Table 2.

Basement Membrane Coverage of Endothelial Sprouts on Tumor Blood Vessels^*

	RIP-Tag2 pancreatic islet tumors	MCa-IV mammary carcinomas	Lewis lung carcinomas
Length of CD31-immunoreactive endothelial sprouts (range)	13 ± 1 μm (5–35 μm)	32 ± 5 μm (11–61 μm)	20 ± 4 μm (5–49 μm)
Length of type IV collagen-immunoreactive sleeve on endothelial sprouts (range)	23 ± 2 μm (7–41 μm)	54 ± 9 μm (14–90 μm)	30 ± 5 μm (9–75 μm)
Length of portion of type IV collagen-immunoreactive sleeve beyond tip of CD31-staining	10 ± 3 μm	22 ± 3 μm	10 ± 3 μm
Length of type IV collagen-immunoreactive sleeve on endothelial sprouts expressed as % of length of CD31-stained sprout	175 ± 18%	170 ± 7%	153 ± 18%
Length of pericyte processes on sprouts expressed as % of length of CD31-stained sprout†	180%	162%	138%
Percent of sprouts with naked tips (length of naked tips of sprouts)	5% (4.8 μm)	5% (5.5 μm)	5% (4.2 μm)

*Endothelial sprouts, identified as thin, tapered, blind-ended CD31-immunoreactive projections away from the vascular axis, were analyzed in 80-μm sections of tumors double-stained for type IV collagen and CD31 immunoreactivities. Naked tips represent the distal-most region of CD31 staining of sprouts with no detectable type IV collagen immunoreactivity. No sprouts were identified on blood vessels of normal pancreatic islets. Values are expressed as means ± SE (n = 4 mice per group). Ten sprouts were examined in each mouse. Ranges are values for 40 sprouts per group.

^†Pericyte data from Morikawa et al. (2002). 10