Sample information and PCR amplification of human CYP2D6 gene
A total of 75 relapsed cases of P. vivax were enrolled, including 71 cases of one relapsed event, 2 cases of two relapse events, and two case of three relapsed events. The geographic information of 75 non-relapsed cases is listed in Table 1. The male-female ratio was 2.66:1, and the proportions of cases imported from Myanmar, Africa countries, Laos and Yunnan were 96.7% (145/150), 1.3% (2/150), 0.7% (1/150) and 1.3% (2/150), respectively. The odds ratio of relapsed was calculated in different age groups and genders. Statistical significance was determined in age groups of 5-20 years (OR = 0.380; 95% CI: 0.986~0.146) and 21-60 years (OR = 2.471:95% CI: 5.417~1.127). Thusly, the risk of relapsed of vivax malaria in patients aged 5-20 years was significantly lower, whereas the risk of relapsed in the patients aged 21-60 was 2.471 times higher than that of other groups. No significant correlation between other age groups or gender and the relapsed of vivax malaria was determined.
Table 1. Information of 150 vivax malaria cases for amplification of CYP2D6 gene exon1-9 fragments
|
Characteristic
|
Years
|
Total
|
2014
|
2015
|
2016
|
2017
|
2018
|
Number
|
18
|
27
|
20
|
10
|
75
|
150
|
1.Relapse times
|
|
|
|
|
|
|
0
|
0
|
0
|
0
|
0
|
75
|
75
|
1
|
17
|
25
|
19
|
10
|
0
|
71
|
2
|
0
|
1
|
1
|
0
|
0
|
2
|
3
|
1
|
1
|
0
|
0
|
0
|
2
|
2.Agea
|
|
|
|
|
|
|
0-4
|
0
|
0
|
1
|
2
|
1
|
4
|
5-20
|
0
|
2
|
4
|
1
|
16
|
23
|
21-60
|
17
|
25
|
15
|
6
|
51
|
114
|
above 60
|
1
|
0
|
0
|
1
|
7
|
9
|
3.Genderb
|
|
|
|
|
|
|
Male
|
16
|
22
|
14
|
6
|
51
|
109
|
Female
|
2
|
5
|
6
|
4
|
24
|
41
|
4.Infection sourcec
|
|
|
|
|
|
|
Myanmar
|
16
|
26
|
18
|
10
|
75
|
145
|
Africa
|
1
|
0
|
1
|
0
|
0
|
2
|
Laos
|
0
|
1
|
0
|
0
|
0
|
1
|
Yunnan indigenous
|
1
|
0
|
1
|
0
|
0
|
2
|
a Years old, b Numbers, c Identified by epidemiological investigation
|
PCR amplification of two segments of human CYP2D6 gene, including exons1-4 and exons5-9, was performed by using 156 blood samples of relapsed cases and 75 blood samples of non-relapsed cases. The amplification products showed clear bands at >2000bp, which were considered as the target bands (Fig 2.).
a b
Fig 2. PCR products of CYP2D6 gene fragment. a M: DNA marker; 1: blank control of PCR for the first round; 2: blank control of PCR for the second round; 3-6: The amplification products of exon 1-4 in CYP2D6 gene; b M: DNA marker; 1: blank control of PCR; 2-5: The amplification products of exon 5-9 in CYP2D6 gene;
Polymorphism analysis of CYP2D6 gene coding region
Locus polymorphism of CDS chain
The PCR amplification products of CYP2D6 gene from the relapsed cases and non-relapsed cases were trimmed, and the CDS chains containing the complete exon1-9 (total length = 1491bp) were obtained from all samples. A total of 150 CDS chains (Genbank accession number: MT339044-MT339193) were selected from the 75 relapsed case and 75 non-relapsed cases, and were then compared with wild-type sequence (NC: 000022.11). Base substitutions at 12 loci, such as c.31 and c.100, were determined (Table 2). The proportions of third-base and first-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. The mutation loci of the relapsed cases accounted for 91.7% (11/12), whereas the mutation loci of non-relapsed cases accounted for 66.7% (8/12) (Table 2)
Table 2. Polymorphism Comparison of Relapse cases and Non-relapse cases in the coding domain of CYP2D6 Genes from 1st aa to 497th aa
|
Relapse cases
|
Non-relapse cases
|
Loci
|
Codon changea
|
Amino acid change
|
|
Loci
|
Codon changea
|
Amino acid change
|
c.31
|
GTG>ATG
|
V11M
|
|
--
|
--
|
--
|
c.100
|
CCA>TCA
|
P34S
|
|
c.100
|
CCA>TCA
|
P34S
|
c.271
|
CTG>ATG
|
L91M
|
|
c.271
|
CTG>TTG
|
L91L
|
c.281
|
CAC>CGC
|
H94R
|
|
--
|
--
|
--
|
c.294
|
ACC>ACG
|
T98T
|
|
c.294
|
ACC>ACG
|
T98T
|
c.297
|
GCC>GCT
|
A99A
|
|
--
|
--
|
--
|
c.336
|
TTC>TTT
|
F112F
|
|
c.336
|
TTC>TTT
|
F112F
|
c.408
|
GTC>GTG
|
V136V
|
|
c.408
|
GTC>GTG
|
V136V
|
c.505
|
GGT>AGT
|
G169S
|
|
--
|
--
|
--
|
--
|
--
|
--
|
|
c.801
|
CCC>CCA
|
P267P
|
c.886
|
CGC>TGC
|
R296C
|
|
c.886
|
CGC>TGC
|
R296C
|
c.1457
|
AGC>ACC
|
S486T
|
|
c.1457
|
AGC>ACC
|
S486T
|
a DNA base highlighted in bold indicates the occurrence of SNP.
|
|
|
|
|
|
|
|
|
Polymorphism of haplotypes
A total of 24 CYP2D6 haplotypes (Hap_1~Hap_24) were clustered amongst the 150 vivax malaria samples; 17 haplotypes were observed in the sequences of relapsed patients (π=0.0015, He= 0.8191). 15 haplotypes were observed in the sequences of non-relapsed patients (π = 0.0014 and He = 0.8065). 8 haplotypes, including Hap_2, Hap_3, Hap_4, Hap_5, Hap_6, Hap_7, Hap_14 and Hap_17, were found in both relapsed and non-relapsed patients. Haplotype and frequency were shown in Fig. 3. Hap_3 was wild-type sequence, and the rest were mutant sequence. Hap_2 accounted for the largest proportion of 36.6% (55/150), followed by 15.3% for Hap_3 (23/150), 8.6% for Hap_4 (13/150), 6.6% for Hap_5 (10/150), 8.0% for Hap_6 (12/150), 6.6% for Hap_7 (10/150) and 4.6% for Hap_21 (7/150). The proportions of Hap_1, Hap_8, Hap_9, Hap_10, Hap_12, Hap_13, Hap_15, Hap_16, Hap_18, Hap_19, Hap_20, Hap_22, Hap_23, and Hap_24 were as low as 0.7% (1/150). The proportions of Hap_11, Hap_14 and Hap_17 haplotypes were 1.3% (2/150).
Fig 3. 9 exons are indicated by numbered boxes with DNA polymorphisms indicated on top. Predicted amino acid
changes are indicated below; no change is synonymous mutation. R: Relapsed cases; N: Non-relapsed cases.
Haplotypes of CYP2D6 gene coding region and their association with relapse
We selected the Hap_2 ~ Hap_6 haplotype sequences which had more than one CDS and were found in relapsed and non-relapsed patients to calculate the odds ratio (Table3). Among them, only two haplotypes, Hap_4 (OR=0.159; 95% CI: 0.746 ~ 0.034) and Hap_6 (OR=5.615; 95% CI: 26.577 ~ 1.186) showed statistical significance (P ˂0.05). Hence, Hap_4 indicates reduced risk of relapsing. By contrast, Hap_6 suggests 5.615-fold higher risk of relapse than other haplotypes. The OR values of Hap_2, Hap_3, Hap_5 and Hap_7 showed no statistically significant difference (P>0.05) (Table 3).
Table 3 Analysis of the relationship between haplotypes and relapse
|
Hap
|
Relapse cases(n)
|
Non-relapse cases(n)
|
OR
|
95%CI
|
P
|
Upper limit
|
Low limit
|
Hap_2
|
28
|
27
|
1.059
|
2.058
|
0.545
|
0.865a(NS)
|
Hap_3
|
8
|
15
|
0.478
|
1.206
|
1.189
|
0.113a(NS)
|
Hap_4
|
2
|
11
|
0.159
|
0.746
|
0.034
|
0.009a(S)
|
Hap_5
|
8
|
2
|
4.358
|
21.258
|
0.894
|
0.050a(NS)
|
Hap_6
|
10
|
2
|
5.615
|
26.577
|
1.186
|
0.016a(S)
|
Hap_7
|
7
|
3
|
2.471
|
9.944
|
0.614
|
0.190a(NS)
|
n: number of cases; NS: not significant; S: significant; a: Chi-square test
|
Enzyme activity prediction of significant haplotypes and their diploids
The enzyme activity of Hap_6 and Hap_4 genotypes, which were validated to implicate in recurrence of vivax malaria, was predicted. As for the 12 CDS chains in Hap_6 (Table 4) from 10 relapsed patients and 2 non-relapsed patients, the diploids in relapsed cases presented 4 genotypes, including *1/*2, *2/*2, *2/*41 and *1/*41. Of the c.408, c.886, c.1023+36 and c.1457 mutations, only c.886 and c.1023+36 were diploid alleles. Whether the diploid of c.886 showed mutant heterozygosity (C/T) or mutant homozygosity (T/T), as long as it co-existed with mutant heterozygosity of c.1023+36, the genotype score of CYP2D6 couldn’t reach 2.0. In this case, the score was 1.5 and the enzyme activity of CYP2D6 belonged to “NM” (Table 4). The genotypes of *1/*2 and *1/*41 were found in the non-relapsed cases. However, *1/*41 was also determined in the relapsed cases and reached 1.5 in genotype score. Hence its enzyme activity was predicted as "NM" level. Moreover, *1/*2 genotype which showed mutant heterozygous (C/T) only at c.886 reached the peak genotype score of 2.0 and was therefore predicted as “NM” level (Table 4).
Table 4 Prediction of enzyme activities of Hap-6 and Hap-4 genotypes
|
A. Hap_6
|
Genotypes
|
Relapse times
|
No. of samples
|
Mutation loci and their diploid
|
|
Caudle's method
|
c.408
|
c.886
|
c.1023+36a
|
c.1457
|
|
Score of diploid
|
Degree
|
NC_000022.11
|
--
|
--
|
G/G
|
C/C
|
G/G
|
G/G
|
|
2
|
NM
|
*1/*2
|
0
|
1
|
C/C
|
C/T
|
G/G
|
C/C
|
|
2
|
NM
|
*1/*41
|
0
|
1
|
C/C
|
C/T
|
G/A
|
C/C
|
|
1.5
|
NM
|
*1/*2
|
1
|
1
|
C/C
|
C/T
|
G/G
|
C/C
|
|
2
|
NM
|
*2/*2
|
1
|
5
|
C/C
|
T/T
|
G/G
|
C/C
|
|
2
|
NM
|
*2/*41
|
1
|
3
|
C/C
|
T/T
|
G/A
|
C/C
|
|
1.5
|
NM
|
*1/*41
|
1
|
1
|
C/C
|
C/T
|
G/A
|
C/C
|
|
1.5
|
NM
|
B. Hap_4
|
Genotypes
|
Relapse times
|
No. of samples
|
Mutation loci and their diploid
|
|
Caudle's method
|
c.100
|
c.408
|
--
|
c.1457
|
|
Score of diploid
|
Degree
|
NC_000022.11
|
--
|
--
|
C/C
|
G/G
|
--
|
G/G
|
|
2
|
NM
|
*10/*10
|
0
|
2
|
T/T
|
C/C
|
--
|
C/C
|
|
0.5
|
IM
|
*1/*10-1
|
0
|
3
|
C/T
|
C/C
|
--
|
G/C
|
|
1.25
|
NM
|
*1/*10-2
|
0
|
6
|
C/T
|
C/C
|
--
|
C/C
|
|
1.25
|
NM
|
*1/*10-2
|
1
|
2
|
C/T
|
C/C
|
--
|
C/C
|
|
1.25
|
NM
|
a: Intronic position c.1023+36 (Intron 2988) has been considered as defining allele *41 because the predictivity of this mutation for impaired enzyme function has been demonstrated recently
|
13 CDS chains in Hap_4 was found in 2 relapsed and 11 non-relapsed patients (Table 4). In the relapsed cases, *1/*10-2 was the only diploid genotype. Mutations of *1/*10-2 at loci c.100, c.408 and c.1457 were heterozygosity (C/T), homozygosity (C/C) and homozygosity (C/C), respectively. Such pattern resembled the genotype of 54.6% (6/11) of the non-relapsed cases, and reached the genotype score of 1.25, which indicates "NM" level for CYP2D6 enzyme activity (Table 4). The remaining 5 non-relapsed cases were classified as genotype *10/*10 and *1/*10-1. The diploids of genotypes *10/*10 at c.100, c.408 and c.1457 loci were mutant homozygous (T/T), (C/C) and (C/C), respectively, reaching the lowest genotype score of 0.5. It was predicted that the enzyme activity of CYP2D6 belongs to "IM" level (Table 4). In 3 cases of *1/*10-1 genotype, mutation heterozygosity (C/T) was found at c.100 loci; its genotype score of 1.25 suggests "NM" level of CYP2D6 enzyme activity (Table 4).