|Home | About | Journals | Submit | Contact Us | Français|
Alport glomerular disease is associated with dysregulation of pro-inflammatory cytokines and matrix metalloproteinases, promoting progressive glomerulonephritis. Changes in composition and structure of Alport GBM resulting from mutations in type IV collagen genes likely alter cell adhesion and cell signaling. Enhanced biomechanical strain on the capillary tuft, resulting from a thinner and less crosslinked GBM may be a source of insult which contributes to gene dysregulation. To test this we subjected cultured podocytes to cyclic biomechanical strain. We observed robust induction of MMP-3, −9, −10, and −14, but not MMP-2 or MMP-12. IL-6 was induced by biomechanical strain, and neutralizing antibodies against IL-6 attenuated induction of MMP-3 and MMP-10. Alport mice given L-NAME salts, which resulted in a significant rise in systolic blood pressure, showed Induction of MMP-3, MMP-10, and IL-6 in glomeruli relative to normotensive Alport mice. Hypertensive Alport mice also had elevated proteinuria, and more advanced GBM disease histologically and ultrastucturally. Collectively these data suggest MMP and cytokine dysregulation may constitute a maladaptive response to biomechanical strain in Alport podocytes, and that this response may contribute to the mechanism of glomerular disease initiation and progression.
The glomerular basement membrane (GBM) in Alport syndrome is irregularly thickened and thinned, with a multilaminar or “basketweave” appearance that is unique to the disease and a definitive diagnostic test for Alport syndrome. Immunostaining shows extracellular matrix deposits in the thickened regions.(1) More recently it was shown that the thickened regions are more permeable to injected ferritin than the non-thickened regions of the GBM. (2) This property is consistent with a partially degraded matrix network, suggesting proteolytic damage may contribute to focal thickening of the Alport GBM. The type IV collagen network in the Alport GBM is comprised entirely of α1(IV) and α2(IV) chains. This network contains fewer interchain crosslinks than wild type GBM (3), and is more susceptible to proteolytic degradation by endogenously expressed matrix metalloproteinases. (4) Glomerular MMPs have been shown to be induced in glomeruli of Alport mice, and blocking the activity of specific MMPs has been shown to ameliorate the progression of glomerular pathology. (4–6)
The delayed onset in Alport syndrome suggests that the glomerulus is subjected to chronic insult that eventually results in activation of the disease process. There are at least two potential sources of insult. One is the persistence of the embryonic-like basement membrane and its failure to program podocyte cell signaling appropriately. This could result from a GBM lacking the α3/α4/α5 collagen network for binding and activating the αVβ3 collagen receptor, for example. (7) A second explanation is the distinct physical properties (thinner, less cross-linked) of the Alport GBM, which is likely more elastic, subjecting podocytes to elevated biomechanical strain even under normal glomerular blood pressure. As the disease progresses and nephron mass is lost, glomerular hypertension develops, further exacerbating the biomechanical strain and the effector functions influenced by it.
Cell culture systems for studying the influence of biomechanical strain have relied on specialized instruments where cells are plated on elastic membranes which can be stretched precisely using computer assisted programs. These systems have been used extensively in studies of skin, cardiomyocytes, and bone cells, where biomechanical strain would be expected to have important biological consequences. The varied studies have provided a wealth of information regarding how cell adhesion, cytoskeletal dynamics, and cell signaling machinery respond to strained adhesive interactions between cells and extracellular matrix. (8) More recently glomerular podocytes have been studied as a means of understanding how glomerular hypertension contributes to progressive glomerular diseases. These studies noted influences of mechanical strain on maladaptive responses such as activation of the tissue angiotensin system (9) and induction of Secreted Protein Acidic and Rich in Cysteine (SPARC), (10) changes that promote glomeruloschlerosis. Conversely, adaptive responses of podocytes to mechanical strain were also documented, where induction of osteopontin enhances cytoskeletal reorganization and protects against strain-induced podocyte loss. (11) The relevance of these findings was often confirmed in vivo using hypertensive mouse or rat models. (9,10)
In this study, we used in vitro and in vivo approaches to examine a potential role for biomechanical strain in regulating expression of matrix metalloproteinases known to contribute to GBM destruction in Alport mouse models. The results suggest that biomechanical strain leads to maladaptive gene regulation in Alport mice, which may constitute an important mechanism for GBM disease initiation and progression.
α3β1 integrin binding to the α5 chain of laminin 521 is a principal adhesive interaction between podocytes and the GBM. Earlier studies of biomechanical strain responses in podocytes were conducted on plates pre-coated with collagen I, which is not a component of GBM. (9–11) To determine whether matrix influences expression of MMP-3, MMP-10, and IL-6 in cultured podocytes, differentiated wild type podocytes were cultured on plastic plates pre-coated with either collagen I, placental laminin, or a coating of collagen I followed by a second coating with placental laminin (Flexcell plates are pre-coated with collagen-1). After 48 hours cells were harvested and basal levels of expression of MMP-3, MMP-10, and IL-6 were examined by real time RT-PCR. The results in Figure 1 show that cells cultured on placental laminin had lower basal levels of mRNA expression for MMP-3, MMP-10, and IL-6 compared to cells cultured on collagen I.
We employed the Flexcell system to subject podocytes to a 10% cyclical stretch for 15 hours on elastic wells coated with collagen I/ placental laminin. RNA was analyzed by conventional RT-PCR for MMP-2, MMP-3, MMP-9, MMP-10, MMP-12, and MMP-14. The results in Figure 2 show MMP-2 and MMP-12 mRNA expression are not influenced by mechanical stretch. In contrast, MMP-3, −9, −10, and 14 are robustly induced by stretch. Cyclophilin and GAPDH were used as controls and showed no response and consistent loading.
Interleukin-6 has been implicated in the regulation of stromelysins in other systems, (12,13) and IL-6 is induced early in Alport glomeruli (unpublished observation). Based on this, we focused our efforts on characterizing the stretch-mediated response of the stromelysins (MMP-3 and MMP-10) and IL-6. We performed mechanical stretch response experiments and analyzed RNA by real time RT-PCR using Taqman probes for stromelysins 1 and 2 (MMP-3 and MMP-10) as well as IL-6 normalized to GAPDH. Figure 3A shows robust stretch-mediated induction of MMP-3, MMP-10, and IL-6 in podocytes.
IL-6 might contribute to induction of MMP-3 and MMP-10 mRNAs through an autocrine loop mechanism. To test this we added IL-6 neutralizing antibodies to the media prior to applying mechanical strain and found it partially blocked induction of both MMP-3 and MMP-10 mRNAs (Figure 3B). To further test this, we treated differentiated podocytes cultured on collagen/placental laminin with recombinant IL-6. As shown in Figure 3C, mRNAs encoding both MMP-3 and MMP-10 (but not IL-6) are induced by recombinant IL-6.
The loss of adhesion resulting from biomechanical stresses is known to effect changes in the actin cytoskeleton. It is well established that α-actinin-4 interacts with synaptopodin. (14) Synaptopodin functions in regulating the actin bundling activity of α-actinin-4, a key regulator of actin organization in the podocyte. (15,16) We examined the cytoskeletal organization in wild type podocytes cultured on laminin by dual immu-nostaining with antibodies specific for either α-actinin-4 or synaptopodin, or with phalloidin to stain the actin cytoskeleton directly. Figure 4 (A–D) shows that wild type cells have fine filamentous staining for actin (4D) with discrete regions of co-localization for α-actinin-4 and synaptopodin at the peripheral processes of the cells (4A–C). Following 15 hours of 10% cyclic biomechanical stretch disruption of filamentous synaptopodin and α-actinin-4 co-localization (Figure 4E–G) and disorganized actin filaments (H) were observed, suggesting strain-mediated disruption of the actin cytoskeleton.
To examine the effects of mechanical strain in vivo, we induced a state of hypertension by administering N-nitro-L-arginine methyl ester salts (L-NAME), a method widely employed in both rat and mouse models. (17,18) We used the CODA-2 (Kent Scientific) non invasive tail cuff system for measuring blood pressure. The C57 Bl/6 X-linked Alport mouse model was employed because the slower disease progression in this background (19,20) allowed blood pressure measurements to be made in fully grown mice, providing better consistency across longitudinal measures. Baseline measures of systolic blood pressure were around 135 +/− 12 mm of Hg, consistent with published results using intra-arterial methodologies. (21) L-NAME treated mice showed consistently increased systolic blood pressure to 170 +/− 16 mm of Hg in both wild type and Alport mice. Figure 5 shows the results of a longitudinal study. The salt treated animals also showed higher systolic blood pressures and urinary albumin levels (Figure 5 panel I). In Figure 5 panel II we express the data showing albuminuria as a function of systolic blood pressure. The correlation of higher blood pressure corresponding to higher albuminuria is independent of the age of the mice. When all of the data for all ages is plotted together, a highly significant correlation coefficient shows a statistically significant relationship between systolic blood pressure and albuminuria. It is notable that L-NAME-treated mice with lower blood pressure also showed lower albuminuria.
Elevated albuminuria likely indicates more advanced glomerular pathology in salt-treated hypertensive Alport mice. To test this, we performed immunofluorescence immunostaining of 10-week old wild type, Alport, and salt-treated Alport (treated from 6 to 10 weeks of age) using antibodies specific for fibronectin or the α2 chain of laminin. Fibronectin staining, which is restricted to the mesangial matrix, shows significantly expanded mesangium in salt treated Alport mice relative to Alport mice given no salt (Figure 6, panel A–C). Laminin α2, which is normally restricted to the mesangial matrix in the glomerulus, is known to accumulate in the GBM of Alport mice. (22,23) In salt treated Alport mice, we observed massive accumulation of laminin α2 in nearly every portion of every capillary loop (Figure 6, Panel F), whereas in no salt Alport animals, there was punctate GBM immunostaining for laminin α2 (panel E), consistent with less advanced glomerular disease compared to salt treated mice. These findings were confirmed by transmission electron microscopy (Figure 6). Salt treated Alport mice invariably showed extensive glomerular basement membrane damage with irregular thickening and multilamination, whereas no-salt treated animals showed much less damage. None of these effects were observed in salt-treated wild type mice (data not shown).
Next we examined the in vivo relevance of biomechanical strain-mediated induction of MMP-3, MMP-10, and IL-6 in cultured podocytes. We used the 129 Sv strain for this experiment because of the rapid disease progression compared to the C57 Bl/6 background, (20) resulting in consistent molecular changes related to glomerular disease. The kinetics of MMP-3, MMP-10 and IL-6 mRNA induction in Alport mice on the 129 Sv background (data not shown) predicted the window between 4 and 5 weeks of age would be most informative. We treated wild type mice or Alport mice with salt or no salt from 4 to 5 weeks of age. Figure 7 I shows that, like the C57Bl/6 X-linked Alport mice, the salt treated 129 Sv autosomal Alport mice show elevated systolic blood pressure and elevated levels of proteinuria relative to animals given no salt. Wild type mice occasionally showed microalbuminuria in response to salt treatment as well, suggesting some of the elevated proteinuria might be due to direct effects of hypertension on glomerular filtration. Glomeruli were isolated from these mice and RNA analyzed by real time PCR for MMP-3, −10, −12, and IL-6 mRNAs. MMP-12 was analyzed because it is known to be induced as a function of Alport glomerular disease progression, (5) but was not found to be induced by biomechanical strain in our cell culture studies (Figure 2). The results in Figure 7 II show that salt treated Alport mice have significantly elevated levels of MMP-3, MMP-10, and IL-6 mRNA compared to no-salt Alport mice. MMP-12 was elevated in both groups, but not significantly different in salt versus no salt groups. This is consistent with biomechanical stretch experiments in cultured podocytes (Figure 2). In wild type mice, salt treatment did not result in significant increases on any of the transcripts analyzed, suggesting that the biomechanical influence of hypertension is more pronounced in the Alport mice than wild type mice. We did not observe a statistically significance in lifespan in salt treated versus no-salt Alport mice, likely owing to the accelerated pace of renal disease progression on the 129 Sv/J background.
The glomerular podocyte foot processes form an adhesive interface basally with the GBM via integrin/cell matrix interactions and laterally with the complex of proteins comprising the slit diaphragm. These adhesions are integrated with the cytoskeleton, and signal through a variety of mechanisms that likely participate in functional maintenance of these cellular and extracellular structures. In diseases that result in glomerular hypertension, the adhesive interface will be under abnormal biomechanical strain, which could conceivably alter the receptor/GBM interactions and signaling. Others have demonstrated that biomechanical strain of podocytes in culture results in activation of adaptive (protective) responses such as changes in the actin cytoskeleton, (24,25) as well as maladaptive (destructive) responses such as the induction of SPARC (10) or the induction of osteopontin. (11,26)
A common cause of Glomerular hypertension is the loss of nephron mass resulting from progressive glomeruloschlerosis. (27) Thus the problem of hypertensive influence on glomerular homeostasis applies, in most instances, to later phases of glomerular disease progression. Alport syndrome presents a somewhat unique twist to this paradigm. Because the Alport GBM is thinner (owing to the absence of the type IV collagen α3/α4/α5 network) and contains fewer interchain disulfide crosslinks, (3) it is likely to be more elastic than wild type GBM. Under normotensive conditions, therefore, the influences of biomechanical strain may be present prior to nephron loss. These influences would be further exacerbated by nephron loss and glomerular hypertension as a function of progressive glomeruloschlerosis. Thus the influences of biomechanical strain on Alport podocytes might be viewed as a chronic insult that contributes to both the initiation and progression of glomerular disease.
In earlier work, we and others have shown that a number of matrix metalloproteinases (MMPs) are induced in glomeruli from Alport mice, and that the presence of elevated MMP activity is linked to the deterioration of glomerular function in these mice. (4–6) In this paper we show that many of the MMPs (MMP-3, −9, −10, and −14) are induced in podocytes when cells are exposed to biomechanical strain. There were exceptions, however (MMP-2 and −12). Affymetrix data from 129 sv Alport mice at an early disease state (4 weeks of age) compared to wild type littermates showed IL-6 was significantly induced (unpublished observation). Given that earlier studies in other systems showed IL-6 associated with stromelysin induction, (28,29) and that IL-6 can be induced by biomechanical strain, (30,31) we surmised that in vitro biomechanical strain studies might involve induction of genes by both direct (adhesion dependent) and indirect (autocrine activation by induced cytokine activity) mechanisms. This is indeed what we observed, since IL-6 neutralizing antibodies partially ameliorate strain-mediated induction of MMP-3 and MMP-10 (Figure 3B) and since recombinant IL-6 induces MMP-3 and MMP-10 in resting differentiated podocyte cultures (Figure 3C).
In vivo correlates show a direct association of proteinuria and systolic blood pressure, and exacerbated disease progression (based on matrix accumulation In the GBM and GBM dysmorphology) in hypertensive Alport mice relative to non-hypertensive Alport mice. IL-6, MMP-3, and MMP-10 were markedly induced in glomeruli from hypertensive Alport mice relative to non-hypertensive Alport mice, while MMP-12 was unaffected in the two groups. This is consistent with the biomechanical strain studies using cultured podocytes, suggesting the biomechanical strain platform may provide in vivo relevance in determining the specific cellular mechanism underlying strain mediated induction in these cells.
The connection between biomechanical strain, adhesive interactions, cell signaling, and matrix metalloproteinase gene regulation has been demonstrated previously in both fibrochondrocytes and vascular endothelial cells. (32,33) This mechanism has not been explored in glomerular podocytes and renal glomeruli, where the connection would be particularly relevant to Alport glomerular pathology, given the established link between MMP activity and glomerular disease progression. (3–6) There are compelling clues in the literature. These include conditional knockout mutations for integrin α3, integrin β1, and the tetraspanin CD151 (known to bind to integrin α3 resulting in the high affinity ectodomain conformer of α3β1 integrin), which directly affect podocyte cell adhesion to laminin 521 in the GBM, (34–38) Integrin linked kinase, which influences cytoskeletal organization and cell signaling through the adhesive interactions of α3β1 integrin and its GBM ligand, (39,40) and α-actininin-4, which influences actin cytoskeletal dynamics and ILK/nephrin interactions as part of the adhesion interactome complex. (41) All of these mutations result in a common GBM phenotype including multilaminated irregular thickening associated with podocyte foot process effacement (the Alport GBM phenotype). Collectively, these data support a model where disruption of the adhesive interactions between α3β1 integrin and laminin 521 disrupts the actin cytoskeleton and activates adhesion-dependent signaling mechanisms (ILK and/or FAK) potentially resulting in maladaptive responses.
These maladaptive responses may cause focal degradation of the GBM, resulting in enhanced permeability (proteinuria) and progression of glomeruloschlerosis. Nephron loss further exacerbates glomerular hypertension, resulting in progressive acceleration to end stage renal failure. This hypothesis is supported by the observation that ramipril therapy, which lowers blood pressure (and hence should decrease biomechanical strain) significantly slows the rate of glomerular disease progression in Alport mice, and increases lifespan by more than 100%. (42) It will be of interest to determine whether these components of the adhesive interface and signal transduction machinery are mechanistically linked to strain-mediated dysregulation of genes involved in the pathobiology of Alport syndrome.
Alport mice on the 129 Sv background were developed here (1) and bred in house. Littermates were used as controls. C57 Bl/6 Alport mice were described previously (19) and were obtained from the Jackson Laboratories (Bar Harbor, Maine). These animals were bred as homozygotes, and age matched C57 Bl/6 mice from the Jackson Laboratories repository were used as controls. All experiments were performed under an approved IACUC protocol and animals were maintained in accordance to USDA standards. Every effort was made to minimize stress and discomfort.
Wild type podocyte cultures were established and cultured as described previously. (5) Differentiated podocytes were qualified by immunofluorescence using primary antibodies against CD2AP (H-290), WT-1 (C-19), VWF (H-300), α-actinin-4 (N.17) all from Santa Cruz Biotechnology, (Santa Cruz, CA.), synaptopodin (SE-19, Sigma), and Nephrin (extracellular domain) a gift from R. Kalluri. Podocyte cultures are requalified frequently to assure consistency of the phenotype. (43)
Untreated 6-well BIOFLEX plates (Flex International Corp., Hillsborough, NC) were coated as follows: 3 mg/ml rat tail collagen type I (BD Biosciences, Bedford, MA.) was diluted to 50 µg/ml in filter-sterilized 0.02 N HOAc. 1.25 mls of the collagen I solution was incubated on each well for 1 hour at room temperature. The wells were then aspirated, gently rinsed two times with DPBS (-Ca++, -Mg++), briefly air dried, and either used or stored at 4°C. 100 µg of Laminin from human placenta (Sigma, St. Louis, MO.) was diluted in 12 mls of DPBS (-Ca++, -Mg++) and 1 ml placed in each of six collagen I-coated well at 37°C for 2 hours. The wells were gently rinsed 3 times with DPBS and used or stored at 4°C.
Conditionally immortalized wild-type murine podocytes were grown under non-permissive conditions (39°C, no γ-interferon) for 10 days and plated onto Bioflex 6-well plates coated with type I collagen/ placental laminin. Cells were plated in 5% FCS containing media at densities that result in 20–40 % confluency. 0.5% FCS media was placed on the cells the next day. 48 hours later the media were changed and the podocytes were exposed to a regimen of 60 cycles of stretch and relaxation per minute at a frequency of 0.5 Hz with a maximum amplitude of 10% radial surface elongation. The Flexercell Strain Unit FX4000 (Flexcell International Corp., Hillsborough, NC.) was used to induce stretch/relaxation for 15 hours according to manufacturer’s directions. Cells grown identically, but not exposed to stretch, were used as controls.
Sample preparation and Real-time RT-PCR was performed on a TaqMan ABI 7000 Sequence Detection System (Applied Biosystems). RT-PCR primers and conditions were as previously described. (5,6) TaqMan reagents for MMP-10 (Mm01168400_g1), IL-6 (Mm99999064_m1) and rodent GAPDH containing both the primers and FAM-probes and VIC-probe were purchased from Applied Biosystems.
Semi-quantitative analysis of mRNA transcripts for MMPs (−2,−3, −9, −10, −12, −14 and IL-6) was performed by RT-PCR as described previously. (44) Oligonucleotide primer pairs were as follows: for GAPDH, 5’-GGT GAA GGT CGG AGT CAA CGG ATT TGG TCG-3’ and 5’-GGA TCT CGC TCC TGG AAG ATG GTG ATG GG-3’ (236 bp target size); for MMP-2, 5’- CCT GAT GTC CAG CAA GTA GAT GC-3’ and 5’-TTA AGG TGG TGC AGG TAT CTG G-3’ (760 bp target size); for MMP-3, 5’-TGT ACC CAG TCT ACA AGT CCT CCA-3’ and 5’-CTG CGA AGA TCC ACT GAA GAA GTA G-3’ (685 bp target size); for MMP-9, 5’-TTC TCT GGA CGT CAA ATG TGG-3’ and 5’-CAA AGA AGG AGC CCT AGT TCA AGG-3’ (414 bp target size); for MMP-10, 5’- CTT CAG ACT TAG ATG CTG CCT A - 3’ and 5’-CAG GAG CAA CTA ATC ATT CTG AC −3’ (533 bp target size) ; for MMP-12, 5’- AAG CAA CTG GGC AAC TGG ACA ACT C-3’ and 5’- GAG ATA CAA AGA AAT GAT GGA TGC −3’ (631 bp target size); and for MMP-14, 5’-GTG ATG GAT GGA TAC CCA ATG C-3’ and 5’- GAA CGC TGG CAG TAA AGC AGT C- 3’ and for IL-6, 5’- CTT ATC TGT TAG GAG AGC ATT GG −3’ and 5’- GAG ATA CAA AGA AAT GAT GGA TGC −3’ (389 bp target size). All PCR products were confirmed by DNA sequencing.
Urine samples were analyzed for albumin by PAGE gel analysis followed by staining with Coomassie blue and scanning densitometry. Data was normalized to urinary creatinine which was measured using the QuantiChrom Creatinine assay kit (DICT-500) according to the manufacturers instructions (BioAssay Systems, Hayward, CA). Samples from 5 mice per group (salt treated) and 3 mice per group (non-salt treated) were run in triplicate, and the mean values for each measurement plotted.
We administered N-nitro-L-arginine methyl ester salts (L-NAME) at 50 µg/g body weight per day in the drinking water. Blood pressure was monitored using the CODA-2 (Kent Scientific, Torrington, CT) non invasive tail cuff system. For C57 Bl/6 Alport mice, 5 animals were given L-NAME salt, and 3 were given normal water starting at 5 weeks of age. Urine samples were collected at weekly intervals along with blood pressure measurements and analyzed in triplicate for proteinuria. Animals were euthanized at 10 weeks of age for analysis of glomerular pathology. For 129 Sv/J mice, we treated 3 wild type and 3 Alport mice with salt or no salt from 4 to 5 weeks of age. Glomeruli were isolated using magnetic beads as previously described. (5)
Stretch and non-stretch wild type podocytes were immunostained for α-actinin-4 (N-17) and synaptopodin (SE-19), Santa Cruz Biotechnology (Santa Cruz, CA), or FITC-phalloidin (Molecular Probes) as previously described. (5) Images were captured with a Zeiss AxioPlan 2IF MOT microscope interfaced with a LSM510 META confocal imaging system.
Data are expressed as mean ± SD. Differences between means were tested for significance using Student’s t-test. Differences were considered significant at the level of P < 0.05.
Supported by NIH R01 DK55000 to DC, RO1 DK49461 to SS.
We are grateful to John (skip) Kennedy for assistance I preparing figures, Walt Jesteadt, Ph.D. for help with data analysis, and Tom Barger for assistance with transmission electron microscopic analysis. Confocal microscopy was conducted at the Integrative Biological Imaging Facility at Creighton University, Omaha, NE. This facility, supported by the C.U. Medical School, was constructed with support from C06 Grant RR17417-01 from the NCRR, NIH.
Daniel T. Meehan, Boys Town National Research Hospital, Omaha Nebraska.
Duane Delimont, Boys Town National Research Hospital, Omaha Nebraska.
Linda Cheung, Boys Town National Research Hospital, Omaha Nebraska.
Marisa Zallocchi, Boys Town National Research Hospital, Omaha Nebraska.
Steven C. Sansom, University of Nebraska Medical Center, Department of Cellular and Integrative Physiology, Omaha, NE.
J. David Holzclaw, University of Nebraska Medical Center, Department of Cellular and Integrative Physiology, Omaha, NE.
Velidi Rao, Boys Town National Research Hospital, Omaha Nebraska.
Dominic Cosgrove, Boys Town National Research Hospital, Omaha Nebraska.