Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Drug Metabolism and Personalized Therapy

Official journal of the European Society of Pharmacogenomics and Personalised Therapy

Editor-in-Chief: Llerena, Adrián

Editorial Board: Benjeddou, Mongi / Chen, Bing / Dahl, Marja-Liisa / Devinsky, Ferdinand / Hirata, Rosario / Hubacek, Jaroslav A. / Ingelman-Sundberg, Magnus / Maitland-van der Zee, Anke-Hilse / Manolopoulos, Vangelis G. / Marc, Janja / Melichar, Bohuslav / Meyer, Urs A. / Nair, Sujit / Nofziger, Charity / Peiro, Ana / Sadee, Wolfgang / Salazar, Luis A. / Simmaco, Maurizio / Turpeinen, Miia / Schaik, Ron / Shin, Jae-Gook / Visvikis-Siest, Sophie / Zanger, Ulrich M.

4 Issues per year


CiteScore 2016: 1.40

SCImago Journal Rank (SJR) 2016: 0.413
Source Normalized Impact per Paper (SNIP) 2016: 0.537

Online
ISSN
2363-8915
See all formats and pricing
More options …
Volume 33, Issue 1

Issues

Carboxylesterase 1 genes: systematic review and evaluation of existing genotyping procedures

Henrik Berg Rasmussen
  • Corresponding author
  • Institute of Biological Psychiatry, Mental Health Centre Sct. Hans, Copenhagen University Hospital, Boserupvej 2, 4000 Roskilde, Denmark, Phone: + 45 3864 2284, Fax: +45 3864 2300
  • Department of Science and Environment, Roskilde University, Universitetsvej 1, DK-4000 Roskilde, Denmark
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Majbritt Busk Madsen
  • Institute of Biological Psychiatry, Mental Health Centre Sct. Hans, Copenhagen University Hospital, Roskilde, Denmark
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/
Published Online: 2018-02-10 | DOI: https://doi.org/10.1515/dmpt-2017-0023

Abstract

The carboxylesterase 1 gene (CES1) encodes a hydrolase that metabolizes commonly used drugs. The CES1-related pseudogene, carboxylesterase 1 pseudogene 1 (CES1P1), has been implicated in gene exchange with CES1 and in the formation of hybrid genes including the carboxylesterase 1A2 gene (CES1A2). Hence, the CES1 region is complex. Using in silico PCR and alignment, we assessed the specificity of PCR-assisted procedures for genotyping CES1, CES1A2 and CES1P1 in studies identified in PubMed. We identified 33 such studies and excluded those that were not the first to use a procedure or lacked sequence information. After this 17 studies remained. Ten of these used haplotype-specific amplification, restriction enzyme treatment or amplicon sequencing, and included five that were predicted to lack specificity. All procedures for genotyping of single nucleotide polymorphisms in eight studies lacked specificity. One of these studies also used amplicon sequencing, thus being present in the group above. Some primers and their intended targets were mismatched. We provide experimental evidence that one of the procedures lacked specificity. Additionally, a complex pattern of segmental duplications in the CES1 region was revealed. In conclusion, many procedures for CES1, CES1A2 and CES1P1 genotyping appear to lack specificity. Knowledge about the segmental duplications may improve the typing of these genes.

This article offers supplementary material which is provided at the end of the article.

Keywords: carboxylesterase 1; published genotyping procedures; specificity assessment

Introduction

Carboxylesterase 1 (CES1) is a hydrolase implicated in the metabolism of several commonly used drugs [1]. This enzyme may also have a role in endogenous functions, including lipid metabolism [2]. The gene encoding CES1 is located on chromosome 16 with the carboxylesterase 1 pseudogene 1 (CES1P1), also designated carboxylesterase 1A3 gene (CES1A3), and carboxylesterase 1 pseudogene 2 (CES1P2) in its vicinity [3].

CES1 is subjected to structural variation, which is manifested in the presence of a hybrid in some individuals [4]. This hybrid, designated the carboxylesterase 1A2 gene (CES1A2) or CES1P1VAR, appears to have been generated by a double crossover event and has the promoter, exon 1 and a portion of intron 1 from CES1P1, while its remainder derives from CES1 [4]. Some individuals harbor CES1 as well as CES1A2 with the CES1-derived segment in CES1A2 representing a duplication. In order to distinguish CES1 from CES1A2, it has been designated CES1A1.

A previous study reported deletion of CES1 to be rare [4], but the Database of Genomic Variants [5] lists a large number of deletion and duplication variants in the region harboring CES1 and CES1P1. These variants have not been further characterized, and since they have been identified using different platforms, including single nucleotide polymorphism (SNP) arrays with different densities, they may be identical. Nonetheless, it is possible that yet-unknown structural variants exist in the region harboring CES1 and CES1P1. If so, this further adds to the complexity of this region.

Besides the duplication of a CES1 segment, which has led to the formation of CES1A2, other types of exchange of nucleotide segments between CES1 and CES1P1 appear to have taken place. The CES1 variants CES1A1b and CES1A1c are examples of this [6], [7], [8]. CES1A1c is also known as CES1VAR, whereas CES1A1b is identical or closely related to the variant that recently was named CES1SVAR [6]. CES1A1c and CES1A1b are characterized by having exon 1, with or without the regions flanking this exon, replaced by the homologous segment of CES1P1 [7]. CES1A1c is a common variant, as it has been determined at a frequency of about 0.18 in a group of subjects with different ethnic backgrounds, while CES1A1b was found at a frequency below 0.01 in the same group of subjects [6]. Besides these variants, there is a CES1A2 haplotype that has acquired a promoter segment with two overlapping Sp1 recognition sites from CES1 and therefore has significantly higher transcriptional activity than other CES1A2 haplotypes [9].

The gene duplication permits classification of CES1 into two major structural haplotypes, both having a single copy of CES1 and one copy of either CES1P1 or CES1A2 [4]. Based on the presence or absence of the subtype CES1A1c, these two haplotypes can be sub-classified into a total of four haplotypes, designated A, B, C and D (Figure 1). This complex structure presents challenges in the analysis of the genes in the region.

Haplotypes A−D of CES1. There are four haplotypes of CES1 called A−D. They differ by the presence or absence of CES1A2, CES1P1 and the CES1 variant known as CES1A1c. Haplotype A is the wild-type haplotype. Haplotypes B and D both harbor CES1A2; the haplotypes C and D are defined by the presence of CES1A1c. CES1A2 is a hybrid of CES1A1 and CES1P1. CES1A1c is a variant of CES1 in which a segment containing exon 1 and its flanking regions has been replaced by the corresponding segment of CES1P1. In CES1A1b, another subtype of CES1, the CES1P1-derived segment is shorter. Since this subtype is rare, it is not depicted in the figure. Vertical bars denote exons. CES1-related sequences are colored in saffron yellow, and CES1P1-related sequences are reddish.
Figure 1:

Haplotypes A−D of CES1.

There are four haplotypes of CES1 called A−D. They differ by the presence or absence of CES1A2, CES1P1 and the CES1 variant known as CES1A1c. Haplotype A is the wild-type haplotype. Haplotypes B and D both harbor CES1A2; the haplotypes C and D are defined by the presence of CES1A1c. CES1A2 is a hybrid of CES1A1 and CES1P1. CES1A1c is a variant of CES1 in which a segment containing exon 1 and its flanking regions has been replaced by the corresponding segment of CES1P1. In CES1A1b, another subtype of CES1, the CES1P1-derived segment is shorter. Since this subtype is rare, it is not depicted in the figure. Vertical bars denote exons. CES1-related sequences are colored in saffron yellow, and CES1P1-related sequences are reddish.

Both CES1A1 and CES1A2 have been redesignated as CES1 in the recommended nomenclature for mammalian carboxylesterase gene families [10]. However, since CES1A2 is a hybrid of CES1 and CES1P1, we prefer to maintain its designation as CES1A2. We consider the designations CES1 and CES1A1 as being synonymous but prefer the former, which is the name approved by the HUGO Gene Nomenclature Committee [11].

Segmental duplications are chromosomal elements with sizes ranging from 1 to 400 kb that typically share a degree of sequence identity of more than 90% and are present at several locations in the genome [12]. These elements may be important to consider in the design of genotyping assays since they may affect assay specificity. So far, no attention has been focused upon segmental duplications in the design of assays for genotyping CES1, CES1P1 or their hybrid, CES1A2.

Inaccurate assays appear to have been responsible for the incorrect assignment of SNPs in CES1 and CES1P1. This includes p.Asp260fs, which initially was assigned to CES1 [13] but reassigned to CES1A2 after being examined by a more specific analysis [14]. Similarly, −816A>C (rs3785161) was previously believed to reside in CES1A2, but later reassigned to CES1P1 [15]. Moreover, different studies have reported different CES1 variants and allele frequencies [16], [17], perhaps reflecting the lack of assay specificity, although different ethnic background cannot be excluded as a contributor to these differences. Taken together, it seems that the complex structure in the region harboring CES1 has not been taken into consideration sufficiently in the design of assays for the examination of this gene.

The aims of this study were to provide an overview of published procedures for genotyping CES1, CES1A2 and CES1P1, assess their specificity and hint to an improved design of such genotyping procedures.

Methods

We performed literature searches in PubMed using the terms “CES1 AND polymorphism”, “CES1 AND PCR”, “Carboxylesterase 1 AND polymorphism” and “CES1P1” to identify publications reporting genotyping and detection of variants of CES1, CES1A2 and CES1P1. All relevant studies that had been published before December 31, 2016, were identified. When available, we extracted primer and probe sequences from these studies.

Primer sequences were subjected to in silico polymerase chain reaction (PCR) analysis using Primer-BLAST, which permits specificity assessment of primer pairs with reference genomes as templates [18]. Additionally, a sequence can be entered into the program as template in the absence of this sequence in the human reference genome assembly or other reference assemblies. We used the standard setting of Primer-BLAST, which ignores targets with six or more mismatches between the primer and a target.

Three different input templates for the in silico PCR analysis were used. The first of these was constructed by combining the CES1A2 sequence AB210090.1 (promoter region, exon 1 and the flanking intronic region) with AB119998.1 (all exons and introns but only a short portion of the 5′ end) to provide the latter with more sequence information in its 5′ end. A second input template was made by supplementing the above input template with the 5′ flanking sequence of the CES1P1 reference sequence from the Homo sapiens chromosome 16, GRCh38.p2 Primary Assembly (NC_000016.10). This was necessary since there is yet no sequence information available for the CES1A2-carrying haplotype more upstream than what is provided by AB210090.1. The third input template was produced by combining the CES1A1c sequence AB210089.1 (promoter region, exon 1 and flanking intronic region) with the reference sequence of CES1 (NCBI Reference Sequence: NG_012057.1) from the Homo sapiens chromosome 16, GRCh38.p2 Primary Assembly to obtain a full-length sequence of CES1A1c.

Primer and probe sequences in the selected publications were aligned with the reference sequences of CES1 and CES1P1 from the Homo sapiens chromosome 16, GRCh38.p7 Primary Assembly. Also included in the alignments were AB119998.1 and AB210090.1, which served as reference sequences for CES1A2. Moreover, AB210089.1 was used as reference sequence for CES1A1c in the alignments. All alignments were done using Clustal Omega [19]. Since CES1A1b is extremely rare, this variant was not included in the in silico analysis or the alignments.

Based on in silico PCR and the alignments, we predicted product outcomes for the primers and compared them with the intended targets in the selected studies. We reported all predicted product outcomes in the chromosomal region harboring CES1, CES1P1 and CES1P2 followed by identification of product outcomes with complete match to the primer pairs under examination. We assumed that only primer pairs with a complete match to a template were capable of amplifying, thus disregarding incomplete matches between a primer pair and a template.

Samples of DNA from 869 Danish subjects were genotyped for CES1 rs71647871 (p.Gly143Glu) in 96-well microplates by a previously described 5′ exonuclease-based assay [20], a TaqMan® assay relying on primer and probe sequences from another study [13]. These sequences are partially or completely identical with the homologous sequences in CES1A2 and CES1P1, thus not being specific for CES1. Therefore, samples found to harbor the minor allele of rs71647871 were subjected to amplification of a fragment that contained this polymorphism using a long-range PCR specific for CES1 followed by Sanger sequencing [21].

Using the UCSC browser, we searched for segmental duplications on chromosome 16 in the region spanning from 55,719,925 to 55,838,186 and containing CES1P2, CES1P1 and CES1 with the genomic coordinates referring to the human genome assembly GRCh38/hg38 [22]. The locations and sizes of the identified segmental duplications were downloaded.

Results

PubMed searches using the terms “CES1 AND polymorphism”, “Carboxylesterase 1 AND polymorphism”, “CES1 AND PCR” and “CES1P1” identified 36, 27, 25 and 13 publications in English, respectively. There was an overlap of hits between the search terms since different search terms could identify the same publication. However, removal of duplications ensured that each publication appeared only once. After a review of the identified publications, we excluded those that did not report the use of a procedure for genotyping of CES1, CES1P1 or the hybrid CES1A2. Also a genome-wide study reporting the association of CES1 with the pharmacokinetics of dabigatran was excluded [23]. Using this approach, 33 different studies were identified.

The identified studies were arranged by genotyping procedure (Table 1). Subsequently, we briefly described the principles underlying these procedures (Table 2). Seven of the procedures for genotyping of CES1, CES1A2 or CES1P1 applied allele-specific amplification and 11 used the 5′ exonuclease-based assay (TaqMan® assay). Fourteen of the studies for genotyping SNPs in CES1, CES1A2 and CES1P1 were based on sequencing the amplified products. PCR combined with restriction fragment length polymorphism analysis (PCR-RFLP) was used by seven studies, while four used other procedures. Several of the studies applied more than one procedure and were therefore classified into more than one group.

Table 1:

Procedures used for the identification and genotyping of variants of CES1, CES1A2 and CES1P1.

Table 2:

Principles of assays used in the determination of variants of CES1, CES1A2 and CES1P1.

The procedures based on amplification followed by sequencing were designed with different purposes including genotyping of specific SNPs, identification of CES1A2 and the variant CES1A1c or the detection of novel SNPs. For these purposes, both short PCR fragments and fragments obtained by long-range PCR were used. Fragments obtained by long-range PCR were often subjected to a secondary PCR to amplify exons, thus increasing the amount of particular interesting regions and facilitating their sequencing.

The PCR-RFLPs were used to identify and distinguish the four CES1 haplotypes A–D or to genotype CES1 p.Gly143Glu, exploiting that the Glu allele has an AluI restriction site, which is absent in the Gly allele [40]. In order to identify the CES1 haplotypes A−D by PCR-RFLP, the ratios of the intensities of the restriction enzyme-treated amplified fragments were determined after electrophoretic separation. Allele-specific amplification was used for genotyping CES1A2/CES1P1 −816 A>C and for distinguishing between the four CES1 haplotypes. The 5′ exonuclease-based approach was used both for genotyping the SNPs and determining the total number of copies of CES1 and CES1A2.

We now excluded those studies that were not the first to use a method in the genotyping of CES1, CES1P1 and CES1A2, which left us with 23 different studies (Table 3, Supplementary Table 1). Sequence information permitted specificity assessment of assays from 17 of the studies. This included 10 studies that used PCR-RFLP for the identification of the CES1 haplotypes A−D or amplification followed by sequencing of amplicons disregarding sequencing aimed at the genotyping of specific SNPs (Table 4, Supplementary Table 1).

Table 3:

Specificity of procedures used for identification of variants and genotyping of CES1, CES1A2 and CES1P1.

Table 4:

In silico assessment of procedures based on sequencing of PCR products and analyses aimed at determining CES1 haplotypes A−D.

Five of the above ten studies each reported the use of one or more amplification primer pairs, which we predicted to cause unintentional amplification, i.e. off-target product outcomes. In two of the five studies, namely the studies by Marsh et al. [16] and Yamada et al. [17], the majority of the primer pairs were predicted to amplify CES1 as well as CES1A2. Coamplification of CES1P1 was predicted to occur in the procedures used by Marsh et al. [16], Tanimoto et al. [7] and Yoshimura et al. [9] for the amplification of CES1 or CES1A2. The study by Zhu et al. [14] reported the development of two long-range PCRs for amplification of fragments of about 14 kb but the first of these, designed for amplification of CES1A1, was predicted to lack the ability to amplify the variant CES1A1c. On the other hand, the predicted product outcome of the other long-range PCR, which was intended to amplify CES1A2 and CES1P1 without distinguishing between them, included CES1A1c. Accordingly, CES1A1c would be misinterpreted as being CES1A2 or CES1P1 by the procedure designed by Zhu et al. [14].

The genotyping of SNPs in CES1 and CES1P1/CES1A2 using 5′ exonuclease-based assays, amplification followed by sequencing, allele-specific amplification and PCR-RFLP in eight different studies [6], [13], [24], [25], [28], [35], [38], [40] all predicted the production of nonspecific amplicons (Table 5, Supplementary Table 1). The predicted amplicons produced by the 5′ exonuclease-based assay used by Zhu et al. for the genotyping of CES1 p.Gly143Glu included CES1A2 in addition to the desired CES1 amplicon [13]. The amplicons in the PCR-RFLP designed to identify CES1 143Glu in the study by Nemoda et al. [40] were predicted to contain CES1A2 and CES1P1 fragments, which do not harbor the AluI restriction site, in addition to the intended fragment of CES1 with the polymorphic AluI site. There were mismatches between one or more of the primers and their intended targets in the studies by Marsh et al. [16], Fukami et al. [4], Yoshimura et al. [9] and Soria et al. [24].

Table 5:

In silico assessment of procedures developed to target specific single nucleotide polymorphisms in CES1, CES1A2 and CES1P1.

Several studies used sequence information from CES1P1 to design primers in the 5′ flanking region of CES1A2 since only a small portion of this CES1A2 region has been sequenced [9], [15], [16], [17]. Hence, we also used CES1P1 sequence information in the specificity assessment of these primers. Notably, the CES1A2 sequence GenBank: AB210090.1 only contained 927 nucleotides before the start of the first coding exon. Lack of sequence information in the 3′ end of CES1A2 prevented us from assessing whether CES1A2 was coamplified by PCRs designed to amplify CES1 in the studies of Marsh et al. [16] and Cha et al. [37] (Supplementary Table 1).

Analysis of the data from the 96-well microplates used in the 5′ exonuclease-based genotyping of rs71647871 (p.Gly143Glu) revealed two genotype clusters. This included a cluster with high intensity for the probe detecting the wild-type allele and low intensity for the variant allele-detecting probe, thus being consistent with a wild-type genotype, and a cluster of samples with high intensity for both the probes. We found that 839 and 30 of the examined samples fell into the former and latter cluster, respectively. Sanger sequencing identified 28 of the samples from the latter of the two clusters as heterozygotes, while two were identified as homozygotes for the variant allele. Results from one of the 96-well microplate are shown in Figure 2.

Allelic discrimination plot from 5′ exonuclease-based analysis of rs71647871, p.Gly143Glu (A) and Sanger sequencing (B). The allelic discrimination plot of samples genotyped for rs71647871, a G>A substitution, showed the presence of two genotype clusters depicted in red and green color, respectively, besides non-template controls in black. The cluster in red consisted of 89 samples with high fluorescence intensity for the probe designed to detect the wild-type allele and low fluorescence intensity for the variant allele-detecting probe, suggesting that these samples were GG homozygotes. The cluster in green consisted of five samples with high fluorescence intensity for both the probes consistent with a GA genotype. Sanger sequencing confirmed the heterozygosity of four of these samples but identified the fifth as an AA homozygote, that is, a homozygote for the variant allele, besides confirming that the red cluster consisted of GG homozygotes.
Figure 2:

Allelic discrimination plot from 5′ exonuclease-based analysis of rs71647871, p.Gly143Glu (A) and Sanger sequencing (B).

The allelic discrimination plot of samples genotyped for rs71647871, a G>A substitution, showed the presence of two genotype clusters depicted in red and green color, respectively, besides non-template controls in black. The cluster in red consisted of 89 samples with high fluorescence intensity for the probe designed to detect the wild-type allele and low fluorescence intensity for the variant allele-detecting probe, suggesting that these samples were GG homozygotes. The cluster in green consisted of five samples with high fluorescence intensity for both the probes consistent with a GA genotype. Sanger sequencing confirmed the heterozygosity of four of these samples but identified the fifth as an AA homozygote, that is, a homozygote for the variant allele, besides confirming that the red cluster consisted of GG homozygotes.

We identified seven pairs of segmental duplications with sizes ranging from 1297 to 43,576 bp (Figure 3, Supplementary Table 2). Both sequences of four of these pairs were located in the region of interest, which spanned from CES1P2 to CES1 on chromosome 16 and included both these. One of the segmentally duplicated sequences in each of the three remaining pairs was located on chromosome 22. A palindrome was also identified that spanned from position 55,776,426 to 55,820,004 on chromosome 16, i.e. an inverted-reverse sequence with the same read from 5′ to 3′ on one strand and 3′ to 5′ on the other.

Segmental duplications in the chromosomal region harboring CES1, CES1P1 and CES1P2. The two segments in each of the duplicated pairs are shown in the same color. A palindrome sequence is in gray color. CES1P2, CES1P1 and CES1 are all in brown color. The genomic coordinates are listed below (GRCh38/hg38). Note that two sequences in a segmentally duplicated pair may differ in size as a result of insertions or deletions in one of the two sequences in the pair.
Figure 3:

Segmental duplications in the chromosomal region harboring CES1, CES1P1 and CES1P2.

The two segments in each of the duplicated pairs are shown in the same color. A palindrome sequence is in gray color. CES1P2, CES1P1 and CES1 are all in brown color. The genomic coordinates are listed below (GRCh38/hg38). Note that two sequences in a segmentally duplicated pair may differ in size as a result of insertions or deletions in one of the two sequences in the pair.

Discussion

We systematically reviewed and evaluated procedures for genotyping CES1 (CES1A1), CES1A2 and CES1P1 and made two major observations. First, a large fraction of the procedures under evaluation was predicted to lack in specificity. Second, segmental duplications in the region harboring CES1, CES1A2 and CES1P1 may be important to take into consideration in the design of assays for the examination of these genes.

The lack in specificity most commonly reflected that the presence of CES1A2 and the high degree of sequence homology between CES1 and CES1P1 had not been taken sufficiently into account. The Homo sapiens CES1A2 sequence with all exons was deposited in GenBank® and given the accession number AB119998.1 in 2003. Hence, information about the existence of CES1A2 was available to most, if not all, of the studies that we reviewed, although it was not described in scientific publications until 2007 [1] and 2008 [4]. The absence of a complete match between primers and their intended targets in some of the procedures may affect the amplification efficiency and specificity.

The segmental duplications in the region harboring CES1 are of importance to the design of primers for amplification of fragments from this region since a primer pair located within a segment of this type of duplication may coamplify the corresponding fragment in the other segment of the duplicated pair. This will lead to production of mixed amplicons consisting of intentional as well as unintentional amplicons. An approach to address this in the genotyping of CES1, CES1A2 and CES1P1 is to design primer pairs that amplify across the border of two different segmental duplications. We also suggest avoiding the palindrome in the region harboring CES1 in the design of primers, since a primer targeting such an element may act as both a forward and reverse primer.

The coamplification of undesired fragments has the potential to compromise the specificity in a genotyping assay. Specifically, the coamplification of two highly related but different fragments may lead to the appearance of false heterozygotes reflecting the formation of off-target amplicons that interfere with the genotyping of the desired fragment and result in incorrect genotype determination. False heterozygosity due to the unintentional coamplification of a pseudogene has been recognized in the analysis of other genes [45], [46] including CYP2D6 [46] and could be a problem to several of the procedures that we evaluated.

To resolve the complex structure in the region harboring CES1, selective amplification of fragments of this gene and CES1A2 followed by post-PCR analysis is required. Since this is a time-consuming procedure with low sample throughput, other procedures have been used for genotyping p.Gly143Glu and rs3815583 (−75 T>G) in CES1 and CES1A2/CES1P1 −816 A>C, including the 5′ exonuclease-based procedure [13], [28], [30], [31], allele-specific amplification [24], [25] and PCR-RFLP [40].

Our in silico assessment of the 5′ exonuclease-based procedure for genotyping p.Gly143Glu in CES1 designed by Zhu et al. [13] suggested that fragments of CES1A2 and CES1P1 are coamplified. These coamplified fragments are not targeted by the hydrolysis probe for the detection of CES1 143Glu since they do not contain the nucleotide defining this allele. However, they produce wild-type signals that may compromise the accuracy of the assay by shifting the balance between wild-type and variant signals. We confirmed this experimentally by showing that a 5′ exonuclease-based procedure for genotyping of CES1 p.Gly143Glu was not able to distinguish between samples heterozygous and homozygous for the variant allele. Similarly, the primers that coamplified a fragment of CES1 in the 5′ exonuclease-based assay designed by Geshi et al. [28] to identify the CES1A2 C allele of −816 A>C, may give rise to an increased variant signal with a risk of misinterpretation of the genotyping results [28].

The lack of ability to distinguish CES1P1 from CES1A2 is also a limitation to the allele-specific PCR developed by Ding et al. to genotype CES1A2/CES1P1 −816 A>C [25]. Additionally, the PCR-RFLP reported by Nemoda et al. [40] for genotyping p.Gly143Glu in CES1, which was based on the treatment of amplicons with AluI, has limitations. One such limitation is that CES1A2 and CES1P1 fragments, which are non-polymorphic and without the AluI restriction site, are coamplified. This may affect the ability to distinguish CES1 143Glu homozygotes from individuals heterozygous for this allele. Taken together, the increased sample throughput has its costs since the amplification of undesired products may decrease sensitivity, thus potentially leading to false-negative findings.

Our interpretation of the in silico-based prediction of the amplification product outcomes was conservative since we assumed that only primers with complete match to a template was able to amplify, thus disregarding templates with incomplete primer matches. However, templates with an incomplete match to a primer are often able to support amplification depending upon a variety of factors including the type of DNA polymerase, annealing temperature, length of the annealing step and location of the mismatch in the primer [47], [48]. Accordingly, the number of undesired targets is not necessarily confined to those with a complete primer match, suggesting that the lack in specificity may be significantly more common than predicted by our assessment.

The complexity of the chromosomal region harboring CES1 and its pseudogenes raises concerns about the accuracy of their reference sequences and the assembly in the chromosomal region of the genomic build that harbors these sequences. Notably, it is possible there are other structural variants than CES1A2 in the region concerned. Since, the genotyping assays, which we assessed, relied on the reference sequences of CES1, CES1A2 and CES1P1, inaccuracies in these sequences may have major consequences to the design of assays for their genotyping. Such inaccuracies would also have affected our assessment of the assays in question.

Since only a limited portion of the 5′ and 3′ flanking regions of the haplotype carrying CES1A2 has been sequenced, several studies designed primers for amplification of this hybrid based on CES1P1 sequence data under the assumption of sequence identity between these two in their 5′ and 3′ flanking regions. So far, it is uncertain whether sequence differences between CES1A2 and CES1P1 in their 5′ and 3′ flanks, including differences resulting from genetic variation in these sequences, have affected the specificity of the primers used for the amplification of these genes.

Whether a lack in the specificity of the procedures used for genotyping CES1, CES1A2 and CES1P1 in the studies under examination, which included population genetics and pharmacogenetics studies, has influenced their conclusions is not known. However, we believe that our study has unraveled the need for a reappraisal of the variation in CES1, CES1A2 and CES1P1 in different populations. It would also be important to know whether genotyping inaccuracies could have led to false positive associations or lack in the ability to detect true associations in the pharmacogenetic studies of these genes, thus potentially affecting the conclusions in the studies concerned. Clues to this may be obtained by comparison of findings done by studies using different genotyping procedures.

It is important to realize that experimental evaluation of the specificity of a large number of the primer pairs in the studies under examination may not be possible. In particular, this includes the primer pairs targeting the region from intron 6 and downstream in CES1 and CES1A2, for which long-range specific PCRs have not been developed.

We did not take SNPs and other types of short genetic variation in CES1, CES1A2, CES1P1 and CES1P2 into consideration in the assessment of primer specificity. A major reason for this was that the structural variation and the existence of two CES1-related pseudogenes appeared to be the major determinants of the lack of specificity of the primers in the procedures that we assessed. However, we cannot exclude that SNPs and other types of short genetic variation have also influenced primer specificity. This includes short variation in the CES1-related pseudogenes, which might affect the binding and product outcome of primers designed to amplify CES1, and vice versa.

Well-characterized reference samples of DNA that can be shared and used in the development, validation and quality control of genotyping assays are available for several drug-metabolizing enzymes [49]. So far, there are no such reference samples for CES1 and CES1A2. Development of well-characterized reference samples for these genes will be necessary in the design of assays for their genotyping.

To conclude, a variety of PCR-based procedures have been used for CES1 and CES1A2 genotyping. We predicted the production of unintentional amplicons for an unexpectedly high number of these procedures with potential implications for assay specificity and robustness. Our study also suggests that it is pivotal to take the segmental duplications in the region harboring CES1 into account in the design of assays for the typing of this gene and its hybrid CES1A2.

Acknowledgments

The project INDICES (INDIvidualised drug therapy based on pharmacogenomics: focus on carboxylesterase 1, CES1) aims at developing strategies for individualized treatment with methylphenidate and angiotensin-converting enzyme inhibitors. It is supported by grant 10-092792/DSF from the Danish Council for Strategic Research “Programme Commission on Individuals, Disease and Society”.

References

  • 1.

    Hosokawa M, Furihata T, Yaginuma Y, Yamamoto N, Koyano N, Fujii A, et al. Genomic structure and transcriptional regulation of the rat, mouse, and human carboxylesterase genes. Drug Metab Rev 2007;39:1–15. PubMedWeb of ScienceCrossrefGoogle Scholar

  • 2.

    Ross MK, Streit TM, Herring KL. Carboxylesterases: dual roles in lipid and pesticide metabolism. J Pestic Sci 2010;35:257–64. Web of SciencePubMedCrossrefGoogle Scholar

  • 3.

    Karolchik D, Hinrichs AS, Kent WJ. The UCSC genome browser. Curr Protoc Bioinformatics 2012;Chapter 1:Unit1.4. Google Scholar

  • 4.

    Fukami T, Nakajima M, Maruichi T, Takahashi S, Takamiya M, Aoki Y, et al. Structure and characterization of human carboxylesterase 1A1, 1A2, and 1A3 genes. Pharmacogenet Genomics 2008;18:911–20. CrossrefWeb of SciencePubMedGoogle Scholar

  • 5.

    MacDonald JR, Ziman R, Yuen RKC, Feuk L, Scherer SW. The database of genomic variants: a curated collection of structural variation in the human genome. Nucleic Acids Res 2014;42:D986–92. Web of SciencePubMedCrossrefGoogle Scholar

  • 6.

    Sanford JC, Wang X, Shi J, Barrie ES, Wang D, Zhu H-J, et al. Regulatory effects of genomic translocations at the human carboxylesterase-1 (CES1) gene locus. Pharmacogenet. Genomics 2016;26:197–207. Google Scholar

  • 7.

    Tanimoto K, Kaneyasu M, Shimokuni T, Hiyama K, Nishiyama M. Human carboxylesterase 1A2 expressed from carboxylesterase 1A1 and 1A2 genes is a potent predictor of CPT-11 cytotoxicity in vitro. Pharmacogenet. Genomics 2007;17:1–10. Google Scholar

  • 8.

    Wang X, Wang G, Shi J, Aa J, Comas R, Liang Y, et al. CES1 genetic variation affects the activation of angiotensin-converting enzyme inhibitors. Pharmacogenomics J 2016;16:220–30. Web of ScienceCrossrefPubMedGoogle Scholar

  • 9.

    Yoshimura M, Kimura T, Ishii M, Ishii K, Matsuura T, Geshi E, et al. Functional polymorphisms in carboxylesterase1A2 (CES1A2) gene involves specific protein 1 (Sp1) binding sites. Biochem Biophys Res Commun 2008;369:939–42. PubMedCrossrefGoogle Scholar

  • 10.

    Holmes RS, Wright MW, Laulederkind SJF, Cox LA, Hosokawa M, Imai T, et al. Recommended nomenclature for five mammalian carboxylesterase gene families: human, mouse, and rat genes and proteins. Mamm Genome 2010;21:427–41. Web of ScienceCrossrefPubMedGoogle Scholar

  • 11.

    Rasmussen HB, Madsen MB, Hansen PR, INDICES Consortium. Nomenclature for alleles of the human carboxylesterase 1 gene. Pharmacogenet. Genomics 2017;27:78–80. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 12.

    Samonte RV, Eichler EE. Segmental duplications and the evolution of the primate genome. Nat Rev Genet 2002;3:65–72. CrossrefPubMedGoogle Scholar

  • 13.

    Zhu H-J, Patrick KS, Yuan H-J, Wang J-S, Donovan JL, DeVane CL, et al. Two CES1 gene mutations lead to dysfunctional carboxylesterase 1 activity in man: clinical significance and molecular basis. Am J Hum Genet 2008;82:1241–8. PubMedWeb of ScienceCrossrefGoogle Scholar

  • 14.

    Zhu H-J, Brinda B, Froehlich TE, Markowitz JS. A discriminative analytical method for detection of CES1A1 and CES1A2/CES1A3 genetic variants. Pharmacogenet. Genomics 2012;22:215–8. Google Scholar

  • 15.

    Sai K, Saito Y, Tatewaki N, Hosokawa M, Kaniwa N, Nishimaki-Mogami T, et al. Association of carboxylesterase 1A genotypes with irinotecan pharmacokinetics in Japanese cancer patients. Br J Clin Pharmacol 2010;70:222–33. Web of SciencePubMedCrossrefGoogle Scholar

  • 16.

    Marsh S, Xiao M, Yu J, Ahluwalia R, Minton M, Freimuth RR, et al. Pharmacogenomic assessment of carboxylesterases 1 and 2. Genomics 2004;84:661–8. PubMedCrossrefGoogle Scholar

  • 17.

    Yamada S, Richardson K, Tang M, Halaschek-Wiener J, Cook VJ, Fitzgerald JM, et al. Genetic variation in carboxylesterase genes and susceptibility to isoniazid-induced hepatotoxicity. Pharmacogenomics J 2010;10:524–36. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 18.

    Ye J, Coulouris G, Zaretskaya I, Cutcutache I, Rozen S, Madden TL. Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction. BMC Bioinformatics 2012;13:134. CrossrefWeb of SciencePubMedGoogle Scholar

  • 19.

    Li W, Cowley A, Uludag M, Gur T, McWilliam H, Squizzato S, et al. The EMBL-EBI bioinformatics web and programmatic tools framework. Nucleic Acids Res 2015;43:W580–4. Web of SciencePubMedCrossrefGoogle Scholar

  • 20.

    Nelveg-Kristensen KE, Madsen MB, Torp-Pedersen C, Køber L, Egfjord M, Hansen T, et al. Prognostic impact of carboxylesterase 1 gene variants in patients with congestive heart failure treated with angiotensin-converting enzyme inhibitors. Pharmacogenet Genomics 2016;26:169–77. CrossrefWeb of ScienceGoogle Scholar

  • 21.

    Bjerre D, Rasmussen HB, INDICES Consortium. Novel procedure with improved resolution and specificity for amplification and differentiation of variants of the gene encoding carboxylesterase 1. Pharmacogenet Genomics 2017;27:155–8. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 22.

    Speir ML, Zweig AS, Rosenbloom KR, Raney BJ, Paten B, Nejad P, et al. The UCSC genome browser database: 2016 update. Nucleic Acids Res 2016;44:D717–25. PubMedCrossrefWeb of ScienceGoogle Scholar

  • 23.

    Paré G, Eriksson N, Lehr T, Connolly S, Eikelboom J, Ezekowitz MD, et al. Genetic determinants of dabigatran plasma levels and their relation to bleeding. Circulation 2013;127:1404–12. PubMedCrossrefWeb of ScienceGoogle Scholar

  • 24.

    Soria NW, Belaus A, Galván C, Pasquali MA, Velez P, del Carmen Montes C, et al. A simple allele-specific polymerase chain reaction method to detect the Gly143Glu polymorphism in the human carboxylesterase 1 gene: importance of genotyping for pharmacogenetic treatment. Genet Test Mol Biomarkers 2010;14:749–51. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 25.

    Ding X-L, Deng Y-L, Zhang J, Miao L-Y. Mutation-sensitive molecular switch method to detect CES1A2 mutation in the Chinese Han and Yao populations. Genet Test Mol Biomarkers 2011;15:659–62. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 26.

    Johnson KA, Barry E, Lambert D, Fitzgerald M, McNicholas F, Kirley A, et al. Methylphenidate side effect profile is influenced by genetic variation in the attention-deficit/hyperactivity disorder-associated CES1 gene. J Child Adolesc Psychopharmacol 2013;23:655–64. PubMedCrossrefWeb of ScienceGoogle Scholar

  • 27.

    Xie C, Ding X, Gao J, Wang H, Hang Y, Zhang H, et al. The effects of CES1A2 A(−816)C and CYP2C19 loss-of-function polymorphisms on clopidogrel response variability among Chinese patients with coronary heart disease. Pharmacogenet Genomics 2014;24:204–10. Web of ScienceCrossrefPubMedGoogle Scholar

  • 28.

    Geshi E, Kimura T, Yoshimura M, Suzuki H, Koba S, Sakai T, et al. A single nucleotide polymorphism in the carboxylesterase gene is associated with the responsiveness to imidapril medication and the promoter activity. Hypertens Res 2005;28:719–25. PubMedCrossrefGoogle Scholar

  • 29.

    Tarkiainen EK, Backman JT, Neuvonen M, Neuvonen PJ, Schwab M, Niemi M. Carboxylesterase 1 polymorphism impairs oseltamivir bioactivation in humans. Clin Pharmacol Ther 2012;92: 68–71. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 30.

    Bruxel EM, Salatino-Oliveira A, Genro JP, Zeni CP, Polanczyk GV, Chazan R, et al. Association of a carboxylesterase 1 polymorphism with appetite reduction in children and adolescents with attention-deficit/hyperactivity disorder treated with methylphenidate. Pharmacogenomics J 2013;13:476–80. PubMedCrossrefWeb of ScienceGoogle Scholar

  • 31.

    Lewis JP, Horenstein RB, Ryan K, O’Connell JR, Gibson Q, Mitchell BD, et al. The functional G143E variant of carboxylesterase 1 is associated with increased clopidogrel active metabolite levels and greater clopidogrel response. Pharmacogenet Genomics 2013;23:1–8. Web of ScienceCrossrefPubMedGoogle Scholar

  • 32.

    Tarkiainen EK, Holmberg MT, Tornio A, Neuvonen M, Neuvonen PJ, Backman JT, et al. Carboxylesterase 1 c.428G>A single nucleotide variation increases the antiplatelet effects of clopidogrel by reducing its hydrolysis in humans. Clin Pharmacol Ther 2015;97:650–8. Web of SciencePubMedCrossrefGoogle Scholar

  • 33.

    Tarkiainen EK, Tornio A, Holmberg MT, Launiainen T, Neuvonen PJ, Backman JT, et al. Effect of carboxylesterase 1 c.428G>A single nucleotide variation on the pharmacokinetics of quinapril and enalapril. Br J Clin Pharmacol 2015;80:1131–8. CrossrefWeb of ScienceGoogle Scholar

  • 34.

    Hoh BL, Gong Y, McDonough CW, Waters MF, Royster AJ, Sheehan TO, et al. CYP2C19 and CES1 polymorphisms and efficacy of clopidogrel and aspirin dual antiplatelet therapy in patients with symptomatic intracranial atherosclerotic disease. J Neurosurg 2016;124:1746–51. CrossrefPubMedGoogle Scholar

  • 35.

    Zhu H-J, Langaee TY, Gong Y, Wang X, Pepine CJ, Cooper-DeHoff RM, et al. CES1P1 variant −816A>C is not associated with hepatic carboxylesterase 1 expression and activity or antihypertensive effect of trandolapril. Eur J Clin Pharmacol 2016;72:681–7. CrossrefWeb of SciencePubMedGoogle Scholar

  • 36.

    Suzaki Y, Uemura N, Takada M, Ohyama T, Itohda A, Morimoto T, et al. The effect of carboxylesterase 1 (CES1) polymorphisms on the pharmacokinetics of oseltamivir in humans. Eur J Clin Pharmacol 2013;69:21–30. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 37.

    Cha YJ, Jeong HE, Shin JG, Kim EY, Yu KS, Cho JY, et al. Genetic polymorphisms of the carboxylesterase 1 (CES1) gene in a Korean population. Transl Clin Pharmacol 2014;22: 30–4. CrossrefGoogle Scholar

  • 38.

    Shi J, Wang X, Eyler RF, Liang Y, Liu L, Mueller BA, et al. Association of oseltamivir activation with gender and carboxylesterase 1 genetic polymorphisms. Basic Clin Pharmacol Toxicol 2016; 119:555–61. CrossrefWeb of SciencePubMedGoogle Scholar

  • 39.

    Shi J, Wang X, Nguyen J-H, Bleske BE, Liang Y, Liu L, et al. Dabigatran etexilate activation is affected by the CES1 genetic polymorphism G143E (rs71647871) and gender. Biochem Pharmacol 2016;119:76–84. CrossrefPubMedWeb of ScienceGoogle Scholar

  • 40.

    Nemoda Z, Angyal N, Tarnok Z, Gadoros J, Sasvari-Szekely M. Carboxylesterase 1 gene polymorphism and methylphenidate response in ADHD. Neuropharmacology 2009;57:731–3. Web of ScienceCrossrefPubMedGoogle Scholar

  • 41.

    Friedrichsen M, Poulsen P, Wojtaszewski J, Hansen PR, Vaag A, Rasmussen HB. Carboxylesterase 1 gene duplication and mRNA expression in adipose tissue are linked to obesity and metabolic function. PloS One 2013;8:e56861. PubMedWeb of ScienceCrossrefGoogle Scholar

  • 42.

    Hue-Roye K, Vege S. Principles of PCR-based assays. Immunohematology 2008;24:170–5. PubMedGoogle Scholar

  • 43.

    Chuang L-Y, Yang C-H, Tsui K-H, Cheng Y-H, Chang P-L, Wen C-H, et al. Restriction enzyme mining for SNPs in genomes. Anticancer Res 2008;28:2001–7. PubMedGoogle Scholar

  • 44.

    Jurinke C, Denissenko MF, Oeth P, Ehrich M, van den Boom D, Cantor CR. A single nucleotide polymorphism based approach for the identification and characterization of gene expression modulation using MassARRAY. Mutat Res 2005;573:83–95. CrossrefPubMedGoogle Scholar

  • 45.

    Brown JT, Lahey C, Laosinchai-Wolf W, Hadd AG. Polymorphisms in the glucocerebrosidase gene and pseudogene urge caution in clinical analysis of Gaucher disease allele c.1448T>C (L444P). BMC Med Genet 2006;7:69. CrossrefPubMedGoogle Scholar

  • 46.

    Riffel AK, Dehghani M, Hartshorne T, Floyd KC, Leeder JS, Rosenblatt KP, et al. CYP2D7 sequence variation interferes with TaqMan CYP2D6 (*) 15 and (*) 35 genotyping. Front Pharmacol 2015;6:312. Web of SciencePubMedGoogle Scholar

  • 47.

    Ayyadevara S, Thaden JJ, Shmookler Reis RJ. Discrimination of primer 3′-nucleotide mismatch by taq DNA polymerase during polymerase chain reaction. Anal Biochem 2000;284:11–8. PubMedCrossrefGoogle Scholar

  • 48.

    Zhang J, Li K, Liao D, Pardinas JR, Chen L, Zhang X. Different applications of polymerases with and without proofreading activity in single-nucleotide polymorphism analysis. Lab Investig J Tech Methods Pathol 2003;83:1147–54. CrossrefGoogle Scholar

  • 49.

    Pratt VM, Zehnbauer B, Wilson JA, Baak R, Babic N, Bettinotti M, et al. Characterization of 107 genomic DNA reference materials for CYP2D6, CYP2C19, CYP2C9, VKORC1, and UGT1A1: a GeT-RM and association for molecular pathology collaborative project. J Mol Diagn JMD 2010;12:835–46. CrossrefGoogle Scholar

Supplemental Material:

The online version of this article offers supplementary material (https://doi.org/10.1515/dmpt-2017-0023).

About the article

INDICES Consortium: For members of the consortium, see Supplementary Material


Received: 2017-06-28

Accepted: 2017-11-06

Published Online: 2018-02-10

Published in Print: 2018-03-28


Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.

Research funding: Danish Council for Strategic Research “Programme Commission on Individuals, Disease and Society”, grant 10-092792/DSF.

Employment or leadership: None declared.

Honorarium: None declared.

Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.


Citation Information: Drug Metabolism and Personalized Therapy, Volume 33, Issue 1, Pages 3–14, ISSN (Online) 2363-8915, ISSN (Print) 2363-8907, DOI: https://doi.org/10.1515/dmpt-2017-0023.

Export Citation

©2018 Walter de Gruyter GmbH, Berlin/Boston. Copyright Clearance Center

Supplementary Article Materials

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

Comments (0)

Please log in or register to comment.
Log in