Aggregatibacter actinomycetemcomitans produces a toxin, which associates this bacterium with periodontitis affecting young individuals (Haubek et al., 2008; Henderson et al., 2010; Könönen and Muller, 2014; Fine et al., 2020). Since the toxin causes lethal effects on leukocytes of different types, it is considered a leukotoxin (LtxA). The production and release of LtxA is affected by alterations in the promoter region of the ltxCABD operon (Kolodrubetz et al., 1989; Spitznagel et al., 1991). Of the seven serotypes of A. actinomycetemcomitans (a-g), a specific variant of serotype b has a deletion of 530 base pairs (bp) in the promoter region (Brogan et al., 1994). This variant is genetically different from those having a full-length promoter, and is since long referred to as the JP2 genotype (Haubek, 2010). As a consequence, all variants of A. actinomycetemcomitans with a full-length leukotoxin gene promoter are typically referred to as non-JP2 genotype. In serotype b strains, also additional ltxCABD promoter types have been identified, with either deletions of alternative regions of the promoter, or with insertions in it (He et al., 1999; Claesson et al., 2015; Claesson et al., 2020).

The JP2 genotype of A. actinomycetemcomitans serotype b is highly leukotoxic (Tsai et al., 2018), and carriers of this specific type are at substantial risk for rapid loss of tooth-supporting tissues, including alveolar bone (Bueno et al., 1998; Haraszthy et al., 2000; Haubek et al., 2008; Burgess et al., 2017). While the JP2 genotype appears to be distributed only in serotype b of A. actinomycetemcomitans, the non-JP2 genotype is found among all serotypes (a-g). Non-JP2 genotype strains of A. actinomycetemcomitans are generally low-leukotoxic. However, a subgroup of non-JP2 genotype strains of A. actinomycetemcomitans serotype b that exhibit high leukotoxicity, and has enhanced association with periodontal disease progression, has been identified (Asikainen et al., 1995; Höglund Åberg et al., 2014).

In addition to leukotoxin promoter typing, A. actinomycetemcomitans strains have been genotypically characterized by a range of other DNA-based methods (Asikainen and Karched, 2008; Kittichotirat et al., 2016). One of them is arbitrarily-primed (AP)-PCR (van Belkum, 1994), which has revealed that all hitherto tested JP2 genotype strains belong to the same AP-PCR banding pattern group (AP-PCR type 1), whereas for non-JP2 strains a number of different AP-PCR types have been identified (Höglund Åberg et al., 2014; Claesson et al., 2017). Interestingly, a subgroup of non-JP2 genotype strains of A. actinomycetemcomitans serotype b that expresses enhanced levels of LtxA also belongs to AP-PCR type 1, and hence appears to be genetically related to the JP2 genotype of A. actinomycetemcomitans (Höglund Åberg et al., 2014; Johansson et al., 2017). Moreover, all hitherto assessed A. actinomycetemcomitans strains of serotype b belonging to AP-PCR type 1 have the property in common that they carry the gene cagE, which with very few exceptions appears to be present in this AP-PCR type only, and as the gene is absent in the other A. actinomycetemcomitans serotypes, it could be used for diagnostic purposes (Johansson et al., 2017; Johansson et al., 2019).

AP-PCR characterization mirrors genetic variability within bacterial species without requiring the analysis of specific genes, whereas Multilocus Sequence Typing (MLST) is a molecular typing method that is used to characterize bacterial strains in terms of pathogenicity and dissemination (Dogan et al., 1999b; Haubek et al., 2007; Jolley and Maiden, 2014). This is a highly discriminatory typing method based on sequencing of a number of housekeeping genes. MLST of bacterial strains renders so called sequence types (ST), which can be further used in dissemination studies. As judged by the MLST-based characterization, the JP2 genotype of serotype b emerged from the Mediterranean part of Africa about 2400 years ago, and has then subsequently been spread worldwide (Haubek et al., 2007; Rylev and Kilian, 2008).

In the present study, we have carried out MLST-characterization of non-JP2 genotype strains of A. actinomycetemcomitans serotype b from two of our clinical collections, isolated from individuals living in Ghana and Sweden, respectively (Höglund Åberg et al., 2012; Claesson et al., 2017). Thereby, we have been able to compare the genetic variability of non-JP2 genotype strains of A. actinomycetemcomitans serotype b collected from carriers living in two from each other separated continents, to assess if geographic dissemination of such strains may exist. In addition, our aim was to study whether any correlation could be observed between AP-PCR/cagE genotype and MLST–based sequence types of non-JP2 genotype strains of A. actinomycetemcomitans serotype b. To this end, the MLST pattern of each strain was compared with their genetic discrepancies earlier revealed by AP-PCR and cagE genotyping (Höglund Åberg et al., 2014; Claesson et al., 2017; Johansson et al., 2017; Johansson et al., 2019).

In this work, we have used MLST to investigate the genetic diversity of 56 non-JP2 genotype strains of A. actinomycetemcomitans serotype b, with 41 strains originating from Ghana and 13 from Sweden.

Interestingly, serotype b includes a genetic variant that can be found both among JP2 genotype, and among some non-JP2 genotype strains. We have earlier demonstrated that these strains belong to the same AP-PCR type (i.e., AP-PCR type 1) (Johansson et al., 2017). While JP2 genotype strains of A. actinomycetemcomitans previously have been characterized by MLST, corresponding studies of non-JP2 strains of serotype b have been scarcely reported on (Haubek et al., 2007). Based on the MLST analysis of the 56 non-JP2 genotype serotype b strains, we suggest that the overall mutation rate of this genotype is rather low. Comparing the nine identified MLST sequence types with their corresponding AP-PCR genotype, we found both similarities and discrepancies, as the strains were distributed in different clusters based on the two genotyping methods.

In this study, although the mutation rate within the non-JP2 genotype strains appeared to be low as evidenced by MLST, discrepancies between the strains of Ghanaian and Swedish origin were detected. This was emphasized by the five unique SNPs among the Ghanaian strains, and seven among the strains isolated from individuals living in Sweden, despite that the number of assessed Swedish strains was more than three times lower. This indicates that the Swedish strains may represent a more diverse group. The reason for this could probably be that the Ghanaian collection of strains was sampled from a more homogeneous group of individuals with regards to age and geographic localization (Höglund Åberg et al., 2012). The notion that A. actinomycetemcomitans serotype b that colonizes the Swedish population, are more prone to mutations is another possible explanation.

A. actinomycetemcomitans strain HK908 is one of few non-JP2 genotype serotype b strains that has been analyzed previously by MLST (Haubek et al., 2007). This strain was in that study described as non-serotypeable. As HK908 in the study by Haubek and coworkers (Haubek et al., 2007) was used to illustrate the evolutionary stage, where non-JP2 genotype A. actinomycetemcomitans serotype b evolved into the JP2 genotype, we chose to include this strain also in the present study, after confirmation that it did in fact belong to serotype b (Johansson et al., 2017). For further comparisons, also the non-JP2 genotype serotype b strain, HK 912 was included as a reference. In the study by Haubek and coworkers (Haubek et al., 2007), HK908 was demonstrated to represent a breaking point, dividing JP2 and non-JP2 genotype strains of A. actinomycetemcomitans of serotype b into two different clusters. In this work, HK908 was instead used to illustrate clustering among non-JP2 genotype strains. As the deletion of an “A” at nt position 56 in tbpA was observed only in non-JP2 genotype serotype b strains in the previous study (Haubek et al., 2007), we here used the presence and the absence of this deletion as a means to divide the 56 non-JP2 genotype serotype b strains into two major clusters, A and B, respectively. This led to the observation that strains of Ghanaian and Swedish origin were differentially distributed in these clusters, again indicating that there is a genetic diversity between the two collections of non-JP2 genotype strains of A. actinomycetemcomitans serotype b from Ghana and Sweden, respectively.

Different types of PCR-based methods for the characterization of bacterial genomes have been developed during the last decades (van Belkum, 1994; Jolley and Maiden, 2014; Bonk et al., 2018; Wade and Prosdocimi, 2020). However, similar results will not necessarily be obtained, when bacterial collections are characterized with other PCR-based methods (Tsang et al., 2017). When we earlier characterized non-JP2 genotype strains of A. actinomycetemcomitans, serotype b by AP-PCR, we found more than 10 different banding patterns. The strains were grouped into AP-PCR types 1, 2 and others (referred to as AP-PCR type 3 in this work) (Claesson et al., 2017; Johansson et al., 2019). Among serotype b, strains of AP-PCR type 1 are in general more leukotoxic than the other AP-PCR types, and they are more frequently associated with periodontitis affecting young individuals (Dogan et al., 1999a; Höglund Åberg et al., 2014). Thus, albeit AP-PCR–based analyses of A. actinomycetemcomitans mainly reflects the general genetic diversity within this bacterial species, this methodology can be valuable for diagnostic purposes, to identify highly leukotoxic strains of serotype b. However, evidently AP-PCR is also a time-consuming method, which is associated with difficulties in standardization (Meunier and Grimont, 1993; Power, 1996; Tyler et al., 1997; Suzuki et al., 2003).

As evidenced from our present study, the MLST scheme implemented appeared not to be a highly discriminatory method to identify AP-PCR type 1, as strains of this AP-PCR type could be found within three different ST-types (ST 2, ST 3 and ST 88). Interestingly, AP-PCR type 1 A. actinomycetemcomitans strains of serotype b, isolated from both Ghana and Sweden, regardless being of JP2 or non-JP2 genotype, carry a gene that with very few known exceptions is present only in this AP-PCR type, i.e., cagE (Johansson et al., 2017; Johansson et al., 2019). As discussed earlier (Johansson et al., 2017; Johansson et al., 2019), the cagE gene product per se is most likely not responsible for the increased leukotoxicity of AP-PCR type 1 strains. However, to identify whether carriers are infected with a more virulent A. actinomycetemcomitans serotype b, i.e., AP-PCR type 1, which includes both JP2 and non-JP2 genotypes, detection of the cagE gene by PCR-based methodology appears to represent a more efficient tool than MLST.

Although based on a limited number of assessed A. actinomycetemcomitans strains in this work, especially those collected from carriers in Sweden, there is a clear indication of a geographic dissemination across continents of non-JP2 genotype A. actinomycetemcomitans serotype b. Detection of the cagE gene is probably the currently known, most straightforward diagnostic procedure for detection of patients carrying the virulent AP-PCR type 1 of A. actinomycetemcomitans, serotype b. The cagE gene as a marker can be used to identify both JP2 and non-JP2 genotype strains of A. actinomycetemcomitans serotype b of this AP-PCR type, with PCR or qPCR, and cagE-specific oligonucleotide primers.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

Conceptualization, RC, AJ, and JO. Methodology, RC, AJ, AE, and JO. Validation, RC, AJ, CH, AE, and JO. Formal analysis, RC, AJ, and JO. Investigation, RC, AJ, and JO. Resources RC, AJ, CH, DH, and JO. Data curation, RC, AJ, and JO. Writing—original draft preparation, RC, AJ, and JO. Writing—review and editing, RC, AJ, CH, AE, DH, and JO. Visualization, RC, AJ, DH, and JO. Supervision, AJ, RC, and JO. Project administration, RC, AJ, and JO. funding acquisition, AJ and JO. All authors contributed to the article and approved the submitted version.

This work was supported by TUA grants from the Region of Västerbotten, Sweden (AJ; 7003193 and JO; 7002667), and from Insamlingsstiftelsen, Medical Faculty, Umeå University, Sweden.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.