Hemorrhage Caused by Snake Venom Metalloproteinases: A Journey of Discovery and Understanding

Abstract

The historical development of discoveries and conceptual frames for understanding the hemorrhagic activity induced by viperid snake venoms and by hemorrhagic metalloproteinases (SVMPs) present in these venoms is reviewed. Histological and ultrastructural tools allowed the identification of the capillary network as the main site of action of SVMPs. After years of debate, biochemical developments demonstrated that all hemorrhagic toxins in viperid venoms are zinc-dependent metalloproteinases. Hemorrhagic SVMPs act by initially hydrolyzing key substrates at the basement membrane (BM) of capillaries. This degradation results in the weakening of the mechanical stability of the capillary wall, which becomes distended owing of the action of the hemodynamic biophysical forces operating in the circulation. As a consequence, the capillary wall is disrupted and extravasation occurs. SVMPs do not induce rapid toxicity to endothelial cells, and the pathological effects described in these cells in vivo result from the mechanical action of these hemodynamic forces. Experimental evidence suggests that degradation of type IV collagen, and perhaps also perlecan, is the key event in the onset of microvessel damage. It is necessary to study this phenomenon from a holistic, systemic perspective in which the action of other venom components is also taken into consideration.

Keywords: basement membrane; capillary vessels; hemorrhage; metalloproteinases; snake venom; type IV collagen; viperids.

Figure 1

Pathological effects induced by a hemorrhagic snake venom metalloproteinase (SVMP) in capillary vessels. (A) Electron micrograph of a capillary vessel from muscle tissue injected with saline solution. Normal ultrastructure is observed in endothelial cell, including the presence of pynocytotic vesicles, and basement membrane (arrow). (B) Micrograph of a section of tissue injected with the hemorrhagic SVMP BaP1, from the venom of Bothrops asper. Notice prominent damage of endothelial cell, with loss of pynocytotic vesicles, distention and thinning of the cell, and rupture of cell integrity at one point (arrowhead). The basement membrane is absent along most of the periphery of the capillary (arrow). An erythrocyte (E) and a neutrophil (N) are observed inside the vessel. Magnification: 17,000× (A); and 10,000× (B). Reproduced by [47], Copyright 2006, Elsevier.

Figure 2

Two-step model to explain the mechanism of action of hemorrhagic SVMPs. The experimental evidence collected suggests that capillary vessel damage induced by hemorrhagic SVMPs occurs by a two-step mechanism. In the first step, SVMPs bind to and hydrolyze critical structural components of the basement membrane of capillary vessels, particularly type IV collagen and perlecan, and possibly other molecules that link the basement membrane to the fibrillar extracellular matrix. The cleavage of key peptide bonds of basement membrane components results in the mechanical weakening of this scaffold structure. As a consequence, in the second step, the biophysical hemodynamic forces normally operating in the microcirculation, i.e., hydrostatic pressure, which largely determines wall tension, and shear stress, induce a distention of the vessel wall, until the capillary is eventually disrupted, with the consequent extravasation of blood. Reproduced by [68], Copyright 2011, Elsevier.

Figure 3

Differential hydrolysis of basement membrane components in vivo by hemorrhagic and non-hemorrhagic snake venom metalloproteinases (SVMPs). Non-hemorrhagic SVMP (leucurolysin a-leuc) and hemorrhagic SVMP (BaP1) were injected in the muscle of mice. After 15 min, animals were euthanized, and tissue was collected and homogenized. Supernatants of homogenates were separated by SDS-PAGE, and transferred to nitrocellulose membranes for immunodetection with either anti-laminin (A); anti-nidogen (B); anti-type IV collagen (C); and anti-endorepelin (perlecan) (D) antibodies, and with anti-GAPDH as loading control. (C) Control muscle injected with saline solution. A chemiluminiscent substrate was used to detect the reactions. Densitometric analysis was then performed. The molecular mass of various markers (in kDa) is shown at the right of the gels. A clear difference is observed between these SVMPs in the patterns of degradation of type IV collagen and endorepelin (perlecan). Reproduced by [80], Copyright 2011, PLOS.

Figure 4

Immunolocalization of snake venom metalloproteinases (SVMPs) with vascular basement membrane on mouse cremaster muscle. Groups of mice were euthanized, and the cremaster muscle was dissected out. The isolated muscles were incubated for 15 min with equi-hemorrhagic amounts of three different SVMPs: BaP1 (PI SVMP, 30 µg), BlatH1 (PII SVMP, 3.5 µg) or CsH1 (PIII SVMP, 15 µg) labeled with Alexa Fluor^® 647 (blue). Control tissues were incubated with the SVMPs without labeling and no fluorescence was detected. Whole tissues were fixed with 4% paraformaldehyde and immunostained with anti-collagen IV following the secondary antibody labeled with Alexa Fluor 488 (green). Tissues were visualized in a Zeiss LSM 5 Pascal laser-scanning confocal microscope. Three-dimensional reconstitution of the images was carried out using IMARIS ×64 7.4.2 software. (A) Distribution of the SVMPs in the cremaster muscle tissue. Scale bar represents 150 µm. (B) Localization of SVMPs in capillary vessels in the cremaster. Scale bar represents 20 µm. White areas represent co-localization of the SVMPs (blue) with collagen IV (green) of vascular basement membrane in capillaries. Notice the predominant localization of PII and PIII SVMPs in the vasculature, whereas PI SVMP localizes in a more widespread fashion in the tissue. Reproduced by [85], Copyright 2015, PLOS.