Oxidative Stress and Antioxidants in the Diagnosis and Therapy of Periodontitis

Oxidative stress has been implicated in the pathogenesis of numerous diseases. However, large interventional studies with antioxidants failed to show benefits in the prevention or treatment of cardiovascular diseases, cancer, or diabetes mellitus. Numerous clinical studies have confirmed the association of oxidative stress markers and periodontitis. Technical and biological variability is high for most of the analyzed markers and none of them seems to be optimal for routine clinical use. In a research setting, analysis of a palette of oxidative stress markers is needed to cover lipid peroxidation, protein oxidation, and the antioxidant status. The source of reactive oxygen species and their role in the pathogenesis of periodontitis remains unclear. Interventional experiments indicate that oxidative stress might be more than just a simple consequence of the inflammation. Small studies have confirmed that some antioxidants could have therapeutic value at least as an addition to the standard non-surgical treatment of periodontitis. A clear evidence for the efficiency of antioxidant treatment in large patient cohorts is lacking. Potentially, because lowering of oxidative stress markers might be a secondary effect of anti-inflammatory or antibacterial agents. As the field of research of oxidative stress in periodontitis gains attraction and the number of relevant published papers is increasing a systematic overview of the conducted observational and interventional studies is needed. This review summarizes the currently available literature linking oxidative stress and periodontitis and points toward the potential of adjuvant antioxidant treatment, especially in cases where standard treatment fails to improve the periodontal status.


INTRODUCTION
Oxidative stress is both, a pathomechanism involved in numerous inflammatory diseases causing damage to lipids, nucleic acids and proteins-oxidative distress, as well as an important physiological process that enables the immune system to cope with microorganisms and intracellular cell signaling-oxidative eustress (Sies et al., 2017). Which edge of the sword is the dominant depends on the delicate balance between the production of reactive oxygen/nitrogen species and the antioxidant capacity of the tissue. The physiological functions of free radicals have been neglected for years and so, much more is known about the pathological role of oxidative stress. A variety of free radicals is produced and interacts with a variety of substrates. This leads to a palette of biomarkers that can be used for the assessment of oxidative stress-induced damage (Frijhoff et al., 2015).
Oxidative stress is usually defined as a disbalance of the production of free radicals and antioxidant mechanisms (Kopáni et al., 2006). However, free radicals are not a simple negative byproduct of oxygen metabolism. They are involved in immune responses, liver metabolism, but also in intracellular signaling pathways (Espinosa-Diez et al., 2015). The mechanisms involved in the physiological intracellular role of free radicals include the modulation of cysteine residues of redox-sensitive enzymes and other regulatory proteins (Finkel, 2011;Russell and Cotter, 2015). It has been hypothesized that under physiological conditions even high concentrations of any one primary reactive oxygen or nitrogen species does not lead to oxidative damage, as the cell has preventive and reparative tools to cope with the radical. The reactions of superoxide, nitric oxide, and other primary reactive species are reversible and are ideal for intracellular signaling (Weidinger and Kozlov, 2015). Thus, measuring oxidative stress using any one marker can lead to wrong interpretations. This might include the hope that administration of antioxidants will effectively treat diseases associated with oxidative stress. As recently hypothesized, one of the prototypic oxidative stress diseases-diabetes mellitus might be actually a consequence of deficiency of reactive oxygen species rather than of oxidative damage (Watson, 2014). It should not escape our notice that this proposed mechanism might suggest that antioxidants could increase the risk rather than prevent metabolic diseases.
Periodontitis is an inflammatory disease affecting supporting structures of the teeth leading at the end to loss of alveolar bone and teeth (Kinane et al., 2017). The main causative factor are microorganisms that colonize the subgingival dental plaque inducing an inflammatory host response. The inflammation affects, however, also the surrounding healthy tissue ultimately leading to the destruction of the periodontium (Kinane et al., 2011). Although lipopolysaccharide and proteolytic enzymes are essential in periodontitis, exaggerated inflammatory response, genetic predisposition, smoking, bad oral hygiene, and malnutrition are important in pathogenesis of periodontitis as well (Laine et al., 2012).
The role of oxidative stress in periodontitis has been postulated already decades ago (Shapira et al., 1991;Chapple, 1997). However, the suggested involvement was not clear. Some studies showed that leukocytes from patients with periodontitis are exhausted and have a low oxidation activity (Loesche et al., 1988), other studies pointed toward higher production of free radicals by leukocytes from periodontitis patients (Kimura et al., 1993). The contradictory findings from these studies might be related to the dynamics of the mechanisms during the pathogenesis of the disease, but they might also be explained by the different forms of periodontitis (Biasi et al., 1999).
The term oxidative stress is vague, similarly, antioxidants might affect many processes not directly related to free radical generation or action (Niki, 2016). This makes the interpretation of studies focusing on oxidative stress in periodontitis difficult. A systematic review of these clinical studies and animal experiments might, thus, be needed.

BIOMARKERS OF OXIDATIVE STRESS AND PATHOGENESIS OF PERIODONTITIS
Surprisingly, many observational studies analyzing oxidative stress in patients with periodontitis had relatively consistent results with higher oxidative stress markers in either saliva or blood and/or decreased antioxidant status in comparison to controls. The summary table of the identified studies can be found in Table 1. One of the largest observational studies has shown that the antioxidant status in blood, analyzed as vitamin C, bilirubin, and calculated total antioxidant capacity was inversely associated with mild and severe periodontitis (Chapple et al., 2007b). The more severe periodontitis, the clearer the association. Additionally, in a subgroup of neversmokers the antioxidants seemed to protect against development of periodontitis. Others smaller studies confirmed these results. Total antioxidant capacity was lower in plasma or serum of patients with chronic periodontitis (Chapple et al., 2002;Brock et al., 2004;Konopka et al., 2007). Similarly, superoxide dismutase activity  along with catalase and glutathione peroxidase activity (Tonguç et al., 2011) as important contributor to the total antioxidant capacity was found to be lower in periodontitis. In line with these results from blood, in the majority of studies, the antioxidant status was lower also locally-in saliva. The total antioxidant status/potential/capacity is in general the ability of a tissue to resist artificially induced oxidative stress, but studies differ in the analytical approaches. Nevertheless, the local antioxidant capacity was found to be lower in saliva from patients with an aggressive form of periodontitis when compared to chronic periodontitis (Acquier et al., 2017). Patients with chronic periodontitis have lower antioxidant capacity than control patients (Zhang et al., 2015;Ahmadi-Motamayel et al., 2017). Although the total antioxidant capacity is not a specific marker of antioxidant power, uric acid, glutathione peroxidase (Miricescu et al., 2014), and reduced glutathione (Gumus et al., 2009) as specific antioxidants were reported to be significantly lower in saliva of patients with chronic or aggressive periodontitis. On contrary, activities of the major antioxidant enzymes were found to be higher in chronic periodontitis patients in all investigated samples, i.e., plasma, erythrocytes, and in the gingival tissue (Panjamurthy et al., 2005). Similarly, SOD2 and GPX1 genes were overexpressed in the gingiva of chronic periodontitis patients (Duarte et al., 2012). Lactoferrin, myeloperoxidase and interleukin 1 beta were all positively correlated with the clinical markers of periodontal damage (Wei et al., 2004). However, whether such associations of higher antioxidant and pro-inflammatory response are a consequence or cause of severe periodontitis cannot be judged only from observations.
Regarding the analyzed markers of oxidative stress, a comparison of the published studies is complicated, if not impossible due to the huge variability of measured markers.   Malondialdehyde, 8-hydroxydeoxyguanosine (Konopka et al., 2007;Almerich-Silla et al., 2015), protein carbonyls (Nguyen et al., 2016;Atabay et al., 2017), thiobarbituric acid reacting substances (Borges et al., 2007), nitric oxide, advanced oxidation protein products, lipid peroxidation products (Gomes et al., 2017), 8-isprostanes (Hickman et al., 2011) were all higher in patients with periodontitis. The most commonly measured markers of oxidative stress seem to be malondialdehyde and thiobarbituric acid reacting substances pointing toward oxidative damage of lipids, especially, lipid membranes. Lipid peroxidation was higher in saliva (Tothova et al., 2013), serum (Tonguç et al., 2011), and in the gingival tissue (Panjamurthy et al., 2005) of patients with chronic or aggressive periodontitis. Correlational studies confirmed a positive association of these markers with periodontal status scores (Chapple et al., 2007b;Tamaki et al., 2014b). The less commonly measured markers related to oxidative damage as mitochondrial DNA (Canakci et al., 2006), micronuclei and nuclear abnormalities (Zamora-Perez et al., 2015), as well as a leukocyte telomere length shortening (Masi et al., 2011) were all higher in periodontitis as well.
Oxidative stress was found to be involved in the pathogenesis of many diseases besides periodontitis. Virtually, almost all inflammatory diseases lead to increased oxidative stress. This in turn can trigger more damage to the tissues, not excluding the gingival tissue and, thus, worsening periodontitis. There are several studies describing oxidative stress in systemic diseases with regard to the periodontal status. Nguyen et al. (2016) investigated patients with the acute coronary syndrome with or without chronic periodontitis, patients with periodontitis and healthy controls. Lipid, protein, and DNA oxidation markers were higher in periodontitis than in the control group (Nguyen et al., 2016). Another study with rheumatoid arthritis and chronic periodontitis confirmed higher oxidative stress markers in plasma and lower antioxidant capacity in both groups when compared to healthy control. However, if both comorbidities were present, there was no further enhancement of oxidative stress (Sezer et al., 2013).
Taken together, observational cross-sectional studies confirmed the association of oxidative stress and periodontitis. Higher oxidative stress and lower antioxidant status can be detected in plasma, saliva as well as in the gingival crevicular fluid of patients with various clinical forms of periodontitis. These findings support the use of body fluids, but especially the non-invasive diagnostic fluid saliva, as suitable sample types for diagnostics or monitoring the course of periodontitis. Current data do not support the use of a single oxidative stress marker. It is likely that a set of markers covering both, oxidative damage and antioxidative status, will be needed. The low specificity of oxidative stress markers calls for caution when interpreting the results even if more than one marker is used. The inter-individual and intra-individual variability of the analyzed markers is very high. This prevents their use at the level of individual diagnostics. The overview of the published observational studies shows the enormous heterogeneity of the patient populations as well as the used analytical tools. In a metaanalysis focusing on systematic oxidative stress, it was shown that higher malondialdehyde and nitric oxide as well as lower total antioxidant capacity of plasma/serum characterize patients with periodontitis in comparison to controls (Liu et al., 2014). Based on our overview, taken into account systemic and local oral biomarkers of oxidative stress, it is clear that there is a need for both, the use of a wider palette of markers to analyze oxidative stress and its causes in more detail, and the introduction of new biomarkers with a better sensitivity/specificity profile in specific subgroups of patients. Of special importance is the small sample size in most studies. A collaborative effort with a multi-center recruitment of patients and a standardized consensus protocol in the pre-analytical and analytical phase is highly needed.

OXIDATIVE STRESS AND TREATMENT OF PERIODONTITIS
The implication of oxidative stress in the pathogenesis of cardiovascular diseases and cancer as the major causes of death in the combination with the widespread availability of dietary antioxidants started a hype that was supported by the hypothesis that aging is caused by oxidative stress (Harman, 1956;Finkel and Holbrook, 2000). However, the hype was quickly over as large clinical studies revealed that antioxidants or at least antioxidant vitamins were not able to prevent any of the diseases of aging (Coulter et al., 2006;Sesso et al., 2008;Myung et al., 2013). Some studies even revealed a slight but increased risk in patients taking antioxidants in preeclampsia (Rumbold et al., 2006) and lung cancer (Alpha-Tocopherol, 1994). One metaanalysis showed that taking antioxidant may even increase the all-cause mortality by 5% (Bjelakovic et al., 2007). This might be related to the terminological and mechanistic confusion about antioxidants, which might have a very indirect effect on the production or effects of reactive oxygen species. In an era of evidence-based medicine the clear conclusion is that antioxidant dietary supplements should undergo clinical evaluation before marketing similarly to other medicinal drugs (Bjelakovic et al., 2012). Recent mechanistic experiments shed light on the details how antioxidants may stimulate tumor growth (Sayin et al., 2014) or increase the risk of metastasis (Piskounova et al., 2015). Antioxidants are highly variable in their mechanism and structure. Diseases and patients vary even more. So, negative results from oncology or cardiology should not be generalized to other diseases including periodontitis. It is very likely that the role of oxidative stress changes during disease progression and, thus, the potential antioxidant treatment affecting not only oxidative damage but also the inflammatory process might have different affects at various stages of the complex pathogenesis. The issue of antioxidants that mostly do not affect free radicals and their action in vivo, but rather interfere with variable cellular signaling pathways has been reviewed in a recently published paper (Azzi, 2017).
Many published studies have analyzed the effects of periodontitis treatment on oxidative stress ( Table 2). Studies of special interest are those with covariates or comorbidities that were taken into account. A trial by Guentsch et al. (2008) examined healthy subjects and patients with periodontitis further  divided into smokers and non-smokers. While smokers with periodontitis displayed highest malondialdehyde concentration and highest glutathione peroxidase activity along with lowest total antioxidant capacity in saliva. The non-surxgical treatment helped to normalize the values regardless of the smoking status (Guentsch et al., 2008). Several other trials confirmed the consistency of these findings (Chapple et al., 2007a;Abou Sulaiman and Shehadeh, 2010;Hendek et al., 2015). On the other hand, a severe systemic condition such as type 2 diabetes may lead to inefficiency of non-surgical therapies to improve the periodontal status or oxidative stress (Koromantzos et al., 2012). In general, most of the studies clearly show an improvement of oxidative damage after standard treatment of periodontitis. The studies, however, vary greatly regarding sample types, markers measured, and most cohorts were very small and highly variable. Thus, making relevant conclusions or recommendations for the clinical dentistry is difficult, if not impossible. Future research efforts should focus on the lack of uniformity and standardization as well as on the issue of the low informative value of small patient cohorts.
Although the causality of the association between oxidative stress and periodontitis is everything but clear, some clinical studies already tested antioxidants in periodontitis. One small, placebo controlled, randomized, and double-blind study showed, that single application of lycopene gel to periodontitis patients improved clinical attachment and decreased oxidative stress in gingival crevicular fluid (Chandra et al., 2013). Nevertheless, a published systematic review on antioxidant treatment of periodontitis revealed that a consistent effect in randomized clinical trials was found only for lipophile antioxidants such as lycopene and vitamin E, but not for hydrophile antioxidants such as vitamin C (Muniz et al., 2015). This might be related to the vulnerability of lipids to oxidative damage, but also to mitochondria as the site of effect of some antioxidants. Interestingly, antioxidants targeting directly mitochondria have been shown to be effective in decreasing inflammatory activity and organ damage in animal model of sepsis (Lowes et al., 2013).

ANIMAL EXPERIMENTS
The high number of observational and interventional studies analyzing the association between oxidative stress and periodontitis indicates that there are many open questions that cannot be answered by more and more clinical studies. Many of the questions need controlled conditions in experiments. The number of animal experiments analyzing the role oxidative stress in periodontitis is small, but it increases. Periodontitis is mostly induced by ligature placement around the first molars of the animals or local injection of periodontal pathogens or their toxins (Genco et al., 1998;Fine, 2009;Oz and Puleo, 2011). In such a rat model, it was shown that periodontitis leads to an increased production of reactive oxygen species and markers of oxidative damage (Ekuni et al., 2010). In addition, oxidative stress induced by periodontitis seems to be associated with the dynamics and severity of the periodontal inflammation (Bosca et al., 2016).
Some animal experiments focus on the distant effects of periodontitis that seem to be mediated by oxidative stress. It was shown that mitochondrial DNA is oxidatively modified in the liver, kidney, heart, and brain of rats with induced periodontitis . Another similar experiment by the same group has shown that periodontitis might worsen ethanolinduced liver damage (Tomofuji et al., 2008). The oxidative damage to the heart, but also endothelial dysfunction and resulting atherosclerosis induced by experimental periodontitis can be prevented by antioxidant treatment (Ekuni et al., 2009;Ozdem et al., 2017;Saito et al., 2017). A majority of the published experiments focus on the use of antioxidants such as vitamin C (Tomofuji et al., 2009b), N-acetylcysteine (Toker et al., 2009), or resveratrol (Tamaki et al., 2014a), but also drugs with an antioxidant activity beyond their main mechanism of action (de Araujo Junior et al., 2013;Culic et al., 2014;Oktay et al., 2015). Dietary interventions have also been investigated-high cholesterol diet seems to worsen periodontitis (Tomofuji et al., 2006). On contrary, the phytoestrogen genistein and cacaoenriched diet was shown to be protective against periodontal damage and oxidative stress induced by periodontitis in mice and rats, respectively (Tomofuji et al., 2009a;Bhattarai et al., 2017). The most promising candidate drug at least according to animal experiments is melatonin. This amphiphile molecule has an optimal distribution in the tissues and can, thus, reach the periodontal tissues even after systemic administration (Köse et al., 2017). Of special clinical relevance is the induction of periodontitis in diabetic rats since periodontitis is a common complication of diabetes in humans. Melatonin was able to prevent alveolar bone loss also in this experimental model (Kose et al., 2016). However, regarding the mechanism of action, it is not clear whether the protective effect of melatonin is due to its direct antioxidant characteristics or due to its immunomodulatory effects that might reduce oxidative stress indirectly (Kara et al., 2013). This uncertainty is not specific for melatonin. Any antioxidant might affect the immune response and, thus, have anti-inflammatory properties. Beyond established systemic antioxidants novel approaches with a local periodontal application are tested (Saita et al., 2016). Experimental tools such as genetically engineered mice that produce luciferase under the regulation of transcription factors related to oxidative stress and antioxidant response have been developed and might greatly improve our understanding of the role of oxidative stress in periodontitis (Kataoka et al., 2016). The new treatment options together with new and improved models could be very helpful in the fight against this widespread and serious disease.

CONCLUSION
The role of oxidative stress in periodontitis is not clear despite decades of research. Numerous studies have been published showing the potential of oxidative stress markers for screening, diagnosis or monitoring of the disease, but none is in routine clinical use. Similarly, animal experiments, as well as most of the interventional studies in patients, indicate that antioxidant treatment should be effective in the therapy of periodontitis, but no such treatment has been approved. The lack of translation could be either due to the lack of strong evidence for the clinical usefulness or due to obstacles in the application of the results including the low or absent commercial interest from major stakeholders. From a research perspective, an important issue is the lack of specificity-both, in diagnostics and treatment. Not even the source of free radical production is clear. While some studies point toward neutrophils (Katsuragi et al., 2003), others show that bacteria actively producing reactive oxygen species might contribute to oxidative stress in periodontitis (Huycke et al., 2002;Vlkova and Celec, 2009). It is of crucial importance that the number of conducted animal experiments in this field is increasing, especially of those focusing on the dynamics of oxidative stress during disease progression. The antioxidant treatment might be effective only in a subset of patients during a specific stage of periodontitis. The shift from pure observations to interventions and animal experiments that can be followed in the published literature in the recent years is highly positive and should bring this field of research closer to true clinical applications despite chronic lack of financial support and human resources.

AUTHOR CONTRIBUTIONS
L'T has analyzed the data from the literature, prepared tables, and drafted the manuscript; PC has designed the review, conducted the literature search, and drafted the manuscript.