Modification of the penicillin-binding-protein 3 as a source of resistance to broad-spectrum cephalosporins in Escherichia coli

Multidrug-resistant Gram-negative bacteria are one of the major threats to human health, largely due to the dissemination of some high-risk successful multidrug-resistant clones.¹ Of particular concern is the resistance to ß-lactams, because those antibiotics are the most extensively used antimicrobial ones, accounting for half of all prescriptions in Europe.² Resistance to broad-spectrum cephalosporins is of particular concern, as related infections require usage of carbapenems considered as last-resort antibiotics supposed to be preserved. Although the production of β-lactamases is the most common mechanism of resistance to broad-spectrum β-lactams in Gram-negative bacteria, and particularly in Enterobacterales, other mechanisms have also been reported including efflux, permeability changes and modification of PBP targets that contribute to this resistance.³

Our study was initiated by the isolation in March 2023 of an Escherichia coli strain recovered from a urine sample of an 86-year-old female patient hospitalized in Zürich, Switzerland, suffering from recurrent urinary tract infections. Corresponding antibiotic regimen were either cotrimoxazole or fosfomycin, but no ß-lactam has been used. Susceptibility testing by disc diffusion and broth microdilution showed that this isolate was resistant to broad-spectrum cephalosporins (ceftazidime, cefotaxime, cefepime) and to aztreonam, but remained susceptible to ceftazidime-avibactam and carbapenems, according to the EUCAST 2023 breakpoints.⁴ In addition, it was resistant to fluoroquinolones, nitrofurantoin, cotrimoxazole and fosfomycin.

Multilocus sequence typing performed as described in ref.⁵ identified that strain as belonging to ST1193, that has been reported as an emerging high-risk multidrug-resistant global clone.¹ Resistance to broad-spectrum cephalosporins (ceftazidime, cefotaxime, cefepime) could not be explained by common mechanisms of resistance such as broad-spectrum ß-lactamases. Indeed, the Rapid ESBL NP test remained negative and PCR assays for all ESBL-encoding genes also remained negative.⁶ Likewise, PCR for all known AmpC β-lactamase encoding genes remained negative.

Whole-genome sequencing was therefore performed, using a short-read technology (MiSeq platform, Illumina, San Diego, CA, USA),⁷ and allowed the identification of mutations in gyrA/parC (gyrA S83L, D87N; parC S80I), responsible for fluroquinolone resistance,¹ and an 8-bp insertion and subsequent disruption of nfsA, conferring resistance to nitrofurantoin.⁸ Regarding β-lactam resistance, the gene encoding the narrow-spectrum TEM-1 was the sole ß-lactamase gene identified, that explained the resistance to penicillins but not to the broad-spectrum cephalosporins.⁹ However, a four amino-acid insertion (YRVP) just after the amino-acid residue 333 (close to the β-lactam binding pocket) of the PBP-3 sequence was identified.

To investigate the role of this PBP-3 insertion in the resistance to cephalosporins, a wild-type PBP-3 was amplified from E. coli ATCC25922 reference strain using primers PBP3_F (5′-CCACGGAAAAGCTGCAAATG-3′) and PBP3_R (5′-CATCGGTCGCCTCATCTTTC-3′) and cloned into plasmid pCR-Blunt II-TOPO,⁷ then the corresponding recombinant plasmid transformed into the E. coli clinical isolate. Hence, recovery of susceptibility to broad-spectrum cephalosporins was observed in the complemented strain (Table 1).

Table 1.

MICs of β-lactams and non-β-lactams for the clinical strain, clinical strain with an empty vector (pTOPO) and clinical strain complemented with a recombinant plasmid possessing a gene encoding a non-mutated PBP-3 (pPBP-3)

Mic (µg/mL)	N3940	N3940 (PTOPO)	N3940 (PPBP-3)
TIC	>256	>256	>256
AMX	>256	>256	>256
TEM	128	128	64
CAZ	8	8	1
CZA	8	8	2
CTX	4	4	1
FEP	4	4	1
FOX	128	128	128
FDC	0.25	0.25	0.25
ATM	8	8	1
AZA	8	8	1
ETP	<0.06	<0.06	<0.06
MEM	<0.06	<0.06	<0.06
IMI	<0.06	<0.06	<0.06
NTF	>256	—	—
CIP	64	—	—
KAN	1	—	—

TIC, ticarcillin; AMX, amoxicillin; TEM, temocillin; CAZ, ceftazidime; CTX, cefotaxime; FEP, cefepime; FOX, cefoxitin; FDC, cefiderocol; CZA, ceftazidime-avibactam; AZA, aztreonam-avibactam; ATM, aztreonam; ETP, ertapenem; MEM, meropenem; IMI, imipenem; NTF, nitrofurantoin; CIP, ciprofloxacin; KAN, kanamycin.

—, not evaluated.

A similar PBP-3 modification has been recently evidenced among NDM-1 and NDM-5 E. coli isolates already been reported,^1,10 with either YRIN or YRIK 4-amino-acid insertions also identified after position 333. It was shown that these PBP-3 modifications were responsible for resistance to aztreonam and consequently also to the aztreonam-avibactam combination, which is a significant concern as those isolates were co-producing NDM-like carbapenemases along with AmpC-type CMY ß-lactamases, resulting in high-level resistance to all ß-lactams including carbapenems.

Overall, we showed that the YRVP insertion in the PBP-3 was involved in the decreased susceptibility to cephalosporins, such as ceftazidime, cefotaxime, cefepime and aztreonam, a phenotype observed without production of any acquired broad-spectrum ß-lactamase. This implies that recovery of an increased susceptibility cannot be expected when considering combinations made of ß-lactam/ß-lactamase inhibitor such as ceftazidime-avibactam, as highlighted here with an MIC at 8 mg/L either for ceftazidime and for ceftazidime-avibactam (a value being at the exact EUCAST breakpoint for the latter, whereas that of ceftazidime is actually at 1 mg/L). Zhang et al. previously highlighted that PBP-3 modifications so far described did not confer resistance to CAZ-AVI but recommended accurate monitoring.¹¹ Our observations are fully in accordance with this warning. In addition, the isolate investigated here belonged to the ST1193, recently considered to be an emerging high-risk multidrug-resistant global clone responsible for urinary tract and bloodstream infections.¹

Funding

This work was funded by the University of Fribourg, Switzerland.

Transparency declarations

None to declare.