Tryptophan metabolism in brain tumors — IDO and beyond

Abstract

Metabolism of the essential amino acid tryptophan is a key metabolic pathway that restricts antitumor immunity and is a drug development target for cancer immunotherapy. Tryptophan metabolism is active in brain tumors including gliomas and promotes a malignant phenotype and contributes to the immunosuppressive tumor microenvironment. In recent years, improved understanding of the regulation and downstream function of tryptophan metabolism has been significantly expanded beyond the initial in vitro observation that the enzyme indoleamine-2,3-dioxygenase 1 (IDO1) promotes the depletion of intracellular tryptophan. Here, we revisit the specific roles of tryptophan metabolites in regulating brain functioning and neuronal integrity as well as in the context of brain tumors. This review summarizes recent developments in identifying key regulators, as well as the cellular and molecular effects of tryptophan metabolism with a particular focus on potential therapeutic targets in glioma

Introduction

Metabolism of the essential amino acid tryptophan is increasingly recognized as an important endogenous metabolic feedback loop that contributes to the regulation of magnitude, duration and cellular composition of immune responses. Tryptophan metabolism by cancer cells and/or cancer-associated stromal cells contributes to the suppression of antitumor immune responses.

Tryptophan metabolism also mediates primary and adaptive resistance to immunotherapeutic treatment with immune checkpoint inhibitors [1••]. These observations have resulted in the development of a broad array of drugs that aim to modulate or inhibit tryptophan metabolism. The first molecular target of small molecule inhibitors was the enzyme indoleamine-2,3-dioxygenase 1 (IDO1) that catalyzes the rate-limiting conversion of tryptophan into kynurenine. Tryptophan-2,3-dioxygenase 2 (TDO2) is another drug target that catalyzes tryptophan into kynurenine that is upregulated in brain tumors and other types of cancers. Moreover, the expression of interleukin-4-induced-1 (IL4I1), an enzyme that metabolizes tryptophan into indole derivatives, is inversely associated with overall survival in patients with low grade glioma (LGG) and glioblastoma (GBM) [2••] and represents yet another therapeutic target for cancer [3]. Recognition that the immunosuppressive effects of tryptophan metabolism are mediated not only by local tryptophan depletion [4,5] but also through the activation of immunosuppressive tryptophan catabolites including kynurenine, kynurenic acid, cinnabarinic acid, indole-3-pyruvic acid, and indole-3-aldehyde binding to the transcription factor aryl hydrocarbon receptor (AHR) [2••,6••,7,8,9••,10•,11], has initiated the development of selective AHR inhibitors.

While drugs that interfere with tryptophan metabolism are moved to clinical trials for cancer immunotherapy — with initial phase III results associated with negative outcomes [12,13] — tryptophan metabolism in brain tumors has been a particular focus of research in the past for several reasons. First, brain tumors and in particular, malignant glioma, are characterized by a profoundly immunosuppressive microenvironment that diminishes immunotherapy-based therapeutic effects [14•]. Here, activation of tryptophan metabolism and downstream effector mechanisms, including AHR-mediated transcription, have been shown to be major contributors [2••,6••,9••]. Second, tryptophan metabolism in the central nervous system (CNS) is highly compartmentalized due to selective expression of pathway enzymes, cell type-specific metabolic vulnerability and selective permeability of the blood brain barrier (BBB) for tryptophan metabolites [15,16•]. Third, tryptophan metabolism has profound effects on brain function and contributes to neuropsychiatric and neurodegenerative disorders through perturbation of neuronal function [16•]. When reflecting on tryptophan metabolism in brain tumors it is imperative to recognize its specific role in the brain (Figure 1).

Figure 1.

Tryptophan metabolism in the brain.

Tryptophan (Trp) catabolites regulate neurological and immunological functions in health and disease. Top left: overview of neuropsychiatric effects of Trp catabolites. Bottom left: Blood-brain barrier and glial cells determine local transport and metabolism of Trp and Trp-catabolites. Top right: Pleiotropic enzymatic digestion of exogenous Trp by brain tumor cells. Bottom right: Mechanisms for enzymatic digestion of Trp and biological effects of downstream catabolites. Kyn: kynurenine, KynA: kynurenic acid, QA: quinolinic acid, 3-HAA: 3-hydroxy-anthranilic acid, 3-HK: 3-hydroxykynurenine, IDO1: indoleamine-2,3-dioxygenase, IL4I1: Interleukin-4-Induced-1, I3P: indole-3-pyruvic acid, TDO2: tryptophan-2,3-dioxygenase, QPRT: quinolinic acid phosphoribosyltransferase, AHR: aryl hydrocarbon receptor, NAD: nicotinamide adenine dinucleotide, WARS: tryptophanyl-tRNA synthetase.

Tryptophan metabolism in the brain and its possible implications for brain tumors

Tryptophan metabolism in the brain is separated from systemic tryptophan metabolism by the BBB. In particular, tryptophan, kynurenine and 3-hydroxykynurenine are actively transported across the BBB via the large amino acid transporter whereas kynurenic acid and quinolinic acid are not [15]. In the brain, the production of tryptophan metabolitesishighly cell type-specific. Astrocytes primarily produce the neuroprotective metabolite kynurenic acid through kynurenine amino transferase enzyme activity, while microglial cells produce quinolinic acid from 3-hydroxy-anthranilic acid (3-HAA) through 3-HAA-3,4-dioxygenase (HAAO) activity [16•]. Unbalanced tryptophan levels or the dysregulation of downstream metabolites have been implicated in a broad range of CNS disorders including depression, schizophrenia and neurodegeneration [16•,17]. These pathologies are largely driven by the neuroactivity of specific tryptophan metabolites. Quinolinic acid mediates excitotoxicity and neurodegeneration by binding of the N-methyl-d-aspartate (NMDA) receptor. Conversely, kynurenic acid is an antagonist of ionotropic glutamate receptors and has been implicated in cognitive dysfunction in schizophrenia. For this reason, inhibitors of enzymes producing kynurenic acid and quinolinic acid, respectively, are being explored for the treatment of neuropsychiatric disorders [16•]. Importantly, both kynurenic acid and quinolinic acid are produced by gliomas [2••,6••,18,19] and may be involved in the frequent comorbidities seen in brain tumor patients. For instance, the high levels of kynurenic acid may contribute to the cognitive dysfunction in brain tumor patients. The control of neuronal excitability and seizures is another effect of tryptophan metabolite exposure.

Quinolinic acid administration causes seizures in animal models [20] and is antagonized by kynurenic acid [21]. Several lines of evidence suggest a role for quinolinic acid and kynurenic acid including: i) a nonsense mutation of the enzyme aminocarboxymuconate semialdehyde decarboxylase that likely results in channeling of tryptophan metabolism towards quinolinic acid and leads to familial cortical myoclonic tremor and epilepsy [22], ii) a ketogenic diet that reduces seizure frequency in pediatric epilepsy patients [23] and increases blood levels of kynurenic acid [24], as well as iii) the anticonvulsant levetiracetam that is widely used in brain tumor patients and inhibits the release of quinolinic acid but stimulates kynurenic acid production in interferon-gamma (IFN-γ)-treated rat cortical astrocytes [25]. Hence, the differential effects of kynurenine pathway metabolites on neuroactivity are important to consider as a potential cause of brain tumor-associated neuronal dysfunction resulting in the clinical manifestation of cognitive dysfunction and/or seizures. Depression, a comorbidity that is frequently observed among brain tumor patients [26] has also been linked with altered brain tryptophan metabolism; namely a shift away from the serotonergic, and instead, toward the kynurenine pathway [27,28]. Glioma patients suffering from depression have a higher level of tryptophan uptake in the frontal cortex and thalamus than patients without depression [29], which may reflect increased tryptophan metabolism in these brain areas.

Taken together, alterations in CNS tryptophan metabolism in patients with brain tumors may contribute to the multiple comorbidities of this disease (Figure 1).

In addition, the serotonergic branch of Trp metabolism can also modulate neuropsychiatric functions as it produces diverse neuroactive substances, including serotonin and melatonin [17]. Interestingly, several links have been discovered between the KP and the serotonergic pathway: The KP enzymes IDO1/2 participate in the serotonergic pathway as they are able to degrade both serotonin and melatonin [30]. Moreover, N-acetylserotonin acts as a positive allosteric modulator of IDO1 leading to increased kynurenine production [31•]. As serotonin and other monoamines are implicated in neurogenesis and development, a potential role of serotonin is discussed in gliomagenesis [32]. However, while treatment of glioblastoma patients with selective serotonin reuptake inhibitors (SSRI) benefits from a favorable safety profile and previous study results supporting their general use in cancer, it was not shown to improve overall survival [33].

IDO1 in brain tumors

IDO1 immunohistochemistry of primary brain tumors and brain metastases has revealed an expression pattern that is largely confined to tumor-infiltrating or resident myeloid cells and tumor-associated endothelial cells [6••,34]. Although it is tempting to speculate that endothelial IDO1 expression contributes to the immunosuppressive environment in the perivascular niche, studies in normal brain endothelial cells have demonstrated that IDO1 is involved in regulating vascular tone during inflammation through the production of kynurenine [35]. TCGA analyses and animal experiments have shown that IDO1 gene expression in the brain and in primary brain tumors increases with age and is associated with poor outcomes and resistance to immunotherapy with immune checkpoint inhibitor treatment [36•]. Considering preclinical evidence [37] a phase 1 clinical trial is underway to evaluate the safety of the IDO enzyme inhibitor BMS-986205, one of five IDO inhibitors currently under clinical evaluation [38] in patients with GBM and in combination with irradiation and the programmed cell death protein-1 (PD-1) antibody nivolumab (ClinicalTrials.gov identifier NCT04047706). Taking into account the high safety profile of this treatment combination, NRG Oncology is current in the final stages of approval for translating this into a phase II/III randomized multisite trial in the United States for patients with wildtype IDH GBM (Table 1).

Table 1.

Clinical trials with IDO1 inhibitors in brain cancer

IDO1 inhibitor	Indication	Age	Combination partner	Trial ID	Development phase	Status
Indoximod	Glioblastoma multiforme, glioma, gliosarcoma, malignant brain tumor	16 years and older	Temozolomide, Bevacizumab, Stereotactic Radiation	NCT02052648	Phase I/II	Completed
Indoximod	Glioblastoma multiforme, glioma, gliosarcoma, malignant brain tumor, ependymoma, medulloblastoma, diffuse intrinsic pontine glioma, primary CNS tumor	3 years to 21 years	Temozolomide, Conformal Radiation, Cyclophosphamide, Etoposide	NCT02502708	Phase I	Completed
Indoximod	Glioblastoma, medulloblastoma, ependymoma, diffuse intrinsic pontine glioma	3 years to 21 years	Partial Radiation, Full-dose Radiation, Temozolomide, Cyclophosphamide, Etoposide, Lomustine	NCT04049669	Phase II	Recruiting
PF-06840003	Oligodendroglioma, astrocytoma, malignant glioma	18 years and older	Not annotated	NCT02764151	Phase I	Terminated, Reardon et al. [81]
BMS-986205	Glioblastoma	18 years and older	Nivolumab, Radiation Therapy, Temozolomide	NCT04047706	Phase I	Recruiting
Epacadostat	Glioma, Glioblastoma	18 years and older	INCMGA00012, Bevacizumab, Radiation Therapy	NCT03532295	Phase II	Recruiting
INCBO24360 (Epacadostat)	Glioblastoma	18 years and older	Nivolumab	NCT03707457	Phase I	Terminated

TDO2 in brain tumors

For a long period of time, TDO2 was thought to be specifically expressed in the liver to regulate systemic tryptophan levels and energy homeostasis through the provision of nicotinamide adenine dinucleotide (NAD). However, it is now recognized that TDO2 is upregulated in several types of cancer including brain tumors [6••,39] and is inversely associated with survival outcomes in patients with glioma [6••,40,41] and meningioma [42,43]. Constitutive TDO2 expression in GBM is driven by a CCAAT/enhancer-binding protein β (C/EBPb)-binding site in the TDO2 promoter and sustained by interleukin-1b via C/EBPb in a mitogen-activated protein kinase (MAPK)-dependent fashion [6••,41]. Prostaglandins also upregulate TDO2 expression and enzyme activity in malignant glioma via activation of prostaglandin E receptor-4 (EP4) [40]. In contrast and unlike in liver cells, steroids limit TDO2 expression in glioma [44]. Moreover, hypoxia suppresses TDO2 expression both in the liver and in glioblastoma [45,46]. After discovering that TDO2 and IDO1 are both expressed by cancer cells and co-contribute to tumor-induced immunosuppression, developmental drug research has aimed to identify specific TDO2 and dual IDO1/TDO2 inhibitors for the application to patients with brain tumors. Similar to the inhibitors for IDO1, TDO2 inhibitors also increase the antitumor efficacy of immune checkpoint inhibitors in preclinical tumor models [47•]. Selective TDO2 and IDO1/TDO2 dual inhibitors have been developed [48]. However, the dual IDO1/TDO2 inhibitor M4112 that inhibits IDO1 ex vivo, does not significantly reduce plasma kynurenine levels [49].

Tryptophan metabolism and protein synthesis

IDO1-expressing and TDO2-expressing brain tumors have a high demand for tryptophan uptake to compensate for the IDO1/TDO2-mediated tryptophan consumption. Positron emission tomography (PET) imaging with α-[¹¹C]-methyl-l-tryptophan (AMT) or new tracers containing fluorine-18 demonstrate increased tryptophan uptake and/or metabolism in patients’ brain tumors [29,50]. Importantly, this imaging strategy may guide the clinical evaluation of drugs targeting tryptophan metabolism [51]. If the demand cannot be met by increased import, depletion of intracellular tryptophan results in incapacitation of protein synthesis through activation of the integrated stress response via general control non-derepressible-2 (GCN2) kinase, leading to phosphorylation of the eukaryotic translation initiation factor 2α (eIF2α) and induction of activating transcription factor 4 (ATF4). While GCN2 activation was initially implicated in the proliferation arrest of CD8⁺ T cells induced by IDO1-positive dendritic cells (DCs) [52], recent evidence in experimental glioma models suggests that GCN2 is required for CD8⁺ T cell survival under conditions of tryptophan deficiency [53]. Particularly in the brain, however, tryptophan levels are rapidly replenished through diet, thus limiting the physiological relevance of in vitro studies. In addition, free tryptophan is loaded on tRNA by tryptophanyl-tRNA synthetase (WARS) and tryptophan-dependent protein synthesis is maintained by upregulation of WARS via IFN-γ and/or GCN2-p-eIF2α-ATF4 signaling in glioblastoma [54]. Moreover, IFN- γ-induced tryptophan depletion has recently been demonstrated to lead to ribosomal frame shifting in melanoma, resulting in the generation of aberrant peptides that are presented on HLA-I molecules and can prime T cells, thereby possibly enhancing anti-tumor immunity [55••] (Figure 2).

Figure 2.

Effects of tryptophan deprivation.

Low intracellular (ic) tryptophan (Trp) levels can mediate pro-tumorigenic (blue background) and anti-tumorigenic effects (green background). Mechanisms described in glioma models are highlighted in black letters, mechanisms described in other systems are highlighted in gray. Abbreviations: GCN2: general control non-derepressible-2, p-eIF2α: phosphorylated eukaryotic translation initiation factor 2α, ATF4: activating transcription factor 4, WARS: tryptophanyl-tRNA synthetase, IFN-γ: interferon-gamma.

Tryptophan as a precursor for NAD

If all enzymes of the kynurenine pathway are present, tryptophan is ultimately metabolized into the energetic substrate NAD. NAD is an important cofactor for enzymes such as the DNA repair proteins poly(ADP-ribose) polymerases (PARPs). In brain tumors, quinolinic acid, a downstream tryptophan metabolite, is converted into NAD by glioma cells through the expression and activity of quinolinic acid phosphoribosyltransferase (QPRT). The quinolinic acid-producing enzyme HAAO is confined to tumor-infiltrating microglia indicating that gliomas exploit microglial-derived quinolinic acid as an alternative source of replenishing intracellular NAD pools as a resistance mechanism against DNA-damaging treatments such as alkylating chemotherapy and irradiation [19,56]. Interestingly, upregulation of QPRT and ensuing NAD production also mediates resistance to combined treatment with the histone deacetylase inhibitor panobinostat and the proteasome inhibitor bortezomib in glioma patients [57], further underscoring how tryptophan catabolism can contribute to higher NAD levels in glioma (Figure 3). Higher NAD levels could also contribute to a higher activity of one of the Trp-degrading enzymes, IDO1. The reduced form of NAD, NADH, has been shown to react with ferric IDO1 under aerobic conditions, fostering the formation of the active, ferrous dioxygen adduct form of IDO1 thus increasing its enzymatic activity [58]. This reaction in which IDO1 acts as an NADH oxidase links the downstream Trp-metabolic product NADH to an initial enzyme of the KP in a feed-forward loop.

Figure 3.

Tryptophan metabolism in glioma yields NAD that can mediate therapy resistance.

Collaborative degradation of tryptophan (Trp) in glioma by microglia and glioma cells yields nicotinamide adenine dinucleotide (NAD) that can mediate therapy resistance in glioma. Therapeutic agents induce quinolinic acid phosphoribosyltransferase (QPRT), NAD and poly(ADP-ribose) polymerases (PARPs) that consecutively mediate resistance to therapy, for example, the histone deacetylase inhibitor panobinostat and the proteasome inhibitor bortezomib. Abbreviations: KP: kynurenine pathway, 3-HAA: 3-hydroxy-anthranilic acid, HAAO: 3-HAA-3,4-dioxygenase, QA: quinolinic acid, RT: radiotherapy, TMZ: temozolomide.

AHR in brain tumors

There is increasing evidence that the immunosuppressive effects of tryptophan metabolism in infection, auto-immunity and tumor immunity are mediated by activating the AHR. The AHR is a cytosolic transcription factor that is highly expressed at environmental interfaces such as the skin and gut [59,60] and is responsible for the detoxification of polyaromatic hydrocarbons. Ligand binding causes AHR translocation from the cytoplasm into the nucleus resulting in heterodimerization with the AHR nuclear translocator (ARNT) [61]. The transcriptional complex binds dioxin or xenobiotic response elements (DRE or XRE) in the promoter regions of AHR target genes. In addition to the genes that encode detoxifying enzymes including cytochrome p450 oxidases, several additional AHR targets include immuno-regulatory genes. In brain tumors and other types of cancer, both AHR and AHR target genes are upregulated and correlate with the expression of IDO1 and TDO2 indicating that the production of kynurenine and its downstream metabolites are responsible for AHR activation [2••,6••]. More recently, IL4I1 was identified as an additional AHR-activating enzyme in cancer including gliomas. IL4I1 produces the AHR-activating ligands indole-3-pyruvic acid, kynurenic acid and indole-3-aldehyde and suppresses antitumor immune response [2••]. Being AHR target genes themselves, the three tryptophan-catabolic enzymes IDO1, TDO2 and IL4I1 are linked to AHR in a positive feedback loop [2••,62•,63] such that AHR induces or upregulates its own ligands.

In glioblastoma cells, AHR promotes motility mediated by kynurenine and indole-3-pyruvic acid as well as clonogenic survival [2••,6••]. AHR activation enhances transforming growth factor beta signaling in glioma cells [64], which may in part underlie the pro-migratory and immunosuppressive effects of AHR. Aside from its direct effects on cancer cells, AHR also influences diverse immune cell functions. AHR functions as a master regulator [65••] that controls differentiation and effector functions of T cells [66,67], macrophages [10•,68], DCs [68], natural killer cells [69] and B cells [70–72]. In glioblastoma, AHR contributes to the regulation of tumor-associated macrophage (TAM) recruitment and functioning [9••]. In these cells, AHR activation enhances the expression of C-C motif chemokine receptor 2 (CCR2), the receptor for glioma cell-derived C-C motif chemokine ligand 2 (CCL2) and upregulates the ectonucleotidase CD39. Together with CD73 [73], CD39 generates adenosine that causes CD8⁺ T cell dysfunction [9••]. Also, immune checkpoint molecules such as programmed cell death protein–ligand 1 (PD-L1) on TAMs [9••] and PD-1 on T cells [8] are induced by AHR activity.

Moreover, AHR activation reduces the viability of CD8⁺ T cells [74]. AHR has been implicated in the differentiation of naïve CD4⁺FoxP3⁻ T cells into inducible CD4⁺FoxP3⁺ regulatory T cells (Tregs) [75,76] and is a potent transcriptional driver of interleukin (IL)-10 expression in diverse immune cell subsets [70,77–79]. In DCs, AHR dampens the induction of pro-inflammatory cytokines resulting in tolerogenic DCs that foster Treg differentiation [76]. Moreover, AHR reduces MHC class II expression and limits CD4⁺ T cell priming [10•]. Depending on its duration [66], AHR activation drives either T helper (TH) 17 cell or Treg differentiation [80]. Through its effects on tumor cells as well as select other immune cell subsets, AHR activation may importantly contribute to immune escape and malignancy in brain cancer (Figure 4).

Figure 4.

Tryptophan catabolism promotes tumor cell malignancy and immune suppression via activation of the aryl hydrocarbon receptor (AHR).

The tryptophan (Trp)-catabolic enzymes (TCEs) indoleamine-2,3-dioxygenase 1 (IDO1) and tryptophan-2,3-dioxygenase 2 (TDO2) initiate Trp degradation to the AHR-activating metabolites kynurenine (Kyn), kynurenic acid (KynA) and cinnabarinic acid (CA), while interleukin-4-induced-1 (IL4I1) generates the AHR agonists indole-3-pyruvic acid (I3P), indole-3-aldehyde (I3A) and KynA. Upon ligand binding the AHR translocates to the nucleus and binds to AHR nuclear translocator (ARNT). The heterodimer binds to xenobiotic response elements (XRE) in the promoters of AHR target genes increasing their expression. AHR activation enhances glioma cell malignancy. Furthermore, several immune cell subsets, including CD8⁺ T cells, regulatory T cells (Treg), Type 1 regulatory T cells (Tr1), regulatory B cells (Breg), natural killer (NK) cells and tumor-associated macrophages (TAM) are influenced by AHR activation. TAMs express the ectonucleotidases CD39 and CD73 that collaboratively produce adenosine which limit T cell function. Abbreviations: Interleukin (IL)-10, Programmed cell death protein-1 (PD-1), PD-ligand 1 (PD-L1), C-C motif chemokine receptor 2 (CCR2), C-C motif chemokine ligand 2 (CCL2).

Conclusion

Tryptophan metabolism is active in brain tumors and promotes their malignant phenotype through mechanisms beyond IDO1-mediated tryptophan degradation. Therapeutic development of strategies interfering with tryptophan metabolism in brain tumors needs to take into account the specific regulation and function of tryptophan metabolites in controlling neuronal function and integrity. The identification of novel key regulators and downstream mediators of tryptophan metabolism as well as repurposing of drugs targeting tryptophan metabolism in neuropsychiatric disorders offers novel targets and opportunities for therapeutic intervention.