NDM‑1–carrying Acinetobacter in Hospital Wastewater: Genomic Findings from Shenzhen

A culture-based genomic survey of pre-disinfection hospital wastewater from Shenzhen, China, reports recovery of carbapenem-resistant Acinetobacter isolates in which blaNDM-1 was a consistent feature.

Across a one-year sampling frame spanning five tertiary hospitals, investigators collected 60 24-hour composite wastewater samples and carried 34 blaNDM-1-positive Acinetobacter isolates forward for detailed characterization. The article reports that all 34 of these isolates carried blaNDM-1. The authors then organize results around blaNDM-1 mobilization, Acinetobacter group composition, and genomic relatedness across hospitals and seasons.

Sampling focused on untreated, pre-disinfection wastewater using 24-hour composite collection at multiple hospitals, followed by selective culture and downstream whole-genome sequencing as described by the investigators. Susceptibility testing was presented alongside genomic findings: the study reports meropenem MIC50/90 values of 32/64 mg/L and imipenem MIC50/90 values of >128/>128 mg/L in the analyzed Acinetobacter set. For last-line agents, the article also reports that 21% of isolates (7/34) were categorized as colistin resistant. In the authors’ description, this phenotypic profile aligns with a multidrug-resistant characterization for the wastewater-derived isolates.

Genomic localization analyses in the article place blaNDM-1 on plasmids in most isolates, with 29 of 34 (85.3%) reported as plasmid-borne and the remainder chromosomal. Among the blaNDM-1 plasmid vehicles, the authors report predominance of Rep_3 family plasmids, highlighting R3‑T28 as the most frequently observed subtype. These plasmid-focused analyses also describe co-occurrence of blaOXA-58 on blaNDM-1–positive plasmids in 18 of 29 cases (62.1%) and note that the pairing was most often observed on the R3-T28 backbone. The authors describe these recurring plasmid backbones as a repeated vehicle within this wastewater collection.

Despite the shared carbapenemase signal, the authors describe a heterogeneous Acinetobacter population: the 34 isolates were assigned to 11 species-level groups, with A. junii (10 isolates) and the A. towneri group (9 isolates) reported as the most prevalent categories. Relatedness was explored with core‑genome SNP analysis, which the article uses to highlight pairs of highly similar isolates detected across hospitals within a season (1–2 SNPs) as well as a closely related pair within one hospital sampled across seasons (19 SNPs). The authors interpret these patterns as spatio-temporal intermixing of related lineages in wastewater, alongside multi-species recovery.

At the sequence-context level, the article describes blaNDM-1 as embedded in a conserved ISAba125-associated, Tn125‑like core with variable flanking insertion sequences, and provides examples of neighboring genes that recur near the locus, including ble, trpF, and groES/groEL. In their discussion and conclusions, the authors describe hospital wastewater as a potential hotspot for dissemination of carbapenem resistance and state that their findings support routine genomic surveillance under a One Health framework.

Key Takeaways:

• The article reports universal blaNDM-1 carriage among the 34 carbapenem-resistant Acinetobacter isolates analyzed from pre-disinfection hospital wastewater.
• Most blaNDM-1 was reported as plasmid-borne, with Rep_3 plasmids (including R3-T28) representing the predominant backbones in this collection; R3-T28 was the most frequent blaNDM-1 vehicle among plasmid-borne isolates (15/29, 51.7%).
• The dataset combined multi-species recovery, close SNP relatedness signals, and a conserved ISAba125/Tn125-like blaNDM-1 context, which the authors discuss within a One Health genomic surveillance framing.