Control of microvascular oxygen pressures in rat muscles comprised of different fibre types

Abstract

In response to an elevated metabolic rate , increased microvascular blood–muscle O₂ flux is the product of both augmented O₂ delivery and fractional O₂ extraction. Whole body and exercising limb measurements demonstrate that and fractional O₂ extraction increase as linear and hyperbolic functions, respectively, of . Given the presence of disparate vascular control mechanisms among different muscle fibre types, we tested the hypothesis that, in response to muscle contractions, would be lower and fractional O₂ extraction (as assessed via microvascular O₂ pressure, P_mvO2) higher in fast- versus slow-twitch muscles. Radiolabelled microsphere and phosphorescence quenching techniques were used to measure and P_mvO2, respectively at rest and across the transition to 1 Hz twitch contractions at low (Lo, 2.5 V) and high intensities (Hi, 4.5 V) in rat (n = 20) soleus (Sol, slow-twitch, type I), mixed gastrocnemius (MG, fast-twitch, type IIa) and white gastrocnemius (WG, fast-twitch, type IIb) muscle. At rest and for Lo and Hi (steady-state values) transitions, P_mvO2 was lower (all P < 0.05) in MG (mmHg: rest, 22.5 ± 1.0; Lo, 15.3 ± 1.3; Hi, 10.2 ± 1.6) and WG (mmHg: rest, 19.0 ± 1.3; Lo, 12.2 ± 1.1; Hi, 9.9 ± 1.1) than in Sol (rest, 33.1 ± 3.2 mmHg; Lo, 19.0 ± 2.3 mmHg; Hi, 18.7 ± 1.8 mmHg), despite lower and in MG and WG under each set of conditions. These data suggest that during submaximal metabolic rates, the relationship between and O₂ extraction is dependent on fibre type (at least in the muscles studied herein), such that muscles comprised of fast-twitch fibres display a greater fractional O₂ extraction (i.e. lower P_mvO2) than their slow-twitch counterparts. These results also indicate that the greater sustained P_mvO2 in Sol may be important for ensuring high blood–myocyte O₂ flux and therefore a greater oxidative contribution to energetic requirements.

The relationship between blood flow and metabolic rate is strong and has been well characterized for both the whole body and the exercising limbs. In the exercising human, leg and are linearly related by the equation , where S represents the slope (5.3) and I the intercept (2.8 l min⁻¹; data from Knight et al. 1992). This equation stipulates that fractional O₂ extraction (or arteriovenous O₂ difference, ) must increase as a hyperbolic function of according to the approximation equation, (Knight et al. 1992; Poole, 1997; see Whipp & Ward, 1982 for original derivation of this equation). However, human leg muscles are comprised of a complex mosaic of type I and type II fibres and, to date, methodological limitations have precluded determination of whether the (or O₂ delivery, )-to- relationship differs between the different muscle fibre types. This information is important because the -to- ratio determines the microvascular O₂ pressure (P_mvO2), which in turn facilitates blood–tissue O₂ flux and thus strongly impacts cellular metabolic regulation.

Free of the methodological limitations noted above, animal studies indicate that during both rest and exercise there exists a pronounced stratification of across different muscles and muscle fibre types (e.g. Armstrong & Laughlin, 1983; Poole et al. 2000; reviewed by Laughlin et al. 1997). This heterogeneity of probably arises, in part, from differential specific control of arteriolar vasodilation for different fibre types. For example, arterioles from slow-twitch (type I) fibres have a greater endothelial nitric oxide synthase (eNOS) mRNA expression and a higher sensitivity and maximal responsiveness to endothelium-dependent vasodilation than their fast-twitch counterparts (Delp et al. 2000; Wunsch et al. 2000; Woodman et al. 2001). In addition, first-order arterioles isolated from slow-twitch muscles demonstrate a reduced sensitivity (i.e. less vasoconstriction) to noradrenaline (Delp, 1999) compared to their fast-twitch counterparts. Moreover, in contrast to fast-twitch muscle, virtually no contraction-induced ‘sympatholysis’ (Thomas et al. 1994) is present in the slow-twitch soleus. Based on the augmented contraction-induced hyperaemia noted in the slow-twitch soleus versus the fast-twitch white and mixed gastrocnemius (reviewed by Laughlin et al. 1997), it is generally to be expected that (and thus /) will be higher in muscles comprised of slow-twitch fibres. Indeed, we have demonstrated, during moderate intensity contractions, that is elevated relative to in slow (soleus) compared with fast-twitch (peroneal) muscle (Behnke et al. 2003; McDonough et al. 2004). It has also been shown that exercise intensity exerts a profound influence upon muscle (Armstrong & Laughlin, 1983). Of particular interest are the findings that both muscle fibre type and oxidative capacity strongly impact both the intensity-dependent pattern of distribution (Armstrong & Laughlin, 1983) and at maximal exercise (Poole et al. 2000).

In light of the above findings, the present investigation was designed to determine the relationships among and P_mvO2 across a broad range of metabolic demands (rest, and low and high stimulation intensities) in muscles selected specifically to span the range of fibre types present in mammalian muscle (soleus, mixed gastrocnemius and white gastrocnemius). Given the heterogeneity in between these muscles, we wished to test the specific hypothesis that P_mvO2 would fall to lower levels during contractions in fast-twitch (type II, mixed and white gastrocnemius) compared to slow-twitch muscles, with the greatest effect manifest in the white gastrocnemius (type IIb) muscle. In addition, we hypothesized that the kinetics of P_mvO2, across the rest–contraction transient, would be significantly faster (indicative of poorer -to- matching) in both fast-twitch gastrocnemius muscles compared with the slow-twitch soleus.

Methods

Ethical approval was granted by the Kansas State University Institutional Animal Care and Use Committee (IACUC) and all procedures were conducted according to National Institutes of Health guidelines.

Surgical preparation

All rats (female Sprague-Dawley, n = 20, wt = 317 ± 10 g) were anaesthetized prior to experimentation with pentobarbitone sodium (40 mg kg⁻¹i.p. to effect). In order to ensure that an adequate anaesthetic plane was maintained throughout the experiment, ocular and pedal reflexes were tested. If indicated, pentobarbitone was administered in a supplemental dosage (5–10 mg kg⁻¹) as needed. The carotid and tail (caudal) arteries were catheterized, with the use of an introducer, with polyethylene tubing (PE-10 connected to PE-50). This allowed for the infusion of the phosphorescent probe (palladium meso-tetra (4-carboxyphenyl) porphine dendrimer (R2); 15 mg kg⁻¹), measurement of arterial blood pressure (Digi-Med BPA model 200, Louisville, KY, USA) and withdrawal of arterial blood for blood gas measurement (Nova Stat Profile M, Waltham, MA, USA). In addition, the catheters allowed us to measure muscle blood flow via the radio-labelled microsphere technique.

The muscles chosen for the present experiment (soleus, Sol; mixed gastrocnemius, MG; and white gastrocnemius, WG) were chosen on the basis of their fibre type composition and oxidative enzyme activities (Delp & Duan, 1996). The Sol is a postural muscle whose primary functions are plantar flexion and ankle stabilization. Soleus is composed primarily of slow-twitch fibres (84% type I, 7% type IIa and 9% type IId/x) and has a citrate synthase activity (CSa; the marker of oxidative capacity used herein) of 21.3 μmol min⁻¹ g⁻¹. The MG is a powerful muscle of plantar flexion and, as such, has a primarily fast-twitch fibre composition (3% type I, 6% type IIa, 34% type IId/x and 57% type IIb), with a CSa of 25.7 μmol min⁻¹ g⁻¹. The WG is a plantar flexor of low oxidative capacity (CSa, 8.1 μmol min⁻¹ g⁻¹) and has a fibre type composition that is purely fast-twitch in nature (8% type IId/x, 92% type IIb; Delp & Duan, 1996). Each muscle was exposed for P_mvO2 measurements in the following manner. For experiments on the MG and WG, the leg was shaved and the skin and fascia overlying the ‘calf’ was opened in the sagittal plane. After the skin was opened, the tibial nerve was isolated and a stimulating electrode was attached. The ground electrode was attached distally, near the Achilles tendon. For experiments on the Sol, the skin and fascia overlying the peroneal muscle group were opened (coronal plane). After these were opened, the peroneal muscle was reflected so as to facilitate access to the Sol (Behnke et al. 2003). Care was taken to minimize the extent of the surgery in all cases. The exposed tissue was superfused with a Krebs–Henseleit bicarbonate-buffered solution (38°C, equilibrated with 5% CO₂–N₂ balance) and body temperature was maintained at ∼38°C by use of a heating pad.

Contractions protocol

Prior to beginning the experimental protocol, the R2 probe was introduced into the blood via the arterial catheter. The rat was then moved to a purpose-built ergometer and secured so as to limit movement, but not impede ventilation or cardiodynamics. To limit leg movement to the plantar flexors, the ankle was pinned to a wooden support, such that the ankle was fixed (same position each time). The phosphorimeter light guide was then positioned (within 1–3 mm) above the belly of the muscle of interest. Approximately 15 min later, the Sol, MG or WG was stimulated at 1 Hz for 3 min (2 ms pulse duration) using a Grass S88 stimulator (Quincy, MA, USA) at two different contraction intensities, 2.5 V (Lo) and 4.5 V (Hi). These contraction intensities were chosen on the basis of an incremental contraction test (+1 V min⁻¹) and correspond to approximately 30 and 65% of the stimulation voltage that produced a minimal P_mvO2 (P. McDonough, B. J. Behnke, T. I. Musch & D. C. Poole; unpublished observations). These stimulation voltages were chosen simply to span a relatively large intensity range and could be taken to resemble, in certain respects, low and moderately heavy intensity exercise. At the conclusion of the experimental protocol all animals were killed with an overdose of pentobarbitone sodium (>80 mg kg⁻¹).

Muscle blood flow and oxygen consumption

Muscle blood flow and oxygen consumption were measured in a separate set of experiments as previously described (Behnke et al. 2002a, 2003). Briefly, was determined using the radiolabelled microsphere technique (Musch & Terrell, 1992) and was measured at rest and just before the cessation of the 3 min contractions protocol in all three muscles and expressed as milliliters of blood per minute per 100 g tissue (ml min⁻¹ (100 g)⁻¹). Three different radiolabelled, 15 μm diameter microspheres (⁴⁶Sc, ⁸⁵Sr and ¹⁴¹Ce; New England Nuclear, Boston, MA, USA) were agitated via sonication and ∼2.5 × 10⁵ microspheres were injected into the ascending aorta at the specified time point. Tissue radiation counts were performed using a gamma scintillation counter (Packard Auto Gamma Spectrometer, Cobra model 5003, Perkin-Elmer Life and Analytical Sciences, Shelton, CT, USA). The correct placement of the carotid catheter in the aorta was verified after the experiment, while adequate mixing of microspheres was verified via inspection of kidney blood flows, with a difference <15% between R and L kidney being considered acceptable.

Muscle was determined in the following fashion. Arterial O₂ content (C_aO2) was measured directly (carotid arterial blood) and mixed venous O₂ content was calculated from P_mvO2 (since P_mvO2 is a valid approximation of mixed venous partial pressure of O₂ (P_O2; McDonough et al. 2001) using the rat O₂ dissociation curve (constructed using an n value (Hill coefficient) of 2.6, the measured [Hb], P₅₀ (the P_O2 at which Hb is 50% Saturated with O₂) of 38 and an O₂ carrying capacity of 1.39 ml O₂ (g Hb)⁻¹). was then calculated via the principle of mass balance using the Fick equation (i.e. ). Muscle O₂ diffusion index was defined as /P_mvO2 and, as such, provides an index of diffusive O₂ transport per unit of driving pressure.

Principle and measurement of phosphorescence quenching

The principles of phosphorescence quenching have been detailed previously (Poole et al. 1995; McDonough et al. 2001; Behnke et al. 2003). Specifically, the Stern-Volmer relationship (Rumsey et al. 1988) describes the nature of the quantitative relationship between probe phosphorescence and P_mvO2:

which, rearranged to solve for P_mvO2, gives:

where t is the lifetime of the phosphorescence at the prevailing O₂ tension, t₀ the lifetime at a P_mvO2 of ‘zero’ and k_Q is the quenching constant of the probe (mmHg s⁻¹).

For the R2 probe, t₀ is 601 μs and k_Q 409 mmHg s⁻¹ and, under the physiological conditions studied herein (Lo et al. 1997), P_mvO2 is wholly dependent upon the lifetime of the phosphorescence decay, which is inversely proportional to the prevailing P_O2.

Importantly, R2 is restricted to the intravascular space within the muscle, which allows measurement of microvascular O₂ tension (P_mvO2; (Poole et al. 2004)). A Frequency Domain Phosphorimeter (PMOD 1000; Oxygen Enterprises, Ltd, Philadelphia, PA, USA) was employed, with the light guide focused on a circular area of exposed muscle of ∼2 mm diameter. Within this area (principally composed of capillary blood, since this compartment constitutes the majority of intramuscular blood volume; Poole et al. 1995), a sample is obtained up to ∼500 μm deep. The PMOD 1000 modulates the excitation frequencies between 100 Hz and 20 kHz, which can measure P_mvO2 values from 0 to 240 mmHg. P_mvO2 was measured continuously, with data reported at 2 s intervals throughout.

Citrate synthase measurement

Citrate synthase was measured in duplicate from homogenates prepared from the Sol, MG and WG muscles according to the methodology of Srere (1969). Citrate synthase activity (CSa) was measured using a spectrophotometer (Spectramax 190 Molecular devices plate reader) in 300 μl aliquots at 30°C. Activity levels were expressed as μmol per minute per gram wet weight.

Curve fitting and statistical analysis

For the P_mvO2 data, curve fitting was accomplished using KaleidaGraph software (version 3.5; Synergy Software, Reading, PA, USA) and was performed on each data set using a one-component model:

and a more complex two-component model:

where P_mvO2(t) is the P_mvO2 at any time t, P_mvO2(BL) is the P_mvO2 at the beginning (i.e. baseline) of the stimulation protocol, Δ₁ and Δ₂ are the amplitudes of the fast and slow P_mvO2 components, TD₁ and TD₂ are the independent time delays and τ₁ and τ₂ are the time constants for each component. Goodness of fit was determined by three criteria: (1) the coefficient of determination (i.e. r²); (2) the sum of the squared residuals; and (3) visual inspection and analysis of the residual fit to a linear model. Mean response time (MRT) was calculated using the formula from MacDonald et al. (1997):

where Δ₁ and Δ₂, TD₁ and TD₂ and τ₁ and τ₂ are as defined above and Δ_tot=Δ₁+Δ₂. If a one-component model provided the best fit, the MRT equation reduced to the following:

In addition, the relative rate of change in P_mvO2 (dP_O2/dt) was defined as the ΔP_mvO2/τ for the on-transient to contractions (Behnke et al. 2003).

P_mvO2 values during resting and steady-state contractions (e.g. baseline and Δ), as well as modelling dependent results (e.g. TD, τ and MRT), were analysed using standard analysis of variance techniques between muscles (Sol, MG and WG). When a significant F value was demonstrated by the ANOVA, a Student–Newman–Keuls (SNK) post hoc test was performed to determine differences among mean values. Pearson product–moment correlations were performed upon selected variables. Statistical significance was accepted at P≤ 0.05.

Results

Microvascular oxygen pressure

Lo-contraction protocol (Lo)

At rest, microvascular O₂ pressure (P_mvO2) values were highest in the Sol, intermediate in MG and lowest in WG (Table 1). The nadir in P_mvO2 was significantly lower in both MG and WG than in Sol. In addition, both MG and WG frequently undershot the final steady-state P_mvO2 during the transition (Table 1 and Fig. 1A). This undershoot of final steady-state P_mvO2 occurred in ∼50% of the MG and WG preparations, but never occurred in Sol (Table 1). In terms of dynamics, for both MG and WG, the P_mvO2 mean response time (MRT) was significantly faster compared to Sol (Table 1). This faster MRT was due to a significantly shorter time constant (τ) and/or time delay (TD) in MG and WG compared to Sol (Table 1). However, the ΔP_mvO2 was significantly greater in Sol compared to both MG and WG (Table 1), such that the relative rate of P_mvO2 fall (dP_O2/dt, equal to dP_mvO2/τ) was not different between muscles (Table 1).

Table 1.

Microvascular P_O2 during Lo- and Hi-contraction protocols

	Sol	MG	WG
Lo-contraction protocol
Baseline PmvO2, i.e. resting (mmHg)	33.1 ± 3.2	22.5 ± 1.0*	19.0 ± 1.3*†
Nadir PmvO2 (mmHg)	19.0 ± 2.3	13.5 ± 1.5*§	11.6 ± 1.2*§
Steady-state PmvO2 (mmHg)	19.0 ± 2.3	15.3 ± 1.3*	12.2 ± 1.1*
PmvO2 undershoot (mmHg)	0	2.9 ± 0.7*	1.4 ± 0.2*
Occurrence of undershoot (%)	0	63	43
ΔPmvO2 (mmHg)	14.2 ± 1.4	9.0 ± 1.5*	7.3 ± 1.4*
Time delay (s)	19.6 ± 3.3	7.1 ± 0.8*	9.4 ± 3.3*
Time constant (s)	32.5 ± 4.2	14.7 ± 4.2*	22.1 ± 2.3*
ΔPO2/dt (mmHg s−1)	0.5 ± 0.1	0.8 ± 0.2	0.4 ± 0.1
MRT (s)	52.0 ± 6.2	21.8 ± 4.4*	31.5 ± 3.5*
Hi-contraction protocol
Baseline PmvO2 (mmHg)	29.2 ± 1.5	25.1 ± 1.1*	19.5 ± 1.6*†
Nadir PmvO2 (mmHg)	18.7 ± 1.8	8.9 ± 1.2*‡	9.3 ± 1.1*
Steady-state PmvO2 (mmHg)	18.7 ± 1.8	10.2 ± 1.6*‡	9.9 ± 1.1*
PmvO2 undershoot (mmHg)	0	3.5 ± 1.8*	1.3 ± 0.5*
Occurrence of undershoot (%)	0	43	43
ΔPmvO2 (mmHg)	10.4 ± 1.3‡	16.2 ± 1.7*‡	10.3 ± 1.8‡
Time delay (s)	12.9 ± 4.9	5.3 ± 1.0*	4.6 ± 0.8*
Time constant (s)	26.0 ± 5.9	6.9 ± 1.2*‡	11.3 ± 2.3*‡
dPO2/dt (mmHg s−1)	0.5 ± 0.1	3.2 ± 0.9*‡	1.1 ± 0.3‡
MRT (s)	38.9 ± 5.7‡	12.1 ± 1.5*‡	16.2 ± 2.8*‡

Values are presented as means ± s.e.m.

^*Significant difference from Sol;

^†denotes significant difference from MG;

^‡denotes significant difference between Lo and Hi;

^§indicates difference between nadir and steady-state value.

Figure 1. Microvascular P_O2 response to contractions in Sol, MG and WG.

A, microvascular O₂ partial pressure (P_mvO2) responses for soleus (Sol), mixed (MG) and white gastrocnemius (WG) in response to the Lo-contraction protocol. Note that P_mvO2 of Sol is higher than that of MG and WG throughout the contraction protocol and particularly throughout the transient. Thin line, real data; hatched line, model fit. Note the undershoot of end-contraction P_mvO2 in the two fast-twitch muscles. B, P_mvO2 responses for Sol, MG and WG in response to the Hi-contraction protocol. Note, again, the substantially higher P_mvO2 in Sol compared to the two fast-twitch muscles. Also note, again, the undershoot of P_mvO2 in the two fast-twitch muscles.

Hi-contraction protocol (Hi)

Like the Lo protocol, both the nadir and the steady-state P_mvO2 values were lower in MG and WG than in Sol (Table 1). Furthermore, an undershoot of final steady-state P_mvO2 values again occurred in MG and WG only. In terms of dynamics, MRT was significantly faster in MG and WG compared to Sol (Table 1 and Fig. 1B), due to both a shorter TD and faster τ (Table 1).

Lo- versus Hi-contraction protocol

While baseline P_mvO2 was not different between experimental conditions, both the nadir P_mvO2 and the steady-state P_mvO2 were significantly lower in Hi versus Lo for MG only (Table 1). However, MRT was significantly shorter for all three muscles in Hi compared to Lo (Table 1). Moreover, while ΔP_mvO2 was significantly greater in Hi versus Lo for MG and WG, ΔP_mvO2 was significantly less in Hi versus Lo for Sol (Table 1). Since τ was significantly shorter for MG and WG in Hi versus Lo, the greater ΔP_mvO2 resulted in a significantly greater dP_O2/dt in these two muscles in Hi versus Lo conditions, which was not the case for Sol (Table 1).

Muscle blood flow and oxygen delivery

Lo-contraction protocol

Muscle blood flow was significantly greater in Sol than in either MG or WG at rest and during contractions, while was greater only at rest in WG compared to MG (Fig. 2; all P < 0.05). In the absence of altered C_aO2, these differences were reflected in oxygen delivery (; Fig. 3: theoretical schema; Fig. 4: actual data), as was greater in Sol versus MG and WG at rest (5.8 ± 1.4 versus 0.7 ± 0.1 and 1.3 ± 0.2 ml min⁻¹ (100 g)⁻¹; Sol versus MG and WG) and during contractions (12.3 ± 3.0 versus 7.2 ± 1.2 and 8.1 ± 1.9 ml min⁻¹ (100 g)⁻¹; Sol versus MG and WG; all P < 0.05). Again, as for , at rest, was higher in WG than in MG (P < 0.05). The increase of (Fig. 2) and from rest to Lo was significant in all three muscles (P < 0.05).

Figure 2. Muscle blood flow () for Sol, MG and WG at rest in response to the Lo- and Hi-contraction protocols.

* Significantly different from Lo; † denotes significant difference from Lo; ‡ denotes significantly greater compared to both fast-twitch muscles (P < 0.05).

Figure 3. Theoretical diagram of the relationship between the convective ( difference) and diffusive () determinants of oxygen uptake.

The intersections of the lines (a, b, c and d) gives the and the y-intercept . These relationships permit analysis of O₂ transport differences among fibre types (see Fig. 4A–C). The line a–b shows how can increase via an increase in alone, while the vertical distance a–c shows how an increase in and will impact . Finally, the line a–d demonstrates how an increase in alone will serve to increase .

Figure 4. Convective and diffusive determinants of oxygen transport.

A, graphical representation of the relationship between the convective and diffusive determinants of for the Sol (see Fig. 3 legend for detailed explanation). Note that Sol increases its predominantly through an increase in with a relatively modest increase in . B, graphical representation of the same relationship in MG. Note how the MG relies to a much greater extent upon an increase in to increase . C, graphical representation of the above relationships for WG. Note that WG does not increase either or very much and so does not increase to nearly the extent demonstrated by MG and Sol.

Hi-contraction protocol

Similar to Lo, was significantly elevated in Sol compared to both MG and WG during contractions. As regards , Sol was elevated versus both MG and WG during contractions (16.1 ± 2.6 versus 9.6 ± 1.4 and 9.6 ± 1.6 ml min⁻¹ (100 g)⁻¹; Sol versus MG and WG). As in Lo, (Fig. 2) and were increased from rest in all muscles (Fig. 2; P < 0.05).

Lo- versus Hi-contraction protocols

Both and increased from Lo to Hi (Fig. 2) for all muscles, while and were significantly greater in Sol compared to both MG and WG for both Lo and Hi (Fig. 2).

Muscle oxygen consumption

Lo-contraction protocol

Muscle oxygen consumption was significantly higher for Sol versus MG and WG at rest (3.1 ± 0.4 versus 0.5 ± 0.02 and 1.1 ± 0.03 ml O₂ min⁻¹ (100 g)⁻¹; Sol versus MG and WG) and during contractions (10.2 ± 0.6 versus 6.4 ± 0.2 and 7.6 ± 0.1 ml O₂ min⁻¹ (100 g)⁻¹; Sol versus MG and WG; Fig. 5; all P < 0.05). However, in contrast to the situation for and was significantly higher in WG compared to MG during contractions (P < 0.05).

Figure 5. Muscle oxygen uptake () for Sol, MG and WG in response to both the Lo- and Hi-contraction protocols.

* Significant difference from rest; † denotes significant difference from Lo; ‡ denotes significant difference from MG and WG; and § denotes significant difference from MG only (all P < 0.05).

Hi-contraction protocol

was significantly higher for Sol versus both MG and WG during contractions (13.5 ± 0.5 versus 9.0 ± 0.2 and 9.2 ± 0.1 ml O₂ min⁻¹ (100 g)⁻¹; Sol versus MG and WG; Fig. 5; all P < 0.05). However, in contrast to the situation at rest and during the Lo-contraction protocol, was not different between WG and MG during Hi (Fig. 5).

Lo- versus Hi-contraction protocol

was significantly elevated in Hi versus Lo in all muscles (Fig. 5). This increase in was due to simultaneous (albeit of differing magnitude) increases in both and (Fig. 4).

Muscle diffusion index

Lo-contraction protocol

At rest, muscle diffusion index was greater in Sol than in WG and both were greater than in MG (Table 2). This difference was eradicated during the contraction period, when was not different between muscles during Lo (Table 2 and Fig. 4). In all three muscles, was increased from rest to Lo.

Table 2.

Instantaneous diffusing capacity for O₂ (D_O2m) during Lo- and Hi-contraction protocols

	Sol	MG	WG
Rest DO2m (ml (100 g)−1 mmHg−1 min−1)	0.11 ± 0.01	0.02 ± 0.001‡	0.06 ± 0.004‡§
Lo DO2m (ml (100 g)−1 mmHg−1 min−1)	0.57 ± 0.07*	0.52 ± 0.06*	0.69 ± 0.06*
Hi DO2m (ml (100 g)−1 mmHg−1 min−1)	0.75 ± 0.07*	1.14 ± 0.17*†‡	1.10 ± 0.15*†‡

Values are presented as means ± s.e.m.

^*Significant difference from rest;

^†denotes significant difference between Lo and Hi;

^‡denotes significant difference from Sol;

^§denotes significant difference from MG.

Hi-contraction protocol

During Hi, was increased from rest (Table 2). In addition, was greater in both MG and WG (which were not different) compared to Sol (Table 2).

Lo- versus Hi-contraction protocol

was increased from Lo to Hi in both fast-twitch muscles (MG and WG). However, was not different in Sol between Lo and Hi (Fig. 4).

Citrate synthase activity

Citrate synthase (CSa) was significantly greater in Sol (24.3 ± 0.5 μmol min⁻¹ g⁻¹) than either MG (17.9 ± 3.2 μmol min⁻¹ g⁻¹) or WG (11.0 ± 0.7 μmol min⁻¹ g⁻¹). In addition, CSa was significantly greater in MG than in WG (P < 0.05).

Discussion

The major original findings of this investigation include the following. (1) At rest and during Lo- and Hi-intensity stimulation, P_mvO2 was significantly lower in both fast-twitch muscles (MG and WG) compared with slow-twitch muscle (Sol), despite the fact that was lower. Thus, these fast-twitch muscles demonstrated a greater fractional O₂ extraction, a stratagem that necessitates a more pronounced increase in , in addition to an increased ATP contribution from non-aerobic sources (i.e. phosphocreatine (PCr) hydrolysis and glycolysis; Wilson, 1994). (2) P_mvO2 kinetics at the onset of contractions were more rapid (i.e. reduced MRT, Table 2) for MG and WG than for Sol. Moreover, there was a significant transient undershoot of P_mvO2 below the steady-state response in both MG and WG that was absent in Sol, an observation previously noted only in the pathological conditions of diabetes (Behnke et al. 2002b) and heart failure (Diederich et al. 2002; Behnke et al. 2004) for the mixed fibre type spinotrapezius muscle. In total, these findings suggest that the slow-twitch Sol, as compared to the fast-twitch MG and WG, is better able to defend P_mvO2 (and thus the O₂ driving pressure into the myocyte) and adapt (i.e. make adjustments in oxidative metabolism) to metabolic transients in a manner which minimizes the reliance upon non-aerobic energy sources and, thus, the reduction in cellular energy state.

Effects of contraction intensity on P_mvO2

The present investigation is the first study to investigate the effect of muscular contractions of different intensity upon P_mvO2 dynamics in skeletal muscle. The results indicate that a higher contraction intensity (i.e. Hi versus Lo) results in: (1) a further lowering of P_mvO2 in fast- (MG and WG) but not in slow-twitch muscle (Sol); and (2) a speeding of MRT in all muscles. In this regard, Sol appears to have the ability to adjust to an increase in energetic demands with little (rest-to-Lo transition) or no increase (Lo-to-Hi transition) in fractional O₂ extraction. Thus, Sol exhibits a proportionally larger increase in , while the two fast-twitch muscle portions demonstrate a relatively modest response and therefore mandate a greater increase in fractional O₂ extraction (i.e. ↑; Fig. 4). Furthermore, since ΔP_mvO2 was significantly different between muscle fibre types during both contraction protocols (Table 1), the relative rate of fall in P_mvO2 (dP_O2/dt) was compared in an effort to normalize the speed with which O₂ extraction increases among muscles. When compared in this fashion, the different strategies by which slow- and fast-twitch muscles adapt to an increase in energetic demands are clearly evident. Specifically, while dP_O2/dt was the same for both Lo and Hi for Sol, dP_O2/dt was significantly larger in the Hi protocol for both MG and WG compared to Lo (Table 1). Thus, these findings indicate that the slow-twitch Sol adjusted to an increase in energetic demand with little (or no) additional reduction in P_mvO2 and a less pronounced speeding of P_mvO2 kinetics. Oppositely, the two fast-twitch muscles (WG and MG) relied upon rapid and large changes in fractional O₂ extraction (i.e. fast P_mvO2 kinetics and an increased ).

While the present study is the first to describe the changes in P_mvO2 and its kinetics with an increase in contraction intensity in vivo, Kindig et al. (2003b), using an isolated single fibre preparation, noted that extracellular P_O2 (P_eO2; essentially a static external O₂ source) fell faster and to progressively lower values in fast- versus slow-twitch Xenopus laevis muscle as stimulation intensity was increased. In addition, ‘peak’ was ∼20% higher in the slow-twitch fibres. Furthermore, they noted that a reduction in the initial P_eO2 (again loosely analogous to ) progressively reduced the MRT for the fall in P_eO2 (Kindig et al. 2003a). Thus, the findings of Kindig et al. (2003a, b) are consistent with those of the present study where we demonstrate a decreased extracellular O₂ pressure-head in fast-twitch muscle that is accentuated with increased contraction intensity. A reduction in this critical O₂ driving pressure (P_mvO2 herein) would be expected to reduce according to Fick's law of diffusion (Wagner et al. 1991) and consequently stimulate substrate level metabolism (i.e. greater PCr degradation and increased glycogenolysis/glycolysis) to a much greater degree (Wilson et al. 1977) in fast- (MG and WG) versus slow-twitch muscles (Sol). Additionally, the results of both the present study and those of Kindig et al. (2003a, b) show that this effect will be exacerbated with an increase in contraction intensity in fast-, but not slow-twitch muscle.

Basis for, and consequences of, differing P_mvO2 responses among muscles

It is becoming increasingly apparent that the interrelationship between muscle and in the steady state and across the rest–contractions transient is complex and, at best, only partly understood. At present, our understanding of the control of is perhaps less contentious than that of . Specifically, steady-state is a function of basal O₂ requirements plus any additional work/tension-induced O₂ demands, which are principally determined by the myofilament cross-bridge activity and muscle ‘efficiency’ (reviewed by Kushmerick, 1983; Poole, 1997), while kinetics in healthy muscle are considered to be determined primarily by what has been termed muscle oxidative enzyme inertia (reviewed by Grassi, 2000). With respect to vascular control, there exist myriad mechanisms that include mechanical (muscle pump), neural (sympathetic), propagated (conducted vasodilation) and vasomotor control (e.g. endothelium dependent and independent), which are thought to contribute to the hyperaemia of exercise and increased (reviewed by Thomas & Segal, 2004). As mentioned in the Introduction, Sol is endowed with a superior ability to undergo a more rapid and greater contraction-induced hyperaemia than MG or WG (Delp et al. 2000; Wunsch et al. 2000; Woodman et al. 2001). The present investigation illustrates the functional consequences of this elevated hyperaemic response (i.e. increased P_mvO2 and reduced reliance upon O₂ extraction) and is in agreement with research wherein an increased during contractions led to a better maintenance of cellular energy state and reduced fatigue (Hogan et al. 1992; Haseler et al. 1998).

Irrespective of P_mvO2 kinetics, P_mvO2 and therefore the pressure driving O₂ diffusion into the myocyte, is uniformly lower across the rest-to-exercise transition in fast- versus slow-twitch locomotory muscle (Behnke et al. 2003; present study). This would impair O₂ flux in MG and WG, insofar as that flux is determined by the capillary-to-myocyte P_O2 gradient (Wagner, 1995). However, the capillary-to-myocyte P_O2 gradient acts in concert with the diffusion characteristics of the muscle to facilitate capillary-to-myocyte O₂ flux. is believed to be determined by: (1) the capillary haematocrit and capillary length per fibre volume, the product of which gives the number of red blood cells adjacent to a muscle fibre at a given time (Federspiel & Popel, 1986; Groebe & Thews, 1990); and (2) the size of the ‘functionally O₂-carrier depleted region’, which will be reduced at lower intramyocyte P_O2 values such that increases (Honig et al. 1997). Figure 4 demonstrates the different strategies by which P_mvO2 and interact to generate the measured in fast- versus slow-twitch muscle and how these relationships change from Lo to Hi stimulation intensities. In the Sol, P_mvO2 is relatively high and capillary-to-myocyte O₂ flux increases from Lo to Hi mostly as a result of increased convective O₂ delivery (see Sol, Fig. 4A). Alternatively, in MG and WG, if decreased P_mvO2 reflects a reduced intracellular P_O2 (and it likely does) the saturation state of myoglobin will be lower and O₂ flux will be improved via an increased O₂ conductance of myoglobin as the thickness of the functionally O₂-carrier depleted region is decreased (i.e. enhanced , see MG and WG, Fig. 4B and C). Thus, large reductions in P_mvO2 across the transient (particularly during Hi; see Figs 4A, B and C) lead to a greater reliance upon to meet O₂ demand (MG and WG), while high convective O₂ delivery obviated this requirement in Sol. This greater reliance upon increased and decreased P_mvO2 (and therefore intracellular P_O2), in concert with the finding that P_mvO2 falls faster and to lower levels (sometimes decreasing transiently below steady-state levels) in fast- versus slow-twitch locomotory muscles (see Figs 1 and 6) may have profound functional consequences and may help explain several key physiological and pathophysiological observations. For example, as mentioned above, a reduced microvascular O₂ pressure-head would act to impair both blood-to-myocyte and intracellular O₂ flux. In turn, a low intracellular P_O2 may serve to slow kinetics (Behnke et al. 2002a), thus elevating the O₂ deficit and forcing a greater reliance upon substrate level phosphorylation. Ultimately, this will cause a greater perturbation (ΔPCr, ΔADP_free, ΔPi and ΔH⁺) of the intracellular milieu and accelerate glycogen depletion and the onset of fatigue (Wilson et al. 1977). It is possible, therefore, that the slowed kinetics and more pronounced intracellular perturbations found under conditions which predispose the individual towards recruitment of fast-twitch fibres (i.e. heavy/severe exercise (Whipp & Mahler, 1980; Barstow & Mole, 1991), diabetes (Behnke et al. 2002b) or heart failure (Diederich et al. 2002; Behnke et al. 2004)] may result, in part, from the altered P_mvO2 profiles characteristic of this fibre type.

Figure 6. Relationship between steady-state (SS)P_mvO2and the mean response time (MRT) for the fall inP_mvO2for a variety of different muscles.

In ascending order of MRT the muscles are: MG, WG, Peroneal muscle group, Spinotrapezius and Sol. Data for Spino are from McDonough et al. (2001) and for Peroneal are from McDonough et al. (2004). Note that fast P_mvO2 kinetics are positively correlated with a lower steady-state P_mvO2 and thus reduced pressure gradient for oxygen transfer from the microvasculature into the myocyte.

Model considerations

As mentioned above, with respect to muscle function, the matching of to as reflected by P_mvO2 is extremely important because it sets the O₂ pressure-head that drives blood-to-tissue O₂ transfer. However, one limitation of the present investigation was the inability to follow the dynamics of in separatum at the onset of contractions and so only the resting and steady-state were measured. Consequently, we have refrained from making mechanistic interpretations regarding the profile beyond the fact that, to produce the observed profile of P_mvO2 in fast- versus slow-twitch muscle (i.e. faster fall to a lower contracting value in fast-twitch muscle), dynamics must be slower and the dynamic range reduced compared to . When the required technology is developed, it would obviously be important to explore these issues further. In addition, since the phosphorimeter only covers a small area of the muscle being studied (see Methods), it could be argued that our results are only applicable to the specific section of muscle sampled. However, we have published several papers that demonstrate a remarkable between-animal consistency of P_mvO2 (Behnke et al. 2001, 2003; Diederich et al. 2002; Geer et al. 2002; present study), which suggests that the P_mvO2 values measured herein are indicative of an ‘average’P_mvO2 for that muscle. Lastly, the determinants of muscle diffusing capacity are complex. Structural variables, such as capillary length per fibre volume and capillary-to-fibre surface contact, in combination with functional indices of capillary haemodynamics (e.g. capillary red blood cell haematocrit, flux and velocity), which are considered important determinants of blood-to-myocyte O₂ transfer (Federspiel & Popel, 1986; Groebe & Thews, 1990; Mathieu-Costello et al. 1991), could lend critical insight into the recruitment of O₂ diffusing capacity during exercise transients. Unfortunately, these variables could not be resolved, due to technical limitations, in the present investigation.

Conclusions

The regulation of the -to- relationship appears to be fundamentally different in the fast- versus slow-twitch muscles examined herein, with P_mvO2 falling faster and to lower levels in fast-twitch muscles (Fig. 6). This finding is consistent with the lower exercise-induced hyperaemia during submaximal contractions reported in certain fast-twitch muscles and may help to explain the presence of slowed kinetics and greater metabolic perturbation found during physiological and pathophysiological conditions that recruit a greater fast-twitch fibre population.