Paternal exposure to benzo(a)pyrene induces genome-wide mutations in mouse offspring

Abstract

Understanding the effects of environmental exposures on germline mutation rates has been a decades-long pursuit in genetics. We used next-generation sequencing and comparative genomic hybridization arrays to investigate genome-wide mutations in the offspring of male mice exposed to benzo(a)pyrene (BaP), a common environmental pollutant. We demonstrate that offspring developing from sperm exposed during the mitotic or post-mitotic phases of spermatogenesis have significantly more de novo single nucleotide variants (1.8-fold; P < 0.01) than controls. Both phases of spermatogenesis are susceptible to the induction of heritable mutations, although mutations arising from post-fertilization events are more common after post-mitotic exposure. In addition, the mutation spectra in sperm and offspring of BaP-exposed males are consistent. Finally, we report a significant increase in transmitted copy number duplications (P = 0.001) in BaP-exposed sires. Our study demonstrates that germ cell mutagen exposures induce genome-wide mutations in the offspring that may be associated with adverse health outcomes.

Keywords: Genomics; Mutation.

Fig. 1

Experimental design. Twenty MutaMouse males received 28 daily doses of 0 or 100 mg/kg BaP. The sperm from six controls and six BaP-exposed males were collected 42 days after exposure and analyzed for the presence of lacZ mutations and microsatellite mutations, and for mutation spectrum analysis. Twelve randomly selected control and BaP-exposed males were mated with untreated C57Bl/6 females 3 and 42 days after the end of the exposure and their offspring analyzed for the presence of single nucleotide variants (SNVs) and copy-number variants (CNVs)

Fig. 2

Mutation counts and spectra in the control and BaP offspring as measured by WGS. a Each circle, triangle, and square represent one offspring from the control (n = 5 animals), post-mitotic (n = 6 animals), and mitotic (n = 6 animals) exposure groups, respectively. The dashed lines connect the embryonic and de novo mutations for the respective animals. For analytical purposes, we called the offspring above the dark blue lines (above the means of the BaP-exposed group) as “affected” by BaP, and those below “unaffected.” b Separating the BaP offspring into two groups (affected, and unaffected) shows that the additional mutations in the affected mice match the mutation fingerprint of BaP indicated with asterisks (G:C → T:A, G:C → C:G; Supplementary Data 1; Supplementary Data 2; generalized estimating equations P ≤ 0.0027). Furthermore, there is an increase in G:C → A:T mutations (shown in asterisks, generalized estimating equations P = 0.0010). In our previous analyses, we observed a decrease in proportion of this mutation type, but the absolute mutation frequency is increased^,. Overall, the mutations induced by BaP occur at guanine nucleotides as expected. Source data can be found in Supplementary Data 8

Fig. 3

Identification and characterization of de novo CNVs. a Example of a 533-kb duplication. Red diamonds show the DNA probes with higher offspring signal (elevated Cy5 at CNV locus) and the yellow diamonds show the DNA probes where offspring DNA is the same quantity as the parental DNA. The red glowing line shows the average log₂ ratio (0.61) for the mouse with a CNV while the yellow glowing lines show the average log₂ ratio for the six litter mates. The log₂ ratio of zero indicates no change in copy number. b The copy number measured by qPCR revealed that the putative CNV was a true de novo event because it was present in multiple tissues with a copy number of three, and the CNV was not present in the parents or sibling. c Mate-pair sequencing revealed that the CNV came from the sire because the paternal SNPs were present at a higher allele fraction within the CNV locus. d Mate-pair sequencing identified the breakpoint locations by mapping discordant mate-pairs against the genome (mapping at chromosome 10 in orange and 4 in pink). Split reads mapping to both chromosome locations simultaneously allowed for the breakpoint sequence to be determined. In this example, the CNV was facilitated by microhomology of two base pairs (CT)

Fig. 4

Size and types of de novo CNVs characterized in control and BaP animals. Blue squares indicate deletions and red circles indicate duplications or insertions. The red dotted line signifies that the two insertion CNVs are from the same animal and were derived from the same mutation event. Post-mitotic and mitotic refer to matings occurring 3 and 42 days after the end of BaP exposure, respectively. Source data can be found in Supplementary Data 9