Effect of Carotid Sinus Stimulation on Cardiac Output and Peripheral Vascular Resistance during Changes in Blood Volume Distribution in Man

In ten healthy subjects (mean age 29.6 years) the hemodynamic response to carotid sinus stimulation (neck suction - 40 mmHg) was studied under control conditions and during peripheral pooling of blood (lower body negative pressure). Heart rate, arterial and central venous pressure, cardiac output and forearm blood flow were measured. The time sequence of the heart rate response was studied separately in six healthy subjects. During control conditions, carotid sinus stimulation induced a significant decrease in arterial pressure and heart rate. The blood pressure decrease mainly reflected a reduction in cardiac output, total peripheral vascular resistance being essentially unchanged. However, in the skeletal muscle, represented by a forearm segment, vascular resistance decreased significantly. During lower body negative pressure (LBNP) the same stimulation of the carotid sinus induced a significantly greater fall in mean arterial pressure even though the reduction in cardiac output was slightly smaller on the average than in the control condition. The heart rate increased, probably secondary to a time dependent increase in heart rate elicited by the continuous LBNP stimulus. Total peripheral vascular resistance decreased significantly during LBNP, the reaction likewise differing significantly from that in the control condition. Thus the augmented blood pressure response was due to a more pronounced vasodilatation when the carotid sinus was stimulated during lower body negative pressure. The results indicate that the hemodynamic changes elicited by carotid sinus stimulation are modified by changes in the distribution of blood volume and in the tone of resistance vessels.     

Possible contribution of central gamma-aminobutyric acid receptors to resting vascular tone in freely moving rats

Previous studies have shown that central administration of GABA (gamma -aminobutyric acid), an inhibitory neurotransmitter, preferentially reduces hindquarters and carotid vascular resistances but not renal and coeliac vascular resistances in conscious rats. This study tested the hypothesis that these preferential actions of central GABA receptors are related to differences between vessels in resting autonomic vascular tone in freely moving rats. Rats were chronically implanted with intracisternal cannulas and/or electromagnetic probes to measure regional blood flows. In response to GABA administration, the changes in vascular resistance (arterial blood pressure/regional blood flow) of the hindquarters (n = 23) and carotid (n = 12) vascular beds were significantly and negatively correlated with basal vascular resistance. No such relationship was found for the renal (n = 21), coeliac (n = 13) and superior mesenteric (n = 23) vascular beds. This finding indicates that the responsiveness to GABA of brainstem pathways controlling the hindquarters and carotid vascular beds co-varies with resting resistance in hindquarters and carotid vessels. A similar analysis was performed, correlating the ongoing vascular resistance of each vessel with its response to ganglionic blockade by chlorisondamine. In this case, a significant negative correlation was also found for the hindquarters (n = 26) and carotid (n = 15) vascular beds, but not for the coeliac (n = 17) or superior mesenteric (n = 19) vessels. Together, these findings suggest that central GABA receptors accessible from the cisterna magna preferentially affect two vascular beds which, in the freely moving rat, show resting autonomic vascular tone.     

Redistribution of cardiac output during grooming of the rat

Changes in blood flow in the carotid, superior mesenteric, or renal artery or terminal aorta during grooming were measured in the conscious rat with a chronically implanted electromagnetic flow probe. Arterial pressure was measured simultaneously with an indwelling catheter. During grooming, carotid flow almost doubled and arterial pressure rose slightly, while hindquarter (terminal aortic) flow decreased by an average of about 20%. Superior mesenteric flow decreased occasionally and renal flow remained almost unchanged. Heart rate increased by an average of about 10%. An increase in carotid flow during grooming was similarly observed after cervical sympathectomy. After ganglion blockade with hexamethonium, grooming did not decrease hindquarter flow any more. These findings suggest that during grooming a substantial amount of blood flow is switched from the hindquarters to the carotid areas, to achieve a redistribution of cardiac output largely by excitation of the respective regional sympathetic and parasympathetic fibers and that sympathetic vasoconstrictor fibers supplying different vascular areas are controlled by separate pathways from separate neuron groups.     

Influence of carotid baroreceptors on different components of the vascular system

1. Reflex changes in wall tension of the lateral saphenous vein of one hind limb, the splenic veins and capsule, and the resistance vessels of the other hind limb caused by changes in baroreceptor activity were measured in vagotomized dogs under thiopentone-chloralose anaesthesia.2. Three different methods were used to alter pressure in one or both carotid sinuses. (1) Both carotid sinuses were vascularly isolated and filled with fully oxygenated Krebs-Ringer bicarbonate solution (pH 7.4) from a reservoir in which the pressure could be altered at will. (2) One sinus was denervated, and the contralateral sinus was perfused with arterial blood at different flow rates. (3) One sinus was denervated, and the innervated sinus was perfused with arterial blood at constant flow, the pressure being altered by changing the outflow resistance.3. The left saphenous vein was perfused at constant flow with autologous blood; changes in perfusion pressure were used as a measure of changes in veno-motor activity. The right common iliac artery was perfused at constant flow to measure changes in resistance vessel activity. Blood flow through the spleen was temporarily arrested, trapping a fixed volume of blood in the organ. Under these conditions, changes in splenic vein pressure were a measure of changes in smooth-muscle tension in the splenic capsule and veins.4. In order to assess the responses to baroreceptor stimulation in terms of alterations in sympathetic nerve traffic to different components of the peripheral vascular system, ;frequency-response curves' were constructed for spleen, saphenous vein, and limb resistance vessels by electrical stimulation of the splenic nerves and lumbar sympathetic chains.5. The saphenous vein showed no consistent response to changes in baroreceptor activity. Reduction in carotid sinus pressure from 180 to 100 mm Hg caused an increase in venous pressure in the isovolumetric spleen and in the iliac artery perfusion pressure. These results were confirmed by electrical stimulation of the carotid sinus nerve. Whereas the peak responses of the limb resistance vessels corresponded to an increase in lumbar sympathetic nerve traffic of 6-10 c/s, the maximal splenic responses were equivalent to an increase in splenic nerve traffic of 1-4 c/s. These results are consistent with selective autonomic nervous control of different components of the peripheral vascular system.     

Lumbar sympathetic nerve activity and hindquarter blood flow during REM sleep in rats

The present study aimed to investigate the response of lumbar sympathetic nerve activity (LSNA) to the onset of rapid eye movement (REM) sleep and its contribution to the regulation of muscle blood flow during REM sleep in rats. Electrodes for the measurements of LSNA, electroencephalogram, electromyogram and electrocardiogram and a Doppler flow cuff for the measurements of blood flow in the common iliac and mesenteric arteries, also catheters for the measurements of systemic arterial and central venous pressures were implanted chronically. REM sleep resulted in a step increase in LSNA, by 22 ± 9% (mean ± s.e.m. , P < 0.05), a reduction of iliac vascular conductance, by −16 ± 3% ( P < 0.05) and a gradual increase in systemic arterial pressure, reaching a maximum value of 8.1 ± 2.0 mmHg ( P < 0.05) at 89 s after onset of REM sleep, while mesenteric vascular conductance increased simultaneously by 5 ± 2% ( P < 0.05). There was a significant (Pearson's correlation coefficient = 0.94, P < 0.05) inverse linear relationship between LSNA and the iliac blood flow. Unilateral lumbar sympathectomy blunted the reduction of iliac blood flow induced by the onset of REM sleep. The present observations suggest that the onset of REM sleep appears to be associated with a vasodilation in viscera and a vasoconstriction in skeletal muscle, such that systemic arterial pressure increases during REM sleep in rats.     

Peripheral blood flows during colorectal distension in anaesthetised dogs

Distension of the descending colon elicits reflex cardiovascular responses, including increases in heart rate and arterial blood pressure. To study the relative contribution of vasoconstriction in individual vascular beds to this reflex response, experiments were performed on seven dogs anaesthetised with chloralose and instrumented with electromagnetic flowmeters around the superior mesenteric, the left renal and the left external iliac arteries. The colorectal portion of the intestine was distended at constant pressure (36.6 mm Hg, 4.9 kPa mean; range 25–50 mm Hg, 3.3–6.7 kPa) with warm Ringer solution for periods of 2 min. After a set of control distensions, the experiments were performed whilst the reflex rise in arterial pressure was prevented by removal of blood from the arterial tree. In control distensions arterial pressure increased by 11.3±1.5 mm Hg, 1.51±0.12 kPa (mean±SEM). In distensions at constant arterial pressure, peripheral blood flows were altered to different extents in the three territories studied: vascular resistance increased by 30.8±5.6% (P<0.01) in the mesenteric, by 4.1±1.5% (P<0.03) in the renal, and by 15.2±6.8% (NS) in the external iliac bed. We conclude that colorectal distension may reflect activation of a function-specific pathway of the sympathetic nervous system, which leads to much greater vasoconstriction in the splanchnic circulation than in renal or musculocutaneous circulations.     

Reflex limb dilatation following norepinephrine and angiotensin II in conscious dogs.

Effects of intravenous and intra-arterial norepinephrine (NE) and angiotensin II (AN) were compared in 18 conscious dogs instrumented with Doppler or electromagnetic flow probes on the iliac, mesenteric, and renal arteries, and catheters in the aorta and iliac arteries. NE and AN administered intravenously constricted the mesenteric and renal beds, and constricted the iliac bed when administered directly into the iliac artery. In contrast, intravenous NE and AN caused striking reflex increases in iliac flow and reductions in iliac resistance, respectively, in 12 of 18 dogs studied. The reflex iliac dilatation was not prevented by beta blockade with propranolol, cholinergic blockade with atropine, or prostaglandin synthetase inhibition with indomethacin. However, the responses were abolished by either phentolamine, 1 mg/kv iv, or after local blockade of the limb with either phentolamine, 0.5 mg/kg, or with tripelennamine, 2 mg/kg. The dilatation was not prevented by either bilateral carotid sinus and aortic nerve section or by bilateral vagotomy alone, but was prevented by a combination of these procedures. Thus, intravenous NE and AN cause striking reflex iliac dilatation in the limb in the conscious dog; the afferent arc of this reflex involves both arterial baroreceptor and vagal pathways, while the efferent mechanism involves an interaction of alpha-adrenergic and histaminergic receptors.     

The role of the carotid body chemoreceptors and carotid sinus baroreceptors in the control of cerebral blood vessels

1. Cerebral blood flow was measured in 17 baboons, anaesthetized with pentobarbitone, paralysed with gallamine and mechanically ventilated and in which the right sinus and both aortic nerves had been cut and the left carotid sinus vascularly isolated. Later in each experiment, the head was artificially perfused with femoral arterial blood via the innominate artery.2. Stimulation of the carotid body chemoreceptors with venous blood invariably caused a rise in regional cerebral blood flow whether the head was naturally or artificially perfused. This response was almost completely abolished if the VIIth cranial nerves were cut intracranially.3. Regional cerebral blood flow varied inversely with carotid sinus pressure.4. After the remaining (left) sinus nerve had been cut, the cerebral vascular response to hypoxia was negligible and the response to hypercapnia was markedly reduced. Blood flow then varied with perfusion pressure.5. These results provide further evidence that cerebral blood vessels are reflexly controlled and that the peripheral arterial receptors are involved. Their action is most conspicuous in the vascular response to hypoxia and together with intrinsic factors in the cerebral vascular bed, they determine the size of the vascular response to changes in CO(2) and pressure.     

Blood pressure, heart rate, and regional resistance in behavioral defense

Rats were instrumented with arterial catheters and directional pulsed Doppler flow probes for measurement of arterial blood pressure, heart rate, and blood flow in the renal, mesenteric, and hindquarter vascular beds. When tested as intruders in a resident-intruder aggression test, subjects responded to resident attack with species-characteristic defensive behavior and the “defense reaction” pattern of increased heart rate, renal resistance, and mesenteric resistance, and decreased hindquarter resistance. Blood pressure was variable, but sustained increases in blood pressure were rarely observed. The maintenance of blood pressure during species-specific defensive behavior can be contrasted with the sustained pressor responses observed in the centrally elicited defense reaction. The combination of pulsed Doppler technology and the residentintruder paradigm appears to be a promising method for investigating the mechanisms of cardiovascular adjustment to behavioral defense.     

Blood flow redistribution during spontaneous wheel walk of the rat

Changes in regional blood flow and arterial pressure in the rat during spontaneous walk in a wheel were observed. An electromagnetic flow probe was implanted around the carotid, superior mesenteric, or renal artery, or the terminal aorta and a catheter for pressure measurement was inserted into the terminal aorta or the common carotid artery. The wheel had a diameter of 35 cm and rotated passively as the rat walked. When hindquarter (terminal aortic) flow increased markedly during wheel walk, carotid flow decreased, superior mesenteric flow decreased or remained unchanged, and renal flow did not change. Arterial pressure remained almost unchanged and heart rate increased an average of about 10%. Semiquantitative considerations indicated that arterial pressure was maintained in the face of the profuse increase in hindquarter flow during wheel walk by an increase in cardiac output rather than shifts of blood flow from other regions.     

Regional hemodynamic responses to exogenous catecholamines and vasoactive peptides in conscious rats.

The vasoactive hormones vasopressin, angiotensin II, adrenaline, and noradrenaline may all be released after central neural stimulation, but our knowledge of their relative actions on regional vascular resistances is incomplete. A comprehensive account of these patterns will help our understanding of their contributions to centrally generated patterns of blood flow. In the present study, regional blood flows and resistances of five major arteries were measured during sequential intravenous infusions of a range of doses of each substance in conscious rats. Vasopressin strongly increased superior mesenteric, hindquarters, carotid, and renal resistance but did not affect celiac artery resistance until the highest infusion rate. Both angiotensin II and noradrenaline increased resistance, although to different extents, in all vascular beds except the hindquarters, which was unaffected. Adrenaline infused at pressor rates markedly increased superior mesenteric resistance while moderately increasing renal, carotid and celiac arterial resistances. At subpressor rates, however, adrenaline increased celiac resistance but decreased hindquarters vascular resistance without significantly affecting the other beds. It is concluded that each vasoactive substance released into the systemic circulation causes its own characteristic pattern of vascular responses. This information should be useful for understanding the humoral basis of hemodynamic responses to central stimuli.     

Acute effects of captopril on blood pressure and regional blood flows in two types of renovascular hypertensive rats in conscious state

Mean arterial pressure and three regional blood flows--renal, superior mesenteric, and hindquarter flows--were monitored in conscious normotensive control rats (NCR), in two-kidney, one-clip (two-kidney), and in one-kidney, one-clip (one-kidney) hypertensive rats. After administering captopril, an angiotensin-converting enzyme inhibitor, mean arterial pressure decreased in all three groups. However, lowering of arterial pressure was not statistically significant in any of the groups. Renal flow increased significantly in the intact artery of two-kidney hypertensive rats, whereas it did not significantly increase in the clipped artery of one-kidney hypertensive rats and in the intact artery of NCR. Superior mesenteric flow increased significantly only in one-kidney hypertensive rats. Hindquarter flow did not significantly change in the three groups. Regional resistance reduced significantly in not only renal but also in superior mesenteric vascular area in two-kidney hypertensive rats, whereas it did not reduce significantly in these two vascular areas in one-kidney hypertensive rats and in NCR. The present findings show that the renin-angiotensin-mediated vasoconstriction plays a role in not only renal but also in superior mesenteric vascular area in two-kidney hypertensive rats, whereas it hardly plays a role in these three vascular beds in one-kidney hypertensive rats.     

Effects of nitroglycerin on cardiac function and regional blood flow distribution in conscious dogs.

The effects of intravenous infusion of nitroglycerin (NTG), 8 and 32 microgram/kg.min for 7 min, and of sublingual NTG, 1.2 mg, were examined on direct and continuous measurements of systemic, coronary, and regional hemodynamics, left ventricular (LV) dimensions, pressures, and myocardial contractility in conscious dogs. NTG induced sustained reductions in LV dimensions and transient increases in heart rate and dP/dt, and decreases in mean arterial pressure. Initially NTG increased cardiac output and flows to the coronary, mesenteric, renal, and iliac beds, while systemic and regional vascular resistances fell. Later, cardiac output, cardiac work, and mesenteric and iliac flows fell significantly below control, and significant vasoconstriction in the systemic as well as mesenteric, iliac, and coronary beds was observed at a time when LV end-diastolic dimensions were still significantly reduced. Peripheral vasoconstriction was not observed with systemic NTG in deafferented dogs or when NTG, 1 microgram/kg.min, was infused intra-arterially into the iliac bed. Thus, systemic NTG induces a biphasic response consisting of initial arteriolar vasodilation followed by vasoconstriction in the mesenteric, iliac, coronary and systemic beds, which is presumably due to longer lasting effects on preload and to secondary reflex responses to the drug.     

Effects of haemorrhage on the distribution of the peripheral blood flow in the rabbit

1. The effects of bleeding unanaesthetized rabbits by 26% of their blood volume on the blood flow in the portal, renal, muscle and skin beds were investigated in normal animals, in animals without functioning autonomic effectors, and in animals with section of the carotid sinus and aortic nerves.2. In animals without functioning autonomic effectors there was progressive vasodilatation during the 4 hr following haemorrhage, which differed markedly in the different regional beds studied. The dilatation was greatest in the portal bed, followed by kidney and skin, but there was no significant change in the vascular resistance in muscle.3. In normal animals with intact reflexes the vascular resistance either increases or remains at control values, suggesting that reflex constrictor effects oppose locally acting dilator mechanisms. During the 4 hr following haemorrhage reflex vasoconstrictor effects were greatest in kidney, followed by muscle, portal bed and skin.4. In animals with section of the carotid sinus and aortic nerves reflex constrictor effects were absent in the portal, muscle and skin beds, but significant vasoconstriction was still evident in the renal bed, though of smaller magnitude than in normal animals. The results suggest that the arterial baroreceptors are a major source of reflex activity following haemorrhage, but that other reflexogenic zones contribute to the renal effects in normal animals.     

Circulatory effects of carotid sinus stimulation and changes in blood volume distribution in hypertensive man

In 8 patients with moderate hypertension and 8 normotensive subjects an attempt was made to study the circulatory effects of high and low pressure baroreceptor stimulation. Intrathoracic low pressure receptors were stimulated by changes in blood volume distribution using lower body negative pressure (LBNP) and lower body positive pressure (LBPP). The carotid sinus was stimulated by sinusoidal neck suction. Blood pressure, central venous pressure, heart rate, cardiac output and forearm blood flow were recorded. During LBNP and LBPP changes in central blood volume, reflected in changes in central venous pressure, induced significantly greater changes in cardiac output and forearm blood flow in the hypertensive subjects. In both normotensive and hypertensive subjects mean arterial blood pressure was essentially unchanged during LBNP and a slight increase was found during LBPP. Heart rate and blood pressure response to stimulation of the carotid sinus decreased with increasing resting mean arterial pressure. The results suggest impairment of reflex adjustments, via arterial baroreceptors, possibly in particular to dynamic stimuli, rather than via intrathoracic “low pressure” baroreceptors in subjects with moderate hypertension.     

Effects of graded pulsatile pressure on the reflex vasomotor responses elicited by changes of mean pressure in the perfused carotid sinus-aortic arch regions of the dog

1. In the anaesthetized dog, the carotid sinuses and aortic arch were isolated from the circulation and separately perfused with blood by a method which enabled the mean pressure, pulse pressure and pulse frequency to be varied independently in each vasosensory area. The systemic circulation was perfused at constant blood flow by means of a pump and the systemic venous blood was oxygenated by an extracorporeal isolated pump-perfused donor lung preparation.2. We have confirmed our previous observations that under steadystate conditions the vasomotor responses elicited reflexly by changes in mean carotid sinus pressure are modified by alterations in carotid sinus pulse pressure, whereas those evoked by changes of mean aortic arch pressure are only weakly affected by modifications of aortic pulse pressure.3. When the carotid sinus and aortic arch regions are perfused in combination at constant pulse frequency (110 c/min), the relationship between mean carotid sinus-aortic arch pressure and systemic arterial perfusion pressure is dependent on the size of the pulse pressure.4. Increasing the pulse pressure alters the curve relating the mean carotid sinus-aortic arch pressure to systemic arterial perfusion pressure in such a way that the perfusion pressure is lower at a given carotid sinus-aortic arch pressure within the range 80-150 mm Hg. The larger the pulse pressure, up to about 60 mm Hg, the greater the fall in systemic arterial perfusion pressure. Above a mean carotid sinus-aortic arch pressure of about 150 mm Hg, alterations of pulse pressure have little effect.5. There is a family of curves representing the relation between mean carotid sinus-aortic arch pressure and systemic vascular resistance, depending on the pulse pressure.     

Interplay among carotid sinus, cardiopulmonary, and carotid body reflexes in dogs.

Interactions among vascular reflexes evoked from carotid sinuses, carotid bodies, and cardiopulmonary region were examined in anesthetized, atropinized, and respired dogs with aortic nerves cut. The carotid sinuses were perfused at 220, 150, and 40-50 mmHg; the chemoreceptors were stimulated by perfusion with hypoxic hypercapnic blood. Cardiopulmonary vasomotor inhibition was interrupted by vagal cold block. Measurements were made of arterial blood pressure and of kidney and hindlimb vascular resistance. At sinus pressures less than 170-160 mmHg, cardiopulmonary vasomotor inhibition increased with increase in blood volume. At high sinus pressure, interruption of this augmented cardiopulmonary inhibition was as ineffective in changing vascular resistance as interruption of the lesser inhibition present during normovolemia. Chemoreceptor stimulation increased the response to vagal block at intermediate but not at high or low sinus pressure. The studies demonstrate the dominant role of the carotid sinus reflex when the three systems interact and the ineffectiveness of chemoreceptor stimulation when carotid or cardiopulmonary inhibition is maximal.     

Differences in brain and liver blood flow during and after acute blood loss in rats with different resistance to circulatory hypoxia

Variations of blood flow and vascular resistance in the common carotid arteries and of blood flow in the hepatic artery and portal vein are examined during and after acute massive blood loss in rats with low and high resistance to circulatory hypoxia. In rats with low resistance, arterial pressure and the rates of cerebral and hepatic blood macro- and microflow, which have decreased during blood loss, continued to fall during the posthemorrhagic period. After cessation of bleeding, a transient arterial pressure rise to 70 mm Hg is observed in rats with high resistance, while the blood flow via carotid arteries increases to 65% of its initial value, being maintained at this level throughout the period of changes in carotid vascular resistance; intrinsic hepatic arterial blood flow increases to 115% of baseline value, while the portal vein blood flow and hepatic microflow increase to 75%. Translated fromByulleten' Eksperimental'noi Biologii i Meditsiny, Vol. 123, No. 3, pp. 253–257, March, 1997     

Vascular area with marked resistance elevation in one-clip, two-kidney renovascular hypertensive rats

Blood flow in the terminal aorta or superior mesenteric artery was recorded in conscious one-clip, two-kidney renovascular hypertensive rats and normal control rats with a chronically implanted electromagnetic flow probe. An indwelling catheter for direct measurement of arterial pressure was inserted into the terminal aorta in rats with a flow probe around the superior mesenteric artery. In the hindquarter area supplied by the terminal aorta, the elevation of vascular resistance in hypertensive rats in relation to normal rats was close to the average over the whole body. However, it was about 40% more intense than the average in the superior mesenteric area. Quantitatively, the hindquarter and superior mesenteric beds in hypertensive rats contributed about a quarter and a third of the decrease in total vascular conductance, respectively. Plots of superior mesenteric flow per body weight against arterial pressure at rest for 51 measurements in 9 hypertensive rats revealed an inverse relation between these two variables. These findings indicate that elevation of vascular resistance in the superior mesenteric area and probably the splanchnic area in general play an important role in one-clip, two-kidney renovascular hypertension.     

Interactions between carotid sinus mechanoreceptor and chemoreceptor reflex loops

Summary Our experiments were designed to evaluate the combined influence of the mechanoreceptor and chemoreceptor control loops on the cardiovascular and respiratory systems. Both carotid sinus areas were isolated by a complete blind sac preparation in chloralose anesthetized dogs in which a flow probe had been implanted previously on the ascending aorta. The following variables were continuously monitored: central aortic pressure, heart rate, cardiac output, vascular resistance of a hind limb, tidal volume, respiratory rate, and oxygen consumption. While one sinus was submitted to step changes of pressure, the opposite sinus was kept at constant mean pressure. In addition, one sinus was perfused with either oxygenated of deoxygenated blood at constant flow rate.These studies show a definite interaction between the control loops initiated by the carotid sinus mechanoreceptors and those initiated by the chemoreceptors. The influence of the chemoreceptors is most marked in the peripheral beds, where it shifts the whole response curve of the mechanoreceptor control loop upward. This response accounts for most of the upward shift of the pressure response curve.On the other hand, the mechanoreceptor loop affects the ventilation, but this effect is range dependent, being most marked for the low pressure perfusion of the carotid sinus.This work has been supported in part by PHS Research grant HL 11747