Enrichment of branched chain amino acid transaminase 1 correlates with multiple biological processes and contributes to poor survival of IDH1 wild-type gliomas

Results

BCAT1 is highly expressed in GBM versus lower grade gliomas

To evaluate the general expression pattern of BCAT1 in gliomas, we compared the mRNA and protein levels of BCAT1 between different tumors, grades, histological and molecular subtypes through several online databases. The mRNA expression of BACT1 in 33 types of tumor tissues (see the extension of tumor abbreviations in Supplementary Table 1) and its paired control tissues were shown in Figure 1A, 1B. Compared to other tumors, glioblastoma (GBM) owns one of the highest levels of BCAT1 expression. Notably, the mRNA level of BCAT1 in GBM is significantly higher than that in LGG. The Human Protein Atlas and TCGA databases also confirmed that gliomas own a high expression of BCAT1 among pan-cancers (Figure 1C, 1D). Collectively, these data indicate that BCAT1 is enriched in glioma tissues and upregulated in GBMs. Furthermore, we took advantage of clinical human glioma samples to detect the protein pattern of BCAT1 and IDH1 R132H in glioma tissues (Figure 1E, 1F). The results showed that BCAT1 is increasingly expressed in glioma tissues corresponding to increasing tumor grade (Figure 1G), and compared to IDH1 wild-type diffuse glioma patients (Figure 1H), the patients that had gained IDH1 R132H mutation showed a lower percentage of tumor cells with detectable BCAT1 expression. In addition, the data from GSE16011, GSE4290, REMBRANDT, TCGA and CGGA datasets reconfirmed our previous findings and detailed that BCAT1 is preferentially expressed in classical and mesenchymal subtype and less expressed in proneural subtype glioblastoma (Supplementary Figure 1).

Figure 1. The expression pattern of BCAT1 in pan-cancer and gliomas. (A, B) The mRNA expression profiles of BCAT1 in 33 tumors and paired normal tissues were shown in bodymap and dot plot. (C) The expression pattern of BCAT1 protein in multiple cancer tissues from the Human Protein Atlas. (D) The BCAT1 expression in multiple cancer tissues from TCGA RNAseq. (E) The entire immunohistochemical staining of BCAT1 and IDH1 R132H in tissue microarray. (F) Representative images of BCAT1 and IDH1 R132H staining in three cases of glioma patients. (Case1: GBM, WHO Grade IV; Case2: Anaplastic oligodendroglioma, WHO Grade III; Case3: GBM, WHO Grade IV) (Scale: 100um) (G) Quantitative analysis of BCAT1 immunohistochemical staining in different grades of gliomas. (H) Quantitative analysis of IDH1 R132H immunohistochemical staining in IDH1 wild-type and mutant diffuse gliomas.

High BCAT1 expression is coupled with the poor survival of IDH1 wild-type gliomas

To determine the prognostic value of BCAT1 gene expression and its association with IDH1 status in glioma patients, Kaplan-Meier (K-M) survival curves with known IDH1 wildtype ratio was performed with data from the TCGA and CCGA RNA sequencing datasets. The results showed that the overall survival (OS) time of glioma patients with higher BCAT1 expression (TCGA, IDH1WT:67.8%; CGGA, IDH1WT:74.5%) is shorter than patients with lower BCAT1 expression (TCGA, IDH1WT:5.9%; CGGA, IDH1WT:14.8%) (Figure 2A, 2D). Moreover, in LGG, higher BCAT1 expression (TCGA, IDH1WT:35.7%; CGGA, IDH1WT:41.8%) is also connected with a worse overall survival time than lower BCAT1 expression (TCGA, IDH1WT:1.7%; CGGA, IDH1WT:7.1%) (Figure 2B, 2E), while in GBM, there is no significant difference between BCAT1 higher (TCGA, IDH1WT:100%; CGGA, IDH1WT:91.2%) and lower groups (TCGA, IDH1WT:86.5%; CGGA, IDH1WT:68.4%) (Figure 2C, 2F). Moreover, we found that IDH1 wild-type gliomas showed a higher expression of BCAT1 than mutant types in both TCGA and CGGA datasets (Figure 2G, 2H), and further internal comparison of IDH1 mutant gliomas showed that 1p19q codeleted IDH1 mutant gliomas owns a lower expression than non codeleted 1p19q forms with a significant difference in the CGGA dataset (Figure 2I, 2J). Those results implied that BCAT1 contributes to the poor survival of IDH1 wild-type glioma patients.

Figure 2. High BCAT1 expression represents poor survival of IDH1 wild-type glioma patients. Representative survival plots of BCAT1 in glioma patients with IDH wild-type proportion attached from TCGA (A–C) and CGGA (D–F) datasets. (G, H) Boxplots showing the distribution of BCAT1 expression in glioma patients according to IDH1 status from TCGA and CGGA datasets. (I, J) Boxplots showing the distribution of BCAT1 expression in IDH1 mutant, 1p19q codeleted glioma patients and IDH1 mutant, 1p19q non codeleted glioma patients from TCGA and CGGA datasets.

BCAT1 is an independent prognostic factor related to higher tumor malignancy in glioma patients

To further delineate the impact of BCAT1 gene expression on the patients’ treatment response, we conducted an overview of BCAT1 expression together with several genetic alterations which are known to be related with progression of glioma. In the TCGA dataset, patient groups of younger age, IDH mutation, 1p/19q deletion, ATRX mutation and MGMT methylation, are associated with lower BCAT1 expression (P < 0.05). Treatment with radiotherapy and additional pharmaceutical therapy was also more likely occurred in BCAT1 low expression patients (P < 0.05) (Table 1). The results are consistent with CGGA dataset (Table 2), in which low expression of BCAT1 is related with younger age, IDH mutation, 1p/19q codeletion and chemoradiotherapy. Besides, patients with high BCAT1 expression are also more likely to develop tumor recurrence. The result of COX regression analysis (Table 3) in the CGGA cohort indicates IDH mutation status, 1p/19q codeletion status and the expression of BCAT1 can be considered as independent clinical prognostic factors. We obtained the same result from COX regression analysis in the TCGA cohort. According to all of the above evidences, high BCAT1 expression is associated with several clinicopathological parameters and could serve as an independent prognostic predictor for glioma patients.

Table 1. The association between BCAT1 expression and clinical features in TCGA dataset.

	BCAT1 LOW	BCAT1 HIGH	Test value	P value
Age (≤60/>60)	71/6	56/21	10.105	0.001
Gender (M/F)	43/34	37/40	0.936	0.333
Grade (LGG/GBM)	76/1	59/18	17.351	0.000
IDH Mutation (Y/N)	4/73	49/28	58.257	0.000
1p/19q deletion (Y/N)	51/26	67/10	9.281	0.002
MGMT Methylation (Y/N)	8/69	34/43	22.131	0.000
ATRX Mutation (Y/N)	32/45	58/19	18.074	0.000
Radiotherapy (Y/N)	28/49	19/58	2.480	0.115
Additional pharmaceutical therapy (Y/N)	37/40	34/43	0.235	0.628

Table 2. The association between BCAT1 expression and clinical features in CGGA dataset.

	BCAT1 LOW	BCAT1 HIGH	Test value	P value
Age (≤60/>60)	235/11	211/36	14.587	0.000
Gender (M/F)	141/105	138/109	0.105	0.746
Grade			-10.205	0.000
2	97	34
3	116	67
4	33	146
IDH Mutation (Y/N)	213/33	62/185	188.893	0.000
1p19q deletion (Y/N)	96/150	9/238	92.043	0.000
PRS_type (Primary/Recurrent)	163/83	139/108	5.178	0.023
Radiotherapy (Y/N)	192/54	212/35	5.044	0.025
Chemotherapy (Y/N)	167/79	201/46	11.851	0.001

Table 3. Multivariate survival analysis of overall survival probabilities with respect to BCAT1 expression based on TCGA and CGGA datasets.

Variables	Univariate analysis			Multivariate analysis
	HR	95%CI	P value	HR	95%CI	P value
TCGA (n=620)
Age	5.731	4.248-7.732	0.000	2.658	1.912-3.697	0.000
Gender	0.889	0.673-1.173	0.405	-	-	-
IDH status	0.102	0.075-0.138	0.000	0.251	0.148-0.425	0.000
1p19q	0.248	0.158-0.390	0.000	0.482	0.265-0.875	0.016
ATRX	0.460	0.333-0.636	0.000	0.692	0.437-1.095	0.116
MGMT	0.305	0.230-0.404	0.000	1.026	0.734-1.436	0.879
BCAT1	4.506	3.283-6.184	0.000	1.626	1.104-2.396	0.014
CGGA (n=465)
Age	2.079	1.447-2.989	0.000	1.215	0.833-1.772	0.312
Gender	1.101	0.862-1.407	0.442	-	-	-
Radiotherapy	1.379	0.970-1.962	0.073	-	-	-
Chemotherapy	1.606	1.177-2.192	0.003	1.382	1.008-1.895	0.045
IDH status	0.274	0.213-0.353	0.000	0.503	0.365-0.694	0.000
1p19q	0.283	0.191-0.420	0.000	0.545	0.354-0.841	0.006
BCAT1	3.513	2.687-4.593	0.000	1.895	1.355-2.649	0.000

BCAT1 correlates with apoptosis, hypoxia, and angiogenesis processes in gliomas

To illuminate the biological features of glioma with different BCAT1 expression, we examined the correlation of BCAT1 expression with angiogenesis, apoptosis and hypoxia marker genes to evaluate the relationship between BCAT1 expression and tumor microenvironment. As shown in functional heatmap analyses, glioma-derived BCAT1 expression was positively correlated with biomarker gene expression of apoptosis, hypoxia and angiogenesis in both the TCGA and CGGA datasets (Figure 3A, 3B). Similarly, Pearson correlation analysis was performed to retest the correlation of interest gene sets with BCAT1 expression in LGG and GBM based on TCGA and CGGA datasets. After screening several apoptosis markers including CASP3, CASP7, CASP8, BCL2, RELA and BAX; hypoxia markers including HIF1A, EPAS1, ARNT, LDHA, SLC2A1 and CA9; angiogenesis markers including VEGFA, KDR, ANGPT2, BAI1, ANGPT1 and FLT1 as shown by correlograms, the results were consistent with the heatmap analysis. BCAT1 expression was positively correlated with apoptosis, hypoxia and angiogenesis markers (Figure 3C–3N). Interestingly, it seems BCAT1 is tighter associated with these genes in GBM than in LGG (Figure 3C–3N). These results suggested that BCAT1 is associated with the apoptosis, hypoxia and angiogenesis processes in gliomas, especially in GBM.

Figure 3. BCAT1 is correlated with apoptosis, hypoxia and angiogenesis processes in gliomas. (A, B) Heatmaps showing the expression patterns of angiogenesis, apoptosis and hypoxia markers in glioma patients according to BCAT1 expression based on TCGA and CGGA datasets. (C–E) The relation between BCAT1 expression and apoptosis, hypoxia and angiogenesis markers in LGG of TCGA dataset. (F–H) The relation between BCAT1 expression and apoptosis, hypoxia and angiogenesis markers in GBM of TCGA dataset. (I–K) The relation between BCAT1 and apoptosis, hypoxia and angiogenesis markers in LGG of CGGA dataset. (L–N) The relation between BCAT1 and apoptosis, hypoxia and angiogenesis markers in GBM of CGGA dataset.

High BCAT1 expression is correlated with higher glycolysis level in gliomas

The tumor metabolic phenotype is characterized by preferential dependence on glycolysis, and glycolytic enzymes are also known to be associated with the worse or better prognosis in various cancers. To evaluate the relationship between BCAT1 and glycolysis in gliomas, we firstly performed the Pearson correlation analysis using both TCGA and CGGA datasets. After screening several glycolytic enzymes including LDHA, PGAM2, TIGAR, PKM, HK2, we found that BCAT1 expression was positively correlated with glycolysis level as shown by Circos plots (Figure 4A, 4B). Moreover, IDH1 wild-type gliomas displayed a significantly increased expression of glycolytic enzyme genes (LDHA, HK2, PGAM2, PGK1, PKM, TIGAR) compared to IDH1-mutant gliomas as shown in heatmaps (Figure 4C, 4D). Finally, we evaluated the prognostic value of these glycolytic genes based on the two glioma datasets to determine their effect on glioma patients’ survival. As shown in Figure 4E–4J, higher expression of glycolytic enzyme genes predicted worse overall survival (OS) in glioma based on the TCGA dataset. Similarly, a strong correlation between higher expression of glycolytic enzyme genes and shorter patient OS was also observed in the CGGA dataset (Figure 4K–4P). These findings indicate that BCAT1 might synergize with glycolytic enzymes in the worse outcome of IDH1 wild-type gliomas.

Figure 4. BCAT1 is correlated with glycolysis process in gliomas. (A, B) The correlation of BCAT1 with glycolytic enzymes in TCGA and CGGA datasets. (C, D) Heatmap showing the mRNA expression of BCAT1 and related glycolytic enzymes in IDH1 wild-type and mutant gliomas based on TCGA and CGGA datasets. (E–J) Survival plots for BCAT1 correlated glycolytic enzymes in TCGA glioma patients. (K–P) Survival plots for BCAT1 correlated glycolytic enzymes in CGGA glioma patients.

High BCAT1 expression is accompanied by increased immunosuppressive status in the tumor

The immunological tumor microenvironment of diffuse gliomas differs in regards to IDH mutation status. IDH1 wild-type gliomas display a more immunosuppressive tumor microenvironment including a more prominent immune cell infiltration and higher PD-L1 expression [22]. To further explore the role of BCAT1 in the immunosuppressive status of gliomas, Circos plots were performed based on the correlation between BCAT1 and six immune checkpoints (PD-L1, PD-L2, B7H3, B7H4, TIM-3, Gal-9). The results showed that BCAT1 synergized well with other immune checkpoint molecules in gliomas, which were observed in both TCGA and CGGA datasets (Figure 5A, 5B). The Pearson correlation analysis further showed that BCAT1 was positively associated with M2 macrophage markers (CD163, FCGR1A, MRC2, PTPRC) and Treg markers (CCR7, CD3, CD4, IL2RA) in both TCGA (Figure 5C–5J) and CGGA (Figure 5K–5R) datasets. Then, mutual relationship analysis of BCAT1 related immune cells was further conducted to identify their clusters according to IDH status and tumor grades. The results demonstrated that M2 macrophages and Tregs were preferentially involved in IDH1 wild-type and higher grades gliomas (Figure 5S, 5T). Overall, these data suggest BCAT1 is correlated with the immunosuppressive signature of IDH1 wild-type gliomas.

Figure 5. BCAT1 is correlated with immunosuppressive status in gliomas. (A, B) The correlation between BCAT1 and immune checkpoints in the TCGA and CGGA datasets. (C–F) The correlation between BCAT1 and M2 markers in the TCGA dataset and (G–J) CGGA dataset. (K–N) The correlation between BCAT1 and Treg markers in the TCGA dataset and (O–R) CGGA dataset. (S, T) Heatmap showing the mRNA expression pattern of BCAT1 related immune cell markers in the TCGA and CGGA datasets, shown respectively for IDH1 mutant and wild-type gliomas and for glioma grades.

Author Contributions

Concept and design: LY, XGF and FY; Development of methodology: LY, XGF and JBL; Acquisition and analysis of data: LY, XGF; Writing, review and editing: LY and XGF; Review and critical comments: MN, JZ and XJY; Supervision: XJY and MN; Project administration: XJY; Funding acquisition: XJY and MN.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Funding

The work was funded by Beijing-Tianjin-Hebei Basic Research Cooperation Project (No. 19JCZDJC64200), National Natural Science Foundation of China (No. 81872063) to XJY; The Swedish Cancer Society (2019, contract 19 0316 Pj01H), The Cancer Society in Stockholm (2018-181223), The Swedish Research Council (VR-MH 2018-02452) to MN; Tianjin Research Innovation Project for Postgraduate Students (2019YJSB105) and Chinese Government Scholarship from China Scholarship Council (201906940041) to LY.

References

• 1. Ostrom QT, Gittleman H, Truitt G, Boscia A, Kruchko C, Barnholtz-Sloan JS. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2011-2015. Neuro Oncol. 2018; 20:iv1–v86. https://doi.org/10.1093/neuonc/noy131 [PubMed]
• 2. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, Belanger K, Brandes AA, Marosi C, Bogdahn U, Curschmann J, Janzer RC, Ludwin SK, et al, and European Organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups, and National Cancer Institute of Canada Clinical Trials Group. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005; 352:987–96. https://doi.org/10.1056/NEJMoa043330 [PubMed]
• 3. Buckner J, Giannini C, Eckel-Passow J, Lachance D, Parney I, Laack N, Jenkins R. Management of diffuse low-grade gliomas in adults - use of molecular diagnostics. Nat Rev Neurol. 2017; 13:340–51. https://doi.org/10.1038/nrneurol.2017.54 [PubMed]
• 4. Sulkowski PL, Corso CD, Robinson ND, Scanlon SE, Purshouse KR, Bai H, Liu Y, Sundaram RK, Hegan DC, Fons NR, Breuer GA, Song Y, Mishra-Gorur K, et al. 2-hydroxyglutarate produced by neomorphic IDH mutations suppresses homologous recombination and induces PARP inhibitor sensitivity. Sci Transl Med. 2017; 9:eaal2463. https://doi.org/10.1126/scitranslmed.aal2463 [PubMed]
• 5. Louis DN, Perry A, Reifenberger G, von Deimling A, Figarella-Branger D, Cavenee WK, Ohgaki H, Wiestler OD, Kleihues P, Ellison DW. The 2016 world health organization classification of tumors of the central nervous system: a summary. Acta Neuropathol. 2016; 131:803–20. https://doi.org/10.1007/s00401-016-1545-1 [PubMed]
• 6. Foote MB, Papadopoulos N, Diaz LA Jr. Genetic classification of gliomas: refining histopathology. Cancer Cell. 2015; 28:9–11. https://doi.org/10.1016/j.ccell.2015.06.014 [PubMed]
• 7. Eckel-Passow JE, Lachance DH, Molinaro AM, Walsh KM, Decker PA, Sicotte H, Pekmezci M, Rice T, Kosel ML, Smirnov IV, Sarkar G, Caron AA, Kollmeyer TM, et al. Glioma groups based on 1p/19q, IDH, and TERT promoter mutations in tumors. N Engl J Med. 2015; 372:2499–508. https://doi.org/10.1056/NEJMoa1407279 [PubMed]
• 8. Waitkus MS, Diplas BH, Yan H. Isocitrate dehydrogenase mutations in gliomas. Neuro Oncol. 2016; 18:16–26. https://doi.org/10.1093/neuonc/nov136 [PubMed]
• 9. Calvert AE, Chalastanis A, Wu Y, Hurley LA, Kouri FM, Bi Y, Kachman M, May JL, Bartom E, Hua Y, Mishra RK, Schiltz GE, Dubrovskyi O, et al. Cancer-associated IDH1 promotes growth and resistance to targeted therapies in the absence of mutation. Cell Rep. 2017; 19:1858–73. https://doi.org/10.1016/j.celrep.2017.05.014 [PubMed]
• 10. Zhang Z, Chan AK, Ding X, Li Y, Zhang R, Chen L, Liu Y, Wang Y, Xiong J, Ng HK, Yao Y, Zhou L. Glioma groups classified by IDH and TERT promoter mutations remain stable among primary and recurrent gliomas. Neuro Oncol. 2017; 19:1008–10. https://doi.org/10.1093/neuonc/nox042 [PubMed]
• 11. Aibaidula A, Chan AK, Shi Z, Li Y, Zhang R, Yang R, Li KK, Chung NY, Yao Y, Zhou L, Wu J, Chen H, Ng HK. Adult IDH wild-type lower-grade gliomas should be further stratified. Neuro Oncol. 2017; 19:1327–37. https://doi.org/10.1093/neuonc/nox078 [PubMed]
• 12. Cho HR, Jeon H, Park CK, Park SH, Kang KM, Choi SH. BCAT1 is a new MR imaging-related biomarker for prognosis prediction in IDH1-wildtype glioblastoma patients. Sci Rep. 2017; 7:17740. https://doi.org/10.1038/s41598-017-17062-1 [PubMed]
• 13. Lee JH, Cho YR, Kim JH, Kim J, Nam HY, Kim SW, Son J. Branched-chain amino acids sustain pancreatic cancer growth by regulating lipid metabolism. Exp Mol Med. 2019; 51:1–11. https://doi.org/10.1038/s12276-019-0350-z [PubMed]
• 14. Gu Z, Liu Y, Cai F, Patrick M, Zmajkovic J, Cao H, Zhang Y, Tasdogan A, Chen M, Qi L, Liu X, Li K, Lyu J, et al. Loss of EZH2 reprograms BCAA metabolism to drive leukemic transformation. Cancer Discov. 2019; 9:1228–47. https://doi.org/10.1158/2159-8290.CD-19-0152 [PubMed]
• 15. Thewes V, Simon R, Hlevnjak M, Schlotter M, Schroeter P, Schmidt K, Wu Y, Anzeneder T, Wang W, Windisch P, Kirchgäßner M, Melling N, Kneisel N, et al. The branched-chain amino acid transaminase 1 sustains growth of antiestrogen-resistant and ERα-negative breast cancer. Oncogene. 2017; 36:4124–34. https://doi.org/10.1038/onc.2017.32 [PubMed]
• 16. Wang Y, Zhang J, Ren S, Sun D, Huang HY, Wang H, Jin Y, Li F, Zheng C, Yang L, Deng L, Jiang Z, Jiang T, et al. Branched-chain amino acid metabolic reprogramming orchestrates drug resistance to EGFR tyrosine kinase inhibitors. Cell Rep. 2019; 28:512–25.e6. https://doi.org/10.1016/j.celrep.2019.06.026 [PubMed]
• 17. Chang IW, Wu WJ, Wang YH, Wu TF, Liang PI, He HL, Yeh BW, Li CF. BCAT1 overexpression is an indicator of poor prognosis in patients with urothelial carcinomas of the upper urinary tract and urinary bladder. Histopathology. 2016; 68:520–32. https://doi.org/10.1111/his.12778 [PubMed]
• 18. Raffel S, Falcone M, Kneisel N, Hansson J, Wang W, Lutz C, Bullinger L, Poschet G, Nonnenmacher Y, Barnert A, Bahr C, Zeisberger P, Przybylla A, et al. BCAT1 restricts αKG levels in AML stem cells leading to IDHmut-like DNA hypermethylation. Nature. 2017; 551:384–88. https://doi.org/10.1038/nature24294 [PubMed]
• 19. Tönjes M, Barbus S, Park YJ, Wang W, Schlotter M, Lindroth AM, Pleier SV, Bai AH, Karra D, Piro RM, Felsberg J, Addington A, Lemke D, et al. BCAT1 promotes cell proliferation through amino acid catabolism in gliomas carrying wild-type IDH1. Nat Med. 2013; 19:901–08. https://doi.org/10.1038/nm.3217 [PubMed]
• 20. McBrayer SK, Mayers JR, DiNatale GJ, Shi DD, Khanal J, Chakraborty AA, Sarosiek KA, Briggs KJ, Robbins AK, Sewastianik T, Shareef SJ, Olenchock BA, Parker SJ, et al. Transaminase inhibition by 2-hydroxyglutarate impairs glutamate biosynthesis and redox homeostasis in glioma. Cell. 2018; 175:101–16.e25. https://doi.org/10.1016/j.cell.2018.08.038 [PubMed]
• 21. Suh EH, Hackett EP, Wynn RM, Chuang DT, Zhang B, Luo W, Sherry AD, Park JM. In vivo assessment of increased oxidation of branched-chain amino acids in glioblastoma. Sci Rep. 2019; 9:340. https://doi.org/10.1038/s41598-018-37390-0 [PubMed]
• 22. Berghoff AS, Kiesel B, Widhalm G, Wilhelm D, Rajky O, Kurscheid S, Kresl P, Wöhrer A, Marosi C, Hegi ME, Preusser M. Correlation of immune phenotype with IDH mutation in diffuse glioma. Neuro Oncol. 2017; 19:1460–68. https://doi.org/10.1093/neuonc/nox054 [PubMed]
• 23. Genetic and Epigenetic Mechanisms Promote IDH1-Mutant Glioma Progression. Cancer Discov. 2016; 6:OF14–OF14. https://doi.org/10.1158/2159-8290.CD-RW2015-232
• 24. Ohgaki H, Burger P, Kleihues P. Definition of primary and secondary glioblastoma—response. Clin Cancer Res. 2014; 20:2013. https://doi.org/10.1158/1078-0432.CCR-14-0238 [PubMed]
• 25. Bai H, Harmancı AS, Erson-Omay EZ, Li J, Coşkun S, Simon M, Krischek B, Özduman K, Omay SB, Sorensen EA, Turcan Ş, Bakırcığlu M, Carrión-Grant G, et al. Integrated genomic characterization of IDH1-mutant glioma malignant progression. Nat Genet. 2016; 48:59–66. https://doi.org/10.1038/ng.3457 [PubMed]
• 26. IDH inhibitors target common glioma mutation. Cancer Discov. 2019; 9:992. https://doi.org/10.1158/2159-8290.CD-ND2019-007 [PubMed]
• 27. Rohle D, Popovici-Muller J, Palaskas N, Turcan S, Grommes C, Campos C, Tsoi J, Clark O, Oldrini B, Komisopoulou E, Kunii K, Pedraza A, Schalm S, et al. An inhibitor of mutant IDH1 delays growth and promotes differentiation of glioma cells. Science. 2013; 340:626–30. https://doi.org/10.1126/science.1236062 [PubMed]
• 28. Waitkus MS, Diplas BH, Yan H. Biological role and therapeutic potential of IDH mutations in cancer. Cancer Cell. 2018; 34:186–95. https://doi.org/10.1016/j.ccell.2018.04.011 [PubMed]
• 29. Chen YY, Ho HL, Lin SC, Hsu CY, Ho DM. Loss of BCAT1 expression is a sensitive marker for IDH-mutant diffuse glioma. Neurosurgery. 2019; 85:335–42. https://doi.org/10.1093/neuros/nyy338 [PubMed]
• 30. Silva LS, Poschet G, Nonnenmacher Y, Becker HM, Sapcariu S, Gaupel AC, Schlotter M, Wu Y, Kneisel N, Seiffert M, Hell R, Hiller K, Lichter P, Radlwimmer B. Branched-chain ketoacids secreted by glioblastoma cells via MCT1 modulate macrophage phenotype. EMBO Rep. 2017; 18:2172–85. https://doi.org/10.15252/embr.201744154 [PubMed]
• 31. Conway ME, Hull J, El Hindy M, Taylor SC, El Amraoui F, Paton-Thomas C, White P, Williams M, Ellis HP, Bertoni A, Radlwimmer B, Hutson SM, Kurian KM. Decreased expression of the mitochondrial BCAT protein correlates with improved patient survival in IDH-WT gliomas. Brain Pathol. 2016; 26:789–91. https://doi.org/10.1111/bpa.12385 [PubMed]
• 32. Panosyan EH, Lasky JL, Lin HJ, Lai A, Hai Y, Guo X, Quinn M, Nelson SF, Cloughesy TF, Nghiemphu PL. Clinical aggressiveness of malignant gliomas is linked to augmented metabolism of amino acids. J Neurooncol. 2016; 128:57–66. https://doi.org/10.1007/s11060-016-2073-5 [PubMed]
• 33. Togashi Y, Shitara K, Nishikawa H. Regulatory T cells in cancer immunosuppression - implications for anticancer therapy. Nat Rev Clin Oncol. 2019; 16:356–71. https://doi.org/10.1038/s41571-019-0175-7 [PubMed]
• 34. Lin C, He H, Liu H, Li R, Chen Y, Qi Y, Jiang Q, Chen L, Zhang P, Zhang H, Li H, Zhang W, Sun Y, Xu J. Tumour-associated macrophages-derived CXCL8 determines immune evasion through autonomous PD-L1 expression in gastric cancer. Gut. 2019; 68:1764–73. https://doi.org/10.1136/gutjnl-2018-316324 [PubMed]
• 35. Wen ZF, Liu H, Gao R, Zhou M, Ma J, Zhang Y, Zhao J, Chen Y, Zhang T, Huang F, Pan N, Zhang J, Fox BA, et al. Tumor cell-released autophagosomes (TRAPs) promote immunosuppression through induction of M2-like macrophages with increased expression of PD-L1. J Immunother Cancer. 2018; 6:151. https://doi.org/10.1186/s40425-018-0452-5 [PubMed]
• 36. Hartley GP, Chow L, Ammons DT, Wheat WH, Dow SW. Programmed cell death ligand 1 (PD-L1) signaling regulates macrophage proliferation and activation. Cancer Immunol Res. 2018; 6:1260–73. https://doi.org/10.1158/2326-6066.CIR-17-0537 [PubMed]
• 37. Hatzioannou A, Banos A, Sakelaropoulos T, Fedonidis C, Vidali MS, Köhne M, Händler K, Boon L, Henriques A, Koliaraki V, Georgiadis P, Zoidakis J, Termentzi A, et al. An intrinsic role of IL-33 in T_reg cell-mediated tumor immunoevasion. Nat Immunol. 2020; 21:75–85. https://doi.org/10.1038/s41590-019-0555-2 [PubMed]
• 38. Yi L, Zhou X, Li T, Liu P, Hai L, Tong L, Ma H, Tao Z, Xie Y, Zhang C, Yu S, Yang X. Notch1 signaling pathway promotes invasion, self-renewal and growth of glioma initiating cells via modulating chemokine system CXCL12/CXCR4. J Exp Clin Cancer Res. 2019; 38:339. https://doi.org/10.1186/s13046-019-1319-4 [PubMed]
• 39. Yi L, Guo G, Li J, Fan X, Li T, Tong L, Liu P, Wang X, Yuan F, Yu S, Huang Q, Yang X. IKBKE, a prognostic factor preferentially expressed in mesenchymal glioblastoma, modulates tumoral immunosuppression through the STAT3/PD-L1 pathway. Clin Transl Med. 2020; 10:e130. https://doi.org/10.1002/ctm2.130