Passive transfer of fibromyalgia symptoms from patients to mice

Abstract

Fibromyalgia syndrome (FMS) is characterized by widespread pain and tenderness, and patients typically experience fatigue and emotional distress. The etiology and pathophysiology of fibromyalgia are not fully explained and there are no effective drug treatments. Here we show that IgG from FMS patients produced sensory hypersensitivity by sensitizing nociceptive neurons. Mice treated with IgG from FMS patients displayed increased sensitivity to noxious mechanical and cold stimulation, and nociceptive fibers in skin-nerve preparations from mice treated with FMS IgG displayed an increased responsiveness to cold and mechanical stimulation. These mice also displayed reduced locomotor activity, reduced paw grip strength, and a loss of intraepidermal innervation. In contrast, transfer of IgG-depleted serum from FMS patients or IgG from healthy control subjects had no effect. Patient IgG did not activate naive sensory neurons directly. IgG from FMS patients labeled satellite glial cells and neurons in vivo and in vitro, as well as myelinated fiber tracts and a small number of macrophages and endothelial cells in mouse dorsal root ganglia (DRG), but no cells in the spinal cord. Furthermore, FMS IgG bound to human DRG. Our results demonstrate that IgG from FMS patients produces painful sensory hypersensitivities by sensitizing peripheral nociceptive afferents and suggest that therapies reducing patient IgG titers may be effective for fibromyalgia.

Keywords: Autoimmune diseases; Neurological disorders; Neuroscience; Pain.

Figure 1. Passive transfer of hypersensitivities from fibromyalgia patients to mice.

Administration of IgG (8 mg on 4 consecutive days) from each of 8 different FMS patients (P1–P8) significantly reduced the withdrawal threshold in the paw-pressure test (A–F) compared with IgG from healthy control subjects (HC1–HC6). The paw withdrawal latency in the cold-plate test was reduced by IgG from 7 of 8 patients (G–L). Data points are mean ± SEM of n = 6 mice in A, C, E, G, I, and K; n = 5 in D, F, J, and L; and n = 4 in B and H. *P < 0.05, **P < 0.01, ***P < 0.001, FMS IgG compared with HC IgG; 2-way repeated measure ANOVA followed by Sidak’s correction. ^†P < 0.05, ^††P < 0.01, ^†††P < 0.001, compared with the naive preinjection value at time zero; 2-way repeated measure ANOVA followed by Dunnett’s test.

Figure 2. FMS IgG produces polymodal abnormalities.

FMS IgG increased the sensitivity to punctate stimulation with von Frey filaments (A). The threshold for leg withdrawal in response to pressure applied to the thigh (using a Randall-Selitto device) was reduced by FMS IgG compared with HC IgG (B). Female and male mice are affected equally by FMS IgG in the paw-pressure test (C) and the cold-plate test (D). Mechanical hypersensitivity produced by either 1, 2, or 4 injections of 8 mg FMS IgG (E), and by single injections of 2, 4, or 8 mg (F). The front paw grip strength is reduced by FMS IgG compared with HC IgG (G). Data points are mean ± SEM or individual measurements. *P < 0.05, **P < 0.01, ***P < 0.001, FMS IgG compared with HC IgG; 2-way repeated measure ANOVA followed by Sidak’s correction (A, C, D, and G). Data in B were analyzed by unpaired, 2-tailed t test. Data in E and F were compared to the naive preinjection value at time zero by 2-way repeated measure ANOVA followed by Dunnett’s test.

Figure 3. Passive transfer of hypersensitivity by IgG pooled from multiple patients.

Visual analog pain scores (VAS, A) and pressure-pain thresholds (PPT, B) in 2 pools of FMS patients and healthy control (HC) subjects. Pool 1, n = 8 FMS and n = 12 HC; Pool 2, n = 14 FMS and n = 10 HC. **P < 0.01, ***P < 0.001 by Mann-Whitney U test. Administration of IgG pooled from FMS patients produced mechanical (C) and cold (D) hypersensitivity in mice compared with pooled HC IgG, 4 days after the first injection. Pool 1, n = 6 mice per group; Pool 2, n = 12; line and whiskers indicate mean ± SEM. **P < 0.01, ***P < 0.001 by unpaired, 2-tailed t test. Time course of mechanical (E) and cold (F) hypersensitivity following administration of pooled IgG (Pool 1). *P < 0.05, FM vs. HC IgG; 2-way repeated measure ANOVA with Sidak’s correction. ^†P < 0.05, ^††P < 0.01, compared with day zero; 2-way repeated measure ANOVA followed by Dunnett’s test. Data in E and F are mean ± SEM of 6 mice per group.

Figure 4. IgG-depleted serum is inactive.

IgG from an FMS patient, but not IgG-depleted serum from the same patient or pooled from FMS or HC cohorts, produced mechanical (A) and cold (B) hypersensitivities. **P < 0.01, ***P < 0.01, FM vs. HC IgG; 2-way repeated measure ANOVA followed by Sidak’s correction. Data points are mean ± SEM of 6 mice per group.

Figure 5. Passive transfer of FMS IgG decreases locomotor activity.

Visual analog pain scores (VAS, A) and pressure-pain thresholds (PPT, B) reported by FMS patients and healthy control (HC) subjects in Pool 3. Following transfer of HC or FMS IgG to mice, locomotor activity was assessed over a 24-hour period using a Comprehensive Animal Lab Monitoring System. The total number of recorded movements was similar between HC IgG and FMS IgG mice during the day phase (low activity), but FMS IgG–injected mice showed less activity during the night phase (high activity) (C). The night phase was divided into 3 phases: 18:00–22:00 hours (D), 22:00–02:00 hours (E), and 02:00–06:00 hours (F), indicated by the dotted lines. Mice injected with FMS IgG displayed significantly less locomotor activity during the peak activity phase (E) compared with HC IgG–injected mice. Data points are mean ± SEM, FMS n = 14, HC n = 11, n =12 mice per group. *P < 0.05, ***P < 0.001 by Mann-Whitney U test (A and B), 2-way ANOVA followed by Bonferroni’s correction (C), or unpaired t test (D–F).

Figure 6. Passive transfer of FMS IgG sensitizes nociceptors.

The mechanical activation thresholds of (A) Aδ- (AM) and (B) C-mechanonociceptors (CM) were reduced in preparations from mice treated with FMS compared with HC IgG (n = 22–27 single units). *P < 0.05 by 1-tailed Mann-Whitney U test. (C) The example trace illustrates a mechanical threshold response (evoked by the minimum force required to elicit at least 2 spikes) in a CM unit. (D) The proportion of cold-sensitive CM units (CMCs) was increased in preparations from mice treated with FMS IgG (21 of 29 units responded to cold) compared with HC IgG (13 of 28 units responded to cold). *P < 0.05 by 1-sided Fisher’s exact test. (E) The cold-activation thresholds of CMC fibers did not differ between FMS and HC preparations (n = 13–21). P > 0.05 by 1-tailed t test. (F) The example trace illustrates a cold-evoked response in a CMC fiber. (G) Application of FMS IgG (200 μg/mL) to isolated DRG neurons loaded with Fura-2 was without effect on [Ca²⁺]_i in all 870 examined neurons (identified by their response to 50 mM KCl). The red trace illustrates the average time course of the displayed 230 neurons.

Figure 7. FMS IgG accumulates in the DRG and binds satellite glial cells.

Following IgG injection into mice, Western blot analysis detected FMS IgG in the DRG but little to no IgG in spinal cords (SC) or brains (A) (pooled IgG, 8 mg per day for 4 consecutive days, tissue collected after last injection). FMS IgG, but not HC IgG, accumulates in the DRG 14 days after the first IgG injection (B and C). Human IgG is red and DAPI is blue. Human IgG immunoreactivity in the neuron-rich area was quantified by assessing the percentage area that was immunoreactive for human IgG and the mean pixel intensity of human IgG. Percentage area and pixel intensity were normalized to the DAPI signal (n = 9–10; data points are median ± 95% CI). **P < 0.01, ***P < 0.001 by 1-way ANOVA followed by Tukey’s post hoc test. FMS IgG immunoreactivity does not colocalize with neuronal NeuN staining but does colocalize with satellite glial cells (SGCs) (glutamate synthase–expressing [GS-expressing] cells), some macrophages (Iba1-expressing cells) and blood vessels (CD31-expressing cells), and myelinated fiber tracts (myelin basic protein [MBP] staining), but not to myelinated fibers in the DRG (D). To further delineate between SGCs and neuronal membranes, FMS IgG immunoreactivity colocalization was compared with GS and TrkA (a membrane receptor expressed by a subset of nociceptors). FMS IgG colocalizes with GS-expressing SGCs (white arrows) but may also infrequently bind to TrkA-positive neuronal cell membranes (white triangles) (E). Scale bars indicate 50 μm, except the high-magnification image scale bar and scale bar in E, which indicate 25 μm.

Figure 8. FMS IgG increases signs of satellite glial cell activity in vivo but does not drive systemic inflammation.

DRG from FMS IgG–injected mice have increased GFAP immunoreactivity (A), which is indicative of increased satellite glial cell activity, compared with HC IgG injected mice when the percentage area of GFAP immunoreactivity and GFAP mean pixel intensity are quantified and normalized to the DAPI signal (B). Gfap and s100b gene expression is elevated in the DRG of mice injected with FMS IgG compared with HC IgG (C). The number of Iba1-immunoreactive macrophages was unchanged, as was the percentage area of Iba1 immunoreactivity, when comparing HC IgG– and FMS IgG–injected mice (D and E). Gene expression of Aif1 (Iba1 gene) and Itagm (gene for CD11b, another macrophage marker) was elevated in FMS IgG–injected mice compared with HC IgG–injected mice (F). Scale bars: 50 μm. qPCR data were normalized to Hprt1 expression analyzed using the 2^–ΔΔCt method. Serum levels of TNF (G), CXCL1 (H), IL-2 (I), IL-5 (J), IL-6 (K), IL-10 (L), and IFN-γ (M) were measured and there were no differences between groups. The dashed lines indicate the lower limit of quantification (LLOQ) and the dotted lines indicate the lower limit of detection (G–M). Line and whiskers indicate mean ± SEM (n = 4–6). Differences between FMS IgG and HC IgG were analyzed with Mann-Whitney U test (B–F). Cytokine levels were compared between saline, HC, and FMS with 1-way ANOVA followed by Tukey’s post hoc test for each analyte (G–M) except IL-6, which was not analyzed statistically because most values were below the LLOQ. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 9. FMS IgG binds to satellite glial cells and to neurons in vitro.

Live cells were incubated with FMS IgG or HC IgG to examine only cell surface binding. In satellite glial cell (SGC) cultures (A), FMS IgG labeled a greater percentage of cells than HC IgG when analyzed per animal (B) and by the total number cells (C). The immunoreactivity, analyzed as signal intensity (integrated density), was higher in cells incubated with FMS IgG than HC IgG (D). FMS IgG labeled SGCs (GS⁺) and non-SGCs (GS^–) to a greater extent (percentage, E) and with a higher intensity (F) than HC IgG. FMS and HC IgG labeling of neuronal cultures (G) was not different when considering the percentage of cells labeled per animal (H), but a difference was observed when the total number of cells was considered (I). The signal intensity was higher for cells exposed to FMS IgG than HC IgG (J). FMS IgG labeled neurons (βIII-tubulin⁺) and non-neurons (βIII-tubulin^–) to a greater extent than HC IgG (K). The signal intensity of FMS IgG binding to neurons was greater than FMS IgG binding to non-neuronal cells and HC IgG binding to all cells (L). FMS IgG and IB4 colocalization (M) indicates that FMS IgG binds neuronal cell membranes. All scale bars: 20 μm. Data points are the percentage of cells bound by HC or FMS IgG (B and H). In D, F, J, and L data points are the integrated density of individual cells across 3 experiments. Bar and whiskers indicate mean ± SEM (n = 3 individual experiments). *P < 0.05; ***P < 0.001 by unpaired t test (B, D, H, and J), χ² test (C, E, I, and K), or Kruskal-Wallis test with Dunn’s post hoc test (F and L).

Figure 10. FMS IgG transfer decreases intraepidermal nerve fiber density.

Intraepidermal nerve fibers (IENFs) were identified in the glabrous hind-paw skin with an anti–PGP 9.5 antibody (A). The number of IENFs crossing from the dermis to the epidermis was decreased following transfer of FMS IgG compared with HC IgG 14 days after the first injection (B) (pooled, 8 mg per day for 4 consecutive days). Scale bar: 20 μm. Data points are mean ± SEM (n = 7). *P < 0.05 by unpaired t test.

Figure 11. FMS IgG binds human DRG.

FMS IgG bound human DRG tissue sections more intensely than HC IgG when assessed by integrated density normalized to DAPI (A and B). The indicated IgG is in white in the top row and DAPI is in blue in the bottom row of A. High-magnification images of FMS IgG (white) demonstrate colocalization (purple) with GFAP-immunoreactive satellite glial cells (green) and NF-200–immunoreactive neurons (green) (C). Scale bars: 50 μm (A) and 20 μm (C). Data are mean ± SEM (n = 5 independently stained slides). ***P < 0.001 by unpaired t test.