Combinatorial interaction between CCM pathway genes precipitates hemorrhagic stroke

Abstract

Intracranial hemorrhage (ICH) is a particularly severe form of stroke whose etiology remains poorly understood, with a highly variable appearance and onset of the disease (Felbor et al., 2006; Frizzell, 2005; Lucas et al., 2003). In humans, mutations in any one of three CCM genes causes an autosomal dominant genetic ICH disorder characterized by cerebral cavernous malformations (CCM). Recent evidence highlighting multiple interactions between the three CCM gene products and other proteins regulating endothelial junctional integrity suggests that minor deficits in these other proteins could potentially predispose to, or help to initiate, CCM, and that combinations of otherwise silent genetic deficits in both the CCM and interacting proteins might explain some of the variability in penetrance and expressivity of human ICH disorders. Here, we test this idea by combined knockdown of CCM pathway genes in zebrafish. Reducing the function of rap1b, which encodes a Ras GTPase effector protein for CCM1/Krit1, disrupts endothelial junctions in vivo and in vitro, showing it is a crucial player in the CCM pathway. Importantly, a minor reduction of Rap1b in combination with similar reductions in the products of other CCM pathway genes results in a high incidence of ICH. These findings support the idea that minor polygenic deficits in the CCM pathway can strongly synergize to initiate ICH.

Fig. 1

Rap1b is required for vascular integrity in vivo. (A,B) Whole-mount in situ hybridization of 24 hpf zebrafish using probes for rap1b (A) and vecdn (VE-cadherin) (B). (C-F) Transmitted light images of the heads of animals injected with 8 nanograms of control (C,D) or rap1b (E,F) morpholino (MO). ICH in the rap1b MO-injected animals (black arrows) (E,F). (G) Quantitation of the percentage of animals developing ICH by 48 hpf after injection of: (1) control MO, (2) rap1b MO, (3) control MO + hRap1b mRNA, (4) rap1b MO + hRap1b mRNA, (5) control MO + fli1:hRap1b DNA, (6) rap1b MO + fli1:hRap1b DNA, (7) rap1b MO + fli1:hRap1b DNA. The amount of MO (in nanograms) and DNA and RNA constructs (in picograms) injected per animal, and the total number of embryos injected and scored (N #) are shown. Anterior is to the left, and dorsal is either above (A-C,E) or coming out of the plane of the page (D,F). Bars, 250 μM.

Fig. 2

Rap 1b is required for proper endothelial junction formation in vivo and in vitro. (A-D) Transmission electron micrographs of endothelial cell-cell junctions in control (A) and rap1b morpholino-injected (B-D) animals. Normal junctions (arrowhead) between endothelial cells in control morphants (A). Poorly formed junctions (arrowheads) between endothelial cells in rap1b morphants (B,C). Gaps between endothelial cells (arrow) are also frequently observed in rap1b morphants (D). (E) Transwell permeability across a HUVEC endothelial monolayer. (F-I) Immunohistochemical staining of control (F,H) or Rap1b siRNA-treated (G,I) HUVEC monolayers, probed for VE-cadherin (F,G) or beta-catenin (H,I). Arrows in (G) and (I) indicate gaps and disruption of junctional organization of markers in Rap1b siRNA-treated monolayers. Bars, 500 nm (A-C), 2 μm (D) and 20 μM (F-I).

Fig. 3

Synergistic effects of defects in rap1b and downstream CCM pathway genes pak2a and β-pix. (A) Quantitation of the percentage of animals developing ICH by 48 hpf after injection of high or low doses of rap1b, pak2a, β-pix or control morpholinos (MO). (B) Quantitation of the percentage of animals developing ICH by 48 hpf following injection of rap1b MO, control MO + fli1:pak2a DNA or rap1b MO + fli1:pak2a DNA. The amount of MO and/or DNA construct injected per animal (in nanograms) and the number of embryos injected and scored (N #) are noted.

Fig. 4

Synergistic effects of defects in rap1b and upstream CCM genes. Quantitation of the percentage of animals developing enlarged hearts and loss of circulation (orange bars) or ICH (blue bars) by 48 hpf after injection of high or low doses of rap1b, ccm1 or control morpholinos (MO). The amount of MO injected per animal (in nanograms) and the number of embryos injected and scored (N #) are noted.