Developing Breath-Based Diagnostics for Lung Infections Using Secondary Electrospray Ionization-Mass Spectrometry (SESI-MS)

Bacteria produce unique combinations of volatiles that can be used to identify the genus and species, and in many cases the strain or serovar. The ability to identify bacteria by their volatilomes has generated great expectations for rapid and non-invasive clinical tests that are able to diagnose and identify infections in situ, particularly for diagnosing lung infections via breath analysis. A few studies have demonstrated that breath volatiles can be used to diagnose infection. While promising, these studies mostly focus on detection of only one specific bacterial species, distinguishing between uninfected and infected patients. The ideal breath diagnostic, however, would be used to identify the pathogen responsible for the infection. We aim to advance breath-based diagnostics by using the volatile fingerprinting technique of SESI-MS to identify murine lung infections caused by H. influenzae, K. pneumoniae, L. pneumophila, M. catarrhalis, P. aeruginosa , S. aureus, and S. pneumoniae, to understand how the infection breathprints change over time, and how host volatiles contribute to diagnosis. Background: Arsenic, a known toxicant, is regulated at concentrations >10 µg/l in U.S. public drinking water. However, in rural areas many residents rely on unregulated private wells for their household water. Previous studies indicate that breastmilk arsenic concentrations are low even in areas with very high groundwater arsenic, suggesting that breastfeeding may be protective against arsenic exposure relative to the use of powdered infant formulas reconstituted with household tap water. Objective: Determine how infant consumption of breastmilk and formula contributes to exposure to arsenic during early infancy. Methods: We estimated arsenic exposure via breastmilk and formula for a subpopulation (n=72) of the New Hampshire Birth Cohort (NHBC), a prospective cohort of >1000 mother-infant pairs who use a private, unregulated well as the household water source. During pregnancy, we collected a sample of home tap water and during the 6-week postpartum appointment we collected a sample of infant urine and 3-day food diary. We also collected a breastmilk sample from 9 mothers. All samples were analyzed for arsenic using inductively coupled plasma mass spectrometry (ICP-MS). We used regression models to evaluate potential predictors of infant urinary arsenic and estimated exposure for exclusively breastfed and exclusively formula fed infants using NHBCS data. Results: Feeding mode was a strong, statistically significant predictor of urinary arsenic. The median concentration for infants who were mostly breastfed (0.24 µg/L) was 2-3 times lower compared to infants fed mostly formula (0.86 µg/L) and those …