Typical development of synaptic and neuronal properties can proceed without microglia in

Animals

All experiments were performed under licenses approved by the UK Home Office according to the Animals (Scientific Procedures) Act, having been approved by the University of Edinburgh Local Ethical Review Board. This study used Csf1r^{∆FIRE/∆FIRE} mice and control Csf1r^+/+ animals, which were generated as littermates by heterozygous Csf1r^∆FIRE/+ inter-crossing. The creation and characterization of the Csf1r^{∆FIRE/∆FIRE} mice is described elsewhere²³. Mice were on a mixed background: 64–68% homozygous for C57BL/6J, 22–25% CBA and 7–11% heterozygous, with approximately 2% unattributable (Mini Mouse Universal Genotyping Array). Animals were housed in communal cages (maximum six per cage) with ad libitum water and food in a normal 12 h light–dark cycle. Both male and female mice were used throughout the study. Experimenter and data analyzer were blind to the genotype. Sample size was estimated using 3R (reduction, replacement and refinement) principles, with sample size numbers calculated to achieve satisfactory power based on the variance and effect size in published data^{3,4,6,7,8,10,12,38}.

Mouse pup innate motor tasks

Mice aged P3–P10 underwent two behavioral experiments to assess motor function: (1) righting reflex and (2) negative geotaxis. Each mouse completed both the righting reflex and negative geotaxis tasks every day. In between the tasks, pups were returned to the heated cage to rest. When the session was over, pups were placed in their home cage and the dam was observed for signs of stress. Mice pups spent a maximum of 1 h separated from the dam.

Innate righting reflex task (P3–P7)

Mice were placed on their backs on a soft platform, with their paws held together for 3 s. Each mouse was released and the time taken for the pup to return to the prone position was recorded. A trial was defined as successful if mice righted themselves with all four paws on the floor. Mice were given three 15-s trials with a 20-s interval for recovery. All mice successfully managed to right themselves within the given time.

Negative geotaxis task (P3–P10)

This is the innate ability of rodents to recognize their orientation on a slope and to turn around so they are facing uphill. Mice were placed head pointing downwards on a 30-degree incline. Mice were placed so that all four paws were touching the surface of the incline. Each pup was released and the time taken for it to turn uphill was recorded. A hand was placed approximately 10 cm below the mouse to catch it if it fell. If a mouse fell, it was replaced back on the slope in the starting position and given another trial. This was repeated until the mouse successfully reorientated itself or until a maximum of ten attempts (nine falls) were made or until 90 s had passed. If the pup showed signs of stress, such as urination or defecation, the trial ended early. The number of failed attempts and the duration to success were recorded.

IHC

Mice were anesthetized with sodium pentobarbital and perfused transcardially with ice-cold PBS followed by 4% paraformaldehyde (PFA) in PBS. Brains were removed from the skull and postfixed in 4% PFA for 24 h. Brains were cryo-protected in 30% sucrose overnight and then mounted on a freezing microtome in optimal cutting temperature compound, where coronal 50-µm brain sections were prepared. Brain slices were placed in a well plate and washed with 0.1 M phosphate buffer and PBS. Slices were then blocked from unspecific binding using 10% normal goat serum (NGS), 0.3% Triton X-100 and 0.05% Na azide and PBS for 1 h at room temperature. Then, the primary antibody solution containing 10% NGS, 0.3% Triton X-100 and 0.05% Na azide and PBS was added (1:1,000 dilution Iba1, cat. no. ab283319, Abcam; 1:500 dilution Iba1, cat. no. 019-019741, Wako; 1:1,000 dilution GFAP, cat. no. CH22102, Neuromics; 1:1,000 dilution Aldh1l1, cat. no. 702573, Invitrogen; 1:500 dilution SYP, cat. no. 102-002, Synaptic Systems). Slices were incubated for 1 h at room temperature followed by 24 h at 4 °C. The well plate was removed and allowed to come to temperature before the secondary antibody solution was added (10% NGS, 0.3% Triton X-100, 0.05% Na azide and PBS) for 24 h. Then, slices were washed with PBS and 0.1 M phosphate buffer for 1 h before being mounted on glass slides and cover slipped. Images were obtained using a Leica SP8 confocal microscope. 1,024 × 1,024 pixel images of the CA1 (Bregma anteroposterior: −1.75 to −1.9 mm) and the dorsal LGN (Bregma anteroposterior: −2.25 to −2.5 mm) were taken using the ×10 (numerical aperture (NA) = 0.45) and ×20 (NA = 0.8) objectives. Multiple images were taken for each animal. To analyze astrocyte density, images were opened using the ImageJ software. For cell counts, a counting grid consisting of 200 × 200 µm squares was superimposed over the stratum radiatum of the CA1. Using the cell counter tool, the number of 4′,6-diamidino-2-phenylindole (DAPI)⁺, GFAP⁺ and ALDH1L1⁺ cells were quantified. Each DAPI, GFAP and ALDH1L1 cell was counted only if the cell body was within the grid and over 50% of the protrusions were within the grid. The average cell density was measured for each animal and reported as the number of GFAP and ALDH1L1 per 100 µm² and as the percentage of total DAPI cells.

Bulk, sc-RNA-seq and snRNA-seq

Bulk tissue RNA extraction was carried out as described in ref. ²³. Briefly, brains from saline-perfused mice were dissected, snap-frozen and subsequently disrupted in the Precellys 24 Homogenizer (Bertin Instruments). RNA isolation was performed using the RNeasy Plus Mini Kit (QIAGEN). RNA-seq reads were mapped to the primary assembly of the mouse reference genome using the STAR RNA-seq aligner v.2.7.11 (ref. ⁴¹). Tables of per-gene read counts were generated from the mapped reads using featureCounts⁴². Differential gene expression was then performed in R using DESeq2 (ref. ⁴³). For bulk RNA-seq, 6–7 mice per genotype were analyzed. To sort astrocytes (4–5 mice per genotype), mouse neocortices were collected then dissociated with Adult Brain Dissociation Kit (Miltenyi Biotec) on a gentleMACS Octo Dissociator using program 37C_ABDK_01. Dissociated samples were then treated with debris and red blood cell removal steps to obtain cell suspensions. For FACS, cells were resuspended to a final volume of 100 μl 0.1% phosphate buffer and incubated with fluorescence-conjugated antibodies at 4 °C for 30 min. The following antibodies were used: ACSA2 APC (1:200 dilution, cat. no. 130-116-245, Miltenyi Biotec) and O4 PE (1:100 dilution, 30-117-357, Miltenyi Biotec). ACSA2 APC⁺ and O4-PE⁻ live single cells were selected as the astrocyte population. These were then loaded onto a 10x Chromium Controller. For snRNA-seq, neocortices were acutely collected, flash-frozen and then stored at −80 °C. A nucleus isolation kit (cat. no. NUC201-1KT, Sigma-Aldrich), dithiothreitol (DTT) (cat. no. R0861), RNase inhibitor (cat. no. AM2694 Thermo Fisher Scientific), 30 µm cell strainer (Partec CellTrics) and 30-G insulin needles (cat. no. 324826, BD) were used. Cryopreserved tissue samples were thawed on ice. Lysis was performed with PURE buffer, 0.1 M DTT, 10% Triton X-100 and RNase inhibitor. The samples were mechanically dissociated in lysis buffer using a P-1000 pipette ten times, and a further three times with insulin needles. The lysate was mixed with a sucrose cushion, which was prepared using PURE 2M sucrose, sucrose cushion buffer, 0.1 M DTT and RNase inhibitor. This mixture was then filtered through a 30-µm cell strainer. In a new low-binding tube, 200 µl of sucrose cushion was added and carefully overlaid with 560 µl of the filtrate. The samples were centrifuged for 45 min at 4 °C. The supernatant was discarded and the pellet was resuspended in Dulbecco’s PBS (DPBS) (cat. no. 14190-094) + 0.5% BSA with RNase inhibitor. The suspension was then centrifuged at 1,000g, repeated and finally resuspended in 200 µl Dulbecco’s PBS + 0.5% BSA with RNase. Sorting of nuclei (six mice per genotype) was done using flow cytometry. Gating was set based on size and granularity using forward and side scatter to capture singlets and remove debris. Nuclei were stained with DAPI for detection. Quality control was conducted using Luna FX7 with Acridine Orange/Propidium Iodide; and the input nuclei number was 20,000 per sample. For library preparation, nuclear suspensions were loaded onto a 10x Chromium Controller (10x Genomics). Single-nucleus transcriptomic amplification and library preparation were conducted with the Chromium Single Cell 3′ v.3.1 Reagent Kit (10x Genomics). Libraries were first sequenced on the iSeq 100 System (cat. no. 20021532, Illumina) using the iSeq 100 i1 Reagent v.2 (300 cycle) Kit (cat. no. 20031371). Library molarity for sequencing was calculated using the Qubit double-stranded DNA quantification results and the fragment size information from the Bioanalyzer results. Libraries were normalized to 10 nM and equal volumes were pooled and diluted for sequencing. The PhiX Control v.3 (cat. no. FC-110-3001) library was spiked into the run at a concentration of ~4% to help with cluster resolution and facilitate troubleshooting. iSeq data were then analyzed so that pools could be rebalanced for subsequent deep sequencing. Deep sequencing was performed on the NextSeq 2000 platform (cat. no. SY-415-1002, Illumina) using the NextSeq 1000/2000 P3 Reagents (100 cycles) v.3 Kit (cat. no. 20040559). The PhiX Control v.3 library was spiked in at a concentration of ~1%. For single-cell and single-nucleus sequencing data, sequencing reads were mapped to the mouse genome; per-cell, per-gene count matrices were produced using 10x CellRanger v.7.0.1 (ref. ⁴⁴). Quality control, normalization and clustering of data were performed using the Seurat R package, v.4.4.0 (ref. ⁴⁵). For snRNA-seq, ambient RNA was estimated and removed using the SoupX R package v.1.6.2 (ref. ⁴⁶). Doublets were identified and removed using the scDblFinder R package v.1.10.1 (ref. ⁴⁷). Pseudobulk differential expression analysis was performed by summarizing single-cell gene expression profiles at the individual level using the aggregateBioVar R package v.1.6.0 (ref. ⁴⁸); then differentially expressed genes between genotypes were calculated using DESeq2 v.1.36.0.

Slice preparation and patch-clamp and field potential recordings

Acute slices were prepared as described previously⁴⁹. Once cut, slices were transferred to a holding chamber containing carbonated sucrose-artificial cerebrospinal fluid (CSF) (whole-cell recordings) or aCSF (field recordings). Slices were allowed to recover for 30 min or longer at 35 °C until needed. For the electrophysiological recordings, slices were transferred to a submerged recording chamber, perfused with carbonated aCSF⁴⁹ at a flow rate of 3–6 ml min⁻¹ at 30–31 °C.

For…