European Journal of Medical Genetics 57 (2014) 1e4

Contents lists available at ScienceDirect

European Journal of Medical Genetics journal homepage: http://www.elsevier.com/locate/ejmg

Experimental research

Evidence of the involvement of the DHFR gene in nonsyndromic cleft lip with or without cleft palate Marcella Martinelli a, *, Ambra Girardi a, Francesca Cura a, Francesco Carinci b, Paolo Giovanni Morselli a, Luca Scapoli a a b

Department of Experimental, Diagnostic and Specialty Medicine, University di Bologna, Via Belmeloro 8, 40126 Bologna, Italy Department of Morphology, Surgery and Experimental Medicine, University of Ferrara, Via Luigi Borsari, 46, 44121 Ferrara, Italy

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 October 2013 Accepted 10 December 2013 Available online 20 December 2013

Studies aimed at evidencing genetic causes for neural tube defect (NTD) occurrence have often provided the inspiration for orofacial cleft aetiology investigations. The correlation between the two congenital malformations is provided by the similar incidence timing and the involvement of structures localized in the midline of the embryo. This connection is corroborated by the existence of a number of genes involved in both malformations. In this article, we considered the dihydrofolate reductase (DHFR) gene, previously seen implicated in NTDs, as a candidate for cleft lip with or without cleft palate (CL/P) risk. Four SNPs mapping on the DHFR gene were genotyped for 400 Italian CL/P triads, using TaqManÒ approach. The rs1677693 provided evidence of association, even if at borderline level (P value 0.049). In particular, the variant allele seems to have a protective effect OR ¼ 0.80 (95% C.I. 0.64e0.99). Moreover, the combination of rs1677693(A)-rs1650723(G) alleles showed a significant association OR 0.64 (95% C.I. 0.47e0.86) (P value ¼ 0.006). This represents the first attempt to demonstrate a role for DHFR in CL/P aetiology, howbeit the study of such gene deserves a deepening. Ó 2013 Elsevier Masson SAS. All rights reserved.

Keywords: Association Cleft lip with or without cleft palate DHFR Polymorphism

1. Introduction The role of folate in risk reduction of orofacial clefts has been supported by several studies [Botto et al., 2004; Shaw et al., 1995; van Rooij et al., 2004]. Folic acid is a water soluble B-vitamin that plays a crucial role in embryonic development. In fact, it is essential for DNA stability maintenance, being involved in DNA synthesis, repair, and methylation. Alterations involving genes responsible for each step of the folate pathway can lead to aberrations in organogenesis that result in congenital malformations such as neural tube defects or oral clefts. This research group previously demonstrated a maternal role for 5,10 methylenetetrahydrofolate reductase (MTHFR) gene and the involvement of transcobalamin 2 (TCN2) in non-syndromic cleft lip with or without palate (NS-CL/P) onset in an Italian population [Martinelli et al., 2001, 2006]. NS-CL/P is a common birth defect with a complex aetiology depending on both genetic and environmental factors. Different combinations of these factors may be responsible for the clinical variability of the cleft,

* Corresponding author. Tel.: þ39 051 2094106; fax: þ39 051 2094110. E-mail addresses: [email protected], [email protected] (M. Martinelli). 1769-7212/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.ejmg.2013.12.002

while their population frequency may account for the variation of incidence observed in different areas of the globe. Ingested folate needs to be chemically modified in order to become a one carbon donor for cell metabolism. Firstly, folate turns to dihydrofolic acid (DHF), it is then converted into tetrahydrofolate (THF) by the dihydrofolate reductase enzyme (DHFR) and finally converted into N5, N10-methylene THF. Decreased DHFR activity can potentially lead to a lower production of the active form of folic acid e the carrier of one carbon unit e possibly compromising a variety of fundamental cell activities such as nucleotide synthesis, DNA and histone methylation, that are essential for cell proliferation and regulation of gene expression [Chen et al., 1984]. Alteration of folate metabolism appears to modify the risk of at least two congenital malformations: neural tube defects (NTDs) and orofacial clefts (OFC). These embryogenesis anomalies occur almost at the same time of development and both involve the embryo midline structures. Not surprisingly, several variants of genes of the folate pathway have been involved in both NTDs and OFC [Blanton et al., 2011; Martinelli et al., 2001, 2006; Molloy et al., 2009]. Parle-McDermott and colleagues found that mothers with a 19 bp intron deletion polymorphism mapping in DHFR gene present a reduced risk of having children affected by NTD, and that this variant correlates with increased DHFR mRNA levels [ParleMcDermott et al., 2007].

2

M. Martinelli et al. / European Journal of Medical Genetics 57 (2014) 1e4

Until today, there have not been any studies evaluating the possible role of DHFR in OFC. On this occasion, a family-based association study was performed to test if DHFR polymorphisms could influence the risk of NS-CL/P. 2. Materials and methods 2.1. Sample study A sample group of 400 unrelated Italian probands with nonsyndromic CL/P and their parents was enrolled, for a total of 1175 individuals. Among patients, 286 cases were considered sporadic or non-familial, as no other relative shares the same malformation; while 114 cases had a positive familial history. In order to exclude known teratogenic influences, probands and their relatives were interviewed regarding their family history, medical history, and exposure to suspected known clefting drugs. Presence of additional defects made CL/P probands ineligible for the study. The study was approved by the ethical committee of hospitals involved in patient enrolment and complied with the Helsinki Declaration’s Ethical Principles for Medical Research Involving Human Subjects. After informed consent, peripheral venous blood samples were drawn from all participating family members, conserved in EDTA, and DNA extraction was performed, as previously described [Carinci et al., 1995].

3. Results A family-based association approach was adopted to test the involvement of DHFR genetic polymorphisms in CL/P aetiology. Genotype frequencies in probands and parents were distributed according to the HardyeWeinberg equilibrium law and no Mendelian errors were detected. The hypothesis of association between alleles or haplotypes and CL/P was tested with a likelihood ratio approach implemented in the UNPHASED program. Table 1 shows results of allelic association analysis. The rs1677693 provided evidence of association even if at borderline level (P value 0.049). The variant allele appeared to have a protective effect OR ¼ 0.80 (95% C.I. 0.64e0.99). Haplotype analysis of the DHFR locus confirmed association with CL/P with a greater confidence level. Indeed, the overall association test with the combination of rs1677693 and rs1650723 provided a global P value of 0.016. Table 2 reports the association test for each haplotype of this marker combination. The haplotype rs1677693(A)-rs1650723(G) provided the strongest signal of association OR 0.64 (95% C.I. 0.47e0.86) (P value ¼ 0.006). A permutation test, to allow multiple testing corrections over all the 33 association tests carried out in the study, was performed with 1000 replicates. The best P value for association (P ¼ 0.006) was adjusted to P ¼ 0.047, thus confirming the evidence of association between DHFR polymorphisms and the occurrence of NSCL/P.

2.2. Markers A genome region including DHFR gene (reference sequence NM_000791) and additional 20 kbp of upstream and downstream sequences, was considered to interrogate Hapmap database (Phase 3 release 2). This region included 26 SNPs with a MAF  0.05 in Caucasians. The Haplotagger tool embedded in Haploview v4.2 was used to select tagging SNPs maximizing SNP prediction accuracy [de Bakker et al., 2005]. The selected 4 SNPs (rs380691, rs11742668, rs1677693, and rs1650723) tagged 25 out of 26 SNPs with a max r2  0.87 based on the Hapmap data. The selected 4 SNPs tagged 25 out of 26 SNPs with a MAF  0.05 with a max r2  0.87 based on the Hapmap data (Phase 3 release 2). SNPs were typed using the TaqMan SNP Genotyping Assay (Assay-On-Demand ID: C___652007_10, C___32167966_10, C___3103231_10, and C___3103203_10) on a 7500 Sequence Detection System (Life Technologies, Foster City, CA) following the manufacturer’s protocol.

4. Discussion Histone and DNA methylation are epigenetic modifications involved in cell differentiation and gene expression regulation: crucial functions during embryonic development. The rapid growth and the successive fusion of maxillary processes and palatal shelves during early embryogenesis are critical steps in orofacial development, both requiring an effective methyl synthesis and availability. DHFR plays an essential role in folate pathway and in one-carbon metabolism and its activity could be essential in the maintenance of optimal DNA synthesis and epigenetic modifications. In fact, the role of DHFR is to convert dihydrofolate into tetrahydrofolate, a methyl group shuttle required not only for the de novo synthesis of thymidylic acid and certain amino acids, but especially for purine production. For this reason we hypothesized that the DHFR gene could be a potential candidate in clefting.

2.3. Statistical analysis The distribution of genotypes in both proband and parent groups was tested for deviations from the HardyeWeinberg equilibrium using Pearson’s c2 test. A family-based allelic association and haplotype analysis were performed with a likelihood ratio approach by the Unphased software v3.1.5 in a Windows Vista operative system [Dudbridge, 2008]. Two, three and four marker combinations were considered for haplotype analysis. A rare haplotype frequency threshold of 0.02 was adopted since likelihood ratio statistics may be sensitive to rare haplotypes. An overall association analysis, that tested whether any haplotypes are associated (global P value), as well as a specific test for each haplotype were carried out. The last test is a score test for a difference in risk between one haplotype and all the others pooled together. This is in contrast with the odds ratios, which are shown in relation to a single reference haplotype.

Table 1 Family based allele association analysis between DHFR SNPs and CL/P. SNP

Allele

T

NT

T-Freq NT-Freq Chisq P value

rs380691

T C

529 524 201 206

0.72 0.28

0.72 0.28

0.09

0.766

rs11742688

C T

689 691 41 39

0.94 0.06

0.95 0.05

0.05

0.816

rs1677693

C A

506 471 222 257

0.70 0.30

0.65 0.35

3.86

0.049

rs1650723

G A

579 573 145 151

0.80 0.20

0.79 0.21

0.16

0.690

Odds-R ref. 0.97 (0.76e0.22) ref. 1.06 (0.67e1.7) ref. 0.80 (0.64e0.99) ref. 0.95 (0.73e1.2)

T: estimated count in affected offspring; NT: estimated number of alleles not transmitted to affected offspring.

M. Martinelli et al. / European Journal of Medical Genetics 57 (2014) 1e4

3

Table 2 Haplotype association analysis. Haplotype M1

M2

T T C

C T C C C T

M3

T T T C

T T T T T C

NT

T-Freq

NT-Freq

Chisq

P valuea

OR(95%C.I.)

468.0 41.0 195.0

470.0 38.0 196.0

0.66 0.06 0.28

0.67 0.05 0.28

0.01 0.12 0.00

0.908 0.726 0.952

ref. 1.09(0.68e1.74) 1.00(0.79e1.27)

442.0 217.0 41.0

416.0 247.0 37.0

0.63 0.31 0.06

0.59 0.35 0.05

2.05 2.89 0.22

0.152 0.089 0.637

ref. 0.83(0.66e1.04) 1.04(0.65e1.67)

478.5 11.5 86.5 129.5

434.9 19.1 124.1 127.9

0.68 0.02 0.12 0.18

0.62 0.03 0.18 0.18

5.75 1.76 7.52 0.00

0.016 0.185 0.006 0.950

ref. 0.53(0.24e1.16) 0.64(0.47e0.86) 0.92(0.70e1.23)

226.0 206.0 40.0 190.0

224.0 228.0 34.0 176.0

0.34 0.31 0.06 0.29

0.34 0.34 0.05 0.27

0.01 1.66 0.53 0.77

0.904 0.198 0.467 0.380

ref. 0.90(0.69e1.17) 1.20(0.72e1.99) 1.08(0.81e1.44)

M4

C A C C C A A

T

G A G A

C C T C

C T C C

C C C C T

C C A A C

G A G A G

417.8 11.2 84.2 127.8 41.0

386.7 17.4 117.3 125.7 35.0

0.61 0.02 0.12 0.19 0.06

0.57 0.03 0.17 0.18 0.05

2.92 1.35 6.19 0.02 0.51

0.088 0.246 0.013 0.891 0.473

ref. 0.58(0.26e1.29) 0.67(0.49e0.92) 0.95(0.71e1.27) 1.08(0.66e1.74)

C C C C T C

C C A A C C

G A G A G G

212.9 10.1 81.1 120.9 40.0 185.0

203.9 17.1 107.1 117.9 33.0 171.0

0.33 0.02 0.12 0.19 0.06 0.28

0.31 0.03 0.16 0.18 0.05 0.26

0.30 1.91 4.04 0.05 0.73 0.79

0.581 0.167 0.044 0.830 0.392 0.374

ref. 0.61(0.27e1.39) 0.75(0.53e1.07) 1.00(0.72e1.39) 1.09(0.65e1.82) 1.11(0.82e1.49)

T: estimated count in affected offspring; NT: estimated number of haplotype not transmitted to affected offspring. a Each haplotype was compared to all the others pooled together; M1: alleles at rs380691; M2: alleles at rs11742688; M3: alleles at rs1677693; M4: alleles at rs1650723.

Our data, obtained using a family-based association approach, evidenced an association with borderline significance for the SNP rs1677693. The variant allele A at this marker seems to reduce the risk of CL/P. Something similar was evidenced by Parle-McDermott and colleagues [Parle-McDermott et al., 2007] for the carriers of a 19bp deletion in DHFR intron 1 in spina bifida case mothers. The strongest association signal was obtained in the haplotype association analysis: OR 0.64 (95% C.I. 0.47e0.86) (P value ¼ 0.006), for the combination rs1677693(A)-rs1650723(G). The findings reported in the present investigation led us to attribute a role in NS-CL/P onset to the enzyme coded by the DHFR gene. How the variant rs1677693 allele could reduce the risk of CL/P is a matter of speculation, because no additional information is available about the effect of the investigated polymorphism. Considering that the association level of haplotypes was higher compared to those of the alleles of each single SNP, the typed SNPs were very likely not directly involved in risk variation but instead linked with untyped causative polymorphisms. This investigation represents the first attempt until now to link the DHFR gene to the clefting aetiology, indeed, none genome-wide association study has ever proposed the region containing the DHFR gene as candidate in the onset of cleft. The discrepancy between published data and our results, can be explained by variances in genetic background among people who belong to different racial and ethnic groups [Croen et al., 1998; Vanderas, 1987], or by intrinsic limitations related to the study design. Indeed, in a multiple testing situation, such as a genome-wide association study, rejection of null hypothesis requires extremely stringent threshold of significance. However, even if our results do not support evidence of a major role played by DHFR in NS-CL/P onset, we cannot exclude that other polymorphisms in DHFR gene or in neighbouring loci could be

involved in the aetiology of the disease. Thus, in order to further explore the real role of this candidate gene in NS-CL/P, replication studies and additional investigations are needed. Acknowledgements We are indebted to the families participating in the study for their invaluable contribution, as well as to all the personnel involved in clinical data and specimen collection. This work was supported in part by a grant from the Association InterethnosInterplast Italy. References Blanton SH, Henry RR, Yuan Q, Mulliken JB, Stal S, Finnell RH, et al. Folate pathway and nonsyndromic cleft lip and palate. Birth Defects Res A Clin Mol Teratol 2011;91:50e60. Botto LD, Olney RS, Erickson JD. Vitamin supplements and the risk for congenital anomalies other than neural tube defects. Am J Med Genet C Semin Med Genet 2004;125C:12e21. Carinci F, Pezzetti F, Scapoli L, Padula E, Baciliero U, Curioni C, et al. Nonsyndromic cleft lip and palate: evidence of linkage to a microsatellite marker on 6p23. Am J Hum Genet 1995;56:337e9. Chen MJ, Shimada T, Moulton AD, Cline A, Humphries RK, Maizel J, et al. The functional human dihydrofolate reductase gene. J Biol Chem 1984;259:3933e 43. Croen LA, Shaw GM, Wasserman CR, Tolarova MM. Racial and ethnic variations in the prevalence of orofacial clefts in California, 1983-1992. Am J Med Genet 1998;79:42e7. de Bakker PI, Yelensky R, Pe’er I, Gabriel SB, Daly MJ, Altshuler D. Efficiency and power in genetic association studies. Nat Genet 2005;37:1217e23. Dudbridge F. Likelihood-based association analysis for nuclear families and unrelated subjects with missing genotype data. Hum Hered 2008;66:87e98. Martinelli M, Scapoli L, Pezzetti F, Carinci F, Carinci P, Stabellini G, et al. C677T variant form at the MTHFR gene and CL/P: a risk factor for mothers? Am J Med Genet 2001;98:357e60.

4

M. Martinelli et al. / European Journal of Medical Genetics 57 (2014) 1e4

Martinelli M, Scapoli L, Palmieri A, Pezzetti F, Baciliero U, Padula E, et al. Study of four genes belonging to the folate pathway: transcobalamin 2 is involved in the onset of non-syndromic cleft lip with or without cleft palate. Hum Mutat 2006;27:294. Molloy AM, Brody LC, Mills JL, Scott JM, Kirke PN. The search for genetic polymorphisms in the homocysteine/folate pathway that contribute to the etiology of human neural tube defects. Birth Defects Res A Clin Mol Teratol 2009;85:285e94. Parle-McDermott A, Pangilinan F, Mills JL, Kirke PN, Gibney ER, Troendle J, et al. The 19-bp deletion polymorphism in intron-1 of dihydrofolate reductase (DHFR) may decrease rather than increase risk for spina bifida in the Irish population. Am J Med Genet A 2007;143A:1174e80.

Shaw GM, Lammer EJ, Wasserman CR, O’Malley CD, Tolarova MM. Risks of orofacial clefts in children born to women using multivitamins containing folic acid periconceptionally. Lancet 1995;346:393e6. van Rooij IA, Ocke MC, Straatman H, Zielhuis GA, Merkus HM, SteegersTheunissen RP. Periconceptional folate intake by supplement and food reduces the risk of nonsyndromic cleft lip with or without cleft palate. Prev Med 2004;39:689e94. Vanderas AP. Incidence of cleft lip, cleft palate, and cleft lip and palate among races: a review. Cleft Palate J 1987;24:216e25.