Detection of high prevalence of Plasmodium falciparum histidine-rich protein 2/3 gene deletions in Assosa zone, Ethiopia: implication for malaria diagnosis

Background Rapid diagnostic tests (RDTs) targeting histidine rich protein 2(HRP2) are widely used for diagnosis of Plasmodium falciparum infections. Besides PfHRP2, the PfHRP3 antigen contributes to the detection of P. falciparum infections in PfHRP2 RDTs. However, the performance HRP2-based RDT is affected by pfhrp2/3 gene deletions resulting in false-negative test results. The objective of this study was to determine the presence and prevalence of pfhrp2/3 gene deletions including the respective flanking regions among symptomatic patients in Assosa zone, Northwest Ethiopia. Methods A health-facility based cross-sectional study was conducted in febrile patients seeking a malaria diagnosis in 2018. Blood samples were collected by finger-prick for microscopic examination of blood smears, malaria RDT, and molecular analysis using dried blood spots (DBS) prepared on Whatman filter paper. A total of 218 P. falciparum positive samples confirmed by quantitative PCR were included for molecular assay of pfhrp2/3 target gene. Results Of 218 P. falciparum positive samples, exon 2 deletions were observed in 17.9% of pfhrp2 gene and in 9.2% of pfhrp3 gene. A high proportion of deletions in short segments of pfhrp2 exon1-2 (50%) was also detected while the deletions of the pfhrp3 exon1-2 gene were 4.1%. The deletions were extended to the downstream and upstream of the flanking regions in pfhrp2/3 gene (above 30%). Of eighty-six PfHRP2 RDT negative samples, thirty-six lacked pfhrp2 exon 2. Five PfHRP2 RDT negative samples had double deletions in pfhrp2 exon 2 and pfhrp3 exon2. Of these double deletions, only two of the samples with a parasite density above 2000 parasite/µl were positive by the microscopy. Three samples with intact pfhrp3 exon2 in the pfhrp2 exon2 deleted parasite isolates were found to be positive by PfHRP2 RDT and microscopy with a parasite density above 10,000/µl. Conclusion This study confirms the presence of deletions of pfhrp2/3 gene including the flanking regions. Pfhrp2/3 gene deletions results in false-negative results undoubtedly affect the current malaria control and elimination effort in the country. However, further countrywide investigations are required to determine the magnitude of pfhrp2/3 gene deletions and its consequences on routine malaria diagnosis.

public health problem [1]. Plasmodium falciparum and Plasmodium vivax co-exist in Ethiopia [2]. Plasmodium falciparum is responsible for most of the malaria-associated deaths in malaria endemic countries. Accurate malaria diagnostic tools and prompt treatment play an important role in reducing morbidity and mortality of malaria [3].
To date, the approach to malaria diagnosis depends on the specific requirements of the malaria program and the dynamics of malaria transmission. Microscopy and rapid diagnostic tests (RDTs) are the most common front-line diagnostic tools in clinical settings. Molecular diagnostic methods are recently used in countries with low-transmission setting and/or approaching malaria elimination phase [4].
Malaria RDTs targeting histidine-rich protein-2 (HRP2) and lactate dehydrogenase (LDH) are commonly used for diagnosis of P. falciparum and non-falciparum malaria parasites, respectively. In Ethiopia, where P. falciparum predominates, PfHRP2 RDT is widely used [5]. However, a number of factors would affect the performance of PfHRP2 RDT. Among others, the lack of HRP2 antigen production in P. falciparum parasites due a deletion of parts or all of the pfhrp2 gene [6], affects the sensitivity and specificity of HRP2 RDTs. Besides, PfHRP2, the PfHRP3 antigen produced by P. falciparum has been shown to share high homology with PfHRP2 [7]. This PfHRP3 antigen may contribute to the detection of P. falciparum infections in PfHRP2 RDTs, but the performance of these RDTs could also be affected if there is a deletion of the pfhrp3 gene in P. falciparum [8].
Interestingly, partial or entire deletion of pfhrp2 and pfhrp3 genes in P. falciparum populations could result in false negative RDT-based diagnosis of malaria [6], which would delay timely prescription of anti-malarial drugs and allow the rapid spread of these genotypes from untreated patients into the communities. Therefore, P. falciparum populations containing these variants of the pfhrp2 and pfhrp3 genes would be selected and would remain a major threat to existing malaria case management and control efforts [9].
In recent years, deletion of pfhrp2 and/or pfhrp3 genes have been reported in different geographical regions including the Peruvian Amazon Basin [10], Brazil [11], India [12], and Senegal [13]. The effect of pfhrp2 and pfhrp3 gene deletion on the performance of PfHRP2RDTs have also been evaluated in a few East African neighbouring countries of Ethiopa, such as Eritrea [14], Sudan [15], and Kenya [16]. Therefore, it is very important to explore the presence and prevalence of pfhrp2/3 gene deletions in the Ethiopian P. falciparum populations. Moreover, there are no reports of pfhrp2/3 genes deletions in Assosa zone in Northwest Ethiopia. Thus, objective of this study was to assess the presence and prevalence of pfhrp2/3 genes deletion as well as the respective flanking regions among P. falciparum isolates in this area.

Study area
A cross-sectional study was conducted in clinical setting during low and high transmission seasons at four selected health facilities: Assosa, Bambasi, Kurmuk and Sherkole Health Centres in Assosa Zone, Benishangul-Gumuz Regional State, Northwest Ethiopia from November to December 2018. The study area was selected using a simple random sampling technique among eight districts in the Assosa zone. Most of the district has a high intensity of transmission according to the annual parasite incidence report [2]. Majority of the population of the districts are permanent residents and depends mainly on agricultural and mining work. The study area is located on the northwestern border of Ethiopia where there is a local cross-border market between the people of Sudan and Ethiopia. In line with this, there is import and export of the malaria parasite between the two countries. Recently, there is a confirmed report of the pfhrp2/3 gene deletion in Sudan [15]. Thus, this study was design to assess the presence and magnitude of the pfhrp2/3 gene deletion on the northwestern border of Ethiopia. The geographic locations of the study sites was taken by handheld GPS (Garmin GPS 73, United State) and the map was generated using ArcGIS version 10.0 software (Fig. 1).

Blood sample collection
Demographic data were collected at the time of enrollment from clinically suspected malaria patients. Simultaneously, in a single finger-prick, capillary blood sample was collected from each study participant for thin and thick blood smear preparation for microscopy, and Malaria RDT. Dried blood spots (DBSs) were spotted on Whatman ™ FTA ™ filter paper (GE Healthcare, Piscataway, NJ, USA) for molecular assay.

Detection of P. falciparum by microscopy and malaria RDT
Thin and thick blood smears were prepared and stained with Giemsa stain working solution diluted in 10% buffer for microcopy detection of Plasmodium parasites. Briefly, Giemsa working stain solution prepared by a stock volume of Giemsa with nine volumes of buffered water with pH 7.2. Thin blood smear fixed with methanol solution and staining with Giemsa working stain for 10 min on the staining rack. Thin smears were used to identify Plasmodium parasites species while the thick smears were examined to detect and measure the density of the parasite against 200 white blood cells (WBC), assuming mean WBC count was 8000/μL, as per WHO recommendations [17]. The Care Start ™ malaria RDT (Pf/PV (HRP2/ PLDH, RMVM-02591) was performed and interpreted based on the manufacturer's instructions in the package insert.

DNA extraction and quantitative PCR
Genomic DNA was extracted from blood DBS using Chelex-Saponin method as described previously [18]. Plasmodium falciparum identification were confirmed by SYBR green based quantitative PCR (qPCR) assay by targeting the 18S rRNA gene using species-specific primers as described elsewhere [19]. Absolute quantification approach was used to quantify unknown samples by qPCR. Serially diluted known quantity of parasite DNA amplified to generate a standard curve. Then, quantifications were performed by comparing Ct-value for unknown samples against this a standard curve.

Amplification of pfhrp2 and pfhrp3 genes
Of the 499 malaria suspected individuals selected for molecular analysis, a total of 218 qPCR confirmed P. falciparum positive samples with Ct-value < 37 were included for pfhrp2/3 molecular analysis to minimize false negative finding of pfhrp2/3 associated with low density infection in this study (Additional file 1). This study protocol included both the positive and negative samples by microscopy and PfHRP2 RDT to reduce underestimation of pfhrp2/3 gene deletion due to submicroscopic samples.
After confirmation of P. falciparum positive samples by qPCR, pfhrp2 and pfhrp3 gene analysis was performed by amplification of two amino acids coding segments (exon2 and exon1-2). The short segment, exon1-2, spans the end of exon1 to the start of exon2. The longest segment, exon 2, consists of the entire exon2 which is the main amino acid coding region of pfhrp2 and pfhrp3 genes. PCR amplifications were performed on upstream (PF3D7_0831900 (MAL7P1_230) and downstream (PF3D7_0831700 (MAL7P1_228) flanking region of pfhrp2 gene. Likewise, PCR amplifications were also performed on upstream PF3D7_1372100 (MAL13P1_475) and downstream (PF3D7_1372400 (MAL13P1_485) flanking regions of pfhrp3 gene. The study flow chart for molecular analysis of pfhrp2 and pfhrp3 is indicated in Additional file 1.
PCR amplification of pfhrp2, pfhrp3 genes and their respective flanking regions were carried out in the total reaction mixture of 25 μl using 2X Go Taq hot start green master mix (Integrated DNA Technologies, Inc, Coraville, USA). 0.4 μM each of forward and reverse primer, 3 μl of template DNA for each pfhrp2 exon2, pfhrp3 exon2 and pfhrp3 exon1-2, 2 μl of template DNA for each pfhrp2 exon1-2, MAL7P1_230 and MAL13P1_475, 1 μl of template DNA for each MAL13P1_485 and MAL7P1_228 were used based on PCR conditions and primer sequences as described previously [9,20,21]. As a positive control, 3D7 (MRA-102G) was used for PCR amplification of pfhrp2, pfhrp3 and their respective flanking genes. Likewise, DNA of laboratory strains Dd2 (MRA-150G) and HB3 (MRA155G) were used as negative control for the PCR amplifications of pfhrp2 and pfhrp3 including their flanking genes, respectively [10]. All pfhrp2/3 PCR negative samples were repeated to confirm the absence of the target gene (pfhrp2/3 exon 2 and exon 1-2) including flanking regions.
The amplified fragments of pfhrp2, pfhrp3 and their flanking genes were separated by electrophoresis on a 2% agarose gel by staining with SYBR safe in a Tris Borate EDTA (TBE) buffer. The DNA bands were visualized in a UV transilluminator and the expected amplicons size compared to 1 kb DNA ladder. Primer sequences, PCR conditions and expected amplicon size of pfhrp2/3 genes are indicated in Additional file 2.

Data analysis
The proportion of pfhrp2 /3 gene deletions were analysed using Statistical Package for Social Sciences (SPSS) version 20. P value < 0.05 were considered as significant.

qPCR, microscopy and PfHRP2 RDTs
A total of 218 P. falciparum positive samples by qPCR were included based on inclusion criteria for molecular analysis of pfhrp2, pfhrp3 gene and their flanking gene. Of the study participants positive for P. falciparum by qPCR, 56% (122/218) and 62.8% (137/218) of the positive cases were female and in the youngest age group (5-24 years), respectively. PfHRP2 RDT and microscopy detected approximately 60% of infections that were detectable by qPCR. There is statistical difference in positivity rate of PfHRP2 RDTs with respect to gender (P = 0.028) and age group (P = 0.05) ( Table 1).

Deletions in pfhrp2 and pfhrp3 genes
Among the 218 P. falciparum positive isolates, exon 2 deletions were observed in 17.9% (39/218) of pfhrp2 gene and in 9.2% (20/218) of pfhrp3 gene. Likewise, deletions of the pfhrp2 exon1-2 gene exhibited in 50% (109/218) of the isolates while deletions of the pfhrp3 exon1-2 gene were 4.1% (9/218). The deletions extended to   The pattern of deletions in pfhrp2/3 genes and their flanking regions varies according to the study sites. The high proportion of deletions of the pfhrp2 exon2 gene were observed in 37.5% (6/16) of the isolates collected at the Assosa health centre ( Table 2).

Pfhrp2/3 genes deletions by transmission seasons
A comparable pfhrp2/3 genes deletions were observed in both high and low transmission seasons. The frequency of pfhrp2 exon 2 deletion was 19.9% (30/167) in high transmission season and 17.7% (9/51) in the low transmission season. The frequency of pfhrp3 exon 2 deletion was 8.4% (14/167) and 11.8% (6/51) in high and low transmission seasons, respectively. The extent of deletions of pfhrp2/3 exon2 varied by the study site. High proportion deletion of pfhrp2 exon2 (46.2%) was observed in the Assosa district during the high transmission season followed by 28.6% of pfhrp2 exon2 deletion in Bambasi district during low transmission season (Fig. 2).

PfHRP2 RDT and microscopy compared to PCR results of pfhrp2 and pfhrp3
Of eighty-six PfHRP2 RDT negative samples, 41.9% (36/86) samples lacked pfhrp2 exon 2 as compared to 10.5% (9/86) deletions observed in pfhrp3 exon 2 (Table 3). Of thirty-six pfhrp2 exon 2 PCR negative samples, thirty-three samples were submicroscopic unlike the three samples with the parasite densities above 1000 /µl (Additional file 4). A total of five PCR double negative samples for pfhrp2 exon 2 and pfhrp3 exon2 gene were also PfHRP2 RDT negative (Fig. 3). Among those five PCR double negative results of pfhrp2 exon 2 and pfhrp3 exon2 gene, two PfHRP2 RDT negative samples with parasite density above 2000 parasite/µl were microscopy positive while three of them were submicroscopic (Table 4). On the other hand, three pfhrp3 exon2 PCR positive isolates lacked the pfhrp2 exon2 gene. These three samples had a parasite density greater than 10,000 / µl and were positive both by PfHRP2 RDT and microscopy in the absence of the pfhrp2 gene, suggesting the contribution of intact pfhrp3 exon2 to reduce false negative PfHRP2 RDT despite a deletion in the pfhrp2 gene (Additional file 5).

PfHRP2 RDT results against PCR results of pfhrp2/3 flanking regions
Out of eighty-six PfHRP2 RDT negative samples, more than 80% of the samples lacked MAL7P1_230, MAL13P1_475, MAL7P1_228 and MAL13P1_485 in both the upstream and downstream pfhrp2/3 flanking region (Additional file 6). Out of eight-six double deletions of MAL7P1_230 and MAL13P1.475, sixty-six negative samples were found by PfHRP2 RDT and sixty-five negative samples by microscopy. Of one hundred twelve double deletions of MAL7P1_228 and MAL13P1.485, a total of eighty-one samples by PfHRP2 RDT and eightytwo samples by microscopy were negative (Fig. 3).

Discussion
This study confirmed the presence of a considerable number of pfhrp2/3 gene deletions in P. falciparum clinical isolates in the Assosa zone, Northwest Ethiopia. Partial deletions of pfhrp2, pfhrp3, and the flanking genes were observed in the majority of the parasite populations while complete deletion of pfhrp2/3 gene occurred only in two of P. falciparum clinical isolates. On the other hand, 37.2% (81/218) of the clinical isolates did not have deletion in exon 2 and exon 1-2 of pfhrp2/3 region including the flanking regions.
Of 218 positive samples for P. falciparum, microscopy and RDT were able to detect almost 60% of the infections detected by qPCR. If these large number of false negatives individual remain undetected and untreated, they could likely serve as malaria reservoirs and accelerate the onward transmission of malaria in the community [22]. High proportion of PfHRP2 RDT false-negative results in this study is suggested to be due to the deletion of pfhrp2/3 gene in the P. falciparum populations.
Out of a total of 39.5% (86/218) PfHRP2 RDT negative samples in this study, a high proportion of PCR confirmed deletions were observed in pfhrp2 /3 exon2 and exon1-2, and upstream and downstream flanking regions of pfhrp2/3 compared to WHO pfhrp2/3 deletion threshold level [38]. Approximately forty-two percent of (36/86) PfHRP2 RDT negative samples lacked exon 2 of pfhrp2, the main amino acid coding region of the pfhrp2 gene. Of these PCR negative Pfhrp2 exon 2 samples, only three of the samples with a parasite density above 1000 / µl were microscopic positive while the rest were submicroscopic. Moreover, high prevalence of deletion was detected in short segmented of pfhrp2 exon1-2(50%) and pfhrp3 exon1-2(4.1%). High proportion of deletions (> 30%) was also extended to the downstream and upstream of the flanking regions in pfhrp2/3 gene. The complete and partial pfhrp2/3 genes deletions observed in this study would undoubtedly impact the diagnostic performance of PfHRP2 RDTs.
According to the WHO protocol, a deletion of the pfhrp2/3 gene in a certain region will be suspected if there is an increase in discordant reports between routine diagnostic tools, such as microscopy positive and PfHRP2 RDT negative [9]. The protocol for this study reiterates that submicroscopic infections should not be overlooked when it comes to pfhrp2/3 gene deletions. Therefore, positive and negative microscopic results were included in this study to reduce underestimation of the true prevalence of pfhrp2/3 gene deletion due to submicroscopic infection. As a result, thirty-three samples with the pfhrp2 deletion gene were found to be submicroscopic in this study and included only those samples with a Ct value below 37 using qPCR (and hence had > 9 parasite density/µl) to minimize the risk of finding false deletion results associated with a single copy of the target gene (Pfhrp2) and low-density infection. P. falciparum positive clinical isolates with pfhrp2/3 gene deletion may selectively increase if there is no continuous monitoring associated with submicroscopic infections and PfHRP2 RDT negative results. Those parasites with submicroscopic parasite densities with pfhrp2/3 gene deletions could be potentially infectious to mosquitoes and contribute to ongoing malaria transmission which could in turn challenge malaria control and prevention.
In this study, 58.1% (50/86) false negative PfHRP2 RDT results were found to be pfhrp2 exon2 positive by PCR. The likely cause of these false negative PfHRP2 RDT results could be due to low levels of parasitaemia that could be below the detection threshold of PfHRP2 RDT in those submicroscopic infections and absence of PfHRP2 antigen due to host immune response in the high transmission areas [39].
All PCR double negative samples in pfhrp2/3 genes and their flanking regions were PfHRP2 RDT negatives in this study. The number of PCR double negative samples of pfhrp2 exon 2 and pfhrp3 exon 2 were low compared to previous studies in Peru [10] and Ghana [27]. In the reported study, discordant results between PfHRP2 RDT and microscopy were observed in five PCR double negative samples of pfhrp2 exon 2 and pfhrp3 exon 2. Two    The contribution of intact pfhrp3 gene to PfHRP2 RDT positive results is noted in the study under report. Even though only three pfhrp3 exon2 PCR positive samples were lacking pfhrp2 exon2, these three samples were confirmed positive by PCR and microscopy with parasite density above 10,000/µl in the absence of pfhrp2 gene suggesting the contribution of pfhrp3 exon2. This finding indicated that the structural homology between epitopes of PfHRP2 and PfHRP3 antigens allows cross-reaction of the monoclonal antibodies of PfHRP2 with PfHRP3 [40], which is in agreement with the previous studies [16,27]. Indeed, PfHRP2 RDT positive test result with intact pfhrp3 exon2 might underestimate the true prevalence of pfhrp2 gene deletion in this study.
However, it is yet unclear what causes partial or complete deletion of pfhrp2/3 gene in P. falciparum isolates in Assosa zone in Ethiopia. Previous studies have indicated that deletion of pfhrp2 in Dd2 and pfhrp3 genes in HB3 laboratory strains was resulted from deletion at distinct and unlinked positions on chromosome 8 and 13, respectively [10,41,42]. Thus, selection may act differently on the two gene regions leading to different frequency of deletions. A possible explanation for the presence of a considerable number of deletions of pfhrp2/3 gene in this region may be the selective pressure of P. falciparum isolates with pfhrp2 deletion. This study is located in the Assosa zone, on the border with Sudan, which has a high transmission of malaria. Consequently, intense and routine screening of malaria with a large number of RDTs may pose selective pressure on P. falciparum and results in a considerable number of parasites carrying pfhrp2/3 gene deletion in this region. Furthermore, in the high transmission setting, polyclonal infection is common. pfhrp2/3 deleted and non-deleted strains may co-exist within an infected individual and result in positive PfHRP2 RDT [33]. Future study merits expanded samples covering broader geographic regions of Ethiopia and in-depth investigations for the detection of pfhrp2/3 deletion in polyclonal infections of P. falciparum using a multiplex qPCR and its effect on malaria diagnosis. Further investigations of the PfHRP2 antibodies in the plasma of P. falciparum patients infected with pfhrp2/3 deleted and non-deleted strains will clarify the impact of gene variation on the host immunogenic responses.

Limitations
This study has certain limitations. Measurement of PfHRP2 antigen in the plasma was not performed. The molecular methods of this study could not show pfhrp2/3 gene deletion in polyclonal infections of P. falciparum. This study was conducted in a limited geographic area.

Conclusion
This study confirms the presence of more than 5% of pfhrp2/3 gene deletions in P. falciparum isolates in the study area according to the WHO protocol [9]. The high proportion of PfHRP2 RDT false-negative results due to the deletion of pfhrp2/3 gene could affect malaria control and elimination efforts in the country. Hence, further country-wide assessment on the magnitude of pfhrp2/3 deletions is important before considering alternative malaria diagnostic strategies.