An immunocompetent human kidney on-a-chip model to study renal inflammation and immune-mediated injury

12 Kidney damage and dysfunction is an emerging health issue worldwide resulting in high morbidity and 13 mortality rates. Numerous renal diseases are recognized to be driven by the immune system. Despite 14 this recognition, the development of targeted therapies has been challenging as knowledge of the 15 underlying mechanism and complex interactions remains insufficient. Recent advancements in the 16 field offer promising avenues for exploring the interplay between renal cells and immune cells and 17 their role in the development of renal inflammation and diseases. This study describes the 18 establishment of a human immunocompetent 3D in vitro co-culture model of the proximal tubule in a 19 high-throughput microfluidic platform that can be used to study renal functionality and inflammatory 20 processes. 21 The model incorporated RPTEC in the top compartment and HUVECs in the bottom compartment 22 cultured under flow and in direct contact with a collagen -I ECM gel resulting in the formation of 23 polarized tubular structures. As an immune component, human primary monocytes of different donors 24 were added to the lumen of the endothelium. Renal inflammation was successfully induced using 25 complement activated serum (CAS) as evident by epithelial morphological changes, increased 26 expression of adhesion molecules, release of pro -inflammatory cytokines, and reduced epithelial 27 viability. Realtime migratory behavior of monocytes showed increased extravasation and migration 28 towards the ECM and Renal compartment upon exposure to CAS with donor -to-donor differences 29 observed. Finally, immune modulatory compounds showed efficacious inhibition of monocyte 30 migration under inflammatory conditions in the microfluidic co-culture model. 31 A successful co-culture model was established and can be applied to study renal functionality in health 32 and disease but also for drug screening due to the compatibility of the platform with automation and 33 relatively high throughput. Overall, the described proximal tubule model has high potential to fill the 34 gap that currently exists to study renal inflammation preclinically. 35

36 Kidney damage and associated dysfunction is an emerging health issue worldwide resulting in high 37 morbidity and mortality rates. 1 Chronic kidney disease (CKD) is a progressive irreversible condition 38 affecting 10-15% of the population in western countries often leading to end -stage renal disease 39 (ESRD), requiring dialysis or kidney transplantation. 2 Over the years, treatment, and management of 40 CKD and ESRD has evolved but despite these advances, they often address patient symptoms rather 41 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint than the underlying cause of disease, with significant side effects that can further compromise quality 42 of life. 3 The lack of specific treatments for CKD worldwide urges the need for efficient preclinical 43 research tools with significant predictive and translational capacity to increase our understanding of 44 the pathophysiology of renal diseases and to support the drug development pipeline. 45 There are a multitude of risk factors and diseases underlying CKD with distinct pathogenic mechanisms 46 including diabetes, hypertension, infections, (drug -induced) toxicity, and immunological disorders. 4,5 47 Regardless of the different underlying mechanisms, one of the hallmarks in renal diseases is 48 inflammation which contributes to disease development and progression.6–8 Inflammation is a process 49 aiming to detect and fight harmful pathogens and subsequently promote tissue repair and recovery. 50 However, low-grade persistent inflammation has been demonstrated as a major pathogenic factor in 51 many renal diseases and has been recognized as both a cause and a consequence. 6 Persistent 52 inflammation is defined by low to moderate levels of circulating inflammation markers and the extent 53 and effects can vary with the cause of injury. Activation of the complement system has been shown to 54 play a crucial role in persistent inflammation.9 The complement system, an integral component of the 55 innate immune system, consists of a complex network of proteins and can become activated through 56 three separate pathways. Complement activation results in recruitment and activation of immune cells 57 like monocytes and macrophages. Abnormalities in number and relative proportions of these cells 58 have been described previously in CKD and s pecific monocyte subsets have been associated with 59 significant loss of kidney function.10 These immune cells can enter the tissue and release cytokines and 60 other signaling molecules that further drive the inflammatory response.11 Inappropriate activation of 61 this tightly regulated process can contribute to cell damage, fibrosis, and declined renal function.9,12,13 62 Developing therapies that target inflammation or specifically the complement system has been of 63 significant interest and progress has been made in identifying novel targets and developing potential 64 therapies.14–16 Eculizumab, a monoclonal antibody targeting complement factor 5 (C5), showed 65 hematological and renal improvements as well as discontinuation of dialysis in most patients with 66 atypical hemolytic uremic syndrome (aHUS) and currently is the first -line treatment option. 17,18 67 Furthermore, novel anti-inflammatory drugs targeting CCR2, IL -1β, IL-33, ASK1 and IL -6 are or have 68 been under development for specific renal diseases.19 Still, most anti-inflammatory drugs fail in phase 69 2 clinical trials due to insufficient efficacy or increased risk of adverse outcomes such as cardiovascular 70 events. Hence, more preclinical research is required to further expand our understanding of these 71 complex interactions in the chronic inflammatory microenvironment allowing improved patient 72 stratification and selection of optimal clinical endpoints, all supporting the development of anti -73 inflammatory therapies. 74 Most research in the field focuses on the proximal tubule (PT) as it is vulnerable to injury due to its 75 high metabolic activity and exposure to high solute concentrations. PT dysfunction is linked to many 76 acute and chronic kidney diseases. 20,21 Numerous animal models are available to study kidney 77 functionality and these generated valuable insights and discoveries to the field. 22,23 Nevertheless, 78 ethical concerns and other limitations including high costs and resources but more importantly, the 79 ‘’translational gap’’ urges the need for alternative methods. Major advances have been made over the 80 years in developing more sophisticated and relevant in vitro models of the proximal tubule. These 81 include 3D cell culturing,24,25 stem cell and organoid cultures,26–28 3D bioprinting,29–31 and organ-on-a-82 chip technology.32–35 Despite these advances, there is a demand for an in vitro model that integrates 83 rapid and scalable readouts while maintaining physiological relevance. This should include features 84 such as co -culture, fluid flow, and the absence of an artificial membrane to effectively study the 85 proximal tubule in both health and disease. 86 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint Here, we report an immunocompetent human proximal tubule and vasculature microfluidic co-culture 87 model to study specific aspects of kidney function and inflammation and test possible therapeutic 88 interventions at scale . As a starting point, a previously established co -culture model was used , 89 composed of human renal proximal tubule epithelial cells (RPTECs) and human umbilical vein 90 endothelial cells (HUVEC) cultured as tubular structures against a collagen-I extracellular matrix (ECM) 91 in the OrganoPlate 3-lane 40.36 The model was transferred to a microfluidic prototype plate, called the 92 OrganoPlate 3-lane Narrow ECM, containing the same layout but with different channel dimensions 93 including a smaller ECM channel. To establish an immun ocompetent model, human primary 94 monocytes were added to the lumen of the endothelial vessel. The tri-culture model revealed barrier 95 formation, polarization, and expression of tight junction markers. Renal inflammation was induced 96 using complement activated serum (CAS) and effects were assessed by determining marker expression, 97 cell viability, cytokine release, and monocyte adhesion and migration. Finally, monocyte donor 98 variability and the effect of immune modulatory compounds on monocyte migration and adhesion 99 were assessed. Overall, we described the development of a human immunocompetent in vitro model 100 of the proximal tubule that can be used to study renal functionality and inflammatory processes and 101 has high potential to fill the gap that currently exists to study renal inflammation preclinically. 102

103 2D Cell Culture 104 Human renal proximal tubule epithelial cells (RPTECs, SA7K clone, MTOX1030; Sigma) were cultured 105 on PureCol -coated T75 flasks (431464U; Corning) in MEME alpha modification (M4526; Sigma) 106 supplemented with RPTEC complete supplement (MTOXRCSUP; Sigma), L-glutamine (2.33 mM, G7513; 107 Sigma), gentamicin (28 µg/ml, G1397; Sigma) and amphotericin B (14 ng/ml, A2942; Sigma), further 108 referred to as RPTEC complete medium. To coat the T75 flasks, PureCol (5005-B; Advanced BioMetrix) 109 was diluted 1:30 in cold Hank’s balanced salt solution (HBSS, H6648; Sigma) and incubated for 20 min 110 at 37℃. The leftover solution was aspirated after which the flasks were ready. RPTECs were used for 111 experiments up to passage 3. 112 Human umbilical vein endothelial cells (HUVECs, C2519A; Lonza) were cultured in T75 flasks (Nunc™ 113 EasyFlask™, 156499; Thermo Scientific) in MV2 medium (C -22221; PromoCell) supplemented with 114 Growth Medium MV2 SupplementMix (C-39226; PromoCell), and penicillin-streptomycin (1%, P4333; 115 Sigma), further referred to as endothelial complete medium. HUVECs were used for experiments up 116 to passage 5. 117 Cells were cultured in a humidified incubator (37℃, 5% CO2) and maintained by adding fresh medium 118 every 2-3 days. Cultures were routinely tested for mycoplasma contamination and found negative. 119 Monocyte isolation from human PBMCs 120 Human peripheral blood mononuclear cells (PBMCs) were obtained from StemExpress (PBMNC100C, 121 100M per vial). PBMCs were thawed, taken up in Roswell Park Memorial Institute (RPMI) 1640 basal 122 medium (11875093; Thermo Scientific) supplemented with 10% fetal bovine serum (FBS, 16140-071; 123 Gibco), pelleted (300g, 5 min), and washed with phosphate-buffered saline (PBS; 20012068; Sigma) 124 containing 2% FBS. Monocytes were isolated from PBMCs by CD14 positive selection using the 125 EasySep™ Human CD14 Positive Selection Kit (18058; Stemcell Technologies) according to 126 manufacturer’s protocol. In short, PBMCs were pelleted (300g, 5 min) and resuspended in EasySep ™ 127 Buffer (20144; Stemcell Technologies) at a concentration of 1 x 10 8 cells/mL and incubated with 100 128 µL selection antibody cocktail per mL of sample for 10 min at RT. Next, 12 µL RapidSpheres ™ per mL 129 of sample was added and incubated for an additional 3 min at RT. Cells were placed into the magnet, 130 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint incubated for 10 min at RT after which the supernatant was removed and the pellet was resuspended 131 in EasySep™ buffer. This step was repeated three times. Finally, monocytes were washed with PBS 132 containing 2% FBS, pelleted (300g, 10 min, low brake), taken up in Cryostor® CS10 freezing medium 133 (07930; Stemcell Technologies) at 5 x 106 cells/mL and stored at -150°C. 134 OrganoPlate Culture (microfluidic cell culture) 135 For all experiments, the OrganoPlate 3 -lane 40 Narrow ECM platform was used (MIMETAS , the 136 Netherlands, custom design). This is a custom-made microfluidic prototype plate utilizing the same 137 chip format as the OrganoPlate 3-lane 40 platform but with different channel dimensions including a 138 smaller ECM channel. Specifically, dimensions of the top and bottom perfusion channels are 300 µm x 139 165 µm (w x h), of the ECM (middle) channel 200 µm x 165 µm (w x h) and phaseguides had dimensions 140 of 50 µm x 55 µm (w x h). The protocol for ECM loading and cell seeding was followed as previously 141 described.37,38 In short, each observation window column was filled with 50 µL of Hank’s balanced salt 142 solution (HBSS, 55037C; Sigma) per well to prevent dehydration and provide optical clarity. An 143 extracellular matrix (ECM) composed of 4 mg/mL rat -tail collagen I (3447-020-01; AMSbio), 100 mM 144 HEPES (15630 -122; Sigma), and 3.7 mg/mL NaHCO 3 (S5761; Sigma) was prepared and 1 µl was 145 dispensed into the gel inlet followed by 10 min static incubation at 37 ℃. After polymerization of the 146 ECM gel, 20 µL of HBSS was added to the gel inlet, and the plate was incubated overnight in a 147 humidified incubator (37℃, 5% CO2). The next day, RPTECs were harvested using Accutase® solution 148 (A6964; Sigma), pelleted (140g, 5 min) and resuspended in RPTEC complete medium at a density of 149 10,000 cells/µL. Subsequently, a 2 µL cell suspension was injected into the inlet of the top medium 150 channel and placed at a 75° angle for 4.5 hours in the incubator to allow the cells to adhere to the 151 ECM. After attachment of the cells, RPTEC complete medium was added to the remaining in- and outlet 152 wells of the top and bottom perfusion channels and the plate was placed horizontally on an interval 153 rocker (OrganoFlow, OFPR-L; MIMETAS, 7° inclination, 8 -minute interval) in a humidified incubator 154 (37℃, 5% CO 2) enabling a passive, bidirectional flow through the perfusion channels .39 Upon flow 155 application, the RPTECs proliferated and started lining all surfaces of the perfusion channel and formed 156 a confluent tubule. Medium was refreshed every 2 -3 days by aspirating all media from the in - and 157 outlet wells and replacing it with fresh RPTEC complete medium. 158 To establish the co-culture, HUVECs were added to the bottom perfusion channel of the microfluidic 159 chip 3 days after RPTEC seeding. HUVECs were detached using 0.025% Trypsin-EDTA (1X) solution (CC-160 5012; Lonza), neutralized with Trypsin Neutralizing Solution (CC-5002; Lonza), pelleted (200g, 5 min), 161 and resuspended in endothelial complete medium at a density of 10,000 cells/µL. Media from all in - 162 and outlet wells of the bottom perfusion channel was aspirated and 2 µL HUVEC cell suspension was 163 injected into the inlet well. Due to loss of capillary forces by pre -wetting the channels with culture 164 media, 1 µL of the cell suspension was taken out of the outlet well using a pipette to guide the HUVECs 165 through the perfusion channel. This 1 µ L cell suspension was added to the inlet well again and this 166 step was repeated 2-3 times. After incubating the plate for 60 min at an angle of 75° at 37 ℃ for cell 167 adherence, endothelial basal medium (CellBiologics) supplemented with growth factor supplement kit 168 (H1168; CellBiologics) was added to the in- and outlet wells of the bottom channel. Subsequently, the 169 media of the top channel was aspirated and replaced with fresh RPTEC complete medium. The 170 OrganoPlate was placed back on the interval rocker in a humidified incubator (37 ℃; 5% CO 2) to 171 promote tubule formation. Polarized confluent tubules were established on day 6, at which they could 172 be used for exposures and/or readouts. Co-cultures were used for experiments up to day 10. 173 Inflammatory trigger addition and compound exposure 174 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint On day 6 of the OrganoPlate culture, the RPTEC and HUVEC tubules were exposed to media containing 175 an inflammatory trigger and/or a potentially protective compound. To this end, Cell Biologics basal 176 media supplemented with L-glutamine, Antibiotic-Antimycotic solution, VEGF, Heparin, EGF, FGF and 177 hydrocortisone from the growth factor supplement kit (H1168; Cell Biologics) and 2% human AB serum 178 (SM-612-HIS; BioIVT) was prepared and filtered using a 0.2 µm filter. This medium is further referred 179 to as assay medium. To induce an inflammatory environment, assay medium was supplemented with 180 5% (v/v) cobra venom activated human complement serum (CAS, CVF-NHS; CompTech). As a negative 181 control, assay medium was included. Media from the top channel (RPTECs) was aspirated and replaced 182 with assay media w/ or w/o CAS (70 µ L/well) and incubated for 4 hours on the interval rocker in the 183 incubator to establish a concentration gradient in the chip . To study the effect of compounds that 184 target either the trigger or the monocytes directly, compound A and B were tested, respectively. Both 185 compounds were dissolved in assay medium. Compound A was added to the Renal, the Donor or to 186 both compartments at a concentration of 25 µg/mL. Compound B was added to the Renal 187 compartment and tested at 1, 10 and 50 µg/mL. For all compound conditions, 5% CAS (v/v) was 188 present in the Renal compartment of the chip. After 4-hour incubation, monocytes were added to the 189 Donor compartment, followed by addition of assay medium +/ - compound A (70 µL/well). Cultures 190 were exposed for 96 hours on a rocker platform (7° angle, 8-min interval) in the incubator (37℃; 5% 191 CO2). Phase-contrast and fluorescent images were acquired every 24 hours using the ImageXpress XLS 192 Micro High Content Imaging System (Molecular Devices). 193 Monocyte cell labeling and seeding 194 Primary human monocytes were thawed, taken up in RPTEC 1640 medium supplemented with 10% 195 FBS, pelleted (200g, 5min) and labelled with CellTracker Orange CMRA (C34551; Thermo Scientific) by 196 addition of 5mL CMRA working solution. A 10 mM stock was prepared by dissolving 50 µg CMRA in 9.1 197 µL dimethyl sulfoxide (DMSO, D8418; Sigma). Next, a working solution of 5 µM in RPMI 1640 basal 198 medium was prepared and warmed up to 37℃. After 30 min incubation at 37℃ in the dark, 5 mL RPMI 199 1640 + 10% FBS was added, cells were pelleted (200g, 5 min) and resuspended in assay medium at a 200 density of 25,000 cells/µL. 201 Media from all in- and outlet wells of the bottom perfusion channel was aspirated and 2 µL monocyte 202 cell suspension was injected into the inlet well. The same procedure as described for seeding the 203 HUVECs was followed to guide the monocytes through the channel. Images of the first timepoint were 204 taken, after which the plate was placed at a 75° angle for 10 min at RT to sediment the cells against 205 the HUVEC-ECM interface. Assay medium +/- compound (70 µL/well) was added to the bottom channel 206 and the plate was placed on a rocker platform (7° angle, 8-min interval) in the incubator (37℃; 5% CO2) 207 for 96 hours. 208 To prevent significant monocyte loss during handling, all pipette tips and culture tubes were coated 209 with a buffer solution. A buffer containing 10 mM Tris HCL (H5121; Promega), 0.1 mM EDTA (AM9260G; 210 Invitrogen) and 0.1% bovine serum albumin (BSA, A2153; Sigma) in sterile Milli-Q water was prepared 211 and used to coat all tips and tubes that were going to be in contact with monocytes. Pipette tips were 212 dipped in the buffer to coat the outside, while the inside was coated by taken up some buffer solution 213 using a pipette. Buffer solution was added to culture tubes, swirled around to cover the full surface, 214 followed by aspiration of the leftover solution. Coating was performed at RT in a sterile laminar flow 215 cabinet. 216 Monocyte tracking and quantification 217 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint To track the monocytes in the chip over time, fluorescent images were taken at 0, 24, 48, 72, and 96h 218 after addition of the cells. The images were acquired using the ImageXpress XLS Micro High Content 219 Imaging System (Molecular Devices) at 4 X magnification in combination with dichroic/emission 220 fluorescent filters for TRITC. After imaging, the plate was placed back on the interval rocker in the 221 incubator (37 ℃; 5% CO 2). Monocyte migration was determined using Fiji (ImageJ) software as 222 previously described.40 In short, a rolling ball background correction was applied to improve the signal-223 to-noise ratio of the acquired images. Rectangle selections were made to determine the number of 224 monocytes in each of the following compartments separately: the Donor compartment ( bottom 225 channel; endothelium), the ECM compartment, and the Renal compartment (top channel; epithelium). 226 Next, an automated thresholding approach was applied that created binary masks of the fluorescently 227 labeled monocytes followed by applying a particle detection algorithm to outline and label individual 228 monocytes. Additionally, monocytes present in the in- and outlet well of the Renal compartment were 229 counted manually and added to the number of monocytes of the Renal compartment. Values were 230 normalized to the number of monocytes present in the Donor compartment at the start of the assay 231 (0h) and represented as percentages. 232 Transepithelial/transendothelial electrical resistance 233 Transepithelial/transendothelial electrical resistance (TEER) was measured to determine the integrity 234 and permeability of the RPTEC and HUVEC barriers in the OrganoPlate. To this end, an automated 235 multichannel impedance spectrometer designed for use with the OrganoPlate was deployed 236 (OrganoTEER, MI-OT-1; MIMETAS) as previously described.41 In short, the electrode board was cleaned 237 with 70% ethanol (85825.360; VWR Chemicals) and was left to dry for 60 minutes in a laminar flow 238 cabinet. Single-point baseline measurements of culture were taken on day 6 . Therefore, 50 µL HBSS 239 was added to the in- and outlet well of the ECM (middle) channel and 50 µL assay medium was added 240 to the in - and outlets wells of the top and bottom channel. The plate was left static at room 241 temperature (RT) for 30 minutes to equilibrate before measurement. Next, the plate was placed in the 242 OrganoTEER device in a laminar flow cabinet and point impedance measurements were performed at 243 RT by frequency sweep from 1000 Hz to 150 kHz (amplitude 100; precision 0.5). Data analysis was 244 performed using the OrganoTEER software, which automatically generated the TEER values per chips 245 in ohms ( Ω). By multiplying these values with the surface area of the epithelial/endothelial ECM 246 interface, normalized values in Ω*cm2 were obtained. 247 WST-8 enzymatic assay 248 The enzymatic activity of the cultures after exposures +/-CAS was determined using the WST-8 assay 249 (96992; Sigma). Culture medium in the OrganoPlate was replaced with 25 µL WST -8 solution diluted 250 1:11 in assay medium in each in - and outlet of the top and bottom perfusion channel. The plate was 251 incubated for 30 min in the incubator (37°C, 5% CO2) on the rocker platform (7° angle, 8-min interval), 252 followed by absorbance measurements at 450nm using a Multiskan FC Microplate Photometer 253 (51119000; Thermo Scientific). As a positive control, 5% Triton X-100 (T8787; Sigma) (v/v) in assay 254 medium was added to a few chips before addition of the WST-8 solution and incubated for 10 min on 255 a rocker platform in the incubator. Measurements of the perfusion in- and outlets were corrected for 256 the background signal, separately for the Renal and Donor compartment. Data were normalized to the 257 no trigger control for either the Renal or Donor compartment and presented in percentages. 258 LDH release 259 Supernatants from the top channel containing the RPTEC tubule and from the bottom channel 260 containing the HUVEC vessel were collected separately after 96 -hour incubation. Samples of the in - 261 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint and outlet of the same chips were pooled. Lactate dehydrogenase (LDH) release was determined using 262 the Lactate Dehydrogenase Activity Assay Kit (MAK066; Sigma) according to manufacturer’s protocol. 263 Briefly, 2 µL sample was added in duplicate to a black glass bottom 384-well plate followed by addition 264 of 18 µL LDH assay buffer. A NADH standard curve (20 µL/well) was added to the plate in duplicate and 265 the plate was centrifuged (200g, 1 min) to have all solutions at the bottom of the plate. Next, 20 µ L 266 Master Reaction Mix was added to each well, placed static for 1 min at RT after which the absorbance 267 at 450 nm was measured using the Multiskan FC Microplate Photometer (Thermo Scientific). A 268 measurement was taken every 2 minutes for 50 minutes. Data from the timepoint before a sample 269 resulted in a higher measurement than the values of the highest concentration of the NADH standard 270 curve was used for analysis. Background subtraction was performed and LDH release (nmol/min per 271 mL) was determined using the following formula: 272 LDH release= NADH (nmol) x sample dilution factor reaction time (min) x sample volume (ml) 273 Cytokine release 274 Medium samples after 96-hour incubation in the OrganoPlate were collected as previously described 275 and used to determine the release of interleukin 6 (IL-6). Additionally, media samples of each condition 276 prepared and used for exposure on day 6 of culture were stored ( -80°C) and used for baseline 277 measurements (referred to as 0h samples). Release was determined by using the human IL-6 DuoSet 278 ELISA kit (DY206 -05; R&D systems) in combination with ELISA Ancillary Reagent Kit 2 (DY008; R&D 279 systems) according to manufacturer’s protocol. In short, undiluted 0-hour samples and 10x diluted 96-280 hour samples were added to a 96-well plate coated with the IL-6 capture antibody. Washing steps were 281 performed using an automated plate washer (MultiFlo FX; BioTek) and absorbance was measured at 282 450 nm and 570 nm using the Multiskan FC Microplate Photometer (Thermo Scientific). To correct for 283 optical imperfections, values of the 570 nm reading were subtracted from the 450 nm reading. IL -6 284 concentrations were calculated, corrected for the dilution factor, and presented in pg/mL. 285 Immunocytochemistry 286 Cultures in the OrganoPlate were fixed using 3.7% formaldehyde (252549; Sigma) in PBS for 10 min, 287 followed by two washing steps with PBS for 5 min. Subsequently, the cultures were permeabilized with 288 0.3% Triton X-100 (T8787; Sigma) in PBS for 10 min. Cultures were washed with 4% FBS in PBS and 289 incubated with blocking solution (2% FBS, 2% bovine serum albumin [BSA, A2153; Sigma], and 0.1% 290 Tween-20 [P9416; Sigma] in PBS) for 45 min. Primary antibody solution was incubated for 2,5-3h, after 291 which the secondary antibody was incubated for 30 min . An overview of the used antibodies can be 292 found in Table 1. Nuclei were stained with Hoechst 33342 (1:2000, H3570; Thermo Scientific) and a 293 few chips additionally with ActinRed ™ 555 reagent (R37112; Thermo Fisher). All incubation and 294 washing steps were performed on a rocker platform (7° angle, 2 -min interval) at RT. Cultures were 295 stored in PBS and transferred to a confocal high content imaging microscope for automated imaging 296 (Micro XLS-C, Molecular Devices). Images were acquired at 10 X magnification at 2 µm increments 297 along the height of the microfluidic channel. Image processing was performed using ImageJ to create 298 3D reconstructions and maximum projections. 299 Statistics and Data analysis 300 Images were processed and analyzed using Fiji (ImageJ) software. Data analysis was performed using 301 Excel (Microsoft Office 365; Microsoft Corp.) and GraphPad Prism (version 9.4; GraphPad Software 302 Inc.). All data are expressed as the mean ± standard deviation (SD) . Two treatment groups were 303 compared by unpaired t-test. Multiple group comparisons were analyzed by one-way ANOVA, two-way 304 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint ANOVA or mixed -effects analysis with Sidak’s or Tukey’s post hoc test. Statistical significance was 305 indicated by one or more asterisks and considered at p < 0.05. Independent experiments are indicated 306 by N, and replicates per experiments representing individual chips are indicated by n. 307 Table 1 Antibodies used for immunofluorescent staining. 308 Primary antibodies Antibody Supplier Cat. no Dilution Host CD31 Dako M0823 1:100 Mouse VE-cadherin Abcam Ab33168 1:1000 Rabbit ICAM-1 R&D systems BBA3 1:100 Mouse ZO-1 Thermo Scientific 61-7300 1:125 Rabbit CD45 R&D systems MAB1430 1:200 Mouse Acetylated Tubulin Sigma T6793 1:2000 Mouse Isotype control antibodies Antibody Supplier Cat. no Concentration Mouse IgG Thermo Scientific 08-6599 0.5 µg/ml Rabbit IgG Thermo Scientific 08-6199 0.5 µg/ml Secondary antibodies Antibody Supplier Cat. no Dilution Goat anti-mouse 488 Thermo Scientific A-11001 1:250 Donkey anti-rabbit 647 Sigma SAB4600177 1:250 309

310 Establishment of a tri-culture model of the proximal tubule 311 The renal proximal tubule is a critical part of the nephron responsible for blood filtration and regulating 312 the composition of urine, which is often affected and linked to many acute and chronic renal diseases.20 313 The renal proximal tubule consists of polarized epithelial cells that display a brush border on the apical 314 side and are supported by a basement membrane separating the tubular epithelial cells from the 315 interstitium. To maintain the microenvironment required for proper tubule functioning, blood vessels 316 play a crucial role and are near the epithelial cells to facilitate oxygen, nutrients, and removal of waste 317 products. Additionally, the blood vessel can facilitate the extravasation of immune cells in response to 318 an inflammatory stimulus as key aspect in many renal diseases (Figure 1C).42 319 The human proximal tubule environment was modeled using epithelial, endothelial, and immune cell 320 components grown against an ECM gel (Figure 1D). To establish this model, a collagen-I ECM gel was 321 patterned into the middle channel of the OrganoPlate 3 -lane Narrow ECM to study different aspects 322 of kidney function and inflammation (Figure 1A). The OrganoPlate 3-lane Narrow ECM consists of 40 323 microfluidic chips embedded in a standard 384-well microtiter plate. Each chip is comprised of three 324 channels that join in the center of the chip: two medium perfusion channels and a gel channel in the 325 middle in which an extracellular matrix (ECM) gel can be patterned through the presence of the 326 phaseguides (Figure 1B).43,44 The ECM gel compartment in this plate is half the width of the standard 327 OrganoPlate 3-lane 40. This allows more efficient migration of immune cells from the endothelial 328 compartment to the proximal tubule. Upon gelation of the ECM gel, RPTECs were seeded in the top 329 channel against this ECM gel. Next, HUVECs were added to the bottom channel. After several days in 330 culture both cell types formed confluent tubules, after which primary human monocytes were 331 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint introduced into the lumen of the endothelial vessel ( Figure 1E). As a result, a renal microfluidic tri -332 culture model was created including a proximal epithelial tubule (top) and an endothelial vessel with 333 fluorescently labeled monocytes (bottom) as demonstrated in Figure 1F. The development of this tri-334 culture model over time is demonstrated in supplementary Figure 1. 335 Characterization of the tri-culture model 336 After establishing the renal tri -culture model, expression of different epithelial and endothelial 337 markers was examined to assess the physiological relevance using an immunofluorescence -based 338 approach in combination with high-content microscopy. A confocal 3D reconstruction of the tri-culture 339 model confirmed the tubular structure and lumen formation of both the RPTEC and HUVEC cultures 340 (Figure 2A ). Actin and acetylated tubulin, visualizing the cytoskeleton and the primary cilia 341 respectively, were observed in both cell types. This could also be observed at higher magnification max 342 projection images (Figure 2B,C). Furthermore, RPTEC and HUVEC tubules demonstrated tight junction 343 formation as shown by immunofluorescence staining of ZO-1 (Figure 2B,C) and expression of adherens 344 junction markers VE-cadherin and CD31 in the HUVEC tubule ( Figure 2D,E). The presence of human 345 primary monocytes was shown by expression of CD45, a cell surface marker of nucleated 346 hematopoietic cells involved in regulation and activation of the immune system (Figure 2F). 347 After observing expression of adherens - and tight junction markers in the renal tri -culture, barrier 348 formation of the epithelial (RPTEC) and endothelial (HUVEC) tubules was examined using TEER 349 (transepithelial/endothelial electrical resistance) measurements. Barrier tightness was measured on 350 day 6 and showed a TEER of 9.4 ± 2.5 and 7.7 ± 3.9 Ω∙cm2 for RPTEC and HUVEC, respectively. 351 Assessing immune-mediated damage 352 Complement activation as well as recruitment and infiltration of immune cells such as monocytes are 353 key steps of an inflammatory response in the proximal tubule which can lead to significant damage. 354 This complex process involves cellular activation, release of pro-inflammatory cytokines, compromised 355 cell viability, and loss of cells. 45,46 Here, we examined the immune -mediated effects of renal 356 inflammation in the renal tri -culture model. To this end, renal inflammation was induced using 5% 357 complement activated serum (CAS) which was added to the RPTEC channel and allowed to form a 358 gradient in the chip (Figure 3A i, ii). Fluorescently labeled monocytes were added to the lumen of the 359 HUVEC vessel and were tracked in real-time using fluorescent microscopy (Figure 3A iii, iv). 360 To evaluate the immune-mediated effects on each part of the culture, the chip was divided into the 361 following compartments for analysis: Donor (HUVEC channel), ECM (ECM channel) and Renal (RPTEC 362 channel) ( Figure 3B). CAS exposure resulted in morphological alterations in the RPTEC channel as 363 demonstrated by cell clustering and slight tubular contraction observed with phase contrast imaging 364 after 96 hours compared to no trigger ( Figure 3C, supplementary Figure 2). Furthermore, increased 365 monocyte infiltration into the ECM compartment was observed upon exposure to CAS. 366 Immune-mediated upregulation of both acetylated tubulin, a marker for primary cilia, and ICAM -1 367 (intercellular adhesion molecule) within the RPTEC tubule were seen as well ( Figure 3D,E). A slight 368 increase in ICAM -1 expression was observed in the HUVEC vessel upon CAS exposure ( Figure 3E). 369 Expression of VE-Cadherin and CD31 in the HUVEC vessel (Donor), a marker for adherens junction and 370 vascular adhesion molecule respectively, did not seem to be affected by exposure to CAS (Figure 3F,G). 371 Cell culture supernatant was collected from the Donor and Renal compartment separately after 96h 372 exposure and used to assess the levels of interleukin 6 (IL-6). This cytokine plays a multifaceted role in 373 the immune system and has an influence on pro-inflammatory signaling, recruitment of immune cells, 374 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint directly impacting epithelial cells, and regulating tubulointerstitial fibrosis.47 As shown in Figure 3H, no 375 distinction in the detected IL-6 levels was found within the Donor compartment between conditions, 376 but the levels were significantly increased in the Renal compartment after exposure to CAS. IL-6 levels 377 at the start of the assay (0h) were found to be 200-2000 times lower compared to the end of the assay 378 (96h) (supplementary Figure 3). 379 LDH is a cytoplasmic enzyme that is being released into the culture supernatant due to compromised 380 plasma membranes and serves as an indicator for cell damage. 48 Immune-mediated injury was 381 confirmed by the increased LDH release in the inflamed renal compartment compared to the non -382 triggered control after 96h exposure ( Figure 3I). Supernatant collected at the start of the assay (0h 383 timepoint) showed comparable LDH levels between untriggered and triggered conditions 384 (supplementary Figure 3 ). Finally, the metabolic cell activity as an indicator of cell viability was 385 assessed using a WST -8 assay.49 No significant differences were observed in the Donor and Renal 386 compartment between inflamed and non -inflamed conditions ( Figure 3J). Overall, this tri -culture 387 setup with inflammatory trigger provided important insights into the complex processes involved in 388 renal inflammation and allowed us to study different aspects. 389 Monocyte migration under healthy and inflamed renal conditions 390 Monocytes and monocyte -derived cells have distinct roles in the proximal tubule under normal 391 conditions and are actively involved in the immune response during inflammation. 50,51 Here, we 392 examined the behavior of primary human monocytes in the context of renal inflammation by 393 determining their migratory behavior and compared it to non-inflamed culture conditions. 394 Primary human monocytes were labeled with a fluorescent Celltracker dye, added to the lumen of the 395 HUVEC vessel and tracked in real-time. The location of the monocytes within the chip was determined 396 at different timepoints and this enabled studying monocyte extravasation and recruitment towards 397 the site of inflammation. 398 Monocyte migration was examined over time by fluorescent imaging and confirmed elevated 399 migration in the inflamed condition compared to the non-triggered control condition (Figure 4A). The 400 development of renal inflammation, monocyte extravasation, and subsequent migration was shown 401 to be a dynamic process as captured in supplementary Figure 4 (video 1). Monocytes were observed 402 to patrol over the vessel wall before extravasating into the ECM, after which monocytes travelled 403 different distances with several eventually reaching the Renal compartment and some even returning 404 to the endothelial tubule. Furthermore, CAS exposure resulted in more, but also faster monocyte 405 migration as observed in both phase contrast and fluorescent imaging. 406 The fluorescent images were used to quantify monocyte adhesion to the endothelial vessel, one of the 407 first steps for monocyte trafficking in response to inflammatory signals. Here, the percentages of 408 monocytes that were present at the start of the assay in the non -triggered and CAS -triggered 409 conditions relative to the no trigger control were determined ( Figure 4B). Results showed that there 410 were no significant differences between the groups. Next, quantification of the images also allowed us 411 to study monocyte migration by determining the number of cells in each compartment of the chip at 412 different timepoints. The percentage of monocytes in the ECM and Renal compartment over time in 413 the non-triggered and CAS-triggered conditions relative to the number of monocytes at the start of 414 the assay in the Donor compartment are depicted in Figure 4C and 4D. Exposure to CAS significantly 415 increased monocyte migration into both compartments at all timepoints between 24-96 hours 416 compared to the non-triggered condition. Upon exposure to CAS, there is a steep significant increase 417 of monocytes migrating into the ECM compartment after 24 hours which stabilizes over time or even 418 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint slightly decreases (Figure 4C). This is likely linked to the observed linear increased monocyte migration 419 into the Renal compartment (Figure 4D). The distribution of the monocytes in the chip after 96 -hour 420 exposure to no trigger or CAS showed a significant difference in the Renal compartment (Figure 4E). 421 Assessing donor variability in the renal inflammation model 422 To determine the robustness and reliability of the renal inflammation model and its applicability for 423 personalized medicine and drug development, it is crucial to understand the effect of donor 424 differences. To examine donor variability, primary human monocytes from 4 different donors were 425 used to establish the renal tri -culture model and were challenged with CAS to induce inflammation. 426 Monocyte adhesion and migration were assessed and quantified using fluorescent imaging over time. 427 Monocyte adhesion at the start of the assay showed no significant differences between the no trigger 428 and CAS-exposed conditions for all donors ( Figure 5A). However, the variability within each donor is 429 different where donor 1 showed the highest variability and donor 4 the least. Next, the percentage of 430 monocytes in each compartment was determined relative to the number of monocytes in the Donor 431 compartment at the start of the assay (0h). Figures 5B-D show the results of the 96-hour timepoint in 432 the Donor, ECM, and Renal compartment, respectively. At this timepoint, clear differences between 433 donors could be observed in each of the compartments. Donor 3 revealed increased presence of 434 monocytes in the Donor compartment compared to the other donors. Upon exposure to CAS, a 435 decreased trend could be observed for all donors, and was significantly different for donor 3 (Figure 436 5B). In the ECM compartment, an increased percentage of monocytes was observed upon exposure to 437 CAS for all donors, and this was significantly increased for donors 2 and 3. Finally, the effect of CAS was 438 homogeneously observed in the Renal compartment as shown by significantly elevated migration for 439 all four donors. The biggest difference between no trigger and CAS was observed for donor 4. An 440 overview with data of the other timepoints of donor 1, 3 and 4 can be found in supplementary Figure 441 5. These graphs revealed further differences in dynamics of monocyte adhesion and migration over 442 time between donors. 443 Effect of immune modulatory compounds on monocyte migration 444 The renal inflammation assay demonstrated to be an effective tool to determine monocyte behavior 445 under healthy and inflamed conditions. Therefore, the applicability of the assay to evaluate effects of 446 immune modulatory compounds on monocyte migration was tested. Two compounds, namely 447 compound A and B were evaluated in the presence of 5% CAS where compound A targets the 448 inflammatory trigger and compound B targets the monocytes directly. 449 Based on the compound characteristics, different exposure strategies were applied. Different 450 concentrations of compound A were tested, and an optimal concentration was selected for further 451 experiments (data not shown). Here, the effect of adding compound A to the Renal compartment, 452 Donor compartment, or to both compartments was determined by using monocyte donor 4. 453 Compound B was added to the Renal compartment and tested at three different concentrations using 454 monocyte donor 1. 455 Both compounds showed significant inhibition of monocyte migration displayed in Figure 6B and 6D 456 by the decreased percentage of monocytes reaching the Renal compartment when exposed to both 457 CAS and compound A or B. Results showed that all the treatment administration methods tested for 458 compound A significantly inhibited monocyte migration into the Renal compartment after 96h 459 exposure ( Figure 6B). Also, migration into the ECM compartment was significantly inhibited upon 460 exposure to all compound A treatment methods as compared to CAS at 96h (Figure 6A). For compound 461 B, all three tested concentrations resulted in equivalent migration inhibition into the Renal 462 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint compartment (Figure 6D). On the contrary, monocyte migration into the ECM compartment was not 463 affected by compound B ( Figure 6C). Results of the Donor, ECM, and Renal compartment over time 464 can be found in supplementary Figure 6. In summary, these results showed that treatment with 465 immune modulatory compounds can diminish monocyte migration in the light of renal inflammation 466 in the tri-culture model. 467

468 Numerous renal diseases are recognized to be driven by the immune system, where dysregulated 469 immune responses are key contributors to their pathogenesis. Despite this recognition, the 470 development of targeted therapies has been challenging as knowledge of the underlying mechanism 471 and complex interactions remains insufficient. Recent advancements in the field offer promising 472 avenues for exploring the interplay between renal cells and immune cells and their role in the 473 development of renal inflammation and diseases. This study describes the establishment of a human 474 immunocompetent 3D in vitro co-culture model of the proximal tubule in a high -throughput 475 microfluidic platform that can be used to study renal functionality and inflammatory processes. The 476 model is especially valuable for its ability to incorporate monocytes, an understudied cell type that is 477 becoming increasingly relevant to renal diseases in CKD and beyond.52–54 478 The model incorporate d RPTEC in the top compartment and HUVECs in the bottom compartment 479 cultured under flow and in direct contact with a collagen -I ECM gel resulting in the formation of 480 polarized tubular structures. As an immune component, human primary monocytes of different donors 481 were added to the lumen of the endothelium. Renal inflammation was successfully induced using 482 complement activated serum as evident by epithelial morphological changes, increased expression of 483 adhesion molecules, release of pro-inflammatory cytokines, and reduced epithelial viability. Realtime 484 migratory behavior of monocytes showed increased extravasation and migration towards the ECM and 485 Renal compartment upon exposure to CAS with donor-to-donor differences observed. Finally, immune 486 modulatory compounds showed efficacious inhibition of monocyte migration under inflammatory 487 conditions in the microfluidic co-culture model. 488 The microfluidic platform used in this study is a custom-made prototype plate utilizing the same format 489 as the OrganoPlate 3-lane 40 platform but with different channel dimensions including a smaller ECM 490 channel. Accordingly, it facilitated enhanced migration of monocytes from the endothelium towards 491 the epithelium within a set timeframe (data not shown). Compared to migration studies in Transwell 492 systems, a microfluidic platform allowed for better control of the microenvironment, flexibility to 493 adjust experimental complexity, improved visualization and analysis, and higher throughput. 55–57 For 494 example, the flow rate and concentration gradients c ould be precisely controlled and adjusted to 495 recapitulate dynamic environmental cues. Furthermore, perfusion flow was generated without the use 496 of pumps by placing the OrganoPlate on an interval rocker, generating gravity -driven fluid flow in a 497 bidirectional manner. With the used rocker settings, a peak shear stress of 1.2 dyne/cm 2 was 498 established in both channels.58 This falls within the in vivo estimated range of 0.3 to 1.2 dyne/cm2 for 499 the human proximal tubule but is lower compared to the reported range of 4 to 12 dyne/cm2 for 500 endothelial cells.59,60 Even though the flow is bidirectional, fluid flow is an important cue for proximal 501 tubule cells and has been shown to increase polarization, transporter functionality, and morphology.61–502 63 Besides fluid shear stress, direct cell-ECM interaction also contributes to improved proximal tubule 503 characteristics,64 which is possible in the used microfluidic platform as there is no artificial membrane. 504 In line with these reports, we showed polarized 3D tubular structures with a clear lumen and 505 expression of actin and acetylated tubulin indicating the presence of the cytoskeleton and primary 506 cilia. 507 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint Barrier formation to control the passage of molecules through the cell layer is another crucial 508 characteristic of the proximal tubule.65,66 In our model, barrier formation was shown by expression of 509 tight and adherens junction markers ZO -1, in both RPTEC and HUVEC, and VE -cadherin and CD31 in 510 the HUVEC. Tightness of the barrier was assessed using TEER measurements and revealed an average 511 TEER of 9.4 ± 2.5 Ω∙cm2 for the RPTEC which is in line with the reported in vivo range of 6.6 to 11.6 512 Ω∙cm2.67,68 For HUVECS, an average TEER of 7.7 ± 3.9 Ω∙cm2 was observed . Direct in vivo TEER 513 measurements of HUVECs or blood vessels in general are lacking. However, in vitro HUVEC TEER values 514 have been reported to range from 10 to 1000 Ω∙cm2 and vary based on the experimental conditions 515 and environmental factors. 69,70 Although beyond the scope of this study, it would be interesting to 516 study the barrier function in real -time under inflammatory conditions through timelapse TEER 517 measurements as shown by Ehlers et al. for vascular inflammation.71 518 Key pathogenic factors of a pro -inflammatory response in renal tubules are complement activation 519 and immune cell infiltration both driving tubulointerstitial damage. 72,73 To induce renal inflammation 520 in the model, complement activated serum was added to the Renal compartment and allowed to form 521 a gradient in the chip. Morphological alterations in the RPTEC tubules were observed upon exposure 522 to CAS. Renal tubular epithelial cells can produce but also activate complement via the alternative 523 pathway.74,75 In turn, the tubular epithelium becomes injured and/or activated resulting in the 524 synthesis and release of pro -inflammatory cytokines, including TNF α and IL -6, release of reactive 525 oxygen species, and increased synthesis of matrix proteins.76 In line with these reports, we observed 526 significantly increased IL -6 release from the Renal compartment as well as increased ICAM -1 and 527 acetylated tubulin expression in RPTECs in the presence of CAS. Increased expression of ICAM -1 on 528 renal tubular epithelial cells under inflammatory conditions has been reported and can potentially 529 serve as a biomarker to predict disease progression.77–79 ICAM-1 can be expressed on both the apical 530 and basolateral side of the proximal epithelium and has an important role in regulating the response 531 to infiltrating immune cells. 80 In addition to increased expression in RPTECs, elevated ICAM -1 532 expression was also observed in the HUVEC tubule in accordance with literature indicating its role in 533 regulating transendothelial migration in response to inflammation.81 534 Acetylated tubulin is found in primary cilia, but also in the centrioles and flagella of epithelial cells, and 535 is regulating cell polarization, proliferation, development, and migration. 82 Dysregulated tubulin 536 acetylation has been associated with tubular cell dysfunction, and if not treated can advance to tubular 537 apoptosis and necrosis. 83 The increased expression and altered appearance of acetylated tubulin 538 observed in the RPTEC tubule indicate early signs of tubular damage. 539 Finally, viability of the cultures was determined by detecting released LDH and measuring enzymatic 540 cell activity using WST-8. Cells undergoing apoptosis, necrosis or other forms of damage will release 541 LDH from the cytoplasm into the supernatant as a signal of impaired cell membranes. 84 Under 542 inflammatory conditions, we observed significantly increased LDH in the Renal compartment 543 suggesting damaged RPTECs. This is not the case for the Donor compartment containing the HUVEC 544 culture. On the contrary, no clear effect upon CAS exposure was detected in the WST-8 assay. A possible 545 explanation for this might be that epithelial cells undergoing programmed cell death maintain 546 metabolic activity during certain stages of the process. Sustained metabolic activity during early stages 547 of apoptosis has been crucial for ensuring proper regulation of the controlled breakdown of cellular 548 components.85 To further investigate immune-mediated effects on the culture viability and to improve 549 our understanding of the type of cell death involved, it would be interesting to prolong the experiment 550 to check the viability at later stages and include other readouts to determine caspase activity and DNA 551 fragmentation.86 552 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint During complement activation, the anaphylatoxins C3a and C5a are produced and establish a gradient 553 primarily aimed to induce chemotaxis.73 These and other complement proteins activate resident and 554 infiltrating cells leading to production and secretion of pro -inflammatory cytokines that further 555 stimulate leukocyte recruitment and infiltration. 87 One of the first responders recruited to the 556 inflammatory tubulointerstitial microenvironment are monocytes, which adhere to activated 557 endothelium and transmigrate into the interstitial space. To study monocyte behavior in normal and 558 inflammatory conditions, primary human monocytes were fluorescently labeled and perfused through 559 the lumen of the HUVEC tubule. With the established workflow in our setup, we were able to visually 560 detect and quantify the number, location, and movement of monocytes in real-time. No differences in 561 monocyte adhesion to the endothelial vessel wall at the start of the assay was observed between the 562 triggered and non-triggered group. This was observed with each of the four monocyte donors. One 563 explanation could be that endothelial activation occurs gradually upon exposure to complement 564 proteins and requires additional time for the synthesis and expression of adhesion molecules.88 On the 565 contrary, significant increased monocyte migration was observed in the triggered condition towards 566 the ECM and Renal compartment. Overall, the impact of CAS was consistently observed across all 567 tested monocyte donors, indicating its efficacy as a robust trigger for evaluating monocyte migration 568 in the context of renal inflammation. 569 To examine the robustness of the assay, it is crucial to understand the effect of donor differences. 570 Hence, four monocyte donors were tested and differences in the number of adhering monocytes to 571 the endothelial vessel wall were observed, both at the beginning of the assay and over time. Moreover, 572 CAS proved to be a potent inducer of migration; however, the extent of migration and location of the 573 monocytes varied among donors. Monocytes can be classified in three major subpopulations, namely, 574 classical, intermediate, and non-classical. Distinguishment is mainly based on the expression of CD14 575 and CD16 on the surface and their function in homeostasis or disease including inflammatory response 576 and migratory potential. 89 This heterogeneity exists between the subsets but has also been 577 demonstrated between healthy and disease conditions and between individuals.89 This could explain 578 the observed differences in adhesion and migration between monocyte donors in our assay. 579 Understanding the role of each of the subtypes in the context of renal inflammation, may provide new 580 avenues for targeted therapies and personalized medicine. In an initial test, we examined the 581 migratory behavior of isolated monocyte subsets in triggered and non-triggered conditions (data not 582 shown). Although shown to be feasible, more time and effort is required to obtain an optimized 583 workflow. Additionally, further characterization of the morphology of the migrated monocytes into the 584 ECM channel could provide valuable information on the subsets as well as the differentiation towards 585 macrophages and the role of subtypes in renal inflammation.90 586 The imaging compatibility of the OrganoPlate enabled us to not only track and quantify monocyte 587 migration but also to examine morphological alterations in the epithelial and endothelial tubules and 588 their interaction with monocytes over time. As shown in supplementary figure 4, timelapse videos 589 were captured of the triggered and non-triggered cultures and revealed monocyte patrolling behavior 590 across the endothelial wall and differences in movements towards the Renal compartment. Using 591 advanced data analysis software or AI provides a potential next step to support our knowledge of 592 interactions in this complex microenvironment and detect abnormalities associated with disease 593 conditions. 594 To provide initial proof -of-concept data that this model can be used to assess investigational 595 therapeutics and serve as a preclinical predictive tool, the cultures were exposed to two immune 596 modulatory compounds under inflammatory conditions. Targeting different components of the 597 inflammatory environment, either the trigger or the monocytes, both compounds showed significant 598 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint reduction in monocyte migration into the Renal compartment. In addition to assessing the number of 599 migrated monocytes, this assay would also enable detection of the tendency of migration, such as 600 speed, distance, optimal time window, and location. Interestingly, the effect of both compounds varies 601 within the ECM channel, where compound A significantly decreased monocyte migration and 602 compound B did not exhibit the same effect. The level of migration between the no trigger and CAS 603 conditions varies between the compounds as different monocyte donors were utilized. To fully validate 604 the effect of the compounds, it would be advisable to evaluate them on multiple monocyte donors. 605 This could also provide valuable information for patient stratification in later stages of drug 606 development. Using patient -derived monocytes in the assay would enable the assessment of 607 migratory capacity within a renal inflammatory environment and facilitate personalized testing of 608 various (novel) therapeutics. 609 Crosstalk between the proximal tubule epithelium and peritubular capillaries is crucial for the 610 exchange of signals in the tubulointerstitium and has been shown to influence renal function both in 611 health and disease.91 Although the distance in our model exceeds the approximate distance of the in 612 vivo human tubular interstitium92, which generally does not surpass the diameter of a single cell, it is 613 smaller compared to other microphysiological co-culture models.93–95 Additionally, the current size of 614 the tubulointerstitial space in the Narrow ECM plate, represented by the ECM channel, facilitated 615 accurate monocyte quantification and enhances visual interpretability. To study the tubulo -vascular 616 crosstalk, secretion of soluble factors can be determined in our model as shown by the release of IL-6 617 and LDH. Further investigation of these factors with omics or ELISA analysis can advance our 618 understanding of the molecules secreted and altered under specific (disease) conditions, potentially 619 revealing novel targets for drug development, or identifying new biomarkers. 620 The versatility of the assay in modulating the system and controlling various environmental factors 621 including cell ratios, stimulus, exposure strategies, and real -time imaging represents significant 622 advantages which cannot be achieved with animal models. By exchanging the disease trigger, the 623 model could be applied for studying other specific renal diseases related to AKI or CKD, such as 624 ischemia-reperfusion injury. Finally, the complexity of the model can be tailored and further increased 625 by incorporation of fibroblasts or other immune cells, such as T cells (supplementary Figure 7) or 626 neutrophils. 627 In conclusion, we successfully established a human immunocompetent 3D in vitro co-culture model 628 comprising renal proximal tubule epithelial cells, endothelial cells, and perfused monocytes in a high-629 throughput microfluidic platform. We were able to induce renal inflammation using complement 630 activated serum and detect immune-mediated damage and associated monocyte migratory behavior. 631 Finally, we demonstrated the use of the assay for assessing effects of immune modulatory compounds 632 in the context of renal inflammation. The model can be applied for fundamental research on renal 633 functionality in health and disease but also for drug screening due to the platform’s compatibility with 634 automation and relatively high throughput. Overall, the described proximal tubule model has high 635 potential to fill the gap that currently exists to study renal inflammation preclinically. 636 Authorship contributions 637 Linda Gijzen: conceptualization, methodology, formal analysis, investigation, writing – original draft, 638 supervision, project administration Marleen Bokkers: methodology, formal analysis, investigation, 639 writing – review and editing Richa Hanamsagar: methodology, resources Thomas Olivier: software 640 Todd Burton: resources Laura Tool: formal analysis, investigation Mouly Rahman: methodology, 641 resources, writing – review and editing John Lowman: conceptualization, methodology, project 642 administration Virginia Savova: conceptualization, methodology, resources, writing – review and 643 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint editing, supervision Terry Means: conceptualization, methodology, resources, writing – review and 644 editing, supervision Henriette Lanz: conceptualization, writing – review and editing, supervision 645

646 We would like to thank Frederik Schavemaker for generating artist impressions of the model. We would 647 like to extend our gratitude to Shinji Kasahara, David Habiel and Vladimir Litvak for their contributions 648 and input to this work. 649 Disclosure and funding 650 The authors declared the following potential conflicts of interest with respect to the research, 651 authorship, and/or publication of this article: L.G., M.B., T.O., T.P.B., L.M.T., J.L., and H.L.L. are or were 652 employees of Mimetas BV, the Netherlands. T.K.M., V .S., M.F.R., and R.H. are or were employees of 653 Sanofi US at the time this work was performed . OrganoPlate, OrganoFlow, and OrganoTEER are 654 registered trademarks of Mimetas BV. 655 This research project was supported by funding from Sanofi US. 656

657 1. Bikbov B, Purcell CA, Levey AS, et al. Global, regional, and national burden of chronic kidney disease, 1990 –2017: a 658 systematic analysis for the Global Burden of Disease Study 2017. The Lancet . 2020;395(10225):709 -733. 659 doi:10.1016/S0140-6736(20)30045-3 660 2. Hill NR, Fatoba ST, Oke JL, et al. Global Prevalence of Chronic Kidney Disease – A Systematic Review and Meta -661 Analysis. PLoS One. 2016;11(7):e0158765. doi:10.1371/journal.pone.0158765 662 3. Ruiz-Ortega M, Rayego-Mateos S, Lamas S, Ortiz A, Rodrigues -Diez RR. Targeting the progression of chronic kidney 663 disease. Nat Rev Nephrol. 2020;16(5):269-288. doi:10.1038/s41581-019-0248-y 664 4. Romagnani P , Remuzzi G, Glassock R, et al. Chronic kidney disease. Nat Rev Dis Primers . 2017;3(1):17088. 665 doi:10.1038/nrdp.2017.88 666 5. Haroun MK, Jaar BG, Hoffman SC, Comstock GW, Klag MJ, Coresh J. Risk Factors for Chronic Kidney Disease. Journal 667 of the American Society of Nephrology. 2003;14(11):2934-2941. doi:10.1097/01.ASN.0000095249.99803.85 668 6. Mihai S, Codrici E, Popescu ID, et al. Inflammation -Related Mechanisms in Chronic Kidney Disease Prediction, 669 Progression, and Outcome. J Immunol Res. 2018;2018:1-16. doi:10.1155/2018/2180373 670 7. Hojs R, Ekart R, Bevc S, Hojs N. Markers of Inflammation and Oxidative Stress in the Development and Progression 671 of Renal Disease in Diabetic Patients. Nephron. 2016;133(3):159-162. doi:10.1159/000447434 672 8. Stenvinkel P , Heimbürger O, Paultre F, et al. Strong association between malnutrition, inflammation, and 673 atherosclerosis in chronic renal failure. Kidney Int. 1999;55(5):1899-1911. doi:10.1046/j.1523-1755.1999.00422.x 674 9. Ricklin D, Reis ES, Lambris JD. Complement in disease: a defence system turning offensive. Nat Rev Nephrol . 675 2016;12(7):383-401. doi:10.1038/nrneph.2016.70 676 10. Naicker SD, Cormican S, Griffin TP , et al. Chronic Kidney Disease Severity Is Associated With Selective Expansion of a 677 Distinctive Intermediate Monocyte Subpopulation. Front Immunol. 2018;9. doi:10.3389/fimmu.2018.02845 678 11. Tian S, Chen SY . Macrophage polarization in kidney diseases. Macrophage (Houst). Published online March 30, 2015. 679 doi:10.14800/macrophage.679 680 12. Thurman JM. Complement and the Kidney: An Overview. Adv Chronic Kidney Dis . 2020;27(2):86 -94. 681 doi:10.1053/j.ackd.2019.10.003 682 13. Willows J, Brown M, Sheerin NS. The role of complement in kidney disease. Clinical Medicine. 2020;20(2):156-160. 683 doi:10.7861/clinmed.2019-0452 684 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 14. Donate-Correa J, Luis-Rodríguez D, Martín-Núñez E, et al. Inflammatory Targets in Diabetic Nephropathy. J Clin Med. 685 2020;9(2):458. doi:10.3390/jcm9020458 686 15. Harris CL, Pouw RB, Kavanagh D, Sun R, Ricklin D. Developments in anti-complement therapy; from disease to clinical 687 trial. Mol Immunol. 2018;102:89-119. doi:10.1016/j.molimm.2018.06.008 688 16. Rayego-Mateos S, Valdivielso JM. New therapeutic targets in chronic kidney disease progression and renal fibrosis. 689 Expert Opin Ther Targets. 2020;24(7):655-670. doi:10.1080/14728222.2020.1762173 690 17. Licht C, Greenbaum LA, Muus P , et al. Efficacy and safety of eculizumab in atypical hemolytic uremic syndrome from 691 2-year extensions of phase 2 studies. Kidney Int. 2015;87(5):1061-1073. doi:10.1038/ki.2014.423 692 18. Greenbaum LA, Fila M, Ardissino G, et al. Eculizumab is a safe and effective treatment in pediatric patients with 693 atypical hemolytic uremic syndrome. Kidney Int. 2016;89(3):701-711. doi:10.1016/j.kint.2015.11.026 694 19. Hofherr A, Williams J, Gan LM, Söderberg M, Hansen PBL, Woollard KJ. Targeting inflammation for the treatment of 695 Diabetic Kidney Disease: a five -compartment mechanistic model. BMC Nephrol . 2022;23(1):208. 696 doi:10.1186/s12882-022-02794-8 697 20. Chevalier RL. The proximal tubule is the primary target of injury and progression of kidney disease: role of the 698 glomerulotubular junction. American Journal of Physiology -Renal Physiology . 2016;311(1):F145 -F161. 699 doi:10.1152/ajprenal.00164.2016 700 21. Nakhoul N, Batuman V. Role of Proximal Tubules in the Pathogenesis of Kidney Disease. In: ; 2011:37 -50. 701 doi:10.1159/000313944 702 22. Hosszu A, Kaucsar T, Seeliger E, Fekete A. Animal Models of Renal Pathophysiology and Disease. In: ; 2021:27 -44. 703 doi:10.1007/978-1-0716-0978-1_2 704 23. Liang J, Liu Y . Animal Models of Kidney Disease: Challenges and Perspectives. Kidney360. 2023;4(10):1479-1493. 705 doi:10.34067/KID.0000000000000227 706 24. Secker P . RPTEC/TERT1 cells form highly differentiated tubules when cultured in a 3D matrix. ALTEX. Published online 707 2018:223-234. doi:10.14573/altex.1710181 708 25. Tröndle K, Rizzo L, Pichler R, et al. Scalable fabrication of renal spheroids and nephron-like tubules by bioprinting and 709 controlled self-assembly of epithelial cells. Biofabrication. 2021;13(3):035019. doi:10.1088/1758-5090/abe185 710 26. Yousef Yengej FA, Jansen J, Ammerlaan CME, et al. Tubuloid culture enables long -term expansion of functional 711 human kidney tubule epithelium from iPSC -derived organoids. Proceedings of the National Academy of Sciences . 712 2023;120(6). doi:10.1073/pnas.2216836120 713 27. Schumacher A, Roumans N, Rademakers T, et al. Enhanced Microvasculature Formation and Patterning in iPSC –714 Derived Kidney Organoids Cultured in Physiological Hypoxia. Front Bioeng Biotechnol . 2022;10. 715 doi:10.3389/fbioe.2022.860138 716 28. Schutgens F, Rookmaaker MB, Margaritis T, et al. Tubuloids derived from human adult kidney and urine for 717 personalized disease modeling. Nat Biotechnol. 2019;37(3):303-313. doi:10.1038/s41587-019-0048-8 718 29. King SM, Higgins JW, Nino CR, et al. 3D Proximal Tubule Tissues Recapitulate Key Aspects of Renal Physiology to 719 Enable Nephrotoxicity Testing. Front Physiol. 2017;8. doi:10.3389/fphys.2017.00123 720 30. Lin NYC, Homan KA, Robinson SS, et al. Renal reabsorption in 3D vascularized proximal tubule models. Proceedings 721 of the National Academy of Sciences. 2019;116(12):5399-5404. doi:10.1073/pnas.1815208116 722 31. van Genderen AM, Valverde MG, Capendale PE, et al. Co -axial printing of convoluted proximal tubule for kidney 723 disease modeling. Biofabrication. 2022;14(4):044102. doi:10.1088/1758-5090/ac7895 724 32. Zhu Y , Shi Z, Ding W, Li C. On-chip construction of a fully structured scaffold-free vascularized renal tubule. Biomed 725 Microdevices. 2023;25(1):8. doi:10.1007/s10544-023-00648-0 726 33. Bas-Cristóbal Menéndez A, Du Z, van den Bosch TPP , et al. Creating a kidney organoid-vasculature interaction model 727 using a novel organ-on-chip system. Sci Rep. 2022;12(1):20699. doi:10.1038/s41598-022-24945-5 728 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 34. Sakolish C, Chen Z, Dalaijamts C, et al. Predicting tubular reabsorption with a human kidney proximal tubule tissue-729 on-a-chip and physiologically-based modeling. Toxicology in Vitro. 2020;63:104752. doi:10.1016/j.tiv.2019.104752 730 35. Kroll KT, Mata MM, Homan KA, et al. Immune-infiltrated kidney organoid-on-chip model for assessing T cell bispecific 731 antibodies. Proceedings of the National Academy of Sciences. 2023;120(35). doi:10.1073/pnas.2305322120 732 36. Vormann MK, Tool LM, Ohbuchi M, et al. Modelling and Prevention of Acute Kidney Injury through Ischemia and 733 Reperfusion in a Combined Human Renal Proximal Tubule/Blood Vessel -on-a-Chip. Kidney360. 2022;3(2):217-231. 734 doi:10.34067/KID.0003622021 735 37. Gijzen L, Yousef Yengej FA, Schutgens F, et al. Culture and analysis of kidney tubuloids and perfused tubuloid cells -736 on-a-chip. Nat Protoc. 2021;16(4):2023-2050. doi:10.1038/s41596-020-00479-w 737 38. Trietsch SJ, Israëls GD, Joore J, Hankemeier T, Vulto P . Microfluidic titer plate for stratified 3D cell culture. Lab Chip. 738 2013;13(18):3548-3554. doi:10.1039/c3lc50210d 739 39. Gijzen L, Marescotti D, Raineri E, et al. An Intestine -on-a-Chip Model of Plug -and-Play Modularity to Study 740 Inflammatory Processes. SLAS Technol. Published online 2020. doi:10.1177/2472630320924999 741 40. de Haan L, Suijker J, van Roey R, et al. A microfluidic 3D endothelium -on-a-chip model to study transendothelial 742 migration of T cells in health and disease. Int J Mol Sci. 2021;22(15). doi:10.3390/ijms22158234 743 41. Nicolas A, Schavemaker F, Kosim K, et al. High throughput transepithelial electrical resistance (TEER) measurements 744 on perfused membrane-free epithelia. Lab Chip. 2021;21(9):1676-1685. doi:10.1039/d0lc00770f 745 42. Imig JD, Ryan MJ. Immune and inflammatory role in renal disease. Compr Physiol . 2013;3(2):957 -976. 746 doi:10.1002/cphy.c120028 747 43. Vulto P , Podszun S, Meyer P , Hermann C, Manz A, Urban GA. Phaseguides: a paradigm shift in microfluidic priming 748 and emptying. Lab Chip. 2011;11(9):1596. doi:10.1039/c0lc00643b 749 44. Trietsch SJ, Israëls GD, Joore J, Hankemeier T, Vulto P . Microfluidic titer plate for stratified 3D cell culture. Lab Chip. 750 2013;13(18):3548. doi:10.1039/c3lc50210d 751 45. Andrade-Oliveira V, Foresto-Neto O, Watanabe IKM, Zatz R, Câmara NOS. Inflammation in Renal Diseases: New and 752 Old Players. Front Pharmacol. 2019;10. doi:10.3389/fphar.2019.01192 753 46. Schnaper HW. The Tubulointerstitial Pathophysiology of Progressive Kidney Disease. Adv Chronic Kidney Dis . 754 2017;24(2):107-116. doi:10.1053/j.ackd.2016.11.011 755 47. Su H, Lei CT, Zhang C. Interleukin -6 Signaling Pathway and Its Role in Kidney Disease: An Update. Front Immunol. 756 2017;8. doi:10.3389/fimmu.2017.00405 757 48. Legrand C, Bour JM, Jacob C, et al. Lactate dehydrogenase (LDH) activity of the number of dead cells in the medium 758 of cultured eukaryotic cells as marker. J Biotechnol. 1992;25(3):231-243. doi:10.1016/0168-1656(92)90158-6 759 49. Tominaga H, Ishiyama M, Ohseto F, et al. A water-soluble tetrazolium salt useful for colorimetric cell viability assay. 760 Analytical Communications. 1999;36(2):47-50. doi:10.1039/a809656b 761 50. Wang Y , Harris DCH. Macrophages in Renal Disease. Journal of the American Society of Nephrology . 2011;22(1):21-762 27. doi:10.1681/ASN.2010030269 763 51. Swaminathan S, Griffin MD. First responders: understanding monocyte-lineage traffic in the acutely injured kidney. 764 Kidney Int. 2008;74(12):1509-1511. doi:10.1038/ki.2008.555 765 52. Arazi A, Rao DA, Berthier CC, et al. The immune cell landscape in kidneys of patients with lupus nephritis. Nat 766 Immunol. 2019;20(7):902-914. doi:10.1038/s41590-019-0398-x 767 53. Mohan C, Zhang T, Putterman C. Pathogenic cellular and molecular mediators in lupus nephritis. Nat Rev Nephrol. 768 2023;19(8):491-508. doi:10.1038/s41581-023-00722-z 769 54. Nachiappa Ganesh R, Garcia G, Truong L. Monocytes and Macrophages in Kidney Disease and Homeostasis. Int J Mol 770 Sci. 2024;25(7):3763. doi:10.3390/ijms25073763 771 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 55. Li J, Lin F. Microfluidic devices for studying chemotaxis and electrotaxis. Trends Cell Biol . 2011;21(8):489 -497. 772 doi:10.1016/j.tcb.2011.05.002 773 56. Boyden S. THE CHEMOTACTIC EFFECT OF MIXTURES OF ANTIBODY AND ANTIGEN ON POLYMORPHONUCLEAR 774 LEUCOCYTES. J Exp Med. 1962;115(3):453-466. doi:10.1084/jem.115.3.453 775 57. Zhang C, Barrios MP , Alani RM, Cabodi M, Wong JY . A microfluidic Transwell to study chemotaxis. Exp Cell Res . 776 2016;342(2):159-165. doi:10.1016/j.yexcr.2016.03.010 777 58. Vormann MK, Gijzen L, Hutter S, et al. Nephrotoxicity and Kidney Transport Assessment on 3D Perfused Proximal 778 Tubules. AAPS Journal. 2018;20(5). doi:10.1208/s12248-018-0248-z 779 59. Raghavan V, Rbaibi Y , Pastor -Soler NM, Carattino MD, Weisz OA. Shear stress -dependent regulation of apical 780 endocytosis in renal proximal tubule cells mediated by primary cilia. Proceedings of the National Academy of 781 Sciences. 2014;111(23):8506-8511. doi:10.1073/pnas.1402195111 782 60. dela Paz NG, Walshe TE, Leach LL, Saint -Geniez M, D’ Amore PA. Role of shear-stress-induced VEGF expression in 783 endothelial cell survival. J Cell Sci. 2012;125(4):831-843. doi:10.1242/jcs.084301 784 61. Ross EJ, Gordon ER, Sothers H, et al. Three dimensional modeling of biologically relevant fluid shear stress in human 785 renal tubule cells mimics in vivo transcriptional profiles. Sci Rep. 2021;11(1):14053. doi:10.1038/s41598-021-93570-786 5 787 62. Vriend J, Peters JGP , Nieskens TTG, et al. Flow stimulates drug transport in a human kidney proximal tubule -on-a-788 chip independent of primary cilia. Biochimica et Biophysica Acta (BBA) - General Subjects. 2020;1864(1):129433. 789 doi:10.1016/j.bbagen.2019.129433 790 63. Jang KJ, Mehr AP , Hamilton GA, et al. Human kidney proximal tubule-on-a-chip for drug transport and nephrotoxicity 791 assessment. Integrative Biology. 2013;5(9):1119-1129. doi:10.1039/c3ib40049b 792 64. Lacueva-Aparicio A, Lindoso RS, Mihăilă SM, Giménez I. Role of extracellular matrix components and structure in 793 new renal models in vitro. Front Physiol. 2022;13. doi:10.3389/fphys.2022.1048738 794 65. Brown D, Stow JL. Protein trafficking and polarity in kidney epithelium: from cell biology to physiology. Physiol Rev. 795 1996;76(1):245-297. doi:10.1152/physrev.1996.76.1.245 796 66. FANNING AS, MITIC LL, ANDERSON JM. Transmembrane Proteins in the Tight Junction Barrier. Journal of the 797 American Society of Nephrology. 1999;10(6):1337-1345. doi:10.1681/ASN.V1061337 798 67. Boulpaep E, Seely J. Electrophysiology of proximal and distal tubules in the autoperfused dog kidney. American 799 Journal of Physiology-Legacy Content. 1971;221(4):1084-1096. doi:10.1152/ajplegacy.1971.221.4.1084 800 68. Liang M, Ramsey CR, Knox FG. The paracellular permeability of opossum kidney cells, a proximal tubule cell line. 801 Kidney Int. 1999;56(6):2304-2308. doi:10.1046/j.1523-1755.1999.00787.x 802 69. Dewi BE, Takasaki T, Kurane I. In vitro assessment of human endothelial cell permeability: effects of inflammatory 803 cytokines and dengue virus infection. J Virol Methods. 2004;121(2):171-180. doi:10.1016/j.jviromet.2004.06.013 804 70. Srinivasan B, Kolli AR, Esch MB, Abaci HE, Shuler ML, Hickman JJ. TEER Measurement Techniques for In Vitro Barrier 805 Model Systems. SLAS Technol. 2015;20(2):107-126. doi:10.1177/2211068214561025 806 71. Ehlers H, Nicolas A, Schavemaker F, et al. Vascular inflammation on a chip: A scalable platform for trans-endothelial 807 electrical resistance and immune cell migration. Front Immunol. 2023;14. doi:10.3389/fimmu.2023.1118624 808 72. KHAN TN, SINNIAH R. Role of complement in renal tubular damage. Histopathology. 1995;26(4):351 -356. 809 doi:10.1111/j.1365-2559.1995.tb00197.x 810 73. Fearn A. Complement activation in progressive renal disease. World J Nephrol . 2015;4(1):31. 811 doi:10.5527/wjn.v4.i1.31 812 74. Biancone L, David S, Pietra V Della, Montrucchio G, Cambi V, Camussi G. Alternative pathway activation of 813 complement by cultured human proximal tubular epithelial cells. Kidney Int . 1994;45(2):451 -460. 814 doi:10.1038/ki.1994.59 815 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 75. Naughton MA, Botto M, Carter MJ, Alexander GJ, Goldman JM, Walport MJ. Extrahepatic secreted complement C3 816 contributes to circulating C3 levels in humans. J Immunol. 1996;156(8):3051-3056. 817 76. David S, Biancone L, Caserta C, Bussolati B, Cambi V, Camussi G. Alternative pathway complement activation induces 818 proinflammatory activity in human proximal tubular epithelial cells. Nephrology Dialysis Transplantation . 819 1997;12(1):51-56. doi:10.1093/ndt/12.1.51 820 77. Daniel L, Sichez H, Giorgi R, et al. Tubular lesions and tubular cell adhesion molecules for the prognosis of lupus 821 nephritis. Kidney Int. 2001;60(6):2215-2221. doi:10.1046/j.1523-1755.2001.00055.x 822 78. Shimizu H, Yisireyili M, Higashiyama Y , Nishijima F, Niwa T. Indoxyl sulfate upregulates renal expression of ICAM-1 via 823 production of ROS and activation of NF -κB and p53 in proximal tubular cells. Life Sci . 2013;92(2):143 -148. 824 doi:10.1016/j.lfs.2012.11.012 825 79. Wen L, Zhao Z, Li F, Ji F, Wen J. ICAM-1 related long noncoding RNA is associated with progression of IgA nephropathy 826 and fibrotic changes in proximal tubular cells. Sci Rep. 2022;12(1):9645. doi:10.1038/s41598-022-13521-6 827 80. Zhang XL, Selbi W, de la Motte C, Hascall V, Phillips A. Renal Proximal Tubular Epithelial Cell Transforming Growth 828 Factor-β1 Generation and Monocyte Binding. Am J Pathol . 2004;165(3):763 -773. doi:10.1016/S0002 -829 9440(10)63339-1 830 81. Bui TM, Wiesolek HL, Sumagin R. ICAM-1: A master regulator of cellular responses in inflammation, injury resolution, 831 and tumorigenesis. J Leukoc Biol. 2020;108(3):787-799. doi:10.1002/JLB.2MR0220-549R 832 82. Han SJ, Kim JH, Kim JI, Park KM. Inhibition of microtubule dynamics impedes repair of kidney ischemia/reperfusion 833 injury and increases fibrosis. Sci Rep. 2016;6(1):27775. doi:10.1038/srep27775 834 83. Nekooki-Machida Y , Hagiwara H. Role of tubulin acetylation in cellular functions and diseases. Med Mol Morphol . 835 2020;53(4):191-197. doi:10.1007/s00795-020-00260-8 836 84. Kumar P , Nagarajan A, Uchil PD. Analysis of Cell Viability by the Lactate Dehydrogenase Assay. Cold Spring Harb 837 Protoc. 2018;2018(6):pdb.prot095497. doi:10.1101/pdb.prot095497 838 85. Zhao L, Hao Y , Tang S, Han X, Li R, Zhou X. Energy metabolic reprogramming regulates programmed cell death of 839 renal tubular epithelial cells and might serve as a new therapeutic target for acute kidney injury. Front Cell Dev Biol. 840 2023;11. doi:10.3389/fcell.2023.1276217 841 86. Kari S, Subramanian K, Altomonte IA, Murugesan A, Yli -Harja O, Kandhavelu M. Programmed cell death detection 842

a systematic review and a categorical comparison. Apoptosis. 2022;27(7-8):482-508. doi:10.1007/s10495-843 022-01735-y 844 87. Tesch GH. Diabetic nephropathy – is this an immune disorder? Clin Sci . 2017;131(16):2183 -2199. 845 doi:10.1042/CS20160636 846 88. Fischetti F, Tedesco F. Cross-talk between the complement system and endothelial cells in physiologic conditions and 847 in vascular diseases. Autoimmunity. 2006;39(5):417-428. doi:10.1080/08916930600739712 848 89. Williams H, Mack C, Baraz R, et al. Monocyte Differentiation and Heterogeneity: Inter -Subset and Interindividual 849 Differences. Int J Mol Sci. 2023;24(10):8757. doi:10.3390/ijms24108757 850 90. Wang Y , Harris DCH. Macrophages in Renal Disease. Journal of the American Society of Nephrology . 2011;22(1):21-851 27. doi:10.1681/ASN.2010030269 852 91. Tasnim F, Zink D. Cross talk between primary human renal tubular cells and endothelial cells in cocultures. American 853 Journal of Physiology-Renal Physiology. 2012;302(8):F1055-F1062. doi:10.1152/ajprenal.00621.2011 854 92. Lemley K V., Kriz W. Anatomy of the renal interstitium. Kidney Int. 1991;39(3):370-381. doi:10.1038/ki.1991.49 855 93. Chapron A, Chapron BD, Hailey DW, et al. An Improved Vascularized, Dual -Channel Microphysiological System 856 Facilitates Modeling of Proximal Tubular Solute Secretion. ACS Pharmacol Transl Sci . 2020;3(3):496 -508. 857 doi:10.1021/acsptsci.9b00078 858 94. Shaughnessey EM, Kann SH, Azizgolshani H, Black LD, Charest JL, Vedula EM. Evaluation of rapid transepithelial 859 electrical resistance (TEER) measurement as a metric of kidney toxicity in a high -throughput microfluidic culture 860 system. Sci Rep. 2022;12(1):13182. doi:10.1038/s41598-022-16590-9 861 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 95. Yin L, Du G, Zhang B, et al. Efficient Drug Screening and Nephrotoxicity Assessment on Co-culture Microfluidic Kidney 862 Chip. Sci Rep. 2020;10(1):6568. doi:10.1038/s41598-020-63096-3 863 864 865 866 867 Figure 1. Establishment of the tri-culture kidney model in the OrganoPlate 3-lane Narrow ECM. (A) The 868 OrganoPlate 3-lane, a culture platform comprised of 40 microfluidic cell culture chips embedded in a 869 standard 384-well microtiter plate. ( B) Schematic image of the horizontal and vertical view of one 870 microfluidic cell culture chip consisting of 3 channels each connected to an in- and outlet well. These 871 3 channels join in the center of the chip (observation window; OW) and exist of 2 perfusion channels 872 and an extracellular matrix (ECM) gel channel in the middle separated by phaseguides. ( C) Simplistic 873 schematic representation of the human proximal tubule composition. On top, there is a tubular 874 structure composed of epithelial cells from the proximal tubule, featuring a lumen (yellow) through 875 which the glomerular filtrate flows. Adjacent to the proximal tubule, there is an endothelial vessel 876 (red) carrying blood enriched with immune cells like monocytes (purple), separated by a basement 877 membrane composed of extracellular matrix proteins (blue). (D) 3D artist impression of the human tri-878 culture kidney model in the microfluidic chip comprising human renal proximal tubule epithelial cells 879 (RPTEC; yellow), a collagen I ECM gel (blue), human umbilical vein endothelial cells (HUVEC; red) and 880 monocytes (green). (E) Seeding strategy for establishing the tri-culture model. A collagen I ECM gel is 881 patterned into the middle channel of the chip (i), followed by the addition of RPTECs in the top 882 perfusion channel (ii). Next, the HUVECs are added to the bottom perfusion channel (iii) and after 883 formation of confluent tubular structures in both channels, monocytes are added to the lumen of the 884 endothelial vessel in medium (iv). ( F) Phase contrast with fluorescent overlay image of the model 885 comprising a RPTEC tubule in the top perfusion channel and a HUVEC vessel with perfused 886 fluorescently labeled monocytes in the bottom perfusion channel at day 6 of culture, right after adding 887 the monocytes. Scale bars in white = 200 µm. 888 889 890 891 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 892 Figure 2. Characterization of the kidney on -a-chip model. (A) 3D reconstruction of a confocal z -stack 893 showing both tubular structures with clear lumen grown against a collagen I ECM in the 3-lane Narrow 894 ECM plate at 10X magnification. Cultures were stained for acetylated tubulin (red), actin (yellow), and 895 DNA (blue). (B,C) Representative immunofluorescent max projections of the RPTEC or HUVEC cultures 896 respectively, stained for acetylated tubulin (green), zonula occludens I (ZO-1; red), and DNA (blue). (D) 897 Max projection of the HUVEC culture stained for VE-cadherin (red) and DNA (blue). (E) Max projection 898 of the HUVEC culture stained for CD31 (green) and DNA (blue). ( F) Max projection indicating part of 899 the HUVEC tubule containing monocytes and ECM compartment with a phaseguide in between. 900 Cultures were stained for CD45 (red) and DNA (blue). (G) Epithelial and endothelial barrier function in 901 the model was assessed by measuring TEER at day 6 of culture (n=9-10). All cultures were fixed on day 902 10 of culture. Scale bars in white = 50 µm (A,D-F) or 25 µm (B,C). 903 904 905 906 907 908 909 910 911 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 912 Figure 3. Assessing immune-mediated effects of renal inflammation in the tri -culture model. (A) To 913 induce inflammation in the model, a trigger comprised of 5% complement activated serum (CAS) in 914 culture medium was added to the lumen of the RPTEC tubule in the top channel (i) and incubated for 915 4 hours to establish a gradient (ii). Next, fluorescently labeled monocytes were added to the lumen of 916 the endothelial vessel (iii) and migration was tracked by collecting fluorescent images at different 917 timepoints (iv). ( B) Schematic representation of the model indicating the nomenclature for each 918 culture channel, namely the donor compartment (bottom perfusion channel), the ECM compartment 919 and the renal compartment (top perfusion channel). (C) Phase contrast images of the tri-culture model 920 exposed to no trigger (culture medium) or to 5% CAS for 96h. (D) Acetylated tubulin (green) and DNA 921 (blue) immunofluorescent staining in the Renal compartment in non -triggered and CAS -exposed 922 conditions. ( E) Representative immunofluorescent max projections of the Renal and Donor 923 compartment exposed to no trigger or CAS for 96h and stained for ICAM-1 (red) and DNA (blue). (F, G) 924 Max projection of the Donor compartment in non -triggered and CAS-exposed conditions stained for 925 VE-cadherin (red) or CD31 (green) and DNA (blue) (H) Secretion of interleukin 6 (IL-6) was assessed in 926 the Renal and Donor compartment in the non -triggered and CAS -exposed conditions at 96h post -927 exposure. (I) Lactate dehydrogenase (LDH) release in supernatant collected from the Donor and Renal 928 compartment after 96h exposure to no trigger or CAS. (J) Enzymatic activity of the cultures in the Donor 929 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint and Renal compartment exposed to no trigger or CAS was determined using WST-8. Triton X-100 was 930 included as a positive control. Graphs showing mean ± standard deviation and individual chips. Data 931 figure D-I monocyte donor 2; data figure J monocyte donor 1. IL-6: n=5, LDH: n=18-20, WST-8: n=2-11. 932 Statistical analysis was performed using multiple unpaired t tests for IL-6 and LDH or two-way ANOVA 933 with Sidak’s multiple comparisons test for WST-8. ***P < 0.001, ****P < 0.0001. All cultures were fixed 934 on day 10 of culture. Scale bars in white = 200 µm (C), 50 µm (D,E) or 100 µm (F,G). 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 953 954 Figure 4. Assessing monocyte migration under healthy or inflamed renal conditions. (A) Fluorescent 955 images of the tri-culture model indicating the location of the monocytes at different timepoints in the 956 no trigger or CAS-exposed condition. Monocytes were labeled using Celltracker Orange CMRA (TRITC). 957 To visualize the distance/location of the monocytes within a chip over time, each specific timepoint is 958 displayed with a different color. (B) Number of monocytes in the Donor compartment at the start of 959 the assay (0h) normalized to the no trigger control. (C,D) Percentage of monocytes that migrated into 960 the ECM or Renal compartment of the chip relative to number of monocytes at start of the assay in 961 the Donor compartment depicted over time comparing no trigger against CAS-exposed condition. (E) 962 Percentage of monocytes present in each compartment of the chip (Donor, ECM or Renal) after 96h in 963 the no trigger or CAS -exposed conditions. Graphs showing mean ± standard deviation or individual 964 chips, N=2, n=10-22 chips. Statistical analysis was performed using unpaired t test (B), mixed -effects 965 analysis (C,D), or two-way ANOVA (E); ns = not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P 966 < 0.0001. Scale bars in white = 200 µm. 967 968 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 969 Figure 5. Evaluating migration and adhesion of different monocyte donors. (A) Number of monocytes 970 in the Donor compartment at the start of the assay (0h) normalized to the no trigger control separately 971 for each monocyte donor. (B-D) Percentage of monocytes present in each compartment of the chip 972 (Donor, ECM or Renal) after 96h exposure to no trigger or to 5% CAS represented for 4 different 973 monocyte donors. Graphs showing mean and individual chips. Donor 1: N=2, n -10-22; Donor 2: N=1, 974 n=7-9; Donor 3: N=1, n=4 -10; Donor 4: N=2, n=6 -7. Statistical analysis was performed using two-way 975 ANOVA; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 976 977 978 979 980 981 982 983 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint 984 Figure 6. Effect of immune modulatory compounds on monocyte migration. Compound A was added 985 to the Renal compartment, Donor compartment or to both compartments of the chip and tested using 986 monocyte donor 4. Compound B was tested at 3 different concentrations (#1, 2 or 3), added to the 987 Renal compartment, and tested using monocyte donor 1. ( A,C) Percentage of monocytes present in 988 the ECM compartment after exposure to no trigger, CAS, or CAS + Compound A or B respectively after 989 96h. (B,D) Percentage of monocytes present in the Renal compartment after exposure to no trigger, 990 CAS, or CAS + Compound A or B after 96h. Graphs showing mean and individual chips. Compound A: 991 N=1, n=5-6; Compound B: N=1, n=2-8. Statistical analysis was performed using one-way ANOVA; ns = 992 not significant, *P<0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. 993 .CC-BY-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted June 12, 2024. ; https://doi.org/10.1101/2024.06.11.598417doi: bioRxiv preprint