The role of the oral microbiome in smoking-related cardiovascular risk: a review of the literature exploring mechanisms and pathways

Researchers collected whole mouth samples by swabbing multiple surfaces of the mouth [38] or collecting an alcohol- [43, 47‐50] or saline-based [34] mouthwash to analyze microbial populations across multiple oral microenvironments. Similar to whole mouth samples, saliva microbiome composition may be influenced by other oral microbial niches, but sample collection occurred through spontaneous passive [36, 37, 39, 40, 45, 51‐54] or stimulated [46] salivary fluid collection in the reviewed studies. In saliva or buccal mucosa samples of cigarette smokers, the majority of studies demonstrated an increased relative abundance (RA) of Actinomyces [36, 37, 48], Actinomyces species [46, 52], Aggregatibacter [53], Bacteroides [52], and Lachnoanaerobaculum [39] versus non-smokers. Cigarette smoking also increased the RA of Alloprevotella species [39], Campylobacter [36, 43, 53], Dialister [34], Eubacterium [52] and Eubacterium species [39, 52] in saliva and whole mouth samples versus subjects who did not smoke. Conversely, several studies reported a decreased RA of Acinetobacter species [39], Bifidobacterium [47], Catonella [39], Capnocytophaga [40, 45], Cardiobacterium [36, 45], Granulicatella [38, 48, 51], Lactococcus [40, 47], Leptotrichia [34, 48, 49, 51], Pseudomonas orientalis [39], Selenomonas [39, 52], and Selenomonas species [39] in saliva and whole mouth samples of cigarette smokers versus non-smokers. The RA of Actinobacillus had different associations with cigarette smoking, where the RA of Actinobacillus was increased in saliva samples [51] but decreased in whole mouth samples [34, 47] in cigarette smokers (versus non-smokers). There were also conflicting results on the influence of cigarette smoking on the RA of Fusobacterium and Fusobacterium species in the whole mouth, buccal mucosa and saliva samples. For example, several studies reported both an increased RA of genus-level Fusobacterium and Fusobacterium species (specifically F. nucleatum; [35, 46, 48, 52, 54]) and decreased RA of Fusobacterium and Fusobacterium species (specifically F. periodonticum; [34, 37, 46, 49, 51]) in cigarette smokers versus non-smokers (see Additional file 2: Table S1 for all smoking-associated oral bacteria responses in the reviewed studies).

Two studies by the same group [41, 55] explored the impact of smoking on the oral microbiome using tongue dorsum samples. These studies reported that cigarette smoking was associated with a decrease in the RA of Alloprevotella, Campylobacter, Cardiobacterium, Capnocytophaga, Fusobacterium, Eubacterium, and Lachnoanaerobaculum versus non-smoking subjects [41]. Two independent groups reported that smoking was associated with a decrease in Peptostreptococcus and Catonella in tongue dorsum samples, compared to non-smokers [41, 45]. Cigarette smoking was also associated with an increased RA of species-level Eubacterium brachy, Eubacterium nodatum, Eubacterium saphenum, Filifactor alocis, Fusobacterium nucleatum, and Mogibacterium timidum in tongue dorsum samples compared to non-smokers [41, 55].

Supragingival and subgingival plaque samples were collected by soft tissue removal [42] or by biofilm [56, 57] and direct plaque [58, 59] collection using sterilized paper points in patients with and without periodontal disease, depending on the study design (Table 1). Cigarette smoking was associated with an increased RA of Bifidobacterium [42, 58] and Campylobacter rectus [59], Eikenella corrodens [59], Fusobacterium [42], Granulicatella [42], and Selenomonas sputigena [59] in subgingival plaque samples and an increased RA of Haemophilus parainfluenzae [57] in supragingival samples versus non-smoking subjects. Conversely, the RA of Capnocytophaga ochracea [59] and Pseudomonas [42] in subgingival plaque samples were decreased in cigarette smokers versus non-smokers.

Specific bacterial taxa had similar associations with cigarette smoking in multiple oral microbiome sampling sites (i.e. a shared increase or decrease in RA in response to cigarette smoking in more than one oral microbiome site; Table 2 and Fig. 1A). For example, cigarette smoking was associated with an increased RA of Atopobium and Atopobium species [34, 36, 37, 41, 45, 55, 58], Rothia and Rothia species [36, 38, 41, 44, 46, 52, 55], Lactobacillus [50, 58], Mycoplasma and Mycoplasma hyosynoviae [34, 41, 52], Megasphaera and Megasphaera micronuciformis [34, 36, 37, 39, 41, 43, 46, 55], Tannerella and Tannerella forsythia [39, 41, 52], Corynebacterium and Corynebacterium species [41, 57], Prevotella and Prevotella species [34, 36‐40, 44, 46, 52, 55], Streptococcus and Streptococcus species [41, 44, 55, 58], Porphyromonas and Porphyromonas species [34, 46, 55], Treponema and Treponema species [34, 42, 45, 48, 52, 55, 59], and Veillonella or Veillonella species [36‐38, 43, 44, 46, 48, 50, 55, 58], compared to non-smokers. Interestingly, the RA of Lactobacillus and Rothia were also increased in subjects exposed to secondhand smoke [47]. Cigarette smoking was associated with a decreased RA of Gemella [37, 38, 40, 41, 43, 45, 47], Haemophilus and H. haemolyticus [34, 37, 41, 43, 45, 47, 55], Neisseria and Neisseria species [34, 36‐38, 41, 43, 47, 50, 52, 53, 55, 58], Bergeyella [37, 43, 47, 48], Oribacterium and Oribacterium species [39, 41, 46, 52], Lautropia and L. mirabilis [34, 36, 37, 41‐43, 47, 52, 55, 58] in most oral sites versus subjects who did not smoke. An exception to the previous inverse association between bacterial RA and cigarette or e-cigarette smoking was an increased RA of L. mirabilis in supragingival [57] samples and increased Oribacterium in the whole mouth [48] and subgingival plaque [42] samples in smokers compared to non-smokers. Additionally, in cigarette smokers (versus non-smokers), the RA of Actinobacillus was increased in saliva samples [51] but decreased in whole mouth samples [34, 47]. Not all taxa had agreement across studies when reporting the association between RA and cigarette smoking. For example, in the whole mouth and saliva samples, there were almost an equal number of studies reporting a decreased [40, 43, 49, 51] and increased [37‐39, 46, 49] RA of Streptococcus or Streptococcus species, and the association between smoking and Prevotella RA was equivocal in subgingival plaque samples, with one study reporting an increase [58] and the other a decrease [59] in RA versus non-smoking subjects. Current cigarette use was associated with a decreased RA of Bergeyella, Haemophilus, Lautropia, and Neisseria compared to former smokers in whole mouth samples [47], indicating that abstinence from smoking might reverse the smoking-associated decrease in RA of these taxa.

Evidence has demonstrated that periodontal disease creates an inflammatory oral environment in which mediators may transmit to the systemic circulation. Systemic diseases such as diabetes mellitus [77, 78] and CVD [79] have also been associated with increases in oral inflammation and periodontal pathology through disease mechanisms and treatment modalities [79], suggesting a bidirectional relationship between local (oral) and systemic inflammatory processes and health. The RA of bacterial taxa associated with periodontal disease, including Porphyromonas, Treponema, Tannerella, Campylobacter, and Prevotella spp. were positively associated with cigarette and e-cigarette smoking in many studies (Table 3). Additionally, in studies evaluating shared oral and coronary artery plaques, atherosclerotic plaques were found to contain DNA from periodontal pathogens, indicating transmission of bacteria from the oral cavity to the heart [68, 71]. The potential link between the periodontal disease pathogens and increased cardiovascular risk or CVD has been studied extensively [32, 60, 62, 80]. Evidence has demonstrated that periodontal disease creates an inflammatory oral environment in which mediators may transmit to the systemic circulation [60].

Lipopolysaccharide (LPS): LPS, an endotoxin that is commonly found in the cell wall of gram-negative bacteria, has been demonstrated to trigger the inflammatory cascade resulting in the systemic inflammation that increases cardiovascular risk and CVD [81, 82]. There is a strong connection of periodontitis to cardiometabolic disorders, mediated by LPS, that has been reviewed in detail [83]. In the reviewed studies, LPS biosynthesis pathway genes were upregulated in subgingival plaque sample of e-cigarette smokers and tongue dorsum samples of cigarette smokers [15, 55]. Additionally, many bacteria that had positive associations with LPS biosynthesis genes had increased RA levels in the oral microbiome of cigarette and e-cigarette smokers [81, 84]. For example, Alloprevotella tannerae was increased in saliva samples of e-cigarette smokers [46], Filifactor was increased in tongue dorsum samples of cigarette smokers [41, 55], Fusobacterium nucleatum was increased in saliva samples of e-cigarette and cigarette smokers [46, 52, 54]; and Prevotella was increased in saliva, whole mouth, buccal, and tongue dorsum samples of cigarette and e-cigarette smokers [34, 38‐40, 44, 46, 55, 58]. Other bacteria associated with LPS biosynthesis that were increased in smokers included Porphyromonas endodontalis in tongue dorsum and saliva samples of cigarette and e-cigarette smokers [46, 55], Veillonella in saliva, whole mouth, buccal, tongue dorsum, and subgingival plaque samples of cigarette and e-cigarette smokers [15, 38, 43, 44, 46, 48, 50, 55, 58], Treponema in saliva, whole mouth and subgingival plaque samples of cigarette smokers [34, 48, 52, 55, 59], and Parvimonas in saliva and subgingival plaque samples of cigarette and e-cigarette smokers [15, 46]. The RA of genus-level Parvimonas was also positively correlated with LPS biosynthesis genes in the oral microbiome [70]. Additionally, in one study, the LPS gene in commensal biofilm samples was upregulated 22.8-fold with nicotine-free vapor and 10.1-fold in nicotine-containing vapor in e-cigarette users versus non-smokers [15].

C-Reactive Protein, CRP: CRP is another biomarker associated with inflammation and cardiovascular risk [85]. Plasma CRP levels can reliably predict the prognosis of atherosclerosis, heart failure, and other CVD, even in otherwise asymptomatic individuals [86]. Plasma CRP levels have shown to be higher in cigarette smokers versus non-smokers in studies with [60] and without [87] inclusion of oral bacteria measures. Increased CRP transcription in the liver have been associated with increased levels of pro-inflammatory cytokines IL-6, IL-1, and TNF-α, suggesting another cardiovascular risk mechanism linked to inflammation and mediated by smoking-associated changes in the oral microbiome and elevated cytokine levels [86]. As there has been a longstanding connection between oral infection and septicemia, especially in immunocompromised patients [88, 89], oral bacterial translocation to the liver and systemic circulation through the gut is another potential mechanism that deserves future research. CRP can also indirectly activate TNF-α, reactive oxygen species, and IL-1ß in apoptotic pathways leading to a persistent systemic inflammatory state [86, 90]. Parvimonas was positively correlated with plasma CRP in subjects with asymptomatic and symptomatic atherosclerosis in previous research [70]. In the reviewed studies of smoking and the oral microbiome, both cigarette and e-cigarette smoking were associated with an increased RA of Parvimonas in subgingival plaque and saliva samples [15, 46]. Collectively, these associations provide evidence that cardiovascular risk may be mediated through cigarette and e-cigarette smoking-associated oral microbiome changes that promote a systemic inflammation driven by LPS and CRP. Future mechanistic research will elucidate the pathways underlying oral microbiome-associated LPS and CRP responses to cigarette smoking so interventional research studies can be designed to lower cardiovascular risk and CVD in these patients.

Cytokines: Oral bacteria that had significant positive associations with cigarette and e-cigarette smoking were also associated with several salivary pro-inflammatory cytokines. For example Porphyromonas correlated positively with proinflammatory cytokines such as IL-2, IL-13, IL-8, and IL-1ß; Lachnoanaerobaculum with IL-2, IL-4, IL-8, IL-13, IL-10, IL-12p70, and IFN-ɣ; Parvimonas and Mogibacterium with IL-8 and IL-1ß; and Fusobacterium with IL-1ß [46]. Streptococcus and Rothia (increased in several oral sites of cigarette smokers) and Selenomonas (increased in subgingival plaque samples of cigarette smokers) also had positive associations with pro-inflammatory cytokines IL-1ß, IL-2, IL-4, IL-6, IL-7, IL-9, IL-12 and IL-17 [91]. The RA of Johnsonella (increased RA in e-cigarette smokers) was positively associated with IL-1ß levels [46]. Conversely, Peptostreptococcus (decreased in the whole mouth and tongue dorsum samples of cigarette smokers versus non-smokers [38, 41]) was negatively associated with salivary IL-8 and IL-1ß [46]. Finally, two bacteria (Porphyromonas gingivalis and Fusobacterium nucleatum) that play pivotal roles in periodontal disease and act on macrophages, neutrophils, and monocytes to induce TNF-α, IL-6, and IL-8 production [92] were increased in smokers’ tongue dorsum [55] and saliva samples [46, 52, 54], suggesting another link between smoking, periodontal disease and inflammation-mediated cardiovascular risk. It is important to note that some of the bacteria that were positively associated with smoking and LPS biosynthesis were not periodontal pathogens but genera that are common to the healthy oral microbiome habitat (i.e. Streptococcus and Veillonella). This suggests that an imbalance of “healthy” oral-associated bacteria as a result of smoking may also contribute to increased inflammation in addition to an increased RA of periodontal pathogens. Overall, upregulation of common and unique inflammatory cytokines in salivary and gingival crevicular fluids in both e-cigarette cigarette smokers suggests that smoking alters the oral microbiome in a manner associated with increased inflammation that plausibly contributes to the increased CVD risk in this population.

Increases or decreases in local (i.e. oral) and systemic amino acid levels through modulation of amino acid metabolism and biosynthesis pathways is another possible mechanism in which smoking may influence cardiovascular risk through oral microbiome-associated bacteria. Amino acids are the foundational molecules for protein catabolism and are involved in the biosynthesis of important molecules including hormones, neurotransmitters, and coenzymes. Gut microbial metabolites have been strongly associated with amino acid and lipid metabolism [93], but relationships between similar oral bacteria and associated amino acid or lipid metabolic functions is not as well known.

Tyrosine is derived from phenylalanine and is a nonessential amino acid involved in protein synthesis and signal transduction, while tryptophan is an essential amino acid also involved in the production and maintenance of enzymes and neurotransmitters [94]. In many of the evaluated studies, smoking was associated with an increase in the RA of Rothia in saliva, whole mouth, tongue dorsum, and buccal samples [38, 41, 44, 46, 52, 55]. Associational analyses demonstrated that oral microbiome communities with higher levels of Rothia were more likely to display greater utilization of tyrosine and tryptophan [55], which could lead to lower levels of these amino acids. Salivary composition changes and increased incidence of periodontitis, both associated with smoking, contribute to an oral environment where gram-positive facultative anaerobes and amino-acid degrading bacteria like Prevotella proliferate [95]. Similarly, smoking was associated with an increased RA of Prevotella across all of the oral sampling sites (Table 2). Therefore, reductions in oral amino acid levels via smoking-associated bacterial changes may occur through known mechanisms such as changes in oral environment or from other mechanisms that will be identified as more work is performed on the functional implications of oral bacteria alterations in response to e-cigarette and cigarette smoking.

It is not known if smoking-associated changes in oral amino acid levels are directly associated with plasma levels, but connections between oral disorders and decreased plasma amino acid levels have been reported [96, 97]. This topic will be of great interest to establishing connections between the oral microbiome and systemic disease as many plasma amino acids are associated with cardiovascular risk and CVD. For example, tryptophan has been thought to be inversely associated with cardiovascular risk, as serum tryptophan levels have predicted lower risk of CVD events and death related to CVD in observational studies [98, 99]. Decreased serum tyrosine levels were also reported in patients with stroke [100] and hypertension [101], and increased tyrosine intake was associated with lower peripheral and central blood pressure [102]. Low plasma serum tryptophan concentrations are associated with increased CRP and other pro-inflammatory cytokines previously discussed, subsequently increasing risk for CVD [99, 103]. Therefore, an increased understanding of pathways linking smoking to the modulation of amino acid-associated oral microbiota and the influence these bacteria have on host-microbial protein–protein interactions and systemic cardioprotective amino acids such as tyrosine or tryptophan is an exciting avenue for future research.

Bacteria of the oral microbiome that are impacted by smoking may also modulate cardiovascular risk through associations with lipids implicated in CVD. For example, the RA of Streptococcus salivarius was increased in saliva and tongue dorsum samples of cigarette smokers versus non-smokers [39, 46, 55], and was also positively associated with oleic acid biosynthesis pathways in a separate in-vitro study of cultured S. salivarius and metabolites [104]. Circulating levels of oleic acid, a fatty acid, are not indicative of dietary intake, and may involve hepatic synthesis by the enzyme Stearoyl-CoA desaturase-1. Interestingly, circulating levels of oleic acid were independently associated with increased rates of CVD and all-cause mortality in a large Multi-Ethnic Study of Atherosclerosis [105], and oleic acid was also positively associated with diastolic blood pressure in other research [106]. The RA of Lactobacillus, which was increased in whole mouth and subgingival plaque samples of cigarette smokers [50, 58], was positively associated with plasma Apolipoprotein B levels in a separate study using linear regression analysis [70]. Apolipoprotein B is being increasingly acknowledged as a main contributor to atherosclerotic CVD [107], and large cohorts studies of have found oxidized lipids like Apolipoprotein B to be associated with coronary artery disease [108]. Finally, platelet activating factor, a potent phospholipid activator, has been reported to increase in both tobacco and e-cigarette smokers through the release of free radicals, creating oxidative stress to the intercellular environment [109, 110]. The RA of Prevotella was found to be positively associated with platelet activating factor in an in vitro experiment in mice [111], and the RA of Prevotella was also increased in saliva, whole mouth, and subgingival plaque samples of cigarette smokers versus non-smokers [34, 37, 38, 40, 46, 58], suggesting a potential link between smoking, Prevotella and platelet activating factor. Platelet activating factor has been associated with worse stroke outcomes through its thrombotic and platelet aggregation properties in addition to promoting oxidative stress and systemic inflammation [112]. While all of these relationships point to a possibility that smoking may contribute to CVD through modulation of specific oral microbiome-associated bacteria that have relationships with amino acid and lipid levels, these studies are associational in nature. Future research confirming the above relationships and testing associations with plasma amino acid and lipid levels will inform mechanisms and pathways potentially underlying the relationship between smoking and the oral microbiome and their combined impact on cardiovascular risk.

Nitric oxide is a known vasodilator that decreases cardiovascular risk through reduction of blood pressure and inhibition of oxidative stress, platelet aggregation and leukocyte adhesion [113, 114]. Nitric oxide is produced endogenously by nitric oxide synthases from the amino acid l-arginine and molecular oxygen. Since human cells lack nitrate (NO₃) reducing capability, commensal oral bacteria have been identified for their role in converting dietary NO₃ to nitrite (NO₂), which can be further reduced to nitric oxide [113, 115]. This alternate nitric oxide production mechanism is important for maintenance of cardiovascular integrity during periods of hypoxia. This NO₃–NO₂-nitric oxide entero-salivary pathway is an important pathway in which the oral cavity influences systemic physiological processes and could prove to be a potential therapeutic target. Alterations in oral bacterial communities induced by antibacterial mouthwash use, had been documented to decrease nitric oxide levels and was associated with alterations in blood pressure [116, 117]. Furthermore, studies have shown that dietary nitrate supplementation using nitrate rich foods (such as green leafy vegetables) could enhance nitric oxide production. Similar to nitric oxide, carbon monoxide is a gas that possesses vasodilatory properties, has documented interactions with the microbiome (particularly the gut microbiome), and its levels are potentially influenced by dietary consumption including fiber, leafy greens, carbohydrates and proteins [118‐120]. Periodontal disease-associated alterations in oral bacteria and consequential impacts of nitric oxide and cardiovascular risk through blood pressure modulation have also been previously reviewed in the literature [17]. Nevertheless, these relationships have not been as thoroughly explored for smoking-associated modulation of oral bacteria or if other small gas production (like carbon monoxide) is modulated by smoking and/or the oral microbiome and how these associations might impact subsequent cardiovascular risk. Multidirectional interactions between vasodilatory gas and associated signaling pathways with the microbiome and host are exciting avenues for future research that potentially holds promising therapeutic benefits for cardiovascular risk and disease.

In the reviewed studies, there were significant associations between smoking and nitrate-reducing bacteria (i.e., Rothia and Neisseria [24, 121]) and bacteria that are inversely associated with nitrate and nitric oxide (Prevotella and Veillonella [24, 121]). Interestingly, the two major nitrate-reducing bacteria in the oral microbiome had disparate RA associations with smoking. For example, there was agreement across several studies that the RA of Neisseria was decreased in saliva, whole mouth, tongue dorsum, and subgingival plaque samples [34, 36‐38, 41, 43, 47, 50, 52, 53, 55, 58], while the RA of Rothia increased in whole mouth, saliva, tongue dorsum and subgingival plaque samples of cigarette and e-cigarette smokers versus non-smokers [36, 38, 41, 44, 46, 52, 55, 58]. Whether the smoking-associated increase in Rothia RA is a compensatory mechanism in response to smoking-induced decreased nitric oxide levels resulting from oral environment changes (acidity, salivary production etc.) and reduced Neisseria levels or is a completely separate mechanism is unknown. Future research with direct nitric oxide measures in response to smoking-associated changes in oral microbiome RA of nitrate-reducing bacteria will continue to inform these mechanisms and pathways. Unlike associations between nitrate-reducing bacteria and smoking, there was equally as strong but consistent evidence that smoking was associated with an increased RA of bacteria that possess negative relationships with nitric oxide production and availability. For example, the RA of Prevotella was increased in saliva, whole mouth, tongue dorsum and buccal samples [34, 36‐40, 44, 46, 52, 58], and Veillonella increased in saliva, whole mouth, tongue dorsum, buccal, and subgingival plaque samples of e-cigarette and cigarette smokers versus non-smokers [15, 36‐38, 41, 43, 44, 46, 48, 50, 55, 58]. A previous study evaluating the RA of Prevotella and Veillonella after nitrate administration reported a decrease in both Prevotella and Veillonella after 10 days of dietary supplementation with nitrate-rich beetroot juice [122]. Nitrate supplementation also lowered blood pressure in two studies using dietary nitrate interventions [113, 123], and the one study that measured oral microbiome responses to nitrate supplementation reported greater plasma NO₂ levels were associated with higher RA of Rothia and Neisseria and lower RA of Prevotella and Veillonella [113]. Taken together, this evidence indicates smoking-mediated responses of nitric oxide-associated oral bacteria plausibly impact cardiovascular risk, and dietary nitrate supplementation may be a way to mitigate progression of CVD through modulation of oral bacteria.