NEW YORK (GenomeWeb News) – A genome-wide association study appearing online yesterday in the Proceedings of the National Academy of Sciences has started to catalog genetic variants that influence peptide levels in the blood proteome.
A team based at Uppsala University in Sweden and the Netherlands' Leiden University brought together genotype information on more than 1,000 individuals from Sweden with mass spectrometry-based profiling of peptide levels in their blood. Results from the discovery and validation stage of the GWAS indicated that heritability influenced peptide levels for a significant proportion of the proteins considered.
When they delved into some of the apparent SNP-peptide connections, the researchers narrowed in on variants that appear to affect peptide levels through regulatory mechanisms or sequence changes — bolstering hopes of further characterizing the set of so-called "protein quantitative trait loci," or pQTLs, that impact protein profiles and abundance in the blood proteome.
Based on findings from the current proteomics-focused GWAS, those involved in the analysis argued that clinically useful information could come from future research looking at larger and larger sets of peptides and their relationship to SNPs and other forms of variation in the genome.
"[T]he SNP array data should be complemented by sequence data to include rare variants and structural variation," the study's senior author Ulf Gyllensten, a researcher with Uppsala University's immunology, genetics, and pathology department, and his co-authors explained.
"Nevertheless, the results of our study demonstrate the presence of genetic variants with a strong impact on the protein profile," they said, "and the ability to annotate their functional relevance in affecting the level of gene products."
Though an ever growing tally of SNPs are being linked to human diseases or traits through GWAS, the researchers argued that there is still a ways to go in understanding these variants and their biological effects — be they regulatory, coding, splice site-related, or so on.
Similarly, some studies have started to track down SNPs that seem to influence gene expression or metabolite profiles in individual's blood or urine, they added, while similar studies on proteomic profiles have lagged behind due to difficulties in efficiently measuring levels of many peptides.
"The main limitation in the identification of pQTL has been access to high-throughput methods in proteomics to study the abundance of individual proteins in human clinical samples, which can be applied to the analysis of large cohorts," authors of the study noted.
For their new analysis, the researchers relied on high-resolution, high-throughput mass spectrometry to find and quantify up to 1,056 peptides — amino acids representing 163 different proteins — in blood samples from 1,060 Swedish individuals enrolled through the Northern Sweden Population Health Study. In addition, DNA from each participant's blood sample was genotyped at SNPs across the genome using Illumina microarrays.
After quality control analyses, researchers were left with mass spec and genotyping data for 1,029 of the individuals — information that they used for looking at genetic and other factors affecting peptide levels in blood.
Along with peptides that had levels varying with an individual's sex, body mass index, or age, the team tracked down almost 200 peptides from 57 proteins with blood levels that appeared to be at least partly heritable.
A discovery GWAS involving Swedish NSPHS participants from a parish known as Karesuando highlighted a slew of SNPs suspected of serving as pQTLs for a subset of the peptides profiled. The team narrowed in on variants in and around 32 peptides for validation testing on participants in a nearby village called Soppero, verifying that at least 11 of the peptides (comprising five proteins) had blood proteome levels tied to nearby SNPs.
Effects for some of the SNPs appeared to be as a consequence of regulatory changes — in some instances affecting multiple peptides from the same protein. Other variants appear to influence peptide levels through changes to the peptide sequence itself.
An even wider set of peptides had apparent links to SNPs when the team considered data from all 1,029 Swedish individuals as a group. There, peptides from 25 proteins exhibited protein levels that surged or waned in concert with genetic variants.
These included SNPs previously implicated in Alzheimer's disease or macular degeneration, they noted, hinting that future studies of this nature could prove useful for improving researchers' understanding of disease biology or perhaps for finding clinically informative variants.
"Abundance and peptide composition of a protein plays an important role in the etiology, diagnosis, and treatment of a number of diseases," Gyllensten and co-authors explained.
"A better understanding of the genetic impact on the plasma proteome is therefore important for evaluating potential biomarkers," they continued, "and therapeutic agents for common diseases."