Comprehensive analysis of indels in whole-genome microsatellite regions and microsatellite instability across 21 cancer types

Abstract

Microsatellites are repeats of 1- to 6-bp units, and approximately 10 million microsatellites have been identified across the human genome. Microsatellites are vulnerable to DNA mismatch errors and have thus been used to detect cancers with mismatch repair deficiency. To reveal the mutational landscape of microsatellite repeat regions at the genome level, we analyzed approximately 20.1 billion microsatellites in 2717 whole genomes of pan-cancer samples across 21 tissue types. First, we developed a new insertion and deletion caller (MIMcall) that takes into consideration the error patterns of different types of microsatellites. Among the 2717 pan-cancer samples, our analysis identified 31 samples, including colorectal, uterus, and stomach cancers, with a higher proportion of mutated microsatellite (≥0.03), which we defined as microsatellite instability (MSI) cancers of genome-wide level. Next, we found 20 highly mutated microsatellites that can be used to detect MSI cancers with high sensitivity. Third, we found that replication timing and DNA shape were significantly associated with mutation rates of microsatellites. Last, analysis of mutations in mismatch repair genes showed that somatic SNVs and short indels had larger functional impacts than germline mutations and structural variations. Our analysis provides a comprehensive picture of mutations in the microsatellite regions and reveals possible causes of mutations, as well as provides a useful marker set for MSI detection.

Figure 1.

Pattern of somatic indels in microsatellite regions. (A) Mutation rate of each microsatellite; 7,650,128 microsatellites were sorted by the proportion of mutated samples. The red box indicates informative microsatellites defined in this study (proportion of mutated samples 0.001 or more). (B) Mutation rate among different microsatellites. Microsatellites are classified based on sequence of unit and length of each microsatellite in the reference genome. (C) Comparison of mutation rate of microsatellites and number of somatic SNVs and indels in different types of cancer. MSI samples are shown in red. (D,E) Pattern of insertions (positive change in repeat length in x-axis) and deletions (negative change in repeat length in x-axis) in microsatellites of the MSI (D) and MSS (E) samples. Correlation between the mutation rate of microsatellites and the number of somatic SNVs. Pearson product-moment correlation; r = 0.19, P-value = 1.6 × 10⁻²³. MSI samples are shown in red. (F) Correlation between the mutation rate of microsatellites and the number of somatic SNVs. Pearson product-moment correlation; r = 0.19, P-value = 1.6 × 10⁻²³. MSI samples are shown in red. (G) Correlation between the mutation rate of microsatellites and the number of somatic indels. Pearson product-moment correlation; r = 0.97, P-value < 1.0 × 10⁻²⁰⁰. MSI samples are shown in red.

Figure 2.

Analysis of mutation rate of each microsatellite. The mutation rates of 198,578 informative microsatellites were analyzed. (A–D) Association of replication timing with insertion and deletion rates. The edges of the boxes represent the 25th and 75th percentile values. The whiskers represent the most extreme data points, which are no more than 1.5 times the interquartile range from the boxes. (A) Insertions in MSS samples (χ² test; P-value < 1 × 10⁻²⁰⁰); (B) insertions in MSI samples (χ² test; P-value < 1 × 10⁻²⁰⁰); (C) deletions in MSS samples (χ² test; P-value < 1 × 10⁻²⁰⁰); (D) deletions in MSI samples (χ² test; P-value < 1 × 10⁻²⁰⁰). (E–H) Association of DNA shape with mutation rate. The top 1000 A/T microsatellites with 10- to 30-bp length were used for the analysis. The microsatellites were divided into the three categories based on the mutation rate. (E) Helix twist (HelT) of insertions in MSS samples; (F) the ·OH Radical Cleavage Intensity (ORChID2) of insertions in MSS samples; (G) Propeller Twist (ProT) of deletions in MSI samples; (H) slide of deletions in MSI samples. In this figure, we divided the microsatellites with mutation rates (0–0.001, 0.001–0.003, and greater than 0.003 for MSS; 0–0.4, 0.4–0.8, and greater than 0.8 for MSI) and showed the DNA shape values. The arrows show base positions with significant association between the DNA shape values and mutation rates (Supplemental Table S4).

Figure 3.

Highly mutated microsatellite markers. Result of clustering analysis with the top 20 microsatellites. (A) Colon/rectum (CR) cancer; (B) stomach (ST) cancer; (C) uterus (UT) cancer. Mutation status for the 20 microsatellites in each sample are shown. Colors indicates mutation status of each microsatellite (red, mutated; blue, unmutated; and white, unanalyzed owing to low depth). (D) Result of validation study in the independent CR and UT cancer cohort (n = 48). MSI status was defined by BAT25, BAT26, NR21, NR24, MONO27, D5S346, D17S250, and D2S123. Standard microsatellite markers, 18 highly mutated microsatellites, and homopolymers in coding regions were analyzed by MiSeq (Supplemental Fig. S19). Number of mutated samples in the MSI samples are shown in the bar plot (right). Colors indicates mutation status of each microsatellite (orange, mutated; white, unmutated; and gray, not tested). (MS) Microsatellite.

Figure 4.

Proportion of mutated samples in coding microsatellites. (x-axis) Proportion of mutated samples in the MSI samples; (y-axis) proportion of mutated samples in the MSS samples. COSMIC cancer genes are shown in red.

Figure 5.

Mutation in mismatch repair (MMR) and proofreading genes. (A) Colon/rectum (CR) cancer; (B) stomach (ST) cancer; (C) uterus (UT) cancer; (D) other MSI cancers. LI, liver; SK, skin; OV, ovary; PA, pancreas; and KI, kidney. Mutated genes and mutation rate of microsatellites are shown.

Figure 6.

Comparison of mutational signatures between the MSI and MSS samples. (A) CR cancer; (B) ST cancer; (C) UT cancer. Signatures showing significant difference between the MSI and MSS samples are shown in rectangles in the legends (Wilcoxon signed-rank test, q-value < 0.05).