Brain injury environment critically influences the connectivity of transplanted neurons

Abstract

Cell transplantation is a promising approach for the reconstruction of neuronal circuits after brain damage. Transplanted neurons integrate with remarkable specificity into circuitries of the mouse cerebral cortex affected by neuronal ablation. However, it remains unclear how neurons perform in a local environment undergoing reactive gliosis, inflammation, macrophage infiltration, and scar formation, as in traumatic brain injury (TBI). To elucidate this, we transplanted cells from the embryonic mouse cerebral cortex into TBI-injured, inflamed-only, or intact cortex of adult mice. Brain-wide quantitative monosynaptic rabies virus (RABV) tracing unraveled graft inputs from correct regions across the brain in all conditions, with pronounced quantitative differences: scarce in intact and inflamed brain versus exuberant after TBI. In the latter, the initial overshoot is followed by pruning, with only a few input neurons persisting at 3 months. Proteomic profiling identifies candidate molecules for regulation of the synaptic yield, a pivotal parameter to tailor for functional restoration of neuronal circuits.

Fig. 1.. Transplanted neurons develop mature morphologies and synaptic structures within a cortical stab injury.

(A and B) Schematic and timeline of the experimental procedure. (C) Representative confocal image of a GFP graft, 5 dpt, showing an almost complete overlap with the immature neuronal marker Dcx (n = 4). (D) Confocal images of a representative transplantation site in the mouse cortex inflicted by a SW injury, 5 wpt [see all nuclei stained with 4′,6-diamidino-2-phenylindole (DAPI); n = 5]. The boxed area highlights dendritic branches of RFP transplanted neurons. (E) Example image of a single optical section in the transplant shows colocalization of RFP with Cux1, a marker of upper layer cortical identity (n = 2). (F) Z-stack projection of an example of grafted neurons and respective high-magnification insets shows the extension of apical dendrites as well as profuse basal dendrites (1) and axonal arborizations (2) from the cell bodies (n = 5). Notice the appearance of spines and boutons in dendrites and axons, respectively (arrowheads in the highlighted neurite). Scale bars, 100 μm (C and D, left), 50 μm (D, right; E; and F, left), and 10 μm (F, right). Ctx, cortex; Hipp, hippocampus.

Fig. 2.. Input connectivity of neuronal transplants in SW and intact cortex.

(A) Molecular tools and rationale of brain-wide monosynaptic tracing. (B) Example of a starter neuron (RFP⁺/GFP⁺): a neuron within the RFP transplant (RFP⁺) that has been infected by the GFP RABV. (C) Experimental timeline. (D and E) Local and brain-wide inputs (GFP-only) to transplants, traced at 4 wpt (n = 4/5). (D) Schematics depict brain regions that innervate the transplant. Red grading in sagittal sections and thickness of the lines in three-dimensional connectograms reflect the connectivity ratio for a given connection. (E) Color-coded connectivity ratios for transplants in SW or intact cortex (n = 4/5, respectively), as well as for endogenous neurons. The data shown for endogenous neurons (Endo) have been published before (5) and are used here solely for comparison. Note the excessive connectivity in SW and scarce connectivity in intact, as compared to the native. (F) Quantification of Vis-Vis and dLGN-Vis connectivity (n = 4/5; *P < 0.05 using Mann-Whitney test). (G) Presynaptic neurons (GFP⁺) in the dLGN of the thalamus. Scale bars, 50 μm (B) and 100 μm (D and G). See table S1 for abbreviations. Contra, contralateral; Ipsi, ipsilateral.

Fig. 3.. Gliosis in the various experimental groups and input connectivity of neuronal transplants in LPS-induced cortex.

(A) Timeline (top) and confocal images of the visual cortex of mice treated with LPS (1 or 3 mg/kg), or inflicted with a SW injury, and controls (intact, i.e., naïve mice and contralateral SW, as an additional control to SW). Sections were immunostained for microglia and reactive astrocyte markers (Iba1 and GFAP, respectively; bottom) [n = 4 for all conditions except LPS (1 mg/kg), where n = 2 mice were analyzed]. Note that the cellular response with the highest concentration of LPS is much closer to that observed in the SW-inflicted cortex. Insets are zoomed on the bottom row for appreciation of the differences in cellular morphology and GFAP expression. (B) Mean gray value of all pixels, calculated using Z-stack projections of similar thickness, for intact, LPS (3 mg/kg), and SW groups (n = 4; note that cortical layer 1 was excluded from the selected area). (C) Timeline and analysis of the brain-wide monosynaptic input to grafts in the brain of LPS (3 mg/kg)–treated mice as compared to those in the brain of intact mice (n = 5/6, respectively). Scale bars, 50 μm (A). See table S1 for abbreviations. Con, contralateral; i.p., intraperitoneal; wpi, weeks post-injury (SW)/ injection (LPS).

Fig. 4.. Comprehensive proteome analysis of SW-injured and LPS-inflamed visual cortex.

(A) Timeline for tissue punch collection. (B and C) Volcano plots showing log₂ mean abundance ratio and corresponding log₁₀ P value comparing SW-injured (n = 10) (B) and LPS-treated (n = 10) (C) with intact control (n = 20) cortical tissue (derived from 10, 5, and 10 mice, respectively, as we can consider both hemispheres from LPS and intact brains). Up-regulated proteins are in the pink area, and down-regulated proteins are in the blue area of the plots. (D) Selection of enriched gene ontology (GO) terms (biological process) of significantly enriched proteins in SW versus intact cortex. For the complete set of GO terms, see data S2. (E) Heatmap shows the significantly regulated proteins in the SW cortex, along with their regulation in the LPS cortex. (F) Venn diagram depicts differentially regulated proteins that overlap or are exclusive for each condition. (G) Selection of proteins that are exclusively up- or down-regulated in the SW-injured cortex. Highlighted are proteins further tested by immunostaining (see fig. S5). See data S2 for protein and GO analysis. wpi, weeks post-injury (SW)/injection (LPS).

Fig. 5.. Comparison of early and late connectivity shows transience in cortical stab injuries.

(A) Experimental timelines. (B) Local inputs (GFP-only) to transplants, traced at 1 and 3 mpt. (C) Color-coded brain-wide connectivity at 1 or 3 mpt in SW (n = 4/6, respectively). Decreased connectivity at 3 mpt shows that many of the early synaptic connections have been pruned. (D) Distribution and strength of single host-graft connections evidenced by the thickness of the lines between the graft (yellow) and each area (green). Scale bar, 100 μm (B). See table S1 for abbreviations.