 |
||||
| 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 |
Rapidly progressive glomerulonephritis (RPGN) is an important nephrologic emergency. Based on underlying immunohistology, RPGN is usually categorized into 3 groups: immune complex–associated RPGN, antibasal membrane antibody–associated RPGN (Goodpasture’s disease), and so-called pauciimmune RPGN with the two partially overlapping entities of Wegener’s granulomatosis and microscopic polyangiitis. In the latter variant, antineutrophil cytoplasmic antibodies (ANCA) (1,2) directed against either proteinase 3 (PR3) or myeloperoxidase (MPO) are found in the great majority of cases. At least in the case of MPO, the elegant experiments of Xiao et al. (3) documented that these antibodies are pathogenetic. A certain proportion of patients with RPGN, 20% according to a study in Wessex (4), markedly less in our experience, fail to present with ANCA. It is for the latter type of ANCA-negative RPGN that the experiments of Ding et al. (5) are of particular clinical relevance, but the implications of the identified mechanisms go considerably beyond this narrow issue and provide more general insights into a novel mechanism of acutely progressive glomerular injury.
The conventional concept postulates that in RPGN antineutrophil antibodies interact with polymorphonuclear cells in the intracapillary space and secondarily induce crescent formation (inside
out). In contrast, in the model established by Ding et al. (5) the podocyte has been identified as the primary culprit, suggesting in this model the alternative sequence outsidein.
The authors used CRE-lox technology to selectively deactivate the gene coding for the von Hippel-Lindau (vHL) gene product in mouse podocytes. The vHL protein plays a crucial role in oxygen sensing (6): under normoxic conditions proline molecules of the oxygen sensing protein HIF1
are oxydised so that hypoxia-inducible factor 1 (HIF1) is recognized, captured by vHL, and delivered to the proteasome for proteolytic breakdown (7). In brief, vHL acts as a negative regulator. Under hypoxic conditions the prolines of HIF1 are no longer oxydised, HIF1 accumulates, dimerizes with HIF1
, and increases transcription of downstream genes, thus triggering responses such as increased glucose uptake, glycolysis, angiogenesis, vascular remodeling, and many other processes (8).
Loss of vHL function, as produced in the experiment of Ding et al. (5), will imitate the effects of hypoxia and result in the constitutive expression of known HiF target genes.
Despite the loss of the vHL gene in the podocytes, the mice were healthy up to the age of 4 wk. At the age of 3 wk, the mice developed proteinuria and dilated capillary loops were seen by histologic analysis. At 4 wk, however, the explosive onset of hematuria, proteinuria, and renal insufficiency was noted. Histologic examination now showed crescentic glomerulonephritis with prominent fibrin deposition and fibrinoid necrosis. ANCA-specific antibodies were not seen.
It is currently widely assumed that crescents are derived from proliferation of parietal epithelium with an admixture of infiltrating macrophages, but in a series of studies Ding et al. identified podocytes as the main component of the crescents (5).
In this experiment it appears that podocytes lost their terminally differentiated postmitotic phenotype and reentered the cell cycle as shown by pulsing with bromodeoxyuridine. The most important question then was to identify the downstream genes activated in this model of a hypoxia analog in the genetically manipulated podocytes. As expected, examination of isolated glomeruli identified increased HiF1 and HiF2 protein levels. Gene expression profiling confirmed that downstream target genes of the vHL-HiF pathway were upregulated, among others the chemokine receptor CXCR4 and its only ligand, stromal-derived factor (SDF1/CXCligand12).
At this point some background information is helpful. The chemokine receptor CXCR4 entered the limelight because of its relation to HIV. HIV enters target cells via the viral envelope glycoprotein (Env) in a coordinated interaction with the CD4 receptor and a chemokine coreceptor (CCR5 or CXCR4) (9,10). This knowledge has led to promising efforts to develop CXCR4 inhibitors (11). To understand its role in renal disease, however, it is relevant to draw attention to its role in the crosstalk between tumor cells and their microenvironment (12). Mesenchymal or marrow-derived stromal cells, which constitute a large proportion of nonneoplastic cells in the tumor microenvironment, secrete the respective ligand SDF-1, which interacts with CXCR4 and attracts tumor cells. The importance of this system is illustrated by the fact that both SDF-1 and CXCR4 knockouts are embryonic lethal (13). SDF-1 is thought to play an important role in aspects such as metastatic tumor spread, tumor cell survival, tumor angiogenesis, and others. In analogy to the above connection to HIV, great hope is currently placed into the potential benefit from CXCR4 antagonists for the treatment of malignant tumors. SDF-1 expression in tumors is triggered by hypoxia (14) and promotes tumor cell proliferation (15) and stem cell trafficking (16), of interest with respect to the findings in the kidney as described below. The role of the SDF-1/CXCR4 system goes far beyond oncology: SDF-1 has been identified, for instance, as an agent promoting negative inotropy in the heart (17), and the role of this system in RPGN may not come as a complete surprise after it was documented in kidney organogenesis (18).
By immunostaining and real-time PCR it was shown that SDF-1 was upregulated in the podocytes of mice with conditional vHL knockout. To prove the functional importance of this system, neutralizing anibodies to the receptor of SDF-1 (i.e., CXCR4) were administered before the onset of overt renal disease. This delayed, but failed to prevent, the onset of RPGN.
Ding et al. (5) went one step further. To assess whether the SDF-1/CXCR4 system fully accounts for the phenotype, or is only one aggravating factor, they generated transgenic mice expressing CXCR4 selectively in their podocytes. This maneuver caused proliferation of podocytes as well as some glomerular disease, and at 6 mo of age caused hematuria and proteinuria, but failed to reproduce the full-blown scenario of RPGN, suggesting that other systems are also involved in the pathogenesis of vHL knockout–associated RPGN.
Finally, to provide evidence for the functional relevance in human disease, the authors examined glomeruli of patients with RPGN using real-time PCR and immunostaining. They found upregulation of HiF target genes including CXCR4, increased immunostaining for CXCR4, and, as sign of podocyte dedifferentiation, reduced staining for synaptopodin.
Presumably because the well-perfused glomerulus appeared to be an unlikely target for hypoxic damage, the HiF/vHL system had not been on the agenda of renal investigators, in contrast to the tubulointerstitium, where the great importance of hypoxia had long been postulated (19) and recently confirmed by elegant experiments (20). Of course, the conditional knockout of vHL is not hypoxia per se. It is of interest that RPGN does not appear to be more frequent in vHL disease and the coincidence of renal cell carcinoma, a state of known (albeit local intratumoral) mutation of the vHL gene, may be no more than coincidence of frequent conditions (21,22).
The merit of this paper by Ding et al. is to have drawn attention to a potentially very important system in the pathogenesis of glomerular disease, which so far had not been very high on the agenda of nephrologists.
|
 |
Footnotes |
|---|
|
References |
|---|
 Top References References |
|---|
Despite the colonization of the external part of the urethra by bacteria, the urinary tract is usually free from microbial colonization for reasons that today become increasingly clear.
In the past it had been assumed that the absence of bacterial colonization resulted from mechanical clearance of organisms by voiding, by shedding of colonized epithelial cells, by microbe-binding proteins such as Tamm-Horsfall protein, by influx of and phagocytosis by neutrophils, as well as by adaptive immunity, i.e., bactericidal antibodies and effector immune cells (1). The classic experiments of Norden et al. (2), however, showed that when phosphorous-labeled Escherichia coli bacteria were introduced into the urinary bladders of female guinea pigs, >99.9% of the bladder inoculum was rapidly excreted but 0.1% of E. coli remained attached to the bladder wall, a quantity sufficient to perpetuate infection if bacterial growth were uninhibited (3). In a series of brilliant experiments, Norden et al. documented that bacteria attached to the bladder mucosa were rapidly killed as a result of intrinsic antibacterial activity of the epithelial cells within minutes, independent of polymorphonuclear leukocytes. With clairvoyance the authors postulated "that a substance (yet unidentified) exists in the bladder wall which is used... as the intact bladder exerts its antibacterial activity."
Recently it has become clear that, apart from adaptive immunity, i.e. bactericidal antibodies and effector immune cells, the antibacterial defense by epithelial cell layers (skin, gastrointestinal, respiratory, and urogenital tracts) is the task of a phylogenetically ancient system that is found in plants, insects, and vertebrates (but obviously not in bacteria), namely inducible antimicrobial peptides, which were first detected in silk moths (3) and Drosophila melanogaster (4), and later in mammalian species (5). These molecules are characterized by clusters of hydrophobic and cationic amino acids organized in what has been called an "amphipathic design" (6), which disrupts the bacterial membranes, the outermost leaflet of which differs strikingly from that of plants and animals. It is disrupted by insertion of inducible antibacterial peptides at micromolar concentrations. Amazingly, bacteria have not developed resistance to these peptides. This observation gave rise to great hopes that the golden bullet against multiresistant bacteria was around the corner, hopes that so far have not materialized.
In humans, there are two major families of inducible antibacterial substances. The first class is the cationic -defensin family coded for by a number of genes (7,8). Human -defensin1 (HBD1) has been demonstrated in urogenital tissues (9,10), including the epithelial cells of the loop of Henle, of distal tubules, and of collecting ducts (9). The molecule, interestingly, had originally been isolated from human hemodialysate of patients with renal failure (11). During human kidney infections, the inducible defensin is strikingly elevated (12) and, conversely, mice genetically manipulated to lack the murine homolog have bacteria in their urines (10).
The second class, investigated in this study by Chromek et al., are the cathelicidins. In contrast to the -defensin family, only one single gene codes for cathelicidin in humans, CAMP (cathelicidin antimicrobial peptide), and in mice, CRAMP (cathelicidin rodent antimicrobial peptide). The human cathelicidin precursor (hCAP18) contains a C-terminal cationic antimicrobial peptide domain that is activated by cleavage from the N-terminal cathelin portion of the propeptide. The precursor is synthesized and stored in secondary granules of neutrophils, but also is found in other cells exposed to microbes, such as epithelia of the mouth, tongue, esophagus, intestine, cervix, vagina (13), lung (14), as well as salivary, sweat, or mammary glands (15,16). The precursor is processed to release the active antimicrobial peptide LL37, which also attracts neutrophils, monocytes, and T cells through a G-protein–coupled receptor for formyl peptide receptor-like 1 (FPRL1) (17). In addition to their bactericidal activity, LL37 also modulates several aspects of immune function (18,19).
What is the evidence that cathelicidin is involved in the defense against urinary tract infection in humans? The authors examined the cathelicidin concentration in the urine of healthy children and children with acute urinary tract infections. Low concentrations were found in normal urine; in children with urinary tract infections the concentration was higher but not tightly correlated to the number of leukocytes, suggesting another source of cathelicidin.
In a next step, the authors therefore examined noninfected human kidney tissue by PCR and ELISA. They found both CAMP mRNA and cathelicidin peptide. The peptide was, however, in the tubule lumen and the parenchymal cells did not stain for cathelicidin.
To assess the response of renal tissue to infection, the authors incubated human renal cortex with uropathogenic E. coli; 5 min after exposure to the bacteria a rapid increase of CAMP mRNA was seen, suggesting that cathelicidin is the acute emergency reaction postulated by Kass (2). The increase in CAMP mRNA lessened after 135 min, either as a result of less synthesis or increased breakdown of the mRNA. In contrast, the active peptide LL37 continued to be released into the media.
Secretion of cathelicidin during urinary tract infection could be confirmed in mice; this model showed that both leukocytes and cells other than leukocytes are the source of the peptide.
To prove that cathelicidins can indeed kill uropathogenic E. coli, the bacteria were exposed to micromolar concentrations of LL37 and its murine homolog, both of which were bactericidal.
The role of cathelicidin to protect against urinary tract infection was assessed in mice with an intact or deleted CAMP gene. The end point was the number of bacteria attached to the bladder 1 h after infection, a time point before neutrophils had entered the urinary space. The number of adhering bacteria was greater in the CAMP knockout mice. It was not influenced by whether the mice had normal neutrophil counts or neutropenia after treatment with neutrophil-specific monoclonal antibody, excluding a major role of neutrophils in this early response.
A final observation arguing for an important in vivo role of cathelicidin was that, in children with more invasive upper urinary tract infections, the E. coli in their urine were more resistant against LL37.
These observations are strong arguments that, by increased synthesis and secretion of cathelicidin, resident uroepithelial cells play an important neutrophil-independent role in the protection against uropathogenic bacteria. In later stages, neutrophils appear to be an additional source of cathelicidin. The relative roles of the two inducible antibacterial peptides (i.e., -defensin and cathelicidin) in the protection against bacterial invasion of the urinary tract require further studies. Presumably both are acute emergency systems to ward off bacterial invasion before the above-mentioned slower mechanisms take over (1).
Interestingly, the authors noted that the early release of cathelicidin did not depend on an increase in mRNA, possibly the result of unblocking translation of mRNA into peptide. The assumption that translation is not tightly linked to transcription would also be compatible with the observation that the peptide could not be detected in the renal cells despite the presence of mRNA. The same has also been observed in the gut (20) and may be a useful adaptive mechanism, as antimicrobial peptides at high concentrations are cytotoxic for eukaryotic cells (21).
In retrospect, these experiments are an impressive confirmation of the hypothesis of Kass (2) that uroepithelial cells play an important sentinel role in the defense against urinary tract infection.
|
References |
|---|
|
TopReferencesReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
HOME
CURRENT ISSUE
ARCHIVES
JASN Express
ONLINE SUBMISSION
AUTHOR INFO
EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP |
