Copy number flexibility facilitates heteroresistance to increasing antibiotic pressure and threatens the beta-lactam pipeline

Abstract

Our understanding of the rapid adaptation of bacteria to resist novel drugs is growing beyond known mechanisms such as mobile genetic elements and mutation selection. Heteroresistance (HR) is a form of antibiotic resistance where a phenotypically unstable minority resistant subpopulation co-exists with a susceptible population. We sought to uncover the mechanism of heteroresistance to cefiderocol, a novel β-lactam developed to resist β-lactamases including extended-spectrum-β-lactamases (ESBLs), which has been recently reported but poorly understood. We observe HR to cefiderocol among clinical isolates collected before its use. The resistant subpopulation in Enterobacter is a continuum; increasing copy number of a gene encoding an ESBL ineffective against cefiderocol mediates increased resistance in decreasing numbers of cells. We then pursued the factors that control the magnitude of amplification. We observe that ESBL activity correlates with the level of amplification, and thus that increased copy number can compensate for poor enzymatic activity. A Klebsiella isolate from a clinical treatment failure also demonstrates amplification, highlighting the potential relevance of this β-lactamase gene amplification-mediated HR. These data provide insights into factors controlling dynamics of HR and how bacteria can use gene amplification to flexibly confront new antibiotic threats.

Fig. 1. Increased bla_SHV-5 copy number generates cefiderocol heteroresistance in Enterobacter strain RS.

a Population analysis profile (PAP) of strains RS and E. cloacae Mu1197 plated on Mueller-Hinton agar (MHA) containing cefiderocol; the proportion of surviving colonies is quantified relative to MHA containing 0 cefiderocol, the mean and standard deviation are shown from two independent experiments with 4 biological replicates each. b Timekill of strain RS: growth in media containing 16 μg/mL cefiderocol over time and quantification of surviving colony forming units on MHA (RS total population) and MHA containing 2 μg/mL cefiderocol (resistant subpopulation) from a single representative replicate. c Quantification of the cefiderocol-resistant subpopulation of strain RS after growth in media alone (baseline), subcultured into 16 μg/mL cefiderocol (CFDL) and grown for 24 h, and subcultured every 24 h into fresh media without cefiderocol. At the end of each growth, an aliquot was diluted and plated onto MHA containing 2 μg/mL cefiderocol to quantify the resistant subpopulation, from a single representative replicate. d The region of the RS chromosome amplified in cells surviving growth in 16 μg/mL cefiderocol is shown, with gene abundance shown below. e PAP on cefiderocol of RS, the Δbla_SHV-5 and Δamplified region (ΔNF29_02990-NF29_03045) isogenic mutants, and the Δbla_SHV-5 mutant complemented with bla_SHV-5 at the native site; the mean and standard deviation are shown from two independent experiments with 3 biological replicates each. f, g bla_SHV-5 gene abundance quantified from DNA (f) or bla_SHV-5 transcript abundance quantified from RNA (g) by qPCR from colonies collected from MHA containing 32 μg/mL cefiderocol, normalized to no cefiderocol for each replicate from the strains indicated. Log-transformed data with mean and quartiles shown, from two experiments with 6 biological replicates each; one-way ANOVA, [F (3, 44) = 254.6 for (f), F (3, 44) = 90.01 for (g)] with Sidak’s multiple comparisons test. h Representative anti-myc immunoblot from total cellular lysate from colonies collected from MHA containing 0 or 32 μg/mL cefiderocol of strain RS Δbla_SHV-5::bla_SHV-5_myc. SHV-5_myc is 30.09 kD following cleavage of the signal sequence. i anti-myc immunoblot band density analysis with Li-Cor analysis, showing the relative ratio of the band intensity from samples grown on 32 μg/mL cefiderocol relative to no drug, from six biological replicates across two independent experiments. Full blots and additional images corresponding to (h, i) are shown in Supplementary Fig. 5. j, k Outcomes of infection of Galleria larvae and cefiderocol treatment, j The frequency of the subpopulation enumerated on agar containing 2 μg/mL cefiderocol with mean and standard deviation shown. k bla_SHV-5 gene abundance from CFU collected from agar containing 2 μg/mL cefiderocol, normalized to infection inoculum, mean and standard deviation are shown. For (i, j), each symbol indicates a single larva, from two independent experiments, n = 13 for PBS-treated and n = 17 for cefiderocol-treated. Data are log-transformed, two-tailed t-test with Welch’s correction [t = 5.854, df = 27.95 for (j) and t = 7.656, df=16.97 for (k)]. Source data are provided as a Source Data file.

Fig. 2. Amplification generates a subpopulation continuum responsive to inhibition.

a PAP of RS plated on cefiderocol, with the proportion of surviving colonies in the black line on the left y-axis, with mean and standard deviation shown. bla_SHV-5 abundance was quantified from the samples collected at each concentration by qPCR and log10 transformed. Each dot indicates a biological replicate, with median and quartiles indicated graphed on the right y-axis, from two independent experiments; total n = 6 for PAP and n = 7 for qPCR, one-way mixed-effects ANOVA analysis with Geisser-Greenhouse correction and Dunnett’s multiple comparison test, F (2.387, 13.85) = 112.1. b Details of the populations collected on 8 µg/mL or 32 µg/mL cefiderocol are shown. c A model of the RS population: as the concentration of cefiderocol increases, the proportion of the surviving cells decreases. The cefiderocol resistance of the surviving resistant cells increases with an increase in the bla_SHV-5 abundance. d PAP of strain RS or the isogenic Δbla_SHV-5 mutant plated on cefiderocol or cefiderocol and 4 µg/mL clavulanate. The mean and standard deviation are shown from two independent experiments with 3 biological replicates each. e Strain RS was plated on MHA containing cefiderocol and clavulanate as indicated. The proportion of the surviving population is the line graph with the left y-axis, with means and standard deviation. The corresponding log10 transformed gene abundance is graphed on the right y-axis, where each point indicates a biological replicate from two independent experiments with n = 9 total biological replicates, one-way mixed-effects ANOVA, F (2.742, 21.15) = 181.8 with Geisser-Greenhouse correction and Dunnett’s multiple comparison test. Source data are provided as a Source Data file.

Fig. 3. Inhibition of β-lactamase activity leads to a compensatory increase in bla_SHV-5 copy number.

a PAP of Enterobacter RS strains plated on cefiderocol. Genotypes are Δbla_SHV-5, Δbla_SHV-5::bla_SHV-5 (referred to as bla_SHV-5 (WT))_, and Δbla_SHV-5::bla_{SHV-5 M69I} (bla_{SHV-5 M69I}). Shown are the means and standard deviation of three independent experiments with 3 biological replicates each. b bla_SHV-5 and bla_{SHV-5 M69I} amplification from surviving colonies on 4 µg/mL cefiderocol in (a), each point indicates a replicate from three independent experiments, total n = 10 for bla_SHV-5 (WT) and n = 15 for bla_{SHV-5 M69I}. Data are log10 transformed, one-way ANOVA with Sidak’s correction for multiple comparisons, F (3, 46) = 6.471. c, d RS strain bla_SHV-5 and bla_{SHV-5 M69I} were plated on MHA containing cefiderocol (lightest and bottom-most line in c and d), colonies collected from 4 µg/mL cefiderocol, and grown in broth containing 4 µg/mL cefiderocol. The cultures were plated on MHA plus cefiderocol (next darker line), collected from and grown in 8 µg/mL cefiderocol, and plated on MHA plus cefiderocol (next darker line). Colonies were collected from and grown in 16 µg/mL cefiderocol, plated on MHA plus cefiderocol (darkest line), and DNA was extracted. ebla_SHV-5 and bla_{SHV-5 M69I} gene abundance quantified by qPCR and log10 transformed following growth in 16 µg/mL cefiderocol from c and d, with mean above each group. c–e show data of bla_SHV-5 from 17 replicates across 4 experiments and bla_{SHV-5 M69I} from 14 replicates across 4 experiments. c, d show the mean and standard deviation, while (e) shows each replicate as a dot. For (e), two-tailed unpaired t-test (t = 5.423, df = 29). Source data are provided as a Source Data file.

Fig. 4. Cefiderocol/β-lactamase inhibitor exposure increases amplification and results in a resistant subpopulation.

a bla_SHV-5 abundance measured by qPCR and log10 transformed after serial exposure to cefiderocol (CFDL)/clavulanate (Clav.), with mean above each group. RS was plated on MHA alone or with 8 µg/mL cefiderocol and 0.06 µg/mL clavulanate, and surviving colonies were collected, and bla_SHV-5 abundance was quantified. A portion was subcultured into media with 8 µg/mL cefiderocol and 0.06 µg/mL clavulanate and grown for 24 h. The culture was plated on MHA containing 16 µg/mL cefiderocol and 0.25 µg/mL clavulanate. After 24 h, the colonies that grew were collected, and bla_SHV-5 abundance was quantified. A portion was subcultured into media with 16 µg/mL cefiderocol and 0.25 µg/mL clavulanate and grown for 24 h. The culture was plated on MHA containing 8 µg/mL cefiderocol and 4 µg/mL clavulanate. After 24 h, colonies that grew were collected, and bla_SHV-5 abundance was quantified. b PAP of RS plated on cefiderocol and 4 µg/mL clavulanate after growth in broth at each concentration of cefiderocol and clavulanate described in (a). a, b show the mean of results from two independent experiments with 5 biological replicates each, shown as individual dots in (a), standard deviation is shown in (b). For (a), * indicates p < 0.05 and ****p < 0.0001 by RM one-way ANOVA with Geisser-Greenhouse correction of log-transformed data, F (1.890, 17.01) = 288.1, between columns (cefiderocol/clavulanate), F (9, 27) = 1.342 between rows (paired replicates), with Sidak’s correction for multiple comparisons. Source data are provided as a Source Data file.

Fig. 5. β-lactamase amplification in a heteroresistant isolate from a patient with failed cefiderocol therapy.

a Population analysis profile of patient isolates on MHA containing cefiderocol, mean and standard deviation are shown from three independent experiments with n = 10 total. A patient was transferred to intensive care with septic shock and positive cultures with an NDM-5 metallo-β-lactamase producing K. pneumoniae. Disk diffusion testing identified the isolate as cefiderocol susceptible, and cefiderocol therapy was initiated but failed. b Depiction of the bla_NDM-5 encoding contig in the genome of the K. pneumoniae patient isolate. c bla_NDM-5 gene abundance following growth in broth containing 0 or 16 μg/mL cefiderocol, with median and interquartile range shown, of log10 transformed data. Each symbol indicates a biological replicate from two independent experiments, with n = 10 total for each group, two-tailed unpaired t-test of log10 transformed data, t = 9.420, df = 9.000, with Welch’s correction. Source data are provided as a Source Data file.

Fig. 6. Heteroresistance threatens the β-lactam/β-lactamase inhibitor development pipeline.

In vitro β-lactamase enzymatic assays determined that cefiderocol was resistant to extended-spectrum β-lactamases (ESBL) and most carbapenemases tested. Broth microdilution (BMD) antimicrobial susceptibility testing assigned low cefiderocol minimum inhibitory concentrations (MIC) to both susceptible (S) and heteroresistant (HR) isolates, the latter of which harbored such a low frequency of resistant cells (e.g., 1 in 100,000) that they did not alter the overall MIC. BMD could detect only conventional resistance (R) in which 100% of the cells exhibit phenotypic resistance. HR is expected to cause treatment failure in patients because cefiderocol treatment selects for the resistant subpopulations, which have amplifications of ESBL genes. These resistant cells become predominant during cefiderocol therapy and then can mediate treatment failure. In agreement with this model, unexpectedly high rates of treatment failure were observed in Phase III testing of cefiderocol, which correlated with rates of cefiderocol HR in surveillance studies. Taken together, the inability of the tests employed by the current antibiotic development pipeline (enzymatic assays and BMD) to detect HR leads drugs to which there are high rates of HR to progress to clinical testing, where treatment failures may be observed. Incorporating testing for HR in the antibiotic development pipeline could potentially make the process more efficient and efficacious.