The podocyte-specific knockout of palladin in mice with a 129 genetic background affects podocyte morphology and the expression of palladin interacting proteins

Proper and size selective blood filtration in the kidney depends on an intact morphology of podocyte foot processes. Effacement of interdigitating podocyte foot processes in the glomeruli causes a leaky filtration barrier resulting in proteinuria followed by the development of chronic kidney diseases. Since the function of the filtration barrier is depending on a proper actin cytoskeleton, we studied the role of the important actin-binding protein palladin for podocyte morphology. Podocyte-specific palladin knockout mice on a C57BL/6 genetic background (PodoPalldBL/6-/-) were back crossed to a 129 genetic background (PodoPalld129-/-) which is known to be more sensitive to kidney damage. Then we analyzed the morphological changes of glomeruli and podocytes as well as the expression of the palladin-binding partners Pdlim2, Lasp-1, Amotl1, ezrin and VASP in 6 and 12 months old mice. PodoPalld129-/- mice in 6 and 12 months showed a marked dilatation of the glomerular tuft and a reduced expression of the mesangial marker protein integrin α8 compared to controls of the same age. Furthermore, ultrastructural analysis showed significantly more podocytes with morphological deviations like an enlarged sub-podocyte space and regions with close contact to parietal epithelial cells. Moreover, PodoPalld129-/- of both age showed a severe effacement of podocyte foot processes, a significantly reduced expression of pLasp-1 and Pdlim2, and significantly reduced mRNA expression of Pdlim2 and VASP, three palladin-interacting proteins. Taken together, the results show that palladin is essential for proper podocyte morphology in mice with a 129 background.


Introduction
More than 10% of the people worldwide are suffering from chronic kidney disease (CKD) and the tendency is still increasing [1]. In more than 75% of the diseased kidneys, a specific cell type in the filtration unit of the kidney, the podocyte, is damaged or lost [2]. mice (Charles River Laboratories, Wilmington, MA, USA) and the "speed congenics" approach [27]. After 8x backcrossing, the new mouse line was established.
Podocyte-specific palladin knockout mice with 129 background (PodoPalld129-/-) and heterozygous Cre-recombinase expression were used for experiments. Mice without Cre-recombinase expression were used as controls (PodoPalld129+/+). Experiments were done with 6 and 12 months old male PodoPalld mice (at least n = 3 of each group). Genotyping of mice was performed with Phire 1 Animal Tissue Direct PCR Kit (Finnzymes/Thermo Fisher Scientific) in accordance to the manufacturer's instructions using primers shown in S1 Table. All prerequisites of the German animal protection law were met and experiments were performed in accordance with the guidelines of the federal agencies in Mecklenburg-Western Pomerania (LALLF M-V). The responsible ethics committee within the LALLF M-V approved the experiments with mice. For kidney removal, mice were sacrificed by the use of barbiturate.

Histology staining
The samples were dehydrated and embedded into paraffin by standard procedures. Paraffin sections (4 μm) were performed on a Leica SM 2000R (Leica Microsystems). After deparaffinization, sections were rehydrated and PAS and H&E stainings were performed by standard procedures. Sections were mounted in Eukitt (Fluka/Sigma-Aldrich, St. Louis, MO, USA) and imaged with an Olympus BX50 microscope (Olympus Europe, Hamburg, Germany).

Immunofluorescence staining and immunohistochemistry of kidney sections
After deparaffinization, sections of mouse kidneys were rehydrated and unmasked in citrate buffer (0.1 M, pH 6.0) by heating for 5 min in a pressure cooker.

RNA extraction and qRT-PCR analysis
Samples from glomeruli were processed in Tri-Reagent (Sigma-Aldrich) according to the manufacturer's instructions. For cDNA synthesis, 1 μg of isolated total RNA was reverse transcribed using the QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany). Quantitative real-time PCR (qRT-PCR) was performed on a LightCycler 1 Nano (Roche Diagnostics GmbH, Mannheim, Germany) using iTaq Universal SYBR Green Supermix (Bio-Rad Laboratories GmbH, Hercules, CA, USA) and primers see S1 Table. For qRT-PCR analysis we used the following number of mice per group: n = 5 of PodoPalld +/+ mice at 6 and 12 months of age, n = 6 and n = 7 of PodoPalld-/-mice at 6 and 12 months of age. Every control was compared to every PodoPalld-/-mouse for both age groups. The data were analyzed by standard methods [29] and the relative mRNA expression was calculated by normalizing values to the housekeeping gene Rpl32.
Asterisks indicate statistically significant differences (p<0.05) based on unpaired Student's t-test or Mann-Whitney U test between the Rpl32 and targets of interest using GraphPad Prism 8 (GraphPad, La Jolla, CA, USA). For the tests, n replicates (n�5) were used and it was always checked for prerequisites such as normal distribution and similar variance between the measured groups.

Glomerular morphology analysis of mouse kidney
For transmission electron microscopy, kidneys were embedded in EPON 812 (SERVA, Heidelberg, Germany). Ultrathin sections were cut and contrasted with 5% uranyl acetate and lead citrate. All grids were examined with a LIBRA 1 120 transmission electron microscope (Carl Zeiss Microscopy, Jena, Germany).
Scanning electron microscopy was performed according to Artelt et al. [30]. Furthermore, the presence of glomerular abnormalities was investigated more precisely using Richardson's (Azur II/ Methylene blue) stained semithin sections of mouse kidneys. Glomeruli were categorized into (i) glomeruli with normal morphology, (ii) dilated capillaries and (iii) affected podocytes (podocytes with cyst and enlarged sub-podocyte space). The presence of dilated capillaries was verified by quantitative analysis of the capillary area per glomerulus on semithin sections.
We analyzed at least n = 3 mice of each group with a total of at least 30 glomeruli. Asterisks indicate statistically significant differences (p<0.05) based on two-way ANOVA analysis with FDR correction using GraphPad Prism 8.

Confirmation of the podocyte-specific palladin knockout
To confirm that the backcrossed PodoPalld129-/-mice show a podocyte-specific knockout for palladin, we analyzed the animals by immunohistochemistry (IHC) and qRT-PCR. The IHC staining of PodoPalld129-/-mice revealed that the podocytes were palladin-negative at the age of 6 as well as of 12 months in contrast to the controls (S1A Fig). A faint signal for palladin (6 months: 0.26 ±0.14, 12 months: 0.19±0.05; mean±SD) was detected by qRT-PCR of isolated glomeruli which is due to a slight contamination with small vessels that still express palladin as shown in S1 Fig.

PodoPalld129-/-mice show dilated capillaries as well as a reduced number of mesangial cells
We studied the glomerular morphology of two groups of mice, one group with an age of 6 months and the other with an age of 12 months by histological and ultrastructural analysis. All glomeruli of the PodoPalld129-/-mice developed severe dilatation of the capillaries as shown by a Periodic Acid Schiff (PAS) and Hematoxylin and Eosin (H&E) staining ( Fig 1A and S2  Fig). We found that 38.1±9.4% (PodoPalld129-/-) vs. 16.5±5.2% (control; mean±SD, p<0.01) of the glomeruli at 6 month and 38.0±5.6% (PodoPalld129-/-) vs. 31.9±1.8% (control; mean ±SD) at 12 months developed a dilatation of their capillaries (n = 3 animals and >60 analyzed glomeruli per group, Fig 1B). Quantitative analysis of the capillary area per glomerular area underlined that the capillaries of PodoPalld129-/-glomeruli were significantly enlarged (6 months: 0.46±0.06, p<0.05; 12 months: 0.50±0.07, p<0.05; mean±SD) compared to the control glomeruli (6 months: 0.36±0.02, 12 months: 0.40±0.03; mean±SD) (n�3 animals and >30 glomeruli per group, Fig 1B). Since the dilatation of the glomerular tuft could be caused by a lower number of mesangial cells, we stained for the specific integrin α8 that is expressed in mesangial cells. Interestingly, we observed a marked reduction of the integrin α8 signal in 6 as well as in 12 months old PodoPalld129-/-mice compared to corresponding controls ( Fig 1C) suggesting that palladin might influence the development or the survival of mesangial cells. The reduction of the Itga8 expression in PodoPalld129-/-mice was verified by Western blot (Fig 1D).

The loss of palladin results in podocyte foot process effacement
Since PodoPalld129-/-mice developed morphological abnormalities we studied the expression of slit membrane proteins and foot process morphology in detail. Immunohistological staining showed that the slit membrane protein nephrin was significantly reduced in 6 as well as in 12 months old PodoPalld129-/-mice compared to the controls (Fig 3A), whereas expression of the nephrin mRNA was unchanged (6 months: 1.05±0.27, 12 months: 0.98±0.26; mean±SD) (S3 Fig). The same results were received for the slit membrane protein podocin (mRNA level 6 months: 1.09±0.32, 12 months: 1.07±0.37; mean±SD) (S3 Fig). A severe effacement of the podocyte foot processes was demonstrated by ultrastructural analysis using transmission electron microscopy ( Fig 3B) and scanning electron microscopy (SEM, Fig 3C). Furthermore, quantification of the foot process area confirmed an increased podocyte foot process effacement in 6 months as wells as 12 months old PodoPalld129-/-mice compared to the corresponding controls (Fig 3D). Podocyte-specific loss of palladin does not result in proteinuria or albuminuria measured by dip stick and SDS electrophoresis (S5 Fig).

PLOS ONE
Podocyte-specific knockout of palladin in mice with a 129 genetic background induces a severe phenotype

Palladin knockout affects the expression of actin-binding proteins
Since podocyte morphology is highly dependent on the actin-cytoskeleton with their actinbinding proteins, we analyzed the role of palladin on the expression of essential actin-and

Discussion
Palladin, an essential actin-binding and regulating protein, is ubiquitously expressed in mammals [8,12]. Recently, our group has shown that palladin is specifically expressed in podocytes, a post mitotic cell type in the kidney [22] where it plays an important role in cross-linking and stabilization of actin filaments [8,31]. Additionally, we demonstrated that palladin is essential for a proper 3D morphology of podocytes as well as for the glomerular tuft formation in vivo, especially after the challenge of the mice with nephrotoxic serum (NTS), which is a well-established kidney disease model [23,32].
It is well known that the genetic background of the mouse strain has an important influence on the severeness and development of kidney disease. As it was nicely shown by Steppan and colleagues, mouse strains have distinct vascular properties [33]. They have found that the blood pressure seems to be similar in 129S and C57BL/6 strains, however there are significant differences in vascular properties which might influence the severeness and onset of kidney diseases. Therefore, it is important to study the role of proteins in different and good characterized mouse strains.
Here we studied the influence of palladin not only in the mainly used C57BL/6 mice strain, which is often described to be robust against kidney damage, but also in the often more sensitive 129 genetic mouse strain [24][25][26]. For this purpose, we backcrossed the palladin knockout mice to the 129 background and analyzed the animals at an age of 6 months and 12 months. We have found that PodoPalld129-/-mice were more affected by the palladin knockout compared with PodoPalldBL/6-/-mice. In 6 month old knockout mice with a C57BL/6 background, approximately 20% of glomeruli showed dilated capillaries [23]. In contrast, glomeruli of 6 months old PodoPalld129-/-mice showed a significant increased dilatation of the glomerular tuft (38±9.4%) compared to the control mice (16.5±5.2%; mean±SD, p<0.01) at the same age. Interestingly, the number of glomeruli with dilated capillaries did not increase further in 12 months old PodoPalld129-/-mice, indicating that the glomerular phenotype developed already during the first 6 months. This analysis underlines the importance of the genetic background of animals with a specific knockout because it severely influences the phenotype.
To identify the reason for such a dilatation in PodoPalld129-/-mice, we investigated the presence of mesangial cells. Mesangial cells are contractile cells and play a key role in capillary loop formation [34]. Since there is an interaction and cross-talk between mesangial cells, endothelial cells and podocytes [34], we examined whether the palladin knockout in podocytes also has an indirect effect on mesangial cells. To investigate this we stained kidney sections with an antibody specific for mesangial cells against integrinα8 [35] and found a significant This could also be observed by loss of mesangial cells in early stages of glomerulonephritis/-sclerosis when mesangiolysis and capillary expansion occurred [36,37]. Since we found only a faint staining for this specific integrin in the glomeruli of Podo-Palld129-/-mice in contrast to the control mice at the same age, we hypothesized that here the number of mesangial cells is reduced. Whether this is due to a developmental failure or caused by the loss of mesangial cells after birth is still unknown.
Beside a dilatation of the glomerular tuft, we found that the majority of PodoPalld129-/podocytes developed an enlarged sub-podocyte space as well as cysts compared to the controls of the same strain and to podocytes of the PodoPalldBL/6-/-mice [23]. Ultrastructural analysis of the foot processes by scanning and transmission electron microscopy revealed that nearby 50% of the analyzed PodoPalld129-/-podocytes in mice with an age of 12 months developed such a severe phenotype. Moreover, this phenotype was already described by Kriz and colleagues in a variety of models [38,39]. Based on descriptions in different podocyte-related diseases [40][41][42], our results lead to two assumptions. First, this phenotype could be caused by an imbalanced growth of the glomerular tuft due to a low number of mesangial cells and/or by podocyte hypertrophy to compensate lost podocytes which might be characteristic for this specific strain. Second, it could be that difference is vascular stiffness observed by Steppan and colleagues [33] leads to a higher risk of developing glomerular hypertension, especially already at a young age.
Furthermore, PodoPalld129-/-mice at 12 months of age showed more effaced foot processes compared to 6 months old mice. This finding is in nice agreement with the quantification of podocyte foot process morphology by super resolution microscopy as already described [30]. Here we could show that 6 and 12 months old PodoPalld129-/-mice possess a significant reduction of the filtration slit density (FSD) resulting insignificantly more effaced podocyte foot processes compared to the littermates [30]. In addition, in this study we could show that protein expressions of the slit diaphragm proteins nephrin and podocin were significantly downregulated in PodoPalld129-/-mice compared to corresponding controls.
Beside the morphological findings, we observed autophagosomes in PodoPalld129-/-podocytes which is a hallmark of autophagy, a self-repair mechanism of post mitotic cells [43]. Interestingly, patients suffering from IgA nephropathy and membranous nephropathy, both podocyte-related kidney diseases, showed also an increase of autophagosomes in podocytes [44,45].
Moreover, ultrastructural analysis revealed an increased number of contacts between podocytes and parietal epithelial cells (PECs) in PodoPalld129-/-glomeruli. This is of specific interest since it is known that contacts between podocytes and PECs trigger the formation of tuft adhesions that are first committed lesions for focal segmental glomerulosclerosis. [46][47][48].
The podocyte foot process morphology is highly dependent on an intact actin cytoskeleton. Therefore, we studied the influence of palladin on the actin cytoskeleton in these specific knockout mice. Since it is described that palladin has specific binding-sites for proteins which are involved in actin dynamics and stability like Lasp-1 [14], Pdlim2 [19], ezrin [12], and VASP [13], we investigated the effect of the palladin knockout on the expression of these proteins. Although palladin is known to recruit Lasp-1to actin stress fibres and that the palladin knockdown in HeLa cells resulted in a reduction of Lasp-1 [14,49], we have found that the Lasp-1 mRNA expression was unchanged in PodoPalld129-/-podocytes. However, the phosphorylation of Lasp-1 was significantly reduced. This might influence the stability of the podocyte foot processes, since it was already shown that phosphorylation of Lasp-1 reduces the binding to F-actin in vitro [16,50].
Furthermore, we have detected a significant down-regulation of Pdlim2 in podocytes of PodoPalld129-/-mice. Pdlim2 was already shown to be essential for the stability of the actin cytoskeleton in cultured podocytes as well as was regulated in patients suffering from glomerulopathies [20]. Interestingly, the Pdlim2-interacting partner Amotl1 [20], a protein which regulates the actin dynamic [51], was significantly up-regulated in PodoPalld129-/-glomeruli. Probably, Amotl1 is able to compensate the reduced Pdlim2 expression.
Another important actin-binding protein, VASP, which also regulates actin dynamics, was significantly down-regulated in 6 and 12 months old PodoPalld129-/-mice, indicating that palladin directly influence this protein via the binding site. In contrast, we observed no difference in the expression of ezrin, an actin-binding and plasma membrane cross-linking protein [52], between control and PodoPalld129-/-mice.
Taken together, this study demonstrates that palladin has an impact on the expression of the interacting proteins Pdlim2, pLasp and VASP and is important for capillary tuft formation and podocyte morphology.