Rapid phenotypic stress-based microfluidic antibiotic susceptibility testing of Gram-negative clinical isolates

Maxim Kalashnikov,Alexis F. Sauer-Budge,Christine McBeth,Marc Mueller,Jennifer Campbell,Jean C. Lee,Andre Sharon

doi:10.1038/s41598-017-07584-z

Abstract

Bacteremia is a life-threatening condition for which antibiotics must be prescribed within hours of clinical diagnosis. Since the current gold standard for bacteremia diagnosis is based on conventional methods developed in the mid-1800s—growth on agar or in broth—identification and susceptibility profiling for both Gram-positive and Gram-negative bacterial species requires at least 48–72 h. Recent advancements in accelerated phenotypic antibiotic susceptibility testing have centered on the microscopic growth analysis of small bacterial populations. These approaches are still inherently limited by the bacterial growth rate. Our approach is fundamentally different. By applying environmental stress to bacteria in a microfluidic platform, we can correctly assign antibiotic susceptibility profiles of clinically relevant Gram-negative bacteria within two hours of antibiotic introduction rather than 8–24 h. The substantial expansion to include a number of clinical isolates of important Gram-negative species—Enterobacter cloacae, Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa—reported here underscores the broad utility of our approach, complementing the method’s proven utility for Gram-positive bacteria. We also demonstrate that the platform is compatible with antibiotics that have varying mechanisms of action—meropenem, gentamicin, and ceftazidime—highlighting the versatility of this platform.

Highlights

The staggering rise in antibiotic-resistant bacteria combined with the sharp decrease in the development of new antibiotics have combined to create a scenario in which many bacterial infections cannot be treated effectively
Molecular methods (e.g., PCR) for determining antibiotic resistance following positive blood culture exist and can be useful during epidemics. These tests are inappropriate for routine diagnosis because 1) susceptibility is not always correlated with genetic markers, 2) genetic markers have not been identified for all antibiotic-resistant bacterial strains[9,10,11], and 3) selective pressure leads to a high frequency of genetic mutation[12]
To interrogate the response of bacteria under stress to antibiotics, it is necessary to immobilize them on the floor of our microfluidic platform

Summary

Introduction

The staggering rise in antibiotic-resistant bacteria combined with the sharp decrease in the development of new antibiotics have combined to create a scenario in which many bacterial infections cannot be treated effectively. The standard process for determining antibiotic susceptibility for blood-borne pathogens, from patient specimen to final result, takes 48–72 h This massive bottleneck contributes largely to the widespread prescription of broad-spectrum antibiotics—given the complete dearth of diagnostic information at early time points, attending medical staff have little choice but to prescribe broad-spectrum antibiotics in order to preserve the life of the patient[1, 6, 7]. Molecular methods (e.g., PCR) for determining antibiotic resistance following positive blood culture exist and can be useful during epidemics. These tests are inappropriate for routine diagnosis because 1) susceptibility is not always correlated with genetic markers, 2) genetic markers have not been identified for all antibiotic-resistant bacterial strains[9,10,11], and 3) selective pressure leads to a high frequency of genetic mutation[12]. Because there are billions of blood cells present in the sample, bacteria must either be purified away from these cells (sample preparation) or multiplied to a much higher concentration (blood culture)

Methods

Results

Conclusion