Brazilian Wastewater Harbors High-Risk Carbapenem-Resistant Superbugs, Genomes Reveal

Deep inside the pipes of four wastewater treatment plants in São Paulo state, Brazil, scientists have found something that should make clinicians and public health officials sit up: carbapenem-resistant bacteria carrying some of the most feared resistance genes in modern medicine, including one enzyme that appears to exist nowhere else on Earth. A new genomic study, published in Environmental Monitoring and Assessment, dissects the DNA of twelve carbapenemase-producing Enterobacterales recovered from raw sewage and reveals a convergence of resistance determinants, mobile genetic elements, and internationally notorious bacterial clones circulating outside hospital walls.

Carbapenems are last-resort antibiotics, deployed when other drugs fail against serious infections. When bacteria produce carbapenemases—enzymes that chemically dismantle these drugs—treatment options collapse to a handful of toxic or imperfect alternatives. The World Health Organization has classified carbapenem-resistant Enterobacterales among its critical priority pathogens, and in Brazil, Klebsiella pneumoniae alone accounts for a substantial share of carbapenem-resistant bloodstream infections. What makes the new findings alarming is not simply that resistant bacteria were found in wastewater, but which bacteria they were and what genetic weaponry they carried.

The research team, led by João Pedro Rueda Furlan of the Federal University of Paraíba and Universidade Federal de São Paulo, collected 500-milliliter samples of raw influent wastewater from four treatment plants between April and May 2024: Ribeirão Preto, Riacho Grande, Parque Novo Mundo, and Barueri. The samples were filtered, cultured in broth, and plated onto selective media containing meropenem and vancomycin to suppress ordinary susceptible organisms. Colonies were identified by mass spectrometry and then subjected to whole-genome sequencing on an Illumina platform, with draft genomes assembled de novo and interrogated using bioinformatic tools for resistance genes, plasmid replicons, virulence factors, and sequence types.

The twelve isolates comprised nine K. pneumoniae strains, two Enterobacter kobei, and one Enterobacter asburiae. Every single strain was multidrug resistant, with meropenem minimum inhibitory concentrations exceeding 64 milligrams per liter—a level far beyond what any clinically achievable drug concentration could overcome. Seventy-five percent were resistant to colistin, one of the few remaining salvage antibiotics, with minimum inhibitory concentrations ranging from 1 to greater than 64 milligrams per liter. Only ceftazidime-avibactam, a newer beta-lactam–beta-lactamase inhibitor combination, and tigecycline retained activity across the board. Phenotypic testing with boronic acid inhibitors pointed to serine carbapenemases, and genomic analysis confirmed the presence of blaKPC-2 in eleven of twelve strains, blaGES-5 in two E. kobei strains, and—remarkably—both blaKPC-2 and blaGES-16 cohabiting in a single E. kobei isolate.

The K. pneumoniae strains belonged to sequence types ST11 and ST307, clones recognized worldwide as high-risk lineages responsible for healthcare-associated outbreaks. ST11 appeared at multiple treatment plants and on different collection dates, indicating repeated environmental introductions rather than a single contamination event. Within ST11, two distinct capsular types were detected—KL64 and KL15—suggesting ecological plasticity and contributions from different source populations feeding into the sewage network. ST11-KL64 is typically associated with highly transmissible, virulent, carbapenemase-producing sublineages, while ST11-KL15 reflects broader environmental and community dissemination. ST307, found with capsular type KL102 at more than one plant, is strongly healthcare-associated yet demonstrably capable of spreading through urban wastewater systems.

The virulence profiles of these environmental isolates add a troubling dimension. All K. pneumoniae strains carried the mrk gene cluster encoding type 3 fimbriae, which mediate adhesion, colonization, and biofilm formation. The ST11 strains harbored the yersiniabactin siderophore locus embedded within integrative conjugative elements—ICEKp3 in the KL15 strain and ICEKp10 in the KL64 strain—machinery that enhances iron acquisition, immune evasion, and resistance to oxidative stress. Most strikingly, the ST11-KL64 strain also carried the colibactin gene cluster, specifically the clb3 variant, which encodes a genotoxin linked to DNA damage and disruption of epithelial barriers. The ST307-KL102 strains displayed the hypermucoviscous phenotype, an overproduction of capsule that shields the bacterium from host immune responses. In short, these are not attenuated environmental curiosities; they are bacteria with the genetic toolkit to cause serious human disease.

The Enterobacter isolates told a different but equally significant story. The E. kobei strains belonged to ST520, an international clone previously recovered from coastal waters in Brazil and from clinical samples in China, India, and Spain. The E. asburiae strain belonged to ST384, a lineage with no previous reports in the literature, suggesting that environmental genomic surveillance can uncover bacterial diversity that clinical sampling has entirely missed. Enterobacter species generally rely on basic fitness traits rather than specialized virulence systems, and the E. kobei strains carried no classical virulence determinants, while the E. asburiae strain carried adhesion and iron-acquisition genes consistent with an opportunistic lifestyle.

Beneath the species-level findings lies the real engine of concern: the mobile genetic elements shepherding the carbapenemase genes from cell to cell. In most K. pneumoniae strains, blaKPC-2 sat within the Tn4401a transposon, the classical vehicle that has driven the global spread of KPC-type resistance. One strain, however, carried blaKPC-2 within the Tn4401i isoform, an uncommon variant previously reported in Brazilian clinical and environmental K. pneumoniae but whose epidemiological risk remains unclear. In the Enterobacter strains, blaKPC-2 was embedded in structures involving ISKpn27 and ISKpn6 insertion sequences alongside Tn3-family transposases—platforms that may represent early stages of mobilization pathways before stabilization in clinically dominant vehicles. Meanwhile, blaGES-5 and blaGES-16 were both nestled within class 1 integrons, evolutionary capture systems whose intI1 integrase can seize, excise, and rearrange gene cassettes, promoting the accumulation and horizontal transfer of resistance genes across bacterial hosts and ecological niches.

The blaGES-16 gene deserves particular attention. GES-16 differs from GES-5 by a single amino acid substitution, glutamine 38 replaced by glutamic acid, yet it functions as a carbapenemase. First described in clinical Serratia marcescens from Rio de Janeiro and later found in Aeromonas veronii from hospital wastewater in Paraná, the gene appears to be geographically restricted to Brazil. A search of public genome databases conducted in January 2026 identified only five genomes carrying blaGES-16, all from Brazilian sources, spanning Aeromonas hydrophila, A. veronii, E. kobei, Enterobacter soli, and Pseudomonas aeruginosa from different states. Three of these genomes also carried KPC-2. In the new study, blaGES-16 was found on both chromosomal and plasmid contigs, consistently associated with class 1 integron machinery and flanked by additional resistance genes, including blaOXA-17, which encodes an extended-spectrum beta-lactamase. This integron association suggests the gene is being actively maintained and potentially mobilized within Brazilian bacterial populations.

The broader resistome of the isolates reinforced the picture of comprehensive multidrug resistance. All strains carried acquired and intrinsic genes conferring resistance to beta-lactams, fluoroquinolones, aminoglycosides, tetracyclines, sulfonamides, and phenicols. Extended-spectrum beta-lactamase genes blaCTX-M-2 and blaCTX-M-15 were present, along with rmtB, a 16S rRNA methyltransferase that obliterates the activity of aminoglycosides such as amikacin. Fluoroquinolone-resistant strains harbored the canonical GyrA S83I and ParC S80I mutations, while colistin-resistant K. pneumoniae strains carried multiple mutations in polymyxin resistance determinants, including disruption of the mgrB gene by an ISKpn14 insertion element and a deleterious R256G substitution in the PmrB sensor kinase. Genes conferring tolerance to quaternary ammonium disinfectants and to arsenic, copper, mercury, and silver were also detected, hinting at selection pressures beyond antibiotics alone. Plasmid replicons of the IncF type predominated, with IncN-pST15 and IncP6 also identified—plasmid types previously documented carrying blaKPC-2 in environmental K. pneumoniae and E. coli from southeastern Brazil.

The authors frame their findings within a One Health perspective: wastewater treatment plants are not merely passive conduits but active reservoirs and mixing vessels where clinically relevant resistance determinants, high-risk clones, and mobile genetic elements converge and evolve. When treatment is inadequate, resistant bacteria and their gene cargo can be discharged into receiving waters, bridging the clinical-environmental interface. Raw wastewater, the authors argue, functions as an early warning signal of high-risk lineages circulating within the human population—detectable before those lineages necessarily appear in hospitals. Continuous genomic surveillance in extra-hospital settings, they conclude, is essential to monitor emerging resistance genotypes and catch the early signs of clinically relevant dissemination. The complete genome sequences are publicly available under BioProject PRJNA1216268, offering researchers worldwide a resource for tracking these Brazilian lineages as the story of antimicrobial resistance continues to unfold at the intersection of human health and the environment.

Subject of Research: Genomic surveillance of carbapenemase-producing Enterobacterales in Brazilian wastewater

Article Title: Genomic characterization of carbapenemase-producing Enterobacterales from wastewater reveals the convergence of KPC-2 and GES-16 in Brazil

Article References: Furlan, J. P. R., Bueno, G. C., Sousa-Carmo, R. R., Silva, R. L. O., Barbosa, M. R. F., Sato, M. I. Z., Lincopan, N., & Schenkman, S. (2026). Genomic characterization of carbapenemase-producing Enterobacterales from wastewater reveals the convergence of KPC-2 and GES-16 in Brazil. Environmental Monitoring and Assessment, 198(9), Article 1014. https://doi.org/10.1007/s10661-026-15867-0

Image Credits: AI Generated

DOI: 10.1007/s10661-026-15867-0

Keywords: antimicrobial resistance, carbapenemase-producing Enterobacterales, Klebsiella pneumoniae, wastewater surveillance, KPC-2, GES-16, mobile genetic elements, class 1 integrons, whole-genome sequencing, One Health, Brazil, high-risk clones