Unveiling P. vivax invasion pathways in Duffy-negative individuals

Summary Vivax malaria has long been thought to be absent from sub-Saharan Africa owing to the high proportion of individuals lacking the Duffy antigen receptor for chemokines (DARC) in their erythrocytes. The interaction between P. vivax Duffy-binding protein (PvDBP) and DARC is assumed to be the main pathway used by merozoites to invade reticulocytes. However, the increasing number of reports of vivax malaria cases in genotypically Duffy-negative (DN) individuals has raised questions regarding the P. vivax invasion pathway(s). Here, we show that a subset of DN erythroblasts transiently express DARC during terminal erythroid differentiation and that P. vivax merozoites, irrespective of their origin, can invade DARC+ DN erythroblasts. These findings reveal that a large number of DN individuals may represent a silent reservoir of deep P. vivax infections at the sites of active erythropoiesis with low or no parasitemia, and it may represent an underestimated biological problem with potential clinical consequences in sub-Saharan Africa.


INTRODUCTION
More than one-third of the world's population is affected by vivax malaria, an acute debilitating disease caused by Plasmodium vivax and transmitted by female Anopheles mosquitoes.In 2021, there were an estimated five million vivax malaria cases worldwide, mainly in Asia, Southeast Asia, South America, the Western Pacific, Eastern Mediterranean regions, Eastern Africa, and Southern Africa. 1 For a long time, vivax malaria was thought to be absent in sub-Saharan Africa.3][4][5] Later, Duffy blood groups were identified, 6,7 and it was demonstrated that the absence of the Duffy antigen receptor for chemokines (DARC) on the surface of human erythrocytes, caused by a single-point mutation in the GATA-1 promoter sequence of ackr1À, was associated with natural resistance to vivax malaria. 8Pioneering studies in the 1980s and the 1990s identified the P. vivax Duffy-binding protein (PvDBP) as a specific ligand for the N-terminal extracellular domain (ECD1) of the host DARC and confirmed that the interaction between PvDBP and DARC is critical for reticulocyte invasion. 9,102][13] Together, these findings led to a scientific paradigm whereby P. vivax recognizes reticulocytes through interactions with PvRBP2b-CD71 and PvRBP2a-CD98 and invades genotypically Duffy-positive (DP) DARC+ reticulocytes through interactions between PvDBP and DARC. 14owever, this paradigm has been called into question.With the advent of nucleic acid amplification tests (NAATs), an increasing number of P. vivax infections have been reported in genotypically Duffy-negative (DN) sub-Saharan African populations.6][17][18] In parallel, P. vivax has been shown to preferentially invade CD71 + immature red blood cells (i.e., reticulocytes) and erythroid precursors, found primarily in extravascular erythropoietic tissues, such as the bone marrow, [19][20][21][22][23][24][25][26][27] spleen, [28][29][30] and tissues where extramedullary erythropoiesis can occur. 31These observations demonstrate that anatomic sites of active erythropoiesis provide niches where pathogenic P. vivax biomass increases may occur, implying that vivax infections are more common than suggested by the presence of vivax parasites in peripheral blood. 15ere, we investigated the infection of genotypically DP and DN erythroblasts with P. vivax merozoites in vitro to explore and identify host receptors that P. vivax uses to invade erythroblasts.We performed in vitro erythropoiesis assays of genotypically DN and DP erythroid progenitors to critically evaluate terminal erythroid differentiation and CD71 and DARC expression profiles over time.We then conducted in vitro invasion assays to further investigate whether DARC expressed on genotypically DN erythroblasts is functional for P. vivax invasion.Our key findings were that a subset of genotypically DN erythroblasts transiently express DARC during terminal erythroid differentiation and that this subpopulation is susceptible to P. vivax merozoite invasion, providing biological evidence that P. vivax can invade a subset of DN erythroblasts that transiently express DARC as a functional receptor for invasion.
We further confirmed the expression of DARC in DP and DN erythroblasts and donor erythrocytes by western blot analysis of cytoplasmic protein extracts.We detected differential expression of DARC in both DP and DN erythroblasts during the early stages of differentiation (Figures 3A and 3B).In DN cells, DARC expression was highly variable among the donors (DN001, DN002, DN003, and DN005), whereas DARC expres-sion remained less variable among DP donor cells (DP001, DP003, DP004, and DP007) throughout differentiation.The relative amount of DARC protein was significantly lower in DN erythroblasts at days 10 and 12 of terminal differentiation compared with DP erythroblasts (p = 0.03, Mann-Whitney test) (Figures 3A and 3B).The DARC expression pattern in DN cells during in vitro erythroid differentiation was consistent with the observation that DN erythrocytes from the peripheral blood of the donor were not DARC+ (Figure 3A; red blood cells from DN005 donor). 33This finding was supported by fluorescence microscopy, which showed that only a few DN erythroblasts were DARC+ at day 9 terminal differentiation (Figure 3C).
Hence, we conclude from these experiments that DN erythroblasts clearly express DARC during erythroid differentiation, albeit at lower levels compared with DP erythroblasts.

P. vivax merozoites can invade DN erythroblasts during terminal erythroid differentiation
To determine whether the subset of genotypically DN erythroblasts expresses functional DARC and CD71, the receptors for PvDBP and PvRBP2b that may allow P. vivax merozoite invasion, we isolated CD34+ HSPCs from the peripheral blood of two additional DP (DP007 and DP010) and two DN (DN005 and DN006) healthy donors (Table 1).CD34+ cells underwent expansion and differentiation in vitro (Figure 4A).Both DP and DN erythroid progenitors differentiated and reached the terminal differentiation stage, as described above.In both DP and DN erythroblasts, the expression profiles of DARC and CD71 were consistent with our previous findings.At day 7 of terminal differentiation, we thawed and cultured parasite isolates obtained from two Malagasy patients and seven (A) The diagram illustrates the two-phase in vitro culture system for the isolation of CD34+ cells followed by the expansion and differentiation of erythroid progenitors.(legend continued on next page) Ethiopian patients infected with P. vivax (Table S1; Figure 4A).After approximately 24-30 h of in vitro maturation (which corresponds to day 8 of terminal differentiation for DP and DN erythroblasts) (Figures S5A and S5B), each P. vivax parasite culture was cocultured with either DP or DN erythroblasts for 24-48 h.We then assessed the infection of DP and DN erythroblasts by P. vivax using light microscopy.Approximately 50,000 cells were cytospun onto glass microscope slides and stained with May-Grunwald-Giemsa stain.Of the 22 co-cultures, we detected P. vivax infection in both DP and DN co-cultures (Table S1).We observed very few parasites, either at ring stages or at more mature stages, with large chromatin dots and hemozoin (Figures 4B-4G, S6C, and S7).No Schuffner's dots were observed in the infected erythroblasts.
We confirmed the infection of DP and DN erythroblasts with P. vivax parasites using fluorescence microscopy.About 3 3 10 6 erythroblasts per condition were stained with P. vivax 18S rRNA fluorescence in situ hybridization (FISH) probe, anti-Plasmodium HSP70, anti-DARC, and anti-CD71 antibodies.Of the 22 co-cultures (11 with DP and 11 with DN erythroblasts), P. vivax infections were detected in both DP and DN co-cultures (Figures 4B-4G, S6, and S7).Dual staining with the 18S rRNA FISH probe and anti-Plasmodium HSP70 antibody confirmed the presence of P. vivax parasites in DP and DN co-cultures (Figures 4B and 4C).Almost half of the P. vivax isolates invaded both DP and DN erythroblasts (5/11), 2/11 invaded neither, 3/11 invaded only DP erythroblasts, and 1/11 invaded only DN erythroblasts, suggesting that the ability to invade is critically dependent on the viability of the cryopreserved P. vivax parasites (Table S1).However, as only one to three infected erythroblasts were detected per invasion assay, we were unable to compare the number of P. vivax-infected cells in the invasion assays between DP and DN erythroblasts.As observed by light microscopy, both P. vivax parasites in DP and DN erythroblasts had a similar profile, in that we could observe the parasite nuclei close to the erythroblast nuclei, sometimes surrounded by signals corresponding to the P. vivax 18S rRNA probe and/or HSP70 monoclonal antibody (mAb), together with the expression of CD71 and DARC on the surface of the erythroblasts (Figures 4B-4G, S6, and S7).We found that all DP and DN erythroblasts infected with P. vivax were CD71+ and DARC+.
These results provide biological evidence that P. vivax can invade DN erythroblasts and that transiently expressed DARC is a potentially functional receptor for invasion.

DISCUSSION
Here, we provide key insights into P. vivax invasion pathways in genotypically DN patients by investigating the terminal erythroid differentiation of DN erythroid progenitors and their potential to express DARC and CD71 and present strong biological evidence that P. vivax merozoites, regardless of their origin, can invade DN erythroblasts.
The expression profile of DARC in DP erythroblasts was consistent with previous observations that approximately half of the erythroblasts expressed DARC on day 12. 34 Interestingly, we found that a small population of DN erythroblasts (<5% erythroblasts on day 12) also express DARC.This suggests that although a single-point mutation in the GATA-1 site of the ackr1 gene promoter significantly reduces the number of DARC+ cells, 33 transient expression of DARC may occur in a small subpopulation of DN erythroid cells ($16-fold lower than the number of genotypically DP cells) and supports P. vivax infection.This observation was confirmed by the detection of DARC by western blotting in cytoplasmic protein extracts from DN erythroblasts but not in extracts from DN peripheral blood erythrocytes.Taken together, these data suggest that the GATA-1 binding motif of the ackr1 promoter may be required for the sustained expression of DARC during erythroid differentiation in DP cells, whereas a single-point mutation in the GATA-1 motif in DN cells results in a decrease in the number of cells expressing DARC at terminal differentiation.
In vivo, the maturation of orthochromatic erythroblasts into reticulocytes involves various changes, including enucleation, hemoglobinization, remodeling of the membrane, and expression or depletion of certain proteins. 35Here, we clearly show that DN erythroblasts are deficient in maintaining their ability to express DARC at the end of terminal erythroid differentiation when orthochromatic erythroblasts enucleate to produce reticulocytes or when reticulocytes migrate to peripheral blood and mature into erythrocytes.However, in this study, we were unable to provide conclusive evidence that the subset of DN erythroblasts expressing DARC gives rise to DN reticulocytes expressing DARC.Indeed, we cannot exclude the possibility that the threshold detection of the DARC protein by western immunoblotting, which is based on the measurement of total cytoplasmic proteins, may have failed to detect a small subset of DARC+ cells among the majority of DARC reticulocytes or erythrocytes due to the presence of large amounts of other proteins, such as hemoglobin.
Overall, we observed that the number of DP and DN erythroblasts expressing DARC during terminal erythroid differentiation varied widely between donors, probably due to the involvement of multiple regulatory elements controlling mRNA expression of the ackr1 locus, among others, that may influence gene regulation.DARC expression profiles contrasted with CD71 expression profiles in both DP and DN erythroblasts, consistent with previous observations. 368][39] Given the specific tropism of P. vivax merozoites for CD71+ immature reticulocytes 13 and the importance of the PvRBP2b-CD71 interaction in the recognition of DP reticulocytes prior to invasion, 11 our data confirmed that P. vivax  infections are restricted to CD71+ erythroid progenitors, erythroblasts, and/or immature reticulocytes.Since the development of functional invasion assays requires large numbers of erythroblasts, we decided to establish co-cultures on day 8 for terminal differentiation (Figure 4A).DP and DN erythroblasts were co-cultured with either Malagasy or Ethiopian merozoites of P. vivax.When DN erythroblasts were cocultured with P. vivax parasites, only DARC+/CD71+ DN erythroblasts were invaded by the parasites.These observations suggest that DN erythroblasts express functional DARC and that P. vivax merozoites can use the PvDBP-DARC invasion pathway, irrespective of the origin and genetic background of the P. vivax strains.This was confirmed as there was no evidence of P. vivax invasion of erythroid cells that were not positive for DARC.Therefore, this finding suggests that P. vivax has not evolved an alternative invasion pathway to overcome Duffynegativity. 40,41However, definitive evidence through the use of an anti-DARC mAb to prevent erythroblast invasion by P. vivax is not provided in our study.These additional experiments were not performed due to the low number of infected cells per assay (approximately 1-3 infected cells) and the inability to accurately quantify the difference in the number of infected cells between the two culture conditions (with and without mAb anti-DARC).
Overall, the expression of DARC assessed in erythroid progenitors from DN donors may confirm the accumulating evidence of P. vivax endemicity across the DN population and provide evidence that P. vivax may be a prominent biological problem in sub-Saharan Africa. 15,42,43The results obtained here suggest that the biomass of P. vivax shifts away from the peripheral blood into the extravascular spaces of the bone marrow and other tissues, where immature reticulocytes can form or accumulate.This echoes old and recent reports of P. vivax presence in the bone marrow and the spleen, which have been observed by autopsy, aspiration, biopsy, or acute infection following bone marrow transplantation. 15,30Therefore, anatomic sites of active erythropoiesis, including bone marrow, spleen, liver, and other tissues in which extramedullary hematopoiesis can occur (as seen in many pathologies such as chronic and deep anemia), may represent a hidden reservoir of P. vivax parasites, [19][20][21][22][23][24][25][26][27][28][29][30][31] suggesting that P. vivax parasites may have established stable transmission at low levels in sub-Saharan Africa. 156][17][18] Importantly, these data suggest that the extent of the prevalence of P. vivax infection in DN African population is currently unknown, as almost all evidence of P. vivax infection comes from the conventional detection of parasites in Giemsa-stained blood smears from peripheral blood.The hypothesis that P. vivax infections are more widespread than can be deduced from peripheral parasitemia is supported by serological surveys that have shown that P. vivax seropositivity in sub-Saharan Africa ranges from 13% to 53% (reviewed in Baird 15 ).Diagnosis of vivax malaria based on the microscopic examination of Giemsa-stained blood smears, rapid diagnostic tests, or even NAATs has probably missed many instances of vivax malaria cases, especially in sub-Saharan Africa where DN Africans predominate and in which tropism away from peripheral blood may be more pronounced.Therefore, there is a need for novel diagnostic tools capable of detecting deeper infections to update the current landscape of vivax malaria distribution, frequency, and assessment of clinical impact in sub-Saharan Africa.
On a more positive note, our findings also have important implications for the development of therapeutic approaches and vaccines against P. vivax malaria.Indeed, evidence that P. vivax merozoites invade DARC+ DN erythroblasts suggests that the leading vaccine candidate targeting the PvDBPIIbinding domain of PvDBP in P. vivax merozoites [44][45][46] could be used to prevent vivax malaria in both DP and DN populations.
Although our experimental work represents a breakthrough in our understanding of P. vivax invasion pathways, it has several limitations.First, the high inter-individual variability observed has prevented us from fully characterizing DN erythroblasts in vitro.Second, the invasion of erythroblasts by P. vivax merozoites, in our hands, resulted in few infected cells, probably suggesting that the successful invasion of DN erythroblasts by P. vivax merozoites is rare.This low number of infected cells prevents us from quantifying and evaluating the inhibition of P. vivax invasion into erythroblasts by anti-DARC antibodies to exclude the potential existence of alternative invasion. 40,41,47,48An alternative approach could be to perform in vitro assays using Duffy-null erythroblasts from immortalized erythroid cell lines (ackr1 knockout).As the invasion of erythroblasts by P. vivax merozoites is highly dependent on culture conditions such as the viability of P. vivax parasites from cryopreserved parasite samples and parasitemia, performing invasion assays with the use of fresh clinical P. vivax isolates from ex vivo field studies could also overcome these technical problems.Third, we do not have complete evidence that in vitro invasion assays fully replicate the in vivo invasion mechanism of P. vivax.Nevertheless, we provide here insights into how genotypically DN erythroblasts can be infected with P. vivax merozoites.This finding reveals that a large number of DN individuals may represent a silent reservoir of deep P. vivax infections at the sites of active erythropoiesis with very low or no parasitemia and an underestimated biological problem with potential clinical consequences in sub-Saharan Africa.

INCLUSION AND DIVERSITY
We support inclusive, diverse, and equitable conduct of research.mouse antibody (actine, Cell Signaling) at room temperature for 60min.Proteins were detected and visualized using the chemiluminescent detection assay Clarity Max Western ECL (Bio-Rad) and the Chemidoc imaging system (Bio-Rad).Protein quantification was performed using the Image Lab software version 6.1 (Biorad).

Received
Fluorescence microscopy RNA FISH and antibody staining assay Cells were harvested and stained to detect the surface expression of DARC, CD71, and P. vivax-infected erythroblasts using fluorescence microscopy (RNA Stellaris FISH probes specific to P. vivax 18S RNA and anti-Plasmodium HSP70 mouse polyclonal antibody).Briefly, 3310 6 cells were harvested after 24h and 48h of co-culture and washed with PBS (1X) + 0.2% BSA.The cells were fixed in PBS (1X) supplemented with 4% paraformaldehyde and 0.0075% glutaraldehyde at room temperature for 30 min.The cells were then blocked in PBS (1X) + 3% BSA at room temperature for 60 min.They were then incubated with anti-DARC rabbit monoclonal antibody (Thermo Fisher Scientific) and/or anti-CD71 mouse monoclonal antibody (Thermo Fisher Scientific) at 37 C for 60 min in the dark.The cells were washed twice in PBS (1X) + 1% BSA and incubated with Alexa Fluor 555-conjugated anti-rabbit IgG goat monoclonal antibody and Alexa Fluor 488-conjugated anti-mouse IgG goat monoclonal antibody (Invitrogen) at room temperature for 60 min in the dark.The cells were washed twice in PBS (1X) + 1% BSA and permeabilized in PBS (1X) supplemented with 0.1% Triton (X-100) at room temperature for 15 min.Cells were incubated with HSP70 mouse polyclonal antibody (gift from Olivier Silvie, Biology and Immunology of Malaria Research Unit, CIMI) at 37 C for 60 min in the dark, and then washed twice in PBS (1X).Cells were incubated with Alexa Fluor 680-conjugated anti-mouse IgG goat monoclonal antibody (Invitrogen) at room temperature for 60 min in the dark and then washed twice in PBS (1X).The cells were then hybridized using a set of RNA Stellaris FISH probes specific to P. vivax 18S RNA and conjugated to quasar-670 (Stellaris) at 37 C overnight.Cells were washed twice in wash buffer A, incubated at 37 C for 30 min in the dark, and then washed once in wash buffer B. Cells were spun on glass slides at 1000rpm for 3 min using a Cytospin centrifuge (Shandon) and left to dry.They were counterstained using Prolong Gold antifade mountant with diamidino phenylindole (DAPI) (Thermo Fisher Scientific) and covered with coverslips (Epredia).Fluorescence images were captured with a computer-assisted Zeiss Axio Imager.M2 upright microscope equipped with a Plan-Apochromat 63x/1.4NA Oil objective lens (Carl Zeiss).DAPI images (blue) were collected with a 365 nm excitation filter and a 445/50 nm emmission filter HE Green Fluorescent Prot images (green) with a 475/20 nm excitation filter and a 530/25 nm emmission filter, HE Ds Red images (orange) with a 545/25 nm excitation filter and a 605/70 nm emmission filter ), Cy5 images (red) with a 643/30 nm excitation filter and a 690/50 nm emmission filter.The exposure times per channel were kept constant for all the samples to ensure that the intensities could be compared with each other.Brighfield images were collected at exposure time 24.4 ms.Images were captured with a Hamamatsu Camera Orca Flash 4.0lt, controlled, and analyzed using Zen blue software v. 3.7 (Zeiss).A pre-processing pipeline was applied to all 2D images to remove noise, increase contrast, and adjust the dynamic range of the image intensities.For quantitative imaging, we considered the intensity differences when the MFI (Mean Fluorescence Intensity = MFI of the region of interest -MFI of background) was at a minimum x1.5 fold higher with antibody and probe staining.

Statistical analysis
Statistical analyses were performed using GraphPad Prism (GraphPad Software Version 9).Categorical variables were described using frequencies, percentages, and 95% confidence intervals.Chi-square or Fisher's exact tests were used to assess differences in proportion.Continuous variables were summarized as means and standard deviations.The Relative Standard Deviation (RSD) was used to measure the deviation of the proportions disseminated around the mean.For comparisons between two groups, the paired two-tailed t-test or Mann-Whitney test was used.One-way ANOVA variance was performed to compare the three groups.Analysis of covariance (ANCOVA) was used to evaluate changes in the proportion of erythroid cell stages during terminal erythroid differentiation.Statistical significance was considered when the p-value was <0.05.
next page)

Figure 3 .
Figure 3.A subset of genotypically Duffy-negative (DN) erythroblasts expresses DARC during terminal erythroid differentiation (western immunoblotting and immunofluorescence microscopy data) For a Figure360 author presentation of this figure, see https://doi.org/10.1016/j.chom.2023.11.007.(A) Representative western blot analysis of cytoplasmic protein extracts of erythroid cells from one DP (DP007) and one DN (DN005) donor at days 0, 2, 4, 6, 8, 10, and 12 of terminal differentiation and in erythrocytes from peripheral blood for the presence of DARC and actin (control).Bar graphs represent the relative quantification density of the DARC protein levels for DP007 donor (blue) and DN005 (red), compared with the actin control in each condition.(B) Additional western blot analyses of cytoplasmic protein extracts of erythroid cells from three DP (DP001, DP003, and DP004) and three DN (DN001, DN003, and DN002) donors at days 0, 10, and 12 (DP001/DN001), and at days 10 and 12 (DP003/DN003 and DP004/DN002).Bar graph represents the fold-change in the relative quantification density (and the standard deviation of the mean) of the DARC protein levels for DP donors (DP001, DP003, and DP004, in blue, used as the reference set at 1) and DN donors (DN001, DN002, and DN003, in red), compared with the actin control in each condition.The data were obtained from cytoplasmic protein extracts of erythroid cells collected at days 10 and 12. Error bars show the standard deviation.The relative amount of DARC protein was significantly lower in DN erythroblasts at days 10 and 12 of terminal differentiation compared with DP erythroblasts (p = 0.03, Mann-Whitney test).(C) Fluorescence microscopy of paraformaldehyde-fixed DP (DP007 and DP010) and DN (DN005 and DN006) cells on day 9 of terminal differentiation.Bright-field images are shown on the left, and images with DNA (stained with DAPI, blue) and DARC (stained with anti-DARC antibody, orange) are shown on the right.Objective: 633.

Table 1 .
Duffy genotyping of healthy blood donors