Multiple endocrine neoplasia type 1: a new germline “homozygous” variant (c.201delC) caused by detection errors

Multiple endocrine neoplasia type 1 (MEN1), characterized by neoplasia in multiple endocrines and nonendocrine tissues, is an autosomal dominant hereditary tumor syndrome with a prevalence of about 1/30,000. The three major endocrine tissues affected by tumors in MEN1 are the parathyroid (95%), enteropancreatic neuroendocrine (50%), and anterior pituitary (40%). A diagnosis of MEN1 may be established in an individual by one of three criteria: on the basis of the occurrence of two or more primary MEN1-associated endocrine tumors (i.e. parathyroid adenoma, enteropancreatic tumor, and pituitary adenoma); the occurrence of one of the MEN1-associated tumors in a first-degree relative of a patient with a clinical diagnosis of MEN; and identification of a germline MEN1 mutation in an individual, even if they are asymptomatic and have not yet developed serum biochemical or radiological abnormalities indicative of tumor development [1]. Besides, MEN1 patients could present with many other hormone-secreting, hormone nonsecreting, and nonendocrine tumors, such as adrenal cortical tumors, foregut carcinoids (bronchial, thymic or of the gastric enterochromaffin-like cells), facial angiofibromas, truncal collagenomas, lipomas, meningiomas, Barrett’s esophagus, leiomyoma (uterine in females or in the esophagus), and ependymoma [2]. Approximately 25% of MEN1 patients die of a malignant gastrinoma (gastrointestinal neuroendocrine tumors) or foregut carcinoid tumor [2].

The related gene MEN1, which encodes the protein MENIN [3, 4], was first reported in 1988 [5] and is located on chromosome region 11q13 [3, 4]. It is composed of 10 exons that encode a 610 or 615 amino acid nuclear protein. MENIN interacts with the activator protein 1 (AP1) transcription factor JunD to inhibit JunD-activated transcription [6], which is related to cell growth regulation, cell cycle, genome stability, and synaptic plasticity. Detection of MEN1 gene variants helps to distinguish MEN1 syndrome from other solitary tumors, such as somatic loss of heterozygosity (LOH) on chromosome 11q13, which is observed in 5–50% of such non-hereditary common sporadic tumors [7]. MEN1 gene variants are distributed throughout the coding region, which has no obvious hot spots and no obvious genotype/phenotype correlation with tumor spectrum or clinical characteristics. More than 1,000 variants have been reported, including frameshift deletions or insertions (40%), nonsense (20%) and missense (25%) variants, splice site changes (8%), deletions or insertions (6%), and large deletions (1%) [7]. Thus, 70–75% of the MEN1 gene variants are inactivating, causing premature protein truncation. However, many variants (such as rare missense in nonfamilial cases) can be classified as variants of unknown clinical significance due to the lack of available assays that could provide evidence for an adverse physiological consequence of such variants [8].

In this study, Sanger sequencing was performed on the proband who had a pancreatic tumor, parathyroid tumor, adrenal tumor, and suspicion of gastrinoma. The suspected MEN1 variant site in exon 2 was screened using the combination of bioinformatics and public database. The results indicated that a variant in an intron could affect the sequencing results and led to the occurrence, at first, of homozygotes. The three family members who were detected as homozygotes in the 1st and 2nd times, were found to be heterozygotes by resequencing. A new MEN1 germline variant [NM_130802: c.201delC (p.Ala68Profs*51) on Chr11:64577381 on assembly GRCh37] was found, which was the genetic pathogenesis of MEN1 in this family.

In this study, we found a proband with typical tumor symptoms of MEN1, including a pancreatic tumor, parathyroid tumor, adrenal tumor, and suspicion of gastrinoma, as well as pituitary changes. Subsequently, Sanger sequencing was used to test the proband for the MEN1 gene. We found a new gene variant in exon 2 of MEN1, which could lead to these clinical symptoms. Due to the heritability of MEN1, the proband’s family members were tested as well. Subsequently, we found that this Chinese family had MEN1 and we found a new germline variant of NM_130802 [c.201delC (p.Ala68Profs*51) on Chr11:64577381 on assembly GRCh37].

As we all know, the MEN1 gene plays a role as a growth suppressor in MEN1 tumorigenesis. In the germline, an inactivated variant in the MEN1 gene on chromosome 11 causes MEN1 syndrome. About 90% of MEN1 cases are usually inherited from affected parents; the other 10% of cases are due to a de novo variant [10]. The loss of heterozygotes at the MENIN site on chromosome 11q13 showed biallelic inactivation of MEN1 [15]. Knudson’s two-hit model of MEN1 gene tumorigenesis [3, 4, 16] is supported by the harmful germline variants observed in the MEN1 kindreds and the loss of heterozygosity observed in the tumors of MEN1 patients in a MEN1 family. Since the cloning of the MEN1 gene in 1997, more than 1,000 families have been reported as MEN1 [17]. Marini et al. reported an analysis of MEN1 variants in 410 patients’ germlines. It was found that there were 99 different variants, of which 41 were frameshift, 26 missense, 13 nonsense, 11 splice site variants, 4 in-frame small deletions, and 4 large intragenic deletions across one exon [18]. In 2008, Lemos and Thakker conducted a comprehensive analysis of 1336 MEN1 variants that had been reported in the first decade after the identification of the MEN1 gene [17]. Paola Concolino et al. reported 208 new germline variants of MEN1 from 2007 to 2015 [19]. We found the MEN1 genetic variant c.201delC (p.Ala68Profs*51) to be a new germline variant site in exon 2 that had not yet been reported.

In the MEN1 gene, the most mutated exons are 2, 9, and 10. In particular, in exons 2 and 10, the most common type of variant is a frameshift [19]. Most of the frameshift and nonsense variants cause protein truncation, resulting in the loss of functional domains, including NLSs located at the C-terminal segment. In the past few years, six new intron variants have been found in MEN1 patients. One of the intron variants (IVS3 + 18 C > T) relates to a c.1546-1547insC variant in a Chinese MEN1 family, which was reported by Zha et al. [20]. Zhiwei Ning et al. [14] found a germline variant of a heterozygous G to A variation at the nucleotide position-1 of intron 5 (c.825-1G > A or IVS5-1G > A) in a MEN1 family.

Interestingly, a very rare situation in this family was that the proband and her daughter were shown to have the homozygous variant in exon 2 when the variant was detected during the first and second times. This result could not be explained by the Mendelian law of inheritance. An animal experiment confirmed that the homozygous variant of MEN1 could lead to death in rats [21]. Obviously, this “homozygous” variant might have been a mistaken diagnosis.

The first reason that we considered was that the large fragment deletion in the gene, which led to the DNA single strand, was detected twice in the Sanger sequencing. Gross deletions, usually detected by the MLPA technique [22‐25], are the rarest kind of MEN1 variant. The complete MEN1 gene deletion has been considered by different authors [22, 26, 27]. The first (exons 1–3) [24, 25], central (exons 5–6) [23], and final (exons 8–10) regions of the gene [28] have been described by other gross deletions. Beijers et al. [13] were the first to report on a family with combined germline and somatic mosaicism for MEN1. They used MLPA to analyze the father and found both germline and somatic mosaicism of MEN1. However, the MLPA results showed that there was no large fragment deletion in the genes of the three MEN1 patients in this study.

Faced with this weird “homozygous” variant result, we considered secondly the primer problem. In the former two detections in which we got the suspicious results, we had designed the forward primer in intron 1 and the reverse primer in intron 2 in order to contain all the bases in exon 2. In the third detection, we re-designed both the forward and reverse primers in exon 2 and close to the variant site. We got the gene detection results that not only complied with Mendelian law of inheritance, but also made sense in her genetic family. Thus, we believe that there were some mistakes exist in the first two times gene detections. The possible reason for this situation is that the gene sequence corresponding to primer most probably not compatible with the intronic variant(s) present in this family. In other words, the primer pairs stated for first two gene detections are actually located in regions with a relatively dense variant presence (intron).

We tried to find the variant site or large fragment deletion range in intron 1. However, after too many times experiments, we still can’t accurately detect the variant site in intron. This may be due to the fact that there are relatively dense variants site in the introns, and we cannot design a suitable primer pair to perform the sanger sequence. Since introns have limited meaning in the structure of translation products, they do not code for proteins and will not cause changes in phenotype. Therefore, in this study, we giving up to continue searching for variant site in intron. It could be useful for further scientific research, but useless for the clinical diagnosis and treatment of MEN1 to proband. But in general, it explained the reason for the genetic test error. When faced with this situation again, clinicians could consider the presence of this situation and solve this issue by re-designing the primer closed to the variant site in exon.

There are two limitations for this study. Firstly, the proband has temporarily refused the invasive examinations (including CT or ultrasound-guided needle biopsy) and surgery for any of her tumors, so we could not obtain tumor tissue for further genetic heterozygosity testing, pathological diagnosis, or cell biology testing. Secondly, about the adrenal tumor, we didn’t check the catecholamines because this proband never have hypertension. Thirdly, due to the fact that there are relatively dense variants site in the introns, although we performed too many times experiment try to find out the accurate intron variant site that led to the detection error, it ended in failure. And considering the introns were limited meaningful for the structure of translation products and clinical diagnosis and treatment, we are finally giving up to find the accurate variant site in intron.

This is the first report of a Chinese family with a new MEN1 germline variant in exon 2 (c.201delC (p.Ala68Profs*51)). It can improve the identification of clinical forms of MEN1 and can be used to diagnose the disease at an early stage. In addition, when the detection error present, the clinician should re-design the detection primer close to the variant site could help find the real genetic changes.