Cord blood administration induces oligodendrocyte survival through alterations in gene expression

Abstract

Oligodendrocytes (OLs), the predominant cell type found in cerebral white matter, are essential for structural integrity and proper neural signaling. Very little is known concerning stroke-induced OL dysfunction. Our laboratory has shown that infusion of human umbilical cord blood (HUCB) cells protects striatal white matter tracts in vivo and directly protects mature primary OL cultures from oxygen glucose deprivation (OGD). Microarray studies of RNA prepared from OL cultures subjected to OGD and treated with HUCB cells showed an increase in the expression of 33 genes associated with OL proliferation, survival, and repair functions, such as myelination. The microarray results were verified using quantitative RT-PCR for the following eight genes: U2AF homology motif kinase 1 (Uhmk1), insulin-induced gene 1 (Insig1), metallothionein 3 (Mt3), tetraspanin 2 (Tspan2), peroxiredoxin 4 (Prdx4), stathmin-like 2 (Stmn2), myelin oligodendrocyte glycoprotein (MOG), and versican (Vcan). Immunohistochemistry showed that MOG, Prdx4, Uhmk1, Insig1, and Mt3 protein expression were upregulated in the ipsilateral white matter tracts of rats infused with HUCB cells 48h after middle cerebral artery occlusion (MCAO). Furthermore, promoter region analysis of these genes revealed common transcription factor binding sites, providing insight into the shared signal transduction pathways activated by HUCB cells to enhance transcription of these genes. These results show expression of genes induced by HUCB cell therapy that could confer oligoprotection from ischemia.

Fig 1. OLs differentiate into the mature phenotype

Photomicrographs show immunofluorescent staining of OL cultures at selected time points following PDGF-AA withdrawal. (A) 6 hrs after withdrawal, NG2 (red) and O4 (green) colocalized in OLs that exhibited both bipolar and immature morphology, as indicated by the lateralized orientation of processes and the relatively low number of processes, respectively. (B) At 36 hrs, NG2-positive OLs (red) expressed MBP (green) and contained greater numbers of processes, indicating that this withdrawal period was sufficient for differentiation into the mature phenotype. Scale bars = 50µm.

Fig 2. HUCB cells decrease LDH release from OLs subjected to 24 hrs OGD

Media from OL cultures subjected to OGD-only contained elevated levels of LDH compared to media from normoxic controls, demonstrating OGD-induced cellular injury. OL cultures subjected to OGD were rescued by co-incubation with HUCB cells, as LDH release was reduced back to levels of normoxic controls (*p < 0.01, n=7).

Fig 3. Affymetrix gene array fold changes are confirmed by qRT-PCR

HUCB cell treatment of OLs exposed to 24 hrs OGD significantly increased gene expression of MOG, Insig1, Prdx4, Mt3, Tspan2, Stmn2, Uhmk1 and Vcan (A–H) as compared to OLs subjected to OGD alone (* p < 0.05, n = 5). Additionally, HUCB cell treatment of OLs exposed to normoxia increased the expression of MOG (A), Insig1 (B), Prdx4 (C), Stmn2 (F), and Vcan (H) as compared to non-treated normoxic controls (^ p < 0.05 n = 5). Under OGD conditions, OL expression of Mt3 (D), Tspan2 (E) and Stmn2 (F) were significantly reduced in non-treated cells compared to both normoxic groups (# p < 0.05, n = 5).

Fig 4. HUCB cells reduce infarct volume

HUCB cells provide neuroprotection when given systemically 48hrs post-stroke. Photomicrographs depict Fluoro-Jade staining of coronal rat brain sections at time points 54, 72 and 96 hrs post-MCAO. Infarct volume remained constant in MCAO only groups at 54 hrs (A), 72 hrs (B), and 96 hrs (C) post MCAO. Whereas HUCB cell administration reduced infarct volume at 72 hrs (E) and 96 hrs (F) post-stroke (* p < 0.05, # p < 0.01, respectively n = 4) while not significantly different from sham operated animals (G–I) (p > 0.05) at respective time points. Bar graph (G) shows the percent volume quantification of the ipsilateral (stroked) hemisphere compared to the contralateral (non-stroked) hemisphere for each group.

Fig 5. HUCB cells rescue OLs of the external capsule following ischemic insult

O4 immunoreactivity was abundant throughout the ipsilateral external capsule of animals treated with HUCB cells 48 hrs post-MCAO (A, D). Vehicle (B) and sham-operated (E) controls also expressed O4, though immunoreactivity was sparsely distributed and less prominent compared to HUCB cell-treated animals. Quantification showed that HUCB cell treatment significantly increased O4 immunoreactivity relative to both vehicle-treated and sham-operated controls (* p < 0.01, n = 3). Scales bar = 50 µm. Arrows points to O4 positive staining.

Fig 6. HUCB cells increase white matter Uhmk1 expression following ischemic insult

HUCB cell treatment (A,B) 48 hrs post-MCAO significantly increased Uhmk1 expression in the ipsilateral hemisphere of the external capsule compared to vehicle (C,D) and sham-operated (E, F) controls (* p < 0.05, n = 3). Low magnification scale bars = 100 µm; high magnification inset scale bars = 20 µm.

Fig 7. HUCB cells alter white matter protein expression following ischemic insult

Photomicrographs show increased expression of Prdx4 (A), Mt3 (C), MOG (E) and Insig1 (G) in the ipsilateral hemisphere of animals treated with HUCB cells 48 hrs post-MCAO compared to vehicle-treated controls (B,D,F,H, respectively). No differences were observed in the expression of Tspan (I,J) or Vcan (K,L) in response to HUCB cell treatment. Scale bars = 50 µm. Arrows points to positive staining.

Fig 8. Immunohistochemical quantification of white matter protein expression

HUCB cell treatment 48 hrs post-MCAO resulted in increased expression of Uhmk1 (A), Prdx4 (B), Mt3 (C), MOG (D), and Insig1 (E) in the ipsilateral external capsule compared to vehicle-treated and sham-operated controls (*p < 0.05, #p < 0.01 n = 3). No significant differences were detected for Tspan2 (F) or Vcan (G).

Fig 9. Prdx4, Uhmk1, Insig1 and Mt3 colocalized with OL marker RIP

Photomicrographs depicts immunoflourescent double-labeling of OL specific antibody RIP (A, D, G, J) and antibodies generated against Prdx4 (B), Mt3 (K), Insig1 (E), and Uhmk1 (H). RIP and Insig1 are colocalized (F) in OL membranes, whereas Prdx4 (C), Mt3 (L), and Uhmk1 (I) are cytoplasmically localized. Scale bars = 50 µm. Arrows points to positive staining.

Fig 10. Prdx4 is expressed in astrocytes but not microglia/macrophages following ischemic insult

Double-label Immunohistochemistry for Prdx4 (A) and CD11b (B) shows that Prdx4 is not expressed in CD11b-positive microglia/macrophages (C) contained within the ipsilateral external capsule. Prdx4 (E) and GFAP (D) colocalization shows astrocytic expression of Prdx4 (F) within the white matter following ischemic insult. Scale bars = 100 µm.

Fig 11. Mt3, Uhmk, and Insig1 are not expressed in microglia/macrophages or astrocytes following ischemic insult

Immunofluorescent double-labeling shows that while expression is evident in the ipsilateral external capsule following MCAO, Mt3, Uhmk1, and Insig1 did not colocalize with CD11b-positive microglia/macrophages (A, C, E) or GFAP-positive astrocytes (B, D, F). Scale bars = 50 µm.