Patient genetics is linked to chronic wound microbiome composition and healing

Author summary Chronic, or non-healing, wounds represent a costly burden to patients, and bacterial infection of wounds is an important driver of chronicity. A variety of bacterial species often occur in chronic wounds, but it is unknown why certain species are observed in some wound infections and not others. In this study, genetic variation of wound clinic patients was compared to the bacteria observed in their infected wounds. Through these comparisons, genetic variation in theTLN2andZNF521genes was found to be associated with both the number of bacteria observed in wounds and the abundance of common pathogens (primarilyPseudomonas aeruginosaandStaphylococcus epidermidis). Moreover,Pseudomonasinfected wounds were found to have fewer species present and wounds with fewer species were slower to heal. Furthermore, patient genes associated with microbiomes commonly encode proteins known to be important for cellular structures important to healing and to which bacteria directly interact. Experimental investigation of one such gene,TLN2, identified genotype-dependent differences in the expression of functionally different versions ofTLN2that is hypothesized to shape differences in cellular adhesion structures. Finally, a new statistical approach is presented in which patient biomarkers are used to predict the number of species observed during infection. Overall, our results describe how patient genetic variation influence the types of bacteria likely to infect an individual as well as influence healing. The clinical importance of microbiomes to the chronicity of wounds is widely appreciated, yet little is understood about patient-specific processes shaping wound microbiome composition. Here, a two-cohort microbiome-genome wide association study is presented through which patient genomic loci associated with chronic wound microbiome diversity were identified. Further investigation revealed that alternativeTLN2andZNF521genotypes explained significant inter-patient variation in relative abundance of two key pathogens,Pseudomonas aeruginosaandStaphylococcus epidermidis. Wound diversity was lowest inPseudomonas aeruginosainfected wounds, and decreasing wound diversity had a significant negative linear relationship with healing rate. In addition to microbiome characteristics, age, diabetic status, and genetic ancestry all significantly influenced healing. Using structural equation modeling to identify common variance among SNPs, six loci were sufficient to explain 53% of variation in wound microbiome diversity, which was a 10% increase over traditional multiple regression. Focusing onTLN2, genotype atrs8031916explained expression differences of alternative transcripts that differ in inclusion of important focal adhesion binding domains. Such differences are hypothesized to relate to wound microbiomes and healing through effects on bacterial exploitation of focal adhesions and/or cellular migration. Related, other associated loci were functionally enriched, often with roles in cytoskeletal dynamics. This study, being the first to identify patient genetic determinants for wound microbiomes and healing, implicates genetic variation determining cellular adhesion phenotypes as important drivers of infection type. The identification of predictive biomarkers for chronic wound microbiomes may serve as risk factors and guide treatment by informing patient-specific tendencies of infection.