Mlh1 heterozygosity and promoter methylation associates with microsatellite instability in mouse sperm

Abstract DNA mismatch repair (MMR) proteins play an important role in maintaining genome stability, both in somatic and in germline cells. Loss of MLH1, a central MMR protein, leads to infertility and to microsatellite instability (MSI) in spermatocytes, however, the effect of Mlh1 heterozygosity on germline genome stability remains unexplored. To test the effect of Mlh1 heterozygosity on MSI in mature sperm, we combined mouse genetics with single-molecule PCR that detects allelic changes at unstable microsatellites. We discovered 4.5% and 5.9% MSI in sperm of 4- and 12-month-old Mlh1+/− mice, respectively, and that Mlh1 promoter methylation in Mlh1+/− sperm correlated with higher MSI. No such elevated MSI was seen in non-proliferating somatic cells. Additionally, we show contrasting dynamics of deletions versus insertions at unstable microsatellites (mononucleotide repeats) in sperm.

Spermatogenesis involves high levels of cell proliferation. Spermatogonial stem cells either self-renew or undergo 9-11 mitotic divisions to produce spermatocytes, which subsequently undergo meiotic cell divisions to produce haploid sperm cells (male gametes) (11). Hence, a mature sperm cell is a product of numerous rounds of DNA replication.
DNA replication is inherently mutagenic. In eukaryotes, during each round of replication, DNA polymerases α, δ and ε make on average less than 1 × 10 −5 replication errors per nucleotide (12)(13)(14). The intrinsic proofreading activity of DNA polymerase corrects most of these errors, and post-replication, nearly all remaining errors are repaired by MMR.
An average of 1.8 × 10 −10 mutations per nucleotide is introduced into the mouse genome during every cell division (15). Short tandem repeat sequences in the genome called microsatellites are particularly prone to replication errors, by a process known as polymerase "slippage". DNA polymerase often erroneously inserts or bypasses individual repeat units at microsatellites, resulting in small indel loops between the parental DNA strand and the newly replicated daughter strand. If left unrepaired, indel loops give rise to mutant alleles of novel microsatellite repeat array lengths. This molecular phenotype is known as microsatellite instability (MSI). During spermatogenesis, several rounds of DNA replication take place before cells enter the meiosis. It follows that any MSI detected in sperm is likely pre-meiotic, A c c e p t e d M a n u s c r i p t 4 that is, it originates in spermatogonia which is the only cell type in the testis to undergo extensive cell proliferation.
Here, we investigate how Mlh1 heterozygosity affects MSI in sperm cells, and also assess spleen MSI to obtain a germline versus somatic MSI comparison. Further, by assaying Mlh1 promoter methylation status in Mlh1 +/sperm, we establish a correlation between MSI and Mlh1 promoter methylation in the germline. In addition, we establish estimates of the contribution of insertions and deletions to sperm MSI.

Mice, genotyping and tissue collection
The mice used in this study were A c c e p t e d M a n u s c r i p t 6 Further homogenization was done using a 20G needle and syringe. Thereafter, column-based DNA extraction was performed according to the manufacturer's instruction. A c c e p t e d M a n u s c r i p t 8

2.
For an allele to be considered as mutant, both the highest peak and the stutter peaks should shift as a single unit. Shift of the highest peak alone was not scored as a mutant.

3.
If a wildtype and (apparently) mutant allele co-occurred in a single PCR, the reaction was scored as wildtype. Non-wildtype peaks were presumed to result from replication slippage during the early rounds of PCR, and thus considered artifacts.
For each of the three microsatellite loci assayed, MSI was separately scored for insertions and deletions. MSI rate was calculated as follows: MSI % = (total no. of single repeat-unit shifts observed/total DNA molecules analyzed) ×

Mlh1 promoter methylation analysis by methylation-specific PCR (MSP)
Methylation status of the Mlh1 promoter in sperm cells and splenic cells were tested using MSP assay (35

Statistical analysis
Unpaired t-test was used to test the differences in MSI rates between the groups. Two-tailed P values < 0.05 were considered to be statistically significant.

Mlh1 +/sperm cells display MSI at mononucleotide repeats
We assayed sperm MSI by single-molecule PCR to investigate the effects of Mlh1 heterozygosity on germline microsatellite stability. We assessed sperm DNA of 4-and 12month-old Mlh1 +/mice. Sperm DNA from age-matched Mlh1 +/+ littermates (or from closely related matings) was used as controls. We tested MSI at three microsatellites: two mononucleotide repeats A27 and A33 ( Figure 1A) and one dinucleotide repeat D14Mit15 (Supplementary figures 1-2). MSI was scored separately for insertions and deletions.
While the dinucleotide D14Mit15 repeat was stable (Supplementary figure 1), both mononucleotide repeats displayed MSI in Mlh1 +/sperm. Mlh1 +/sperm showed substantially more 1-bp deletions than age-matched wildtype sperm (which also showed low levels of 1-bp deletions), at both 4-and 12-month time points (Figure 1B). Both Mlh1 +/+ and Mlh1 +/-A c c e p t e d M a n u s c r i p t 10 sperm showed an increase in deletions with age ( Figure 1C). This increase was significant  Figure 1C). One 4-month old Mlh1 +/mouse (indicated by an arrow in Figure 1C) showed higher deletions (8%) in sperm compared to other Mlh1 +/mice in the same age group. With Grubbs' test this mouse was categorized as an outlier (P < 0.05) and was omitted from statistical analyses.
In Mlh1 +/+ sperm, insertions were more common than deletions at both time points (P= 0.002 for insertions versus deletions comparison for both 4-and 12-month time point).
Compared to age-matched Mlh1 +/+ mice, sperm from Mlh1 +/mice showed fewer insertions at both time points (Figure 1C). Insertions were predominantly single repeat unit (i.e. 1-bp) in size (Figure 1B), and there was no considerable change in insertion% with age in Mlh1 Figure 1C).

Mlh1 promoter methylation is frequent in sperm of Mlh1 +/mice and contributes to MSI
We used MSP to test Mlh1 promoter methylation status in Mlh1 +/sperm, and to investigate whether germline Mlh1 promoter methylation correlates with germline MSI. The Mlh1 promoter in a given sample was scored as methylated if an amplification product (143 bp in size) was detected by PCR with methylation-specific primers. A representative gel image of the MSP assay is shown in Figure 2A.
None of Mlh1 +/+ mice assayed showed Mlh1 promoter methylation in sperm (Figure   2A). In Mlh1 +/mice, Mlh1 promoter methylation was detected in 67% (4 out of 6), and 83% A c c e p t e d M a n u s c r i p t 11 (5 out of 6) sperm DNA samples at 4-and 12-month time points, respectively (Figure 2A).
Mlh1 promoter methylation was associated with elevated deletions (Figure 2B), but not insertions (Supplementary figure 3) in sperm.
MSP was also performed in spleen. All Mlh1 +/mice with Mlh1 promoter methylation in sperm displayed Mlh1 promoter methylation in spleen, while those without promoter methylation in sperm did not (Supplementary figure 4). Mlh1 promoter methylation was not observed in spleen of Mlh1 +/+ mice.

MSI is higher in Mlh1 +/sperm than in Mlh1 +/spleen
We also performed the SM-PCR based MSI assay in spleen, which enabled us to compare germline versus somatic MSI for each mouse. As in sperm, the dinucleotide locus D14Mit15 was also stable in spleen (Supplementary figure 5), and therefore tissue-specific MSI was compared only for mononucleotide repeats. Both sperm and spleen of wildtype mice showed only baseline levels of deletions at mononucleotide microsatellites. In Mlh1 heterozygotes, the increase in deletions was near-exclusive to sperm, the exception being spleen DNA in the outlier mouse (Figure 3). Deletions in Mlh1 +/sperm were significantly higher than in spleen (2.5-and 3.2-fold at 4-and 12-month time points, respectively, Figure 3; p-values= 0.0087 and 0.0052).
In wildtype mice, both sperm and spleen DNA at both time points had an insertional burden (Figure 3). The substantial decrease in insertions, seen in Mlh1 +/sperm compared to Mlh1 +/+ sperm, was not observed in Mlh1 +/spleen (Figure 3).
A c c e p t e d M a n u s c r i p t

Discussion
There is emerging evidence that phenotypically normal somatic tissues from individuals with inherited MMR heterozygosity display MSI (37,38), raising the question whether germline cells of such individuals also exhibit MSI. Further, Mlh1 promoter methylation in sperm of LS patients has been reported (21,39,40)