Biomedical subjects

A Comtois

Publications and source records attributed to A Comtois.

18 recordsLinked to original sources

PubMed2000

A pathophysiological study of abdominal organs following intraperitoneal injections of chloral hydrate in rats: comparison between two anaesthesia protocols.

Chloral hydrate (CH) is used as an anaesthetic agent in laboratory rats. Side effects occurring with high concentrations have mainly occurred in abdominal organs. The objective of the present study was to minimize these side effects following intraperitoneal administration of CH using lower concentrations. Animals were evaluated using different procedures including a general necropsy, intraperitoneal white cell counts, histology and duodenal peristalsis and acetylcholine-induced contractions. Results clearly show that lower concentrations of CH while keeping the same anaesthetic dose (400 mg/kg) will minimize the irritancy of CH on abdominal organs while providing the same level of anaesthesia.

Abdominal Muscles↗

PubMed1999

Dystonin-deficient mice exhibit an intrinsic muscle weakness and an instability of skeletal muscle cytoarchitecture.

Dystonia musculorum (dt) was originally described as a hereditary sensory neurodegeneration syndrome of the mouse. The gene defective in dt encodes a cytoskeletal linker protein, dystonin, that is essential for maintaining neuronal cytoskeletal integrity. In addition to the nervous system, dystonin is expressed in a variety of other tissues, including muscle. We now show that dystonin cross-links actin and desmin filaments and that its levels are increased during myogenesis, coinciding with the progressive reorganization of the intermediate filament network. A disorganization of cytoarchitecture in skeletal muscle from dt/dt mice was observed in ultrastructural studies. Myoblasts from dt/dt mice fused to form myotubes in culture; however, terminally differentiated myotubes contained incompletely assembled myofibrils. Another feature observed in dt/dt myotubes in culture and in skeletal muscle in situ was an accumulation and abnormal distribution of mitochondria. The diaphragm muscle from dt/dt mice was weak in isometric contractility measurements in vitro and was susceptible to contraction-induced sarcolemmal damage. Altogether, our data indicate that dystonin is a cross-linker of actin and desmin filaments in muscle and that it is essential for establishing and maintaining proper cytoarchitecture in mature muscle.

Actins↗

PubMed1998

Endotoxin-induced skeletal muscle contractile dysfunction: contribution of nitric oxide synthases.

The aims of this study were to assess the role of nitric oxide (NO) and the contribution of different NO synthase (NOS) isoforms in skeletal muscle contractile dysfunction in septic shock. Four groups of conscious rats were examined. Group 1 served as control; group 2, 3, and 4 were injected with Escherichia coli endotoxin [lipopolysaccharide (LPS), 20 mg/kg i.p.] and killed after 6, 12, and 24 h, respectively. Protein expression was assessed by immunoblotting and immunostaining. LPS injection elicited a transient expression of the inducible NOS isoform, which peaked 12 h after LPS injection and disappeared within 24 h. This expression coincided with a significant increase in nitrotyrosine formation (peroxynitrite foot-print). Muscle expression of the endothelial and neuronal NOS isoforms, by comparison, rose significantly and remained higher than control levels 24 h after LPS injection. In vitro measurement of muscle contractility 24 h after LPS injection showed that incubation with NOS inhibitor (S-methyliosothiourea) restored the decline in submaximal force generation, whereas maximal muscle force remained unaffected. We conclude that NO plays a significant role in muscle contractile dysfunction in septic animals and that increased NO production is due to induction of the inducible NOS isoform and upregulation of constitutive NOS isoforms.

Animals↗

PubMed1998

Diaphragm sarcolemmal injury is induced by sepsis and alleviated by nitric oxide synthase inhibition.

Endotoxemia is associated with impaired diaphragm contractility, and increased nitric oxide (NO) production has recently been implicated in this phenomenon. However, the precise nature of sepsis-related alterations in diaphragm myofiber function remains unclear. We tested the hypothesis that enhanced NO synthesis during sepsis produces diaphragm sarcolemmal injury with attendant abnormalities of myofiber membrane electrophysiology. Two different rat sepsis models were employed: acute (4 h) intraarterial endotoxin (LPS; 20 mg/kg) and subacute (24 h) peritonitis induced by cecal ligation and perforation (CLP). Diaphragm damage occurred after both LPS and CLP, as indicated by hyperpermeability of myofibers to a low molecular weight tracer dye, which is normally unable to penetrate the sarcolemma. Sarcolemmal injury was significantly correlated with reductions in the resting membrane potential (Em) of single diaphragm myofibers. Western analysis revealed increased diaphragmatic expression of the inducible isoform of NO synthase (iNOS) after LPS and CLP. An inhibitor of NOS activity, LNMMA, significantly decreased morphologic as well as electrophysiologic signs of myofiber membrane injury and dysfunction. Therefore, we conclude that both acute endotoxemia and subacute peritonitis models of sepsis lead to significant sarcolemmal damage and altered Em in diaphragm myofibers. These changes appear to be mediated, at least in part, through the pathway of increased nitric oxide production.

Animals↗

PubMed1998

Ontogenesis of nitric oxide synthases in the ventilatory muscles.

Nitric oxide (NO) acts as an endogenous mediator in mature skeletal muscle. In this study, we investigated the regulation of the endothelial (eNOS) and neuronal (nNOS) isoforms of nitric oxide synthase (NOS) in skeletal-muscle development (rat diaphragm). Muscle NOS activity, nNOS and eNOS protein, and mRNA expressions were markedly increased during the late gestational and early postnatal periods. Expression of both isoforms, however, declined progressively thereafter. Similarly, argininosuccinate lyase and argininosuccinate synthetase, both involved in the recycling of L-citrulline to L-arginine, were expressed at high levels in rat embryonic and neonatal diaphragms, with gradual reduction in their expression during late postnatal development. Immunostaining revealed extensive nNOS expression at the sarcolemma in neonatal and mature diaphragms, whereas eNOS expression was limited to the endothelium. Both neonatal and adult diaphragms expressed an alternatively spliced nNOS isoform with an insert of 34 amino acids between exons 16 and 17. In vitro-generated muscle force rose significantly after NOS inhibition in both neonatal and adult diaphragms, but the magnitude of force augmentation was larger in adult than in neonatal diaphragm. These results indicate that constitutive NOS isoforms are developmentally regulated in skeletal muscles, suggesting multiple roles for NO in developing and mature skeletal-muscle fibers.

Aging↗

PubMed1996

Effect of adenosine and intracellular GTP on KATP channels of mammalian skeletal muscle.

We investigated the action of adenosine and GTP on KATP channels, using inside-out patch clamp recordings from dissociated single fibers of rat flexor digitorum brevis (FDB) skeletal muscle. In excised patches, KATP channels could be activated by a combination of an extracellular adenosine agonist and intracellular Mg2+-ATP and GTP or GTP-gamma-S. The activation required hydrolyzable ATP and could be partially reversed with Mg2+, suggesting that it may involve a G-protein dependent phosphorylation of KATP channels. We found that KATP channels of the rat FDB could not be activated by Mg2+-ATP alone or by Mg2+-ATP in the presence of extracellular adenosine. Patches whose channel activity had been 'rundown' by Ca2+ could not be recovered by adenosine, GTP or Mg2+-ATP. KATP channels activated by adenosine receptor agonists had a similar ATP sensitivity to those under control conditions; but adenosine appears to be able to switch these KATP channels from an inactive to an active mode.

Adenosine↗

PubMed1996

Modulation of muscle contractility during fatigue and recovery by ATP sensitive potassium channel.

The activity of ATP-sensitive potassium channels of skeletal muscle is controlled by changes in the bioenergetic state of the cell. These channels are inactive in unfatigued muscle and become activated during fatigue. It has been postulated that ATP-sensitive potassium channels shorten the action potential duration, increase the potassium efflux and contribute to the decrease in force during fatigue. Although blocking ATP-sensitive potassium channels during fatigue prolongs the action potential duration and decreases the potassium efflux as expected, it does not affect the rate of fatigue development as observed from the decrease in tetanic force. Even though such results are not consistent with the hypothesis that ATP-sensitive potassium channels contribute to the decrease in force during fatigue, a reduced capacity of skeletal muscles to recover their tetanic force following fatigue is also observed when ATP-sensitive potassium channels are blocked during fatigue, suggesting that these channels have a myoprotective effect. It is thus possible that removing this myoprotection during fatigue results in deleterious effects which counteract the expected slower decrease in force. However ATP-sensitive potassium channel openers also fall to affect the rate of fatigue development. Therefore, the results obtained so far do not support the hypothesis that ATP-sensitive potassium channels contribute to the decrease in force during fatigue.

Adenosine Triphosphate↗

PubMed1995

Effect of tolbutamide on the rate of fatigue and recovery in frog sartorius muscle.

The goal of this study was to determine how blocking ATP-sensitive K+ channels with tolbutamide affects the excitability and contractility of intact frog sartorius muscle during fatigue development. Fatigue was elicited with one tetanic contraction every sec for 3 min. During fatigue the resting potential decreased by 10 mV although the action potential overshoot remained constant. The addition of 2 mmol.liter-1 tolbutamide 60 min before fatigue did not modify the effect of fatigue on the resting potential and action potential overshoot. During fatigue development the half-repolarization time of control muscles increased by 0.26 msec in control muscles, although it increased by 0.77 msec in the presence of 2 mmol.liter-1 tolbutamide; the difference was significant. The decrease in force during fatigue development was not affected by 2 mmol.liter-1 tolbutamide (added 60 min before fatigue), whereas the recovery of force after fatigue was slower in tolbutamide- exposed muscles than in control muscles. Addition of 2 mmol.liter-1 tolbutamide after 5 min of recovery reduced the recovery rate of the resting potential and half-repolarization time, but did not affect the recovery of tetanic force during the first 40 min. Our results are consistent with the hypothesis that ATP-sensitive K+ channels are activated during fatigue development and that they contribute to the repolarization phase of action potentials, but they do not support the hypothesis that ATP-sensitive K+ channels contribute to the decrease in force.

Action Potentials↗

PubMed1994

The effect of K+ on the recovery of the twitch and tetanic force following fatigue in the sartorius muscle of the frog, Rana pipiens.

The goal of this study was to investigate how an increase in the extracellular K+ (K0+) concentration immediately after fatigue affects the recovery of the resting potential, the twitch and tetanic contraction of frog sartorius muscle to further understand the role of K+ in the mechanism of fatigue. Resting potentials were measured with conventional microelectrodes. Twitch and tetanic contractions were elicited by field stimulation. All muscles were fatigued with tetanic contractions at a rate of one contraction per second for 3 min while being exposed to 3 mmole l-1 K0+. During fatigue development the resting potential decreased by 16 mV (control group and pH0 7.2, extracellular pH), while the decrease in the twitch force was 32.8%, compared to 79.3% for the tetanic force, and 84.6% for the maximum rate of force development of the tetanus. Fatigued muscles were also unable to maintain a plateau phase during a tetanus: force declined by 14.8% during this phase. During the recovery period under control conditions (3 mmole l-1 K0+), all four parameters returned to their pre-fatigue values, the recovery of the plateau phase was the fastest (10 min), while that of the twitch force was the slowest (80 min). When K0+ was increased to 7.5 or 9.5 mmole l-1 immediately after fatigue, the recovery rate of the tetanic force and plateau phase was reduced. The maximum rate of force development of the tetanus, however, recovered at a faster rate than control muscles. The recovery of the twitch force was also increased above that of control when K0+ was increased to 9.0 mmole l-1 (a concentration which maximally potentiates the twitch force of unfatigued muscle). Frog sartorius muscles were also tested at pH0 6.4, a pH0 which inhibits force recovery. At that pH0 the effects of K0+ were similar to those observed at pH0 7.2. It is concluded that the role of K+ in muscle fatigue is more complex and may not involve just a contribution to the decrease in force during fatigue development, but may also contribute to an increase in force development under some conditions.

Animals↗

PubMed1994

An ATP-sensitive potassium channel blocker decreases diaphragmatic circulation in anesthetized dogs.

The goal of this study was to determine whether in the dog ATP-sensitive K+ channels blocked with glibenclamide affect diaphragmatic blood flow [phrenic arterial blood flow (Qpa)] during both spontaneous breathing at rest and increased diaphragmatic activity. A control group (no glibenclamide; n = 4) and an experimental group (50 mg/kg of glibenclamide; n = 5) were studied. During spontaneous breathing at rest, Qpa was 15.0 ml.min-1 x 100 g-1 and decreased by 5% in the presence of glibenclamide. Diaphragmatic pacing (30 min-1) generated by phrenic nerve pacing produced an initial diaphragmatic tension-time index of 0.25 in both groups. A 50% decay in transdiaphragmatic pressure was reached at 165 s in the experimental group compared with 421 s in the control group. Diaphragmatic pacing increased Qpa by 46% in the experimental group and 65% in the control group, yielding a 63% greater vascular resistance in the experimental group. Phrenic vein K+ content at rest was unchanged by the presence of glibenclamide, being 3.6 +/- 0.16 mmol/l compared with 3.5 +/- 0.19 mmol/l in the control group. Phrenic nerve pacing in the control group produced a 13% increase in phrenic vein K+ content, whereas in the experimental group a 16% decrease was observed. We suggest that ATP-sensitive K+ channels play an important role in the modulation of Qpa.

Adenosine Triphosphate↗

PubMed1993

Tolbutamide, but not glyburide, affects the excitability and contractility of unfatigued frog sartorius muscle.

The goal of this study was to characterize the effects of tolbutamide and glyburide, two known KATP channel blockers, on intact, unfatigued sartorius muscle fibres of the frog, Rana pipiens. Tetanic contractions were elicited by field stimulation with 200 ms long train of pulses (0.5 ms, 6 V, 140 Hz). Resting and action potentials were measured using conventional microelectrodes. At pHo 7.2 (extracellular pH), the tetanic force was unaffected by 0.5 mM and 1.0 mM tolbutamide, but at 2.0 mM it decreased by 15.5 +/- 1.0%. The effect of tolbutamide on the tetanic force was significantly greater at pHo 6.4: all three tolbutamide concentrations caused a significant decrease in tetanic force, being 62.3 +/- 9.4% at 2 mM. In the presence of tolbutamide a large number of fibres became unexcitable at pHo 6.4, but not at pHo 7.2. Glyburide at 10 microM, on the other hand, caused a 5-7% decrease in tetanic force at both pHo 6.4 and 7.2, but no further decreases in tetanic force were observed when the glyburide concentration was increased up to 100 microM. Unlike tolbutamide, glyburide did not affect the excitability of muscle fibres, but significantly prolonged the repolarization phase of action potentials, especially at pHo 6.4. We suggest that several of the tolbutamide effects reported in this study cannot be accounted for by a direct effect on KATP channels, and that the large decrease in membrane excitability and muscle contractility in the presence of tolbutamide must seriously be taken into consideration when this channel blocker is used to study the physiological role of KATP channels in intact muscle fibres.

Action Potentials↗

PubMed1992

Optimal diaphragmatic blood perfusion.

The intrabreath time course of phrenic artery blood perfusion (Qpha) was studied in five anesthetized dogs. The diaphragm was paced with submaximal levels of stimulation at various duty cycles (DC) to achieve tension-time index below and above the fatigue threshold (0.03-0.60). Left Qpha was measured via Doppler technique during control (inactive diaphragm) and during two submaximal levels of bilateral phrenic nerve stimulation sustained for 1 min. Measurements were done when Qpha reached steady state in each run. The frequency of pacing of each run was 10/min, and the DC ranged from 0.1 to 0.9 in 0.1 increments. Shortening of costal and crural segments was measured by sonomicrometry. It was found that Qpha during the diaphragmatic contraction phase (QphaC) was a sigmoidal function of DC and was not affected by the levels of transdiaphragmatic pressure (Pdi) explored (34-64% of maximal Pdi). Qpha during the diaphragmatic relaxation phase (QphaR) was a parabolic function of the DC, reaching an optimal value at DC of approximately 0.3 at any given Pdi. QphaR increased significantly with the preceding level of Pdi. QphaT (the sum of QphaC and QphaR) was a parabolic function of DC, reaching peak values at DC of 0.4-0.6 and then decreasing. This function was similar at two levels of Pdi. Post-pacing hyperemia was directly related to tension-time index greater than 0.20.

Animals↗

PubMed1991

Chemical activation of thin-fiber phrenic afferents. 2. Cardiovascular responses.

To assess the effects of groups III and IV (thin-fiber) phrenic afferents on arterial pressure, heart rate, and distribution of cardiac output, we injected capsaicin into phrenic arteries of in situ isolated and innervated left diaphragms of dogs anesthetized with chloralose, vagotomized, and mechanically ventilated. Blood flow in the ascending aorta, common carotid, renal, superior mesenteric, and femoral arteries was measured by electromagnetic and Doppler flow probes. Injection of 1 mg capsaicin into the left phrenic artery produced congruent to 15% increase in mean arterial pressure and congruent to 7% increase in heart rate with no change in aortic flow. Phrenic arterial flow decreased by 64%, renal arterial flow by 16%, and superior mesenteric arterial flow by 10%, whereas carotid flow increased by 13% and flow to the right gastrocnemius muscle did not change. Mean arterial pressure, heart rate, and blood flow distribution (with the exception of the decline in phrenic blood flow) returned to baseline within 60 s of the injection. Injection of 1.5 mg capsaicin into the right isolated and innervated gastrocnemius produced congruent to 35% increase in mean arterial pressure, 17% rise in heart rate, and no change in aortic blood flow. Phrenic and carotid arterial flow rose by 240 and 41%, respectively, whereas renal and superior mesenteric flow declined by 50 and 20%, respectively. In conclusion, thin-fiber phrenic afferents have an excitatory effect on arterial pressure and heart rate. They redistribute blood flow away from the renal and intestinal vascular beds and toward the carotid vascular bed. On the other hand, the cardiovascular reflex from thin-fiber phrenic afferents seems less potent than that from limb muscle afferents.

Afferent Pathways↗

PubMed1991

Restriction of regional blood flow and diaphragmatic contractility.

We have tested the hypothesis that the diaphragmatic head-to-head arterial anastomosis system should maintain adequate diaphragmatic function even during occlusion of some of its arteries. In six anesthetized open-chest dogs, left phrenic vein blood flow (Qphv) was measured by pulsed Doppler flowmetry. Contractility was measured by sonomicrometry in the left costal and crural diaphragm. The diaphragm was paced for 15 min by continuous bilateral supramaximal phrenic nerve stimulation. In five separate runs the following arteries were occluded at minute 5: 1) left phrenic artery, 2) internal mammary artery (IMA), 3) left phrenic artery and IMA, 4) descending aorta, and 5) descending aorta and IMA. Occlusion was then released at minute 10 of the run. In runs 1-3 there were no changes in contractility in costal or crural diaphragm and no changes in Qphv. However, in runs 4 and 5, Qphv decreased to 55.2 +/- 7.4 and 24.0 +/- 6.5% of control values, respectively. In run 4, percent maximum shortening from functional residual capacity (%LFRC) of the crural diaphragm decreased by 39.1%, while %LFRC of the costal diaphragm increased by 41.4% and abdominal pressure decreased by 47.0%. In run 5, abdominal pressure decreased by 53.5% and %LFRC of the crural and costal diaphragm decreased by 45.5 and 5.8%, respectively. Also relative postocclusion hyperemia was greater in run 5 (64.8%) than in run 4 (40.2%).(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗

PubMed1990

Contraction-dependent modulations in regional diaphragmatic blood flow.

Blood flow (Q) of the diaphragm was measured simultaneously with Doppler probes placed on diaphragmatic veins and an artery and by direct volumetric measurements obtained from cannulation of diaphragmatic blood vessels. The Doppler converting coefficients obtained were 6.27, 7.25, 4.21, and 41.07 ml.min-1.kHz-1 for left phrenic artery flow (Qpha), phrenic vein flow (Qphv), internal mammary vein flow (Qimv), and azygos vein flow (Qazv), respectively. The time course of Qpha, Qphv, Qimv, and Qazv after imposed patterns of diaphragmatic contraction was measured in nine anesthetized dogs. Each pattern consisted of various combinations of transdiaphragmatic pressure (Pdi), frequency of pacing (f), and duty cycle obtained by bilateral phrenic nerve stimulation. The dogs were prepared with chests open and loosely casted abdomens. Qpha, Qphv, Qimv, and Qazv were measured at rest (control, passive diaphragm, mechanical ventilation) and at two submaximal levels of stimulation (30 and 60% of Pdimax). The f was 10 or 30 cycles/min and the duty cycle was 0.25, 0.50, and 0.75. The results show 1) Qpha, Qphv, Qimv, and Qazv reached stable values (equilibration) after 30-36 s of pacing; 2) the steady Qpha, Qphv, and Qimv were linearly related to Pdi, and they were related by a parabolic function to duty cycle, whereas Qazv was not significantly affected by Pdi and increased linearly as a function of the duty cycle; 3) the diaphragmatic blood drainage was approximately 60% through the intercostal veins leading into the azygos trunk, 25% through the phrenic vein, and 15% through the internal mammary vein during pacing of the diaphragm at a duty cycle of 0.50 and 60% Pdimax; and 4) for a given pacing pattern, Qpha and Qphv increased with f, but Qimv and Qazv did not.

Animals↗

PubMed1990

Effect of separate hemidiaphragm contraction on left phrenic artery flow and O2 consumption.

Phrenic arterial blood flow has been shown to increase during bilateral phrenic nerve stimulation (BPNS). However, the role of unilateral phrenic nerve stimulation [left (LPNS) or right (RPNS)] on the blood flow and O2 consumption of the contralateral hemidiaphragm is not known and is explored here. In six anesthetized, mechanically hyperventilated dogs, left phrenic arterial blood flow (Qlpha) was measured (Doppler technique). Supramaximal (10 V, 30 Hz, 0.25-ms duration) LPNS, RPNS, and BPNS at a pacing frequency 15/min and duty cycle of 0.50 were delivered in separate runs. Left hemidiaphragmatic blood samples for gas analyses were obtained by left phrenic venous cannulation. During RPNS, Qlpha and left hemidiaphragmatic O2 consumption (VO2ldi) did not change significantly compared with control. During LPNS and BPNS, there was a significant increase in Qlpha and VO2ldi (P less than 0.01). There was no significant difference in Qlpha and VO2ldi between LPNS and BPNS (P greater than 0.05). We conclude 1) that there is a complete independence of left-right hemidiaphragmatic circulation both at rest and during diaphragm pacing and 2) that during unilateral stimulation transdiaphragmatic pressure is not related to diaphragmatic blood flow.

Animals↗

PubMed1988

Effect of transdiaphragmatic pressure partitioning on activation and fatigue of the diaphragm.

Fatiguing contractions of the diaphragm during inspiratory resistive loading are accompanied by a predictable rate of decay in the high/low (H/L) frequency ratio of the diaphragmatic EMG (EMG-DI) when the esophageal and gastric pressure (Pes and Pga) components of transdiaphragmatic pressure (Pdi) are equal. However, Pes and Pga do not contribute equally to Pdi under a number of clinical and physiologic conditions. We therefore tested the effect of varying the Pes and Pga contribution to Pdi on the EMG-DI using an esophageal electrode in 5 normal men during fatiguing contractions. Falls in the H/L occurred in all subjects regardless of the relative contribution of Pes and Pga to Pdi at a tension-time index of 0.29 +/- 0.1 (mean +/- SE). However, the time constant of decay of the H/L varied widely among subjects when Pes predominated (173.9 +/- 45.7 s; coefficient of variation, 58.7%) or Pga predominated (78.4 +/- 19.4 s; coefficient of variation, 55.2%), whereas it was consistent among subjects when Pes and Pga were equal (72.3 +/- 3.8 s; coefficient of variation, 11.6%). In addition, a significant relationship was found between the mean integrated activity of the EMG-DI compared with maximum and the time constant of decay of the H/L (r = 0.57, p less than 0.03). We conclude that the differences in the rate of decay of the H/L was at least partly a result of differences in the relative activation of the crural diaphragm at the same Pdi.

Adult↗

PubMed1987

Anatomy of diaphragmatic circulation.

The diaphragmatic circulation was studied in 48 mongrel dogs weighing 10-35 kg by injecting acrylic coloring into the arteries and veins of the diaphragm. The phrenic arteries and internal mammary arteries were found to anastomose head to head, forming an internal arterial circle around the medial leaflet of the diaphragm tendon. This arterial circle emitted vascular branches that traveled between muscle fibers toward the periphery of the diaphragm. These branches anastomosed with vessels of the intercostal arteries to form costophrenic arcades all along the fibers of the crural and costal diaphragms. The intercostal arteries were anastomosed to one another by small vessels within the muscular diaphragm, thus forming an arterial ring around the insertions of the diaphragm on the ribs. The venous drainage has an anatomic distribution similar to that observed on the arterial side, but with the additional presence of valves that could play a role in directing blood flow.

Animals↗