The pro-algesic effect of γ-aminobutyric acid (GABA) injection into the masseter muscle of healthy men and women

1 Introduction

In the central nervous system, γ-amino-butyric acid (GABA) decreases neuronal excitability by acting on a ligand-gated chloride channel, the GABA_A receptor and a G-protein coupled receptor, the GABA_B receptor [1], [2], [3], [4]. Activation of pre-synaptic GABA_A receptors depolarizes the central endings of afferent fibers; a process that results in decreased release of neurotransmitters from terminal endings in the central nervous system [1], [5], [6], [7], [8], [9], [10]. Whether GABA depolarizes or hyperpolarizes the peripheral endings of nerve fibers is not known, however, limited evidence suggests that GABA is much less effective than the excitatory amino acid glutamate at exciting rat masseter muscle afferent fibers [11].

Activation of peripheral GABA_A receptors in the rat can attenuate nociceptive input [12], [13], [14]. Injections of glutamate made into the rat temporomandibular joint evoke reflex jaw muscle activity [15] which can be attenuated by co-injection of GABA in a concentration dependent manner [12]. The effect of GABA was inhibited by bicuculline, a GABA_A receptor antagonist. Subcutaneous administration of low concentration muscimol, a selective GABA_A receptor agonist, to the rat paw also suppressed nocifensive responses, while a dose five times higher resulted in increased nocifensive responses to formalin [14]. Together, these results in animals suggest that activation of peripheral GABA_A receptors may result in a local analgesic or hyperalgesic effects in a concentration related manner.

Intramuscular injection of glutamate (0.5 M, 0.2 mL) into the masticatory muscles of healthy human subjects produces pain of moderate intensity (4–6/10) that lasts for 10–15 min [13], [16], [17], [18], [19], [20], [21], [22], [23]. This pain is reported as being more intense by women than by men [11], [22]. Injection of glutamate at a higher concentration (1.0 M, 0.2 mL) can also produce a longer-lasting (~90 min) mechanical sensitization of the masseter muscle which is similar in both sexes [22]. In humans, glutamate-induced pain and mechanical sensitization can be attenuated by local injection of ketamine, which indicates that they are mediated, in part, through activation of peripheral N-methyl-D-aspartate (NMDA) receptors [17], [19], [20], [24]. Masseter muscle biopsies from healthy subjects have identified NMDA receptor expression in a subgroup of sensory nerve fibers [25]. As a result of these properties of intramuscular glutamate injection, it has been used to model acute masseter muscle pain and sensitivity reported by patients suffering from a myofascial temporomandibular disorder [19].

The purpose of the present study was to see if findings in rats of a GABA_A mediated antinociceptive effect could be translated into healthy human subjects. Human subjects were given injections of GABA at concentrations that had been shown to reduce nociceptive input from the rat temporomandibular joint [12], [13]. However, since injection of these substances into the healthy human temporomandibular joint was not feasible for ethical reasons, the study was instead conducted using intramuscular injections into the masseter muscle. It was hypothesized that injections of GABA alone into the muscle would produce no more pain than injection of saline, and that injection of GABA with glutamate would attenuate glutamate-evoked muscle pain in a concentration-related manner that was enhanced by the GABA_A receptor positive allosteric modulator, lorazepam.

2 Methods

The experimental protocol was approved by the North Denmark Ethics Committee (reference no. N-20160037) and carried out according to the Helsinki Declaration and the IASP guidelines. Written informed consent was obtained from all participants.

Thirty healthy volunteers (15 men and 15 women) were recruited for study one, and 30 additional healthy volunteers (15 men, 15 women) for study two. Two subjects, both women, withdrew from study two. Baseline demographics for subjects in both studies are provided (Table 1). Both studies were randomized, placebo-controlled, double-blinded, and had a crossover design. Subjects were eligible to participate in the study if they were between 20 and 40 years of age and free from ongoing or chronic pain. Subjects who were pregnant, or intended to become pregnant, were breast feeding, had signs or symptoms of any serious systemic diseases including malignancies or high blood pressure, required chronic administration of psychiatric, analgesic or other medications that might influence their response to pain, reported any recreational drug or alcohol use, reported a previous neurologic, musculoskeletal or mental illnesses or lacked the ability to cooperate, were excluded from participation in the study.

2.1 Study one

This project was designed to test whether intramuscular injection of GABA alone is painful and/or alters responses to mechanically-induced muscle pain. Each subject attended two sessions with a minimum interval of 1 week between sessions (Fig. 1A). Treatments were assigned randomly, and neither the subject nor the tester was aware of the content of the injections.

Fig. 1:

(A) The drawing shows the structure of study I. (B) The drawing shows the structure of study II.

Two injections were given in each session with an interval of at least 1 h between injections. Injections were made into the right, then subsequently into the left masseter muscle in each session. Subjects were instructed to continuously rate their pain after the injections. Bilateral masseter muscle pressure pain threshold (PPT) and pressure pain tolerance (PPTOL) as well as maximal jaw-opening (MJO) were assessed at baseline and periodically after injections (Fig. 1A). After each injection, subjects were also asked to draw their perceived distribution of pain on a picture of the profile of the face.

3 Results

3.1 Study 1

Injection of GABA was more painful than injection of isotonic saline, but overall, pain evoked by GABA was mild (<3/10). Injection of GABA 0.5 M was rated more painful than injection of either isotonic saline or GABA 0.05 M (Fig. 2A). This was reflected in significantly higher peak and AUC values for GABA 0.5 M with or without lorazepam (Fig. 2B). The addition of lorazepam with GABA 0.5 M further increased pain ratings compared with GABA 0.5 M. Peak pain ratings for the combination of GABA and lorazepam were significantly higher than GABA 0.5 M alone (Fig. 2B). There were no statistically significant differences between male and female subjects in these parameters.

Fig. 2:

(A) The line graphs illustrate the mean masseter muscle pain intensity produced by injection of GABA with or without lorazepam, compared to saline in the 30 subjects. (B) The bar graphs indicate the mean area under the pain curve (AUC), peak and duration of pain produced by injection of the substances indicated. There was a significant concentration-related increase in GABA-evoked overall (F=6.223, p<0.001) and peak (F=19.237, p<0.001) muscle pain, and a non-significant increase in the duration of pain. The addition of lorazepam to GABA 0.5 M injections resulted in a significantly higher peak pain rating than GABA 0.5 M alone, which suggests that pain was being mediated through activation of GABA_A receptors. Asterisks: p<0.05 compared to saline control; Error bars: SE.

There were no significant effects of any of the injected substances on PPT, PPTOL or MJO over the time course of the experiment (Fig. 3). There were no significant sex-related differences in the baseline values of these three parameters (Table 1).

Fig. 3:

(A) The line and scatter plot shows the mean relative pressure pain threshold (PPT) normalized to baseline (−5 min). There was no significant effect of any of the injections on PPT. (B) The bar graphs indicate the mean relative pressure pain tolerance (PPTOL) 5 and 45 min after masseter muscle injections. There was no significant effect of any of the injections on the PPTOL. (C) The line and scatter plot shows the mean maximal jaw opening (MJO). There was a significant decrease in MJO over time (F=3.426, p=0.011), but no significant effect of treatment or treatment time interaction. Error bars: SE.

Significant treatment effects similar to those found for average pain were found for pain area (Fig. 4). Injections of GABA 0.5 M with and without lorazepam resulted in subject drawings of significantly greater pain area than injection of isotonic saline or GABA 0.05 M. Female subjects drew significantly larger areas of pain after injection of GABA 0.5 M than did male subjects.

Fig. 4:

(A) The image shows composite pain area drawings from the 30 subjects. Pain was generally localized to the site of injection with all treatments. (B) The bar graphs show the mean pain area. There was a significant effect of treatment (F=11.489, p<0.001). The drawn pain areas for GABA 0.5 M with or without lorazepam were significantly larger than those drawn for saline. There was no significant difference between areas drawn for GABA 0.05 M and saline or between GABA 0.5 M with lorazepam and without lorazepam. Asterisks: p<0.05 compared to saline control; Error bars: SE.

3.2 Study 2

Injection of glutamate 0.5 M alone as an internal control produced moderate pain, with average ratings of just under 4. Overall and peak pain produced by the initial glutamate injection was reported as significantly greater by women than by men (Table 2).

Table 2:

Glutamate-evoked pain.

VAS	Women	Men
AUC	1,722±463	808±178
Peak	4.8±0.5	3.6±0.5
Duration (s)	571±120	395±75

1. The mean (± SE) area under the pain curve (AUC), peak and duration of muscle pain evoked by the first injection of glutamate in men and women over four sessions are shown. Women reported significantly greater overall pain (AUC) and higher intensity of pain (peak) than did men. Bold text: Students t-test, p<0.05.

Repeated injection of glutamate evoked pain of similar intensity and duration (Fig. 5A). The addition of GABA 0.5 M to the glutamate in the second injection significantly increased overall, peak and duration of pain reported (Fig. 5B). However, the addition of lorazepam to the glutamate/GABA 0.5 M injection did not further increase pain.

Fig. 5:

(A) The line graphs illustrate the mean masseter muscle pain intensity produced by injection of glutamate (time 0) followed 30 min later by glutamate with GABA (0.05 or 0.5 M) or GABA 0.5 M with lorazepam in 28 healthy subjects. Repeat injection of glutamate alone (control) evoked relatively reproducible pain responses. The addition of GABA 0.5 M, with or without lorazepam, in the second injection increased the pain intensity compared to injection of glutamate alone. (B) The bar graphs indicate the mean relative area under the pain curve (AUC), relative peak and relative duration of pain produced by injection of the substances indicated. There was a significant effect of treatment on overall pain (AUC) (F=4.094, p=0.009), peak pain (F=3.021, p=0.035), and duration of pain (F=3.444, p=0.021). The addition of GABA 0.5 M significantly increased pain compared to control injections. The addition of lorazepam lowered pain ratings compared to GABA 0.5 M without lorazepam. GABA 0.05 M had no effect on glutamate-evoked muscle pain. Asterisks: p<0.05 compared to saline control; Error bars: SE.

Analysis of the pain parameters also revealed a significant interaction between sex and treatment. In men, the addition of GABA 0.5 M with glutamate in the second injection evoked significantly greater overall and peak pain than glutamate alone (Fig. 6). However, in women, the addition of GABA 0.5 M with or without lorazepam in the second injections did not significantly alter pain ratings compared to glutamate alone. When men and women were compared, men reported a significantly greater enhancement of their pain by GABA 0.5 M added to glutamate than women (Fig. 6B).

Fig. 6:

(A) The line graphs illustrate the mean masseter muscle pain intensity produced by injection of glutamate (time 0) followed 30 min later by glutamate with GABA 0.5 M in men (n=15) and women (n=13). Note that the initial injection of glutamate alone produced substantially less pain in men than in women (see Table 2). Further, the graphs illustrate that in men, there was a substantial increase in pain ratings when glutamate and GABA were injected together, whereas in women, the pain responses appear almost identical. (B) The bar graphs indicate the mean relative area under the pain curve (AUC), relative peak and relative duration of pain produced by injection of the substances indicated in men (black) and women (white). There was a significant interaction between sex and treatment for AUC (F=5.005, p=0.003). In men, the combination of GABA 0.5 M with glutamate significantly increased AUC, peak pain, and duration of pain compared to glutamate alone. In women, none of these injections had significantly altered pain ratings compared with glutamate alone. Asterisks: p<0.05 men compared with women; Error bars: SE.

There were no significant treatment effects on PPT, PPTOL or MJO over the time course of the experiment (Fig. 7). The only significant sex-related difference found was in the baseline PPT value, which was significantly higher in men than in women. No other sex-related differences in these parameters were identified.

Fig. 7:

(A) The line and scatter plot shows the mean relative pressure pain threshold (PPT) normalized to the pre-injection baseline. There was no significant effect of any of the second injections on PPT. (B) The bar graphs indicate the mean relative pressure pain tolerance (PPTOL) 45 min after the second masseter muscle injections. There was no significant effect of any of the injections on the PPTOL. (C) The line and scatter plot shows the mean maximum jaw opening (MJO). There was no significant effect of any of the second injections on MJO. Error bars: SE.

There were no significant treatment effects on pain area. There was also no sex-related difference in average pain area for the initial glutamate injection when areas from men and women were compared.

4 Discussion

Preclinical studies have reported that activation of peripheral GABA_A receptors can exert either analgesic or algesic effects, depending on the concentration used [12], [13], [14]. The present study found that injection of GABA alone into the masseter muscle of healthy humans resulted in concentration-related reports of mild pain that was further increased by lorazepam, a GABA_A receptor positive allosteric modulator. These results suggest that intramuscular injection of GABA excites muscle nociceptors through activation of GABA_A receptors. Subsequent experiments examined the effect of GABA administration on pain evoked by intramuscular injection of glutamate, which provokes pain responses through activation of peripheral NMDA receptors [17]. These experiments indicated that GABA can, in a concentration-related manner, increase glutamate-evoked masseter muscle pain. This pro-algesic effect of GABA was significantly greater in men than in women. The effect of GABA injections in either sex are apparently short lasting, as they did not significantly affect mechanical nociception within 5 min of an intramuscular injection. Taken together, these results suggest that GABA can act through the peripheral GABA_A receptor in humans to provoke and enhance masseter muscle pain.

The peripheral effect of elevated GABA concentrations on nociception has previously been investigated in the skin, joint and oral cavity of rats. Injection of GABA into the temporomandibular joint of male rats resulted in a concentration related declination in the magnitude of the glutamate-evoked TMJ-jaw muscle reflex in male rats [12], [13]. This effect could be reversed by bicuculline, a GABA_A receptor antagonist, but not phaclofen, a GABA_B receptor antagonist, which indicates it was mediated through activation of GABA_A receptors. However, subcutaneous injection of the GABA_A receptor agonist muscimol was shown to differentially modulate responses in the formalin model of cutaneous inflammatory pain which depended on concentration [14]. At low concentration, muscimol exerted analgesic effects, whereas at increased concentration it enhanced formalin-evoked nocifensive behavior. More recently, oral administration of muscimol to the tongue was shown to significantly increase the mechanical thresholds of tongue afferent fibers compared to vehicle, but only after the tongue had been heated with 60 °C water [26]. It was also shown in this study that 95% of afferent fibers innervating the rat tongue mucosa express GABA_A receptors [26]. GABA containing oral rinses have been recently employed in humans to test their effect on burning pain induced by topical application of capsaicin [27]. In these experiments, capsaicin was applied to the tongue to provoke burning pain, and the effect GABA mouthwashes examined. GABA rinses did not alter the peak pain, but did shorten the time for healthy human subjects to stop feeling the burning pain [27]. Taken together, these previous results demonstrate that peripheral GABA_A receptor activation can exert very different effects that are dependent on the tissue to which it is applied and the concentration used.

In the spinal cord and trigeminal sensory nucleus, GABA acts via the GABA_A receptor to mediate primary afferent depolarization; a presynaptic inhibitory mechanism [5], [6], [7], [8], [9]. This is thought to occur because the chloride reversal potential in central endings of primary afferent fibers is more depolarized than the resting membrane potential. Indeed, in vivo experiments in the rat have shown that sustained application of GABA to the dorsal root ganglion neurons results in biphasic depolarization as long as GABA is present [28], [29], [30]. However, the environment that central terminals and sensory ganglion neurons are in may be quite different from that of their peripheral endings in various tissues, and thus it can only be speculated that the resting membrane potential in these endings is similar to that measured in sensory ganglion neurons. The finding that intramuscular injection of GABA was painful, and that GABA could enhance glutamate-evoked muscle pain is suggestive that GABA can depolarize peripheral afferent endings, at least in masticatory muscle, and that this mechanism contributes to its algesic effects. Why then, in other tissues at different concentrations, does GABA administration result in analgesia? The analgesic effect of GABA_A receptor activation in preclinical studies has been speculated to be due to a current shunt or depolarization block [12], [14], [26]. It has been shown that GABA produces greater depolarizations of Aβ and Aδ afferent fibers, than of C fibers [6]. There is a good temporal relationship between the firing of a population of Aδ afferent fibers in rats and the change in pain in human subjects, after injection of glutamate into the masseter muscle [31]. This may mean that much of the acute pain response evoked by glutamate when injected into the human masseter muscle is coded by the firing of Aδ afferent fibers and thus that GABA, through its stronger depolarizing action on myelinated fibers acts to increase pain intensity reports. A recent report also indicated that GABA has the ability to activate neuromuscular nicotinic receptors at high concentration, which could lead to low level muscle contraction [32]. In the present study, subjects commonly reported that the muscle felt tight after GABA and glutamate were co-injected. If GABA induces low levels of muscle contraction, this may also contribute to increased pain reports [33]. Indeed, this difference might also help explain why GABA increased glutamate-evoked pain when injected into the human masseter muscle, but decreased glutamate-evoked nocifensive response when injected into the rat temporomandibular joint.

The effects of GABA can also be mediated through activation of G-protein linked GABA_B receptors, which are expressed by around one third of trigeminal ganglion neurons that innervate the rat masseter muscle (Cairns, unpublished results). Baclofen, a GABA_B receptor selective agonist has been shown to exert suppressive effects on the responses of vagal and pelvic afferent fibers to mechanical stimulation in ferrets and rats, respectively [34], [35]. Baclofen has also been shown to increase a transient and a sustained potassium current in trigeminal ganglion neurons, which results in membrane hyperpolarization and decreased excitability [36]. However, GABA injection into the human masseter muscle did not alter PPT or PPTOL, either alone or in combination with glutamate. This suggests that if sensory afferent fibers that innervate the human masseter muscle express GABA_B receptors, their activation by injection of GABA does not result in a detectable change in mechanical sensitivity.

Both healthy men and women were recruited into the present study to test whether there are sex-related differences in the effect of intramuscularly injected GABA. While no sex-related differences were found for pain AUC, duration or intensity after injection of GABA alone into the masseter muscle, it was found that GABA significantly enhanced glutamate-evoked masseter muscle pain in men but not in women. It is a consistent finding that injection of glutamate 0.5 M into the masseter muscle evokes significantly more pain in women than in men (Table 2) [19], [22], [37]. However, the combination of GABA 0.5 M with glutamate evoked muscle pain of relatively similar intensity in men and women (Fig. 5). If one simply adds the intensity of pain produced by GABA alone (Fig. 1B) to the pain produced by glutamate alone (Table 1) for men, it is a pretty good estimate of the pain reported when GABA 0.5 M and glutamate were injected together in men. A similar estimation for women, suggests that peak pain intensity produced by the combination GABA 0.5 M and glutamate should have been around 6. It is not clear why pain produced by injection of glutamate and GABA together were not also additive in women. It is possible this sex-related difference in the enhancement of glutamate-evoked pain by GABA is merely a methodological artifact.

In conclusion, the present study was unable to translate findings of analgesic actions of activation of peripheral GABA_A receptors from animal models into healthy human subjects. Indeed, the findings support the currently accepted concept that activation of GABA_A receptors on sensory primary afferent fibers leads to depolarization and increased excitability. However, it is important to recognize that the actions of peripherally administered GABA, with regard to algesic versus analgesic effect, appear to differ depending on the tissue tested (oral mucosa versus skeletal muscle). The mechanistic basis for these differences in the effect of GABA remain to be determined.