MTAP loss: a possible therapeutic approach for glioblastoma

In normal cells, MTAP metabolizes 5’-Methylthioadenosine (MTA), a by-product of polyamine biosynthesis, to produce adenine and 5-methylthioribose-1-phosphate (MTR-1-P) [13]. MTR-1-P is converted to methionine by utilizing a series of intermediate steps, and adenine is converted to adenosine 5′-monophosphate (AMP) by utilizing adenine phosphoribosyltransferase (APRT). AMP is also produced by de novo purine biosynthesis (Fig. 1). Adenosine triphosphate (ATP) is generated from AMP providing cell energy.

Hence, in MTAP-deficient cells, adenine is not produced from MTA. As a result, for AMP production, cells completely rely on de novo purine biosynthesis. Also, cells become more sensitive to inhibitors of de novo purine biosynthesis [14, 15] and methionine starvation [9]. These inhibitors include L-alanosine, 6-mercaptopurine, 6-thioguanine (6-TG). This phenomenon is used for selective killing of MTAP deleted cells. The cells are treated with MTA, followed by a high dose of purine analogs like L-alanosine or 6-TG. In normal cells, adenine blocks the conversion of the analog to its toxic nucleotide by phosphoribosylation with 5-phosphoribosyl-1-pyrophosphate (PRPP). However, in MTAP deleted cells, due to lack of adenine, PRPP is present at a high level and converts the analog to its toxic nucleotide, hence killing the tumor cell [16] (Fig. 1).

On the other hand, polyamine biosynthesis connects methionine cycle and one-carbon metabolism [17]. Decarboxylation of s-adenosyl-methionine (SAM) removes one-carbon unit from the methionine cycle. It produces polyamines and MTA. MTA carries the one-carbon unit. With the help of folate derivatives, the one-carbon unit returns back to the methionine cycle for further synthesis of methionine and SAM [18]. Thus MTAP pathway is linked to the folate metabolism pathway through the core component methionine [19, 20] and opens up the possibility of antifolate therapy in treating GBM [21].

The gene encoding MTAP is located adjacent to cyclin-dependent kinase inhibitor 2A (CDKN2A), cyclin-dependent kinase inhibitor 2B (CDKN2B), and CDKN2B-AS1 in the genomic locus 9p21.3. MTAP loss is associated with the deletion of CDKN2A in the 9p21 locus [22, 23]. MTAP deletion is prevalent in approximately 15% of all solid tumors [20]. In gliomas, deletion of the 9p21 is more frequent in grade IV glioma as compared to grade I glioma [24]. Based on the microarray analysis, Suzuki et al. have shown that MTAP is one of the most frequently deleted genes in glioblastoma [25]. Across 32 TCGA PanCancer Atlas studies, data accessed using cBioPortal [26, 27], MTAP loss is the maximum in GBM based on alteration frequency (deep deletion) and GBM has the maximum alteration frequency for deep deletion (> 40% of the cases); in terms of RNA expression data, GBM has the least median expression for MTAP. Besides GBM, MTAP loss is seen in a wide variety of cancers including mesothelioma, bladder urothelial carcinoma, and pancreatic carcinoma among others, and MTAP is also considered as a tumor suppressor gene [7, 28]. MTAP loss in GBM has its impact on multiple aspects—co-deletion of CDKN2A, prognosis, metabolism pathway, immunosuppressive profile, and cancer cell stemness [4] among others.

MTAP is frequently co-deleted with CDKN2A in a wide variety of cancers such as malignant pleural mesothelioma [29, 30], non-small cell lung cancer [30, 31], and gliomas [32, 33]. Apart from reporting MTAP as most frequently deleted gene, Suzuki et al. have also shown that MTAP and CDKN2A are co-deleted in 15 cases accounting for 50% of the total glioblastoma cases analyzed [25]. GBM, and lower grade glioma (LGG) datasets have shown that MTAP and CDKN2A are co-deleted in more than 40% of the cases in GBM; however, the co-deletion occurs in approximately 8% of the cases in LGG, which include astrocytoma, oligoastrocytoma and oligodendroglioma. GBM had the least expression for both MTAP and CDKN2A compared to LGG. Also, the expression correlation of MTAP and CDKN2A is higher in GBM compared to LGG, being almost twice as highly correlated in GBM.

Homozygous deletion of MTAP is one of the most frequent genetic alterations in GBM. In gliomas, loss of MTAP expression is associated with 9p21 locus deletion and as shown based on the data in cBioPortal it is often co-deleted with CDKN2A in many cancers. Analysis of copy number alteration in the TCGA-GBM dataset by Menezes et al. [7] has shown that there was significantly lower MTAP gene expression in the homozygous deleted group compared to the normal (p < 0.01). Also based on the glioma subtype analysis, they have shown that all the subtypes showed more than 50% MTAP gene expression loss except for the G-CIMP (Glioma-CpG Island Methylator Phenotype) subtype, and the classical subtype showed the highest frequency of loss of MTAP gene expression. It is reported that G-CIMP + subtypes usually co-occur with IDH mutation and have better prognosis [34]. According to the stratified analysis based on different grades of glioma, MTAP expression loss in the high-grade glioma subgroup was almost two-fold greater than in the lower-grade glioma subgroup, although there was no association between MTAP expression and clinicopathological features of patients such as gender and age. Interestingly and contrary to other authors, Menezes et al. have shown that reduced MTAP expression, which is a result of higher MTAP loss frequency in GBM, is associated with better prognosis in adult glioblastoma; although MTAP loss does not affect cell line proliferation, invasion, and migration as shown in their study results based on in vitro models [7]. Menezes et al. further showed that MTAP does not have strong biological importance in gliomas using in silico and in vitro models. Moreover, they have shown that MTAP neither has a clinical impact on gliomas nor does it act as a canonic tumor suppressor gene [7]. Based on the TCGA-GBM data analysis, the results of the study have shown that the loss of MTAP expression is due to its loss and not due to its promoter methylation. This is unlike the findings of another study by Hansen et al. [4] that showed an association of MTAP methylation and loss of expression although with a low coefficient of correlation. It is also to be noted that MTAP loss is not observed in some cancers like prostate cancer [18]. Additional studies may be required to explore this aspect in further detail.

Previous studies [35, 36] have shown that GBM cells utilize multiple strategies to escape the immune surveillance and create an immunosuppressive environment. Hence GBM is characterized by its immunosuppressive nature. MTAP loss results in the accumulation of MTA. The metabolite MTA is associated with cellular context-dependent mechanisms such as targeting the Akt signal pathway, downregulation of TNFα response and host inflammatory response among other associations. Hansen et al. investigated the link between MTAP loss and GBM microenvironment and has shown that loss of MTAP correlates with an immunosuppressive profile in GBM [3]. The results of the study also showed that MTAP loss correlates with differential expression of genes regulating innate or adaptive immune response in experimental cell models and in GBM samples as well. Specifically, MTAP null cells showed lower expression of HLA genes and reduced expression of inflammatory cytokines.

Also, low MTAP expression affects immunosuppressive molecular profiles and has correlations with different immune cells. These correlations manifest in several ways, including alteration in immune cell populations, such as higher M2 macrophages, decreased proportions of γδT cells, and fewer activated CD4 cells which indicates the immunosuppressive context. The study by Hansen et al. concludes with the suggestion that MTAP loss contributes to an immunosuppressive microenvironment in GBM cells. MTAP status should be an important consideration when exploring the immune states and devising immunotherapy-based approaches for GBM treatment.

Hansen et al. have shown that a deficiency in MTAP influences the DNA methylome by dysregulation of glioma cell epigenome and promotion of “stemness” in GBM cells [4]. MTAP deficiency leads to increased expression of PROM1/CD133 which results in enhanced tumorigenicity of GBM cells and also promotes glioma stem-like cell (GSC) formation. This is associated with poor prognosis in GBM patients. Although expression of CD133 is associated with poor clinical outcomes in GBM patients and PROM1 expression is required for defining stem cells giving rise to cancer, there are additional markers as well that are identified for different cellular features including cellular characteristics, origins and hierarchy. The results of the study by Hansen et al. support that MTAP loss contributes to the genesis and/or maintenance of the stem-like cancer cells and also regulates specific characteristics of these cells in GBM pathogenesis. These results provide evidence for the role of MTAP loss in GBM pathogenesis, as they have shown that MTAP loss is at the core of the two main components of GBM pathogenesis, which are aberrant DNA methylation, a key feature of cancer cells, and the GBM cell stemness. The findings of this study support and reveal a significant concept of linking aberrant metabolism leading to epigenetic alterations in tumor cells [37, 38]. Since GBM pathogenesis is promoted by MTAP loss, this provides a unique opportunity for GBM therapeutics. Hansen et al. suggest that purine deprivation-based therapy may be a unique approach for the treatment of MTAP-null GBM as they are vulnerable to purine starvation [4].

Different studies have shown varying results for MTAP loss in pediatric glioma. Based on glioma cell line data, Menezes et al. have shown that there is no loss of MTAP expression in pediatric GBM cell lines, which is in contrast to 50% of cell lines showing loss of MTAP expression in adult glioma cell lines [7]. However, Frazao et al. have shown deletion of MTAP in pediatric gliomas [32]. Frazao et al. also suggested that co-deletion of MTAP and CDKN2A may have therapeutic relevance in pediatric glioma. Further research is required to find additional details and also study the impact of MTAP loss in the context of pediatric glioma. This will facilitate targeted therapy approaches in pediatric glioma as well.

Besides glioblastoma, MTAP loss is also relevant in multiple other cancers as shown across 32 TCGA PanCancer Atlas Studies, data accessed using cBioPortal. Furthermore, MTAP loss is also significant in rare tumors such as chordoma. Chordoma is a rare, aggressive and invasive cancerous tumor of the bone involving the base of the skull, spine and sacrum [39]. Chordoma cases have poor prognosis and are a part of a major group of bone and soft tissue tumors called sarcomas. Previously published studies have shown that MTAP loss has implications on chordoma [28, 40]. Other studies have also shown the association of MTAP with different cancers such as head and neck carcinoma [41], lung cancer [42], prostate [18], colorectal [43], and breast cancer [44]. Even though MTAP is associated with several cancers, the functional relevance and the biological mechanism are yet to be fully understood [45].

There have been only a few inhibitors that are currently undergoing clinical trials in the context of MTAP. In this section, we have reviewed these specific inhibitors, and relevant clinical trials for the drugs and/or inhibitors in the context of MTAP are listed in Table 1.

Methionine adenosyltransferase II alpha (MAT2A) is expressed in most tissues and cancer cells and is a key enzyme that utilizes methionine to produce S-adenosyl methionine (SAM) in both normal and cancer cells (Fig. 1) [46]. SAM is converted back to MTA completing the cycle (Fig. 1). AG-270 is a safe, tolerable and first-in-class oral MAT2A inhibitor that reduces the proliferation of cancer cells and tumors that lack MTAP [47]. AG-270, currently undergoing a clinical trial (NCT03435250), has been shown to inhibit MAT2A thus blocking the conversion of methionine to SAM and reducing SAM production to a lower level compared to normal. Consequently, the reduction of SAM levels slows down the growth of the cancer cells. The growth of MTAP-deleted cancer cells is blocked as a result of MAT2A inhibition which lowers PRMT5-dependent mRNA splicing and prompts DNA damage [20]. Tumors that have MTAP and p16 co-deletion are sensitive to inhibition of MAT2A and that makes MAT2A an attractive lethal target for MTAP-deleted cancers.

IDE397 is another MAT2A inhibitor that is currently being evaluated in a clinical trial (NCT04794699). IDE397 is a differentiated small molecule inhibitor with great potential for MTAP-deleted cancer patients. IDE397 has demonstrated single-agent anti-tumor activity across various solid tumors including non-small cell lung cancer, gastric and bladder cancer among others.

Besides AG-270 and IDE397 which inhibit MAT2A, in the clinical trial NCT03666988, Fedoriw et al. have shown GSK3368715 as a potential type I PRMT inhibitor [48]. The deficiency of MTAP results in impaired PRMT5 activity and thus sensitizes cancer cells to GSK3368715. To inhibit tumor growth, GSK3368715 works synergistically with PRMT5 inhibitor, GSK3326595. GSK3368715 works by altering exon usage and has strong anti-cancer activity. MTAP loss results in accumulation of PRMT5 inhibitor and this correlates sensitivity to GSK3368715.

Falchook et al. have used PRT811, another PRMT5 inhibitor for treating advanced gliomas. PRT811 is a molecule which worked as a potent, selective brain penetrant PRMT5 inhibitor in animal models of brain tumors [49]. In the clinical trial (NCT04089449), Falchook et al. observed tolerance of PRT811. The inhibition of PRMT5 and anti-tumor activity was observed after multiple doses of PRT811 [50].

AMG 193 is another PRMT5 inhibitor which preferentially binds with MTA-bound state of PRMT5 which is abundant in MTAP-deficient tumors. Clinical trial NCT05094336 uses AMG 193 to check efficacy in advanced MTAP-null solid tumors [51].

TNG908 is also a PRMT5 inhibitor and is being orally administered in the clinical trial NCT05275478 to treat MTAP-deleted advanced solid tumors. TNG908 can bind with PRMT5-MTA complex leading selective inhibition of PRMT5 in MTAP-deleted tumors [52].

MRTX1719 is another inhibitor which targets PRMT5-MTA complex and inhibits PRMT5 activity in animal model of MTAP-deleted tumors [53]. It is being tested in human subjects in the clinical trial NCT05245500.

MTA together with 6-thiogunaine (6-TG) provides selective treatment for cancers with MTAP loss [54]. Therapeutic effects utilizing 6-TG may also be enhanced in MTAP-deficient tumors, when combined with other agents such as methotrexate or pralatrexate. Pemetrexed is an anti-folate drug that inhibits nucleic acid synthesis and disrupts folate-dependent metabolic processes required for cell replication. Pemetrexed also inhibits multiple other enzymes such as thymidylate synthase (TS) and glycinamide ribonucleotide formyl transferase (GARFT) among others in the folate pathway and hence has more clinical significance compared to methotrexate. Several clinical trials are on-going which use Pemetrexed for different types of cancers like chordoma (NCT03955042), and MTAP deficient urothelial cancer (NCT03744793, NCT05335941).

Clinical trials having NCT identifiers NCT00062283 and NCT00075894 use anitibiotic l-alanosine as an inhibitor for MTAP-deficient tumor cells. L-alanosine inhibits purine synthesis by blocking de novo purine synthesis pathway [55]. These trials were initiated in 2003 and 2004, however no results were posted at the time of writing this review.