Preliminary evidence for conserved transcriptional response to adversity in adults with temporomandibular disorder.

Temporomandibular disorder (TMD) is one of the most common orofacial pain conditions. Alteration in immune functioning is one promising biological mechanism underlying pain in TMD. However, there is a gap in the understanding of molecular bases contributing to altered immune functioning in these patients.
OBJECTIVES: In the current study, we investigated whether individuals with TMD would exhibit differential activity of 3 specific transcription factors involved in inflammatory (nuclear factor-kappa B, NF-kB), antiviral (interferon-regulatory factors, IRF), and sympathetic (cAMP response element-binding protein, CREB) processes using a promoter-based bioinformatics analysis, which is characterized as the "Conserved Transcriptional Response to Adversity."
METHODS: Adults with TMD (n = 19) and without (n = 17) underwent a standardized clinical examination for TMD. A blood sample was collected for genome-wide transcriptional RNA profiling. Bioinformatic analyses tested for differential prevalence of proinflammatory and antiviral transcription factor activity in core promoter sequences from all genes showing >1.2-fold differential expression in TMD vs controls.
RESULTS: Promoter-based bioinformatic analyses of genome-wide transcriptome profiles confirmed upregulation of genes bearing response elements for proinflammatory transcription factor (NF-kB, P = 0.002) and downregulation of genes with response elements for IRF (P = 0.037) in patients with TMD relative to controls. Results also indicated upregulated activity of CREB in patients with TMD (P = 0.08), consistent with increased activity of the sympathetic nervous system.
CONCLUSION: These results provide initial support that the regulation of immune pathways is altered in individuals with TMD. A shift of transcriptional resources to a proinflammatory state may be driven by psychosocial stress and contributes to symptoms associated with TMD.

1. Introduction

Chronic temporomandibular disorder (TMD), characterized by muscle and/or joint pain of the masticatory system lasting greater than 3 months, affects up to 5% of the population.[37,38] Although a number of psychobehavioral (somatic awareness, stress, and sleep[13,16,24,32,33,36,39]) stressors contribute to TMD, alterations in immune system has been proposed as a key biological mechanism in chronic pain and TMD.[7,29,31] However, there is a current gap in the understanding of molecular basis contributing to alterations in immune system in individuals with chronic pain. The “Conserved Transcriptional Response to Adversity” (CTRA) profile, characterized by an upregulation of genes bearing response elements for the proinflammatory transcription factor nuclear factor-kappa B (NF-kB) and a downregulation of innate antiviral response genes bearing response elements for interferon-regulatory factors (IRF), is one molecular mechanism linking psychosocial and behavioral stress to negative health outcomes through alterations in the immune system.[9,10]

Using a promoter-based bioinformatics analyses of RNA from peripheral blood leukocytes, the CTRA profile can be assessed by quantifying the differential representation of “response elements,” or a sequence of DNA within a promoter permitting binding of transcription factors, in a set of differentially expressed genes empirically associated with an outcome of interest (TMD). Studies have repeatedly found a shift to a more proinflammatory and less Type I interferon-regulated bias within the transcriptome of circulating leukocytes, which is driven by extended periods and high levels of psychosocial, behavioral, and/or environmental stress, and consequent increases in activity of the sympathetic nervous system (cAMP response element-binding protein, CREB). This proinflammatory/anti-interferon transcriptome bias, in turn, can contribute to negative health outcomes related to stress.[4,8,21,23,30]

For the current study, we investigate whether the activity of 3 specific CTRA-related transcription factors, as indicated by expression of genes bearing response elements for NF-kB, IRF, and CREB, differed as a function of TMD status. This is important for 3 reasons. First, little is known about the relationship of CTRA to chronic pain states because only one study to date has tested whether the CTRA profile is present in patients with irritable bowel syndrome.[19] Second, pain conditions such as TMD are highly sensitive to psychosocial stress, and it is possible that the relationship between stress and pain involves changes in proinflammatory gene regulation. Third, the CTRA profile may offer a unique window into the functioning of the cellular components of the immune system, which may not be readily observed with peripheral measurements of cytokines. The current study tested the hypothesis that individuals with TMD would exhibit elevated CTRA expression, which will be characterized by a differential expression of inflammation (upregulated) and antiviral (downregulated) genes bearing response elements for transcription factors, compared with controls. We also explored other aspect of the CTRA profile including differential expression of genes underlying the sympathetic nervous system (SNS)-responsive CREB signaling pathway because altered β-adrenergic signaling has been implicated in TMD.

2. Methods

Detailed information about participants and methods can be found in Supplementary Information (available at http://links.lww.com/PR9/A89).

2.1. Participants

After informed consent and health history review, all participants underwent a standardized clinical examination (DC-TMD) by a calibrated investigator for TMD.[5,34] In participants with TMD (n = 19) and pain-free healthy controls (HCs, n = 17), a blood sample was collected for gene expression profiling (Fig. 1A). Information about sex, age, and history of smoking and alcohol consumption were collected and used as covariates (Table 1).

Figure 1.

Collection, processing, and bioinformatic analysis of transcriptome outcomes between adults with and without DC-TMD confirmed diagnosis of TMD. (A) Blood samples were collected with Tempus Blood RNA Tube. (B) When all samples were collected, RNA was extracted and isolated with Tempus Spin RNA Isolation Kit. (C) Genome-wide mRNA profiling using a high-efficiency 3′ mRNA-targeted transcript counting assay (Lexogen QuantSeq 3′ FWD) with multiplex cDNA sequencing on an Illumina HiSeq 4000 instrument in the UCLA Neuroscience Genomics Core Laboratory. (D) Gene expression was quantified and screened to exclude transcripts showing minimal variation in expression (SD < 0.5 log2 unit). All genes showing 1.2-fold differential expression in TMD vs Controls (adjusted by covariates) served as input into Transcription Element Listening System (TELiS) promoter-based bioinformatic analyses. (E) The components of the CTRA profile were explored based on empirical differences in genes attributed to differential activity of inflammatory-, interferon-, and sympathetic-related transcription factors using TELiS. TMD, temporomandibular disorder.

Table 1

Participant characteristics.

Mean (SD) or %	TMD* (n = 19)	HCs (n = 17)	Cohen's d
Study covariates†
Age (Years)	31.2 (6.4)	27.4 (6.1)	0.62
BMI (kg/m2)	25.8 (7.7)	25.2 (5.1)	0.09
Sex (% female)	78.9%	88.2%	—
Race (% NHW)‡	89.5%	58.8%	—
Alcohol consumption (%yes)	42.1%	17.6%	—

All participants with TMD meet criteria for mixed facial pain diagnosis (myofascial and arthralgia).

Although the study did not restrict enrollment into the study for smoking, none of the enrolled participants report a history of smoking.

Significant group differences: P < 0.05.

TMD, temporomandibular disorder.

2.2. Gene expression profiling and analysis

After RNA extraction (Fig. 1B), genome-wide transcriptional profiling of RNA samples was performed (Fig. 1C). For the gene expression analysis, we first excluded genes with minimal level (or variation) in expression and then identified a set of genes with a >1.2-fold differential expression in TMD vs controls (Supplemental Tables, available at http://links.lww.com/PR9/A89). Due to the limited size of the sample and the hypothesis-testing nature of this study, we did not perform any exploratory/discovery analysis or statistical testing of individual gene transcripts, but rather this gene set served as input to the Transcription Element Listening System (TELiS) promoter-based bioinformatic analyses[12] to explore associated transcription factor activity (Fig. 1D) and the CTRA profile (Fig. 1E). As in previous research, point estimates of differential expression served as input because previous research has found such lists to generate more reliable downstream bioinformatics results than gene lists derived from statistical p-/q-values.[17,35,47]

3. Results

3.1. Covariates

Sex (χ2 = 0.56, P = 0.46), history of alcohol use (χ2 = 2.5, P = 0.11), BMI (t = 0.3, P = 0.79), and age (t = 1.9, P = 0.07) were similar between the 2 groups. Differences were observed with race (χ2 = 4.5, P = 0.03) with a higher percentage of Whites in the TMD group. None of the participants reported a history of smoking (current or past), so this variable was not included as a covariate in the transcriptome analysis.

3.2. Transcription factor activity

Genome-wide transcriptional profiling by RNA sequencing identified 1173 gene transcripts showing >1.2-fold differential expression in TMD vs HCs (80 upregulated and 1093 downregulated). TELiS bioinformatics analysis of transcription control pathways (Fig. 2) indicated significant upregulation of genes with response elements for the proinflammatory transcription factor NF-kB (2.138 log2 TFBM ratio ± 0.694 standard error, P = 0.002) and significant downregulation of genes with IRF response elements (−1.135 ± 0.541, P = 0.037) in adults with TMD compared with HCs. Results also indicated a nonsignificant trend toward upregulation of the SNS-responsive CREB signaling pathway in patients with TMD (1.092 ± 0.624, P = 0.08).

Figure 2.

Gene expression profiling of components of the CTRA Profile in adults with and without TMD. Data are represented as log2-transformed ratios of transcription factor binding motifs (TFMB) for proinflammatory (nuclear factor-kappa B, NF-kB), antiviral (interferon-regulatory factor, IRF), and sympathetic (cAMP response element-binding protein, CREB) transcription factors of differentially expression genes, which showed a >1.2-fold difference in average transcript abundance between adults with TMD and controls. Because the study was interested in differences in molecular mechanisms (eg, transcription factors), we did not explore if individual genes were different between TMD and HCs. IRF, interferon response factor; TMD, temporomandibular disorder.

4. Conclusions

In this small pilot study, we found preliminary evidence that patients with TMD show a CTRA-characteristic shift in gene regulation within the basal leukocyte transcriptome including greater proinflammatory activity, reduced antiviral activity, and greater adrenergic signaling. These 3 transcription control pathways were targeted a priori based on previous hypotheses linking pain to stress biology and inflammation and were quantified by asymmetrical distribution of binding sites for the targeted transcription factors in upregulated vs downregulated genes.[4,8,21,23,30] Consistent with the classic CTRA profile, patients with TMD showed upregulation of genes bearing response elements for the proinflammatory transcription factor NF-kB, suggesting elevated levels of NF-kB activation and a propensity for proinflammatory responses. In addition, patients with TMD showed downregulation of genes bearing response elements for interferon regulatory factors (IRF), the primary inverse component of the CTRA profile.[9,10] These results were observed after controlling for demographic and behavioral factors that might otherwise confound relationships between TMD and gene expression.

Evidence for proinflammatory shift in our current sample is supportive of a growing literature about the presence of low-grade inflammation (eg, basal inflammation and exaggerated inflammatory activity)[1-3,40,41] in chronic pain. Although some of this literature is conflicting,[1] low-grade inflammation can be driven by toll-like receptor-4 (TLR4) signaling,[6,18] which induces the transcription factor NF-kB. Our preliminary data support the presence of low-grade inflammation, as reflected by NF-kB, in TMD. Also, limited evidence suggests that interferon signaling pathways can distinguish individuals with pain,[27] and infections have been shown to be predictive of widespread pain.[22] Finally, these CTRA-characteristic shifts may be related to altered β-adrenergic signaling due to chronic activation of the SNS by psychobehavioral stressors[11,14,15,20,39,48]. Recent studies have highlighted favorable impacts of β-adrenergic antagonist medication in TMD.[42,43] Our preliminary data found directional evidence for CREB activation in TMD, which would be consistent with increased SNS activity in TMD.

Although findings provide initial support for the hypothesis of altered inflammatory and antiviral pathway activity in individuals with TMD, the study is preliminary and has several limitations. First, our current study was not large enough to support any hypothesis-free genome-wide discovery analysis. Consequently, any attempts to interpret the role of specific, differentially expressed genes will require a larger sample size. Second, our gene expression was measured in peripheral blood samples. Although the observed transcriptional shift occurred in total blood leukocytes, future studies should collect information about the leukocyte subset composition of peripheral blood (eg, by flow cytometry) and clarify whether other tissues associated with TMD also show similar transcriptional shifts.

Overall, a shift of transcriptional resources of immune cells in individuals with TMD to a proinflammatory state may contribute to somatic symptoms (pain and fatigue), although it is also possible that chronic pain causes SNS activation (through stressful experience). In addition to limiting healing of damaged tissue of the temporomandibular joint and muscle,[26] this shift to a proinflammatory state of peripheral immune cells in individuals with TMD may result in a direct[45,48] and/or indirect (eg, synovial fluid of the temporomandibular joint[25]) enhancement of neuronal sensory processing. Preclinical evidence suggests that elevated levels of peripheral inflammation lead to activation of glial cells in peripheral (eg, trigeminal ganglia) and central nervous systems,[46] which can increase release of inflammatory mediators and amplify neuronal excitability, which in turn leads to an increased hypersensitivity to painful stimuli within these neuronal structures.[28,44] Future research blocking the pathways identified here and/or manipulating pain or distress may help clarify the causal mechanisms involved. Despite these limitations, our results identify new candidate molecular pathways that can be targeted in future research examining pathogenesis of TMD pain.

Disclosures

G. Schulert has received consulting fees from Novartis and SOBI. The remaining authors have no conflicts of interest to declare.