Rapid antimicrobial sensitivity testing by single cell nanoscale optical interference

10 Growing antimicrobial resistance (AMR) is a serious global threat to human health, with 11 estimates of AMR leading to 10 million deaths per year and costing the global economy 12 $100tn by 2050. Current methods to detect resistance include phenotypic antibiotic 13 sensitivity testing (AST) which measures bacterial growth and is therefore hampered by slow 14 time to result (~12-24 hours). Therefore new rapid phenotypic methods for AST are urgently 15 needed. Here we describe a novel method for detecting phenotypic antibiotic resistance in 16 ~45 minutes, capable of detecting single bacteria. The method uses a sensitive laser and 17 detector system to measure nanoscale optical interference of single bacterial cells present in 18 media, with simple sample preparation. This provides a read out of bacterial antibiotic 19 resistance by detecting growth (resistant) or death (sensitive), much faster than current 20 methods. We demonstrate the potential of this technique by determining resistance in both 21 lab and clinical strains of E. coli, a key species for clinically burdensome urinary tract 22 infections. This work provides the basis for a simple and fast diagnostic tool to detect 23 antibiotic resistance in bacteria, reducing the health and economic burdens of AMR. 24 25

on multiple factors which effect growth rates, including inoculant concentration, strain, and 93 temperature, for example. We therefore normalise the data to the baseline before the addition 94 of antibiotic when comparing between experiments (S baseline ) (SI Figure 3b). 95

96
To obtain a systematic readout of antibiotic sensitivity across experiments, including multiple 97 strains and antibiotics, we obtain a normalised measure of bacterial growth as follows. We 98 All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/679399 doi: bioRxiv preprint define antibiotic sensitivity as r sensitivity : the ratio of S baseline and 45 minutes post-antibiotic 99 treatment (S antibiotic ), shaded blue in Figure 3a. r sensitivity provides a binary readout of 100 sensitivity, r sensitivity ≤ 1indicates cell death or inhibition of bacterial growth, and sensitivity to 101 the antibiotic in solution; r sensitivity > 1 indicates bacterial growth, and therefore resistance to 102 the antibiotic used. This method allows for both bactericidal and bacteriostatic antibiotics to 103 be used, as r sensitivity < 1 indicates a decrease in cell number, or cell death (bactericidal); 104 r sensitivity = 1 would indicate inhibition of growth, but little cell death (bacteriostatic). For 105 Figure 3a with ampicillin, r sensitivity = 0.5 for the green strain (sensitive) and r sensitivity = 1.  Table 1). This study demonstrates the ability of this method to successfully 132 All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. Vertical deflection data (nm) was recorded on JPK Nanowizard 3 software at 20 kHz 198 sampling frequency. This raw data (SI Figure 5a) was then processed in 800 second "chunks" 199 All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/679399 doi: bioRxiv preprint using analysis code written in Matlab. This code applies a Savitzky-Golay finite impulse 200 response (FIR) smoothing filter of polynomial order 2 to the data, with a filtering frequency 201 of 101 Hz (SI Figure 5b). A Savitzky-Golay smoothing filter was chosen as this function can 202 filter noisy data effectively without removing high frequency data. 203

204
To identify the number of bacterial crossings, both local maxima and minima were identified, 205 as bacteria moving through the laser was observed to cause both peaks and dips in the signal 206 (SI Figure 5c, peaks labelled with blue triangles). A "Peak Finder" function was used to 207 identify local minima/maxima in the signal, where a "peak" was defined as having a 208 threshold drop of at least 0.5 nm on each side. This was to ensure that only the larger peaks 209 were counted, which correspond to bacteria moving across the laser. Smaller "noise" seen in 210 the signal was not attributed to actual bacterial crossings, but could be due to partial 211 crossings, or a change of orientation of bacteria within the laser during a crossing. This 212 threshold peak prominence value of 0.5 nm was applied empirically across all files when 213 carrying out the analysis to remove any bias of identifying peaks in the signal.    The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/679399 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/679399 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/679399 doi: bioRxiv preprint