| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Renal Immunopathology Laboratory, Associazione Nuova Nefrologia, Fondazione D’Amico per la Ricerca sulle Malattie Renali, c/o Department of Nephrology and Immunology, San Carlo Hospital, Milan, Italy;
Microscopy and Image Analysis Consortium, Department of Neurosciences and Biomedical Technologies, Milano Bicocca University, Monza, Italy;
Molecular Medicine, Biomedicum Helsinki, Helsinki, Finland;
Protein Chemistry Unit, Biomedicum Helsinki, Helsinki, Finland;
|| Medical Policlinic, University of Munich, Munich, Germany;
¶ Sanofi-Aventis, Frankfurt, Germany;
# Neurobiology of Learning Unit, Universita’ Vita-Salute S. Raffaele, Segrate, Italy; and
** Biology of Neurodegenerative Diseases Laboratory, Mario Negri Institute, Milan, Italy
1Correspondence: Renal Immunopathology Laboratory, Associazione Nuova Nefrologia and Fondazione D’Amico per la Ricerca sulle Malattie Renali, c/o San Carlo Borromeo Hospital, via Pio II, 3, Milan 20153, Italy. E-mail mp.rastaldi{at}fastwebnet.it
ABSTRACT
Although patients with chronic renal failure are increasing worldwide, many aspects of kidney biology remain to be elucidated. Recent research has uncovered several molecular properties of the glomerular filtration barrier, in which podocytes, highly differentiated, ramified cells that enwrap the glomerular basement membrane, have been reported to be mainly responsible for filter’s selectivity. We previously described that podocytes express Rab3A, a GTPase restricted to cell types that are capable of highly regulated exocytosis, such as neuronal cells. Here, we first demonstrate by a proteomic study that Rab3A in podocytes coimmmunoprecipitates with molecules once thought to be synapse specific. We then show that podocytes possess structures resembling synaptic vesicles, which contain glutamate, coexpress Rab3A and synaptotagmin 1, and undergo spontaneous and stimulated exocytosis and recycling, with glutamate release. Finally, from the results of a cDNA microarray study, we describe the presence of a series of neuron- and synapse-specific molecules in normal human glomeruli and confirm the glomerular protein expression of both metabotropic and ionotropic glutamate receptors. These data point toward a synaptic-like mechanism of communication among glomerular cells, which perfectly fits with the molecular composition of the glomerular filter and puts in perspective several previous observations, proposing a different working hypothesis for understanding glomerular signaling dynamics.— Rastaldi, M. P., Armelloni, S., Berra, S., Calvaresi, N., Corbelli, A., Giardino, L. A., Li, M., Wang, G. Q., Fornasieri, A., Villa, A., Heikkila, E., Soliymani, R., Boucherot, A., Cohen, C. D., Kretzler, M., Nitsche, A., Ripamonti, M., Malgaroli, A., Pesaresi, M., Forloni, G. L., Schlöndorff, D., Holthofer, H., D’Amico, G. Glomerular podocytes contain neuron-like functional synaptic vesicles.
Key Words: synaptic communication • exocytosis • endocytosis • visceral epithelial cells • renal glomerulus
DESPITE A LOT OF RESEARCH efforts and the dramatic worldwide increase of patients affected by chronic renal failure, several aspects of glomerular physiology still await a clarification. The last advances in glomerular biology have highlighted the central role of podocytes in the maintenance of the glomerular filtration barrier and revealed precious details of their complex molecular structure (reviewed in Ref. 1
).
From morphological and biochemical studies, it has become evident that podocytes have some similarities with neuronal cells. Both cells are highly arborized, have a common cytoskeletal organization, a common machinery for process formation (2)
, and share several expression-restricted proteins, such as nephrin (3)
, densin (4)
, glomerular epithelial protein 1 (GLEPP1) (5)
, the amino acid transporters CAT3 and EAAT2 (6)
, the cytoskeletal proteins synaptopodin (7)
and drebrin (8)
, and the RNA processing protein Sam68-like mammalian protein 2 (9)
.
We have contributed to this line of research by reporting that podocytes possess the small GTPase Rab3A (10)
, a protein involved in processes of highly regulated exocytosis and most abundant in neurons, in which it modulates the vesicle fusion step at the presynaptic membrane (11)
. The presence of Rab3A in podocytes raises the question of its function in these cells, in which highly regulated exocytosis has so far not been considered to play a role.
In the present study, we have performed a proteomic analysis of molecules that coimmunoprecipitate with Rab3A from extracts of normal glomeruli and found protein molecules once thought to be synapse specific. On the basis of these results, we have carried out a detailed analysis of podocyte vesicles expressing Rab3A and found that they contain glutamate and coexpress synaptotagmin 1 and glutamate transporters. Furthermore, these vesicular structures undergo spontaneous exocytosis and endocytosis with glutamate release, and these processes can be enhanced by the application of 
-latrotoxin. These results suggest that podocytes contain intracellular vesicular organelles that behave like functional synaptic vesicles.
MATERIALS AND METHODS
Immunoprecipitation
Glomeruli from normal mouse kidneys were isolated by sieving, as described (12)
, and further purified manually under a stereomicroscope.
Protein extracts were prepared from isolated glomeruli and from normal mouse total brain by sonication in modified Ripa Buffer (10 mM Tris HCl pH 7.5, 10 mM EDTA, 0.5% Nonidet P-40, 0.5% DOC, 0.1% SDS) with addition of protease inhibitor-cocktail (Roche Applied Science, Monza, Italy).
Before immunoprecipitation, 400 µg of each lysate were precleared with protein G immobilized on agarose (Sigma-Aldrich, Milan, Italy) in 500 µl IP buffer (1% Triton x100, 75 mM NaCl, 50 mM, HEPES pH 7.4, 1 mM EGTA, 1 mM Na3VO4, protease inhibitor) for 1 h at 4°C to remove proteins that might nonspecifically bind to the beads. Precleared lysates were immunoprecipitated for 2 h at 4°C with 5 µg of rabbit polyclonal anti-Rab3A antibody (Ab) (Abcam, Cambridge, UK), which had been first conjugated with protein G beads. The immunoprecipitates were pelleted by centrifuging 13,000 rpm for 10 min at 4°C and washed 3 times with IP buffer. A specificity-control with rabbit IgG (Zymed, Histoline, Milan, Italy) and a negative control without Ab were added for each assay to test specificity of immunoprecipitation. Western blot analysis with a mouse monoclonal anti-Rab3A Ab (Synaptic System, Göttingen, Germany) in nonreducing condition was performed to confirm the accuracy of immunoprecipitation.
Sample preparation for MALDI-TOF-MS analysis
Protein concentration of the immunoprecipitates was measured by the Bradford method, and 20-µg protein samples were loaded on a 10% sodium dodecyl polyacrylamide (SDS-PAGE) gel, separated by electrophoresis and silver-stained (13)
. Six stained protein bands with approximate sizes ranging from 30 to 100 kDa were excised and cut into pieces. The stain was removed with 15-mM potassium ferricyanide in 50-mM sodium thiosulfate followed by washes with water, shrinking with acetonitrile (ACN), and drying in a vacuum centrifuge. Proteins were reduced with 20 mM dithiothreitol in 0.1 mM NH4HCO3 for 15 min at room temperature and water with ACN shrinking in between, followed by final ACN shrinking and drying in vacuum. Ingel trypsin digestion was performed with 0.05 µg/µl sequencing-grade modified trypsin (Promega, Madison, WI) in 10 mM NH4CO3/10%ACN on an ice bath for 10 min, followed by overnight incubation at 37°C. The digested peptides were extracted with 0.1%TFA/60%ACN and desalted by reverse phase material (ZIP TIP C18, Millipore, Bedford, MA), packed into a gel-loading pipette tip according to manufacturer’s instructions.
MALDI-TOF-MS analysis
Fingerprinting for extracted peptides was performed with a Biflex matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry (2-GHz digitizer; Bruker, Rheinstetten, Germany) at a local protein chemistry unit (Biomedicum Helsinki). Positive ion reflector mode was used with an accelerating voltage of 19,000 V and delayed extraction of 2 ns. Internal peptide calibration standards (Bruker Daltons, Bremen, Germany) were applied to obtain higher peptide mass accuracy. Results were analyzed by Profound search engine (http://prowl.rockefeller.edu/profound_bin/WebProFound.exe) and confirmed by Western blot analysis, using a rabbit antiglycine binding protein (Vinci-Biochem, Vinci, Florence, Italy) and a rabbit antisynaptotagmin 1 (Abcam).
Immunofluorescence studies
An indirect immunofluorescent method was performed on 5-µm-thick acetone-fixed tissue sections, on 1-µm-thick paraformaldehyde-fixed semithin sections, and on acetone-fixed cultured cells.
The following primary antibodies were used for the study: 1) podocyte characterization: Rabbit antimouse nephrin (intracellular domain) (#035, provided by H. Holthofer, University of Helsinki, Finland), Mouse antisynaptopodin (Progen, Heidelberg, Germany), Rabbit anti-WT1 (C-19) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), Mouse antismooth muscle actin alpha isoform (Zymed), Rat antimouse CD31/PECAM (Abcam), Mouse antipan cytokeratin (Abcam); 2) neuronal cell markers: rabbit anti AMPA glutamate receptor 1 (Chemicon International, Prodotti Gianni, Milan, Italy), rabbit anti AMPA glutamate receptor 2 (Chemicon International), rabbit anti-GRM7 (Abcam), rabbit anti-GRM8 (Abcam), mouse anti-NMDA1 receptor (Synaptic System), rabbit anti-NMDA1 receptor (Abcam), rabbit antiglutamate (Abcam), mouse anti-Rab3A (Synaptic System), rabbit anti-SAP102 (Abcam), rabbit antibassoon (Synaptic System), rabbit antipiccolo (Synaptic System), rabbit antisynapsin 1 (Abcam), mouse antisynaptophysin (Abcam), mouse antisynaptotagmin 1 (p65, C terminal domain, provided by Dr. A. Malgaroli, Università Vita-Salute S. Raffaele, Milan, Italy), rabbit antisynaptotagmin 1 (syt 1, N-terminal domain, provided by Dr A. Malgaroli, Università Vita-Salute S. Raffaele, Milan, Italy), rabbit antisyntaxin 1A (Synaptic System), rabbit antivacuolar proton pump (VATPase, 116 kDa subunit, Synaptic System), mouse anti-VGlut1 (Vesicular Glutamate Transporter 1, Synaptic System).
As secondary antibodies, we used the following, all from Molecular Probes (Invitrogen, S. Giuliano Milanese, Italy): Alexa Fluor 488 goat anti-rabbit IgG, Alexa Fluor 546 goat anti-rabbit IgG, Alexa Fluor 488 goat antimouse IgG, Alexa Fluor 546 goat antimouse IgG, Alexa Fluor 488 goat antimouse IgG highly cross adsorbed, Alexa Fluor 546 goat antimouse IgG highly cross adsorbed, Alexa Fluor 488 chicken antirat IgG highly cross adsorbed.
Specificity of Ab labeling was demonstrated by the lack of staining after substituting proper control immunoglobulins (rabbit primary Ab isotype control and mouse primary Ab isotype control, both from Zymed, and Rat IgG1 negative control, from Serotec, Kidlington Oxford, UK) for the primary antibodies.
Positive controls were performed by applying the stainings on brain tissue sections and primary neuron cell cultures. Positive and negative controls were run concurrently.
Slides were mounted with VectaShield aqueous mounting medium (Vector Laboratories, DBA Italia SRL, Milan, Italy). Images were acquired using a digital video camera (Leica DC 250, Leica Italia, Milan, Italy) connected to a Leica DM IRB microscope (Leica), using the Leica IM1000 software for image storing and the Leica Q-Fluor software to acquire double stainings.
Immunogold electron microscopy
An indirect immunogold labeling procedure was performed on ultrathin frozen kidney sections, as described previously (10)
. Briefly, after blocking, the material was incubated with the primary monoclonal mouse antiglutamate, then by the secondary 10-nm gold-conjugated goat antimouse secondary Ab (Aurion, DBA, Milan, Italy). Specificity of Ab labeling was demonstrated by the lack of staining after substituting proper control immunoglobulins (Zymed) for the primary Ab.
Podocyte cell cultures
Kidneys taken from 7- to 10-day-old normal mice were decapsulated, washed in Ca2+ and Mg2+ free Hanks medium, and treated with collagenase type I AS (Sigma-Aldrich) 1.5 mg/ml, for 1 min at 37°C. The reaction was stopped by growth medium consisting of Dulbecco’s modified Eagle medium: F12 supplemented with 10% FCS, 5 µg/ml transferrin, 10–7M hydrocortisone, 5 ng/ml sodium selenite, 0.12 U/ml insulin, 100 µg/ml penicillin, 100 µg/ml streptomycin, 2 mM L-glutamine (Sigma-Aldrich).
Glomeruli isolated by sieving and further manually purified, were seeded in culture flasks precoated with collagen type IV (Sigma-Aldrich) at 37°C in 5% CO2 atmosphere. On days 4 to 5 podocyte growth started and allowed by day 8 to detach glomeruli using trypsin-EDTA. First passage podocytes, which resulted in >90% pure as judged by light microscopy inspection, were seeded on flasks and chamber slides. Cell characterization was performed by immunohistochemistry, using podocyte (nephrin, synaptopodin, WT1), epithelial (cytokeratins), smooth muscle (alpha-smooth muscle actin), and endothelial cell (CD31) markers.
Alpha-latrotoxin assay
Alpha-latrotoxin-stimulated glutamate release was detected by an enzymatic assay (14)
, based on the following reaction that occurs in the presence of glutamate dehydrogenase (GDH): Glutamate– + NAD+ + H2O 
ketoglutarate2– + NADH + NAD4+ + H+. Briefly, to this purpose, podocytes obtained from normal mouse glomeruli were plated and characterized. Before measurements, cells were thoroughly washed and resuspended for 1 h at 37°C in DME incubation buffer (109.5 mM NaCl, 5.3 mM KCl, 5.5 mM glucose with 1 mM MgCl2, 20 mM HEPES). The medium was further supplemented with GDH (Sigma 60 U/ml) and NAD+ (Sigma 1 mM), and incubated for 5 min. Then, -latrotoxin was added at subnanomolar (0.5 nM) or nanomolar (2.5 nM) concentration and spectrophotometric increase of optical density (OD) due to an increase of NADH was monitored at 340 nm.
The results were confirmed by using the EnzyPlus L-Glutamic Acid kit (Diffchamb AB, Vastra Frolunda, Sweden), a colorimetric assay based on the same reaction.
In inhibition experiments, 200 nM Bafilomycin A1 (Sigma), a V-ATPase inhibitor, that dissipates the electrochemical proton gradient necessary for glutamate uptake and storage into vesicles (15)
, was added for 30 min at 37°C before -latrotoxin stimulation.
In further experiments, the following dopamine and acetylcholine (Ach) agonists were applied: apomorphine hydrochloride (dopamine agonist, 5 µM), cytisine (nicotinic receptor agonist, 100 µM), and pilocarpine hydrochloride (muscarinic receptor agonist, 3 µM).
Human material
Kidney tissue from transplant living donors was collected in a European multicenter study, the European cDNA Bank (ERCB), obtained after informed consent of the patients and with acknowledgment of the local ethical committees.
Microdissection and RNA isolation
Microdissection of material stored in RNAlater into glomerular and tubular fragments was performed manually under a stereomicroscope using two dissection needle holders (16)
. Total RNA was isolated from microdissected glomeruli and from tubular fragments using a commercially available silica gel-based isolation protocol (RNeasy mini kit; Qiagen, Hilden, Germany). RNA quality and quantity were controlled by microfluid electrophoresis using the RNA 6000 LabChip on a 2100 Bioanalyzer (Agilent Technologies, Waldbronn, Germany).
Target preparation and microarray analysis
15–30 ng of total RNA was reverse-transcribed in a 10-µl reaction with 0.5-µl SuperScript II (Invitrogen, Karlsruhe, Germany, 200 U) and 5 pmol of T7-(dT)24 primer (5'-GGCCAGTGAATTGTAATACGACTCACTATAGGGAGA-(dT)24–3') in 5x first-strand buffer (Invitrogen), 1 µl 100 mM DTT (Invitrogen, Carlsbad, CA), 0.5 µl 10 mM dNTP (Amersham Pharmacia, Freiburg, Germany), 0.5 µl T4gp32 (Amersham Pharmacia, 8 mg/ml), 0.25 µl RNase inhibitor (RNasin, Promega, Mannheim, Germany, 40 U/µl) at 42°C for 50 min, then 10 min 45°C, 10 min 50°C and 15 min for 70°C. The second-strand synthesis was performed at 16°C for 2 h, in the presence of E. coli enzymes, DNA polymerase I (Invitrogen, 40 U), DNA ligase (Invitrogen, 10 U), RNase H (Roche Applied Science, Basel, Switzerland, 1 U/µl) and 1X second-strand buffer (Invitrogen). The double-stranded cDNA was blunt-ended using 5 U of T4 DNA polymerase (Invitrogen) purified by phenol/chloroform extraction and transcribed using the MEGAscript T7 kit (Ambion Europe, Huntingdon, UK) as described by the manufacturer. The unlabeled cRNA was purified using RNeasy minicolumn (RNeasy kit; Qiagen, Hilden, Germany) followed by a quality check using the Bioanalyzer mRNA pico assay, showing cRNA length. The cRNA was precipitated and dissolved in RNase free water. For the second round of cDNA synthesis 200–300 ng of cRNA was reverse-transcribed in a 10-µl reaction as described above using 0.45 µg pd(N6) random hexamer primer. RNase H (2 U) was then added and incubated for 30 min at 37°C followed by 3 min at 95°C. The second-strand synthesis was performed at 16°C for 2 h in the presence of 50 pmol of T7-(dT)24 primer, 5 U DNA Polymerase I, 1 U RNase H, 15 nMol dNTPs and 1x second-strand buffer (Invitrogen). Double-strand cDNA was blunt ended, purified by phenol/chloroform extraction and subjected to in vitro transcription using the BioArray High Yield RNA transcript labeling kit (Enzo Laboratories, Farmingdale, NY) as described by the manufacturer. The biotin-labeled cRNA was purified using RNeasy minicolumn (RNeasy kit; Qiagen) followed by a quality check using the Bioanalyzer mRNA nanoassay. The fragmentation, hybridization, staining and imaging procedures were performed according to the Affymetrix Expression Analysis Technical Manual (Affymetrix UK Ltd, High Wycombe, UK).
RESULTS
Proteomic analysis of molecules that coimmunoprecipitate with Rab3a extracted from normal glomeruli
Rab3A was immunoprecipitated from normal mouse kidney glomeruli and, for comparison, from normal mouse brain (Fig. 1
A). After SDS-PAGE electrophoresis and silver staining, 6 bands with approximate sizes ranging from 30 to 100 kDa, selected on the basis of their apparent overexpression in glomerular immunoprecipitates, were trypsin digested and analyzed by mass spectrometry. MALDI-TOF analysis did not demonstrate coimmunoprecipitation of Rab3A with any specific podocyte protein. It did however show the coimmunoprecipitation of Rab3A with synaptotagmin 1 (Syt1) and glycine-, glutamate-, thienylcyclohexylpiperidine-binding protein (Gly-BP), that are molecules involved in neuronal processes of synaptic transmission, and with Rab GDP-dissociation inhibitor, which belongs to the cellular cycle of Rab3A (Fig. 1B
). Western blot experiments confirmed the presence of both Syt1 and Gly-BP in the immunoprecipitates (Fig 2
A, B), as well as in glomerular and primary podocyte protein extracts (Fig 2C, D
); immunohistochemistry allowed their visualization at glomerular level (Fig 2E, F
).
|
|
Glutamate-containing vesicles are detected in podocytes
Foot processes of podocytes have been described to contain vesicular structures (17)
, and we have shown by immunogold electron microscopy (EM) some of these vesicles to be positive for Rab3A (10)
. In the brain of Rab3A null mice the prevalent defects occur in glutamatergic synapses (18)
. In addition, previous data have shown that a glutamate transport system by specific amino acid carriers is present in cultured podocytes (6)
.
On the basis of this observation, it is reasonable to think that podocytes might contain a vesicular carrier for glutamate, and vesicles might be loaded with this neurotransmitter. Indeed, by immunocytochemical analysis, we have found that Rab3A-positive vesicles are also strongly positive for glutamate. Such a precise colocalization of glutamate and Rab3A in vesicular structures was mainly found along podocyte processes (Fig. 3
). EM performed on kidney sections and cultured podocytes confirmed the presence of a number of vesicles located in podocyte processes (Fig 4
A, B), and immunogold EM showed glutamate particles contained in these vesicular structures (Fig 4C
). Glutamate and glutamate receptors were also demonstrated by immunohistochemistry and Western blot analysis (Fig. 5
).
|
|
|
In parallel experiments we tested for the presence of glutamate carriers. This analysis uncovered the presence in podocytes of the vesicular glutamate transporter VGlut1 (Fig. 6
), and of the vacuolar proton pump (V-ATPase) (Fig. 7
), suggesting that podocytes are able to recruit glutamate into vesicles, as it happens in glutamatergic neurons.
|
|
Spontaneous vesicle exocytosis and recycling in podocyte foot processes
From these results, we hypothesized that the vesicular structures seen in podocyte foot processes might undergo exocytic/endocytic cycling. The presence of Syt1, identified by proteomic analysis, indicated that vesicular fusion and recycling should be detected by the uptake of specific antibodies directed against the lumenal epitope of this vesicular protein. In neuronal cells, during exocytosis, lumenal components of synaptic vesicles are briefly exposed at the surface of presynaptic terminals: the addition of anti-Syt1 antibodies to the culture medium results in their binding to exposed lumenal epitopes followed by their internalization inside recycling synaptic vesicles (19)
. Therefore, cultured podocytes were incubated with rabbit anti-Syt1 (anti-N-terminal lumenal domain) antibodies for a 2-h period. Cells were then fixed and processed for standard immunocytochemistry where the rabbit antibodies were detected using fluorescent-tagged secondary antibodies to identify sites of uptake (Fig. 8
A). On the basis of the occurrence of fluorescent puncta in all examined podocytes, we can conclude that these antibodies were efficiently internalized.
|
Furthermore, after fixation, cells were probed with a monoclonal antibody directed against the cytosolic portion of Syt1 (Fig 8B
). The precise colocalization of the two anti-Syt1 markers (Fig 8C
) definitely demonstrated that staining occurs by specific interaction with synaptic-like vesicles, and not by a general mechanism of pinocytosis.
When the same experiment was repeated after a 24 h incubation of the cells with the complete (150 kDa) tetanus toxin (TeNT), which inhibits neurotransmitter release, the uptake of the antilumenal epitope of synaptotagmin (Fig 8E
) was completely abolished in 89% of the cells.
Glutamate exocytosis from podocytes vesicles: regulation by -latrotoxin
Alpha-latrotoxin is a presynaptic neurotoxin widely used to trigger neurotransmitter release from synaptic vesicles. The application of increasing concentrations of -latrotoxin to presynaptic nerve terminals leads to discrete dose-dependent effects on synaptic transmission. At low subnanomolar concentrations, -latrotoxin stimulates mild neurotransmitter release, due to individual vesicle fusion events, without evident morphological changes of synaptic terminals. The addition of higher concentrations, in the nanomolar range, causes instead a strong burst of neurotransmitter release followed by continuous steady state release, with clear morphologically detectable depletion of synaptic vesicles (20)
.
Applying subnanomolar (0.5 nM) and nanomolar (2.5 nM) -latrotoxin concentrations on cultured podocytes, we found that glutamate release was actually taking place with kinetics that precisely parallel those described in neurons (Fig. 9
A). Confirming the similarities with neurons, a disappearance of glutamate immunostaining from podocyte processes was evident only after incubation with nanomolar -latrotoxin doses.
|
The same experiment, repeated using dopamine and Ach agonists, showed that only the muscarinic agonist pilocarpine was able to induce some glutamate exocytosis, although the effect was milder if compared to that obtained with -latrotoxin (Fig 9B
).
Considering that at nanomolar concentrations, binding of -latrotoxin to specific receptors is necessary for the toxin to insert into biological membranes (20)
, the exocytic response we obtained on cultured podocytes suggests that -latrotoxin receptors must be present on the surface of these cells. Furthermore, glutamate released by -latrotoxin must be contained in vesicular structures because the effect was abolished by the preincubation of podocytes with the V-ATPase inhibitor bafilomycin (Fig. 9A
).
These results were confirmed by experiments in which recycling vesicles were stained with the styryl dye FM1–43 (Molecular Probes). Styryl dyes (21)
reversibly insert into the surface of lipid membranes and have no fluorescent properties in aqueous solution, but they become intensely fluorescent on membrane binding, allowing a labeling of recycling vesicles that is easily detectable by fluorescence microscopy. Actually, primary podocytes accumulated the styryl dye FM1–43 a few seconds after a first -latrotoxin stimulus and almost completely discharged the dye after a second -latrotoxin stimulation (Fig 10
A–C), features that confirm the existence in these cells of a synaptic-like vesicle compartment whose exocytosis and recycling can be regulated.
|
Because regions of neurons labeled with styryl dyes have strong immunoreactivity for synaptic vesicle proteins (21)
, we used the fixable variant of FM1–43 for a further experiment, demonstrating that FM1–43 fluorescent areas were also labeled by an Ab directed against the synaptic vesicle protein synaptophysin (Fig 10D-I
).
Microarray screening of normal human glomeruli
To get a global view of the molecules involved in synaptic transmission that are possibly synthesized by normal human glomeruli, we generated microarray expression data from microdissected glomeruli of transplant living donor biopsies. These data (Supplemental Table 1) show that normal human glomeruli do transcribe not only all the molecules that we have demonstrated in cultured cells but also those of a complete series of synaptic molecules, neurotransmitter receptors, and neuron-specific molecules. Confirming some of these data at protein concentration, we detected by immunohistochemistry and Western blot analysis synaptophysin, synapsin, piccolo, bassoon, and syntaxin in cultured podocytes and in normal kidney glomeruli.
Furthermore, comparing the microarray glomerular data with the data obtained by the microdissected tubulointerstitial compartment of the same living donor biopsies, we could appreciate that about 100 of these molecules appear specifically transcribed in glomeruli (Supplemental Table 2), showing not detectable levels of expression in the interstitial compartment despite its content in innervating fibers.
DISCUSSION
Glomerular podocytes are highly differentiated cells with a complex ramified structure: few major processes departing from the cell body give rise on both sides to many fine pedicels or foot processes, which interdigitate with corresponding projections of neighbor cells to completely enwrap the glomerular basement membrane. Specialized adhesions are interposed among foot processes, forming the so-called slit diaphragm, and between the basal domain of foot processes and the glomerular basement membrane. It has been recently described by several authors (reviewed in 22) that both the slit diaphragm and the basal domain of the foot processes are highly dynamic signaling domains, and our present results strongly suggest that signals can be triggered in a synaptic-like way.
Our investigation shows in fact that podocytes possess vesicular structures, which from both the molecular and functional point of view strictly resemble synaptic vesicles: they are composed by synaptic vesicle molecules, among them Rab3A, synaptotagmin 1, synapsin 1 and synaptophysin, are able to undergo spontaneous and regulated exocytosis/endocytosis and release glutamate, which can accumulate into vesicles thanks to the presence of vesicular glutamate transporters.
Cells need to talk to each other, and more differentiated, highly specialized cells have developed sophisticated means of communication that reach the highest expression in neuronal cells, but are not limited to them. The concept of an immunological synapse provides a very good example of specialized communication, and the functional expression of glutamate-signaling molecules has been described for several nonneuronal cell types, mainly but not only belonging to the endocrine system, where glutamate likely acts as an extracellular signaling mediator (23)
.
It is our opinion that, given the complexity of podocyte structure and the continuous stimulation and stress by blood pressure and blood contents, these cells are likely in the need of a precise and fast modality of communication among themselves and with the other glomerular cells.
Furthermore, our data seem to prove that normal glomeruli express a wide array of neuron-specific molecules, and recent data have actually shown that a depolarization, due to an increase of free intracellular calcium, does occur by exposure of podocytes to dopamine and ACh via specific receptors (24
, 25)
, suggesting for podocytes an even more neuron-like behavior.
On the basis of this view, the slit diaphragm should be regarded as a synaptic adhesion, composed by synaptic adhesion molecules. Considered in this way, many already described features fall into place, starting from the morphological aspect of the slit diaphragm, which is a modified adherens junction where staggered cross-bridges extend from the slit walls to a longitudinal central filament, generating its typical zipper-like structure. Furthermore, the cytoplasm opposite the points of attachment of the diaphragm to the cell membrane, shows an increased density, sometimes asymmetrical (26)
. The molecular composition of the slit diaphragm also fits well with a synaptic adhesion. It is known that neuron-neuron synapses bear at least cadherin-like and immunoglobulin (Ig) superfamily-like adhesion molecules (27)
. The slit diaphragm is precisely composed by both Ig-like molecules, such as nephrin and NEPH1, and cadherin molecules, such as P-cadherin, and the proto-cadherin FAT (1)
.
As for their function, apart from promoting the stability of synapses, many data support the role of synaptic adhesion molecules in target recognition, that is, to help in choosing the right partners from a network of processes (28)
. Establishing interdigitations (29)
, differentiating podocyte foot processes possibly use the same mechanism to find their right place and partner, and in this respect, it seems worth noting that SYG-1, the C. elegans ortholog of NEPH-1, is a molecule that has been isolated in a genetic screen for mutants defective in synaptic positioning (30)
.
Our "synaptic view" also fits with the adhesive properties of the basal cell membrane (the sole plate) of foot processes with the glomerular basement membrane, the structure making direct contact with both endothelial and mesangial cells. In this case, a kind of neuromuscular junction comes to mind, first because of the presence of a basal lamina, second because of its composition by integrins, dystroglycans (31)
, and especially agrin, considered to be a critical nerve-derived organizer of postsynaptic differentiation in neuromuscular junctions (32)
. Agrin is produced by podocytes and is localized in the glomerular basement membrane (33)
. Furthermore, the so-called synaptic laminin (s-laminin, or laminin ß2 chain) is selectively expressed in the whole organism only by the synaptic basement membrane of neuromuscular junctions and by the kidney glomerular basement membrane (34)
.
In conclusion, we believe that our hypothesis of synaptic transmission in podocytes may help to better explain several already known glomerular physiological features and surely offers a different perspective for understanding glomerular cell signaling that may be important for advancing of knowledge of glomerular cell behavior. This is especially needed because chronic renal failure is increasing worldwide, whereas there are few therapeutic options, and those that are available mainly consist of generic immunosuppressive drugs or symptomatic treatments, because of the still limited knowledge of kidney cell biology.
ACKNOWLEDGMENTS
The research has been performed in the EU Framework V Program "Progressive Renal Disease" QLRT–2001–01215.
Received for publication September 10, 2005. Accepted for publication January 3, 2005.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  N. Guillen, M. A. Navarro, C. Arnal, E. Noone, J. M. Arbones-Mainar, S. Acin, J. C. Surra, P. Muniesa, H. M. Roche, and J. Osada Microarray analysis of hepatic gene expression identifies new genes involved in steatotic liver Physiol Genomics, May 13, 2009; 37(3): 187 - 198. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sato, J. Xu, S. Okuyama, L. B. Martinez, S. M. Walsh, M. T. Jacobsen, R. J. Swan, J. D. Schlautman, P. Ciborowski, and T. Ikezu Spatial Learning Impairment, Enhanced CDK5/p35 Activity, and Downregulation of NMDA Receptor Expression in Transgenic Mice Expressing Tau-Tubulin Kinase 1 J. Neurosci., December 31, 2008; 28(53): 14511 - 14521. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Saleem, J. Zavadil, M. Bailly, K. McGee, I. R. Witherden, H. Pavenstadt, H. Hsu, J. Sanday, S. C. Satchell, R. Lennon, et al. The molecular and functional phenotype of glomerular podocytes reveals key features of contractile smooth muscle cells Am J Physiol Renal Physiol, October 1, 2008; 295(4): F959 - F970. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Holthofer Molecular architecture of the glomerular slit diaphragm: lessons learnt for a better understanding of disease pathogenesis Nephrol. Dial. Transplant., August 1, 2007; 22(8): 2124 - 2128. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Asanuma, K. N. Campbell, K. Kim, C. Faul, and P. Mundel Nuclear relocation of the nephrin and CD2AP-binding protein dendrin promotes apoptosis of podocytes PNAS, June 12, 2007; 104(24): 10134 - 10139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Bertelli, M. Regoli, L. Fonzi, R. Occhini, S. Mannucci, L. Ermini, and P. Toti Nestin Expression in Adult and Developing Human Kidney J. Histochem. Cytochem., April 1, 2007; 55(4): 411 - 421. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
60 kDa and 50 kDa, respectively, in both samples. C, D) Western blot analyses on protein lysates from normal mouse brain, glomeruli (glom), and cultured podocytes (podo) using a rabbit anti-GlyBP (C) and a rabbit anti-Syt1 (D), allow the detection of bands of 
