Lack of quadruple and quintuple mutant alleles associated with sulfadoxine-pyrimethamine resistance in Plasmodium vivax isolates from Brazilian endemic areas.

BACKGROUND AND
OBJECTIVE: Brazil is responsible for a large number of Plasmodium vivax cases in America. Given the emergence of P. vivax parasites resistant to chloroquine and the effectiveness of antifolates in vivax malaria treatment together with a correlation between mutations in P. vivax dhfr and dhps genes and SP treatment failure, the point mutations in these genes were investigated.
METHODS: Blood samples from 54 patients experiencing vivax malaria symptomatic episodes in the Amazonian Region were investigated. Genomic DNA was extracted using a DNA extraction kit (QIAGENTM). Nested polymerase chain reaction (PCR) amplification was carried out followed by Sanger sequencing to detect single nucleotide polymorphisms (SNPs).
FINDINGS: All tested isolates showed non-synonymous mutations in pvdhfr gene: 117N (54/54, 100%) and 58R (25/54, 46%). Double mutant allele 58R/117N (FRTNI, 28%) was the most frequent followed by triple mutant alleles (58R/117N/173L, FRTNL, 11%; 58R/61M/117N, FRMNI, 5% 117N/173L, FSTNL, 4%) and quadruple mutant allele (58R/61M/117N/173L, FRMNL, 2%). A single mutation was observed at codon C383G in pvdhps gene (SGKAV, 48%).
CONCLUSION: No evidence of molecular signatures associated with P. vivax resistance to SP was observed in the Brazilian samples.

Plasmodium vivax is the most geographically widespread human malaria parasite. It is prevalent mainly outside Africa including Asia, South and Central America, and the Middle East. In the Americas, the burden of vivax malaria mostly affects Venezuela and Brazil. In Brazil, malaria transmission occurs almost entirely (> 99% of the registered cases) within the northern Brazilian Amazon Region where both P. falciparum and P. vivax infections co-exist. In this area, P. vivax is the predominant species, responsible for 89% of 194,409 malaria cases reported in 2017. Nowadays, falciparum malaria is treated with a 3-day fixed Artesunate+Mefloquine combination, according to Brazilian National Malaria Program guidelines, and a radical cure for P. vivax malaria is achieved with 25 mg/kg of CQ base for three days (maximum adult dose, 1.5 g for three days), combined with a short hypnozoitocidal regimen of 0.5 mg/kg/day of primaquine (PQ) base (maximum daily dose, 30 mg/day) for seven days in patients that weighed below 70 kg. As subtherapeutic PQ doses may lead to relapse in overweight patients, weight-adjusted PQ doses are now recommended in Brazil for patients over 70 kg.

P. falciparum resistance to chloroquine (CQ) observed in the 1980s greatly contributed to the emergence of falciparum malaria outbreaks across Amazon. P. vivax resistance to CQ occurred later in 1989 in Papua New Guinea and CQ monotherapy was ineffective. Following this seminal observation, numerous cases of CQ resistance were reported in Southeast Asia and South America, , thus complicating the current international efforts for malaria control and elimination, and signalling the need for alternative drugs for vivax malaria treatment.

Antifolates, most notably sulfadoxine-pyrimethamine (SP), have been used as anti-malaria for P. falciparum treatment throughout the world because this combination is inexpensive, relatively safe, and requires only a single dose course treatment. SP had been available in Brazil since 1960s to treat CQ-resistant falciparum malaria but SP-resistant P. falciparum isolates appeared since 1990; SP is not used for malaria therapy in Brazil. Although resistant to antifolates, P. falciparum treatment has been well documented in many parts of the world, and P. vivax chemoresistance to SP is scarcely studied.

Sulfadoxine and pyrimethamine are competitive inhibitors of dihydropteroate synthase (dhps) and dihydrofolate reductase (dhfr), the two major proteins involved in folate biosynthesis pathway . Polymorphisms in these two genes are the major factors associated with SP resistance.

Data on pvdhfr and pvdhps genotypes are available for many Southeast Asian countries. Such reports remain limited for some P. vivax endemic areas, notably South America. In Brazil, only one study characterising polymorphisms in pvdhfr gene was documented8) and there is no report on the frequency of single nucleotide polymorphism (SNP) in dhps gene in P. vivax clinical isolates from Brazilian endemic areas.

Given the emergence of P. vivax CQ resistant parasites and the effectiveness of antifolates in malaria vivax treatment together with a strong correlation between mutations in P. vivax dhfr and dhps genes and SP treatment failure, the present paper reports an investigation on the pattern of point mutations in pvdhfr and pvdhps genes in Brazilian isolates.

MATERIALS AND METHODS

Parasites isolates and DNA extraction - Blood samples from Amazon Region (Acre, Amapá, Amazonas, Rondônia and Pará) were collected from 54 patients presenting with vivax malaria from 2010 to 2016 at the Laboratório de Doenças Febris Agudas, INI-IPEC, Fiocruz, the Reference Clinical Laboratory for Malaria in the Extra-Amazon to the Brazilian Ministry of Health. All the clinical isolates were diagnosed as single P. vivax infections by light microscopic examination of Giemsa’s solution-stained blood smears and by P. vivax cysteine-proteinase target gene polymerase chain reaction (PCR). The parasitaemia ranged from 960 to 19160 parasites/µL. All malaria patients presented with clinical signs and/or symptoms of uncomplicated malaria, such as fever, headache, and chills, and the baseline characteristics were similar. No significant difference in parasitaemia was observed among the studied Brazilian localities and all the Brazilian endemic states were hypoendemic malaria areas.

Genomic DNA was extracted using a commercially available DNA extraction kit (QIAGENTM, Frankfurt, Germany), following the manufacturer’s instructions. This study was performed according to the protocols previously approved by the Ethical Research Committees of Fiocruz (32839013.6.00005248). Patients were treated with CQ plus PQ, according to the Brazilian Ministry of Health recommendation for uncomplicated vivax malaria treatment and were followed up to 42 days. No treatment failure was detected during this period.

Nested PCR and electrophoresis - Nested PCR amplification of pvdhfr and pvdhps were carried out as described previously. Ten point mutations were investigated: F57L/I, S58R, T61M, S117T/N and I173F/L for pvdhfr, and S382A, C383G, K512M/T/E, A553G and V585G for pvdhps. PCR products were analysed by ethidium bromide-stained agarose-gel (2%) electrophoresis.

DNA sequencing and SNPs detection - The 632 bp and 767 bp fragments generated by amplification of pvdhfr and pvdhps, respectively, were extracted and purified from gel using the Wizard® SV Gel and PCR Clean-Up System (Promega, Madison, WI, USA) commercial kits. Briefly, the amplified fragments were sequenced using BigDye Terminator cycle sequencing ready reaction version 3.1 and ABI Prism DNA analyser 3730 (Applied Biosystems) at the Genomic Platform/PDTIS/Fiocruz. The direct DNA sequencing from PCR products were compared with the reference Sal I sequence of pvdhfr (GenBank X98123) and pvdhps (GenBank AY186730.1). Forward and reverse sequences were analysed using the free software, Bioedit Sequence Alignment Editor version 7.2.5. PCRs and DNA sequencing were randomly repeated to check possible sequence errors introduced during these stages.

RESULTS

All the 54 isolates sequenced for pvdhfr gene showed non-synonymous mutations: 117N (54/54; 100%) and 58R (25/54; 46%) mutant alleles were more frequent, while 173L (9/54; 17%) and 61M (4/54; 7%) were detected at lower frequencies. Mutation at position 57L was not found (Table I). The most common single mutant allele was 117N (27/54; 50%). This single mutant was more frequent in Acre (10/15; 66%), Amazonas (11/23; 52%) and Pará states (4/8; 50%), compared to Rondônia state (1/7; 14%), where double 58R+117N mutant was dominant (Table II). Independent of the year collection, Amazonas state showed the highest number of pvdhfr gene mutations (23/54; 42,5%), followed by Acre (15/54; 27,7%, Para (8/54; 15%) and Rondônia (7/54, 13%) (Tables III-VI). Apparently in 2011, Acre presented more pvdhfr gene mutations (7/15; 47%) than Amazonas (2/23; 8,6%) (Tables III-IV), but this difference could be related to the smaller number of Amazonas samples collected in 2011, because when percentages are compared instead of figures, 100% of Amazonas (2/2) and Acre samples (7/7) presented mutations in 2011.

TABLE I

Plasmodium vivax dhfr and dhps amino acid changes in 54 P. vivax isolates from Brazilian endemic areas

Gene	SNPs	Prevalence N (%)
dhfr	58R	25 (46)
	61M	4 (7)
	117N	54 (100)
	173L	9 (17)
dhps	383G	26 (48)

SNPs: single nucleotide polymorphisms.

TABLE II

Number of alleles in dhfr and dhps genes observed among 54 Brazilian Plasmodium vivax isolates, according to sampling location

Gene	SNPs	Amazonas (n = 23)	Acre (n = 15)	Amapá (n = 1)	Pará (n = 8)	Rondônia (n = 7)
dhfr	117N	12	10	-	4	1
	58R/117N	8	2	-	-	5
	117N/ 173L	-	-	-	1	1
	58R/117N/173L	1	3	1	1	-
	58R/61M/117N	1	-	-	2	-
	58R/61M/117N/173L	1	-	-	-	-
dhps	383G	13	8	-	5	-

SNPs: single nucleotide polymorphisms.

TABLE III

Number of alleles in dhfr and dhps genes observed among 23 Plasmodium vivax isolates from Amazônia, according to year of blood collection

Genotype	Mutation codon	2010	2011	2012	2013	2015	2016	Total
Dhfr	117N	5	-	1	-	4	2	12
	58R/117N	1	2	1	-	3	1	8
	58R/117N/173L	-	-	-	1	-	-	1
	58R/61M/117N	-	-	1	-	-	-	1
	58R/61M/117N/173L	-	-	1	-	-	-	1
	TOTAL	6	2	4	1	7	3	23
Dhps	383G	5	-	4	1	2	1	13

TABLE VI

Number of alleles in dhfr gene observed among seven Plasmodium vivax isolates from Rondônia, according to year of blood collection

Genotype	Mutation codon	2010	2011	2014	Total
Dhfr	117N	1	-	-	1
	58R/117N	-	2	3	5
	117N/173L	-	-	1	1
	TOTAL	1	2	4	7

TABLE IV

Number of alleles in dhfr and dhps genes observed among 15 Plasmodium vivax isolates from Acre, according to year of blood collection

Genotype	Mutation codon	2011	2013	2014	2015	2016	Total
Dhfr	117N	4	-	-	4	2	10
	58R/117N	1	-	1	-	-	2
	58R/117N/173L	2	1	-	-	-	3
	TOTAL	7	1	1	4	2	15
Dhps	383G	5	-	-	2	1	8

The double 58R/117N allele (FRTNI, 28%) was the most common allele, contrasting with the frequencies of other dhfr double, triple, or quadruple mutant alleles, with lower frequencies: 58R/117N/173L (FRTNL, 11%), 58R/61M/117N (FRMNI, 5%), 117N/173L (FSTNL, 4%), and 58R/61M/117N/173L (FRMNL, 2%). In all localities, wild-type pvdhfr (FSTSI) was not observed (Table VII). The 58R/117N double mutant allele was detected in Acre (2/15; 13%), Rondônia (5/7; 71%), and Amazonas (8/23; 35%) while the 117N+173L only in Pará and Rondônia. The triple mutant allele 58R/117N/173L was found in all localities, except Rondônia, and the quadruple mutant 58R/61M/117N/173L was observed only in one isolate collected from Amazonas state (1/23; 4%) (Table VIII). The frequencies of double, triple, or quadruple mutants were not related to the year of collection (Tables III-VI).

TABLE VII

Deduced dhfr and dhps haplotype profiles in 54 Plasmodium vivax isolates from Brazilian endemic areas

Gene	Haplotypes	N	%
dhfr	FSTNI	27	50
	FRTNI	15	28
	FSTNL	2	4
	FRTNL	6	11
	FRMNI	3	5
	FRMNL	1	2
dhps	SGKAV	26	48
	SCKAA (wild type)	28	52

Bold letters: mutated codons; F: codon 57; S: codon 58; T: codon 61; S: codon 117; I: codon 173.

TABLE VIII

Prevalence of Plasmodium vivax dhfr and dhps amino acid changes in 54 P. vivax isolates from Brazilian endemic areas

Gene	Mutants	SNPs	Prevalence N (%)
dhfr	Single	S117N	27 (50)
	Double	S58R/S117N	15 (28)
		S117N/ I173L	2 (4)
	Triple	S58R/S117N/I173L	6 (11)
		S58R/T61M/S117N	3 (5)
	Quadruple	S58R/T61M/S117N/ I173L	1(2)
dhps	Single	C383G	26 (48)

SNP: single nucleotide polymorphisms.

Concerning pvdhps gene in 26 out of 54 (48%) isolates only a single mutation at codon C383G was detected. No other mutations, including 382A, 512M, 553G, and 585C, were found. The wild-type SCKAV (52%) and single haplotype SGKAV (48%) were observed at similar frequencies. The single mutant 383G was observed in isolates from Amazonas (13/23, 56%), Acre (8/15, 53%) and Pará (5/8, 62%) but not in isolates from Rondônia state (0/7) (Table II). Once again frequencies of pvdhps gene mutations were not related to the year of collection (Tables III-VI).

Combining pvdhfr and pvdhps alleles, only one haplotype (FRTNI for pvdhfr and SGKAV for pvdhps) was seen in three of the four study sites with a higher frequency in Amazonas state (where one pvdhfr quadruple mutant was detected) (Table IX). No pvdhfr or pvdhps quadruple or quintuple mutant haplotype, which might result in poor clinical response against antifolate drugs, was detected in any of the Brazilian localities investigated.

TABLE IX

Percentage of double mutant dhfr / single mutant dhps in 54 Plasmodium vivax samples according to Brazilian states

Amazônia	Acre	Pará
NP (%)	NP (%)	NP (%)
13 (24)	8 (15)	5 (9)

NP: number of positive.

DISCUSSION

Mutations in pvdhfr and pvdhps genes have been found to be associated with antifolate drug resistance. Both in vivo and in vitro assays suggested that these molecular markers may provide information about the trends of SP resistance in P. vivax. Here, we investigated SP resistance in vivax isolates by seeking specific point mutations in pvdhfr and pvdhps genes.

It has been postulated that pvdhfr 117N mutation might occur first, followed by S58R mutation. In this study, pvdhfr S117N was detected in all isolates followed by 58R (74%), 173L (17%), and 61M (7%) polymorphisms, supporting that S117N mutation is the first step in drug selection process. These data are similar to other observations done in areas where P. falciparum and P. vivax parasites co-exist. ,

The predominance of S117N followed by the double mutant 58R/117N (28%) was also analogous to those reported in India, Afghanistan, China, Nepal, Thailand, Colombia, , French Guiana and Brazil. The triple 58R/117N/173L pvdhfr mutant, not seen in P. vivax samples from Southeast Asian, where non-synonymous mutation in codon 173 comprises the change of I by F generating the 173F allele, was here detected in Amazonas, Acre, Amapá and Pará states and also in P. vivax parasites from French Guiana , and Amazonas, Brazil. Conversely, the non-synonymous mutation at position F57L not recorded in this study was exclusively reported in Southeast Asian samples; findings that could reflect different drug pressure history and selective processes in the old and new worlds. In fact, the genetic similarity of 173L SNP recorded for P. vivax parasites from two neighbouring South-American countries Brazil and French Guiana, reinforce the possible existence of geographic subdivision of differentP. vivax parasites in samples from the old and new worlds.

Concerning the pvdhps gene, previous data indicated that mutations were mainly detected at codons A383G and A553G , , , and suggested that these mutations alone could be responsible for reduced sensitivity to sulfa and sulfones. , In the present work, the wild-type (52%) and the mutated codon 383G (48%) were detected at similar frequencies among P. vivax isolates, similar to reports from Thai-Cambodian (53%), Thai-Myanmar border (47%) and Indonesia (50%). Whereas, in a Colombian study investigating polymorphisms in pvdhps, the wild-type was the most frequently detected (71.6%); the same was true in India (79%) and also in Thai - Cambodian border (74%). Therefore, the pvdhps wild-type allele seems to be common in malaria endemic areas of the world, probably due to a low SP drug selection in the sympatric P. vivax populations of these countries. However, in Brazil, for example, SP or its analogues have been used for fever and antimicrobial therapy and, in this way, there continues to be a lengthy selection pressure for SP-resistant strains of P. vivax resulting to low frequencies of wild-type pvdhps parasites.

Amazonas state recorded the highest number of pvdhfr and pvdhps mutations. This finding could not be attributed to differences of antimalarial drug usage in Brazilian states because the malaria treatment in Brazil is the same all over the country. Besides that, SP has never been recommended for vivax malaria treatment and SP has been excluded from P. falciparum treatment since 1989. Thus, it is more reasonable suppose that more mutations were found in Amazonas due to the highest number of samples examined from this locality, as only one sample from Amazonas was from a border area of the Amazon with Acre - the second state that showed the greatest number of mutations. A study with a representative number of Amazonian state cases may help answer this question.

In conclusion, we found no molecular strong evidence of P. vivax SP resistance in recently collected Brazilian samples. As mutations in P. vivax dhps and dhfr genes provide a valuable tool for epidemiological surveillance of SP resistance, the prevalence of point mutations on these genetic markers of SP resistance should be assessed for providing information for future treatment policy with alternative antifolate drugs because of the appearance and dispersion of CQ resistance in malaria endemic areas.

TABLE V

Number of alleles in dhfr and dhps genes observed among eight Plasmodium vivax isolates from Pará, according to year of blood collection

Genotype	Mutation codon	2010	2011	2013	2015	2016	Total
Dhfr	117N	1	-	-	1	2	4
	117N/173L	-	1	-	-	-	1
	58R/117N/173L	-	-	1	-	-	1
	58R/61M/117N	-	-	-	2	-	2
	TOTAL	1	1	1	3	2	8
Dhps	383G	1	-	-	1	2	5