Duplex sequencing provides detailed characterization of mutation frequencies and spectra in the bone marrow of MutaMouse males exposed to procarbazine hydrochloride

Abstract

Mutagenicity testing is an essential component of health safety assessment. Duplex Sequencing (DS), an emerging high-accuracy DNA sequencing technology, may provide substantial advantages over conventional mutagenicity assays. DS could be used to eliminate reliance on standalone reporter assays and provide mechanistic information alongside mutation frequency (MF) data. However, the performance of DS must be thoroughly assessed before it can be routinely implemented for standard testing. We used DS to study spontaneous and procarbazine (PRC)-induced mutations in the bone marrow (BM) of MutaMouse males across a panel of 20 diverse genomic targets. Mice were exposed to 0, 6.25, 12.5, or 25 mg/kg-bw/day for 28 days by oral gavage and BM sampled 42 days post-exposure. Results were compared with those obtained using the conventional lacZ viral plaque assay on the same samples. DS detected significant increases in mutation frequencies and changes to mutation spectra at all PRC doses. Low intra-group variability within DS samples allowed for detection of increases at lower doses than the lacZ assay. While the lacZ assay initially yielded a higher fold-change in mutant frequency than DS, inclusion of clonal mutations in DS mutation frequencies reduced this discrepancy. Power analyses suggested that three animals per dose group and 500 million duplex base pairs per sample is sufficient to detect a 1.5-fold increase in mutations with > 80% power. Overall, we demonstrate several advantages of DS over classical mutagenicity assays and provide data to support efforts to identify optimal study designs for the application of DS as a regulatory test.

Keywords: Duplex sequencing; Error-corrected sequencing; Genetic toxicology; Mutagenesis; Transgenic rodent assay; ecNGS.

Fig. 1

Duplex Sequencing mutation frequencies (MF) in the bone marrow of MutaMouse males at various doses of PRC. Bars represent the MF (mutations per bp) for each animal using MF_Min or MF_Max. Data labels indicate total mutation count per animal. X-axis indicates PRC dose group (mg/kg-bw/day). Asterisks indicate a significant difference (p < 0.05 relative to the control in the average MF across animals)

Fig. 2

Spontaneous and PRC-induced MF by Duplex Sequencing target ordered from highest MF_Min in the high PRC dose group to the lowest. Data are mean MF ± SEM mutations per bp, for each target (n = 6), separated by dose (mg/kg-bw/day). Intergenic targets are labelled in red and genic targets are in black. Asterisks indicate a significant increase in the high PRC dose relative to the controls. The double asterisk indicates a significant increase in the high and middle PRC doses relative to controls. (Generalized linear mixed model, p < 0.05)

Fig. 3

Fold change in MF between PRC high-dose group and VC for the 20 Duplex Sequencing targets as well as for the lacZ gene (red bar). Estimated using a general linear mixed model. DS targets are listed along the X-axis. Errors bars are SEM. Asterisks indicate the fold change was a significant increase in MF from VC for MF_Max. Double asterisks indicate significance for both MF_Max and MF_Min. The targets chr17, chr14, chr1, chr11, chr8, and chr10 were within range of the lacZ fold change

Fig. 4

Pearson’s correlation analysis between lacZ mutant frequency and Duplex Sequencing using MF_Min (A) or MF_Max (B) for MutaMouse males exposed to PRC. Data are presented as the individual MFs for each animal

Fig. 5

95% confidence intervals (CI) based on BMD model averaging analyses. A log₁₀ BMD for a 50% increase of lacZ mutant frequency and the Duplex Sequencing MF, using either MF_Min or MF_Max. B log₁₀ BMD of lacZ mutant frequency and Duplex Sequencing MF_Max for each genomic target. DS targets are organized from smallest BMD to highest. Target chr1.2 failed to produce a BMD due to poor dose response

Fig. 6

Mutation spectrum of controls and PRC dose groups measured by Duplex Sequencing in the bone marrow of MutaMouse males. A Mutation subtypes are presented as the average MF^Min ± SEM. B Mutation subtypes are represented by the proportion of total mutations. Values are mean ± SEM. Plots represent MF_Min. Mutation subtypes include the six single nucleotide variants using a pyrimidine reference, multi-nucleotide variants (mnv), small insertion (ins), and small deletion (del) mutations. Asterisks indicate a significant increase in MF from controls

Fig. 7

The power to detect a 50% increase in MF by Total Duplex Base Pairs. Results are displayed for a sample size of 6 animals per group (dark blue), 4 animals per group (blue) and 3 animals per group (light blue). A Sample variance of 0.062 (based on current study). B Sample variance of 0.1. Red line indicates the threshold for > 80% power and p < 0.05 (color figure online)