Advances in the Microbiology of Aggressive Periodontitis

It had long been claimed that Aggregatibacter actinomycetemcomitans is strongly associated with localized aggressive periodontitis. In particular, A. actinomycetemcomitans has even been involved in the pathogenesis of aggressive periodontitis. Later it had been realized that this bacterium occurs frequently also in other forms of periodontitis and even gingivitis. And, other bacteria, such as Porphyromonas gingivalis, may be involved in the pathogenesis of both localized and generalized aggressive periodontitis.

One of the upcoming issues of Periodontology 2000 is dedicated to all aspects of, what has turned out, very controversial aggressive periodontitis. Eija Könönen at Turku University and I had been asked to scrutinize the voluminous literature on the microbiology of aggressive periodontitis [1]. The following is a brief summary of the role of A. actinomycetemcomitans as a causative agent and new findings of open-ended molecular techniques regarding the microbiome of localized and generalized aggressive periodontitis.

Criteria for Causality

Sir Bradford Hill had described about 50 years ago criteria for possible causal relationship of certain risk factors for complex chronic diseases (Hill 1965), see here [pdf]. These criteria ultimately proved that smoking causes some forms of lung cancer. Among them, strength of association (including consistency and specificity), temporality of events (the causal factor has to be present before the onset of the disease) and experimental evidence (intervention trials) are of great importance as is for instance the biological gradient (dose-response relationship), while plausibility, coherence with laboratory findings, and analogy are not, or not so much [2].

Recently, Howick et al. (2009) have re-arranged Hill’s criteria [3], see here [pdf], putting much emphasis on direct evidence for a causal relationship which comprises the size of the effect (after adjusting for possible confounding), the appropriate temporal and/or spatial proximity (the cause precedes the effect and the effect occurs after a plausible interval; the cause occurs at the same site as the intervention), and dose-responsiveness and reversibility. What they call mechanistic evidence (evidence for biological, chemical, or mechanical mechanisms of action; and parallel evidence (coherence, replicability, similarity) would therefore not suffice to prove causality.

Methods and Results (I)

A systematic literature search was done in PubMed using the following focused question:

“Do prospective studies exist which may indicate a causal relationship between Aggregatibacter actinomycetemcomitans and the development of periodontitis at young age?”

Search terms were: (“actinobacillus actinomycetemcomitans”[MeSH Terms] OR (“actinobacillus”[All Fields] AND “actinomycetemcomitans”[All Fields]) OR “actinobacillus actinomycetemcomitans”[All Fields] OR (“aggregatibacter”[All Fields] AND “actinomycetemcomitans”[All Fields]) OR “aggregatibacter actinomycetemcomitans”[All Fields]) AND (“aggressive periodontitis”[MeSH Terms] OR (“aggressive”[All Fields] AND “periodontitis”[All Fields]) OR  “aggressive periodontitis”[All Fields]) AND (“longitudinal studies”[MeSH Terms] OR (“longitudinal”[All Fields] AND “studies”[All Fields]) OR “longitudinal studies”[All Fields] OR “prospective”[All Fields]), which yielded 63 papers published between 1986 and 2011. 

Fifty-eight papers were not further considered: six which dealt with the immune response to periodontal pathogens, nine case reports, three cross-sectional studies, eleven articles describing findings made in periodontitis patients, 28 therapy studies, two review articles and one letter.

Three prospective studies report associations between A. actinomycetemcomitans at baseline and development of aggressive periodontitis were read in full. In addition, two longitudinal studies which assessed A. actinomycetemcomitans as a risk factor for the development of periodontitis in adolescents or young adults (not mentioning explicitly aggressive periodontitis) were also included. So, in total five studies were read in full and data extracted, Höglund Åberg et al. (2009), Fine et al. (2007), Haubek et al. (2008), van der Velden et al. (2006) and Müller et al. (1996) [4].

The paper by Höglund Åberg et al. (2009) is a follow-up of thirteen subjects after 16 years. Each subject had been diagnosed with radiographic alveolar bone loss in the primary or mixed dentition. Each had harbored oral A. actinomycetemcomitans. Sixteen years later, all subjects had gingivitis but in only three out of 13 subjects, attachment loss with pocketing and radiographic bone loss had occurred. Checkerboard DNA-DNA hybridization and cultivation of subgingival plaque yielded A. actinomycetemcomitans in twelve out of 13 subjects, mostly at low levels (<10E5 CFU/genomes). In all subjects, other periodontal organisms were found as well. Authors conclude that “presence of A. actinomycetemcomitans and early bone loss in the primary dentition does not necessarily predispose the individual to periodontal attachment loss in the permanent dentition.”

Fine et al. (2007) had screened 1075 students, 11-17 years old, clinically and microbiologically for A. actinomycetemcomitans. Of those who carried the organism, 38 were followed for at least 1 year (test group). Fifty-eight age- and gender-matched controls without A. actinomycetemcomitans were followed up as well. It is noteworthy that authors consider 36 out of 38 test subjects as “periodontally healthy” although they presented at baseline with up to one 5-mm pocket without attachment loss; and as “borderline” healthy, despite the fact that they presented with two or more 5-mm pockets with <2 mm attachment loss. The remaining two subjects of the test group were considered “potentially diseased”; they presented with pocket depth of 6 or more millimeter with attachment loss of 2 mm or more. Among the 58 control subjects, 55 were periodontally healthy while three were “borderline” healthy (two or more 5-mm pockets with <2 mm attachment loss). At recall after at least 1 year, eight students presented with bone loss visible on bitewing radiographs. All were A. actinomycetemcomitans-positive. Two harbored the JP2 clone. Four had been classified at baseline as healthy, two as “borderline” healthy and two as “potentially” diseased. No A. actinomycetemcomitans-negative control subject presented with bone loss at any recall. Note that, in order to calculate the relative risk (which is not given in the original article) of carrying A. actinomycetemcomitans at baseline to develop aggressive periodontitis, the two subjects potentially diseased were not considered in the table below.

fine-2007

After adding 0.5 to each cell, the relative risk (RR) may be estimated as (6.5/36.5)/(0.5/58.5), which is about 20.9. For a 95% confidence interval (CI) the reader is referred to standard procedures, see, for instance, here. In the above case it is very large: (1.21; 360). Authors conclude that “the detection of A. actinomycetemcomitans in periodontally healthy children can serve as a risk marker for initiation of localized juvenile [aggressive] periodontitis.”

However, given the sparseness and heterogeneity of the data (the two groups seem to differ clinically at the outset; so, some carriers of A. actinomycetemcomitans might have already developed periodontitis), evidence must be considered weak.

Haubek et al. (2008) enrolled a population sample of 682 periodontally healthy adolescents in Morocco. The mean age at baseline examination was 12.5 years. The subjects were classified according to A. actinomycetemcomitans carrier status. Re-examinations took place after two years. Presence of the especially leukotoxic JP2 clone of A. actinomycetemcomitans which has been frequently found in West Africa, was assessed as Attachment loss of 3 mm or more at at least one periodontal site indicated the development of aggressive periodontitis.

haubek-2008

Carriers of the JP2 clone alone (RR 18.0; 95% CI 7.8 – 41.2), or together with other clones of A. actinomycetemcomitans (RR 12.4; 95% CI 5.2 – 29.9) were at considerably higher risk for periodontal attachment loss of 3 mm or more. Those with only other clones of A. actinomycetemcomitans than JP2 had a less increased risk (RR 3.0; 95% CI 1.3 – 3.7).

Van der Velden et al. (2006) conducted a longitudinal clinical and microbiological study over 15 years of all 15-25-year-old inhabitants of a governmental tea estate in western Java, Indonesia. At the outset in 1987, 255 subjects were examined. Follow-up exams took place in 1994 and 2002. Among 101 non-cases at baseline, attachment loss of 2 mm or more after 7 years was associated with A. actinomycetemcomitans (odds ratio 4.30, 95% CI 1.17 – 15.75) and the number of sites with subgingival calculus (1.40; 1.10 – 1.79) at baseline. After 15 years,the development of aggressive periodontitis (2 or more sites with pockets of 5 mm or more and attachment loss of 6 mm or more), which occurred in 20% of all 128 subjects available, was not associated with any of the three target bacteria A. actinomycetemcomitans, P. gingivalis, and Prevotella intermedia. Authors conclude that subgingival calculus and the presence of A. actinomycetemcomitans at basline would put a young subject at risk for developing periodontitis. Note, however, that subgingival calculus might be regarded as evidence that the process which eventually leads to loss of attachment had already commenced. In that regard, subgingival calculus may be regarded a consequence, not the cause, of periodontitis. In that case, progression is likely to be erroneously associated with any bacteria present, since temporality of events is not granted.

Müller et al. (1997) conducted a 1-year clinical and microbiological follow-up of 105 male conscripts of the German Armed Forces with gingivitis. A. actinomycetemcomitans was selectively cultivated from subgingival plaque, saliva and mucosal surfaces. The limit for detection was very low, 30-300 CFU. A. actinomycetemcomitans was recovered from 28% and 29% subjects at baseline and follow-up, respectively, generally in very low numbers. Active periodontal disease occurred in 7 subjects and was defined as an increase of periodontal pocket depth of at least 3 mm in one or more sites, and a significant increase in mean pocket depth and a significant skew of distribution of pocket depth alterations. In an adjusted multiple logistic regression model, A. actinomycetemcomitans was not associated with active disease (OR 0.29, 95% CI 0.03-3.38), in contrast to heavy smoking (14.2, 1.29-155).

Very recently, Hujoel et al. (2013) conducted a systematic review where they tried to identify evidence from longitudinal studies as regards onset of periodontitis (not necessarily aggressive periodontitis) that a panel of putative periodontal pathogens maybe responsible [5]. Authors identified three of the above five studies (Haubek et al. 2008, Fine et al. 2007, van der Velden et al. 2006) all dealing with A. actinomycetemcomitans. Apparently, no study has ever shown a possible causal role of major suspected pathogens such as P. gingivalis, Tannerella forsythia, P. intermedia, Treponema denticola, Fusobacterium nucleatum, Eubacterium nodatum, etc.

Conclusions (I)

There is strong evidence that the highly leukotoxic clone of A. actinomycetemcomitans is associated with the onset of aggressive periodontitis in Morocco. Claimed evidence for significant associations in other populations may be regarded controversial. Main caveats include (i) violation of the principle of temporality of events (if periodontitis had already commenced, any bacterium found in the pocket might be associated with disease progression), (ii) the selected cohorts (minorities, low social class, high infant mortality), and (iii) generally limited number of selected target microorganisms.

Methods and Results (II)

A systematic literature search in PubMed was done regarding published studies on the microbiome of aggressive periodontitis using checherboard DNA-DNA hybridization, 16S rRNA genomic analyses and the Human Oral Microbe Identification Microarray (HOMIM).

The search terms: (“aggressive periodontitis”[MeSH Terms] OR “aggressive”[All Fields] AND “periodontitis”[All Fields] AND “aggressive periodontitis”[All Fields] AND DNA-DNA[All Fields] AND checkerboard[All Fields] OR (“aggressive periodontitis”[MeSH Terms] OR “aggressive”[All Fields] AND “periodontitis”[All Fields] AND “aggressive periodontitis”[All Fields] AND HOMIM[All Fields] OR (“aggressive periodontitis”[MeSH Terms] OR “aggressive”[All Fields”] AND “periodontitis”[All Fields] AND “aggressive periodontitis”[All Fields] AND 16S[All Fields] AND rRNA[All Fields]) yielded 35 papers published between 1999 and 2012.

Thirty papers were not considered: an abstract dealing with Papillon-Lefevre syndrome, 18 papers dealing with limited numbers of traditional or “new” pathogens or herpesviruses, or case reports; and twelve therapy studies. The five remaining studies described the main portion of the subgingival microbiota in aggressive periodontitis. Two additional articles that employed checkerboard DNA-DNA hybridization and two articles that dealt with 16S rRNA clonal analysis of subgingival plaque in aggressive periodontitis were not identified by the search terms but also read in full. Thus, nine studies were considered and data extracted [6].

For the purpose of this brief summary of the systematic review, three representative of the above studies are described in some detail.

Faveri et al. (2009 examined 120 subjects, 15 with localized periodontitis and 25 with generalized aggressive periodontitis, 30 with chronic periodontitis, as well as 30 healthy adults and 20 healthy young subjects. Nine individual subgingival plaque samples were  in each subject, three from shallow sites (probing depth of 3 mm or less), intermediate sites (4-6 mm probing depth) and deep sites (pockets of 7 mm or more) each. Thirty-eight species were identified with whole genomic DNA probes. When samples from patients with localized aggressive periodontitis were compared with those from young healthy subjects, T. forsythia and P. gingivalis were the most numerous and prevalent (% sites with 10E6 genomes or more) species in localized aggressive, and higher mean counts of T. forsythia, P. gingivalis, Campylobacter gracilis, E. nodatum, and P. intermedia were observed in localized aggressive periodontitis. The comparison of the three disease and two healthy groups revealed higher proportions of red-complex bacteria [7] and E. nodatum in disease groups, h
igher proportions of A. actinomycetemcomitans in localized and generalized aggressive periodontitis, and higher proportions of the purple, yellow and green (except A. actinomycetemcomitans) complexes and Actinomyces spp. in healthy groups. Higher proportions of the red-complex bacteria and lower proportions of the purple-, yellow- and green-complex bacteria (except A. actinomycetemcomitans) were observed with increasing pocket depth. And finally, higher proportions of A. actinomycetemcomitans were observed in intermediate and deep pockets in localized aggressive periodontitis. Comparisons among the three disease groups revealed higher proportions of A. actinomycetemcomitans in localized aggressive periodontitis; of F. nucleatum ssp. polymorphum and P. gingivalis in generalized aggressive periodontitis, and of Actinomyces naeslundii in chronic periodontitis. In deep pockets, the microbiota was quite similar in all disease groups. Further observations in localized aggressive periodontitis patients included a negative correlation of age with proportions of A. actinomycetemcomitans and a positive correlation with the proportions of the red-complex bacteria. Proportions of A. actinomycetemcomitans were negatively correlated with the proportions of red-complex bacteria.

Teles et al. (2010) examined 31 patients with generalized aggressive periodontitis and 25 healthy adults. The age range of both groups was 18-30 years. Individual subgingival samples were obtained from mesiobuccal aspects of each tooth in two random quadrants. Forty species were identified with whole genomic DNA probes. Significantly higher proportions of Campylobacter rectus, Streptococcus anginosus, E. nodatum, T. forsythia, P. gingivalis and T. denticola were observed in generalized aggressive periodontitis. Species which were found in periodontal health in significantly higher proportions were Capnocytophaga gingivalis, Capnocytophaga ochracea, Eikenella corrodens, Fusobacterium periodonticum, Eubacterium saburreum, Gemella morbillorum, Neisseria mucosa, and Propionibacterium acnes. Further observations among individual samples included five clusters with the predominance of Actinomyces spp., or members of the orange or red complexes, and the green complex, including A. actinomycetemcomitans (eight samples in two patients), at high counts in 221 samples from generalized aggressive periodontitis patients. In periodontally healthy subjects, variation of microbial composition was even greater.

Rylev et al. (2011) sampled subgingival plaque from a limited number of aggressive periodontitis patients and healthy controls from Morocco. Four patients had aggressive periodontitis (two harboring the highly leukotoxic JP2 clone of A. actinomycetemcomitans), and four subjects were healthy (two harbored non-JP2 A. actinomycetemcomitans). Two subjects in each group had been monitored clinically and microbiologically for four years. Sequence analysis of 16S rRNA yielded 2717 validated sequences. One-hundred-seventy-three bacterial taxa (between 24 and 71 taxa per subject) were identified. Thirty-nine % of these taxa had not previously been described.  The JP2 clone in the 2 JP2-positive subjects represented 1% and 6.7% of sequenced clones, respectively. The periodontal microbiome of both diseased and healthy subjects was very complex comprising taxa belonging to eight bacterial phyla: TM7, Proteobacteria, Firmicutes, Synergistes, Actinobacteria, Spirochaetes, Bacteroidetes, and Fusobacteria. Notably, the TM7 phylum (3 – 21.2% of cloned sequences) included a previously undetected cluster of 31 sequences distributed in 17 sequence variants. Particularly high levels of another phylotype of TM7 were found in the two patients with aggressive periodontitis (20.6% and 10.8% of the sequences) who did not carry the JP2 clone of A. actinomycetemcomitans. No single taxon, except for the JP2 clone and taxa detected in only a single individual, was exclusively detected in patients with aggressive periodontitis. The authors speculate that disease activity might be terminated in some patients with aggressive periodontitis as a result of elimination of the JP2 clone by high anti-leukotoxin serim IgG titers in blood.

Conclusions (II)

Since the first recognition, about 35 years ago, of A. actinomycetemcomitans in periodontal lesions of aggressive periodontitis, the current picture has become much more complicated. Carefully conducted longitudinal studies aiming at the identification of specific microorganisms involved in the initiation of the disease are still rare. In many cases of those who are susceptible for aggressive periodontitis, the disease may have its onset in the primary dentition. A. actinomycetemcomitans may join the host-compatible green complex, comprising Capnocytophaga spp. and E. corrodens, and P. intermedia. In established disease and after spread of periodontal lesions, the microbiota is becoming much more complex and soon resembles that of generalized chronic periodontitis with the red-complex bacteria seemingly dominating the subgingival ecosystem. The composition of the subgingival microbiota among subjects with aggressive periodontitis usually differs considerably. Host-compatible microorganisms such as streptococci, Actinomyces spp. and Capnocytophaga spp. are largely reduced anyway. New technologies based on DNA sequencing and advanced bioinformatics has the potential to reveal the complexity of subgingival microbiota. So far, in aggressive periodontitis the role of novel and yet-not-cultivable bacteria has not yet been elucidated.

agp

Brief summary of current understanding of the development of subgingival microbiota in aggressive periodontitis. (a) In healthy children, the flora mainly consists of the yellow- and green-complex bacteria and Actinomyces spp. (b) In the case of gingivitis, P. intermedia frequently colonizes the area at an early stage. (c) A. actinomycetemcomitans may join and cause, in susceptible subjects, early alveolar bone loss in the deciduous and mixed dentition. (d) Once pocketing has occurred and lesions have spread, the flora is becoming very diverse, resembling that of chronic periodontitis. In many cases, the red-complex bacteria predominate while in others, A. actinomycetemcomitans is found in large numbers.

Brief summary of current understanding of the development of subgingival microbiota in aggressive periodontitis. (a) In healthy children, the flora mainly consists of the yellow- and green-complex bacteria and Actinomyces spp. (b) In the case of gingivitis, P. intermedia  frequently colonizes the area at an early stage. (c) A. actinomycetemcomitans may join and cause, in susceptible subjects, early alveolar bone loss in the deciduous and mixed dentition. (d) Once pocketing has occurred and lesions have spread, the flora is becoming very diverse, resembling that of chronic periodontitis. In many cases, the red-complex bacteria predominate while in others, A. actinomycetemcomitans is found in large numbers.

Notes

[1] Könönen E, Müller HP. Microbiology of aggressive periodontitis. Periodontol 2000 2013; in press.

[2] Hill AB. The environment and disease: association or? Proc Roy Soc Med 1965; 58: 295-300. The order Hill had listed his criteria when discussing them one after another had probably caused confusion, not only among dentists. That Hill’s criteria are applied in dentistry, as could be noted when reading at least some of the proceedings of the recent joined workshop of the American Academy and the European Federation of Periodontology on the relationship between periodontitis and systemic diseases and conditions,  seems to be a new development. For more than 30 years, dentists have preferred to follow Socransky’s criteria of 1979 for identifying causal agents in dental caries and periodontal disease, a largely weakened version of Robert Koch’s classic postulates for proving that a specific organism causes a well-defined infection.

[3] Howick J, Glasziou P, Aronson JK. The evolution of evidence hierarchies: what can Bradford Hill’s guidelines for causation contribute? J R Soc Med 2009; 102: 186-194.

[4] Höglund Åberg C, Sjödin B, LakioI,Pussinen PJ, Johansson A, Claesson R. Presence of Aggregatibacter actinomycetemcomitans in young individuals: a 16-year clinical and microbiological follow-up study. J Clin Periodontol 2009; 36: 815-822. Fine DH, Karkowitz K, Furgang D, Fairlie K, Ferrandiz J, Nasri C, McKiernan M, Gunsolley J. Aggregatibacter actinomycetemcomitans and its relationship to initiation of localized aggressive periodontitis: longitudinal cohort study of initially healthy adolescents. J Clin Microbiol 2007; 45: 3859-3869. Haubek D, Ennibi OK, Poulsen K, Vaeth M, Poulsen S, Kilian M. Risk of aggressive periodontitis in adolescent carriers of the JP2 clone of Aggregatibacter actinomycetemcomitans in Morocco: a prospective longitudinal cohort study. Lancet 2008; 371: 237-242. Van der Velden U, Abbas F, Armand S, Loos BG, Timmerman MF, van der Weijden GA, van Winkelhoff AJ, Winkel EG. Java project on periodontal diseases. The natural development of periodontitis: risk factors, risk predictors and risk determinants. J Clin Periodontol 2006; 33: 540-548. Müller HP, Eger T, Lobinsky D, Hoffmann S, Zöller L. A longitudinal study of Actinobacillus actinomycetemcomitans in army recruits. J Periodontal Res 1997; 32: 69-87.

[5] Hujoel P, Zina L, Cunha-Cruz J, Lopez R. Specific infections as the etiology of destructive periodontal disease: a systematic review. Eur J Oral Sci 2013; 121: 2-6.

[6] Faveri M, Figueiredo LC, Duarte PM, Mestnik MJ, Mayer MP, Feres M. Microbiological profile of untreated subjects with localized aggressive periodontitis. J Clin Periodontol 2009; 36: 739-749. Ximenez-Fyvie IA, Almaguer-Flores A, Jacobo-Soto V, Lara-Cordoba M, Moreno-Borjas JY, Alcantara-Marui E. Subgingival microbiota of periodontally untreated Mexican subjects with generalized aggressive periodontitis. J Clin Periodontol 2006; 33: 869-877. Rescala B, Rosalem W Jr, Teles RP, Fischer RG, Haffajee AD, Socransky SS, Gustafsson A, Figueredo CM. Immunological and microbiologic profiles of chronic and aggressive periodontitis subjects. J Periodontol 2010; 81: 1308-1316. Teles RP, Gursky LC, Faveri M, Rosa EA, Teles FR, Feres M, Socransky SS, Haffajee AD. Relationships between subgingival microbiota and GCF biomarkers in generalized aggressive periodontitis. J Clin Periodontol 2010; 37: 313-323. Carvalho RP, Mesquita JS, Bonomo A, Elsas PX, Colombo AP. Relationship of neutrophil phagocytosis and oxidative burst with the subgingival microbiota of generalized aggressive periodontitis. Oral Microbiol Immunol 2009; 24: 124-132. Heller D, Varela VM, Silva-Senem MX, Torres MC, Feres-Filho EJ, Colombo AP. Impact of systemic antimicrobials combined with anti-infective mechanical debridement on the microbiota of generalized aggressive periodontitis: a 6-month RCT. J Clin Periodontol 2011; 38: 355-364. Faveri M, Mayer MP, Feres M, de Figueiredo LC, Dewhirst FE, Paster BJ. Microbiological diversity of generalized aggressive periodontitis by 16S rRNA clonal analysis. Oral Microbiol Immunol 2008; 23: 112-118. Hutter G, Schlagenhauf U, Valenza G, Horn M, Burgemeister S, Claus H, Vogel U. Molecular analysis of bacteria in periodontitis: evaluation of clone libraries, novel phylotypes and putative pathogens. Microbiology 2003; 149: 67-75. Rylev M, Bek-Thomsen M, Reinholdt J, Ennibi OK, Kilian M. Microbiological and immunological characteristics of young Moroccan patients with aggressive periodontitis with and without detectable Aggregatibacter actinomycetemcomitans JP2 infection. Mol Oral Microbiol 2011; 26: 35-51.

[7] Based on community ordination and cluster analysis of checkerboard DNA-DNA hybridization results of more than 13,000 subgingival plaque samples from subjects and patients with periodontal disease color-coded bacterial complexes had been described by Socransky et al. (1998) which are considered established. The red complex comprises periodontal pathogens Porphyromonas gingivalis, Tannerella forsythia and Treponema denticola. The orange complex contains, among others, members of the genuses Prevotella, Campylobacter, Fusobacterium, and Eubacterium. Actinomyces spp., the purple (Veillonella parvula, Actinomyces odontolyticus), yellow (Streptococcus spp.) and most of the green complex (Eikenella  corrodens, Capnocytophaga spp., but not Aggregatibacter actinomycetemcomitans) are considered host-compatible.

11 August 2013 @ 7:10 pm.

Last modified August 15, 2013.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s