 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
Nephrology beyond JASN |
In the past it had been thought that the role of the kidney in the control of BP was adequately described by the paradigm of the volume–renin relationship (1). The concept implies that hypertension in general and hypertension specifically in patients with renal disease is the result of sodium retention and the ensuing hypervolemia in the presence of an inappropriately high activity of the renin-angiotensin system. While the concept is certainly valid and in agreement with the data (2), it doesn’t fully describe the pathogenesis of hypertension in kidney disease (3). This issue has become much more complex in recent years with the recognition of the important role of sympathetic overactivity (4–6), which is reversible with nephrectomy (5,7), of impaired nitric oxide (NO) synthesis (8,9) and endothelial cell–dependent vasodilatation (10), as well as of structural remodeling of the vasculature with anatomically fixed elevation of peripheral vascular resistance (11,12).
The issue of sympathetic dysfunction in renal disease is particularly complex. It has been known for a long time that plasma catecholamine concentrations are elevated in renal failure (13,14), but the data were difficult to interpret because in uremia presynaptic reuptake of catecholamines is impaired (15), as is the 
- (16,17) and 
-adrenergic (15,18) response to catecholamines. Nevertheless the dramatic BP response of renal patients to sympathicoplegic medication (19) suggests an important role on BP for sympathetic overactivity, which has now been proven beyond doubt using the methodological gold standard of microneurography (5,20,21).
All of a sudden there is a new player in the field of circulating catecholamines. This issue has been given an entirely new twist in the paper by Xu et al. (22), an impressive example of the power of modern molecular techniques and rational investigative strategies. The authors started with the hypothesis that the kidney may have endocrine functions other than the known secretion of renin, erythropoietin, 1,25(OH)2D3, and others. To prove this hypothesis they screened libraries of the Mammalian Gene Collection Project to identify potential candidates for a novel renal secretory product that met the following a priori criteria: (1) protein with <20% similarity to known proteins, (2) presence of a signal peptide, but (3) no transmembrane domain (which would fix the product to the cell rather than permit secretion). To cut a long story short, they identified one 37.8 kilodalton protein (with several ancillary bands) that was mainly, although not exclusively, expressed in the kidney and which they called "renalase." In situ hybridization revealed expression in glomeruli and proximal tubules (but also in cardiomyocytes and other tissues). That this product can indeed be secreted, at least by renal cells, was proven by detecting the product in plasma and urine of healthy individuals. To permit such measurements, the authors first generated a transcriptionally active fragment by PCR and transfected a cell line to obtain the product in the culture medium as proof for secretion. They then detected the product by Western blot in the plasma of healthy individuals, but not in the plasma of uremic patients—an argument for, although not definite proof of, secretion by the kidney. This fact was surprising, because the protein is expressed not only in the kidney, but also, although less intensely, in the heart, skeletal muscle, and small intestine.
The next question was to identify the function of this novel protein. The authors identified an amine oxidase domain in the renalase molecule. This led to experiments documenting that renalase metabolized catecholamines: dopamine > epinephrine > norepinephrine. The enzymatic activity was specifically inhibited by a renalase antibody. This is indeed a remarkable finding, because all flavin-adenin-dinucleotide (FAD)–containing oxidases known so far are intracellularly anchored and, in contrast to renalase, not secreted, circulating enzymes.
Finally, as proof of the principle that this novel FAD-containing amine oxidase, which was capable of degrading catecholamines in vitro, did affect hemodynamic parameters in vivo, recombinant renalase was injected into Sprague-Dawley rats: Within <30 seconds systolic, much less diastolic, BP decreased transiently (4 min), but dramatically (–23.5% systolic BP). This was accompanied by a decrease in the rate of change of left ventricular pressure (dP/dt) and of maximal left ventricular pressure as evidence of impaired cardiac pumping function.
Obviously the kidney is always good for a surprise! These experiments define a novel secretory product of the kidney with dramatic acute hemodynamic effects. The future will show what the long-term effects of renalase are, whether its secretion is regulated and controlled, whether the (presumably not secreted) renalase in other tissues has a physiologic role, whether in the long run counterregulation modifies the effect of circulating renalase, whether it is part of a regulatory circuit, and whether renalase has a role in renal disease. It would be premature to make strong statements on its role at this early stage, but it is easy to predict that in the near future this novel endocrine product of the kidney will be intensely investigated experimentally and in renal patients.
Footnotes
Address correspondence to: Prof. Eberhard Ritz, Department Internal Medicine, Division of Nephrology, Bergheimer Strasse 56a, D-69115 Heidelberg, Germany. Phone: 49-0-6221-601705 or 49-0-6221-189976; Fax: 49-0-6221-603302; E-mail: Prof.E.Ritz{at}t-online.de
References
Microalbuminuria was originally described in diabetics and was recognized to be a predictor of renal and cardiovascular (CV) risk (1–3). Damsgaard (4) analyzed albumin excretion in type 2 diabetics and noted—as an unexpected chance finding—that a significant relationship between albuminuria and mortality existed in her nondiabetic control group as well. More recently, many studies confirmed this finding and provided solid support for the concept that microalbuminuria is a strong predictor of CV (5–8) as well as renal (9,10) risk in nondiabetic patients. In the Copenhagen study, albuminuria was not only an independent predictor of CV events (11) but was also strongly correlated to CV and overall mortality (12). The PREVEND study showed that in the general population potential causal factors that might have explained microalbuminuria such as hypertension or diabetes mellitus were absent in the great majority of microalbuminuric individuals (7). Obviously, measurement of urinary albumin as an independent risk predictor and screening for this parameter at least in high-risk populations has considerable public health importance (13)
It is quite doubtful that the current definition of microalbuminuria is optimal for risk prediction (14). The definition had originally been established in diabetics and is currently accepted in nondiabetics as well. Recently it has been shown, however, that even urine albumin concentrations in the high normal range are predictive of CV risk both in diabetic (15) and in nondiabetic patients (5,6).
According to the hypothesis proposed by Remuzzi and Bertani (16), it is now commonly accepted and supported by solid evidence (17–19) that proteinuria is a valid target for treatment in patients with manifest diabetic (18–20) and nondiabetic nephropathy (17). Reduction of proteinuria is associated with less progression of renal disease and also associated with fewer cardiovascular events (21).
Until recently, however, no evidence had been available to indicate whether a similar relationship extends into the low range of urinary albumin excretion. This missing link has now been provided by two independent studies: a small, prospective study (22) and a retrospective analysis (23) of the huge Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) study (24). Both studies provide clear evidence that intervention with an angiotensin-converting enzyme inhibitor (22) or an angiotensin receptor blocker (25) reduces the risk of a CV event.
In the LIFE study, 8206 subjects with high CV risk and left ventricular hypertrophy were randomized to receive either the -blocker atenolol or the angiotensin receptor blocker losartan and were then followed for 4.8 yr. The urinary albumin/creatinine ratio was measured at baseline and annually thereafter. What is remarkable about the post hoc analysis of Ibels (23) is the fact that the response of the albumin excretion rate to the intervention predicted the frequency of future CV events. For this analysis the patient population was divided according to median baseline albuminuria and median 1-yr albuminuria. Of note, the median baseline value of 1.21 mg albumin/mmol creatinine is far below the threshold of microalbuminuria (>3.5 mg/mmol). The change in albuminuria was significantly correlated to the composite primary endpoint (CV death, nonfatal stroke, and nonfatal myocardial infarction), as well as to its individual components. The composite primary endpoint was lowest (5.5%) in individuals with low baseline/low 1-yr albuminuria, intermediate in individuals with low baseline/high 1-yr (8.6%) as well as with high baseline/low 1-yr albuminuria (9.4%), and highest in individuals with high baseline/high 1-yr albuminuria (13.5%). It is important that this was not explained by a confounding effect of in-treatment BP as assessed in a complex Cox proportional hazard model with time-varied albuminuria.
This finding of Ibels (23) is reminiscent of the relationship between the reduction of proteinuria and the regression of renal risk and mortality which had been observed in diabetic patients with heavy proteinuria and advanced nephropathy (18,19,21).
What Are the Clinical Implications?
The observation provides a strong argument for monitoring albuminuria. If albuminuria is not lowered satisfactorily, it would make sense to uptitrate antihypertensive treatment, specifically blockade of the renin-angiotensin system, which in the LIFE study reduced albuminuria more than the -receptor blockade (23). Furthermore, one should then also consider treatment of modifiable risk factors, e.g., administration of statins (26), cessation of smoking, and in the future possibly insulin sensitizers as well.
This type of observation does not prove causality. It proves only that there is a very tight correlation between albuminuria and cardiovascular risk. It is nevertheless plausible to assume that albuminuria, i.e., disturbed permselectivity of the glomerulus, is somehow linked to vascular dysfunction (Steno hypothesis). Very suggestive evidence for this idea is provided by the observation that albuminuria is associated with vascular leakiness (27), i.e., albumin escape from the plasma space into the interstitial space, and by evidence of endothelial cell dysfunction and microinflammation (28) associated with, and even preceding the onset of, microalbuminuria. Elucidation of the molecular details of the pathogenetic link between podocyte and endothelial cell is a fascinating challenge to future research.
References
Since the seminal observations of Davies (1) and later of van der Woude (2), it is well known that antineutrophil cytoplasmic antibodies (ANCA) are found in approximately 80% of patients with pauci-immune necrotizing/crescentic glomerulonephritis and with systemic small vessel vasculitis. The most common epitopes recognized by ANCA are myeloperoxidase (MPO) and proteinase 3 (PR3) (3), the former more frequently, but not uniquely, in Wegener’s granulomatosis, the latter mostly in microscopic polyangiitis (4). ANCA stimulate cytokine-primed neutrophils and monocytes, causing for instance respiratory burst, release of toxic granule constituents such as oxygen metabolites and proteinases, as well as endothelial cell injury (5). However, it had long remained controversial whether ANCA are just a marker of disease or have a causal role in the genesis of necrotizing and crescentic glomerulonephritis.
The elegant animal model of Xiao et al. (6) had provided definite proof that ANCA were both necessary and sufficient to cause necrotizing and crescentic glomerulonephritis, as well as systemic necrotizing arteritis and hemorrhagic pulmonary capillaritis. For this purpose the authors used MPO knockout mice for which mouse MPO is a "foreign" substance. They immunized these mice with mouse MPO to obtain mouse MPO antibodies. In a second step, mice which were unable to launch an immune response against such mouse MPO antibodies, i.e., knockout mice with a deletion of the recombinase activating gene 2 and therefore devoid of functional B and T cells, received purified mouse anti-MPO IgG or control IgG by intravenous injection. This maneuver reproduced the features of microscopic polyangiitis, i.e., focal necrotizing and crescentic glomerulonephritis with fibrinoid necrosis, crescent formation, and absence or paucity of glomerular IgG deposits. The same outcome was seen when mouse anti-MPO splenocytes were injected. These experiments with injection of mouse anti-MPO IgG or adoptive transfer of mouse anti-MPO splenocytes left no doubt that MPO antibodies alone were able to cause the disease in the absence of B or T cells.
What had remained unclear, however, was the cellular target through which MPO antibodies mediated glomerular and vascular injury. To chase down the culprit and to provide evidence that indeed neutrophils—and not for instance endothelial cells, as postulated by others—were the target of the MPO antibodies and the key effector cells, the authors now recently carried out an additional experiment. They used the model of the neutropenic mouse. After injection of NIMP-R14, a rat monoclonal antibody, the mice were selectively depleted of circulating neutrophils, the number decreasing from 14% to 1%. Neutropenic mice or control mice received a low (50 µg/g body weight) or high (50 µg/g body weight on day 0 and day 3) dose of anti-MPO IgG by intravenous injection. Control mice received an injection of bovine serum albumin. Five days after injection of anti-MPO IgG, mice without neutrophil depletion developed hematuria, proteinuria, and leucocyturia. Renal histology documented focal glomerular necrosis and glomerular crescents. Glomerular neutrophil phenotyping revealed glomerular infiltration by neutrophils in foci of inflammation and necrosis as well as in a few afferent arterioles, while monocytes/macrophages were mainly observed in glomerular crescents. With the higher dose, blood urea nitrogen was increased as well, but interestingly necroses were seen on average only in 17.8 ± 7.8% of glomeruli.
All these lesions were attenuated or even absent in neutropenic mice receiving the anti-MPO IgG. The observation of no glomerular injury in the absence of neutrophils points to the neutrophil as the key effector cell in the induction of the acute glomerular injury of MPO-induced glomerulonephritis, although an ancillary role of monocytes is not completely excluded. The smoking gun points to the neutrophil, and this fits in nicely with past observations documenting the ability of human MPO ANCA and PR3 ANCA to activate and degranulate neutrophils causing the release of toxic oxygen metabolites (7), proteases, nitric oxide (8) and inflammatory cytokines (9), facilitating attachment and killing of endothelial cells (10,11).
Why is the observation of Xiao et al. important? It points to maneuvers interfering with ANCA-induced neutrophil recruitment and activation as potential therapeutic targets, if this can be achieved safely and without inducing harm.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
HOME
CURRENT ISSUE
ARCHIVES
JASN Express
ONLINE SUBMISSION
AUTHOR INFO
EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP |
