Profiling Temporal Changes of the Pineal Transcriptomes at Single Cell Level Upon Neonatal HIBD

Abstract

Neonatal hypoxic-ischemic brain damage (HIBD) often results in various neurological deficits. Among them, a common, yet often neglected, symptom is circadian rhythm disorders. Previous studies revealed that the occurrence of cysts in the pineal gland, an organ known to regulate circadian rhythm, is associated with circadian problems in children with HIBD. However, the underlying mechanisms of pineal dependent dysfunctions post HIBD remain largely elusive. Here, by performing 10x single cell RNA sequencing, we firstly molecularly identified distinct pineal cell types and explored their transcriptome changes at single cell level at 24 and 72 h post neonatal HIBD. Bioinformatic analysis of cell prioritization showed that both subtypes of pinealocytes, the predominant component of the pineal gland, were mostly affected. We then went further to investigate how distinct pineal cell types responded to neonatal HIBD. Within pinealocytes, we revealed a molecularly defined β to α subtype conversion induced by neonatal HIBD. Within astrocytes, we discovered that all three subtypes responded to neonatal HIBD, with differential expression of reactive astrocytes markers. Two subtypes of microglia cells were both activated by HIBD, marked by up-regulation of Ccl3. Notably, microglia cells showed substantial reduction at 72 h post HIBD. Further investigation revealed that pyroptosis preferentially occurred in pineal microglia through NLRP3-Caspase-1-GSDMD signaling pathway. Taken together, our results delineated temporal changes of molecular and cellular events occurring in the pineal gland following neonatal HIBD. By revealing pyroptosis in the pineal gland, our study also provided potential therapeutic targets for preventing extravasation of pineal pathology and thus improving circadian rhythm dysfunction in neonates with HIBD.

Keywords: astrocyte; hypoxic-ischemic brain damage; microglia; pineal gland; pinealocyte; pyroptosis; single cell RNA sequencing.

FIGURE 1

Transcriptomic characterization of pineal gland at single cell level post neonatal HIBD (A,B) Uniform Manifold Approximation and Projection (UMAP) visualization of rat pineal gland cells collected at 24 and 72 h post sham or neonatal HIBD (C) Heatmap of expression values of characteristic genes for distinct cell types.

FIGURE 2

Cell composition and proliferation of the pineal gland post neonatal HIBD (A,D) Relative composition of distinct cell types in total pineal cells (%) at 24 h (A) or 72 h (D) post sham or neonatal HIBD, respectively (B,E) Feature plots of Ki67 among distinct cell types of the pineal gland at 24 h (B) or 72 h (E) post sham or neonatal HIBD, respectively (C,F) Representative images of transverse sections of the pineal gland with immunostaining against Ki67 and quantification at 24 h (C) or 72 h (F) post sham or neonatal HIBD, respectively. Scale bar: 200 μm n = 3,3 for sham and HIBD animals, five sections crossing the pineal gland were used for individual animals, Student’s t test, n. s, no statistical significance.

FIGURE 3

Identification of cell type prioritization in the pineal gland upon neonatal HIBD (A,C) UMAP visualization of distinct cell types colored by Augur cell type prioritization at 24 h (A) or 72 h (C) post sham or neonatal HIBD, respectively (B,D) Values of the area under the receiver operating characteristic curve (AUC) of distinct cell types colored by Augur cell type prioritization at 24 h (B) or 72 h (D) post sham or neonatal HIBD, respectively.

FIGURE 4

Transcriptome changes in pinealocyte subtypes post neonatal HIBD (A,E) Feature plots of a pinealocyte specific transcriptional factor (Lhx4) over total pineal populations at 24 h (A) or 72 h (E) post sham or neonatal HIBD, respectively (B,F) Violin plots showing expression differences of characteristic genes (Tph1 and Asmt) at 24 h (B) or 72 h (F) post sham or neonatal HIBD, respectively (C,G) Volcano plots highlighting genes with significant expressional changes between sham and HIBD in both subtypes at 24 h (C) or 72 h (G) post sham or neonatal HIBD, respectively. Red dots: < adjusted p value (0.0001) and >2 fold; Blue dots: < adjusted p value (0.0001) with <2 fold; Gray dots: no significance (D,H) Quantification of pineal RNA expression of Asmt and Aanat at 24 h (D) or 72 h (H) post sham or neonatal HIBD, respectively. n = 5,5 for sham and HIBD animals, fold changes of RNA expression were first normalized to those of GAPDH, and then to those in the sham. Student’s t test, ** or *, p < 0.01 or 0.05.

FIGURE 5

Expression changes of characteristic genes in distinct astrocyte subtypes post neonatal HIBD (A,C) Feature plots of an astrocyte specific transcriptional factor (Sox9) over total pineal populations at 24 h (A) or 72 h (C) post sham or neonatal HIBD, respectively (B,D) Bubble plots showing expression levels of reactive astrocyte related genes at 24 h (B) or 72 h (D) post sham or neonatal HIBD, respectively. Note, the size of dots represents the abundance of cells expressing of a given gene; the color scale represents normalized expression level for a given gene across all cells (E–G) Representative images of transverse sections of the pineal gland with immunostaining against Vimentin and quantification (G) at 24 h (E) or 72 h (F) post sham or neonatal HIBD, respectively. Scale bar: 200 μm n = 5,5 for sham and HIBD animals, five sections crossing the pineal gland were used for individual animals, Student’s t test, **, p < 0.01.

FIGURE 6

Expression changes of characteristic genes in distinct microglia subtypes post neonatal HIBD (A,C) Feature plots of a microglia specific transcriptional factor (Lyl1) over total pineal populations at 24 h (A) or 72 h (C) post sham or neonatal HIBD, respectively (B,D) Bubble plots showing expression levels of inflammation related genes at 24 h (B) or 72 h (D) post sham or neonatal HIBD, respectively (E) Quantification of pineal RNA expression of Ccl3 and Ccl4 at 24 h or 72 h post sham or neonatal HIBD, respectively. n = 5,5 for sham and HIBD animals, fold changes of RNA expression were first normalized to those of GAPDH, and then to those in the sham. Student’s t test, **, p < 0.01.

FIGURE 7

Microglial pyroptosis post neonatal HIBD (A,C) Feature plots of a pyroptosis specific factor (Caspase-1) over total pineal populations at 24 h (A) or 72 h (C) post sham or neonatal HIBD, respectively (B,D) Bubble plots showing expression levels of pyroptosis and apoptosis related genes at 24 h (B) or 72 h (D) post sham or neonatal HIBD, respectively (E–F) Western blotting (E) with quantification (F) of NLRP-3, Capase-1, GSDMD, and GSDMD-N. in F, for each lane, the expression of a given protein is first normalized to that of β-actin. ** or *, p < 0.01 or 0.05 (sham and HIBD 24 h), ## (sham and HIBD 72 h), p < 0.01, n = 3 for each condition, one-way-ANOVA with Bonferroni’s multiple-comparisons post hoc test.