Chester Clinic

 

98-071

Published May 3, 1999 Inhibition of T Cell Proliferation by Macrophage
Tryptophan Catabolism
By David H. Munn,*‡ Ebrahim Shafizadeh,* John T. Attwood,*Igor Bondarev,* Achal Pashine,* and Andrew L. Mellor* From the *Institute of Molecular Medicine and Genetics and the ‡Department of Pediatrics, Medical College of Georgia, Augusta, Georgia 30912 We have recently shown that expression of the enzyme indoleamine 2,3-dioxygenase (IDO) dur-ing murine pregnancy is required to prevent rejection of the allogeneic fetus by maternal T cells.
In addition to their role in pregnancy, IDO-expressing cells are widely distributed in primaryand secondary lymphoid organs. Here we show that monocytes that have differentiated underthe influence of macrophage colony-stimulating factor acquire the ability to suppress T cell pro-liferation in vitro via rapid and selective degradation of tryptophan by IDO. IDO was induced in macrophages by a synergistic combination of the T cell–derived signals IFN-g and CD40-ligand. Inhibition of IDO with the 1-methyl analogue of tryptophan prevented macrophage-mediated suppression. Purified T cells activated under tryptophan-deficient conditions wereable to synthesize protein, enter the cell cycle, and progress normally through the initial stagesof G1, including upregulation of IL-2 receptor and synthesis of IL-2. However, in the absenceof tryptophan, cell cycle progression halted at a mid-G1 arrest point. Restoration of tryptophan to arrested cells was not sufficient to allow further cell cycle progression nor was costimulationvia CD28. T cells could exit the arrested state only if a second round of T cell receptor signal-ing was provided in the presence of tryptophan. These data reveal a novel mechanism by whichantigen-presenting cells can regulate T cell activation via tryptophan catabolism. We speculate that expression of IDO by certain antigen presenting cells in vivo allows them to suppress un-wanted T cell responses.
macrophage • indoleamine 2,3-dioxygenase • T cells • tryptophan • macrophage Certain macrophages (Møs)1 and possibly other subsets In the course of our studies, we found that MCSF-derived of APCs suppress T cell responses (1, 2). Immunosup- Møs were capable of rapidly and selectively depleting the pressive APCs have been hypothesized to play an important essential amino acid tryptophan from cocultures and that role in maintaining peripheral T cell tolerance. We have this depletion occurred only in response to attempted T cell previously shown that Møs that differentiate in vitro un- activation. Møs are known to possess the inducible trypto- der the influence of macrophage colony-stimulating factor phan-degrading enzyme indoleamine 2,3-dioxygenase (IDO), (MCSF) acquire the ability to suppress T cell proliferation which catalyzes the initial and rate-limiting step in the metab- (3, 4). This attribute was not constitutively present but olism of tryptophan along the kynurenine pathway (5–8). It rather was invoked only in response to attempted T cell ac- has been postulated that the role of IDO is to inhibit prolif- tivation. Suppressor activity was restricted to specific Mø eration of eukaryotic intracellular pathogens (9–13) or tu- phenotypes (e.g., the phenotype produced by MCSF), with mor cells (14) by depriving them of tryptophan. At the other phenotypes supporting normal T cell activation (3).
time of this study, however, no role had been proposed for Taken together, these characteristics suggested that the in- IDO in regulating T cell responses. Recently, we have re- hibitory properties of MCSF-derived Møs might reflect a ported that IDO expression in placenta is critically involved physiologic system for regulating T cell activation. How- in preventing rejection of the allogeneic fetus by maternal ever, the mechanism of this inhibition was unknown.
T cells (15). The current study tests the hypothesis thattryptophan depletion via IDO is the mechanism by which MCSF-derived Møs inhibit T cell activation in vitro and Abbreviations used in this paper: CD40L, CD40 ligand; IDO, indoleamine identifies a tryptophan-sensitive cell cycle arrest point dur- 2,3-dioxygenase; Mø, macrophage; MCSF; macrophage colony-stimulat-ing factor.
ing T cell activation.
J. Exp. Med.  The Rockefeller University Press • 0022-1007/99/05/1363/10 $2.00Volume 189, Number 9, May 3, 1999 1363–1372http://www.jem.org Published May 3, 1999 Materials and Methods
Recombinant human MCSF was the gift of Genet- Validation studies showed this assay to be linear in the range of ics Institute, Cambridge, MA. Recombinant human IFN-g was 0.1–100 mM, with an estimated threshold sensitivity of 0.05 mM.
the gift of Genentech, South San Francisco, CA. Recombinant Where it was desirable to show that tryptophan depletion in human CD40 ligand (CD40L) homotrimer was the gift of W.
cultures was due to IDO activity, culture supernatants were as- Fanslow, Immunex Corp., Seattle, WA. The IDO inhibitor sayed by HPLC for the presence of kynurenine. IDO catalyzes 1-methyl-d,l-tryptophan (16) was purchased from Aldrich Chem- the oxidation of tryptophan to N-formylkynurenine, which in ical Co. 6-nitro-tryptophan (17) was synthesized by D. Boykin, Møs is rapidly converted into kynurenine (22) and then to other Georgia State University, Atlanta, GA, using a modification of downstream metabolites (7). With the exception of tryptophan the method of Moriya et al. (18). Polyclonal antiserum against oxygenase, which is found only in hepatocytes, IDO is the only human IFN-g was obtained from Biosource International. All enzyme capable of degrading tryptophan along the kynurenine other reagents were obtained from Sigma Chemical Co. unless pathway (8). Thus, the appearance of kynurenine in cultures was unambiguous evidence of functional IDO activity. However, be- Cell Isolation and Culture. Human peripheral blood mono- cause kynurenine can be converted into other downstream me- cytes and lymphocytes were isolated from healthy volunteer do- tabolites, this assay was not quantitative. Where quantitative data nors by leukocytapheresis and counterflow centrifugal elutriation, were required, the tryptophan depletion assay described above following appropriate informed consent under a protocol ap- proved by our Institutional Review Board. Monocytes (.95% HPLC assays were performed by the Medical College of purity by cell surface markers) were cultured in 96-well plates as Georgia Molecular Biology Core Facility. Samples were prepared previously described (4) using RPMI 1640 with 10% newborn by extracting 150 ml culture supernatant with 1 ml methanol.
calf serum (Hyclone) plus MCSF (200 U/ml).
Precipitated proteins were removed by centrifugation and the su- T cell activation studies in cocultures were performed as previ- pernatant dried under vacuum. Samples were resuspended in 100 ml ously described (4), using the above medium supplemented with initial mobile phase (deionized water) and an aliquot injected an additional 5% FCS. In brief, Møs (5 3 104 cells/well) were al- onto a C-18 column (Phenomenex Luna C-18; 250 3 4.6 mm; lowed to differentiate for 4–6 d in MCSF, and then autologous 5 mm). Samples were eluted with a linear gradient of acetonitrile lymphocytes (2 3 105 cells/well) were added along with mito- in water (0–80% over 20 min), and absorbance was measured at gen. The mitogens used in this study were anti-CD3 mAb (100 254 nm. Standards for tryptophan, kynurenine, and 1-methyl- ng/ml, clone OKT3; American Type Culture Collection) and tryptophan were run with each assay to establish retention times.
staphylococcal enterotoxin B (5 mg/ml; Sigma Chemical Co.).
In preliminary validation studies, the identity and purity of each Both gave equivalent results; the data shown are from anti-CD3 peak was confirmed by mass spectroscopy.
unless otherwise specified. T cell proliferation was assessed by Protein Synthesis and Amino Acid Analysis. Total protein syn- standard thymidine incorporation assay as described (3). When thesis was measured as incorporation of tritiated leucine (4 mCi/ml) T cell activation was studied without Møs, fresh autologous over 24 h. TCA-insoluble proteins were precipitated and washed monocytes were added (1:4) as nonsuppressive accessory cells.
three times in 5% TCA, and the precipitate was analyzed by liquid Conditioned medium from cocultures of T cells and Møs was scintillation counting. Amino acid concentrations in culture su- prepared by harvesting supernatant 48 h after T cell addition.
pernatants were measured by HPLC in our clinical Neonatal Nu- Conditioned medium was then used to support a second round trition Laboratory.
of T cell activation. Mitogen and other additives were prepared Møs were harvested with EDTA and total RNA in tryptophan-free buffers.
prepared. Sample RNA (1 mg) was reverse transcribed with avian A chemically defined, serum-free medium (19) selectively de- myeloblastosis virus (AMV)-RT, and a 182-bp fragment amplified ficient in tryptophan was prepared using tryptophan-free RPMI with the following primers: forward, bp 237–254 of the pub- 1640 (Select-amine kit; GIBCO BRL) supplemented with in- lished sequence (23); reverse, bp 402–418, spanning exons 3–4.
sulin (10 mg/ml), iron-saturated transferrin (5 mg/ml), and BSA Product formation was assessed by agarose gel electrophoresis and (1 mg/ml ultra-pure grade; measured concentration of free trypto- ethidium bromide staining. PCR product was isolated from the phan ,5 nM). Preliminary validation experiments confirmed that gel and reamplified with internal primers to confirm specificity.
T cell proliferation in this medium was undetectable but was Flow Cytometry. Two-color FACS® analysis was performed comparable to serum-based medium when tryptophan was using directly conjugated mAbs as previously described (24). T added. To study T cells in the absence of Møs, T cells were acti- lymphocytes were identified by gating on CD3-positive cells, and vated using anti-CD3 mAb adsorbed onto plastic tissue culture expression of CD69, CD25, and CD71 was measured in the sec- wells (0.5 mg/cm2 in bicarbonate buffer, pH 9) plus soluble anti- CD28 mAb (1 mg/ml; PharMingen).
Experiments for all figures were replicated at least Tryptophan and IDO Assays. The tryptophan-degrading activ- three times, and representative data are shown. Data points were ity of Møs reflects a multifactored combination of IDO expres- measured in triplicate and the mean reported. Error bars show sion, tryptophan transport into the cells, and intracellular condi- standard deviation. Where SD was ,10%, error bars have been tions that posttranslationally affect enzyme activity (20). Therefore, omitted for clarity. Comparisons of multiple groups within a sin- when tryptophan depletion was the outcome of interest, we mea- gle experiment were by ANOVA.
sured the rate of disappearance of tryptophan from culture super-natants over time. Tryptophan was assayed using the method ofBloxam and Warren (21). Proteins were precipitated with 10% TCA and free tryptophan assayed after conversion to norharmanusing formaldehyde and FeCl Cocultures of Møs and T Cells Are Selectively Depleted of 3. The reaction product was mea- sured spectrofluorometrically (excitation 360 nm, emission 460 Supernatants were harvested from cocultures nm) and compared against a standard preparation of tryptophan.
of Møs and mitogen-activated T cells after 48 h. Fresh Regulation of T Cells by IDO

Published May 3, 1999 lymphocytes were suspended in conditioned medium andactivated with additional mitogen. Fig. 1 shows that condi-tioned medium completely failed to support T cell prolifer-ation (,1% of the proliferation in fresh medium). How- Figure 2.
Dose–response rela- ever, the addition of tryptophan to conditioned medium tionship to tryptophan for T cell fully restored its ability to support T cell proliferation, indi- proliferation. Tryptophan was ti- cating that tryptophan was the only component that had trated in coculture-conditioned been depleted. Consistent with this finding, amino acid medium (prepared as described inFig. 1) and proliferation of T cells analysis of conditioned media showed that all other es- measured after 72 h.
sential amino acids were present, and only tryptophan wasundetectable (data not shown). Titration of reagent trypto- To confirm the presence of IDO activity, culture superna- phan into conditioned medium gave a half-maximal con- tants were assayed for kynurenine. As shown in Fig. 3 C, centration for T cell proliferation of 0.5–1 mM (Fig. 2), depletion of tryptophan was accompanied by a correspond- compared with a measured concentration of tryptophan in ing increase in kynurenine production, confirming the pres- coculture-conditioned medium of ,50 nM (the detection ence of functional IDO activity.
limit of our assay). Control-conditioned media from Møs Inhibition of IDO Prevents Mø-mediated Suppression of T alone, from cocultures of Møs 1 T cells without mitogen, We next asked whether pharmacologic inhibition or from T cells activated with fresh monocytes instead of of IDO could prevent suppression of T cells in cocultures.
Møs all supported T cell proliferation comparably to freshmedium (90–140% of control; n 5 3–4/group).
Expression of IDO by MCSF-derived Møs. tryptophan elimination were measured by coincubatingMøs and T cells with mitogen for 24 h to allow upregula-tion of the tryptophan depletion pathway and then addingfresh tryptophan and following its disappearance. As shownin Fig. 3 A, tryptophan was eliminated by first-order kinet- ics with a half-life of 2–3 h. The initial rate of eliminationwhen tryptophan was not limiting was up to 20,000 pmol/106 cells/h. This far exceeded the consumption attributableto cellular metabolism (see control, Fig. 3 A), as Møs with- Figure 3.
Elimination kinetics of out activated T cells depleted tryptophan at a rate of 300 6 tryptophan in cocultures and expression 130 pmol/106 cells/h (cumulative measurement obtained of IDO by MCSF-derived Møs. (A) over 7 d; data not shown). This implied that the majority MCSF-derived Møs were cultured for of tryptophan depletion by activated Møs was due to an in- 24 h with autologous T cells, either with ducible system, which we suspected was IDO.
(d) or without (j) anti-CD3 mAb. Themedium was then replaced with fresh Consistent with this finding, abundant IDO mRNA was medium and supernatant from replicate detectable by RT-PCR in Møs after activation, whereas cultures harvested at the times shown. Tryptophan concentration was as- before activation, IDO message was undetectable (Fig. 3 B).
sayed spectrofluorometrically as described in Materials and Methods. (B)IFN-g–inducible IDO mRNA in MCSF-derived Møs. RT-PCR show-ing IDO expression in MCSF-derived Møs before (lane 4) and after (lanes1–3) activation for 24 h with recombinant IFN-g. Starting RNA for thereverse transcriptase reaction in lanes 1–3 was from 20,000, 2000, and 200activated Møs, respectively, and from 20,000 unactivated Møs in lane 4.
Figure 1.
Lane 5 shows amplification of human IDO plasmid template giving the tioned medium is selectively de- expected 182-bp product. (C) HPLC analysis of Mø culture supernatants pleted of tryptophan. Human showing degradation of tryptophan and production of kynurenine.
monocytes were allowed to dif- MCSF-derived Møs were preactivated for 24 h with IFN-g to induce ferentiate for 5 d in MCSF. Then, IDO expression, and then the spent medium was replaced 90:10 with T cells were added and activated fresh medium. Trace 1 shows the analysis of supernatant immediately af- with anti-CD3 mAb. Condi- ter adding fresh medium (time 0); trace 2 shows the conditioned medium tioned medium was harvested 24 h later. The number of Møs in these experiments was kept low so that from cocultures after 48 h and some tryptophan would be detectable at the end of the assay. The traces then used to support a second shown represent the portion of the elution gradient between 28 and 42% round of activation with fresh T acetonitrile (minutes 7.00–10.50), during which kynurenine (K) and cells. Replicate cultures were sup- tryptophan (T) appeared. The peak labels are positioned at the points at plemented with individual amino which the purified standards eluted, which were within 63 s of the cor- acids to the concentrations nor- responding sample peak. Compounds present in culture medium that also mally found in RPMI 1640. Con- absorbed at OD254 (unlabeled peaks) were readily resolved from tryp- trol cultures received either fresh tophan and kynurenine, and the T and K peaks were confirmed by mass medium (CTL) or no supple- spectroscopy (see Materials and Methods). The experiment shown used ment (PBS). Proliferation was purified Møs activated with recombinant ligands; identical results were measured by thymidine incorpo- obtained when Møs were activated in coculture with T cells plus mito- ration after 72 h.
gen. One of four experiments is shown.

Published May 3, 1999 Figure 4.
Inhibition of IDO activity prevents Mø-mediated suppression. (A) 1-methyl-tryptophan inhibits Mø IDO enzyme activity. MCSF-derived Møs were activated with IFN-g for 24 h to induce IDO expression, andthen fresh medium was added as described in Fig. 3, along with 1-methyl-tryptophan (1 mM). Supernatants wereanalyzed by HPLC for tryptophan (T) and kynurenine (K) immediately after the addition of fresh medium (trace1) and 24 h later (trace 2). The 1-methyl-tryptophan peak (M) is off scale at the settings used. Control cultures forthese experiments (Møs with IFN-g but without 1-methyl-tryptophan) uniformly had .90% reduction in tryp-tophan at hour 24, with a corresponding increase in kynurenine, as shown in Fig. 3. The experiment shown usedpurified Møs activated with recombinant ligands; identical results were obtained when Møs were activated in co-culture with T cells plus mitogen. (B) 1-methyl-tryptophan prevents T cell suppression in cocultures. T cells wereadded to MCSF-derived Møs and activated with anti-CD3 mAb. Replicate cultures were treated with varying concentrations of 1-methyl-tryptophan. Proliferation was measured after 72 h by thymidine incorporation. Controls (s) show proliferation by T cellswithout Møs at the highest concentration of inhibitor used (there was no effect of inhibitor on T cells alone throughout the range of concentrations shown). (C) A second inhibitor of IDO activity, 6-nitro-tryptophan, showed similar reversal of Mø-mediated inhibition of T cells. Experimental designas in B. (D) Supplementation with high concentrations of tryptophan prevents Mø-mediated suppression. Møs were seeded at low density (CC-lo; 5 3104 cells/well) and high density (CC-hi; 2 3 105 cells/well) and the medium supplemented with 53 the normal tryptophan concentration. Proliferationwas measured at hour 72. Controls show proliferation by Møs alone (M) and T cells alone (T).
The compound 1-methyl-tryptophan has been reported to Tryptophan-degrading Activity Is Synergistically Induced by be a potent competitive inhibitor of IDO activity when Early Signals of T Cell Activation. Møs did not degrade tryp- tested in vitro using purified enzyme (16, 17). To deter- tophan simply as a result of contact with T cells. Rather, mine whether this agent could inhibit IDO activity in in- there was an obligate requirement that the T cells attempt to tact Møs, we added 1-methyl-tryptophan to activated Mø activate (Fig. 3 A). In light of the existing studies implicating cultures. As shown in Fig. 4 A, the presence of 1-methyl- IFN-g as an inducer of IDO (25–27), we suspected that tryptophan markedly reduced the degradation of trypto- IFN-g from activating T cells might be the signal for IDO phan by Møs, and this was accompanied by a correspond- induction. Consistent with this idea, low but detectable levels ing inhibition of kynurenine production (e.g., compare the of IFN-g were present in cocultures within 4–6 h of T cell ratio of tryptophan to kynurenine after 24 h in Fig. 3 C), activation, coincident with the time that tryptophan degrada- confirming that the target of the inhibitor was IDO.
tion began (Fig. 5 A). Neutralizing antibodies against IFN-g Functionally, the addition of 1-methyl-tryptophan to co- reduced the induction of tryptophan-degrading activity (Fig.
cultures abrogated the ability of Møs to suppress T cell prolif- 5 B) and reduced suppression of T cells by Møs (Fig. 5 C), eration in a dose-dependent manner (Fig. 4 B). Although supporting a role for IFN-g in the signaling pathway. How- this finding was consistent with the proposed role for IDO in ever, the dose–response relationship using recombinant Mø-mediated suppression, it might in theory indicate an un- IFN-g revealed that relatively high concentrations of IFN-g anticipated immunostimulatory role for 1-methyl-tryptophan were required for full induction of tryptophan-degrading ac- itself. To exclude this possibility, we synthesized a second ana- tivity (Fig. 5 D). We therefore asked whether there was an logue of tryptophan, 6-nitro-tryptophan, which has also been additional signal that might act in concert with IFN-g.
reported to inhibit purified IDO enzyme in vitro (17). As CD40L is upregulated early in T cell activation and is known shown in Fig. 4 C, 6-nitro-tryptophan also prevented Mø- to act synergistically with IFN-g to activate other Mø func- mediated suppression in a dose-dependent fashion (Fig. 4 C).
tions (28). Fig. 5 D shows that CD40L exerted marked syn- Finally, we tested the effects of supplemental tryptophan on ergy with IFN-g, shifting the dose–response curve for IFN-g suppression. As shown in Fig. 4 D, high levels of tryptophan one to two orders of magnitude so that significant tryptophan did prevent suppression of T cells, provided that the number depletion began at IFN-g concentrations of ,1 U/ml.
of Møs in cocultures was kept low. At our usual concentra- Effect of Tryptophan Deprivation on T Cell Protein and DNA tions of Mø, it proved impossible to supplement with suffi- We have previously shown that T cells acti- cient tryptophan to overcome its rapid degradation. Thus, by vated in coculture with MCSF-derived Møs initially enter the use of two pharmacologic inhibitors of IDO and by tryp- the cell cycle but arrest before the first G1/S transition (4).
tophan supplementation, the mechanism of T cell suppression We therefore asked whether a comparable phenomenon in our system appeared to be depletion of tryptophan by IDO.
occurred when T cells were activated in the absence of Regulation of T Cells by IDO Published May 3, 1999 Figure 5.
IFN-g and CD40L act synergistically to induce IDO. (A) MCSF-derived Møs were cocultured with T cells and anti-CD3 mAb. Following lymphocyte addition, culture supernatants were harvested at the times shown and assayed for IFN-g (j, left axis) and tryptophan concentration (d, rightaxis). (B) Møs and T cells were cocultured with mitogen in the presence of various concentrations of neutralizing anti–IFN-g antiserum. Tryptophan con-centration in culture supernatants was determined after 18 h. A low density of Møs was used for these experiments so as not to obscure the effect ofIFN-g. (C) Møs were cultured at a range of seeding densities as shown, and then T cells and anti-CD3 mAb were added either with (j) or without (d)neutralizing antibodies to IFN-g (100 neutralizing U/ml). Antibodies to IFN-g reduced the effectiveness of Møs in suppressing T cells, particularly whenthe number of Møs was limiting. (D) MCSF-derived Møs were cultured for 24 h with various concentrations of recombinant IFN-g, either in the pres-ence (j) or absence (m) of recombinant CD40L (500 ng/ml). At the end of the activation period, culture supernatants were assayed for the concentra-tion of tryptophan remaining. The single round point shows tryptophan degradation in response to CD40L alone.
tryptophan. Purified T cells (without monocytes or Møs) and the time to entry into S phase determined. Control were cultured in tryptophan-free medium using immobi- cells, cultured with tryptophan throughout, reproducibly en- lized anti-CD3 plus anti-CD28 mAb as activating stimuli.
tered S phase 28–32 h after initial TCR engagement (times In this system, T cells stimulated in the presence of trypto- are reported as 4-h ranges to reflect the limit of precision of phan activated normally, whereas T cells stimulated with- the assay). In contrast, T cells that had been preactivated out tryptophan arrested before entry into the first S phase, under tryptophan-free conditions required only 12–16 h to as shown by the complete absence of DNA synthesis (Fig. 6).
enter S phase after tryptophan was added (Fig. 9 A), indi- This arrest was not due to an absence of protein synthesis, cating that significant progression through G1 had occurred as T cells without tryptophan successfully upregulated CD69, in the absence of tryptophan. The tryptophan-sensitive ar- CD25 (high-affinity IL-2 receptor), and CD71 (transferrin rest point was stable, with T cells surviving .72 h in the receptor) (Fig. 7) and secreted IL-2 and IFN-g (Fig. 8), all absence of tryptophan with no loss of viability. When tryp- of which require new protein synthesis (29). Total protein tophan was added to arrested cells, the time of entry into synthesis, measured as incorporation of radiolabeled leucine S phase was consistently 12–16 h, regardless of whether during the first 24 h of activation, continued at a rate 40– cells had been preactivated for 36, 48, or 72 h without 55% of controls (n 5 3; see Materials and Methods), despite tryptophan. This suggested that the arrest occurred at a spe- the absence of exogenous tryptophan. Nonetheless, no en- cific point in G1 and that this position in the cell cycle was try into S phase occurred. Thus, T cells activated in the ab- maintained until tryptophan was restored.
sence of exogenous tryptophan arrested in a fashion similar From the preceding experiments, we estimated that the to that which we had previously observed in coculture.
tryptophan-independent portion of G1 was z14 h (calcu- Identification of a Tryptophan-sensitive Arrest Point in lated as the difference between the average time to S phase The upregulation of early G1 markers suggested for resting T cells versus the time to S phase for preacti- that some portion of G1 was tryptophan independent. To vated cells). To test this estimate, we deprived T cells of test this hypothesis, T cells were activated for various times in tryptophan during the initial 14 h of activation, then added the absence of tryptophan, and then tryptophan was added tryptophan just before the putative arrest point. As shownin Fig. 9 B, cultures deprived of tryptophan for the first 14 hentered S phase identically to T cells supplied with tryp-tophan throughout, supporting the hypothesis that the ini- Figure 6.
T cells do not enter tial portion of G1 was independent of tryptophan. In addi- S phase in the absence of tryp- tional experiments (not shown), delaying the addition of tophan. T cells were activated tryptophan beyond 14 h introduced a corresponding delay with immobilized anti-CD3 mAb in entry into S phase, supporting the proposed localization plus anti-CD28, either in chemi-cally defined tryptophan-free of the arrest point close to hour 14.
medium (j) or in the same me- T Cells Can Commit to Cell Division in the Absence of Tryp- dium supplemented with 25 mM Resting (G0) T cells require TCR signaling in tryptophan (d). DNA synthesis order to enter G1, but subsequent progression through the was assayed by thymidine incor-poration at the times shown.
cell cycle rapidly becomes TCR independent (for a review Published May 3, 1999 Figure 8.
Production of IFN-g and IL-2 by T cells deprived of tryp- tophan. T cells were activated with anti-CD3 mAb in tryptophan-freemedium (j) or in the same medium supplemented with 25 mM tryp-tophan (d) and the concentration of IFN-g (A) and IL-2 (B) in culturesupernatants determined at the times shown.
of tryptophan. As shown in Fig. 11, despite their previous48-h exposure to anti-CD3, the arrested T cells still re-quired additional TCR signaling plus the presence of tryp-tophan to exit the arrested state. Even costimulation via CD28 was not sufficient to promote cell cycle progressionin the absence of TCR engagement.
In this study, we show that tryptophan catabolism via IDO is the mechanism by which MCSF-derived Møs sup-press T cell proliferation in vitro. We have recently testedthis hypothesis of IDO-mediated T cell suppression in vivo Figure 7.
Expression of activation markers on T cells deprived of tryp- tophan. T cells were activated in tryptophan-free medium using immobi-lized anti-CD3/CD28 (heavy trace), or cultured under identical condi-tions but without anti-CD3/CD28 (light trace). At the times shown,both groups were harvested and stained for expression of activation mark-ers as described in Materials and Methods.
see reference 29). In our system, commitment to TCR-independent cell division was first detectable z6 h afterTCR engagement, and most cells were committed by hour Figure 9.
T cells that have entered the tryptophan-sensitive arrested 12. As shown in Fig. 10, this commitment occurred identi- state retain their position in mid-G1. (A) T cells were activated in tryp- cally regardless of whether tryptophan was present or ab- tophan-free medium using immobilized anti-CD3/CD28 (d). After a sent during the relevant time period. As long as the cells period of preactivation (24–72 h with similar results; 48 h in the experi- were not allowed to arrest (i.e., tryptophan was supplied ments shown), tryptophan was added and the time to entry into S phasedetermined (defined as the initiation of thymidine incorporation). Repli- before the tryptophan-sensitive checkpoint), commitment cate aliquots of cells were activated in tryptophan-containing medium to cell division proceeded normally.
without the 48-h preincubation period (j). Lag time in each case was T Cells Reverse Their Commitment to Cell Cycle Progression defined as the time to initiation of S phase from the point at which cells upon Entering the Arrested State. In contrast to the experi- saw both tryptophan and anti-CD3. The arrow shows that the lag time toS phase was shortened by 12–16 h due to preactivation in the absence of ments shown in Fig. 10, however, once T cells entered the tryptophan, suggesting that this portion of G1 had been accomplished be- arrested state, simply restoring tryptophan was no longer fore the point at which cells arrested. Representative of seven experi- sufficient to allow cell cycle progression. T cells were acti- ments at 36, 48, and 72 h, all showing the same lag time to S phase. (B) vated for 48 h in tryptophan-deficient medium using im- T cells were activated with anti-CD3/CD28 in the presence (j) or ab- mobilized anti-CD3/CD28. The arrested cells were then sence (d) of tryptophan. After 14 h (the time of the putative arrest pointestimated from A), tryptophan was added to the tryptophan-deficient cul- removed from contact with anti-CD3, washed free of anti- tures and entry into S phase determined. T cells rescued at hour 14 CD28, and transferred to medium containing normal levels showed no delay compared with controls.
Regulation of T Cells by IDO Published May 3, 1999 Figure 10.
phology (33–35). These IDO-expressing cells are found at normal commitment to TCR- several putative sites of immune tolerance or privilege, in- independent activation in the ab- cluding thymus, mucosa of the gut, epididymis, placenta, sence of tryptophan. T cells wereexposed to immobilized anti- and the anterior chamber of the eye (32, 33, 36, 37). This CD3/CD28 for 2–12 h in the pattern of widespread expression throughout the immune presence (light bars) or absence system is difficult to reconcile with a simple mechanism of (dark bars) of tryptophan. At the host defense. We hypothesize that IDO expression by APCs times shown, cells were removedfrom contact with anti-CD3. Af- functions to suppress undesirable T cell activation and thus ter transfer, tryptophan was helps maintain peripheral tolerance.
added to the tryptophan-defi- Two models might be proposed by which IDO could cient cultures, and all groups suppress T cells in vivo: it might catalyze the production of were continued out to hour 48.
Cells were transferred in their own conditioned medium without wash- a suppressive metabolite of tryptophan, or it could deplete ing and continued to receive anti-CD28 throughout. At hour 48, all local tryptophan below some threshold level required for groups were assayed for proliferation by thymidine incorporation. The 2-h T cell activation. In repeated experiments, we have been time point (no proliferation after transfer) is included as a control to con- unable to detect any evidence of an immunosuppressive firm that there was no carryover of anti-CD3 into the new cultures.
metabolite in coculture supernatants (Figs. 1 and 2) (4).
Furthermore, our experiments with isolated T cells imply a using the model of allogeneic pregnancy. This model was specific checkpoint in early T cell activation that is sensitive chosen because it has long been recognized as paradoxical to low concentrations of tryptophan. For these reasons, we that the maternal immune system tolerates a genetically for- favor the tryptophan depletion hypothesis.
eign fetus throughout gestation (30). IDO is known to be Implicit in this hypothesis is the assumption that cells ex- expressed in human placenta and has been reported to be pressing IDO in vivo could create a local microenviron- localized to the zone of contact between fetal-derived tis- ment in which tryptophan is low, despite the availability of sues and the maternal immune system (31). Using 1-methyl- ample tryptophan elsewhere. In this regard, it is well estab- tryptophan (described in Fig. 4) as a pharmacologic inhibi- lished that delivery of a substrate into local microenviron- tor of IDO, we have demonstrated that IDO is a required ments is sharply limited by the rate of diffusion (Kd) component of the mechanism by which the allogeneic fe- through the interstitial space (38, 39). In the face of even tus protects itself from rejection by the maternal immune normal metabolic demands, substrate concentrations rap- system and that inhibition of IDO breaks maternal toler- idly fall to undetectable levels within a few cell diameters of ance to the allogeneic fetus (15). In the same report, we also the source of delivery (39). Because the rate of tryptophan showed that pharmacologic inhibition of IDO enhances consumption by IDO-expressing Møs is orders of magni- the activation of autoreactive T cells. Thus, by two mea- tude greater than normal metabolic demands, it is plausible sures—breaking tolerance and enhancing autoreactivity— that such Møs could create local conditions of very low these data support a role for IDO in regulating T cell re- tryptophan concentrations. Although this hypothesis is now sponses in vivo.
speculative with regard to tryptophan, the phenomenon is IDO has previously been viewed primarily as a host de- well documented with regard to, for example, the local hy- fense mechanism, inhibiting proliferation of intracellular poxic state created within muscle tissue during exercise.
pathogens (6, 9–13) or cancer cell lines (14) by depriving Because tryptophan degradation by IDO is much greater them of tryptophan (for a review see reference 8). In these than consumption by metabolic demands (Fig. 3), it is likely settings, the proposed role of IDO has been to eliminate that IDO constitutes the major route of tryptophan deple- the cell's own stores of tryptophan. To our knowledge, no tion by activated Møs. However, IDO could act in combi- role for IDO in regulating the proliferation of adjacent cells nation with other pathways. Møs have a high rate of pro- has been suggested. However, both direct and indirect evi- tein synthesis, and the incorporation of free tryptophan dence indicates that IDO is widely expressed throughout into proteins could contribute to local tryptophan deple- the immune system (32, 33) and, specifically, that it is lo- tion. Indeed, the tRNA synthetase for tryptophan (the WRS calized to a subset of cells with a Mø or dendritic cell mor- gene) is unique among tRNA synthetases in that it is mas-sively induced in Mø lineage cell lines (but not lymphoidlines) by the same signals that induce IDO (40). It has been Figure 11.
T cells require TCR signaling to exit the arrested state.
proposed that this induction allows Møs to compete prefer- T cells were activated for 48 h in entially for tryptophan when the concentration of substrate is the absence of tryptophan using low. Likewise, any pathway that transported tryptophan into Møs, whether for protein synthesis, degradation by IDO, or To simulate loss of contact with incorporation into other biosynthetic pathways, would also the APC, T cells were removedfrom the immobilized anti-CD3, serve to deplete local tryptophan. Thus, IDO could act in washed, and returned to culture in concert with other catabolic pathways to render Møs an ef- medium containing 25 mm tryp- fective local "sink" for tryptophan.
tophan. Upon replating, replicate The proposed tryptophan depletion model gains support cultures received immobilizedanti-CD3, anti-CD28, or both.
from the apparent existence of a cell cycle arrest point sen- Published May 3, 1999 sitive to tryptophan concentration. Although the absence The requirement for a second signal from the TCR in of any essential nutrient is, by definition, incompatible with order to exit the arrested state is an important finding in long-term proliferation, the arrest point we describe appears light of our proposed biologic model. Under this model, more specific than simple protein starvation. First, although T cells that attempt to activate while in contact with an protein synthesis is reduced in the absence of exogenous IDO-expressing APC are inhibited by the local absence of tryptophan, it still occurs at a significant rate, presumably re- tryptophan. In theory, however, once such T cells were flecting a combination of endogenous tryptophan stores and committed to cell division, they could migrate elsewhere recycling of tryptophan from catabolism of endogenous and and complete the activation process under tryptophan- exogenous proteins (41). Yet despite ongoing protein syn- sufficient conditions. The data presented in Fig. 11 show thesis, cell cycle progression is not simply delayed but rather that once T cells have arrested, simply regaining tryptophan is completely arrested. Second, the arrest induced by tryp- is no longer sufficient to allow continued activation. De- tophan deprivation occurs at a reproducible point in the cell spite the fact that T cells would normally have become in- cycle and remains stable once entered, suggesting a regulated dependent of TCR signaling before the tryptophan-sensi- process. Taken together, these attributes suggest a specific, tive checkpoint (Fig. 10), once they enter the arrested state tryptophan-sensitive cell cycle arrest point.
they apparently reverse this commitment and reimpose It has been noted by several groups that deprivation of upon themselves a requirement for a second round of TCR certain amino acids—tryptophan in particular—exerts an signaling. From a biologic standpoint, this would mean that inhibitory effect on cell cycle progression that cannot be a T cell arrested by an IDO-expressing APC would be explained by the effect on protein synthesis (42–45). For obliged to find a second, nonsuppressive APC presenting that reason, it has been suggested that levels of these amino the same antigen in order to exit the arrested state.
acids may function as specific checkpoints regulating cell cy- What would be the fate of an arrested T cell if no such cle progression. However, the biologic significance of such supportive APC could be found? In vitro, we find that ar- amino acid–specific checkpoints and the mechanism by which rested cells undergo progressive apoptosis after several days the levels of amino acids might be manipulated in order to if not rescued by TCR engagement (4). Whether this means regulate T cell activation has remained obscure. We now that they would likewise die in vivo, enter some form of propose a system in which regulation of local tryptophan anergy, or return to a resting state remains to be deter- concentration functions as a means of communication be- mined. However, the arrested state we describe differs from tween APCs and T cells, with APCs regulating the trypto- classical anergy (54) in several interesting respects. First, the phan level via IDO and T cells responding with either acti- cells retain their responsiveness to TCR engagement (Fig.
vation or arrest, depending on the level they detect.
11). Second, costimulation via CD28 is not sufficient to As a strategy to inhibit T cell activation, arresting pro- rescue cells once they arrest. And third, arrested cells die if gression through the cell cycle is not unique to tryptophan not rescued within a relatively brief window of time. Taken metabolism. The immunosuppressive drugs mycopheno- together, these attributes suggest that T cells arrested by tryp- late, rapamycin, and leflunomide all induce a mid-G1 arrest tophan deprivation are not immediately deleted from the in activating T cells, and this is believed to account in repertoire but that they must find a permissive APC and whole or part for their immunosuppressant action (46–48).
complete the activation process if they are to survive.
Recent evidence suggests that T cells require one or more In conclusion, our hypothesis regarding the biologic rounds of cell division to acquire a variety of effector func- role of IDO-expressing APCs is that they are involved in tions (49–51), so inhibiting proliferation may also inhibit maintaining peripheral tolerance to self antigens. Our in functional activity. In our system, it is currently unknown vitro model has focused on MCSF-derived Møs as one how T cells sense the level of tryptophan and trigger cell example of immunosuppressive APCs, but dendritic cells cycle arrest. Tryptophan-sensing systems in bacteria have or other APCs that possess inducible IDO could likewise been well described (52), but comparable systems in eu- be immunosuppressive. We speculate that tryptophan ca- karyotes have not yet been identified. However, mamma- tabolism may constitute a previously unsuspected mecha- lian genes such as tryptophan oxygenase are known to be nism contributing to the regulation of peripheral T cell regulated by changes in tryptophan levels (53), so such sensing systems can be inferred to exist.
The authors thank J.-F. Tsai for expert technical assistance, T. Stoming for developing and performing theHPLC assays, and C. Rossignol and J. Bhatia for amino acid analysis. This work was supported by the National Institutes of Health (grants K08 HL03395, R21 AI44759, andR01 HL60137 to D.H. Munn) and generous support from the Carlos and Marguerite Mason Trust.
Address correspondence to David H. Munn, Medical College of Georgia, IMMAG, Room CA-2010,Augusta, GA 30912. Phone: 706-721-7141; Fax: 706-721-8732; E-mail: [email protected] Received for publication 29 April 1998 and in revised form 2 March 1999. Regulation of T Cells by IDO Published May 3, 1999 1. Fazekas de St. Groth, B. 1998. The evolution of self-toler- log of tryptophan) are competitive inhibitors for indoleamine ance: a new cell arises to meet the challenge of self-reactivity.
2,3-dioxygenase. Arch. Biochem. Biophys. 291:326–333.
Immunol. Today. 19:448–454.
17. Southan, M.D., R.J. Truscott, J.F. Jamie, L. Pelosi, M.J.
2. Banchereau, J., and R.M. Steinman. 1998. Dendritic cells Walker, H. Maeda, Y. Iwamoto, and S. Tone. 1996. Struc- and the control of immunity. Nature. 392:245–252.
tural requirements of the competitive binding site of recom- 3. Munn, D.H., and E. Armstrong. 1993. Cytokine regulation binant human indoleamine 2,3-dioxygenase. Med. Chem. of human monocyte differentiation in vitro: the tumor-cyto- toxic phenotype induced by macrophage colony-stimulating 18. Moriya, T., K. Hagio, and N. Yoneda. 1975. A facile syn- factor is developmentally regulated by interferon g. Cancer thesis of 6-chloro-d-tryptophan. Bull. Chem. Soc. Japan. 48: 4. Munn, D.H., J. Pressey, A.C. Beall, R. Hudes, and M.R. Al- 19. Munn, D.H., and N.K. Cheung. 1987. Interleukin-2 en- derson. 1996. Selective activation-induced apoptosis of pe- hancement of monoclonal antibody-mediated cellular cyto- ripheral T cells imposed by macrophages: a potential mecha- toxicity against human melanoma. Cancer Res. 47:6600– nism of antigen-specific peripheral lymphocyte deletion. J. 20. Sono, M., T. Taniguchi, Y. Watanabe, and O. Hayaishi.
5. Shimizu, T., S. Nomiyama, F. Hirata, and O. Hayaishi.
1980. Indoleamine 2,3-dioxygenase: equilibrium studies of the 1978. Indoleamine 2,3-dioxygenase: purification and some tryptophan binding to the ferric, ferrous, and co-bound en- properties. J. Biol. Chem. 253:4700–4706.
zymes. J. Biol. Chem. 255:1339–1345.
6. Carlin, J.M., E.C. Borden, P.M. Sondel, and G.I. Byrne.
21. Bloxam, D.L., and W.H. Warren. 1974. Error in the deter- 1989. Interferon-induced indoleamine 2,3-dioxygenase ac- mination of tryptophan by method of Denckla and Dewey.
tivity in human mononuclear phagocytes. J. Leukoc. Biol. 45: A revised procedure. Anal. Biochem. 60:621–625.
22. Yamamoto, S., and O. Hayaishi. 1967. Tryptophan pyrrolase 7. Werner, E.R., B. Bitterlich, D. Fuchs, A. Hausen, G. Reib- of rabbit intestine: d- and l-tryptophan-cleaving enzyme or negger, G. Szabo, M.P. Dierich, and H. Wachter. 1987. Hu- enzymes. J. Biol. Chem. 242:5260–5266.
man macrophages degrade tryptophan upon induction by in- 23. Dai, W., and S. Gupta. 1990. Molecular cloning, sequencing terferon-g. Life Sci. 41:273–280.
and expression of human interferon-g-inducible indoleamine 8. Taylor, M.W., and G. Feng. 1991. Relationship between in- 2,3-dioxygenase cDNA. Biochem. Biophys. Res. Commun. 168: terferon-g, indoleamine 2,3-dioxygenase, and tryptophan ca- tabolism. FASEB J. 5:2516–2522.
24. Munn, D.H., A.G. Bree, A.C. Beall, M.D. Kaviani, H. Sabio, 9. Pfefferkorn, E.R. 1984. Interferon g blocks the growth of R.G. Schaub, R.K. Alpaugh, L.M. Weiner, and S.J. Gold- Toxoplasma gondii in human fibroblasts by inducing the host man. 1996. Recombinant human macrophage colony-stimu- cells to degrade tryptophan. Proc. Natl. Acad. Sci. USA. 81: lating factor in nonhuman primates: Selective expansion of a CD161 monocyte subset with phenotypic similarity to pri- 10. Gupta, S.L., J.M. Carlin, P. Pyati, W. Dai, E.R. Pfefferkorn, mate natural killer cells. Blood. 88:1215–1224.
and M.J. Murphy. 1994. Antiparasitic and antiproliferative 25. Koide, Y., and A. Yoshida. 1994. The signal transduction effects of indoleamine 2,3-dioxygenase enzyme expression in mechanism responsible for gamma interferon-induced in- human fibroblasts. Infect. Immun. 62:2277–2284.
doleamine 2,3-dioxygenase gene expression. Infect. Immun. 11. Daubener, W., C. Mackenzie, and U. Hadding. 1995. Estab- lishment of T-helper type 1- and T-helper type 2-like hu- 26. Dai, W., and S.L. Gupta. 1990. Regulation of indoleamine man Toxoplasma antigen-specific T-cell clones. Immunol. 86: 2,3-dioxygenase gene expression in human fibroblasts by in- terferon-g. J. Biol. Chem. 265:19871–19877.
12. Daubener, W., C. Remscheid, S. Nockemann, K. Pilz, S.
27. Chon, S.Y., H.H. Hassanain, and S.L. Gupta. 1996. Cooper- Seghrouchni, C. Mackenzie, and U. Hadding. 1996. Anti- ative role of interferon regulatory factor 1 and p91 (STAT1) parasitic effector mechanisms in human brain tumor cells: response elements in interferon-g-inducible expression of role of interferon-g and tumor necrosis factor-a. Eur. J. Im- human indoleamine 2,3-dioxygenase gene. J. Biol. Chem. 13. Nagineni, C.N., K. Pardhasaradhi, M.C. Martins, B. Detrick, 28. Stout, R.D., and J. Suttles. 1996. The many roles of CD40 in and J.J. Hooks. 1996. Mechanisms of interferon-induced in- cell-mediated inflammatory responses. Immunol. Today. 17:487– hibition of Toxoplasma gondii replication in human retinal pigment epithelial cells. Infect. Immun. 64:4188–4196.
29. Crabtree, G.R. 1989. Contingent genetic regulatory events 14. Aune, T.M., and S.L. Pogue. 1989. Inhibition of tumor cell in T lymphocyte activation. Science. 243:355–361.
growth by interferon-g is mediated by two distinct mecha- 30. Medawar, P.B. 1953. Some immunological and endocrino- nisms dependent upon oxygen tension: induction of tryp- logical problems raised by evolution of viviparity in verte- tophan degradation and depletion of intracellular nicotin- brates. Symp. Soc. Exp. Biol. 7:320–328.
amide adenine dinucleotide. J. Clin. Invest. 84:863–875.
31. Kamimura, S., K. Eguchi, M. Yonezawa, and K. Sekiba.
15. Munn, D.H., M. Zhou, J.T. Attwood, I. Bondarev, S.J.
1991. Localization and developmental change of indoleamine Conway, B. Marshall, C. Brown, and A.L. Mellor. 1998.
2,3-dioxygenase activity in the human placenta. Acta. Med. Prevention of allogeneic fetal rejection by tryptophan catabo- lism. Science. 281:1191–1193.
32. Yoshida, R., T. Nukiwa, Y. Watanabe, M. Fujiwara, F.
16. Cady, S.G., and M. Sono. 1991. 1-methyl-d,l-tryptophan, Hirata, and O. Hayaishi. 1980. Regulation of indoleamine b-(3-benzofuranyl)-d,l-alanine (the oxygen analog of tryp- 2,3-dioxygenase activity in the small intestine and the epi- tophan), and b-[3-benzo(b)thienyl]-d,l-alanine (the sulfur ana- didymis of mice. Arch. Biochem. Biophys. 203:343–351.
Published May 3, 1999 33. Moffett, J., M. Espey, and M. Namboodiri. 1994. Antibodies tryptophan deprivation on L1210 cells in culture. Cancer Res. to quinolinic acid and the determination of its cellular distri- bution within the rat immune system. Cell Tissue Res. 278: 44. Brunner, M. 1973. Regulation of DNA synthesis by amino acid limitation. Cancer Res. 33:29–32.
34. Espey, M., Y. Tang, H. Morse, J. Moffett, and M. Namboodiri.
45. Tobey, R., and K. Ley. 1971. Isoleucine-mediated regulation 1996. Localization of quinolinic acid in the murine AIDS of genome replication in various mammalian cell lines. Cancer model of retrovirus-induced immunodeficiency: implications for neurotoxicity and dendritic cell immunopathogenesis.
46. Cherwinski, H.M., R.G. Cohn, P. Cheung, D.J. Webster, Y.-Z. Xu, J.P. Caulfield, J.M. Young, G. Nakano, and J.T.
35. Espey, M., J. Moffett, and M. Namboodiri. 1995. Temporal Ransom. 1995. The immunosuppressant leflunomide inhibits and spatial changes of quinolinic acid immunoreactivity in lymphocyte proliferation by inhibiting pyrimidine biosynthe- the immune system of lipopolysaccharide-stimulated mice. J. sis. J. Pharmacol. Exp. Ther. 275:1043–1049.
Leukoc. Biol. 57:199–206.
47. Terada, N., K. Takase, P. Papst, A.C. Nairn, and E.W. Gel- 36. Yoshida, R., Y. Urade, K. Nakata, Y. Watanabe, and O.
fand. 1995. Rapamycin inhibits ribosomal protein synthesis Hayashi. 1981. Specific induction of indoleamine 2,3-dioxy- and induces G1 prolongation in mitogen-activated T lympho- genase by bacterial lipopolysaccharide in the mouse lung.
cytes. J. Immunol. 155:3418–3426.
Arch. Biochem. Biophys. 212:629–637.
48. Laliberte, J., A. Yee, Y. Xiong, and B.S. Mitchell. 1998. Ef- 37. Malina, H.Z., and X.D. Martin. 1996. Indoleamine 2,3-dioxy- fects of guanine nucleotide depletion on cell cycle progres- genase: antioxidant enzyme in the human eye. Graefe's Arch. sion in human T lymphocytes. Blood. 91:2896–2904.
Clin. Exp. Ophthalmol. 234:457–462.
49. DeSilva, D.R., K.B. Urdahl, and M.K. Jenkins. 1991. Clonal 38. Casciari, J.J., S.V. Sotirchos, and R.M. Sutherland. 1988.
anergy is induced in vitro by T cell receptor occupancy in Glucose diffusivity in multicellular tumor spheroids. Cancer the absence of proliferation. J. Immunol. 147:3261–3267.
50. Bird, J.J., D.R. Brown, A.C. Mullen, N.H. Moskowitz, 39. Li, C.K. 1982. The glucose distribution in 9L rat brain multi- M.A. Mahowald, J.R. Sider, T.F. Gajewski, C.R. Wang, and cell tumor spheroids and its effect on cell necrosis. Cancer. 50: S.L. Reiner. 1998. Helper T cell differentiation is controlled by the cell cycle. Immunity. 9:229–237.
40. Fleckner, J., P.M. Martensen, A.B. Tolstrup, N.O. Kjeld- 51. Oehen, S., and K. Brduscha-Riem. 1998. Differentiation of gaard, and J. Justesen. 1995. Differential regulation of the hu- naive CTL to effector and memory CTL: correlation of ef- man, interferon inducible tryptophanyl-tRNA synthetase by fector function with phenotype and cell division. J. Immunol. various cytokines in cell lines. Cytokine. 7:70–77.
41. Smith, C.B., G.E. Deibler, N. Eng, K. Schmidt, and L.
52. Babitzke, P. 1997. Regulation of tryptophan biosynthesis: Sokoloff. 1988. Measurement of local cerebral protein syn- trp-ing the TRAP or how Bacillus subtilis reinvented the thesis in vivo: influence of recycling of amino acids derived wheel. Molec. Microbiol. 26:1–9.
from protein degradation. Proc. Natl. Acad. Sci. USA. 85: 53. Knox, W.E., and A.H. Mehler. 1951. The adaptive increase of the tryptophan peroxidase-oxidase system of liver. Science. 42. Dauphinais, C., and W. Waithe. 1977. PHA stimulation of human lymphocytes during amino acid deprivation: protein, 54. Schwartz, R.H. 1996. Models of T cell anergy: is there a RNA, and DNA synthesis. J. Cell. Physiol. 91:357–368.
common molecular mechanism? J. Exp. Med. 184:1–8.
43. Woolley, P.V., R.L. Dion, and V.H. Bono. 1974. Effects of Regulation of T Cells by IDO

Source: http://fs.huntingdon.edu/biology/tdudley/kylee3.pdf