Spinal Microgliosis Due to Resident Microglial Proliferation Is Required for Pain Hypersensitivity after Peripheral Nerve Injury

Abstract

Peripheral nerve injury causes neuropathic pain accompanied by remarkable microgliosis in the spinal cord dorsal horn. However, it is still debated whether infiltrated monocytes contribute to injury-induced expansion of the microglial population. Here, we found that spinal microgliosis predominantly results from local proliferation of resident microglia but not from infiltrating monocytes after spinal nerve transection (SNT) by using two genetic mouse models (CCR2(RFP/+):CX3CR1(GFP/+) and CX3CR1(creER/+):R26(tdTomato/+) mice) as well as specific staining of microglia and macrophages. Pharmacological inhibition of SNT-induced microglial proliferation correlated with attenuated neuropathic pain hypersensitivities. Microglial proliferation is partially controlled by purinergic and fractalkine signaling, as CX3CR1(-/-) and P2Y12(-/-) mice show reduced spinal microglial proliferation and neuropathic pain. These results suggest that local microglial proliferation is the sole source of spinal microgliosis, which represents a potential therapeutic target for neuropathic pain management.

Figure 1. No monocyte infiltration into the dorsal horn of spinal cord after spinal nerve transection

(a–c) Representative confocal images of the ipsilateral spinal cord at post-operative day (POD) 14 following spinal nerve transection (SNT) in double transgenic CCR2^RFP/+:CX3CR1^GFP/+ mice (a). Resident GFP⁺ microglia (CX3CR1^GFP/+) are shown in green and infiltrated RFP⁺ monocytes (CCR2^RFP/+) are shown in red. Dotted boxes show regions of higher magnification in the dorsal horn (DH, b) and ventral horn (VH, c), respectively. Note the obvious appearance of CCR2^RFP/+ cells in the ipsilateral VH (c), but not DH (b). Scale bar is 200 μm (a) and 50 μm (b, c), respectively. Representative images of CCR2^RFP cell infiltration in DH and VH before SNT and at POD7 after SNT are shown in Figures S1a–c and S1f–h. (d–e) CCR2^RFP cell infiltration was found in L4 dorsal root ganglia (DRG, d) and the damaged spinal nerve stump (e) at POD 14 following SNT in CCR2^RFP/+:CX3CR1^GFP/+ mice. Scale bar is 100 μm. Representative images of CCR2^RFP cell infiltration into DRG before SNT and at POD7 after SNT are shown in Figure S1d–e. (f–g) Representative images of the spinal cord DH at POD 14 following SNT in CX3CR1^creER/+: R26^tdTomato/+ reporter mice (f). The dotted box indicates the region that is magnified in g. Resident microglia are tdTomato⁺Iba1⁺ cells and hematogenous monocytes are tdTomato⁻Iba1⁺ cells. There are no tdTomato⁻Iba1⁺ cells in the DH. Scale bar is 150 μm (f) and 50 μm (g), respectively. (n=4 mice per group) Representative images before SNT and at POD3, POD7 following SNT using CX3CR1^creER/+: R26^tdTomato/+ reporter mice are shown in Figure S3.

Figure 2. No monocyte infiltration in the spinal dorsal horn by P2Y12 and CD169 immunostaining

(a) Representative confocal images of double immunostaining showing co-localization of P2Y12 (red) and Iba1 (white) signals in the spinal cord dorsal horn in sham and POD 7 following SNT in WT mice (n=4 per group). Scale bar is 50 μm. (b) Representative confocal images of P2Y12 staining in the spinal dorsal horn showing CX3CR1^GFP positive (green) microglia are also P2Y12⁺ (red) in sham and POD 7 after SNT in CX3CR1^GFP/+ mice (n=4 per group). Scale bar is 50 μm. (c) No P2Y12⁺ cells were found in the damaged nerve stump in CX3CR1^GFP/+ mice and no P2Y12 immunostaining was found in the spinal dorsal horn of P2Y12^−/− mice at POD 7 after SNT (n=4 per group). Scale bar is 50 μm. (d) Representative confocal images of double immunostaining showing co-localization of CD169 (green) and Iba1 (red) signals in the ipsilateral spinal cord at POD 7 in WT mice (n=2–3 mice per group). Scale bar is 100 μm. Note the obvious appearance of CD169⁺ cells (green) in the ipsilateral VH, but not DH. (e) CD169⁺ cell infiltration was also found in L4 dorsal root ganglia (DRG) at POD 7 in WT mice (n = 2–3 mice per group). Scale bar is 50 μm.

Figure 3. Local microglial proliferation dominates spinal microgliosis after spinal nerve transection

(a) Representative confocal images of spinal cord DH showing co-localization of BrdU (green) and Iba1 (red) at POD 7 following SNT. BrdU (i.p. 100mg/kg) was applied immediately after SNT at 2 pulses/day for 7 days (n=4–6 mice per time point). Inset is z-sectioned image with high magnification, showing that BrdU+ signal are located in the nuclei of Iba1⁺ microglia. Scale bar is 200 μm and 10 μm for lower and higher magnification images, respectively. Representative images of the co-localization at POD3 after SNT are shown in Figure S4a. (b) Quantitative summary of the number of BrdU⁺ Iba1⁺ microglia in the ipsilateral DH (lamina I–IV) at POD3 and 7 was dramatically increased after SNT (see also Figure S4a–b for comparison of BrdU⁺ cells between contralateral and ipsilateral DH). Data are presented as mean ± s.e.m. Shown is summarized data for the number of Iba1⁺ and Iba1⁺BrdU⁺ cell per mm² in DH in sham control and at POD3 and 7 after surgery (n = 4–6 mice per time point). (c–d) Microglial proliferation accounted for a majority of proliferating events after SNT. Shown is the percentage of BrdU⁺ Iba1⁺ cells among total Iba1⁺ cells, i.e mitotic microglial cells in total microglia (c), or the percentage of BrdU⁺ Iba1⁺ cells in BrdU⁺ cells, i.e. microglia in total mitotic cells (d) at POD 3 and 7 after SNT (n = 4 mice at POD3 and 6 mice at POD7).

Figure 4. Dynamics of microglial proliferation in the spinal dorsal horn after spinal nerve transection

(a) Representative images of Ki-67 immunostaining in L4 lumbar DH at POD 3 and 5 after SNT (n=6 mice). Scale bar is 150 μm. Co-localization of Ki-67 immunostaining with Iba1⁺ microglia is shown in Figure S5a. (b) The time course of Ki-67 expressing cells, which peaks at POD3, in the DH after SNT. Data are presented as mean ± s.e.m. Values represent the number of Ki-67⁺Iba1⁺ cells (per DH slice) (n = 6 mice per each time point; ***P < 0.001 versus number of Ki67⁺Iba1⁻ cells at the corresponding time point). (c) Time-lapse images (180 min) of a dividing GFP⁺ microglia in acute spinal cord slices at POD 3 following SNT. Arrows denote the dividing microglia [parent cells (white arrows); daughter cells (red arrows)] and time is presented as hr:min (n=4 mice). The dotted box shows the region with magnification. Scale bar is 100 μm and 10 μm for lower and higher magnification images, respectively. Two-photon live imaging of microglia proliferation in the DH from spinal slice at POD3 after SNT is presented in Movie S1.

Figure 5. Microglial proliferation determines neuropathic pain hypersensitivities after spinal nerve transection

Anti-mitotic drug cytosine arabinoside (AraC, i.t., 50ug in 5μl ACSF, twice/day at POD 1–4 and then once/day at POD 5–7) effectively decreases the number of Iba1⁺ microglia after SNT. (a) Representative confocal images of microglia (Iba1 staining) in the ipsilateral spinal cord at POD 7 following SNT in sham, SNT, and SNT+ AraC treated mice (n=4 mice). Scale bar is 200 μm. (b) Summarized data for the number of Iba1⁺ cell per mm² in the ipsilateral dorsal horn (lamina I–IV) from the three groups shown in a (n = 4 mice in sham group; 7 mice in SNT and SNT + AraC groups). Data are presented as mean ± s.e.m.; ***P < 0.001 compared with SNT control; ^##P < 0.01, ^###P < 0.001 compared with sham. (c) Attenuation of mechanical allodynia and thermal hyperalgesia after SNT by intrathecally administered AraC. Data are presented as mean ± s.e.m. (n =7 mice per group; **P < 0.01, ***P < 0.001 compared with SNT control). (d) A positive correlation between the number of microglia in the DH and the extent of allodynia (index of AUC) after AraC administration (n=4 mice in sham group; 7 mice in both SNT and SNT+AraC groups). To further investigate the molecular mechanisms underlying microglial proliferation, we examined microglial proliferation by BrdU labeling (BrdU, i.p. 100mg/kg, 1 pulses/day at POD 1–3) in CX3CR1^−/− mice and in P2Y12^−/− mice, both have reduced allodynia compared with WT mice (Figure S6a–b). (e) Representative confocal images showing double staining of BrdU⁺ (green) and Iba1⁺ (red) cells in the spinal cord DH at POD 7 following SNT in WT, CX3CR1^−/− and P2Y12^−/− mice (n=4 mice). Scale bar is 200 μm. (f) Summarized data showing the reduced number of BrdU⁺ Iba1⁺ cells per mm² in DH in CX3CR1^−/− and P2Y12^−/− mice compared with that in WT mice at POD 7 following SNT. All data are mean ± s.e.m. (n=6–7 mice; *P < 0.05 compared with WT mice). Further analysis of microglial proliferation in P2Y12^−/− mice is shown in Figure S6c–e.