Genetic Polymorphism of Cyp2d6 Gene Coding Region and Association Between Them With Vivax Malaria Relapse in Yunnan Province, China

Background Primaquine is one of two medications able to effectively eliminate the hypnozoites of P. vivax, and therefore has been recommended by World Health Organization (WHO) as the anti-relapsing drug for vivax malaria patients. However, the reduced CYP2D6 enzyme activity caused by gene mutations has been considered to result in the failure of primaquine relapse vivax malaria. Based on the analysis of CYP2D6 gene polymorphisms in vivax malaria cases the current study aims to determine the association between human CYP2D6 polymorphism and the relapse of vivax malaria. Blood samples of vivax malaria cases who received "chloroquine/primaquine eight-day therapy" from 2014 to 2018 in Yunnan Province were collected. The suspected relapsed cases were determined by epidemiology and gene sequence alignment. Human genomic DNA was extracted from the blood samples, nine exons of CYP2D6 gene were amplied by polymerase chain reaction (PCR) and PCR products were then sequenced. The coding DNA sequence (CDS) of CYP2D6 gene was obtained by splicing and aligning with the non-mutation reference sequence. The mutation loci, haplotypes (star alleles) and genotypes of CYP2D6 gene were inferred and their association with vivax malaria relapse were analyzed. A total of 120 blood samples from 45 suspected relapsed cases of vivax malaria and 75 non-relapsed cases were collected and two nested-PCR products (2411bp and 2388bp) containing exon1-4 and exon5-9, were obtained from 119 samples. 119 CDS chains (1491bp length) were formed by splicing and combining sequencing sequences, A total of 12 mutation loci were identied in 238 CDS chains (including 119 chains from direct sequencing and another 119 chains identied from sister chromosome). Among them, the locus at c.886 closest (OR=2.167, 95%CI: 1.104~4.252, P<0.05). were 23 haplotypes (Hap_1~Hap_23)

genotyping method can create more stable results in predicting the phenotype of CYP2D6 enzyme activity [18,[22][23][24][25][26]. The prediction is based on the locus status of CYP2D6 gene [13], or the allelic combination form of genotype [27][28]. However, more than 130 CYP2D6 alleles have been de ned by the Cytochrome P450 Nomenclature Committee at https://www.pharmvar.org/gene/CYP2D6. The prediction of CYP2D6 enzyme activity by performing genotype analysis needs to be further optimized to enhance accuracy and systematisms [29].
Gene mutations of enzymes related to drug metabolism not only affect the clinical consequences of drug metabolism, but can even lead to individual adverse event [13]. Therefore, understanding the speci c CYP2D6 genetic polymorphism in different individuals is the prerequisite to ensure radical treatment of vivax malaria based on personalized primaquine dose [24]. Despite the frequent use of primaquine in malaria endemic areas of Yunnan Province, the systematic survey about genetic polymorphisms of CYP2D6 has been not conducted. In this study, based on the gene sequence method the polymorphism investigation of CYP2D6 gene from vivax malaria cases and the association between human CYP2D6 polymorphisms and relapse of vivax malaria were explored in order to effectively assess the risk of using primaquine to stop malaria transmission or eliminate the source of P. vivax in Yunnan Province.

Ethics statement
The study was approved by the Yunnan Institute of Parasitic Diseases and Ethical Committee. Genetic testing was performed on stored blood samples obtained as part of the routine diagnostic work-up of patients with fever who were suspected to have malaria. Although there was no risk and the data processing after sample collection was done anonymously, informed consent was obtained.

Subjects and blood samples
The subjects were diagnosed cases of P. vivax in Yunnan Province. Mono-infection of P. vivax was con rmed by both microscopy examination and Plasmodium 18S rRNA gene detection (PCR) by Yunnan Province Malaria Diagnosis Referent Laboratory (YPMDRL). All vivax malaria cases were treated by oral Chloroquine therapy (total 1550 mg) during spending three days, followed by an 8-day course of primaquine (22.5 mg/day). The subjects include two groups of cases: (1) Suspected relapsed cases of vivax malaria (SR group). The SR cases were re-diagnosed as P. vivax infection within 28-180 days after initial diagnosis and who has not been exposed to malaria transmission again between two clinical malaria outbreaks through epidemiological investigation [30]. Besides, the cases were con rmed to be infected by the same P. vivax isolates, when the alignment results of sequencing sequences of csp (circumsporozoite surface protein) and msp-1 (merozoite surface protein 1) gene showed not variation, both Variable sites and He being 0, between isolates from initial and recurrence vivax malaria cases. The SR cases were con rmed from the vivax malaria cases diagnosed in Yunnan Province from 2014 to 2018.
(2) Non-relapsed cases of vivax malaria (NR group). The NR cases were selected by simple random sampling from the 143 cases of vivax malaria diagnosed in 2018, who were followed up for one year and had no recurrent history of P. vivax. The infection sources of P. vivax were con rmed according to epidemiological investigation in all the subjects: those without a travel history to epidemic areas outside Yunnan Province within the last 30 days before the onset of malaria were de ned as local cases; those have history of travelling to epidemic regions, such as Myanmar and Africa, were regarded as imported cases.
The blood samples of all the subjects were the rest part of the routine diagnostic work-up provided by the Centers for Disease Control and Prevention in Yunnan Province. The patient's dried blood spots (DBS) samples for genetic analysis were derived from 0.6 ml of venous blood, which were then put in a drying tube and stored at -80 ℃ before DNA extraction.
Extraction of human genomic DNA Three dried blood spots (diameter = 5 mm) were made, and human genomic DNA was extracted according to the manufacturer's instructions of the QIAgen Mini Kit (Germany, QIAamp Company's DNA Mini Kit), according to the manufacture's instruction. The extracted DNA was stored at -20 ℃.

PCR ampli cation of CYP2D6 gene fragment
The primers of CYP2D6 for polymerase chain reaction (PCR) ampli cation were designed by using GenBank (https://www.ncbi.nlm.nih.gov/gene/), reference sequence (ID NC_000022.11) was used as template, following the reaction condition in previous studies [23,31]. The primers for rst-round nested PCR covering the coding region of CYP2D6 gene exons1~4 were: 5'-CCAGTGACAGATAAGGGTGC-3' and 5'-GACGTGGATAGGAGGTACAGAG-3'; the primers for second-round nested PCR were: 5'-GGTGACTTCTCCGACCAGG-3' and 5'-TTCCCAAACCCATCTATGC-3'. The nal ampli cation region was 42131088~42128678, and the expected fragment length of the product was 2411 bp. The primers that cover exons 5~9 of CYP2D6 gene were: 5'-GCCGACTGAGCCCTGGGAGGTAGGTA-3' and 5'-GCTGGGGCCTGAGACTT-3. The ampli cation area was 42126035~42128422, and the ampli cation products had an expected fragment length of 2388 bp. For all the PCR reaction systems, we used 2.6μl DNA template, 14.0μl 2 × Taq PCR hybrid system (QIAGEN, Germany), 0.7μl upstream primer (20umol / L), 0.7μl downstream primers (20umol / L). The total volume was adjusted to 25.0μl with ddH 2 O. PCR reaction conditions were clari ed as follows: 92 to 94 ° C for 2-5 mins; 92 to 94 ° C for 10-30s, 50 to 56 ° C for 15-30s, 68 to 72 ° C for 2 to 3 min, 35 cycles; 68 to 72 ° C for 7 mins. The triplicated parallel repetition was adopted for each PCR reaction. The ampli ed products were observed using 1.5% agarose gel electrophoresis. The positive ampli cation products were sent to Shanghai Meiji Biomedical Technology Co, Ltd. for sequencing by using the dideoxy chaintermination method. Only the sequences that showed identical results in at least two tests were used for subsequent analysis.

Mutation analysis of the coding region in CYP2D6 gene
The sequencing results were aligned by using DNAStar v5.10 and BioEdit v7.0.9.0 software. The obtained all DNA sequences were assessed by using the Basic Local Alignment Search Tool (BLAST, http://blast.ncbi.nlm.nih.gov/Blast.cgi) in the NCBI platform. Sequences with identi cations equals 100% and Query cover above 99% were considered as CYP2D6 gene sequence of human, this chain was identi ed as the rst DNA sequence of each sample in this study. The second DNA sequence from sister chromosome was based on the rst one, which supplements the base pairs that have been shown in the sequencing peak but not been read by the sequencer [32]. The DNA sequence of exon1~9 region of CYP2D6 gene was determined and spliced into the CDS of CYP2D6 gene in the order from exon1 to exon9.MEGA v5.04 software was used to align the CDSs with the non-mutation sequence to con rm the loci of missense mutation and synonymous mutation. The de ned mutation loci were queried for the corresponding ID in Genbank. The diploid of each mutation loci was identi ed by checking the sequencing peak maps (Supplement 2). DnaSP v6.11.01 software was used to identify haplotypes of all CDSs, and calculate the parameters of CDSs diversity index (π) and expected heterozygosity (He). Every haplotype of CDSs chain was given the allele type according to the criteria from Human Cytochrome P450 Allele Nomenclature Committee Database (NM_000106.5) [33]. Those Haplotypes not matched with known CYP2D6 alleles were grouped into the "other" category when calculating odds ratio. The genotype of each sample was composed of the sister CDS chains. The all suballeles under every star (*) allele were combined when calculating the frequency of allele or genotype.
SPSS software (version 21.0; IBN; Chicago; IL) was used to conduct chi-square test on the frequency of genotypes and mutation loci co-existed in SR and NR cases. The odds ratio (OR) value of genotypes and mutation loci with statistical signi cance were calculated by comparing to the relapse rate of P. vivax. The signi cance level was set as P< 0.05; OR 1 indicates that the haplotype is a protective factor, otherwise a risk factor.

Results
Sample information and PCR ampli cation of human CYP2D6 gene A total of 45 SR vivax malaria cases were con rmed by epidemiology and gene sequence alignment (Supplement 3), including 44 cases with one suspected relapsed event, 1 case with three suspected relapsed events. The geographic information of 45 SR cases and 75 NR cases is listed in Table 1. The malefemale ratio was 3:1, and the proportions of cases imported and infected from Myanmar, Africa countries and Yunnan were 97.5% (117/120), 0.8% (1/120) and 1.7% (2/120), respectively. Most of the patients presented only one suspected relapsed episode (97.8%) Locus polymorphism of CDS chains and their association with vivax malaria relapse A total of 119 PCR ampli cation products of CYP2D6 gene were sequenced. The 119 DNA sequencing sequences were trimmed as the rst CDS chains (Genbank accession number: MT339075-MT339193) containing the complete exon1-9 (total length = 1491bp) from all samples. Another 119 sister CDS chains were deduced from the sequencing peak maps. Base substitutions at 12 loci, such as c.31 and c.100, were determined by comparing with nonmutation sequence (NC: 000022.11) in. 238 CDS chains ( Table 2). All the mutation loci were given the corresponding ID from Genbank and no new mutation loci were found. The proportions of third-base and rst-base substitution in the codon triplet were 41.7% (5/12), and the proportion of second-base substitution was 16.6% (2/12). 7 missense mutation loci and 5 synonymous mutation loci were determined. Of these mutation loci, the SR cases accounted for 91.7% (11/12), whereas the mutation loci of NR cases accounted for 66.7% (8/12) ( Table 2).
To our knowledge, the current study is the rst one that attempts to analyze mutation polymorphism in the whole coding region of human CYP2D6 gene to Yunnan Province, thereby verifying its possible association with thwarted CYP2D6 enzyme activity in relapsed vivax malaria patients after primaquine treatment. In this study, the mutation at c. 886 locus of CYP2D6 gene was calculated to increase the vivax malaria relapse risk of 2.167-fold (P 0.05).
Combined with the observation by Wang et al. [39] and Gaedigk et al. [40], that the base substitution at c. 886 locus will lead to the decrease of CYP2D6 enzyme activity. It is inferred that one of the reasons for the relapse of vivax malaria may be relate to the decrease of CYP2D6 enzyme activity caused at the c.886.mutation of CYP2D6 gene in Yunnan Province Among the 23 alleles de ned from 23 haplotypes in all the CDSs chains, the proportion of CYP2D6*10 reached as high as 45.4% (108/238). The composition of CYP2D6 gene alleles is the same as that found by Qian et al. [32] in Chinese Han population, also consistent with previous studies which reported that CYP2D6*10 allele is most frequently detected in Asian populations [31,[35][36]. Although the frequencies of CYP2D6*10 allele de ned from SR group and NR group did show signi cant different, but the frequency of CYP2D6*10 allele was the highest in both groups, and CYP2D6*10 has been an allele leading to the decrease of CYP2D6 enzyme activity. It is suggested that the danger. of vivax malaria infection reservoir should be ignored under wildly existing CYP2D6*10 allele which results in for primaquine not to effectively kill P. vivax hypnozoite of vivax malaria cases in Yunnan Province.
On the other hand, the CYP2D*2 allele was more common in the suspected relapsed patients in this study. This result is not in line with that of Brasil et al. [30], which reported that the detected frequency of CYP2D*2 in the relapsed patients was signi cantly higher in the non-relapsed patients. Such very different results might be attributed to the different subjects chosen between two studies, the subjects in this study were vivax malaria cases who mainly settled in the traditional malaria endemic areas in Yunnan Province, and the long-term malaria epidemic is thought to result in positive screening on human defective genes [41][42][43]. Moreover, in this study all of alleles were identi ed only based on the mutations of CYP2D6 gene exon regions, but it is still different from the allelic recognition including intron mutations [33,44]. Therefore, the homogeneity of CYP2D6*2 allele in the two studies may be the reason why the results between research display different. Although the CYP2D6*2 allele is generally considered to be an allele of CYP2D6 normal enzyme activity, the multiple mutation loci making up CYP2D6*2 contain the mutation at c. 886 locus which has the effect of reducing the activity of CYP2D6 enzyme [39][40]. Therefore, this study may further support the complexity of CYP2D6*2 allele effect on CYP2D6 enzyme activity change [39][40][44][45].
As expected, there were eight genotypes to be found only in the SP group, such as genotype *4/*4, *2/*39, etc., rather than in all vivax malaria cases in this study, and the association between different genotypes of CYP2D6 gene and the relapse of vivax malaria was not found, too. However, in view of the reality, that the genotype *10/*10 still accounted for the largest proportion (30.3%) in the study sample, and found the non-functional genotype (*4/*4) [33] only in SR group. It is once again suggested that Yunnan Province should not ignore the risk of decline in therapeutic e cacy of primaquine caused by the decrease of CYP2D6 enzyme activity in malaria endemic areas. In addition, the ampli cation failure case in this study needs to be further veri ed whether the really existing the full gene deletion (*5/*5) [33], or not the other factors resulting in the failure of PCR ampli cation for CYP2D6 gene.
However, this study has been some limitations. Firstly, the sample size is not large enough, and the detected low-frequency alleles (*4, *m-*x, Supplement 5) of CYP2D6 genes cannot be effectively analyzed. Secondly, only the possibility of P. vivax re-infection was excluded in identi cation of the relapsed vivax malaria patients, but other confounding factors which may also in uence on the killing effect for the P. vivax hypnozoites under using primaquine, such as no excluding out the effect from primaquine-resistance Plasmodium, the quality of primaquine, and the exposure experience to other inhibitors in the CYP2D6 enzyme activity. Therefore, this study could not infer the causal relationship between the recurrence of vivax malaria and the mutation of CYP2D6 gene.
Finally, although the whole coding region polymorphism of CYP2D6 gene was detected and analyzed in this study, but still lacked the research on the others factors, such as the mutation in intron region and the copy number of CYP2D6 gene, which has been considered to also disturb on the activity of CYP2D6 enzyme. So the activity of CYP2D6 cannot be indirectly predicted through CYP2D6 genotype in this study. In view of this de ciency, the research team will introduce multiplex ligation-dependent probe ampli cation (MLPA) and other methods to improve the accurate of identi cation between different human CYP2D6 genotypes and the effect of prediction for CYP2D6 enzyme activity in future, nally in order to provide a molecular support for the safe and effective application of primaquine in Yunnan Province.

Conclusion
In conclusion, this study preliminarily reveals the polymorphisms of CYP2D6 gene coding region and the associate between them with vivax malaria relapse in the population exposed to primaquine in Yunnan Province. It makes up for the lack about risk assessment when primaquine is used in key malaria endemic areas in China, and provides the following ndings: (1) CYP2D6*10 allele accounts for the highest proportion of vivax malaria cases in Yunnan Province. (2) The mutation at c. 886 locus of CYP2D6 gene is most closely related to the relapse of vivax malaria (OR=2.167, P<0.05), and the frequency of CYP2D6*2 allele containing this mutation locus was signi cantly higher in SR (P<0.05). (3) The *4/*4 genotype with null CYP2D6 enzyme function were only found in the SR cases. These results suggest that it could exist the risk that the radical cure effect on vivax malaria cases may be reduced under using primaquine due to the damage of CYP2D6 enzyme activity in Yunnan Province.