| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
REVIEW
1Department of Physiology, Monash University, Melbourne; 2Department of Civil and Environmental Engineering, University of Melbourne, Victoria, Australia; and 3Department of Physiology, Medical College of Wisconsin, Milwaukee, Wisconsin
Submitted 2 April 2008 ; accepted in final form 3 June 2008
ABSTRACT
The kidney is faced with unique challenges for oxygen regulation, both because its function requires that perfusion greatly exceeds that required to meet metabolic demand and because vascular control in the kidney is dominated by mechanisms that regulate glomerular filtration and tubular reabsorption. Because tubular sodium reabsorption accounts for most oxygen consumption (
O2) in the kidney, renal O2 varies with glomerular filtration rate. This provides an intrinsic mechanism to match changes in oxygen delivery due to changes in renal blood flow (RBF) with changes in oxygen demand. Renal O2 is low relative to supply of oxygen, but diffusional arterial-to-venous (AV) oxygen shunting provides a mechanism by which oxygen superfluous to metabolic demand can bypass the renal microcirculation. This mechanism prevents development of tissue hyperoxia and subsequent tissue oxidation that would otherwise result from the mismatch between renal O2 and RBF. Recent evidence suggests that RBF-dependent changes in AV oxygen shunting may also help maintain stable tissue oxygen tension when RBF changes within the physiological range. However, AV oxygen shunting also renders the kidney susceptible to hypoxia. Given that tissue hypoxia is a hallmark of both acute renal injury and chronic renal disease, understanding the causes of tissue hypoxia is of great clinical importance. The simplistic paradigm of oxygenation depending only on the balance between local perfusion and O2 is inadequate to achieve this goal. To fully understand the control of renal oxygenation, we must consider a triad of factors that regulate intrarenal oxygenation: local perfusion, local O2, and AV oxygen shunting.
counter-current exchange; diffusional oxygen shunting; hyperoxia; hypoxia; kidney circulation; kidney disease
5% of hospitalized patients (76). The economic costs of treating patients with kidney disease are staggering. For example, in the United States the 0.7% of Medicare patients requiring dialysis therapy account for 5% of the Medicare budget (102). One of the messages we wish to convey in this review is that to develop better strategies to prevent and treat kidney disease we require a better understanding of the factors that regulate kidney oxygenation. The field of renal oxygenation has a long though intermittent history (58, 61, 117). It has reemerged as a strong focus in recent years (81, 83, 89, 101, 108, 119, 123), and, as we will see, a number of new findings have called into question existing dogma regarding the factors that regulate intrarenal oxygenation and cause dysregulation of intrarenal oxygenation in kidney disease.
In this review, we first detail the unique challenges faced by the kidney in the maintenance of homeostasis of intrarenal oxygenation and the biophysical mechanisms that allow these challenges to be met under physiological conditions at the level of the entire organ and at the scale of the microcirculation. We then consider the contribution of renal tissue hypoxia to the pathogenesis of kidney disease and how, armed with new insights into the mechanisms regulating kidney oxygenation, we might improve diagnostic and therapeutic approaches. We present a case to support the concept that arterial-to-venous (AV) oxygen shunting might make an important contribution to the dynamic physiological regulation of kidney oxygenation and also to the development of renal hypoxia in kidney disease. This is a controversial proposition because much of the evidence supporting a dynamic role of AV O2 shunting is indirect and recent (46, 55) or based largely on theoretical considerations (84). Nevertheless, we believe that its incorporation into conceptual models of kidney oxygenation may foster rapid advances in this field.
The central thesis of this review is that multiple interacting mechanisms operate in concert to allow tight regulation of intrarenal oxygenation and that dysfunction of these mechanisms makes a major contribution to the pathogenesis of kidney disease. Because of the complexities of these multiple interacting systems, we believe much can be learned from an approach that includes both experimental physiology and computational models grounded in established biophysical principles (18, 43, 112).
Unique Challenges for Regulation of Kidney Oxygenation
Matching supply and demand at the whole organ level.
Under normal physiological conditions, 80% of renal oxygen consumption (O2) is used to drive Na-K-ATPase (36, 53). In turn, Na-K-ATPase drives active transport of not only sodium but also dependent transport processes for glucose, amino acids, and other solutes (36). Because tubular transport processes within the kidney are highly load dependent (36), renal O2 changes in proportion with glomerular filtration rate (GFR), providing an intrinsic regulatory system for maintaining homeostasis of intrarenal oxygenation. For example, when arterial pressure is reduced, the GFR falls and the associated decrease in the filtered load of sodium leads to decreased tubular sodium reabsorption and, therefore, decreased renal O2 (113).
The results of early studies (21, 52, 57, 58, 61, 62, 113, 117) in the field of intrarenal oxygenation led to the conventional wisdom that the flow dependence of renal O2 is the predominant, if not only, mechanism mediating homeostasis of intrarenal oxygenation. That is, dynamic physiological regulation of oxygen in kidney tissue was thought to rely entirely on only two factors: 1) delivery of oxygen in renal arterial blood (i.e., renal perfusion) and 2) renal O2 (8, 12; Fig. 1A). For example, Levy (57) found that fractional oxygen extraction by the kidney remained stable across a wide range of renal blood flow (RBF), suggesting that changes in oxygen delivery induced by changes in RBF are directly offset by changes in renal O2. However, in many of these classic studies, RBF was altered by maneuvers that alter O2 independently of changes in RBF. For example, RBF was altered by chronic uninephrectomy (61, 117), by changes in renal perfusion pressure that would greatly alter GFR and tubular load (21, 52, 57, 58, 62, 113), or by cooling the kidney, which would reduce tissue metabolic rate (57). More importantly, we now know that the relationship between renal perfusion and O2 is far more complex than was envisaged at the time of these classic studies and that there is ample opportunity for mismatched changes in renal oxygen supply and demand. For example, RBF and GFR (and so total tubular sodium reabsorption) do not always change in parallel under physiological conditions, as evidenced by changes in filtration fraction in response to both vasoconstrictor and vasodilator factors (78). Moreover, altered efficiency of renal O2 can occur, for example, when the profile of sodium reabsorption along the nephron is altered, since the efficiency of tubular sodium reabsorption differs among nephron segments (8, 36). It can also occur when the bioavailability of nitric oxide changes, since nitric oxide inhibits nephron O2 both by inhibiting sodium reabsorption (86) and by increasing the efficiency of nephron oxygen utilization (119). Vascular wall O2 in the kidney also appears to vary with vascular tone (125), adding an additional layer of complexity to the relationship between RBF and total renal O2. Thus it seems unlikely that oxygen supply and demand in the kidney can be matched under all conditions by a simple relationship between renal O2 and RBF.
|
O2 (Fig. 2). Indeed, these vasoactive agents had remarkably little effect on total renal O2, even though they had profound effects on GFR and sodium reabsorption. Thus maintenance of stable renal O2 under these experimental conditions was likely attributable at least partly to changes in the efficiency of tubular oxygen utilization for sodium reabsorption. However, despite stable renal O2, renal tissue oxygen tension (PO2) remained conspicuously stable as RBF (and so renal oxygen delivery) was varied from 30% below basal to 30% greater than its basal level. Renal venous PO2, however, increased as RBF increased and decreased as RBF fell. Our experiment showed that the flow dependence of renal O2 cannot completely account for dynamic regulation of intrarenal oxygenation under all experimental conditions. Rather, our findings could be explained if AV oxygen shunting in the kidney changes in proportion to RBF (Fig. 3), thus providing an additional mechanism for maintenance of homeostasis of intrarenal oxygenation in the face of physiologically relevant changes in renal perfusion.
|
|
25% of cardiac output (36, 84). In terms of energy mass balance, basal RBF is approximately fivefold greater than basal coronary blood flow, yet at rest the cardiac O2 is approximately double that of the kidneys (83). Therefore, in the absence of specific mechanisms to limit oxygen delivery to renal tissue, PO2 in renal cortical tissue should be high. This would be expected to drive production of reactive oxygen species such as superoxide and lead to oxidative stress, since superoxide production in kidney tissue is highly dependent on oxygen availability (15). However, mammalian kidney tissue is not hyperoxic; cortical PO2 is 15–50 mmHg (6, 11, 55, 56, 120), similar to skeletal muscle PO2 (115), while medullary PO2 can be even lower (PO2 = 5–25 mmHg; Refs. 12, 85). In the mammalian kidney, the phenomenon or AV oxygen shunting appears to act as a structural antioxidant mechanism to blunt delivery of oxygen to renal tissue (84). In the avian kidney, in which there is little opportunity for counter-current exchange of oxygen between arteries and veins, an alternative mechanism appears to operate, with most blood flow to the kidney being of venous origin (34). AV O2 shunting occurs in tissues where arteries and veins are arranged in close proximity in a counter-current fashion (e.g., skeletal muscle and kidney; Refs. 94, 95). Evidence for AV O2 shunting in nonrenal tissues such as skeletal muscle includes the greater PO2 of venous than capillary blood (107) and direct visualization of AV oxygen transfer (49). Mathematical models also support its existence (14, 45, 104, 110). It is driven by the AV PO2 difference and facilitated by the close anatomical association of arteries and veins, as occurs most strikingly in the kidney (31, 32, 80, 84). Nordsletten et al. (80) have recently produced a three-dimensional reconstruction of the renal vasculature using images from high resolution computer tomography. Their data show that an intimate relationship exists between intrarenal arteries and veins along the entire renal circulation up to and including the interlobular vessels (80).
There are three major lines of evidence for the existence of renal AV oxygen shunting. First, nearly 50 yr ago Levy and Sauceda (60), in an ingenious experiment, measured the respective transit times for oxygen and hemoglobin (and thus erythrocytes) across the renal circulation by measuring the optical density of renal venous blood. They labeled hemoglobin by methylation, which increases its optical density, while oxygenation of hemoglobin reduces its optical density. When they administered blood into the renal artery supplemented with oxygen and/or erythrocytes containing methhemoglobin, they found the transit time for oxygen was always less than that for erythrocytes. This can only be explained by the existence of a diffusional shunt for oxygen that bypasses at least some of the circulation (Fig. 4A). Initially, they proposed that AV oxygen shunting was confined to the medullary circulation. However, their experimental findings were reproduced even when the medulla was cooled to reduce medullary perfusion, providing strong evidence that AV shunting of oxygen also occurs in the cortical circulation. Secondly, Schurek et al. (103) found that the PO2 of blood in glomerular capillaries increased relatively little during pure oxygen breathing. Thirdly, and most definitively, Welch and colleagues (120) demonstrated that the PO2 of blood in the renal vein exceeds that of blood in the efferent arteriole (120; Fig. 4B), while Schurek et al. (103) demonstrated that it exceeds that of blood in glomerular capillaries.
|
Determinants of AV oxygen shunting.
Collectively, the new experimental observations we have discussed (46, 55) suggest that regulation of intrarenal oxygenation might depend not only on the matching of changes in RBF with changes in renal O2 but also on the influence of AV oxygen shunting on oxygen delivery to kidney tissue (Fig. 1B). There are likely complex interactions between these three factors.
Renal O2 creates the driving force for AV oxygen shunting, the PO2 gradient between blood in renal arteries and veins, so AV oxygen shunting should be enhanced when renal O2 increases. As discussed earlier, most kidney O2 is attributable to tubular sodium reabsorption, so renal O2 (and thus shunting) should increase when RBF (and so GFR) increases. However, O2 factored for sodium reabsorption can also change under physiological (55) and pathophysiological (89, 119) conditions. This phenomenon likely arises from both changes in the efficiency of sodium reabsorption itself and changes in oxygen utilization for other cellular processes. It has important implications for our understanding of the biophysical basis of AV oxygen shunting, because it allows renal O2 to change in ways both dependent on RBF and independent of RBF.
RBF should also have effects on AV oxygen shunting independent of its effects on renal O2. First, changes in RBF will alter circulatory transit time in the renal circulation and, therefore, the time available for diffusive transport of oxygen from arteries to veins. This will act to reduce AV oxygen shunting when RBF increases, but our findings suggest that it instead increases (Figs. 2 and 3). Currently, we have no adequate explanation for this paradox. One possibility is that diffusion equilibrium is reached at some point in the renal circulation, so that AV oxygen shunting ceases beyond this point. Changes in RBF might then be expected to alter the point at which the diffusion equilibrium is reached; reductions in flow moving the point of diffusion equilibrium to more proximal vascular elements and increases in flow moving the point to more distal vascular elements. This putative mechanism is analogous to that which allows glomerular filtration to be flow dependent under conditions in which glomerular filtration pressure equilibrium is reached at some point along the glomerular capillary network (70). Another possibility is that oxygen consumption by the vascular wall has a profound effect on renal AV O2 shunting. The question of how much oxygen is consumed by the vascular wall is currently a matter of intense controversy (35). Furthermore, in the absence of direct experimental data or quantitative analysis it is unclear what quantitative effects vascular wall O2 might have on AV oxygen shunting. Nevertheless, the vascular wall does consume oxygen in driving vasomotion and endothelial nitric oxide synthesis (105, 114, 125). There is evidence that this produces PO2 gradients across the arterial wall of as much as 25 mmHg in large arterioles (100 µm) and 10 mmHg in small arterioles, at least in skeletal muscle (105). Vasodilatation reduces these gradients, reflecting reduced vascular smooth muscle O2. The resultant increase in oxygen tension at the outside wall of arterial vessels could in turn enhance AV oxygen shunting by increasing the effective PO2 gradient driving shunting. Another way of considering this concept is that vascular wall O2 extracts oxygen from the shunting pathway, so decreasing vascular wall O2 as a fraction of the total kidney O2 should lead to an increase in the contribution AV O2 shunting makes to the various oxygen transport pathways in the kidney. Thus in our experiment where RBF was increased through vasodilatation, shunting may have increased in part because of reduced vascular wall O2. Conversely, when RBF was decreased during vasoconstriction, shunting may have decreased in part because of increased vascular wall O2. This prediction is consistent with our findings (Figs. 2 and 3). However, such predictions must be made with care, as oxygen diffusion between "connected" counter-current flows (Fig. 3) is complex and poorly understood. The PO2 levels in the artery, vein, and wall and, therefore, the driving force for shunting are all linked together with RBF and vascular and tubular O2. These multiple interactions provide great potential for counter-intuitive behavior. Regardless, it is now clear that to understand intrarenal oxygenation we require a deeper knowledge of how the three chief regulatory factors, perfusion, oxygen consumption, and AV oxygen shunting, interact in physiological and pathophysiological settings (Fig. 1B).
Medullary circulation and maintenance of medullary oxygenation. While parts of the renal medulla, especially the papilla, have a high anaerobic capacity, almost all segments of the medullary nephron appear to rely at least in part on oxidative metabolism (16). Of particular note, the renal medullary thick ascending limb, the segment responsible for formation of the medullary interstitial NaCl gradient, has a high respiratory rate and a mitochondrial density similar to cardiac myocytes. However, just as the functions of the renal cortex dictate that this tissue is richly perfused, the functions of the renal medulla dictate that blood flow is limited. The maintenance of a relatively low medullary blood flow (MBF) appears to be critical for maintaining the cortico-medullary solute gradient and, therefore, urinary concentrating mechanisms (88). Furthermore, the level of MBF appears to be a critical determinant of the fine control of tubular sodium reabsorption and, therefore, long-term control of arterial pressure, through regulation of renal interstitial hydrostatic pressure (17, 27, 74). Accordingly, there must be considerable trade-off, in the control of MBF, between the physiological imperatives of the regulation of the cortico-medullary solute gradient and renal interstitial hydrostatic pressure (and so normal tubular function) and the supply of oxygen within the renal medulla.
Blood is supplied to the renal medulla from the vasa recta capillaries that arise from the efferent arterioles of juxtamedullary glomeruli, which comprise 10% of all glomeruli in the kidney (88). Thus while all blood flow to the kidney enters the renal cortex, only 10% of this enters the renal medulla (88). Blood flow in the outer and inner medulla is 40 and 10%, respectively, of that in the cortex (88). The bulk of the direct diffusive supply of oxygen to the medullary nephrons is likely drawn from the plexus of small intermingled capillaries that arise within the renal medulla rather than the vasa recta themselves. The unique anatomy of the renal capillaries has been suggested to underlie much of the susceptibility of the kidney to acute renal ischemic injury, particularly within the renal medulla. Rosen et. al (99) point out that much of the blood supplying the capillaries that feed the medullary tissue with oxygen is derived from the ascending vasa recta, which drain the deeper and less well-oxygenated regions of the renal medulla. In effect, much of the medullary tissue is supplied with a mix of "venous" and arterial blood.
Another factor that limits the delivery of oxygen to the renal medulla is the counter-current arrangement of the major medullary blood vessels, the descending and ascending vasa recta. This allows counter-current diffusion of oxygen from blood flowing into the medulla in the descending vasa recta to blood flowing out of the medulla in the ascending vasa recta, analogous to AV oxygen shunting in the cortex. A mathematical model of this phenomenon indicates it is a major factor limiting oxygen delivery to the renal medulla (126). Medullary hematocrit is also low relative to arterial blood, further limiting medullary oxygen supply (98). As a result of these factors, coupled with the relatively high metabolic demand of the medullary thick ascending limb, small alterations in local blood flow and/or local O2 could potentially result in tissue hypoxia and cellular injury (39).
Regulation of medullary tissue PO2 has been thought to rely heavily on tight control of oxygen delivery through MBF (25, 83). Indeed, there is good evidence that paracrine and autocrine factors act in concert to regulate MBF in the face of altered medullary metabolic activity (25). The relative insensitivity of the medullary circulation to many vasoconstrictor stimuli has also been proposed as an adaptive mechanism that protects the medulla from hypoxic damage (25). However, such a strategy is unlikely to be successful because medullary hypoxia occurs during moderate ischemia of the renal cortex, even when medullary perfusion is maintained at normal levels (85). The mechanisms that link medullary oxygenation with cortical perfusion remain to be determined but could potentially involve AV oxygen shunting (Fig. 5). Cortical hypoxia will lead to reduced oxygen content of blood in interlobular veins draining the renal cortex, which then drain into arcuate and interlobar veins before exiting the kidney via the main renal vein. Thus the driving force for AV oxygen shunting will be increased not only in vascular elements specific to the cortical circulation but also in vascular elements common to both the cortical and medullary circulations (interlobar, arcuate, and proximal interlobular). Thus medullary hypoxia may occur during cortical ischemia, even when MBF is maintained, because oxygen is "stolen" from arterial blood at a point upstream from the divergence of the cortical and medullary circulations by AV oxygen shunting. This hypothesis remains to be formally tested.
|
Oxygen delivery to mitochondria and cellular hypoxia.
Cellular oxygen consumption creates the driving force for cellular oxygen uptake. Therefore, when cellular O2 increases, cellular oxygen uptake increases. However, this also reduces local extracellular PO2 and so limits oxygen diffusion to neighboring cells. A similar phenomenon occurs within cells, with regard to delivery of oxygen to mitochondria. Aw et al. (1) demonstrated that half-maximal oxidation (p50) of cytochrome c + c1 in intact resting renal proximal tubular cells is 3.6 µM oxygen. The p50 in isolated mitochondria however, is much less, occurring between 0.02 and 0.5 µM (i.e., <1 mmHg; Refs. 1, 87). The discrepancy between findings in intact cells and isolated mitochondria is likely accounted for by the presence of oxygen gradients within the cell (1) in large part due to the oxygen sink arising from the mitochondria themselves. That is, as mitochondria utilize oxygen, the oxygen concentration of the surrounding cytosol falls, reducing the diffusion of oxygen further into the cell and to more distant mitochondria. The critical PO2 of renal cells (the extracellular PO2 at which cellular oxygen utilization becomes limited) in vitro has been reported to be between 10 and 20 mmHg (3). Similar values have been reported for renal cortical and medullary tissue in vivo (79). Importantly, the critical PO2 of intact cells is close to the range of tissue PO2 observed in many regions of the kidney, particularly in the medulla, suggesting that even mild bouts of hypoxia are likely to limit oxidative metabolism and alter tubular function in these regions.
Based on the concepts discussed above, it could be expected that increased tubular ion transport activity and enhanced metabolic ATP demand would drive cellular hypoxia by increasing the rate of utilization of oxygen by mitochondria. In agreement with this concept, Aw et al. (1) demonstrated that either the addition of the sodium ionophore nystatin or the mitochondrial uncoupler FCCP significantly increased the p50 for cytochrome oxidation in intact renal proximal tubular cells. Furthermore, the activity of Na-K-ATPase in rat medullay thick ascending limb cells is highly correlated with the susceptibility of these cells to hypoxic injury (24).
Tubular hypertrophy may also limit the diffusion of oxygen in renal epithelial cells. In addition to increasing tubular transport activity, stimuli such as protein loading or reduced renal mass also result in significant cellular hypertrophy, which is in turn accompanied by mitochondrial hypertrophy (44). Thus the increased susceptibility of hypertrophied cells to hypoxic injury may be due not only to an increase in transport related mitochondrial oxygen utilization but also to increased mitochondrial mass. The development of local cellular hypoxia through cellular hypertrophy, in the absence of systemic renal hypoxia, may provide an explanation for the seemingly paradoxical results of some studies investigating the role of tissue hypoxia in the remnant kidney model of chronic kidney disease. Using oxygen mirco-electrodes to measure tissue O2 tension, Priyadarshi et al. (97) demonstrated a 73% increase in both cortical and medullary tissue PO2 in rats subjected to 5/6ths nephrectomy. In contrast, Manotham et al. (72), using the cellular specific hypoxic probe pimonidazole, demonstrated an increase in hypoxic cell staining in the renal cortex of rats in the early phase of the remnant kidney model. As tubular hypertrophy is known to occur in the remaining nephrons in the 5/6ths nephrectomy model, one possibility for the discrepancy between the results of these two studies may be that Manotham et al. observed the development of local cellular hypoxia in response to cellular and mitochondrial hypertrophy, while Priyadarshi et al. observed an increase in systemic renal interstitial PO2 associated with an increase in renal oxygen delivery relative to O2.
To conclude, the kidney is a heavily respiring organ in which oxygen is supplied to the parenchyma via a complex capillary network. PO2 in the cellular microenvironment may depend not only on the relative rate of whole organ oxygen delivery to metabolic demand but may be highly dependent on the barriers to diffusion of oxygen through the parenchyma as well as local rates of cellular oxygen consumption and related micro-oxygen gradients. As regions of local tissue and cellular hypoxia could plausibly occur in the absence of significant whole organ hypoxia, identification and characterization of renal hypoxia may require multiple experimental approaches at multiple scales. Mathematical modeling may assist the planning and interpretation of these experimental approaches.
Reconciling the endocrine (erythropoietin) and exocrine functions of the kidney. In most tissues, blood flow is tightly regulated by tissue oxygenation. For example, hypoxia in tissues such as the brain (7) and skeletal muscle (23) causes vasodilatation and increased perfusion. Similarly, hyperoxia can induce vasoconstriction, particularly in brain tissue, which in turn limits tissue oxygen delivery (Fig. 6). In contrast, RBF is relatively insensitive to the effects of hypoxia and hyperoxia (54, 55; Fig. 6). This makes adaptive sense, since it allows control of renal perfusion to be dominated by the need for regulation of renal excretory function. It also makes adaptive sense for the role of the kidney as the "critmeter" of the body, since it allows changes in arterial blood PO2 to be transmitted to renal tissue and so regulate erythropoietin synthesis and secretion.
|
10% of oxygen delivered in the renal artery, so the oxygen-hemoglobin dissociation curve is relatively steep in the renal cortex. Thus increased fractional oxygen extraction induced by anemia (46) will result in large reductions in renal tissue PO2 and so is a strong signal for increased erythropoietin synthesis. There is also recent evidence that AV oxygen shunting in the kidney is enhanced in anemic states, which would act to augment renal hypoxia and so amplify the signal for erythropoietin synthesis (46).
However, now we are left with a paradox, because changes in perfusion of the cortex and medulla that occur as part of the renal response to challenges to homeostasis of fluid and electrolyte balance should also lead to changes in the delivery of oxygen to renal tissues. Such changes in renal oxygen delivery could potentially confound the maintenance of homeostasis of intrarenal oxygenation and in particular the control of erythropoietin synthesis. Halperin et al. (38) have argued that the "flow-dependence" of renal O2 acts to maintain stable renal tissue PO2 in the face of changes in renal hemodynamics. That is, since most oxygen consumption in the kidney is attributable to tubular sodium reabsorption, kidney O2 often varies with RBF (36), so that kidney tissue oxygenation changes little when RBF changes under physiological conditions. Indeed, this concept is supported by the observations that moderate reductions in RBF have little effect on erythropoietin synthesis (see Ref. 118). However, this analysis relies on the assumptions that the filtration fraction remains relatively constant under physiological conditions and that renal O2 always changes in response to altered RBF in a way that negates the effects of altered oxygen delivery. However, the physiological regulation of the GFR is often associated with changes in filtration fraction (78) and renal O2 does not always vary in direct proportion to RBF (55). Nevertheless, renal tissue PO2 can remain remarkably stable in the face of changes in RBF within the physiological range (i.e., ± 30% of baseline), even if there is a mismatch between changes in oxygen delivery and demand (55, 85). As we have discussed in this review, recent findings (55) suggest that an additional mechanism contributes to dynamic regulation of intrarenal oxygenation in the face of changes in renal perfusion; renal AV oxygen shunting (Figs. 2 and 3).
Intrarenal Oxygenation in Disease States
Over the last three decades medical research has facilitated dramatic advances in the treatment of acute and chronic ischemia and hypoxia in the heart, but the same cannot be said for the treatment of kidney disease (76). This limited success is due in part to the complexity, and our incomplete understanding, of the factors governing intrarenal oxygenation. Nevertheless, some recent advances have been made, in large part because of the recognition that increased oxygen utilization in the kidney, not just reduced oxygen delivery to kidney tissue, contributes to the pathogenesis of kidney disease. Indeed, there is now strong evidence that reduced efficiency of mitochondrial oxygen utilization caused by oxidative stress and reduced nitric oxide bioavalability contributes to renal hypoxia in a diverse range of renal pathologies (89, 119). Further advances should also arise from an increased understanding of the effect of AV oxygen shunting on intrarenal oxygenation in disease states.
Acute renal injury.
A range of factors can lead to acute renal injury, such as renal ischemia/reperfusion (9), allograft rejection (20), endotoxic and hemorrhagic shock (116), severe anemia (116), or hemodilution during medical procedures such as cardiopulmonary bypass (37, 106) and nephropathy induced by chemicals such as radiocontrast agents and other nephrotoxins (41, 101). Often, in a clinical context, a number of these factors are present (116). Renal tissue hypoxia is an important common feature of acute renal injury (101) and is a major driver of the cascade of events leading to cellular injury and vascular and tubular dysfunction (9). The outer medulla is particularly susceptible to hypoxic damage, due both to the relatively large O2 of outer medullary tubular elements (e.g., thick ascending limbs and S3 segments of proximal tubules) and the relatively poor supply of oxygen by the outer medullary vasculature (12, 101).
Until recently, most interest has focused on the role of renal ischemia in the hypoxia associated with acute renal injury, but it is becoming increasingly clear that increased O2 and perhaps also reduced oxygen delivery due to enhanced AV oxygen shunting also contribute. Indeed these three factors make variable contributions, depending on the underlying cause of kidney injury. For example, injection of radiocontrast agents into the kidney circulation causes renal hypoxia, particularly in the medulla (67). This may in part be mediated by reduced local blood flow (65), perhaps secondary to release of endothelin peptides (66). However, there is also evidence that contrast agents can increase blood flow in the outer medulla (40). Radiocontrast-induced medullary hypoxia can be reversed by the loop diuretic furosemide, indicating that it is at least partially dependent on increased oxygen use for tubular transport (40). Similarly, renal tissue hypoxia in anemia (46), after blockade of production of nitric oxide and/or prostanoids (101) and in response to nephrotoxins such as amphotericin (101), appears at least partly due to increased renal O2. Johannes et al. (46) have also provided compelling evidence that enhanced renal AV O2 shunting in anemia contributes to the development of renal hypoxia, presumably because the high affinity of oxygen for hemoglobin normally acts to retard AV oxygen shunting.
Consideration of the importance of the relative roles of the triad of factors (Fig. 1B); renal O2, local perfusion, and AV oxygen shunting in acute renal injury has important implications for its clinical management, which to date has largely been based on the aim to augment GFR to return it to the normal range (101). The problem with this approach is that enhancing GFR will also increase kidney O2 and so potentially worsen intrarenal oxygenation. Rosenberger et al. (101) have championed the potential use of loop diuretics in acute renal failure, since they reduce O2 in medullary tubular elements. This approach is most likely to be successful when at least some level of medullary perfusion and oxygen delivery is maintained and when the majority of tubular ATP production is directed toward powering sodium transport. In pathologies in which medullary hypoxia is prolonged or where active sodium transport is not the chief cause of cellular ATP depletion, such as during mitochondrial dysfunction, loop diuretics may have little efficacy. Indeed, the balance of evidence suggests that these agents have only limited clinical efficacy (2). Consideration has also been given to the need to maintain perfusion of the renal medulla during acute renal injury to maximize delivery of oxygen to medullary tissue. However, as discussed earlier, maintenance of medullary perfusion may not prevent the development of renal hypoxia during moderate to severe cortical ischemia (Fig. 5).
We hypothesize that the development of mismatches in medullary perfusion and tissue oxygen tension may be more deleterious than parallel decreases in perfusion and tissue oxygenation. For example, the maintenance of medullary perfusion in the acutely damaged kidney will likely be associated with the maintenance of glomerular filtration and the production of tubular fluid in the juxtamedullary nephrons that supply the renal medulla. This, in turn, will maintain tubular sodium transport and oxygen consumption in long-looped nephrons. Thus, paradoxically, during conditions of medullary hypoxia cellular ATP depletion may occur more rapidly in medullary nephron segments if perfusion (and so tubular sodium delivery) is maintained than if tubular perfusion and, therefore, ATP usage is reduced.
In summary, hypoxic damage, particularly in the outer medulla, is a hallmark of the pathogenesis of acute renal injury. Medical management of acute renal injury has long focused on the maintenance of renal function (i.e., GFR). We submit that instead it should focus on maintenance of intrarenal oxygenation and minimizing the potential for mismatching of tubular oxygen delivery and energy demand. Thus to successfully prevent or treat acute renal injury we must consider the triad of factors that regulate intrarenal oxygenation (perfusion, O2, and AV oxygen shunting) and in many cases the driving forces of increased cellular O2 (Fig. 1B).
Chronic renal disease.
Compared with acute renal injury, intrarenal oxygenation in chronic renal failure has been relatively little studied. Fine and colleagues (29, 81) have proposed that renal hypoxia is a common pathway in the progression of chronic kidney disease. They have proposed the "two capillary circulations" hypothesis (28), whereby hypoxia results from 1) decreased perfusion in damaged capillaries, and 2) increased O2 associated with hyperfiltration in remaining nephrons. In support of this concept, much evidence now indicates that renal capillary dysfunction is an important factor in the initiation of renal hypoxia and kidney disease (48). Hypoperfusion and rarefaction of renal capillaries have been demonstrated in a number of renal disease models including hypertension (68), aging (111), and reduced renal mass (47, 72, 73). Loss of renal capillary density also occurs after recovery from acute renal injury. For example, renal ischemia/reperfusion is associated with loss of capillary density in both the cortex and medulla (5). Importantly, reduced peritubular capillary perfusion and capillary density after recovery from acute renal injury has been associated with the development of local tissue hypoxia and the progression of renal disease (4, 5, 13). Capillary dysfunction and rarefaction in chronic renal disease, or after acute renal insults, result in regions of local hypoxia and cellular injury in the kidney (109). Local areas of hypoxic injury may stimulate renal inflammation (100), the production of reactive oxygen species (64), and the deposition of extracellular matrix (71), further limiting the supply and diffusion of oxygen within the kidney. Further, progression of renal disease stimulates hypertrophy of remaining functional nephrons increasing the critical nephron PO2 and promoting progression of cellular hypoxia within the kidney and renal disease (77). Given the accumulation of evidence indicating that injury to the renal microvasculature contributes to the initiation and progression of renal disease, future studies aimed at determining the mechanisms underlying microvascular dysfunction and hypoxia in the diseased kidney will undoubtedly be of major clinical importance in the prevention of chronic renal disease.
Renal dysfunction in diabetes and hypertension: the importance of oxidative stress and reduced nitric oxide bioavailability.
The last decade has witnessed a huge growth in our understanding of the roles of nitric oxide and superoxide in the pathogenesis of hypertension- (124) and diabetes-induced end organ damage (26, 89, 96). One of the consequences of renal oxidative stress in these conditions is renal tissue hypoxia (89, 119, 120). There is now strong evidence that this tissue hypoxia arises chiefly because of reduced nitric oxide bioavailability in large part due to nitric oxide quenching by superoxide (89, 119). Nitric oxide can increase renal oxygenation by increasing oxygen delivery through vasodilatation (10) and reducing O2 by inhibiting tubular sodium reabsorption (86) and also by competing with oxygen at the level of cytochrome oxidase within mitochondria (50). However, it is becoming increasingly clear that it is the relative absence of this latter effect of nitric oxide that is most important in the development of renal hypoxia in diabetes and hypertension, because it results in reduced efficiency of oxygen utilization for sodium reabsorption (89, 120).
One issue that has not been considered previously, in relation to chronic renal pathologies associated with oxidative stress and reduced nitric oxide bioavailability, is the effect of the consequent increase in renal tubular O2 on AV oxygen shunting. As we have described earlier in this review, renal O2 provides the driving force for AV oxygen shunting; the difference in oxygen tension between arterial and venous blood. Thus while reduced perfusion and increased O2 can themselves contribute to tissue hypoxia, oxygen delivery is partially dependent on O2 because changes in O2will alter AV oxygen shunting (Fig. 1B). The degree to which this phenomenon contributes to regulation of intrarenal oxygenation in health and disease remains to be determined. Clearly mathematical modeling has an important role to play in uncovering the complex interplay between these processes.
Important recent findings with implications for treatment of kidney disease. A number of recent findings in the field of intrarenal oxygenation have potential implications for the clinical management of chronic kidney disease. First, renal tissue hypoxia in diabetes and multiple forms of hypertension can be ameliorated by scavengers of reactive oxygen species (63, 90, 121, 122), suggesting that such treatments may have a place in preventing renal injury in chronic hypertension and diabetes. Second, increased hepatic arginine metabolism also appears to make a major contribution to reduced renal nitric oxide bioavailability in diabetes and, in turn, associated tissue hypoxia in the renal medulla (92). This observation provides a mechanistic basis for the use of arginine supplementation in diabetic nephropathy. Third, AT1-receptor blockade acutely increases renal cortical PO2 (82) and, if given chronically, upregulates mitochondrial nitric oxide synthase activity in the kidney and reduces respiratory chain activity (93). It also protects against the development of mitochondrial dysfunction in Type 1 diabetes (19). Collectively, these data provide a mechanistic basis for the use of AT1-receptor antagonists to prevent dysregulation of intrarenal oxygenation in chronic renal disease. The potential roles of AT2-receptors in control of intrarenal oxygenation also need to be considered, since Palm et al. (91) recently provided strong evidence that AT2-receptor mediated nitric oxide release sustains renal perfusion and oxygenation in experimental renovascular hypertension.
Conclusion
The kidney faces unique challenges in the maintenance of homeostasis of intrarenal oxygenation. Intrarenal oxygenation is regulated by the maintenance of a fine balance between demand and delivery. At least in the mammalian kidney, this balance could not be achieved without AV oxygen shunting, which allows the functional requirement of the kidney for rich perfusion to be met without the development of tissue hyperoxia (84). The price paid for this protection from tissue hyperoxia is that the kidney is susceptible to tissue hypoxia. More recent observations indicate that renal AV oxygen shunting may not merely be a "structural antioxidant defense mechanism" but may also contribute to the dynamic regulation of intrarenal oxygenation in the face of changes in RBF (55) and also to the development of renal hypoxia in anemic states (46). Renal hypoxia is a salient feature of diverse renal pathologies, including various forms of acute renal injury, chronic renal disease, hypertension and diabetes. Reduced renal perfusion and increased O2 are major contributors to renal hypoxia in these conditions. Theoretical considerations suggest that increased AV oxygen shunting may exacerbate tissue hypoxia in these conditions, but as yet this has not been investigated experimentally or by mathematical modeling. We hope this review has made a case for such studies.
GRANT
This study was supported by grants from the National Health and Medical Research Council of Australia (143603, 143785, and 384101). P. O'Connor is a recipient of an American Heart Association Post-Doctoral fellowship (0625793Z).
FOOTNOTES
Address for reprint requests and other correspondence: R. Evans, Dept. of Physiology, PO Box 13F, Monash Univ., Victoria 3800, Australia (e-mail:Roger.Evans{at}med.monash.edu.au)
* R. G. Evans and P. M. O'Connor contributed equally to this work. 
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  M. H. Little and J. F. Bertram Is There Such a Thing as a Renal Stem Cell? J. Am. Soc. Nephrol., October 1, 2009; 20(10): 2112 - 2117. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
