Genotypic characterization and comparison of Streptococcus mutans in American Indian and Southeast Iowa children

Abstract Early childhood caries is a complex health care issue that has a multifactorial etiology. One aspect of this etiology is the colonization and propagation of acidogenic bacteria at an early age. There have been several bacterial species associated with caries but 1 common species is Streptococcus mutans. Here, we describe genotypic diversity and commonality of Streptococcus mutans recovered from children representing 2 groups with similar socioeconomic demographics: a Northern Plains American Indian Tribe and a Southeast Iowa population. Forty 36‐month‐old American Indian children were selected from a cohort of 239 mothers and children, and forty 2‐ to 5‐year‐old children from Southeast Iowa were selected to compare the genotypic profiles of Streptococcus mutans recovered from each child's plaque. S. mutans isolates were selected from whole mouth plaque samples; DNA was extracted and amplified via AP‐PCR to show specific genotype patterns. These patterns were compared with GelComparIIv6.5 gel analysis software. We found 18 distinct genotypes from 524 isolates; 13 of which were common between the 2 communities. Five genotypes were unique to only the American Indian children while the Southeast Iowa children harbored no unique genotypes. Although the American Indian children had some genotypes that were not present in the Southeast Iowa children, these were not widely distributed among the community. Furthermore, the levels of genotypic diversity and commonality were similar between the 2 populations. This study sets the groundwork for a comprehensive comparison of genotypes and caries among larger subsections of both populations.

Although the development of caries in children is dependent on many variables, as stated previously, colonization of cariogenic bacteria within the oral cavities of these children appears to occur at a very young age (Hughes et al., 2012;Palmer et al., 2010;Parisotto et al., 2010;Tanner, Mathney, Kent, et al., 2011). This can lead to establishment of a plaque community dominated by organisms that drive the caries process. Although prevention and treatment of early childhood caries will involve intervention at many different levels, it is important to understand the development of the cariogenic oral microbial flora. As with any infectious disease, a better understanding of the microorganisms involved will help advance practices in prevention development.
Two different studies on early childhood caries and specific GTs of the cariogenic bacterium, S. mutans, have been conducted, one in an Iowa cohort and one in a Northern Plains American Indian Tribe. Diversity of S. mutans GTs in AI children has been shown by arbitrarily primed polymerase chain reaction (AP-PCR; Lynch et al., 2015).
Although each population has a high level of early childhood caries [46% of Southeast Iowa (SEI) children, 80% of AI children] and a similar percentage of families categorized as low SES (55% SEI, 49% AI), both the living conditions and overall environment are quite different.
Therefore, the purpose of this study was to determine similarities and differences in S. mutans GTs in children with early childhood caries across two different populations from two studies that have been conducted within the same laboratory utilizing the same AP-PCR protocol.

| Study populations and recruitment
Study population 1 (AI) was composed of 239 mothers who were pregnant or who had just given birth were recruited from a Northern Plains American Indian Tribe. All onsite research team members were AI and were under the guidance of a study director who was a senior dental hygienist in the tribe. Whole mouth plaque samples were collected periodically at eight visits from birth until 36 months of age.
Study population 2 (SEI) was composed of one hundred ninety 2to 5-year-old children (mean = 3.6 years) were recruited from the University of Iowa Muscatine Pediatric Dental Clinic. The majority of the children were either Hispanic (~45%) or Caucasian. A half-time study coordinator, who had experience in coordinating studies involving children, traveled to the dental clinic to recruit subjects in person. A small number of the subjects were recruited through Muscatine Head Start.
This investigation was focused on the comparison of genotypic profiles of S. mutans in 40 children from each population. Since the average age of the children in the Iowa population was 3.6 years, in order to have the closest equivalent in age when samples were collected, the S. mutans isolates from the 36-month samples from the AI children were analyzed. SEI children were chosen based on the criteria that each child had at least 2 S. mutans isolates recovered. There were 40 SEI children with at least 2 S. mutans isolates. The 40 AI children in this study were randomly chosen from those who were S. mutans positive at 36 months out of the 239 children in the original birth cohort.

| Consent
For the Northern Plains tribe, the Internal Review Board (IRB) on record was the Aberdeen Area IRB and approval was obtained. The study proposal was also presented to and received approval from the Tribal Research Review Board. For both study populations, approval was obtained from the University of Iowa IRB.

| Clinical examination
For the AI children, the clinical examination process has been detailed in Warren et al (2012). Briefly, caries examinations used dmfs criteria adapted from those used by NHANES (Dye, Tan, Smith, et al., 2007) and were conducted by trained and calibrated dental hygienist-examiners. These examinations were completed using the knee-to-knee method (Nowak & Warren, 2000). For the SEI children, examinations were conducted by one of the three trained and calibrated examiners using portable dental equipment. The study utilized the d1d2-3 caries criteria developed by Warren et al. (2015). With both study populations, examinations were conducted using a halogen examination light and a DenLite® illuminated mirror (Integra Miltex, York, PA). Teeth were dried with either gauze (AI children) or compressed air SEI children) and dental explorers were used to remove debris and to confirm areas of suspected decay.

| SEI samples
Plaque samples were collected using a gentle, noninvasive swabbing technique at the clinic prior to the dental examination. A sterile cotton swab was wiped over all smooth surfaces of the teeth. Swabs were placed into tubes containing reduced transport fluid and transported to the microbiology laboratories on ice. Plates were incubated at 37°C, 5% CO 2 for 72-96 hr.

| Sample processing and isolation of S. mutans
Ten colonies displaying typical S. mutans colony morphology were selected from the mitis-salivarius-kanamycin-bacitracin plate. Isolates were identified by fermentation profile (mannitol, raffinose, salicin, and sorbitol) and arginine decarboxylase activity. If there were fewer than 10 colonies, then all available colonies were selected. The average number of isolates obtained per subject in the SEI and AI cohorts was 4.65 and 8.17, respectively. In the AI population, a significant amount of Streptococcus sobrinus was present as well (approximately 30% of isolates). Isolates from this population were identified via fermentation profile or a second protocol which included preliminary identification by colony morphology and confirmation of species identification via PCR using primers specific to the gtfB (S. mutans) and the gtfI (S. sobrinus) genes . Following a second, single colony isolation onto TSB-YE agar (48 hr, 37°C, 5% CO 2 ) to assure strain purity, isolates were frozen in TSB-YE (10% glycerol) and stored at −80°C.

| DNA extraction and AP-PCR
Isolates were cultured in TSB-YE for 24 hr at 37°C, 5% CO 2 . DNA was extracted using the Epicentre® MasterPure™ Gram Positive DNA Purification Kit (Epicentre, Madison, WI, USA) with the following modifications: (a) 25-ml culture resuspended in 1.8-ml TE, (b) 2-μl Ready-Lyse with 1-hr incubation, (c) 25-min Proteinase K incubation, (d) 1-hr RNase A incubation after Proteinase K incubation, and (e) sample divided into three tubes for DNA precipitation steps. Due to the high volumes of mutans streptococci isolates being processed as part of the AI study, a new identification scheme, including a more rapid DNA extraction protocol, was tested and adopted as detailed by Villhauer et al (2017).
Although both methods of DNA extraction were utilized within these subject sets, the majority of isolates were processed using the rapid DNA extraction protocol. Genotypic diversity was examined by AP-PCR using the primer OPA-2 (5′-TGCCGAGCTG-3′). Each 50-μl PCR reaction contained 2-μl template DNA (50 ng/μl), 5 μl of 10X PCR buffer, 200 μM of dNTP, 7 mM MgCl 2 , 2.5 U Taq polymerase, and 4 μm of OPA-2 primer. S. mutans ATCC 25175 was used as a positive control for all reactions. Amplification was performed in a thermocycler (Eppendorf, Hauppauge, NY) programmed with the following temperature profile: initially 5 min at 94°C, followed by 45 cycles of denaturation at 94°C for 1 min, annealing at 36°C for 1 min, and elongation at 72°C for 2 min. Amplified products were electrophoresed on a 1.5% agarose gel and stained with ethidium bromide. A 100bp DNA ladder served as a molecular size marker on the gels. Gel images were captured using a transilluminator and digital imaging system (Fotodyne, Hartland, WI, USA).
A range of one to two GTs per person was seen in both study populations (Figure 1), with the AI children displaying an average of 1.33 GT per subject compared to 1.15 in the SEI children. The average for the entire subject set was 1.24. AI children had a higher percentage of two GTs than the Iowa children: 32.5% and 15%, respectively.
Eighteen distinct S. mutans GTs were found in these 80 children ( Figure 2). Thirteen of the GTs were detected in both study populations. Five GTs were found to be unique to the AI children, while none were unique to the SEI children. Some GTs were more prevalent in one population than the other (Table 1). In the AI children, GT1 is the most common and was detected in 35% of the subjects. GT8 and GT9 were the most common in the SEI children and were both found in 20% of those subjects. Five hundred twenty-four S. mutans isolates were analyzed from the subjects. The most commonly isolated was GT1, which accounted for 16% of the total isolates, followed by GT2, GT9, and GT11 (each at 10% of total isolates; Figure 3). Distribution of genotypes within the study populations. Each unique genotype was assigned a number (1-18; X axis) and the number of individuals (Y axis) who harbored each particular genotype is represented by the black (American Indian) and/or grey (Southeast Iowa) stacked bars Thornqvist, 2000;Klein, Flório, Pereira, Höfling, & Gonçalves, 2004;Lembo et al., 2007;Li & Caufield, 1995). Although we did not observe greater than two S. mutans GTs in any individual, we must remember that this is only one time point and it is possible and likely that these individuals could have harbored more GTs over time. In fact, the mothers of the AI children we studied sometimes harbored up to four GTs concurrently, and we did see some evidence of GT switching in the mothers and children in the AI group over time (Lynch et al., 2015). Previously published (Lynch et al., 2015) and unpublished results suggest that S. mutans colonization may not be as stable in this particular population as others have reported in other populations (Berkowitz, Jordan, & White, 1975;Köhler, Lundberg, Birkhed, & Papapanou, 2003). Additionally, reports suggest that there could be site specific differences in GT representation within an individual (Grönroos & Alaluusua, 2000). However, this study focused on similarities and differences of GTs in two populations separated both culturally and geographically. We feel that the total plaque samples, collected crosssectionally, are sufficient to determine the major GTs associated with each population.
We have observed large household sizes in the AI cohort. It would be interesting to compare household size and the genotypic diversity   there were no GTs unique to the Iowa children. However, we recovered five GTs that were unique to the AI children. Moreover, there was not a great deal of commonality among the GTs found only in AI children. Therefore, it is possible that these GTs are not widely distributed in the AI population. The fact that there were approximately twice as many S. mutans isolates analyzed from the AI cohort does raise the question of whether that is the reason more GTs were detected within those subjects. While we recognize that this could potentially create the appearance of bias toward greater genotypic diversity in this population, our findings for both populations are consistent with other published studies when looking at genotypic diversity at the single subject level, as stated earlier.
This is the first study we are aware of that compares S. mutans GTs between two different populations separated by geography and ethnicity but with similar SES. However, a very recent paper compared the oral microbiome diversity among Cheyenne and Arapaho individuals in Oklahoma (Ozga, Sankaranarayanan, Tito, et al., 2016). While there was not an emphasis on S. mutans GTs, it is notable that differences between native and non-native subjects were seen in the oral microbiomes in this cross-sectional study. It is intriguing to find S.
mutans GTs in this AI tribe that are not found in Iowa children. It is possible that S. mutans GTs unique to this population contribute to the high level of caries activity observed; however, further investigation is necessary to make that determination.
It is conceivable that there is more genetic variation or that other strep strains (i.e., S. sobrinus) could provide more genetic variation with these S. mutans GTs in the AI community. Analysis of the total plaque microbiome of these two populations may provide more information as to the diversity. Moreover, comparison of S. mutans GT profiles across additional populations is warranted and could provide interesting information on the commonality and uniqueness of specific S.  Distance matrix showing individual strain relatedness via curve-based cluster analysis. The Pearson correlation and Unweighted Pair Group Method using Arithmetic Averages was used. Analyses were performed with GelCompar®IIv6.5 software. Numbers are expressed as percent similarities between isolates of each genotype. As this is a representative sample of one (for each genotype) instead of the full comparison, some of the similarity values are higher than in the full cluster analysis of all isolates from the subject sets, which is expected. The intent behind this smaller, single representative analysis was to see what genotypes are most distinct from the other genotypes within the comparison. The table at the bottom of the figure is a compilation of the data in the matrix showing how unique/distinct each genotype is compared to the others most common GT in this population, there was no instance where it was the dominant GT within an individual subject's S. mutans isolates.
Although the AI children demonstrated greater genotypic diversity, in both colonized individuals and in total isolates, it is unlikely that this could be the cause for greater caries in the AI population. It is more likely that this is the result of greater caries due to the greater likelihood of plaque dominated by S. mutans and similar acidogenic flora.
Several studies have shown that greater S. mutans genotypic diversity is associated with increased tooth decay (Alaluusua, Mättö, Grönroos, et al., 1996;Napimoga, Kamiya, Rosa, et al., 2004) while other studies show the opposite effect (Kreulen, de Soet, Hogeveen, & Veerkamp, 1997), or no difference (Lembo et al., 2007). Paddick et al. demon-strated that in high-caries individuals, genotypic diversity of commensal species, such as Streptococcus oralis and Actinomyces naeslundii is lower than that of caries free individuals (Paddick, Brailsford, Kidd, et al., 2003). It is reasonable to assume that increased genotypic diversity of S. mutans in caries active individuals could simply be a consequence of a higher S. mutans burden in these individuals and/or decreased competition from other species. A possibility remains that certain GTs exhibit specific virulence traits consistent with increased caries risk but Nascimento et al. found no such associations in their study with Brazilian adults (Nascimento, Hofling, & Goncalves, 2004).
Whether certain GTs are more virulent or are associated with increased caries is the topic of ongoing research.