The animals were housed in temperature-controlled chambers and had free access to water and food. Animal facilities in this study were approved by AAALAC International, and all animal handling procedures were approved in advance by the Institutional Animal Care and Use Committee of Kunming Biomed International (license number: KBI K001116020-01, 01). Kunming Biomed International is a professionally qualified non-human primate scientific research institution, and some monkeys from animal rescue stations or cynomolgus monkeys with scientific research needs are bred at the base every year. KBI has primate breeding facilities that meet the standards of the Association for Evaluation and Certification of Laboratory Animal Management (AAALAC). Two ADPKD cynomolgus monkeys were used for our experiments, both living in Yunnan Province, China, who had never been used in any drug or medical experiments until their death.

Due to improper preservation of the specimen, after dissection, the livers and kidneys of the PKD- and PLD-affected cynomolgus monkeys were used for histology analysis. The cross sections were equilibrated at room temperature for 15 min, and the nuclei were dyed with Harris' hematoxylin for 5 min. The slices were rinsed under running water for 5 min, and 1% eosin staining was performed for 5 min. After dehydration in a gradient ethanol bath and washing with xylene, the patches were sealed with neutral gum and observed under the bright field of a microscope (Leica, Germany).

The ultrasound procedure was performed using a GE LOGIQ P5 ultrasound imaging system after the onset of short-acting Zoletil anesthetic (0.1 mg/kg i.m.); then, the animal was safely secured to the examination table. Abdominal ultrasonography of monkey 010139 affected by PKD and PLD was performed after it exhibited depression and poor appetite. The other PKD-affected monkey was diagnosed during necropsy; therefore, no relevant ultrasonography results were available. Animal IDs and health status were confirmed before biopsy procedures. Animals were anesthetized with ketamine, 10 mg/kg i.m., and hair on the area of skin that covered the right chest and the liver lobe was removed using a razor. The imaging probe was placed on the abdomen. Ultrasonographic examinations were conducted as previously described (16). The liver and other organs were imaged transcutaneously using the GE LOGIQ P5 ultrasound imaging system, with 10 L and 4°C probes (GE Healthcare, USA).

Before the ADPKD monkeys died, we collected serum by venipuncture of the saphenous vein using a 5-mm animal prick needle (Goldenrod Animal Lancet, Braintree Scientific). Then, 2 ml of blood was collected in an anticoagulant-free blood collection tube (BD vacutainer, SST II Advance). The monkeys were carefully monitored for 30 min following blood withdrawal to watch for signs of distress. An additional 0.5 ml of blood was collected into an EDTA-K2 tube. Biochemical index levels were determined by Kunming Biomed International Laboratory using a Roche Cobas C501 (Roche, Germany). Hemocytometry was performed on a Sysmex XT-2000i (Sysmex, Japan).

After the ADPKD monkeys died, genomic DNA was extracted from the livers and/or kidneys using the protocol of the Wizard Genomic DNA Purification Kit. DNA concentration was measured with NanoDrop2000 (Thermo Scientific, Germany). Whole-genome resequencing and bioinformatics analysis were performed according to the whole-genome resequencing workflow by the BGI Shenzhen Corporation under the contract terms of project number F20FTSCCKF0459_MONldhR. The qualified DNA samples were randomly fragmented by Covaris, and the fragments were collected by magnetic beads after sample quality control. Adenines were added to the 3′-end of end-repaired DNA fragments before adaptor ligation. The ligation products were then cyclized and amplified using linear isothermal rolling-circle replication and DNA nanoball technology. BGISEQ was used to sequence these DNA libraries. Bioinformatics analysis was performed with the sequencing raw data generated by the sequencing pipeline. In the raw data, low-quality reads were removed first. Then, the Burrows–Wheeler Aligner was used to align the reads to the reference sequence. Alignment information was stored in BAM format files, which were further processed in the following steps: sort by samtools, filter mapping quality, and label duplicate reads. GATK (https://www.broadinstitute.org/gatk/) was used to detect single-nucleotide polymorphisms (SNPs) and small insertions and deletions (Indels) in the post-processed BAM files. Break Dancer (http://breakdancer.sourceforge.net/) was used for structural variants (SVs) calling and Control-FREEC (http://boevalab.inf.ethz.ch/FREEC/index.html#introduction) for copy number variants (CNVs) calling. All variants were annotated by Annovar (17, 18). Mutation pathogenicity prediction analysis was performed on the Mutation Taster website (https://rddc.tsinghua-gd.org/). The cynomolgus monkey whole-genome data reference version is GCF_000364345.1_Macaca_fascicularis_5.0.

Primers for PKD1 and PKD2 target genes were designed using the Premier 6.0 software. The primer sequences were as follows: PKD1-F2: GTGGTGGAGATGGACGTAGAGT; PKD1-R2: CGGAAATAGGGCAGGATTAGCG; PKD1-F3: ACACA GTGGAGCATGTGTACC; PKD1-R3: GACCTTGATGTCCGTGAC CA; PKD1-F4: CGAGTCACCATCACGGATGGT; PKD1-R4: ACCCAGGCAGGCACTATGAGA; PKD2-F3: AATGGTGGTGGAGATGG ACGTA; PKD2-R3: CCCCAGATGTACTCTTTCACTCTAA; GANAB-F1: GAATGTATGAAGGATGACCCAA; GANAB-R: CGCAGGTGAATACTCCAATC. The reaction conditions and PCR system were as follows: 1.5 μl of DNA template (200 ng/μl), 1 μl of each upstream and downstream primer (10 μmol/l), 2 μl of dNTP mix, 1 μl of PrimeSTAR GXL DNA polymerase (TAKARA), 10 μl of 5× PrimeSTAR GXL buffer Mg²⁺, 4 μl of dNTP mixture, and up to 50 μl of nucleic acid-free water; amplification of the PCR reaction: 3 min predenaturation at 95°C, 30 s denaturation at 94°C, 15 s annealing temperature, 30 s extension at 72°C, 35 cycles, 5 min extension at 72°C, and storage at 4°C. After electrophoresis on a 1% agarose gel at 110 V for 30 min and staining, PCR products were detected with a gel imager. Sequencing reactions were prepared using a BigDye^® Terminator kit (Applied Biosystems), and reaction products were sequenced using the ABI 3500Dx Genetic Analyzer (Applied Biosystems, Forster City, USA). Sanger sequencing was performed by Tsingke Biotechnology Co., Ltd.

We screened for variants where the WGS and Sanger sequencing results were completely consistent and clarified the mutation information of these sites in the human genome. The pathology prediction tool on the website (https://rddc.tsinghua-gd.org/tools) is based on a machine-learning approach that uses the XGBoost classifier to predict pathogenicity. The predicted outcome is a number between 0 and 1, with a higher number (closer to 1) indicating a higher probability that the mutation at the locus is pathogenic. Encoded values and qualitative prediction are as follows: 0–0.02: Benign; 0.02–0.1: Likely Benign; 0.1–0.8: Likely Pathogenic; and 0.8–1: Pathogenic.

Diseased monkey 1: We investigated a 17-year-old male PKD- and PLD-affected cynomolgus monkey (Monkey 010139). Veterinarians observed that the monkey's abdomen was bulging and that it exhibited depression and poor appetite. Ultrasonography of 010139 showed an enlarged liver with the presence of multiple cysts that were not clearly separated from the kidneys (Figures 1A–C). We analyzed the serum biochemical indexes of the male cynomolgus monkeys. The serum index results were as follows: creatinine 47 μmol/L, blood urea nitrogen 6.7 mmol/L, aspartate aminotransferase 38 U/L, alanine aminotransferase 27.2 U/L, albumin 15.2 g/L, alkaline phosphatase 4,719 U/L, total protein 53.7 g/L, Na⁺ 136 mmol/L, K⁺ 4.55 mmol/L, CL⁻ 98.4 mmol/L, globulin 38.5 g/L, A/G 0.4, and ketone 2,249 μmol/L. Routine blood examination showed the following values: leukocyte 20.5 × 10⁹/L, erythrocyte 6.01 × 10¹²/L, hemoglobin 119 g/L, hematocrit 38.9%, platelets 364 × 10⁹/L, neutrophils 69.9%, and lymphocytes 24%. Reference values of clinical pathology parameters in cynomolgus monkeys used in this study are listed in Supplementary Table S1 (19, 20). The normal leukocyte range in the blood of cynomolgus monkeys is (11.15 ± 2.85) × 10⁹/L, and the normal value of ketone bodies in the serum of cynomolgus monkeys should be lower than 0.6 mmol/L (19, 20). The above results showed that 010139 had inflammatory infection and ketosis without liver and kidney dysfunction. After a definitive diagnosis of inflammatory infection and ketosis, the first line of treatment is rapid anti-inflammatory therapy and nutritional support. Accordingly, veterinarians administered an intravenous drip comprising ceftriaxone sodium (80 mg/kg) anti-inflammatory therapy, vitamin supplementation, and energy support with rehydration salts.

We performed an autopsy on Monkey 010139 when it died 1 month later despite active treatment, due to refractory infection. Normal liver tissue (indicated by black arrows) had mostly disappeared (Figure 1D), and there were small- to medium-sized liver cysts (diameter between 0.5 and 2 cm) in all liver segments (>100 cysts). Multicystic structures filled with tissue exudate were observed, some of which were light yellow and some were brown (Figures 1D, E). Most of the kidney surface cortex was relatively smooth, with around 12 vesicles; multiple vesicles had spontaneously ruptured, and tissue necrosis was observed (Figure 1F). The results of hematoxylin–eosin staining (H&E) of liver tissue showed diffuse steatosis of hepatocytes, a liver cyst wall lined with a single layer of columnar epithelium, eosinophilic secretions and macrophages in the lumen, pericystic tissue fibrosis, and interstitial lymphocyte infiltration (Figure 1G). The normal hepatic lobular structure was destroyed, and the hepatocyte-regenerated nodules were divided and wrapped into round or oval pseudo-lobules of varying sizes by extensively proliferating fibrous tissue (Figure 1H). Histopathological features of kidney cystic tissue are shown, including tubular dilation, eosinophilic changes in epithelial cells, and foam-like changes in some epithelial cells (Figure 1I). No normal glomerular structure was observed, and interstitial lymphocyte infiltration, multiple cystic cavities, and eosinophilic secretions were seen in some cystic cavities.

Diseased monkey 2: A 20-year-old PKD-affected monkey (Monkey 993563) was diagnosed during the process of dissection after it died spontaneously. A few renal cysts existed in bilateral kidneys (Supplementary Figure S1), and there were no obvious abnormal changes in other organs. Results related to pathological slide analysis were not collected due to improper preservation of the specimen. Neither blood cell analysis nor serum biochemical index analysis indicated abnormalities before it died. The results of serum biochemical indexes were as follows: creatinine 87 μmol/L, blood urea nitrogen 6.5 mmol/L, aspartate aminotransferase 26 U/L, alanine aminotransferase 25.3 U/L, glucose 154 mg/dl, glycated hemoglobin% 4.7, high-density lipoprotein 0.7 mmol/L, triglycerides 1.16 mmol/L, cholesterol 1.4 mmol/L, very low-density lipoprotein 0.53 mmol/L, and low-density lipoprotein cholesterol 30.6 mmol/L. Blood cell tests indicated the following values: leukocyte 9.65 × 10⁹/L, erythrocyte 6.91 × 10¹²/L, hemoglobin 168 g/L, hematocrit 55.7%, platelets 394 × 10⁹/L, neutrophils 35.4%, and lymphocytes 53.4%. None of these indicators were higher than the normal values.

When analyzing the WGS data, autosomal recessive and dominant modes of inheritance were considered. In accordance with previous literature reports, our analysis focused on the WGS data of 31 autosomal recessive and/or dominant PLD- and PKD-related genes (PKD1, PKD2, PKHD1, HNF1B, ALMS1, TSC1, TSC2, TCTN2, B9D1, B9D2, OFD1, MKS1, DYNC2H1, TTC21B, TMEM216, TMEM67, TMEM231, CEP290, ROGRIPIL, CC2D2A, NEK1, GANAB, PRKCSH, UMOD, MUC1, SEC63, LRP5L, DZIP1L, VHL, IFT80, and DNAJB11) (8, 9). Filtering for private variants present as homozygous or heterozygous in two affected monkeys and absent in the genomes of 11 wild-type cynomolgus monkeys revealed no variants shared across all affected monkeys. The WGS data of 11 wild-type cynomolgus monkeys were obtained from literature reports by Wang et al. (21). All variants that refer to indels and SNPs per genome after annotation were compiled and are shown in Table 1. Filtering each case separately for private variants produced candidate protein-changing variants with dominant inheritance in four positions. There are no SV CNVs found in these genes related to PKD and PLD across the WGS data for the two affected monkeys.

In PKD- and PLD-affected monkey 010139, heterozygous missense variants were found in exons of PKD2, GANAB, and UMOD genes. Detected variants include PKD2: (XM_005555387.2: c.452C>T/p. P151L); GANAB: (NM_001285075.1: c.2708T>C/p. V903A); UMOD: (XM_005591391.1: c.970A>T: p. T324S); and UMOD: (XM_005591391.1: c.499A>G: p. T167A), identified by Sanger sequencing. The mutation status of these loci is consistent with the results of next-generation sequencing. The GANAB gene located on chromosome 11q12.3 encodes the alpha subunit of glucosidase II and is a member of the glycosyl hydrolase 31 family of proteins.

We compared the conservation of amino acids at the mutation site across species, GANAB: (NM_001285075.1: c.2708T>C/p. V903A) is highly conserved among multiple species (Figure 2), while the variant in PKD2: (XM_005555387.2: c.452C>T/p. P151L) is leucine in mice, which is inconsistent with the corresponding amino acid in humans and monkeys (Supplementary Figure S2). Based on the conservation analysis of the position of the variants between the cynomolgus monkey and the human genome, we performed pathogenicity prediction analysis of highly conserved locations across species. The prediction analysis method is published in the database website developed by the South China Rare Disease Data Center (RDDC) (https://rddc.tsinghua-gd.org/manual/patho-predic). Encoded value and qualitative prediction are as follows: 0–0.02: Benign; 0.02–0.1: Likely Benign; 0.1–0.8: Likely Pathogenic; and 0.8–1: Pathogenic. Two variants in PKD2: (XM_005555387.2: c.452C>T/p. P151L) and GANAB: (NM_001285075.1: c.2708T>C/p. V903A) were likely to cause severe disease, with pathogenic probabilities of 0.32 and 0.16, respectively. However, variants in the UMOD gene were analyzed as benign.

In the PKD-affected monkey 993563, two missense variants in exon 4 of the polycystin-1 (PKD1) gene were discovered. The variants in PKD1: (XM_015442355.1: c.1144G>C/ p.E382Q) and PKD1: (XM_015442355.1: c.2488A>G/ p. S830G) affect amino acid residues 382 and 830 in PKD1, with moderate effects on the resulting protein. Missense variants affect the PKD/chitinase domain of PKD1, which typically consists of the PKD1-IgG motif. Humans have a similar missense variant in exon 4 (Figure 3). A final missense variant in exon 13 of the PKD1: (XM_015442355.1: c. 5530G>T/ p. A1884S) was found. Amino acid conservation analysis between species at the mutation position and mutation pathogenicity prediction analysis on the RDDC website suggested that PKD1: (XM_015442355.1: c.1144G>C/ p.E382Q) is a pathogenic mutation of 0.14. In humans, PKD caused by missense variants has been reported in many patients (9). The other variant of the PKD1 gene is thought to be benign. Information on the pathogenic loci of the two PKD-/PLD-affected monkeys implicated in PKD and PLD is summarized in Table 2.