Aβ low threshold mechanoreceptors contribute to sensory abnormalities in fibromyalgia

Abstract

Fibromyalgia syndrome (FM) is characterized by widespread pain and fatigue. People living with FM also experience tactile allodynia, cold-evoked pain, paraesthesia and dysaesthesia. There is evidence of small fibre neuropathy and hyperexcitability of nociceptors in FM; however, the presence of other sensory abnormalities suggests involvement of large diameter sensory fibres. The passive transfer of FM IgG to mice causes cold and mechanical hyperalgesia associated with changes in A- and C-nociceptor function. However, whether FM IgG also confers sensitivity to light touch and whether large diameter sensory fibres contribute to symptoms evoked by cold is unknown. Here we demonstrate that the presence of sensory abnormalities such as tingling, correlate with the impact of FM, and that people with FM describe the sensation of cutaneous cooling with neuropathic descriptors such as tingling/pins and needles. We find a causal link between circulating FM IgG and the sensitization of large diameter, Aβ low threshold mechanoreceptors (Aβ-LTMRs) to mechanical and cold stimuli in mice ex vivo and in vivo. In keeping with our experimental observations, a larger proportion of Aβ-LTMRs respond to cold stimulation in people with FM, but in contrast to our results ex vivo, the same fibres display reduced responses to mechanical stimuli. These results expand the pathophysiological role of IgG in FM and will inform future studies of sensory symptoms and pain in people with FM.

Keywords: Aβ low threshold mechanoreceptor; fibromyalgia; passive-transfer; sensory systems.

Figure 1

The presence of self-reported specific neuropathic symptoms (SF-MPQ-2) and ambient temperature preference in a cohort of people with fibromyalgia. (A) Three patients (P1–P3) with severe fibromyalgia (FM) who underwent therapeutic plasma exchange (TPE) described decreased brief pain inventory–impact (BPI-I) scores and (B) increased EQ-5D Visual Analogue Scale values post treatment. Data shown at baseline (−1 month prior) and 1 month post TPE. (C) Example of SF-MPQ-2 question dark pink indicates high scores (≥8), which are indicated as percentages in B. (D) Percentage of total cohort with high scores for particular questions in the neuropathic subset of SF-MPQ-2. These scores all have significant moderate (r > 0.4) positive correlation with overall scores from the patient’s Fibromyalgia Impact Questionnaire (FIQR). (E) In the same cohort, FM patients were asked if ambient air temperature affects their pain (Answer no/yes, female: 12/57 = 69 and male: 1/9 = 10), if yes, they were asked to indicate their preferred ambient temperature. People with FM that affirmed that ambient temperature affected their pain have a preference for warmer ambient temperatures (≥21°C) compared to cooler temperatures (<21°C). Male: 7/9 (77%) and female 46/57 (81%) prefer ≥21°C. (F) FM patients show preference for both cool and warm ambient temperatures—data binned per 2°C.

Figure 2

Administration of fibromyalgia IgG to mice causes hypersensitivity to noxious mechanical and cold stimuli, along with increased responses to innocuous mechanical stimuli. (A) Graphical representation of experimental design. Syringe indicates fibromyalgia (FM)-patient IgG administration [intraperitoneal (i.p.) 8 mg]. Mouse indicates behavioural testing. Signal indicates electrophysiology experiment performed 4–7 days after IgG administration, which coincides with the peak of behavioural response. (B) Mice administered either P1–P3 (yellow, green and blue, respectively) individually or P1/P2 combined (turquoise) IgG (i.p. 8 mg) have reduced paw withdrawal thresholds [healthy control (HC): pre 105 ± 1; post 105 ± 1 g versus FM: pre 110 ± 1; post 85 ± 2 g, P < 0.05, two-sided paired and unpaired t-test, P < 0.001 n = 25–31]. P1 and P2 injected mice were injected for subsequent use in ex vivo experiments. (C) Mice administered with either P1–P3 (as for paw pressure) individually or P1/P2 combined had decreased withdrawal latency to the 10°C cold plate (FM: pre IgG 13.7 ± 0.3 versus post IgG 9.9 ± 0.3 s, two-sided paired t-test, P < 0.0001) compared to healthy control IgG (HC: pre IgG 13.3 ± 0.5 versus post IgG 12.7 ± 0.5 s). (D) Mice injected with FM IgG (pooled P1 and P2—teal and P3—blue) had an increased response rate to a low force (0.07 g) Von Frey hair compared to before treatment (9% before and 38% after, P < 0.05, two-sided paired t-test, P < 0 .05, n = 15) and between HC- and FM-treated mice (HC post 13%, two-sided unpaired t-test, P < 0.05, n = 13). (E) FM IgG administration also significantly increased the response rate to puffed cotton bud on the plantar surface (HC pre 45% and post 47%; FM 60% and post 92%, P = 13–15, P < 0.05). (F) Total score calculated based on responses to four low force stimuli tested 0.07 g, 0.4 g Von Frey hair, puffed cotton bud and fan brush on the glabrous hind-paw. Binary scoring system, 0 no response and 1 response (including rapid lift away from stimulus, flick or shake). The four innocuous stimuli were applied three times to each hind-paw (total, six tests; maximum score possible, 24/24). Mice injected with FM IgG scored higher (post IgG score 13 ± 1and 14 ± 1) than mice injected with HC IgG (post IgG score 7 ± 1).

Figure 3

Fast conducting Aβ-LTMRs ex vivo are mechanically sensitized after fibromyalgia IgG administration. (A) Aβ-slowly adapting (AβSA) conduction velocity (line denotes 10 ms⁻¹; HC: 15.1 ± 0.8 ms⁻¹ FM: 15.3 ± 0.5 ms⁻¹) and (B) mechanical thresholds (HC 0.2 ± 0.03 g; FM 0.21 ± 0.03 g) remained the same between the treatment groups (total HC: n = 23, FM: n = 28). (C) AβSA from FM IgG-treated animals were sensitized to the lowest mechanical force (1 g) applied firing more action potentials (APs) (HC: 37 ± 6 APs FM: 88 ± 11 APs, HC: P < 0.001, two-way ANOVA with Sidak’s multiple comparison, HC: n = 18, FM: n = 28) at a (D) higher peak firing frequency (HC: 474 ± 42 Hz versus FM: 649 ± 42 Hz, P < 0.01, unpaired t-test, n = 18, FM: n = 28) than fibres from healthy control treated animals. (E) Adaption properties were unchanged between treatment groups with FM treating firing more action potentials throughout (P < 0.05 between 2–6 s, two-way ANOVA with Sidak’s multiple comparison). (F) Representative traces of AβSA fibres at 1 g force step. (G) Aβ-rapidly adapting (AβRA) conduction velocities were not significantly different between treatment groups (HC: 12.7 ± 0.7 ms⁻¹ versus FM: 13.2 ± 0.6 ms⁻¹; HC: n = 14 and FM: n = 18). (H) Mechanical thresholds of AβRA were lower in skin taken from FM-treated mice (HC: 0.3 ± 0.08 g versus FM: 0.15 ± 0.05 g, P < 0.05) in particular P1-treated animals (P < 0.001, two-sided unpaired t-test, n = 7). (I) AβRA were sensitized to mechanical stimuli over a range of forces. (J) Peak instantaneous frequency of action potential firing significantly higher at 1 g force step (HC: 338 ± 44 Hz versus FM: 483 ± 38 Hz, P < 0.05, two-sided unpaired t-test, HC: n = 12 and FM: n = 13). (K) Adaption properties were not markedly changed between treatment groups. (L) Representative traces of AβRA at 1 g force step. FM = fibromyalgia; HC = healthy control; LTMR = low threshold mechanoreceptors.

Figure 4

Aβ-slowly adapting fibres from mice injected with fibromyalgia patient IgG fire to cutaneous cooling ex vivo. (A) Representative trace of Aβ-slowly adapting (AβSA) fibres from a healthy control (HC) IgG-treated animal compared to a fibromyalgia (FM) IgG (P1)-treated animal. (B) The proportion of fibres (32%) that fired seven or more action potentials during a 60 s cooling ramp (∼32°C to less than 10°C). These cold sensitive AβSA are present in skin taken from mice injected with both patients’ IgG but not HC IgG. (C) Pie chart indicates the proportion of AβSA that responded to cold by patient sample administered. P1: cold insensitive, represented in grey (5/9 cold sensitive, blue); and P2: cold insensitive, indicated in black (4/19 cold sensitive, blue). Grey and black are then used to identify which patient the AβSA recording came from in both D and E. (D) Sum of number of action potentials fired by the cold sensitive AβSA fibres (FM IgG-treated 87 ± 0 action potentials/60 s stimulus; HC IgG-treated animals 1 ± 0 n = 7–9). Yellow dots represent AβSA recorded from skin of mice injected with P1 and green dots represent fibres from P2 IgG-administered mice. (E) Average temperature threshold of activation of AβSA fibres from FM-injected animals is 21.4°C (±1.8°C). (F) Average firing rate of AβSA over the course of the cold ramp in skin from FM treated mice (n = 9).

Figure 5

Mice injected with IgG from people with fibromyalgia have more large diameter dorsal root ganglion neurons responding to cold in vivo. (A) Representative image of in vivo calcium imaging of dorsal root ganglia (DRG) of mice injected with fibromyalgia (FM) IgG demonstrates calcium influx in larger diameter fibres after submersion of hind-paw in a 10°C water bath, brush and pinch stimuli. Coloured arrows indicate DRG that responded to cold alone (blue), cold and pinch (red), cold and brush (green) and cold, brush and pinch (orange). (B) In healthy control (HC) IgG-treated animals, the majority of DRG neurons that responded to cold temperature with calcium influx were small in diameter. In FM IgG-injected animals, there was an increase in the number of large diameter (>30 µm) DRG neurons responding to cold. (C) Trace of calcium influx during paw immersion in 10°C water in a large diameter (>40 µm) and small diameter (17 µm) neuron. (D) Overall polymodality of all the cold responding neurons remained similar between HC and FM. (E) The average total number of cold responding DRG neurons per animal was increased in mice treated with FM IgG (HC 27 ± 2 and FM 37 ± 3 neurons, P < 0.05, two-sided unpaired t-test, n = 5 HC and FM n = 6 animals). (F) The number and percentage of cold responding neurons that were greater than 30 µm in diameter of each class; cold alone (blue), cold and pinch (red), cold and brush (green) and cold, brush and pinch (orange). (G) A comparison of the diameter of fibres that responded to cold alone or cold and other mechanical stimuli demonstrated that large diameter fibres (line denotes 30 µm) responded to cold across all groups. In HC IgG-treated animals, fewer medium to large diameter neurons responded to brush and cold stimuli (two-sided unpaired t-test, P < 0.01).

Figure 6

People with fibromyalgia have an altered perception of cold stimuli alongside changes to the firing of Aβ-slowly adapting fibres. (A) People living with fibromyalgia (FM) were asked to describe the sensation of cooling during Quantitative Sensory Testing (QST) use different phrases compared to healthy control (HC) participants. Dashed line highlights the reduction in the number of patients saying that the cold stimulus was cold (HC: 6/22 and FM 5/49) and the concurrent increase in number of people who describe the sensation as like ‘pins and needles/tingling’ (HC 2/22 and FM 14/49). (B) Compared to Aβ-slowly adapting (AβSA) afferents recorded from HC (left, black), AβSA in people with FM (right, red) fire fewer action potentials during mechanical indentation of receptive field patients with Von Frey hair. (C) Average number of action potentials fired by AβSA in the first 10 s of the 10–13 g Von Frey stimulation was reduced in FM (97 ± 15 AP, n = 12) compared to HC participants (286 ± 31, n = 18, two-tailed unpaired t-test, P < 0.05). See Supplementary Fig. 2 for a breakdown of response by sex and by site. (D) The number of action potentials binned per second demonstrates that the adaption pattern remained similar between FM and HC AβSA afferents, with a lower number of action potentials across both the dynamic and static phases of indentation. (E) The mean inter-spike interval (ISI) was increased in FM patients (HC 0.05 ± 0.01 s n = 18: FM 0.14 ± 0.03 s n = 11, two-tailed unpaired t-test, P < 0.05) without affecting the (F) coefficient of variance (CoV) of ISI indicated that both FM and HC AβSA fired regularly to mechanical stimuli (HC 0.53 ± 0.27 n = 18: FM 0.65 ± 0.44 n = 11). (G) Representative recordings of two cold sensitive AβSA afferents: (i) an AβSA afferent that did not continuously fire in response to the thermode being placed (indicated by the black arrow) on the receptive field compared to (ii) an AβSA afferent which fired continuously after the thermode was positioned on the receptive field. The time of the dynamic phase of the cooling ramp is shown as the blue line above the trace. In the bottom trace, a clear increase in firing (>Δ20%) was observed during the dynamic phase of cooling. (H) The proportion of cold sensitive AβSA was increased in FM, represented as a percentage (HC 6/15 versus FM 8/9, two-tailed Fisher’s exact test, P < 0.05).