been elevated at some time during that period. Therefore, we looked at the liver tryptophan concentration of adrenalectomized rats at various time intervals after a tryptophan load (Fig. 5). The liver tryptophan concentration is elevated more than 2-fold at 15 min but falls off rapidly and is back to normal by 60 min.
We also performed measurements in vitro on rats pretreated with cortisol (Table I). In this case the experiment was performed in the same way as the in &JO cortisol experiment except that the rats were killed 0.5 hour after the injection of labeled tryptophan. We chose this interval as it was the time of maximum r4C02 production. In cortisol-pretreated rats the r4C in the liver was not significantly changed, but there was a significant decrease in the liver tryptophan concentration.

Liver
Content of [ring-2-'4C]Tryptophan-Measurement of the liver content of tryptophan and 14C will not, alone, give a measure of the specific activity of liver tryptophan, as we have shown previously that not all the r4C is associated with tryptophan (8). Therefore, rats were treated as described in the previous section except that they received 50 PCi per kg instead of 5 &Ji per kg. Liver extracts were chromatographed as described under "Materials and Methods." Fig. 6 shows the chromatograms obtained from two of the normal rats and Table III gives the percentage of radioactivity that was associated with tryptophan for the four adrenalectomized and two normal rats. Calculations Based on in Vivo and in Vitro Measurements-The liver tryptophan concentration increases acutely and returns rapidly to normal after a tryptophan load; in cortisol-treated rats, which have elevated tryptophan pyrrolase, tryptophan concentration is low. Both of these situations would be explained if the liver tryptophan content is small compared with the rate at which it is degraded by tryptophan pyrrolase. Calculations show that this is indeed so.
One-half hour after [ring-2-r4C]tryptophan administration a normal rat is expiring r4C02 at the rate of 1.95% of the administered 14C per hour (Fig. 2). At this time its liver contains 0.119Y0 of the administered r4C per g of liver (Table I). Of this r4C, 70% is in tryptophan (Table III). Thus, using the mean liver weight found of 6.1 =t 0.2 g (12 determinations) we can calculate that the whole liver will contain 0.50% of the administered r4C as labeled tryptophan. This is being broken down to 14C02 which,  in turn, is being expired at the rate of 1.95% of the administered r4C per hour. Thus, in 15 min the amount of 14C02 expired will be equal to the total [r4C]tryptophan content of the liver. The normal tryptophan content of the liver is 8.18 pg per g (Table I), which gives a total liver tryptophan content of 49.8 pg. Thus, if the rate of tryptophan degradation is equivalent to the tryptophan content of the liver per 15 min this gives a rate of tryptophan degradation in viva of 3.3 pg per min for the total liver tryptophan pyrrolase. This estimate is probably too low, as some of the [14C]formate released from the tryptophan by the action of tryptophan pyrrolase and formamidase (aryl-formylamine aminohydrolase, EC 3.5.1.9) will be incorporated into the l-carbon pool. We have found that about one-half of the r4C in labeled formate injected into a rat is released as r4C02 (8). If we assume that one-half of the labeled formate derived in metabolism from tryptophan is released as r4C02, the rate of tryptophan degradation in the whole rat liver would be about 7 pg per min (35 nmol per min). This activity would consume an amount of tryptophan equal to the total of liver tryptophan content in 7.5 min.
The in V&O assay of tryptophan pyrrolase gives the activity of the conjugated enzyme as 0.45 unit (micromoles of kynurenine formed per hour per g of liver). This is equivalent to a total tryptophan catabolizing capacity for the liver of about 10 pg per min (50 nmol per min). DISCUSSION One of the aims of this work was to determine if an increase in tryptophan pyrrolase activity will lead to an increase in tryptophan breakdown.
Our results clearly indicate that tryptophan breakdown is increased in rats whose tryptophan pyrrolase has been induced with cortisol. Although this conclusion is in disagreement with that of Kim and Miller (2) we do not dispute their results. Kim and Miller found that cortisol-treated rats converted 20% more [methylene-14C]tryptophan to i4C02 than control rats, over 6 hours. Collecting the i4C02 over 5 hours we found a somewhat larger, although qualitatively similar value of 36% (Fig.  1). When we used [ring-2-14C]tryptophan this value increased to 100% with a peak rate of i4C02 production half an hour after the labeled tryptophan injection of 170% more than the control (Fig.   2).
We feel that results obtained with the ring-labeled tryptophan are more appropriate to answer the question of the relation of pyrrolase activity to the rate of tryptophan catabolism in viva than those obtained with the compound labeled in the side chain because the ring-14C has a much shorter pathway to i4C02. Thus, the 14C in [ring-2-i4C]tryptophan is converted to labeled formate by the action of tryptophan pyrrolase and formamidase. The pathway from formate to COz is not firmly established, but liver homogenates can oxidase formate to CO2 in a system that requires Mg2f, ATP, NADPf, and tetrahydrofolic acid (15). The pathway from [mefhylene-%]tryptophan is much longer. The i4C is split off kynurenine or 3-hydroxykynurenine as the methyl group of alanine. This is then transaminated to pyruvate which is oxidized to COZ.
We also suggest that results obtained at short time intervals after labeled tryptophan administration are more valid than those obtained by collecting i4C02 over long periods. Our calculations indicate that there is a high rate of turnover of tryptophan in the liver. In such a situation the estimates of the rate of i4C02 production at long time intervals after labeled tryptophan injection would probably depend more 011 the specific activity of the liver tryptophan than on the rate of tryptophan catabolism.
Results obtained using [ring-2-i4C]tryptophan and i4C02 collection for only half an hour indicate that cortisol does increase tryptophan breakdown. However, after cortisol administration the rate of COZ production does not increase as much as the conjugated tryptophan pyrrolase activity measured in vitro. This may be due partly to the decline in liver tryptophan concentration (Table I). The normal liver tryptophan concentration of 8.18 pg per g of liver is equivalent to about 50 pM. However, in vitro tryptophan pyrrolase does not reach saturation until the tryptophan concentration is above 1 mM (16). The experiments in Figs. 3 and 4 show that tryptophan loads, as well as increases in tryptophan pyrrolase activity, raise the rate of tryptophan catabolism. Some of the rats that received tryptophan loads had been pretreated with otMeTrp, a compound which promotes conjugation of the enzyme with heme (14). Thus, the tryptophan load does not increase the conjugation of the enzyme any further after otMeTrp (Table II). Despite this, tryptophan degradation is increased (Fig. 3). Hence we agree with Kim and Miller (2) when they concluded that tryptophan loads could increase tryptophan breakdown even under circumstances in which tryptophan pyrrolase did not increase.
The conclusions reached above depend on the acceptance of i4C02 evolution from [ring-2-Vltryptophan as an index of tryptophan breakdown by tryptophan pyrrolase. Although this has been commonly assumed there is another factor in addition to the rate of tryptophan catabolism that could affect i4C02 production. This is the specific activity of the liver tryptophan, which will depend on the uptake of the radioactive tryptophan into the liver, and its dilution by endogenous cold tryptophan. For this reason we measured the concentrations of tryptophan and radioactivity in the liver (Tables I and II). We also determined by paper chromatography the portion of the label that was attributable to unmetabolized tryptophan (Fig. 6). This decreases with increasing 14C02 production ( Table III). The fact that a higher percentage of the radioactivity is associated with metabolic products when i4C02 release is high is additional evidence that i4C02 release is an index of tryptophan breakdown. Thus, in cortisol-treated rats, 0.5 hour after the labeled tryptophan injection, 67% more of the radioactivity is due to metabolites than in controls. This supports the conclusion that cortisol does increase tryptophan breakdown.
In adrenalectomized rats a tryptophan load causes a significant increase (p < 0.001) in the liver radioactivity (Table II). However, using the values in Table III for the portion of radioactivity that is due to labeled tryptophan, the controls of Table  II would have O.ll%, the tryptophantreated rats 0.10% of the administered 14C per g liver as tryptophan. As there is no change in the radioactive tryptophan or in the tryptophan concentration, the specific activity of the tryptophan remains unchanged. The same is true for the effect of cortisol on normal rats. In this case cortisol decreases both the liver tryptophan and labeled tryptophan concentrations. The specific activity remains the same, and thus 14C02 output alone is a valid index of tryptophan breakdown in both these situations.
From our results we calculated a rate of tryptophan breakdown by tryptophan pyrrolase in normal rats, under the circumstances of our experiments of about 7 pg per min (35 nmol per mm). This value will be only approximate because of assumptions made in calculating it. The main assumption is the percentage of endogenously derived formate that is converted to i4C02; this could well be less than 50%. Thus our value will set an approximate lower limit for the rate of tryptophan catabolism. It is likely that tryptophan is being broken down in the liver at a rate equal to its content in the whole liver every 7.5 min or faster. This means that the liver tryptophan must have a very high turnover rate.
After a tryptophan load the liver tryptophan concentration is back to normal within 30 min (Fig. 5) in contrast to the brain tryptophan which has not started to decline by 1 hour (17). The functional tryptophan pyrrolase activity will be increased after a tryptophan load both by an increase in the conjugation with heme and by an increase in saturation with substrate. In view of the high rate of tryptophan breakdown in the liver relative to its total tryptophan content it may be that the excess tryptophan in the liver is rapidly broken down. After cortisol treatment tryptophan pyrrolase is elevated and the liver tryptophan declines. Further work will be necessary to determine how much of this decline is due to an increased rate of catabolism and how much to other factors.
The liver tryptophan concentration can affect tryptophan pyrrolase activity, and this work indicates that the enzyme may affect the concentration of its substrate. These two effects would act to stabilize the liver tryptophan concentration. Thus a high liver tryptophan concentration increases the enzymic activity to speed tryptophan breakdown, while a high tryptophan pyrrolase activity may lower the liver tryptophan concentration, thus decreasing both the rate of tryptophan catabolism and the active enzyme concentration.
Understanding the factors that control tryptophan breakdown in the liver should help in determining under what circumstances tryptophan pyrrolase can affect other metabolic processes. Thus Bloxam, Warren, and White (18) have shown that the stimulation of gluconeogenesis during starvation (19) is accompanied by a decrease in the liver tryptophan concentration.
As tryptophan metabolites can inhibit gluconeogenesis (20,21) they contend that this is consistent with a regulatory role for tryptophan in gluconeogenesis. Thus, this is one area in which tryptophan pyrrolase may be important.
Other possible areas are control of brain 5HT (22), regulation of liver NAD concentration (4), and, as tryptophan is the least abundant amino acid in the pool available for protein synthesis (23), also protein synthesis.