Identification of novel genetic variants in the mutational hotspot region 14 kb upstream of the LCT gene in a Mexican population.

Several polymorphic loci linked to lactase persistence (LP) have been described, all located in a small mutational hotspot region far upstream (∼14 kb) of the lactase (LCT) gene. One is typically found in Europeans, LCT –13910C > T, several others are found in East Africans and Arabs, e.g. LCT –13907C > G and LCT –13915T > G. The possibility of similar loci, specific to populations in South and Central America, has not received much attention so far. To identify possible novel polymorphisms in the mutational hotspot region, we sampled 158 subjects from a rural area in South-Central Mexico. DNA was isolated from serum, and Sanger sequencing of a 501 bp region spanning the LCT –13910C > T hotspot was successfully performed in 150 samples. The frequency of the European-type LCT –13910 T-allele was q = 0.202, and 35% of the population was thus lactase-persistent (CT or TT). Sixteen novel genetic variants were found amongst 11 of the subjects, all were heterozygotes: seven of the subjects were also carriers of at least one LCT –13910 T-allele. Thus, the mutational hotspot region is also a hotspot in the rural Mexican population: 11/150 subjects carried a total of 16 previously unknown private mutations but no novel polymorphism was found. The relationship between such novel genetic variants in Mexicans and lactase persistence is worthy of more investigation.

Keywords: DNA resequencing; lactose intolerance; lactase persistence; Latin America; SNP; Adult hypolactasia

After the involution of lactase (which hydrolyses lactose) in the gut, which takes place sometime after weaning, milk intake in large quantities causes abdominal symptoms. This state is called lactase non-persistence (LNP) and is commonly referred to as lactose intolerance. LNP is globally predominant. However, by a phenomenon referred to as gene-culture coevolution, some populations in different regions of the world have favored a genetic variant which causes persistence of lactase in the gut even in adult age (Lactase Persistence [LP], as opposed to ancestral non-persistence [LNP]). Such variants not only allow individuals to consume large amounts of milk without experiencing the typical abdominal complaints of lactose intolerance, but also have fundamentally transformed the diets and nutritional habits of the respective human populations since their emergence approximately 10,000 years ago.

The main causative variant in Europeans was shown by a Finnish team [[1]] to reside ∼14 kb upstream of the gene for lactase (LCT). It is a simple C > T transition at LCT –13910 (rs4988235) where C is the globally more prevalent, ancestral allele, associated with LNP, whereas the T allele confers the globally less frequent LP phenotype, mainly in Europeans. Further variants in the close vicinity of the above locus, also linked to the LP phenotype, have since been identified, for example in African and Asian populations [[2]]. We therefore shall refer to this gene region as the 'LP mutational hotspot'. Still, there has been little resequencing work in Latin America. In a recent genotyping study on Brazilian subjects a high frequency of known genetic variants was found at the hotspot, but no novel mutations were screened for [[5]]. Here we report the finding, by DNA resequencing, of 16 novel mutations in a Mexican population.

Serum samples from 158 subjects enrolled in a homocysteine-lowering trial were taken in Yecapixtla, a town of around 40,000 people located 125 km south of Mexico City, of largely Mestizo people (Amerindian-Spanish) with approximately 65% American Indian genes. The mean age was 55.1 years (SD 12.8). Ethics approval was granted by the local ethics committee at Instituto Nacional de Ciencias Médicas y Nutrición 'Salvador Zubirán' and the participants signed an informed consent form.

DNA was extracted from lyophilized serum (0.5 mL) after 400 μL of distilled water was used to reconstitute each lyophilized sample. DNA was extracted from 300 μL of the reconstituted sample with the QIAamp DNA Mini Kit spin procedure, according to the instructions of the manufacturer (QIAGEN Inc., Valencia, CA, USA), yielding 2–6 ng/μL DNA.

For Sanger sequencing of the mutational hotspot region, the PCR primers were: sense 5′-TGCTCATACGACCATG-GAAT-3′ and antisense 5′-CCATGCCATACATTTCCCTTT-3′, the amplicon size was 501 bp. PCR was performed with the HotStarTaq DNA polymerase Kit (QIAGEN Inc.). A 50 μL reaction mixture was used containing 0.2 μmol/L of each primer, 1.25 units of Taq polymerase, 2.0 mmol/L MgCl2, and 0.2 mmol/L each of dGTP, dATP, dTTP and dCTP and 10 μL DNA as template. The PCR conditions were: initial polymerase activation step at 95 °C for 15 min followed by 40 cycles of 94 °C for 30 s, 60 °C for 30 s and 72 °C for 45 s with a final extension step at 72 °C for 7 min. PCR products were purified using the Viogene® PCR-M Clean Up System kit (Viogene – Biotek Corp, USA), according to the manufacturer's instructions.

A 10 μL cycle sequencing reaction was performed using the BigDye® Terminator v3.1 cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) according to manufacturer's instruction. The extension products were purified using BigDye® Xterminator™ Purification Kit (Applied Biosystems, Foster City, CA). Sequencing was performed on an ABI Prism 3130 Genetic Analyzer. Sequence alignment with the canonical LCT upstream sequence (GenBank NM_005915.4) was performed with the SeqScape software, version 2.6 (Applied Biosystems).

Independent verification of the previously described LP-associated polymorphisms was performed using Pyrosequencing technology [[6]], modified as previously described [[7]]. The modified method simultaneously tests for the 'African' LCT –13907C > G (rs41525747) and –13915T > G (rs41380347) polymorphisms as well. Note that in the text we employ the 'consensus numbering' of these mutations based on the location of rs4988235, as originally published. Since then, the canonical position of LCT –13910C > T has shifted and a revised version of the MCM6 GenBank sequence NM_005915.5 in which the mutation lies released.

Fasting blood samples were analyzed for glucose levels pre- and post-50 g lactose ingestion using a desktop glucose analyzer. As per local laboratory standard protocol, an individual with a post-lactose blood glucose rise of greater than 30 mg/dL was considered to be LP.

Sanger sequencing of a 501 bp amplicon spanning the LP mutational hotspot yielded satisfactory results in 150 subjects (assay failure rate 5.1%). A total of 16 novel private mutations were discovered in 11 different subjects (Table 1); all were heterozygous and none was previously registered in the National Center for Biotechnology Information (NCBI) Single Nucleotide Polymorphism database (dbSNP), accessed 17 September 2015. Of these 11 subjects, only four had the LCT –13910 CC genotype and their novel mutations therefore a priori arose on LCT –13910C-haplotypes. Six of the subjects with novel mutations were LCT –13910 CT heterozygotes and it is thus likely that several of these novel mutations also occurred on an LCT –13910C-haplotype. However, one subject (No. 19, Table 1) had the LCT –13910 TT genotype and this novel mutation therefore a priori arose on a LCT –13910 T-haplotype.

Table 1. LCT mutations in a sample of 150 resequenced Mexican subjects.

1 NCBI: National Center for Biotechnology Information; Oral Lactose Load Test; aLNP (lactase non-persistence); bLP (lactase persistence).

2 Note the 'consensus numbering' is used here for the LCT position, relative to the canonical European LP mutation which is defined as LCT –13910C > T.

The LCT –13910 C/T status was independently confirmed by pyrosequencing: 151/158 subjects could be successfully genotyped (assay failure rate 4.4%). Of these, 98 subjects (65%) had the CC genotype and were thus potentially LNP (unless they would be carriers of a novel mutation conferring LP). There were 45 CT heterozygotes, and eight TT carriers. The locus was thus in Hardy-Weinberg equilibrium (χ2 = 0.86; p = .35). None carried an LCT –13907C > G or –13015T > G mutation.

Four of the subjects that carried a novel mutation were available for oral lactose loading test (OLLT). Subject No. 22 was confirmed as LNP (no significant increase (<0.5 mmol/L) of blood glucose), and we can thus conclude that the LCT –13941C > T mutation does not confer lactase persistence. Subject No. 50 was uninformative since she carried a 13910 T-allele and was found to be LP, as expected. Patients Nos. 17 and 21 were also heterozygous carriers of a 13910 T-allele, but were unexpectedly found not to be LP but rather showed a definite LNP pattern in the OLLT.

The 'European' T allele of rs4988235 was present in 35% of this Mexican population (pC = 0.798, qT = 0.202). Published figures from Spanish populations show ∼60% frequency (pC = 0.635, qT = 0.365) of this allele [[8]]. The frequencies of the T allele are perhaps therefore slightly higher than would be expected in a 65% Amer-Indian population; however, they match the prevalence of LNP recently found in Chile [[9]]. Surprisingly, we did not find any evidence that the 16 novel mutations conferred LP; in contrast, it appeared that those in two subjects (see Table 1, #17; –13889T > A and #21; –13890C > T; –13849T > A) cancelled the expected LP effect of the LCT –13910 T allele. This could, mechanistically, be due either to secondary LNP (bowel disorders causing flat mucosa), or that these novel mutations actually interfere with the LP-effect of the 13910 T-allele. The latter could, speculatively, be a novel gene-regulatory mechanism, and future studies on LNP/LP in Mexico should be designed to allow testing of this intriguing hypothesis. However, the OLLT test followed by measuring either blood glucose or breath hydrogen concentrations is prone to both false positive and false negative test results. The best – but unfortunately most demanding – approach would be direct measurements of disaccharidase activities in gut biopsies. Confirmation of our finding is thus required, and if confirmed, further cell biological investigation into how the novel variants might counteract the effect of the LCT –13910 T-allele could be rewarding. In fact, such studies on the functionality of rare variants in vitro are now ongoing [[10]].

The finding of 16 totally novel LP hotspot variants in a Mexican population, at a relatively common frequency (∼10%) confirms the LCT –14 kb region as a mutational hotspot [[2]]. Our findings strongly suggest that in the absence of a food tradition based on a high milk intake, there is not enough selection pressure to force an LP-inducing variant to reach the polymorphism prevalence, which indirectly corroborates the concept of gene-culture coevolution as regards the LCT gene. Further LP hotspot resequencing paralleled with disaccharidase measurements in gut biopsies in Mexican populations may be rewarding, especially as in Europe, LCT genotyping has become an accepted technique for diagnosing LP [[11]], but may not be universally suited, particularly in areas where milk assumes a peripheral role in the home fare cuisine.

We gratefully acknowledge The Research Committee of Örebro County Council and Nyckelfonden Örebro for financial support.

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.