All animals were sourced from the Animal Resource Centre, Murdoch, Western Australia, Australia. Experiments were conducted in accordance with the National Health and Medical Research Council Australian code of practice for the care and use of animals for scientific purposes and approved by the Murdoch University and Macquarie University Animal Ethics Committees.
The lpk mutant arose as a spontaneous mutation from the Lewis (LEW/SsNArc) strain at the Animal Resource Centre, (Perth WA, Australia) . Two strategies were used to obtain DNA for linkage genetic analysis. LPK males and Brown Norway (BN/ssArc) females were crossed and first filial (F1) hybrids intercrossed to produce 152 (LPKxBN) second filial (F2) segregants, with a segregation pattern of LPK:non LPK (42:110), as reported previously . LPK and Wistar Kyoto (WKY/NArc) females were crossed and the F1 hybrids crossed back to the LPK males to produce 139 (LPKxWKY) BC1 segregants with a segregation pattern of LPK:non LPK (72:67) within the F2 generation, being consistent with an autosomal recessive mutation.
Prior to euthanasia animals were weighed, urine collected and systolic blood pressure determined by tail-cuff plesmography, using an average of 3 measurements after the animals were acclimatized to the procedure (NIBP controller, ADI Instruments, Castle Hill, NSW, Australia). The animals were euthanized using CO2/O2. Blood was collected by cardiac puncture and tissues were dissected and weighed. Kidneys were then fixed in 4% formalin in 0.01M phosphate buffer (PB) for subsequent histology and the liver was collected and frozen at -80˚C for genomic analysis. Blood was analysed for serum urea and creatinine, total protein and albumin using a Rx Daytona analyser (Randox Laboratories, Antrum, UK). Micro haematocrit tubes were used to determine packed cell volume and urinary protein to creatinine ratio was determined using a Cobas Mira analyser (Roche Diagnostics, Schweiz, AG).
In order to determine the ratio of cyst area to total kidney area, kidney sections were processed for histology by paraffin embedding, sectioning (4μm) and staining with haematoxylin and eosin (H&E). Images of kidney were captured using a Nikon Dx40 Digital SLR and then digitally enhanced to maximize contrast (Additional file 4: Figure S3). The area of cyst to cross sectional area of the kidney (μm2) was determined using Image J .
DNA was extracted from the liver using the Wizard® Genomic DNA Purification Kit (Promega, Maddison, WI, USA) following the manufacturer’s protocol. The samples were assayed for DNA concentration and contamination using a NanoDrop 1000 Spectrophotometer (ThermoFisher Scientific, Scoresby, Vic, Australia). DNA was stored at -80˚C and working samples diluted to 50ng/μl in MilliQ H2O and stored at -20˚C. Preliminary polymerase chain reaction (PCR) studies of the BC1 using 2-4 simple sequence length polymorphism (SSLP) markers per chromosome (obtained from ) mapped the lpk locus to chromosome 10 between markers D10Rat43 (23.42Mbp) and D10Rat26 (77.09Mbp). Using primer sequences available from Rat Genome Database (RGD)  additional polymorphic DNA markers (17 in the BC1 and 12 in the F2, refer Additional file 5: Table S2), synthesized by Geneworks (Adelaide, SA) were then used to further genotype the animals by PCR. The defining markers obtained from the interval mapping from the BC1 and F2 generation were used for comparative analysis between the rat, mouse and human using the Virtual Comparative Map [VCM, RGD ].
PCR was performed with Taq polymerase 0.055 U/μl (Fisher Biotech, Australia) in 10 μl reaction volume containing forward and reverse primers 0.5μM, dNTPs 0.25mM, MgCl 1.5mM. Cycling conditions consisted of 4 stages. Stage 1: TAQ activation 94˚C, 5 mins. Stage 2: touchdown protocol, 12 cycles of 94˚C for 30s, annealing for 45s commencing at 54˚C with a 0.5˚C drop in temperature per cycle, and elongation at 74˚C for 30s. Stage 3: 32-40 cycles of denaturing at 94˚C for 30s, annealing at 48˚C for 45s, and elongation at 74˚C for 30s. Stage 4: 1 cycle as per stage 3 with a 5 minute extension and sample hold at 14˚C. PCRs were performed on an Applied Biosystems Veriti™ Thermal Cycler.
Genotyping of the animals was carried out on the Protean II xi | XL cell (BioRad) using T = 8% [monomer] and C = 5% [crosslinker] PAGE gels run at 70-80V for 18-22 hours. Gels were post stained with SYBR® Safe (Invitrogen, Aust) for 30 minutes and photographed with Bio-Vision camera with VISION-Capt image acquisition software (v. 15.06, Vilber Lourmat, Germany). The lanes were scored as Homozygote [WKY (A) in the BC1 and BN (A) in the F2 generation & LPK (B)] or Heterozygote (H) (Figure 1).
Coinheritance of the phenotypic parameters with polymorphic DNA markers was analyzed by QTL analysis using Map Manager QTX  and WinCart QTL . The map in the present paper focused on chromosome 10 based on earlier work (unpublished). The QTL was determined by cumulative interval mapping (CIM) with forward and reverse regression using model 6 with a window size of 0.5 cM and walk speed of 1 cM. The threshold was determined using permutations, with permutations set to 1000 and significance 0.05.
DNA was extracted from the liver of three 6-week old male LPK animals and three 6-week old male Lewis genetic controls, derived from the same inbred line from which the mutation arose (LEW/SsNArc) but selected without the mutation, using a DNeasy Blood & Tissue Kit (QIAGEN, Vic, Australia) following the manufacturer’s protocol. DNA was checked for quality on a 2% agarose gel and LPK and Lewis samples were respectively pooled.
A Roche Nimblegen sequence capture array (Roche NimbleGen Inc. Madison, WI, USA) was designed to enhance the region of interest and Illumina sequencing reactions of the ~200 bp DNA regions were carried out by the Australian Genome Research Facility Ltd (AGRF, Gehrmann Laboratories, St Lucia, QLD, Australia) yielding 100 bp single end sequence reads. Alignment of the reads was performed using the open source short read aligner Bowtie (release 0.12.5)  and limited to 2 mismatches. Homozygous and heterozygous SNPs were identified using the freeware Package SHORE . Samples were concurrently analyzed by AGRF QLD using CASAVA Software (version 1.7, Illumina Inc. San Diego, CA, USA) to identify SNPs. Both approaches yielded the same result. Biological network analyses were drawn from published literature on PKD and links to cilia and cell division target genes of interest were refined further using the web based programs Search Tool for the Retrieval of Interacting Genes/Proteins  Cytoscape , Inner Medullary Collecting Duct Proteome Database  and Murine Immortalized Cortical Collecting Duct Cell (mpkCCD) Transcriptome Database .
Sequence read archive accession numbers
Ilumina raw sequence data for the Lewis (LEW/SsNArc) wild type strains and LPK mutant were submitted to the NCBI Sequence Read Archive (SRA) under the submission numbers SRA056808 and SRA056807, respectively.
Direct sequencing of target SNP region
Amplification of the identified lpk SNP region within Nek8 was carried out using standard PCR techniques with the forward and reverse primers ( nk-f3: 5′ TCA AGA TGG TGA TGG TGG 3′) and (nk-r3: 5′ TGA TGT CAC CGT GTA AGG 3′), respectively. Primers were designed using Primer3  to produce a 173 bp product covering the homozygous SNP identified by direct sequencing. Parental BN, WKY & LPK and selected BC1 progeny (animal numbers 85, 87, 90 and 92) were assayed and product visualized on a 1.6% agarose gel. The product was purified using the Promega Wizard® Gel and PCR cleanup system. Thirty ng of product, as determined by spectrometer (Thermo Fisher Scientific, Scoresby, Vic) was used as the template for sequencing using the forward primer only (nk-f3) , with the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, CA, USA) as per the manufacturer’s instructions. The PCR sequencing reaction consisted of 96˚C for 2 mins, followed by 30 cycles of 96˚C for 10s, 48˚C for 10s and 60˚C for 10s. The product was then purified using an EDTA/NaAc ethanol precipitation technique and samples stored at -20˚C. Sequencing was performed on a capillary sequencer (Applied Biosystems 3730xl DNA Analyzer, CA, USA). The chromatograph calls were aligned and visualised using the software Geneious version 5.5 .
Protein structure prediction of the Nek8 protein
The Rattus norvegicus Nek8 protein (NP_ 001099274) was aligned to the homologous proteins of Homo sapiens (NP_835464), Mus musculus (NP_543125), Gallus gallus (XP_415822), Danio rerio (NP_620766), Xenopus tropicalis (NP_001006906), Xenopus laeivis (NP_001090238) Saccoglossus kowalevskii (XP_002730582). The alignment was conducted using MAFFT version 6.814b  using a Blosum62 scoring matrix with the following parameters: 1.53 for the Gap open penalty and 0.123 for the offset value.
RCC1-like repeats for NP_001099274 (R. novegicus Nek8) were analysed by the web based program REP . Secondary protein structure was predicted using the Emboss tool Garnier , whilst the tertiary protein structure prediction and 3D modeling were performed using Phyre2  and ESyPred .
Preliminary work was conducted based on an initial VYxWG motif that early papers described as fingerprinting RCC1 repeats . This was modified to [VILC]xx[WLFC]G, to account for variation within Nek8 and the RCC1 protein. Sequences for the initial blade alignments using MAFFT were determined using the RCC1 motif and selecting 22 amino acids up stream and 25 amino acids downstream. A final blade was selected, being an intervening sequence between blades (non-canonical blade 2). The 7 predicted blades were then aligned using MAFFT against the known RCC1 blades of the Mesocricetus auratus (Golden Hamster [UniProtKB: P23800]) for which a crystal structure is available .
Possible functional sites and domains were determined within the Nek8 protein using the web based program Eukaryotic Linear Motif Resource (ELM: ). Possible functional motifs within domains were then refined based on predicted 3-D structure of the Nek8 protein and in particular the outer regions of the folded protein because these were most likely to be involved in protein-protein interactions.
Conserved G[X]LG motif within the blades of the RCC1 domain
The G[X]LG motif was identified as a key feature of the RCC1 domain of Nek8. In order to increase the specificity of the search, the motif GX(5)GXF(4)G[QRC]LG was used in the web based application Prosite . The proteins containing this sequence were then entered into the web based programme KAAS  to examine possible pathway connections/ functionality associated with the sequence.
Scanning electron microscopy
Four Lewis (standard inbred LEW/CrlBR) and 4 LPK animals aged 10-12 weeks of mixed sex were deeply anaesthetised (sodium pentobarbitone 60 mg/kg i.p.) and transcardially perfused at 110mmHg with heparinized saline (0.2% heparin, Mayne Pharma Pty Ltd, Melbourne, Vic, Australia and 0.9% w/v NaCl2), and then 2.5% (v/v) glutaraldehyde solution (ProSciTech Pty Ltd, Brisbane, QLD, Australia) in 0.1M PB, pH 7.4. Kidneys were dissected, sectioned into coronal slices 1 mm thick, and post-fixed for 24 hours in the same fixative solution. Samples were sectioned into 3 mm x 1 mm wedges, washed in 0.1M PB, post-fixed in 1% (v/v) osmium tetroxide solution, washed in 0.1M PB, dehydrated in graded series of ethanol (30-100%) and critical point dried in a EMITECH K850 Critical Point drier, with CO2 as a transition fluid. Kidney fragments were mounted on the aluminum stubs, covered previously with carbon tabs and coated with gold using EMITECH sputter coater K550.
Imaging and cilia measurement
Samples were viewed with JEOL JSM- 6480 LA Scanning Electron Microscope (Jeol USA, Inc.) and files stored as jpg. To determine cilia length images were opened using Image J software  and cilia length measured using the Freehand tool after calibration of the image scale. Cilia were selected from random fields of distal and collecting tubules. For LPK, total number of cilia measurements was 788 (from each animal average n = 197 ± 6.78) and for Lewis, a total of 159 cilia were measured (from each animal average n = 53.5 ± 30.35). The differences in number of cilia measured from each LPK and Lewis rat was due to cysts, which in the LPK allowed ready access of cilia for SEM from broad fields of cells, while in Lewis samples, cross sections of normal tubules presented only small fields of view and therefore limited numbers of cells with cilia available for assessment. A D’Agostino-Pearson test was used to determine normality of data and a two-tailed Mann–Whitney T-test was used for comparison of cilia length between Lewis and LPK. Results are presented as median and quartiles (25th/ 75th percentile). All analyses were undertaken using the statistical package Prism 5 for Mac (Version 5.0d, GraphPad Software, La Jolla, USA).
Three 10-week old female LPK and three 10-week old female Lewis rats (LEW/CrlBR) were deeply anaesthetised (sodium pentobarbitone 60 mg/kg i.p.) and transcardially perfused at 110mmHg with heparinised saline followed by 4% formalin containing 0.01M PB (pH 7.4) for 45 minutes. Kidneys were removed and post-fixed for 24 hours in the same fixative at 4°C, then stored in 0.01M PB at 4°C until processed. Kidneys were sectioned coronally, paraffin embedded and 10 μm sections cut and mounted onto Superfrost Plus slides, then deparaffinised with xylene and hydrated in a descending series of graded ethanol solutions. Antigen retrieval was performed with 10 mM citrate buffer pH 6 (Sigma Chemical Co., USA). Buffer was heated to boiling, the samples placed in solution and microwaved for 5 min at 720W power and then allowed to cool to room temperature. Sections were then incubated in blocking solution containing tris phosphate buffered saline solution (TPBS-Tx; 10 mM tris base, 0.01 M PB, 0.9% w/v NaCl, 0.1% v/v Triton X 100 and 0.1% w/v sodium azide, pH 7.4) with 10% v/v donkey serum (Sigma-Aldrich, USA) for 1 hour at 20°C. Primary antibodies were diluted in the blocking solution and samples incubated at 4°C overnight. Experiments were initially performed as either single or double labelled experiments using a polyclonal anti-Nek8 antibody (1:200 ), a mouse anti-acetylated α-tubulin antibody (1:5000, T7451, clone 611B-1, Sigma-Aldrich, USA) or mouse anti γ-tubulin (1:500, T5326, clone GTU-88, Sigma-Aldrich, USA). After washing in PBS (3 x 30 min), species-specific fluorescent-labelled secondary antibodies were used to detect primary antibody labeling (donkey anti-rabbit Cy3 [1:500, Jackson Immunoresearch, USA] and donkey anti-mouse Dylight 488 [1:500, Jackson Immunoresearch]). Sections were again washed (3 x 30 min) and coverslips mounted using fluorescent mounting media (Dako, Denmark). Additional experiments were performed in the LPK animals with all three antibodies as a triple label, as γ-tubulin and α-tubulin antibodies identified discrete structural regions (basal body and ciliary shaft, respectively) and could be distinguished by location, despite both being identified by the same secondary anti-mouse Dylight 488 antibody.
The sections were imaged on a Leica TCS SP5 confocal microscope (Leica Microsystems, Wetzlar, Germany), equipped with an argon laser, with excitation and detection wavelengths set for Dylight 488 (excitation 488 detection range 496 to 552 nm) and Cy3 (excitation 514 detection range 555 to 625 nm). For double labelled sections, images were collected using sequential acquisition between stacks. Images were adjusted for brightness and contrast only.