NEW YORK — DNA methylation signatures may be able to distinguish patients with persistent methicillin-resistant Staphylococcus aureus infections from those whose MRSA infections resolve, a new study has found.
Persistent MRSA infections in which the infection does not clear up even with proper treatment can occur in up to 30 percent of MRSA cases and be life threatening. This includes treatment with antibiotics that the isolate appears to be sensitive to in vitro, suggesting there may be an interplay of factors that contribute to infection persistence.
Building on a previous finding of theirs that found ties between a DNA methylation enzyme in people and MRSA treatment outcomes, researchers led by the University of California, Los Angeles' Elaine Reed examined differences in DNA methylation patterns between people who had MRSA that resolved with treatment and those who did not. They reported in the Proceedings of the National Academy of Sciences on Monday that the differences in methylation signatures were linked to specific immune cell genes and could potentially be used to guide treatment.
"The results suggest that DNA methylation epigenotypes in specific immune cells could be used to assess and predict the likelihood of persistence outcomes at the first signs of infection," Reed and her colleague Michael Yeaman, also of UCLA, wrote in an email. "If so, forecasting such outcomes may guide pre-emptive, personalized therapeutic regimens to prevent persistence and improve outcomes."
The researchers recruited a matched cohort of 142 individuals, about half of whom had persistent MRSA and half of whom had MRSA that resolved. As expected, the individuals with persistent infections had higher rates of death, longer hospital stays, and increased metastatic risk.
They collected peripheral blood from the patients for global methylation analysis using reduced-representation bisulfite sequencing and built a classification model based on DNA methylation levels of nearly 750,000 CpG sites. They found that the methylation levels of 276 host CpG sites were significantly associated with persistent or resolved infections. In particular, 160 of the differentially methylated sites were hypomethylated and 116 hypermethylated in persistent MRSA versus resolving MRSA.
Based on these clusters, the researchers developed a classification model. Further analysis additionally suggested that DNA methylation profiles could help determine or even predict persistent MRSA.
A number of the differentially methylated sites the researchers identified fell just upstream or more than 50 kilobases downstream of transcription start sites and a number were enriched in transcription factor binding sites. They particularly noted hypomethylation in the binding sites for CCAAT enhancer binding protein-β (C/ENPβ) and the signal transducer/activator of transcription 1 (STAT1) in patients with persistent disease. Meanwhile, hypomethylation among patients with resolving infections largely focused on glucocorticoid receptor and histone acetyltransferase p300 binding sites. These methylation signatures were particularly enriched in neutrophils, they noted.
The researchers further validated their findings through targeted bisulfite sequencing of differentially methylated sites and regions they uncovered. They additionally analyzed the genetic sequences at those sites to find that epigenetics rather than genetics appear to sway patient responses.
The findings suggested to the researchers that their methylation signatures could be used in the clinic to quickly determine which patients might be more likely to have persistent MRSA infections. "In a nutshell, our findings bring us one large step closer to personalized diagnosis and effective treatment of life-threatening infections," Reed and her colleagues said.
They added they are now conducting a prospective trial to validate whether these DNA methylation signatures can predict persistent MRSA from resolving MRSA bacteremia. In addition, they are examining whether such methylation signatures may be present in other infections.