Deficiency for scavenger receptors Stabilin‐1 and Stabilin‐2 leads to age‐dependent renal and hepatic depositions of fasciclin domain proteins TGFBI and Periostin in mice

Abstract Stabilin‐1 (Stab1) and Stabilin‐2 (Stab2) are two major scavenger receptors of liver sinusoidal endothelial cells that mediate removal of diverse molecules from the plasma. Double‐knockout mice (Stab‐DKO) develop impaired kidney function and a decreased lifespan, while single Stabilin deficiency or therapeutic inhibition ameliorates atherosclerosis and Stab1‐inhibition is subject of clinical trials in immuno‐oncology. Although POSTN and TFGBI have recently been described as novel Stabilin ligands, the dynamics and functional implications of these ligands have not been comprehensively studied. Immunofluorescence, Western Blotting and Simple Western™ as well as in situ hybridization (RNAScope™) and qRT‐PCR were used to analyze transcription levels and tissue distribution of POSTN and TGFBI in Stab‐KO mice. Stab‐POSTN‐Triple deficient mice were generated to assess kidney and liver fibrosis and function in young and aged mice. TGFBI and POSTN protein accumulated in liver tissue in Stab‐DKO mice and age‐dependent in glomeruli of Stabilin‐deficient mice despite unchanged transcriptional levels. Stab‐POSTN‐Triple KO mice showed glomerulofibrosis and a reduced lifespan comparable to Stab‐DKO mice. However, alterations of the glomerular diameter and vascular density were partially normalized in Stab‐POSTN‐Triple KO. TGFBI and POSTN are Stabilin‐ligands that are deposited in an age‐dependent manner in the kidneys and liver due to insufficient scavenging in the liver. Functionally, POSTN might partially contribute to the observed renal phenotype in Stab‐DKO mice. This study provides details on downstream effects how Stabilin dysfunction affects organ function on a molecular and functional level.


| INTRODUC TI ON
During tissue turnover and regeneration, contents of the extracellular matrix, such as collagens and connective tissue polysaccharides are released into the plasma to a certain extent. Those molecules are cleared and degraded by scavenger receptors expressed by liver sinusoidal endothelial cells (LSECs) (Bhandari et al., 2021). Stabilin-1 (Stab1) and Stabilin-2 (Stab2) are the two members of the class H Scavenger Receptors expressed by LSECs, which contain EGF-like, laminin EGF-like (https://www.unipr ot.org/unipr otkb/Q9NY1 5/ entry) and fasciclin domains (PrabhuDas et al., 2017). They each have distinct, but also common ligands, based on the functionality of several domains (Harris & Cabral, 2019;Patten & Shetty, 2019). For example, hyaluronic acid is only bound by Stab2, while Stab1 is unable to ibind hyaluronic acid, likely due to differences in link domain function (Hansen et al., 2005;Politz et al., 2002). Previously, we showed that Stab1/Stab2-double-knockout mice (Stab-DKO) present an impaired kidney function and a decreased lifespan likely due to their impaired scavenging function through LSECs (Schledzewski et al., 2011). Only recently, we were able to show that single Stabilin inhibition reduces atherosclerosis development (Manta et al., 2022), furthermore, anti-Stab1 antibodies are in initial clinical trials as an immunomodulatory cancer treatment (Virtakoivu et al., 2021).
Mechanistically, we were able to show that extracellular matrix proteins Transforming growth factor, beta-induced, 68 kDa (TGFBI) and Periostin (POSTN) are ligands of Stab1 and Stab2, likely due to fasciclin domain interactions. Their plasma concentration is strongly increased in Stab-DKO mice (Manta et al., 2022). Interestingly, Stab1 and Stab2 as well as POSTN and TGFBI are the only four fasciclindomain containing proteins in mammals (Seifert, 2018). POSTN is implicated as a causative agent in a plethora of fibrotic diseases: Knockout of POSTN in mice protects from CCl 4 -induced liver injury (Kumar et al., 2018) as well as unilateral ischemia-reperfusion injury induced kidney fibrosis . Additionally, it is implicated as a biomarker for disease severity in kidney disease (Guerrot et al., 2012;Satirapoj et al., 2014Satirapoj et al., , 2015Wantanasiri et al., 2015).
Interestingly, aged POSTN-KO C57BL/6J mice show ameliorated kidney fibrosis compared to WT C57BL/6J mice after 24 months, while overall survival is unaffected . TGFBI is described as a modulator of cell-collagen interaction and thus seems likely to contribute to fibrous disease pathways (Billings et al., 2002;Hanssen et al., 2003). As TGFBI and POSTN constitute paralogs it seems probable that both proteins contribute in certain diseases (Schwanekamp et al., 2016). Stab1 and Stab2 single inhibition mediates beneficial effects in atherosclerosis development (Manta et al., 2022), anti-Stab1 inhibition is in clinical trials in immuneoncology (Virtakoivu et al., 2021). We here assess the possible side effects of single and double Stabilin deficiency or inhibition on the scavenging of POSTN and TGFBI from the circulation and their tissue distribution.  Oka et al., 2007) to generate Stab-POSTN-Triple deficient mice (Stab-POSTN-TrKO). Male and female mice were pooled for analysis except in Figure S1. Genotyping was performed as described earlier (Oka et al., 2007;Schledzewski et al., 2011).

| Quantification of staining intensities in IF-stainings of glomeruli
Microscopy images of kidney samples were prepared for analysis by using "Rolling Ball" and "Extended Depth of Focus" from the NIS Elements (Version 5) software. Glomeruli were defined as ROIs by Freehand selection in a merged channel view. After importing all ROIs of an image into the ROI manager, quantification was performed on single channel images. Thresholding was performed using the "Otsu dark" Auto Threshold filter. Data were collected for area of the ROI and percentage of area with positive staining signal and later used to calculate total positive staining area of the glomeruli.

| Simple Western
Simple Western was performed according to manufacturer's protocols (Harris, 2015;Lück et al., 2021). Primary antibodies are found in Table S1.

| Western blotting
Kidney and liver tissue were homogenized in RIPA and were analyzed by SDS-Page and Immuno-blotting on PVDF membranes as by the manufacturer protocol (Trans-Blot Turbo Transfer System, Bio-Rad). Incubation with primary antibodies (Table S1) was performed at 4°C overnight at a 1:1000 dilution in 5% nonfat dry milk (Bio-Rad No. 1706404). Incubation of HRP conjugated secondary antibody (different species, Dianova) was performed for 1 h at room temperature. Chemiluminescence (Millipore WBLUF0500) intensity detection was performed with the Azure ® c400 imaging system (Azure ® Biosystems) and quantified using ImageJ 1.53 (Schneider et al., 2012).

| RNA in situ hybridization
PFA-fixed and paraffinized liver and kidney tissue was sectioned at 4 μm. A modified non-isotopic in situ hybridization protocol was carried out using the RNAscope 2.5 HD Red kit (Advanced Cell Diagnostics) following the manufacturer's recommended protocol with specific probes against the positive control mouse Ppib (Cyclophilin B) gene and mouse POSTN or TGFBI (Table S1).
Sections were ultimately stained with DAB and counterstained with hematoxylin.

| Urinalysis
Animals of the same genotype were arranged in groups of 6-8 and followed for an observation period of up to 2 years. Mice were housed in metabolic cages for 24 h to collect urine samples. Urine was analyzed for Protein and Albumin (Cobas c 311 Analyzer; Roche Diagnostics).

| Statistics
Statistical analyses were performed with JMP® 16 and SAS® 9.4M7 (SAS Institute Inc.). For comparisons between two groups, t test was used when the normality assumption was met using the Shapiro-Wilk test. For group comparisons of three groups, the one-way ANOVA was used when the normality assumption was met using the Shapiro-Wilk test. We used the Brown-Forsythe test to check for equal variances. In case of unequal variances, the Welch ANOVA was used. Tukey-Kramer HSD was used as a post hoc tests for groups of three. For group comparisons of four or more groups, the one-way ANOVA was used when the normality assumption was met using the Shapiro-Wilk test. We used the Brown-Forsythe test to check for equal variances. In case of unequal variances, the Welch ANOVA was used. For post hoc analyses after significant ANOVA p-values, Dunnett's test was used. If the normality assumption was not met, we used a Kruskal-Wallis test followed by Steel's multiple comparison test. Reference groups for Dunnett's test or Steel with control were WT mice. Differences between data sets with p < 0.05 in t test, ANOVA, Kruskal-Wallis test and post hoc tests were considered statistically significant. For comparison of old and young animals, Standard Least Squares with emphasis on Effect leverage was used. Effects of age, genotype and a full factorial analysis of Age*Genotype was performed. Test slices were used to compare different ages in a single genotype.
Amplification data were analyzed using qPCRsoft 4.0.8.0 (Analytik Jena). Normalized expression values were calculated using the Pfaffl method considering amplification efficiency values determined by standard curves.

| Stabilin ligand POSTN is more abundant in aged, enlarged Stab-DKO glomeruli and liver tissue compared to WT
Immunofluorescence revealed an increased abundance of Stabilinligand POSTN in glomeruli from aged (12 months old) Stab-DKO mice compared to aged WT mice ( Figure 1a). Quantification revealed Stab1-KO and Stab2-KO glomeruli did not show significant alterations in POSTN abundance compared to WT (Figure 1b). Stab-DKO glomeruli were significantly larger than WT glomeruli, Stab1-KO and Stab2-KO glomeruli diameter did not reach significance ( Figure 1c).

| POSTN expression is unaltered in Stab-DKO tissues, indicating deposition from plasma
To explore the mechanism of increased POSTN abundance, we performed RNA-ISH for POSTN in glomeruli and liver tissue of aged Stab-DKO mice in comparison to WT mice. Furthermore, we performed qRT-PCR of kidney lysate and liver lysate. Here, no difference in the RNA staining pattern and expression levels in kidney and liver ( Figure 2) were found, indicating no increased transcription and thereby no increased production of POSTN in both organs in Stab-DKO mice, suggesting a deposition due to increased plasma levels.

| Genetic ablation of POSTN does not rescue shortened lifespan, glomerulofibrosis and albuminuria observed in Stab-DKO mice
To assess whether POSTN depositions in Stab-DKO glomeruli and Liver fibrosis levels were scored using Sirius-red quantification, which did not show significant differences in Sirius-red positive areas between Stab-DKO and Stab-POSTN-TrKO mice but were elevated compared to WT in both knockout mice lines ( Figure S2b).

| Stabilin ligand TGFBI is highly abundant in diseased glomeruli and livers from Stabilin-deficient animals
Since genetic deletion of POSTN did not rescue the phenotype of double-Stabilin deficiency, we assessed abundance of another newly Since POSTN was found to be more abundant in Stab-DKO livers, we also analyzed liver tissue for TGFBI abundance. Here, similarly to the kidney, we found a pronounced increase in TGFBI levels only in Stab-DKO livers compared to WT in immunofluorescence (Figure 4d,e) and no increased transcription using RNA-ISH and qRT-PCR (Figure 4f).
To confirm antibody specificity, we performed Simple-Western ( Figure S4a) and Western Blotting ( Figure S4b), which revealed increased TGFBI abundance in kidney and liver lysate of Stab-DKO animals. No overt changes were observed in Stab1-KO und Stab2-KO liver and kidney tissue in comparison to WT mice, although a small but significant increase was found in Stab2-KO livers using simple Western, which could not be confirmed with other methods.

| TGFBI depositions occur independently of POSTN in Stab-DKO tissue
Previously, we have demonstrated that POSTN and TGFBI are both

| DISCUSS ION
We here show that scavenging by Stab1 and Stab2 is crucial not only for homeostasis of TGFBI and POSTN plasma levels in mice (Manta et al., 2022), but also to prevent age-dependent accumulation of TGFBI and POSTN in liver and kidney tissue. We hypothesize that POSTN and TGFBI are not produced locally in Stab-DKO livers and kidneys, but are deposited in these organs, possibly due to their high concentrations in Stab-DKO plasma, since their scavenger receptors are missing. This hypothesis is supported by constant mRNA levels in the liver and kidney as shown here by RNA-ISH and qRT-PCR (Figures 2 and 4f). As we have shown previously, transcriptomic data from isolated LSECs show a slight but significant downregulation of POSTN in LSECs from Stabilin deficient animals compared to WT animals (Olsavszky et al., 2021). POSTN and TGFBI staining colocalized in liver and glomeruli and was largely confined to the midzonal, perisinusoidal area in the liver, with a similar pattern compared to Sirius-red positive areas (Figure 5c, Figure S3). We here observed partial co-localization with Collagen 1 staining in glomeruli and liver tissue, indicating a similar, but not identical pattern ( Figure S3).
In mice and humans, there are somewhat contradictory findings regarding age-dependent plasma levels of TGFBI and POSTN.
In mice, it was described that TGFBI and POSTN plasma levels are decreased in aged mice (Yang et al., 2020), while human data show an increase in aged populations (Lehallier et al., 2019). Since aging in mice and humans differ considerably, we can only speculate on the biological relevance of decreased TGFBI and POSTN plasma abundance in aged WT mice. Since we were able to find increased deposition of both Stabilin ligands in glomeruli of aged knockout mice compared to younger mice, we could further underline the hypothesis of a time-dependent deposition from plasma in our models.
Interestingly, ablation of POSTN in Stab-DKO mice did not res- These findings bear consequences for the use of therapeutical inhibition of Stabilins, which for Stab1 is already in Phase 2 clinical trials in cancer (Virtakoivu et al., 2021) and might be a treatment option for atherosclerosis in future studies (Manta et al., 2022). Single deficiency for Stab1 or Stab2 alone did not seem to induce strong damage to kidney function, although a mild increase in TGFBI depositions was observed in young, but not aged Stab2-KO glomeruli. It can be assumed that absence of one of the  and Stab2-KO as well (Olsavszky et al., 2021). Nevertheless, side effects of anti-Stabilin targeted therapies in individuals with decreased renal or liver function cannot be ruled out and should be monitored closely.
To conclude, we were able to further validate our previous findings that the four fasciclin proteins Stab1, Stab2, POSTN, and TGFBI constitute a ligand-scavenger receptor system without any relevant compensation by other scavenger receptors. While targeted inhibition of Stab1 or Stab2 might be beneficial in the context of atherosclerosis prevention, double deficiency leads to premature organ failure likely due to multi-ligand depositions caused by deficient hepatic scavenging.
were responsible for analysis and interpretation of data; T.L., A.R. and C.G. were involved in drafting the manuscript and revising it critically for important intellectual content.

ACK N OWLED G M ENTS
We thank Hiltrud Schönhaber, Maria Muciek, Camela Jost, Stefanie Riester, and Janina Ritz for excellent technical support. Open Access funding enabled and organized by Projekt DEAL. Graphical Abstract: Created with BioRender.com.

FU N D I N G I N FO R M ATI O N
The authors gratefully acknowledge the data storage service SDS@

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors have declared that no conflict of interest exists.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.