Real-time imaging of de novo arteriovenous malformation in a mouse model of hereditary hemorrhagic telangiectasia

Abstract

Arteriovenous malformations (AVMs) are vascular anomalies where arteries and veins are directly connected through a complex, tangled web of abnormal arteries and veins instead of a normal capillary network. AVMs in the brain, lung, and visceral organs, including the liver and gastrointestinal tract, result in considerable morbidity and mortality. AVMs are the underlying cause of three major clinical symptoms of a genetic vascular dysplasia termed hereditary hemorrhagic telangiectasia (HHT), which is characterized by recurrent nosebleeds, mucocutaneous telangiectases, and visceral AVMs and caused by mutations in one of several genes, including activin receptor-like kinase 1 (ALK1). It remains unknown why and how selective blood vessels form AVMs, and there have been technical limitations to observing the initial stages of AVM formation. Here we present in vivo evidence that physiological or environmental factors such as wounds in addition to the genetic ablation are required for Alk1-deficient vessels to develop to AVMs in adult mice. Using the dorsal skinfold window chamber system, we have demonstrated for what we believe to be the first time the entire course of AVM formation in subdermal blood vessels by using intravital bright-field images, hyperspectral imaging, fluorescence recordings of direct arterial flow through the AV shunts, and vascular casting techniques. We believe our data provide novel insights into the pathogenetic mechanisms of HHT and potential therapeutic approaches.

Figures

Figure 1. Endothelial Alk1 deletion from the Alk12loxP allele resulted in postnatal lethality by P5, with AVMs in the brain, lung, and GI tract.

(A–E) Dissection microscopic views of vascular images of control [L1Cre(–);Alk12loxP/2loxP; A and C] and mutant [L1Cre(+);Alk12loxP/2loxP; B, D, and E] P3 mouse brains with latex dye injected into the left ventricle of the heart. (D and E) Magnified views of blood vessels in the hippocampal area. (E) Asterisks indicate peculiar looping of vessels at the distal tips of arteries shunting to veins. A, artery; V, vein. (F–K) Vascular morphology of control (F, H, and J) and mutant (G, I, and K) P3 lungs. (F and G) Dissection microscopic view of pulmonary veins. Double-headed arrows indicate diameter of pulmonary veins. A dashed circle highlights the area of abnormal tangled vessels, and an arrow indicates looping of a dilated vessel (G). (H and I) Gross views of left lung after Evans blue dye injection via jugular veins. Note bumpy, dilated looping vessels in the mutant lungs (I). (J and K) Scanning electron microgram view of corrosion cast, showing lack of microvessels and presence of numerous dilated looping vessels (asterisks) in the mutant lungs (K). (L–Q) Gross views of small intestines of control (L, N, and P) and mutant (M, O, and Q) P3 mice. AV shunts and abnormal anastomosis was observed by latex dye (O) and corrosion cast (indicated by an arrow with asterisk; Q). Scale bars: 2 mm (A and B); 1 mm (C, D, H, I, and N–Q); 300 μm (E); 500 μm (F, G, L, and M).

Figure 2. Global Alk1 deletion in adult stages resulted in lethality with hemorrhages in the lung and GI tract.

(A) Kaplan-Meier survival curve, showing lethality from Alk1 deletion in male and female adult mice. Female mice were more sensitive to Alk1 deletion than males. Control: R26+/+Alk12loxP/2loxP; mutant: R26+/CreERAlk12loxP/2loxP. (B–D) Mutant mice displayed external signs of internal bleeding and anemia, such as pale paws (B), a low pO2 level (C), and darkened feces (D) by 8 days after TM injection. (E) Reduced hematocrit in the mutants 8 days after TM injection. Error bars in C and E indicate SD. (F–H) Superficial blood vessels in the small intestine of the mutant 8 day after TM injecting were visualized by latex dye. Formation of AVM judged by the presence of latex dye in both arteries and veins was apparent in the areas of Peyer’s patches (G). Scale bars: 4 mm (F); 2 mm (G and H). *P < 0.05; **P < 0.001.

Figure 3. Wounding can induce de novo AVM formation in Alk1-deleted adult mice.

Vascular patterns shown by latex dye injected into the left heart of control (R26+/+Alk12loxP/2loxP; A, C, and E) and mutant (R26+/CreERAlk12loxP/2loxP; B, D, and F) mice bearing wounds in the ear (A and B) or dorsal skin (C–F), 8 days after TM injection. The images were taken after clearing in organic solvents. The center of the wound is indicated by an asterisk. Note that only mutant mice developed AV shunts, shown by the presence of latex dye in both arteries and veins. AV shunting and abnormal vascular morphologies were apparent only in the wound areas. Blood vessels away from the wound, indicated by arrows with asterisks (B and D), showed normal appearance. Inset in D shows a magnified view of AV fistulas formed in the rim area of the mutant wound. Arteries and veins are marked by red and light blue lines, respectively, and the AV anastomoses by yellow lines. (E and F) The images in C and D were processed to quantify the area of blood vessels containing the latex dye in the given area. All figure panels are at the same magnification. Scale bars: 5 mm (A); 2 mm (insets in A and B); 1 mm (inset in D).

Figure 4. Intravital images of vascular morphogenesis responding to wound.

The wound was induced and TM given on day 0. The center of the wound is indicated by an asterisk. Representative bright-field images of blood vessels of a control (B) and a mutant (C) mouse during wound healing for 8 days. Corresponding pseudocolor images of Hb(O2) saturation levels on the blood vessels of a control (A) and a mutant (D) mouse are shown. Arteries are distinguished from veins by their thin morphology and a high level of Hb saturation (red) at day 1. While such distinctions in major branches remained through 8 days of wound healing in controls (A and B), arteries and veins were almost indistinguishable at days 7–8 in mutant mice due to extensive remodeling and AV shunting (C and D). The color code to the right of D represents the percentage of Hb(O2) saturation [Hb(O2)sat] in the blood vessels. Scale bar (D, day 8): 1 mm.

Figure 5. Vascular remodeling during development of wound-induced de novo AVM.

A light-microscopic view and the corresponding Hb(O2) pseudocolor map of vascular morphogenesis during wound healing for 8 days in the mutant. Arteries are readily distinguishable from veins by high O2 levels days 1–5 after wounding (A–D). Early stages of AV shunts (dotted circles in B and D) were detected as early as day 3. White arrows in B, C, and E indicate veins containing high Hb(O2) levels. Blue arrows in C and F indicate veins at downstream of the AV shunts, showing dilated vascular morphologies. By day 6 (E–G), Hb(O2) levels in the veins draining from the AV shunts are as high as those in the arteries feeding to AV shunts. It is noteworthy that there were signs of bleeding where AV shunts are established (red arrows in B–D). Ratio of Hb(O2) saturation of veins and arteries in the mutants was elevated as wound healing progressed, while that in control mice remained unchanged (H). Error bars in H indicate SD. Scale bar in A: 1 mm.

Figure 6. Blood flow through patent AV fistulas.

Vascular images by latex dye injection via left heart of control (A) and mutant (B) mice demonstrate AV shunts in the mutants where latex dye was found in both arteries and veins. (C) Direction of blood flow obtained from video recording is indicated by arrows. Yellow asterisks indicate an area of vein at the receiving end of AV shunts showing dilated and tortuous vessel morphology. (D and E) Fluorescence images of blood flow in 2 AVM areas labeled as D and E in C. Black arrows in D indicate the direction of blood flow. Arrowheads indicate an AVM vessels showing leakage of blood. Yellow asterisks in E indicate a venous area having turbulent blood flow and corresponding vessel tortuosity. The blue arrows in D and E indicate the same area of blood vessel. See Supplemental Videos 3 and 4. Scale bars: 1 mm (A and B); 0.5 mm (C); 100 μm (D); 200 μm (E).