Klebsiella pneumoniae ST307 with blaOXA-181, South Africa, 2014–2016

Klebsiella pneumoniae sequence type (ST) 307 is an emerging global antimicrobial drug–resistant clone. We used whole-genome sequencing and PCR to characterize K. pneumoniae ST307 with oxacillinase (OXA) 181 carbapenemase across several private hospitals in South Africa during 2014–2016. The South Africa ST307 belonged to a different clade (clade VI) with unique genomic characteristics when compared with global ST307 (clades I–V). Bayesian evolution analysis showed that clade VI emerged around March 2013 in Gauteng Province, South Africa, and then evolved during 2014 into 2 distinct lineages. K. pneumoniae ST307 clade VI with OXA-181 disseminated over a 15-month period within 42 hospitals in 23 cities across 6 northeastern provinces, affecting 350 patients. The rapid expansion of ST307 was most likely due to intrahospital, interhospital, intercity, and interprovince movements of patients. This study highlights the importance of molecular surveillance for tracking emerging antimicrobial clones.


Selection of Isolates
We selected K. pneumoniae isolates that tested positive by PCR for ST307 (n = 88) to represent different geographic locations, time periods, and specimens. The selection criteria were as follows: n = 5 from a private Alberton hospital A1 (
We tested the performance of the PCR methodologies first as a pilot study and then during implementation. When comparing WGS results with PCR results, the ST307, IncX3 and IS3000-OXA-181 PCR screening displayed 97% correlation with WGS results, furthering the evidence that the PCR approach was highly reliable.

Identification and Susceptibility Testing of Bacterial Isolates
We used matrix-assisted laser desorption ionization-time of flight mass spectrometry for identification testing (Vitek AMS; bioMérieux Vitek Systems Inc., https://www.biomerieuxusa.com) and VITEK 2 instrument (bioMérieux Vitek Systems) for susceptibility testing.

Pulsed-Field Gel Electrophoresis (PFGE)
Genetic relatedness of K. pneumoniae isolates was examined by comparing XbaI digested profiles generated by PFGE analysis using the standardized E. coli (O157:H7) protocol from the Centers for Disease Control and Prevention (CDC), Atlanta, GA, USA (6). The subsequent PFGE analyses were performed on a CHEF-MAPPER XA apparatus (Bio-Rad Laboratories, http://www.Bio-Rad.com). Gel images in tiff format were exported to BioNumerics software version 3.0 (Applied Maths, http://www.applied-maths.com) for analysis. Comparisons were made using the band-based Dice coefficient, which is a binary coefficient measuring similarity based upon common and different bands. Dendrograms were generated using Unweighted Pair Group Method using Arithmetic averages (UPGMA) method with 1.5% position tolerance. DNA relatedness was calculated based on the Dice coefficient and isolates were considered to be genetically related if the Dice coefficient correlation was >80%, which corresponds to the "possibly related (4-6 bands difference)" criteria of Tenover et al. (7). The blaOXA-181 plasmid detection and size estimation was performed using S1-nuclease linearization, followed by PFGE and Southern blot hybridization (8).
A dominant pulsotype with >80% similar PFGE profiles (named pulsotype A), was detected among 310/471 (66%) of OXA-48-like isolates (data not shown). Forty isolates exhibited >70% similarity of PFGE profiles to pulsotype A, suggesting that they were related to pulsotype A and were named AR. The remaining isolates were not clonally related, i.e., exhibited <60% similar PFGE profiles and did not show patterns similar to those of pulsotypes A or AR.
Root-to-tip correlations to test for a temporal signal in the data was performed using TempEST (9), and a significant positive slope for the alignment was present (p<0.001, R = 0.527). BEAST v2.4.7 (10) was used to estimate a timed phylogeny with concatenated recombination-free core SNP alignment. The GTR substitution model was selected based on evaluation of substitution models in bModelTest (11). BEAST analyses were performed using a coalescent constant population model and a coalescent Bayesian skyline model. In addition, a strict clock, with a lognormal prior, and a relaxed clock (both lognormal and exponential) were  Figure 2).

Identification of the blaOXA181 Harboring IncX3 Plasmids
The genome sequences were de novo assembled using plasmidSPAdes (13), followed by extraction of the blaOXA181-harboring contigs, using blastdbcmd. The blaOXA-181 contigs were "BLAST" against PlasmidFinder (8) database. The results showed that blaOXA-181 co-existed on the same contig as a truncated ColKp3 replicon gene, similar to that of p72_OXA181_X3.The identification of p72_OXA181_X3 were investigated using Bandage assembly graph viewer (14) and BLASTn. The plasmid region was identified by BLASTn searching for the p72_OXA181_X3 "core" genes, excluding insertion or repeated sequences.
Enterobacteriaceae with Carbapenemases other than K. pneumoniae from 2014-

16
Eighteen Enterobacter cloacae complex and 7 E. coli with carbapenemases were isolated at Ampath laboratories during 2014-2016. The majority of E. cloacae complex was positive for NDMs (n = 15) followed by OXA-48-like (n = 2) and KPC (n = 1); 2 of the E. coli was positive for NDM and 2 for OXA-48-like. WGS plasmid mining, S1-PFGE and Southern blotting showed that p72_X3_OXA181 was present among OXA-48-like Enterobacter cloacae complex and E. coli.

Infection Prevention and Control Measures
Routine auditing of hospital-acquired infections is part of continuous surveillance at the private hospitals. All carbapenem resistant microorganisms (CRE) detected at Ampath and satellite laboratories were communicated to the treating physician and to the infection prevention control (IP&C) practitioners on a daily basis. When an increase of CRE was noted, the affected hospital implemented CDC guidelines for CRE infection prevention and control (15). Outbreak response teams were identified within each hospital management during information sharing meetings with management and heads of departments. Infection prevention practitioners held weekly information sessions with all the wards and increased hand hygiene practices. Contact precautions training and education on donning and removal of gloves and aprons were monitored. Environmental cleaning was systematically actively pursued.
All patients within the same wards with newly laboratory confirmed CREs were screened using rectal swabs inoculated on chromID Carba (bioMérieux). All high-risk patients were screened on admission and cohorted immediately if identified. Active surveillance cultures consisted of stool, rectal swabs, ostomy output, wound swabs and endotracheal suctions as requested by the IPC team. Isolation wards were identified at each hospital and patients were cohorted. Strict access control to wards for staff and visitors was implemented. All patients were bathed with 2% chlorohexidine daily and when transferred in or out of the facility.