Animals

4-week old male db/m and db/db mice either heterozygous or homozygous for a mutation in the leptin receptor gene (Lepr) gene (BKS. Cg-Dock7m + /+ Leprdb/J) respectively, were purchased from The Jackson Laboratory (Bar Harbor, ME). The db/db mouse exhibits chronic hyperglycemia, hyperphagia, obesity, and insulin resistance and is widely utilized as a model of type 2 diabetes mellitus, whilst the db/m mouse is a suitable non-diabetic control mouse [7]. The number of mice per group was selected based on pilot data, to detect a change in the urinary albumin creatinine ratio of at least 100 g/mol (α = 0.05, 1 - β = 0.8) between the control and RS-supplemented diabetic mice. Mice were acclimatized for a period of 2 weeks, prior to the commencement of diet intervention at 6 weeks of age. Mice were housed in a climate-controlled animal facility with a 12:12 h light-dark cycle and received ad libitum access to mouse chow and water. Each week, mice were weighed and had random blood glucose levels measured using an Accu-Chek® Performa glucometer (Roche Diagnostics, USA). All study protocols were conducted in accordance with the principles and guidelines devised by the Alfred Medical Research & Education Precinct Animal Ethics Committee (AMREP AEC) under the guidelines laid down by the National Health and Medical Research Council of Australia and had been approved by the AMREP AEC (E1487/2014/B).

Diets

At 6 weeks of age, mice commenced a specialty formulated semi-pure resistant starch supplemented diet (SF15-015) containing 25% g/g Hi-maize 1043 (Ingredion, Westchester, IL), equivalent to 12.5% g/g resistant starch type 2 [8] or a custom-made matched control diet (SF15-021). Both diets were prepared by Specialty Feeds (Perth, Western Australia, Australia) and were matched in terms of total caloric content, protein and fat (Supplementary Table 1). The control diet was received by the non-diabetic mice (designated “db/m”). The diabetic mice were randomly allocated by cage to receive either the control diet (designated “db/db”) or the resistant starch supplemented diet (designated “db/db RS”). Mice received these diets ad libitum for a period of 10 weeks. Two cohorts of mice, referred to in this paper as Cohort 1 and Cohort 2, were utilized for different experimental endpoints. Both cohorts received the same diets. The experimental overview for each cohort is demonstrated in Fig. 1A and Fig. 4A, respectively.

Fig. 1: Resistant starch reduces kidney injury in diabetic mice.

A Schematic of study design for Cohort 1. Created with biorender. B Urinary Albumin Creatinine Ratio, C Plasma Creatinine Levels, D 24-h urine output. * = P < 0.05, *** = P < 0.001, **** = P < 0.0001. One-way ANOVA with Tukey’s post hoc test. Data = mean ± SD. n = 10–12.

Metabolic caging, plasma and tissue collection

At experimental week 9 (15 weeks of age), mice in Cohort 1 were housed individually in metabolic cages (Iffa Credo, L’Arbresle, France) for 24 h for urine collection and measurement of water and food intake. Commercially available ELISA tests were utilized for the measurement of urinary albumin (Kit: E90-134, Bethyl Laboratories Inc., USA) and C5a (Kit: DY-1250, R&D Systems, USA). Urinary creatinine was determined via a commercially available assay kit (Kit: 03263991190, Roche Diagnostics Corporation, USA) by a Cobas Integra 400 Plus autoanalyzer (Roche Diagnostics Corporation, USA). At the end of the experimental intervention, mice were anesthetized by intraperitoneal injection of sodium pentobarbitone (100 mg/kg body weight; Euthatal; Sigma-Aldrich, Castle Hill, Australia). Following euthanasia, blood was drawn from the portal vein, treated with sodium citrate (3.2% v/v), centrifuged at 6000 rpm for 6 mins, and plasma was collected and stored at −80 °C. Both kidneys were removed for tissue digestion and flow cytometry.

As both kidneys were utilized in Cohort 1 for flow cytometry to investigate renal immune cell populations, a second cohort of mice was run to investigate the effects of these diets on renal and intestinal histology. At the endpoint, mice were anesthetized by intraperitoneal injection of sodium pentobarbitone (100 mg/kg body weight; Euthatal; Sigma-Aldrich, Castle Hill, Australia) followed by cardiac exsanguination. Cardiac blood was centrifuged at 6000 rpm for 6 mins, and plasma was collected and stored at −80 °C. Kidney sections were fixed in neutral buffered formalin (10% v/v) before being embedded in paraffin. The gastrointestinal tract was dissected, and the mesentery was removed. The total gastrointestinal tract, and then the dissected cecum and colon, were weighed and length was measured. The gastrointestinal tract was flushed with chilled phosphate-buffered saline. Jejunum and ileum sections were fixed in paraformaldehyde (4% v/v) for 24 h before being transferred to 4% sucrose solution and embedded in paraffin. Jejunum, ileum, and colon sections were snap-frozen in liquid nitrogen and stored at −80 °C.

Flow cytometry

Both kidneys were collected from mice in Cohort 1. A single-cell suspension was prepared as follows. Briefly, both kidneys from each mouse were mechanically dissociated (sliced into cubes ~1–2 mm, by scissors) followed by orbital incubation at 37 °C with Collagenase IV (1 mg/ml, Sigma-Aldrich) for 30 min in RPMI 1640 medium with 5% fetal bovine serum. Kidney homogenates were filtered through a 70-μm cell strainer, and then subjected to a Percoll gradient centrifugation in 36% Percoll (Sigma-Aldrich), overlaid onto a 72% Percoll solution and centrifuged at 1000 g with no brake for 20 min at 4 °C. Cells were isolated from the Percoll interface and washed in cold media (RPMI 1640 medium with 5% fetal bovine serum). Erythrocytes were lysed using RBC lysis buffer (BD Biosciences). The resulting single-cell suspension was then treated with Fc Block (anti-CD16/CD32, BD Biosciences, clone 2.4G2) to block non-specific binding of antibodies. Cells were analysed with a panel of antibodies including Brilliant Violet 786–conjugated (BV786) anti-CD45 antibody (BD Biosciences, clone 30-F11, 1/2000), AF700-conjugated anti-CD11b antibody (Biolegend, clone M1/70, 1/1000), BV605-conjugated anti-Ly6C antibody (BD Biosciences, clone AL-21, 1/1000), Fluorescein isothiocyanate-conjugated (FITC) anti-C5aR1 antibody (Cedarlane, clone 20/70, 1/300), PE anti-Ly6G antibody (Tonbo biosciences, clone 1A8, 1/1000). To distinguish between live and dead cells, cells were stained with a Fixable Aqua live dead cell stain kit (Invitrogen, L34966). All antibody and live dead stain concentrations were used according to the manufacturer’s recommendations. Cells were counted with a Tali Image-Based Cytometer (Life Technologies). Analysis was performed on a BD FACSCanto II system using BD FACSDiva software (BD Biosciences).

Plasma analyses and targeted metabolomics

At cull, portal vein blood was collected from mice in Cohort 1. Plasma lipopolysaccharide-binding protein (LBP) was measured using an ELISA (ab269542, Abcam) as per the manufacturer’s instructions. The intra-assay coefficient of variation was 5.2%. Plasma endotoxin was measured using a commercially available assay based on Limulus amoebocyte lysate (A39552, Pierce Thermo Scientific, Rockford, IL, USA). Portal vein blood that had not undergone any prior freeze-thaw cycles was used for targeted metabolomics as previously described [9].

Renal histology

Kidney sections with a thickness of 3 micrometers were subjected to periodic-acid Schiff (PAS) staining. The extent of sclerosis in each glomerulus was evaluated using a subjective 0–4 scale, where 0 represented normal conditions, and higher grades indicated increased sclerotic areas: minimal (grade 1), sclerotic area up to 25%; moderate (grade 2), sclerotic area 26 to 50%; grade 3, sclerotic area 51 to 75%; severe (grade 4), sclerotic area 76 to 100% (severe). The Glomerulosclerosis Index (GSI) was computed using a formula that factored in the number of glomeruli in each grade as previously described [10]. Digital images of the renal cortex were taken with an Eclipse Ci brightfield microscope (Nikon, Melville, NY, USA) at 200x magnification. Evaluations were performed using Image-Pro Plus (version 7.0; Media Cybernetics, Bethesda, MD, USA) and analysis was conducted in a blinded fashion.

Renal immunohistochemistry

Paraffin sections of mouse kidney (4 μm) were immunostained with rabbit anti-fibronectin (AO245, DAKO) at a 1:400 dilution. In brief, endogenous peroxidases were blocked by 3% hydrogen peroxide for 15 min, followed by a pepsin antigen retrieval for 10 min, after which samples were blocked in Dako Superblock (1:10 in TBS) for 20 min. A primary antibody was applied and left overnight at 4 °C. The following day, slides were incubated with biotinylated secondary antibody for 10 min at room temperature. Sections then underwent incubation with Vectastain ABC reagent (Vector Laboratories, CA, USA). Peroxidase activity was identified by reaction with 3,3’-diaminobenzidine tetrahydrochloride (Sigma-Aldrich Pty. Ltd, NSW, Australia). Counterstaining with hematoxylin was done to identify nuclei. Sections were examined under light microscopy (Olympus BX-50; Olympus Optical). All digital quantitation (Image-Pro Plus, v7.0) and assessments were performed in a blinded manner.

Intestinal Histology

Paraffin-embedded ileum and jejunum sections, with a thickness of 5 micrometers, were prepared for analysis. To assess villi length and crypt depth, the sections were stained with hematoxylin and eosin. For the enumeration of goblet cells, the sections were initially stained with 1% Alcian blue for 15 min, followed by periodic acid Schiff staining. Images were captured using an Eclipse Ci brightfield microscope (Nikon, Melville, NY, USA) at a 100x magnification. Image-Pro Plus software (version 7.0, Media Cybernetics, Bethesda, MD, USA) was employed to view the images and measure villi length and crypt depth utilizing the built-in measurement tool. The number of goblet cells per villus was manually counted for each villus. All analyses were performed in a blinded manner.

Quantitative RT-PCR

RNA was extracted from snap-frozen colon, jejunum and ileum samples using TRIzol Reagent (Life Technologies) as previously described [10]. cDNA was synthesized from RNA using M-MuLV Reverse Transcriptase (Thermo Fischer). Expression of zonula occludens-1 (ZO-1; Tjp1) and junctional adhesion molecule A (JAM-A; F11r) was determined using TaqMan reagents (Life Technologies, Carlsbad, CA). For occludin (Ocln), claudin-1 (Cldn1), claudin-2 (Cldn2), claudin-3 (Cldn3), claudin-4 (Cldn4), claudin-5 (Cldn5) and claudin-7 (Cldn7), SYBR Green reagents (Applied Biosystems, California, USA) were utilized. RT-PCR was conducted using either a QuantStudio 3 or 5 Real-Time PCR System (Thermo Fisher). Gene expression was normalized to β-actin (Applied Biosystems) and fold change was calculated relative to db/m CON mice using the ΔΔCT method.

Statistical analysis

Statistical analyses were performed using Graphpad Prism (Version 9.3.1, Graphpad Software, USA). Outliers were assessed and removed using robust regression and outlier removal (ROUT) [11] using a false discovery rate (FDR) value of 0.01. Data were assessed for normality using Kolmogorov-Smirnov test. Normally distributed data were analyzed using one-way ANOVA with Tukey’s post hoc test for multiple comparisons. Non-normally distributed data were assessed using Kruskal-Wallis test with Dunn’s multiple comparisons test. Targeted metabolomics data were log-transformed and analyzed using Metabolanalyst (version 5.0) [12], with a fold change threshold of 2 and an FDR threshold of 0.05.