Point: There is capillary recruitment in active skeletal muscle during exercise.

Focus. The key issue here is whether resting skeletal muscle is fully perfused. We will make the case that it is not, and thus there exists a reserve of unperfused capillaries (capillary reserve) that are recruited to carry flow not only by muscle contraction, but also by insulin. Our background. Our interest in this area arose from efforts to develop noninvasive methods to assess whether insulin affects muscle microvascular perfusion in vivo and whether muscle insulin resistance is at least in part due to impaired perfusion that diminishes insulin and glucose delivery. Since it is well accepted that bulk blood flow to working muscle increases, it was logical to apply our methods to compare exercise and insulin in terms of capillary recruitment. We cite findings obtained using five different approaches that each provides evidence of capillary recruitment. Three of these are entirely noninvasive, the other two involve surgical preparation that may appreciably alter “basal” perfusion. Exercise increases red blood cell occupancy of capillaries in muscle biopsies. Honig and colleagues (8) examined cryosections of denervated dog gracilis muscle and reported that at rest only one-third of the capillaries were perfused with erythrocytes. Muscle contraction (4/min) effectively doubled the number of capillaries containing red blood cells without increasing bulk flow. Near maximal recruitment occurred at frequencies 8 Hz and bulk flow increased dramatically. Certainly vessel tortuosity (with a single vessel crossing a section several times) could have led to underestimation and, conversely, denervation (with resultant high basal flow rates) to overestimulation of basal perfusion. Intriguingly, these observations supported those of Krogh (11) who first proposed a reserve of unperfused capillaries in muscle more than half a century earlier—a cornerstone of the work for which he was awarded the 1919 Nobel Prize. Exercise increased red blood cell movement in exposed muscle. Intravital microscopy of surgically exposed transparent muscle of anesthetized animals has provided evidence for (6, 7, 10, 12, 16, 18) and against capillary recruitment (see accompanying “counterpoint”). The study by Lindbom (12) viewed the rabbit tenuissimus muscle using intravital microscopy and noted that one-third to one-half of all capillaries were perfused based on red blood cell movement and these were homogeneously distributed. Electrical stimulation increased perfusion of muscle and not connective tissue capillaries. The distribution of flow between muscle and connective tissue was influenced by the oxygen tension of the superfusing solution. Thus, as the PO2 increased, muscle capillary flow decreased and, conversely, when the superfusate PO2 decreased, more capillaries were perfused. The dependence of capillary flow in such preparations on PO2 was noted by others earlier (13) and confirmed recently (17). These findings both support the presence of a capillary reserve and emphasize the sensitivity of exposed muscle preparation vasculature to environmental factors. Each of the above two methods are quite invasive, requiring general anesthesia, surgical exposure of the muscle, and, in some preparations, denervation. This makes it particularly difficult to obtain an estimate of true “basal” perfusion, which is required for any estimate of recruitment in response to exercise or other stimuli. An increase in metabolism of the exogenous marker substrate 1-methylxanthine (1-MX). This method was developed in our laboratory (15) and measures the metabolism of 1-MX as it traverses the muscle vascular network from artery to vein. Others have shown that xanthine oxidase, which is responsible for metabolizing 1-MX in muscle, is located predominantly in the capillary endothelium (9). The method has been validated in the pump-perfused isolated rat hindlimb, where the proportion of total flow that is nutritive can be manipulated (1). With this preparation, 1-MX metabolism is proportional to the volume of nutritive flow, is increased during muscle contraction (22) and decreased when nutritive flow was decreased pharmacologically (14). Applying this method in vivo to rats, we showed that field stimulation of muscle in vivo to simulate exercise increased the metabolism of 1-MX, reinforcing the above issues (4). From a number of studies it is clear that 1-MX metabolism, and thus capillary recruitment, is not dependent on bulk blood flow (for review, see Ref. 2 and references therein). Interestingly, we showed that physiological insulin also increased 1-MX metabolism, suggesting that insulin increased microvascular perfusion (15) and, by default, indicating that prior to insulin the muscles were not fully perfused. This method does not require muscle biopsies or surgical exposure and might therefore be expected to provide a more physiological “basal” environment against which to assess the impact of recruitment. Contrast-enhanced ultrasound (CEU) imaging of the muscle microvasculature. With this method an acoustic signal is obtained from intravenously infused microbubbles of inert gas as they track with erythrocytes through the vasculature. The microbubbles are smaller than red blood cells and do not change blood flow rheology. We adapted the method for use with skeletal muscle based on that described by Wei et al. (21), who validated the method to trace myocardial blood perfusion. The acoustic properties of microbubbles results in their bursting when impacted by a high energy ultrasound pulse and simultaneously emitting a signal. By introducing a variable time interval between pulses, a series of images is collected and the contributions of tissue density and large, rapidly filling vessels (arteries, arterioles, veins, and venules) subtracted from the integrated video intensity in a region of interest. From this, a replenishment curve is generated that describes the refilling rate and the volume of microvasculature filled by microbubbles. This method has the advantages of being entirely noninvasive, requiring no anesthesia, and it can be used in humans performing voluntary exercise (as opposed to electrical stimulation). As a result, this method is better suited than the others mentioned above for obtaining a true measure of basal microvascular perfusion and hence to assess increments above basal. J Appl Physiol 104: 889–894, 2008; doi:10.1152/japplphysiol.00779.2007.