Phenotypic characteristics of diabetic kidney involvement.

Understanding phenotypic characteristics of the diabetic kidney is important for the development of therapies to prevent progression of diabetic nephropathy. In addition to glomerular hyperfiltration and kidney growth, major metabolic abnormalities characterize the diabetic kidney. Increased kidney oxygen consumption leads to cortical and medullary hypoxia in diabetes. Decreasing inspired oxygen to 10% reduces pO2, while oxygen consumption remains elevated, lactate increases, and redox potential decreases, but only in the diabetic kidney--a shift to Warburg metabolism.

During the past decade and a half, we have become aware of the large number of persons afflicted with chronic kidney disease.[1] Diabetes mellitus is the dominant cause of ESRD in the USA and therefore contributes greatly to the CKD population of over 20 million individuals in the USA.[1, 2] Slowing the rate of progression of CKD and diabetic nephropathy is of obvious greater importance. In order to determine effective therapies, we need to define better the phenotypic characteristics of CKD and diabetic kidney involvement. This information has been rather slow in coming, with developments primarily deriving from experimental animal models of CKD and diabetes. In 2000, Leon Fine and colleagues suggested that chronic kidney hypoxia might be a phenotypic characteristic of models of kidneys with decreased surviving nephrons.[3] During the past decade, several studies of experimental animal models of CKD and the diabetic kidney have verified the earlier prediction that kidney hypoxia does persist in both CKD[4-6] and early diabetic kidney.[7]

Why should hypoxia develop as a primary characteristic or adaptation to chronic kidney disease and the diabetic kidney? Early studies in the subtotal nephrectomy model demonstrated that peritubular capillary rarefaction does occur [5] raising the question as to whether the hypoxia was primarily ischemic in origin However, this seems less likely given the anatomic arrangement of arterial and venous vessels in the kidney. Studies earlier by Schurek provided evidence for a countercurrent oxygen exchange system within the kidney whereby the cortical pO2 was maintained at relatively low levels of 40–45 mmHg in spite of wide variations in inspired oxygen and arterial oxygen tension from below those values to as high as 500 mmHg.[8] Therefore, the kidney, of all organs, should be able to defend itself from wide variations in inspired oxygen and arterial oxygen tension. Later studies in CKD models and in the early diabetic kidney observed that the lower cortical and medullary oxygen tensions observed were associated with increased oxygen utilization rather than a consequence of decreased oxygen delivery and blood flow. [4,6,7] In general, models of CKD and diabetic kidney involvement have demonstrated elevated levels of nephron filtration rate and reabsorption, but also increased levels of nephron blood flow, making classical “ischemia” unlikely and not a cause for cortical and medullary hypoxia. Greater kidney oxygen consumption could have been attributed to increased tubular reabsorption of NaCl per nephron since both CKD models and early diabetes are associated with glomerular hyperfiltration per nephron unit. However, when kidney oxygen consumption is factored by the quantity of NaCl reabsorbed, oxygen consumption remained increased and progressively over time in the case of the subtotal nephrectomy model of CKD. [4,6,7,9] In the diabetic kidney, primary hyperabsorption by the proximal tubule does not fully explain the increase in kidney oxygen consumption since oxygen consumption factored by NaCl reabsorption is still increased.[7]

Several laboratories have largely agreed that metabolic efficiency in CKD and the diabetic kidney are markedly reduced and overall oxygen consumption increased.[4,6,7,9] However, there is no universal agreement as to the causes of the reduction in metabolic efficiency and increased oxygen costs. This reduction in metabolic efficiency correlates well with the development of kidney fibrosis and declines in kidney hemodynamics.[3,6,9] In CKD and diabetes, gluconeogenesis may be increased and contribute to oxygen requirements because of increased ATP needs to synthesize glucose. However, other etiologies have also been proposed and tested. Inhibition of angiotensin II activity and the restoration of NOS activity normalize oxygen consumption and the prevention of reactive oxygen species also plays a corrective role in both CKD and diabetic kidneys.[4,6] Several other treatments including amplification of hypoxia inducible factor and restoration of the observed reduction in AMPK activity in CKD (using metformin and AICAR) and diabetes have also normalized oxygen consumption and renal hemodynamics.[6,9] It is of major interest that these divergent treatments share not only normalization of oxygen efficiency and renal hemodynamics but also prevent the later development of kidney fibrosis and deterioration of kidney function. [6,9] The mechanism linking increased oxygen consumption, the consequent hypoxia, and the development of kidney fibrosis and declines in kidney function remains largely undetermined. Clearly, the finding of kidney hypoxia is related to increased oxygen utilization and is not due to deficiency in oxygen delivery.

The current study has utilized a combination of microelectrode measurements, BOLD MRI and metabolic assessments to evaluate the role of oxygen sensitivity early in the diabetic kidney.[10] Microelectrode and MRI assessments verify previous observations of significant cortical and medullary hypoxia in early diabetic kidney, presumably related to increased oxygen consumption.[7] When exposed to 10% oxygen, cortical and medullary pO2 declined in both control and diabetic kidneys but oxygen tensions were approximately 10 mmHg lower in the diabetic kidney cortex than in control kidneys at normal and reduced inspired pO2 values. Lactate and pyruvate levels and oxidative metabolism overall were not changed in the control kidneys exposed to 10% oxygen tensions. The major site of oxygen consumption in normal kidneys is the proximal tubule, and the metabolic activity is solely oxidative since glycolytic activity is essentially absent in normal proximal tubules. However, when the diabetic kidneys were exposed to 10% inspired oxygen, the cortical pO2 was reduced to approximately 25 mmHg relative to 35 mmHg in control kidneys. This produced greater changes in lactate and pyruvate and the lactate/pyruvate ratio increased markedly in the diabetic kidney. Total oxidative metabolism was not changed however, suggesting that this finding was not simply the result of increased glycolysis since the other indicator of redox potential, NAD+/NADH, was also reduced in parallel with the increase in lactate/pyruvate ratio. Implicit to the results of this study is that the greater reduction in kidney pO2 as a consequence of reductions in inspired oxygen in the diabetic kidney forms the basis of the major differences in lactate and pyruvate generation and increased LDHA activity, while oxidative activity was largely unchanged. The greater redox potential and lactate response in the diabetic kidney implies a state of greater mitochondrial stress.

Further studies must now differentiate whether these effects are due entirely to the lower cortical and medullary pO2 values or processes specific to the adapted diabetic mitochondria. Would further reductions in inspired pO2 in control kidneys sufficient to reduce oxygen levels to approximately 25 mmHg comparable to diabetic kidneys also lead to major increases in lactate levels, LDHA activity and major reductions in NAD+/NADH ratios? This issue is critical to the conclusion that progression of diabetic nephropathy could be more severe at higher altitude exposure as a consequence of further reductions in tissue pO2. Since oxidative metabolism was not reduced in the diabetic kidney, this may represent Warburg metabolism or other mitochondrial stresses reducing the redox potential of the diabetic kidney with lower inspired oxygen breathing. How much hypoxia and what altitude and inspired oxygen levels are required to trigger this Warburg process? After such determinations, then we may be able to determine why these processes contribute to greater kidney fibrosis. These observations remain correlative and the signal mediating the marked alterations in NAD+/NADH ratio and lactate generation with hypoxia remains to be determined.