Predicting the Severity of Infection with Antibiotic-Resistant Bacteria

Antibiotic resistance is an adaptation through natural selection in which populations of micro-organisms, like bacteria, become resistant to the drugs that we use to treat them.  This is becoming a big problem in public health due to the widespread, overprescribed, and incorrect use of antibiotics.  Hospital acquired infections from antibiotic resistant bacteria are especially common.  The WHO keeps track of antibiotic resistant bacterial infections worldwide, and it also keeps a list of antibiotics that are considered to be most important for human medicine.

One "super bug" in particular that's problematic in hospitals and other health care facilities is methicillin-resistant Staphylococcus aureus (MRSA).  S. aureus is a sphere-shaped (coccus), gram-positive species of bacteria (pictured) naturally found in the respiratory tract and skin of humans that, when pathogenic, causes skin conditions and respiratory problems like sinusitis.  In the pre-antibiotic era, S. aureus was usually fatal.  S. aureus contains a number of surface proteins, called microbial surface components recognizing adhesive matrix molecules (MSCRAMM) that recognize and bind to molecules like collagen, fibrinogen, and fibronectin.  Once bound to these molecules, S. aureus cells can survive, grow, and persist.  During infection, S. aureus produces enzymes, like proteases, lipases, and elastases, that allow the bacteria to invade and destroy host tissues by interfering with the coagulation pathway.  The virulent factors of S. aureus are typically categorized as toxins (causing damage to host tissue) or adhesins (facilitate adherence and invasion of host tissue).  The treatment of choice for S. aureus is through the use of a beta-lactam family antibiotic, which includes penicillins (like methicillin, dicloxacillin, nafcillin, etc) and cephalosporins.  MRSA is a strain of S. aureus that is resistant to beta-lactams.  That means that infections are harder to treat, and people die from these infections.

A new European (open access) study that was made available online yesterday, published in Genome Research, has reported a method for predicting the severity of MRSA infection based on its genome sequence.  Prior to this study, genome sequencing of bacterial isolates from infected patients could determine factors like antibiotic resistance, but more complex phenotypes that involve contributions from a number of genes or epigenetic processes, like virulence, just wasn't possible.  Using a genome-wide association study (GWAS), the authors, led by co-contributors Laabei and Recker, determined the feasibility of predicting virulence through genetic signatures associated with specific phenotypes.

A GWAS is an examination of a number of common genetic variants to see if any variants are associated with any specific trait.  They typically focus on single nucleotide polymorphisms (SNPs, discussed previously here).  They took 90 MRSA isolates, and found that toxicity varies across isolates more than adhesiveness does, and that this toxicity correlates with disease severity.  Out of the 3060 SNPs found in the genome, 100 were identified as being associated with toxicity.  A further 22 specific indels (mutations involving insertions or deletions of DNA bases) were also identified as toxicity-related.  These polymorphisms and indels were located in genes that were involved in metabolism and regulation.

Since the GWAS approach is likely to produce a high number of false positives, the authors looked at the functional effect of these polymorphisms.  Each of the 13 SNPs they focused on was involved in virulence.  Furthermore, looking at epistatic interactions (in which a mutation in one gene masks the phenotypic effect of a mutation at another locus) found that a small number of genetic loci containing SNPs are interacting with numerous other loci.  For example, the authors identified 5 genes that interact with 20 other loci, involved in everything from tRNA synthesis to cell wall synthesis to carotenoid synthesis.  From all of this, the authors were able to identify SNPs/indels that can be used to predict the virulence of a particular strain of MRSA.

This study is a first step in the creation of a predictive model for identifying the most virulent strains of MRSA in a clinical setting, which can influence the approach taken to treat the infection.

   

                                                                                   

(I love Giant Microbes)