Dear Editor,
Deletions of methylthioadenosine phosphorylase (MTAP) are frequent in several malignancies and lead to 5′‐deoxy‐5′‐methylthioadenosine (MTA) accumulation, competing with S‐adenosylmethionine (SAM) for binding to Protein Arginine Methyltransferase 5 (PRMT5) and enhancing tumour sensitivity to PRMT5 inhibitors. 1 , 2 Although MTAP enzyme deficiency has been documented in acute myeloid leukaemia (AML), deletions of the MTAP gene have not been identified in this haematological malignancy. Here we evaluated MTAP downregulation in AML datasets (TCGA and OHSU) 3 , 4 , 5 , 6 , 7 and its associations with clinical and molecular characteristics and patient's clinical outcome.
When comparing the MTAP expression between AML bone marrow (BM) (n = 542) and healthy BM (n = 73), different MTAP probes showed different results (MILE dataset, 8 Figure 1A–F). However, when comparing the GTEx and TCGA datasets on UCSC Xena, 9 we found higher MTAP expression in AML blood (n = 173) than normal blood samples (n = 337) (median‐log2: 4.040 vs. −0.199, p = 6.596e−197, Figure 1G). Thirteen transcripts of MTAP were differentially expressed (Figure 1H). While only one deep deletion case was identified in AML in the TCGA dataset, using Z‐score < −1 to define low expressor MTAP, we found 10.40% and 16.84% of cases have low MTAP expression in the TCGA and OHSU, respectively (Figure 1I).
FIGURE 1.
Methylthioadenosine phosphorylase (MTAP) expression in acute myeloid leukaemia (AML) and normal specimens and the distribution of MTAP z‐score in the OHSU and TCGA datasets. (A–F) MTAP expression in healthy bone marrow (HBM: n = 73) and AML bone marrow (n = 542) in the MILE dataset. Six different probes were used to measure MTAP. (G) MTAP expression in normal whole blood samples (NB: n = 337) and white blood cells in AML patients’ peripheral blood (AML: n = 173) was compared by Welch's t‐test. (H) Expression of 13 MTAP transcripts in normal whole blood samples (NB: n = 337) and white blood cells in AML patients’ peripheral blood (AML: n = 173). Protein coding transcripts are marked by arrows (ENST00000580900.5, ENST00000460874.6, ENST00000427788.2 and ENST00000577563.1). (I) Z‐score of MTAP of all available samples with RNA‐seq data in TCGA and OHSU. (**** p < .0001; *** p < .001; ** p < .01; * p < .05; except for H: * q < 0.05).
To investigate whether low expression of MTAP is associated with specific baseline clinical features in patients with AML, we compared the frequency of MTAP low expression according to diagnosis age, sex, BM blast percentage, white blood cell count, peripheral blasts percentage, cytogenetic risk and molecular risk (Tables S1 and S2). In the TCGA dataset, we found MTAP low expression to be more frequent in older patients (diagnosis age ≥ 65 years) than in younger patients (diagnosis age < 65 years) (18.87% vs. 6.67%, p = .027). MTAP expression levels were found to be lower in patients with AML M3 (n = 16) than AML M2 (n = 38) (p = .030, Figure 2A).
FIGURE 2.
(A) Expression of methylthioadenosine phosphorylase (MTAP) regarding FAB (M0–M7) in M0 (n = 16), M1 (n = 44), M2 (n = 38), M3 (n = 16), M4 (n = 34), M5 (n = 18), M6 (n = 2) and M7 (n = 3). Comparison of MTAP expression in TCGA dataset between (B) 48 patients with NPM1 mutations (MUT) and 125 patients with NPM1 wild type (WT); (C) 32 patients with DNMT3A mutations and 131 patients with DNMT3A wild type. Comparison of MTAP expression in OHSU dataset between (D) 168 patients with NPM1 mutations and 481 patients with NPM1 wild type; (E) 136 patients with DNMT3A mutations and 513 patients with DNMT3A wild type; (F) 203 patients with FLT3 mutations and 446 patients with FLT3 wild type. Dunn's test was used for calculating adjusted P values. (**** p < .0001; *** p < .001; ** p < .01; * p < .05).
We also assessed the association between MTAP low expression and AML molecular characteristics in terms of the presence of certain AML mutations (Tables S3 and S4). We compared the frequencies of FLT3, DNMT3A, NPM1, IDH2, IDH1 and TP53 mutations between low and unaltered/high MTAP patients. In OHSU, patients with MTAP low expression have a lower frequency of FLT3 mutations (17.3% vs. 32.1%, p = .036) and NPM1 mutation (3.8% vs. 30.6%, p < .001) compared with unaltered/high MTAP group. MTAP was expressed at significantly higher levels in patients with NPM1 mutation (median‐log2, TCGA: 1037 vs. 857.5, p = .0018, adjusted‐p = .0112; OHSU: 4.058 vs. 3.606, p < .001, adjusted‐p < .001), DNMT3A mutation (median‐log2, TCGA: 980.3 vs. 844.0, p = .0094, adjusted‐p = .0569; OHSU: 3.887 vs. 3.702, p = .048, adjusted‐p = .288), FLT3 mutation (median‐log2, OHSU: 3.905 vs. 3.670, p = .007, adjusted‐p = .044), compared with patients carrying wild type genes (Figure 2B–F).
Survival analyses showed that the overall survival (OS) of MTAP‐low patients was significantly shorter than that of unaltered/high MTAP patients (MTAP‐low patients vs. MTAP‐ unaltered/high patients; median‐OS [months]: TCGA: 7.5 vs. 20.5, p = .014; OHSU: 10.16 vs. 17.79, p = .02, Figure 3A,B). Additionally, the disease‐free survival (DFS) of MTAP‐low patients was significantly lower than MTAP‐unaltered/high patients in the TCGA dataset (MTAP‐low patients (n = 18) vs. MTAP‐unaltered/high patients (n = 153); median‐DFS (months): 8.200 vs. 17.00, p = .017, Figure 3C). The OHSU dataset does not contain DFS data. We also conducted survival analyses after the exclusion of patients with AML M3, and patients with no available FAB data, due to the favourable prognosis of all‐trans retinoic acid therapy. MTAP‐low patients still have worse outcomes compared with MTAP‐unaltered/high patients (TCGA: MTAP‐low patients (n = 16) vs. MTAP‐unaltered/high patients (n = 139); median OS (months): 7.200 vs. 17.40, p = 0.004, OHSU: MTAP‐low patients (n = 50) vs. MTAP‐unaltered/high patients (n = 349); 10.16 vs. 15.52, p = 0.035, Figure 3D,E). Multivariable analysis (cox‐regression model) showed that MTAP low expression is significantly associated with OS (TCGA: p = 0.026; OHSU: p = 0.031) when adjusted by age, DNMT3A‐mutation, TP53‐mutation, FLT3‐mutation (Table 1).
FIGURE 3.
Association between MTAP expression and AML clinical outcome. (A and B) Overall survival of methylthioadenosine phosphorylase (MTAP) low and MTAP unaltered/high patients in TCGA dataset and OHSU dataset. (C) Disease‐free survival of MTAP low and MTAP unaltered/high patients (n = 171) in TCGA dataset. (D and E) Overall survival of MTAP low and MTAP unaltered/high patients excluding patients with M3‐AML in the TCGA and OHSU dataset. (F–I) Expression of MTAP of 443 initial diagnosis samples, 25 remission samples, 36 relapse samples and 129 residual samples in the OHSU dataset. (J) Paired analysis of 12 samples between initial diagnosis samples and residual samples. (**** p < .0001; *** p < .001; ** p < .01; * p < .05).
TABLE 1.
Multivariable analysis.
| Variable |
|
p | |
|---|---|---|---|
| TCGA | |||
| Low MTAP expression | 2.222 | .0263 | |
| Diagnosis age ≥65 years | 4.568 | <.0001 | |
| DNMT3A mutant type | 1.535 | .1248 | |
| TP53 mutant type | 2.057 | .0396 | |
| Cytogenetic & molecular risk: Intermediate | 2.026 | .0428 | |
| Cytogenetic & molecular risk: Poor | 3.693 | .0002 | |
| FLT3 mutant type | 2.604 | .0092 | |
| OHSU | |||
| Low MTAP expression | 2.155 | .0312 | |
| Diagnosis age ≥65 years | 8.342 | <.0001 | |
| DNMT3A wild type | 0.06665 | .9469 | |
| TP53 mutant type | 6.024 | <.0001 | |
| FLT3 wild type | 1.537 | .1243 | |
When comparing MTAP expression at different disease statuses in the OHSU dataset (initial diagnosis, remission, residual and relapse), we found MTAP expression levels were significantly higher at diagnosis (n = 443) than at remission (n = 25, median‐log2: 3.855 vs. 3.073, p < .001, Figure 3F) or at residual disease (n = 129, median‐log2: 3.855 vs. 3.491, p < .001, Figure 3G). MTAP expression was significantly higher at relapse (n = 36) than at remission (n = 25) (median‐log2: 3.938 vs 3.073, p = .003, Figure 3H) and at residual disease (n = 129) (median‐log2: 3.938 vs. 3.491, p = .002, Figure 3I). Consistently, MTAP expression levels were significantly higher in initial diagnosis compared with that in patients with residual disease when comparing samples from the same patients (n = 12, p = .001, Figure 3J).
Studies using synthetic lethal screens have shown that MTAP‐deleted cells exhibit higher sensitivity to downregulation of PRMT5. MTAP‐deleted cancer cells are particularly vulnerable to further inhibition of PRMT5 by the MTA‐cooperative PRMT5 inhibitor MRTX1719. 2 This therapeutic approach selectively inhibits the PRMT5‐MTA complex in MTAP‐deficient cells.
MTA is generated through polyamine synthesis in which arginine is metabolized into ornithine and then into polyamines. Polyamine metabolism plays a key role in leukaemia stem cell survival and presents a potential therapeutic target in AML. 10 It is plausible that MTA accumulation plays an important metabolic vulnerability in AML cells, and thus MTA levels should be evaluated in patients with AML. Therapeutic strategies that are proven effective in MTAP‐deleted cancers such as PRMT5 inhibitors and MTA‐cooperative PRMT5 inhibitors should be investigated in the context of low MTA AML.
Altogether, our study reveals the association between low MTAP expression and shorter overall survival and the absence of NPM1 mutation in patients with AML. Whether this is a causative association, what underlying mechanism of unfavourable clinical outcomes in patients with low MTAP expression, and whether MTAP low expression may lead to enhanced sensitivity to PRMT5 inhibitors remain to be studied.
AUTHOR CONTRIBUTIONS
Yiyu Xiao and Qianqian Peng: data analysis, validation and visualization, writing‐original draft, reviewing‐editing and methodology. Advaith Maya Sanjeev Kumar: data curation, analyses and validation, methodology. Houda Alachkar: conceptualization, resources, supervision, funding, validation, writing original draft, project administration, writing review and editing. All authors contributed to the article and approved the submitted version.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
FUNDING INFORMATION
This study was supported by the University of Southern California Grant no. NIH‐NCI 1R01CA248381‐01A1 and National Institutes of Health (NIH) Grant no. 5P30CA014089‐45.
ETHICS STATEMENT
Ethical approval was not required for the studies involving humans because this study was conducted on a publicly available dataset. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and the institutional requirements.
Supporting information
Supporting Information
ACKNOWLEDGEMENTS
We acknowledge the funding resources including support from NIH‐NCI 1R01CA248381‐01A1, the University of Southern California, the School of Pharmacy Seed Fund, The Norris Cancer Center pilot fund, STOP Cancer pilot funding and in part by NIH grant 5P30CA014089‐45.
Yiyu Xiao and Qianqian Peng contributed equally to the work
DATA AVAILABILITY STATEMENT
All the data used in this paper are publicly available.
REFERENCES
- 1. Pegg AE, Williams‐Ashman HG. On the role of S‐adenosyl‐L‐methionine in the biosynthesis of spermidine by rat prostate. J Biol Chem. 1969;244(4):682‐693. [PubMed] [Google Scholar]
- 2. Engstrom LD, Aranda R, Waters L, et al. MRTX1719 is an MTA‐cooperative PRMT5 inhibitor that exhibits synthetic lethality in preclinical models and patients with MTAP‐deleted cancer. Cancer Discov. 2023;13(11):2412‐2431. doi: 10.1158/2159-8290.CD-23-0669 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. de Bruijn I, Kundra R, Mastrogiacomo B, et al. Analysis and visualization of longitudinal genomic and clinical data from the AACR Project GENIE Biopharma Collaborative in cBioPortal. Cancer Res. 2023;83(23):3861‐3867. Published online September 5, 2023. doi: 10.1158/0008-5472.CAN-23-0816 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Ceramil E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2(5):401‐404. doi: 10.1158/2159-8290.CD-12-0095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6(269):pl1. doi: 10.1126/scisignal.2004088 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Cancer Genome Atlas Research Network , Ley TJ, Miller C, et al. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013;368(22):2059‐2074. doi: 10.1056/NEJMoa1301689 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Bottomly D, Long N, Schultz AR, et al. Integrative analysis of drug response and clinical outcome in acute myeloid leukemia. Cancer Cell. 2022;40(8):850‐864. e9. doi: 10.1016/j.ccell.2022.07.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Haferlach T, Kohlmann A, Basso G, et al. The clinical utility of microarray‐based gene expression profiling in the diagnosis and sub‐classification of leukemia: final report on 3252 cases from the International MILE Study Group. Blood. 2008;112(11):753. doi: 10.1182/blood.V112.11.753.753 [DOI] [Google Scholar]
- 9. Goldman MJ, Craft B, Hastie M, et al. Visualizing and interpreting cancer genomics data via the Xena platform. Nat Biotechnol. 2020;38(6):675‐678. doi: 10.1038/s41587-020-0546-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Rondeau V, Culp‐Hill R, O'Brien C, et al. Abstract NG07: targeting polyamines metabolism in acute myeloid leukemia stem cells. Cancer Res. 2023;83(7_Supplement):NG07. doi: 10.1158/1538-7445.AM2023-NG07 [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information
Data Availability Statement
All the data used in this paper are publicly available.