Expression of matrix metalloproteinase-9 in gingival tissue biopsy in patients with slowly/ moderately and rapidly progressing periodontitis: An observational study.

BACKGROUND: Matrix metalloproteinases (MMPs) are a group of host-derived zinc-dependent enzymes which mediates the destruction of the extracellular matrix. In periodontitis, there is excess production of MMPs associated with periodontal tissue destruction. The aim of this study was to estimate the level MMP-9 in both active and latent form in gingival tissue (GT) samples collected from periodontitis patients with different rates of progression and compare it with healthy individuals.
MATERIALS AND METHODS: Sixty patients were selected and divided into three groups, 20 each: Group A (slowly/moderately progressing periodontitis), Group B (rapidly progressing periodontitis), and Group C (clinical periodontal health). Plaque index, gingival index, periodontal probing depth (PPD), and clinical attachment level were recorded. GT samples were collected from all 60 patients and MMP-9 expressions were measured using gelatin zymography and western blotting.
RESULTS: Levels of active MMP-9 (aMMP-9) and latent MMP-9 (lMMP-9) were significantly high in both Group A (GA) (aMMP-9: 2.05 arbitrary unit [AU]/lMMP-9: 2.54 AU) and Group B (GB) (aMMP-9: 1.32 AU/lMMP-9: 1.74 AU) when compared to that of Group C (GC) (aMMP-9: 0.93/lMMP-9: 1.08 AU). In GA, levels of aMMP-9 showed a significant correlation with PPD values. No other correlations were found.
CONCLUSION: The levels of aMMP-9 and lMMP-9 were increased in both the types of periodontitis when compared with periodontally healthy individuals. A significant correlation was found between PPD and activities of aMMP-9 in slowly/moderately progressing periodontitis patients. However, further studies are required to confirm these findings. Copyright:

INTRODUCTION

Matrix metalloproteinases (MMPs) are a group of structurally related zinc-dependent neural endopeptidases belonging to the Matrixin family of proteinases.[1] MMPs can degrade extracellular membrane macromolecular components such as elastin, laminin, fibronectin and proteoglycans, and basement membrane collagen at neutral pH.[234] MMPs play a significant role in physiological and pathophysiological processes such as tissue remodeling,[56] epithelial wound closure,[7] inflammation,[7] periodontal disease destruction,[56] rheumatoid arthritis,[89] atherosclerosis,[10] diabetes,[11] osteoporosis,[11] vascular diseases,[10] acquired connective tissue disorder such as anetoderma,[12] and cancer invasion.[13] Typically, the activity of MMPs is tightly regulated by inhibitors like tissue inhibitors of metalloproteinases (TIMPs) and α2-macroglobulin which are endogenous tissue-specific and circulating inhibitors, respectively.[14] Based on the sequence homology and substrate specificity,[15] MMPs can be grouped as (a) collagenases – MMP-1, MMP-8, MMP-13, and MMP-18, (b) gelatinases – MMP-2 and MMP-9, (c) stromelysins – MMP-3, MMP-10, and MMP-11 and (d) Membrane type matrix metalloproteins – MMP-14, MMP-15, MMP-16, MMP-17 and MTI-MMP[16] and (e) Matrilysins – MMP-7 and MMP-26.[1617]

Periodontitis can be defined as “an inflammation of the periodontal tissues resulting in clinical attachment loss, alveolar bone loss, and periodontal pocketing.”[18] Various studies over the decade have shown an association between levels of MMPs and periodontal tissue destruction.[192021] In periodontitis patients, the balance between the MMPs and its inhibitors like α2-macroglobulin and TIMPs are altered in favor of MMPs.[52223] Hence, the level of MMPs is increased in periodontal tissues. The balance between production and degradation of collagen is altered resulting in the destruction of periodontal connective tissue leading to attachment loss and finally tooth loss.[24]

Gelatinases (MMP-2 and MMP-9) are of interest because of their substrate specificity which are determined by their fibronectin- and hemopexin-like domains.[2526] Both MMP-2 and MMP-9 can degrade type IV and type V collagen of the basement membrane. Hence, they are considered very important in the early development of inflammatory conditions like arthritis and in cancer invasion.[2728] In periodontitis, MMP-2 and MMP-9 are involved in the extracellular matrix degeneration of gingival tissue (GT).[4] Particularly MMP-9, which is the major gelatinase found in the gingival crevicular fluid in patients with periodontal disease as early as stage I inflammation.[4] Most of the MMPs, including MMP-9, are synthesized by a variety of inflammatory cells like resident macrophages and polymorphonuclear leukocytes and noninflammatory cells such as fibroblasts, chondrocytes, and keratinocytes. Usually, they are secreted as inactive enzymes or latent forms.[29] These latent forms of MMPs are activated to complete their biological activity in the extracellular compartment or near cell membrane by MMPs themselves[30] or other molecules like plasmin[31] or serine proteinase.[32] Alterations in the expression of latent and active forms on collagenases (MMP-1, -8, and -13) on different degree of gingival inflammation have been established.[33] The aim of the present study was to compare the levels of MMP-9 in both active and latent form in GT extracts collected from patients with periodontitis at different rate of progression and compare it with individuals with well-maintained clinical periodontal health as defined by the American Academy of Periodontology and European Federation of Periodontology in their classification of periodontal and peri-implant diseases and condition.[3435]

MATERIALS AND METHODS

The current study was approved by the institute's ethical committee. It was a cross-sectional human clinical study conducted at the department of periodontics and the laboratory analysis was performed at an outside facility. 73 outpatients belonging to the Kolkata population who were visiting the department of periodontics were analyzed for eligibility for the study. Among those patients, 60 members aged between 18 and 50 years were selected and divided into three groups (20 each), Group A (GA), Group B (GB), and Group C (GC) based on the inclusion and exclusion criteria. The sample size was calculated using nMaster software with previous study reference using power hoc analysis with effect size 90 and interpreted sample size was 20 in each group.[36] Informed consents were obtained from all the participants of the study after explaining about the procedure of the study.

Both GA and GB consist of patients with periodontitis at stages II and III. The difference in GA and GB was in grading the disease. GA consists of periodontitis with grades A and B and on the other hand, GB consists of periodontitis with grade C.[34] GC was the control group and it consists of individuals with well-maintained clinical periodontal health which was characterized by absence of clinical inflammation or minimal levels of clinical inflammation with normal periodontal support as described by Lang and Bartold.[35] Other forms of healthy periodontium namely pristine periodontal health, periodontal disease stability, and periodontal disease remission/control were not included in the study.

Inclusion criteria common for both GA and GB include, (a) interdental clinical attachment level (ICAL) at the site of greatest loss ≥3 mm, (b) radiographic bone loss extending from coronal third (15% to 33%) to middle or apical third of the root, (c) periodontal probing depth (PPD) ≥5 mm, (d) no tooth loss or ≤4 teeth loss due to periodontitis. The difference in the inclusion criteria for GA and GB changes while grading the periodontitis. Inclusion criteria for GA include, (a) slow and moderate rate of progression of periodontitis (% bone loss/age: ≤0.25–1.0), (b) heavy plaque deposits with slight levels of periodontal destruction or destruction was proportionate with plaque deposits. Inclusion criteria for GB include, (a) rate of progression of periodontitis was rapid (% bone loss/age; >1), (b) biofilm deposits and periodontal destruction were incommensurate, i.e., destruction exceeds expectation, (c) specific clinical patterns like molar/incisor pattern, (e) early onset of disease. Exclusion criteria for the study include (a) smokers, (b) individuals with <20 remaining teeth, (c) patients with diabetes or any other systemic disorders or conditions or medication which can affect the periodontium, (d) patients who had undergone any periodontal therapy for the past 6 months, (e) individuals who were not fit for any periodontal surgery [Figure 1].

Figure 1

Flowchart representing the study population: n - number of samples

Plaque index (PI)[37] and gingival index (GI)[38] were used to assess the periodontal health status of the patients. UNC 15 periodontal probe was used to measure PPD on all four sides (buccal, mesial, distal, and lingual/palatal) of the tooth and CAL from the cementoenamel junction to the nearest millimeter. All the clinical parameters were measured in a research setting by a single examiner under the supervision of the head of the department to avoid the random measurement error.[39] Intraexaminer calibration was assessed using kappa statistics. Repeated measurements of PPD of two quadrants in eight patients were used for testing the reliability. The resulting kappa value was 0.86 which was found to be acceptable.

GT samples from periodontally healthy individuals (GC) were collected during gingivectomy procedures performed under asepsis and local anesthesia (LA) to increase the clinical crown for prosthetic or esthetic reasons. For GA and GB, the tissue samples were collected during the Modified Widman flap procedure which was performed as previously described by Ramfjord and Nissle under LA and total asepsis.[40] All the biopsy samples collected for the study were GTs excised during the above-mentioned procedures. Infiltration LA was avoided to prevent the biopsy site contamination with LA. Interproximal sites with probing depth more than 5 mm were selected for sample tissue harvesting in both GA and GB groups. All the patients were prescribed analgesic medication as required postsurgery. All tissue samples collected were divided into two parts, one was preserved in −80°c for further enzymatic analysis and the other was kept in 10% formalin solution for histological and immunofluorescent studies.

The frozen tissue samples were allowed to thaw at room temperature to prepare the phosphate buffer saline extract (PBS ex) which was then used to analyze the active and latent forms of MMP-9. To prepare the extract, the defrosted tissues were suspended in PBS (PBS, 10 mM phosphate buffer pH 7.4, 150 mM NaCl) containing protease inhibitors cocktail (Sigma Chemical Co, St. Louis, MO, USA), minced, and incubated for 10 min at 4°C. The suspension was centrifuged at ×12,000 g for 15 min at 4°C and the supernatant was collected in new centrifuge tubes as PBS extracts.[41] The PBS extracts were preserved at −70°C for future studies.

Proteins in the tissue extracts were estimated using the Lowry method.[42] Gelatin zymography was used to quantitate the activities of latent and active forms MMP-9 using PBS ex of GTs collected from the sample population. Previously established protocols were followed for the gelatin zymography procedure.[41] The activities of latent and active MMP-9 were measured by analyzing the lysis bands at 92 kDa and 88 kDa positions, respectively, using densitometric measurements by lab image software. Western blotting was used to assess the activity of MMP-9 in PBS ex. Histological examination of the excised GT was performed to find the architecture and the amount of inflammation. Immunofluorescence of GT samples was performed to see the MMP-9 expression. All the three procedures were performed by the same steps previously followed.[41]

The statistical analysis of the data collected was done using GraphPad Instat-3 software. Comparison between groups was done using one-way analysis of variance followed by Student–Newman–Keuls test. Data were fitted using Sigma plot represented as means ± standard error of mean Pearson's correlation study was done to determine the correlation between the clinical parameters and MMP-9 levels. P < 0.05 was accepted as level of significance, P < 0.001 was considered as very highly significant, P < 0.01 was considered as highly significant, P < 0.05 was considered as significant. P > 0.05 was considered as nonsignificant (NS).

RESULTS

The current study consists of 60 individuals aged between 18 and 50 years (mean ± standard deviation: 36.86 ± 4.35 years) who were recruited from the outpatient unit of the department of periodontics. The recruited individuals were divided in to three groups based on the inclusion and exclusion criteria. Clinical parameters were recorded, and tissue samples were collected from all the 60 individuals and statistically analyzed. All the clinical parameters (PI, GI, CAL, and PPD) recorded were higher in GA and GB when compared with GC [Table 1]. The activities of active and latent MMP-9s were analyzed using gelatin zymography. In gelatin zymography, the activities of latent MMP-9 (lMMP-9) and active MMP-9 (aMMP-9) were measured by analyzing the lysis bands at 92 kDa and 88 kDa position, respectively, using densitometric image software [Figure 2a]. Western blotting was done for the confirmation of the high levels of MMP-9 obtained by gelatin zymography in GT of all three groups [Figure 2b].

Table 1

Clinical parameters of the participants

Groups	GI	PI	PD	CAL
Group A (n=20)	2.07±0.26	2.13±0.35	6.07±0.89	8.00±1.36
Group B (n=20)	1.20±0.41	1.07±0.26	8.13±1.85	8.60±1.80
Group C (n=20)	0.00	0.06±0.51	1.06±0.51	0.00

GI – Gingival index; PI – Plaque index; PD – Probing depth; CAL – Clinical attachment level, n – Number of samples

Figure 2

(a) Gelatinase zymogram of PBS extracts showing lysis in bands corresponding to 92 kDa (latent MMP-9) and 88 kDa (active MMP-9) in Group A, Group B, and Group C; (b) Western bolting of MMP-9 in Group A, Group B, and Group C. Western blotting of Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was done to confirm equal protein loading in abode blots. Expression of MMP-9 was increased by 2.15-fold in Group A and 1.64-fold in Group B when compared with Group C. PBS – Phosphate buffer saline; MMP-9 – Matrix metalloproteinases

The levels of both aMMP-9 and lMMP-9 were significantly higher in GA and GB when compared with that of GC [Table 2]. In GA, aMMP-9 was found to be almost 2 (2.05 Arbitrary Unit [AU]) folds higher when compared with GC (P < 0.01). In GB, aMMP-9 was 1.32 AU and still higher than GC (P < 0.01). Between GA and GB, the aMMP-9 levels were higher in GA which was statistically significant (P < 0.05). The levels of lMMP-9 were also higher and statistically significant in both GA and GB, when compared with that of GC. In GA, it was almost 2.5 folds (2.54 AU) higher than GC (P < 0.001), and in GB, it was almost 1.5 times (1.74 AU) higher than GC (P < 0.001). In both the cases, the differences were significant. Between GA and GB, the former has higher levels of lMMP-9 which was also statistically significant (P < 0.001).

Table 2

Arbitrary activity of active and latent matrix metalloproteinases - 9 levels (arbitrary unit)

Groups	aMMP-9	lMMP-9
Group A (n=20)	2.05	2.54
Group B (n=20)	1.32	1.74
Group C (n=20)	0.935	1.08

aMMP-9 – Active matrix metalloproteinases -9; lMMP-9 – Latent matrix metalloproteinases -9; n – Number of samples

Pearson's correlation coefficient was performed to analyze the correlation between the different MMP-9 levels and all the clinical parameters recorded. In case of aMMP-9, there was no correlation between GI, PI, and CAL and the levels of aMMP-9 in all the three groups [Table 3]. But a significant correlation was found between PPD and levels of aMMP-9 in GA (P < 0.001). However, the other two groups, i.e., GB and GC showed no such correlation between the probing depth and aMMP-9 levels. There was no correlation found between the lMMP-9 and all the clinical parameters recorded in all the three groups [Table 4].

Table 3

Pearson’s correlation coefficient values with respective P values of active-matrix metalloproteinases -9

Groups		GI		PI		PD		CAL
Group A	PBS ex	r=0.4523	P=0.0905	r=0.4735	P=0.0746	r=0.8879	P<0.0001*	r=0.4806	P=0.0697
Group B	PBS ex	r=0.3030	P=0.2722	r=0.1080	P=0.7017	r=0.1284	P=0.6485	r=0.0232	P=0.9347
Group C	PBS ex	NA	NA	r=0.9081	P=0.0001	r=0.6997	P=0.0053	NA	NA

*Extremely significant (P<0.001). P<0.05 was considered as significant. GI – Gingival index; PI – Plaque index; PD – Probing depth; CAL – Clinical attachment level; PBS ex – Phosphate buffer saline extracts; P – Probability value; r – Pearson’s correlation; NA – Not applicable

Table 4

Pearson’s correlation coefficient values with respective P values of latent matrix metalloproteinases-9

Groups		GI		PI		PD		CAL
Group A	PBS ex	r=0.1153	P=0.7212	r=0.1710	P=0.5951	r=0.3798	P=0.2234	r=0.1738	P=0.5890
Group B	PBS ex	r=0.1539	P=0.6330	r=0.3494	P=0.2656	r=0.5445	P=0.0672	r=0.4270	P=0.1663
Group C	PBS ex	NA	NA	r=0.0074	P=0.9838	r=0.0007	P=0.9838	NA	NA

P<0.05 was considered as significant. GI – Gingival index; PI – Plaque index; PD – Probing depth; CAL – Clinical attachment level; PBS ex – Phosphate buffer saline extracts; NA – Not applicable; P – Probability value; r – Pearson’s correlation

Microscopic analysis of the GT samples from GC revealed normal gingival architecture with few inflammatory cell infiltrations in the connective tissue [Figure 3a]. In GA and GB, the histological examination showed the presence of rete pegs and a larger amount of inflammatory cell infiltration with numerous blood vessels in the connective tissue [Figure 3b and c]. Immunofluorescence of the GT samples showed no strong reactivity in the GC when visualized by indirect fluorescence [Figure 3d]. In GA and GB, high reactivity for MMP-9 expression in both epithelial and connective tissue was seen [Figure 3e and f].

Figure 3

Histology (×40) and immunofluorescence staining of the gingival tissue samples in all the three groups (a and d– Group C; b and e– Group A; c and f– Group B) (arrow marks showing the presence of epithelial rete pegs and inflammatory cells infiltration)

DISCUSSION

A shift from a well-controlled intracellular pathway to a metalloproteinase pathway occurs, when healthy gingiva develops periodontitis.[43] MMPs are secreted as inactive proenzymes (zymogens) and are thought to be activated in the tissue by cleavage of the polypeptide. Hence, MMPs could be available in both inactive and active forms which can be detected using zymographic analysis of gelatinase. Comparing the results of the current study with previous studies was tricky as the inclusion and exclusion criteria which was based on the new classification given by the AAP and EFP employed in this study have never been used in previous studies involving MMPs.[3435] Furthermore, many methods have been employed to estimate the levels of MMPs and the samples collected for this estimation were also different, oscillating from gingival crevicular fluid to blood, making the comparison between studies more and more difficult.

The levels of aMMP-9 were higher in periodontitis (GA – slow/moderate rate of progression) patients when compared with the healthy individuals, which was in accordance with a previous report by Soell et al. In the same study, they also found out that there was no statistically significant difference between the levels of aMMP-9 in health and disease, which was in contrast to the current findings. Even though the samples collected by Soell et al. included GTs, the method to estimate the levels of MMP-9 was different (ELISA test) and no clear definition for periodontitis was given in that study.[44]

A previous report by Kim et al. showed an increase in the levels of MMP-9 in systemically healthy patients with periodontitis when compared with periodontally healthy individuals which was in line with the current observation. The inclusion criteria used for recruiting systemically healthy patients with periodontitis in the former study coincide with GA in the current study. In both the studies, GT samples which were collected in sites with more than 5 mm PPD and used for MMP-9 analysis.[45] In contrast to the current study, Séguier et al. found a significant increase only in the levels of aMMP-9 in periodontitis patients, even though the definition of periodontitis, the sample collected and site of sample collection, and the methods used for the analysis of aMMP-9 levels were similar in both the studies.[46] In a study by Bildt et al., the levels of both active and latent MMP-9 in healthy individuals were more than that of the periodontitis patients which was not in line with the current observation.[47] Bildt et al. have mentioned that the presence of subclinical inflammation and age difference between the samples of the two groups might be the reason for this unusual observation.[47]

Active MMP-9 is short-lived[48] and activations of latent MMP-9 are prevented by endogenous inhibitors like TIMP.[49] This may the reason for the difference in the levels of aMMP-9 (2.05 AU) and lMMP-9 (2.54 AU) in GA. In the current study, the levels of aMMP-9 and lMMP-9 in GB were almost 1.5-fold higher than the healthy group with was also observed by Brakenhoff et al. but in a bit higher than the present level reported (2 times higher) but it was not statistically significant.[36]

Attempts were made to correlate the clinical parameters with active and latent forms of MMP-9 levels in GT samples analyzed with the help of gelatin zymography. A highly significant correlation (r = 0.8879/P < 0.000) was observed between PPD and the level of aMMP-9 in slowly/moderately progressing periodontitis patients. No correlation was seen among other clinical parameters (PI, GI, and CAL) with aMMP-9. No correlations were observed among the lMMP-9 level and all the clinical parameters (PI, GI, PPD, and CAL) recorded.

Unlike the current study, Pozo et al. found no significant correlation between the disease severity and MMP 9 activity in slow/moderately progressing periodontitis-affected patients when zymography was used to assess the MMP-9 activity.[50] Pozo et al. study was an interventional study unlike the current which is an observational one.[50] An observational study by Alfant et al. also has found no such correlation between the levels of MMP-9 in slow/moderately progressing periodontitis and PPD.[51] This can be explained by the fact that the study by Alfant et al. used fluorometric kits to measure the activity of the MMPs.[51] Interestingly in Pozo et al. study, when fluorogenic substrate assay was used to assess the MMP activity, they found a significant correlation between PPD and CAL with the actual MMP activity like the current study. The point to be noted here is that the activity of both MMP-8 and MMP-9 was analyzed in that study.[50]

Rapidly progressing periodontal group showed no correlation between the levels of MMP-9 and all clinical parameters recorded. Ironically, a similar observation was reported by Alfant et al. who used different method to analyze the activity of the MMPs.[51] Observation contrast to the current study was reported by Gonçalves et al. in 2013 where they have found a correlation between probing depth and MMP-9 levels in rapidly progressing periodontitis.[52] One possible explanation for this contrasting observation might be because of the difference in the study type. The previous study and the current study were observational ones where else the latter one was an interventional study, i.e., samples were collected before and after treatment.[52]

Histology of GT samples was done to examine the signs of inflammation at the tissue level. The examination of GT from GC revealed normal architecture with some inflammatory infiltrates which was in accordance with the latest classification of healthy gingiva given by Lang and Bartold which conclude that in well-maintained clinical periodontal health, there can be minimal levels of inflammation.[35] The microscopic analysis of the diseased tissue revealed rete pegs and inflammatory cell infiltration with numerous new blood vessels in the connective tissue. The MMP-9 levels were also high in two diseased groups (GA and GB) which was similar to previous reports.[114749] Hence, it can be hypothesized that the levels of MMP-9 might be found higher in inflamed periodontal tissues. To observe the MMP-9 distribution in GT, immunofluorescence study of tissue samples was done, which claimed that the MMP-9 level was higher in GA and GB than that of GC. Another observation was that MMP-9 was expressed in both epithelial and connective tissue areas in the diseased groups. This observation of the present study indicates that MMP-9 was synthesized by the epithelial cells also, which was in accordance with the study by Smith et al., who also demonstrated that keratinocytes have a role in inflammation mediated by MMP-9.[53] Expression of MMP-9 in this immunofluorescence study was found to be higher in rapidly progressing periodontitis than that of healthy gingiva.[53]

An increase in the activity of MMPs like MMP-9 in slow/moderately and rapidly progressing periodontitis as noted in the current study can be used as a biomarker for the early detection of periodontitis and also as an useful tool to assess the prognosis of the disease.[54] This knowledge can also be used for developing periodontal disease prevention protocols and targeted therapy toward MMPs like MMP inhibitors which in adjunct nonsurgical periodontal therapy could permit enzymatic modulations to control periodontal diseases and peri-implantitis in addition to traditional bacterial control procedures.[54] The major limitation of the current study was that it is an observational study. Future studies should focus on the expression of MMP-9 in healthy GT and compare it with the changes of MMP-9 expression in diseased GT, before and after therapy instead of cross-sectional observational studies like the current one. Such studies can provide more insights about the clinical significance of MMP-9 expression in periodontal diseases. So that newer diagnostic techniques and better treatment approaches can be formulated.

CONCLUSION

The levels of aMMP-9 and lMMP-9 were found higher in patients with moderate severity periodontitis (both slow/moderately and rapidly progressing) when compared to the levels of MMP-9 in clinically healthy individuals. A correlation between the periodontal probing depth and aMMP-9 was found in patients with slow/moderately progressing periodontitis. However, further studies are required to confirm and interpret the clinical significance of the present findings.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.