|Home | About | Journals | Submit | Contact Us | Français|
The application of molecular techniques to characterize clinical kidney biopsies has the potential to provide insights into glomerular diseases that cannot be revealed by traditional renal pathology. The present work is a proof-of-concept approach to test whether proteomic analysis of glomeruli isolated from clinical biopsies by laser capture microdissection can provide unique information regarding differentially-expressed proteins relevant to disease pathogenesis.
The proteomes of glomeruli isolated by laser capture microdissection from biopsies of normal kidneys (living-related donor kidneys) were compared to those from patients with diabetic nephropathy, lupus nephritis, and fibronectin glomerulopathy. Glomerular proteins were extracted, trypsin digested and subjected to liquid chromatography-tandem mass spectrometry for identification and quantitation.
Relative to normal glomeruli, all disease-associated glomeruli showed an increased presence of complement components, a marked decline in podocyte-associated proteins, and a decrease in proteins associated with cellular metabolism. Additionally, fibronectin glomerulopathy glomeruli differed from all the other glomeruli because of a significant accumulation of fibronectin and fibulin.
This study demonstrates that our method acquires reproducible and quantitative proteomic information from laser capture microdissection isolates that can be used to characterize the molecular features of glomerular diseases.
The application of molecular techniques such as genomics, proteomics, or metabolomics to clinical kidney biopsies has the potential of providing significantly more new and unique information regarding glomerular diseases than traditional renal pathology using microscopy. New insights could be relevant to glomerular disease pathogenesis, novel therapeutic targets, or inter-patient variability. While there have been several reports on the use of laser capture microdissection in the analysis of glomerular diseases, these investigations mainly focused on changes in gene expression in specific kidney compartments 1-6. A small number of studies have tested the feasibility of proteomic analysis of isolated glomeruli in animal models 7, 8 or human tissue 9-11. Only one investigation showed that laser capture microdissection glomeruli from a clinical kidney biopsy specimen yielded sufficient material to create a robust proteome. However, none of the proteins were further characterized 10.
The present work was undertaken to more clearly establish whether proteomic analysis of laser capture microdissection glomeruli from clinical kidney biopsies can be used to identify differentially-expressed glomerular proteins that may inform disease pathogenesis. For this proof-of-concept study we feature an examination of laser capture microdissection glomeruli from patients with diabetic nephropathy, lupus nephritis, and fibronectin glomerulopathy. Fibronectin glomerulopathy is a rare disease with large, proteinaceous glomerular deposits, which to date have only been assessed by traditional immunohistochemistry. Thus it is an ideal clinical model to evaluate the extent to which analysis by proteomics can exceed that of immunohistologic methods.
Fibronectin glomerulopathy is a familial glomerulopathy characterized clinically by proteinuria, microscopic hematuria, hypertension and progressive renal dysfunction which can lead to end stage renal disease 12. An autosomal dominant inheritance pattern with age-related penetrance affecting both genders has been demonstrated 13, 14. Histologically, fibronectin glomerulopathy is characterized by lobular glomerular enlargement, obliteration of capillary loops and mesangial expansion due to accumulation of acellular material. These deposits are large homogenous to finely granular electron-dense masses, usually with no distinctive fibrillary substructure. By immunohistochemical staining, the glomerular material stains for fibronectin 15 and only weakly or not at all for IgG, complement products, or other matrix components like laminin and collagen type IV 15. Using antibodies to plasma-derived and cell-derived fibronectin, it has been reported that the fibronectin in the deposits is mainly plasma-derived 15, 16. Therefore it has been postulated that fibronectin glomerulopathy is a primary mesangiopathic glomerulopathy that is caused by abnormal plasma fibronectin deposition, and subsequent disruption of the glomerular architecture and filtration barrier leading to glomerular proteinuria and loss of glomerular filtration rate.
However, fibronectin deposition is not specific for fibronectin glomerulopathy. Intra-glomerular fibronectin can be found in other kidney diseases, such as diabetic nephropathy and lupus nephritis 17-19. These findings suggested that comparing glomerular protein expression of fibronectin glomerulopathy to that of normal kidneys, and kidneys with diabetic nephropathy and lupus nephritis would be a good test for demonstrating the capability of using laser capture microdissection in combination with liquid chromatography-tandem mass spectrometry to characterize kidney diseases in greater detail. The approach used here revealed anticipated differences as well as unexpected similarities between disease states, and also highlighted potential pathogenic mechanisms not previously known or considered for these diseases.
Baseline allograft biopsies from living donor renal allograft recipients (n=3) were used as normal controls. Glomerular disease biopsies included diabetic nephropathy (n=2), lupus nephritis proliferative Class IV (n=3) and pure membranous Class V (n=2), and two patients from a family known to have fibronectin glomerulopathy. The kidney biopsy findings of this family have been previously reported 20. One biopsy was from an affected native kidney, while the second biopsy was from a living unrelated donor renal allograft that developed recurrent disease three years post-transplant. The study was approved by the Ohio State University Institutional Review Board (Protocol number 2009H0179).
Laser capture microdissection was performed using the laser microdissection system from PALM Technologies (Carl Zeiss MicroImaging GmbH, Munchen, Germany) containing a PALM MicroBeam and RoboStage for high-throughput sample collection and PALM RoboMover (PALM Robo software, Version 2.2). Typical settings used for laser cutting were UV-Energy of 52 and UV-Focus of 79-81. laser capture microdissection was done using snap frozen tissue embedded in optimum cutting temperature medium at the time of the biopsy. The frozen tissue blocks were cut into 10 μm sections, mounted on a thermoplastic (polyethylenenapthalate) covered glass slide (Carl Zeiss MicroImaging) and kept at −80°C until use. When needed, sections were briefly thawed and immediately fixed in ice cold 70% ethanol and rinsed in deionized water for 1 minute. After the water rinse the wet sections were stained briefly (5-10 s) with hematoxylin, immediately rinsed, and then dehydrated in graded ethanol. The dehydrated slides were allowed to air dry for immediate capture on the PALM technology microdissection system.
Sections were cut and captured under a 10× ocular lens. Fifteen 10 micron sections from each clinical biopsy case were utilized, and each section contained an average of ten glomeruli. The glomeruli were isolated by dissecting around Bowman's capsule. For each biopsy an average of 75 glomerular cross-sections, equivalent to about 7000 cells were captured. Cut elements were catapulted into 25 μl of 0.5% Rapigest (Waters Corporation, MA, USA), resuspended in 50 mM NH3HCO3, collected in a 0.2 ml tube, and held at −80°C until protein retrieval.
For protein retrieval, samples that had been stored at −80°C were thawed briefly and sonicated (in the original 0.2 ml tube) on the low power setting for a total instrument time of 60 s with a 3 s on 3 s off time interval using a Diagenode Bioruptor (model UCD-200, NJ, USA.). The samples were placed in a boiling water bath for 5 min and cooled to room temperature. Trypsin was added in a ratio of 1:30 trypsin:protein assuming ~2 μg retrieved protein/10,000 isolated cells. After overnight incubation at 37°C with gentle shaking, formic acid was added to a final concentration of 30% and the suspension was incubated for 30 min at 37°C to precipitate the Rapigest. The extracts were dried and the protein digests were re-suspended in 20 μl of a solution of 2% acetonitrile with 0.1% formic acid and sonicated for 1-2 min in a water bath sonicator at 4°C to ensure peptide solubilization. Peptide/protein concentration was obtained at 280 nm absorbance using 1 μl of sample on a Nanodrop ND-1000 spectrometer.
For mass spectrometry analysis, peptides were separated by reversed-phase HPLC (Dionex Ultimate 3000 capillary/nano HPLC system, Dionex, CA, USA) and mass analyzed with a Thermo Fisher LTQ Orbitrap XL (CA, USA), equipped with micro/nanospray ionization sources (Michrom Bioresources Inc., Auburn, CA). HPLC separations were carried out at a flow rate of 2 μL/min on a 0.2mm × 150 mm C18 column (5 μm, 300 Å, Michrom Bioresources Inc). The mobile phases consisted of HPLC grade water and acetonitrile with formic acid added as ion-pairing reagent. A 10-32% gradient was run over 5 hours followed by a 32-90% gradient run over 40 minutes. The heated capillary temperature and electrospray voltage were set at 175°C and 2.0kV, respectively. Protein identifications were obtained using the MassMatrix search engine 21-24. An average of 350 proteins were identified per biopsy, with a range of 250-450 protens.
Immunoperoxidase staining with a monoclonal antibody to fibulin -1 (Santa Cruz Biotechnology, CA, USA) and a polyclonal antibody to podocin (Alpha Diagnostic Intl Inc., TX, USA) was performed (dilution 1:100 and 1: 40 respectively) on formalin-fixed paraffin embedded tissue sections of the fibronectin glomerulopathy biopsies, and additional cases of diabetic nephropathy and proliferative Class IV lupus nephritis (different from the cases used for mass spectrometry). Fibronectin immunostaining was performed in the laboratory of the late Dr. Andrew Herzenberg, Toronto, Canada). The antibody used was a mouse monoclonal antibody directed to the N-terminal of fibronectin. Control tissues used to validate the antibodies were normal kidney, aorta, placenta, and blood vessel walls in the gastrointestinal tract.
Genomic DNA samples were sequenced for fibronectin gene coding region exons 19, 36 and 37, and the complete coding region, including intron-exon boundaries for the fibulin-1 gene. Genomic DNA was extracted from peripheral blood leukocytes using Gentra Puregene Blood Kits (Qiagen, CA, USA) or frozen serum using the QIAamp Investigator kit (Qiagen). DNA amplification was carried out by polymerase chain reaction (PCR). Oligonucleotide primers (Invitrogen, CA, USA) were specifically designed to flank fibronectin exons 19, 36 and 37, and fibulin-1 exons 1-20, according to the gene bank published sequences. Primer sequences are available upon request. PCR products ranged in size from 262-983 base pairs.
PCR products were checked on 2% agarose gel to confirm a single product for direct sequencing with BigDye Terminator Cycle Sequencing Chemistry (Applied Biosystems, CA, USA) using the same PCR primers. Sequencing was performed by 3730 DNA Analyzer by Applied Biosystems.
Protein quantitation was performed by analysis of peptide spectral counts for each protein group identified 25-27. The MassMatrix protein identification search results were regrouped by use of an in-house developed application to create a parsimonious protein list 28. This application regroups protein identifications based on common peptides observed in all sample datasets. Hierarchical clustering of the total peptide matches for each protein group was then performed using Cluster 3.0. The Ingenuity Pathway Analysis tool was used to examine glomerular molecular pathways based on identified proteins (Ingenuity Systems, http://www.ingenuity.com). To evaluate statistically significant difference in protein abundances, the MS/MS data for each sample classification were merged, analyzed by MassMatrix and combined into a parsimonious protein list. The resulting peptide spectral counts for each protein were then normalized by linear scaling such that the total spectral counts were equal. Sample classes were then evaluated pair-wise using the G-test to determine if differences in spectral counts were significant 29.
The pedigree of the fibronectin glomerulopathy family is shown in Figure 1A, and the two family members whose biopsies were studied by laser capture microdissection /proteomics are indicated. It has previously been shown that several fibronectin glomerulopathy kindreds display an A→G single nucleotide polymorphism in heparin-binding domain III of exon 19 of the fibronectin gene 30. This results in a tyrosine→cysteine amino acid change, and may provide a genetic predisposition for fibronectin glomerulopathy. Exon 19 of the fibronectin gene from our patients was sequenced and showed this same heterozygous A→G single nucleotide polymorphism (Fig. 1B). Another affected and one unaffected family member were also sequenced, and the results are shown in Figure 1A.
Renal biopsies from both fibronectin glomerulopathy patients showed prominent glomerular enlargement and diffuse acellular mesangial expansion that was periodic acid-schiff positive and methenamine silver negative (Fig. 2A, B). On ultrastructural examination the glomeruli showed bulky, nondescript granular electron dense material in the mesangium and in the widened subendothelial space along glomerular capillary walls (Fig. 2C). Direct immunofluorescence staining for IgG, IgM, IgA, and C3 did not reveal strong specific staining for any of these immunoreactants. The bulky glomerular deposits did however show staining with the antibody to fibronectin (Fig. 2D).
A heat map of glomerular protein expression obtained from the liquid chromatography-tandem mass spectrometry analysis of the glomerular laser capture microdissection isolates is shown in Figure 3. The two patients with fibronectin glomerulopathy were most closely clustered, followed by the three normal controls. There was more variability among the patients with lupus nephritis and diabetic nephropathy. Compared to normal glomeruli, fibronectin glomerulopathy glomeruli displayed 133 differentially expressed proteins that achieved a pre-determined level of significance (G-score ≥ 5). A subset of these differentially expressed proteins is shown in Table 1. The normalized spectral ratios show the fold-increase (or decrease) of these proteins relative to normal. Table 2 shows the raw spectral counts for each of these proteins and for each individual normal and fibronectin glomerulopathy patient. For comparison, the same proteins were also examined for differential expression relative to normal glomeruli in diabetic nephropathy and lupus nephritis (Table 1). The most striking difference between fibronectin glomerulopathy and normal glomeruli was the extensive level of fibronectin expression in the affected glomeruli. Diabetic nephropathy and lupus nephritis did not show any increase in glomerular fibronectin over control. Additionally, fibulin-1 and fibulin-5 were highly abundant in fibronectin glomerulopathy, and were not found or were minimally detected in normal glomeruli, or glomeruli from patients with diabetic nephropathy and lupus nephritis .
A number of other proteins were altered in fibronectin glomerulopathy, diabetic nephropathy and lupus nephritis. Unexpectedly, many components of the complement pathway were found in fibronectin glomerulopathy and diabetic nephropathy, diseases not associated with systemic complement activation or consumption. Glomeruli from biopsies with both class IV and V lupus nephritis showed complement expression, as expected. In all the disease groups podocyte-associated proteins and anti-oxidant proteins were markedly decreased relative to control. Several serum proteins were increased in fibronectin glomerulopathy, diabetic nephropathy and lupus nephritis compared to control, including clotting pathway components (fibrinogen), amyloid components, and apolipoprotein E.
To verify the laser capture microdissection /proteomic findings, kidney biopsies were stained for fibulin-1 and podocin, as representative proteins that were increased or decreased, respectively, in the fibronectin glomerulopathy kidneys. Fibulin-1 showed mild diffuse staining in the expanded mesangium of patients with fibronectin glomerulopathy (Fig 4B). No fibulin-1 staining was seen in normal control kidneys (Fig. 4A), in the Kimmelstiel-Wilson nodules of diabetic nephropathy (Fig.4C), or in the wire-loop lesions of lupus nephritis (Fig. 4D). It is known that fibulin-1 is present in basement membranes 31, 32, consistent with the staining along the glomerular capillary walls seen in the glomeruli of diabetic nephropathy and lupus nephritis, and less intensely in normal kidneys.
Podocin is known to be expressed in the slit-diaphragms of podocytes 33, and podocin staining was seen in the podocytes along the glomerular capillary loops in normal glomeruli (Fig 5A). Staining was present but less intense in the biopsies with fibronectin glomerulopathy (Fig. 5B) and lupus nephritis (Fig. 5D). Staining was almost negative in the glomeruli with diabetic nephropathy (Fig. 5C).
Given the unexpected similarities in protein expression from three clinically disparate types of glomerular disease, Ingenuity Pathway Analysis was used to functionally annotate glomerular proteins from normal and diseased kidneys (Fig. 6). Representative biologic functions or canonical pathways based on the differentially expressed proteins are shown. The complement system, humoral immunity, and inflammatory response pathways were increased in fibronectin glomerulopathy, diabetic nephropathy and lupus nephritis. Metabolic pathways appeared to be significantly down-regulated in all three glomerular diseases.
The discovery that fibulin was highly represented in the lesions of fibronectin glomerulopathy prompted us to determine if there were unique single nucleotide polymorphisms in the fibulin-1 gene that segregated with fibronectin glomerulopathy. Although a number of single nucleotide polymorphisms were found (data not shown), none appeared to segregate with fibronectin glomerulopathy.
The present work demonstrates the successful application of laser capture microdissection followed by mass spectrometry in the characterization of glomerular proteomes from clinical kidney biopsies. The high degree of clustering within the normal group and the fibronectin glomerulopathy group is an indication of the reproducibility of the method. There was more variability in the diabetic nephropathy and lupus nephritis patient groups, however this is not unexpected. Kidney biopsies in diabetics are generally done in patients with atypical clinical courses, and in lupus nephritis there is a considerable spectrum of pathologic changes within individual lupus nephritis classes. Glomerular immunohistochemistry was used to validate differential glomerular expression of selected proteins identified by the proteomic approach. Additional validation was provided by the finding that several of the differentially expressed proteins revealed by proteomic analysis are consistent with proteins known to be involved in glomerular injury, for example complement activation products and immunoglobulins in lupus nephritis, or fibronectin in fibronectin glomerulopathy.
The most interesting findings from this study are the newly discovered changes in protein expression. This demonstrates that our approach is a robust platform that may inform disease pathogenesis. For example, despite the diversity of glomerular diseases examined, we found a number of glomerular protein changes that were common in the disease glomeruli but not the normal glomeruli, as follows:
1) Several podocyte proteins, particularly those associated with the slit diaphragm, were significantly diminished in fibronectin glomerulopathy, diabetic nephropathy and lupus nephritis. As all of these diseases are proteinuric, and because the attenuated podocyte proteins are important in the maintenance of the glomerular filtration barrier, decreased expression likely contributes to proteinuria. Interestingly, laser capture microdissection followed by microarray analysis of glomeruli from patients with focal segmental glomerulosclerosis showed a down-regulation of mature podocyte genes 5. It is not presently known if the loss of these podocyte proteins and genes is a non-specific response to podocyte injury and/or proteinuria, or is a primary event in the development of proteinuria. The latter scenario would imply a common pathogenic mechanism for different diseases.
2) Complement components were found in fibronectin glomerulopathy, diabetic nephropathy and lupus nephritis. The presence of complement proteins in the glomeruli of lupus nephritis patients is expected, though the patterns of Class IV and Class V are slightly different. Class IV lesions uniquely contained C1q, no C4 protein, and abundant C3 protein. The presence of C3 without C4 suggests significant alternative pathway activation, and is in agreement with our recent report of C3 consumption, but not C4 consumption, being independently associated with renal flare 34. In contrast, Class V lesions contained no C1q but detectable C4 protein, suggesting the involvement of the lectin pathway. Other complement components were also present in Class V lesions, including abundant C9, likely reflecting the presence of the complement membrane attack complex as previously reported 35.
In diabetic nephropathy there are some data that complement may be involved in disease progression, possibly through the lectin binding pathway, and C3 and the membrane attack complex (C5b-9) have been found in glomeruli (reviewed in 36).
The presence of complement proteins in fibronectin glomerulopathy glomeruli is somewhat surprising, because immunohistochemical analysis of fibronectin glomerulopathy reveals inconsistent involvement of immunoglobulins or complement 13. Fibronectin binds C1q via its collagenous tail 37, 38, which could explain the presence of C1q, but this binding does not result in complement activation 39. Complement activation could occur through the alternative pathway, especially if the expression or function of complement regulators is compromised. Interestingly, a genome-wide linkage study identified a fibronectin glomerulopathy genetic susceptibility locus within the regulators of complement activation gene cluster 40. However, no studies to date have identified abnormal expression or function of regulators of complement activation gene cluster proteins in patients with fibronectin glomerulopathy 41. Another possible explanation for the presence of complement proteins in fibronectin glomerulopathy is through glomerular complement expression. This has been demonstrated for C3 by in situ hybridization in glomeruli from lupus nephritis patients, but not from healthy controls 42. Finally, this could simply represent non-specific trapping of complement proteins, but if so this is significantly more than in normal glomeruli (see raw spectral counts, Table 2). The presence of complement proteins in the glomeruli of fibronectin glomerulopathy raises the possibility that complement-mediated damage may be involved in this non-immune glomerular lesion. A recent proteomic evaluation of chronic allograft nephropathy also unexpectedly found complement pathway activation and reached a similar conclusion 43. Limiting complement activation may therefore improve clinical outcomes in diseases not previously thought to involve clinically relevant complement activation.
3) Anti-oxidant proteins were conspicuously under-represented in fibronectin glomerulopathy, diabetic nephropathy and lupus nephritis glomeruli, but the proteomic approach cannot distinguish between a primary defect in anti-oxidant production, or a secondary loss of these proteins. Nonetheless, a reduction in glomerular anti-oxidants leaves the tissue open for oxidant injury.
4) Pathway analysis based on the differentially expressed proteins suggested that compared to normal, metabolic pathways were suppressed in all three types of glomerular disease. This suggests that any type of glomerular injury may attenuate metabolism of glomerular cells, but more importantly, indicates that metabolomic evaluation of glomerular disease may uncover novel mechanisms of injury and repair.
To further demonstrate the utility of laser capture microdissection followed by glomerular proteomics, we studied fibronectin glomerulopathy in more detail. This uncommon glomerular disease results from the deposition of excessive matrix material in glomeruli. The hypothesis that fibronectin glomerulopathy is mechanistically related to abnormal glomerular accumulation of fibronectin is based mainly on evidence from immunohistochemical staining, which has limitations. Immunohistochemical staining can only test constituents for which specific antibodies are available, and can be falsely positive due to nonspecific trapping of antibodies, or falsely negative if antigens are blocked. By isolating glomeruli from kidney biopsies of affected patients, and doing unbiased proteomic analysis of the recovered glomerular proteins, we confirmed that fibronectin is the predominant protein in the lesions of fibronectin glomerulopathy glomeruli. In addition to several common pathways of glomerular injury activated in fibronectin glomerulopathy described above, we found considerable deposition of fibulin in these glomeruli. Among the diseases examined, fibulin accumulation appears to be unique to fibronectin glomerulopathy because it is not present in normal glomeruli, diabetic nephropathy, or lupus nephritis. Interestingly, fibronectin binds to fibulin-1 but not fibulin-5 31, 41, 44, 45. The fibulins appear to regulate several of the molecular activities of fibronectin. For example, fibronectin is important in cell-matrix interactions, and facilitates cell attachment, cell spreading, and cell migration 46. Fibronectin also regulates the cytoskeleton, cellular morphology, and differentiation 30. Fibulin-1 down-regulates fibronectin-facilitated cell spreading, adhesion and motility 47. These effects may occur because fibulin-1 inhibits fibronectin-mediated extracellular signal-regulated kinase activation 47. Although fibulin-5 does not directly bind to fibronectin, it binds to the same β1 integrins as fibronectin (α5β1, α4β1), and like fibulin-1, blocks fibronectin-mediated cell spreading, motility, and proliferation 48.
The role of fibulins plus fibronectin in the pathogenesis of glomerular injury in fibronectin glomerulopathy is not yet known, however fibulins have been associated with other types of kidney disease. Specifically, fibulin-1 was identified as a candidate permeability factor in focal segmental glomerulosclerosis 49, perhaps because it interferes with cell adhesion, and could therefore disrupt the podocyte-basement membrane relationship. Gene profiling of autosomal dominant polycystic kidney disease tissue showed markedly increased fibulin mRNA expression, and the protein was demonstrated in epithelial cells lining cysts 50. Fibulin-1 has been shown to co-localize with ADAMTS-1 and increase its catalytic activity 51. It is conceivable that in fibronectin glomerulopathy the accumulated fibronectin helps localize fibulins to glomeruli, and the fibulins inhibit cell spreading, adhesion, and motility, disrupting mesangial and/or podocyte function and facilitating the development of proteinuria.
In summary, the present study shows that the application of laser capture microdissection followed by liquid chromatography-tandem mass spectrometry analysis of kidney biopsies can provide large-scale quantitative proteomic profiles that may be important for characterizing glomerular diseases as well as determining mechanisms of disease pathogenesis. The ability to more clearly understand underlying molecular events critical to disease expression may open avenues for novel therapeutic interventions.
This work was supported in part by DK077331 (BHR), and RR025755 (MF).
Disclosures/Conflicts of Interest: None