Erythropoietin Modulates Cerebral and Serum Degradation Products from Excess Calpain Activation following Prenatal Hypoxia-Ischemia

Abstract

Preterm infants suffer central nervous system (CNS) injury from hypoxia-ischemia and inflammation - termed encephalopathy of prematurity. Mature CNS injury activates caspase and calpain proteases. Erythropoietin (EPO) limits apoptosis mediated by activated caspases, but its role in modulating calpain activation has not yet been investigated extensively following injury to the developing CNS. We hypothesized that excess calpain activation degrades developmentally regulated molecules essential for CNS circuit formation, myelination and axon integrity, including neuronal potassium-chloride co-transporter (KCC2), myelin basic protein (MBP) and phosphorylated neurofilament (pNF), respectively. Further, we predicted that post-injury EPO treatment could mitigate CNS calpain-mediated degradation. Using prenatal transient systemic hypoxia-ischemia (TSHI) in rats to mimic CNS injury from extreme preterm birth, and postnatal EPO treatment with a clinically relevant dosing regimen, we found sustained postnatal excess cortical calpain activation following prenatal TSHI, as shown by the cleavage of alpha II-spectrin (αII-spectrin) into 145-kDa αII-spectrin degradation products (αII-SDPs) and p35 into p25. Postnatal expression of the endogenous calpain inhibitor calpastatin was also reduced following prenatal TSHI. Calpain substrate expression following TSHI, including cortical KCC2, MBP and NF, was modulated by postnatal EPO treatment. Calpain activation was reflected in serum levels of αII-SDPs and KCC2 fragments, and notably, EPO treatment also modulated KCC2 fragment levels. Together, these data indicate that excess calpain activity contributes to the pathogenesis of encephalopathy of prematurity. Serum biomarkers of calpain activation may detect ongoing cerebral injury and responsiveness to EPO or similar neuroprotective strategies.

Figure 1

Summary of experimental design and major data points. Transient systemic hypoxia-ischemia (TSHI) was performed on embryonic day18 via laparotomy with 60 min of uterine artery occlusion. Pups were born at term (E22). TSHI pups were randomized to receive either EPO (2000 U/kg/dose) or vehicle daily from postnatal day 1 (P1) to P5. Reds dots indicate when samples were collected from serum for quantification, while gray dots represent samples collected from frontal lobes (prior to postnatal day 5), and cortex and white matter (older than P5) for quantification. “p” denotes previously published data.

Figure 2

Calpain digestions of cortex and white matter. A. Representative calpain digestion of P7 sham cortex incubated with αII-spectrin antibodies reveals a primary band at 250 kDa, and a fainter band at 150 kDa. In the presence of calcium-activated µ-calpain, the αII-spectrin is digested into a faint 150 kDa, a heavy 145 kDa and a very faint 120 kDa band. B. Sham P7 cortex digested with calpain shows elimination of the p35 band, with a corresponding increase in the p25 fragment. C. Calpain digestion of P11 sham cortex assayed alongside P11 TSHI cortex probed with αII-spectrin shows a 150/145 kDa band that is similar to a band present in sham cortex digested with activated calpain, but not present in sham cortex without calpain digestion. D. Calpain digestion of P11 sham white matter in the presence of calcium eliminates MBP, while P11 TSHI white matter assayed in an adjacent lane shows reduced expression. E. Calpain digestion of P11 sham white matter in the presence of calcium eliminates pNF and NF, while P11 TSHI white matter assayed in an adjacent lane shows reduced pNF and NF expression.

Figure 3

Excess calpain activity is induced by prenatal TSHI and modulated by EPO treatment. A. Representative western blot of αII-SDPs, and ratio of cleaved to full-length cerebral αII-SDPs at 6 hours (E18+6), 24 hours (E19), 96 hours (P0) and 6 days (P2) after prenatal TSHI compared to sham. A significant increase in αII-SDPs is present at P2. B. The ratio of p25/p35 is increased at P2 following prenatal TSHI. C. Following prenatal TSHI, the ratio of cleaved to full length αII-SDP at P11 is elevated, compared to shams. Postnatal treatment with EPO normalizes the ratio to sham levels (two-way ANOVA with Bonferroni correction). *p<0.05, **p<0.01, ***p≤0.001.

Figure 4

Prenatal TSHI reduces calpastatin expression. A. At P2 immunolabeling with calpastatin antibodies shows loss of prominent cell labeling in the white matter, subplate and cortex following TSHI, compared to sham. By contrast, immunolabeling with µ-calpain antibodies is increased after TSHI. Bar = 20 µm. B. Cortical calpastatin expression is reduced at P7 following prenatal TSHI compared to shams. C. Following prenatal TSHI, calpastatin expression is reduced at P11 compared to shams. Postnatal treatment with EPO restores calpastatin expression to sham levels (two-way ANOVA with Bonferroni correction). *p<0.05, **p<0.01, ***p≤0.001.

Figure 5

Calpain degrades molecules essential for neurodevelopment including KCC2, MBP and NF. A. At P11 KCC2 oligomer expression is reduced following prenatal TSHI, and KCC2 levels are restored following postnatal EPO treatment. B. At P11 in the midst of myelin development, MBP expression is reduced following prenatal TSHI and normalizes following EPO treatment. C. Similarly, at P11 the ratio of pNF/NF is lower following prenatal TSHI, and restored following EPO treatment. D. At P28 in juvenile white matter, pNF/NF ratio at P28 is also reduced following prenatal TSHI, and restored with postnatal EPO treatment. E. Likewise, the white matter MBP levels remain lower following prenatal TSHI, and normalize with EPO treatment (two-way ANOVA with Bonferroni correction). *p<0.05, **p<0.01, ***p≤0.001.

Figure 6

Calpain degradation products are present in serum. Serum levels of 145 kDa αII-SDP are elevated at E19 (A), P0 (B) and P2 (C) (unpaired two-tailed t test). Coomassie staining is used for the loading control. D At P11 serum levels of the 90 kDa KCC2 calpain fragment are elevated following prenatal TSHI, and normalized with postnatal EPO treatment (two-way ANOVA with Bonferroni correction). *p<0.05, **p<0.01, ***p≤0.001.