Control Mechanisms Autoregulation
It should be apparent from this survey that the renal vasculature potentially plays a number of functional roles, many of which are as yet incompletely defined. The control of regional perfusion rates within the kidney and of the intrarenal distribution of blood flow has major consequences not only for overall cardiovascular homeostasis but also for renal function.
An analysis of the determinants of regional intrarenal perfusion rates requires an examination of the effects of perfusion pressure and the influence of a number of endogenous vasoactive factors, including the renin-angiotensin system, norepinephrine and the sympathetic nervous system, the prostaglandins and the products of metabolism.
An analysis of the determinants of regional intrarenal perfusion rates requires an examination of the effects of perfusion pressure and the influence of a number of endogenous vasoactive factors, including the renin-angiotensin system, norepinephrine and the sympathetic nervous system, the prostaglandins and the products of metabolism.
Renal blood flow remains almost constant over a wide range of perfusion pressure, from about 80 to 180 mmHg (Selkurt 1946b); glomerular filtration rate also remains constant over much of this range (Forster & Maes 1947). The constancy has utility in preventing wide swings in the load presented to the tubular system for reabsorption with changes in perfusion pressure.
Kiil et al. (1969b) pointed out that the autoregulatory processes could also play a role in maintaining glomerular filtration rate in the face of the increased tubular pressures associated with high urine flow rates. Acute ureteral obstruction, on the one hand, and massively increased urine flows induced by diuretics, or by an osmotic or water load, on the other, all result in vasodilatation, an increase in renal blood flow and maintained filtration despite increased proximal tubular pressure.
Kallskog and Wolgast( 1975) extended these studies with the demonstration by micropuneture that glomerular capillary pressure was elevated during ureteral obstruction, which they attributed to afferent arteriolar dilatation.
Autoregulation of, renal blood flow occurs within seconds, and is prevented neither by denerva'tion nor by perfusion with fluids free of red blood cells. The increase in vascular resistance which accompanies an increase in perfusion pres~ sure must be the result of intrarenal mechanisms (Thurau & Kramer 1959, Thurau 1964). The constancy of the filtration rate first identified by Forster and Maes (1947) pointed to a preglomerular location of the autoregulatory resistance change, as discussed below.
The autoregulatory capacity of the kidney clearly depends on the state and reactivity of the. arterioles. The functional integrity of the vascular smooth muscle must be involved, since autoregulation is regularly blunted by the administration of a number of unrelated vasodilator agents iwhich act directly on smooth muscle, including papaverine, procaine, chloral hydrate, dopamine, prostaglandins and acetylcholine (Thurau & Kramer 1959, Thurau 1964, Baer et al. 1973).
Consistent with that hypothesis, Schmid et al. (1964) demonstrated that autoregulation became ineffective when the renal vessels lost tone, and was restored by administration of a vasoconstrictor such as norepinephrine. This interpretation is supported by the recent demonstration by Ono et al, (1974) that calcium antagonists, which interfere with smooth muscle function, blunt the renal autoregulatory process.
Many hypotheses have been elaborated in an attempt to account for autoregulation. It is convenient to classify the hypotheses into several broad categories, including the myogenic, vasoactive, metabolic and mechanical. According to the ‘myogenic’ hypothesis the stimulus for the vascular smooth muscle contraction in response to increasing intraluminal pressure is either the transmural pressure itself, or the increase in tangential tension of the vascular wall (Thurau & Kramer 1959, Folkow 1964, Folkow & Langston 1964, Schmid et al. 1964, Johnson 1964).
According to this hypothesis the vascular smooth muscle is pressure or tension sensitive. The vasoactivefactor hypothesis suggests that local vasoactive substances such as angiotensin II or the prostaglandins are released in response to. changes in perfusion pressure, appropriately changing local vascular resistance (Thurau & Levine, 1973, Herbaczynska-Cedro & Vane 1973). Both of these hypotheses in a number of forms are of current interest and will be analysed further.
The other hypotheses are of historical interest only (J ohnson 1964). The mechanical hypothesis suggested that changes in resistance were due not to active smooth-muscle function but rather were due to physical phenomena such as compression of capillaries by changes in transmural pressure or changes in effective blood viscosity through pressure-dependent changes in the dynamic intravascular haematocrit.
The metabolic hypothesis suggested that changes in perfusion induced a local change in the milieu either through changes in the delivery of oxygen or the removal of products of metabolism.
While a definitive conclusion cannot yet be drawn. the weight of available evidence favours the myogenic hypothesis, as first suggested by Bayliss (1902). Certainly the rapidity of the response (Folkow & Langston 1964) and the effects of vasoactive agents are most compatible with a myogenic basis for autoregulation.
The major difficulty with this hypothesis has been in relating the stimulusw-presumably tangential tension in the arteriolar wall-opto the response (Thurau & Levine 1971). Folkow (1964) anticipated this problem, and pointed out that in a multi-unit, phasically contracting smooth muscle, such as a vascular bed, the rate or frequency, rather than the degree, of contraction could be modilied by transmural pressure or tension.
Golenhofen (1965) suggested, similarly but more specifically that stretch activation may synchronize the myogenic auto-oscillations at the pacemaker region of smooth muscle. Alternative sensing mechanisms suggested include a length sensor in parallel with the contractile elements (Speden 1973) or, vascular stretch receptors in the glomeruli (Raeder et a]. 1975), in addition to or instead of transmural pressure, per se.
While considerable and attractive-but still circumstantial-evidence links tubular function to the autoregulatory process, as reviewed below, it is important to remember that many vascular beds autoregulate (Johnson 1964, Folkow 1964). Only the kidney has a macula densa!
Speculation about another type of control mechanism based on information derived from tubular function arose from several sources.
Goormaghtigh (1942) pointed out the proximity of the distal tubule to the vascular structures of the same nephron’s glomerulus and speculated on the possibility of local control. Guyton et al. (1964) were equally impressed with the anatomical arrangement, and on the basis of analogue modelling proposed that renal perfusion and glomerular filtration were directly dependent on the osmolality of tubular fluid reaching the macula densa-which provided a route by which information from the distal nephron might be transferred to the glomerular vasculature.
Thurau (1964) assembled information from a number of sources which supported the possibility that it was the sodium concentration in fluid reaching the macula densa to which the feedback loop was sensitive, and conceptually incorporated the renin-angiotensin system into the loop as the efferent effector mechanism.
Autoregulation of the renal blood flow and glomerular filtration was thus proposed as being both a manifestation of, and an important link in, a sodium-preserving feedback mechanism which involves the distal tubule, its macula densa, the juxtaglomerular apparatus, renin release, angiotensin generation and thus renal vasoconstriction (Thurau 1964).
According to this hypothesis, ina creased sodium reabsorption at the macula densa results in renin production in the juxtaglomerular cells of the afferent arteriole, thus raising the local interstitial angiotensin concentration and consequently resulting in vasoconstriction of the afferent arteriole. Conversely, reduced sodium reabsorption results in reduced renin production and the afferent arterioles thus dilate.
This hypothesis implies that the tone of afferent arterioles would be influenced constantly by the rate of renin release and angiotensin generation and also implies that the autoregulatory response would be mediated by renin release and local angiotensin generation.
Schnermann et al. (1970) provided the first direct evidence that nephron glomerular filtration rate was inversely related to the delivery of sodium chloride to the distal segments of the nephron, supporting the existence of a negative feedback loop. They went on to demonstrate (Schnermann et al. 1973) a striking non-linear relationship between loop of Henle perfusion rate and glomerular hydrostatic pressure.
Glomerular capillary pressure was unresponsive to changes in distal flow delivery from O to 13 nl/min; showed an inverse sigmoid relationship over the flow range l3-32 nl/min; and reached a maximum reduction at 32 nl/min. They suggested a teleologically useful relationship between distal sodium delivery and glomerular filtration rate. Increased sodium delivery to the distal tubule, with the possibility of sodium wasting, will docrease the filtration rate of the nephron involved, Via afferent arteriolar vasoconstriction.
Conversely, with a reduction of the distal sodium load, either intrarenal regulation is abolished or proportional changes in afferent and efferent ’ resistance occur. This would be useful, for example, when the decreased distal sodium load reflected the response to hypovolaemia: an increase in the filtration rate under that circumstance would have unfortunate consequences for the organism.
Further support for this hypothesis came from the demonstration (Granger et al. 1972) that there is suflicient converting enzyme and renin substrate in the juxtaglomerular apparatus to result in local angiotensin generation and thus the local control of single nephron glomerular filtration rate. Moreover, juxtaglomerular renin is responsive to change in distal sodium delivery.
Additional support has come from technical factors in the measurement of single nephron filtration rate, since the standard quantitative micropuncture collection requires the placement of an oil drop in the proximal tubule and thus interrupts flow to the loop of Henle. According to Schnermann’s hypothesis, this should result in an increase in nephron glomerular filtration rate.
Several investigators have demonstrated a difference in single nephron glomerular filtration rate when samples are collected from proximal and distal sites in the rat and dog (Schnermann et al.1970, Davies et al. 1972, Navar et a]. 1974). Further evidence in support of the concept that a distal tubular feedback mechanism participates in the autoregulatory control of glomerular function was provided by the observation that interruption of distal delivery interferes with the autoregulatory response (Navar et al. 1975).
On the other hand, a number of observations are not in accord with this hypothesis. Dilatation of the afferent arteriole which occurs with decreasing perfusion pressure promotes renin release (Vander 1967, Kiil 1975) and an increase in the interstitial concentration of angiotensin II (Bailie et al. 1972). Thus vasodilatation is associated with an increase in the local generation and concentration of angiotensin II.
Moreover, a number of manoeuvrcs which interrupt the renin angiotensin axis, such as antibody specific for angiotcnsin ll (liidc cl al. 1973a), converting enzyme inhibitots ((lagnon or al. 1974), [i-adrenergic blocking agents and angiotcnsin antagonists (Anderson of al. 1975, Carlson & Schramm 1976), do not influence the autoregulatory process.
While it could be argued that these agents did not achieve effective concentrations at the angiotcnsin receptor, Ride et al. (1973a) demonstrated effective arteriolar concentrations of even the large and polar antibody molecules. Certainly the pharmacologic agents, which are much smaller and very diffusible, must have reached effective arteriolar concentrations within the kidney.
The elegant single nephron studies have also faced serious challenge. Morgan (1971) was unable to demonstrate any influence of elevation of loop perfusion rate on early proximal flow rate. Moreover, Maddox et al. (1974b) and Knox et al. (1974) failed to document a difference in single nephron glomerular filtration rate between proximal and distal collection sites in the rat. Furthermore, both of the latter groups of investigators demonstrated that single nephron filtration rate was autoregulated effectively in the absence of flow to the macula densa.
The demonstration by Dev et al. (1974) that sodium intake modified the sensitivity of the tubulo-glomerular feedback loop perhaps accounts for some of the variability in the reports. An elevation of sodium chloride intake strikingly diminished or abolished feedback responses, so that some of the discrepancy in the reports may well have reflected variation in the quantity of sodium ingested by rats in different laboratories, as suggested by Schnermann and Levine (1975).
It is difficult, however, to attribute the well-sus-tained autoregulatory process in rats after proximal tubular obstruction to differences in sodium intake. Taken in all, the evidence suggests that tubule-glomerular feedback, while it may well have important implications for overall kidney function, probably plays at most a minor role in the autoregulatory process.
Given the interest in the possibility that nephron function in different regions of the kidney contributes to the overall function of the kidney, it is not surprising that attention turned to autoregulatory responses in the outer and inner cortex and the medulla. Thurau et a]. (1960) first reported that medullary blood flow was autoregulated less effectively than blood flow in the cortex, a phenomenon which those investigators used to account for the diuresis which accompanies a pressor response (Thurau & Deetjen 1962).
Diminished or absent autoregulation in the medulla was confirmed by Girndt & Ochwadt (1969) and Stumpe et a]. (1969). Conversely, another series of investigators (Aukland 1967, Wolgast 1968, Grangsjo 1968, Loyning 1971, Nissen 1966) demonstrated effective autoregulation of medullary blood flow.
The differences in these studies can be accounted for neither by species difference nOr by technique for measurement. In fact the only common factor appears to be that the investigators who identified complete medullary autoregulation performed their experiments in Scandinavia, Whereas the experiments in which medullary autoregulation was incomplete or absent were performed in Germany.
To complicate matters further, "two American groups have applied the microsphere method to the assessment of regional‘cortical perfusion with a reduction in arterial pressure, and have found that autoregulation is primarily inner cortical (McNay & Abe 1970, Stein et al. 1973a). To the extent that the medulla receives its blood supply from the inner cortex, one might anticipate from these data that medullary perfusion was autoregulated more effectively than that of the cortex.
A number of recent observations may not only account for the differences in the above studies, but perhaps may even account for the role that geography has played in the discrepancies. Again it may be relevant that sodium intake modifies the sensitivity of the tubulo-glomerular feedback loop (Dev et al. 1974), since Kaloyanides et a]. (1974) found that autoregulatory effrCiency was modifled in the isolated dog kidney by sodium intake prior to nephrecto'my.
A high sodium intake combined with the administration of a sodium-retaining hormone, desoxycorticosterone, impaired the capacity of the kidney to autoregulate blood flow and filtration rate. Moreover, Burger et al. (1976) found that the renal vascular response to anaesthesia and trauma was strikingly modihed by sodium intake prior to anaesthesia and that the vascular responses to the anaesthetic agent were mediated by activation of the renin-angiotensin system.
The major vascular response to the anaesthetic agent occurred in the renal cortex. Taken in all, and given the clear geographic association of the different reports, the intriguing possibility that the differences reflect dietary caprice exists. After all, what a laboratory animal eats as his normal diet prior to study merely reflects what the investigator provides.
Forster and Macs ( 1947) utilized the constancy of glomerular filtration rate to calculate relative preglomerular and postglomerular resistance changes over the autoregulatory range of perfusion pressure. They concluded that a reduction in perfusion pressure induces predominant preglomerular dilatation, and that postglomerular resistance is little altered.
Recently, Abe et a]. (1970) reported a phenomenon which they could only account for on the basis of increasing postglomerular resistance during the autoregulatory response. Acetyleholine infused into the renal artery normally induced an increase in renal blood flow without a change in glomerular tiltration rate. When perfusion pressure was reduced to 100 mmHg, both Iiow and glomerular filtration autoregulation were complete.
Acetyleholine infusion at this point increased blood flow, but the increase in blood flow was associated with a large reduction in glomerular filtration rate. Identical effects of vasodilators when superimposed on autoregulatory vasodilatation are evident in the reports of Thurau and Kramer (1959) and Baer et al. (1970). It is difficult to interpret this response in any way other than a progressive, autoregulatory reduction in afferent arteriolar resistance with the reduction in per= fusion pressure, accompanied by a progressive increase in efferent resistance.
At the lower end of the autoregulatory pressure range it seems likely that afferent arteriolar dilatation is maximal, and that administration of the vasodilator at this point therefore induces predominant or sole efferent arteriolar dilatation-and as a consequence reduces filtration rate. This possibility is supported by the recent study of Robertson et al. (1972) who exploited micropuncture of the accessible surface glomeruli in mutant rats to examine single nephron responses to a progressive reduction in perfusion pressure.
They found that autoregulation was complete down to 80 mmHg, and that over this range not only was there a progressive reduction in afferent arteriolar resistance, but that efferent resistance tended to rise. The mechanisms responsible for an efferent vascular resistance increase under these circumstances have not been delineated, but any hypothesis concerning the autoregulatory response has to account not only for a reduction in afferent arteriolar resistance but also for an increase in efferent arteriolar resistance as perfusion pressure falls.
Certainly an increase in efferent resistance is not consistent with the myogenic hypothesis. Possibly two processes occur simultaneously: a dominant myogenic action on the afferent arteriole and a humoral action on the efferent arteriole, which could well be attributable to angiotensin. Evidence of a predominant efferent arteriolar locus for angiotensin action is reviewed below.
Kiil et al. (1969b) pointed out that the autoregulatory processes could also play a role in maintaining glomerular filtration rate in the face of the increased tubular pressures associated with high urine flow rates. Acute ureteral obstruction, on the one hand, and massively increased urine flows induced by diuretics, or by an osmotic or water load, on the other, all result in vasodilatation, an increase in renal blood flow and maintained filtration despite increased proximal tubular pressure.
Kallskog and Wolgast( 1975) extended these studies with the demonstration by micropuneture that glomerular capillary pressure was elevated during ureteral obstruction, which they attributed to afferent arteriolar dilatation.
Autoregulation of, renal blood flow occurs within seconds, and is prevented neither by denerva'tion nor by perfusion with fluids free of red blood cells. The increase in vascular resistance which accompanies an increase in perfusion pres~ sure must be the result of intrarenal mechanisms (Thurau & Kramer 1959, Thurau 1964). The constancy of the filtration rate first identified by Forster and Maes (1947) pointed to a preglomerular location of the autoregulatory resistance change, as discussed below.
The autoregulatory capacity of the kidney clearly depends on the state and reactivity of the. arterioles. The functional integrity of the vascular smooth muscle must be involved, since autoregulation is regularly blunted by the administration of a number of unrelated vasodilator agents iwhich act directly on smooth muscle, including papaverine, procaine, chloral hydrate, dopamine, prostaglandins and acetylcholine (Thurau & Kramer 1959, Thurau 1964, Baer et al. 1973).
Consistent with that hypothesis, Schmid et al. (1964) demonstrated that autoregulation became ineffective when the renal vessels lost tone, and was restored by administration of a vasoconstrictor such as norepinephrine. This interpretation is supported by the recent demonstration by Ono et al, (1974) that calcium antagonists, which interfere with smooth muscle function, blunt the renal autoregulatory process.
Many hypotheses have been elaborated in an attempt to account for autoregulation. It is convenient to classify the hypotheses into several broad categories, including the myogenic, vasoactive, metabolic and mechanical. According to the ‘myogenic’ hypothesis the stimulus for the vascular smooth muscle contraction in response to increasing intraluminal pressure is either the transmural pressure itself, or the increase in tangential tension of the vascular wall (Thurau & Kramer 1959, Folkow 1964, Folkow & Langston 1964, Schmid et al. 1964, Johnson 1964).
According to this hypothesis the vascular smooth muscle is pressure or tension sensitive. The vasoactivefactor hypothesis suggests that local vasoactive substances such as angiotensin II or the prostaglandins are released in response to. changes in perfusion pressure, appropriately changing local vascular resistance (Thurau & Levine, 1973, Herbaczynska-Cedro & Vane 1973). Both of these hypotheses in a number of forms are of current interest and will be analysed further.
The other hypotheses are of historical interest only (J ohnson 1964). The mechanical hypothesis suggested that changes in resistance were due not to active smooth-muscle function but rather were due to physical phenomena such as compression of capillaries by changes in transmural pressure or changes in effective blood viscosity through pressure-dependent changes in the dynamic intravascular haematocrit.
The metabolic hypothesis suggested that changes in perfusion induced a local change in the milieu either through changes in the delivery of oxygen or the removal of products of metabolism.
While a definitive conclusion cannot yet be drawn. the weight of available evidence favours the myogenic hypothesis, as first suggested by Bayliss (1902). Certainly the rapidity of the response (Folkow & Langston 1964) and the effects of vasoactive agents are most compatible with a myogenic basis for autoregulation.
The major difficulty with this hypothesis has been in relating the stimulusw-presumably tangential tension in the arteriolar wall-opto the response (Thurau & Levine 1971). Folkow (1964) anticipated this problem, and pointed out that in a multi-unit, phasically contracting smooth muscle, such as a vascular bed, the rate or frequency, rather than the degree, of contraction could be modilied by transmural pressure or tension.
Golenhofen (1965) suggested, similarly but more specifically that stretch activation may synchronize the myogenic auto-oscillations at the pacemaker region of smooth muscle. Alternative sensing mechanisms suggested include a length sensor in parallel with the contractile elements (Speden 1973) or, vascular stretch receptors in the glomeruli (Raeder et a]. 1975), in addition to or instead of transmural pressure, per se.
While considerable and attractive-but still circumstantial-evidence links tubular function to the autoregulatory process, as reviewed below, it is important to remember that many vascular beds autoregulate (Johnson 1964, Folkow 1964). Only the kidney has a macula densa!
Speculation about another type of control mechanism based on information derived from tubular function arose from several sources.
Goormaghtigh (1942) pointed out the proximity of the distal tubule to the vascular structures of the same nephron’s glomerulus and speculated on the possibility of local control. Guyton et al. (1964) were equally impressed with the anatomical arrangement, and on the basis of analogue modelling proposed that renal perfusion and glomerular filtration were directly dependent on the osmolality of tubular fluid reaching the macula densa-which provided a route by which information from the distal nephron might be transferred to the glomerular vasculature.
Thurau (1964) assembled information from a number of sources which supported the possibility that it was the sodium concentration in fluid reaching the macula densa to which the feedback loop was sensitive, and conceptually incorporated the renin-angiotensin system into the loop as the efferent effector mechanism.
Autoregulation of the renal blood flow and glomerular filtration was thus proposed as being both a manifestation of, and an important link in, a sodium-preserving feedback mechanism which involves the distal tubule, its macula densa, the juxtaglomerular apparatus, renin release, angiotensin generation and thus renal vasoconstriction (Thurau 1964).
According to this hypothesis, ina creased sodium reabsorption at the macula densa results in renin production in the juxtaglomerular cells of the afferent arteriole, thus raising the local interstitial angiotensin concentration and consequently resulting in vasoconstriction of the afferent arteriole. Conversely, reduced sodium reabsorption results in reduced renin production and the afferent arterioles thus dilate.
This hypothesis implies that the tone of afferent arterioles would be influenced constantly by the rate of renin release and angiotensin generation and also implies that the autoregulatory response would be mediated by renin release and local angiotensin generation.
Schnermann et al. (1970) provided the first direct evidence that nephron glomerular filtration rate was inversely related to the delivery of sodium chloride to the distal segments of the nephron, supporting the existence of a negative feedback loop. They went on to demonstrate (Schnermann et al. 1973) a striking non-linear relationship between loop of Henle perfusion rate and glomerular hydrostatic pressure.
Glomerular capillary pressure was unresponsive to changes in distal flow delivery from O to 13 nl/min; showed an inverse sigmoid relationship over the flow range l3-32 nl/min; and reached a maximum reduction at 32 nl/min. They suggested a teleologically useful relationship between distal sodium delivery and glomerular filtration rate. Increased sodium delivery to the distal tubule, with the possibility of sodium wasting, will docrease the filtration rate of the nephron involved, Via afferent arteriolar vasoconstriction.
Conversely, with a reduction of the distal sodium load, either intrarenal regulation is abolished or proportional changes in afferent and efferent ’ resistance occur. This would be useful, for example, when the decreased distal sodium load reflected the response to hypovolaemia: an increase in the filtration rate under that circumstance would have unfortunate consequences for the organism.
Further support for this hypothesis came from the demonstration (Granger et al. 1972) that there is suflicient converting enzyme and renin substrate in the juxtaglomerular apparatus to result in local angiotensin generation and thus the local control of single nephron glomerular filtration rate. Moreover, juxtaglomerular renin is responsive to change in distal sodium delivery.
Additional support has come from technical factors in the measurement of single nephron filtration rate, since the standard quantitative micropuncture collection requires the placement of an oil drop in the proximal tubule and thus interrupts flow to the loop of Henle. According to Schnermann’s hypothesis, this should result in an increase in nephron glomerular filtration rate.
Several investigators have demonstrated a difference in single nephron glomerular filtration rate when samples are collected from proximal and distal sites in the rat and dog (Schnermann et al.1970, Davies et al. 1972, Navar et a]. 1974). Further evidence in support of the concept that a distal tubular feedback mechanism participates in the autoregulatory control of glomerular function was provided by the observation that interruption of distal delivery interferes with the autoregulatory response (Navar et al. 1975).
On the other hand, a number of observations are not in accord with this hypothesis. Dilatation of the afferent arteriole which occurs with decreasing perfusion pressure promotes renin release (Vander 1967, Kiil 1975) and an increase in the interstitial concentration of angiotensin II (Bailie et al. 1972). Thus vasodilatation is associated with an increase in the local generation and concentration of angiotensin II.
Moreover, a number of manoeuvrcs which interrupt the renin angiotensin axis, such as antibody specific for angiotcnsin ll (liidc cl al. 1973a), converting enzyme inhibitots ((lagnon or al. 1974), [i-adrenergic blocking agents and angiotcnsin antagonists (Anderson of al. 1975, Carlson & Schramm 1976), do not influence the autoregulatory process.
While it could be argued that these agents did not achieve effective concentrations at the angiotcnsin receptor, Ride et al. (1973a) demonstrated effective arteriolar concentrations of even the large and polar antibody molecules. Certainly the pharmacologic agents, which are much smaller and very diffusible, must have reached effective arteriolar concentrations within the kidney.
The elegant single nephron studies have also faced serious challenge. Morgan (1971) was unable to demonstrate any influence of elevation of loop perfusion rate on early proximal flow rate. Moreover, Maddox et al. (1974b) and Knox et al. (1974) failed to document a difference in single nephron glomerular filtration rate between proximal and distal collection sites in the rat. Furthermore, both of the latter groups of investigators demonstrated that single nephron filtration rate was autoregulated effectively in the absence of flow to the macula densa.
The demonstration by Dev et al. (1974) that sodium intake modified the sensitivity of the tubulo-glomerular feedback loop perhaps accounts for some of the variability in the reports. An elevation of sodium chloride intake strikingly diminished or abolished feedback responses, so that some of the discrepancy in the reports may well have reflected variation in the quantity of sodium ingested by rats in different laboratories, as suggested by Schnermann and Levine (1975).
It is difficult, however, to attribute the well-sus-tained autoregulatory process in rats after proximal tubular obstruction to differences in sodium intake. Taken in all, the evidence suggests that tubule-glomerular feedback, while it may well have important implications for overall kidney function, probably plays at most a minor role in the autoregulatory process.
Given the interest in the possibility that nephron function in different regions of the kidney contributes to the overall function of the kidney, it is not surprising that attention turned to autoregulatory responses in the outer and inner cortex and the medulla. Thurau et a]. (1960) first reported that medullary blood flow was autoregulated less effectively than blood flow in the cortex, a phenomenon which those investigators used to account for the diuresis which accompanies a pressor response (Thurau & Deetjen 1962).
Diminished or absent autoregulation in the medulla was confirmed by Girndt & Ochwadt (1969) and Stumpe et a]. (1969). Conversely, another series of investigators (Aukland 1967, Wolgast 1968, Grangsjo 1968, Loyning 1971, Nissen 1966) demonstrated effective autoregulation of medullary blood flow.
The differences in these studies can be accounted for neither by species difference nOr by technique for measurement. In fact the only common factor appears to be that the investigators who identified complete medullary autoregulation performed their experiments in Scandinavia, Whereas the experiments in which medullary autoregulation was incomplete or absent were performed in Germany.
To complicate matters further, "two American groups have applied the microsphere method to the assessment of regional‘cortical perfusion with a reduction in arterial pressure, and have found that autoregulation is primarily inner cortical (McNay & Abe 1970, Stein et al. 1973a). To the extent that the medulla receives its blood supply from the inner cortex, one might anticipate from these data that medullary perfusion was autoregulated more effectively than that of the cortex.
A number of recent observations may not only account for the differences in the above studies, but perhaps may even account for the role that geography has played in the discrepancies. Again it may be relevant that sodium intake modifies the sensitivity of the tubulo-glomerular feedback loop (Dev et al. 1974), since Kaloyanides et a]. (1974) found that autoregulatory effrCiency was modifled in the isolated dog kidney by sodium intake prior to nephrecto'my.
A high sodium intake combined with the administration of a sodium-retaining hormone, desoxycorticosterone, impaired the capacity of the kidney to autoregulate blood flow and filtration rate. Moreover, Burger et al. (1976) found that the renal vascular response to anaesthesia and trauma was strikingly modihed by sodium intake prior to anaesthesia and that the vascular responses to the anaesthetic agent were mediated by activation of the renin-angiotensin system.
The major vascular response to the anaesthetic agent occurred in the renal cortex. Taken in all, and given the clear geographic association of the different reports, the intriguing possibility that the differences reflect dietary caprice exists. After all, what a laboratory animal eats as his normal diet prior to study merely reflects what the investigator provides.
Forster and Macs ( 1947) utilized the constancy of glomerular filtration rate to calculate relative preglomerular and postglomerular resistance changes over the autoregulatory range of perfusion pressure. They concluded that a reduction in perfusion pressure induces predominant preglomerular dilatation, and that postglomerular resistance is little altered.
Recently, Abe et a]. (1970) reported a phenomenon which they could only account for on the basis of increasing postglomerular resistance during the autoregulatory response. Acetyleholine infused into the renal artery normally induced an increase in renal blood flow without a change in glomerular tiltration rate. When perfusion pressure was reduced to 100 mmHg, both Iiow and glomerular filtration autoregulation were complete.
Acetyleholine infusion at this point increased blood flow, but the increase in blood flow was associated with a large reduction in glomerular filtration rate. Identical effects of vasodilators when superimposed on autoregulatory vasodilatation are evident in the reports of Thurau and Kramer (1959) and Baer et al. (1970). It is difficult to interpret this response in any way other than a progressive, autoregulatory reduction in afferent arteriolar resistance with the reduction in per= fusion pressure, accompanied by a progressive increase in efferent resistance.
At the lower end of the autoregulatory pressure range it seems likely that afferent arteriolar dilatation is maximal, and that administration of the vasodilator at this point therefore induces predominant or sole efferent arteriolar dilatation-and as a consequence reduces filtration rate. This possibility is supported by the recent study of Robertson et al. (1972) who exploited micropuncture of the accessible surface glomeruli in mutant rats to examine single nephron responses to a progressive reduction in perfusion pressure.
They found that autoregulation was complete down to 80 mmHg, and that over this range not only was there a progressive reduction in afferent arteriolar resistance, but that efferent resistance tended to rise. The mechanisms responsible for an efferent vascular resistance increase under these circumstances have not been delineated, but any hypothesis concerning the autoregulatory response has to account not only for a reduction in afferent arteriolar resistance but also for an increase in efferent arteriolar resistance as perfusion pressure falls.
Certainly an increase in efferent resistance is not consistent with the myogenic hypothesis. Possibly two processes occur simultaneously: a dominant myogenic action on the afferent arteriole and a humoral action on the efferent arteriole, which could well be attributable to angiotensin. Evidence of a predominant efferent arteriolar locus for angiotensin action is reviewed below.