Iba1 deficiency impairs microglial synaptic remodeling and neuronal survival after axonal injury Authors: Koji Sekiguchi, Hirotaka Shoji, Tomoko Shindo, Jumpei Sasabe, Daiki Tokuyasu, Jin Nakahara, Tsuyoshi Miyakawa, Daisuke Ito Published: 2026 ABSTRACT BackgroundMicroglia remodel neuronal circuits in pathological conditions; however, the molecular requirements for these responses and their consequences for motoneuron survival remain unclear.MethodsAif1 (Iba1) knockout mice were generated using CRISPR/Cas9-mediated deletion, and baseline phenotypes and responses to unilateral facial nerve axotomy were assessed using immunohistochemistry, transmission electron microscopy, and single-nucleus RNA sequencing of the facial motor nucleus. Motoneuron survival and nuclear γH2AX foci were evaluated 28 days post-axotomy.FindingsUnder baseline conditions, Iba1−/− mice had reduced body weights and mild behavioral abnormalities compared to wild-type mice. After axotomy, microglial ensheathment of ChAT-positive facial motoneurons was reduced, with fewer neurons showing extensive perisomatic microglial coverage than in Iba1+/+ mice. Ultrastructurally, somatic synapse loss observed after injury in wild-type mice was not detected in Iba1−/− mice, and fewer injured motoneurons were in contact with microglial processes. Single-nucleus transcriptomics showed an exaggerated expansion of an interferon-responsive microglial state in Iba1−/− mice after axotomy, whereas injured motoneurons displayed altered transcriptional programs related to synapse organization and neurotransmission. At 28 days, Iba1−/− mice showed reduced motoneuron survival, lower ChAT expression, and increased nuclear γH2AX foci.InterpretationIba1 supports microglia-neuron cross-talk that enables effective perisomatic remodefling after axonal injury; disruption of this response is accompanied by inflammatory-state shifts and compromised motoneuron survival.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12974-026-03848-6. Background Among central nervous system (CNS) cellular components, microglia—brain resident immune cells—play a central role in neural injury response. Historically viewed as bystanders or contributors to neuroinflammation, microglia are now recognized as dynamic CNS microenvironment regulators. Their activation is a hallmark of various neurological diseases, such as Alzheimer’s disease (AD), amyotrophic lateral sclerosis (ALS), and multiple sclerosis [1]. However, whether microglial activation contributes to neural protection or exacerbates pathology remains debatable. Although degeneration is a key pathological event in various neurological diseases, including demyelinating diseases, neurodegeneration, and CNS injury, the mechanisms governing axonal repair after axonal damage remain unclear. In the late 1960s, the phenomenon of microglial ensheathment and synaptic stripping of motoneurons was first described. After axonal injury, microglia were observed to extend their processes, ensheathe neurons, and physically displace synaptic terminals from the surface of injured motoneurons [2]. This process may act as a protective mechanism by transiently reducing excitatory inputs and creating a permissive environment for neuronal protection [3–5]. However, its biological significance has not been directly examined. Synaptic stripping has been reported in facial nerve axotomy models and at other CNS injury sites, suggesting a fundamental response to axonal damage [6, 7]. Peripheral nerve axotomy, particularly facial nerve axotomy, has long served as a well-established and technically reproducible model for studying motoneuron injury responses, because the lesion can be made unilaterally, does not directly injure the CNS, is well tolerated by the animal, and allows the contralateral facial motor nucleus to serve as an internal control. Within this paradigm, injury to motor axons induces a coordinated central reaction involving not only neuronal gene expression changes but also perineuronal astrocytic hypertrophy, microglial proliferation and migration toward injured motoneuron somata, and remodeling of synaptic coverage [8, 9]. These responses have been interpreted as part of a broader neuron–glia injury program that may influence neuronal survival, regenerative capacity, and circuit reorganization, although their precise causal contributions remain incompletely resolved [8–10]. Despite decades of study, the molecular mechanisms and biological significance of microglial ensheathment and synaptic stripping—and their impact on neuronal survival and repair—remain incompletely defined. Importantly, prior work has emphasized that the magnitude and character of microglial responses after peripheral nerve injury vary with injury type, severity, anatomical location, and species, indicating that microglial activation after axotomy should not be viewed as a monolithic phenomenon [9, 10]. While previous research has focused on modulating upstream microglial activation regulators or pharmacological microglia depletion, findings have been inconsistent [11, 12]. For instance, TREM2 knockout models show variable effects on amyloid pathology and tau aggregation in different AD models [13–15]. CSF1R inhibition reduces microglial numbers but yields conflicting results on plaque burden and neuronal outcomes [16]. However, it is unclear whether microglia are beneficial or detrimental to AD pathology. These discrepancies highlight the dualistic nature of microglia, which appear capable of both neuroprotective and neurotoxic actions, depending on the context. In facial nerve injury models, this context dependence is further underscored by evidence that immune mechanisms both within and outside the CNS can influence facial motoneuron viability after axotomy, including a neuroprotective contribution of CD4 + T-cell-dependent immune responses [17]. To move beyond upstream regulation, we focused on Iba1 (also known as Aif1), a molecule intrinsically involved in microglial cytoskeletal remodeling [3, 18]. Iba1is a calcium-binding protein expressed specifically in microglia and is essential for phagocytosis and membrane ruffling via PLCγ-dependent RAC activation under M-CSF stimulation [19–21]. Monocytes/macrophages lacking Iba1 showed decreased phagocytosis and efferocytosis [22]. Iba1 which functions in a low-level signaling pathway, increases the actin-bundling activity of L-fimbrin, thereby triggering the formation of dynamic structures such as lamellipodia and filopodia [23] (Fig S1). These structures are critical for microglial mobility and close contact with neurons. Although Iba1 is widely used as a microglial marker, its functional and biological significance has been less extensively studied in vivo. Iba1−/− mice, which are conventional knockouts generated by homologous recombination in embryonic stem cells, exhibit reduced microglial ramification, impaired synaptic pruning during development [24], and altered responses in neuroinflammatory models such as experimental autoimmune encephalomyelitis [25]. However, the role of Iba1 in microglia under pathological conditions remains poorly understood. Given its pivotal role in actin dynamics, we hypothesized that Iba1 is a key effector in synaptic stripping, mediating the intimate interaction between microglia and injured neurons. The impairment of this mechanism can disrupt synaptic remodeling and hinder post-axonal injury survival. Moreover, data from human pathological studies have shown that Iba1 expression and microglial morphology are altered in aging and neurodegenerative diseases, further underscoring its potential relevance in disease contexts [26, 27]. In this study, we investigated the role of Iba1 in microglia-neuronal cross-talk following facial nerve axotomy. By combining the well-established axotomy model with genome-edited Iba1−/− animals, we sought to clarify how Iba1 absence affects microglial ensheathment, synaptic stripping, and neuronal survival after injury. Our findings clarify the underexplored downstream mechanisms of microglial action and evaluate whether Iba1 could serve as a therapeutic target to enhance axonal repair in CNS injury. Materials and methods Generation of Iba1-/-mice Iba1 knockout mice were generated using CRISPR/Cas9 technology (Cyagen Biosciences). Guide RNAs (gRNAs) targeting exons 3–8 of the Iba1 gene were designed and introduced to disrupt gene function (Fig. 1a). Genetic modifications were confirmed by PCR amplification followed by Sanger sequencing (Fig. S2a, b). Fig. 1Generation of Iba1-deficient mice (Iba1−/−). A The Iba1 gene was disrupted by deleting exons 3–8 using CRISPR/Cas9 technology. B Representative immunohistochemical images of the cerebral cortex from 8-week-old C57BL/6J mice. Wild-type (Iba1+/+) and Iba1 knockout (Iba1−/−) mouse samples were stained with anti-Iba1 (green), P2RY12 (magenta), and DAPI (blue). Images demonstrate microglial distribution and morphology in the cerebral cortex. Scale bar: 20 μm. C Representative images showing the Sholl analysis procedure in the hippocampus. From left to right: P2RY12 immunofluorescence image of a representative microglial cell, grayscale image, skeletonized image, and an example of concentric circles (5.0–20.0 μm radius, 2.5 μm intervals) overlaid on the skeletonized image. Yellow dots indicate intersections between microglial processes and the circles. Scale bar: 10 μm. D, E Quantification of Sholl analysis of hippocampal microglia from Iba1+/+ and Iba1−/− mice at 8 weeks (d) and 20 months (e) of age. Mice (n = 3 per group) were analyzed in three fields of view, with 10 cells per field, yielding a total of 90 microglia per group. Data are presented as mean ± standard error of the mean. Two-way analysis of variance followed by Šídák’s multiple comparisons test was performed. No significant differences are detected at 8 weeks, whereas at 20 months Iba1+/+ mice exhibit a significantly higher number of intersections at the 20.0-µm radius compared with Iba1−/− mice (**p = 0.0069) Generation of Iba1-deficient mice (Iba1−/−). A The Iba1 gene was disrupted by deleting exons 3–8 using CRISPR/Cas9 technology. B Representative immunohistochemical images of the cerebral cortex from 8-week-old C57BL/6J mice. Wild-type (Iba1+/+) and Iba1 knockout (Iba1−/−) mouse samples were stained with anti-Iba1 (green), P2RY12 (magenta), and DAPI (blue). Images demonstrate microglial distribution and morphology in the cerebral cortex. Scale bar: 20 μm. C Representative images showing the Sholl analysis procedure in the hippocampus. From left to right: P2RY12 immunofluorescence image of a representative microglial cell, grayscale image, skeletonized image, and an example of concentric circles (5.0–20.0 μm radius, 2.5 μm intervals) overlaid on the skeletonized image. Yellow dots indicate intersections between microglial processes and the circles. Scale bar: 10 μm. D, E Quantification of Sholl analysis of hippocampal microglia from Iba1+/+ and Iba1−/− mice at 8 weeks (d) and 20 months (e) of age. Mice (n = 3 per group) were analyzed in three fields of view, with 10 cells per field, yielding a total of 90 microglia per group. Data are presented as mean ± standard error of the mean. Two-way analysis of variance followed by Šídák’s multiple comparisons test was performed. No significant differences are detected at 8 weeks, whereas at 20 months Iba1+/+ mice exhibit a significantly higher number of intersections at the 20.0-µm radius compared with Iba1−/− mice (**p = 0.0069) To obtain a uniform genetic background, heterozygous mice were backcrossed with C57BL/6J mice for three generations. The final generation of heterozygotes was intercrossed to produce homozygous Iba1−/− and wild-type (Iba1+/+) mice. Littermate Iba1+/+ mice served as controls in all experiments. Only male mice were used in all experiments in this study. All mice were housed in a specific pathogen-free facility under standard conditions with a 12-h light/dark cycle, controlled temperature (23 ± 1.0℃), and humidity (55 ± 7.0%). Mice had ad libitum access to food and water. Facial nerve axotomy and perioperative procedures Facial nerve axotomy was performed on adult mice. The mice were anesthetized by intraperitoneal injection of a mixed anesthetic cocktail consisting of medetomidine (0.75 mg/kg), midazolam (4 mg/kg), and butorphanol (5 mg/kg). They were then placed in the prone position, and the surgical area was disinfected with 70% ethanol. A small right postauricular incision was made to expose the facial nerve at the stylomastoid foramen. The main trunk of the facial nerve was identified and completely transected using fine forceps and microsurgical scissors. The wound was closed using a cyanoacrylate adhesive (Aron Alpha, Toagosei Co.) and disinfected. The mice were monitored postoperatively until full recovery from anesthesia and housed under standard conditions with ad libitum access to food and water. Tissues were collected at 7 days post-axotomy to assess the acute-phase response and at 28 days post-axotomy to evaluate the recovery phase. Perioperative analgesia and monitoring Anesthesia was induced and maintained during surgery using a three-drug combination of medetomidine (0.75 mg/kg), midazolam (4 mg/kg), and butorphanol (5 mg/kg). No additional postoperative analgesic was administered. Animals were carefully monitored for seven days after surgery for signs of pain or distress. No unexpected adverse events occurred. Tissue collection and preparation Mice were perfused with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) using a volume equivalent to twice their body weight. The brains were post-fixed in 4% PFA overnight and then cryoprotected by immersion in 20% sucrose solution for 1 day, followed by 30% sucrose solution on the following day. The tissues were embedded in Tissue-Tek® O.C.T. Compound (Sakura Finetek, 4583), frozen, and sectioned coronally at 12-µm thickness using a Leica cryostat (CM3050S). Immunohistochemistry and immunofluorescence For immunostaining, frozen sections were brought to room temperature, and well-preserved sections were selected under a microscope and rehydrated in 1× PBS. Antigen retrieval was performed using 1× Target Retrieval Solution (pH 6.0, Dako) followed by autoclaving at 105℃ for 5 min. Sections were permeabilized with PBS containing 0.25% Triton X-100 for 10 min and washed three times with PBS containing 0.1% Tween-20 (5 min each at room temperature). Sections were then blocked in a donkey- or goat-serum-based blocking solution for 1 h at room temperature. Primary antibodies were applied in blocking solution and incubated overnight at 4℃: Anti-choline acetyltransferase (ChAT; 1:100, cat. no. AB144P, Sigma-Aldrich), anti-GABA(A) α1 receptor (extracellular; 1:400, cat. no. AGA-001, Alomone Labs), GFAP (clone 2.2B10; 1:200, cat. no. 13–0300, Thermo Fisher Scientific), anti-phosphohistone H2AX (Ser139; 1:200, clone JBW301, cat. no. 05-636, Sigma-Aldrich), anti-Iba1 (1:500, cat. no. 019-019741, Wako), anti-P2RY12 (1:100, cat. no. 848001, BioLegend), anti-P2Y12 (1:100, cat. no. AS-55043 A, AnaSpec), anti-Ret (1:250, cat. no. 14556, Cell Signaling Technology), anti-TMEM119 (1:500, cat. no. ab209064, Abcam), anti-VGLUT1 (1:500, cat. no. 135–303, Synaptic Systems), and anti-VGLUT2 (EPR21085; 1:500, cat. no. ab216463, Abcam). After washing with PBS containing 0.1% Tween-20, sections were incubated with secondary antibodies (1:200, cat. no. A-11034, A-21424, A-11055, A-31572, A-31571, Thermo Fisher Scientific; cat. no. ab150155, Abcam) and DAPI (3:1000) for 1 h at room temperature. The sections were then mounted using Fluoromount-G, coverslipped, and dried in the dark. Imaging and image analysis Cross-sectional images of mouse brain tissue were obtained using an all-in-one fluorescence microscope (BZ-X810, KEYENCE, Osaka, Japan) equipped with an optical sectioning module (BZ-H4XF, KEYENCE, Osaka, Japan). For direct comparisons between genotypes or experimental conditions within the same assay, sections were processed and imaged under matched conditions. Image analysis was performed using ImageJ/Fiji. Unless otherwise specified, background fluorescence was measured from off-tissue regions and subtracted before quantification. Cortical P2RY12 quantification To assess basal cortical P2RY12 immunoreactivity in 8-week-old mice, images were acquired from the cerebral cortex under identical imaging settings across genotypes. P2RY12 fluorescence intensity was quantified within cortical regions of interest after subtraction of off-tissue background. In parallel, the P2RY12-positive area fraction was measured by applying a uniform threshold to all images obtained under matched conditions and dividing the thresholded P2RY12-positive area by the total cortical ROI area. These measurements were used to compare cortical P2RY12 signal preservation between Iba1+/+ and Iba1-/- mice. Hippocampal Sholl analysis of microglial ramification Sholl analysis was performed to assess the branching complexity of P2RY12-positive microglia in the hippocampus of mice at 8 weeks and 20 months of age. A representative workflow is shown in Fig. 1c. Briefly, P2RY12 fluorescence images were processed in ImageJ/Fiji and converted into skeletonized images. The center of each cell was defined at the center of the nucleus, and concentric circles were drawn with radii from 5 to 20 μm at 2.5-µm intervals. The number of intersections between microglial processes and each circle was manually counted at each radius for each cell. Microglia from mice (n = 3 per group) were evaluated across three fields of view, with 10 cells per field, yielding a total of 90 microglia per group. Quantification of microglial perisomatic coverage in the facial motor nucleus Microglial coverage of facial motoneuron somata was quantified using two complementary approaches. First, ChAT-positive (ChAT+) facial motoneurons were identified in images acquired at the focal plane in which nucleoli were visible within DAPI-stained nuclei, and each motoneuron soma was manually outlined as a region of interest (ROI). Based on the extent of surrounding P2RY12-positive microglial contact, motoneurons were classified into five categories: I, absent; II, punctate; III, segmental (< 1/3 of the soma circumference); IV, segmental (1/3–2/3); and V, extended (> 2/3). More than 100 motoneurons were analyzed from n = 4 Iba1+/+ and n = 4 Iba1-/- mice. Second, perisomatic microglial signal intensity was quantified using an annulus-based ROI method. For each ChAT+ motoneuron, the soma ROI was expanded outward by 3 μm to generate a donut-shaped perisomatic ROI. P2RY12 fluorescence intensity within this perisomatic region was measured and normalized to the background P2RY12 intensity obtained from regions of the same image free of neuronal cell bodies or obvious artifacts. A similar annulus-based analysis was performed using TMEM119 immunostaining as an independent microglial marker. For intensity-based quantification, 7–10 motoneurons were analyzed per animal and summarized at the animal level. Quantification of ChAT+ motoneuron survival after axotomy At 28 days after unilateral facial nerve axotomy, surviving ChAT+ motoneurons were quantified in the facial motor nucleus on the axotomized (Ax) and contralateral (CL) sides. For each mouse, surviving cell density (cells/mm²) was calculated for the Ax and CL sides, yielding one summary value per side per animal (Iba1+/+ and Iba1−/−; n = 8 mice per genotype). Quantification of γH2AX-positive cells γH2AX immunofluorescence was evaluated at 7 and 28 days after axotomy. A neuron was scored as γH2AX-positive if its DAPI-defined nucleus contained one or more discrete punctate γH2AX foci that were clearly distinguishable from nucleolar signal and background. Nuclei showing diffuse pan-nuclear staining without discrete foci were not counted as positive. Only ChAT+ neurons with a visible nucleus were included in the analysis. ChAT+ facial motoneurons were identified by soma morphology within the facial motor nucleus. For each genotype and time point, neurons were sampled from both the Ax and CL sides. For γH2AX analysis, the outcome measure was the proportion of ChAT+ neurons containing γH2AX foci per animal. Transmission electron microscopy Seven days after facial nerve axotomy, mice were deeply anesthetized and transcardially perfused with phosphate buffer at a volume equal to their body weight to remove blood, followed by fixation with 2.5% glutaraldehyde and 2% PFA in phosphate buffer at a volume three times their body weight. Brains were post-fixed overnight in the same fixative at 4℃. The following day, the brains were coronally sectioned using a vibratome (VT1200S, Leica Microsystems, Germany) to identify the facial motor nucleus. Once the region was confirmed, sagittal sections encompassing the nucleus were excised to prepare tissue blocks for electron microscopy. Tissue blocks were processed for standard transmission electron microscopy, including post-fixation in osmium tetroxide, dehydration through a graded ethanol series, and embedding in epoxy resin. Ultrathin sections were cut using an ultramicrotome (EM UC7, Leica Microsystems, Germany) equipped with a diamond knife, mounted on copper grids, stained with uranyl acetate and lead citrate, and examined under a transmission electron microscope (JEM-1400Plus, JEOL, Japan). Facial motoneurons were selected based on the presence of a nucleus and a long-axis diameter ≥ 20 μm. For each genotype (Iba1+/+ and Iba1−/−) and side (axotomized [Ax] and contralateral [CL]), 20 neurons were evaluated in two mice (n = 2 per genotype). The presence or absence of microglial contact and the number of synaptic structures were assessed for each neuron. Synapses were defined as structures with clearly visible synaptic membranes, clustered synaptic vesicles, and a prominent postsynaptic density. The number of synapses was analyzed using the Kruskal–Wallis test, whereas the presence or absence of microglial contact was analyzed using the chi-square test. Single-nucleus RNA sequencing (snRNA-seq) and data processing snRNA-seq was performed using the 10× Genomics Chromium system with the Flex Gene Expression workflow to analyze gene expression in brainstem tissue from Iba1+/+ and Iba1−/− mice after facial nerve axotomy or sham operation. The brainstem tissue was rapidly dissected without perfusion, immediately frozen in liquid nitrogen, and stored in liquid nitrogen until further processing. Frozen tissues were fixed using a formaldehyde-containing fixation buffer and dissociated into nuclei using a Liberase enzyme mixture and gentleMACS Octo Dissociator (Miltenyi Biotec), following the 10× Genomics Flex Gene Expression protocol. Approximately 2 × 107 nuclei were obtained from 100 ng of tissue, and 1 × 106 nuclei per sample were subjected to probe hybridization. Subsequent steps—including sample pooling, generation of single-nucleus emulsions, and sequencing library preparation—were performed using the Chromium Fixed RNA Kit for Mouse Transcriptome (10x Genomics, #1000496). The resulting sequencing libraries were quality-controlled using a Qubit dsDNA Assay Kit (Thermo Fisher Scientific) and a TapeStation D1000 System (Agilent Technologies), and sequencing was performed on an Illumina NovaSeq 6000 platform. Initial data processing, including demultiplexing, alignment, barcode counting, and clustering, was performed using the Cell Ranger pipeline (10x Genomics, version 7.0.1). Volcano plot analysis Differential gene expression analysis was performed using Loupe Browser 8 (10x Genomics), based on the gene expression matrices generated by the Cell Ranger pipeline (v7.0.1). Statistical significance was determined by the built-in differential expression testing. Genes with an adjusted p-value (padj < 0.05) and absolute log2 fold-change (|log2FC| ≥ 1) were classified as DEGs. A volcano plot was generated using ggplot2 and ggrepel in R (v4.2.0) to visualize the distribution of DEGs. Genes with significant upregulation (log2FC > 0, padj < 0.05) were colored red, whereas significantly downregulated genes (log2FC < 0, padj < 0.05) were colored blue. Genes that did not meet these thresholds were represented in gray. The top 30 genes with the lowest p-values were labeled in the plot. GO enrichment analysis To investigate the functional relevance of the DEGs, GO enrichment analysis was conducted using clusterProfiler (v4.0). Upregulated and downregulated genes were analyzed separately, and gene symbols were converted to Entrez IDs using org.Mm.eg.db (v3.13.0). GO enrichment was performed for the following three categories:Biological processMolecular functionCellular component Biological process Biological process Molecular function Molecular function Cellular component Cellular component Significant GO terms were identified using the Benjamini-Hochberg correction (q-value < 0.05). GO enrichment visualization GO enrichment results were visualized using dot plots, where the GeneRatio (ratio of DEGs associated with a given GO term) was plotted against the GO term descriptions. Dot size represented the number of genes per term, and color indicated the log-transformed adjusted p-value (-log10(p.adjust)). Upregulated and downregulated genes were plotted separately. All analyzes were performed using R (v4.2.0), and figures were saved in TIFF format at a resolution of 300 dpi for publication-quality output. Behavioral assessments Behavioral phenotyping was performed in a single longitudinal cohort of Iba1−/− and wild-type mice that entered the study at 8 weeks of age and was followed across the behavioral battery. Thus, although the full battery extended from 2 to 9 months of age, genotype comparisons within each individual assay were made between age-matched mice tested contemporaneously. Animals were sequentially assessed using a series of behavioral paradigms, including general health and neurological screening, light/dark transition, open field, elevated plus maze, hot plate, social interaction, rotarod, three-chamber social approach, acoustic startle response and prepulse inhibition, Porsolt forced swim, T-maze spontaneous alternation, Barnes maze, tail suspension, contextual and cued fear conditioning, and home-cage social interaction tests. All apparatuses were thoroughly cleaned with 70% ethanol and hypochlorous acid solution to minimize potential olfactory confounders. The experiments were conducted during the light phase (09:00–17:00). Detailed experimental procedures are provided in the Supplementary Methods. Statistical analysis Statistical analyzes were performed using Prism (version 9.5.1; GraphPad) and R (version 4.4.1). All tests were two-tailed with a significance threshold of α = 0.05 unless otherwise stated. Data are presented as mean ± SD or mean ± SEM as indicated in the corresponding figure legends. The experimental unit was defined a priori for each assay (typically one animal); when multiple measurements were obtained from the same animal (e.g., multiple cells or fields of view), values were summarized within each animal before group-level inference unless explicitly stated otherwise. For comparisons between two independent groups, unpaired two-tailed Student’s t-test was used for approximately normally distributed data; otherwise, non-parametric Mann–Whitney U tests were applied. For paired within-animal comparisons (e.g., axotomized vs. contralateral sides), paired t-tests or Wilcoxon signed-rank tests were used as appropriate. For comparisons among ≥ 3 groups, one-way ANOVA (parametric) or Kruskal–Wallis tests (non-parametric) were used, followed by multiple-comparisons procedures as specified in the figure legends (e.g., Šídák’s or Dunn’s tests). For planned families of simple comparisons, p-values were adjusted using the Holm method where indicated. No exclusion criteria were applied other than predefined technical failures, such as failed axotomy, tissue processing failure, or unusable imaging due to artifacts. No animals or data points were excluded unless explicitly stated; the exact n for each analysis is reported in the figure legends. For repeated-measures outcomes, genotype was treated as a between-subject factor and time/trial/block as a within-subject factor; post-hoc simple main effects were examined when interactions were detected. For imaging-based quantifications, the animal was considered the independent biological replicate; multiple cells/fields per animal were treated as technical replicates and summarized to a single per-animal value for statistical testing unless otherwise stated. Sample size No formal a priori power calculation was performed. Sample sizes were based on previous studies using this model and feasibility constraints. Exact n for each experiment and analysis is reported in the corresponding Methods subsection and figure legends. Blinding Investigators were not blinded to genotype during allocation, conduct of experiments, outcome assessment, or data analysis. Randomization No randomization was used for group allocation because genotype determined group membership. To minimize confounding, experiments were performed with interleaved genotypes when feasible, and sample processing and imaging were balanced across batches. Outcome measures Outcome measures included microglial perisomatic coverage, perisomatic synapse number by EM, motoneuron integrity/survival after axotomy, and transcriptional responses by snRNA-seq. The primary outcome measure for hypothesis testing was motoneuron survival at 28 days post-axotomy. Data deposition and materials sharing Data availability The authors confirm that the data supporting the findings of this study are available within the article and its Supplementary material. Additional datasets generated and analyzed during the current study are available from the corresponding author(s) upon reasonable request. Code availability The custom code used to reproduce the results shown in Fig. 5 and Supplementary Figures S11–S16 is available in a public repository: https://github.com/sekikoji54-crypto/single-nucleus-RNA-seq-code. Results Generation and characterization of Iba1-deficient mice To investigate Iba1 function, we generated Iba1 knockout (Iba1−/−) mice using CRISPR/Cas9-mediated genome editing. Exons 3–8 of the Iba1 gene were deleted to disrupt the gene function (Fig. 1A). Representative genotyping results confirmed successful generation of Iba1−/− mice (Fig. S2a). Two deletions—1521 bp and 9 bp—were further validated by Sanger sequencing (Fig. S2b). To confirm the absence of Iba1 protein at the tissue level, we performed immunohistochemistry on cerebral cortex sections from 8-week-old mice. Compared with Iba1+/+ mice, Iba1−/− mice showed complete Iba1 imoreactivity loss, whereas P2RY12 staining remained detectable, indicating preserved microglial identity despite Iba1 deletion (Fig. 1B). Quantitative analysis further showed that neither cortical P2RY12 fluorescence intensity nor P2RY12-positive area fraction differed significantly between Iba1+/+ and Iba1−/− mice at 8 weeks of age (Fig. S3a, b). Sholl analysis of hippocampal microglial ramification showed no significant differences between Iba1+/+ and Iba1−/− mice at 8 weeks (Fig. 1D). However, in 20-month-old mice, Iba1+/+ microglia exhibited a significantly higher number of Sholl intersections at the 20.0-µm radius compared with Iba1−/− microglia (p = 0.0069; Fig. 1E). To assess Iba1 influence on synaptic formation, frozen brain sections of Iba1−/− and Iba1+/+ mice were fluorescently stained for VGLUT2, a marker of synaptic vesicles, and the signal intensity in the hippocampal dentate gyrus was analyzed. No obvious signal intensity differences were identified between the Iba1+/+ and Iba1−/− mice (Fig. S4), indicating no major Iba1 deficiency effect on VGLUT2-positive synaptic formation in these mice. Because Iba1 is expressed during osteoclast differentiation and activation, comprehensive static and dynamic bone histomorphometry was performed. However, no genotype-dependent differences were observed between Iba1−/− and Iba1+/+ mice (Fig. S5). Behavioral and physical assessments of Iba1-/- mice To determine the neurological, cognitive, and affective consequences of global Iba1 loss, we conducted a comprehensive behavioral test battery. Iba1−/− mice exhibited a significant reduction in body weight compared with Iba1+/+ controls (p = 0.038), whereas no significant differences were observed in body temperature, forelimb grip strength, or performance in the hanging wire test (Fig. 2A and S6a-c). In addition, a battery of behavioral tests in Iba1−/− mice between 2 and 9 months of age revealed subtle behavioral alterations, including a trend toward impaired social recognition and memory in the three-chamber social novelty preference test (Fig. 2B), potential deficits in cognitive stress responses on day 2 of the forced swim test (Fig. 2c), and reduced spatial memory retention 1 month after training in the Barnes maze (Fig. 2D-F). In contrast, no genotype-dependent differences were observed between Iba1−/− and Iba1+/+ mice in motor coordination, locomotor activity, anxiety-like behavior, depressive-like behavior, or other social behaviors, as evaluated across the remaining behavioral and physical assessments (Fig. S6–9). These results suggest that Iba1 loss causes minor behavioral abnormalities in young mice. Fig. 2Body weight, social novelty preference, and spatial memory in Iba1-deficient mice. A Body weight of Iba1+/+ and Iba1−/− mice at 10 weeks. Bars show mean ± standard error of the mean (SEM), and dots indicate individual mice (n = 20 per genotype; unpaired two-tailed Student’s t-test). B Three-chamber social novelty preference test. Left, time spent in the chamber containing the already-investigated mouse, Stranger 1 (S1), versus the chamber containing the novel stranger, Stranger 2 (S2). Right, time spent around S1 versus S2. Iba1+/+ mice showed a significant preference for the novel conspecific S2, whereas Iba1−/− mice showed no significant preference for S2 (p = 0.0526). Data are expressed as mean ± SEM (n = 20 per genotype); novelty preference was assessed by paired two-tailed t-tests within each genotype, and genotypes were compared by unpaired two-tailed t-tests. C Porsolt forced swim test. Immobility and distance traveled were measured on day 1 and day 2. Data are expressed as mean ± SEM (n = 20 per genotype); two-way repeated-measures analysis of variance (ANOVA) with genotype and time as factors, followed by post-hoc tests. D–F Barnes maze test. D Errors, latency, distance, and omissions during acquisition trials. E, F Time spent around the target hole in probe tests on days 1 and 28 days after training. Reduced time in the target zone on day 28 in Iba1−/− mice indicates impaired long-term spatial memory. Data are expressed as mean ± SEM (n = 20 per genotype); acquisition was analyzed by two-way repeated-measures ANOVA, and probe tests were analyzed by paired two-tailed t-tests for comparisons between the target and other holes within each genotype and by unpaired two-tailed t-tests for comparisons between genotypes. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 Body weight, social novelty preference, and spatial memory in Iba1-deficient mice. A Body weight of Iba1+/+ and Iba1−/− mice at 10 weeks. Bars show mean ± standard error of the mean (SEM), and dots indicate individual mice (n = 20 per genotype; unpaired two-tailed Student’s t-test). B Three-chamber social novelty preference test. Left, time spent in the chamber containing the already-investigated mouse, Stranger 1 (S1), versus the chamber containing the novel stranger, Stranger 2 (S2). Right, time spent around S1 versus S2. Iba1+/+ mice showed a significant preference for the novel conspecific S2, whereas Iba1−/− mice showed no significant preference for S2 (p = 0.0526). Data are expressed as mean ± SEM (n = 20 per genotype); novelty preference was assessed by paired two-tailed t-tests within each genotype, and genotypes were compared by unpaired two-tailed t-tests. C Porsolt forced swim test. Immobility and distance traveled were measured on day 1 and day 2. Data are expressed as mean ± SEM (n = 20 per genotype); two-way repeated-measures analysis of variance (ANOVA) with genotype and time as factors, followed by post-hoc tests. D–F Barnes maze test. D Errors, latency, distance, and omissions during acquisition trials. E, F Time spent around the target hole in probe tests on days 1 and 28 days after training. Reduced time in the target zone on day 28 in Iba1−/− mice indicates impaired long-term spatial memory. Data are expressed as mean ± SEM (n = 20 per genotype); acquisition was analyzed by two-way repeated-measures ANOVA, and probe tests were analyzed by paired two-tailed t-tests for comparisons between the target and other holes within each genotype and by unpaired two-tailed t-tests for comparisons between genotypes. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 Microglial ensheathment of facial motoneurons is altered in knockout mice after axotomy To evaluate the role of Iba1 in microglia under pathological conditions, we used a well-established method for inducing microglial activation: facial nerve axotomy. Immunohistochemical analysis of the facial motor nuclei revealed distinct changes in response to axotomy (Fig. 3A). Low-magnification (40×) images showed that the axotomized side exhibited reduced intensity of ChAT immunoreactivity—a marker for motoneurons—compared with the contralateral side, reflecting a pathological response in motoneurons, consistent with a previous report. In contrast, the intensity of P2RY12, a microglial marker, increased on the axotomized side. High-magnification (400×) images further demonstrated that microglia were tightly associated with motoneuron somata on the axotomized side in both genotypes. However, in Iba1−/− mice, P2RY12 signals appeared more diffuse and less concentrated around the neuronal somata, suggesting defective microglial ensheathment of motoneuron cell bodies compared with those in Iba1+/+ mice. Fig. 3Immunohistochemical analysis of the pons in Iba1+/+ and Iba1−/− mice. A Representative immunostaining of the pons in Iba1+/+ and Iba1−/− mice at low (40×) and high (400×) magnifications. Sections were stained with DAPI (blue), ChAT (green), and P2RY12 (magenta). The right side of the images corresponds to the axotomized hemisphere. Scale bar: 500 μm (40×), 20 μm (400×). B Visual classification of microglial stripping around ChAT positive neuronal somata based on P2RY12 immunoreactivity. Representative images show ChAT (green), P2RY12 (magenta), and merged images including DAPI (blue). Category I indicates absent microglial contact around the soma. Category II shows punctate contacts. Categories III–IV indicate segmented contacts, stratified by the fraction of the soma circumference covered (III, < 1/3; IV, 1/3–2/3). Category V represents extensive microglial coverage. Scale bar: 10 μm. C Proportion of motoneurons in each microglial-coverage category defined in (b) (I–V). Data are shown as pooled cell counts across animals (n = 4 mice per genotype; total 142 cells for Iba1+/+ and 148 cells for Iba1−/−). D Bar graph showing the proportion of high-coverage motoneurons (Categories IV–V) in each animal. Dots represent individual animals (n = 4 mice per genotype); bars and error bars indicate mean ± s.d. Groups were compared using the two-tailed Mann–Whitney U test (p = 0.0286). E Bar graph showing the ratio of perisomatic P2RY12 fluorescence intensity (3-µm annulus around each ChAT+ soma) normalized to background P2RY12 intensity. For each animal, values were calculated as the mean of 7–10 motoneurons, and statistical testing was performed at the animal level (n = 4 mice per genotype). Dots represent individual animals; bars and error bars indicate mean ± s.d. Groups were compared using the two-tailed Mann–Whitney U test (p = 0.0286). F Bar graph of ChAT fluorescence intensity in neuronal somata in the contralateral (CL) and axotomized (Ax) facial motor nuclei for each genotype. Dots represent individual animals (n = 4 mice per condition); bars and error bars indicate mean ± s.d. Group comparisons were performed using the Kruskal–Wallis test followed by Dunn’s multiple-comparisons test (adjusted p values: Iba1+/+ CL vs. Iba1+/+ Ax, p = 0.1251; Iba1+/+ CL vs. Iba1−/− CL, p = 0.9391; Iba1+/+ Ax vs. Iba1−/− Ax, p > 0.9999; Iba1−/− CL vs. Iba1−/− Ax, p = 0.0374) Immunohistochemical analysis of the pons in Iba1+/+ and Iba1−/− mice. A Representative immunostaining of the pons in Iba1+/+ and Iba1−/− mice at low (40×) and high (400×) magnifications. Sections were stained with DAPI (blue), ChAT (green), and P2RY12 (magenta). The right side of the images corresponds to the axotomized hemisphere. Scale bar: 500 μm (40×), 20 μm (400×). B Visual classification of microglial stripping around ChAT positive neuronal somata based on P2RY12 immunoreactivity. Representative images show ChAT (green), P2RY12 (magenta), and merged images including DAPI (blue). Category I indicates absent microglial contact around the soma. Category II shows punctate contacts. Categories III–IV indicate segmented contacts, stratified by the fraction of the soma circumference covered (III, < 1/3; IV, 1/3–2/3). Category V represents extensive microglial coverage. Scale bar: 10 μm. C Proportion of motoneurons in each microglial-coverage category defined in (b) (I–V). Data are shown as pooled cell counts across animals (n = 4 mice per genotype; total 142 cells for Iba1+/+ and 148 cells for Iba1−/−). D Bar graph showing the proportion of high-coverage motoneurons (Categories IV–V) in each animal. Dots represent individual animals (n = 4 mice per genotype); bars and error bars indicate mean ± s.d. Groups were compared using the two-tailed Mann–Whitney U test (p = 0.0286). E Bar graph showing the ratio of perisomatic P2RY12 fluorescence intensity (3-µm annulus around each ChAT+ soma) normalized to background P2RY12 intensity. For each animal, values were calculated as the mean of 7–10 motoneurons, and statistical testing was performed at the animal level (n = 4 mice per genotype). Dots represent individual animals; bars and error bars indicate mean ± s.d. Groups were compared using the two-tailed Mann–Whitney U test (p = 0.0286). F Bar graph of ChAT fluorescence intensity in neuronal somata in the contralateral (CL) and axotomized (Ax) facial motor nuclei for each genotype. Dots represent individual animals (n = 4 mice per condition); bars and error bars indicate mean ± s.d. Group comparisons were performed using the Kruskal–Wallis test followed by Dunn’s multiple-comparisons test (adjusted p values: Iba1+/+ CL vs. Iba1+/+ Ax, p = 0.1251; Iba1+/+ CL vs. Iba1−/− CL, p = 0.9391; Iba1+/+ Ax vs. Iba1−/− Ax, p > 0.9999; Iba1−/− CL vs. Iba1−/− Ax, p = 0.0374) To qualitatively evaluate perisomatic microglial coverage, ChAT-positive (ChAT+) motoneurons were classified into five categories based on the extent of P2RY12-positive microglial contact (Fig. 3B). The distribution of motoneurons across these categories differed between Iba1+/+ and Iba1−/− mice (Fig. 3C-D), with Iba1+/+ mice displaying a greater proportion of motoneurons with extensive microglial coverage (two-tailed Mann–Whitney U test, p = 0.0286). Quantitative analysis supported these findings. Iba1+/+ mice exhibited significantly higher normalized P2RY12 intensity in this region than Iba1−/− mice, indicating denser microglial contact (Fig. 3E). Similar trends were observed when TMEM119 was used as an alternative microglial marker (Fig. S10). Neuronal integrity after axotomy To assess motoneuron integrity, we measured ChAT expression intensity—which is known to decrease in response to motoneuron injury [28, 29]—and compared the axotomized and contralateral sides 7 days after surgery. ChAT intensity tended to decrease on the axotomized side in Iba1+/+ and was significantly reduced in Iba1−/− mice compared the contralateral side (Fig. 3F). Iba1−/− mice exhibited a greater ChAT intensity reduction than Iba1+/+ mice (Fig. 3F), suggesting an exacerbated cholinergic phenotype loss in the absence of Iba1. Microglial contact and synapse number To confirm the immunohistochemical findings, we performed facial motoneuron electron microscopy. On the axotomized side of Iba1+/+ mice, microglia were in close contact with the motoneuron somata, extending pseudopodia and stripping synapses (Fig. 4A–B). Iba1−/− mice displayed infrequent direct microglia-motoneuron contact (Fig. 4C,DC). The number of remaining synapses on the facial motor neurons (Fig. 4E) showed that, while there was no difference between the Iba1+/+ and Iba1−/− groups on the contralateral side, the Iba1+/+ mice exhibited a significant reduction in synapse number on the axotomized side compared with the contralateral side (p = 0.0017). In contrast, no such reduction was evident in Iba1−/− mice. The proportion of facial motoneurons in contact with microglia also differed between the four groups (p = 0.0027; Fig. 4F). Taken together, Iba1-deficient microglia impaired synaptic stripping. Fig. 4Microglial contact and synapse number in facial motoneurons (MN) after axotomy. A, B Electron micrographs of facial MN from the axotomized side of Iba1+/+ mice. (a) Low-magnification image at 3000× showing an MN soma; yellow box indicates the area enlarged in (b). Scale bar: 5.0 μm. B Higher-magnification image at 5,000× showing microglia (MGL, blue) closely apposed to the MN (yellow), with presynaptic terminals (red) interposed between them. Scale bar: 500 nm. C, D Electron micrographs of facial MNs from the axotomized side of Iba1−/− mice. C Low-magnification image at 3000× of an MN soma; yellow box indicates the area enlarged in (d). Scale bar: 5.0 μm. D Higher-magnification image at 5,000× showing an MGL process (blue) near the MN (yellow) without direct contact. Scale bar: 1.0 μm. E Quantification of the number of synapses per facial MN cell body across four groups: Iba1+/+-contralateral (CL), Iba1+/+-axotomy (Ax), Iba1−/−-CL, and Iba1−/−-Ax. Data are presented as mean ± standard error of the mean. **p < 0.01, ***p < 0.001; ns, not significant (Kruskal–Wallis test followed by Dunn’s multiple-comparisons post hoc test). F Percentage of facial MNs with or without microglial contact in each group. Contact proportion was compared across groups using pairwise Fisher’s exact tests (two-sided) with Holm correction following a significant omnibus test (χ² (3) = 14.17, p = 0.0027). Significant difference was observed for Iba1+/+-Ax vs. Iba1−/−-CL (odds ratio = 10.27, 95% confidence interval 1.75–60.09, Holm-adjusted p = 0.0195); other comparisons were ns Microglial contact and synapse number in facial motoneurons (MN) after axotomy. A, B Electron micrographs of facial MN from the axotomized side of Iba1+/+ mice. (a) Low-magnification image at 3000× showing an MN soma; yellow box indicates the area enlarged in (b). Scale bar: 5.0 μm. B Higher-magnification image at 5,000× showing microglia (MGL, blue) closely apposed to the MN (yellow), with presynaptic terminals (red) interposed between them. Scale bar: 500 nm. C, D Electron micrographs of facial MNs from the axotomized side of Iba1−/− mice. C Low-magnification image at 3000× of an MN soma; yellow box indicates the area enlarged in (d). Scale bar: 5.0 μm. D Higher-magnification image at 5,000× showing an MGL process (blue) near the MN (yellow) without direct contact. Scale bar: 1.0 μm. E Quantification of the number of synapses per facial MN cell body across four groups: Iba1+/+-contralateral (CL), Iba1+/+-axotomy (Ax), Iba1−/−-CL, and Iba1−/−-Ax. Data are presented as mean ± standard error of the mean. **p < 0.01, ***p < 0.001; ns, not significant (Kruskal–Wallis test followed by Dunn’s multiple-comparisons post hoc test). F Percentage of facial MNs with or without microglial contact in each group. Contact proportion was compared across groups using pairwise Fisher’s exact tests (two-sided) with Holm correction following a significant omnibus test (χ² (3) = 14.17, p = 0.0027). Significant difference was observed for Iba1+/+-Ax vs. Iba1−/−-CL (odds ratio = 10.27, 95% confidence interval 1.75–60.09, Holm-adjusted p = 0.0195); other comparisons were ns Iba1 loss skews axotomy-evoked microglial states toward interferon-responsive programs To investigate the transcriptional response of motoneurons to axonal injury in Iba1-deficient conditions, we performed single-nucleus RNA sequencing (snRNA-seq) of facial motor nuclei from Iba1+/+ and Iba1−/− mice subjected to either a sham operation or facial nerve axotomy (n = 3 per group; 12 mice in total; Fig. 5A). Dimensionality reduction using t-SNE revealed distinct nucleus clustering (Fig. 5B). The cell types were annotated based on unsupervised clustering and marker gene expression profiles (Fig. S11b). Fig. 5Single-nucleus RNA sequencing reveals genotype-dependent transcriptional changes in motoneurons after axotomy. A Schematic overview of the experimental workflow. Single-nucleus RNA sequencing was performed on facial motor nuclei from Iba1+/+ and Iba1−/− mice with or without facial nerve axotomy (n = 3 per group). B t-SNE plot showing nucleus clusters obtained from all samples. Cell types were annotated based on unsupervised clustering and marker gene expression (Fig. S11 shows marker gene expression profiles). C Volcano plot showing differentially expressed genes in ChAT-positive neurons between Iba1−/−axotomy and Iba1+/+-axotomy groups. Differentially expressed genes in Iba1−/− and Iba1+/+ mice under sham conditions were excluded to isolate axotomy-specific effects. Points are colored by direction of change (upregulated in Iba1−/−-axotomy, red; downregulated, blue; others, gray), and gene labels are limited to the top 30 by −log10(p). D Gene ontology (GO) enrichment analysis of genes that were upregulated in Iba1−/−-axotomy vs. Iba1+/+-axotomy. Enriched terms are categorized into biological process (BP), molecular function (MF), and cellular component (CC). Dot color represents the adjusted p-value (− log10 scale), and size indicates the number of associated genes. Only the top 10 non-redundant terms per category (q < 0.05) are shown. E GO enrichment analysis of genes downregulated in Iba1−/−-axotomy vs. Iba1+/+-axotomy. Enriched terms are categorized as indicated in (d). Functional annotations point to impaired synaptic structure and neurotransmission-related pathways Single-nucleus RNA sequencing reveals genotype-dependent transcriptional changes in motoneurons after axotomy. A Schematic overview of the experimental workflow. Single-nucleus RNA sequencing was performed on facial motor nuclei from Iba1+/+ and Iba1−/− mice with or without facial nerve axotomy (n = 3 per group). B t-SNE plot showing nucleus clusters obtained from all samples. Cell types were annotated based on unsupervised clustering and marker gene expression (Fig. S11 shows marker gene expression profiles). C Volcano plot showing differentially expressed genes in ChAT-positive neurons between Iba1−/−axotomy and Iba1+/+-axotomy groups. Differentially expressed genes in Iba1−/− and Iba1+/+ mice under sham conditions were excluded to isolate axotomy-specific effects. Points are colored by direction of change (upregulated in Iba1−/−-axotomy, red; downregulated, blue; others, gray), and gene labels are limited to the top 30 by −log10(p). D Gene ontology (GO) enrichment analysis of genes that were upregulated in Iba1−/−-axotomy vs. Iba1+/+-axotomy. Enriched terms are categorized into biological process (BP), molecular function (MF), and cellular component (CC). Dot color represents the adjusted p-value (− log10 scale), and size indicates the number of associated genes. Only the top 10 non-redundant terms per category (q < 0.05) are shown. E GO enrichment analysis of genes downregulated in Iba1−/−-axotomy vs. Iba1+/+-axotomy. Enriched terms are categorized as indicated in (d). Functional annotations point to impaired synaptic structure and neurotransmission-related pathways At the whole-microglia level, we did not observe a pervasive shift in DEG burden across Iba1+/+-sham, Iba1+/+-axotomy, Iba1−/− -sham, and Iba1−/− -axotomy groups. The effect sizes were generally modest, and cytokine/chemokine/growth-factor transcripts were broadly comparable among the groups (Fig. S12a-e). The absence of broader transcriptional changes directly attributable to Iba1 is plausible, given that Iba1 operates at the lowest tier of the signaling cascade and is not expected to directly drive transcriptional regulation. The expression of the Iba1 homolog Aif1l [30], which can exert compensatory effects, remained unchanged (Fig. S12e). We re-clustered the dataset after extracting microglia based on annotation markers (Fig. S13a–h and S14). This yielded four clusters (C0–C3), with C0 comprising the majority of all groups (Fig. S13i-j). C1 cluster proportion increased from sham to axotomy and expanded more prominently in Iba1−/− mice than in Iba1+/+ mice (Fig. S13j). To contextualize this cluster, we cross-referenced published transcriptional programs. Guided by Tay et al. [31], the C1 cluster corresponds to microglial populations known to expand after axotomy (Fig. S13k). Within AD-related disease-associated microglia (DAM) frameworks, the C1 cluster mapped predominantly to an interferon-responsive microglia (IRM) program with partial DAM features (Fig. S13l–n). Trajectory analysis consistently supported the transition from homeostatic microglia (C0) to the IRM-like C1 cluster along lineage 1 (Fig. S13o-p) . Although C2 and C3 clusters contained fewer cells, GO enrichment suggested distinct roles. The C2 cluster was enriched for antigen processing/presentation and interferon pathways, whereas the C3 cluster was enriched for phagocytic, lysosomal, and lipid-metabolic programs, consistent with a repair- or clearance-oriented DAM-like state (Fig. S14). Therefore, Iba1−/− mice exhibited an exaggerated inflammatory bias (IRM-like C1 cluster expansion), potentially at the expense of progression toward repair or clearance programs. Iba1 knockout alters gene expression in ChAT+ motoneurons after axotomy To assess axotomized motoneuron dynamics under microglial control, we extracted ChAT+ nuclei from each group and conducted DEG analysis. A volcano plot comparing Iba1−/− and Iba1+/+ mice under axotomy conditions revealed significant transcriptional differences (Fig. 5C). GO analysis of genes upregulated in Iba1−/− motoneurons (Fig. 5D) revealed enrichment in biological processes, such as cognition, learning or memory, and hormone metabolic regulation, suggesting enhanced neuromodulatory and stress-responsive pathways. Molecular function terms included neurotransmitter receptor activity, G protein-coupled receptor binding, and active transmembrane transporter activity, indicating possible alterations in synaptic signaling and membrane transport. Cellular component terms such as membrane rafts, neuronal cell bodies, and dopaminergic synapses suggest the reorganization of key functional compartments. In contrast, genes downregulated in Iba1−/− motoneurons (Fig. 5E) were enriched for biological processes, including synapse organization, postsynaptic membrane assembly, and GABAergic synapse regulation, indicating impairments in synaptic maintenance. Molecular function terms such as glutamate receptor activity, potassium ion channel activity, and GABA receptor activity indicate ionotropic signaling pathway suppression. The cellular component category included terms such as postsynaptic membrane, dendrite, axon terminus, and synaptic vesicle, suggesting a widespread transcriptional downregulation of genes critical for synaptic integrity and neurotransmission. Collectively, these findings demonstrate that defective microglial ensheathment due to Iba1 deficiency disturbs injury-induced transcriptional programs in motoneurons, promoting aberrant neuromodulatory signaling while impairing structural and functional components essential for synaptic connectivity and neuronal communication. Iba1 loss diverts microglial signaling from adhesion to inflammation To probe the microglia-motoneuron crosstalk, we applied NicheNet using microglia as sender cells and ChAT+ nuclei as receivers (Fig. S15). A receptor dot plot confirmed that ChAT+ neurons expressed the expected entry points—integrins (Itgb1, Itga3/6/7, Itgb4/5, and Itgav) and Aplp2—supporting their capacity to decode microglial cues (Fig. S15b-c). In Iba1+/+ mice (axotomy vs. sham), the top-ranked microglial ligands upregulated after injury converged on ECM/adhesion programs—Tgfbi/Mmp2/Anxa2—predicted to signal through the integrin set and Aplp2. A prominent MHC-I (H2-D1/H2-K1/H2-Q) → Aplp2 axis was also recovered (Fig. S15d, f), indicating adhesion-oriented remodeling of the injured synaptic milieu. However, in Iba1−/− mice (axotomy vs. sham), the ligand spectrum shifted toward immune/inflammatory and synaptic-remodeling cues, with Adam17, Ebi3, B2m, Cd72, and Agrn among the top signals (Fig. S15e, g). While the MHC-I → Aplp2 route remained strong, the overall routing of microglial outputs diverted from ECM/adhesion (Iba1+/+) to immune/inflammatory (Iba1−/−) cues. These communication patterns were consistent with IRM-like C1 cluster expansion in Iba1−/− mice and aligned with the axotomy-induced transcriptional shifts observed in ChAT+ neurons (enhanced neuromodulatory/stress programs with concomitant synaptic organization suppression). Therefore, NicheNet supports a model in which Iba1 deficiency biases microglia toward inflammatory signaling at the expense of adhesion-guided repair, providing a mechanistic link between the microglial state and maladaptive motoneuron gene expression. Alterations in inhibitory and excitatory synaptic markers after axotomy Given that GO analysis indicated differences in inhibitory synaptic pathways between Iba1+/+ and Iba1−/− mice, we examined GABRA1, the α1 GABA(A) receptor subunit, as a representative inhibitory synaptic marker (Fig. S16). In Iba1+/+ axotomy, GABRA1 immunoreactivity became punctate, forming discrete puncta. In Iba1−/− motoneurons, such puncta are evident even without injury. Axotomy decreased total GABRA1 fluorescence intensity (membrane + cytoplasmic) in motoneurons of Iba1+/+ and Iba1−/− mice (p < 0.0001; Fig. S17a and S18a). Regarding subcellular distribution, Iba1+/+ mice showed no detectable redistribution between perimembrane and cytoplasmic pools, whereas Iba1−/− mice exhibited a significant shift toward the plasma membrane after axotomy (p < 0.0001; Fig. S17a and S18b-c), consistent with altered receptor localization under Iba1-deficient conditions. To assess excitatory synaptic inputs, we analyzed VGLUT1 expression, a presynaptic glutamatergic terminal marker. In Iba1+/+ mice, VGLUT1 expression surrounding motoneurons was not significantly affected by axotomy. However, Iba1−/− mice exhibited a marked perisomatic VGLUT1 signal reduction on the axotomized side (Fig. S17b and S18d-e), suggesting a loss or reorganization of excitatory synaptic contacts in the absence of Iba1. Prompted by our snRNA-seq finding that Ret was upregulated in Iba1−/− microglia after axotomy compared with Iba1+/+ microglia, we assessed whether injury-responsive RET signaling was also altered at the motoneuron level (Fig. S16). Next, we evaluated the expression of RET, a receptor tyrosine kinase in the glial cell line-derived neurotrophic factor pathway implicated in neuronal survival after motor nerve injury [32, 33]. In Iba1+/+ mice, RET expression increased in motoneurons on the axotomized side relative to those on the contralateral side, indicating a compensatory response to injury. Conversely, Iba1−/− mice showed elevated baseline RET expression even on the contralateral side, with a further increase ipsilaterally after axotomy, but the upregulation was lower on the axotomized side (Fig. S17c and S18f-g). Therefore, RET expression regulation is altered under Iba1-deficient conditions, potentially reflecting differences in injury-induced neuroprotective signaling. Iba1 deficiency reduces motoneuron survival after axotomy To evaluate neuroprotection by Iba1-dependent microglial ensheathment after axotomy, we assessed the number of ChAT+ survivor motoneurons in brainstem sections. 28 days post-injury (Fig. 6A). The axotomized and contralateral sides were compared in both groups (Fig. 6B). Iba1−/− mice showed fewer ChAT+ cells on the axotomized side compared with the contralateral side. Moreover, the ratio of cell numbers (axotomized side/average of contralateral side) was significantly lower in Iba1−/− than in Iba1+/+ mice (p = 0.013; Fig. 6C), indicating that Iba1 deficiency-induced defective microglial ensheathment decreased motoneuron survival after axotomy. Fig. 6Reduced motoneuron survival and nuclear damage in Iba1-/- mice following facial nerve axotomy. a Representative 40× immunofluorescence images of ChAT-stained facial motor nuclei at 28 days after axotomy in Iba1+/+ (upper) and Iba1-/- (lower) mice. The left and right sides of each image correspond to the contralateral (CL) and axotomized (Ax) hemispheres, respectively. Scale bar: 100 μm. b Quantification of ChAT-positive motoneuron density (cells/mm²) on the Ax and CL sides (Iba1+/+ and Iba1-/-; n = 8 mice per genotype). Each dot represents one mouse (one summary value per side); bars indicate mean ± standard deviation (SD). Because normality was uncertain, non-parametric tests were pre-specified. Within-genotype paired comparisons (Ax vs. CL) used the Wilcoxon signed-rank test. Between-genotype comparisons within each side (Iba1+/+ vs. Iba1-/- at Ax; Iba1+/+ vs. Iba1-/- at CL) used the Mann–Whitney U test. A genotype × side interaction was assessed on difference scores by comparing (Ax−CL) between genotypes with a Mann–Whitney U test. P-values were Holm-adjusted across the four planned comparisons (two-tailed α = 0.05). Effect sizes are reported as rank-biserial r (paired) and Cliff’s δ (unpaired). Asterisks denote adjusted significance (* p < 0.05; ** p < 0.01). c Axotomy impact is expressed as the per-animal ratio of axotomized-to-contralateral cell density (Ax/CL, unitless). Each dot represents one mouse; bars indicate mean ± SD. Ratios < 1 indicate fewer ChAT-positive neurons per unit area on the Ax side. Group comparisons of Ax/CL ratios (Iba1+/+ vs. Iba1-/-) were performed using the Mann–Whitney U test (two-tailed; Holm-adjusted where applicable). Asterisks denote adjusted significance (* p < 0.05). (d) Quantification of ChAT immunofluorescence intensity in facial motoneurons, calculated as the ratio of the mean fluorescence on the Ax side to that on the CL side (Ax/CL) within each animal. e Representative immunofluorescence images of facial motoneurons co-stained for ChAT (green), DAPI (cyan), and γH2AX (magenta) on the Ax side in Iba1+/+ and Iba1-/- mice. Yellow arrowheads (triangles) indicate cells identified as γH2AX-positive motoneurons (ChAT⁺ neurons with condensed nuclear γH2AX foci). Scale bar: 20 μm. f Proportion of ChAT⁺ facial motoneurons with nuclear γH2AX foci at 7 days post-axotomy on the CL and Ax sides in Iba1+/+ and Iba1-/- mice (Iba1+/+: n = 5; Iba1-/-: n = 5). Each dot represents one animal; bars indicate mean ± SD. Statistical analysis was performed using the Kruskal–Wallis test followed by Dunn’s post-hoc comparisons (genotype × side). g Proportion of ChAT⁺ facial motoneurons with nuclear γH2AX foci at 28 days post-axotomy on the contralateral (CL) and axotomized (Ax) sides in Iba1+/+ and Iba1-/- mice (Iba1+/+: n = 5; Iba1-/-: n = 6). Each dot represents one animal; bars indicate mean ± SD. Statistical analysis was performed via Kruskal–Wallis test followed by Dunn’s post-hoc comparisons (genotype × side) Reduced motoneuron survival and nuclear damage in Iba1-/- mice following facial nerve axotomy. a Representative 40× immunofluorescence images of ChAT-stained facial motor nuclei at 28 days after axotomy in Iba1+/+ (upper) and Iba1-/- (lower) mice. The left and right sides of each image correspond to the contralateral (CL) and axotomized (Ax) hemispheres, respectively. Scale bar: 100 μm. b Quantification of ChAT-positive motoneuron density (cells/mm²) on the Ax and CL sides (Iba1+/+ and Iba1-/-; n = 8 mice per genotype). Each dot represents one mouse (one summary value per side); bars indicate mean ± standard deviation (SD). Because normality was uncertain, non-parametric tests were pre-specified. Within-genotype paired comparisons (Ax vs. CL) used the Wilcoxon signed-rank test. Between-genotype comparisons within each side (Iba1+/+ vs. Iba1-/- at Ax; Iba1+/+ vs. Iba1-/- at CL) used the Mann–Whitney U test. A genotype × side interaction was assessed on difference scores by comparing (Ax−CL) between genotypes with a Mann–Whitney U test. P-values were Holm-adjusted across the four planned comparisons (two-tailed α = 0.05). Effect sizes are reported as rank-biserial r (paired) and Cliff’s δ (unpaired). Asterisks denote adjusted significance (* p < 0.05; ** p < 0.01). c Axotomy impact is expressed as the per-animal ratio of axotomized-to-contralateral cell density (Ax/CL, unitless). Each dot represents one mouse; bars indicate mean ± SD. Ratios < 1 indicate fewer ChAT-positive neurons per unit area on the Ax side. Group comparisons of Ax/CL ratios (Iba1+/+ vs. Iba1-/-) were performed using the Mann–Whitney U test (two-tailed; Holm-adjusted where applicable). Asterisks denote adjusted significance (* p < 0.05). (d) Quantification of ChAT immunofluorescence intensity in facial motoneurons, calculated as the ratio of the mean fluorescence on the Ax side to that on the CL side (Ax/CL) within each animal. e Representative immunofluorescence images of facial motoneurons co-stained for ChAT (green), DAPI (cyan), and γH2AX (magenta) on the Ax side in Iba1+/+ and Iba1-/- mice. Yellow arrowheads (triangles) indicate cells identified as γH2AX-positive motoneurons (ChAT⁺ neurons with condensed nuclear γH2AX foci). Scale bar: 20 μm. f Proportion of ChAT⁺ facial motoneurons with nuclear γH2AX foci at 7 days post-axotomy on the CL and Ax sides in Iba1+/+ and Iba1-/- mice (Iba1+/+: n = 5; Iba1-/-: n = 5). Each dot represents one animal; bars indicate mean ± SD. Statistical analysis was performed using the Kruskal–Wallis test followed by Dunn’s post-hoc comparisons (genotype × side). g Proportion of ChAT⁺ facial motoneurons with nuclear γH2AX foci at 28 days post-axotomy on the contralateral (CL) and axotomized (Ax) sides in Iba1+/+ and Iba1-/- mice (Iba1+/+: n = 5; Iba1-/-: n = 6). Each dot represents one animal; bars indicate mean ± SD. Statistical analysis was performed via Kruskal–Wallis test followed by Dunn’s post-hoc comparisons (genotype × side) Next, we compared axotomized-side ChAT fluorescence intensity in facial motoneurons, normalized to mean contralateral-side levels. In Iba1+/+ mice, the normalized ChAT intensity was 0.94 (95% confidence interval [CI] 0.90–0.98), indicating ChAT expression recovery. However, in Iba1−/− mice, it was 0.88 (95% CI 0.86–0.91), which was significantly lower than that in Iba1+/+ mice (p = 0.021; Fig. 6D). Furthermore, axotomy induced γH2AX accumulation and the formation of nuclear foci in facial motoneurons, which were separated from the nucleolus (Fig. 6E). γH2AX foci represent chromatin domains enriched in phosphorylated histone H2AX at sites of DNA double-strand breaks, a process closely linked to cell death. At day 7 post-axotomy, the proportion of γH2AX-positive neurons was comparable between Iba1+/+ and Iba1−/− mice (Fig. 6F); however, by day 28, Iba1−/− mice exhibited a higher proportion. The higher prevalence of these foci in Iba1−/− mice suggests that motoneurons lacking microglial support have increased long-term vulnerability. Therefore, microglial motoneuron ensheathment enhances neuronal protection after axotomy. On day 28 post-axotomy, GFAP-assessed astrocytic reactivity showed no genotype-dependent differences between Iba1+/+ and Iba1−/− mice (Fig. S19). Discussion The present study involved a systematic analysis of Iba1−/− mice to reveal the role of the EF-hand calcium-binding protein Iba1. We found that Iba1-mediated ensheathment and synaptic stripping are critical for axotomized motoneuron survival. By targeting the downstream signaling pathways involved in actin bundling and membrane ruffling during microglial signaling, we uncovered a neuronal-protective role of activated microglia amid their diverse and complex functions. Our findings should be interpreted in the context of a long-standing literature showing that peripheral motor axotomy elicits a coordinated perineuronal glial reaction, characterized by early astrocytic activation and microglial proliferation, migration, and close apposition to injured motoneurons. This general framework has been well described in facial, hypoglossal, and spinal motoneuron systems, but the precise functional significance of these structural interactions remained unsettled [8, 9]. Using Sholl analysis in adult mice, Lituma et al. revealed that Iba1−/− microglia exhibit decreased branching complexity compared with wild-type microglia, which is consistent with our findings [24]. In juvenile mice (P16–P19), Iba1−/− microglia show decreased branching, but this phenotype is reversed during maturation, and microglial morphology in adult Iba1−/− mice at 3 months is more complex than that in wild-type microglia. In our study, we performed Sholl analysis in genome-edited knockout mice at 8 weeks and 20 months of age and observed decreased branching only at 20 months, indicating an inconclusive Iba1 effect on homeostatic microglia morphology. Lituma et al. also reported significant behavioral changes in adult Iba1−/− mice aged 3–5 months [24], including increased locomotor activity, impaired object-recognition memory, and reduced sociability, whereas anxiety-like behavior appeared largely normal in standard assays. In our systematic behavioral assessment of 2–9-month-old mice, Iba1−/− mice displayed behavioral abnormalities in addition to weight loss. Specifically, we observed impaired social recognition and memory in the three-chamber social novelty preference test, potential cognitive deficits under stressful conditions in the second-day forced swim test, when mice are re-exposed to the inescapable water and must recall and adapt to the previous stressful experience, and reduced spatial memory retention in the Barnes maze 1 month after training. Consistent with Lituma et al., both studies therefore support abnormalities in hippocampus-dependent spatial memory and social-recognition processes, while anxiety-like behavior is preserved. Notably, despite employing a broader panel of tasks, we did not detect genotype-dependent differences in locomotor activity, motor coordination, depressive-like behavior, or other social behaviors (Fig. S6–9). Prior studies using Csf1r∆FIRE/∆FIRE mouse pups with selective microglia depletion reported no significant changes in hippocampal synapse number, synaptic maturation, and motor skills [34]. Although these phenotypes vary by mouse model and age, our findings suggest that microglia play limited, context-dependent roles in brain development under physiological conditions. However, microglia are more likely to exert a critical influence under pathological conditions. Indeed, peripheral axotomy models have historically served as tractable systems for studying neuron–glia interactions as the injury provokes robust central responses without direct mechanical damage to the brain parenchyma or overt disruption of the blood–brain barrier [17, 35]. For instance, microglia interact closely with axotomized neurons and accumulate at injury sites alongside reactive astrocytes to promote sustained inflammatory responses in traumatic brain injury (TBI) models [36]. To date, studies of microglial synapse pruning have relied mainly on knockdown models of TREM2 [37], complement C1q [38], MER tyrosine kinase [39, 40], and fractalkine receptor (CX3CR1) [41, 42]. These models demonstrate increased spinal density and reduced microglial engulfment. Although these molecules are involved in apoptotic cell phagocytosis [43–46], they also exert multifaceted effects on microglial physiology, transcriptomic profiles, and inflammatory responses [47–52]. Therefore, target gene deletion alone is inappropriate for dissecting specific pruning mechanisms [53]. By targeting Iba1, located downstream in the microglial signaling cascade, we affected synapse stripping without marked microglial transcriptomic alterations. Our data revealed the role of Iba1 in direct microglia-motoneuron interactions following axotomy. Although we observed substantial transcriptional changes in ChAT+ axotomized motoneurons under Iba1-deficient conditions, Iba1-deficient microglia showed no major transcriptional changes. The limited changes in mRNA expression in activated microglia may reflect that Iba1 functions in a low-level signaling pathway regulating actin bundling during membrane ruffling and phagocytosis, and thus has minimal impact on transcription. A related observation is that Lgals3 (galectin-3) was among the few genes reduced in Iba1−/− sham microglia (Fig. S12b). Previous work has reported that galectin-3 is not induced in activated microglia in the facial nucleus after facial nerve axotomy, in contrast to its robust induction after ischemic injury [54]. Thus, the reduction of Lgals3 in our sham comparison may reflect a subtle baseline shift in microglial state, although its contribution to the axotomy phenotype in our model remains uncertain. Furthermore, in axotomized motoneurons, immunohistochemical and ultrastructural analyzes consistently demonstrated markedly reduced microglial adhesion to neuronal somata in Iba1−/− mice compared with Iba1+/+. Electron microscopy further confirmed significantly impaired synaptic stripping in Iba1−/− mice, indicating that Iba1 activation is crucial for pathological microglia–neuron interactions and microglial shield formation. This interpretation is also compatible with prior work suggesting that classical “synaptic stripping” should not be viewed as a single binary event but rather as part of a broader, context-dependent remodeling process involving neuronal intrinsic changes, astrocytic ensheathment, and microglial motility and positioning. Earlier depletion or anti-mitotic studies did not always support an obligatory role of microglial proliferation in synapse removal, highlighting that distinct components of the glial response may have separable functions [8, 9]. GO analysis revealed that axotomy-induced upregulated genes in Iba1−/− motoneurons were significantly enriched in cognition/learning/memory and hormonal signaling pathways, and in molecular functions including transmembrane signaling receptor activity and hormone activity (Fig. 5E). A major enrichment involved membrane rafts/microdomains and cell-adhesion molecule binding, indicating altered membrane-associated signaling complexes and cell–cell interaction dynamics in axotomized motoneurons under Iba1 deficiency. Transcriptomics and immunohistochemistry indicated disrupted inhibitory (GABAergic) synapse distribution, consistent with reduced inhibitory inputs onto axotomized motoneurons in Iba1−/− mice. Mechanistically, axotomy in Iba1+/+ mice activated a microglial adhesion/ECM program with IRM-like cluster expansion, whereas Iba1 deficiency shifted microglial transcription toward inflammatory signaling, further enlarging the IRM-like population. Consistent with NicheNet predictions, microglial ligands upregulated in Iba1+/+ mice (Tgfbi/Mmp2/Anxa2) converged on neuronal integrins and Aplp2, indicating adhesion-oriented remodeling. In Iba1−/− mice, top ligands shifted to immune/inflammatory or synaptic-remodeling cues (Adam17/Ebi3/B2m/Cd72/Agrn), whereas a strong MHC-I → Aplp2 axis persisted. These observations support a model wherein an impaired microglial adhesion/ensheathment of injured motoneurons in Iba1−/− mice is compensated for by an inflammatory bias, paralleling neuronal GO signatures (downregulation of synaptic structure and receptor pathways; Fig. 5E) and inhibitory synaptic organization deficits. The facial nerve axotomy literature also indicates that motoneuron viability after injury is influenced not only by local glial remodeling but also by immune–neural crosstalk, including CD4+ T-cell-dependent neuroprotective mechanisms. In this broader framework, our data support the view that Iba1-dependent microglial ensheathment constitutes one structural component of a protective post-injury niche and that disruption of this component is associated with poorer neuronal survival [17]. This shift likely contributes to—or is at least consistent with—the poorer 1-month post-axotomy motoneuron survival and γH2AX-positive cell increase, indicating heightened cellular stress. At 28 days post-injury, motoneuron survival was significantly reduced in Iba1−/− mice. Although microglia have diverse functions and display both neuroprotective and neurotoxic effects, our findings—using a mouse model wherein Iba1−/− microglia show no robust cytokine transcriptional changes—demonstrate that microglial ensheathment protects axotomized motoneurons. Using three-dimensional electron microscopy and electrophysiological recordings, Chen Z. et al. demonstrated that lipopolysaccharide-activated microglia displace inhibitory GABAergic synapses in cortical neurons [6]. Pharmacological suppression of microglial activation by minocycline abolished this lipopolysaccharide-induced neuroprotection. Thus, they proposed that the microglial displacement of inhibitory synapses is neuroprotective after TBI, a finding that strongly aligns with our results. Several limitations should be acknowledged. First, only male mice were used in this study. Because sex may influence injury responses and motoneuron survival, the generalizability of our findings to female mice remains uncertain. Future studies including both sexes will be needed to clarify whether Iba1-dependent microglia–motoneuron interactions differ between sexes. Second, although the upregulation of Iba1 homolog Aif1l was not evident in microglia, we employed a conventional Iba1 knockout model, and compensatory mechanisms during development may have masked or altered primary Iba1 functions in adult microglia. Conditional or inducible knockout strategies will be required to elucidate the temporal and cell-type-specific roles of Iba1. Third, while transcriptomic and behavioral analyzes revealed no prominent phenotypes under normal conditions in Iba1−/− mice, altered immunoreactivity of ChAT, GABRA1, VGLUT1, and RET was observed in unaxotomized facial motoneurons (Fig. S18), suggesting subphenotypic differences caused by Iba1 deficiency. Fourth, although our findings highlight critical microglia-motoneuron interactions in a murine axotomy model, Iba1 functional relevance in human microglia remains unclear. In addition, prior work has emphasized that glial responses after peripheral axotomy are influenced by injury severity, anatomical system, species, and possibly the hormonal milieu; accordingly, caution is warranted in extrapolating the present findings beyond this specific facial nerve transection paradigm [9, 10, 55]. Human microglia heterogeneity has been extensively reported in AD studies. While DAM represents the predominant cellular state in mouse models, human studies have revealed a more complex situation with diverse transcriptional states that extend beyond those identified in mice [56]. Species-specific differences in microglial behavior and gene expression necessitate further studies to investigate whether similar Iba1-dependent mechanisms contribute to human CNS neuronal maintenance and repair. Overall, our findings indicate that Iba1 is essential for microglial shielding of motoneurons, preserving synaptic integrity and functional stability after injury. Without Iba1, impaired microglia–neuron interactions may trigger synaptic instability and abnormal compensatory transcriptional programs in motoneurons following injury. Notably, microglial transcription changed little between Iba1+/+ and Iba1−/−, whereas axotomized motoneurons showed marked transcriptional shifts, suggesting that microglial ensheathment loss—not altered microglial gene expression—primarily drives the neuronal response via intercellular interactions. Neurological disease therapeutic strategies (AL002 INVOKE-2 phase 2 trial and DNL-919 [TAK-920]) targeting upstream signaling in microglia have not hitherto yielded favorable outcomes [57]. This study clarified the microglia’s neuroprotective role by isolating their motility from their multifunctionality through the loss of Iba1, which is at the lowest signal transduction level. Iba1 could serve as a new target for selectively controlling microglial membrane ruffling and motility. Therefore, our findings highlight the essential role of Iba1 in modulating microglia–motoneuron interactions following nerve injury and underscore its potential as a therapeutic target for promoting neuroprotection. Conclusions Iba1 deficiency impairs perisomatic microglial remodeling after facial nerve axotomy, reducing microglial ensheathment of ChAT+ motoneuron somata and attenuating ultrastructural features of synaptic stripping that are evident in Iba1+/+ mice, thereby identifying Iba1 as a key downstream effector that supports microglia–motoneuron contact and cytoskeleton-dependent remodeling of the motoneuron. Single-nucleus RNA sequencing further indicated that, despite broadly similar microglial transcriptomes between Iba1+/+ and Iba1−/− conditions after axotomy and no major divergence in cytokine-related transcriptional programs, Iba1−/− mice showed an increased representation of an interferon-responsive microglial state (IRM) alongside pronounced genotype-dependent transcriptomic changes in injured motoneurons involving synapse organization and neurotransmission. Together, these multimodal data support a model in which Iba1 tunes microglial effector mechanisms for perisomatic remodeling and synaptic stripping without globally rewiring cytokine transcription, and that disruption of this selective remodeling axis is associated with compromised motoneuron integrity and survival at 28 days post-axotomy, as reflected by reduced motoneuron survival, lower ChAT expression, and increased nuclear γH2AX foci at this time point. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.