A germline mutation in Rab43 gene identified from a cancer family predisposes to a hereditary liver-colon cancer syndrome

The family with a history of hereditary cancers consisted of two persons with colorectal cancer and four persons with liver cancer spanning in three consecutive generations. Blood samples were collected from available family members except the fourth generation members because of their young ages and other concerns. For the proband’s deceased father diagnosed with liver cancer, DNA and liver tissue were obtained from the pathological archives of a local hospital where the patient was treated. All living participants involved in this research signed written informed consent forms. This research was approved by the Harbin Medical University ethics committee in accordance with Helsinki Declaration.

Genomic DNA was extracted from peripheral leukocytes by DNA purification Kit (Qiagen) according to the manufactrue’s instructions. DNA concentrations, DNA purity and integrity were assessed by spectrophotometry (Nanodrop) and agarose gel electrophoresis and met the DNA sequencing requirements. The gDNA library was prepared using Ion Xpress Plus Fragment Library Kit (Thermo Fisher Scientific). Exome enrichment was performed by SureSelect™ target enrichment system (Illumina) in accordance with the manufacturer’s instructions. The enriched DNA samples were sequenced via paired 126 bp sequencing using a Hiseq2500 Sequencing System (Illumina).

A binary sequence alignment map (BAM) file was generated by comparison of the paired DNA samples. Somatic SNVs and indels were detected by comparison of the paired BAM files. The sequence reads were aligned to the human reference genome (UCSChg19) using the Burrows-Wheeler Aligner (BWA) with default parameters. Variants were identified using the Genome Analysis Toolkit (GATK). SNVs and indels were analyzed using Haplotype Caller and data were filtered by variant quality score recalibration. Relevant mutations in all of the genes were subsequently prioritized manually.

Exome sequencing results for prioritized variants were validated using specific primers for amplification by polymerase chain reaction and Sanger sequencing. Segregation analysis of the prioritized variants was performed in additional family members (including a deceased member with hepatocellular cancer) whenever genomic DNA was available.

IHC was carried out on paraffin-embedded sections of colon, liver cancer and their adjacent non-cancerous tissues obtained through surgery to investigate the expression levels of endogenous Rab43 with primary antibodies against human Rab43. A total of fifteen cancerous and fifteen adjacent non-cancerous tissue sections (slides) were analyzed for each liver or colon cancer patient. Each slide was blinded read under light microscopy by two board-certified pathologists. Ten randomly picked view fields from each slide were assessed. The staining intensity and proportion of positively stained cells were evaluated as described previously with minor modifications [7]. The intensity of Rab43 staining (brown color) was scored as follows: 0, negative; 1, weak; 2, moderate; 3, strong; And the percentage of positively stained cells was scored as follows: 0, < 5%; 1, 5–25%; 2, 26–50%; 3, 51–75%; 4, > 75%. The sum of each intensity score multiplied by its corresponding positive percentage score was regarded as the Rab43 staining score for the viewed field and assigned as follows: - for (0); + for (1–2); ++ for (3–5); +++ for (6–8); ++++ for (9–12).

The human cancer cell lines used in our research including HeLa (CCL-2), A549 (CCL-185), Skov-3 (HTB-77) and HepG2 (HB-8065) were purchased from American Type Culture Collection (ATCC, VA, USA). Cell line Huh7 (BNCC337690) was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). These cell lines were free of mycoplasma contamination as tested by a PCR method prior to our research. Cells were maintained in Dulbecco’s Modified Eagle Medium (Gibco, NY, USA) supplemented with 10% fetal bovine serum (Gibco), 100 unit/ml penicillin and 0.1% (w/v) streptomycin (Gibco) in a 5% CO₂ humidified incubator at 37 °C.

The human Crispr-Cas9 guide RNA was designed based on the coding gene of Rab43 (GenBank accession number: AY166852). The primer pair for gRNA was: Rab43-cri-F3: 5′-caccggctactaccgcagtgccaat-3′, Rab43-cri-R3: 5′- aaacattggcactgcgg tagtagcc-3′. The primer pair was annealed using the following parameters: denatured at 95 °C for 5 min and cooled down to room temperature. The annealed primer pair was cloned into the Lenti-Crispr-V2 vector (a gift from Dr. Feng Zhang of MIT, USA) through BsmI restriction sites. The gRNA cloning was verified by Sanger sequencing. HepG2 cells were transfected with the plasmid Lenti-Crispr-V2-Rab43–3 by lipofectamin-3000 according to the manufacturer’s instruction. The stable transfected cells were selected by puromycin, and the resulting cell clones with desired mutations in Rab43 gene were verified by Sanger sequencing and Western blotting.

Wild type Rab43 gene was amplified from a gift plasmid (from Dr. John Cox of University of Tennessee) containing the wild type ORF of Rab43 by using two following specific primers with XhoI and EcoRI restriction sites. Rab43-cpXhoI-F: 5′-aattctcgagatggcagggccgggccca-3′, Rab43-cpEcoRI-R: 5′-aattgaattctcagcacccgcagcc ccag-3′. Wild type Rab43 ORF was subcloned into pcDNA3.1 (−)-3 × Flag via XhoI and EcoRI sites. Site-directed mutagenesis using following primers with the desired mutations (F: 5′- ctatgacatcctgtgtaccattgagacgtctg-3′, R: 5′-cagacgtctcaatggtacacaggat gtcatag-3′) was performed for Rab43 A158T mutant generation. The Huh-7 cells transfected with pcDNA3.1-(−)-3 × Flag-Rab43 A158T were selected by G418. Following primer pair was used for generating Crispr resistance Rab43 ORF in pcDNA3.1-(−)-3 × Flag plasmid. Cri-resi-3F: 5′-agctactatagatcagctgcagccgccatccttg-3′, Cri-resi-3R: 5′-caaggatggcggctgcagctgatctatagtagct-3′.

Cells were seeded at a density of 500 cells/well in 6-well plates. After 2 weeks, the cells were fixed with 4 ml of 4% paraformaldehyde for 15 min, and stained with 0.2% crystal violet for 10 min. The numbers of visible colonies with more than 50 cells per colony were counted. Colony formation was calculated from three independent experiments.

Cells were seeded in 96-well plates at the density of 1000 cells/well and incubated for the next day. The cell proliferation assays (MTT) were performed on day 1, 3 and 5 after seeding. Experiments were carried out in triplicate.

Cell migration assay was performed by using wound healing method as described previously [8]. Briefly, cells were seeded into a six-well plate at a density of 5 × 10⁵/well and allowed to culture overnight. Cells were then treated with mitomycin C (final concentration at 50μg/ml, Sigma) to inhibit cell proliferation for 2 h before scratches were generated by applying a 200 μL pipette tip. Cells were washed twice with PBS buffer in order to remove all the dislodged cells. Photographs were taken at 0 h, 24 h and 48 h after wounding. Migration of the cells was measured by determining the wound healing areas of control cells versus Rab43 overexpressed Huh7 cells. Assays were performed in triplicate.

Cells were lysed in RIPA lysis buffer. The lysates were sonicated and centrifuged at 12,000 rpm and 4 °C for 10 min. Cell lysates were quantified using BCA protein assay kit (Pierce) and normalized for protein content by RIPA buffer. After separation on SDS-PAGE, the proteins were then transferred to PVDF membrane (Millipore). The membranes were incubated with the primary antibody (1:1000) probing Rab43 (Santa Cruz), Akt (Cell Signaling), p-Akt (Ser473) (Cell Signaling),, phos-pP70S6K (Cell Signaling) or GAPDH (Sigma) over night at 4 °C and incubated with the fluorescein conjugated secondary antibodies (1:5000) for 1 h at room temperature. The proteins were detected and photographed using Odyssey imaging system (Li-COR biosciences). GAPDH was used as loading control. The signal intensity was quantified using the Image J software (NIH).

We identified a large Chinese family with four generations living in Northeast of China with a clear history of hereditary liver and colon cancers. As of the date when blood samples were collected, a total of two colon cancers and four liver cancers and one colon adenoma have been diagnosed spanning the first three consecutive generations (Fig. 1). Those diseases display an obvious autosomal dominant pattern of inheritance. Furthermore, there is no affected member in the fourth generation most likely due to their relative young ages.

We sequenced the whole-exome of the proband who was diagnosed with colon cancer at the age of 54. Meanwhile, his mother’s whole-exome sequencing was performed in parallel for comparison. In addition, we also compared our sequencing results with the hg19 genome database maintained by University of California Santa Cruz as a third party reference. We used six bioinformatics tools to select for deleterious mutations, including PhyloP (score > 0.85), SIFT (score < 0.05), PolyPhen (score > 0.85), GERP (score > 2), Mutation Taster (score > 0.5), and LRT (score > 0.9), and only those mutations with four or more positive deleterious predictions were further considered. Eventually, we obtained 13 SNVs from 13 genes (Table 1) and 16 Indels from 15 genes (Table 2). Among these candidate genes, based on available information including their known functions, potential protein interactions and predicted effects on protein structure, we focused on the variant of Rab43 for further analysis. There is a C > T transition occurred on the anti-sense strand of one Rab43 allele in the genome of the proband, this mutation leads to substitution of alanine by threonine at the 158 residue on Rab43 protein. The heterozygous mutation status of Rab43 gene in the proband was confirmed by Sanger sequencing (Fig. 2a). The segregation of the mutant allele was analyzed in additional family members except all members from the fourth generation of that family. While the patient’s healthy mother and two sisters are homozygous for wild type Rab43 alleles, the genome of his father who died from liver cancer indeed harbors one of the mutant allele of Rab43 (Fig. 2b-d). These data demonstrated the mutant allele of Rab43 is clearly co-segregated with the familial cancer syndrome, though the segregation information is not available in the fourth generation currently. In addition, we performed protein structure analysis by using PyMol software [9] and found there is no overall structural change with the threonine substitution of alanine at residue 158 on Rab43 protein (Fig. 2e-f). There are many known oncogenic mutations that would interfere with the functions rather than the overall structures of the affected proteins. One prominent example of such mutations is the oncogenic Ras G12 V mutant. A previous study reported the overall structures between the GDP bound form of the normal Ras protein and the GDP bound form of the Ras G12 V are identical, suggesting it is the side chain on Valine12 rather than the conformation changes would impair the intrinsic GTPase activity associated with Ras [10]. In the case of Rab43 A158T, we propose it is the hydroxyl group on Threonine 158 rather than any structural changes that would enhance the functions associated with Rab43 protein, hinting strongly the identified Rab43 variant might be oncogenic with a gain of function.

Rab43 is a Rab family related small GTPase and regulates the sorting of a subset of protein cargo through the medial Golgi [11‐13]. Similar to mutations occurred on oncogene Ras [14, 15], the Rab43 mutation identified in our study might be oncogenic. In order to determine the functional consequence of Rab43 A158T, we subcloned wild type Rab43 gene and generated Rab43 A158T mutant gene by site directed mutagenesis. We screened an array of human cancer cell lines and found the expression levels of Rab43 are varied dramatically (Fig. 3a). Ectopically expressed both wild type Rab43 and Rab43 A158T mutant proteins in liver cancer cell line Huh7 with a very low level of endogenous Rab43 protein enhanced cell growth and proliferation (Fig. 3b-c), stimulated the colony formation on cultured plates (Fig. 3d-e) and promoted the migration of the transfected Huh7 cells in vitro (Fig. 3f-g). More strikingly, compared to the effects of the expressed wild type Rab43, the expressed Rab43 A158T mutant protein tends to confer more malignant transformation capacity to the transfected Huh7 cells (Fig. 3c-g). Interestingly, overexpression of Rab43A158T in another liver cancer cell line HepG2 with a high level of endogenous Rab43 expression has minimal effects on the growth, proliferation, colony formation and migration of the transfected HepG2 cells, indicating Rab43 A158T variant can substitute the function of WT Rab43 once wild type Rab43 is insufficient or lost completely. Taken together, our results strongly support Rab43 A158T variant possesses oncogenic activity and plays a driving role in the onset of the studied cancer syndrome, consistent with the autosomal dominant pattern of inheritance observed in that family.

Rab43 has been implicated in glioma development previously [16], but the underlying mechanism is unclear. To further elucidate the molecular mechanism of Rab43 variant mediated tumorigenesis, we eliminated the Rab43 expression in HepG2 cells by Crispr-Cas9 gene editing technology targeting the exon 2 of Rab43 gene (Fig. 4a) and established a HepG2 Rab43 knockout cell line with 3 different reading frame shift mutations in the coding region of Rab43 gene, which cause complete loss of endogenous Rab43 protein expression in the Crispr-Rab43 HepG2 cells (Fig. 4a-b, for detailed Sanger sequencing results see Additional file 1: Figure S1). To our surprise, the Rab43 knockout HepG2 cells are enlarged and flattened with slow growth (Fig. 4c-d). Some knockout cells exhibit multi-nuclei, indicating problems in cell dividing and cell cycle progression. These observations suggest certain level of Rab43 is required for normal cell growth and cell cycle progression. To further reveal the molecular mechanism underlying Rab43 mediated regulation, the Akt activation in control (V2 vector transfected) and Crispr-Rab43 HepG2 cells was examined under normal culture conditions. We found the basal level of Akt activation was barely detectable in Crispr-Rab43 cells without any stimulation (Fig. 4e), rendering the reasonable comparison of Akt activation between the two cell lines unfeasible. Thus, we switched to look at the activation of Akt following insulin stimulation. Despite the fact that little activated Akt was detected in the insulin stimulated Rab43 knockout cells, Akt activation was dramatically increased in wild type Rab43 control cells following insulin stimulation, whereas the levels of total Akt protein in both cell lines were almost equal (Fig. 4e-g). Moreover, a similar pattern of the phosphorylation of p70S6K, a downstream target of Akt signaling, was observed (Fig. 4f-g), suggesting Akt-mTOR-p70S6K signaling is impaired upon loss of Rab43. Since it is well known that Rab43 involves in ER-Golgi trafficking, our data point to a potential role of protein trafficking in cell transformation and tumorigenesis.

To further confirm the relevance of identified Rab43 variant to the onset of the studied familial cancer syndrome, we performed immunohistochemistry in colon tissues of the 54Y proband with colon cancer and in liver tissues of his farther died from liver cancer. We found high levels of Rab43 protein occurred in both colon and liver cancer tissues, whereas much less of Rab43 was detected in their adjacent non-cancerous tissues (Fig. 5a-c). Eventually, a total of 15 cancerous or non-cancerous tissue sections (slides) from the proband or his father were analyzed, the overall expression difference of Rab43 protein in between cancer tissues and non-cancerous tissues is statistically significant (Fig. 5b-d) (p < 0.001). These results clearly demonstrated there is a positive correlation between Rab43 protein expression level and the development of the familial liver- colon cancer syndrome. Given the fact that the proband and his father were heterozygous carriers of the Rab43 mutant allele, we are not sure whether the high levels of Rab43 observed in cancer tissues is contributed by elevated expression of Rab43 A158T alone. Nevertheless, the positive correlation

between Rab43 and cancer development in the family strongly hints Rab43 A158T mutant might be the driving force for the onset of that familial liver-colon cancer syndrome.