Potential Risk for Localized Aggressive Periodontitis in African American Preadolescent Children

Abstract

Purpose

This pilot study evaluated the potential risk for localized aggressive periodontitis (LAgP) in African-American (AfAm) pre-adolescent children by detection of the potential periodontal pathogen Aggregatibacter actinomycetemcomitans (Aa) using PCR and microbiome analysis of oral samples.

Methods

Twenty-one pre-adolescents (age range 10.7–13.1 years) were recruited, for this IRB approved study. Oral examination included limited periodontal examination determining bleeding index (BOP) and periodontal probing (PD). An oral mucosa sample of was used for analysis.

Results

Nine of 21 children were positive for Aa (Aa+) by PCR. The Aa+ group, had significantly higher proportion of teeth with BOP and PD >4mm than the Aa− group (p=0.014 and 0.006 for percent BOP and percent PD equal to or greater than four mm, respectively, Mann Whitney Test). No significant differences in microbe abundance or composition were found from comparison of Aa + versus Aa− samples.

Conclusions

Detection of Aa from preadolescent AfAm children was associated with early surrogates for periodontal inflammation (i.e., BOP and PD>4mm). Although none of these children were diagnosed with LAgP, PCR targeting Aa could be a risk factor for LAgP. Further study is indicated however, the usefulness of PCR in dental practice setting to assess risk, may be cost-effective for early diagnosis and prevention of LAgP.

Introduction

Periodontal disease is a serious problem in underserved populations in the US,¹ and the resulting tooth loss effects are far reaching, leading to compromised dietary quality among older individuals², poor esthetic self-image, poor dental development, and overall poor health³. This problem is related to other oral health disparities, especially in underserved African American (AfAm) communities. Reid⁴ reported that much of the excess risk for untreated caries in the AfAm population can be explained by disparate income, education, employment, and dental insurance status. Wu⁵ studied oral health disparities and found that AfAm had significantly more decayed teeth and that fewer filled teeth, and that among those with some teeth, they were missing more teeth than Caucasians. Also, AfAm had fewer dental check ups than Caucasians. Sabbah⁶ found higher probabilities of poor oral health (gingival bleeding, periodontitis, and tooth loss) among AfAm, than in Caucasians, all associated with education and income levels. Borrell⁷ showed that race/ethnicity, education, and neighborhood socioeconomic conditions were associated with periodontitis with AfAm twice as likely to have periodontitis as Caucasians. Besides the chronic periodontitis that causes tooth loss among older adults, AfAm adolescents also suffer from a greater incidence (about 16 times greater than Caucasians⁸) of an aggressive periodontal disease, namely Localized Aggressive Periodontitis (LAgP), that results in severe bone and tooth loss^8–10. This disease manifests at preteen ages as incidental gingival attachment loss in approximately 10 percent of AfAm, a precursor of tooth loss. The etiology of LAgP is complex and includes associations with specific bacterial pathogens that greatly increase the risk of disease progression and teenage tooth loss¹¹. These findings demonstrated for the first time that AfAm and Hispanic adolescents that carried oral Aggregatibacter actinomycetemcomitans (Aa) bacteria had a one in five chance of developing serious bone loss at an early age that leads to tooth loss.

Polymerase chain reaction (PCR) has revolutionized the ability to detect specific bacterial species from biologic samples such as from the oral cavity. Using Aa–specific primers, it is possible to detect the presence of this potential pathogen from relatively small samples by PCR. New technologies have also been described to assess the overall microbe composition using sequence analysis of 16S ribosomal genes. Bacterial genera (even some species level detection) present in a sample are obtained by Next Gen DNA sequencing coupled with bioinformatic computing platform.

The Uniontown, Alabama community is located within the so-called “Black Belt,” a term used to describe the dark fertile soil. This area was historically a rural farming center. Today it remains underdeveloped and extremely poor. This area is home to a population made up of approximately ninety percent AfAm. The median income for a household in 2014 was $13,800¹². There is currently limited dental care available to youths in this community, which requires travelling thirty miles for dental care. Because of the risk for development of LAgP, this project reports a pilot study to identify individuals at risk among a cohort of AfAm (aged 10–13) children by screening for infection with Aa using PCR technology. We also wanted to determine if the overall composition of the oral microbiome was influenced by the presence or absence of Aa.

Methods and Materials

Population and information

A convenience sample of twenty-nine preadolescent-aged, unrelated African-American fifth and sixth grade children at the local elementary school in Uniontown, Alabama that responded to a parental letter that was sent from school were recruited to participate in this study. The study was approved by the University of Alabama at Birmingham (UAB) Institutional Review Board. Parents gave informed consent, and assent was obtained for participation by the children. Inclusion criteria were: healthy children with no condition requiring prophylactic or long-term antibiotic use and parental willingness to participate for the study. The children received anticipatory guidance, a professional fluoride treatment, customized oral hygiene guidance upon dental examination, and oral sample collection.

Limited periodontal exams that involved, collecting periodontal parameters (e.g. probing depth and noting bleeding on probing) from permanent maxillary anterior teeth, mandibular incisors, and all first molars were conducted by a Periodontist (PJ). Examinations were done using a light source, air, and periodontal probe (Michigan Probe, Henry Schein, Melville, NY, USA). No radiographs were obtained or used in clinical assessment of periodontal oral health. Height and weight were recorded for calculation of the body mass index (BMI), which was compared to standard national values for percentile based on age and sex from the CDC¹³.

Oral Sample Collection and Processing

DNA isolation

A sterile tongue depressor was used to gently scrape the cheeks and tongue of subjects, and then placed in a sterile tube containing reduced transport fluid (RTF). The sample tubes were placed on ice for transportation to the laboratory and stored at four degrees Celsius overnight before processing. Bacterial DNA was extracted using the ArchivePure DNA Yeast & Gram-+ kit (5′ PRIME Inc., Gaithersburg, MD, USA) Quantification of DNA was performed using a Nanodrop 1000 (Thermo Scientific, Wilmington, DE, USA). The threshold quality of DNA was set at a minimum value of 1.7. The DNA obtained was frozen at −20 degrees Celsius until submitted used for PCR, and for microbiome analysis.

Detection of Aa by PCR

PCR using 50 ng of DNA and 300 nM each of primers specific for a fragment of the Aa strain 624 (GenBank: CPO12959.1) tubulin binding protein. Amplification was carried out with the OneTaq Hot Start 2X Master Mix (New England BioLabs, Ipswich, MA, USA) in a T100 Thermal Cycler (BioRad, Hercules, CA, USA). The sequences of the forward and reverse primers were:

5′-ATGACAGAACACACGGAACAAGCACCG-3′

5′-GCAAATGCA ATTAAGCAGCTTATGGAAATTAAAGGTTTGCG-3′, respectively resulting in a 331-bp amplicon. The specificity of all primers used were confirmed through National Center for Biotechnology Information (NCBI) (Align Sequences Nucleotide BLAST, done May 10, 2016) focusing on known human and human commensal oral bacteria DNA sequences. PCR products were evaluated on 8.0 percent polyacrylamide gels for electrophoresis and stained with ethidium bromide using standard procedures.

Microbiome

Microbiome analysis was done using methods previously reported¹⁴. Briefly, following isolation of DNA from oral samples as described above, PCR using unique barcoded primers were used to amplify the 251 bp V4 region of the 16S rRNA gene. PCR products from the individual samples were electrophoresed on an agarose gel and visualized by UV illumination, then excised from the gel, purified, and sequenced using the NextGen sequencing Illumina MiSeq platform. Collection, analysis and annotation of sequence information derived from microbiome studies was as described by Kumar et al 2014¹⁴. Briefly, sequences were grouped into operational taxonomic units (OTUs) using the clustering program UCLUST at a similarity threshold of 97 percent ¹⁵. The Ribosomal Database Program (RDP) classifier, trained using the Greengenes (v13.8) 16S rRNA database, ¹⁶ was used to make taxonomic assignments for all OTUs at confidence threshold of 80 percent (0.8) ¹⁷. The resulting OTU table included all OTUs, their taxonomic identification, and abundance information. OTUs whose average abundance was less than 0.005 percent were filtered out. OTUs were then grouped together to summarize taxon abundance at different hierarchical levels of classification (e.g. phylum, class, order, family, genus, and species). Multiple sequence alignment of OTUs was performed with PyNAST ¹⁸.

Statistics

Data was analyzed by independent sample t-test assuming equal variance between groups using IBM SPSS Statistics Vs 22. The level for significance was set at p < 0.05. For the microbiome data, differences of taxa abundance and alpha diversity between two groups were measured using Kruskal-Wallis. The p values were corrected by the Benjamini-Hochberg FDR procedure for multiple comparisons and considered significant at p < 0.05. Differences in microbiota between groups were measured using PERMANOVA (weighted UniFrac distance).

Results

Informed consent was obtained from 29 subjects from whom examination data and samples had been collected. However, due to technical processing errors, samples from the first eight subjects did not result in sufficient quantity or quality of DNA to perform PCR. Therefore, only results from 21 subjects are reported. Demographic data for these AfAm subjects were 42 percent male, 10.7–13.1 years old (average age 11.8 years). Nine of the 21 oral samples were found to be positive for Aa (Fig. 1) by PCR. Table 1 presents the demographic and clinical data for the 21 subjects. Analysis for Aa negative compared to Aa positive (Aa carriers) subjects revealed no significant association with age, sex, BMI, or BMI percentile (T-test, data not shown). However, the percent of sites with bleeding on probing (BOP) and pocket depth (PD) that was greater than or equal to four mm were significantly greater among Aa carriers (Fig. 2, p = 0.014, and Fig. 3, p = 0.006, respectively).

Fig. 1.

Average Percent of sites that resulted in bleeding on probing (BOP). Scatterplot (■, %BOP, 6 sites probed per tooth for permanent incisors and first molar teeth) with Mean (●), Median (—), and 95% confidence interval (shaded box) for 12 subjects that did not have Aa detected by PCR analysis of oral samples (+) and 9 subjects that were Aa positive (+).

Aa + group significantly higher by t-test (p = 0.014).

Table 1.

Demographic, Clinical, and Laboratory Data for Twenty-one Subjects

Subject #	Aa PCR	Age (yrs)	Gender	BMI	BMI %tile	BOP,*	% PD ≥4mm*
1	−	10.8	F	18.3	64	15.6	4.2
2	−	11.3	M	24.8	95	0.0	1.0
3	−	12.3	F	29.2	98	5.2	7.3
4	−	11.4	M	31.8	99	25.0	14.6
5	−	10.7	F	31.3	99	1.0	6.3
6	−	11.0	F	28.3	98	25.0	10.7
7	−	13.1	M	ND	ND	7.8	12.2
8	−	12.2	M	31.5	98	5.2	6.3
9	−	11.1	F	20.2	77	16.7	11.5
10	−	11.3	M	21	88	14.6	8.3
11	−	11.7	F	17	52	10.4	10.4
12	−	12.7	M	17	26	36.7	25.6
13	+	11.8	F	21	81	46.9	38.5
14	+	10.8	F	17.2	47	20.8	4.2
15	+	12.6	F	15.8	10	15.6	24.0
16	+	11.4	M	19.6	79	21.9	25.0
17	+	12.2	F	18.4	53	18.8	11.5
18	+	12.7	M	34.2	99	53.3	17.5
19	+	12.8	F	23.4	89	27.1	15.0
20	+	12.5	F	22.1	85	21.1	23.3
21	+	11.4	M	28	98	25.0	29.8

BMI = Body Mass Index, percentile is based on US National Average for Adolescent based on age/sex, CDC¹³

BOP = bleeding on probing, percent of sights that bled following measurement of pocket depth.

% PD ≥4mm = percent of probing depths greater than or equal to 4 mm.

ND: missing height data, therefore could not calculate BMI.

^*indicates P<0.05 by t-test, assuming equal variance, see Figs. 2 and 3.

Fig. 2.

Average Percent of pocket depths greater than 4 mm. Scatterplot (■, %PD>4, 6 measurements per tooth for permanent incisors and first molar teeth) with Mean (●), Median (—), and 95% confidence interval (shaded box) for 12 subjects that did not have Aa detected by PCR analysis of oral samples (−) and 9 subjects that were Aa positive (+).

Aa + group significantly higher by t-test (p = 0.006).

Fig. 3.

Proportion of phylums among total OTUs. Scatterplot (■) of the proportion of the total operational taxonomic units (OTU) for each of five most common phylums for microbiome analysis, with Mean (●), Median (—), and 95% confidence interval (shaded box) for 12 subjects that did not have Aa detected by PCR analysis of oral samples (−) and 9 subjects that were Aa positive (+). No significant association at p < 0.05 (Kruskal-Wallis test) at the phylum level of bacterial identification was associated with Aa presence in samples from PCR.

Twenty-one oral DNA samples were used for microbiome analysis. PERMANOVA analysis resulted in 12 genus or species with significant differences (P <0.05) in the between Aa negative (Aa−) and Aa positive (Aa+) groups, however following multiple comparison correction, these findings were not supported. The most commonly identified (i.e., generally, most samples had greater than five percent of OTU from the pylum) were Firmicutes, Bacteroidetes, Actinobacteria, Proteobacteria, and Fusobacteria, greatest to least, respectively (Fig. 3). The proportion of the Proteobacteria phylum containing Aa (nor any other phylums identified) was (were) not significantly associated with positive Aa identification by PCR.

Discussion

In this pilot study, we found that almost one-half of the pre-adolescent children screened, were determined to be Aa+ carriers by PCR and that this group exhibited significant increases in periodontal inflammatory indicators (i.e., bleeding on probing and greater than four mm pocket depths). More severe symptoms are typically the crucial determining criteria needed for a diagnosis of LAgP, however, early detection of inflammation could be an important indicator of disease susceptibility. Since this study involves a cohort of children in the mixed dentition phase, with many teeth actively erupting, it was difficult to obtain accurate readings of the position of the CEJ that would be necessary to determine actual attachment losses. We used PD and BOP as potential indicators of early disease, and it must be interpreted with caution without more detailed assessment for periodontal disease. As has been reported by others¹¹, mucosal scrapings are useful to sample the oral cavity for the presence of Aa. Sampling of local sites were not done for this pilot study but would be important for follow-up with subgingival samples from sites that show signs of inflammation. Therefore, more rigorous prospective clinical studies would be needed to establish the efficacy of early Aa detection as a screening tool for LAgP risk.

The composition and relative abundance of microbiota did not significantly correlate with the presence or absence of Aa, suggesting that the presence of Aa does not alter the overall composition of the oral microbiome. This finding is consistent with the probability that Aa is very low abundance, especially in the absence of overt LAgP. Shaddox, et. al., study of the microbiota of subgingival samples using Human Oral Microbe Identification Microarray analysis (Forsyth Institute, Boston MA) from African American children reported that Aa was found in significantly greater abundance in children with Aggressive Periodontitis; however, even children with no evidence of disease had some Aa¹⁹. Further studies using gingival sulcus samples and more sensitive methodology, such as HOMIM, could provide additional quantitative and qualitative information relating to specific site and bacteria colonization for assessment of risk for LAgP. Fine, et al. ²⁰ used this approach and found that a consortium of bacteria that included Aa was seen prior to increased attachment loss and bone loss. This study included the finding that Aa was necessary but not sufficient for disease but did serve as a useful marker of susceptibility, confirming that Aa maybe a useful predictor of susceptibility. Therefore, other bacterial species associated with LAgP may be necessary to be included in screening tests. The PCR and Microbiome methods used in this pilot study are well suited to increase the scope of bacterial species screening.

The findings reported herein are preliminary and have some limitations that are acknowledged. This pilot study consisted of a convenience sample of children whose parents self selected based on interest in participation. Although there could be some inherent bias, these children reside in a predominantly AfAm disparate community that has been reported as a high dental caries risk and limited access to dental care²¹. Epidemiological risk for LAgP in poverty areas such as the Black Belt counties of the southeastern US is lacking. Therefore, this community is an important population for follow-up studies.

Because of the association of LAgP in an adolescent AfAm population, scientific advancement in this subject area aimed at preventing disease through early diagnosis could result in improved oral health for an underserved population. The screening of children could identify a “lead-time” for the detection of Aa infection before disease is manifested. Early identification of this disease susceptibility would facilitate the design of clinical protocols for the early diagnosis & treatment, and potentially the prevention of disease. Additional studies aimed at “detecting” Aa using PCR as a screening tool to formulate a Dental Practice-based Research Network (DPBRN) protocol for identification of patients at risk for periodontal disease in practice settings could serve as a baseline for future longitudinal studies that would determine whether or not more aggressive therapies (e.g., antibiotics) would be effective to stave off the aggressive periodontal destruction.

Conclusions

1. The molecular detection of A. actinomycetemcomitans in oral samples from preadolescent African American children in this convenience sample was associated with early surrogates for periodontal inflammation (i.e., bleeding on probing and pocket depth greater than four mm).

2. Although none of these children were diagnosed with Localized Aggressive Periodontitis, PCR targeting A. actinomycetemcomitans could be a risk factor for Localized Aggressive Periodontitis.

3. Further clinical study is needed to determine if PCR detection of A. actinomycetemcomitans or other species is/are a useful early risk assessment tool for Localized Aggressive Periodontitis in African American Children before the disease is manifested.