DIPG-33. CHARACTERIZING THE NEURO-VASCULAR UNIT IN DIFFUSE INTRINSIC PONTINE GLIOMA

Diffuse intrinsic pontine glioma (DIPG) is a childhood brainstem tumor with a median overall survival of eleven months. Lack of chemotherapy efficacy may be related to an intact blood-brain-barrier (BBB). In this study we aim to compare the neuro-vascular unit (NVU) of DIPG to healthy pons tissue. End-stage DIPG autopsy samples (n=5) and age-matched healthy pons samples (n=22), obtained from the NIH NeuroBioBank, were immunohistochemically stained for tight-junction proteins claudin-5 and zonula occludens-1 (ZO-1), basement membrane component laminin, and pericyte marker PDGFRβ. Claudin-5 stains were also used to determine vascular density and diameters. In DIPG, expression of claudin-5 and ZO-1 was reduced, and claudin-5 was dislocated to the abluminal side of endothelial cells. Laminin expression at the glia limitans was reduced in both pre-existent vessels and neovascular proliferation. In contrast to healthy pons, no PDGFRβ expression was detected. The number of blood vessels in DIPG was significantly reduced compared to healthy pons, 13.9±11.8/mm2 versus 26.3±14.2/mm2, respectively (P<0.01). Especially the number of smaller blood vessels (<10µm) was significantly lower (P<0.01). Distribution of larger blood vessels (≥10µm) did not differ between groups (P=0.223). Mean vascular diameter was 9.3±9.9µm for DIPG versus 7.7±9.0µm in healthy pons (P=0.016). Our study demonstrates evidence of structural changes in the NVU in end-stage DIPG. Chemotherapeutic inefficacy could be the result of reduced vascular density. However, further research is needed to determine meaning and extent of these changes and to determine whether these observations are caused by the tumor or the result of treatment.

Characterization of cathepsin S exosites that govern its elastolytic activity

We have previously determined that the elastolytic activities of cathepsins (Cat) K and V require two exosites sharing the same structural localization on both enzymes. The structural features involved in the elastolytic activity of CatS have not yet been identified. We first mutated the analogous CatK and V putative exosites of CatS into the elastolytically inactive CatL counterparts. The modification of the exosite 1 did not affect the elastase activity of CatS whilst mutation of the Y118 of exosite 2 decreased the cleavage of elastin by ∼70% without affecting the degradation of other macromolecular substrates (gelatin, thyroglobulin). T06, an ectosteric inhibitor that disrupt the elastolytic activity of CatK, blocked ∼80% of the elastolytic activity of CatS without blocking the cleavage of gelatin and thyroglobulin. Docking studies showed that T06 preferentially interacts with a binding site located on the Right domain of the enzyme, outside of the active site. The structural examination of this binding site showed that the loop spanning the L174N175G176K177 residues of CatS is considerably different from that of CatL. Mutation of this loop into the CatL-like equivalent decreased elastin degradation by ∼70% and adding the Y118 mutation brought down the loss of elastolysis to ∼80%. In addition, the Y118 mutation selectively reduced the cleavage of the basement membrane component laminin by ∼50%. In summary, our data show that the degradation of elastin by CatS requires two exosites where one of them is distinct from those of CatK and V whilst the cleavage of laminin requires only one exosite.

Pericytes promote selective vessel regression to regulate vascular patterning

Blood vessel networks form in a 2-step process of sprouting angiogenesis followed by selective branch regression and stabilization of remaining vessels. Pericytes are known to function in stabilizing blood vessels, but their role in vascular sprouting and selective vessel regression is poorly understood. The endosialin (CD248) receptor is expressed by pericytes associated with newly forming but not stable quiescent vessels. In the present study, we used the Endosialin−/− mouse as a means to uncover novel roles for pericytes during the process of vascular network formation. We demonstrate in a postnatal retina model that Endosialin−/− mice have normal vascular sprouting but are defective in selective vessel regression, leading to increased vessel density. Examination of the Endosialin−/− mouse tumor vasculature revealed an equivalent phenotype, indicating that pericytes perform a hitherto unidentified function to promote vessel destabilization and regression in vivo in both physiologic and pathologic angiogenesis. Mechanistically, Endosialin−/− mice have no defect in pericyte recruitment. Rather, endosialin binding to an endothelial associated, but not a pericyte associated, basement membrane component induces endothelial cell apoptosis and detachment. The results of the present study advance our understanding of pericyte biology and pericyte/endothelial cell cooperation during vascular patterning and have implications for the design of both pro- and antiangiogenic therapies.