Roles of individual prolyl-4-hydroxylase isoforms in the first 24 hours following transient focal cerebral ischaemia: insights from genetically modified mice

Abstract

This study investigated the function of each of the hypoxia inducible factor (HIF) prolyl-4-hydroxylase enzymes (PHD1–3) in the first 24 h following transient focal cerebral ischaemia by using mice with each isoform genetically suppressed. Male, 8- to 12-week old PHD1−/−, PHD2+/− and PHD3−/− mice and their wild-type (WT) littermate were subjected to 45 min of middle cerebral artery occlusion (MCAO). During the experiments, regional cerebral blood flow (rCBF) was recorded by laser Doppler flowmetry. Behaviour was assessed at both 2 h and 24 h after reperfusion with a common neuroscore. Infarct volumes, blood–brain barrier (BBB) disruption, cerebral vascular density, apoptosis, reactive oxygen species (ROS), HIF1α, and glycogen levels were then determined using histological and immunohistochemical techniques. When compared to their WT littermates, PHD2+/− mice had significantly increased cerebral microvascular density and more effective restoration of CBF upon reperfusion. PHD2+/− mice showed significantly better functional outcomes and higher activity rates at both 2 h and 24 h after MCAO, associated with significant fewer apoptotic cells in the penumbra and less BBB disruption; PHD3−/− mice had impaired rCBF upon early reperfusion but comparable functional outcomes; PHD1−/− mice did not show any significant changes following the MCAO. Production of ROS, HIF1α staining and glycogen content in the brain were not different in any comparison. Life-long genetic inhibition of PHD enzymes produces different effects on outcome in the first 24 h after transient cerebral ischaemia. These need to be considered in optimizing therapeutic effects of PHD inhibitors, particularly when isoform specific inhibitors become available.

Figure 1. Summary of rCBF during MCAO in PHD1^−/−, PHD2^+/− and PHD3^−/− mice and their WT littermates

Upon MCAO, rCBF dropped significantly below 30% of baseline levels. On reperfusion, the rCBF in PHD2^+/− mice returned faster to preischaemic levels than their WT littermates (B). In contrast, the rCBF in PHD3^−/− mice returned slower to baseline levels than their WT littermates (C) whereas there was no difference in rCBF return in PHD1^−/− mice and their WT littermates (A) (*P < 0.01, post hoc testing of rm-ANOVA).

Figure 2. Neuroscores assessed at both 2 and 24 h after MCAO in PHD1^−/−(A), PHD2^+/− (B), PHD3^−/− (C) mice and their WT littermates

There were no significant difference in neuroscore between PHD1^−/− and PHD3^−/− mice and their WT littermates at either 2 h or 24 h after reperfusion, but at both time points the PHD2^+/− mice had significantly better neuroscores than their WT littermates (median with IQR; *P < 0.05, Kruskall–Wallis test with Dunn's multiple comparison test).

Figure 3. Histological analysis of infarct volume and the volume of mouse IgG extravasation

The quantification of infarct volume using the Nissl and H&E stains was similar (A and B). Infarct volumes showed a trend to be smaller in PHD2^+/− mice (open bar) and PHD1^−/− mice (open bar) compared with the corresponding WT littermates (filled bars). In contrast, there was no difference in infarction volume between PHD3^−/− (open bar) and their WT littermates (filled bar) (D). Mouse IgG extravasation located in the ischaemic hemisphere (A–C). The volume of blood–brain barrier disruption assessed by the IgG extravasation was significantly smaller in the PHD2^+/− mice (open bar) than their WT littermate (filled bar) whereas this was not seen in either PHD1 or PHD3 comparisons (*P < 0.05, t test, E).

Figure 4. Apoptotic cell detection on mouse 6 μm brain sections

A and B, representative Nissl and ApopTag stains. The box in A indicates the area shown in B. Apoptotic cells were more prominent in the penumbra than in the ischaemic core (B). The ratio of TUNEL positive cells to DAPI positive nuclei in the penumbra of PHD2^+/− mice (open bar) was significantly higher than those of the WT mice (filled bar), while the ratio was not different between PHD1^−/−, PHD3^−/− and their WT littermates (*P < 0.05, t test, C).

Figure 5. CD31 immunostains of mouse brain sections

A, the number of CD31-positive structures with the morphology of capillaries was quantified in sections obtained from mice after 45 min MCAO and 24 h reperfusion. The PHD2^+/− mice (open bar) had significantly more CD31-positive structures than their WT littermates (filled bar), whereas this was not seen in either PHD1 or PHD3 comparisons (*P < 0.05, t test). B and C, representative CD31 immunohistochemistry images from PHD2^+/− (B) and wild-type control (C) mice, respectively.

Figure 6. HIF1α detection on 6 μm mouse brain sections from a wild-type animal ipsilateral hemisphere following MCAO (A), contralateral hemisphere following MCAO (B), ipsilateral hemisphere following sham operation (C), and contralateral hemisphere following sham operation (D)

At this time point the sections from PHD1^−/−, PHD2^+/− and PHD3^−/− animals showed similar staining pattern to those of the wild-type animals.