Quantitative Analysis of the Effects of MYC Overexpression on the Mutation Rate in Human Cells Using the PIG-A Gene

Abstract

Abstract 4406It has been hypothesized that the early acquisition of genomic instability is essential to account for the high number of mutations commonly identified in human cancers. Previous reports have demonstrated that MYC overexpression results in gene amplification. Given the importance of MYC in Burkitt's lymphoma, we investigated the effects of MYC on spontaneous inactivating somatic mutations, using an assay that we have developed, based on the PIG-A gene. PIG-A is X-linked, thus a single mutation can disrupt its function. PIG-A encodes an enzyme required for the biosynthesis of glycosylphosphatidylinositol (GPI), which is required for the expression of a set of membrane proteins on the cell surface (e.g., CD48, CD52, CD55, and CD59). Expansion of PIG-A mutant stem cells occurs in PNH, and it is known from this disorder that a very broad spectrum of mutations can inactive PIG-A. Using PIG-A, we have previously determined the mutation rate (μ) in EBV transformed B-lymphoblastoid cell lines (BLCLs) from normal human donors, and we have demonstrated genomic instability in cell lines derived from hematologic malignancies. Here we have measured μ in P493, a cell line derived from normal human B cells, which overexpress MYC (normalized levels 17 × higher than BLCLs). P493 can also be induced to express lower levels of MYC, resulting in a lower growth rate, using tetracycline and estradiol. To measure μ, pre-existing mutants are first eliminated from the culture by staining the cells with an antibody specific for CD59 and sorting to collect the upper 50th percentile of the population. This purified GPI(+) population is then returned to culture for 3 weeks, with careful cell counts taken to estimate the number of cell divisions (d) occurring in vitro. After expansion, the frequency (f) of mutants arising in vitro is measured by first incubating with a mixture of mouse antibodies specific for CD48, CD52, CD55, and CD59, followed by a rabbit anti-mouse PE conjugate, followed by a FITC conjugated antibody specific for a transmembrane protein (CD45 or HLA-DR). Live cells are identified based on exclusion of propidium iodide and based on FITC fluorescence; GPI(−) cells are defined as having <4% of the mean PE fluorescence as the GPI(+) population. f is calculated as # GPI(−) cells / total gated live cells. The mutation rate is then calculated as μ = f /d. In initial experiments, 6 separate cultures of P493 were analyzed, demonstrating a median μ of 15 × 10−7 mutations per cell division (range 11 – 33 × 10−7). For comparison, the positive control, a hypermutable mantle cell lymphoma cell line (HBL2), demonstrated a μ of 674 × 10−7, and five BLCLs derived from normal donors demonstrated a median μ value of 37 × 10−7 (range 11 to 57 × 10−7). In parallel experiments, P493 was also grown in the presence of tetracycline and estradiol to reduce MYC: in 5 separate experiments, P493 growing in these conditions demonstrated a median μ of 6 × 10−7. However, in 2 subsequent experiments, we observed a higher μ under “lower MYC” conditions: 81 × 10−7 (SEM ±15 × 10−7) vs 13 × 10−7 (±2.2 × 10−7) and 65 × 10−7 (±6.2 × 10−7) vs 21 × 10−7(±3.3 × 10−7). We then isolated 6 clones of P493: for 5 clones, under “high MYC” conditions, μ was similar to the previous experiments (26 × 10−7, SEM ±3.4 × 10−7), and for 3 of these clones μ went up in the “lower MYC” conditions by 5 to 9 fold. For 2 of these clones, μ was similar in the “high MYC” and “lower MYC” conditions. One of the 6 clones, surprisingly, did not adopt a lower growth rate in the presence of tetracycline and estradiol and demonstrated a reversed pattern: μ was 70 × 10−7 under “high MYC” conditions and 16 × 10−7 under “lower MYC” conditions. We next analyzed 7 cell lines derived from Burkitt's neoplasms which also overexpressed MYC (mean normalized levels ~ 3 × higher compared with normal BLCLs). Here we found a bimodal pattern, as we have previously described for primary acute lymphoblastic leukemia samples: for 4 of the cell lines μ was < 25 × 10−7, whereas 3 samples exhibited μ values comparable to the hypermutable positive control: 229 × 10−7, 562 × 10−7, and 1786 × 10−7 respectively. We conclude that overexpression of MYC is not sufficient to produce hypermutability in human B cells, based on this assay for gene-inactivating mutations. Hypermutability–as detected here–is a common feature of cell lines derived from Burkitt's neoplasms but does not seem to be necessary for its development. Disclosures:No relevant conflicts of interest to declare.