Edinburgh Research Explorer Genetic diversity and transmissibility of imported Plasmodium vivax in Qatar and three countries of origin

Malaria control program in the Arabian Peninsula, backed by adequate logistical support, has interrupted transmission with exception of limited sites in Saudi Arabia and sporadic outbreaks in Oman. However, sustained influx of imported malaria represents a direct threat to the above success. Here we examined the extent of genetic diversity among imported P. vivax in Qatar, and its ability to produce gametocytes, compared to parasites in main sites of imported cases, the Indian subcontinent (india) and East Africa (Sudan and Ethiopia). High diversity was seen among imported P. vivax in Qatar, comparable to parasites in the Indian subcontinent and East Africa. Limited genetic differentiation was seen among imported P. vivax , which overlapped with parasites in India, but differentiated from that in Sudan and Ethiopia. Parasite density among imported cases, ranged widely between 26.25–7985934.1 Pv18S rRNA copies/ µ l blood, with a high prevalence of infections carried gametocytes detectable by qRT-PCR. Parasitaemia was a stronger predictor for P . vivax gametocytes density (r = 0.211, P = 0.04). The extensive diversity of imported P. vivax and its ability to produce gametocytes represent a major threat for re-introduction of malaria in Qatar. The genetic relatedness between P. vivax reported in Qatar and those in India suggest that elimination strategy should target flow and dispersal of imported malaria into the region. its population structure. Thus, the present study examined the source of imported P. vivax into the transmission free setting of Qatar, assessed its genetic complexity and its ability to produce gametocytes and transmit malaria. In addi- tion, we compared the extent of diversity and genetic relatedness of imported parasites into Qatar with that in three countries of origin of immigrant malaria into Qatar, in Africa (Sudan and Ethiopia) and the Indian Subcontinent (India). Such knowledge would allow regional and national control programs to develop targeted policies to reduce the size of infectious reservoir, define the source of outbreaks and limit the risk of re-introduction of malaria.

programs in the region has prompted the health authorities in GCC countries to shift strategies towards malaria elimination and prevention of re-introduction in receptive areas 5 . However, the above success is vulnerable to a possible resurgence of local transmission caused by imported parasites via asymptomatic travellers from malaria-endemic areas 6,7 . The risk of malaria resurgence depends on the combined effect of receptivity and vulnerability. Receptivity is a function of the presence of local vectors and environmental conditions that favour malaria transmission, while vulnerability reflects the probability of importation of malaria parasites into a country 8 . The region is highly vulnerable to re-introduction of malaria, given the high volume of immigration from malaria endemic regions. This has been implicated in the persistence of high rates of imported malaria in many GCC countries 9 . For example, while a significant reduction in locally acquired malaria cases was observed between 2000 and 2014 in Saudi Arabia, imported cases remained consistent averaging between 250 and 830 cases per year 6,10 . Imported malaria is also common in Qatar, Bahrain, Kuwait and the UAE 7,11-13 , and has been implicated in repeated malaria outbreaks in Oman 14 . In Qatar, malaria transmission has been interrupted since 1970 15 , however, a constant influx of imported malaria by the large number of immigrant workers from the Indian Subcontinent, and Sub-Saharan Africa, poses a risk for re-introduction 11,16 . Receptivity of the country to a re-introduction of malaria is evident by the presence of two potential vectors, Anopheles stephensi and A. multicolour 17 . The proportion of all imported malaria reported amongst travellers in Qatar has increased over the past decade, with P. vivax being the main prevalent parasite 11 .
The impact of imported malaria, often as asymptomatic infection, on local epidemiology and local transmission is not well defined in Qatar and surrounding regions. However the ability of asymptomatic P. falciparum and P. vivax infections to produce gametocytes is well documented 18,19 . Currently there is no clear information on the actual magnitude of asymptomatic parasite reservoir in areas where malaria has been eliminated in the GCC countries and its population structure. Thus, the present study examined the source of imported P. vivax into the transmission free setting of Qatar, assessed its genetic complexity and its ability to produce gametocytes and transmit malaria. In addition, we compared the extent of diversity and genetic relatedness of imported parasites into Qatar with that in three countries of origin of immigrant malaria into Qatar, in Africa (Sudan and Ethiopia) and the Indian Subcontinent (India). Such knowledge would allow regional and national control programs to develop targeted policies to reduce the size of infectious reservoir, define the source of outbreaks and limit the risk of re-introduction of malaria.

Results
Demographic characteristics of imported malaria in Qatar. Out  Spearman's rank-order correlation test showed a positive relationship between the copy numbers of Pv18srRNA and Pvs25 transcripts (Correlation Coefficient, r = 0.211, P = 0.04), suggesting an association between parasitaemia and gametocyte density (Fig. 2).    There was an extensive diversity among imported P. vivax in Qatar (H E = 0.782) similar to that seen in endemic sites in Sudan (H E = 0.726), Ethiopia (H E = 0.684) and India (H E = 0.797) ( Table 2). A total of 142 alleles in all loci were found among imported P. vivax in Qatar compared to 96, 93 and 87 alleles in Sudan, Ethiopia and India, respectively. However, many private alleles were seen in some loci in all sites, imported parasites, Qatar (n = 48), Ethiopia (n = 34), Sudan (n = 33) and India (n = 30) ( Table 2).
The predominant allele at each locus in each isolate was used to construct multi-locus genotypes (haplotypes). Each P. vivax isolate, among imported cases as well as parasites collected in Sudan, Ethiopia and India, carried a distinct multi-locus genotype, with exception of 5 pairs (10 isolates) in Sudan each shared a distinct genotype.
Linkage disequilibrium. The multi-locus genotype (haplotypes) data was used to estimate the standard index of association (I S A ). This determines whether there is an association between alleles on different loci, and if the above high levels of diversity observed in P. vivax populations could be explained by frequent genetic exchange.
Limited linkage disequilibrium (LD) was seen among the imported cases in Qatar I S A = 0.0050 while a relatively higher LD value was seen among parasites in Sudan (0.0958) compared to parasites from Ethiopia (0.0367) and India (0.0265) ( Table 3). These values, suggest that a particular combination of alleles on different chromosomes exist at a higher frequency than that expected among randomly mating parasites.

Population differentiation.
To assess the genetic relatedness among imported P. vivax in Qatar, the isolates were divided into three populations (India [n = 132], Pakistan [n = 89] and Sudan [n = 24]) based on the nationality of patients. Pairwise analysis showed a close relationship between parasites from Pakistan and India (F ST = 0.0008), compared to that seen between parasites from Sudan and India (F ST = 0.0037) or Sudan and Pakistan (F ST = 0.0011). Nonetheless, imported P. vivax from the three countries can be considered as one population. Similar close genetic relatedness was seen between imported P. vivax in Qatar, treated as one population, and isolates collected in India (F ST = 0.0594). However, imported P. vivax in Qatar was slightly diverged from local parasite in Sudan (F ST = 0.1435) and Ethiopia (F ST = 0.1896).
The above results were supported by Principal Component Analysis (PCoA) (Fig. 3). The Ethiopian P. vivax population was distanced from the imported parasites in Qatar, while the Sudanese and Ethiopian parasites were slightly overlapped (Fig. 3). Analysis of molecular variance (AMOVA) indicated that most of the genetic variation (86%) was contained within parasites in each region and only 14% can be explained by differences between regions.
Structure analysis showed a clear separation between the imported P. vivax in Qatar and the East African isolates (Sudan and Ethiopia) when K = 2. However, when K = 3 there was some overlap between Sudanese isolates and imported cases in Qatar in agreement with the F ST analysis. However, when K = 4, a clear overlap between the imported parasites to Qatar and parasite in Sudan and Ethiopia (Fig. 4A), suggesting a likely existence of 4 sub-populations (Fig. 4A). Moreover, re-analysis of the data using the imported cases as single population revealed a high relationship between the isolates collected in India with the imported P. vivax in Qatar compared to the parasites in Sudan and Ethiopia (Fig. 4B).   Moreover, the minimum number of clones (MNC), estimated as the minimum number of alleles per locus, across all loci, was relatively low among imported P. vivax from India (1.6), Sudan (0.7), and Pakistan (1.2) compared to that seen among endemic parasites in Sudan (2.4), Ethiopia (2.00) and India (2.00).

Discussion
The GCC countries have achieved great success in the fight against malaria. Four out of the six states accomplished complete interruption of malaria transmission and a malaria-free status, while limited sites in southwest Saudi Arabia 10 and sporadic outbreaks in Oman 14 represent the final stage in the battle to eliminate endemic malaria in the region. However, the above success is threatened by the high influx of imported malaria by young immigrants from endemic countries in the Indian subcontinent (India, Pakistan) and East Africa (Sudan).
The findings of the present study are in line with previous reports, demonstrating that the Indian Subcontinent is the most likely source for imported P. vivax malaria into GCC countries while Africa is the main origin of P. falciparum malaria. For some decades, the oil-rich region has been a popular destination for temporary labour workers, seeking employment opportunities. The high migratory turnover of immigrants from malarious areas is accompanied by increased reports of imported malaria in GCC countries 9 , reflecting a high vulnerability for re-introduction of malaria in areas where the disease has been eliminated. Though, receptivity, defined as the presence of the Anopheles vector, is limited as the result of current efficient vector control programs 8,20 . Nonetheless, autochthonous malaria cases were seen near livestock farms and building sites, where a large number of immigrants work, and environmental conditions favour vector habitats 14

. Analysis of a recent malaria outbreak in Oman identified
Anopheles culicifacies as the vector involved 14 . Therefore, to minimize vulnerability, to re-introduction of malaria, a stringent surveillance system, based on highly sensitive molecular tools, should be structured to ensure timely detection and management of outbreaks seeded by imported malaria.
The vulnerability to re-introduction of local transmission is emphasised by the high prevalence of gametocytes carriage among imported P. vivax in Qatar. Out of 95 P. vivax isolates successfully examined by RT-PCR, 83 (87.37%) harboured gametocytes evident by detectable transcripts of Pvs25 gene. This accords well with the findings of the high prevalence of low-density gametocyte carriage among P. vivax infection in many endemic sites 21 . P. vivax isolates examined in the present study were obtained from clinical cases. However, imported malaria often exists as asymptomatic low-level parasitaemia detectable only by PCR 22 that may easily be missed by traditional microscopy and serological surveillance methods. An additional feature of asymptomatic infection, of particular relevance to prevention of re-introduction in areas where malaria has been eliminated, is its extended longevity. Asymptomatic P. falciparum infection can persist for several years, and discovered only when an individual develops clinical malaria 23 . Clinical presentations of P. falciparum malaria among immigrants have been found to delay by as long as 8 years after patients have left malaria-endemic areas 24 . Thus, asymptomatic parasite carriers can sustain long-term threat to malaria elimination programs.
The presence of gametocyte among low-density asymptomatic parasite carriers is well documented 25 . Amplification-based techniques are sensitive enough to detect and quantify gametocytes as low as 0.02-10 gametocytes per microliter 25,26 . Such low gametocytes densities are unlikely to be detected by microscopy and can make significant contributions to transmission 23,27 . In addition, parasites can increase gametocyte conversion rate in response to clues for transmission opportunity 25,27,28 . For example, the reappearance of Anopheles mosquitoes following the long dry period in areas of seasonal transmission can promote surge in gametocyte densities among asymptomatic parasite carriers 26 . This may explain the success of imported asymptomatic malaria to raise regular outbreaks in receptive sites, with low mosquito density in some GCC countries such as Oman 14  . This is consistent with previous reports on diversity of P. vivax in the Indian Subcontinent 30 and East Africa 31 . The high level of diversity seen among imported P. vivax in Qatar reflects a constant flow of importation of diverse strains from multiple origins. This hypothesis is supported by the close genetic relatedness of imported P. vivax in Qatar and local parasite in India as shown by F ST and PCoA analysis (Fig. 3), and some overlap with parasites in East Africa. The introduction of novel lineages into the region can increase effective population size (Ne) and enhance the parasite diversity, as there is a direct relationship between the expected level of diversity and Ne 32 . In addition, the appearance of drug resistance in countries of origin such as India 33 , Pakistan 34 and Ethiopia 35 can augment the ability of the parasite to establish and evolve in the face of current control measures, if local transmission arises.
The extent of diversity among imported P. vivax is higher than that seen amongst P. falciparum reported by immigrants in Qatar at the same time of the present study 36 . This is in line with data from different endemic areas that examined sympatric P. vivax and P. falciparum populations 37 . This has been attributed to the fact that the global P. vivax population is older, efficient in transmission, very diverse and less structured than P. falciparum 37,38 . This is conistsnt with the findings of low level of genetic differentiation between imported P. vivax in Qatar and local parasites in India (F ST = 0.0594), and moderate differentiation between imported parasites in Qatar and that in East Africa (Fig. 3). Thus, analysis of imported P. vivax in transmission-free areas and autochthonous malaria in the GCC countries, can predict the genetic relatedness and the likelihood of success of integrated intervention strategy applied across the whole region.
In summary, the present study highlighted the vulnerability of GCC countries to resurgence of malaria via parasite carried by immigrants from endemic areas. Imported P. vivax infection in Qatar is characterised by a high rate of genetic diversity and ability to produce gametocytes to transmit disease. The genetic diversity and complexity of imported parasites can lead to appearance of novel genotypes, that can escape current treatment regimen should local transmission starts. However, the risk of receptivity and malaria resurgence is limited, as a result of adequate vector control programs, despite the presence of local vectors and ecological conditions favourable to malaria transmission 39 .

Study sites and P. vivax isolates. A total of 583 patients reported to Hamad Medical Corporation (HMC)
and Alkhor Hospital; Hamad General Hospital in Al-Doha city and Alkhor Hospital in the north area are the two main hospitals within HMC that assigned to admit patients with malaria, were examined microscopically for malaria between January 2013 and October 2016. All cases were diagnosed using conventional microscopic examination of Giemsa stained thick (100 fields) and thin blood (1000 RBCs). A total of 448 (76.8%) were found to be positive for the malaria parasites.
All subjects were interviewed using a structured questionnaire to collect demographic information, including age, gender, nationality, history of travel and treatment. Patients were given treatment as per the Ministry of Public Health, Qatar.
The present study focused on analysis of 265 isolates obtained from patients with P. vivax infection, in Qatar. In addition, we also used 137 confirmed infected samples from Sudan, collected in Whatman qualitative filter paper,  35 . Samples were collected after obtaining a written informed consent from the patients or guardians. All experiments in this study were performed in accordance with relevant guidelines and regulations of the above institutes.
Extraction of DNA and RNA. DNA was extracted from 200 µl of venous blood from the imported P. vivax isolates in Qatar and parasites collected in India, using QIAamp DNA Blood Mini Kit as described by the manufacturer (QIAGEN,CA, USA). DNA was isolated from blood collected on filter paper (Whatman qualitative filter paper, grade 3) from Sudan and Ethiopia using Chelex extraction method 40 .
RNA was extracted from 100 µl blood of P. vivax isolate obtained from imported cases in Qatar using SV Total RNA Isolation kit (Promega, UK). Conventional PCR was used to confirm the absence of co-extracted genomic DNA, and then RNA samples were converted to cDNA using the High Capacity cDNA Reverse Transcription Kit (ThermoFisher, UK).
Detection and quantification of total parasitaemia and gametocytes among imported P. vivax in Qatar. Species identification of P. vivax was confirmed using species-specific PCR as described by 41 . Total parasitaemia was quantified as copy number of the 18S rRNA transcripts in reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) assays based on TaqMan probe chemistry, primers and probes sequences are provided in Table 4 19 . Pv18s rRNA copy numbers were estimated using in-plate standard curve generated from a 10-fold serial dilution (starting with 3.79 × 10 7 copies/µl, limit of detection 0.37 copy/µl) of a purified PCR product. Gametocytes were detected and quantified using RT-qPCR, targeting transcripts of Pvs25 gene, as described elsewhere 21 , with the use of modified probes and primers sequences ( Table 4). PCR products were analysed using ABI 3130XL Genetic analyser (Applied Biosystems,UK) with relative to an internal reference Genescan 500 LIZ (Applied Biosystems, UK). GeneMapper v4.0 (Applied Biosystems, UK) for allele calling and quantification of peak height 43 .

Data analysis. Multiplicity of Infection (MOI).
The multiplicity of infection (MOI) refers to presence of more than one allele per isolate at least in one locus. To avoid overestimation of MOI, multiple alleles per locus were scored if electrophoretic peaks corresponding to minor alleles were ≥33% the height of the predominant allele 44 . The minimum number of alleles across the 9 loci in each sample was calculated and this index value represented the minimum number of clones (MNC) (genotypes) per isolate. Then the average MNC for all isolates in each region was calculated 44 . Population genetic indices. Genetic diversity parameters were calculated for the entire dataset using GenAlex v6.5 45 . This included determining the number of alleles per locus, and expected heterozygosity (H E ). These two parameters were used to assess the level of polymorphism at each locus and determine diversity. H E was calculated using the formula for 'unbiased heterozygosity' also termed haploid genetic diversity, where n is the number of isolates and p the frequency of each different allele at a locus 46 .
The predominant allele in each of the examined 9 microsatellites was used to construct multi-locus haplotypes to determine whether the P. vivax populations in different regions comprised a single panmictic population with a high degree of genetic exchange. Multilocus linkage disequilibrium (LD = non-random association of allele among loci) of the alleles at pairs of loci was estimated as the standard index of association (I S A ) using the program LIAN version 3.5 47 . The software tests for independent assortment of alleles by determining the number of loci at which each pair of MLGs differs, and from the distribution of mismatch values a variance V D (the variance of the number of alleles shared between all pairs of haplotypes observed in the population) is calculated, which is then compared with the variance expected for LE, termed V e . The null hypothesis that V D = V e is tested Name Sequence (5′-3′) PV18s-qFW TCT AGC TTA ATC CAC ATA ACT GAT AC PV18s-qRV  CCR AAG CAA AGA AAG TCC TT   VIV 18s-probe  6-Fam-TCG TAT CGA CTT TGT GCG CAT TTT GCT-BHQ-1   Pvs25-qFW  AAG TGT AYG TGT AAC GAA GGG   Pvs25-qRv  TAT ACA CTG GCC AAA TTC CC   Pvs25-qPr  6-Fam-CGC ATG CTT TGC CTA GGG TTT CTT TCT-Tamra   Table 4. Primers and probes used for qRT-PCR of Pv18s rRNA and Pvs25.
SCIEntIFIC REPORTS | (2018) 8:8870 | DOI:10.1038/s41598-018-27229-z by a Monte Carlo simulation and a parametric method and the results provide 95% confidence limits, which are denoted L MC and L PARA , respectively. If there is limited or no association between alleles at different loci, indicating panmixia, a value close to zero is obtained, whereas if association is detected, a value significantly greater than 0 is obtained, indicating non-panmixia 47 . The variance of pairwise difference (V D ) between the data and that predicated for panmixia 14 and L (L MC & L PARA ), were calculated in order to test the hypothesis of panmixia. Therefore, when the V D value exceeds the L value, LD is indicted and the null hypothesis of panmixia is discarded. When the V D is less than L, LE is indicated and the null hypothesis of panmixia is accepted 47 . The genetic association between population pairs was evaluated by Wright's fixation index (F ST ) using Gene pop web interface 48 . Molecular variance (AMOVA) was calculated to estimate the variation within and between population and Principal component analysis (PCoA) was used to visualise the relationship between MLGs using GenAlEx 6.5 45 .
Structure software was used to elucidate the genetic structure and to detect the most likely number of clusters (K) 49 . STRUCTURE analysis runs were performed exploring K from 1 to 6 (10 iterations each), consisting of a burn-in period of 10,000 iterations followed by 100,000 Markov Chain Monte Carlo (MCMC) iterations, assuming a mixture model and correlated allele frequencies. K was defined according to 49 by calculating the rate of change of K, ΔK using STRUCTURE HARVESTER v0.6.94 50 . Additional analysis of structure results was conducted using CLUMMP 51 and DISTRUCT 52 . CLUMMP aligns the cluster assignment across replicate analyses while DISTRUCT performs a graphical display of the aligned cluster assignments.