Subpopulations of stressedY. pseudotuberculosispreferentially survive doxycycline treatment within host tissues

Abstract

Severe systemic bacterial infections result in colonization of deep tissues, which can be very difficult to eliminate with antibiotics. It remains unclear if this is because antibiotics are not reaching inhibitory concentrations within tissues, if subsets of bacteria are less susceptible to antibiotics, or if both contribute to limited treatment efficacy. To determine the concentration of doxycycline (Dox) present within deep tissues following treatment, we generated a fluorescent transcriptional reporter derived from the tet operon to specifically detect intracellular tetracycline exposure at the single bacterial cell level. Dox exposure was detected in the spleen 2 hours after intraperitoneal injection, and by 4 hours post-injection, this treatment resulted in a significant decrease in viable Yersinia pseudotuberculosis in the spleen. Nitric oxide-stressed bacteria preferentially survived treatment, suggesting stress was sufficient to alter Dox susceptibility. Many bacteria (~10%) survived a single dose of Dox, and the antibiotic accumulated at the periphery of microcolonies to growth inhibitory concentrations until 48 hours post-treatment. After this timepoint, antibiotic concentrations decreased and bacterial growth resumed. Dox-treated mice eventually succumbed to the infection, albeit with significantly prolonged survival relative to untreated mice. These results indicate that Dox delivery by intraperitoneal injection results in rapid diffusion of inhibitory concentrations of antibiotic into the spleen, but stressed cells preferentially survive drug treatment, and bacterial growth resumes once drug concentrations decrease. This fluorescent reporter strategy for antibiotic detection could easily be modified to detect the concentration of additional antimicrobial compounds within host tissues following drug administration.ImportanceBacterial infections are very difficult to treat when bacteria spread into the bloodstream and begin to replicate within deep tissues, such as the spleen. Subsets of bacteria can survive antibiotic treatment, but it remains unclear if this survival is because of limited drug diffusion into tissues, or if something has changed within the bacteria, promoting survival of some bacterial cells. Here, we have developed a fluorescent reporter to detect doxycycline (Dox) diffusion into host tissues, and show that Dox impacts the bacterial population within hours of administration, and inhibits bacterial growth for 48 hours. However, bacterial growth resumes when antibiotic concentrations decrease. Subsets of bacteria, stressed by the host response to infection, survive Dox treatment at a higher rate. These results provide critical information about the dynamics that occur within deep tissues following antibiotic administration, and suggests subsets of bacteria are predisposed to survive inhibitory concentrations of antibiotic before exposure.

Modifying TIMER, a slow-folding DsRed derivative, for optimal use in quickly-dividing bacteria

“…Recently, we have shown that NO-exposed Hmp + cells preferentially survive doxycycline treatment in our mouse model of infection, which suggested the Hmp + cells may represent a slow growing subpopulation of bacteria (28). Our in vitro results were consistent with this, where exposure to NO was sufficient for TIMER42 signal accumulation.…”

“…It remains unclear if the protective expression of hmp by peripheral cells comes at a fitness cost, and if this stress promotes the formation of a slow-growing subset of bacteria at the periphery of microcolonies. The preferential survival of hmp-expressing cells during antibiotic treatment supported this hypothesis (28), however our results here will suggest that NO exposure occurs transiently within host tissues, and that Hmp + cells retain the ability to divide.…”

“…(Fig 8D). The half-life of this destabilized mCherry protein is approximately 35 minutes in culture (28), and assuming the turnover rate is similar within host tissues, this marks a very recent exposure to NO. This mosaic of expression suggests that while many cumulative cells were exposed to NO and responded with hmp expression, this process is transient and actively occurring in only a few cells at a time.…”

“…We were surprised to find that NO stress did not correlate with TIMER42 signal accumulation within host tissues, and hypothesized this could be due to temporal differences in when bacteria were exposed to NO and when tissues were collected. We believe that bacteria respond to NO much earlier during infection (approximately 48h p.i., (22,28)), and it was possible that bacteria were no longer actively responding to stress. We returned to the acidified nitrite method of introducing NO stress, since this likely results in an initial burst of NO, and then slower NO generation (36).…”

“…Phmp::mCherry::gfp was a control construct generated with stable mCherry and stable GFP proteins, and we confirmed these signals overlapped during growth in tissue (Fig 8D). The Phmp::mCherry-ssrA::gfp construct contained destabilized mCherry (28) and stable gfp, to allow us to detect bacteria that had responded to NO as some point during infection (GFP + ) compared to bacteria that were actively responding to NO when the tissue was harvested and fixed (mCherry + ). Using this reporter, we found very few bacterial cells were actively responding to NO when tissue was harvested (Fig 8D), suggesting that other cells may have responded to NO, recovered, and resumed growth, explaining why they no longer exhibited TIMER42 signal accumulation.…”