Extensive Deletion and Sequence Variation of Plasmodium Falciparum Histidine Rich Protein 2/3 (pfhrp2/3) Genes in Ethiopia: Implication for RDT-based Malaria Diagnosis and Control

Background: Despite remarkable malaria reduction in recent years, malaria remains a public health problem in Ethiopia. With the introduction of rapid diagnostic tests (RDTs), malaria diagnosis has been transformed. However, the Plasmodium falciparum histidine rich protein 2 ( hrp 2) that is targeted by the most widely used RDTs is prone to genetic mutations and gene deletions as observed in recent years. Patients infected with P. falciparum malaria parasites with a deletion in hrp 2/3 gene locus would remain undetected and results in ‘false-negatives’, which are not treated. Undoubtedly, these untreated infected patients are at risk of developing complicated disease and may further fuel parasite transmission. Hence, molecular targeting of the region across exons and anking genes has been used to provide greater conrmatory evidence of gene deletions. This study was initiated to determine pfhrp2 /3 gene deletions including the anking regions. Methods: A cross-sectional study was conducted to determine the prevalence of hrp 2/3 genes deletion.Finger-prick blood samples were collected from a total of 64 febrile patients attending Adama Malaria Diagnostic Centre in 2015. Thick and thin blood smears were prepared for microscopic slide readings, and parasitaemias were determined. Blood samples were spotted onto lter for parasite DNA extraction. Results: From a total of 64 microscopically and PCR conrmed P. falciparum infections, 50 were successfully analyzed for deletion of pfhrp2 , pfhrp3 and anking regions. Extensive deletions were observed in the pfhrp2 gene with all 50(100%) isolates presenting a deletion. This deletion extended downstream towards the Pf3D7 0831900 (MAL7PI.230) gene in 11/50 (22%) cases. In contrast, only 2/50 (4%) of samples had deletions for the upstream gene Pf3D7 0831700 (MALPI.228). Similarly, all isolates had deleted pfhrp3 anking where 40% samples showed absence of this region. However, the deletion also extended toward the upstream region Pf3D7 081372100 (MAL13PI.475). The deleted pfhrp 3 genomic region also extended upstream to the region Pf3D7 081372100 (MAL13PI.475) with 49/50 (95%) of the isolates exhibiting absence of the locus. Conclusion: As patient recruitment was not done on pfhrp 2/3-based RDTs, it is impossible to know if the isolates would test negative or positive in the absence of hrp 2/3 genes. Indeed, the sequence variation and high frequencies of deletion in pfhrp2 and pfhrp3 genes in Ethiopian isolates most likely will have a negative inuence on the performance of currently used pfhrp2 RDTs. This study conrms the presence of that P. falciparum parasite population with extensive deletions of the pfhrp 2 and pfhrp3 genes in Ethiopia and this calls for a countrywide surveillance to determine the extent of these deletions and its effect on routine malaria diagnosis.


Introduction
In Ethiopia, Plasmodium falciparum and Plasmodium vivax co-exist with relative proportions of 0.60 and 0.40 of malaria cases, respectively [1,2]. Nearly 68% of the landmass, where 60% of the population lives, is favourable for malaria transmission [1,3]. Depending on the local malaria transmission intensity or endemicity class, the epidemiology of malaria varies geographically [4]. Like in many most of Africa, the incidence of malaria has been substantially reduced in Ethiopia which also reported a substantial 40% reduction in incidence by the year 2020 [5]. Ethiopia is on track to achieve the 2020 milestone by reducing the incidence of malaria by 40% and also aligns with the World Health Organization (WHO) Global Technical Strategy (GTS) by intensifying existing malaria control activities to eliminate malaria by 2030 [5]. Much of this can be accredited to the national malaria control and prevention strategies, which have introduced insecticide treated nets, insect residual spray and a change in drug policy. Essentially, rapid diagnostic tests (RDTs) have been introduced to improve early diagnosis of malaria in remote areas where microscopic examination of blood smears seems impractical.
RDTs are immune-chromatographic tests that are designed to detect malaria speci c antigens such as lactate dehydrogenase or aldolase for pan-malaria diagnosis and pfhrp2 for P. falciparum speci c diagnosis [6]. Since the rst report detailing the deletion of pfhrp2 in Peru and its impending consequences was published in 2010, several studies have shown the global spread of parasites lacking pfhrp2 gene and its anking regions [7]. In Eritrea, P. falciparum lacking histidine-rich protein has become a major threat to malaria control programs [8,9]. As they are not detected by pfhrp2/3-based RDTs and remain untreated, parasites carrying the pfhrp2/3 gene deletions have a tness advantage posing a challenge to progress made in malaria control and elimination.
Particularly in South America, Peru has shown a high prevalence of these parasites with 41% deletion in pfhrp2 and 70% deletion in pfhrp3 gene, this was also con rmed by another study which showed an increase in prevalence from 20.7 to 40.6% within two years, in another region of the country [10]. Similarly, a study on P. falciparum outbreak on the northern Paci c coast of Peru (2010-2012) using 54 samples showed all 100% samples lacked the pfhrp2 gene [11]. Following this, several countries have reported occurrence of pfhrp2 and pfhrp3 gene deletions [9,[12][13][14][15][16][17][18][19][20][21]. For highly malaria endemic Africa, the prevalence of the deletion from recent report have generally been low or absent. However, very high prevalence has been reported for Eritrea, which shares a border with Ethiopia. So far, no studies have been conducted in Ethiopia to survey or detect the status of pfhrp2 deletions in the country. Undoubtedly, parasites with deleted hrp2/3 genes will be a signi cant setback for malaria elimination in this region since RDT-based diagnostics are a key pillar in effective malaria management and control. Following the WHO recommendation for pfhrp2/3 surveys and surveillance activities to target neighbouring countries where deletions have been reported, this study investigates the extent of pfhrp2/3 deleted P. falciparum parasites in Ethiopia. Molecular targeting the region across exons and anking genes would provide greater con rmatory evidence of gene deletions.

Sample collection and diagnosis of malaria
Finger-prick blood samples were collected from a total of 64 febrile patients attending Adama Malaria Diagnostic Centre in 2015. Thick and thin blood smears were prepared for microscopic slide readings, and parasitaemias were determined. Blood samples were spotted onto lter for parasite DNA extraction.

Parasite Dna Extraction, Pcr Analysis And Dna Sequencing
Parasite DNA was extracted from blood spots on lter paper using chelex extraction methods as described elsewhere [22].
Detection and ampli cation of hrp2/3 genes Successful ampli cation of pfhrp2 and pfhrp3 was made for a total of 50 samples con rmed both by microscope. Exon 2 of both genes was ampli ed to demonstrate the presence or absence of the pfhrp2 and pfhrp3 genes and their anking regions using published protocols and primers (Additional le 3). Brie y, pfhrp3 gene (475 to 485) was ampli ed using the following PCR conditions: denaturing at 95 0 C for 3 minutes followed by 94 0 C for 15 seconds (36 cycles), annealing at 55 0 C for 15 seconds (35 cycles), elongation 60 0 C for 1:3o minutes (36 cycles) and cooling at 16 0 C. Ampli cation of the pfhrp2 genes (475 to 485) was made as shown in Table 1. Table 1 Outer and nested PCR conditions used to amplify pfhrp2 gene.

Hrp2 -Outer
Temp Time Process Step 1 95 3 min Denaturation Step 2 94 15 sec Step 3 55 15 sec Annealing Step 4 60 1:30 min Elongation Step 5 72 5 min Step 6 16 ∞ Cooling Hrp2-Nested Temp Time Process Step 1 95 3 min Denaturation Step 2 94 15 sec Step 3 55 15 sec Annealing Step 4 60 1 min Elongation Step 5 72 5 min Step 6 16 ∞ Cooling For ampli cation of pfhrp2, two different primer sets were used. The rst one targets the repeat region which harbours most of the variable part of the gene (Additional le 3). The second primer set targets an intron region and for this primer set 5 samples gave adequate PCR products and their sequences were determined. Successful PCR products were puri ed by the GeneJet PCR Cleanup Kit from Thermo Fisher Scienti c and sent for sequence determination at Euro ns genomics, Germany. Sequences were analysed by the 4peaks program (A. Griekspoor and Tom Groothuis, nucleobytes.com).

Ethical Issues
The study was approved by Aklilu Lemma Institute of Pathobiolohy, Addis Ababa University, Institutional review Board. Written consent and/or assent were obtained from each study participants.

PCR con rmation of P. falciparum infections
Febrile patients travelled from various areas to Adama Malaria Diagnostic Centre for malaria diagnosis. The male: female ratio was 3.1:1. Participant's mean age was 25.2 years (range 11-48). A total of 50 specimens met the inclusion criteria for this study by yielding positive results for ampli cation for both 18S rRNA gene.
Using 18S rDNA PCR, all 50 specimens were positive for P. falciparum. Additionally, all samples were ampli ed for the pfmdr1 gene. The minimum parasite density reported by microscopy was 400 parasites/µL, which would indicate that this is the threshold for detection by the eld microscopists.
The samples contained different AT repeat sequences (one sample with 10 repeats, and four samples with 17 repeats) (data not shown) suggesting that the deletion in this region did not involve the entire region.
Indeed, the positive PCR products were from a small part covering an intron region where these repeats are found (Additional le 3).
The results for pfhrp3 and anking regions were comparable. Here, only the repeat region was analysed and all isolates had deletions in the pfhrp3 gene (100%) and like pfhrp2 this deletion extended towards the downstream anking region Pf3D7 13272400 (MAL13PI.485) where 40% samples shown absence of this region. However, the extension of the deletion was more prevalent towards the upstream region Pf3D7 081372100 (MAL13PI.475), where 49/50 (95%) of the isolates exhibited absence of the loci. The extension of deletions in pfhrp2 and pfhrp3 genes can be viewed from the Additional les 1 and 2.
By targeting six regions in the hrp2/3 and anking regions, different deletion patterns were observed in Ethiopian P. falciparum clinical samples (Table 2). A greater proportion of parasite isolates had deleted the gene located 3' of pfhrp2, PF3D7_0831900, compared to the anking gene 5', PF3D7_0831700. In sharp contrast, the 5' anking gene PF3D7_1372100 (upstream of pfhrp3) showed more deletions than the 3' anking region PF3D7_1372400 region (downstream of pfhrp3). As such the most common pattern exhibited in the isolates was the presence of the the two anking regions for pfhrp2 in combination with the downstream anking region for hrp3. This was followed by isolates that had deleted the downstream anking region for hrp3 but retained the two anking for hrp2. Notably, only one isolate showed the presence for all four anking regions. Table 2 Results of PCR ampli cation of pfhrp2, pfhrp3 and their respective anking genes in P. falciparum samples collected in Ethiopia.
No. PF3D7_0831700 HRP2 PF3D7_0831900 PF3D7_1372100 HRP3 PF3D7_13724 20 The association between parasite density and hrp2 deletion was not evaluated due to the absence of hrp2/hrp3 positive samples.

Discussion
RDTs have become extremely essential for implementing early diagnosis and prompt effective treatment to substantially reduce the incidence rate of malaria in Africa. Nowadays, RDTs are the most widely used malaria diagnostics especially in areas where microscopic diagnosis is impractical for several reasons.
However, variation in the performance of RDTs variation has been chronicled, probably driven by known deletion polymorphism targeting pfhr2/hrp3 loci. However, the prevalence and dynamics of pfhrp2 deleted strains has not been extensively investigated, especially in Africa. This study is one of the rst reports showing an extensive deletion of pfhrp2/3 genes in clinical isolates from Ethiopia. This was evident for pfhrp2 and its structural homolog pfhrp3 as well as their respective anking genes. In order to rule out the possibility that this was caused by primers we used alternative primers with different binding sites and ampli cation conditions. In addition, we used another locus, pfmdr1, to rule out the possibility of low DNA quality being an issue. Furthermore, we used 18ssRNA to con rm microscopically determined P. falciparum positive specimens.
From all the six targets of pfhrp2/3 and anking regions (Table 2), at least one locus was ampli ed for the 50 studied samples from a total of 64 con rmed P. falciparum clinical samples. In all 50 samples, pfhrp2/3 genes had deletions which continued to their anking regions although deletions here were to a lesser degree. The lack of prior data on pfhrp2/3 deleted parasites in the country limits our investigation on whether these deletions are recent events or were present prior to the introduction of pfhrp2-based RDTs. For Peru, for instance, the appearance of these parasites was evident before the introduction of RDTs, which perhaps show the strength of selective forces on these deletions [7]. In contrast, a recent mathematical modelling study showed that the use of pfhrp2-based RDTs is su cient to select P. falciparum parasites lacking this protein.
Interestingly, in Ethiopia pfhrp2-based RDTs are far more popular than pLDH-based RDTs which could be because of their higher sensitivity for P. falciparum diagnosis. However, the fact that both pfhrp2 and pfhrp3 deletions were detected in all of our samples probably indicate that this could have been a result of a recent selection. Furthermore, parasites lacking both genes are quite rare to nd in other African countries, though this is quite prevalent in South American countries [7,10]. Indeed, a recent whole genome study on P. falciparum isolates from 15 African countries has shown a highly divergent Ethiopian P. falciparum population, de ning a genomic background that could likely dictate a different response to selective forces on the parasite [23]. A similar neighbouring parasite population in Eritrea also reported very high frequencies of these deletions [8,9].
RDT-based treatment could be a determining factor to selectively clear non-deleted infections and hence increase the rate of spread of parasites with deletions [24]. However, the bene ts of this mutation to the parasites and if RDT-guided treatments may have selected for pfhrp2-deleted mutants are yet to be determined. In Eritrea where malaria prevalence is quite similar to Ethiopia, pfhrp2-negative parasites had lower genetic diversity compared with that of pfhrp2-positive parasites and formed a closely related cluster probably caused by selection by use of pfhrp2-based RDTs [7]. Given that the observation of partial or complete deletion of the pfhrp2 gene in 2010 in South America sparkled the recommendation against the use of pfhrp2-based RDTs in these areas, the results here are relevant for future policy in Ethiopia [25][26][27]. This needs to take in consideration some limitations, such as the absence of data on diagnostics outcome for the samples based on RDTs that targets the pfhrp2/3 genes. The fact that patient recruitment was not made using pfhrp2/3-based RDTs while the samples were collected makes it di cult to determine if the isolates would test negative and/or positive in the absence of pfhrp2/3 genes. Further studies using microscopy and pfhrp2/3-based RDTs are required at large scale to determine the extent of pfhrp2/3 gene deletions and the role the deletion could play in the test results in Ethiopia. The samples were also collected from one region (health centre) of the country, warranting future and extensive studies to further inform the national malaria control program (NMCP) on diagnostic approaches in Ethiopia as the elimination agenda in pursued. The routine use of RDTs may allow the parasites to escape detection by pfhrp2-based RDTs and may be selected to expand in the population. Worth noting is that although the pfhrp2/3 genes were deleted and/or absent, the respective anking regions were ampli ed and they may contribute to reactivity of pfhrp2/3-based RDTs. Indeed, the sequence variation and deletion in Pfhrp2/3 genes in Ethiopian isolates may not likely to negatively in uence performance of currently used pfhrp2 RDTs given that the RDTs targeting hrp2/3 genes are massively used in the area.
Conclusions P. falciparum parasite populations with deletions of the pfhrp2 and pfhrp3 genes are present in Ethiopia and further large scale studies are required to identify the prevalence of deletion and its effect on current RDTs-