Investigation of CDH1 germline mutations in Turkish patients with Kaposi’s sarcoma

Patient collection

The study was conducted in accordance with the Declaration of Helsinki [15]. The Ethics Committee of Istanbul University accepted the study (Approval No. 1490, September 22, 2011). CDH1 gene mutations were investigated in 25 patients diagnosed with KS who presented to the Oncology Institute of Istanbul University clinics between 2016 and 2020. DNA samples were isolated from the patients’ peripheral blood mononuclear cells, and each exon was amplified with primers specific for the CDH1 gene. Patients’ records were examined for viral etiologic factors, Human immunodeficiency virus (HIV), Kaposi’s sarcoma-associated herpesvirus (KSHV), Hepatitis B surface antigen (HbsAg), hepatitis B core antibody (ANTI-HBc), anticytomegalovirus (ANTI-CMV), Epstein-Barr virus capsid antigen (EBV VCA), human herpesvirus 1 (HHV-1), human herpesvirus 2 (HHV-2), and Varicella-zoster virus (ANTI-VARI ZOS) are listed in Table 1. After the Dye Terminator Cycle Sequencing (DTCS) reaction, the gene regions loaded into the sequence analyzer were analyzed and scored.

The median age of patients diagnosed with KS was 64 years (46–76 years). 28% (7/25) of patients are female, and 72% (18/25) are male. While 44% (11/25) of the patients smoke, 32% (8/25) use alcohol. 24% (6/25) of patients have diabetes mellitus, hypertension, and chronic heart failure. 8% (2/25) of patients have an iatrogenic subtype, and 92% (23/25) have a classic subtype of KS. One extremity involvement was observed in 48% (12/25) of the patients, 2–5 in 36% (9/25), >5 16% (4/25).

5A family history of cancer was noted in 52% (13/25) of patients. Recurrence was observed in 76% (19/25) of patients. 56% of the patients have patches, 36% have plaque, and 8% have nodule histopathological subtype.

In addition, benign variants were detected in 84% (21/25) of patients, Variant of uncertain significance (VUS) in 12% (3/25), and pathogenic variants in 4% (1/25) of KS patient.

Leukocytes isolation

First, 20 mL of the patients’ peripheral blood mononuclear cells were collected and placed in a tube containing 3 mL of Ficoll solution (Sigma-Aldrich, Darmstadt, Germany). The peripheral blood mononuclear cells were diluted with phosphate-buffered saline (AppliChem, Darmstadt, Germany). Peripheral blood mononuclear cells diluted with phosphate-buffered saline were centrifuged at 1,910 rpm for 30 min at room temperature. Cells were frozen at −80 °C (Sanyo, Moriguchi, Japan) for 24 h before removal and storage in a tank of liquid nitrogen (Thermo Fisher Scientific, Gothenburg, Sweden) for long-term storage.

DNA isolation

DNA isolation was performed according to the protocol of the Quick DNA Kit (Zymo Research, CA, USA). First, lysis buffer (AppliChem, Darmstadt, Germany) was added to leukocytes isolated from peripheral blood. The lysed leukocytes isolated from peripheral blood were transferred to a DNA filter and centrifuged to obtain the DNA adhering to the filter. The DNA filter was washed with wash solutions and centrifuged. A separation solution was added to dissolve the filter adhering to the DNA and washed. Eppendorf tubes were placed under the DNA filters for the final centrifugation. The DNA of peripheral blood mononuclear cells from KS patients was then transferred to tubes and prepared for quality control and further testing.

Agarose gel electrophoresis

1 µl of DNA samples from the patient’s peripheral blood mononuclear cells were mixed with 5 µl of bromophenol blue (Sigma-Aldrich, Darmstadt, Germany) loading buffer and loaded onto a 1.5% agarose gel (Invitrogen, MA, USA). The current was set at 150 V, and the gel was allowed to run for 25 min. The DNA of the isolated peripheral blood mononuclear cells was visualized under UV light (Vilber Lourmat, Marne-la-Vallée Cedex 3, France).

RT-PCR reaction and purification

DNA from peripheral blood mononuclear cells of patients were isolated in a 0.2 mL tube containing 0.5 μLTaq polymerase enzyme (Thermo Fisher Scientific, CA, USA), 0.5 μL dNTP (Thermo Fisher Scientific, CA, USA), MgCl (Thermo Fisher Scientific, CA, USA), 0.5 μL forward and reverse primer (Oligomer, Istanbul, Turkey), 39.5 μL dH₂O, 5 μL PCR buffer using a thermal cycler (BioRad T100TM Thermal Cycler, WI, USA) according to the appropriate conditions. The PCR product was stored at −20 °C (Bosch, Turkey) for later use in DTCS. Tubes were placed in a thermal cycler. The following program was set at 98 °C for initial denaturation for 30 s, at 60 °C for denaturing for 30 s, at 72 °C for annealing for 15 s, at 72 °C for extension for 10 min, and holded at 4 °C.

DTCS reaction and purification

The High Pure PCR Product Purification Kit (Sigma-Aldrich, Darmstadt, Germany) was used according to the company’s instructions. The purified PCR products were run on an agarose gel (Sigma Aldrich, Darmstadt, Germany) in the column after purification.

The DTCS reaction was performed with the purified PCR product for 25 ng per kilobase. 8.0 µL of DTCS Quick Start Kit master mix, 2.0 µL of Sequencing Primer 1.6 pmol/μL, and 8.3 µL of dH₂O were added as PCR Reaction Content.

PCR product purification and Sanger sequencing

After the DTCS reaction, a second purification was performed. After this purification, agarose gel electrophoresis was performed to verify the presence of PCR products and their suitability for sequencing. PCR products from the isolated peripheral blood mononuclear cells of KS patients were then prepared for sequencing using a Beckman Coulter CEQ 8000 (Beckman Coulter CEQ-8000 Analyzer, CA, USA). For sequence analysis, PCR products were stopped immediately after the DTCS reaction with a 1:2:2 glycogen-sodium-EDTA-sodium acetate mixture (AppliChem, Darmstadt, Germany). Subsequently, 60 µL of 95% cold ethanol (Sigma Aldrich, Darmstadt, Germany) was added to the PCR products and centrifuged for 15 min at 4 °C and 14,000 rpm (VWR MICRO-STAR 17/17R Leuven, Belgium). The supernatant was removed, and 200 µL of 70% cold ethanol (Sigma Aldrich, Darmstadt, Germany) was added to the pellet and centrifuged at 14,000 rpm for 5 min. The alcohol was collected, and the pellet remaining at the bottom of the tube was dried using a vacuum concentrator (CentriVap Benchtop Vacuum Concentrators, Kansas City, USA) to evaporate the alcohol. 40 µL of sample loading solution (SLS) buffer was used to dissolve the dry pellet. The thawed material was added to the instrument on the plate for Beckman Coulter 8000 (GenomeLab™ GeXP, CA, USA) sequence analysis. Beckman Coulter CEQ 8000 Analysis Software was used to analyze the data. The results were evaluated after comparing the analyzed samples with the reference gene sequence.

This study used the genome sequence of Homo sapiens (human) GRCh37 (hg19) from the Genome Reference Consortium as a control.

Reading of sequencing results

The presence of PCR products was detected by agarose gel electrophoresis, and the purified PCR products were analyzed by direct Sanger sequencing. Sanger reads were manually evaluated for all queried variants to emulate techniques typically used in a clinical setting. The Genome Analyzer (Beckman Coulter CEQ-8000 Analyzer, CA, USA) was used to examine the sequence analysis data generated by the capillary electrophoresis instrument and the electropherogram image was validated by comparison with the reference gene sequences. This study used the H. sapiens (human) GRCh37 (hg19) genome assembly from the Genome Reference Consortium as a control. The ALAMUT [16], dbSNP [17], SIFT [18], and ClinVar [19] databases, as well as in silico analyses, were used to investigate the functional effects of variants and to predict the effects of amino acid substitution variations on the human genome in the study.

Statistical analysis

All statistical analyses were performed using the SPSS program (SPSS version 27; SPSS Science, Chicago, IL, USA), with a p-value of 0.05 considered statistically significant. The chi-square, ANOVA, and Fisher-exact tests were used to evaluate statistical significance. Kaplan-Meier analysis was used to determine the survival of KS patients according to mutation subtype.

In silico analyses of the missense variants

A total of 16 variants of the CDH1 gene were identified: HET c.531 + 10G > C, HET c.688-20C > A, HET c.1008 + 12G > T, HOM c.100916G > A, HET c.1137 + 144_1137 + 145dupT, HOM c.1320 + 39C > T, HET c.1849G > A, HET c.193713T > C, HET c.2164 + 17dupA, HET c.2245C > T, HET c.2268T > C, HET c.2287G > C, HET c.2296-14T > C, HET c.2439 + 19A > G, HET c.2439 + 161dupT, HET c.2471C > T. In silico analyses of these variants using five different prediction tools have shown conflicting results. We searched the Human Genome Mutation Database (HGMD), ALAMUT, Single Nucleotide Polymorphism Database (dbSNP), Sorting Intolerant From Tolerant (SIFT), and Clinically observed variants (ClinVar) databases, which classified variants as benign, likely benign, or uncertain significance [1], [2], [3], [4, 7]. The pathogenic mutation HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5 was found in a KS patient with a family history of gastric and breast cancer (patient ID: 25). HGMD considered HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5 as a potential disease-causing mutation. In a previous study by Kaurah and colleagues, the missense mutation HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5 was predicted to be a pathogenic variant with the prediction software SIFT [3]. Moreover, and they proposed this variant as a novel pathogenic variant with a family history of gastric cancer in a Colombian family [22]. Aggregation and invasion analyses were tested in CHO cells to validate the pathogenicity of the novel missense mutation HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5 by Kaurah and colleagues as these analyses are common approaches to compare the function of epithelial cadherin missense variants with wild-type epithelial cadherin, which increases cell aggregation and prevents cell invasion [23, 24]. In addition, Kaurah et al. described HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5 as a known CDH1 mutation [22]. In our research group, we also found the CDH1 HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5 variant in KS patients with a family history of gastric cancer and, and it may have an impact on the initiation and metastasis of KS tumors with a family history of gastric and breast cancer.

In addition, the patient in whom the pathogenic variant HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5 was detected had five affected extremities and a high number of lesions compared with other KS patients. Because KS is an epithelial lesion resembling gastric cancer and the CDH1 gene has been associated with gastric malignancies, the CDH1 gene was investigated in KS patients. Consistent with the literature, pathogenic variant (HET c.2245C > T, p.(Arg749Trp) rs776975632 NM _004360.5) was detected in one of the 25 KS patients with a family history of cancer in the study [22]. Since patient with this pathogenic mutation in the CDH1 gene had more active lesions and a more aggressive disease course KS patients should have germline test of CDH1.

Previously, the missense mutation HET c.1849G > A, p. (Ala617Thr) rs33935154 was discovered as a somatic pathogenic variation in endometrial cancer and identified as a pathogenic germline mutation in a patient with diffuse gastric cancer [25, 26]. In our study, the variation HET c.1849G > A, p. (Ala617Thr) rs33935154 was detected in one patient. This variation affects a conserved region encoding one of the calcium-binding motifs in the extracellular part of CDH1. Since the presence of calcium ions stabilizes the active form of the protein, these calcium-binding motifs are functionally important. Because of their position, this mutation has been associated with developing an unstable intercellular protein complex. Suriano et al. discovered this germline mutation in two African American female patients with diffuse gastric cancer who were 43 years old at the time [24]. In a study by Valente and colleagues, the identical germline variant was found in 6% (10/165) of African American patients with ductal or mixed breast cancer [27]. Here in this study, the variant HET c.1849G > A p. (Ala617Thr) rs33935154 was detected in a patient with a family history cancer, and this variant may be investigated in terms of pathogenicity in the etiology of KS in future studies of KS patients.

On the other hand, immunosuppression, HHV-8, advanced age, diabetes, and corticosteroid treatment have been reported as the most important causes of KS. In this study, a KS patient with a variant of HET c.1849G > A, rs33935154, was also found to have diabetes and hypertension in his late 60 s.

In the present study, five exonic variants (HET c.1849G > A rs33935154, HET c.2245C > T rs776975632, HET c.2268T > C rs1567515480, HET c.2287G > C, HET c.2471C > T rs1450920019) were detected in KS patients. These patients, in whom exonic variants were detected, had more plaque lesions, which are a sign of early-stage KS [25]. Patients with the exonic variant HET c.2245C > T, p.(Arg749Trp) rs776975632 had a higher number of affected extremities and increased number of lesions compared with patients with intronic variants.

NHKS is a multifocal vascular endothelial carcinoma of KS that is quite rare and not associated with HIV, as in all patients studied in this study. We reported the CDH1 variants in the etiology and prognosis of KS in only 25 HIV-negative KS cases. Blood DNA testing has its limitations. However, basal transcription of the genome is known to occur at low levels, and blood has been found to contain approximately 80% of all human coding sequences. In addition, this study was modest in scope and relied on secondary data from clinical diagnostic laboratories. Further studies are needed to better characterise the variants detected in this study in many patients with KS.

Secondary primary malignancies (SPMs) are also a severe problem for people with KS. SPMs can occur in KS patients up to 15 years after the diagnosis of CKS. However, the guidelines have not yet established the mechanism for detecting SPMs [26]. Maruyama et al. reported that more than 280 patients with KS had solid malignancy related to head and neck, esophagus, stomach, duodenum, or colorectal cancer. As SPMs are crucial complications in patients with KS, these malignancies should be detected as early as possible [27]. We believe that life-threatening SPM in patients with CKS should be screened by genetic analysis at early stage.

It has been shown that in patients who are considered immunocompromised, inherited genetic mutations may also be the cause of elderly diseases, such as KS. Since genetic disorders may cause disease in immunosuppressed patients similar to KS, patients should be screened in this direction. More comprehensive genetic screening is recommended for individuals with HIV-independent KS with a family history of cancer. In addition, further large-scale studies, and particularly multicenter cohort studies, will provide further insight into the role of CDH1 in carcinogenesis and in KS patients with a family history of cancer.