Glomage : A multimodal platform for high-content morphological and RNA profiling of glomeruli in zebrafish and mouse models

16 Chronic kidney disease affects over 800 million people worldwide, but therapeutic options 17 remain limited. The glomerulus, with its highly specialized podocytes, forms the functional 18 core of blood filtration. However, studying this complex three -dimensional kidney 19 compartment is difficult, which has slowed progress in identifying drugs. Zebrafish larvae 20 are a powerful model for kidney research and drug discovery, but histological staining of 21 individual larvae is time-consuming, and scalable methods for glome rular analysis have 22 been lacking. Here, we establish a high -throughput workflow for isolating, staining, 23 imaging, as well as extracting RNA from hundreds of zebrafish glomeruli simultaneously. 24 The workflow preserves the three -dimensional architecture, enab les quantitative 25 assessment of podocyte numbers, and was validated in a podocyte-specific injury model. 26 Moreover, we adapted the protocol for mouse glomeruli, allowing reliable detection of the 27 age-related loss of podocytes. Glomage provides a systematic, cross-species platform 28 for glomerular analysis, opening new opportunities for kidney research and drug 29 discovery. 30 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 3

31 Chronic kidney disease affects an estimated 10% of the global population and is a leading 32 cause of morbidity and mortality worldwide.1 Its prevalence continues to rise due to aging 33 populations, hypertension, and diabetes, highlighting the urgent need to understand how 34 morphological and gene expres sion changes impact kidney function .2 The glomeruli, 35 highly specialized filtration units of the kidney, consist of parietal epithelial cells, 36 mesangial cells, fenestrated capillary endothelia, the glomerular basement membrane, 37 and morphologically complex podocytes that together ensure effective blood filtration. 38 Podocytes represent a unique population of postmitotic epithelial cells, that envelop the 39 glomerular capillaries with interdigitating foot processes. By forming the slit diaphragm 40 and expressing key proteins such as nephrin, together with transcription factors like 41 Dach1, they play a pivotal role in maintaining the size - and charge selectivity of the 42 glomerular filtration barrier .3–5 Despite its critical role, the glomerular structure remains 43 difficult to study in depth. Traditional histological approaches using paraffin - or 44 cryosections are indispensable for assessing morphological changes and protein 45 expression, yet they provide only two -dimensional snapshots of a three -dimensional 46 structure. Moreover, loss of podocytes, whether through apoptosis or detachment, is 47 irreversible and is a major driver of disease progression, making podocyte number a 48 crucial indicator of kidney health and a strong predictor of renal outcomes. However, 49 conventional methods may under - or overestimate podocyte counts due to inherent 50 sampling bias.6–8 In contrast to classical metho ds, advanced imaging approaches such 51 as optical clearing combined with confocal or light-sheet microscopy have enabled whole-52 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 4 glomerulus visualization but are time -intensive, technically demanding, and require 53 specialized equipment, limiting their scalabili ty.9–11 These challenges have slowed 54 progress towards standardized, high-throughput and deep glomerular analysis. 55 Zebrafish larvae have emerged as a versatile vertebrate model for kidney biology and 56 disease due to their rapid development, optical transparency, and genetic accessibility.12–57 14 Furthermore, approximately 70% of human genes have at least one zebrafish ortholog, 58 emphasizing its utility as a model for human disease .15 The larval zebrafish pronephros 59 closely mimics the architecture and filtration function of the mammalian glomerulus, 60 making it a powerful model for studying e.g. podocytopathies and to identify drugs to treat 61 kidney disease.16–22 62 The zebrafish pronephros consists of only a single glomerulus attached to a pair of 63 tubules. While tissue sections and whole-mount imaging of single zebrafish larvae provide 64 valuable insights, they are not yet suitable for high-throughput applications.23 Molecular 65 studies face similar challenges: isolating sufficient, pure glomerular RN A from zebrafish 66 larvae has been technically difficult and time-consuming, so most transcriptomic analyses 67 rely on whole -larva samples, masking glomerular or podocyte -specific expression 68 patterns.24,25 69 To address these limitations, we developed a protocol for batch isolation of intact and 70 whole larval zebrafish glomeruli, enabling fast and scalable immunofluorescence, 3D-71 high-resolution imaging, an unbiased and automated podocyte count as well as 72 downstream molecular analyses. This protocol was applied in a larval model of focal 73 segmental glomerulosclerosis (FSGS) , where it effectively identified structural damage, 74 enabled a reliant quantification of podocyte loss and detection of volumetric changes. The 75 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 5 same staining and quantification protocol was successfully applied to isolated mouse 76 glomeruli, confirming the age-dependent podocyte loss. 77 In summary, this protocol enables scalable isolation and analysis of whole glomeruli 78 across animal and disease models, supporting advances in glomeru lar research, 79 biomarker discovery, and drug screening. 80 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 6

81 Glomage: High-throughput isolation of larval zebrafish glomeruli 82 To enable high -throughput analysis of pronephric zebrafish glomeruli, we developed a 83 protocol for their batch isolation and immunofluorescent labeling. For this purpose, we 84 generated a nphs2:eGFP (Podo:GFP) zebrafish line, which expresses enhanced green 85 fluorescence protein (eGFP) endogenously and exclusively in glomerular podocytes (Fig. 86 S1). The first step of the protocol is the fixation of the larvae at 5 days post fertilization 87 (dpf) in 2% paraformaldehyde (PFA) for 90 min at room temperature (RT) , which is a 88 critical step preceding tissue dissociation. The procedure was carried out using 100 fixed 89 eGFP-positive larvae as the standard sample size . The larvae were placed in a tube 90 together with ~100 µl ceramic beads with a size of 1 mm and were dissociated with a 91 Fast-Prep-24TM device for 13 seconds with 4 m/s , resulting in the release of glomeruli 92 (Fig. 1A). The efficiency of dissociation is highly dependent on the fixation parameters 93 such as PFA concentration, incubation time and temperature as wel l as dissociation 94 intensity and time . Insufficient fixation or excessive dissociation le ad to glomerular 95 damage, compromising tissue integrity . Moreover, extended fixation times can result in 96 reduced glomerular yield and in a loss of antigenicity of native tissue as it was also 97 described in the literature .26,27 To prevent adhesion of isolated glomeruli to plastic 98 surfaces, all materials were precoated with fetal bovine serum (FBS) since glomeruli are 99 highly charged.28,29 Following larval dissociation, intact glomeruli were collected under a 100 fluorescence stereomicroscope using a gel-loading pipette and subsequently transferred 101 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 7 102 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 8 onto a cell culture insert with a polyethylene terephthalate (PET) membrane 103 (mesh size: 8 µm, Greiner ThinCert®). The inserts were transferred into a 24-well plate, 104 in which the immunofluorescence staining was performed. The staining procedure was 105 based on previously established protoc ols with minor modifications, 16,30,31 which are 106 described in detail in the methods section. The minimum volume of each (antibody) 107 solution was 500 µl per well to ensure that the PET membrane carrying glomeruli 108 remained fully covered. At the end of the staining procedure, the insert membrane was 109 excised with a scalpel and mounted o nto a glass slide for microscopy (Fig. 1A). For 110 automated imaging, low-magnification overview maps (1.25x objective) were acquired to 111 facilitate navigation to individual glomeruli with higher -magnification objectives (20x or 112 60x, Fig. 1B-D). 113 At the end of the protocol, 3 sets of 100 larvae were used as input. Manual counting of 114 glomeruli on PET membranes resulted in 68 (SD: 2) isolated, stained, and mounted whole 115 glomeruli. Given that zebrafish larvae only possess a single glomerulus, this corresponds 116 to a 68% recovery rate. 117 Figure 1: Batch isolation of whole zebrafish glomeruli. PFA-fixed Podo:GFP larvae are dissociated using ceramic beads, and fluorescent, intact glomeruli are collected and transferred to mesh baskets. Immunofluorescence staining is performed in a 24-well plate, and the cut mesh containing the stained glomeruli is mounted onto a glass slide (A). An overview image acquired with a 1.25x objective shows the mesh in brightfield and the fluorescent glomeruli ( B). The corresponding eGFP-only image illustrates all glomeruli (C). A higher magnification using a 20x

reveals intact, peanut-shaped larval zebrafish glomeruli ( D). Using 100 larvae per isolation resulted in a mean of 68 (SD: 2, n=3) isolated, stained, and mounted whole zebrafish glomeruli per round (E). Scale bars: 1 mm (B, C), 100 µm (D). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 9 Glomage: A platform for 3D immunofluorescence analysis of larval glomeruli 118 For antibody staining, we consistently used antibodies and working concentrations that 119 had already been validated and optimized in established protocols for zebra fish larval 120 cryosections (Fig. S2). They were therefore expected to perform similarly in native, non–121 paraffin-embedded whole glomeruli. 122 To demonstrate the applicability of this protocol, we used antibodies against podocyte -123 specific proteins, such as the slit diaphragm proteins nephrin and podocin; the 124 transcription factor Pax2a, which is specific for parietal epithelial cells (PECs) and 125 proximal tubule epithelial cells (PTECs) ; Ehd3, a marker of glomerular endothelial cells; 126 and laminin, a major component of the glomerular basement membrane (GBM). Positive 127 antibody signals alongside endogenous eGFP expression and nuclei can be seen in 128 single optical sections of the stained glomeruli (Fig. 2A-E, Movie S1). 129 For each glomerulus, optical z-stacks were acquired at 1 µm intervals, yielding a 130 volumetric range of 2 5–60 µm, followed by maximum intensity projections (MIPs) . This 131 approach enabled 3D reconstruction of the whole slit diaphragm architecture in larval 132 zebrafish glomeruli (Figure 2A′, B′ , Movie S2 and S3 ). This procedure also revealed the 133 localization of Pax2a-positive PECs surrounding the glomerulus . Pax2a also labelled 134 PTECs of the two attached proximal tubules, which in some cases remained connected 135 to the glomeruli (Fig. 2C′). In addition, the staining visualized Ehd3-positive, fenestrated 136 endothelial cells within the glomeruli consistent with their known expression pattern (Fig. 137 2D′).32 Additionally, labeling of F -actin with fluorescence -conjugated phalloidin showed 138 the brush border of the convoluted proximal tubule in 3D, including the short neck 139 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 10 140 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 11 segment which connects the glomerulus to the tubule. The neck segment was 141 characterized by reduced apical microvillus density (Fig. S3, Movie S4). 142 Taken together, all five antibodies produced specific and reproducible staining patterns 143 in whole-mount glomeruli, demonstrating that this protocol enables, for the first time, rapid 144 and efficient immunolabeling of multiple glomerular ta rget structures across large 145 numbers of intact whole glomeruli. 146 Glomage: Rapid quantification of total podocyte number in whole larval glomeruli 147 Because podocyte number changes with both disease progression and organismal age, 148 precise quantification of these postmitotic cells is essential. A reduction below a critical 149 threshold compromises glomerular and renal function, ultimately resulting in proteinuria, 150 the leakage of high-molecular-weight proteins into the urine.33 To establish this analysis, 151 z-stacks of isolated glomeruli from Podo:GFP larvae were analyzed. Since eGFP is a 152 protein of less than 60 kDa an d can therefore diffuse into the nucleus, podocyte nuclei 153 can be readily identified by their exclusive nuclear labelling with eGFP (Fig. 3A-D). To 154 address the challenge of three-dimensional nuclear segmentation, particularly the risk of 155 counting the same n ucleus multiple times across optical slice s, an AI-based image 156 analysis software was used to segment and quantify eGFP-positive nuclei. By applying 157 Figure 2: Visualization of whole zebrafish glomeruli using different antibodies in the Podo:GFP strain. Antibodies against the slit diaphragm proteins nephrin and podocin show a continuous, linear staining along the filtration slits in single optical sections ( A, B) and in corresponding maximum intensity projections (3D, A′, B′). Staining with a Pax2a antibody reveals the coverage of the glomerulus by nuclei of parietal epithelial cells (arrows), and identifies two populations of proximal tubule cells connected to the glomerulus (arrowheads, C, C′). An Ehd3 antibody specifically labels fenestrated endothelial cells ( D, D′), while the laminin staining highlights the basement membranes ( E, E′). Podocytes are endogenously labeled with eGFP (green), nuclei (blue) are displayed in the 2D panels and omitted in the 3D merge panels. Scale bars: 10 µm. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 12 158 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 13 an eGFP intensity threshold after segmentation of all nuclei within and surrounding each 159 glomerulus, it was possible to reliably determine the number of podocytes per glomerulus 160 from full z-stacks. Analysis of 18 glomeruli revealed an average of 88.40 podocytes per 161 glomerulus (SD: 13.23) in zebrafish larvae at 5 dpf (Fig. 3E-H). 162 Therefore, Glomage allows for the first time the robust and scalable quantification of 163 podocyte number in transgenic zebrafish larvae without the use of antibodies. Accidental 164 co-isolation of adjacent tissue do es not interfere with the analysis, as only podocytes 165 express eGFP in this specific zebrafish strain. 166 Glomage: Podocyte quantification in podocyte injury models 167 To evaluate the functional applicability of this novel podocyte quantification method, we 168 used the nphs2:NTR-mCherry (Cherry) transgenic zebrafish strain, which expresses the 169 bacterial enzyme nitroreductase (NTR) and mCherry exclusively in podocytes under the 170 control of the podocin promoter (Fig. S1).34 Upon exposure of zebrafish larvae to 50 nM 171 nifurpirinol (NFP) in the tank water at 4 dpf, podocyte injury develops as NFP is converted 172 into a toxin by the NTR within podocytes, mimicking key features of mammalian 173 FSGS.30,31,35,36 Since mCherry fluorescence in podocytes was markedly reduced after 24 174 hours of NFP treatment, experiments were terminated at this time point to avoid complete 175 podocyte loss.16 We analyzed three groups: (i) a control group treated with 0.2% dimethyl 176 Figure 3: Establishment of the total podocyte count. Confocal image stacks of isolated zebrafish glomeruli show eGFP-positive podocyte nuclei in individual optical sections (A–C) and in the corresponding projection (D). AI-based segmentation visualizes the podocyte cytoplasm (green), eGFP-positive nuclei (blue), and eGFP -negative nuclei (red) ( E). Panel F shows the spatial distribution of eGFP -positive (blue) and eGFP -negative (red) nuclei within the glomerulus. All eGFP-positive nuclei are used to define the total number of podocytes in a single glomerulus (G). Podocyte nuclear volume is color-coded, ranging from 6.7 µm³ (purple/blue) to 83.3 µm³ (red). Automated analysis of n=18 zebrafish glomeruli at 5 days post fertilization (dpf) revealed an average of 88.40 podocytes per glomerulus (SD: 13.23) (H). Scale bars: 10 µm. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 14 sulfoxide (DMSO), (ii) an injury group treated with 50 nM NFP, and (iii) a group co-treated 177 with NFP and the small molecule NM_187 to prevent disease progression. After 24 hours, 178 the Glomage procedure was performed, and glomeruli were stained for the slit diaphragm 179 protein podocin. 3D-image analysis of healthy control glomeruli displayed a continuous, 180 meandering pattern of podocin along the glomerular filtration barrier and a homogeneous 181 mCherry fluorescence in podocytes (Fig. 4A). In contrast, NFP treatment induced clear 182 morphological signs of podocyte injury, including a punctate slit diaphragm staining 183 pattern, pseudocyst formation, intracellular vacuoles, and a fragmented or 184 inhomogeneous mCherry signal (Fig. 4B). Co-treatment with NM_187 fully rescued these 185 features: Podocin staining returned to a linear pattern, and mCherry fluorescence was 186 homogeneously distributed (Fig. 4C). Our automated quantification pipeline was 187 successfully adapted to work with mCherr y instead of eGFP as a nuclear marker, 188 confirming its fluorophore-independence. Analysis of podocyte nuclear volume revealed 189 that NFP treatment significantly reduced the average nuclear volume (mean: 33.44 µm³, 190 SD: 2.4) compared to controls ( mean: 48.46 µm³, SD: 2.82). Co-treatment of the larvae 191 with NM_187 significantly restored nuclear volume ( mean: 39.81, SD: 1.84 , Fig. 4D). A 192 deeper analysis of individual optical slices revealed abundant small mCherry-positive 193 nuclear fragments in the NFP group, represe nting apoptotic bodies. These were absent 194 in the control and NM_187 co-treatment groups (Fig. S4). To exclude apoptotic fragments 195 from the unbiased podocyte count, a volume threshold was set at one standard deviation 196 below the mean nuclear volume of the co ntrol group. The same threshold was applied 197 consistently across all groups and all nuclei exceeding this value were defined as nuclei 198 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 15 from viable podocytes. After applying th e threshold, podocyte counts in healthy Cherry 199 larvae (mean: 95.6; SD: 12.90) closely matched those in Podo:GFP larvae (mean: 88.40, 200 SD: 1 3.23, Fig. 3H ). NFP treatment resulted in a significant loss of viable podocytes 201 Figure 4: Histology and podocyte counts in healthy, NFP -treated, and NFP + NM_187 - treated Cherry larvae. Healthy control larvae expressing mCherry and the nitroreductase (NTR) in podocytes show a linear Podocin staining pattern along the slit diaphragm ( A). Targeted podocyte injury with NFP for 24 hours induces structural damage and disruption of the slit diaphragm (B). Treatment with the compound NM_187 prevents podocyte injury and preserves both podocyte integrity and the linear Podocin expression pattern ( C). Podocyte injury significantly reduces nuclear volume, which is rescued by NM_187 treatment (DMSO me an: 48.46 µm³, SD: 2.82, NFP mean: 33.44 µm³, SD: 2.4), NFP + NM_187 mean: 39.81, SD: 1.84, D). Automated AI -based segmentation and quantification reveal a significant reduction in the number of viable podocytes per glomerulus following specific injury (me an: 48.2; SD: 12.14), compared to healthy DMSO -treated controls (mean: 95.6; SD: 12.9). Treatment with NM_187 completely prevents podocyte loss (mean: 95.3; SD: 19.72) (E). Statistical analysis was performed with n=10 glomeruli in each group using one-way ANOVA followed by Tukey’s multiple comparisons test and p≤0.05 was considered statistically significant. Scale bars: 10 µm. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 16 (mean: 48.2, SD: 12.14), whereas co -treatment with NM_187 fully prevented podocyte 202 loss, with final counts comparable to healthy controls (mean: 95.30, SD: 19.72, Fig. 4E). 203 This is the first demonstration that differences in podocyte number in zebrafish larvae can 204 be determined in injury and in response to treatment with a drug that inhibits the 205 development of larval FSGS, highlighting the potential of this approach for drug screening 206 experiments and toxicity assessments. 207 Glomage: Volumetric analysis of healthy and injured larval glomeruli 208 To further assess the morphological impact of NFP and the potential protective effect of 209 NM_187 on podocyte structure, we performed volumetric analysis of glomerular podocyte 210 cytoplasm in whole glomeruli of Cherry larvae. Confocal z -stacks were used to render 211 three-dimensional, contiguous volumes engulfed by mCherry -positive podocyte 212 structures (Fig. 5 A-C´). Quantitative analysis revealed a significant reduction in 213 glomerular mCherry volume in NFP -treated larvae (m ean: 26,4 1 µm³, SD: 4,8 9) 214 compared to controls (mean: 38,05 µm³, SD: 4,00). Notably, co -treatment with NM_187 215 significantly restored podocyte volume to near-control levels (mean: 35,58 µm³, SD: 8,07; 216 Fig. 5D). In addition to volume changes, the sphericity of glomeruli was assessed in which 217 a value of 1 describes a perfect sphere. The sphericity of the mCherry-positive volume 218 was significantly reduced in both NFP (mean: 0.44, SD: 0.05) and NFP + NM_187-treated 219 larvae (mean: 0.51, SD: 0.07) relative to contro ls (mean: 0.60, SD: 0.08; Fig. 5E ), 220 indicating a loss of normal glomerular morphology that was only partially rescued by 221 NM_187. 222 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 17 223 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 18 Glomage: Isolation of pure glomerular RNA 224 In zebrafish larvae, isolating single organs in large numbers within a short time frame and 225 with sufficient purity remains challenging. As a result, most studies rely on whole -larva 226 analysis. To expand the utility of our batch glomerulus isolation protocol beyond imaging 227 applications, we developed a modified workflow to obtain pure glomerular mRNA. In this 228 approach, larvae from the Cherry strain were dissociated without PFA fixation under mild 229 conditions (1 s with 4 m/s) , and glomeruli were manually selected under fluorescence 230 guidance. Several wash steps and stringent visual quality control were applied to 231 eliminate adjacent or contaminating cells, which are critical for RNA purity. RNA was 232 extracted from 50 or 100 isolated glomeruli using a single-cell RNA isolation kit, followed 233 by complementary DNA (cDNA) synthesis and RT-PCR as well as RT -qPCR analysis 234 (Fig. 6A). Gene expression was assessed using primers for eef1a1a (housekeeping 235 gene), nphs2 (podocyte-specific), and mCherry (transgene, podocyte -specific). To 236 assess sample purity, we included markers specific to other organs: gnat1 (eye-specific), 237 myl7 (heart-specific), and ctrb1 (pancreas-specific). 238 Figure 5: Volumetric depiction and analysis of DMSO, NFP and NFP + NM_187 -treated Cherry larvae. The continuous mCherry-positive podocyte cytoplasm in glomeruli from confocal z-stacks ( A–C) was used to render a three -dimensional, contiguous volume ( A′–C′). Quantification revealed a significant reduction in mCherry-positive volume in NFP-treated larvae (mean: 26,406 µm³, SD: 4,889) compared to DMSO controls (mean: 38,053 µm³, SD: 4,004). Co-treatment with NM_187 significantly restored mCherry volume to near-control levels (mean: 35,580 µm³, SD: 8,070; D). Glomerular sphericity analysis showed a significant reduction in roundness in both NFP (mean: 0.44, SD: 0.05) and NFP + NM_187-treated larvae (mean: 0.51, SD: 0.07) compared to controls (mean: 0.60, SD: 0.08, E). For each group, n=10 whole glomeruli were analyzed. Statistical analysis was performed using one -way ANOVA followed by Tukey’s multiple comparisons test. p≤0.05 was considered statistically significant. Scale bars: 10 µm. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 19 In conventional RT-PCR (Fig. 6B, Fig. S5), the non-glomerular markers gnat1, myl7, and 239 ctrb1 were not detectable in isolated glomerular RNA, whereas they were readily 240 Figure 6: Quantitative glomerular RNA isolation and downstream application in podocyte injury. Unfixed Cherry larvae were dissociated and glomerular RNA was isolated using a single- cell RNA kit at 5 dpf. Subsequent cDNA synthesis and RT -PCR as well as RT -qPCR was performed using podocyte and non -podocyte specific primers ( A). Primers detecting nphs2 (podocin, podocyte-specific), mCherry (podocyte-specific) and gnat1 (eye), myl7 (heart) ctrb1 (pancreas) were used to assess sample purity. cDNA from whole larvae were compared to cDNA derived from 100 isolated glomeruli. RT-PCR results show absence of non-glomerular markers (gnat1, myl7, ctrb1) in isolated glomeruli, while nphs2 is strongly enriched compared whole-larva levels ( B). RT -qPCR shows that nphs2 levels are substantially higher in isolated glomeruli, showing a 68.9 -fold change (SD: 1.1) in 50 glomeruli and 11 48.8-fold (SD: 256.7) in 100 glomeruli compared to whole larvae (1.0-fold, SD: 0.03, C). In NFP-induced injury, nphs2 (0.22- fold, SD: 0.06) and mCherry (0.06-fold,SD: 0.00) are downregulated. Co-treatment with NM_187 significantly restores expression (nphs2: 0.69-fold, SD: 0.29, mCherry: 0.70-fold, SD: 0.19, D). For RT -qPCR experiments, three independent isolations were compared (n=3). Statistical analysis was conducted by using two-way ANOVA followed by Bonferroni-post-hoc test. p≤0.05 was considered statistically significant. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 20 amplified from whole-larva samples. This confirms that the isolation workflow yields highly 241 pure glomerular material without detectable contamination from other organs. In contrast, 242 the podocyte-specific marker nphs2 was robustly expressed in glomerular samples and 243 only slightly expressed in whole larvae samples, validating the specificity of the isolation 244 procedure (Fig. 6B, Fig. S5). 245 RT-qPCR analysis demonstrated a strong enrichment of nphs2 transcripts in isolated 246 glomeruli compared to whole larvae. The mean relative nphs2 expression in 50 isolated 247 glomeruli was 68.9-fold (SD: 1.1), and in 100 glomeruli 1148.8-fold (SD: 256.7), whereas 248 whole-larva samples served as baseline (1.00 -fold, SD: 0.03, Fig. 6C). These results 249 indicate a more than thousand-fold enrichment of podocyte-specific transcripts in isolated 250 glomeruli while maintaining high RNA purity and absence of non -glomerular gene 251 expression. 252 To further validate the approach in a functional context, nphs2 and mCherry expression 253 were analyzed in a podocyte injury model. Larvae were treated with 0.2% DMSO, 50 nM 254 NFP, or 50 nM NFP and NM_187. Upon NFP treatment, nphs2 levels dropped markedly 255 to 0.22-fold (SD: 0.06), and mCherry decreased to 0.06-fold (SD: 0.00), consistent with 256 severe podocyte injury and loss of podocyte -specific gene expression compared to the 257 DMSO control. Co-treatment with NFP and NM_187 significantly restored expression of 258 nphs2 to 0.69-fold (SD: 0.29) and mCherry to 0.70-fold (SD: 0.19), indicating a protective 259 or restorative effect of NM_187 on podocyte integrity and transcriptional activity (Fig. 6D). 260 Together, these results demonstrate that our workflow additionally enables high-purity 261 RNA extraction suitable for sensitive transcriptional analyses and that it can reliably detect 262 transcriptional responses in podocyte injury models and treatments. 263 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 21 Glomage: Application of batch staining to isolated mouse glomeruli 264 After establishing the staining, imaging, and quantification pipeline in zebrafish larvae, we 265 next adapted the method to isolated mammalian glomeruli. To this end, glomeruli were 266 isolated from nephrin:CFP mice, expressing cyan fluorescence protein (CFP) exclusively 267 in podocytes,37 by using a sieving and differential adhesion approach which is already 268 published.38 Shortly after isolation, glomeruli were fixed in PFA, and their density in 269 suspension was assessed under a stereomicroscope. To avoid overloading the culture 270 inserts and especially the PET membrane, a volume corresponding to approximately 50–271 100 glom eruli was transferred into each insert. The staining and imaging protocol 272 developed for zebrafish glomeruli was then applied to mouse glomeruli (Fig. 7 A). Low-273 magnification images of the cut PET membrane revealed a high density (>50) of intact, 274 CFP-positive glomeruli after staining (Fig. 7B, C). Higher magnification MIPs 275 demonstrated successful immunolabeling with an antibody against the slit diaphragm 276 protein nephrin, while CFP fluorescence in podocyte cytosol was well preserved (Fig. 7D, 277 E). Single optical slices from the z-stacks further confirmed tissue integrity, with podocyte 278 foot processes clearly identifiable in the CFP channel and the nephrin signal localized 279 between them (Fig. 7F-I). 280 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 22 Together, these findings demonstrate that the Glomage batch-staining and imaging 281 pipeline can be directly transferred to mammalian glomeruli. 282 Figure 7: Isolation and staining of whole mouse glomeruli. Glomeruli from n ephrin:CFP reporter mice were isolated using a sieve -based protocol combined with differential adhesion, followed by fixation. Fixed glomeruli were transferred to mesh baskets and subjected to immunofluorescence staining in a 24-well plate. The stained mesh was cut and mounted onto a glass slide (A). A confocal overview acquired with a 1.25x objective shows the mesh ( B) with CFP-positive glomeruli (cyan, C). Projections of stacks acquired with a 60x objective reveal podocyte cytoplasm (cyan, D) and immunostaining for the slit diaphragm protein Nephrin (magenta). Nuclei are shown in yellow ( E). A single optical section demonstrates the linear Nephrin staining pattern (magenta, F) and podocytes in cyan ( G, H). Higher magnification highlights the preserved ultrastructure of the tissue, including resolved podocyte foot processes by confocal microscopy (I). Scale bars: 1 mm (B, C), 10 µm (D–H), 3 µm (I). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 23 Glomage: Total podocyte count in glomeruli of young and aged mice 283 As a proof of concept, we applied our total podocyte quantification pipeline to isolated, 284 stained glomeruli from 6 -month-old (young) and 18 -month-old (aged) mice. Podocyte 285 nuclei were identified using an antibody against the transcription factor Dach1, which is 286 highly expressed in adult podocytes, while the slit diaphragm was labeled with an 287 antibody against nephrin in z-stacks of whole glomeruli (Fig. 8A). MIPs of glomeruli from 288 young mice showed continuous nephrin coverage of the glomerular filtration barrier and 289 a high number of Dach1-positive podocyte nuclei (Fig. 8B). In contrast, aged glomeruli 290 exhibited two distinct pathological phenotypes: (1) a mild phenotype characterized by 291 partial nephrin coverage with a punctate staining pattern and a reduced number of Dach1-292 positive podocytes (Fig. 8C); (2) a severe phenotype with a nearly complete loss of the 293 nephrin signal as well as a striking reduction or complete absence of Dach1 -positive 294 podocytes, respectively (Fig. 8D). In severely affected glomeruli, the nephrin staining was 295 essential to confirm the identification of actual glomeruli, as Dach1-positive nuclei were 296 entirely absent in some cases. Unbiased quantification of Dach1 -positive cells from 30 297 glomeruli (three young and three aged mice) revealed a significant reduction in 298 differentiated Dach1-positive podocytes in 18-month-old mice compared to 6 -month-old 299 controls (Fig. 8E). 300 301 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 24 302 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 25 These results demonstrate that the Glomage podocyte count can be effectively 303 transferred to mammalian glomeruli, enabling rapid and reliable determination of total 304 podocyte number. 305 Figure 8: Immunofluorescence of whole glomeruli and podocyte counts in young and aged mice. Glomeruli from 6 -month-old (young) and 18 -month-old (aged) mice were isolated and stained for Nephrin and Dach1 using the batch immunofluorescence protocol ( A). MIPs of glomeruli from young mice showed a continuous, linear Nephrin pattern (yellow) and a high density of Dach1 -positive podocyte nuclei (magenta, B). Glomeruli form aged mice exhibited either a mild phenotype, with partially disrupted Nephrin labeling and a reduced podocyte density (C), or a severe phenotype, with marked structural deterioration and near -complete loss of Nephrin and Dach1 signals ( D). Quantification of Dach1 -positive nuclei revealed a significant reduction in podocyte number in aged mice ( mean: 21.4, SD: 16.2, n=28) compared to young controls (mean: 74.6, SD: 15.2, n=29 ( E)). Data were analyzed by a two -tailed Mann-Whitney test; p ≤ 0.05 was considered significant. Scale bars: 20 µm. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 26

306 Podocytes are crucial cells for maintaining the integrity of the filtration barrier in the 307 kidney.39 Damage or loss of podocytes leads to a breakdown of this barrier, allowing high 308 molecular weight proteins such as albumin to pass into the urine – a life-shortening and 309 life-threating situation.40 In approximately 75% of patients with chronic kidney disease, 310 postmitotic podocytes are affected, underscoring their critical role and vulnerability.41 This 311 circumstance makes it necessary to explore new approaches in drug discovery and 312 analysis. Zebrafish larvae play an important role in this process, as they represent an 313 outstanding model that enables rapid and reliable investigation of kidney development, 314 physiological function, and drug identification.42–44 Additionally, the zebrafish is becoming 315 an increasingly importan t model, as there is a global trend away from mammalian 316 experiments toward simpler organisms and organoids. 317 Since zebrafish have a high reproduction rate and the larvae are optical transparent, they 318 are ideally suited for in vivo high-throughput screening approaches.16,45,46 Nevertheless, 319 the subsequent histological analysis of z ebrafish larvae remains a significant and time -320 consuming challenge that has not yet been solved until now. 321 In the present study, we introduce Glomage as a novel high-throughput histological 322 technique designed to accelerate and improve glomerular research. The method is based 323 on batch staining of isolated zebrafish glomeruli , and we further demonstrate that it can 324 also be applied to mice, making it a highly versatile. 325 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 27 Currently, entire zebrafish larvae, which are only a few millimeters in length, are 326 processed individually for histological analysis following experimental procedures. Due to 327 the loss of endogenous fluorophores during paraffin embedding, transgenic larvae are 328 preferably prepared as cryosections, a technically demanding and time -consuming 329 process. 330 Furthermore, each section must be cut precisely through the single glomerulus of each 331 larva. To overcome these limitations, we developed a method , which allows for a 332 significantly faster and more efficient histological workflow. 333 The Glomage procedure reliably releases whole intact glomeruli from a large number of 334 pre-fixed larvae in a short period of time, it takes only one hour for 100 larvae. In contrast, 335 a skilled operator embeds, orientates and generates cryosections from four larvae in one 336 hour. Therefore, the Glomage sample preparation is 25-fold faster and less technically 337 demanding. Additionally, because the samples require no cryoprotection and are not 338 prone to freezing-induced morphological changes, the protocol better preserves its native 339 state.47 340 Another pioneering step of the Glomage method is the incubation with antibodies in mesh 341 baskets containing large numbers of glomeruli. In standard stainings of cryosections , 342 100 µl of antibody solution is required per larva. A theoretical staining of 50 sections would 343 thus requires around 5 ml of antibody solutions. In contrast, the Glomage approach allows 344 the simultaneous staining of over 50 glomeruli using only 500 µl of antibody solution, 345 followed by mounting on a single glass slide, highlighting the protocol's remarkable 346 efficiency. Overall, processing 100 larvae yielded approximately 70 stained and mounted 347 glomeruli within one and a half days, demonstrating and scalability of the workflow. 348 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 28 An important feature of the glomerulus is its three-dimensional architecture, which cannot 349 be fully represented in standard histological sectio ns. To date, analyses of whole 350 glomeruli of larval zebrafish rely on either in vivo multi-photon microscopy or whole-mount 351 immunostainings.36,48–51 Although these techniques are powerful tools, they have notable 352 limitations. In vivo multiphoton microscopy requires specialized equipment and skilled 353 operators, and it does not allow antibody -based detection of specific proteins. Whole -354 mount immunofluorescence is labor -intensive, often affected by limited antibody 355 penetration and background fluorescence, and neither method is easily scalable. 356 Here we show that by using antibodies against key podocyte-specific proteins such as 357 nephrin and podocin, we were able to reconstruct the entire slit diaphragm of larval 358 zebrafish glomeruli in three-dimensions. Staining for Pax2, a transcription factor specific 359 for parietal epithelial cells , allows the three-dimensional visualization of the Bowman’s 360 capsule. 361 Beside this, the determination of the number of podocytes in glomeruli is a highly valuable 362 diagnostic parameter for kidney health and disease. 52–54 Although th e number of 363 podocytes has already been determined in mouse glomeruli, the corresponding number 364 for the larval zebrafish model was previously unknown. To address this, our protocol 365 exploits the free nuclear diffusion of endogenous fluorescent reporters such as eGFP or 366 mCherry, making it highly specific for podocytes and antibody-independent. By combining 367 the Glomage workflow with AI-assisted, unbiased segmentation of eGFP-positive nuclei 368 in our podocyte reporter strain s, we can now quantify the number of podocytes rapidly 369 and reproducibly. Using this approach, we quantified podocyte numbers in larvae at 5 370 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 29 dpf. On average, these larvae possess 90-100 podocytes, closely matching the podocyte 371 count observed in healthy adult mouse glomeruli.9,10 372 For the first time, this technique allows qui ckly longitudinal analysis of podocyte number 373 dynamics during development or disease, and it can be readily adapted to any glomerular 374 cell type with a suitable nuclear marker or reporter. 375 As a proof -of-concept of this strategy , we used the podocyte counting pipeline for our 376 established FSGS-like model in zebrafish larvae. For this, Cherry larvae were exposed to 377 the prodrug NFP which induces a podocyte-specific injury, resulting in many hallmarks of 378 FSGS such as podocyte loss, foot process effacement, glomerular matrix accumulation, 379 proteinuria, and edema formation.16,29,30 After 24 hours of NFP treatment, we counted the 380 number of podocytes in healthy, diseased and diseased larvae that received 381 pharmacological treatment. Larval FSGS-induction resulted in a significant loss of viable 382 podocytes and w e found that concurrent treatment of the NFP -treated larvae with the 383 compound NM_187 completely prevented podocyte loss. 384 Furthermore, Glomage enables volumetric analysis of the mCherry - or eGFP -positive 385 volume in larval glomeruli. In accordance with podocyte loss, NFP induced a significant 386 reduction and the co-treatment with NM_187 a complete rescue of the mCherry-positive 387 volume. This novel readout might be additionally used in the future to assess podocyte 388 hypertrophy or further volumetric changes upon injury or drug treatment. 389 Beside the use of Glomage for imaging purposes, we developed a protocol for the 390 extraction of glomerular RNA. Compared to a whole-larvae RNA isolation, this approach 391 yielded in a strong enrichment of glomerular and podocyte-specific transcripts, while 392 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 30 markers of other organs such as pancreas, hea rt, or eye s were undetectable . This 393 indicates a high purity of glomerular RNA. 394 With this modified technique, we measured the downregulation of podocyte -specific 395 genes upon podocyte injury as well as a protective effect of NM_187 on mRNA levels. 396 The close agreement between imaging and mRNA results demonstrates that both 397 approaches can be effectively combined, allowing podocyte imaging and RNA analysis 398 from the same batch of larvae which could also be used for RNA_Seq and other omic 399 technologies. 400 To study whether Glomage has also the potential for other animal species, we next 401 applied it to isolated glomeruli from mouse tissue . First, glomeruli were obtained by a 402 sieving and differential adhesion technique from a nephrin:CFP reporter strain .38 After 403 this, all isolated glomeruli were processed with the same antibody staining procedure in 404 mesh baskets as reported for zebrafish glomeruli . Here, we have shown by using an 405 antibody against the slit-diaphragm protein nephrin, that the Glomage workflow is directly 406 transferable to mouse tissue. 407 As proof -of-concept, we used the technique also in the context of age -associated 408 podocyte loss. To address this this, we isolated and analyzed glomeruli from young and 409 aged mice by using the transcription factor Dach1 as a podocyte-specific nuclear marker. 410 We found that aged mice exhibited a significant reduction in Dach1-positive podocyte 411 number compared with young animals, consistent with previous reports and supporting 412 the validity of this method.9,10 413 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 31 In summary, c ompared to already established protocols, Glomage applied to mouse 414 tissue offers several advantages: the entire workflow from living animal to stained and 415 mounted glomeruli can be completed within two days . N o animal perfusion, tissue 416 clearing, or prolonged antibody incubation on thick slices is required, steps that typically 417 take up to two weeks.9 Furthermore, the method does not require specialized microscopy 418 techniques such as light -sheet or super -resolution imaging. Additionally, the number of 419 glomeruli obtained from one kidney is sufficient for multiple staining procedures, allowing 420 the contralateral kidney to be used for standard histology and omics analysis. Moreover, 421 this approach can be extended to established glomerular injury models such as 422 nephrotoxic serum nephritis or puromycin aminonucleoside nephritis. 423 In summary, the Glomage workflow combines accuracy, scalability, and cross -species 424 applicability with remarkable simplicity and velocity using only standard laboratory 425 equipment. T his accessibility allows a wide range of laboratories to implement high -426 throughput, three -dimensional analyses of glomeruli, supporting studies from basic 427 podocyte biology to translational drug screening, and enabling rapid reproducible insights 428 into kidney health and disease. 429 430 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 32

431 Zebrafish husbandry 432 Zebrafish were maintained under standard conditions as previously described. 55 The 433 transparent nphs2:eGFP (Podo:GFP) line expresses eGFP specifically in pronephric 434 podocytes and was generated by crossing TG( nphs2:GAL4-VP16) (own outcross) with 435 TG(UAS:eGFP) (a kind gift from Prof. Brand, Dresden), resulting in the genotype 436 TG(nphs2:GAL4-VP16);TG(UAS:eGFP) mitfa W2/W2 (Fig. S1, A –C). The nphs2:NTR-437 mCherry (Cherry) line expresses both mCherry and the nitroreductase in podocytes and 438 was used for experiments involving targeted podocyte injury. The genotype of this line is 439 TG(nphs2:GAL4-VP16);TG(UAS:Eco.nfsB-mCherry) mitfa W2/W2 (a kind gift from Prof. 440 Weibin Zhou, Fig. S1, D–E).56,57 Fluorophore expression was verified at 4 dpf prior to the 441 start of each experiment under 0.1 mg/ml tricaine anesthesia (MS -222, E10521, Merck, 442 Darmstadt, Germany). 443 All experiments with zebrafish larvae were terminated at 5 dpf and conducted in 444 accordance with the guidelines of the local regulatory authorities. 445 Isolation and collection of larval zebrafish glomeruli 446 100 Podo:GFP larvae were fixed in 2% PFA for 1.5 hours at room temperature. After 447 washout, larvae were transferred in PBS to a screw cap tube prepared with 100 µl ceramic 448 beads. Both the beads and the inner surface of the screw cap were pre-coated with FBS 449 (Thermo Fisher Scientific, Dreieich, Germany). PBS, ceramic beads, and the 100 larvae 450 were adjusted to a total volume of 1 ml and dissociated using a dissociator (Fast-Prep-451 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 33 24TM, MP Biomedicals, Eschwege, Germany) at 4 m/s for 20 s. The resulting suspension 452 was then transferred, along with additional PBS to a FBS-coated petri dish (Greiner Bio-453 One, Kremsmünster, Austria). Fluorescent and morphologically intact glomeruli were 454 manually picked under a fluorescence stereomicroscope (SMZ18, Nikon, Düsseldorf, 455 Germany) using FBS -coated gel -loading pipette tips (Biozym, Hessisch -Oldendorf, 456 Germany) and tran sferred to a second FBS -coated petri dish containing PBS. After a 457 second round of quality control, glomeruli were transferred to FBS -coated cell culture 458 inserts (ThinCert®, Greiner Bio-One, Cat. No. 662638) with a mesh bottom (pore size: 8 459 µm) placed in a 24-well plate (Greiner Bio-One). 460 Immunofluorescence of whole larval zebrafish glomeruli 461 All solutions were applied by gently rinsing through the mesh, allowing liquids to pass 462 while retaining the glomeruli. Glomeruli resting on the mesh were washed twice with PBS 463 containing 0.1% Triton X-100 (PBST). Permeabilization was carried out by incubating the 464 samples in 0.3% PBST for 5 minutes, followed by antigen retrieval in 1× sodium dodecyl 465 sulfate ( Biomol, Hamburg, Germany ) buffer in PBS for 5 minutes. After was hing with 466 PBST, blocking solution was applied for 45 minutes. Primary antibodies were diluted in a 467 total volume of 500 µl and incubated overnight at 4°C . The following rabbit primary 468 antibodies and dilutions were used: anti -zebrafish nephrin 1:2000 (Innova gen, Lund, 469 Sweden), anti-podocin 1:7500 (Sigma-Aldrich, Taufkirchen, Germany, Cat. No. PO372), 470 anti-Pax2 1:1000 (Abcam, Cambridge, UK, Cat. No. ab38738), anti -Ehd3 1:200 (Sigma-471 Aldrich, Cat. No. HPA049986), and anti-Laminin 1:100 (Sigma-Aldrich, Cat. No. L9393). 472 After antibody incubation, glomeruli were washed four times for 5 minutes each with 473 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 34 PBST on a horizontal shaker. This was followed by a 1 -hour incubation with an Alexa 474 Fluor 647-conjugated donkey anti -rabbit secondary antibody (Thermo Fisher Scienti fic) 475 at a dilution of 1:800. Glomeruli were subsequently washed four times with PBST. Nuclei 476 were counterstained using HOECHST 33342 (Sigma -Aldrich), and F -actin was labeled 477 with Phalloidin-546, applied together with the nuclear stain. After two additional washes 478 with PBST and three final washes with A. dest., the mesh was carefully cut usi ng a 479 scalpel, grasped with forceps, and mounted on a glass slide containing a drop of Mowiol 480 (Roth) for embedding with a 22 x 22 mm cover slit. 481 Imaging of whole zebrafish glomeruli 482 All glomeruli were imaged using an FV3000 confocal microscope (Evident, Tokyo, Japan) 483 equipped with a 1.25x air objective (NA: 0.04), a 20x air objective (NA: 0.8), and a 60x oil 484 immersion objective (NA: 1.5). The 1.25x objective was used to image the entire mesh 485 for the generation of a map, using brightfield and either 488 nm or 561 nm excitation 486 depending on the expressed fluorophore (488 nm for eGFP, 561 nm for mCherry). This 487 map enabled automated navigation of the 20x and 60x objective to individual glomeruli. 488 For each glomerulus, z-stacks were acquired with the 60x objective and a step size of 1 489 µm. The zoom factor used for acquisition ranged from 2 x to 3.5 x, depending on the 490 individual orientation of the zebrafish glomerulus in Mowiol. 491 Cryosections and immunofluorescence 492 Zebrafish larvae (Podo:GFP) were fixed in 2% PFA at 5 dpf for 1.5 hours at room 493 temperature. Larvae were incubated overnight in 15% sucrose (Sigma-Aldrich) in PBS 494 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 35 after fixation. Following this, larvae were snap frozen in a 1:1 solution of 15% sucrose in 495 PBS and TissueTek O.C.T. compound (Sakura Finetek Europe, AV, Netherlands) in liquid 496 nitrogen. 5 µm cryosections were prepared using a CM 1950 cryostat (Thermo Fisher 497 Scientific) and stained with the same primary and secondary antibodies a t previously 498 described concentrations. Sections were mounted in Mowiol and images were captured 499 with the Evident FV3000 system using a 60x water objective (NA: 1.2). 500 Image processing 501 All imaging data were processed using the open -source software FIJI.58 Custom-written 502 macros were employed to automatically generate single slices an d maximum intensity 503 projections (MIPs) from z -stacks. The number of glomeruli per mesh was determined 504 manually. Two-channel hyperstacks (nuclei + eGFP or nuclei + mCherry) were used for 505 glomerular reconstruction, quantification of eGFP - or mCherry-positive podocyte nuclei 506 and nuclear volume using Imaris (Bitplane, version 10.2.0). Nuclei were segmented using 507 a custom trained AI network based on manually labelled nuclei (Fig. 3) or threshold 508 segmentation (Fig. 4, 5, 8) followed by watershed -based separation of touching objects 509 (region growing, seed size 2 µm) and filtering for total volume. eGFP or mCherry positive 510 nuclei were identified by an intensity mean threshold filter. The podocyte volume was 511 determined by threshold segmentation and volume filtering. 512 Induction of larval FSGS and compound treatment 513 Specific podocyte injury was induced by treating 100 Cherry larvae with 50 nM nifurpirinol 514 (NFP, Sigma-Aldrich) in 0.5x E3 medium for 24 hours from 4 to 5 dpf. The healthy control 515 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 36 group received 0.2% DMSO, whereas the intervention group was treated with 50 nM NFP 516 in combination with the compound NM_187 for the same duration. NM_187 is an 517 experimental small molecule. Its structure and mechanism of action are currently not 518 disclosed due to intellectual propert y restrictions. Larvae were subsequently used for 519 glomerular isolation, immunofluorescence, and confocal imaging (as described above), 520 or processed fresh for RNA isolation. 521 RNA isolation and PCR of zebrafish glomeruli 522 Non-fixed Cherry zebrafish larvae were dissociated as described above at 4 m/s for 1 s. 523 Glomeruli were collected under fluorescence guidance, and RNA was isolated following 524 two rounds of quality control using the Single Cell RNA Purification Kit (Norgen Biotek, 525 Thorold, ON, Canada) after manuf acturer`s instructions . RNA from 10 whole larvae, 526 processed in parallel, served as controls. Reverse transcription was carried out with the 527 QuantiTect® Reverse Transcription Kit (Qiagen, Hilden, Germany), followed by RT–PCR 528 using DreamTaq DNA Polymerase (T hermo Fisher Scientific) for 34 cycles. RT-qPCR 529 was performed as previously described on a QuantStudio™ 3 Real-Time PCR System 530 (Thermo Fisher Scientific) using the iTaq Universal SYBR Green Supermix (Bio -Rad).59 531 Raw Ct-values were normalized to eef1a1 as reference gene and to whole larvae or to 532 the 0.2% DMSO treatment group, as indicated respectively. Ct-values ≤ 38 were excluded 533 from analysis. As negative controls we included no -template controls (NTC) as well as 534 no-reverse-transcriptase controls ( -RT). All measurements were r un in triplicates and 535 primer sequences are provided in the supplemental material (Table S1). 536 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 37 Mouse husbandry 537 For this study, we used transgenic nephrin:CFP mice, in which CFP is expressed 538 specifically in podocytes under the control of a nephrin promoter fr agment. The 539 generation of this mouse line involved insertion of the CFP cDNA into the EcoRI site of 540 the NPXRS nephrin construct, as previously described. 60,61 Animals were housed under 541 standardized condi tions (21 °C, 60% humidity, 12:12 h ours light–dark cycle) with ad 542 libitum access to food and water. Experimental procedures were carried out on 6-month-543 old or 18-month-old mice. All animal work was conducted in compliance with national 544 animal welfare legislation with prior approval from the competent local authority. 545 Isolation and immunofluorescence of mouse glomeruli 546 Glomeruli were isolated using the differential adhesion method as described by Wang et 547 al.. 38 At the final step, glomeruli were fixed in 2% PFA for 20 minutes at room temperature 548 in a 15 ml conical tube (Sarstedt, Nümbrecht, Germany). After fixation, PFA was removed 549 by washing, and all glomeruli from one mouse were resuspended in 2 ml PBS. 50 µl of 550 the suspension was inspected under a stereomicroscope (brightfield) to assess the 551 density of glomeruli in the solution . An appropriate volume of the suspension was then 552 transferred onto cell culture inserts (Greiner Bio-One) to achieve an estimated number of 553 ≥50 glomeruli per mesh. Immunofluorescence staining, embedding, and imaging were 554 performed identically to the procedure used for isolated zebrafish glomeruli. The following 555 antibodies and concentrations were used for the immunoflu orescence on mouse 556 glomeruli: Anti-Nephrin 1:1000 (Progen, Heidelberg, Germany, Cat. No.: GP-N2), Anti-557 Dach1 1:500 (Sigma -Aldrich, Cat. No.: HPA012672). The same Alexa Fluor 647 -558 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 38 conjugated donkey anti-rabbit (Thermo Fisher Scientific) and a Cy3-conjugated donkey-559 anti guinea pig (Jackson ImmunoResearch, Newmarket, United Kingdom) were used as 560 secondary antibodies. 4′,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich) was used for 561 the nuclear counterstain. 562 Statistical analysis 563 Data analysis and graph generation were performed using GraphPad Prism 9.1.2 564 (GraphPad Software, San Diego, CA, USA). Normality was assessed using the Shapiro-565 Wilk test. For comparisons between two groups, either an unpaired two -tailed Student’s 566 t-test or a Mann -Whitney U test was applied. For comparisons between more than two 567 groups, a one-way ANOVA followed by Tukey’s multiple comparisons test or a Kruskal -568 Wallis test with Dunn’s post hoc correction was used, as appropriate. A p -value ≤0.05 569 was considered statistically significant. 570 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 39

571 1. Bikbov, B. et al. Global, regional, and national burden of chronic kidney disease, 572 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. The 573 Lancet 395, 709–733 (2020). 574 2. Foreman, K. J. et al. Forecasting life expectancy, years of life lost, and all-cause and 575 cause-specific mortality for 250 causes of death: reference and alternative scenarios 576 for 2016-40 for 195 countries and territories. Lancet 392, 2052–2090 (2018). 577 3. Pavenstädt, H., Kriz, W. & Kretzler, M. Cell Biology of the Glomerular Podocyte. 578 Physiological Reviews 83, 253–307 (2003). 579 4. Miyaki, T. et al. Three-dimensional imaging of podocyte ultrastructure using FE-SEM 580 and FIB-SEM tomography. Cell Tissue Res 379, 245–254 (2020). 581 5. Kocylowski, M. K. et al. A slit -diaphragm-associated protein network for dynamic 582 control of renal filtration. Nat Commun 13, 6446 (2022). 583 6. Bai, X. Y. & Basgen, J. M. Podocyte Number in the Maturing Rat Kidney. Am J Nephrol 584 33, 91–96 (2011). 585 7. White, K. E. & Bilous, R. W. Estimation of podocyte number: A comparison of 586 methods. Kidney International 66, 663–667 (2004). 587 8. Puelles, V. G. et al. Design-based stereological methods for estimating numbers of 588 glomerular podocytes. Annals of Anatomy - Anatomischer Anzeiger 196, 48 –56 589 (2014). 590 9. Puelles, V. G. et al. Validation of a Three-Dimensional Method for Counting and Sizing 591 Podocytes in Whole Glomeruli. J Am Soc Nephrol 27, 3093–3104 (2016). 592 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 40 10. Shi, J. et al. Quantifying Podocyte Number in a Small Sample Size of Glomeruli with 593 CUBIC to Evaluate Podocyte Depletion of db/db Mice. Journal of Diabetes Research 594 2023, 1–12 (2023). 595 11. Puelles, V. G., Combes, A. N. & Bertram, J. F. Clearly imaging and quantifying the 596 kidney in 3D. Kidney International 100, 780–786 (2021). 597 12. Drummond, I. A. et al. Early development of the zebrafish pronephros and analysis of 598 mutations affecting pronephric function. Development 125, 4655–4667 (1998). 599 13. Drummond, I. A. & Davidson, A. J. Zebrafish Kidney Development. in Methods in Cell 600 Biology vol. 100 233–260 (Elsevier, 2010). 601 14. Schindler, M. & Endlich, N. Zebrafish as a model for podocyte research. American 602 Journal of Physiology-Renal Physiology 326, F369–F381 (2024). 603 15. Howe, K. et al. The zebrafish reference genome sequence and its relationship to the 604 human genome. Nature 496, 498–503 (2013). 605 16. Schindler, M. et al. A Novel High -Content Screening Assay Identified Belinostat as 606 Protective in a FSGS-Like Zebrafish Model. J Am Soc Nephrol 34, 1977–1990 (2023). 607 17. Schindler, M., Blumenthal, A., Moeller, M. J., Endlich, K. & Endlich, N. Adriamycin 608 does not damage podocytes of zebrafish larvae. PLoS ONE 15, e0242436 (2020). 609 18. Yu, T. et al. Identification of renal stem cells in zebra fish. Sci Adv 11, eadx5296 610 (2025). 611 19. Müller, T. et al. Non-muscle myosin IIA is required for the development of the zebrafish 612 glomerulus. Kidney International 80, 1055–1063 (2011). 613 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 41 20. Bolten, J. S., Pratsinis, A., Alter, C. L., Fricker, G. & Huwyler, J. Zebrafish ( Danio rerio 614 ) larva as an in vivo vertebrate model to study renal function. American Journal of 615 Physiology-Renal Physiology 322, F280–F294 (2022). 616 21. Siegerist, F. et al. The differential expression of MAGI2 in glomerulopathies and its 617 application as a molecular discriminator of podocytopathies. J Transl Med 23, 701 618 (2025). 619 22. Bondue, T. et al. Evaluation of the efficacy of cystinosin supplementation through 620 CTNS mRNA delivery in experimental models for cystinosis. Sci Rep 13, 20961 621 (2023). 622 23. Taimatsu, K. et al. Comprehensive 3D Imaging of Whole Zebrafish Using a Water -623 Based Clearing Reagent for Hard Tissues. Zebrafish 22, 65–75 (2025). 624 24. Siegerist, F. et al. Evaluation of endogenous miRNA reference genes across different 625 zebrafish strains , developmental stages and kidney disease models. Sci Rep 11, 626 22894 (2021). 627 25. Mattias, F. et al. Regulation of the transcriptome, miRNAs, and alternative splicing in 628 a FSGS zebrafish injury model. Preprint at https://doi.org/10.1101/2025.07.09.663814 629 (2025). 630 26. Arber, D. A. Effect of Prolonged Formalin Fixation on the Immunohistochemical 631 Reactivity of Breast Markers: Applied Immunohistochemistry & Molecular Morphology 632 10, 183–186 (2002). 633 27. Konno, K., Yamasaki, M., Miyazaki, T. & Watanabe, M. Glyoxal fixation: An approach 634 to solve immunohistochemical problem in neuroscience research. Sci Adv 9, 635 eadf7084 (2023). 636 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 42 28. Dekan, G., Gabel, C. & Farquhar, M. G. Sulfate contributes to the negative charge of 637 podocalyxin, the major sialoglycoprotein of the glomerular filtration slits. Proc. Natl. 638 Acad. Sci. U.S.A. 88, 5398–5402 (1991). 639 29. Kerjaschki, D. Caught flat-footed: podocyte damage and the molecular bases of focal 640 glomerulosclerosis. J Clin Invest 108, 1583–1587 (2001). 641 30. Klawitter, M. et al. Investigating FSGS-like injury in zebrafish larvae by nifurpirinol: 642 efficacy and molecular insight. Am J Physiol Renal Physiol 327, F463–F475 (2024). 643 31. Hansen, K. U. I. et al. Prolonged podocyte depletion in larval zebrafish resembles 644 mammalian focal and segmental glomerulosclerosis. FASEB j. 34, 15961 –15974 645 (2020). 646 32. George, M. et al. Renal thrombotic microangiopathy in mice with combined deletion 647 of endocytic recycling regulators EHD3 and EHD4. PLoS One 6, e17838 (2011). 648 33. Wanner, N. et al. Unraveling the role of podocyte turnover in glomerular aging and 649 injury. J Am Soc Nephrol 25, 707–716 (2014). 650 34. Wan, X. et al. Loss of Epithelial Membrane Protein 2 Aggravates Podocyte Injury via 651 Upregulation of Caveolin-1. Journal of the American Society of Nephrology 27, 1066–652 1075 (2016). 653 35. Siegerist, F., Blumenthal, A., Zhou, W., Endlich, K. & Endlich, N. Acute podocyte injury 654 is not a stimulus for podocytes to migrate along the glomerular basement membrane 655 in zebrafish larvae. Sci Rep 7, 43655 (2017). 656 36. Siegerist, F., Zhou, W., Endlich, K. & Endlich, N. 4D in vivo imaging of glomerular 657 barrier function in a zebrafish podocyte injury model. Acta Physiol (Oxf) 220, 167–173 658 (2017). 659 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 43 37. Kindt, F. et al. A novel assay to assess the effect of pharmaceutical compounds on 660 the differentiation of podocytes. Br J Pharmacol 174, 163–176 (2017). 661 38. Wang, H. et al. A simple and highly purified method for isolation of glomeruli from the 662 mouse kidney. Am J Physiol Renal Physiol 317, F1217–F1223 (2019). 663 39. Loreth, D., Sachs, W. & Meyer‐Schwesinger, C. The Life of a Kidney Podocyte. Acta 664 Physiologica 241, e70081 (2025). 665 40. Nagata, M. Podocyte injury and its consequences. Kidney International 89, 1221–666 1230 (2016). 667 41. Wiggins, R. -C. The spectrum of podocytopathies: A unifying view of glomerular 668 diseases. Kidney International 71, 1205–1214 (2007). 669 42. Poureetezadi, S. J. & Wingert, R. A. Little fish, big catch: zebrafish as a model for 670 kidney disease. Kidney International 89, 1204–1210 (2016). 671 43. Kramer-Zucker, A. G., Wiessner, S., Jensen, A. M. & Drummond, I. A. Organization 672 of the pronephric filtration app aratus in zebrafish requires Nephrin, Podocin and the 673 FERM domain protein Mosaic eyes. Developmental Biology 285, 316–329 (2005). 674 44. Peng, Z. et al. Somites are a source of nephron progenitors in zebrafish. Nat Commun 675 16, 6914 (2025). 676 45. Gehrig, J., Pandey, G. & Westhoff, J. H. Zebrafish as a Model for Drug Screening in 677 Genetic Kidney Diseases. Front. Pediatr. 6, 183 (2018). 678 46. Bolten, J. S. et al. Zebrafish (Danio rerio) larvae as a predictive model to study 679 gentamicin-induced structural alterations of the kidney. PLoS ONE 18, e0284562 680 (2023). 681 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 44 47. Adam, M., Fei Hu, J., Lange, P. & Wolfinbarger, L. The effect of liquid nitrogen 682 submersion on cryopreserved human heart valves. Cryobiology 27, 605–614 (1990). 683 48. Endlich, N. et al. Two-photon microscopy reve als stationary podocytes in living 684 zebrafish larvae. J Am Soc Nephrol 25, 681–686 (2014). 685 49. Siegerist, F., Blumenthal, A., Zhou, W., Endlich, K. & Endlich, N. Acute podocyte injury 686 is not a stimulus for podocytes to migrate along the glomerular basement membrane 687 in zebrafish larvae. Sci Rep 7, 43655 (2017). 688 50. Ichimura, K. et al. Developmental Localization of Nephrin in Zebrafish and Medaka 689 Pronephric Glomerulus. J Histochem Cytochem. 61, 313–324 (2013). 690 51. Gerlach, G. F. & Wingert, R. A. Zebrafish pronephros tubulogenesis and epithelial 691 identity maintenance are reliant on the polarity proteins Prkc iota and zeta. 692 Developmental Biology 396, 183–200 (2014). 693 52. Puelles, V. G. & Bertram, J. F. Counting glomerul i and podocytes: rationale and 694 methodologies. Current Opinion in Nephrology and Hypertension 1 (2015) 695 doi:10.1097/MNH.0000000000000121. 696 53. Haruhara, K. et al. Podocyte density as a predictor of long -term kidney outcome in 697 obesity-related glomerulopathy. Kidney International 106, 496–507 (2024). 698 54. Ding, F. et al. Accelerated podocyte detachment and progressive podocyte loss from 699 glomeruli with age in Alport Syndrome. Kidney International 92, 1515–1525 (2017). 700 55. Endlich, N. et al. The transcription factor Dach1 is essential for podocyte function. J 701 Cell Mol Med 22, 2656–2669 (2018). 702 56. Wan, X. et al. Loss of Epithelial Membrane Protein 2 Aggravates Podocyte Injury via 703 Upregulation of Caveolin-1. J Am Soc Nephrol 27, 1066–1075 (2016). 704 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 45 57. Zhou, W. & Hilde brandt, F. Inducible Podocyte Injury and Proteinuria in Transgenic 705 Zebrafish. Journal of the American Society of Nephrology 23, 1039–1047 (2012). 706 58. Schindelin, J. et al. Fiji: an open -source platform for biological -image analysis. Nat 707

9, 676–682 (2012). 708 59. Ristov, M.-C. et al. The ShGlomAssay Combines High -Throughput Drug Screening 709 With Downstream Analyses and Reveals the Protective Role of Vitamin D3 and 710 Calcipotriol on Podocytes. Front. Cell Dev. Biol. 10, 838086 (2022). 711 60. Cui, S., Li, C., Ema, M., Weinstein, J. & Quaggin, S. E. Rapid Isolation of Glomeruli 712 Coupled with Gene Expression Profiling Identifies Downstream Targets in Pod1 713 Knockout Mice. Journal of the American Society of Nephrology 16, 3247–3255 (2005). 714 61. Wong, M. A., Cui, S. & Quaggin, S. E. Identification and characterization of a 715 glomerular-specific promoter from the human nephrin gene. American Journal of 716 Physiology-Renal Physiology 279, F1027–F1032 (2000). 717 718 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 46

719 The authors are very gra teful for excellent technical assistance of Jurij Barmenkov , 720 Claudia Weber and the outstanding zebrafish as well as mouse husbandry by Oliver 721 Zabel and Steffen Prellwitz. Image analysis of S.S.L. was conducted at the Center for 722 Microscopy and Image Analysis of the University of Zurich. Selected elements of Figures 723 1, 6, 7, and 8 were created using BioRender.com. 724 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 47 Funding 725 This work was supported by the Federal Ministry of Education and Research (BMBF, 726 grant 01GM2202B, STOP-FSGS) awarded to N.E.. Additionally, funding was provided by 727 the Federal Ministry for Economic Affairs and Climate Action (BMWi, grant 16KN077229, 728 project title: Alterna Tier-vivoPod). Furthermore, generous support was received from the 729 Dr. Gerhard Büchtemann Fund, Hamburg, Germany, and the Südmeyer Foundation for 730 Kidney and Vascular Research (“Südmeyer Stiftung für Nieren - und Gefäßforschung”). 731 S.S.L. is supported by the Swiss National Science Foundation (project 310030_236370) 732 and a project grant from the Theiler-Haag Foundation. 733 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 48 Author contributions 734 Conceptualization: M.S., N.E. 735 Methodology: M.S., S.M.B., T.L. 736 Software: M.S., T.L., S.S.L. 737 Validation: M.S., T.L. 738 Formal analysis: M.S., T.L., S.S.L. 739 Investigation: M.S., T.L., S.M.B. 740 Resources: S.S.L., N.E. 741 Data curation: M.S.,T.L., S.S.L. 742 Writing – original draft: M.S. 743 Writing – review & editing; T.L., S.S.L. N.E. 744 Visualization: M.S., S.S.L. 745 Supervision: N.E. 746 Project administration: N.E. 747 Funding acquisition: N.E. 748 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 49 Competing Interests 749 N.E. is CEO of the NIPOKA GmbH, Greifswald, Germany. 750 N.E. and M.S. filed a patent that is related to this study. 751 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 50 Data and materials available 752 All data needed to evaluate the conclusions are present in the manuscript and/or the 753 Supplemental Materials. 754 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint 51 Supplementary Material 755 The corresponding supplementary information can be found in the PDF file: 756 Supplemental Material Glomage 757 758 Other Supplementary Material for this manuscript includes the following: 759 Movie S1: https://doi.org/10.6084/m9.figshare.30384343 760 Movie S2: https://doi.org/10.6084/m9.figshare.30384454 761 Movie S3: https://doi.org/10.6084/m9.figshare.30384496 762 Movie S4: https://doi.org/10.6084/m9.figshare.30384502 763 Movie S5: https://doi.org/10.6084/m9.figshare.30384526 764 Movie S6: https://doi.org/10.6084/m9.figshare.30425632 765 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 31, 2025. ; https://doi.org/10.1101/2025.10.31.684126doi: bioRxiv preprint