A Large-Scale Analysis of the Relationship of Synonymous SNPs Changing MicroRNA Regulation with Functionality and Disease

Yuchen Wang, Chengxiang Qiu, Qinghua CuiInternational Journal of Molecular Sciences 2015 · DOI

TL;DR

This paper challenges the long-held assumption that synonymous mutations (which don't change amino acid sequences) are neutral. By analyzing large-scale genomic data, the researchers found that a significant portion of these "silent" SNPs can alter microRNA (miRNA) binding sites, leading to functional consequences and increased disease risk. This suggests a novel mechanism by which synonymous mutations contribute to human health and evolution.

What Problem Does This Paper Address?

For decades, the field of molecular biology has largely considered synonymous mutations to be neutral. The reasoning is straightforward: if a DNA change doesn't alter the amino acid sequence of a protein, it's unlikely to affect the protein's function and therefore unlikely to impact an organism's phenotype or disease susceptibility. This has led researchers to primarily focus on non-synonymous mutations when studying genetic variations and their roles in disease. However, recent findings have hinted that this view might be too simplistic. Some synonymous mutations have been shown to influence mRNA splicing, stability, and translation efficiency. This paper tackles the question of whether there are broader, more general mechanisms by which synonymous mutations can exert functional effects, specifically by investigating their impact on microRNA (miRNA)-mediated gene regulation.

Methods

The study employs a large-scale bioinformatic analysis to investigate the functional impact of synonymous single nucleotide polymorphisms (SNPs). The core of their approach involves:

  • Predicting miRNA Binding Sites: They used the TargetScan and miRanda tools to identify synonymous SNPs that could alter miRNA binding sites within the coding regions of genes. This involved generating sequences for both alleles of a given SNP and scanning them for potential miRNA interactions.
  • Analyzing Selection Signatures: To assess if these miRNA-regulating SNPs are under evolutionary pressure, they analyzed the Integrated Haplotype Score (iHS), a statistic that detects recent positive selection. This was done across three human populations: Asian (ASN), European (CEU), and African (YRI).
  • Investigating Gene Expression: They correlated genotype data with gene expression data from lymphoblastoid cell lines (LCLs) to see if synonymous SNPs affecting miRNA regulation also impacted gene expression levels.
  • Examining Disease Association: They leveraged data from the NHGRI GWAS Catalog and the ClinVar database to determine if synonymous SNPs that alter miRNA regulation are disproportionately associated with human diseases and traits compared to those that do not.
  • Extending the Model: The researchers also extended their analysis to non-synonymous SNPs, investigating whether miRNA-mediated regulation plays a role in their functional consequences and disease associations, even for those predicted to be "benign."

Datasets

The study utilized several publicly available datasets:

  • Synonymous and Non-synonymous SNPs: Obtained from the UCSC genome browser (dbSNP 137), comprising 47,039 synonymous and 48,768 non-synonymous SNPs.
  • iHS Scores: Sourced from the haplotter website for three populations: ASN, CEU, and YRI.
  • Transcriptome and Genotype Data: From lymphoblastoid cell lines (LCLs) of 60 CEU individuals, obtained from the lab website of Emmanouil T. Dermitzakis.
  • Human Disease and Trait Loci: From the NHGRI GWAS Catalog and the ClinVar database.

Key Insights and Ideas

  • Synonymous SNPs are Not Always Silent: The most significant insight is that a substantial fraction of synonymous SNPs can actively influence gene regulation through their impact on miRNA binding. This overturns the traditional view of their neutrality.
  • miRNA Regulation as a Novel Mechanism: The paper proposes miRNA-mediated gene regulation as a key, and previously underappreciated, mechanism by which synonymous mutations can lead to functional consequences and disease.
  • Coding Regions are Important miRNA Binding Sites: The study highlights that miRNA binding sites are not confined to the 3' UTR but are also frequently found in coding regions, making synonymous SNPs in these regions particularly relevant for miRNA regulation.
  • Functional SNPs are Under Selection: Synonymous SNPs that alter miRNA regulation show a tendency to be under recent positive selection, suggesting they have been favored by evolution due to their functional impact.
  • Beyond Deleterious: Benign SNPs Matter Too: The extension of the model to non-synonymous SNPs suggests that even "benign" mutations (those not predicted to change amino acid sequence) can have functional consequences through miRNA regulation, implying that traditional classification of SNPs might be incomplete.

Key Results

  • Prevalence of miRNA-Altering Synonymous SNPs: Approximately 45.9% of all synonymous SNPs were predicted to alter miRNA binding sites.
  • Association with Positive Selection: Synonymous SNPs changing miRNA regulation showed a statistically significant tendency to be under recent positive selection in Asian (p=0.03) and African (p=0.02) populations, but not European.
  • Impact on Gene Expression: Genes hosting synonymous SNPs that alter miRNA regulation were more likely to show a significant correlation between genotype and expression level (10.5%) compared to genes without such SNPs (6.8%), with an odds ratio of 1.6.
  • Disease and Trait Association: Synonymous SNPs changing miRNA regulation were significantly more likely to be associated with human diseases and traits (OR = 1.6, p = 0.02 from GWAS data; OR = 1.3, p = 9.0 × 10⁻¹¹ from ClinVar data).
  • Non-synonymous SNPs and miRNA Regulation: A significant portion (45.7%) of non-synonymous SNPs were also found to alter miRNA-mediated gene regulation, and these also showed increased association with disease and positive selection.

Limitations and Future Work

  • Predictive Tools Limitations: The study acknowledges that current miRNA target prediction tools (like TargetScan and miRanda) have limited consistency and may not capture all true miRNA-changing SNPs. Improvement in these tools is crucial for better prediction accuracy.
  • Experimental Validation Needed: While the bioinformatic analysis provides strong correlative evidence, direct experimental validation of the functional consequences of these miRNA-altering synonymous SNPs is still needed.
  • Understanding Non-Canonical Binding: The exact molecular mechanisms by which miRNAs regulate target genes through non-canonical binding sites in coding regions require further investigation.
  • Broader Evolutionary Implications: The findings suggest that the concept of synonymous SNPs as neutral controls in evolutionary analyses may need to be revisited.

The authors propose that future work should focus on developing better tools for predicting miRNA regulation changes by coding region SNPs and experimentally validating these predictions. They also emphasize the need to consider both synonymous and benign non-synonymous SNPs alongside deleterious ones when assessing functional consequences and disease risk.

Summary updated

Abstract

Historically, owing to not changing amino acid composition of protein sequences, synonymous mutations are commonly assumed to be neutral during evolution and therefore have no effect on the phenotype and disease. Here, based on observations from large-scale analysis of genomic data, we predicted the putative synonymous SNPs that could result in functional consequences and disease risk through changing the microRNA-mediated gene regulation. We found that nearly half of the synonymous SNPs could affect protein expression by changing microRNA regulation in human genome and these SNPs significantly prefer to be associated with human diseases and traits. The synonymous SNPs changing microRNA-mediated gene regulation tend to be more under recent positive selection, prefer to affect gene expression, and implicate in human disease. We conclude that the miRNA-mediated regulation changes could be a potential mechanism for the contributions of synonymous SNPs to protein functions and disease risks.
BibTeX
@article{wang2015large,
  author = {Yuchen Wang and Chengxiang Qiu and Qinghua Cui},
  title = {{A Large-Scale Analysis of the Relationship of Synonymous SNPs Changing MicroRNA Regulation with Functionality and Disease}},
  journal = {International Journal of Molecular Sciences},
  year = {2015},
  doi = {10.3390/ijms161023545},
}
Figure