| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Division of Nephrology and
Division of Infectious Diseases, Mount Sinai School of Medicine, New York, New York, USA; and
Division of Nephrology, Juntendo University School of Medicine, Tokyo, Japan
1Correspondence: Mt. Sinai School of Medicine, Division of Nephrology, Box 1243, One Gustave L. Levy Pl., New York, NY, USA 10029. E-mail: lewis.kaufman{at}mssm.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: HIVAN • pseudocrescent • collapsing glomerulopathy
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
. Expression of HIV-1 gene products in podocytes induces these changes in both human HIVAN (4
, 5)
and in the HIV-1 transgenic (Tg26) mouse model of HIVAN (6)
.
In collapsed glomeruli in HIVAN, hypertrophied and proliferating podocytes reorganize themselves into aggregates that fill an enlarged Bowman’s space. In severe disease, these crowding podocytes are called "pseudocrescents", which differ from traditional crescents in that they are composed of clustered abnormal podocytes and do not contain the typical components of glomerular crescents such as spindle or inflammatory cells (3)
. Furthermore, in HIVAN, podocytes not only demonstrate foot process fusion, but they also often lack primary processes such that the cell body adheres directly to the GBM. These cells are sometimes detached from the GBM and assume a cuboidal shape, often containing cytoplasmic protein resorption droplets (3)
.
The highly specialized structure and function of the podocyte, including its intricate foot processes, are highly dependent on complex cytoskeletal machinery that include the formation of a precise actin structure (7)
. HIV-1 infection of podocytes induces rearrangement of the actin filaments (8)
, such that the cells are more elongated than control cells. The normal stress fiber pattern of actin filaments disappears, and actin filaments are mislocalized to the cell membrane surface.
The distinctive response of podocytes to HIV-1 infection is the result of a complex network of interactions involving both viral and host proteins. We have recently begun to characterize these pathways. For example, podocyte proliferation and dedifferentiation require activation of the Src-dependent Stat3 and MAPK pathways via the nef gene of HIV-1 (9)
. Many of the other host responses to HIV infection, however, have not yet been elucidated.
Previously, we reported that sidekick-1 (sdk-1), a transmembrane protein of the Ig superfamily, is highly up-regulated in vitro in HIV-1 conditionally immortalized transgenic podocytes and in vivo in glomeruli of both the Tg26 (
gag-pol) HIVAN mouse model and from human kidney biopsies (10)
. Sdk-1 and its ortholog sidekick-2 (sdk-2) each consist of a large extracellular domain containing 6 Ig motifs followed by 13 fibronectin type III repeats, a single transmembrane domain, and a short cytoplasmic tail (11)
. Sdk-1 and sdk-2 function as homophilic adhesion molecules, and cells expressing sdk-1 or sdk-2 exhibit a strong preference to interact exclusively with cells expressing the same sidekick isoform (11
, 12)
. Sdk-1 and sdk-2 are highly expressed in many organs during embryogenesis, and their expression stabilizes developing neuronal synapses and guides axons to specific synapses in developing chicken retinas (11)
.
We hypothesized that sdk-1 up-regulation in podocytes might alter podocyte architecture by changing the intercellular binding affinity of infected cells and that subsequent clustering of infected cells might ultimately lead to pseudocrescent formation in HIVAN. To test this hypothesis, we analyzed the effects of HIV-1 infection on podocyte adhesion and investigated the role of sdk-1 in inducing these changes. We also examined the effects of stable sdk-1 overexpression in murine podocytes on cellular morphology and cytoskeletal architecture.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
previously isolated from the "Immortomouse" was grown at 33°C in RPMI 1640 medium containing 10 U/ml mouse 
-IFN (Sigma, St. Louis, MO, USA). A conditionally immortalized HIV-1 podocyte cell line (14)
generated by crossing an HIV-1 transgenic mouse (Tg26) with the "Immortomouse" was propagated in an identical fashion. Cells were allowed to differentiate by transferring them to type-I collagen (BD Biosciences, San Jose, CA, USA) coated flasks at 37°C and removal of -IFN from the media to allow for degradation of the temperature-sensitive SV40 T-antigen. All podocytes were allowed to differentiate for at least 7 d prior to use in experiments.
Viral transduction
DsRed2 coding sequence was amplified by polymerase chain reaction (PCR) using pIRES2-DsRed2 plasmid (BD Biosciences) as template with a sense primer containing an SphI or a BamHI site (ccacatgcatgcaatggcctcctccgagaacgtc or cgcggatccatggcctcctccgagaacgtc) and an antisense primer carrying an EcoRV site (ggctaggatatcctacaggaacaggtggtggcg). The plasmid pNL4–3:G/P-DsRed2 was generated by digesting full-length pNL4–3 with SphI and MscI and then ligating the amplified DsRed2 PCR product after digestion with SphI and EcoRV. Primers were designed to ensure that DsRed2 would be expressed in the gag open reading frame. Control VVCW-DsRed2 plasmid was generated by digesting native VVCW (15)
(gift of Dr. Luca Gusella) with BamHI and EcoRV and ligating in amplified DsRed2 after digestion with BamHI and EcoRV.
The plasmids pNL4–3:G/P-DsRed2 and VVCW-DsRed2 were then used to generate VSV-G-pseudotyped virus for infection of undifferentiated murine podocytes (33°) as described previously (16)
. Concentrated virus was produced by ultracentrifugation of virus-containing media. Each virus was titrated to a multiplicity of infection of 3 as determined by immunofluorescence. After 96 h, infected podocytes were trypsinized and cells were replated in media without -IFN at 37° to allow for differentiation.
Northern blotting
Total RNA was harvested from infected podocytes after differentiation for 10 d. Twenty µg of RNA from each sample was resolved on 1.2% agarose/formaldehyde gel and transferred to a 0.45 µm membrane (Biodyne, Sarasota, FL, USA). This membrane was probed with a 1.75 kb StuI-StuI fragment of sdk-1 cDNA that had been labeled with P32. rRNA bands were analyzed to ensure equal loading of RNA.
Generating sdk-1 stable podocyte transfectants
Full-length sdk-1 cDNA was cloned into the BamHI and SmaI sites of pIRES2-eGFP vector (BD Biosciences) to generate pIRES2-sdk-1-eGFP. This vector allows for high expression of sdk-1 under the constitutively active cytomegalovirus (CMV) promoter and biscistronic expression of enhanced green fluorescent protein (eGFP). Proliferating podocytes were transfected with this sdk-1 expression vector or with native vector control using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Stably transfected cells were selected in serial passages using G418 antibiotic at 400 µg/ml, which is 100% lethal to untransfected cells. Clonal populations of pIRES2-sdk-1-eGFP and pIRES2-eGFP podocytes were selected by limiting dilution and on the basis of eGFP expression levels. The clone with the highest eGFP expression was used in experiments. Stably transfected cells were allowed to differentiate as described above, except in the presence of lower concentrations of G418 (200 µg/ml).
Western blotting
Podocytes were lysed in BugBuster (Novagen, San Diego, CA, USA) supplemented with complete protease inhibitor cocktail (Roche, Mannheim, Germany) and 1 µl/ml of nuclease (Novagen). To isolate membrane fractions, lysed samples were centrifuged at high speed for 30 min. Cellular lysates were removed and pellets were resuspended in 1% SDS and solubilized via sonication. The solubilized samples were separated by SDS-PAGE and transferred to Immobilon-P transfer membranes (Millipore, Bedford, MA, USA). Western blots were performed using 1:5000 dilution of polyclonal anti-sdk-1 antibody (10)
followed by 1:10000 dilution of horseradish peroxidase-labeled goat anti-rabbit IgG (Kirkegaard and Perry, Gaithersburg, MD, USA).
Immunocytochemistry
Podocytes were stably transfected with sdk-1 expression vector or vector alone as described above and then grown and differentiated on type-1 collagen-coated glass coverslips. Cells were fixed in 2% paraformaldehyde with 4% sucrose in PBS for 5 min followed by permeabilization with 0.3% Triton for 10 min. For actin staining, alexa fluor-594 labeled phalloidin (Molecular Probes, Eugene, OR, USA) was used per the manufacturer’s protocol. Staining for sdk-1 was done using anti-sdk-1 antibody (10)
at a dilution of 1:40. Primary antibodies were detected using Cy3-labeled goat anti-rabbit IgG (Kirkegaard and Perry).
Mixing studies
Fully differentiated pNL4–3:G/P-DsRed2 infected, VVCW-DsRed2 infected, or pIRES2-sdk-1-eGFP stably transfected podocytes were trypsinized and washed in Hanks’s balanced salt solution (HBSS) without calcium or magnesium. Cells were plated in HBSS at 0.5 x 104 cells/ml on 24-well plates precoated with BSA and then rotated on a gyratory shaker for 10 min. Aggregation was stopped with the addition of glutaraldehyde. Cells were then examined by immunofluorescence microscopy, and the total number of cell interactions was manually counted. To confirm data, mixing studies were repeated twice using separate batches of freshly infected cells. Results were analyzed for statistical significance using the unpaired, two-tailed t test and a P-value of <0.05 was considered significant.
Generation and characterization of shRNA constructs
The short hairpin RNA (shRNA) vector was constructed by annealing complementary 61-mer oligonucleotides containing the 19-nucleotide target sequence in both the sense and antisense orientation separated by a 9-nucleotide spacer. The 19 mer sequence (aagaccgcgtggtgattaa) targets a region corresponding to the sixth Ig motif of sdk-1 and is predicted to be specific only for sdk-1 as determined by basic local alignment search tool (BLAST) database searches. The annealed oligonucleotide was inserted under the control of the human H1 promoter in the self-inactivating pVIRHD/E lentiviral vector (17)
derived from VVPW/enhanced GFP (EGFP) (gift of Dr. Luca Gusella). The resulting lentiviral vector, which also expresses the reporter GFP from the constitutive murine phosphoglycerate kinase promoter, was packaged into VSV-G pseudotyped virions using the methods previously described. The pVIRHD/EsiLuc lentivector vector (17)
(gift of Dr. Luca Gusella) expressing shRNA to Luciferase was used as a control to ensure that any observed effects are due to expression of the sdk-1 shRNA sequence.
To characterize the shRNAs, 293T cells were cotransfected with both a sdk-1 expression construct (pcDNA3.1/sdk-1) and either sdk-1 or control shRNA constructs using Effectene transfection reagent (Qiagen, Valencia, CA, USA), according to the manufacturer’s protocol. Transfection was done with a 10x higher concentration of the corresponding shRNA construct relative to the sdk-1 expression construct. Forty-eight hours post-transfection, cells were lysed using radio-immuno-precipitation assay (RIPA) buffer (Santa Cruz Biotechnology, Santa Cruz, CA, USA) supplemented with 1x protease inhibitor cocktail (Roche). Western blotting was performed as described previously. Membranes were stripped and reprobed with an antiactin antibody (Sigma) to normalize for protein loading.
Dispase assays
A dispase-based dissociation assay (18
19
20)
was modified for our purposes and performed as follows. Wild-type and HIV-1 podocytes were plated sparsely on 6-well plates and allowed to differentiate for 7–10 d. Twenty-four to forty-eight hours after reaching 100% confluence, cells were washed with 1x PBS and then incubated for greater than 30 min with dispase enzyme (2.4 U/ml) (Roche) at 37°C. Cell monolayers were carefully scraped to separate any remaining adherent cells from the dish. The cell sheets were carefully transferred using a wide-tip transfer pipette to a conical tube. Tubes were then inverted on a rocking platform for 5 min. Subsequently, 200 µl of each sample was removed, trypsin was digested, and cell number was determined using a hemocytometer. Total volume of the cell aggregate suspension was adjusted such that the concentration of cells between the two samples was equal. Each aggregate mixture (500 µl) was then plated in a 6-well plate. The number of cell aggregates containing >10 adherent cells was determined by manually counting using a dissection microscope. The number of cellular aggregates was tabulated by an independent observer who was blinded to sample identity. To perform dispase assays using shRNA, HIV-1 podocytes were infected with pseudotyped virus encoding either shRNA against sdk-1 or control as described in "Viral transduction" prior to analysis. Unpaired t tests were used to compare the number of aggregates in each population during three experiments and a P-value of <0.05 was used to determine statistical significance.
Immunohistochemistry
Rabbit polyclonal antibodies against sdk-1 were generated against the intracellular peptide ESEASDSDYEEALPK. This epitope was chosen because of its predicted antigenicity and its specificity for sdk-1 and not sdk-2 as determined by sequence analysis and BLAST database searches. Peptide synthesis, innoculation, antiserum collection, and affinity purification were performed by Open Biosystems (Huntsville, AL, USA). To characterize antibody specificity, 293T cells were transfected with either a sdk-1 or sdk-2 expression vector or native vector alone (PcDNA3.1, Invitrogen). Western blotting was performed on cellular lysate extracted from transfected cells.
Paraffin-embedded kidney sections were prepared from HIV-1 transgenic mice (Tg26) with 3–4+ proteinuria on urine dipstick. Paraffin blocks of kidneys from CD2AP null mice (21)
and adriamycin treated mice (22
, 23)
were gifts from Dr. Erwin Bottinger and Dr. Ali Gharavi, respectively. After dewaxing and rehydrating the slides, endogenous peroxide was blocked by incubation with 0.3% H2O2 in methanol for 30 min. Slides were then boiled for 20 min in sodium citrate buffer (ph 8.0) (Zymed, Burlingame, CA, USA) followed by 3 cycles of 2 min periods of irradiation in a microwave oven. Endogenous biotin was then blocked using an avidin/biotin blocking kit (Vector Labs, Burlingame, CA, USA). Sections were incubated overnight with undiluted affinity purified anti-sdk-1 antibody at 4°C followed by incubation with biotinylated goat-anti-rabbit antibody (Vector) at a dilution of 1:100. Vectastain Elite ABC kit and AEC Substrate kits (Vector) were used per the manufacturer’s directions to produce the color reaction. Control slides using preimmune serum for anti-sdk-1 were performed in parallel to ensure specificity of the anti-sdk-1 antiserum.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
G/P-DsRed2 (Fig. 1
A), a gag/pol-deleted virus derived from the same proviral construct used to generate the Tg26 HIV-1 transgenic model, or VSV-VVCW-DsRed2, a control lentivirus. Because infection of podocytes in vitro by HIV-1 is inefficient, both viruses were pseudotyped with the vesicular stomatitis virus glycoprotein (VSV-G) and concentrated to maximize transduction efficiency. Sdk-1 was highly up-regulated in the HIV-1 infected podocytes compared to controls (Fig. 1B
). This demonstrates that HIV-1 directly induces sdk-1 expression and is consistent with our previous findings, which showed significant up-regulation of sdk-1 in conditionally immortalized HIV-1 transgenic podocytes derived from the Tg26 HIVAN mouse model (10)
.
|
Sdk-1 overexpressing podocytes grow in aggregates and have a disorganized actin cytoskeleton
We next generated two stably transfected podocyte cell lines, one carrying an sdk-1 expression vector (pIRES2-sdk1-GFP) and the other control podocyte cell line carrying the native vector alone (pIRES2-GFP), both in the absence of HIV-1. Only early passage podocytes were used, and they were allowed to differentiate at 37°C for at least 7 d prior to performing experiments. Podocytes transfected with pIRES2-sdk1-GFP expressed high levels of sdk-1 protein compared to control podocytes as demonstrated by Western blotting of membrane protein fractions (Fig. 2
A). Immunocytochemical analysis of the sdk-1 transfected podocytes demonstrated sdk-1 protein localization at the plasma membrane, primarily at the intercellular plasma membranes between sdk-1 expressing podocytes (Fig. 2B
).
|
Morphologically, the control podocytes were similar to wild-type podocytes, possessing multiple branches to form delicate focal contacts with adjacent cells (Fig. 2C
, left panel). The sdk-1 expressing podocytes, however, grew in large aggregates with a simplified cell shape, less intricate-appearing processes, and loss of focal contacts between adjacent cells (Fig. 2C
, right panel).
Since the complex shape of the podocyte foot process formation is dependent on an elaborate network of actin filaments and actin-associated proteins including synaptopodin and alpha actinin-4 (7)
, we studied the effect of sdk-1 expression on actin filament formation. When grown sparsely on glass coverslips, sdk-1 expressing podocytes (Fig. 3
B) have a disorganized centrifugal distribution of actin filaments compared with control transfected podocytes (Fig. 3A
). This centrifugal distribution of actin is strikingly similar to the pattern seen in HIV-1 transgenic podocytes derived from the Tg26 HIVAN mouse model (Fig. 3C
). These observed differences in actin cytoskeletal architecture were more prominent when podocytes were grown sparsely on glass coverslips compared with plastic dishes.
|
HIV-1 infection increases podocyte adhesion to sdk-1 expressing cells
To assess whether HIV-1 infected cells could stimulate sufficient sdk-1 to alter cellular adhesion properties, we infected wild-type podocytes with either HIV-1 (VSV-pNL4–3:G/P-DsRed2) or control lentivirus (VSV-VVCW-DsRed2). We also generated stably transfected sdk-1 podocytes expressing GFP as is described above. To quantify adhesion, we performed mixing assays using a modified protocol initially described by Hayashi et al. (12)
. Podocytes were trypsinized and single cell suspensions were mixed in calcium and magnesium-free buffer on 24-well plates. Because sdk-1 mediated adhesion is independent of divalent cations (12)
, the use of cation-free buffer decreases potential confounding effects of many other cation-dependent adhesion molecules (i.e., cadherins). After gentle mixing, cells were fixed and then examined by immunofluorescence microscopy to count the number of cell interactions.
Sdk-1 stably transfected podocytes formed numerous large, green clumps when analyzed alone in the mixing assay (data not shown). When an equal number of sdk-1 transfected podocytes (green) and control-infected podocytes (red) were mixed, the control-infected cells remained separate from sdk-1 podocyte aggregates (Fig. 4
A, left panel). However, when HIV-infected podocytes (red) were mixed with the sdk-1 expressing podocytes (green), a large number of cell aggregates contained both red and green cells (Fig. 4A
, right panel). The total number of red fluorescent cells participating in green aggregates was counted and found to be significantly greater for HIV-infected compared with control-infected podocytes (P<0.01) (Fig. 4B
).
|
Cell-cell adhesion is increased in HIV-1 podocytes
Podocyte intercellular adhesion was measured by releasing monolayers with dispase (18
19
20)
, a neutral protease that degrades collagen type IV and to a lesser degree collagen type I. HIV-1 transgenic and wild-type podocytes were plated sparsely on six-well plates precoated with type I collagen and allowed to differentiate until 100% confluent (>7 d). One day after reaching confluence, wild-type and HIV-1 transgenic podocyte cell sheets were incubated with dispase, released from the substrate, and briefly shaken for a fixed time. Differences in cell-cell adhesion were evident immediately after release of the monolayers even before mechanical dissociation (Fig. 5
A). HIV-1 expressing podocytes separated as large aggregates or sheets, whereas wild-type podocytes detached only as single cells or as smaller aggregates. Substrate-released wild-type podocytes were further dissociated when subjected to mechanical stress, whereas HIV-1 podocyte sheets remained in large aggregates (Fig. 5B
). The degree of dissociation was quantified by counting the number of aggregates containing 10 cells or greater. HIV-1 podocytes showed dramatically higher numbers of aggregates than did wild-type podocytes (P<0.005).
|
Sdk-1 expression contributes to podocyte–podocyte adhesive strength in HIV-1 podocytes
To determine whether sdk-1 expression has a role in the increased podocyte adhesion induced by HIV-1, we studied the effects of sdk-1 inhibition on adhesion of HIV-1 transgenic podocytes. We generated an shRNA lentiviral vector targeting the sixth Ig motif of sdk-1. This shRNA is highly effective at inhibiting sdk-1 expression when cotransfected into 293T cells with an sdk-1 expression construct (Fig. 6
A). A control shRNA vector targeting luciferase had no effect on sdk-1 expression. HIV-1 transgenic podocytes were then infected with pseudotyped virus encoding either sdk-1 or control shRNA. 96 h postinfection, cells were trypsinized and plated on collagen-coated six-well plates and incubated at 37°C to induce cellular differentiation. Immediately prior to performing the dispase assay, the confluent differentiated podocytes demonstrated equivalently high rates of transduction as determined by eGFP expression (Fig. 6B
). After separation from the culture flask using dispase, cell sheets were subjected to mechanical dissociation on a rocking platform. The number of HIV-1 podocyte aggregates (>10 adherent cells) was significantly reduced in cells infected with an shRNA construct specific for sdk-1 as compared to cells infected with a control shRNA (P<0.01) (Fig. 6C
). This result suggests a direct role for sdk-1 in contributing to increased podocyte adhesion in the setting of HIV-1 infection.
|
Sdk-1 is expressed in podocytes in the murine HIVAN model
We tested the specificity of the affinity purified anti-sdk-1 peptide antibody by Western blotting using cellular lysates from 293T cells transfected with either a sdk-1 or sdk-2 expression construct. The antibody specifically detected sdk-1 without cross-reactivity against sdk-2 (Fig. 7
A). Immunostaining using this antibody on kidney sections from nephrotic Tg26 HIVAN mice revealed sdk-1 expression in proliferating cells in a glomerular pseudocrescent (Fig. 7B
, bottom panel). Staining of a serial section of the same glomerulus using preimmune serum was negative (Fig. 7B
, top panel). Sdk-1 expression was also occasionally detected in podocytes in collapsed glomeruli in the HIVAN mouse model but was not detected in glomeruli in either genetic (CD2AP null mice) or toxin mediated (adriamycin nephropathy) forms of podocyte injury (Fig. 7C
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The disease occurs almost exclusively in Blacks and Hispanics; in fact, only sickle cell disease is more closely linked to Black race than is HIVAN (25)
. Moreover, in the HIV transgenic mouse model, presence and severity of renal disease is highly dependent on genetic background (26)
. This implies that host genetic pathways downstream of infection are critical to disease development.
Direct infection of podocytes by HIV-1 induces podocyte injury manifested by proliferation, dedifferentiation, podocyte rearrangement, and loss of an organized actin cytoskeleton, all of which culminate in the CG observed in HIVAN (1
2
3)
. Previously, we reported that sdk-1 is highly up-regulated in podocytes in response to HIV-1 expression (10)
. We have also shown that sdk-1 is a homophilic adhesion molecule and that this adhesion is mediated by the extracellular Ig motifs binding to each other (12)
.
In these studies, we demonstrate that HIV infection of wild-type podocytes causes significant up-regulation of sdk-1. We also show that HIV infection can alter the interaction of adjacent podocytes, inducing aggregation via up-regulation of sdk-1. Furthermore, the increased podocyte intercellular adhesion seen in HIV-1 transgenic podocytes was partially reversed by reducing the expression of sdk-1. In vivo, sdk-1 is expressed in proliferating podocytes in the Tg26 mouse model of HIVAN. Taken together, our data suggest that sdk-1 overexpression in podocytes in response to HIV-1 can alter the adhesive properties of adjacent podocytes and that this change may contribute to podocyte rearrangement that is characteristic of glomerular pseudocrescent formation.
Sdk-1 overexpression in podocytes in vitro causes podocytes to grow in large aggregates with a simplified cellular morphology and a disorganized actin cytoskeleton. This centrifugal distribution of actin is similar to what is seen in HIV-1 infected podocytes. Within podocyte aggregates, sdk-1 is expressed primarily at intercellular plasma membranes at sites of cellular adhesion. Sdk-1 contains a highly conserved PDZ binding domain at its extreme carboxyl-terminus (11)
. In podocytes, many critical slit diaphragm molecules including nephrin, JAM-4, Neph-1, and FAT contain PDZ binding domains (27
28
29
30)
. Like sdk-1, all of these proteins are transmembrane adhesion molecules, and nephrin, JAM-4, and Neph-1 are also members of the Ig superfamily. These proteins interact with cytoplasmic anchoring proteins, which serve as adapters to link them to the actin cytoskeleton. For example, Neph-1 and JAM-4 interact with PDZ domains on ZO-1 and MAGI-1, respectively (28
, 31)
. MAGI-1 also interacts with the actin binding proteins 
actinin-4 and synaptopodin thus connecting JAM-4 to the actin cytoskeleton (32)
. ZO-1 similarly links Neph-1 to the actin cytoskeleton via an interaction with catenin (33)
. Nephrin interacts with a large complex of proteins including podocin, CD2AP, Neph-1, MAGI-1, and Nck adapter protein (27
, 34
35
36)
. This intricate protein complex allows for multiple links between nephrin and the actin cytoskeleton. Mutations in many of the proteins in these complexes lead to rearrangement of the actin cytoskeleton, foot process effacement, and renal failure. We suspect that sdk-1 also interacts indirectly with the actin cytoskeleton via its PDZ binding domain and unidentified cytoplasmic linker proteins. Future experiments should help to elucidate this pathway.
Sdk-1 and sdk-2 are highly expressed in many organs during organogenesis and play a critical role in mediating cell adhesion during pattern formation. During neuronal development, sidekicks function as neural guidance molecules, directing growing axons to specific synapses (11)
. Sdk-1 and sdk-2 localize specifically to the synaptic cleft in nonoverlapping neurons such that a sdk-1 expressing presynaptic neuron preferentially forms a synapse with a postsynaptic neuron also expressing sdk-1; likewise, presynaptic cells that express sdk-2 target sdk-2 positive laminae (11)
. Adhesion by sidekick proteins across the synapse is important to stabilize this developing neuronal circuit. In Drosophila, a mutation in the sidekick gene resulted in the development of extra photoreceptor cells in the eye (37)
, consistent with sidekick’s participation in cell-cell interaction during eye development.
In glomerular pseudocrescents, dedifferentiated podocytes express lower levels of differentiation markers such as synaptopodin and WT-1 (2
, 38)
and often reacquire genes that are usually expressed only during development (i.e., cytokeratin and pax-2) (38
39
40
41)
. Moreover, these podocytes may express genes that are not normally ever expressed by the podocyte (i.e., desmin or CD68) (38
, 39
, 42)
. Interestingly, sdk-1 is highly expressed in the ureteric bud (10)
, not in podocyte precursors, during nephrogenesis. Therefore, the expression of sdk-1 by the podocyte in HIVAN appears to involve de novo expression of a predominantly developmentally expressed gene. This implies that HIV infection induces both dedifferentiation and dysregulation of the podocyte.
The podocyte phenotype in the CG is unique in that dedifferentiated podocytes reenter the cell cycle and divide. This is in stark contrast to most other forms of proteinuric renal disease where podocyte depletion is a critical factor in mediating disease progression. Our findings suggest that sdk-1 is expressed in podocytes in HIVAN but is not expressed in podocytes in disease models characterized by decreased podocyte number and podocyte apoptosis. This implies that sdk-1 is not a pathogenic factor in all forms of podocyte injury but is specific to the proliferative phenotype present in CG.
Interestingly, inhibition of sdk-1 in HIV-1 transgenic podocytes did not significantly affect cellular proliferation, anchorage independent growth, or organization of the actin cytoskeleton. The pathways responsible for these changes are complex, requiring multiple upstream signaling events. Further work is required to elucidate the pathways responsible for sdk-1 up-regulation by HIV-1. HIV-1 nef is a likely candidate since a single gene expression construct carrying this HIV gene can induce podocyte dedifferentiation and proliferation both in vitro (16)
and in mouse models (43)
. Vpr may also be important for HIVAN pathogenesis, although its role is less clear. One of the most intriguing observations is that only sdk-1, and not sdk-2, is up-regulated by HIV-1 infection. Thus, better characterizing sidekick regulation is an important goal in understanding both HIVAN pathogenesis and normal embryogenesis.
|
ACKNOWLEDGMENTS |
|---|
Received for publication October 17, 2006. Accepted for publication December 25, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
