CD11c-YFP (Itgax-Venus) mice and C57BL/6 mice were purchased form The Jackson Laboratory (USA) and Junbiotech (Gyeonggi, Korea). CD11c-YFP mice were used for imaging experiments, while C57BL/6 mice were used for histological analysis and immunofluorescence staining for F4/80. To induce acute kidney injury, CD11c-YFP mice were intraperitoneally injected with LPS (5 µg/g), whereas the control group (n=7) was injected with the same volume of 0.9% saline. The LPS-treated mice were pretreated with either MCC950 (10µg/g, intraperitoneally, n=9) or 0.9% saline (n=7) 1 hour before LPS administration. Mice were euthanized, and kidneys were harvested 24 hours after LPS treatment. Mice received 10 μg of Alexa Fluor ^® 647 anti-mouse CD31 antibody (BioLegend^®, 102516, 0.5 μg/μl) via the tail vein 10 min before euthanasia. Blood urea nitrogen and serum creatinine levels were determined using the VetTest 8008 (IDEXX Laboratories, Westbrook, ME, USA). All animal experiments were performed in compliance with the guidelines of the Animal Research Ethics Committee of Kyung Hee University Hospital at Gangdong (KHNMC AP 2019-014).

Kidney tissues were fixed with 10% neutral buffered formalin at 4°C for 24 hours and embedded in paraffin. Paraffin blocks were cut into 4-μm-thick sections and stained periodic-acid sciff (PAS) reagent. After staining with PAS, morphologic changes were assessed. Tubular epithelial cell vacuolar deformation/hypertrophy and luminal occlusion occurred during renal tubular injury. We assessed tissue injury using the semiquantitative tubular injury score used in previous studies (15, 16). The areas of the tubular epithelial cell vacuolar deformation, loss of brush border, tubular dilation, cast formation and cell lysis were calculated. A score of 0 indicated no change; scores of 1, 2, 3 or 4 indicated an injury section involving <25, 25–50, 50–75, or >75% of the area, respectively. A minimum of ten different fields of vision were randomly selected using a light microscope (magnification x400) and an average score was calculated. For immunofluorescence staining, cryosections were air-dried for 15 min, then primary antibodies against the following proteins were used: F4/80 (rat, 1:50, BioRad, USA). The slides were then exposed to AlexaFluor 488-conjugated goat anti-Rat IgG (1:200) secondary antibodies (Thermo Fisher Scientific, USA). Slides were counterstained with DAPI to delineate the nuclei. Representative images were taken with confocal microscopes LSM-700 (Carl Zeiss, Germany). The number of F4/80+ cells in the tubulointerstitium was counted in ten consecutive fields under high-power fields (x63 oil lens).

RNA was extracted from the kidney tissue using the Total RNA Isolation Kit (Macherey-Nagel, Düren, Germany) and was reverse-transcribed into complementary DNA (cDNA) using random primers (Promega, Madison, WI, USA), deoxynucleotide mix (Takara Bio Inc., Shiga, Japan), and Maloney-murine leukemia virus reverse transcriptase (Mbiotech Inc., Gyeonggi, Korea). Thereafter, real-time PCR was run on the Applied Biosystems^® StepOnePlus™ System with a final volume of 20 µl containing 1 µl of cDNA, 5–10 pmol of each of the sense and antisense primers, 10 µl of Power SYBR Green Master Mix (Applied Biosystems, Foster City, CA, USA), and nuclease-free water. The primer sequences used 18S: 5’-GTAACCCGTTGAACCCCATT -3’, 5’ -CCATCCAATCGGTAGTAGCG -3’, IL-6: 5’-ATGGATGCTACCAAACTGGAT -3’, 5’-TGAAGGACTCTGGCTTTGTCT-3’, TGF-beta: 5’-CAACAATTCCTGGCGTTACCTTGG-3’,5’GAAAGCCCTGTATTCCGTCTCCTT-3’, CTGF: 5’ -GAGTGTGCACTGCCAAAGAT -3’ 5’GGCAAGTGCATTGGTATTTG -3’. Each sample was run in duplicate, and data were analyzed using the ∆∆CT method with normalization to 18S expression.

We cleared the kidneys with the Tissue Clearing Kit from Binaree (Daegu, Korea). Briefly, formalin-fixed CD11c-YFP mouse kidneys were sectioned into 2-mm-thick tissues. The sections were transferred into the Fixing solution at 4°C for 6 hours and then into the Tissue clearing solution A at 42°C. After 4 days, kidneys were washed with 3mL of Washing solution for 8 hours at 42°C. Next, tissues were transferred in 2mL of Clearing solution B for 2 days at 42°C. Likewise, kidneys were washed with 3mL of Washing solution for 8 hours at 35°C, immersed in 3mL of Mounting solution at 35°C, and maintained for more than 12 hours. After washing, tissue clearing process was continued in the solution B at 42°C for another 2 days. The sections were stored in the Mounting Solution until imaged.

We used the light sheet fluorescence microscope (Lightsheet Z.1, Carl zeiss, Seoul Center, Korea Basic Science Institute) equipped with an EC Plan-Neofluar 5x/0.16 objective and clarity chamber (5x, n:1.45). The laser wavelengths used were 488 nm and 638 nm, and the laser intensity was measured under the same conditions. Whole kidney images were constructed from 6~9 tiles with 10% overlap and merged into a single 3D image stack using Arivis software (Arivis AG, Munich, Germany). Three-dimensional rendering of LSFM and laser scanning microscopy data was performed by Imaris 9.5.0 (Bitplane, Switzerland). Quantitative analysis of LSFM data was performed using 3D stacks of 161 images with a size of 863 × 863 × 650 μm acquired from five fields of view in each group. Selected 3D stacks included 40 to 50 glomeruli.

MPM images were acquired using a multiphoton microscope (LSM780 NLO, Zeiss, Germany) with a dual band femtosecond pulse laser system (Spectra Physics InSight DS+, USA) at Seoul Center in Korea Basic Science Institute. For 3D reconstruction of MPM images, ImageJ (National Institutes of Health, Bethesda, MD, USA) and Imaris were used. MPM images were quantitatively analyzed using five 3D stacks of 75 images with a size of 354 × 354 × 150 μm for each group. Single plane data was analyzed with five representative images selected from MPM image stacks. Maximum intensity projection (MIP) images were generated from the stacks of z-plane images using the Z project function of ImageJ. The area of CD11c+ cells was calculated as (CD11c-YFP+ area/total area) × 100. In addition, 3D stacks of 49 MPM images (100 × 100 × 100 μm) acquired from five fields in each group were analyzed for morphometric analysis of dendrites.

Three-dimensional quantitative analysis was performed using Imaris modules. The CD11c-YFP signals were surface-rendered using Imaris surface module. Background subtraction (local contrast) was used, and the surface detail was set to 0.8 μm (smoothness). The upper threshold was determined automatically, and the lower threshold was manually selected to be 10% of the upper threshold. The surface objects for glomeruli were selectively reconstructed from CD31 signals using a diameter of the largest sphere of 65 μm (background subtraction; surface detail, 1.5 μm). Surface objects incorrectly selected as glomeruli were removed using the surface editing tools. The glomerular volume was measured by dividing the total volume of glomerular surface objects by the total number of glomeruli. After surface rendering, unspecific background signals were removed by setting voxels outside surface to zero.

The total number of dendrites and dendrite segments was quantified within the selected 3D stacks. A single cell is defined by its corresponding surface object. Dendrites were selected using autopath algorithm (no loop) of Imaris filament module based on the local contrast; dendrites are defined as sequences of vertices and edges with branching, and dendrite segment indicates the part between branch points (or end points). The starting point (largest diameter) and seed point (thinnest diameter) of dendrites were set to 6 μm and 3 μm, respectively. The thresholds for the starting points and seed points were automatically determined. The threshold for the starting points was adjusted if there were multiple or no starting points in each cell. The total number of dendrites was estimated from the number of dendrite terminal parts provided from the Imaris filament module. Dendrite branch level was hierarchically assigned according to the diameter calculations of individual dendrite segments and averaged by the number of dendrite segments. All parameters of surface and filament objects were exported from Imaris for statistical analysis.

Isolated MNPs from kidneys were incubated with surface markers anti-mouse CD86 (PE/Cy7; BD Biosciences) and MHC class II (PE/Dazzle 594; BD Biosciences). Cells were analyzed on a fluorescence-activated cell sorting (FACS) Caliber flow cytometry system (BD Biosciences), and flow cytometric data were analyzed using FlowJo software version 10 (TreeStar, Ashland, Oregon).

All values are expressed as the mean ± standard error of the mean (SEM) The results were analyzed using the Kruskal-Wallis nonparametric test for multiple comparisons. Significant differences in the Kruskal-Wallis test were confirmed by the Wilcoxon rank sum and Mann-Whitney tests (used to compare mean differences). P values < 0.05 were considered statistically significant.

We used three animal groups for this study: the control, saline-treated LPS, and MCC950-treated LPS groups (Figure 1A). The saline-treated LPS group showed higher levels of blood urea nitrogen (BUN) (17.4 ± 2.3 vs. 106.3 ± 12.3 mg/dL; p < 0.05, Figure 1B) and serum creatinine than the control group (0.11 ± 0.07 vs. 0.56 ± 0.15 mg/dL; p < 0.05, Figure 1C). MCC950 treatment significantly reduced the BUN and creatinine levels compared with those of the saline-treated LPS group (106.3 ± 12.3 vs. 67.0 ± 29.1 mg/dL; p < 0.05, 0.56 ± 0.15 vs. 0.31 ± 0.16 mg/dL; p < 0.05). We also analyzed the tissue expression of interleukin-6 (IL-6), connective tissue growth factor (CTGF), and transforming growth factor-β1 (TGF-β1) to determine the inflammatory and profibrotic processes caused by LPS treatment. The expression of IL-6, CTGF, and TGF-β1 was significantly higher in the saline-treated LPS group than in the control group (Figures 1D–F). MCC950 treatment decreased the mRNA expression levels of IL-6 and CTGF in the kidney. The inhibition of NLRP3 by MCC950 ameliorated LPS-induced renal injury and inflammatory cytokine gene expression and resulted in a lower tubular injury score in the MCC950-treated LPS group than in the saline-treated LPS group (Figures 1G, H). Immunostaining for F4/80, a marker of monocytes, showed an increased number in the saline-treated LPS group compared to the control group, while there was no difference between the saline-treated and MCC950-treated LPS groups (Figures 1I, J).

The main strength of LSFM is rapid, wide-field image acquisition at a high spatiotemporal resolution, enabling whole-organ imaging (17). We applied LSFM to visualize the spatial distributions of CD11c+ cells and vasculatures in transversely sectioned whole mouse kidneys after clearing (Figure 2A, whole kidney). CD11c-YFP+ cells within a 2-mm-thick whole kidney image were mainly distributed in the cortex and, in particular, in the tubulointerstitial area. An Alexa Fluor 647-conjugated antibody to the endothelial cell adhesion molecule CD31 stained vasculature including glomeruli. Merged images of CD31-stained and CD11c-YFP+ cells were suitable for visualizing renal MNPs and glomeruli distributed in the renal cortex. In the saline-treated LPS mouse kidneys, the glomerular volume measured by CD31 staining was significantly decreased but was not affected in the MCC950-treated LPS group (Figures 2B, C). This finding could be related to the systemic vasoconstrictive effect after LPS administration. We compared the spatial distribution of CD11c-YFP+ cells and vasculature between the LPS-treated and control groups in 3D tissue blocks from cleared kidneys (Figures 2D, E). The total number and spatial density of CD11c-YFP+ cells were significantly increased in the saline-treated LPS group. Although the spatial cellular density appeared to be decreased in the MCC950-treated LPS group, there was no significant difference in the number of CD11c-YFP+ cells between the two groups. In addition, denser clusters of CD11c-YFP+ cells were found in certain sites, such as periglomerular and perivascular lesions, in the saline-treated LPS group compared to the control group or the MCC950-treated LPS group (Figure 2F).

MPM enables deep optical imaging at a subcellular resolution with reduced phototoxicity (18). We applied MPM to analyze renal MNPs, including deep-level specimens, with a 3D approach. In a single-plane 2-dimensional (2D) image, similar to the current histological approach, there was no significant difference in the CD11c-YFP+ area between the saline-treated LPS and MCC950-treated LPS groups (Figures 3A, B). Through the MIP technique with the highest attenuation values on every view throughout the 150µm onto a single 2D image (Figures 3C, D), we found that the projection area of CD11c+ MNPs in the saline-treated LPS group was higher than that in the control and MCC950-treated LPS groups. This difference was more obvious in 3D images. In 3D images, the number of renal MNPs could be directly measured (Figures 3E, F). MPM imaging demonstrated that CD11c-YFP+ cells were more densely interspersed within millimeter-thick tissue of the saline-treated LPS group than the control and MCC950-treated LPS groups. These results suggested that MPM combined with kidney clearing provides direct quantitative information on renal MNPs and covers a deeper range of the kidney.

We also investigated the expression of surface molecules on CD11c+ cells with FACS analysis to confirm the functional changes in MNPs. The number of CD11c+ cells was increased in the saline-treated LPS group compared with the control group (Figure 4A). The number of CD11c+ cells in the MCC950-treated LPS group was not significantly reduced compared with that in the saline-treated group, as shown by FACS (Figure 4B). MHC class II molecules are expressed on antigen-presenting cells such as DCs, and their main function is to present processed antigens to CD4+ T lymphocytes. CD86 is a costimulatory molecule involved in the activation of naïve T cells. The proportions of CD11c^highCD86^high and CD11c^highMHCII^high cells were significantly increased in the saline-treated LPS group compared to the control group (0.13 vs. 2.26% and 0.06 vs. 2.68%, respectively, Figures 4C–F). However, MCC950 treatment significantly reduced the expression of CD86 and MHC class II molecules on CD11c+ cells (2.26 vs. 0.25% and 2.68 vs. 0.13%, respectively). FACS analysis showed significant changes in renal CD11c+ cell activity rather than the number of renal CD11c+ cells after LPS injury.

CD11c+ cells have multiple dendrites on the cell body. Dendrites have been shown to be engaged in efficient antigen uptake, antigen presentation, and interaction with T cells (19–21). We measured the number of dendrites and dendrite segments and the branch levels of CD11c+ cells to confirm that the morphological changes of MNPs reflect functional activation after LPS by using the Imaris filament tool. In 3D imaging, CD11c+ cells had multiple dendrites with branching, especially in the saline-treated LPS group compared to the control group (Figures 5A, B). Quantitative measurement of the dendrite and segment numbers and the number of branch levels also revealed increases in the saline-treated LPS group (Figures 5C–E). However, NLRP3 inhibition by MCC950 treatment prevented these morphological changes due to LPS-induced injury.