Genetic variation in the Duffy blood group among vivax malaria patients and its impact on disease susceptibility

Abstract

Introduction

Plasmodium vivax is a widely distributed human malaria parasite in Ethiopia. Known for its ability to form hypnozoites and early gametocytogenesis, it promotes transmission and challenges global elimination efforts. The red blood cell reticulocyte stage is the primary target, and entry is facilitated primarily by the interaction between the Duffy-binding protein (PvDBP) and the Duffy Antigen Receptor for Chemokines (DARC). This study aimed to evaluate genetic variation in the Duffy blood group among P. vivax malaria patients and its effect on illness susceptibility.

Method

A facility-based cross-sectional study was carried out between January and June 2024 at four sites, including Arba Minch General Hospital, Dil Fana Primary Hospital, Shecha Health Center, and Weze Health Center in Arba Minch. Molecular screening was performed via SYBR Green qPCR, and Duffy genotyping via a TaqMan-based qPCR assay for 485 microscopy-confirmed P. vivax-infected samples. Demographic and clinical data were stored on an open data kit mobile application and analyzed via SPSS.

Result

Among the 485 samples, 93.8% were mono-P. vivax infections, whereas the remaining 6.2% were mixed P. vivax and P. falciparum infections. Relatively more males (58.4%) aged between 15 and 24 years participated in this study. Almost all the study participants were Duffy positive, with 75.1% (364/485) being heterozygous and 24.5% (119/485) being homozygous. Two of the study participants (0.4%) were Duffy-negative individuals, and both had mixed P. vivax and P. falciparum infections.

Conclusion

This study revealed a low prevalence of the Duffy-negative genotype, all of which presented with mixed infections characterized by low parasitemia. This finding indicates that Duffy negativity is not an absolute barrier to P. vivax infection, suggesting the existence of a possible alternative invasion pathway. Further research targeting alternative invasion pathways is recommended to better understand this phenomenon.

Background

Plasmodium vivax is widely distributed and is capable of causing serious and life-threatening illnesses [1]. In the past, P. vivax malaria was believed to be a benign infection, but it is now recognized as a serious global health concern because its high rates of morbidity and mortality significantly impact public health [2]. In 2023, of 263 million malaria cases worldwide, 3.5% were caused by P. vivax [3]. In the same year, Ethiopia contributed ~ 9.5 million malaria cases and was classified as one of the five countries with the heaviest malaria burdens [3], with approximately 30% of these cases attributed to P. vivax [4].

The process by which P. vivax enters reticulocytes was previously thought to depend primarily on Duff binding protein (PvDBP) and Duffy antigen receptor for chemokine (DARC) interaction [5, 6]. The Duffy-negative phenotype results from a single nucleotide mutation (67 $T\to C$; thymine (T) is replaced by cytosine (C) at 67th position) in the GATA-1 binding site of the FY gene promotor, which disrupts transcription and consequently silences DARC expression in red blood cells (RBCs) [5, 7–11]. This phenotype is highly prevalent in sub-Saharan Africa, particularly West and central Africa, where Duffy negativity was reported in 95–100% [12, 13]. In concurrent, studies from different regions of Ethiopia report frequencies ranging between 2.9–35% [6, 14–16].

Duffy negativity conferred protection against erythrocyte invasion by P. vivax, which normally requires DARC binding to enter red cells [9, 13, 17]. However, this paradigm has been challenged by several reports of P. vivax in Duffy-negative individuals [18]. This makes it inconclusive concerning the classical understanding of P. vivax invasion, suggesting that the PvDBP-DARC interaction is not the sole pathway involved [6, 19]. Emerging evidence suggests that other parasite ligands, such as PvEBP/DBP2 and PvRBP2b, are involved in Duffy-independent invasion [20, 21]. In addition, immature erythroid cells from Duffy-negative donors can transiently express DARC, providing another pathway for merozoite entry [19].

P. vivax infections in Duffy-negative individuals are typically characterized by low, often submicroscopic, parasitemia, which can evade detection by routine diagnosis [7, 9, 22]. In addition to its unique genetic features, such as hypnozoite formation and early gametocytogenesis, Duffy negativity reduces invasion efficiency but does not fully prevent infection, allowing silent reservoirs that hinder elimination efforts [6]. It is clear that Duffy-negative phenotypes are no longer a barrier to P. vivax infections [18, 23], and new strategies are needed to address these emerging challenges. Priority should be given to understanding alternative invasion pathways and identifying parasite ligands and host receptors involved, which may provide novel targets for vaccines or therapeutics [7, 9, 24]. Enhanced molecular surveillance is also critical for detecting low-density and asymptomatic P. vivax infections, particularly in Duffy-negative populations where transmission is likely underestimated [25]. Therefore, this study aimed to assess the genetic variation in the Duffy blood group among P. vivax malaria patients and its impact on vivax malaria susceptibility.

Materials and methods

Study area and sample collection

Facility based cross-sectional study was conducted between January and June 2024 at four study sites in Arba Minch town, including Arba Minch General Hospital, Dil Fana Primary Hospital, Shecha Health Center, and Weze Health Center. Arba Minch is located in Gamo zone, South Ethiopia Regional state in the southern part of the country (altitude 1285 m; Latitude and Longitude: 6.041588 and 37.562034; Fig. 1). Arba Minch has a total population of approximately 201,049 residents, and covers an estimated area of 55 km² [26]. The district lies within a malaria-endemic area characterized by a tropical climate, with both lowland and midland altitudes that favor continuous malaria transmission. Malaria in Arba Minch is highly seasonal and unstable, often leading to localized epidemics, influenced by rainfall and temperature cycles, and an average annual min/max temperatures are approximately 17.3 °C–30.4 °C [27]. The annual rainfall is estimated approximately 889.7 mm peaking in April and October [28].

Fig. 1.

Map showing sample collection sites: Arba Minch town, southern Ethiopia, Ethiopia, 2025

Written informed consent and assent were obtained from patients exhibiting clinical signs of malaria and suspected of malaria who sought diagnosis at one of the four participating health facilities, and those confirmed with P. vivax infection. Once microscopically and RDT confirmed with P. vivax infection, the participants were enrolled and 3–4 ml of venous blood was drawn from each participant using a vacutainer needle and collected into a 4 ml EDTA-coated anticoagulated tube. Dried blood spot (DBS) was prepared and used for the molecular analysis of malaria species and Duffy antigen genotyping. Data were collected using the Open Data Kit (ODK) mobile application and uploaded daily to the Ethiopian Public Health Institute server. Data completeness was reviewed each day, and every microscopically confirmed P. vivax case was cross validated using a malaria rapid diagnostic test (RDT) prior of blood collection. All laboratory activities performed in accordance with established standard operating procedures (SOPs). Each participants obtained appropriate treatment following the microscopic result of the recruited Health facilities laboratory.

Malaria diagnosis

Thick and thin blood films were prepared, thoroughly dried at ambient temperature and stained with 10% Giemsa solution. The stained slides were examined under oil immersion at the study sites, and all positive slides were subsequently reexamined by WHO-certified microscopist. Parasite counting was carried out as follows by counting the parasite in the thick blood film against 200 WBCs, assuming a total white blood cell count of 8000/μl [29]. The parasitemia was then classified as low (< 1000 parasites per microliter of blood), intermediate (1000–4999 parasites per microliter of blood) or high (> 5000 parasites per microliter of blood) [30]. To improve the accuracy of species identification, RDTs were carried out alongside microscopy using the SD Bioline malaria Ag Pf/Pv (05FK80) in each four participating health facility. This assay detects histidine-rich protein II (HRP-II) of P. falciparum and lactate dehydrogenase (pLDH) of P. vivax.

Genomic DNA from both human and parasite sources was extracted from a 6 mm diameter DBS via the saponin/Chelex method and subsequently eluted in 100 μl of TE buffer (Tris–EDTA) [31]. The eluted DNA was stored at −20 °C until analysis. Primers targeting P. vivax and P. falciparum were used to amplify the 18S rRNA specific to P. vivax and P. falciparum, respectively [15]. The PCR amplifications were carried out in a mixture containing 10 μl of SYBR Green qPCR Master mix (2x) (Thermo Fisher), 7 μl of nuclease-free water, 0.5 μl of each primer, and 2 μL of DNA template with a final volume of 20 μl on a QuantStudio Real-Time PCR detection system. The reactions were performed with initial denaturation at 95 °C for 3 min, followed by 45 cycles at 94 °C for 30 s, 55 °C for 30 s, and 68 °C for 1 min with a final 95 °C for 10 s. This was followed by a melting curve step of temperatures ranging from 65 °C to 95 °C with 0.5 °C increments to determine the melting temperature of each amplified product [15]. Positive controls were included in each assay for P. vivax Pak Chong (MRA-342G) and Nicaragua (MRA-340G) and for P. falciparum HB3 (MRA-155) and 7G8 (MRA-926). Uninfected samples and water were used as negative controls [16].

Molecular identification of the Duffy genotypes

DARC genotyping was performed via a TaqMan-based qPCR assay to examine the point mutation (c.1–67 T > C; rs2814778) in the GATA-1 transcription factor-binding site of the DARC gene on the surface of reticulocytes. The following primers/probes (forward: 5’-GGCCTGAGGCTTGTGCAGGCAG-3’; reverse: 5’-CATACTCACCCTGTGCAGACAG-3’) and dye-labeled probes (FAM-CCTTGGCTCTTA[C]CTTGGAAGCACAGG-BHQ; HEX-CCTTGGCTCTTA[T]CTTGGAAGCACAGG-BHQ) were used. For each PCR, 7 μl of TaqMan Fast Advanced Master mix (Thermo), 0.54 μl of each forward and reverse primer (10 nM), 0.54 μl of probe (10 nM), and 2 μl of DNA template were used. The thermal cycling was performed with initial denaturation at 95 °C for 2 min, followed by 45 cycles at 95 °C for 3 s and 58 °C for 30 s. A no-template control was used in each assay. The Fy genotypes were determined via an allelic discrimination plot, which analyzed the fluorescent signal emitted from the allele-specific probes [15].

Statistical analyses

Descriptive statistics, including means, SDs, and frequencies, were estimated via SPSS version 26. Chi-square tests or Fisher’s exact tests were used to compare frequency data. Associations between Duffy genotypes and different variables were assessed, and two-sided p values less than 0.05 were considered statistically significant.

Ethics statement

Ethical approval was obtained from the Ethiopian Public Health Institute Institutional Review Board (Ref No: IRB/531/2023), Addis Ababa, Ethiopia. The official letter was written to the Zonal Health Department of Arba Minch. Written informed consent and assent were obtained from all study participants and/or parents or guardians.

Results

Molecular screening of Plasmodium species

A total of 485 PCR-confirmed P. vivax infections were included for Duffy genotyping. Of these, 58.4% of the participants were male, and 60.5% were between the ages of 15 and 24 years. Most participants were from Gamo ethnicity (93.8%), and 94.2% were urban residents (Table 1). Over 90% reported fever (mean temperature: 37.8 ± 1.1 °C), joint pain, chills, excessive sweating, and appetite loss. Four participants aged 15–24 years were admitted to the inpatient ward and presented with a full spectrum of malaria symptoms, excluding abdominal pain. They presented parasite densities exceeding 5,000 p/µl of blood, comprising three mono-P. vivax infections and one mixed infection.

Table 1.

Sociodemographic characteristics of the study participants (n = 485)

Variables	Category	Frequency (n)	Percent (%)
Gender	Female	202	41.6
	Male	283	58.4
Total		485	100
Resident Location	Bere-Edgetber	39	8.0
	Gurba	64	13.2
	Limat	125	25.8
	Nechsar	49	10.1
	Shecha	48	9.9
	Sikela	97	20.0
	Shara Chano	32	6.6
	Others (Rural Resident)	31	6.4
Total		485	100
Age group	< 14	17	3.5
	15–24	295	60.8
	25–34	100	20.6
	> 35	73	15.1
Total		485	100
Clinical presentation	Arba Minch General Hospital	142	29.3
	Dil Fana Primary Hospital	211	43.5
	Weze Health Center	100	20.6
	Shecha Health Center	32	6.6
Total		485	100
Bed net usage	No	364	75.1
	Yes	121	24.9
Total		485	100

Near to ninety four percent (93.8%, 455/485) of the participants were confirmed to be P. vivax-mono infected, while the remaining 6.2% (30/485) had P. falciparum and P. vivax mixed infections by qPCR. A 6.2% discrepancy was observed between the microscopic and molecular diagnoses.

Distribution of Duffy genotypes

All 485 samples were successfully genotyped, revealing 483 (99.6%) Duffy-positive and 2 (0.4%) Duffy-negative cases. Both Duffy negative (CT value; 30.8 and 31.9) individuals were females from Nechsar and Shecha kebeles of Arba Minch town. The percentage of heterozygous Duffy-positive individuals (75.4%) was threefold greater than that of homozygous (24.6%) Duffy-positive individuals. The highest proportion of heterozygous (26.4%, 96/364) and homozygous (24.4%, 29/119) Duffy-positive individuals were from Limat kebele (Table 2). Among the 30 PCR-confirmed mixed P. falciparum and P. vivax infections, 2 (6.7%) were Duffy negative, and 28 (TT: 17.9% and CT: 82.1%) were Duffy positive (Table 3).

Table 2.

Analysis of Duffy status by residence, study site, study participants, gender and age group (n = 485)

Variables	Category	Homozygous Duffy negative n (%)	Heterozygous Duffy positive n (%)	Homozygous Duffy positive n (%)
Gender	Female	2 (100.0)	144 (39.6)	56 (47.1)
	Male	0	220 (60.4)	63 (52.9)
Total		2	364	119
Resident Location	Bere-Edgetber	0	37 (10.2)	2 (1.7)
	Gurba	0	46 (12.6)	18 (15.1)
	Limat	0	96 (26.4)	29 (24.4)
	Nechsar	1 (50.0)	32 (8.8)	16 (13.4)
	Shecha	1 (50.0)	32 (8.8)	14 (11.8)
	Sikela	0	77 (21.2)	20 (16.8)
	Shara Chano	0	24 (6.6)	8 (6.7)
	Others (Rural Resident)	0	19 (5.2)	12 (10.1)
Total		2	364	119
Age group	< 14	0	12 (3.3)	5 (4.2)
	15–24	1 (50.0)	223 (61.3)	71 (59.7)
	25–34	1 (50.0)	74 (20.3)	25 (21.0)
	> 35	0	55 (15.1)	18 (15.1)
Total		2	364	119
Recruitment hospital	Arba Minch G/Hospital	1 (50.0)	115 (31.6)	26 (21.8)
	Dil Fana Pr/Hospital	1 (50.0)	153 (42.0)	57 (47.9)
	Weze Health Center	0	74 (20.3)	26 (21.8)
	Shecha Health Center	0	22 (6.0)	10 (8.4)
Total		2	364	119
Bed net usage	No	2 (100.0)	281 (77.2)	80 (67.2)
	Yes	0	83 (22.8)	39 (32.8)
Total		2	364	119

Table 3.

Distribution of mono-P. vivax and mixed P. falciparum and P. vivax infections by Duffy genotype

Variables		Species
		Mono Pv* n (%)	Mixed (Pf† + Pv) n (%)	Total n (%)
Duffy genotype	Negative	0	2 (6.7)	2 (0.4)
	Positive	455 (100.0)	28 (93.3)	483 (99.6)
	Total	455	30	485
Duffy positive	Homozygous (TT)	114 (25.1)	5 (17.9)	119 (24.6)
	Heterozygous (CT)	341 (74.9)	23 (82.1)	364 (75.4)
	Total	455	28	483

^*Pv = Plasmodium vivax † Pf = Plasmodium falciparum

Asexual and sexual parasitaemia comparison by Duffy genotype

The two Duffy-negative patients with mixed P. falciparum and P. vivax infections had low asexual parasitemia counts (Table 4). Microscopically confirmed P. vivax gametocytes were observed in one of these Duffy-negative participants. In contrast, the mean asexual parasite density in homozygous and heterozygous Duffy-positive patients was 6095 p/μl (95% CI 4334–7926 p/μl) and 5832 p/μl (95% CI 4614–7246 p/μl), respectively, which was greater than that in Duffy-negative patients (532 p/μl; 95% CI 240–824 p/μl). Approximately 84.8% of Duffy-positive participants had gametocytemia with less than 1000 p/µl of blood, and approximately two-thirds of them were heterozygous Duffy-positive.

Table 4.

Distribution of the Duffy genotype by asexual-stage parasite density

Variables		Duffy genotype				Chi-Square	p value
		Duffy negative n (%)	Heterozygous Duffy positive n (%)	Homozygous Duffy positive n (%)	Total n (%)
Asexual parasitemia	Low	2 (100)	132 (36.3)	28 (23.5)	162 (33.4)	12.4	0.015
	Intermediate	0	127 (34.9)	57 (47.9)	184 (37.9)
	High	0	105 (28.8)	34 (28.6)	139 (28.7)
	Total	2	364	119	485

Discussion

In regions where Duffy negativity is common, P. vivax malaria was traditionally thought to be rare or absent [32, 33]. However, our findings indicate that Duffy-negative individuals can still acquire P. vivax infection, albeit at a lower frequency than Duffy-positive individuals [33–37]. The occurrence of P. vivax among Duffy-negative individuals may suggest ongoing adaptation in parasite invasion strategies or host susceptibility factors. Although Duffy antigen expression is diminished in mature red blood cells, the high levels of DARC and the transferrin receptor CD71 expressed on immature reticulocytes, which are found largely within the bone marrow and spleen, could be major targets of invasion in Duffy-negative individuals [38, 39].

Although the proportion was too low, the current study confirmed the frequent detection of P. vivax infections among Duffy-negatives in Africa [36, 37, 40, 41], in contrast with findings in Ghana where P. vivax was not detected from Duffy-negatives [42]. Both the two Duffy-negative patients in this study were coinfected with P. falciparum and P. vivax parasites, which is consistent with prior studies from Southwest Ethiopia showing that mixed-species infections were more common than P. vivax-only infections among Duffy-negative patients [14]. It is plausible that Duffy-negative erythrocytes limit P. vivax invasion, resulting in low parasitemia especially when infection occurs as mono-infection. In mixed infections, however, P. falciparum may facilitate clinical presentation and increase the likelihood of detecting coexisting P. vivax parasites. Nevertheless, this observation contrasts with another study showing that mono-P. vivax infections were three times more common than mixed infections in Duffy-negative individuals [6, 43]. Mixed infections in Duffy-negative individuals may complicate control measures. The lower visibility of P. vivax infections in clinical settings can lead to underdiagnosis and suboptimal treatment in mixed cases, exacerbating morbidity and hindering malaria elimination efforts [18, 44, 45]. Additionally, the adaptation of P. vivax to infect Duffy-negative individuals raises concerns over potential changes in transmission dynamics and challenges for elimination initiatives aimed primarily at P. falciparum.

Misdiagnosis of P. vivax poses a challenge in malaria-endemic Africa, as infections can be mistakenly identified as P. ovale owing to morphological similarities, the dominance of P. falciparum in clinical settings, and relatively low parasitemia. Our present findings revealed that the mean asexual parasite density in Duffy-positive individuals was greater than that in Duffy-negative individuals. This was supported by several studies reported that P. vivax infections in Duffy-negative individuals are associated with decreased parasitemia and are often symptom free [6, 37, 40]. This may reflect the parasite's limited ability to invade and multiply in RBCs within the peripheral circulation, making mild or asymptomatic infections more challenging to detect. The presence of gametocytes enables ongoing transmission and poses a significant barrier to malaria elimination programs [6]. Notably, one of the Duffy-negative individuals in this study harbored microscopically confirmed P. vivax gametocytes in the bloodstream, highlighting the potential role of Duffy-negative carriers as silent reservoirs that sustain P. vivax transmission in the area [15]. Additionally, owing to the low parasitemia commonly associated with P. vivax infections in Duffy-negative individuals, standard diagnostic techniques such as microscopy or rapid diagnostic tests are often inadequate. Molecular diagnostic tools such as qPCR demonstrate increased sensitivity, enabling the detection of submicroscopic infections that would otherwise go unnoticed [6, 46, 47]. This advancement is crucial for uncovering the prevalence and dynamics of mixed infections in vulnerable populations, emphasizing the pathways through which P. vivax maintains its presence in Duffy-negative hosts.

In contrast to previous findings in Ethiopia, where heterozygous Duffy-positives had relatively higher parasite densities than homozygous Duffy-positives [15, 48], the currently obtained results showed that homozygous Duffy-positives had higher asexual and sexual parasite density than heterozygous Duffy-positives, supported by previous evidence from Ethiopia [6] and Papua New Guinea [49]. This finding highlighted the advantages of being heterozygous for reducing parasite adhesion, thereby parasite density and is linked to less susceptibility to clinical vivax malaria [50, 51].

Overall, this study confirms that Duffy negativity does not confer complete protection against P. vivax infection. These infections often present with low parasitemia, representing potential hidden reservoirs for transmission. These findings underscore the need for further research to elucidate the alternative invasion pathways utilized by P. vivax in Duffy-negative individuals and their implications for malaria transmission and control strategies. A deeper understanding of P. vivax transmission biology and gametocyte dynamics through infectivity studies and in vitro experiments, particularly in Duffy-negative populations, might aid in the development of targeted interventions for P. vivax control in Africa. Sensitive molecular detection tools, such as PCR, are also recommended in areas where Duffy-negative and Duffy-positive genotypes coexist.

Author contributions

AA 1, JP, and EL conceptualized and formulated the study design and mobilized the resources. MT, AA 2, DM and FG were responsible for the survey, selection, and collection of samples from the study participants. MT, LB, TTS, NK, and BFD were responsible for conducting the experiments and data analyses. MT wrote the original draft manuscript, and AA 1, JP, EL, LB, BF, TS, DM, AA 2, FG, SM, AH, GT 1, GT 2, and MH reviewed and edited the draft paper. All the authors have read and approved the final version of the article. (AA1–Abnet Abebe, AA2–Adugna Abera, GT1–Getachew Tollera, GT2–Geremew Tasew)

Funding

This research is supported by NIH/NIAID R01AI173171 and R01AI162947.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.