Assessment of Kidney Function in Mouse Models of Glomerular Disease

The use of murine models to mimic human kidney disease is becoming increasingly common. Our research is focused on the assessment of glomerular function in diabetic nephropathy and podocyte-specific VEGF-A knock-out mice; therefore, this protocol describes the full kidney work-up used in our lab to assess these mouse models of glomerular disease, enabling a vast amount of information regarding kidney and glomerular function to be obtained from a single mouse. In comparison to alternative methods presented in the literature to assess glomerular function, the use of the method outlined in this paper enables the glomerular phenotype to be fully evaluated from multiple aspects. By using this method, the researcher can determine the kidney phenotype of the model and assess the mechanism as to why the phenotype develops. This vital information on the mechanism of disease is required when examining potential therapeutic avenues in these models. The methods allow for detailed functional assessment of the glomerular filtration barrier through measurement of the urinary albumin creatinine ratio and individual glomerular water permeability, as well as both structural and ultra-structural examination using the Periodic Acid Schiff stain and electron microscopy. Furthermore, analysis of the genes dysregulated at the mRNA and protein level enables mechanistic analysis of glomerular function. This protocol outlines the generic but adaptable methods that can be applied to all mouse models of glomerular disease.


Introduction
The use of murine models to mimic human kidney disease is becoming increasingly common. Such murine models include spontaneous models such as spontaneously hypertensive rats (SHR) 1 , streptozotocin (STZ)-induced diabetic rats and mice 2 , and the db/db type II diabetic mice 3 , genetically engineered models such as primary podocyte-specific focal segmental glomerular sclerosis (FSGS) models 4 , the podocyte-specific vascular endothelial growth factor A (VEGF-A) knock-out (VEGF-A KO) model 5 , and Alport syndrome models 6 , and acquired models such as the 5/6 nephrectomy 7 and the unilateral ureteral obstruction (UUO) model 8 . In order to assess the different aspects of glomerular function in these models, several techniques are available. The purpose of this method paper is to demonstrate a comprehensive work-up that should be performed in mouse models of kidney disease in order to fully assess glomerular function.
The rationale behind the use of this method is that it enables the glomerular phenotype to be fully evaluated from multiple aspects. This includes assessing the glomerular permeability, both to protein and to water, the glomerular structural abnormalities, and changes in the expression/ splicing of mRNAs and proteins essential for normal glomerular function. By using this method, the researcher is able to determine the kidney phenotype of the model and assess the mechanism as to why the phenotype develops. This is vital information on the mechanism of disease, which is required when examining potential therapeutic avenues in these models.
In the literature, it is a common occurrence to be presented with a mouse model of glomerular disease where the phenotype is determined by an increased level of albumin in the urine. However, there is evidence to suggest that a single method to determine glomerular function is not always effective; measuring the urinary albumin excretion rate or the urinary albumin creatinine ratio (uACR) only provides information on total renal function, and not of the individual glomeruli. Previous studies have demonstrated that the permeability can vary in different glomeruli from the same kidney 5,9,10 . In addition, assessment of the permeability of individual glomeruli is a more sensitive way of assessing glomerular function; the technique of measuring the individual glomerular water permeability (L p A/V i ) has shown to be more sensitive to changes in glomerular function than measuring the uACR 9 . This assay is beneficial in mouse models that are resistant to proteinuria, such as those on a c57BL/6 background 11 . The advantage of the present method paper is that it examines both the total renal permeability to albumin as well as the individual glomerular permeability to water.
Examination of glomerular structural abnormalities is often assessed by a battery of stains such as Periodic Acid Schiff (PAS), trichrome, and silver stains. These enable a trained renal pathologist to evaluate the level of renal disease via a scoring method. Although all good methods, changes to the glomerular macro-structure are not always observed in acute kidney injury models 1. Collect urine at the experimental baseline and at regular intervals (once weekly to once monthly) up until the experimental endpoint. 2. Set up mouse metabolic cages with water and enrichment diet. Place mice (male, aged 6 -8 weeks) in individual cages for 6 h in a quiet room. 3. Return mice to regular housing and collect urine from the empty cages. A minimum of 50 µL is required.
NOTE: If the mouse does not produce urine in the given time, repeat on another day in a warmer room. 4. Centrifuge the urine at 500 x g for 10 min. Collect the urine and retain sediment to assess podocyte loss. Store the urine at -20 °C in the short-term at this point. 5. Dilute the urine into 1% bovine serum albumin (BSA) in 1x Tris buffered saline (TBS pH 7.5) at a 1:500 to 1:10000 dilution, depending on the severity of the albuminuria. NOTE: The end volume should be >400 µL. Optimize to determine the right dilution at each time point. 6. Quantify the urinary albumin concentration using a mouse albumin enzyme linked immunosorbent assay (ELISA) per the manufacturer's instructions. 1. Briefly, coat the ELISA plate, which is optimized for protein binding, with 100 µL of the anti-mouse albumin primary antibody (10 µg/mL in 0.5 M carbonate-bicarbonate pH 9.6) for 1 h at room temperature. 2. Wash excess antibody from the wells five times with 1x TBS plus 0.05% Tween (pH 8.0), and add 200 µL of blocking solution (1x TBS with 1% BSA pH 8.0) overnight at room temperature. Prepare the following materials: isoflurane, small ethylenediaminetetraacetic acid (EDTA)-coated blood tubes, 23 -25G needles, 5 mL EDTA coated syringes, 10 mL glass vials, 10 mL plastic vials, 0.5 mL plastic tubes, disposable tissue molds, dry ice, liquid N 2 , mouse surgical tools, and optimal cutting medium (OCT). 3. Place the mouse under deep anesthesia, which is verified by the mouse being non-responsive to a needle prick to the foot pad, using an isoflurane chamber, or equivalent routes of terminal anesthesia such as injectable anesthetic agents (pentobarbital, 50 mg/kg intraperitoneal [IP]; avertin, 240 mg/kg IP), or carbon dioxide (CO 2 ) exposure (75% CO 2 /25% O 2 ). 4. Cull mouse via cardiac puncture into the left ventricle and collect as much blood as possible. Transfer to the EDTA-coated blood tube for up to 4 h. If preferred, mice can be culled via cervical dislocation with care taken not to rupture the jugular vein. 5. Dissect out the kidneys through the abdomen and wash in ice cold 1x PBS. 6. To examine the cortical glomeruli, remove one pole of kidney cortex and cut into 1 mm 3 pieces. To examine the deep juxta-medullary glomeruli, repeat the same technique with tissue from the medulla. Place in 5 mL of 2.5% glutaraldehyde solution in a glass EM vial. Store at 4 °C. CAUTION: 2.5% glutaraldehyde solution: toxic, sensitizer, irritant; use in a fume cabinet NOTE: process within 1 month for best results. 7. For histology, remove the upper third of a kidney, to ensure both cortical and juxta-medullary glomeruli will be present, and fix in 5 mL of 4% paraformaldehyde at 4 °C for 24 h. Transfer to 5 mL of 70% EtOH for 24 hours before embedding in paraffin. CAUTION: 4% PFA: fixative, use in fume cabinet 8. For immunofluorescence, place a third of the kidney, to ensure both cortical and medullary glomeruli will be present, into the tissue mold and coat in OCT. Place on dry ice to freeze and store at -80 °C. 9. For protein and RNA, place 3 x 2 mm 3 pieces of kidney cortex into 0.5 mL plastic tubes and snap freeze in liquid N 2 . Store at -80 °C. For long-term tissue storage for RNA, place tissue in 5 volumes of RNA stabilization solution and store at -80 °C. 10. For isolation of glomeruli, slice up the remaining kidney tissue and place in 5 mL of mammalian Ringer's solution with 1% BSA on ice.
Prepare to sieve glomeruli immediately.

Plasma Creatinine
NOTE: Plasma creatinine can be up-regulated in renal disease, indicating a reduction in the filtration capacity of the glomeruli. The blood urea nitrogen (BUN) levels can also be assessed, although the protocol is not described here.
1. Centrifuge the blood sample at 500 x g for 15 min at 4 °C. 2. Collect the plasma, which can be stored at -20 °C in the short-term at this point. 3. Quantify the plasma creatinine concentration using a chemical creatinine assay per the instructions above for urinary creatinine in protocol 1.11. 4. Determine the creatinine concentration of each sample by reading the plate at an absorbance of 490 nm before and after the addition of the acid solution. NOTE: The difference between these absorbance values is directly proportional to the creatinine concentration in each sample. A standard curve is generation from the standards. If the technical repeats have a CV value greater than 5%, repeat the assay for those samples.

Isolation of Glomeruli
NOTE: Glomeruli can be isolated to assess the permeability of individual glomeruli ex vivo, as well as the expression of specific protein and mRNA markers of glomerular disease.
As the bits of kidney are pushed through, remove the top sieve and proceed to do the same on the next. Repeat until only the 100 µm and 70 µm sieves remain. 3. Transfer the glomerular harvest retained by the 100 µm and 70 µm sieves to 10 mL of fresh mammalian Ringer's solution with 1% BSA, on ice. NOTE: If the number of glomeruli per mL Ringer's solution is few, reduce the volume of Ringer's solution used to collect the glomeruli from the last two sieves.
. Please refer to Figure 1 for a detailed diagram of the set up. 2. Prepare the following solutions: mammalian Ringer's solution with 1% BSA (pH 7.4) and mammalian Ringer's solution with 8% BSA (pH 7.4).
Warm both to 37 °C. 3. Pull micropipettes from glass capillary tubes (optical density: 1.2 mm). Generate a 5 -8 µm aperture tip by cutting the micropipette under a microscope. 4. Use the glomerular L p A/V i rig to catch intact individual glomeruli that are free of Bowman's capsule and tubular fragments onto the micropipette using suction. A detailed summary of the oncometric assay is found in Salmon et al 10 .
In brief, once a glomerulus is caught and secured on the suction micropipette, begin recording the video of the glomerulus under the microscope. 5. Firstly, equilibrate the glomerulus in the 1% BSA Ringer's solution for 30 s before switching the perifusate to the concentrated 8% BSA Ringer's solution for 10 s. Then switch the perifusate back to 1% BSA Ringer's solution and stop the recording. 6. Wash the glomerulus away and repeat the process for 10 -15 glomeruli per mouse. Ensure the perifusate flow rates are identical and not to fast (10 mL/min) so as not to distort the glomerular structure. 7. Measure the initial rate of glomerular shrinkage to calculate the glomerular water permeability (L p A) normalized to the glomerular volume (V i ).
Detailed information regarding the analysis can be found in Salmon et al 10 .

Periodic Acid Schiff (PAS) Stain
NOTE: The PAS stain will highlight the basement membranes of glomerular capillary loops and the tubular epithelium. It enables detailed visualization of the glomerular cells, mesangial matrix and potential expansion, and potential changes of the GBM (i.e., thickening and irregularities).
1. Section the paraffin-embedded, PFA-fixed kidney cortex using a microtome at 5 µm thickness onto poly-L-lysine coated slides. Dry at 37 °C for 1 h. Ensure the section does not contain any folds or holes, which can distort the morphology under the microscope. Wash slides in running tap water for 5 min. 5. Counterstain with Hematoxylin for 3 s before thoroughly rinsing slides in running tap water for 15 min. NOTE: Some optimization may be required to determine the optimal time for Hematoxylin staining. 6. Dehydrate slides using the reverse of the deparaffinization protocol in step 6.2. Finish with xylene. 7. Air dry slides and mount with xylene-based mounting media. 8. Image on a light microscope at 400X magnification to assess glomerular structures. Evaluate the following: thickening and irregularities of the GBM, collapsing of capillary loops, fibrotic tissue, sclerosis, cellular proliferation (endothelial, podocyte, and mesangial, or inflammatory cells infiltrating the tuft). NOTE: For a comprehensive evaluation of glomerular pathophysiology, lesions elsewhere in the kidney should be evaluated, such as in the tubules.

Transmission Electron Microscopy (TEM)
NOTE: TEM allows the examination of ultra-structural abnormalities in the kidney, such as the GBM, podocyte foot processes, and endothelial fenestrations, which are not visible with light microscopy. This is important in models where renal damage is not so pronounced (i.e., no albuminuria and major structural abnormalities).
1. For GBM, insert a fixed digital grid (10 x 10) over the 6200X micrograph and measure the thickness of the GBM at the point where the grid lines cross the GBM. Measure from the basal endothelial cell membrane to the basal podocyte foot process cell membrane in a perpendicular tangent to the endothelial cell membrane using the straight line tool. Determine the mean measurement for each glomerulus from 10 individual measurements. 2. For the endothelial fenestration number, measure the length of the GBM present in the 6200X micrograph and count the number of endothelial fenestrations per unit length of GBM. Take an average from at least 4 micrographs per glomerulus. 3. For the podocyte foot process width, insert a fixed digital grid (10 x 10) over the 6200X micrograph. Measure the width of the podocyte foot processes that cross the grid lines. Measure the width at the widest part of the foot process where it meets the GBM; ensure the line is perpendicular to the tangent of the podocyte basal membrane. Determine the mean measurement for each glomerulus from 10 individual measurements. 4. For podocyte slit width, insert a fixed digital grid (10 x 10) over the 6200X micrograph. Measure the width of the podocyte slit diaphragms that cross the grid lines. This is the point where the foot processes are closest together, at the widest part of each foot process, from podocyte membrane to membrane. Ensure the measurement is perpendicular to the tangent of the podocyte basal membrane. Determine the mean measurement for each glomerulus from 10 individual measurements. 5. For the number of podocyte foot processes, measure the length of the GBM present in the 6200X micrograph and count the number of podocyte foot processes per unit length of GBM. Take an average from at least 4 micrographs per glomerulus. 6. For the sub-podocyte space coverage, see detailed method in Neal et al. 14 .
6. Using the 940X micrographs, examine glomeruli for the presence of abnormal structure, deposits, and infiltrates by eye.

Immunofluorescence for Podocyte and Endothelial Markers
NOTE: Immunostaining allows visualization of the protein expression patterns, such as endothelial capillary loops, which can collapse in glomerular disease.
1. Whilst the kidney cortex is still frozen, thoroughly grind in 3 mL of phenol reagent using a pestle and mortar. If using glomerular extracts, add 1 mL of phenol reagent and homogenize the sample for 30 s. CAUTION: TRIzol reagent: irritant; use in fume cabinet 2. Perform an RNA extraction using the method described by Chomczynski and Sacchi 16 . NOTE: Commercial RNA extraction kits are available as an alternative to this method. 3. Assess the quantity and quality of RNA obtained using one of the various methods available. RNA is aliquoted and stored at -80 °C at this point. Avoid repeat freeze thawing. NOTE: If new to this method, check the quality of the RNA before proceeding to the next step by running the RNA on an agarose gel to ensure a clear 28S and 18S ribosomal band. Expect to recover between 2 to 5 µg of RNA using this method. 4. DNase treat 1 µg of RNA (make volume up to 10 µL with RNase-free water plus 1 µL of DNase and 1 µL of DNase buffer) for 1 h at 37 °C.
Stop the reaction with 1 µL of DNase stop solution at 65 °C for 10 min. 5. Add 0.5 µL of oligo (dT) and random primers. Incubate at 70 °C for 10 min. 6. Immediately quench on ice for 5 min. 7. Add the following; MMLV reverse transcriptase enzyme (400 U; replace with DEPC H 2 O in the RT -control sample), MMLV buffer (1x), dNTP mix (0.5 mM), and ribonuclease inhibitor (40 U); make up to 50 µL with DEPC water. 8. Incubate reaction mix at 37 °C for 1 h followed by 95 °C for 5 min to deactivate the enzyme.
NOTE: To generate a higher yield of cDNA, incubate at 37 °C for up to 3 h. 9. Assess the quantity and quality of cDNA using the various methods available. 10. Use PCR to assess the mRNA expression and splicing patterns of genes hypothesized to be dysregulated in the glomerular disease model.
The protocol will vary depending on the gene of interest.

Representative Results
Urine was collected using metabolic cages from wild type (WT), inducible podocyte-specific VEGF-A knock out (VEGF-A KO), and VEGF-A KO X Neph-VEGF 165 b mice (VEGF-A KO mice that over-express the human VEGF-A 165 b isoform in the podocytes in a constitutive manner). Upon measurement of the urinary albumin creatinine ratio at weeks 0, 4, 10, and 14 after doxycycline induction of VEGF-A KO, VEGF-A KO mice developed progressive albuminuria by 10 weeks compared to WT littermate controls. The absolute values can be seen in Figure 2A, and the normalized to the baseline of each mouse values in Figure 2B. However, albuminuria in not observed in the VEGF-A X Neph-VEGF-A 165 b mice (Figure 2), indicating that VEGF-A 165 b is protective in the model of albuminuria 5 .
The glomerular L p AV i was measured in individual glomeruli sieved from WT, VEGF-A KO, and VEGF-A X Neph-VEGF-A 165 b kidneys. An example of how a glomerulus is caught and the shrinkage observed when perifused with 8% BSA is shown in Figure 3A. This shrinkage is then used to determine the glomerular L p A/V i for each glomerulus (Figure 3B). VEGF-A KO mice had a significantly increased glomerular L p A/V i at 14 weeks post VEGF-KO induction compared to WT control glomeruli. Although lower in VEGF-A X Neph-VEGF-A 165 b mice, the increased glomerular L p A/V i was not prevented by over-expression of VEGF-A 165 b at 14 weeks 5 .
PAS staining of kidney cortex sections 14 weeks after induction of VEGF-A KO did not reveal any glomerular structural abnormalities in the VEGF-A KO or VEGF-A X Neph-VEGF-A 165 b mice ( Figure 4A). However, upon analysis of the glomerular ultra-structure via EM, VEGF-A KO mice had developed an increased GBM width, decreased number of endothelial fenestrae, decreased SPS coverage, and increased average podocyte slit width (Figure 4C-4F). The average podocyte foot process width and number of slits remained unchanged (Figure 3B and 3G).
Over-expression of VEGF-A 165 b in the VEGF-A KO mice prevented the changes to the GBM and slit width (Figure 4C and 4F). However, VEGF-A 165 b had no effect on the altered fenestrae number and SPS coverage (Figure 4D and 4E) 5 .
RT-PCR performed on RNA extracted from sieved glomeruli revealed that the human VEGF-A 165 b mRNA is only present in the VEGF-A KO X Neph-VEGF-A 165 b mice ( Figure 5A). When extracting protein from sieved glomeruli and assessing the levels of proteins via Western blotting, the glomerular protein expression of VEGFR-2 was found to be decreased in VEGF-A KO mice, which was prevented by over-expression of VEGF-

Discussion
This protocol describes a full kidney work-up that should be carried out in mouse models of glomerular disease, enabling a vast amount of information regarding kidney and glomerular function to be obtained from a single mouse. The critical steps in each method allow for detailed functional, structural, and mechanistic analysis of glomerular function, including assessment of the permeability of the kidneys as a whole (uACR and plasma creatinine measurements), the permeability of individual glomeruli (glomerular L p A/V i ), examination of the structural alterations (PAS, Trichrome blue, and EM), protein localization (IF), and glomerular gene expression (RT-PCR and Western blotting). These methods are key to the full assessment of glomerular function in mouse models of renal disease.
When assessing the permeability of the GFB, many studies have opted to use the conventional uACR or 24 h albumin excretion rate as an effective measure 17,18 . Although these techniques allow assessment of the GFB permeability as a whole, it does not allow for individual glomerular permeability assessment and variation amongst glomeruli. Previous studies have found measurement of the glomerular L p A/V i to be a more sensitive measure of changes to the GFB permeability 5,9 . Indeed, in the representative results demonstrated in this paper, at 14 weeks post induction of VEGF-A KO, VEGF-A KO X Neph-VEGF-A 165 b mice have a significantly lower uACR compared to VEGF-A KO mice; however, this result is not reflected in the glomerular L p A/V i measurements, where VEGF-A 165 b did not significantly prevent increases in the GFB permeability (Figure 1 and Figure 2) 5 . This shows the importance of using multiple assays to assess both the kidney permeability and the permeability of individual glomeruli. Furthermore, the glomerular L p A/V i oncometric assay suggests that the permeability of individual glomeruli from the same kidney can vary greatly, especially in disease models 5,10,19 . One limitation to measuring the glomerular L p A/V i is that it can only be performed at the experimental end-point; thus, regular uACR measurements are required to give an indication of the experimental end-point.
In addition to assessing the functional phenotype, the present method also encourages assessment of the structural and ultrastructural phenotype. This can be done using a selection of stains such as PAS, trichrome, and silver stains; each to assess different aspects of the glomerular morphology. In acute models of glomerular disease, which is often the case in mouse models, is can be difficult to detect any major structural abnormalities using these stains unless you are a trained renal pathologist. Therefore, carrying out EM is suggested to assess the ultrastructure of the GFB, which allows the quantitative measurement of parameters such as the GBM, endothelial fenestrae size and number, and podocyte characteristics. Such measurements require minimal training to perform and enables the investigator to determine the cell-types/ structures affected in a disease model. In the example shown in the representative results, the VEGF-A KO mouse was found to be a mild model of glomerular disease, thus, no major structural abnormalities were present upon PAS staining. However, podocyte-specific VEGF-A KO did induce changes to the GBM, podocytes, and endothelial cells when examining the glomerular ultra-structure 5 . Unfortunately, the preparation of the kidney for EM described in the present method does not enable detection of the endothelial glycocalyx, which is also known to have significant effects on the permeability of the GFB 19 . In order to accurately measure the glycocalyx depth, the kidney should be perfuse-fixed with 2.5% glutaraldehyde with 1% Alcian blue for endothelial glycocalyx labelling, as described in Oltean et al 19 .
Once the functional and structural phenotype have been assessed, the expression/activation patterns of different genes and pathways can then be assessed specifically in the glomeruli. Prior ultra-structural assessment could give some information regarding the cell types/glomerular structures involved, indicating whether podocyte-or endothelial-specific genes/pathways should be examined. For example, in the representative results from the VEGF-A KO mice, a reduction in the endothelial fenestrae number was observed ( Figure 3D); therefore, the glomerular protein expression of an endothelial marker known to be involved in the VEGF-A pathway was examined; VEGFR-2 ( Figure 4B) 5 . In addition to the expression of proteins in the glomeruli, their localization can also be visualized using IF. In a study by Zhang et al 20 , podocyte-specific overexpression of GLUT1 was confirmed in the podocytes by IF co-localizing the increased GLUT1 with podocin.
In comparison to alternative methods presented in the literature to assess glomerular function, the use of the method outlined in this paper to assess kidney function in mouse models of glomerular disease enables the glomerular phenotype to be fully evaluated from multiple aspects. By using this method, the researcher is able to determine the kidney phenotype of the model and assess the mechanism as to why the phenotype develops. This vital information on the mechanism of disease is required when examining potential therapeutic avenues in these models. This method can be easily applied to future investigations into glomerular function both in the assessment of disease phenotypes and potential therapeutics.