Introduction
Toxigenic Corynebacterium diphtheriae is the main aetiological agent of diphtheria, which manifests itself as either respiratory or cutaneous disease. Respiratory diphtheria is life-threatening; it is characterised by fever, neck swelling, and sore throat, which can lead to breathing and swallowing difficulties. Cutaneous diphtheria results in non-healing ulcer-like skin lesions and can act as a source of respiratory diphtheria in vulnerable people [Reference Truelove, Keegan, Moss, Chaisson, Macher, Azman and Lessler1]. Until the establishment of widespread vaccination, diphtheria posed a great threat to public health in Australia [Reference Meinel, Kuehl, Zbinden, Boskova, Garzoni, Fadini, Dolina, Blümel, Weibel, Tschudin-Sutter, Widmer, Bielicki, Dierig, Heininger, Konrad, Berger, Hinic, Goldenberger, Blaich, Stadler, Battegay, Sing and Egli2–Reference Timms, Nguyen, Crighton, Yuen and Sintchenko5]. The C. diphtheriae vaccination utilises the diphtheria toxoid and helps in protection against toxin-associated clinical disease. In Australia, vaccines containing the diphtheria toxin are included in the National Immunisation Program, where doses are funded for children at 2, 4, 6, and 18 months, and again at 4 years of age. Further doses are provided in adolescence as part of school immunisation programmes, as well as for pregnant women, but, although recommended, these are not funded for adults. Vaccination does not prevent the carriage of C. diphtheriae but is estimated to reduce onward transmission by 60% and may reduce the risk of severe disease [Reference Truelove, Keegan, Moss, Chaisson, Macher, Azman and Lessler1].
In 2022, outbreaks of both cutaneous and respiratory diphtheria cases caused by C. diphtheriae were reported in some European countries [Reference Kofler, Ramette, Iseli, Stauber, Fichtner, Droz, Zihler Berner, Meier, Begert, Negri, Jachmann, Keller, Staehelin and Grützmacher6, Reference Badenschier, Berger, Dangel, Sprenger, Hobmaier, Sievers, Prins, Dörre, Wagner-Wiening, Külper-Schiek, Wichmann and Sing7]. Genomic analysis illustrated several multi-locus sequence type (MLST) profiles amongst isolates in these outbreaks, suggesting that they are not associated with a single clonal event but rather linked to risk factors including migrant movement and low vaccination rates in particular populations. Similarly, an increase in diphtheria cases caused by multiple genetic lineages of C. diphtheriae since 2017 in Yemen has also been reported, where vaccination efforts were disrupted by civil war [Reference Badell, Alharazi, Criscuolo, KAA, Lefrancq, Bouchez, Guglielmini, Hennart, Carmi-Leroy, Zidane, Pascal-Perrigault, Lebreton, Martini, Salje, Toubiana, Dureab, Dhabaan and Brisse8].
As a result of ongoing vaccination efforts in Australia, diphtheria is rare, with generally fewer than 10 cases notified each year. Typically, cases are sporadic and associated with acquisition outside of Australia. There have been fewer than 10 cases of respiratory diphtheria reported in Australia since 2001, with two fatal cases occurring in 2011 and 2018 in the state of Queensland [Reference Winkler, Dey, Quinn, Pourmarzi, Lambert, McIntyre and Beard4, Reference Forde, Henderson, Playford, Looke, Henderson, Watson, Steen, Sidjabat, Laurie, Muttaiyah, Nimmo, Lampe, Smith, Jennison, McCall, Carroll, Cooper, Paterson and Beatson9, Reference Grigg, Hogan, Hosein, Johns, Jennison and Subedi10]. The detection of the toxin gene in C. diphtheriae isolates is notifiable under Queensland public health guidelines [11]; however, non-toxin gene-bearing C. diphtheriae is regularly isolated from patients, primarily from wound swabs but also from throat swabs where it is a commensal organism [Reference Timms, Nguyen, Crighton, Yuen and Sintchenko5, Reference Doyle, Mazins, Graham, Fang, Smith and Jennison12]. Whilst occasional detections of toxin gene-bearing C. diphtheriae were reported previously, since 2020, North Queensland (NQ) (defined as the region of Queensland north of the tropic of Capricorn) has seen an increase in locally acquired toxin gene-bearing diphtheria cases, with a significant surge during 2022 [Reference Hempenstall, Short, Marquardt, Fisher and Johnson13]. Between January and August 2022, 21 patients with a residential address in NQ had toxin gene-positive C. diphtheriae isolated compared to 13 patients in the preceding 5 years (2017–2021). A further three interstate cases in 2022 were identified as having epidemiological links to the region.
Here we report on sequence typing of toxin gene-positive C. diphtheriae isolates collected in Queensland between January 2017 and August 2022 and investigate the genetic diversity between them and non-toxin gene-bearing isolates recovered in the NQ region during this period.
Materials and methods
Isolate collection and DNA extraction
The clinical isolates included in this study are listed in Supplementary Table S1. These comprised a selection of 131 C. diphtheriae isolates received by the Queensland Health Public Health Microbiology Reference Laboratory between January 2017 and August 2022, including 51 toxin gene-bearing isolates from across Queensland and 80 non-toxin gene-bearing C. diphtheriae from the NQ region. A single toxin gene-bearing ST381 isolate from 2013 was also included in the analysis, bringing the study sample to 132 isolates in total. DNA was extracted from isolates grown overnight at 37°C on horse blood agar (Edwards Group Holdings, Murrarie, Australia), using the QiaSymphony DSP DNA Mini kit (Qiagen, Hilden, Germany), according to the manufacturer’s protocol.
DNA library and whole genome sequencing
Whole genome sequencing was performed on the isolates listed in Supplementary Table S1 using the procedure described in our previous studies [Reference Graham, Hiley, Rathnayake and Jennison14, Reference Pintara, Jennison, Rathnayake, Mellor and Huygens15]. Briefly, DNA was prepared for sequencing using the Nextera XT kit (Illumina, San Diego, USA) and sequenced on the NextSeq500 using the NextSeq 500 Mid Output v2 kit (300 cycles) (Illumina), according to the manufacturer’s instructions. Sequence reads for the C. diphtheriae isolates were trimmed with Trimmomatic v0.36 [Reference Bolger, Lohse and Usadel16] and quality checked by FastQC v0.11.5 and MultiQC v1.1 [Reference Wingett and Andrews17]. Sequences were de novo assembled into contigs using the SPAdes assembler v3.12.0 [Reference Bankevich, Nurk, Antipov, Gurevich, Dvorkin, Kulikov, Lesin, Nikolenko, Pham, Prjibelski, Pyshkin, Sirotkin, Vyahhi, Tesler, Alekseyev and Pevzner18]. Sequence files have been submitted to the European Nucleotide Archive under project accession number PRJEB58646.
Data analysis
Phylogenetic analysis was performed using single nucleotide polymorphism (SNP)-based typing. Core SNPs were determined for the ST381 reference strain CD38 (SRR6816591, pubMLST id914) with Snippy v4.3.6, using the default settings (available at https://github.com/tseemann/snippy). Core SNPs from each sample were aligned using Snippy-core. A Maximum Likelihood tree was generated from the SNP alignment, using Fast Tree v2.1.10. Interactive tree of life (iTOL) v5.6 was used for the visualization of the phylogenetic tree [Reference Letunic and Bork19]. MLST was performed in RidomSeqSphere+8.4.1 (Ridom, Munster, Germany) according to the scheme described in [Reference Bolt, Cassiday, Tondella, DeZoysa, Efstratiou, Sing, Zasada, Bernard, Guiso, Badell, Rosso, Baldwin and Dowson20], available at https://bigsdb.pasteur.fr/diphtheria/. Abricate (https://github.com/tseemann/abricate) was used to search for the presence of acquired antimicrobial resistance genes using the resfinder database [Reference Chen, Zheng, Liu, Yang and Jin21]. Penicillin-binding protein (pbp) genes were investigated using the allele scheme available at https://bigsdb.pasteur.fr/diphtheria/, and by manual interrogation of the pbp2B gene for mutations described in [Reference Hennart, Panunzi, Rodrigues, Gaday, Baines, Barros-Pinkelnig, Carmi-Leroy, Dazas, Wehenkel, Didelot, Toubiana, Badell and Brisse22].
Results
Of the 51 toxin gene-positive isolates from all Queensland cases between January 2017 and August 2022, 42 (82.3%) were recovered from wound swabs and 9 (17.6%) from throat swabs. Ten (19.6%) of the toxin gene-positive strains were thought to be acquired from outside of Australia based on provided travel history, while the remainder were from cases that had not reported any travel outside of Australia. A single isolate with ST379 had a non-functional toxin gene as previously described [Reference Doyle, Mazins, Graham, Fang, Smith and Jennison12].
In early 2022, an increase in toxin gene-bearing C. diphtheriae cases associated with NQ was observed, with 24 cases identified between January and August 2022 compared to 6 cases in the region in each of 2020 and 2021. The majority of isolates in 2022 were from wound swabs (n = 18), with six from throat swabs. Of these six patients, two presented with classic respiratory diphtheria symptoms, including toxin-mediated complications such as formation of pseudo-membrane, myocarditis, or bulbar palsy, and were treated with antitoxin, while four presented with mild respiratory illness. All 24 cases had epidemiological links to NQ during their exposure period; no international travel nor links to overseas travellers were identified on investigation and all cases were therefore considered locally acquired. Fourteen of the 24 cases were detected following public health interventions; 2 were close contacts and 1 was a casual contact (identified through contact tracing), with another 11 cutaneous cases detected through enhanced clinical surveillance in an area where cases had occurred.
MLST analysis
MLST analysis of the 132 C. diphtheriae isolates (80 non-toxin gene-bearing isolates from the NQ region and 52 toxin gene-bearing isolates from across Queensland) identified 27 different sequence types (STs). Toxin gene-positive strains isolated prior to 2020 exhibited a range of different STs; however, from January 2020 to August 2022, 29/34 (88.3%) of the toxin gene-positive isolates were ST381. This ST had been relatively rare in Queensland before this time, with only three representatives being isolated prior to 2020, and all were associated with travel to islands in the southwestern Pacific. Three ST381 isolates were identified in 2020 and three in 2021 (Supplementary Table S1). Numbers of ST381 C. diphtheriae increased markedly in 2022 with 23 isolated between January and August 2022. All of these ST381 isolates were recovered from patients who either were residents of NQ or had epidemiological links to this region.
Phylogenetic analysis
When all sequenced isolates were analysed by SNP analysis, clustering based on STs was observed, with the ST381 clone forming a group distinct from the other STs included in the analysis (Figure 1). The analysis showed that the ST381 isolates from 2020 to 2022 were more closely related to each other than to ST381 cases from before 2020, with more than 120 SNPs present between the pre-2020 isolates and the 2020–2022 isolates (Figure 2). Among the 2020–2022 ST381 isolates, there were fewer than 20 SNPs present between isolates, with no temporal differences in clustering (isolates from different years showed the same number of SNPs between them as isolates from the same year). Further analysis of SNP differences found that none of the other ST groups were as tightly clustered as the ST381 group.
Maximum likelihood tree built using SNP differences between Corynebacterium diphtheriae isolates. Colours of leaves represent the year isolated and white stars on leaves indicate that the isolate was from the NQ region. ST and AMR genes are indicated to the right of the tree. Branch length represents genetic distance as indicated by the scale bar.
Maximum likelihood tree built using SNP differences between ST381 Corynebacterium diphtheriae isolates. Colours of leaves represent the year isolated and white stars on leaves indicate that the isolate was from the NQ region. Branch length represents genetic distance as indicated by the scale bar.
The C. diphtheriae MLST Institut Pasteur database contains information from two isolates of ST381, one being a Queensland isolate from 2013 included in this study and the other being grown from a cutaneous case in the state of New South Wales (NSW), Australia, from 2015 (id914 used as the reference in the phylogenetic analysis). BURST analysis of the ST381 profile in the C. diphtheriae MLST Institut Pasteur database revealed that the closest sequence type to ST381 is the single locus variant (SLV) ST461. Only one ST461 is in the MLST database – a cutaneous case from 2016 also from NSW.
Other than ST381, the most common ST of isolates from the NQ region was ST39. These strains were toxin gene negative but have been increasingly isolated since 2018, with numbers rising from one in 2018 to eight in 2022 (Figure 3).
Numbers of isolates with different STs by year in (a) toxin gene-positive and (b) toxin gene-negative isolates. STs are shown on the columns as well as in the legend for clarity.
The prevalence of different STs among toxin gene-positive strains detected in Queensland appears to have changed between 2017 and 2022, with some STs present prior to 2022 becoming less frequent, and a decrease in the diversity of STs seen in 2021 and 2022 when compared to previous years (Simpson’s diversity index of 0.07 for 2021–2022 compared to 0.88 for 2017–2020). ST381 was the only lineage among toxin gene-bearing strains between January 2021 and August 2022, accounting for all but one of 27 (96.3%) such strains from the NQ region (Figure 4).
Prevalence of STs per year for toxin gene-bearing Corynebacterium diphtheriae isolates. Circle size represents number of isolates and colours represent ST.
AMR genes
Screening of the sequences for acquired AMR genes identified four isolates with sul1, and one isolate with aph(3′), aph(6)-Id, cmx1 and sul1 (Figure 1).
Screening of sequences for pbp revealed the presence of pbp1A, pbp1B, pbp2A, pbp2B, pbp2C, and pbp4 with alleles as shown in Supplementary Table S1. The pbp2B mutation V535I and the pbp2m gene were not found in any of the isolates.
Discussion
Phylogenetic analysis of C. diphtheriae from the NQ region between January 2017 and August 2022 has demonstrated that the genetic population of C. diphtheriae in this region has changed over time, with STs such as N01 and N05 decreasing and ST39 increasing in the non-toxin gene-bearing C. diphtheriae population, and ST381 emerging as dominant in the toxin gene-bearing population. Isolates of both ST39 and ST381 increased markedly in 2022. At a genomic level, ST39 is not closely related to ST381; so despite their increase in numbers at the same time in the same region, it is unlikely that they are genetically linked, and these increases possibly represent changes in the bacterial population over time. This change is potentially in response to a decrease in the introduction of international strains due to public health interventions in Australia during the COVID-19 pandemic, including border restrictions on international arrivals, which may have allowed local strains to expand. However, as we have only sequenced the non-toxigenic strains from NQ, we cannot comment on whether similar increases in ST39 have occurred in other regions of Queensland, or more widely in Australia.
The STs seen in the toxin gene-positive isolates included in this study differ from the globally dominant STs reported, and in contrast to the increases in C. diphtheriae incidence reported in other countries, which have been polyclonal in nature with a range of STs being reported [Reference Meinel, Kuehl, Zbinden, Boskova, Garzoni, Fadini, Dolina, Blümel, Weibel, Tschudin-Sutter, Widmer, Bielicki, Dierig, Heininger, Konrad, Berger, Hinic, Goldenberger, Blaich, Stadler, Battegay, Sing and Egli2, Reference Kofler, Ramette, Iseli, Stauber, Fichtner, Droz, Zihler Berner, Meier, Begert, Negri, Jachmann, Keller, Staehelin and Grützmacher6–Reference Badell, Alharazi, Criscuolo, KAA, Lefrancq, Bouchez, Guglielmini, Hennart, Carmi-Leroy, Zidane, Pascal-Perrigault, Lebreton, Martini, Salje, Toubiana, Dureab, Dhabaan and Brisse8, Reference Guglielmini, Hennart, Badell, Toubiana, Criscuolo and Brisse23], the increased incidence of toxin gene-positive C. diphtheriae in NQ appears primarily linked to the clonal expansion of ST381 alone. The reason for this is not known; however, the tropical environment of NQ and social determinants, including household overcrowding in the areas where most diphtheria cases have been reported, may be contributing to transmission. Public health interventions in the NQ region in response to the ST381 clonal expansion are continuing.
All three ST381 cases identified in Queensland prior to 2020 were associated with travel to the southwest Pacific islands. Travel history is not documented for the single ST381 case listed in the C. diphtheriae MLST Institut Pasteur database, nor for the isolate of the nearest sequence type, ST461, but both originated from NSW, Australia, in 2015–2016. Our phylogenetic analysis suggests that during or prior to 2020, a particular clone of ST381, which does not appear closely genetically linked to any of the previous ST381 detections, has rapidly expanded in the NQ region, resulting in a cluster of locally acquired toxin gene-bearing cutaneous and respiratory diphtheria cases. There is no evidence of sustained circulation of toxigenic C. diphtheriae in Australia since the early 1990s [Reference Patel, Morey, Butcher, Moore, Brennan and Mollison24].
A difference in ST prevalence was seen between the NQ non-toxin gene-bearing strains and those reported in another Australian study, where ST32 was the most commonly reported non-toxin gene-bearing ST in the state of NSW, which is located to the south of Queensland [Reference Timms, Nguyen, Crighton, Yuen and Sintchenko5]. It may be that in the more urban southern region of Queensland, the C. diphtheriae population is similar to that seen in the southern Australian states and this would make an interesting future study.
Toxin gene-positive and non-toxin gene-bearing strains tended to fall into different STs (Figure 1); however, ST849 and ST516 contained both toxigenic and non-toxigenic strains. This may represent loss or acquisition of the diphtheria toxin phage by these strains, but overall, the genomic results indicate that for the most part, toxigenic and non-toxigenic strains are genomically distinct. The presence of the toxin gene in all ST381 isolates included in this study suggests that clonal expansion of this strain across NQ between 2020 and 2022 did not arise from the local acquisition of the toxin phage by a non-toxigenic strain. Although the production of diphtheria toxin by the toxin gene-positive isolates was not confirmed phenotypically as the Elek test is no longer carried out by any Australian laboratory, a correlation between the presence of the toxin gene and the production of the toxin has been described [Reference Hauser, Popoff, Kiredjian, Boquet and Bimet25].
The presence of antimicrobial resistance markers was not common in the strains analysed, with only four containing genes associated with resistance, all of which were toxin gene positive and from cases who had reported travel. One of these strains (M1911202 and ST469 (Supplementary Table S1)) contained multiple resistance genes, associated with resistance to macrolides, chloramphenicol, and sulfonamides; however, unfortunately, information on the potential phenotypic resistance was not available. This level of antimicrobial resistance is lower than that seen in other countries [Reference Hennart, Panunzi, Rodrigues, Gaday, Baines, Barros-Pinkelnig, Carmi-Leroy, Dazas, Wehenkel, Didelot, Toubiana, Badell and Brisse22, Reference Husada, SDP, Kurniawati, Hendrata, Irawan, Kartina, Puspitasari, Basuki and Ismoedijanto26]. In 2011, there was a fatal case of respiratory C. diphtheriae in Queensland caused by a penicillin-resistant strain carrying the pbp2m gene [Reference Forde, Henderson, Playford, Looke, Henderson, Watson, Steen, Sidjabat, Laurie, Muttaiyah, Nimmo, Lampe, Smith, Jennison, McCall, Carroll, Cooper, Paterson and Beatson9, Reference Grigg, Hogan, Hosein, Johns, Jennison and Subedi10, Reference Hennart, Panunzi, Rodrigues, Gaday, Baines, Barros-Pinkelnig, Carmi-Leroy, Dazas, Wehenkel, Didelot, Toubiana, Badell and Brisse22]. Although several pbp genes were found in the isolates analysed in this study, none had the pbp2m gene present or the V535I mutation in the pbp2B gene that is associated with resistance to penicillin [Reference Hennart, Panunzi, Rodrigues, Gaday, Baines, Barros-Pinkelnig, Carmi-Leroy, Dazas, Wehenkel, Didelot, Toubiana, Badell and Brisse22].
This study reports on a spatially and temporally related increase in toxin gene-positive C. diphtheriae cases in NQ, first noted in 2021 by both relevant public health authorities and the Queensland microbiology reference laboratory. A source of possible bias of this study is that some of the recent increase in toxin gene-positive C. diphtheriae isolates may have been driven by the increased health surveillance in this region leading to an increase in testing and reporting of diphtheria, and increased isolation of diphtheroids that may have previously been dismissed as skin flora.
Conclusions
In conclusion, this genomic analysis of C. diphtheriae has demonstrated that an increase in cases of toxin gene-bearing C. diphtheriae in NQ is likely due to an increase in one ST, ST381. This group clusters closely together with higher genetic relatedness amongst all 2020–2022 ST381 isolates than to ST381 isolates from previous years and is not closely related to non-ST381 isolates or any of the non-toxin gene-bearing strains. Thus, this likely represents the clonal expansion of a toxin gene-bearing strain that has moved into this region rather than the result of an already endemic non-toxigenic strain acquiring the toxin gene.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S0950268823000699.
Data availability statement
Sequence files have been submitted to the European Nucleotide Archive under project accession number PRJEB58646.
Acknowledgments
Thanks to the Queensland Public Health Microbiology Reference laboratory for their assistance with preparing the C. diphtheriae isolates used in this study. We would also like to thank all laboratories that referred the isolates used in this study and the Queensland Health Public Health Units involved in the public health management of these cases.
Author contribution
Conceptualization: A.J., I.U.R., S.S., R.M.G.; Resources: A.J.; Supervision: A.J.; Writing – original draft: A.J., T.M., A.H., C.T., R.M.G.; Writing – review & editing: A.J., T.M., V.F., A.H., C.T., I.U.R., M.B., S.S.; Formal analysis: T.M., A.H., C.T.; Investigation: T.M., V.F., A.H., C.T., M.B., R.M.G.; Data curation: V.F., A.H., C.T., R.M.G.; Methodology: I.U.R., M.B., S.S.
Financial support
The authors received no specific funding for this work.
Competing interest
The authors have no competing interest.
Open access
