1 Introduction

From rest-to-exercise skeletal muscle metabolic rate may increase upto two orders of magnitude [1]. As rhythmic contractions are sustained primarily through oxidative metabolism the ability to distribute blood flow and deliver O2 and substrates [glucose and free fatty acids (FFA)] within and among active skeletal muscles is of paramount importance. Since the seminal microscopy studies of August Krogh nearly a century ago the notion of “capillary recruitment” has been invoked to explain increased blood-muscle O2 flux at the onset of contractions [2]. This notion is founded on the tacit presumption that many capillaries in resting muscle do not sustain red blood cell (RBC) flux and that these capillaries then initiate RBC flux with contractions.

Krogh used high pressure India ink perfusions of resting and contracting muscle to demonstrate that many more capillaries contained ink after contractions [2]. Obvious limitations of this procedure, which were acknowledged by Krogh, included vascular and myocyte distortion and the clumping together of India ink particles which may have preferentially blocked terminal arterioles in the resting condition when their caliber was the smallest. Despite these shortcomings, textbook authors [e.g., 3, 4] and contemporary researchers [e.g., 5–8] use these observations, and Krogh’s O2 diffusion model developed from them, to explain fundamental physiological phenomenon in resting and contracting muscles including: (1) Recruitment of additional capillary surface area. (2) Increased blood-muscle O2, glucose and FFA fluxes. (3) Decreased capillary-mitochondrial diffusion distances.

It is certainly logical that, if there were a substantial reserve capacity for blood-myocyte flux at rest, during exercise when the muscles’ energetic demands increased, that reserve would be recruited. However, evidence is presented herein that conflicts with the existent dogma that, in exercising muscle(s), increased blood-myocyte flux results solely or primarily from the initiation of RBC and plasma flow in previously non-flowing capillaries. Rather, the position is taken that, if RBC and plasma flow already exist in most capillaries in resting muscle (and much recent evidence indicates that this is so), increased blood-myocyte flux (i.e., phenomena 1–3 above) during exercise must occur within already-recruited capillaries. This position forces consideration of new models for blood-myocyte exchange in exercising muscle and reveals important clues into the mechanistic bases for microvascular dysfunction in chronic disease(s) (Fig. 1)

Fig. 1
figure 43_1_117031_1_En

Putative mechanisms that can account for the increase of blood-myocyte O2 (and substrate) flux, from rest-exercise, within existent RBC flowing capillaries without the necessity for “recruitment” of previously non-flowing capillaries. (1) Elevation of RBC numbers (hematocrit) within capillaries. (2) Substantial elevation of blood flow (RBC flux) within originally very low flow capillaries means that capillaries that were unimportant for blood-myocyte flux at rest become important during exercise. (3) Increased RBC velocity and greater fractional O2 extraction mean that additional capillary surface area along the length of the capillary becomes important for blood-myocyte flux. (4) There is the possibility that the endothelial cell surface layer (thought to be crucial in lowering capillary hematocrit below systemic) is modified under high flow conditions. (5) The precipitous fall in intracellular PO2 increases the blood-myocyte gradient and also desaturates myoglobin potentially elevating its ability to transport O2 and therefore intracellular diffusing capacity for O2

Full size image

Notwithstanding the above, the motto of The Royal Society “Nullius in Verba” (Take nobody’s word for it, see it for yourself) exhorts the scientist to challenge dogmatic adherence to ideologies especially when they conflict with rigorous scientific observations. It is pertinent that the most ardent supporters of “capillary recruitment” do not actually observe capillaries in their experiments [8] – a practice that has drawn acerbic criticism from eminent microscopists such as Eugene Renkin [9]. Instead, their conclusions are based upon assumptions that often regard a capillary as a simple switch that is either on or off [e.g., 6, 7].

It is the purpose of this review to present: (1) compelling direct evidence for RBC flux in the majority of capillaries in resting muscle – which precludes the possibility of substantial recruitment of previously non-flowing capillaries during exercise. (2) Recent NIRS evidence that supports little de novo capillary recruitment from rest-to-exercise. (3) Quantitative support that maintenance of RBC flux in the majority of resting muscle is feasible. (4) Putative explanations for apparently contrary evidence. (5) A novel model that accounts for increased blood-myocyte flux without the necessity for de novo capillary recruitment (Fig. 1).

2 RBC Flux in the Majority of Capillaries in Resting Muscle

In conscious rats at rest systemic indicator infusions reveal that almost all capillaries in the muscles examined (diaphragm, soleus, vastus lateralis) evidenced perfusion within 7 s [i.e., < 1 circulation, ref. 10]. This observation is supported by an extensive array of intravital microscopy studies of rat extensor digitorum longus [11], diaphragm [12] and spinotrapezius [13, 14] as well as hamster sartorius and cremaster [15], rabbit tenuissimus [16] and cat sartorius [17] muscles which demonstrate that > 80% of capillaries support RBC flux under resting conditions.

Intravital microscopy techniques have been criticized based upon the surgery required to visualize the muscle of interest, whether all capillaries within the optical field are visible and the necessity for anesthesia. Each of these concerns has been systematically countered as follows: Careful measurements in vasodilated [18] and contracting [14] muscles demonstrated that no more capillaries became visible under hyperemic conditions when additionally “recruited” capillaries would be obvious. Bailey et al. [19] showed that surgical preparation of the spinotrapezius muscle requisite for intravital microscopy did not alter blood flow (radiolabelled microspheres) or the precise matching of blood flow to metabolic rate (microvascular PO2) at rest or during contractions. Regarding anesthesia, the similarity of findings between conscious and anesthetized animals suggests that the majority of the capillaries in resting muscle support RBC flux [10–15, 17]. Additional evidence that smooth muscle function is preserved under anesthesia – which counters the argument that it is a loss of vasomotor tone which accounts for > 80% of capillaries supporting RBC flux – is that muscles of anesthetized animals preserve their blood flow to metabolism (O2 uptake) ratio of ∼6:1 [20] as found in animals and humans performing voluntary exercise [1].

3 NIRS Evidence Supports Little De novo Capillary Recruitment from Rest-Exercise

If one were to accept the notion of substantial capillary recruitment at exercise onset, for example from 15 to 90%, muscle hemoglobin concentration [Hb] would be expected to increase a minimum of 6-fold (and probably much more due to augmented capillary hematocrit). Near infrared spectroscopy (NIRS) interrogation of the exercising human quadriceps reveals that the actual increase of [Hb] is a very modest <30% [21] which may be achieved via augmented capillary hematocrit within capillaries that support RBC flux at rest [22] without de novo recruitment of capillaries.

4 RBC Flux in Most Resting Muscle Capillaries is Mathematically Feasible

The exemplar 70 kg human in whom 45% of body mass (31.5 kg) is muscle has an estimated muscle blood flow of ∼1 L/min (corresponding to 5.4 trillion RBC/min) [3, 4]. Accepting that capillary length and density vary among muscles 1,000 µm and 300/mm2 are reasonable mean values. If we consider that 80% [as in the rat, refs. 13, 14, 23] of these 8.9 billion muscle capillaries sustain RBC flux ∼12 RBC/capillary/s is calculated. This value compares very closely to that actually measured in the rat spinotrapezius muscle [i.e., 15–20 RBC/capillary/s, ref. 14] and suggests that maintenance of RBC flux in the majority of muscle capillaries at rest is not quantitatively unreasonable in humans.

5 Putative Explanations for Apparently Contrary Findings

Experimental evidence for “capillary recruitment” has largely avoided critical examination because of the overall belief in the concept. Although there are serious problems with the interpretation of histological studies [e.g., 8, 24] in the interests of brevity only intravital microscopy studies are dealt with here. Paramount among many important considerations are: (A) Anesthetized animals may be hypovolemic and hypotensive which invokes a reflex vasoconstriction in muscle. This arteriolar vasoconstiction can be extreme and reduce the proportion of flowing capillaries. When relieved by superfusion of vasodilatory substances there is the appearance of capillaries being recruited. (B) If surgical exteriorization is not performed extremely carefully, so as to avoid blunt trauma or stretching, the delicate capillaries can be damaged and RBC flux ceases [23]. (C) Mean O2 pressures within resting muscle are maintained at extremely low levels (< 20 mmHg). Experimental conditions that are hyperoxic relative to this cause reflex vasoconstriction which stops RBC flux in capillaries that would otherwise be flowing at rest. Subsequent lowering of the hyperoxic PO2 into the normal range (or below) then gives the appearance of capillary “recruitment”[5, 8].

6 Conclusions

In physiologically well-controlled, stable preparations that are not damaged, overly stretched or hyperoxic the vast majority of capillaries support RBC flux. Thus, in response to elevations of metabolic rate that occur with contractions the increased blood flow presents primarily as an elevated RBC flux (and hematocrit) within already flowing capillaries (i.e., the concept of longitudinal recruitment as shown in Fig. 1) [14, 25, 26]. In healthy muscle(s) of young individuals this process elevates convective and diffusive O2 delivery so rapidly that microvascular O2 pressures are maintained close to (or above) resting values for the first few seconds of contractions despite immediate and rapid increase in muscle O2 utilization [27]. This is crucial because if muscle O2 delivery is adequate, it places the site of limitation for O2 uptake dynamics within the mitochondrial respiratory chain [28]. In marked contrast to the situation in healthy muscles, chronic disease such as heart failure (and also diabetes) severely reduce the proportion of capillaries supporting RBC flux in muscle at rest and during contractions [29]. This scenario reduces both convective and diffusive O2 delivery and microvascular O2 pressures fall transiently far below those in healthy muscles [30]. As the speed of muscle O2 delivery becomes limiting both muscle and pulmonary gas exchange and their kinetics are impaired [28]. Without an appreciation of in vivo capillary hemodynamics, and the realization that recruitment of previously non-RBC flowing capillaries is not fundamental to increasing blood-myocyte O2 and substrate flux from rest-exercise, resolution of the mechanistic bases for microvascular/muscle dysfunction in disease is severely compromised.