Investigating Metabolic and Immunologic Consequences of Nucleotide Substitutions in Acute Myeloid Leukemia

Background

Stressors, conventionally defined as cell extrinsic factors with the ability to promote clonal evolution, have been implicated as modulators of cancer development and progression. Smoking has previously been linked to unique single-base substitutions (SBSs), which may facilitate the development and progression of acute myeloid leukemia (AML). Similar to smoking in lung cancer, we previously reported unique C>A single-base substitutions (SBSs) in AML. SBS4 mutational signature (COSMIC database) is characterized by C>A transversions and may induce “immunogenicity”. Interestingly, C>A associates with G>A in elderly AML with previous smoking history (Patel, ASH, 2022). In recent years, high body mass index (BMI) has been associated with risk for clonal hematopoiesis (CH). However, a SBS correlation with metabolic parameters (i.e. lipid studies and BMI) remains uncharacterized. In this study, we aimed to: (1) confirm the association of smoking with C>A and G>A SBSs in an expanded cohort of leukemia patients and (2) to investigate metabolic and immunologic factors associated with these SBSs among patients with leukemia.

Methods

After IRB approval, we conducted a retrospective analysis of AML patients (pt) from Baylor and MedStar Georgetown University Hospital (MGUH). Genetic data using next generation sequencing (NGS), as well as demographic and clinical information was obtained. For patients from MGUH, FASTA nucleotide sequences were obtained by subjecting mutations to transcript expression and cDNA alignment in the NCBI gene database. MHC I binding neopeptides were identified by mapping FASTA nucleotide sequences and HLA class I alleles in the Net MHC pan 4.1 database. SBSs with neopeptides that exhibited binding were grouped for further analysis of associations. Descriptive statistics were obtained using SAS software.

Results

Our expanded database included 85 pt with AML that had NGS performed at diagnosis. This increased the number of SBSs in our database from 148 to 213. Median age was 64.8 years (y) (range, 22-89). 42/85 patients (49.4%) had a positive smoking exposure. 41/213 (19.2%) of SBSs were C>T, 14/213 (6.6%) were C>A, 12/213 (5.6%) were G>C, 68/213 (31.9%) were G>A, 12/213 (5.6%) were A>C, 8/213 (3.8%) were A>T, 25/213 (11.7%) were G>T, 13/213 (6.1%) were A>G, 4/213 (1.9%) were T>A, 6/213 (2.8%) were T>C, 8/213 (3.8%) were C>G, and 2/213 (0.9%). 4/80 (5%) of G>A transitions were in pts age <40y, 18/80 (22%) age 40-59 y, and 58/80 (73%) >60y (Fig 1A). Smoking significantly increased association with C>A transversion rate from 2.44% to 12.94% (p=0.003). C>A combined with G>A showed an increased smoking rate of 50.59% vs 29.27%(p=0.002). When compared with patients without a smoking exposure, patients with C>A or G>A mutations and a smoking exposure had significantly higher total cholesterol of 165.2mg/dL compared to 125.8mg/dL (p=0.02) and LDL of 101.8mg/dL compared to 74.5mg/dL (p=0.04). Interestingly, higher frequency of G>T, (previously recognized to be associated with oxidative damage) was observed among AML patients aged 40-59 exhibiting higher body mass index (BMI) (Fig 1B). Given dominance of G>A expression in older AML pt, we investigated MHC class I neopeptide immunogenicity initiated by G>A mutations as a suggestion of functional interference. Among pathogenic G>A variants demonstrating TCR-reactivity, 1/18 (5.6%) vs 7/21 (33.3%) neopeptides were observed in AML pt less than 60 y vs older than 60 y (p = 0.03).

Discussion

Our data confirm the association of smoking exposure with C>A and G>A mutagenicity. Subsequent analysis demonstrated a significant association with increased TC and LDL levels, indicating an altered lipidome may affect the acquisition of these mutations in AML patients with smoking exposure. We have also seen that G>A is enriched in older AML patients. In this pt subgroup, our data suggests that neopeptides derived from G>A SBS retain immunogenic potential. While this has been previously associated with aging, other etiologic associations have yet to be established. Additionally, younger obese AML pt exhibit increased susceptibility for G>T mutagenesis possibly as result of oxidative stress. Continued work is needed to elucidate the spectrum of stressors associated with mutational signatures in AML and their underlying mechanisms, as this will have important biologic implications for preventative and therapeutic interventions.