34 - Regional Transcriptomic and Immunohistochemical Signatures of Neonatal Hypoxic-Ischemic Encephalopathy in Human Brain Tissue

Publication Number: 1029.34

Olivia C. Brandon, University of Washington School of Medicine, Seattle, WA, United States; Thomas Wood, University of Washington, Seattle, WA, United States; Veronica England, University of Washington, Seattle, WA, United States; Sandra E. Juul, University of Washington, Seattle, WA, United States; Raj Kapur, Seattle Children's Hospital, Seattle, WA, United States; Sarah Kolnik, University of Washington - Seattle Children's Hospital, Seattle, WA, United States

Background: Hypoxic-ischemic encephalopathy (HIE) is a leading cause of neonatal mortality and long-term neurodevelopmental impairment, yet treatment options remain limited. Despite extensive preclinical research, region-specific transcriptomic responses to HIE and their relationship to histopathologic injury patterns in human neonatal brain tissue remain poorly understood.

Objective: To identify cell-type and transcriptomic signatures of HIE in human neonatal brain tissue.

Design/Methods: We analyzed paraffin-embedded brain tissue from six term-delivered neonates with HIE displaying varying neuropathological severity. Prior to transcriptomic and immunohistochemical analysis, HIE pathology was scored semi-quantitatively by a board-certified pediatric pathologist based on assessment of hematoxylin-and-eosin-stained (H&E) brain sections (Table 1). Transcriptomic profiling used the NanoString nCounter Human Neuropathology panel (770 genes), complemented by quantitative immunohistochemical (qIHC) staining to assess microglia (Iba-1) and astrocytes (GFAP). The relationship between pathology score and gene expression was determined using linear mixed effect models with random effect by patient.

Results: Clinical history and pathology summary are shown in Table 1. Percent positive Iba-1 staining increased significantly with pathology severity in the cortical gray matter (p=0.03), cerebral white matter (p=0.02), and hippocampus (p=0.03). Percent positive GFAP was also significantly associated with increasing pathology score in the white matter (p=0.02). Microglial number, size, and GFAP-positive astrocyte density were markedly increased in more severely affected regions (Figure 1). Globally, n=166 transcripts were significantly differentially expressed with increasing pathology severity, including the most globally affected transcript being ataxia-telangiectasia mutated (ATM), a key kinase in DNA double-strand break repair (Figure 2). Regionally, n=113 were significant in the hippocampus, n=62 in the white matter, and n=61 in the basal ganglia. Biological pathways related to differentially expressed transcripts included inflammatory responses, catabolic processes, and cell communication (Figure 2).

Conclusion(s): In infants with HIE, microgliosis and astrocytosis increase with local pathology severity, with transcriptomics showing region-specific expression patterns linked to inflammation and cellular stress. These findings provide new insight into mechanisms of injury and resilience and may guide the development of targeted neuroprotective therapies for infants with HIE.

Table 1. Infant characteristics and pathology scores.
NR, not recorded (no electroencephalography); H&E, hematoxylin-and-eosin
*Each brain region was evaluated semi-quantitatively (0-3) for histopathologic features of injury, including edema, reactive astrocytosis/gliosis, vacuolated microglia, hyper-eosinophilic (necrotic) neurons, confluent necrosis, apoptosis, and hemorrhage. Regional pathology scores were summed or averaged across regions to derive composite severity indices. Scores represent the extent or severity of each sample as follows:
- Edema: 0 = none; 1 = narrow rims of pericellular clearing or "washed-out" neuropil; 2 = wide rims of pericellular clearing in some areas; 3 = diffuse wide rims of pericellular clearing.
- Reactive astrocytes/gliosis: 0 = none; 1 = sparse astrocytes with visible cytoplasm; 2 = many astrocytes with visible cytoplasm; 3 = abundant astrocytes with visible cytoplasm.
- Vacuolated microglia: 0 = none; 1 = sparse; 2 = many; 3 = abundant or confluent.
- Hyper-eosinophilic neurons: 0 = none; 1 = sparse; 2 = many; 3 = abundant or confluent.
- Confluent necrosis: 0 = none; 1 = hyper-eosinophilic neurons only; 2 = focal or multifocal necrosis; 3 = diffuse (e.g., laminar) necrosis.
- Apoptosis: 0 = none; 1 = sparse; 2 = many; 3 = abundant.
- Hemorrhage: scored 0-3 according to the severity and regional extent of parenchymal or perivascular blood.

Figure 1. Representative histologic images from regions with less injury (left column) and greater injury (right column).
Hippocampus (A, B; top row), gray and white matter (C, D; middle row), and basal ganglia (E, F; bottom row) shown. The hippocampus is stained with Iba-1 to highlight microglia, while gray and white matter and basal ganglia are stained with GFAP to highlight astrocytes. The number and size of microglial cells and the density of GFAP-immunoreactive astrocytes are markedly increased in areas with more severe hematoxylin-and-eosin (H&E) findings. Insets (20X magnification) show the boxed regions: dentate gyrus of the hippocampus, subcortical white matter beneath a sulcus, and caudate nucleus of the basal ganglia. Less severe examples are from Patient 2 (global pathology score: 19) and the more severe examples are from Patient 6 (global pathology score: 55).

Figure 2. Differential transcript expression globally and regionally relative to pathology severity.
Volcano plots showing the relationship between effect size (β: log₂ fold-change per unit increase in pathology score) and statistical significance (−log₁₀ p-value) for all transcripts analyzed globally (A), in the hippocampus (C), white matter (E), and basal ganglia (G). The global model included adjustment for region. Each point represents an individual transcript; red points indicate transcripts significantly upregulated, and blue points indicate those significantly downregulated with increasing pathology severity. The horizontal dashed line marks the significance threshold (p=0.05). The top five most significantly upregulated and downregulated transcripts are labeled. ShinyGo was used to visualize the biological pathways related to the significantly expressed transcripts globally (B), in the hippocampus (D), white matter (F), and basal ganglia (H).