- Email: [email protected]
Next-generation sequencing analysis of twelve known causative genes in congenital hypothyroidism
Clinica Chimica Acta 468 (2017) 76–80
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Next-generation sequencing analysis of twelve known causative genes in congenital hypothyroidism Xin Fan a,b, Chunyun Fu a,b, Yiping Shen a,b, Chuan Li a,b, Shiyu Luo a,b, Qifei Li a,b, Jingsi Luo a,b, Jiasun Su a,b, Shujie Zhang a,b, Xuyun Hu a,b, Rongyu Chen a,b, Xuefan Gu c, Shaoke Chen a,b,⁎ a b c
Department of Genetic Metabolism, Children's Hospital, Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region, Nanning 530003, People's Republic of China GuangXi Center for Birth Defects Research and Prevention, Nanning 530003, People's Republic of China Endocrinology and Genetic Metabolism of Institute for Pediatric Research, Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine, Shanghai 200092,China
a r t i c l e
i n f o
Article history: Received 8 January 2017 Received in revised form 4 February 2017 Accepted 14 February 2017 Available online 16 February 2017 Keywords: Congenital hypothyroidism Gene mutations Next-generation sequencing China
a b s t r a c t Background: Gene variants have been reported to be associated with congenital hypothyroidism (CH), the purpose of this study was to analyze the mutation spectrum and prevalence of 12 known causative genes (TSHR, PAX8, NKX2.1, NKX2.5, FOXE1, DUOX2, TG, TPO, GLIS3, NIS, SLC26A4 and DEHAL1) in CH in China. Methods: Peripheral venous blood samples were collected from the patients. Genomic DNA was extracted from peripheral blood leukocytes. All exons and their exon-intron boundary sequences of the 12 known CH associated genes in 66 CH patients were screened by next-generation sequencing (NGS). Results: NGS analysis of 12 known CH associated genes revealed that 32 patients (32/66, 48.5%) were detected to have at least one potentially functional variant. 21, 9, 1, 1, 1 and 1 patients were found to have potential pathogenic variants in DUOX2, TG, PAX8, SLC26A4, TSHR and TPO genes, respectively. Novel variants included one DUOX2 and one TPO missense variants of unknown significance (VUS). Conclusion: Our study expands the mutation spectrum of DUOX2 and TPO genes. 48.5% CH patients had at least one potential pathogenic variant. We found relatively high frequency of DUOX2 (31.8%) and TG (13.6%) mutations in our cohort. © 2017 Published by Elsevier B.V.
1. Introduction Congenital hypothyroidism (CH) is one of the most common and preventable endocrine disorder that affects 1 in 2000–4000 newborns [1,2]. Genetic and environmental factors are reported to be associated with CH [3]. Based on genetic factors, CH cases can be classified into two groups: (i) 80%–85% of cases result from thyroid dysgenesis, which has been linked to mutations in TSHR, PAX8, NKX2.1, NKX2.5, GLIS3 and FOXE1 genes [4,5]. (ii) the remaining 15%–20% of CH cases are caused by errors in thyroid hormone synthesis, which is associated with pathogenic variants in TG, TPO, DUOX2, NIS, SLC26A4 and DEHAL1 genes [6–8]. CH can be classified into permanent CH (PCH) and transient CH (TCH) according to the clinical course [9]. PCH refers to a persistent deficiency of thyroid hormone that requires life-long treatment. TCH
refers to a temporary deficiency of thyroid hormone at the early stage of life but thyroid hormone production recovers to normal afterwards. Up to now, the genetic basis of CH remains poorly understood, the mutational spectrum of the CH associated genes and the genotype-phenotype relationships has not been fully established [10,11]. In fact, at least 12 different genes (TSHR, PAX8, NKX2.1, NKX2.5, FOXE1, DUOX2, TG, TPO, GLIS3, NIS, SLC26A4 and DEHAL1) has been associated with CH. By the advent of next generation sequencing (NGS), there has been tremendous progress in facilitating the mutation detection in multiple genes [12,13]. In order to elucidate the molecular basis of CH in China, we performed CH associated gene mutation screening in the 66 CH Chinese cohort using NGS. 2. Materials and methods 2.1. Patients
⁎ Corresponding author at: Department of Genetic Metabolism, Children's Hospital, Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region, Nanning 530003, People's Republic of China. E-mail address: [email protected] (S. Chen).
http://dx.doi.org/10.1016/j.cca.2017.02.009 0009-8981/© 2017 Published by Elsevier B.V.
We enrolled 66 patients (including 35 with TCH and 31 with PCH) who had more comprehensive clinical data and underwent clinical reevaluation after temporary withdrawal of L-T4 therapy at approximately 2–3 years of age. Patients with syndromes or other diseases were
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
77
Table 1 Characteristics of neonates with PCH and TCH. Characteristics
PCH (31) mean ± SD
TCH (35) mean ± SD
Test type
P-value
TSH level at diagnosis (mIU/L) FT4 level at diagnosis (pmol/L) Initial L-T4 dose (μg/day) Birth weight (kg) Gestation week (week)
96.46 3.383 37.04 3.222 39.02
76.32 6.723 26.77 3.126 39.04
t-Test t-Test t-Test t-Test t-Test
0.0005⁎ b0.0001⁎ b0.0001⁎ 0.3662 0.9672
± ± ± ± ±
11.89 2.731 11.28 0.451 1.465
± ± ± ± ±
28.3 3.263 8.233 0.4055 1.396
⁎ P b 0.05.
excluded. Most of patients were initially identified by neonate screening among 356,000 newborns in the Guangxi Zhuang Autonomous Region, China, from June 2012 to June 2014. Newborn screening was done with filter paper for CH between 72 h and 7 days after birth. Blood samples were collected from the heel and the TSH level was measured by time-resolved fluorescence assay (Perkin Elmer, USA). Subjects with increased TSH (TSH ≥ 8 mIU/l) levels observed during neonatal screening were followed-up for further evaluation. Serum TSH and FT4 were determined by electrochemiluminescence assay (Cobas e601, Roche Diagnostics, USA). Diagnosis of CH was based on elevated TSH levels (TSH ≥ 10 mIU/l) and decreased FT4 levels (FT4 b 12 pmol/l). Permanent or transient CH was determined using results of thyroid function tests after temporary withdrawal of L-T4 therapy at approximately 2– 3 years of age. After one month of discontinuation of L-T4 treatment, TSH and FT4 levels were measured in venous blood sample. Individuals showed contiguous dependency on L-T4 were diagnosed with permanent CH. These children were then repeatedly evaluated at regular intervals for 1 to 1.5 year to monitor thyroid function. Those who did not need contiguous L-T4 therapy were diagnosed with transient CH. This study was approved by the local Medical Ethics Committee of the Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region. Informed consent was obtained from the parents of the patients.
2.2. Mutation detection and interpretation Genomic DNA was extracted from peripheral blood leukocytes using QIAamp DNA Blood Mini Kit (Qiagen, Germany) according to the manufacturer's protocol. CH capture panel using Illumina Truseq Custom Amplicon v1.5 kit included 12 known CH genes (TSHR, PAX8, NKX2.1, NKX2.5, FOXE1, DUOX2, TG, TPO, GLIS3, NIS, SLC26A4 and DEHAL1). All coding regions and flanking intronic regions were captured. The prepared sample library was sequenced by Illumina MiSeq instrument using MiSeq Reagent Kit v2, 500-cycles (Illumina Inc., San Diego, CA). Illumina Amplicon Viewer v1.3 and MiSeq Reporter v2.3 software were used for data analysis. In addition, a cohort of 300 ethnicity-matched healthy subjects was used to assess the variant frequencies in normal control. All control subjects had normal FT4 and TSH levels.
3. Results The characteristics of neonates with PCH and TCH are shown in Table 1. There were statistical differences between PCH and TCH regarding TSH levels at diagnosis (Fig. 1A), FT4 levels at diagnosis (Fig. 1B) and initial L-T4 dose (Fig. 1C). There were no statistical differences between
Fig. 1. The differences of TSH levels at diagnosis, FT4 levels at diagnosis, initial L-T4 dose, gestational weeks and birth weight between TCH and PCH. A) BOX plot chart showing the differences between TCH and PCH regarding TSH levels at diagnosis. B) BOX plot chart showing the differences between TCH and PCH regarding FT4 levels at diagnosis. C) BOX plot chart showing the differences between TCH and PCH regarding initial L-T4 dose. D) BOX plot chart showing the differences between TCH and PCH regarding birth weight. E) BOX plot chart showing the differences between TCH and PCH regarding gestation weeks.
78
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
Table 2 32 patients identified with potential pathogenic variants by NGS. Patients
Clinical phenotype
DUOX2
1 2 3
TCH TCH TCH
p.L1343F (het)
4 5 6
TCH TCH TCH
7 8 9 10
TCH TCH TCH TCH
11 12 13 14 15 16 17 18 19
TCH TCH TCH TCH TCH PCH PCH PCH PCH
20
PCH
21
PCH
22 23 24 25 26
PCH PCH PCH PCH PCH
27 28 29 30 31 32
PCH PCH PCH PCH PCH PCH
TG
TPO
PAX8
TSHR
SLC26A4
p.P1012L (het) p.R842X (het) p.E879K (het) p.K530X (hom) p.R1110Q (het) c.3339delC (het) p.K530X(het) p.L1726F (het) p.R411K (het) p.K530X (het) p.R683L (het) p.L1343F (het) p.E879K (het) p.L1343F (het) p.R683L (hom) p.T461I (het) p.K530X (het) p.R1110Q (het) p.R31H (het) P.D222Y (het) p.R683L (hom) p.L1343F (het) p.E879K (het) p.R683L (het) p.A1138D (het) p.R683L (hom) p.L1343F (het) p.A1138D(het) p.P1012L (het) p.P1012L (het) p.T1111R (het) c.3339delC (het) p.R683L (het) p.L1343F (het) p.P1012L (het) p.K530X (het)
p.R109Q (het) p.P1012L (het)
p.K530X (het) p.E879K (het)
p.P1012L (het) p.V233L (het)
PCH and TCH regarding birth weight (Fig. 1D) and gestation weeks (Fig. 1E). Next generation sequencing analysis of 66 CH patients including 31 PCH and 35 TCH revealed that 32 patients (Table 2) (32/66, 48.5%) were detected to have at least one potentially functional variant and all variants were Sanger confirmed (Fig. S1): 8 previously reported and one novel DUOX2 variations in 21 individuals (21/66, 31.8%), four previously reported TG variations in nine individuals (9/66, 13.6%), one previously reported PAX8 variations in one individuals (1/66, 1.5%), one previously reported SLC26A4 variations in one individuals (1/66, 1.5%), one previously reported TSHR variations in one individuals (1/66, 1.5%) and one novel TPO variations in one individuals (1/66, 1.5%). No significant sequence variations were observed in NIS, GLIS3, DEHAL1, FOXE1, NKX2.1
or NKX2.5 gene. The TG variant p.P1012L and DUOX2 variants p.R683L, p.L1343F and p.K530X are highly recurrent in our patients cohort. The present study identified one novel variant c.1232G N A(p.R411K) (Fig. 2) in DUOX2 and one novel variant c.664G N A(p.D222Y) (Fig. 2) in TPO. The two novel variations were not detected in our normal control population and classified as VUS according to our assessment using the ACMG/AMP guideline (Table 3). Among 35 cases of TCH, 8 DUOX2 variations were found in 12 individuals, 3 TG variations were detected in 3 individuals, and no significant sequence variations were observed in the remaining 20 patients with TCH. It was noted that monoallelic and biallelic DUOX2 variants were detected in seven cases and five cases of TCH patients, respectively; and monoallelic TG variants were found in three TCH patients.
Fig. 2. Novel variants in DUOX2 and TPO in 66 CH patients.
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
79
Table 3 Classification and evidence of 2 novel variants. Variant
Gene
Classification PS4 Frequency in 300 normal control
P.D222Y C.664G N T TPO VUS P.R411K C.1232G N A DUOX2 VUS
0 0
PM1 Located in functional domain
PM2 Frequency in East Asian population (and in globe)
PP3 In-silico prediction
Topological domain Topological domain
0.0001157 (0.0000166) 0.0006933(0.00004946)
NA Damaging
Classification and evidence system according to ACMG/AMP variants interpretation guidelines. PS4 = the prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. PM1 = located in a mutational hot spot and/or critical and well-established functional domain. PM2 = absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or ExAC. PP3 = multiple lines of computational evidence support a deleterious effect on the gene or gene product. VUS = Variants of Uncertain Significance; NA = data not available.
Among 31 patients with PCH, seven DUOX2 variations were found in 9 individuals, two TG variations were detected in 6 individuals, 4 patients had one variant in TPO, PAX8, TSHR or SLC26A4, and the remaining 14 PCH patients had no significant sequence variations in the 12 CH associated genes. In the meantime, there were 5, 3, 1, 1 and 1 PCH patients with monoallelic variant in TG, DUOX2, PAX8, TPO and SLC26A4 gene, respectively. Four PCH patients had triallelic DUOX2 variations, and two patients had digenic variants. The characteristics of CH patients with/out variants detected are shown in Fig. 3. There was no statistical difference between the two groups regarding TSH levels at diagnosis (Fig. 3a), FT4 levels at diagnosis (Fig. 3b), initial L-T4 dose (Fig. 3c), birth weight (Fig. 3d) or gestation weeks (Fig. 3e).
4. Discussion It is reported that iodine deficiency in some geographic regions is still a major factor leading to CH [14,15]. Apart from iodine deficiency, there are at least 12 genes known to cause CH mostly following a recessive genetic mode except PAX8 [4]. In the present study, we conducted CH gene panel screening in a cohort of 66 patients with CH from our newborn screening program. As a result, 48.5% (32/66) patients had at least one potential pathogenic variant. We found relatively high frequency of DUOX2 (31.8%) and TG (13.6%) in this sample, no mutation was detected in NIS, GLIS3, DEHAL1, FOXE1, NKX2.1 or NKX2.5 gene. These results demonstrate the efficiency of NGS in performing molecular diagnosis of CH, and variable mutation frequencies of CH associated
Fig. 3. The differences of TSH levels at diagnosis, FT4 levels at diagnosis, initial L-T4 dose, gestational weeks and birth weight between CH patients with variants detected and patients without variants detected. a) BOX plot chart showing the differences between the two groups regarding TSH levels at diagnosis. b) BOX plot chart showing the differences between the two groups regarding FT4 levels at diagnosis. c) BOX plot chart showing the differences between the two groups regarding initial L-T4 dose. d) BOX plot chart showing the differences between the two groups regarding birth weight. e) BOX plot chart showing the differences between the two groups regarding gestation weeks. A: CH patients with variants detected; B: CH patients without variants detected.
80
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
genes across human populations. Currently we do not know if heterozygous variants are contributing the occurrence of CH or digenic mode also contribute to CH (In our study, 2 PCH patients had variations in two CH associated genes.). More sequence data from large population will be needed to understand these questions, also will help to see how big fraction of CH have genetic underpinning rather than environmental factors such as iodine deficiency. It is interesting to note that 31 from 33 (96.8%) CH patients were found to have at least one potential pathogenic variant in dyshormonogenesis gene (DUOX2, TG, TPO and SLC26A4), whereas pathogenic variants in thyroid dysgenesis genes are far less. This is quite different from previous reports that the genetic cause of CH are mostly (80%–85%) due to mutations in thyroid dysgenesis genes such as TSHR, PAX8, NKX2.1, NKX2.5, GLIS3 and FOXE1 [16]. The discrepancy may simply reflect the fact that there is no sufficient data had been generated. It may also suggest that the genetic risk factors differ significantly from one ethnic group to other. Large data set from more comprehensive gene panel in more diverse population will help to reveal the genetic nature of CH. We found that most of monoallelic and biallelic DUOX2 pathogenic variants turned out to be TCH, whereas patients with triallelic DUOX2 pathogenic variants were associated with PCH, the accumulation of mutations contribute to the severity of the disease. We also found monoallelic DUOX2, TG, TPO, SLC26A4 variant in PCH patients. It is possible the patients with heterozygous variant may carry another undetected variant since NGS-based mutation screening does not detect large noncoding intragenic rearrangements or microdeletions involving one or more exons. Indeed, a recent paper demonstrated that deletion of exons accounting for about 70% GLIS3 gene mutation spectrum [5]. There are also 13 PCH who had no significant sequence variations in any of known CH genes, the patients provide an excellent model for studying CH candidate genes. Several limitations should be noted in the study when reviewing our findings. First, we did not carry out functional studies of the two novel variants. Second, the sample size used in this study is relatively small, future studies based on a larger cohort need to be conducted to confirm our findings. In conclusion, 48.5% CH patients had at least one potential pathogenic variant. We found relatively high frequency of DUOX2 (31.8%) and TG (13.6%) mutations in this cohort. Two novel variants were reported, which expanded the DUOX2/TPO mutation spectrum. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cca.2017.02.009.
Contributorship statement X.F. participated in the research design, performed the experiments, analysed the data, wrote and reviewed the manuscript; C.F. and Y.S. analysed the data, wrote and reviewed the manuscript; C.L., S.L., and Q.L. performed the NGS experiments; J.L. and J.S. performed the Sanger sequencing experiments; S.Z., X.H., R.C. and X.G. collected clinical samples and analysed the clinical data; S.C. proposed the research design, analysed the overall data, wrote and reviewed the manuscript.
Competing interests No potential conflicts of interest relevant to this original article were reported by authors.
Funding This study was supported by the National Natural Science Foundation of China (81260126), Guangxi Natural Science Foundation Program (2016GXNSFBA380192; 2012GXNSFAA053174) and Key Projects of Guangxi Health Department (2012025). Data sharing statement No additional data are available. References [1] Z. Heidari, A. Feizi, M. Hashemipour, R. Kelishadi, M. Amini, Growth development in children with congenital hypothyroidism: the effect of screening and treatment variables-a comprehensive longitudinal study, Endocrine 54 (2) (2016) 448–459, http://dx.doi.org/10.1007/s12020-016-1010-x (published Online First: Epub Date). [2] C. Fu, J. Wang, S. Luo, et al., Next-generation sequencing analysis of TSHR in 384 Chinese subclinical congenital hypothyroidism (CH) and CH patients, Clin. Chim. Acta 462 (2016) 127–132, http://dx.doi.org/10.1016/j.cca.2016.09.007 (published Online First: Epub Date). [3] V. Anastasovska, M. Kocova, Ethnicity and incidence of congenital hypothyroidism in the capital of Macedonia, J. Pediatr. Endocrinol. Metab. (2016)http://dx.doi.org/ 10.1515/jpem-2016-0178 (published Online First: Epub Date). [4] C. Fu, R. Chen, S. Zhang, et al., PAX8 pathogenic variants in Chinese patients with congenital hypothyroidism, Clin. Chim. Acta 450 (2015) 322–326, http://dx.doi. org/10.1016/j.cca.2015.09.008 (published Online First: Epub Date). [5] P. Dimitri, A.M. Habeb, F. Gurbuz, et al., Expanding the clinical spectrum associated with GLIS3 mutations, J. Clin. Endocrinol. Metab. 100 (10) (2015) E1362–E1369, http://dx.doi.org/10.1210/jc.2015-1827 (published Online First: Epub Date). [6] A.K. Nicholas, E.G. Serra, H. Cangul, et al., Comprehensive screening of eight known causative genes in congenital hypothyroidism with gland-in-situ, J. Clin. Endocrinol. Metab. (2016), jc20161879. http://dx.doi.org/10.1210/jc.2016-1879 (published Online First: Epub Date). [7] C. Fu, B. Xie, S. Zhang, et al., Mutation screening of the TPO gene in a cohort of 192 Chinese patients with congenital hypothyroidism, BMJ Open 6 (5) (2016), e010719. http://dx.doi.org/10.1136/bmjopen-2015-010719 (published Online First: Epub Date). [8] C. Fu, S. Luo, S. Zhang, et al., Next-generation sequencing analysis of DUOX2 in 192 Chinese subclinical congenital hypothyroidism (SCH) and CH patients, Clin. Chim. Acta 458 (2016) 30–34, http://dx.doi.org/10.1016/j.cca.2016.04.019 (published Online First: Epub Date). [9] G.A. Ford, S. Denniston, D. Sesser, M.R. Skeels, S.H. LaFranchi, Transient versus permanent congenital hypothyroidism after the age of 3 years in infants detected on the first versus second newborn screening test in Oregon, USA, Horm. Res. Paediatr. 86 (3) (2016) 169–177, http://dx.doi.org/10.1159/000448658 (published Online First: Epub Date). [10] V.N. Bas, H. Cangul, S.Y. Agladioglu, et al., Mild and severe congenital primary hypothyroidism in two patients by thyrotropin receptor (TSHR) gene mutation, J. Pediatr. Endocrinol. Metab. 25 (11− 12) (2012) 1153–1156, http://dx.doi.org/10.1515/ jpem-2012-0211 (published Online First: Epub Date). [11] D. Perone, G. Medeiros-Neto, C.R. Nogueira, et al., Analysis of the PAX8 gene in 32 children with thyroid dysgenesis and functional characterization of a promoter variant, J. Pediatr. Endocrinol. Metab. 29 (2) (2016) 193–201, http://dx.doi.org/10. 1515/jpem-2015-0199 (published Online First: Epub Date). [12] S. Ozen, H. Onay, T. Atik, et al., Rapid molecular genetic diagnosis with next-generation sequencing in 46, XY disorders of sex development cases: efficiency and cost assessment, Horm. Res. Paediatr. (2016)http://dx.doi.org/10.1159/000452995 (published Online First: Epub Date). [13] E. Kameta, K. Sugimori, T. Kaneko, et al., Diagnosis of pancreatic lesions collected by endoscopic ultrasound-guided fine-needle aspiration using next-generation sequencing, Oncol. Lett. 12 (5) (2016) 3875–3881, http://dx.doi.org/10.3892/ol. 2016.5168 (published Online First: Epub Date). [14] C. Li, S. Peng, X. Zhang, et al., The urine iodine to creatinine as an optimal index of iodine during pregnancy in an iodine adequate area in China, J. Clin. Endocrinol. Metab. 101 (3) (2016) 1290–1298, http://dx.doi.org/10.1210/jc.2015-3519 (published Online First: Epub Date). [15] M.S. Kibirige, S. Hutchison, C.J. Owen, H.T. Delves, Prevalence of maternal dietary iodine insufficiency in the north east of England: implications for the fetus, Arch. Dis. Child. Fetal Neonatal Ed. 89 (5) (2004) F436–F439, http://dx.doi.org/10.1136/adc. 2003.029306 (published Online First: Epub Date). [16] S. Liu, J. Chai, G. Zheng, H. Li, D. Lu, Y. Ge, Screening of HHEX mutations in Chinese children with thyroid dysgenesis, J. Clin. Res. Pediatr. Endocrinol. 8 (1) (2016) 21–25, http://dx.doi.org/10.4274/jcrpe.2456 (published Online First: Epub Date).
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Next-generation sequencing analysis of twelve known causative genes in congenital hypothyroidism Xin Fan a,b, Chunyun Fu a,b, Yiping Shen a,b, Chuan Li a,b, Shiyu Luo a,b, Qifei Li a,b, Jingsi Luo a,b, Jiasun Su a,b, Shujie Zhang a,b, Xuyun Hu a,b, Rongyu Chen a,b, Xuefan Gu c, Shaoke Chen a,b,⁎ a b c
Department of Genetic Metabolism, Children's Hospital, Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region, Nanning 530003, People's Republic of China GuangXi Center for Birth Defects Research and Prevention, Nanning 530003, People's Republic of China Endocrinology and Genetic Metabolism of Institute for Pediatric Research, Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine, Shanghai 200092,China
a r t i c l e
i n f o
Article history: Received 8 January 2017 Received in revised form 4 February 2017 Accepted 14 February 2017 Available online 16 February 2017 Keywords: Congenital hypothyroidism Gene mutations Next-generation sequencing China
a b s t r a c t Background: Gene variants have been reported to be associated with congenital hypothyroidism (CH), the purpose of this study was to analyze the mutation spectrum and prevalence of 12 known causative genes (TSHR, PAX8, NKX2.1, NKX2.5, FOXE1, DUOX2, TG, TPO, GLIS3, NIS, SLC26A4 and DEHAL1) in CH in China. Methods: Peripheral venous blood samples were collected from the patients. Genomic DNA was extracted from peripheral blood leukocytes. All exons and their exon-intron boundary sequences of the 12 known CH associated genes in 66 CH patients were screened by next-generation sequencing (NGS). Results: NGS analysis of 12 known CH associated genes revealed that 32 patients (32/66, 48.5%) were detected to have at least one potentially functional variant. 21, 9, 1, 1, 1 and 1 patients were found to have potential pathogenic variants in DUOX2, TG, PAX8, SLC26A4, TSHR and TPO genes, respectively. Novel variants included one DUOX2 and one TPO missense variants of unknown significance (VUS). Conclusion: Our study expands the mutation spectrum of DUOX2 and TPO genes. 48.5% CH patients had at least one potential pathogenic variant. We found relatively high frequency of DUOX2 (31.8%) and TG (13.6%) mutations in our cohort. © 2017 Published by Elsevier B.V.
1. Introduction Congenital hypothyroidism (CH) is one of the most common and preventable endocrine disorder that affects 1 in 2000–4000 newborns [1,2]. Genetic and environmental factors are reported to be associated with CH [3]. Based on genetic factors, CH cases can be classified into two groups: (i) 80%–85% of cases result from thyroid dysgenesis, which has been linked to mutations in TSHR, PAX8, NKX2.1, NKX2.5, GLIS3 and FOXE1 genes [4,5]. (ii) the remaining 15%–20% of CH cases are caused by errors in thyroid hormone synthesis, which is associated with pathogenic variants in TG, TPO, DUOX2, NIS, SLC26A4 and DEHAL1 genes [6–8]. CH can be classified into permanent CH (PCH) and transient CH (TCH) according to the clinical course [9]. PCH refers to a persistent deficiency of thyroid hormone that requires life-long treatment. TCH
refers to a temporary deficiency of thyroid hormone at the early stage of life but thyroid hormone production recovers to normal afterwards. Up to now, the genetic basis of CH remains poorly understood, the mutational spectrum of the CH associated genes and the genotype-phenotype relationships has not been fully established [10,11]. In fact, at least 12 different genes (TSHR, PAX8, NKX2.1, NKX2.5, FOXE1, DUOX2, TG, TPO, GLIS3, NIS, SLC26A4 and DEHAL1) has been associated with CH. By the advent of next generation sequencing (NGS), there has been tremendous progress in facilitating the mutation detection in multiple genes [12,13]. In order to elucidate the molecular basis of CH in China, we performed CH associated gene mutation screening in the 66 CH Chinese cohort using NGS. 2. Materials and methods 2.1. Patients
⁎ Corresponding author at: Department of Genetic Metabolism, Children's Hospital, Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region, Nanning 530003, People's Republic of China. E-mail address: [email protected] (S. Chen).
http://dx.doi.org/10.1016/j.cca.2017.02.009 0009-8981/© 2017 Published by Elsevier B.V.
We enrolled 66 patients (including 35 with TCH and 31 with PCH) who had more comprehensive clinical data and underwent clinical reevaluation after temporary withdrawal of L-T4 therapy at approximately 2–3 years of age. Patients with syndromes or other diseases were
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
77
Table 1 Characteristics of neonates with PCH and TCH. Characteristics
PCH (31) mean ± SD
TCH (35) mean ± SD
Test type
P-value
TSH level at diagnosis (mIU/L) FT4 level at diagnosis (pmol/L) Initial L-T4 dose (μg/day) Birth weight (kg) Gestation week (week)
96.46 3.383 37.04 3.222 39.02
76.32 6.723 26.77 3.126 39.04
t-Test t-Test t-Test t-Test t-Test
0.0005⁎ b0.0001⁎ b0.0001⁎ 0.3662 0.9672
± ± ± ± ±
11.89 2.731 11.28 0.451 1.465
± ± ± ± ±
28.3 3.263 8.233 0.4055 1.396
⁎ P b 0.05.
excluded. Most of patients were initially identified by neonate screening among 356,000 newborns in the Guangxi Zhuang Autonomous Region, China, from June 2012 to June 2014. Newborn screening was done with filter paper for CH between 72 h and 7 days after birth. Blood samples were collected from the heel and the TSH level was measured by time-resolved fluorescence assay (Perkin Elmer, USA). Subjects with increased TSH (TSH ≥ 8 mIU/l) levels observed during neonatal screening were followed-up for further evaluation. Serum TSH and FT4 were determined by electrochemiluminescence assay (Cobas e601, Roche Diagnostics, USA). Diagnosis of CH was based on elevated TSH levels (TSH ≥ 10 mIU/l) and decreased FT4 levels (FT4 b 12 pmol/l). Permanent or transient CH was determined using results of thyroid function tests after temporary withdrawal of L-T4 therapy at approximately 2– 3 years of age. After one month of discontinuation of L-T4 treatment, TSH and FT4 levels were measured in venous blood sample. Individuals showed contiguous dependency on L-T4 were diagnosed with permanent CH. These children were then repeatedly evaluated at regular intervals for 1 to 1.5 year to monitor thyroid function. Those who did not need contiguous L-T4 therapy were diagnosed with transient CH. This study was approved by the local Medical Ethics Committee of the Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region. Informed consent was obtained from the parents of the patients.
2.2. Mutation detection and interpretation Genomic DNA was extracted from peripheral blood leukocytes using QIAamp DNA Blood Mini Kit (Qiagen, Germany) according to the manufacturer's protocol. CH capture panel using Illumina Truseq Custom Amplicon v1.5 kit included 12 known CH genes (TSHR, PAX8, NKX2.1, NKX2.5, FOXE1, DUOX2, TG, TPO, GLIS3, NIS, SLC26A4 and DEHAL1). All coding regions and flanking intronic regions were captured. The prepared sample library was sequenced by Illumina MiSeq instrument using MiSeq Reagent Kit v2, 500-cycles (Illumina Inc., San Diego, CA). Illumina Amplicon Viewer v1.3 and MiSeq Reporter v2.3 software were used for data analysis. In addition, a cohort of 300 ethnicity-matched healthy subjects was used to assess the variant frequencies in normal control. All control subjects had normal FT4 and TSH levels.
3. Results The characteristics of neonates with PCH and TCH are shown in Table 1. There were statistical differences between PCH and TCH regarding TSH levels at diagnosis (Fig. 1A), FT4 levels at diagnosis (Fig. 1B) and initial L-T4 dose (Fig. 1C). There were no statistical differences between
Fig. 1. The differences of TSH levels at diagnosis, FT4 levels at diagnosis, initial L-T4 dose, gestational weeks and birth weight between TCH and PCH. A) BOX plot chart showing the differences between TCH and PCH regarding TSH levels at diagnosis. B) BOX plot chart showing the differences between TCH and PCH regarding FT4 levels at diagnosis. C) BOX plot chart showing the differences between TCH and PCH regarding initial L-T4 dose. D) BOX plot chart showing the differences between TCH and PCH regarding birth weight. E) BOX plot chart showing the differences between TCH and PCH regarding gestation weeks.
78
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
Table 2 32 patients identified with potential pathogenic variants by NGS. Patients
Clinical phenotype
DUOX2
1 2 3
TCH TCH TCH
p.L1343F (het)
4 5 6
TCH TCH TCH
7 8 9 10
TCH TCH TCH TCH
11 12 13 14 15 16 17 18 19
TCH TCH TCH TCH TCH PCH PCH PCH PCH
20
PCH
21
PCH
22 23 24 25 26
PCH PCH PCH PCH PCH
27 28 29 30 31 32
PCH PCH PCH PCH PCH PCH
TG
TPO
PAX8
TSHR
SLC26A4
p.P1012L (het) p.R842X (het) p.E879K (het) p.K530X (hom) p.R1110Q (het) c.3339delC (het) p.K530X(het) p.L1726F (het) p.R411K (het) p.K530X (het) p.R683L (het) p.L1343F (het) p.E879K (het) p.L1343F (het) p.R683L (hom) p.T461I (het) p.K530X (het) p.R1110Q (het) p.R31H (het) P.D222Y (het) p.R683L (hom) p.L1343F (het) p.E879K (het) p.R683L (het) p.A1138D (het) p.R683L (hom) p.L1343F (het) p.A1138D(het) p.P1012L (het) p.P1012L (het) p.T1111R (het) c.3339delC (het) p.R683L (het) p.L1343F (het) p.P1012L (het) p.K530X (het)
p.R109Q (het) p.P1012L (het)
p.K530X (het) p.E879K (het)
p.P1012L (het) p.V233L (het)
PCH and TCH regarding birth weight (Fig. 1D) and gestation weeks (Fig. 1E). Next generation sequencing analysis of 66 CH patients including 31 PCH and 35 TCH revealed that 32 patients (Table 2) (32/66, 48.5%) were detected to have at least one potentially functional variant and all variants were Sanger confirmed (Fig. S1): 8 previously reported and one novel DUOX2 variations in 21 individuals (21/66, 31.8%), four previously reported TG variations in nine individuals (9/66, 13.6%), one previously reported PAX8 variations in one individuals (1/66, 1.5%), one previously reported SLC26A4 variations in one individuals (1/66, 1.5%), one previously reported TSHR variations in one individuals (1/66, 1.5%) and one novel TPO variations in one individuals (1/66, 1.5%). No significant sequence variations were observed in NIS, GLIS3, DEHAL1, FOXE1, NKX2.1
or NKX2.5 gene. The TG variant p.P1012L and DUOX2 variants p.R683L, p.L1343F and p.K530X are highly recurrent in our patients cohort. The present study identified one novel variant c.1232G N A(p.R411K) (Fig. 2) in DUOX2 and one novel variant c.664G N A(p.D222Y) (Fig. 2) in TPO. The two novel variations were not detected in our normal control population and classified as VUS according to our assessment using the ACMG/AMP guideline (Table 3). Among 35 cases of TCH, 8 DUOX2 variations were found in 12 individuals, 3 TG variations were detected in 3 individuals, and no significant sequence variations were observed in the remaining 20 patients with TCH. It was noted that monoallelic and biallelic DUOX2 variants were detected in seven cases and five cases of TCH patients, respectively; and monoallelic TG variants were found in three TCH patients.
Fig. 2. Novel variants in DUOX2 and TPO in 66 CH patients.
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
79
Table 3 Classification and evidence of 2 novel variants. Variant
Gene
Classification PS4 Frequency in 300 normal control
P.D222Y C.664G N T TPO VUS P.R411K C.1232G N A DUOX2 VUS
0 0
PM1 Located in functional domain
PM2 Frequency in East Asian population (and in globe)
PP3 In-silico prediction
Topological domain Topological domain
0.0001157 (0.0000166) 0.0006933(0.00004946)
NA Damaging
Classification and evidence system according to ACMG/AMP variants interpretation guidelines. PS4 = the prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. PM1 = located in a mutational hot spot and/or critical and well-established functional domain. PM2 = absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or ExAC. PP3 = multiple lines of computational evidence support a deleterious effect on the gene or gene product. VUS = Variants of Uncertain Significance; NA = data not available.
Among 31 patients with PCH, seven DUOX2 variations were found in 9 individuals, two TG variations were detected in 6 individuals, 4 patients had one variant in TPO, PAX8, TSHR or SLC26A4, and the remaining 14 PCH patients had no significant sequence variations in the 12 CH associated genes. In the meantime, there were 5, 3, 1, 1 and 1 PCH patients with monoallelic variant in TG, DUOX2, PAX8, TPO and SLC26A4 gene, respectively. Four PCH patients had triallelic DUOX2 variations, and two patients had digenic variants. The characteristics of CH patients with/out variants detected are shown in Fig. 3. There was no statistical difference between the two groups regarding TSH levels at diagnosis (Fig. 3a), FT4 levels at diagnosis (Fig. 3b), initial L-T4 dose (Fig. 3c), birth weight (Fig. 3d) or gestation weeks (Fig. 3e).
4. Discussion It is reported that iodine deficiency in some geographic regions is still a major factor leading to CH [14,15]. Apart from iodine deficiency, there are at least 12 genes known to cause CH mostly following a recessive genetic mode except PAX8 [4]. In the present study, we conducted CH gene panel screening in a cohort of 66 patients with CH from our newborn screening program. As a result, 48.5% (32/66) patients had at least one potential pathogenic variant. We found relatively high frequency of DUOX2 (31.8%) and TG (13.6%) in this sample, no mutation was detected in NIS, GLIS3, DEHAL1, FOXE1, NKX2.1 or NKX2.5 gene. These results demonstrate the efficiency of NGS in performing molecular diagnosis of CH, and variable mutation frequencies of CH associated
Fig. 3. The differences of TSH levels at diagnosis, FT4 levels at diagnosis, initial L-T4 dose, gestational weeks and birth weight between CH patients with variants detected and patients without variants detected. a) BOX plot chart showing the differences between the two groups regarding TSH levels at diagnosis. b) BOX plot chart showing the differences between the two groups regarding FT4 levels at diagnosis. c) BOX plot chart showing the differences between the two groups regarding initial L-T4 dose. d) BOX plot chart showing the differences between the two groups regarding birth weight. e) BOX plot chart showing the differences between the two groups regarding gestation weeks. A: CH patients with variants detected; B: CH patients without variants detected.
80
X. Fan et al. / Clinica Chimica Acta 468 (2017) 76–80
genes across human populations. Currently we do not know if heterozygous variants are contributing the occurrence of CH or digenic mode also contribute to CH (In our study, 2 PCH patients had variations in two CH associated genes.). More sequence data from large population will be needed to understand these questions, also will help to see how big fraction of CH have genetic underpinning rather than environmental factors such as iodine deficiency. It is interesting to note that 31 from 33 (96.8%) CH patients were found to have at least one potential pathogenic variant in dyshormonogenesis gene (DUOX2, TG, TPO and SLC26A4), whereas pathogenic variants in thyroid dysgenesis genes are far less. This is quite different from previous reports that the genetic cause of CH are mostly (80%–85%) due to mutations in thyroid dysgenesis genes such as TSHR, PAX8, NKX2.1, NKX2.5, GLIS3 and FOXE1 [16]. The discrepancy may simply reflect the fact that there is no sufficient data had been generated. It may also suggest that the genetic risk factors differ significantly from one ethnic group to other. Large data set from more comprehensive gene panel in more diverse population will help to reveal the genetic nature of CH. We found that most of monoallelic and biallelic DUOX2 pathogenic variants turned out to be TCH, whereas patients with triallelic DUOX2 pathogenic variants were associated with PCH, the accumulation of mutations contribute to the severity of the disease. We also found monoallelic DUOX2, TG, TPO, SLC26A4 variant in PCH patients. It is possible the patients with heterozygous variant may carry another undetected variant since NGS-based mutation screening does not detect large noncoding intragenic rearrangements or microdeletions involving one or more exons. Indeed, a recent paper demonstrated that deletion of exons accounting for about 70% GLIS3 gene mutation spectrum [5]. There are also 13 PCH who had no significant sequence variations in any of known CH genes, the patients provide an excellent model for studying CH candidate genes. Several limitations should be noted in the study when reviewing our findings. First, we did not carry out functional studies of the two novel variants. Second, the sample size used in this study is relatively small, future studies based on a larger cohort need to be conducted to confirm our findings. In conclusion, 48.5% CH patients had at least one potential pathogenic variant. We found relatively high frequency of DUOX2 (31.8%) and TG (13.6%) mutations in this cohort. Two novel variants were reported, which expanded the DUOX2/TPO mutation spectrum. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cca.2017.02.009.
Contributorship statement X.F. participated in the research design, performed the experiments, analysed the data, wrote and reviewed the manuscript; C.F. and Y.S. analysed the data, wrote and reviewed the manuscript; C.L., S.L., and Q.L. performed the NGS experiments; J.L. and J.S. performed the Sanger sequencing experiments; S.Z., X.H., R.C. and X.G. collected clinical samples and analysed the clinical data; S.C. proposed the research design, analysed the overall data, wrote and reviewed the manuscript.
Competing interests No potential conflicts of interest relevant to this original article were reported by authors.
Funding This study was supported by the National Natural Science Foundation of China (81260126), Guangxi Natural Science Foundation Program (2016GXNSFBA380192; 2012GXNSFAA053174) and Key Projects of Guangxi Health Department (2012025). Data sharing statement No additional data are available. References [1] Z. Heidari, A. Feizi, M. Hashemipour, R. Kelishadi, M. Amini, Growth development in children with congenital hypothyroidism: the effect of screening and treatment variables-a comprehensive longitudinal study, Endocrine 54 (2) (2016) 448–459, http://dx.doi.org/10.1007/s12020-016-1010-x (published Online First: Epub Date). [2] C. Fu, J. Wang, S. Luo, et al., Next-generation sequencing analysis of TSHR in 384 Chinese subclinical congenital hypothyroidism (CH) and CH patients, Clin. Chim. Acta 462 (2016) 127–132, http://dx.doi.org/10.1016/j.cca.2016.09.007 (published Online First: Epub Date). [3] V. Anastasovska, M. Kocova, Ethnicity and incidence of congenital hypothyroidism in the capital of Macedonia, J. Pediatr. Endocrinol. Metab. (2016)http://dx.doi.org/ 10.1515/jpem-2016-0178 (published Online First: Epub Date). [4] C. Fu, R. Chen, S. Zhang, et al., PAX8 pathogenic variants in Chinese patients with congenital hypothyroidism, Clin. Chim. Acta 450 (2015) 322–326, http://dx.doi. org/10.1016/j.cca.2015.09.008 (published Online First: Epub Date). [5] P. Dimitri, A.M. Habeb, F. Gurbuz, et al., Expanding the clinical spectrum associated with GLIS3 mutations, J. Clin. Endocrinol. Metab. 100 (10) (2015) E1362–E1369, http://dx.doi.org/10.1210/jc.2015-1827 (published Online First: Epub Date). [6] A.K. Nicholas, E.G. Serra, H. Cangul, et al., Comprehensive screening of eight known causative genes in congenital hypothyroidism with gland-in-situ, J. Clin. Endocrinol. Metab. (2016), jc20161879. http://dx.doi.org/10.1210/jc.2016-1879 (published Online First: Epub Date). [7] C. Fu, B. Xie, S. Zhang, et al., Mutation screening of the TPO gene in a cohort of 192 Chinese patients with congenital hypothyroidism, BMJ Open 6 (5) (2016), e010719. http://dx.doi.org/10.1136/bmjopen-2015-010719 (published Online First: Epub Date). [8] C. Fu, S. Luo, S. Zhang, et al., Next-generation sequencing analysis of DUOX2 in 192 Chinese subclinical congenital hypothyroidism (SCH) and CH patients, Clin. Chim. Acta 458 (2016) 30–34, http://dx.doi.org/10.1016/j.cca.2016.04.019 (published Online First: Epub Date). [9] G.A. Ford, S. Denniston, D. Sesser, M.R. Skeels, S.H. LaFranchi, Transient versus permanent congenital hypothyroidism after the age of 3 years in infants detected on the first versus second newborn screening test in Oregon, USA, Horm. Res. Paediatr. 86 (3) (2016) 169–177, http://dx.doi.org/10.1159/000448658 (published Online First: Epub Date). [10] V.N. Bas, H. Cangul, S.Y. Agladioglu, et al., Mild and severe congenital primary hypothyroidism in two patients by thyrotropin receptor (TSHR) gene mutation, J. Pediatr. Endocrinol. Metab. 25 (11− 12) (2012) 1153–1156, http://dx.doi.org/10.1515/ jpem-2012-0211 (published Online First: Epub Date). [11] D. Perone, G. Medeiros-Neto, C.R. Nogueira, et al., Analysis of the PAX8 gene in 32 children with thyroid dysgenesis and functional characterization of a promoter variant, J. Pediatr. Endocrinol. Metab. 29 (2) (2016) 193–201, http://dx.doi.org/10. 1515/jpem-2015-0199 (published Online First: Epub Date). [12] S. Ozen, H. Onay, T. Atik, et al., Rapid molecular genetic diagnosis with next-generation sequencing in 46, XY disorders of sex development cases: efficiency and cost assessment, Horm. Res. Paediatr. (2016)http://dx.doi.org/10.1159/000452995 (published Online First: Epub Date). [13] E. Kameta, K. Sugimori, T. Kaneko, et al., Diagnosis of pancreatic lesions collected by endoscopic ultrasound-guided fine-needle aspiration using next-generation sequencing, Oncol. Lett. 12 (5) (2016) 3875–3881, http://dx.doi.org/10.3892/ol. 2016.5168 (published Online First: Epub Date). [14] C. Li, S. Peng, X. Zhang, et al., The urine iodine to creatinine as an optimal index of iodine during pregnancy in an iodine adequate area in China, J. Clin. Endocrinol. Metab. 101 (3) (2016) 1290–1298, http://dx.doi.org/10.1210/jc.2015-3519 (published Online First: Epub Date). [15] M.S. Kibirige, S. Hutchison, C.J. Owen, H.T. Delves, Prevalence of maternal dietary iodine insufficiency in the north east of England: implications for the fetus, Arch. Dis. Child. Fetal Neonatal Ed. 89 (5) (2004) F436–F439, http://dx.doi.org/10.1136/adc. 2003.029306 (published Online First: Epub Date). [16] S. Liu, J. Chai, G. Zheng, H. Li, D. Lu, Y. Ge, Screening of HHEX mutations in Chinese children with thyroid dysgenesis, J. Clin. Res. Pediatr. Endocrinol. 8 (1) (2016) 21–25, http://dx.doi.org/10.4274/jcrpe.2456 (published Online First: Epub Date).