This article is mentioned in:

Introduction

Lung cancer is the leading cause of cancer-related mortality worldwide. In 2012, World Health Organization statistical data reported that the numbers of global new cases and mortalities from lung cancer were 1,824,701 and 1,589,800, respectively (1). The majority of lung cancers are diagnosed at an advanced stage, so the 5-year survival rate of lung cancer patients is <15% (1). Based on the histological type, lung cancers are divided into two subtypes: Small cell lung cancer and non-small cell lung cancer (NSCLC). NSCLC accounts for >85% cases, and can be further classified into three main subtypes: Squamous cell carcinoma, adenocarcinoma and large cell carcinoma (2–4).

Previous studies have identified a large quantity of driver genes in lung carcinogenesis, genetic aberrations of which can promote tumor cell proliferation, survival, angiogenesis, invasion and metastasis, thus driving the development of lung cancer (5). For example, epidermal growth factor receptor (EGFR) mutations (6), EML4/ALK fusion genes (7), RAS mutations, BRAF mutations and PI3K/mTOR mutations are frequently observed in lung cancer tissues, and are predictive of response to targeted therapies (8).

The majority of lung cancers are sporadic, and familial cases are extremely rare. Unlike sporadic cancer, germline mutations may drive carcinogenesis in familial cases. For example, functional mutations in BRCA1/2 genes may cause familial breast or ovarian cancer. The family members who carry BRCA1/2 mutations have a 70–80% possibility of developing cancers (9,10). Another example in colorectal cancer demonstrates that DNA mismatch repair gene MSH2 and MLH1 mutations carriers have a high risk of developing colorectal cancer in their 40s (11,12).

To date, only three familial lung cancer cases have been studied. In one study, it was reported that a lung squamous cancer patient carried the EGFR R776H germline mutation, and that the patient's daughter, also a lung cancer patient, carried the same alteration (13). In another study, 3/6 individuals in a Japanese family developed lung adenocarcinoma. The germline EGFR V843I mutation was identified in all the lung adenocarcinoma-affected members of this family (14). In a study by Yamamoto et al, it was found that mutations in HER2 may be the causative variants of familial lung adenocarcinomas (15). The molecular mechanism of familial lung cancer remains largely unknown.

The present study recruited a family of Chinese descent in which multiple members were diagnosed with lung squamous carcinoma. To find the causative mutations, whole exome sequencing was conducted using a peripheral blood sample of one lung squamous carcinoma patient, and certain variants were validated in more samples. This study provides the basis for further studies, in which further hereditary lung cancer members will be sequenced.

Materials and methods

DNA extraction

The selected family were from Northwest China, Shaanxi province. All individuals were diagnosed with lung squamous cell carcinoma and treated at the Department of Thoracic Surgery, Second Affiliated Hospital, Medical School, Xi'an Jiaotong University (Xi'an, China) in August 2011. Peripheral blood samples were collected from 4 individuals, 3 of whom were lung cancer patients, and 1 of whom had no malignant disease history. DNA was extracted using a QIAamp DNA Mini kit (cat. no. 51106) according to the manufacturer's instruction (Qiagen, Hilden, Germany). Histological examination was performed by hematoxylin and eosin staining of the bronchoscopic biopsies, needle aspiration biopsies and surgical specimens.

The experiments were undertaken with the understanding and written consent of each subject, and the study conforms with The Code of Ethics of the World Medical Association (Declaration of Helsinki). This study was performed in accordance with the ethical standards of the Ethics Committee of the Second Affiliated Hospital, Medical School, Xi'an Jiaotong University (Xi'an, China).

Sanger sequencing

Specific primers were designed using Primer 5 (Premier Biosoft International, Palo Alto, CA, USA; Table I). The PCR materials included 0.25 mM dNTPs, 1 unit of HotStarTaq (Qiagen), 1X HotStarTaq buffer (Qiagen) and 0.4 µM primer. PCR conditions were 33 cycles of 95°C for 50 sec, 61°C for 40 sec and 72°C for 60 sec, following initial denaturation at 95°C for 5 min. PCR products were purified using SAP, and used as templates for the sequencing reaction using Big Dye v3.1 (Applied Biosystems Life Technologies, Foster City, CA, USA). Subsequent products were run on the ABI PRISM 3130×1 Genetic Analyzer (Applied Biosystems Life Technologies). Electropherograms were analyzed using Sequence Analysis Software version 5.2 (Applied Biosystems Life Technologies).

Table I. Primer sequences.

Table I.

Primer sequences.

Primer	Sequence
EGFRrs2227983-F	GTGACCCACTCTGTCTCCG
EGFRrs2227983-R	GAAAACCCAAAACCTCCAA
NOTCH4-F	GGCTTCCTGGGTGAGACG
NOTCH4-R	TTGGTTGTGGGTAAGTGGATT
MAD1L18-F	ATGACCAGAGCAGGACCAA
MAD1L1-R	CACAGGGGAGGACAAACC
TNFRSF10A-F	CTGATGAAATGGGTCAACAAA
TNFRSF10A-R	TCCCGAGTAGCCGAGAATA
CDKN1Ars76-F	TGTGAGGTAGATGGGAGCG
CDKN1Ars76-R	GGGGAGGATTTGACGAGTG
CDKN1Ars18-F	GTGTCTAATCTCCGCCGTGAC
CDKN1Ars18-R	GCCTGCCTCCTCCCAACT
CDKN1Ars10-F	GCGTCCTCTTCTTCTTGGC
CDKN1Ars10-R	AAAGGAGAACACGGGATGAG
TP53rs1042522-F	ACCTGTGGGAAGCGAAA
TP53rs1042522-R	GGAAGGGACAGAAGATGACA
ATM-F	AATCTCCTCCTTTCTGCTGC
ATM-R	CCACTCTGCCAAGTATTATGCTAT
NBN-F	GGTTACCTCAGTGCCATTTAC
NBN-R	GCAGTGACCAAAGACCGA
XPC-F	TGGCAGCGGCTCTGATTT
XPC-R	GGCTCTGGTATGGTCTCAAGGTC
PDE4DIP-F	CCCATGACTGAAGAGTTGCT
PDE4DIP-R	CATCTCCCAAGTCCTACAAAG
CLTCL1-F	CAGAAAGATGCTTACCACCC
CLTCL1-R	GCTGCCTGATACTCACCGA
MLL3-F	TTTCATCATCAGCAGACATAAGC
MLL3-R	CACCCTACCTGTTTGGACC
ARID2-F	TTCTTTGCTTTCCATCCTTG
ARID2-R	CCTCCGTTTTACTCCTACCA
CLTCL1-F	TTATTGTTGAAGATAGAAGCAGTT
CLTCL1-R	AAAGTTAGGAGGCAGGTAGG
GNAS-F	CGGAGCCTTCAGTGGTG
GNAS-R	CTGTCGTCCCCAAATCCT
NACA-F	CAAAGGTCCCTCTACCATCTC
NACA-R	CAACAGAAGAAACGGGCAT
POT1-F	ATTTTAGGCAAGAAGTTCCACA
POT1-R	TTGAAAGCGGGAGAATACC
TFRC-F	TTCCTCAAGCCAAATAATGC
TFRC-R	CAAAGGCATAGATAATGTCAGC
NIN-F1	GCTCTGGGATTGAAGGGAC
NIN-R1	CAGGCTGGAGTACAGTGGC

[i] F, forward; R, reverse.

Whole exome sequencing

Whole exome sequencing was performed on the peripheral blood sample of the proband. Firstly, to construct a library, whole exome DNA from the patient's whole blood was treated using Agilent SureSelect Human All ExonV5 kits following the manufacturer's instructions (Agilent Technologies Inc., Santa Clara, CA, USA). Subsequent to the quality test, the qualified library was sequenced as 100-bp paired-end reads on an Illumina Hiseq 2000 platform (Illumina, San Diego, CA, USA) according to the manufacturer's instructions.

Data analysis

For whole exome sequencing, clean data was obtained after filtering the reads of low quality (reads with adapter sequence, reads with proportion of N >10% and reads with low quality base numbers of >5). Burrows-Wheeler transform methods (16) were adopted to map these reads in a human reference (UCSC hg19). Next, the Picard and Genome Analysis Toolkit (GATK) methods (17,18) were adopted for duplicate removal, local realignment and base quality recalibration. Finally, the GATK Unified Genotyper (version 3.0; Broad Institute, Cambridge, MA, USA) was used for single nucleotide variation (SNV)/InDel annotation.

Variants were annotated using the ANNOVAR software tool (http://www.openbioinformatics.org/annovar/). Annotations for function (exonic, intronic and untranslated region), reference genes, exonic function (synonymous, non-synonymous, stop-gain, frameshift and unknown), amino acid changes, 1000 Genomes Project data and dbSNP reference number were performed.

The whole exome sequencing generated large volumes of data, and several filtering criteria were applied to the data set. Firstly, variants with a low quality score (depth <20 or genotype quality <20) were filtered. Secondly, the variants with a reported frequency of >0.01 were filtered. Thirdly, synonymous changes were removed, taking only the protein-altering variants.

Results

Description of the pedigree

The lung cancer family were from Northwest China, Shaanxi province, and four members of the family were diagnosed with lung cancer (Fig. 1). The proband was a 65-year-old male who presented with a tumor in the inferior lobe of the right lung. Histological examination revealed individual cell keratinization, as well as intercellular bridge and squamous pearl formation, which resulted in a diagnosis of lung squamous cell carcinoma, stage IIIa (T3N2M0), according to the National Comprehensive Cancer Network TNM staging system (19). The patient was a heavy smoker, with 1.6 pack-year history of >10 years. Following lobectomy by thoracotomy with lymph node dissection, the patient underwent chemotherapy (120 mg/m2 cisplatin day 1, and 30 mg/m2 vinorelbine days 1 and 8 every 21 days for 4 cycles). A good response to chemotherapy was recorded. The proband had suffered from no respiratory or malignant diseases prior to the lung cancer diagnosis. Two elder brothers of the proband, who were also smokers, had succumbed to lung squamous carcinoma metastasis at the ages of 71 and 69 years old, respectively. One of the proband's nephews had been diagnosed with lung squamous cancer at the age of 51. All the lung cancer cases in the family were histologically confirmed. Other members in the family had no history of tumors or respiratory diseases. Peripheral blood samples were collected from four family members, including three of the cancer patients (I4, I6 and II3) and one healthy family member (II1). II1 was the only surviving healthy control who was elder than at least one of the lung cancer patients (Fig. 1).

Figure 1. Pedigree chart of the family. The circles and boxes represent women and men, respectively.

Whole exome sequencing

Whole exome sequencing was conducted using the peripheral blood sample of the proband, obtaining ~2.0 Gb of data and achieving an average of 60x depth for each targeted base, with 98.3% of the exomic positions covered >10x. In all, 41,210 SNVs and 3,299 InDels were identified. As described in the Materials and methods section, 2,845 variations, including SNVs and InDels, were removed due to a low quality score.

Sanger sequencing technology was applied to validate the variants identified by high throughput sequencing. In all, 10 variants were selected randomly for testing. The sequence results by Sanger sequencing confirmed the variants identified by whole exome sequencing.

Candidate gene validation

To identify the causative variants, the no reported alterations, or the alterations with low frequency (<=0.01, or no reported frequency), and the alterations which were predicted to alter the protein product (non-synonymous SNVs, splice-site mutations and InDels) were selected. In all, 975 SNVs and 1,052 InDels in 767 genes satisfied the aforementioned criteria.

Next, the candidate alterations were further restricted to 11 variants, which were located in genes with definite functions and recorded in the COSMIC driver gene database (http://cancer.sanger.ac.uk/cosmic; Table II) (20).

Table II. Variants for validation.

Table II.

Variants for validation.

Gene	Ref.	Alt.	Amino acid change	Mutation type	1000G MAF	DP	GQ
ARID2	A	G	NM_152641:exon14:c.A1759G:p.S587G	Non-synonymous SNV	0.01	69	99
GNAS	A	G	NM_080425:exon1:c.A484G:p.M162V	Non-synonymous SNV	0.0014	116	99
MLH1	T	A	NM_001167619:exon11:c.T428A:p.V143D	Non-synonymous SNV	0.01	83	99
NACA	T	C	NM_001113203:exon3:c.A1247G:p.K416R	Non-synonymous SNV	0.01	105	99
NIN	G	A	NM_016350:exon28:c.C3856T:p.R1286C	Non-synonymous SNV	0.0014	22	99
POT1	C	G	NM_001042594:exon13:c.G835C:p.D279H	Non-synonymous SNV	0.0027	70	99
TFRC	G	C	NM_001128148:exon6:c.C634G:p.L212V	Non-synonymous SNV	0.01	273	99
CLTCL1	G	T	NM_001835:exon8:c.C1342A:p.L448I	Non-synonymous SNV	0.01	61	99
PDE4DIP	GT	G	NM_001002811:exon2:c.1218delA:p.E406fs	Frameshift deletion	–	169	99
CLTCL1	A	AC	NM_001835:exon23:c.3601_3602insG:p.V1201fs	Frameshift insertion	–	60	99
MLL3	G	GT	NM_170606:exon14:c.2447_2448insA: p.Y816_I817delinsX	Stop-gain SNV	–	373	99

[i] Ref, base in the human reference genome hg19; alt, variant base; 1000G MAF, minor allele frequency in the 1000 Genomes Project; DP, depth (read depth at a certain position); GQ, genotype quality; SNV, single nucleotide variation.

Subsequently, Sanger sequencing was conducted using four peripheral blood samples from this family, from the three patients and one non-cancer family member (II1), to validate the variations in the aforementioned genes. As shown in Fig. 2, for the CLTCL1 L448I mutation, the genotypes of the three affected individuals were heterozygous, while that of the healthy man was homozygous wild-type. For the PDE4DIPE406fs mutation, the three patients carried heterozygous frameshift mutations, while the healthy individual carried the homozygous wild-type. For variations in other genes, namely ARID2, CLTCL1, GNAS, MLH1, NACA, NIN, POT1, TFRC, PDE4DIP and MLL3, the variations did not segregate well.

Figure 2. Sanger sequencing results. The vertical bars are at the mutation position. All the three patients (I4, I6, II3) carried heterozygous mutation in gene CLTCL1 and PDE4DIP, while the healthy individual (II1) carried homozygous wild type.

Discussion

The present study recruited a Chinese family in which four members had developed lung cancer, and is the first time a study of Chinese familial lung cancer has been reported. Lung cancer is one of the most common cancer types all over the world, and a number of previous studies have identified a large quantity of lung cancer driver genes, such as EGFR, Her2, AKT1, NRAS, PIK3CA, BRAF, ALK-fusion and RET-fusion (5–8). Somatic mutations of these genes have been recurrently identified in lung cancer samples (5–8). However, unlike in sporadic lung cancer, the molecular mechanisms behind inherited lung cancer remains largely unknown.

Previous studies reported the germline mutations of EGFR and HER2 as driver mutations in familial lung cancer patients (13–15). In the present study, whole exome sequencing identified 8 SNVs in the EGFR gene. However, 7 of these were synonymous, and 1 was a common variant with a frequency of >0.2. No variants that cause amino acid or splice changes were identified in the HER2 gene. The study also analyzed mutations in other inherited cancer-related genes, namely BRCA1/2, MSH2 and MLH1, but found no non-synonymous mutations in these genes. Thus, an unreported causative variation may drive lung cancer development in this family.

In this study, whole exome sequencing using a peripheral blood sample of one lung cancer member of the family totally identified >2,000 unreported or low frequency non-synonymous variants (975 SNVs and 1,052 InDels). Alterations were validated in 10 definitive cancer driver genes, and two plausible candidate genes, PDE4DIP and CLTCL1, were identified. For the PDE4DIP gene, all the patients carried a frameshift variation, while the non-cancer family member did not. The PDE4DIP protein acts as an anchor phosphodiesterase, localized in the Golgi/centrosome region of the cell. The protein is able to interact with a phosphodiesterase superfamily protein member (21,22). Functional mutations of PDE4DIP have been found to be associated with several disorders. In the study by DeWan et al, it was found that one mutation, I303L, in PDE4DIP may be the causative variant in hereditary asthma (23). CLTCL1 is a member of the clathrin family, which plays essential roles in intracellular traffic and centrosomal stabilization (24). Mutations or the abnormal expression of CLTCL1 have been reported to be associated with breast cancer, meningioma and pulmonary valve stenosis (25). Further functional studies are required to verify whether the mutations in the aforemenioned two genes cause functional change. Other variants in the ARID2, GNAS, MLH1, NACA, NIN, POT1, TFRC and MLL3 genes did not segregate perfectly with lung cancer. It is also possible that one or several alterations that were not validated are the causative mutations for this familial case.

Overall, the present study focused on a Chinese family in which four members had developed lung cancer. Whole exome sequencing was conducted using the peripheral blood sample of the proband, and 11 variants were validated in more samples. Two potentially functional variants were identified in two genes, PDE4DIP and CLTCL1. Although the present study has certain limitations, this study provides useful sources for the further study of hereditary lung cancer.

References