Genetic variations in histidine-rich protein 2 and histidine-rich protein 3 of Myanmar Plasmodium falciparum isolates

Background Malaria rapid diagnostic tests (RDTs) are precious tools to diagnose malaria. Most RDTs used currently are based on the detection of Plasmodium falciparum histidine-rich protein 2 (PfHRP2) in a patient’s blood. However, concern has been raised in recent years that deletion of pfhrp2 in the parasite could affect the accuracy of PfHRP2-based RDTs. In addition, genetic variation in pfhrp2 might influence the accuracy and sensitivity of RDTs. In this study, the genetic variation in pfhrp2 and pfhrp3 in Myanmar P. falciparum isolates was analysed. Methods Blood samples were collected from malaria patients who were infected with P. falciparum in Mandalay, Naung Cho, Tha Beik Kyin, and Pyin Oo Lwin, Upper Myanmar between 2013 and 2015. The pfhrp2 and pfhrp3 were amplified by nested polymerase chain reaction (PCR), cloned and sequenced. Genetic variation in Myanmar pfhrp2 and pfhrp3 was analysed using the DNASTAR program. Comparative analysis of Myanmar and global pfhrp2 and pfhrp3 isolates was also performed. Results One-hundred and two pfhrp2 and 89 pfhrp3 were amplified from 105 blood samples, of which 84 pfhrp2 and 56 pfhrp3 sequences were obtained successfully. Myanmar pfhrp2 and pfhrp3 showed high levels of genetic variation with different arrangements of distinct repeat types, which further classified Myanmar pfhrp2 and pfhrp3 into 76 and 47 haplotypes, respectively. Novel amino acid changes were also found in Myanmar pfhrp2 and pfhrp3, but their frequencies were very low. Similar structural organization was shared by Myanmar and global pfhrp2 and pfhrp3, and differences in frequencies of repeat types and lengths were also observed between and among global isolates. Conclusion Length polymorphisms and amino acid substitutions generated extensive genetic variation in Myanmar pfhrp2 and pfhrp3. Comparative analysis revealed that global pfhrp2 and pfhrp3 share similar structural features, as well as extensive length polymorphisms and distinct organizations of repeat types. These results provide a better understanding of the genetic structure of pfhrp2 and pfhrp3 in global P. falciparum populations and suggest useful information to develop RDTs with improved quality.

Background Myanmar has the majority of malaria cases and deaths in Southeast Asia, but the incidence of the disease has declined dramatically during the last decade and is progressing steadily towards elimination [1]. The annual number of malaria cases in the country has dropped from approximately 700,000 in 2010 to about 85,000 in 2017 [1]. Microscopic examination of a blood smear is the primary diagnostic tool for malaria in Myanmar, but malaria rapid diagnostic tests (RDTs) have become a valuable alternative for use in remote areas where microscopy may not be feasible or where microscopy results would not be available immediately. RDTs offer a practical alternative to microscopy because they do not require a laboratory or special equipment, are simple to use, and provide reliable results in a short time [2,3]. RDTs have been introduced as reliable diagnostic tools in many malariaendemic areas, including Myanmar.
Malaria RDTs are designed to detect one or more Plasmodium antigens in a patient's blood by using specific monoclonal antibodies. Several antigens of Plasmodium falciparum have been utilized in RDTs for malaria detection, including histidine-rich protein 2 (PfHRP2), lactate dehydrogenase (PfLDH) and aldolase [4]. Among these antigens, PfHRP2 is the most widely employed in commercially available malaria RDTs at present, because of its abundant expression in the asexual blood stage of P. falciparum [5][6][7], structural stability [8] and high specificity recognized by multiple antibodies [9][10][11]. PfHRP2 is a protein encoded by pfhrp2, which is located in the subtelomeric region of chromosome 8. This P. falciparumspecific protein is expressed abundantly in the infected erythrocyte surface of the blood stage and early gametocyte stage [12,13]. Some of the PfHRP2-based RDTs can cross-react with a structural homologue, P. falciparum histidine-rich protein 3 (PfHRP3) that is encoded by pfhrp3 and shares high sequence identity and epitope similarity with PfHRP2 [13,14]. The pfhrp2 and pfhrp3 genes are predicted to be derived from a common ancestral gene. As such, these near-duplicate genes may compensate each other in function [7]. However, it has been reported recently that P. falciparum field isolates in some parts of malaria-endemic regions lack pfhrp2 [15][16][17][18][19][20]. Deletion of pfhrp2 could affect the accuracy of PfHRP2based RDTs and lead to false-negative results followed by inappropriate treatment, which in turn causes negative impact for effective control and elimination of malaria. Furthermore, co-deletion of pfhrp2 and pfhrp3 has also been identified [8,21]. In addition to the density of the parasite or the lack of PfHRP2 expression, it has also been suggested that genetic diversity in pfhrp2 and pfhrp3 could affect the sensitivity of PfHRP-based RDTs [22,23]. Therefore, monitoring parasite factors that can affect performance of RDT-based diagnosis is important.
In this study, the genetic variation in pfhrp2 and pfhrp3 of Myanmar P. falciparum isolates was analysed. The diversity of the two genes from global P. falciparum isolates was also investigated comparatively to gain in-depth understanding of the genetic diversity and population structure of global pfhrp2 and pfhrp3.

Study sites and blood sample collection
A total 105 blood samples from malaria patients infected with P. falciparum were collected during a previous study conducted in Myanmar between 2013 and 2015 [24]. The patients were detected during regional malaria surveys, which were conducted in the regions of Naung Cho, Pyin Oo Lwin, Tha Beik Kyin, and Mandalay in Upper Myanmar (Additional file 1: Fig. S1). Malaria infection was diagnosed by microscopic analysis of thin and thick blood smears. Finger-prick blood was taken from P. falciparum-infected patients and spotted in Whatman 3MM filter (GE Healthcare, Maidstone, UK) for confirmation by polymerase chain reaction (PCR) targeting 18S ribosomal RNA (rRNA) gene [24]. Informed consent was obtained from all patients before blood collection. The study protocol was approved by either the Ethics Committee of the Ministry of Health, Myanmar (97/Ethics 2015) or the Biomedical Research Ethics Review Board of Inha University School of Medicine, Republic of Korea (INHA 15-013).

Genomic DNA extraction and amplifications of pfhrp2 and pfhrp3
Genomic DNA was extracted from the dried blood spots using the QIAamp DNA Blood Kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. The primers specific for pfhrp2 and pfhrp3 were designed and used (Additional file 2: Table S1). Both genes were amplified by nested PCR methods. Each PCR was done with thermal cycling conditions: 94 °C for 5 min, and 35 cycles of 94 °C for 1 min, 53 °C for 1 min, and 72 °C for 1.5 min, followed by the final extension at 72 °C for 10 min. In order to minimize the nucleotide mis-incorporation during amplification, Ex Taq DNA polymerase (Takara, Otsu, Japan), which has a proof-reading activity, was used in all PCR steps. Each PCR product was resolved on 1.5% agarose gel, extracted from the gel, and cloned into T&A cloning vector (Real Biotech Corporation, Banqiao, Taiwan). Each ligation mixture was transformed into Escherichia coli DH5α competent cells. Colony PCR was performed to select the positive clones with appropriate inserts. The nucleotide sequences of the cloned pfhrp2 and pfhrp3 were analysed by automatic DNA sequencing with M13 forward and M13 reverse primers by the Sanger method. Plasmids from at least two independent clones from each transformation mixture were sequenced to verify the sequence accuracy. The nucleotide sequences of Myanmar pfhrp2 and pfhrp3 analysed in this study have been deposited in the GenBank database under the accession numbers KX138275-KX138311, MG417056-MG417080, and MT591418-MT591439 for pfhrp2 and KX138312-KX138340 and MT591440-MT591466 for pfhrp3.

Sequence analyses of Myanmar and global pfhrp2 and pfhrp3
The nucleotide and deduced amino acid sequences of Myanmar pfhrp2 and pfhrp3 were analysed using Edit-Seq and SeqMan in the DNASTAR package (DNASTAR, Madison, WI, USA). Genetic variations of pfhrp2 and pfhrp3 in global P. falciparum isolates were also analysed. The pfhrp2 and pfhrp3 sequences deposited in public database were used in this study. The pfhrp2 sequences analysed in this study were from China, India, Sri Lanka, Thailand, Philippines, Cambodia, Vietnam, Central African Republic, Ghana, Haiti, Kenya, Madagascar, Nigeria, Tanzania, French Guinea, Brazil, Honduras, Papua New Guinea, Solomon Islands, and East Timor (Additional file 3: Table S2). For pfhrp3, the sequences from Cambodia, India, Philippines, Kenya, Madagascar, Nigeria, Peru, Colombia, Papua New Guinea, and Solomon Islands were analysed (Additional file 4: Table S3).

Amplification of pfhrp2 and pfhrp3 in Myanmar
Plasmodium falciparum isolates PCR amplification of pfhrp2 and pfhrp3 from 105 P. falciparum-infected blood samples resulted in successful amplification of 102 pfhrp2 and 89 pfhrp3. The approximate sizes of amplified products were highly variable, ranging 100-1000 bp for pfhrp2 and 50-600 bp for pfhrp3. Of these, 84 pfhrp2 and 56 pfhrp3 PCR products were cloned successfully and sequenced for further assessments. The remaining 19 pfhrp2 and 33 pfhrp3 PCR products were excluded from this study because the quality of the amplicons was not adequate for further analysis, despite repeated attempts.

Polymorphic character of Myanmar pfhrp3
Forty-seven distinct haplotypes of pfhrp3 were observed in 56 Myanmar P. falciparum isolates (Fig. 2). Eleven different types of repeat were observed in Myanmar pfhrp3 and each haplotype was constructed with different numbers of the repeats ranging from 2 to 25. Structural features of Myanmar pfhrp3 haplotypes were highly diverse, but all haplotypes shared similar patterns. Most haplotypes started with type 1 repeat (AHHAHHVAD) and terminated with type 4 repeat (AHH). Two exceptions were haplotypes 38 and 46. Haplotype 38 stared with type 1 repeat but finished with type 15 repeat (AHHAH-HAAN). Haplotype 46 began with type 7 repeat (AHHAAD) and terminated with type 4 repeat (AHH). Non-repeat (NR) regions were scattered randomly in the sequences of most haplotypes. The length variation in Myanmar pfhrp3 was caused mainly by repeating numbers of type 16 (AHHAAN), type 17 (AHHADG), or type 18 (AHHDD).

Prevalence of repeat types in Myanmar pfhrp2 and pfhrp3
Overall prevalence of each type of repeat differed in Myanmar pfhrp2 and pfhrp3. Type 2 repeat was the most prevalent and showed high numbers of duplicates 18, and 20 repeats were found commonly in most haplotypes, while the proportions of other types of repeats were variable (Fig. 3). Types 1, 2, 4, and 7 repeats were found in both Myanmar pfhrp2 and pfhrp3. Types 3, 5, 6, 8, 10, 12, 13, and 14 were identified only in Myanmar pfhrp2 and types 11,15,16,17,18,20, and 28 were found only in Myanmar pfhrp3.

Genetic polymorphisms of pfhrp2 and pfhrp3 in global isolates
Comparative analysis of repeat types revealed that types 1, 2, 3, 5, 6, 7, 10, and 12 were found commonly in all global pfhrp2 population ( Table 2). The frequencies of types 1, 2, 6, and 12 were especially high in global pfhrp2. The frequencies of types 3, 5, 7, and 10 differed by country. Type 8 repeat also showed high frequencies in global pfhrp2, but it was not identified in Brazil pfhrp2. Types 4, 13 and 14 were observed at low frequencies in pfhrp2 from some countries, but types 15 and type 19 were found only in pfhrp2 from China, India, Kenya, or the Solomon Islands. A similar variety of frequencies of repeat types was found in global pfhrp3 populations (Table 3). Types 1, 4, 7, 17, 18, and 20 were detected at high frequencies in all global pfhrp3. Type 16 also showed high frequencies in all global pfhrp3, except for Myanmar. Type 15 was found in nearly all global pfhrp3 at relatively high frequencies, but it was not detected in pfhrp3 from Peru, Papua New Guinea and the Solomon Islands. Types 2, 28 and 29 were identified at low frequencies in only several countries, including Myanmar, India, Philippines, Kenya, Madagascar, Peru, or Papua New Guinea. Length variation in pfhrp2 and pfhrp3 was also found in the global P. falciparum population (Fig. 4)

Discussion
Malaria RDTs provide a simple, rapid and reasonably reliable diagnosis of malaria, which could aid proper treatment of the disease and offer significant benefits for malaria control and elimination. Given these advantages, global use of RDTs has been increasing rapidly in many malaria-endemic regions, and having RDTs with high specificity and sensitivity against global isolates is very important. However, genetic variation in the target antigens employed in the RDTs could affect the performance, especially their sensitivity to detect low-density malaria infections [9,23,25]. Understanding the genetic diversity and structure organization of the pfhrp2 and pfhrp3 in global P. falciparum isolates is important because most commercially available P. falciparum RDTs target PfHRP2 expressed solely by the parasite [26]. This study is the first report on genetic analysis of pfhrp2 and pfhrp3 in Myanmar P. falciparum isolates. Sequence analysis of Myanmar pfhrp2 and pfhrp3 suggested extensive genetic variations in the both genes consistent with previous studies for field isolates from various geographical areas [8,9,17,22,23,25,27,28]. Similar to previous results, overall genetic diversity was greater in pfhrp2 than pfhrp3 in Myanmar isolates. Structural organizations of repeat types in Myanmar pfhrp2 and pfhrp3 were highly diverse. Most Myanmar pfhrp2 and pfhrp3 sequences occurred only once in the analysed P. falciparum isolates. Characteristics shared by the isolates were also identified. The majority of Myanmar pfhrp2 started with type 1 repeat and terminated with type 12 repeat. Similarly, the majority of Myanmar pfhrp3 started with type 1 repeat and terminated with type 4 repeat. This conserved structural organization was also identified in global pfhrp2 and pfhrp3 [25,[27][28][29]; the major repeat types found in Myanmar pfhrp2 and pfhrp3 were also the most common repeat types observed in global pfhrp2 and pfhrp3. Although the frequency of each repeat type differed slightly in global isolates, repeat types 1, 2, 3, 6, 7, 8, 10, and 12 were the most common in Myanmar and global pfhrp2 [25,[27][28][29]. The frequencies of repeat types 4, 5, 13, and 14 varied among global isolates. Repeat types 1,4,7,15,16,17,18, and 20 were the most common in Myanmar and global pfhrp3 [25,[27][28][29]. Few repeat types were regional or country specific. For example, repeat types 15 and 19 were found only in limited numbers of pfhrp2 from several countries including China, India, Kenya, or the Solomon Islands. Repeat type 29 was found only in India pfhrp3, and its frequency was low.
Although global pfhrp2 and pfhrp3 shared similar structural organizations, they also displayed differences. The most important differences identified were length polymorphisms, which are caused by variation in the number and arrangement of different repeat types. These length polymorphisms of global pfhrp2 and pfhrp3 may be induced by several molecular mechanisms, such as recombination, slipped-strand mispairing event, gene conversion, and unequal crossover [30][31][32][33]. The effects of length polymorphisms in pfhrp2 and pfhrp3 on diagnostic performance of PfHRP2-based RDTs are not clearly understood, but they could alter the binding affinity of specific monoclonal antibodies and consequently influence the sensitivity of PfHRP2-based RDTs [9,23]. The relationship between the combined length   -of type 2× type 7 repeats in pfhrp2 and the sensitivity of PfHRP2-based RDTs has been studied previously [8,23,25,28,34,35]. Four studies [8,23,34,35] proposed that at low parasitaemia (less than 250 parasites/ µl), false-negative rates increased as combined lengths of type 2 × type 7 repeats decreased. In contrast, two studies [25,28] found that sensitivity of PfHRP2-RDTs was not influenced greatly by combined lengths of type 2× type 7 repeats [25,28]. Myanmar pfhrp2 were classified into two major groups based on the lengths of type 2× type 7 repeats, borderline group (77.4%, less than 43) and group B (19%, ranged from 50 to 100). The impact of repeat length polymorphisms in Myanmar pfhrp2 on the sensitivity of RDTs was not determined in this study because RDT results for all P. falciparum isolates were not available. Further studies to determine the effect of pfhrp2 length polymorphisms on performance of PfHRP2-based RDTs is necessary. Amino acid changes in pfhrp2 and pfhrp3 are another important characteristic that cause genetic polymorphisms in global isolates [27,36]. In total, 17 and 3 amino acid changes were found in Myanmar pfhrp2 and pfhrp3, respectively, although the frequency of these changes was generally low. Most amino acid changes in Myanmar pfhrp2 and pfhrp3 were  The influence of amino acid changes identified in global pfhrp2 and pfhrp3 on the diagnostic performance of PfHRP2-based RDTs is also not clear yet, and therefore further study is required. This study had several limitations. The pfhrp2 and pfhrp3 were not amplified successfully in all Myanmar P. falciparum isolates analysed in this study, a result explained by the poor quality of genomic DNA. Indeed, some of the pfhrp2-negative samples were also negative for P. falciparum merozoite surface protein-1 (pfmsp-1) and pfmsp-2 amplifications. Otherwise, deletion of pfhrp2 and pfhrp3 in the negative samples is also a possibility, but further study to elucidate this is necessary. Although extreme sequence variations in Myanmar pfhrp2 and pfhrp3 were identified, the effect of these variations on the performance of PfHRP2-based RDTs was not elucidated clearly in this study. Further study, including larger sample sizes and RDT negative samples, is needed to understand the effect of genetic diversity and deletion of pfhrp2 and pfhrp3 on the performance of the PfHRP2-RDTs.

Conclusion
Extensive genetic diversity was found in Myanmar pfhrp2 and pfhrp3. Length polymorphisms due to variation in the number and arrangement of different repeat types as well as amino acid changes contributed to the genetic diversities of Myanmar pfhrp2 and pfhrp3. Comparative sequence analysis of global pfhrp2 and pfhrp3 suggests that global pfhrp2 and pfhrp3 share similar structural features, but they also differ in some features. These results may provide a better understanding of the pfhrp2 and pfhrp3 structure in global P. falciparum population and suggest useful information to develop RDTs with improved quality. Further examination of genetic diversity of pfhrp2 and pfhrp3 in diverse global P. falciparum populations with a larger number of isolates also is necessary to better understand the structural nature of the two genes in the global populations.