Dorsal root ganglia CSF1+ neuronal subtypes have different impact on macrophages and microglia after spared nerve injury

Abstract

Background and aims: Colony-stimulating factor 1 (CSF1) is a growth factor secreted by dorsal root ganglia (DRG) neurons important for DRG macrophages and spinal cord (SC) microglia injury-induced proliferation and activation, specifically released after spared nerve injury (SNI). In this study, we investigated if SNI-induced CSF1 expression and perineuronal rings of macrophages around mouse DRG neurons vary between L3-L5 DRG and with the neuronal type, and if the CSF1⁺ neuronal projections at the SC dorsal horns were associated with an increased microglial number in the corresponding laminae.

Methods: Seven days after surgery, L3-L5 DRG as well as their corresponding segments at the SC level were collected, frozen, and cut. DRG sections were double-immunostained using antibodies against CSF1 and NF200, CGRP or IB4, while SC sections were immunostained using a fluorescent Nissl Stain and analyzed for CX3CR1-GFP microglia number and distribution by an in-house ImageJ Plug-in.

Results: Our results showed that SNI-induced CSF1 expression was common for all subtypes of mouse DRG neurons, being responsible for attracting more resident macrophages around them in a DRG-dependent manner, with L4 showing the stronger response and CSF1⁺/NF200⁺ neurons showing the highest incidence. Even though the total number of microglia in the SC ipsilateral dorsal horns increased after SNI, the increase at their specific laminar projection sites did not mirror the incidence of DRG neuronal subtypes among CSF1⁺ neurons.

Interpretation: Taken together, these results contribute to a more comprehensive understanding of the connection between CSF1 and macrophage/microglia response after SNI and emphasize the importance of considering L3-L5 DRG individually when investigating SNI-neuropathic pain pathogenesis in mice.

Keywords: CSF1; macrophages; microglia; neuropathic pain.

FIGURE 1

Spinal cord Rexed laminae analysis algorithm. The figure outlines the main steps of the algorithm used to attribute spinal cord (SC) microglia to Rexed Laminae, based on a fluorescent Nissl stain image of the neuronal bodies (blue) and a CX3CR1⁺ fluorescence image of microglia (green): (1) Identification of the gray matter area, (2) Collection of positional information, (3) Identification and classification of the neurons based on their size and the local density of neurons, (4) Construction of a Voronoi map of the gray matter and color thresholding, (5) ROI construction based on the logic shown on the bottom of the figure, and (6) Identification of microglia and overlay with Rexed lamina ROIs, to obtain (7) the number and density of microglia in each Rexed Lamina.

FIGURE 2

SNI‐induced macrophage expansion is stronger in L4 DRG, compared with L3 and L5 DRG. (A, C and E) Representative fluorescent immunohistochemistry images illustrating CX3CR1⁺ macrophages 7 days after SNI surgery in L3 (A), L4 (C) and L5 (E) DRG (scale bar, 100 μm). (B, D and F) Bar graphs representing the number of CX3CR1⁺ macrophages normalized to Control condition: SNI lesion induced a significant increase compared with Sham condition only in L3 DRG (**p < .01) and L4 DRG (****p < .0001). L4 DRG seems more sensitive, since Sham surgery alone also induced a significant increase compared with Control condition (**p < .01). (G) Bar graph representing the comparison between L3, L4 and L5 DRG, showing that the strongest increase of CX3CR1⁺ macrophages after SNI was in L4 DRG, followed by L3 and L5 (****p < .0001 for L3 vs. L4 and for L4 vs. L5).

FIGURE 3

The number of CSF1⁺ neurons in L3, L4 and L5 DRG varies after SNI. (A, C and E) Representative fluorescent immunohistochemistry images illustrating CSF1⁺ DRG neurons 7 days after SNI surgery in L3 (A), L4 (C), and L5 (E) DRG (scale bar, 50 μm). (B, D, and F) Bar graphs representing the number of CSF1⁺ neurons normalized to Control condition. SNI lesion induced a significant increase compared with Sham condition in all DRG (****p < .0001 for L3 and L4 DRG, and **p < .01 for L5 DRG). L4 DRG seems more sensitive, since Sham surgery alone also induced a significant increase compared with Control condition (**p < .01). (G) Bar graph representing the comparison between L3, L4, and L5 DRG, showing that the strongest increase of CSF1⁺ neurons after SNI was in L4 DRG, followed by L3 and L5 (**p < .01 for L3 vs. L5 and ****p < .0001 for L4 vs. L5).

FIGURE 4

Seven days after SNI, CX3CR1⁺ macrophages cluster around CSF1⁺ neurons in L3, L4 and L5 DRG. (A–C) Representative fluorescent immunohistochemistry images illustrating CSF1⁺ neurons (in red), CX3CR1⁺ macrophages (in green), and merged images, in L3 (A), L4 (B), and L5 (C) DRG. Insets represent high‐magnification images of the selected areas, showing the perineuronal rings of macrophages that surround CSF1⁺ neurons after the lesion (indicated by white arrows) (scale bar, 50 μm). (D) Bar graph showing mean percentages of CSF1⁺ neurons surrounded by rings of CX3CR1⁺ macrophages in all 3 DRG: L4 DRG > L3 DRG > L5 DRG.

FIGURE 5

SNI increased the incidence of CSF1⁺/NF200⁺ DRG neurons and decreased the number of CSF1⁺/CGRP⁺ and CSF1⁺/IB4⁺ DRG neurons. (A, C, and E) Representative fluorescent immunohistochemistry images illustrating CSF1⁺/NF200⁺, CSF1⁺/CGRP⁺, and CSF1⁺/IB4⁺ 7 days after SNI in L3 (A), L4 (C), and L5 (E) DRG (scale bar, 50 μm). (B, D, and F) Bar graphs representing the percentages of CSF1⁺/NF200⁺, CSF1⁺/CGRP⁺, and CSF1⁺/IB4⁺ 7 days after SNI in L3 (B), L4 (D), and L5 (F) DRG. Although there is an increasing trend for CSF1⁺/NF200⁺ and a decreasing trend for CSF1⁺/CGRP⁺ and CSF1⁺/IB4⁺ in all DRG, only in L3 DRG there is a significant decrease for CSF1⁺/IB4⁺ neurons. (G) Bar graph showing mean percentages of CSF1⁺/NF200⁺, CSF1⁺/CGRP⁺, and CSF1⁺/IB4⁺ neurons in all 3 DRG 7 days after SNI. There were no differences between L3, L4, and L5 DRG for each neuronal types.

FIGURE 6

CX3CR1⁺ microglia expansion in the ipsilateral SC (L3–L5 spinal segments) changes from Lamina I to Lamina VII. (A) Representative fluorescent immunohistochemistry images illustrating CX3CR1⁺ microglia (in green), Neurotrace‐labeled neurons (in blue), and merged images (with laminae borders) in Control and 7 days after Sham and SNI in the ipsilateral half of L3–L5 SC segments (scale bar, 100 μm). (B) Bar graph representing the number of CX3CR1⁺ microglia in the ipsilateral half of the SC (L3–L5 spinal segments). SNI induced a significant increase compared with Sham condition (****p < .0001). (C–I) Bar graphs illustrating the number of CX3CR1⁺ microglia from Lamina I to Lamina VII. After SNI, the number of microglia significantly increased in all laminae compared with Sham condition (****p < .0001 for all laminae). (J) Bar graph illustrating the difference between the degree of microglia accumulation in each lamina, 7 days after SNI. The strongest effect was in Lamina I and II, followed by Lamina III, IV, and VII, while the smallest effect was in Lamina V.

FIGURE 7

SNI‐activated DRG neurons increase the number of CX3CR1⁺ microglia more in Laminae I + II. (A–C) Bar graphs representing the number of CX3CR1⁺ microglia in Laminae I + II (A), Laminae III + IV + V (B), and a comparison between them (C). SNI induced an increase in the number of microglia at both nociceptors' projection site (Laminae I + II) and mechanoreceptors' projection site (Laminae III + IV + V) compared with Sham, but the increase was significantly higher in Laminae I + II (****p < .0001). (D) Reconstruction of the ipsilateral half of the L3‐L5 spinal cord (SC) showing how microglia (represented as green dots) distribution varies between spinal segments. In Control mice, these cells were evenly scattered throughout the whole ipsilateral half, from L3 to L5, while in the Sham condition there was a slight accumulation mainly in the dorsal horns of the L4 spinal segment. After SNI injury, microglia accumulated mainly in the ventral horns in L5 region, and in both dorsal and ventral horns in L3 and L4 regions, gradually decreasing toward L2 and L6.