Analysis of capillary geometry in rat subepicardium and subendocardium

The sustained high-energy turnover of cardiac muscle presents a formidable challenge to the O2 delivery systems. One major determinant of blood-tissue gas exchange potential is capillary surface area per volume of muscle fiber, Sv(c,f). Estimation of Sv(c,f) necessitates quantification of capillary orientation. Capillary geometry was analyzed systematically in subepi- (epi) and subendocardium (endo) of glutaraldehyde perfusion-fixed rat heart (n = 4). On 1-micron sections cut rigorously transverse and longitudinal to the muscle fiber axis we determined capillary number per fiber square millimeter on transverse, QA(0), and longitudinal, QA(pi/2), sections, capillary diameter, d(c), fiber cross-sectional area, a(f), and sarcomere length, l. Sv(c,f) was computed as pi.d(c).Jv(c,f), where Jv(c,f) is capillary length per fiber volume determined on the basis of a directional distribution model of capillary segments (Fisher axial). Analysis of capillary density, QA(alpha), in sections taken at angles alpha [from 0 to 90 degrees (pi/2) to fiber axis] showed that the Fisher axial distribution provides a good fit to capillary segment orientation in cardiac muscle. No systematic difference was found in fiber size (epi = 269.7 +/- 28.9, endo = 283.8 +/- 16.3 microns 2), capillary diameter (epi = 4.9 +/- 0.3; endo = 4.5 +/- 0.2 microns), Jv(c,f) (epi = 6,302 +/- 558; endo = 5,957 +/- 492 mm-2), or capillary surface per volume of muscle fiber (epi = 968.1 +/- 76.5; endo = 838.2 +/- 93.0 cm-1) between epi and endo. Contribution of capillary tortuosity and branching to Jv(c,f) ranged from 6-27% (epi) and 8-21% (endo) over the small ranges of l considered (epi = 2.09-2.23; endo = 2.04-2.17 microns).(ABSTRACT TRUNCATED AT 250 WORDS)