Temporal changes in genetic diversity of msp-1, msp-2, and msp-3 in Plasmodium falciparum isolates from Grande Comore Island after introduction of ACT

Background Malaria is still one of the serious public health problems in Grande Comore Island, although the number of annual cases has been greatly reduced in recent years. A better understanding of malaria parasite population diversity and transmission dynamics is critical for assessing the effectiveness of malaria control measures. The objective of this study is to investigate temporal changes in genetic diversity of Plasmodium falciparum populations and multiplicity of infection (MOI) in Grande Comore 10 years after introduction of ACT. Methods A total of 232 P. falciparum clinical isolates were collected from the Grande Comore Island during two sampling periods (118 for 2006‒2007 group, and 114 for 2013‒2016 group). Parasite isolates were characterized for genetic diversity and complexity of infection by genotyping polymorphic regions in merozoite surface protein gene 1 (msp-1), msp-2, and msp-3 using nested PCR and DNA sequencing. Results Three msp-1 alleles (K1, MAD20, and RO33), two msp-2 alleles (FC27 and 3D7), and two msp-3 alleles (K1 and 3D7) were detected in parasites of both sampling periods. The RO33 allele of msp-1 (84.8%), 3D7 allele of msp-2 (90.8%), and K1 allele of msp-3 (66.7%) were the predominant allelic types in isolates from 2006–2007 group. In contrast, the RO33 allele of msp-1 (63.4%), FC27 allele of msp-2 (91.1%), and 3D7 allele of msp-3 (53.5%) were the most prevalent among isolates from the 2013–2016 group. Compared with the 2006‒2007 group, polyclonal infection rates of msp-1 (from 76.7 to 29.1%, P < 0.01) and msp-2 (from 62.4 to 28.3%, P < 0.01) allelic types were significantly decreased in those from 2013‒2016 group. Similarly, the MOIs for both msp-1 and msp-2 were higher in P. falciparum isolates in the 2006–2007 group than those in 2013–2016 group (MOI = 3.11 vs 1.63 for msp-1; MOI = 2.75 vs 1.35 for msp-2). DNA sequencing analyses also revealed reduced numbers of distinct sequence variants in the three genes from 2006‒2007 to 2013‒2016: msp-1, from 32 to 23 (about 28% decline); msp-2 from 29 to 21 (about 28% decline), and msp-3 from 11 to 3 (about 72% decline). Conclusions The present data showed dramatic reduction in genetic diversity and MOI among Grande Comore P. falciparum populations over the course of the study, suggesting a trend of decreasing malaria transmission intensity and genetic diversity in Grande Comore Island. These data provide valuable information for surveillance of P. falciparum infection and for assessing the appropriateness of the current malarial control strategies in the endemic area.


Background
Malaria is a major infectious disease that led to ~ 212 million clinical cases and about 429,000 deaths worldwide in 2016 [1]. Plasmodium falciparum malaria had been widely distributed throughout the Union of Comoros (Grande Comore, Moheli, and Anjouan Islands) and posed a serious impediment to socioeconomic development historically [2]. To effectively control malaria in Comoros, many malaria control measures have been deployed since 2000s, including indoor residual sprayings (IRS), long-lasting insecticide nets (LLINs), artemisininbased combination therapy (ACT), intermittent presumptive treatment (IPT) for all pregnant women, and, particularly, mass drug administration (MDA) of ACT. These malaria control measures have resulted in substantial decrease malaria infection, from 108,260 cases in 2006 to 1072 in 2015 (about 99.0% decline) in Comoros, with no malaria-related deaths. However, despite the great efforts in malaria control, the annual malaria cases increased from 2015 (1072 cases) to 2016 (1372 cases) in Comoros, and the threat of future malaria outbreak remains. Furthermore, malaria transmission intensity differs among the three islands of Comoros (Grande Comore, Moheli, and Anjouan Islands). In Anjouan and Moheli, there was a limited numbers of malaria annual cases during 2014 to 2016 ( To date, several malaria pre-erythrocytic (RTS/S and PfSPZ Vaccine) or erythrocytic (MSP-1, MSP-2, and MSP-3) stage vaccines have been designed to induce immunity against the pre-erythrocytic or erythrocytic stage of the malaria parasites, respectively [3,4]. Although several vaccines are now being tested in clinical Phase I and II trials (MSP-1, MSP-2, and MSP-3) or even have completed the pivotal Phase III clinical testing (RTS/S), the efficacies of these vaccines have been low, with limited impact against clinical malaria [5,6]. One of the difficulties in developing an effective vaccine against P. falciparum parasite is the extensive genetic diversity of vaccine targets allowing parasites with mutated genes to escape from the host's immune response [7,8]. Thus, studying genetic diversity of malaria parasites in endemic areas may provide important information to improve vaccine design. Additionally, the genetic diversity of P. falciparum parasites has been widely used as an indicator of level of malaria transmission intensity in endemic regions, thus serving as a tool to evaluate the effectiveness of malaria control and intervention.
Polymorphic genetic marks, such as microsatellites and genes encoding merozoite surface proteins (msp-1, msp-2, and msp-3) have been widely used for characterization of parasite genetic diversity [9,10]. Currently, only one study described the genetic structure of P. falciparum parasites collected from Comoros Archipelago (Grande Comore, Moheli, Anjouan, and Mayotte) using microsatellite loci [11], showing that microsatellite genotypes of the P. falciparum populations in Grande Comore were substantially different from those in other two islands (Moheli and Anjouan). Currently, no data on temporal changes in genetic diversity of P. falciparum isolates from Grande Comoros after introduction of ACT are available. Herein, the objective of this study is to investigate the dynamics of genetic diversity and multiplicity of infection (MOI) in clinical P. falciparum isolates from Grande Comore during two different periods (2006-2007 and 2013-2016) using polymorphic markers of msp-1, msp-2, and msp-3. The data in this study provide insights on parasite diversity and MOI after various malaria control measures.

Ethics clearance
This study was approved by the Ethics Committees of Comoros Ministry of Health (No. 07-123/VP-MSSPG/ DNS) and Guangzhou University of Chinese Medicine (No. 2012L0816). Blood samples were collected from children after obtaining written informed consent from their parents or legal guardians.

Study sites and sample collection
This study was conducted on the Grande Comore Island, Union of Comoros ( Fig. 1), that is located in the Indian Ocean off the south-east coast of Africa, to the east of Mozambique and north-west of Madagascar (11°00′-12°00′S, 43°10′-43°35′E). This island has an area of 1147 km 2 with about 420,000 inhabitants (2012 estimate). A tropical hot and rainy season occurs from November to April, and a cooler dry season runs from May to October. Annual temperature ranges from 11 to 35 °C and rainfall ranges from 1000 to 3000 mm per year. Malaria transmission on this island is perennial with most of infections occurring during the rainy season. P. falciparum is the dominant malaria species, with occasional Plasmodium malariae and Plasmodium vivax infections [12].
A total of 232 blood samples from microscopically

PCR amplification and allelic analysis of the pfmsp-1, pfmsp-2, and pfmsp-3 genes
Genomic DNA of each blood sample was extracted using Takara DNA Blood Mini Kit according to the manufacturer's instructions (Takara, Kyoto, Japan). Extracted parasite DNA was dissolved in TE buffer (10 mM Tris-HCl, 0.1 M EDTA, pH 8.0) and stored in microfuge tubes at −20 °C. Segments of the pfmsp-1 (block 2), pfmsp-2 (block 3), and pfmsp-3 were amplified using nest PCR, as described previously [10,13]. An initial amplification of the outer regions of the three genes was followed by a nested PCR with sequence specific primer pairs. All the reactions were carried out in final volume of 25 μl containing 10.0 μl of dH 2 O, 0.5 μl of each primer (0.4 µM), 12.5 μl of Taq PCR Mast Mix (2.5 U) following the manufacturer's instructions (Sangon Bio Inc., Shanghai, China) on a S1000 Thermal cycler (Bio-Rad, Hercules, USA). In the primary amplification reactions, 2.0 μl of template genomic DNA were added as a temple. In the nested reaction, 0.5 μl of primary PCR product was added as a temple. The nested PCR products were separated on 2.0% agarose gel (Sangon Bio Inc., Shanghai, China) and visualized under ultraviolet (UV) trans-illumination. A 100-bp DNA ladder was used to determine the size of PCR products (Sangon Bio Inc., Shanghai, China). MOI of the msp-1 or msp-2 genes was calculated by averaging number of amplified bands per positive P. falciparum isolate as described previously [14]. For DNA sequencing, amplified DNA fragments representing different alleles were purified using a PCR purification kit (Takara, Kyoto, Japan). Purified PCR products from selected isolates representing different alleles of msp-1, msp-2, and msp-3 were directly sequenced in both directions with the primers in the secondary PCR using an ABI PRISM3730 DNA sequencer (Sangon Bio Inc., Shanghai, China). The sequences were also used to correct the estimated molecular weight and to confirm the nature of the amplified product.

Discussion
Dramatic reduction in annual malaria cases has been achieved in Grande Comore through the use of ACT for the treatment of uncomplicated P. falciparum patients, ACT-based MDA, and other malaria control interventions. However, malaria continues to be one of most important public health problems on this island, which calls for monitoring changes in drug resistance status and parasite population dynamics. Determining P. falciparum genetic diversity and MOI from field samples is important for understanding the impacts of malaria control measures on parasite populations and for developing strategies to better control malaria infection. The present study investigates the temporal change of genetic diversity and MOI of Grande Comore P. falciparum populations based on msp-1, msp-2, and msp-3 genes that have been used to monitor parasite population widely [10,15].
Previous reports show that polyclonal infection is more common in areas with high endemicity, and 50-100% of infections are polyclonal infections in mesoendemic and holoendemic areas [43][44][45]. Furthermore, a significant association between the complexity of infection and polyclonal infections with the asymptomatic malaria was observed in malaria endemic area of Congo [46]. In the present study, more than 76 and 62% of the samples examined harboured polyclonal infections (two or three allelic types) of the msp-1 and msp-2 gene, respectively, in 2006-2007 group. The frequencies of polyclonal infections were reduced to about 29 and 28%, respectively, in the 2013-2016 group, which again suggests decreasing population diversity and/or transmission intensity. MOI is conventional index to measure of complexity of infection and intensity of transmission. A high MOI value is often observed in a hyperendemic region with high malaria transmission [21,31,47,48]. In the present study, the MOI values decreased from 3.11 to 1.63 for msp-1 and from 2.75 to 1.35 for msp-2, respectively. The findings in this study are similar to those reported in southeastern Senegal [24] and Congo [30]. The present data suggest a progressive decrease of P. falciparum transmission on this island. In fact, according to a report from the Comoros Ministry of Health, the numbers of annual malaria cases in Grande Comore dramatically decreased  . This is in agreement with previous reports in other countries with declining endemicity [49,50]. However, studies from Senegal, Mozambique, and Iran indicated that the introduction of ACT in Congo has reduced the MOI but not the genetic diversity of msp-2 gene among P. falciparum isolates from children living in Southern districts of Brazzaville [30]. Again, haplotype analysis supports reduced genetic diversity and transmission on the Grande Comore island.
The polymorphism of pfmsp-3 is predominantly confined to sequence diversity in the N-terminal domain within the heptad-repeats (insertion/deletion and nucleotide substitutions) [10]. The present study detected both msp-3 K1 and 3D7, but not recombinant type, similar to those reported from Thailand, Papua New Guinea, India, Keyan [10]. Recombinant msp-3 alleles were detected in Iran and African countries at a very low frequency [51,52]. Some of the Grande Comore parasites collected in this study had new msp-3 alleles (K1-3 to K1-8, 3D7-2 and 3D7-3) that have not been reported previously. However, many parasites showed 100% identity with those from Asia, Africa, and South America reported previously, such as Thailand (AOT86948 with K1-1, AOT86951 with K1-2, and AOT86944 with 3D7-1), India (AEI28718 with K1-1, AEI28725 with K1-2, and AEI28765 with 3D7-1), Kenya (AMM75906 with K1-1, AMM75893 with K1-2, and AMM75927 with 3D7-1), Nigeria (CAJ44166 with K1-1, CAJ44194 with K1-2, and CAJ44184 with 3D7-1), China (AAF04099 with K1-1), Indonesia (AAF59914 with K1-2), Papua New Guinea (AAC47670 with K1-2, and AAC47662 with 3D7-1), Vietnam (AAK94780 with K1-2), and Brazil (AFP75269 with K1-2). In the present study, the msp-3 3D7-1 haplotype was the most prevalent in both 2006-2007 and 2013-2016 groups. The present data were in line with the findings from Thailand, India, and Nigeria, where 3D7-1 haplotypes was the most abundant types [10]. In the present study, K1 allelic type was the predominant (66.7%) in 2006-2007 group. The data in the present study are in some degree consistent with the reports of K1 being the most prevalent type in the Thailand [10], Thailand-Myanmar border [53], and Thailand-Cambodia border [53], but is contrast to previous reports from in Thailand-Laos border [53] and Peru [54], with the 3D7 being the most prevalent type. Over the course of 10 years (from 2006 to 2016), the frequencies of K1 type dramatically decreased from 66.7 to 46.5% (P < 0.01), while the 3D7 type dramatically increased from 33.3 to 53.5% (P < 0.01), suggesting that parasites with the msp-3 3D7 type may survive better after introduction of ACTs in Grande Comore. The total number of haplotypes in msp-3 gene changed from 11 in 2006-2007 to 3 in 2013-2016 (a 60% decline), suggesting a decreasing tend in genetic diversity of msp-3 in Grande Comore after 10 years of use of ACT.
The observation of increased frequencies of msp-2 FC27 and msp-3 3D7 allelic types when general population genetic diversity and other allelic types have decreased are interesting, although we do not know the reason for the shift of the alleles. One remote possibility is that the msp-2 FC27 and/or msp-3 3D7 alleles or some unknown genes nearby (linked to msp-2 and/or msp-3) play a role in parasite response to ACT. Parasites carrying these specific alleles/genes can survive better under drug pressure and increase frequency. Another possibility is that the 2013-2016 parasite populations might consist of some parasites carrying these alleles imported from nearby endemic regions after reduction in the original parasite populations. These issues require further investigations.

Conclusion
This study investigated the temporal change in genetic diversity and MOI of P. falciparum populations in Grande Comore Island after the introduction of ACT using the polymorphic genetic markers (MSP-1, MSP-2, and MSP-3). Results from the current study showed that the prevalence of genetic diversity and MOI in msp-1, msp-2, or msp-3 decreased over the course of the study (July 2006 to July 2016). The data in this study suggest a progressive decrease in genetic diversity likely due to lower malaria transmission intensity. The data presented here provide a valuable information for assessing the appropriateness of the current malarial control strategies in this endemic area.