PROJECT ABSTRACT
Antibiotic resistant bacterial infections are a significant, modern epidemic causing >23,000 deaths and two million
illnesses in the US, annually. To treat and prevent antibiotic resistant infections, we must understand the
complementary issues of (i) how new antibiotic resistance emerges and (ii) how these organisms are transmitted.
Traditionally, bloodstream infection-causing antibiotic resistant pathogens are thought to be passed directly to
the bloodstream from patient-to-patient. We propose instead that three of the most prevalent pathogens,
Escherichia coli, Enterococcus faecium, and Enterococcus faecalis are transmitted to and from the gut
microbiome of hospitalized patients, a transformative concept. Recently, we (Bhatt, Banaei) showed that
pathogen reservoirs may ‘lie in wait’ in the gut microbiome of hematopoietic cell transplantation (HCT) patients,
eventually translocating into the bloodstream to cause infections (Tamburini et al, Nature Medicine 2018).
Leveraging a skilled and dedicated multidisciplinary team of experts in microbial genomics tool development
(Bhatt), Clinical Microbiology (Banaei) and Statistics/Genetics (Tang), we will investigate both gut-to-gut (aim 1)
and gut-to-blood transmission (aim 3), and will also test hypothesis that the gut microbiome is a niche where E.
coli and Enterococci can evolve new antibiotic resistance (aim 2). Others have demonstrated that shotgun
sequencing methods can enable exquisite strain tracking of isolated and cultured bacterial pathogens (e.g.
carbapenemase-producing Klebsiella pneumonia and XDR TB). Unfortunately, technical challenges have
prevented simultaneous and accurate complete assembly of microbial genomes, and thus tracking of multiple
organisms’ genomes within complex gut microbiomes. Recently, we overcame this challenge by developing an
innovative “read cloud” sequencing approach to stitch together the genomes of individual bacteria directly from
sequencing of a microbiome sample (Bishara et al, Nature Biotech 2018). We will apply this approach as well as
meta-HiC, which can “link” plasmids to chromosomal DNA, to a longitudinal stool specimen collection from 666
HCT patients cared for on the same hospital ward over 3 years (1552 samples; median 2/patient; 215 patients
with ³3 samples). Leveraging temporal, geospatial and clinical information, we will use this stool bank resource
to test the hypotheses that (i) E. coli and Enterococci are transmitted between gut reservoirs of HCT patients
and (ii) new antibiotic resistance can evolve in these common gut-dwelling pathogens in HCT patients. Thus, we
will identify risk factors for microbiome transmission, quantify emergence of new antibiotic resistance in HCT
patients, and determine hospital-wide transmission networks for three important pathogens. Identifying these
networks will inform methods that hospitals use (such as isolation rooms and use of contact precautions) to limit
bacterial transmission, as well as how clinicians perceive the downsides of antibiotic use in hospitalized patients
by demonstrating the impact of antibiotics on the emergence of new antibiotic resistance.