Environmental enrichment ameliorates perinatal brain injury and promotes functional white matter recovery

Perinatal hypoxia is a major cause of preterm brain injury, prominently affecting periventricular white matter by delaying maturation of oligodendrocytes (OLs), the myelin-forming glia. This maturational arrest leads to lasting myelin abnormalities and neurodevelopmental impairments. Given the plasticity of the developing brain, exogenous environmental factors may modulate recovery, but mechanisms are unclear. The study asks whether environmental enrichment (EE) during critical developmental windows can enhance endogenous repair of oligodendroglial lineage cells and myelination after perinatal hypoxia, and whether such cellular effects translate into functional recovery. It further aims to define which components and timing of EE are necessary, whether de novo oligodendrogenesis is required, and to elucidate oligodendroglial-specific molecular programs underlying EE-induced recovery using TRAP RNA-seq.

• Premature infants’ environments markedly influence cognitive and behavioral outcomes, and environmental stimulation promotes neural plasticity and myelin structural changes affecting function (cited works 9–16).
• EE enhances recovery after various brain injuries in adults and affects OL lineage responses to experience (22–26).
• Rodent models can recapitulate diffuse WM injury from perinatal hypoxia due to developmental timing differences (8,17,18).
• Prior work shows hypoxia causes OL developmental delay; initial OL loss may recover by ~1 month post-injury but myelin abnormalities persist (4–8,17).
• Activity-dependent myelination and rapid generation of new OLs underpin motor skill learning, suggesting experience influences myelin (13–15).
• Social environment affects myelination and outcomes; isolation yields thinner myelin and behavioral deficits, while enriched social settings correlate with improved development in preterm infants (43,44).

• Animal models and groups: Primarily CNP-EGFP mice for OL lineage visualization. Four main groups: Normoxic Standard (NX), Normoxic Enriched (NX-EE), Hypoxic Standard (HX), Hypoxic Enriched (HX-EE). Hypoxia administered P3–P11 at ~10.5% O2. EE initiated at P15 and maintained to P45 unless otherwise specified.
• Environmental enrichment (EE): Combined components including increased socialization (8–12 mice/cage), voluntary physical activity (running wheel), and exposure to novel objects in a complex cage environment.
• Component controls: (a) Activity-only (running wheel, fewer cagemates, no novel objects); (b) Socialization + novel objects (no running wheel).
• Timing paradigms: Early EE (P15–P25), Delayed (P25–P35), Late (P35–P45), Continuous (P15–P45), Delayed & Extended (P25–P55; 30 days total).
• Cell and histology assessments: Immunohistochemistry on 40-μm coronal sections; markers included NG2 and PDGFRα (OPCs), Ki67 (proliferation), CC1 (mature OLs), Olig2 (OL lineage), cleaved caspase-3 (apoptosis), astrocytic marker GFAP, microglial markers IBA1 and CD68, axonal markers SMI31 (phosphorylated neurofilament) and SMI32 (nonphosphorylated). ENPP6 in situ hybridization to identify premyelinating OLs. Cell counts normalized to volume across subcortical WM regions (corpus callosum, cingulum, external capsule).
• Myelination assays: Western blots for MAG and MBP from microdissected subcortical WM at P45. Electron microscopy of corpus callosum at P45 for myelinated axon number and g-ratio analyses (≥100 axons/brain).
• Oligodendrogenesis fate mapping: BrdU injections P11–P15 (post-HX, pre-EE) or P24–P28 (during EE); quantification of EGFP+BrdU+CC1+ OLs at P45.
• Functional behavior: Inclined beam-walking task at P45 with 2-cm and 1-cm beams; foot slips and time to completion measured.
• Genetic requirement for de novo oligodendrogenesis: PDGFRα-CreERT2; R26R-YFP; Myrf flox/flox and Myrf +/+ controls. Tamoxifen P13–P17 to ablate Myrf in OPCs, thereby blocking maturation; confirm recombination by qPCR. Assess OL numbers (YFP+CC1+) and beam-walking performance under HX vs HX-EE.
• Cell type-specific translatome profiling: TRAP in PDGFRα-bacTRAP (OPCs/immature OLs) and CNP-bacTRAP (mature OLs) mice. WM microdissection followed by anti-GFP immunopurification of ribosome-bound mRNAs; RNA-seq (NovaSeq 6000; 50 bp PE). Read trimming with Trimmomatic; alignment with STAR; counts with HTSeq; DESeq2 for DEG analysis (normalized counts >5 in all samples; adjusted p<0.05). Ingenuity Pathway Analysis for functional predictions and upstream regulators. Validation of IP enrichment by qRT-PCR for lineage-specific markers.
• Hippocampal neurogenesis validation under NX-EE: Quantification of Sox2, Nestin, Ki67, and Dcx in dentate gyrus at P45.
• Statistics: Normality by D'Agostino & Pearson. One-way ANOVA with Tukey’s post hoc for multiple comparisons; unpaired t-test where specified. Significance at p<0.05. Sample sizes provided per figure legends.

• EE enhances OPC proliferation and OL maturation after hypoxia: At P30, HX increased OPCs and proliferating OPCs vs NX (OPCs: 6080 ± 167 vs 4090 ± 241 cells/mm³; proliferating OPCs: 5006 ± 279 vs 3243 ± 110). EE further increased both after HX (HX-EE vs HX OPCs: 9284 ± 393 vs 6080 ± 167; proliferating OPCs: 7632 ± 486 vs 5006 ± 279), with no effect in NX.
• Increased mature OLs and myelin proteins with EE after HX: At P45, HX-EE had more CC1+ OLs than HX (30,115 ± 1011 vs 24,302 ± 411 cells/mm³); HX also exceeded NX (24,302 ± 411 vs 17,580 ± 492). Western blots showed higher MAG and MBP in HX-EE vs HX (p = 0.037 and p = 0.031).
• Ultrastructural recovery: EM showed more myelinated axons in HX-EE vs HX (22 ± 2.7 vs 8.7 ± 3.0 axons/field). HX had markedly fewer myelinated axons than NX (8.7 ± 3.0 vs 32.5 ± 1.5). g-ratios were reduced in HX-EE vs HX, indicating thicker myelin.
• Premyelinating OL accumulation after HX: ENPP6+ premyelinating OLs increased in HX vs NX (12,090 ± 1230 vs 3559 ± 244 cells/mm³); reduced in HX-EE vs HX (8095 ± 628 vs 12,090 ± 1230), remaining above NX.
• Axonal cytoskeleton markers: HX decreased phosphorylated neurofilaments (SMI31) at P22 and P30 vs NX (P22: 4.23 ± 0.87% vs 11.08 ± 0.61; P30: 4.03 ± 0.16 vs 7.72 ± 0.23), partially rescued by EE. Nonphosphorylated neurofilaments (SMI32) increased after HX and were rescued by EE by P30. No EE-related changes in astrocyte or microglia densities.
• Oligodendrogenesis timing: BrdU P11–P15 increased new OLs after HX regardless of EE; BrdU during EE (P24–P28) increased new OLs in HX-EE vs HX (23.75 ± 0.55 vs 20.64 ± 0.45 cells/mm³). EE did not increase OL generation under NX conditions, although NX-EE increased hippocampal neurogenesis markers (Sox2, Nestin, Ki67, Dcx).
• Behavioral recovery: HX impaired inclined beam performance; HX-EE reduced foot slips (2-cm: 0.8 ± 0.1 vs 1.4 ± 0.2; 1-cm: 1.6 ± 0.1 vs 2.3 ± 0.2). NX-EE decreased traversal time but not foot slips vs NX.
• EE components: Running wheel alone did not alter OPCs/OLs or myelination nor improve behavior after HX. Increased socialization plus novel objects (without wheel) also failed to affect OL lineage or myelination. A complete EE was required for recovery.
• Critical window: Early EE (P15–P25) modestly increased OLs and lineage cells but did not improve EM or behavior. Delayed (P25–P35), late (P35–P45), or delayed & extended (P25–P55) EE did not improve cellular or behavioral outcomes after HX. Continuous EE from P15–P45 was necessary for robust cellular, ultrastructural, and functional recovery.
• Requirement for de novo oligodendrogenesis: Myrf ablation in OPCs prevented EE-induced increase in new OLs and abolished behavioral improvements after HX. HX-EE Myrf+/+ improved foot slips (2-cm: 2.1 ± 0.1 vs 3.2 ± 0.2; 1-cm: 3.3 ± 0.3 vs 5.6 ± 0.3), whereas HX-EE Myrf fl/fl performed similarly to HX controls, indicating de novo OL maturation is required.
• OPC TRAP RNA-seq: HX vs NX OPCs showed 745 DEGs at P18 (289 up/456 down), 672 at P22 (311 up/361 down), 107 at P30 (65 up/42 down). Early HX predicted decreases in movement/survival/membrane and cytoskeleton, with later predicted increases in cell movement, differentiation, and process formation, suggesting rebound maturation programs. EE vs HX OPCs: 509 DEGs at P18 (277 up/232 down), 363 at P22 (258 up/105 down), 228 at P30 (74 up/154 down). EE induced early increases in synapse/membrane/process formation functions at P18–P22 with decreases by P30, indicating a temporal acceleration of maturation-related programs. Upstream regulator patterns were transiently modulated by EE.
• OL TRAP RNA-seq: HX vs NX OLs had 369 DEGs at P22 (260 up/109 down), 702 at P30 (181 up/521 down), 9 at P45 (8 up/1 down). P22 changes reflected delayed maturation; P30 showed reductions in cytoskeleton organization, microtubule dynamics, neurite growth, and phospholipid metabolism, consistent with impaired myelination and accumulation of premyelinating OLs. EE vs HX OLs: 196 DEGs at P22 (73 up/123 down), 51 at P30 (22 up/29 down), 50 at P45 (40 up/10 down). Notable EE-upregulated genes:
  • OL maturation/myelination: Ctsl, Sema5a, Sytl2; Vcam1, Slc8a3 (NCX3), Daam2.
  • Morphology/cytoskeleton: Kalrn, Lcp1, Nefl (P22); Chn1, Cnr1, Plxna4, Neo1, Plxnb3, Tuba4a (P45).
  • Metabolism/axon support: Pdha1, Dld (P22; PDC components), Slc16a1/MCT1 (P30), Elovl6, Elovl7, Fabp5 (P30; lipid metabolism).

The findings demonstrate that early and continuous environmental enrichment engages oligodendroglial plasticity to counteract hypoxia-induced developmental delays. EE amplifies the endogenous proliferative response of OPCs after injury, accelerates differentiation into mature OLs, and enhances myelin formation and thickness, translating into improved motor coordination on a WM-dependent task. A complete EE—combining social, cognitive, and physical components—is required; individual elements alone are insufficient. The efficacy of EE is tightly time-locked to a critical post-injury developmental window, emphasizing early intervention. Genetic blockade of OL maturation via Myrf ablation abolishes EE benefits, establishing that de novo oligodendrogenesis and subsequent myelination are necessary for functional recovery. Cell type-specific TRAP RNA-seq reveals that EE transiently advances OPC maturation programs and upregulates OL gene networks for myelination, cytoskeletal remodeling, metabolism, and axonal support, offering mechanistic insight into how environmental factors can re-synchronize glial and neuronal development to restore function.

This study identifies environmental enrichment as an effective, time-sensitive intervention that promotes oligodendrogenesis, myelination, and functional recovery after perinatal hypoxia-induced white matter injury. EE must be comprehensive and initiated early and continuously to be effective. De novo oligodendrogenesis is essential for EE-driven behavioral improvements. OPC- and OL-specific TRAP RNA-seq uncovers dynamic, stage-specific molecular programs accelerated by EE, including genes governing differentiation, myelin formation, cytoskeletal remodeling, and metabolic support. These results support targeting myelin plasticity and environmentally driven gene networks as therapeutic avenues for preterm brain injury. Future research should dissect causative pathways among identified candidates, optimize translatable enrichment protocols and timing, and evaluate combinatorial strategies with pharmacologic agents to harness endogenous repair.

• The OPC-focused TRAP RNA-seq alone did not fully explain HX-induced myelination deficits, necessitating OL-focused analyses.
• In OL comparisons (HX-EE vs HX), fewer DEGs limited pathway analysis via IPA, providing minimal results and requiring manual curation of candidate genes.
• Findings are based on mouse models of perinatal hypoxia and EE; translation to human preterm infants requires caution and clinical validation.
• Behavioral assessment centered on an inclined beam task; broader functional domains were not evaluated.
• EE effects were highly dependent on timing and completeness of the paradigm; the precise contributions of each component beyond the all-or-none conclusion remain to be mechanistically dissected.