Novel mouse endothelial cell surface marker is suppressed during differentiation of the blood brain barrier.

Few markers specific for mouse endothelium exist. We describe here one such marker, MECA-32, a monoclonal antibody which shows high specificity for mouse endothelium in both embryonic and mature tissues. The MECA-32 antigen has a M(r) of 50-55 x 10(3) under reducing conditions and M(r) of 100-120 x 10(3) under nonreducing conditions. It is expressed on most endothelial cells in the embryonic and in the adult mouse, with the exception of the brain, skeletal, and cardiac muscle, where it has a more restricted distribution. In skeletal and cardiac muscle only small arterioles and venules express the MECA-32 antigen, while in the brain its expression is negatively correlated with the differentiation of the vasculature to form the blood brain barrier. Interestingly, during embryonic development the antigen occurs on the brain vasculature up to day 16 of gestation (E16), whereupon it disappears. The embryonic brain is an avascular organ anlage which is vascularized by ingrowth of external blood vessels. Differentiation of the vasculature to form the blood brain barrier occurs at approximately E16 in the mouse. This differentiation correlates with the downregulation of MECA-32 antigen expression. Between E12 and E16 MECA-32 detects most endothelial cell surfaces of the blood vessels in the brain. No MECA-32 antigen is found in the brain at E17 or any later stage of development with the exception of the vasculature of the circumventricular organs. The results suggest that MECA-32 antigen expression is temporally and spatially correlated with the development of the blood brain barrier.

Transiently expressed, neural-specific molecule associated with premigratory granule cells in postnatal mouse cerebellum.

A rat monoclonal antibody (OZ42), raised against immature mouse granule cells, recognizes a region of the external granular layer of postnatally developing cerebellar cortex. This region, about three cells thick, is adjacent to the developing molecular layer and contains postmitotic, premigratory granule cells. The OZ42 reactivity commenced near postnatal day 3 (P3), the deep external granular layer was strongly reactive by P10 and this level was maintained while granule cells remained in the external granular layer (approximately P15). Isolated immature granule cells in cytospin preparations specifically reacted with OZ42. Reactivity was extranuclear and was substantially reduced when cells were prepared by trypsinization, suggesting that at least some of the antigen is associated with the outer surface of the plasma membrane. Other postnatal reactivity to OZ42 (P0 to P3) was found in a band of cells in the deep cortical layers overlying the corpus callosum through the entorhinal cortex, terminating adjacent to the hippocampus. Reactivity in some regions of the corpus callosum and anterior commissure was seen from P0 to P5. No reactivity of non-neural tissues was observed at any stage. In the embryo there was extensive staining of the CNS and PNS at E10 and E14, which was largely gone by E16. Weaver mutant mice examined for reactivity to OZ42 showed that the granule cell death and cerebellar disorganization in P10 homozygous mutants was associated with a substantial decrease in OZ42 reactivity in the external granular layer. At P14 and P20, OZ42 reactivity in the weaver external granular layer was restricted to single cells, rather than an entire layer of cells, further indicating that the OZ42 antigen is present on granule cells rather than the substratum. By Western analysis of non-reducing SDS-PAGE gels, OZ42 recognized a single band with the molecular weight between 120 and 145 kD in P10, but not adult cerebellum and BALB/c mice. An OZ42-specific band at 60-70 kD was also seen under reducing conditions and occasionally in non-reducing conditions. These bands were not recognized by antibodies against NCAM, L1 and AMOG. Immunoprecipitation and cross-blocking with antiserum to TAG-1 suggested that OZ42 recognized the same molecule in the mouse cerebellum that has been described in embryonic rat and mouse spinal cord. The developmentally regulated expression of the neural-specific molecule recognized by OZ42 in the postnatal cerebellum suggests it my be involved with the early stages of granule cell axon elongation.

The effect of valvulotomy on the flow rate through the saphenous vein graft: clinical implications.

Potential differences in flow rates between reversed and in situ saphenous vein bypass grafts were evaluated. One hundred ten greater saphenous vein segments containing isolated valves were examined with fiber-optic angioscopy during pulsatile and nonpulsatile flow. Valve competency was determined, and the degree of luminal obstruction caused by the valve during reversed flow was calculated with caliper measurements of the video image. Flow measurements were obtained before and after valvulotomy, in reversed and nonreversed vein orientations. Increased flow rates occurred during pulsatile irrigation only, after valvulotomy in vein segments with diameters less than 2.5 mm (p less than 0.001, Bonferroni t test). In these small-diameter vein segments, the flow rate in reversed valve-intact vein was 94.4 +/- 28.9 ml/min (mean +/- 1 standard deviation), the flow rate in reversed valve-disrupted vein was 136.4 +/- 36.5 ml/min, and the flow rate in nonreversed valve-disrupted vein was 137.8 +/- 31.3 ml/min. In 22 vein segments, luminal obstruction caused by the intact valve was measured angioscopically. A small valve orifice was found to be related to a large increase in flow rate after valvulotomy (p less than 0.02, least-squares regression). In addition, veins with diameters less than 2.5 mm have significantly smaller valve orifices compared with veins with diameters greater than 2.5 mm. These results present important clinical implications as the number of distal extremity reconstructions increases.

Independent developmental regulation of migratory granule neurons and their cerebellar ligand in the mouse.

Interactions between migratory granule neurons and the developing molecular layer of the mouse cerebellum were examined using an in situ binding assay. Single cell suspensions of postnatal granule neurons specifically adhere to unfixed frozen cerebellar tissue sections. We investigated the influence of postnatal age of granule neurons and of tissue on this interaction. Granule neurons from P10 (the time of peak migratory activity) bind preferentially to the molecular layer. Premigratory granule neurons, P5, do not bind age-matched cerebellar tissue. Postmigratory granule neurons, P14 and older, adhere to the molecular and internal granular layers of age-matched and older cerebellar tissue but not to younger tissue. These binding patterns are most simply explained as a single receptor-ligand system in which both elements exhibit independent developmental regulation. Although granule neurons lose the ability to bind with increasing age, the molecular layer ligand retains its capacity for this interaction into adulthood, long after normal migration has ceased.