1. Genetics and Genomics
  2. Microbiology and Infectious Disease

The Dantu blood group prevents parasite growth in vivo: Evidence from a controlled human malaria infection study

  1. Silvia N Kariuki  Is a corresponding author
  2. Alexander W Macharia
  3. Johnstone Makale
  4. Wilfred Nyamu
  5. Stephen L Hoffman
  6. Melissa C Kapulu
  7. Philip Bejon
  8. Julian C Rayner
  9. Thomas N Williams
  10. On behalf of for the CHMI-SIKA Study Team
  1. Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kenya
  2. Sanaria Incorporate, United States
  3. Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, United Kingdom
  4. Cambridge Institute for Medical Research, University of Cambridge, United Kingdom
  5. Institute for Global Health Innovation, Department of Surgery and Cancer, Imperial College London, United Kingdom
https://doi.org/10.7554/eLife.83874
Share this article
Cite this article
  1. Silvia N Kariuki
  2. Alexander W Macharia
  3. Johnstone Makale
  4. Wilfred Nyamu
  5. Stephen L Hoffman
  6. Melissa C Kapulu
  7. Philip Bejon
  8. Julian C Rayner
  9. Thomas N Williams
  10. On behalf of for the CHMI-SIKA Study Team
(2023)
The Dantu blood group prevents parasite growth in vivo: Evidence from a controlled human malaria infection study
eLife 12:e83874.
https://doi.org/10.7554/eLife.83874
  2. Download BibTeX
  3. Download .RIS

Abstract

Background:

The long co-evolution of Homo sapiens and Plasmodium falciparum has resulted in the selection of numerous human genetic variants that confer an advantage against severe malaria and death. One such variant is the Dantu blood group antigen, which is associated with 74% protection against severe and complicated P. falciparum malaria infections in homozygous individuals, similar to that provided by the sickle haemoglobin allele (HbS). Recent in vitro studies suggest that Dantu exerts this protection by increasing the surface tension of red blood cells, thereby impeding the ability of P. falciparum merozoites to invade them and reducing parasite multiplication. However, no studies have yet explored this hypothesis in vivo.

Methods:

We investigated the effect of Dantu on early phase P. falciparum (Pf) infections in a controlled human malaria infection (CHMI) study. 141 sickle-negative Kenyan adults were inoculated with 3.2 × 103 aseptic, purified, cryopreserved Pf sporozoites (PfSPZ Challenge) then monitored for blood-stage parasitaemia for 21 days by quantitative polymerase chain reaction (qPCR)analysis of the 18S ribosomal RNA P. falciparum gene. The primary endpoint was blood-stage P. falciparum parasitaemia of ≥500/μl while the secondary endpoint was the receipt of antimalarial treatment in the presence of parasitaemia of any density. On study completion, all participants were genotyped both for Dantu and for four other polymorphisms that are associated with protection against severe falciparum malaria: α+-thalassaemia, blood group O, G6PD deficiency, and the rs4951074 allele in the red cell calcium transporter ATP2B4.

Results:

The primary endpoint was reached in 25/111 (22.5%) non-Dantu subjects in comparison to 0/27 (0%) Dantu heterozygotes and 0/3 (0.0%) Dantu homozygotes (p=0.01). Similarly, 49/111 (44.1%) non-Dantu subjects reached the secondary endpoint in comparison to only 7/27 (25.9%) and 0/3 (0.0%) Dantu heterozygotes and homozygotes, respectively (p=0.021). No significant impacts on either outcome were seen for any of the other genetic variants under study.

Conclusions:

This study reveals, for the first time, that the Dantu blood group is associated with high-level protection against early, non-clinical, P. falciparum malaria infections in vivo. Learning more about the mechanisms involved could potentially lead to new approaches to the prevention or treatment of the disease. Our study illustrates the power of CHMI with PfSPZ Challenge for directly testing the protective impact of genotypes previously identified using other methods.

Funding:

The Kenya CHMI study was supported by an award from Wellcome (grant number 107499). SK was supported by a Training Fellowship (216444/Z/19/Z), TNW by a Senior Research Fellowship (202800/Z/16/Z), JCR by an Investigator Award (220266/Z/20/Z), and core support to the KEMRI-Wellcome Trust Research Programme in Kilifi, Kenya (203077), all from Wellcome. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. For the purpose of Open Access, the authors have applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission.

Clinical trial number:

NCT02739763

Editor's evaluation

The large genetic association studies conducted in East Africa have shown that the Dantu blood group provides substantial protection against severe malaria since it increases the surface tension of red blood cells making it harder for malaria parasites to invade. In this important work, the authors show that parasite growth is indeed restricted in vivo by testing this hypothesis in adult Kenyan volunteers infected with P. falciparium under careful monitoring. They were able to show convincingly that indeed, parasite growth was reduced amongst Dantu adults.

https://doi.org/10.7554/eLife.83874.sa0

Introduction

Plasmodium falciparum malaria has been the pre-eminent cause of child morbidity and mortality in the tropics and sub-tropics for much of the last 5000 y. As a consequence, it has had a substantial impact on the human genome through the positive selection of multiple polymorphisms that confer a survival advantage against the disease (Williams, 2017). The best studied affect the biology of red blood cells (RBCs), which host malaria parasites for most of their life cycle in humans, the rs334 A>T βs sickle mutation in HBB (Williams, 2016), α-thalassaemia (Williams et al., 2005), and blood group O (Fry et al., 2008) all being important examples.

Recently, we identified a new variant which is associated with high-level protection against severe P. falciparum malaria to a degree that is close to that of sickle cell trait (HbAS), the strongest malaria-protective condition yet described (Malaria Genomic Epidemiology Network, 2014). The rare Dantu blood group antigen, which results from a genetic rearrangement within the glycophorin (GYP) cluster, was shown to confer 74% protection against severe malaria in homozygous individuals (Band et al., 2015; Ndila et al., 2018). Subsequent in vitro studies have suggested that this protection is explained by the resistance of Dantu RBCs to invasion by P. falciparum merozoites (Kariuki et al., 2020), thereby preventing infections from progressing to become severe or ultimately fatal. However, this hypothesis has not been tested directly in vivo to date.

In this study, we have investigated the impact of the Dantu blood group on in vivo P. falciparum parasite growth and clinical disease progression through a controlled human malaria infection (CHMI) study with aseptic, purified, cryopreserved P. falciparum sporozoites (PfSPZ Challenge) conducted in semi-immune Kenyan adults. To the best of our knowledge, this is the first time that CHMI has been used to directly explore the impact of Dantu on parasite growth in vivo.

Methods

Study design and population

The primary aim of the Kenya CHMI study was to investigate the impacts of naturally acquired immunity on early-phase malaria infections (Kapulu et al., 2018; Kapulu et al., 2021). Briefly, 161 healthy adult volunteers living in areas of varying malaria transmission were inoculated by direct venous inoculation (DVI) with 3.2 × 103 P. falciparum sporozoites (PfSPZ) of Sanaria PfSPZ Challenge (NF54) (Roestenberg et al., 2013; Mordmüller et al., 2015). With a sample size of 161 individuals, we had 80% power to detect a single variable with an effect size (r2) of 0.3 that accounts for 15% of the variability in parasite growth, as previously described (Kapulu et al., 2018; Kapulu et al., 2021). Because of its major impacts on both malaria susceptibility (Allison, 1954) and disease progression (Taylor et al., 2012), and in view of results from a previous CHMI with PfSPZ Challenge conducted in Gabon (Lell et al., 2018), recruitment was restricted to those who were negative for both sickle cell trait and disease. After inoculation, venous blood samples were collected twice daily from days 7–14, and then once every day from day 15 until the end of the experiment on day 21, and screened for parasitaemia by quantitative polymerase chain reaction (qPCR) analysis of the P. falciparum 18S ribosomal RNA gene. For each participant, the endpoint was considered met, and anti-malarial treatment administered, when a threshold of 500 P. falciparum parasites/µl was reached. Participants were treated earlier if signs and symptoms were observed in association with blood film positivity at any parasite density, and on day 21 post-inoculation regardless of outcome. While qPCR was used as the primary measurement of parasitaemia because it is much more sensitive than microscopy at low density (Bejon et al., 2006), thick film blood smears were also performed as an additional precaution, and participants were treated if they became blood film positive at any density, an approach that accords with that used in other CHMI studies (Sheehy et al., 2012; Murphy et al., 2012; Hodgson et al., 2014; Kamau et al., 2014; Hodgson et al., 2015; Seilie et al., 2019; Salkeld et al., 2022). Participants were recruited during 2016, 2017, and 2018 into three successive cohorts from three different malaria transmission zones: Kilifi North (no- to low-transmission) and Kilifi South (moderate transmission), both on the coast, and Ahero (moderate to high transmission) in Western Kenya. We tested for antimalarial drugs as described in a previous publication (Kapulu et al., 2021). Briefly, antimalarial drugs were measured retrospectively in all volunteers using samples collected on both the day before the challenge and 8 days after challenge. Plasma samples were tested by liquid chromatography-tandem mass spectrometry in two independent laboratories. Sulfadoxine, pyrimethamine, and chloroquine levels were measured at the Strathmore University in Nairobi, Kenya, while artemether and dihydroartemisinin concentrations were measured at the Mahidol Oxford Tropical Medicine Research Unit in Bangkok, Thailand. The study was conducted at the KEMRI-Wellcome Trust Research Programme in Kilifi, Kenya, and was registered on ClinicalTrials.gov (NCT02739763).

Genotyping for Dantu and other malaria-protective variants

Whole blood samples were collected at the point of recruitment into tubes containing EDTA and stored at –80°C pending batch processing at the end of the study. Genomic DNA was extracted following the manufacturer’s instructions using a QIAmp 96 DNA QIAcube HT kit on a QIAcube HT System (QIAGEN, Manchester, UK). Genotyping for Dantu was performed using ABI TaqMan SNP genotyping Assays-by-Design primers and probes on an ABI 7900HT PCR machine, as previously described (Kariuki et al., 2009). Dantu genotypes were inferred from the rs186873296 FREM3 allele, which is in strong linkage disequilibrium with the Dantu structural rearrangement (Band et al., 2015). The following primer sequence was used for the rs186873296 SNP: ATGTGAAGAAGCTGGGAACCCTGTC[A/G]TACAAGAAATGACAAAGAAAGCTT, with A being the reference allele. For comparative purposes, we also genotyped participants for a range of other polymorphisms that have been reproducibly associated with protection from severe and complicated malaria in other studies. We typed for the common African form of G6PD deficiency, caused by the G6PD c.202T mutation (Clarke et al., 2017), the blood group O mutation in ABO (Rowe et al., 2007) and the rs4951074 allele in ATP2B4 (Malaria Genomic Epidemiology Network, 2014) using TaqMan SNP genotyping assays, and for the -α3.7I deletional form of α+-thalassaemia by gap PCR (Wambua et al., 2006). To aid ease of comparison across all variants tested, the genotype groups are categorised as ‘Homozygous reference’ for individuals with two copies of the reference allele, ‘Heterozygous’ for individuals with one copy of the reference allele and one copy of the derived allele, and ‘Homozygous derived’ for individuals with two copies of the derived allele.

Statistical analysis

We considered three distinct outcomes, each capturing a different aspect of the parasitological and clinical progress of malaria infections: (a) whether infections progressed to reach the pre-defined treatment threshold of 500 parasites/µl; (b) whether or not participants received malaria treatment for either clinical or parasitological reasons; (c) the time from inoculation to treatment in participants who did receive treatment. We made (a) the primary endpoint for this analysis because the hypothesis we were testing concerned in vivo parasite growth rather than susceptibility to symptoms. We conducted between-genotype comparisons both by univariate analysis and by multivariate analysis with adjustment for other malaria-protective genes, anti-schizont antibody concentration, and location of residence, of which the latter two were significantly associated with the same outcomes in an earlier analysis of the same cohort (Kapulu et al., 2022). We used the Fisher’s exact test using the fmsb package (version 0.7.3) in R to test for any differences in the proportions of individuals that reached the pre-defined treatment threshold of 500 parasites/μl and to test for any differences in the proportions of individuals that required treatment across genotype groups. We also compared the treatment outcomes for the heterozygous- and homozygous-derived genotypes, individually, to the homozygous reference genotype by pairwise analysis.

To test for associations between each variant genotype and the treated or untreated categorical outcome, we used multivariate logistic regression using the glm function in the stats package (version 3.6.2) in R using additive models for each variant, where each variant genotype was coded as zero, one, or two copies of the derived allele. Each multivariable model adjusted for other malaria-protective variants, anti-schizont antibody concentration, and location of residence.

We used the Kruskal–Wallis test (stats package version 3.6.2) and the Dunn’s test (FSA package version 0.9.3) to investigate between-group differences in maximum parasitaemia. Finally, we compared time to treatment using Kaplan–Meier survival curves with univariate comparisons across genotype groups performed using the Log-Rank test, and Cox regression models for multivariate analyses, using the survival package (version 3.2.13) in R. All statistical analyses were performed using R V3.6.2 (R Development Core Team, 2017).

Results

Of 161 volunteers recruited to the study, 19 were excluded, either because non-CHMI parasite strains were detected in post-CHMI samples (suggesting the presence of coincidental natural infections) (n = 7), or because antimalarial drugs were detected in pre-CHMI samples (n = 12) (Figure 1) (Kapulu et al., 2021). After these exclusions, data from 142 individuals contributed to the current analysis. Genotyping revealed that 30 of these individuals were either heterozygous or homozygous for the Dantu allele. We did not get Dantu genotype data on one individual due to poor quality of the DNA sample; therefore, Dantu genotype data on 141 out of the 142 participants was used in the downstream analysis (Figure 1).

Figure 1
Download asset Open asset
Study design and participant recruitment.

The Dantu variant protects against P. falciparum growth in vivo

While infections progressed to the point of reaching the pre-determined treatment threshold of 500 parasites/µl in 25/142 (17.6%) of all volunteers, the proportions varied markedly by Dantu genotype. None of the thirty (0.0%) Dantu carriers reached this threshold in comparison to 25/111 (22.5%) of the non-Dantu individuals (p=0.01) (Figure 2, Table 1). The difference in the proportion of Dantu heterozygous (0/27; 0%) and non-Dantu individuals reaching the treatment threshold was strongly significant (p=0.004) but did not reach statistical significance in the case of Dantu homozygotes 0/3 (0.0%) because of the small number of individuals in this group.

Figure 2
Download asset Open asset
The impact of human genotype on parasite growth.

Following inoculation of volunteers with Pf sporozoites, parasitaemia was monitored by quantitative PCR (y-axis) over the full duration of the study (x-axis). The different panels indicate the specific variants that were studied. The red dots indicate only the individuals that exhibited febrile symptoms and met treatment criteria. The different genotype groups for all the variants are categorised as ‘Homozygous reference’ for individuals with two copies of the reference allele (red lines), ‘Heterozygous’ for individuals with one copy of the reference allele and one copy of the derived allele (green lines), and ‘Homozygous derived’ for individuals with two copies of the derived allele (blue lines). αα/αα, no α-thalassaemia; −α/αα, heterozygous α-thalassaemia; −α/−α, homozygous α-thalassaemia. * For G6PD, male and female were combined as C/CC = normal (wild type hemizygous males and homozygous females), CT = carrier females, and T/TT = G6PD-deficient hemizygous males and homozygous females. All volunteers were typed for all variants, and any one individual may carry a mixture of genotypes – the potential confounding effect of this was controlled for in multivariate analysis. The tables adjacent to each plot show the results from Fisher’s exact tests investigating differences in the proportion of participants that reached the pre-defined treatment threshold of 500 parasites/μl (n) compared to the total number within each genotype category (N).

Figure 2—source data 1

Related to Figure 2.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig2-data1-v1.txt
Figure 2—source data 2

Related to Figure 2 table.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig2-data2-v1.xlsx
Table 1
The proportion of participants reaching the pre-defined treatment parasitaemia threshold of 500 parasites/μl by genotype category.
VariantGenotypen/N%p value overallp value homozygous reference vs. heterozygousp value homozygous reference vs. homozygous derived
Dantu rs186873296Non-Dantu (AA)25/11122.50.010.0041
Heterozygous (AG)0/270.0
Dantu homozygous (GG)0/30.0
G6PD +202 rs1050828Homozygous reference (C/CC)20/11018.20.5050.7390.463
Heterozygous (CT)4/1723.5
Homozygous derived (T/TT)1/156.7
ABO rs8176719Non-O13/6420.30.5170.517-
O12/7516.0
α-thalassaemiaHomozygous reference (αα/αα)6/4812.50.4080.3280.29
Heterozygous (−α/αα)14/7020.0
Homozygous derived (−α/−α)5/2123.8
ATP2B4 rs4951074Homozygous reference (GG)11/6217.70.87610.749
Heterozygous (AG)11/5719.3
Homozygous derived (AA)3/2313.0

  1. n = the number of participants that reached the pre-defined treatment threshold of 500 parasites/μl and were treated; N = the total number within each genotype category; αα/αα, no α-thalassaemia; −α/αα, heterozygous α-thalassaemia; −α/−α, homozygous α-thalassaemia. * For G6PD, male and female were combined as C/CC = normal (wild type hemizygous males and homozygous females), CT = carrier females, and T/TT = G6PD -deficient hemizygous males and homozygous females. We used the Fisher’s exact test to investigate differences in the proportions of individuals that reached the pre-defined treatment threshold of 500 parasites/μl, both for global comparisons across genotype groups, and for separate comparisons between genotype pairs, where the proportion of individuals that reached the treatment threshold of 500 parasites/μl in the heterozygous- and homozygous -derived genotypes were compared to the homozygous reference genotype.

Table 1—source data 1

Related to Table 1.

https://cdn.elifesciences.org/articles/83874/elife-83874-table1-data1-v1.txt

Fewer Dantu carriers required malaria treatment

Although a threshold of 500 parasites/μl was considered the primary endpoint of the study and triggered the administration of antimalarial treatment per-protocol, some participants developed symptoms and were therefore treated before reaching this parasitaemia threshold. Only one quarter (7/27; 25.9%) of the Dantu heterozygotes and none (0/3; 0.0%) of the Dantu homozygotes (Figure 3, Table 2) received antimalarial treatment in comparison to 49/111 (44.1%) of the non-Dantu individuals. While this did not reach statistical significance on univariate analysis, we carried out multivariate regression analysis with the treated or untreated status as the dependent variable, and the Dantu variant as well as other genetic variants (as described below), anti-schizont antibody and location of residence as the independent variables. This analysis revealed that Dantu-carrying subjects overall were administered antimalarial treatment 83% less frequently (OR 0.17; 95% CI 0.04–0.55; p=0.007) than non-Dantu individuals (Table 3). In order to address the core issue of whether prior immunity was a confounder in our analysis, we used measurements of antibodies to whole schizont extract as a proxy indicator of transmission setting or ‘malaria exposure’ in our multivariate analyses. We compared anti-schizont antibody levels across Dantu genotype groups and found no differences (p=0.659) (Figure 3—figure supplement 1).

Figure 3 with 1 supplement see all
Download asset Open asset
The impact of each gene variant on the requirement for malaria treatment.

The proportion of individuals in each genotype category that required treatment over the course of the controlled human malaria infection (CHMI) study is shown on the y-axis. The number of treated individuals out of the total number in each genotype group is given in parenthesis above the bar graphs, while the p values from the Fisher’s exact tests comparing the differences in proportions of individuals that required treatment across genotype groups are also given above the bar graphs.

Figure 3—source data 1

Related to Figure 2.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig3-data1-v1.txt
Table 2
The numbers and frequencies of individuals who received treatment before day 21 by genotypic category.
VariantGenotypen/N%Overall p valueHomozygous reference vs. heterozygousHomozygous reference vs. homozygous derived
Dantu rs186873296Non-Dantu (AA)49/11144.10.100.130.26
Heterozygous (AG)7/2725.9
Dantu homozygous (GG)0/30.0
G6PD +202 rs1050828*Homozygous reference (C/CC)46/11041.80.281.000.16
Heterozygous (CT)7/1741.2
Homozygous derived (T/TT)3/1520.0
ABO rs8176719Non-O28/6443.80.39-0.39
O27/7536.0
α-thalassaemiaHomozygous reference (αα/αα)19/4839.60.790.850.79
Heterozygous (−α/αα)30/7042.9
Homozygous derived (−α/−α)7/2133.3
ATP2B4 rs4951074Homozygous reference (GG)26/6241.90.611.000.45
Heterozygous (AG)23/5740.4
Homozygous derived (AA)7/2330.4
Table 2—source data 1

Related to Table 2.

https://cdn.elifesciences.org/articles/83874/elife-83874-table2-data1-v1.txt
Table 3
The association between each gene variant and the requirement for treatment after challenge.
Overall comparison across genotype groupsHomozygous reference vs. heterozygousHomozygous reference vs. homozygous derived
VariantOdds ratios95% CIp-valueOdds ratios95% CIp valueOdds ratios95% CIp value
Dantu rs1868732960.170.04–0.550.0070.200.04–0.830.0390NA – Inf0.990
G6PD +202 rs10508280.600.28–1.190.1571.710.41–6.590.4420.170.02–0.970.074
ABO rs81767190.400.15–1.030.0640.540.18–1.560.259---
α-thalassaemia0.910.46–1.730.7660.970.31–3.040.9570.870.18–3.220.758
ATP2B4 rs49510740.840.43–1.630.6190.530.17–1.580.2660.570.10–2.580.491
Table 3—source data 1

Related to Table 3.

https://cdn.elifesciences.org/articles/83874/elife-83874-table3-data1-v1.txt

Peak parasitaemias were lower in Dantu-carrying participants

The proportion of participants who became PCR-positive at any parasitaemia was similar at 86/111 (77.5%) in the non-Dantu and 20/27 (74.1%) and 2/3 (66.7%) among Dantu heterozygotes and homozygotes, respectively (p=0.745) (Figure 4). However, maximum parasitaemias were considerably higher in non-Dantu than in Dantu-carrying individuals. Peak parasitaemias reached 9694 parasites/μl in the non-Dantu group in comparison with 411 parasites/μl in the Dantu heterozygotes and only 3 parasites/μl among the Dantu homozygotes (non-Dantu vs. Dantu heterozygotes p=0.028; non-Dantu vs. Dantu homozygotes p=0.141; and non-Dantu vs. Dantu heterozygotes and homozygotes combined p=0.009) (Figure 4). Similarly, the median parasitaemia among those who did become PCR-positive was 112 parasites/μl in the non-Dantu, 13 parasites/μl in the Dantu heterozygous, and 2 parasites/μl in the Dantu homozygous groups, respectively, although these differences did not reach statistical significance (non-Dantu vs. Dantu heterozygotes p=0.108; non-Dantu vs. Dantu homozygotes p=0.256; and non-Dantu vs. combined Dantu heterozygotes and homozygotes p=0.068) (Figure 4).

Figure 4
Download asset Open asset
Peak parasitaemias were lower in Dantu variant carriers.

Maximum parasitaemia values for individuals across Dantu genotype groups, with dashed line indicating the treatment threshold of 500 parasites/μl. The table below the figure shows the numbers and frequencies of individuals in each genotype category that were PCR-positive over the course of the controlled human malaria infection (CHMI) study. n = the number of participants that were PCR-positive; N = the total number within the genotype category. Statistical comparisons of proportions of PCR-positive individuals across genotype groups and pairwise comparisons between genotype groups were performed using the Fisher’s exact test. Statistical comparisons of maximum and median parasitaemia between genotype groups were performed using the Kruskal–Wallis test, and post-hoc Dunn’s test for pairwise differences between the genotype groups.

Figure 4—source data 1

Related to Figure 4.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig4-data1-v1.txt
Figure 4—source data 2

Related to Figure 4 table.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig4-data2-v1.xlsx

Time to treatment was significantly longer in Dantu-carrying than non-Dantu individuals

Among the participants who did receive treatment, the time to treatment was significantly longer among Dantu-carrying than non-Dantu individuals. While the univariate comparisons across genotype groups, performed using the Log-Rank test in the Kaplan–Meier survival curves, did not reach statistical significance (Figure 5a), the multivariate Cox regression analysis with adjustments for other malaria-protective variants, anti-schizont antibody concentration, and location of residence showed that time to treatment was significantly longer among Dantu-carrying than non-Dantu individuals, the overall hazard ratio being 0.39 (CI 0.17–0.87; p=0.022) (Figure 5b). A dose-dependent effect of the Dantu genotype was also seen as none of the three Dantu homozygotes required treatment and the time to treatment was significantly longer in Dantu heterozygous than in non-Dantu individuals (HR = 0.41, p=0.042) (Figure 5b).

Figure 5
Download asset Open asset
Time to treatment was longer in Dantu variant carriers.

The impact of each gene variant on time to treatment was analysed by (a) Kaplan–Meier survival curves, with univariate comparisons across genotype groups performed using the Log-Rank test and (b) multivariate Cox regression models, with each variant genotype coded as zero, one, or two copies of the homozygous derived allele in an additive model, adjusting for the other four malaria-protective variants, anti-schizont antibody concentration, and location of residence. Pairwise analysis compared the time to treatment in the heterozygous- and homozygous-derived genotypes to the homozygous reference genotype.

Figure 5—source data 1

Related to Figure 5a.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig5-data1-v1.txt
Figure 5—source data 2

Related to Figure 5b.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig5-data2-v1.txt
Figure 5—source data 3

Related to Figure 5b table.

https://cdn.elifesciences.org/articles/83874/elife-83874-fig5-data3-v1.xlsx

No significant impacts were seen with regard to any of the outcomes under study for any of the remaining RBC polymorphisms

While our primary analysis focused on the Dantu genotype, individuals in malaria-endemic regions often carry more than one malaria protection-associated genotype (Ndila et al., 2018). We therefore typed individuals for α-thalassaemia, blood group O, G6PD deficiency and ATP2B4 alleles, and carried out the same set of analysis as above for each genotype. Each of these genotypic groups also included, by random chance, individuals of different genotypes for the other variants of interest. As the genotyping was conducted at the end of the study, and the study included only a relatively small number of individuals overall, it was not possible to limit each group to those who were reference homozygotes for all the other variants of interest. Instead, we used multivariate analyses to check that any differences seen on univariate analysis were not explained by the presence of other variants or confounders. No significant differences were seen in any of the study outcomes between the different genotype groups (Figures 2, 3 and 5, Tables 1–3). As noted above, the impact of Dantu on malaria treatment was independent of the presence or absence of these other genetic variants in multivariate analysis.

Discussion

Through the analysis of data from a CHMI study with PfSPZ Challenge injection conducted in Kenyan adults, we have shown that the Dantu blood group is associated with prevention of parasite growth in vivo. While more than 20% of Dantu-negative volunteers developed bloodstream malaria infections that reached a pre-defined threshold of 500 parasites/μl following controlled PfSPZ administration, this threshold was not reached by any of the 30 Dantu-positive volunteers. This is the first time that Dantu has been shown to protect against early-stage malaria infections. Our study suggests that Dantu protects against severe and complicated malaria (Ndila et al., 2018; Band et al., 2015) by preventing the disease from becoming established in its earliest phase. This is consistent with recent in vitro observations that have demonstrated a link between Dantu genotype and susceptibility to red blood cell invasion by P. falciparum merozoites, which we previously predicted would lead to reduced parasite growth in vivo (Kariuki et al., 2020). That mechanistic study revealed that the Dantu genotype protects red blood cells from invasion by increasing their surface tension, which reduces the ability of merozoites to deform their surface and hence productively invade. Critically, this tension difference was greater in Dantu homozygotes than heterozygotes, as was the reduction in invasion efficiency, supporting a dose-dependent protective effect. This dose-dependency was similarly reflected in our current in vivo study, as while 44.1% of the Dantu-negative volunteers developed symptoms that precipitated malaria treatment, this occurred in only 25.9% of Dantu heterozygous and in none of the three (0%) Dantu homozygous volunteers. Similarly, the maximum parasitaemia observed was considerably higher in the non-Dantu than in the Dantu heterozygous and homozygous participants.

In contrast to CHMI studies in malaria-naïve populations, where most participants typically develop clinical malaria (Church et al., 1997), less than one-fifth of the participants in this study reached a threshold of >500 parasites/μl. This is probably because most of the participants in our study were residents of malaria-endemic communities and will therefore have been partially immune to the disease. Indeed, in previous analyses of data from the same study we have shown that infection outcome was attributable to the degree of prior exposure as estimated by the titre of anti-schizont antibodies (Kapulu et al., 2022), as well as other measures of exposure including the antibody-dependent phagocytosis of both ring-infected and uninfected erythrocytes from parasite cultures (Musasia et al., 2022) and the breadth of antibodies to P. falciparum Variant Surface Antigens (Kimingi et al., 2022). As such, Dantu genotype was therefore clearly not the only factor at play in determining the clinical outcome in this study. However, our multivariate analysis adjusted for these factors (Kapulu et al., 2022) as well as other malaria-protective genetic variants, and the significant association in that analysis underscores the strong protective effect conferred by Dantu. There were also no differences observed in anti-schizont antibody levels across Dantu genotype groups, suggesting that differences in pre-existing anti-malaria immunity between Dantu and non-Dantu cannot explain the differences seen in this study.

The protective impact of Dantu against in vivo parasite growth was in stark contrast to that of the other genetic factors under study. Although consistent evidence has been found for protective effects against severe malaria by G6PD deficiency, blood group O, the rs4951074 allele in ATP2B4 and α+-thalassaemia in numerous previous studies (Malaria Genomic Epidemiology Network, 2014; Taylor et al., 2012), none of these polymorphisms had any significant impacts on any of the outcomes under investigation in this study. This is probably because, unlike Dantu, none of these conditions have a clearly established impact on merozoite red cell invasion, but instead influence the progress of malaria once the disease has become established. For example, recent studies have shown that α+-thalassaemia has no effect on either red blood cell invasion (Williams et al., 2002) or the development of uncomplicated malaria (Taylor et al., 2012; Wambua et al., 2006), but protects instead against the development of severe and complicated disease through mechanisms that include the reduced expression of red cell surface antigens that result in cytoadhesion (Krause et al., 2012; Opi et al., 2014). Similarly, it had no apparent impact on infectivity in a previous, smaller, CHMI study using PfSPZ Challenge, conducted in Tanzania (Shekalaghe et al., 2014). Because clinical guidelines mean that controlled human malaria challenge studies only ever reach relatively low-density parasitaemias, they may not adequately capture the impacts of genetic factors that influence the later, more severe stages of malaria disease. However, this study shows that they can be helpful in pointing towards the pathways leading to such outcomes for further study by other methods.

In conclusion, this study reveals the power of CHMI studies to deconvolute the malaria-protective effects of naturally occurring human genetic variants, establishes for the first time that the Dantu blood group provides strong protection against in vivo parasite growth, and emphasises the potential of Dantu-phenocopying interventions to limit P. falciparum growth in vivo.

Appendix 1

Members of the CHMI-SIKA Study Team

  • Centre for Geographic Medicine Research (Coast), Kenya Medical Research Institute-Wellcome Trust Research Programme, Kilifi, Kenya: Abdirahman I Abdi, Philip Bejon, Zaydah de Laurent, Mainga Hamaluba, Domtila Kimani, Rinter Kimathi, Kevin Marsh, Sam Kinyanjui, Khadija Said Mohammed, Moses Mosobo, Janet Musembi, Jennifer Musyoki, Michelle Muthui, Jedidah Mwacharo, Kennedy Mwai, Joyce M Ngoi, Omar Ngoto, Patricia Njuguna, Irene Nkumama, Francis Ndungu, Dennis Odera, Donwilliams Omuoyo, Faith Osier, Edward Otieno, Jimmy Shangala, James Tuju, Juliana Wambua, Thomas N Williams

  • Sanaria Inc, Rockville, MD, United States: Yonas Abebe, Peter F Billingsley, Stephen L Hoffman, Eric R James, Thomas L Richie, B Kim Lee Sim

  • Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University Oxford, Oxford, United Kingdom: Sam Kinyanjui, Kevin Marsh

  • Department of Pathology, University of Cambridge, Cambridge, United Kingdom: Peter C Bull

  • Pwani University, P. O. Box 195-80108, Kilifi, Kenya: Sam Kinyanjui, Cheryl Kivisi

  • Epidemiology and Biostatistics Division, School of Public Health, University of the Witwatersrand, Johannesburg, South Africa: Kennedy Mwai

  • Centre for Infectious Diseases, Heidelberg University Hospital, Heidelberg, Germany: Irene Nkumama, Dennis Odera, Faith Osier

  • Center for Research in Therapeutic Sciences, Strathmore University, Nairobi, Kenya: Bernhards Ogutu, Fredrick Olewe, John Ong’echa

  • Center for Research in Therapeutic Sciences, Strathmore University, Nairobi, Kenya: Bernhards Ogutu, Fredrick Olewe, John Ong’echa

Data availability

All data generated or analysed during this study are included in the manuscript and supporting file; Source Data files have been provided for Figures 2, 3, 4, and 5 and Tables 1, 2 and 3.

References

  24. Software
    1. R Development Core Team
    (2017) R: A language and environment for statistical computing
    R Foundation for Statistical Computing, Vienna, Austria.
    https://www.r-project.org
  36. Book
    1. Williams TN
    (2017)
    Host Genetics
    In: Gaur G, editors. Advances in Malaria Research. John Wiley & Sons, Inc. pp. 465–494.

Decision letter

  1. Amy Wesolowski
    Reviewing Editor; Johns Hopkins Bloomberg School of Public Health, United States
  2. Jos W van der Meer
    Senior Editor; Radboud University Medical Centre, Netherlands
  3. Wiebke Nahrendorf
    Reviewer; University of Edinburgh, United Kingdom
  4. Nicholas J White
    Reviewer; Mahidol University, Thailand

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "The Dantu blood group provides high level protection against uncomplicated malaria: evidence from a Controlled Human Malaria Infection study" for consideration by eLife. Your article has been reviewed by 2 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Jos van der Meer as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Wiebke Nahrendorf (Reviewer #1); Nicholas J White (Reviewer #2). The reviewers have some suggestions to take into account when you resubmit.

Reviewer #1 (Recommendations for the authors):

Nice short paper showing that Dantu prevents parasite growth in vivo. I think it is important to be really clear that the mechanism of protection is that red cells are rigid so parasites cannot get in. This makes a lovely clear-cut story so I would recommend changing the title to something like "The Dantu blood group prevents parasite growth in vivo." – simple (and exciting!).

At present the title and some of the wording around uncomplicated malaria is a little misleading: of course if there are no malaria parasites no malaria symptoms are experienced. "Protection from uncomplicated malaria" would to me imply that at identical parasite densities Dantu individuals experience less symptoms than non-Dantu – but this is of course not what the paper shows (and the study would not be powered to do so).

Below a few comments on how the data is presented – my recommendation is to incorporate the key bits of information from Tables in the Figures as it is hard to interpret dissociated information. Also more informative labels and legends (include details of statistical test and n) in the Figures would help them be more self-explanatory methods: more details for genotyping – primer sequences etc. what is the reference allele? more details & references for R packages used for multivariate analysis. define fixed and random effects and all response variables.

Figure 2: More intuitive labelling: in text referred to as "non Dantu" "Dantu heterozygotes" and "Dantu homozygotes" – keep this in Figures for consistency (also for the other red cell polymorphisms). Why does the red group look different between the different polymorphisms? Does that mean e.g. G6PD contains Dantu heterozygotes/homozygotes? Should all polymorphisms be excluded? It was not clear to me what the control group is ("homozygous reference" is not mentioned in methods or elsewhere in text).

To help interpretation the number of participants in each graph should be included. My suggestion would be to incorporate the information of Table 1 into Figure 2: n could be indicated above each existing parasite growth curve graph, a small Figure (a bit like Figure 3) showing proportion of individuals that reached the parasitaemia threshold would fit next to it and could include p values. Bit more information in the legend would help the Figure to be more self-explanatory: "Following inoculation of volunteers with Pf sporozoites parasitaemia was monitored by qPCR. Here split by red cell polymorphisms. The control (red) is xyz…".

I was also not quite clear on the red dots: do they indicate all treated participants – so when they either crossed 500 p/ul threshold or had any parasites + symptoms? In which case why are there some very high parasitaemias with no red dot? Or just febrile participants? Consider if this is the right place for this information given that this Figure is about parasite growth.

Table 1: My suggestion would be to incorporate the key info into corresponding Figure 2. The heading needs to make it to clear that this is related to parasite growth – the "predefined threshold" is 500 parasites/ul. I also think n is the number of participants being treated (not "left untreated" as stated). Please check the p values and indicate the statistical test/multiple testing correction used in legend – how can Dantu p-value Homozygous Reference vs Homozygous Derived be 1…? I imagine a Wilcoxon rank sum exact test (two-tailed) would be most appropriate.

Figure 3: Group labelling like Figure 2 (e.g. "non Dantu"). Heading a bit misleading – the outcome of treatment is not considered. If n in Figure 2 does not need to be repeated here above each bar. Instead incorporate p values from Table 2. In text explain what variables were adjusted for and provide a formula for your multivariate regression analysis: which dependent and independent variables did you use?

Table 2: incorporate information in Figure 3. Like in Table 1 n is the number of treated (not untreated) participants.

Table 3: Again I think "treatment outcome" in title is misleading. The multivariate model and which variables were adjusted for needs to be clearly stated in legend and Results section (currently in methods only but with not enough details).

Figure 4: descriptive title matching subheading in results might be better: "Lower peak parasitaemia in Dantu-carrying participants". Axis labels and dots too small. Information from Table 4 could again be easily contained within this Figure.

Table 4: transfer key information to Figure 4. (or include a small Table in Figure 4)

Figure 5: Group labelling like Figure 2/3, axis labels too small. Condense the information provided within the table or at least highlight the key bits of information.

Figure 5 S1: Group labelling like Figure 2/3/5. I think this should go into Figure 5.

Reviewer #2 (Recommendations for the authors):

This is a very nice study. The result is clear and the data on the other genetic human polymorphisms are also very interesting!

Here are a few points to consider:

1. qPCR is a lot less accurate than microscopy. Was the primary endpoint determined by microscopy also and if so how did the two methods compare?

2. Obvious questions, but I presume that other protective genetic polymorphisms were not linked, and the Dantu blood groups were not clustered in an ethnic group with more "immunity" i.e. from a focus of higher transmission? I think both were addressed -but wish to check.

3. Ideally growth rates would have been calculated but it looks as if the data are too noisy. Is that correct?

4. Have you counted merozoites per schizont and asexual cycle lengths in in-vitro cultures of P. falciparum infected Dantu blood group red cells to confirm there is no growth inhibition ?

5. How were blood samples screened for antimalarial drugs?

https://doi.org/10.7554/eLife.83874.sa1

Author response

Reviewer #1 (Recommendations for the authors):

Nice short paper showing that Dantu prevents parasite growth in vivo. I think it is important to be really clear that the mechanism of protection is that red cells are rigid so parasites cannot get in. This makes a lovely clear-cut story so I would recommend changing the title to something like "The Dantu blood group prevents parasite growth in vivo." – simple (and exciting!).

We thank the reviewer for this helpful comment. Following the reviewer’s suggestion, we have now changed the title to: The Dantu blood group prevents parasite growth in vivo: evidence from a Controlled Human Malaria Infection study.

At present the title and some of the wording around uncomplicated malaria is a little misleading: of course if there are no malaria parasites no malaria symptoms are experienced. "Protection from uncomplicated malaria" would to me imply that at identical parasite densities Dantu individuals experience less symptoms than non-Dantu – but this is of course not what the paper shows (and the study would not be powered to do so).

We agree that our study does not compare symptomatology across genotype groups. We have now removed wording about uncomplicated malaria from the title and the rest of the manuscript as suggested by this reviewer.

Below a few comments on how the data is presented – my recommendation is to incorporate the key bits of information from Tables in the Figures as it is hard to interpret dissociated information. Also more informative labels and legends (include details of statistical test and n) in the Figures would help them be more self-explanatory methods: more details for genotyping – primer sequences etc. what is the reference allele? more details & references for R packages used for multivariate analysis. define fixed and random effects and all response variables.

Thank you for this helpful comment. We have now added details of the primer/probe sequences used to genotype the Dantu marker SNP, as well as the reference allele, to the Methods section of the revised paper on page 6. The multivariate analysis was performed with the glm function in the stats package (version 3.6.2), with the response variables being the Dantu marker SNP, other malaria-protective SNPs, anti-schizont antibody levels and area of residence. We have now included details of this and other packages within R that we used in our analyses to the statistical analysis section on page 7.

Figure 2: More intuitive labelling: in text referred to as "non Dantu" "Dantu heterozygotes" and "Dantu homozygotes" – keep this in Figures for consistency (also for the other red cell polymorphisms).

Thank you for this. We previously used “homozygous reference”, “heterozygous” and “homozygous derived” as a generalised labelling for the genotype categories in each of the different red cell polymorphisms. In the light of this comment, we have now clarified these genotype categories in the legend to Figure 2, while also amending the description for the Dantu group within this Figure for ease of reference. We hope this addresses this point to the reviewer’s satisfaction.

Why does the red group look different between the different polymorphisms?

In this Figure, the group coloured in red are the homozygous reference groups for each gene under investigation. In all cases, this group contains the largest number of individuals, but the precise number of individuals will differ depending on the minor allele frequency of each specific gene – this explains why the red group looks different in each case.

Does that mean e.g. G6PD contains Dantu heterozygotes/homozygotes? Should all polymorphisms be excluded?

As noted above, in this Figure the groups coloured in red are the homozygous reference individuals for each gene under investigation. Each of these genotypic groups will also include (by random chance) individuals of different genotypes for the other variants of interest. As the genotyping was conducted at the end of the study, and the study included only a relatively small number of individuals overall, it was not possible to limit each group to those who were reference homozygotes for all the other variants of interest. Instead, we used multivariable analyses to check that any differences seen on univariate analysis were not explained by the presence of other variants or confounders; these multivariate analyses clearly showed that the effect of Dantu was not due to such confounders. We hope this addresses the reviewer’s query. We have amended the text to address this point on page 11 of the revised manuscript.

It was not clear to me what the control group is ("homozygous reference" is not mentioned in methods or elsewhere in text).

No “control” group was included in this specific analysis. The overall p-values presented in the table were derived using Fisher’s exact tests, that were used to test for differences in the proportions of individuals that reached the pre-defined treatment threshold of 500 parasites/ml across ALL genotype groups – i.e. a global rather than a paired approach. We have added an explanation about the statistical methods involved both in the footnote to Figure 2 and in the Methods section on page 7.

To help interpretation the number of participants in each graph should be included. My suggestion would be to incorporate the information of Table 1 into Figure 2: n could be indicated above each existing parasite growth curve graph, a small Figure (a bit like Figure 3) showing proportion of individuals that reached the parasitaemia threshold would fit next to it and could include p values. Bit more information in the legend would help the Figure to be more self-explanatory: "Following inoculation of volunteers with Pf sporozoites parasitaemia was monitored by qPCR. Here split by red cell polymorphisms. The control (red) is xyz…".

Thank you for this suggestion. We have now modified Figure 2 as suggested to include more information both on each graph and in the legend.

I was also not quite clear on the red dots: do they indicate all treated participants – so when they either crossed 500 p/ul threshold or had any parasites + symptoms? In which case why are there some very high parasitaemias with no red dot? Or just febrile participants? Consider if this is the right place for this information given that this Figure is about parasite growth.

The red dots indicate just the febrile participants (not all of them exhibited febrile symptoms even at the very high parasitaemias). We have now added an explanation about this point to the Figure legend.

Table 1: My suggestion would be to incorporate the key info into corresponding Figure 2. The heading needs to make it to clear that this is related to parasite growth – the "predefined threshold" is 500 parasites/ul. I also think n is the number of participants being treated (not "left untreated" as stated). Please check the p values and indicate the statistical test/multiple testing correction used in legend – how can Dantu p-value Homozygous Reference vs Homozygous Derived be 1…? I imagine a Wilcoxon rank sum exact test (two-tailed) would be most appropriate.

Due to the low sample sizes in the different genotype categories, we used the Fisher’s exact test to compare the proportions of individuals that reached the pre-defined treatment threshold of 500 parasites/ml (see the numbers in the table below). Both global comparisons across genotype groups and pairwise comparisons between genotype groups were done using the Fisher’s exact test. We did not correct for multiple testing in the pairwise analyses because we did separate comparisons of the proportion of individuals that reached the treatment threshold of 500 parasites/ml in the heterozygous and homozygous derived genotypes to the homozygous reference genotype. We have updated the footnote to Author response table 1 to explain the statistical methods that were used in better detail.

Author response table 1
Dantu genotype groupBelow 500 parasites/ulAbove 500 parasites/ulSum
Non-Dantu (AA)8625111
Dantu Heterozygote (AG)27027
Dantu Homozygote303

Contingency table for maximum PCR values categorised as below 500 parasites/ul and above 500 parasites/ul

Figure 3: Group labelling like Figure 2 (e.g. "non Dantu"). Heading a bit misleading – the outcome of treatment is not considered. If n in Figure 2 does not need to be repeated here above each bar. Instead incorporate p values from Table 2. In text explain what variables were adjusted for and provide a formula for your multivariate regression analysis: which dependent and independent variables did you use?

Thank you for this. We have now added more information in this Figure about the numbers of individuals in each genotype group that required treatment, as well as the p-values from the Fisher’s exact tests comparing the differences in proportions of individuals that required treatment across genotype groups. The results from the multivariate regression are listed in Table 3. We have added further details to the Methods section on pages 7 and 8 about the Fisher’s exact test and multivariate regression models used.

Table 2: incorporate information in Figure 3. Like in Table 1 n is the number of treated (not untreated) participants.

Our apologies for this. We have corrected the information on the number of treated participants in the legend and included the p-values comparing the proportions of individuals requiring treatment across genotypes in Figure 3.

Table 3: Again I think "treatment outcome" in title is misleading. The multivariate model and which variables were adjusted for needs to be clearly stated in legend and Results section (currently in methods only but with not enough details).

We agree with this point. We have now edited the title and legend to make it clearer that it is the impact of each gene variant on treatment requirement that we are analysing. We have also added further details to the legend and the Results section on page 9 regarding the multivariate models used.

Figure 4: descriptive title matching subheading in results might be better: "Lower peak parasitaemia in Dantu-carrying participants". Axis labels and dots too small. Information from Table 4 could again be easily contained within this Figure.

We thank the reviewer for this helpful suggestion. We have now edited the subheading of Figure 4 and incorporated the information from Table 4 into this figure accordingly.

Table 4: transfer key information to Figure 4. (or include a small Table in Figure 4)

Thank you. We have now included a small Table in Figure 4.

Figure 5: Group labelling like Figure 2/3, axis labels too small. Condense the information provided within the table or at least highlight the key bits of information.

Figure 5 S1: Group labelling like Figure 2/3/5. I think this should go into Figure 5.

We have now edited Figure 5 into a multi-part figure including Figure 5 S1.

Reviewer #2 (Recommendations for the authors):

This is a very nice study. The result is clear and the data on the other genetic human polymorphisms are also very interesting!

Here are a few points to consider:

1. qPCR is a lot less accurate than microscopy. Was the primary endpoint determined by microscopy also and if so how did the two methods compare?

Because of the clinical guidelines around CHMI studies, participants need to be treated very early, meaning parasites only ever reach a relatively low density. In a previous publication (Bejon et al. Malaria J 2006, new Ref. # 16), parasite density measurements by qPCR and microscopy were compared using serially diluted parasite cultures of known parasitaemia. This study clearly showed that qPCR gives more accurate readings of parasite numbers at low densities, while thick blood films are less reproducible. As a consequence, we used qPCR as the primary endpoint measurement of parasitaemia in our current study, and it is also routinely used as the primary endpoint in CHMI studies conducted by other groups around the world (Refs. # 17 – 23). Thick blood smears were also performed but this was as an additional clinical precaution, with participants being treated if they became blood film positive at any density. We have added more information in the Methods section on page 5 to clarify this point, and added the references discussed here.

2. Obvious questions, but I presume that other protective genetic polymorphisms were not linked, and the Dantu blood groups were not clustered in an ethnic group with more "immunity" i.e. from a focus of higher transmission? I think both were addressed -but wish to check.

While all 3 Dantu homozygotes came from the same location (Kilifi South), as noted in the response to Reviewer 1, we compared anti-schizont antibody levels of all individuals to assess “immunity” and used the readings as a factor in our multivariate analysis. We found no differences in anti-schizont antibody levels, a correlate of pre-existing immunity to malaria, between the genetic groups. We have now included these data in Figure 3 —figure supplement 1.

3. Ideally growth rates would have been calculated but it looks as if the data are too noisy. Is that correct?

This is correct. Growth rates were calculated in a previous study (Kapulu et al. BMC Infect Dis, 2022; Ref. # 21), and they were indeed found to be too noisy to allow for meaningful further analysis. This is because, to a variable extent, all of the participants had been previously exposed to malaria infections and their growth rate measurement data varied very significantly as a result, compared to what you would expect among malaria-naïve individuals such as those used in CHMI malaria vaccine trials. Of course, it would be next to impossible to perform CHMI to explore malaria-protective genotypes, which are enriched in African populations, with completely malaria-naïve individuals. We have not changed the manuscript in this regard but would be happy to add a comment in the discussion if helpful.

4. Have you counted merozoites per schizont and asexual cycle lengths in in-vitro cultures of P. falciparum infected Dantu blood group red cells to confirm there is no growth inhibition ?

This is a good suggestion but, unfortunately, because we used qPCR, rather than microscopy, as our primary method for tracking parasitaemia, we are not able to count merozoites per schizont and asexual cycle lengths in infected Dantu variant red cells in this study. This would require further in vitro experiments rather than this study’s in vivo experimental design, where for clinical reasons, treatment intervention guidelines mean that parasitaemia levels are never sufficiently high to generate such data.

5. How were blood samples screened for antimalarial drugs?

Antimalarial drugs were measured retrospectively in all volunteers at two time points, the day before the challenge and eight days after challenge. Their plasma samples were tested in two independent laboratories: Strathmore University in Nairobi, Kenya, which measured sulfadoxine, pyrimethamine, and chloroquine levels; and Mahidol Oxford Tropical Medicine Research Unit in Bangkok, Thailand, which measured artemether and dihydroartemisinin concentrations. All drug concentrations were measured by liquid chromatography-tandem mass spectrometry. These methods were detailed in a previous publication (Kapulu et al. JCI Insight 2021; Ref. # 21). We have now referenced this paper more clearly in the revised manuscript and have also summarised this in the Methods section on page 6.

In addition to these specific changes we have made in response to the reviewer comments, on reviewing our analyses we note that, unfortunately, we found a minor error in the script underlying the multivariable analysis included in Table 3. We have now corrected that error and re-run the analysis which made no material difference to our initial conclusions. The revised p-values are now shown in the updated table. Finally, we have also made a number of small changes to deal with typographical errors and issues of clarity.

https://doi.org/10.7554/eLife.83874.sa2

Article and author information

Author details

  1. Silvia N Kariuki

    Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
    Contribution
    Conceptualization, Data curation, Formal analysis, Investigation, Visualization, Methodology, Writing - original draft, Writing – review and editing
    For correspondence
    SNKariuki@kemri-wellcome.org
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-0801-5285

  2. Alexander W Macharia

    Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
    Contribution
    Investigation, Methodology
    Competing interests
    No competing interests declared

  3. Johnstone Makale

    Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
    Contribution
    Investigation, Methodology
    Competing interests
    No competing interests declared

  4. Wilfred Nyamu

    Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
    Contribution
    Investigation
    Competing interests
    No competing interests declared

  5. Stephen L Hoffman

    Sanaria Incorporate, Rockville, United States
    Contribution
    Conceptualization, Resources, Methodology, Writing – review and editing
    Competing interests
    SLH and members of the CHMI-SIKA Study Team (YA, PFB, ERJ, TLR, and BKLS; see Author Information) are salaried, full-time employees of Sanaria, the manufacturer of Sanaria PfSPZ Challenge. The authors have no additional financial interests

  6. Melissa C Kapulu

    1. Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
    2. Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, United Kingdom
    Contribution
    Conceptualization, Resources, Investigation, Methodology, Project administration, Writing – review and editing
    Competing interests
    Melissa C Kapulu has received grants from UKRI MRC UK and Wellcome Trust and has received travel expenses for the HIC-VAC Annual Meeting, Greenwood Africa Prize and WHO working group meeting on challenge studies. The author has no other competing interests to declare

  7. Philip Bejon

    1. Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
    2. Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, United Kingdom
    Contribution
    Conceptualization, Resources, Formal analysis, Supervision, Methodology, Writing – review and editing
    Competing interests
    No competing interests declared

  8. Julian C Rayner

    Cambridge Institute for Medical Research, University of Cambridge, Cambridge, United Kingdom
    Contribution
    Conceptualization, Supervision, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-9835-1014

  9. Thomas N Williams

    1. Centre for Geographic Medicine Research (Coast), KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
    2. Institute for Global Health Innovation, Department of Surgery and Cancer, Imperial College London, London, United Kingdom
    Contribution
    Conceptualization, Resources, Supervision, Methodology, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-4456-2382

Funding

Wellcome Trust (107499)

  • Melissa C Kapulu
  • Philip Bejon

Wellcome Trust (216444/Z/19/Z)

  • Silvia N Kariuki

Wellcome Trust (202800/Z/16/Z)

  • Thomas N Williams

Wellcome Trust (220266/Z/20/Z)

  • Julian C Rayner

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. For the purpose of Open Access, the authors have applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission.

Ethics

Clinical trial registration: The study is registered on ClinicalTrials.gov (NCT02739763).

The study was approved by both the KEMRI Scientific and Ethics Review Unit (protocol KEMRI/SERU/CGMR-C/029/3190) in Kenya, and the University of Oxford Tropical Research Ethics Committee (OxTREC; protocol 2-16) in the UK. The study was conducted based on good clinical practice (GCP) following the principles of the Declaration of Helsinki.

Senior Editor

  1. Jos W van der Meer, Radboud University Medical Centre, Netherlands

Reviewing Editor

  1. Amy Wesolowski, Johns Hopkins Bloomberg School of Public Health, United States

Reviewers

  1. Wiebke Nahrendorf, University of Edinburgh, United Kingdom
  2. Nicholas J White, Mahidol University, Thailand

Version history

  1. Preprint posted: September 22, 2022 (view preprint)
  2. Received: September 30, 2022
  3. Accepted: May 11, 2023
  4. Version of Record published: June 13, 2023 (version 1)

Copyright

© 2023, Kariuki et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 461
    Page views
  • 22
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Silvia N Kariuki
  2. Alexander W Macharia
  3. Johnstone Makale
  4. Wilfred Nyamu
  5. Stephen L Hoffman
  6. Melissa C Kapulu
  7. Philip Bejon
  8. Julian C Rayner
  9. Thomas N Williams
  10. On behalf of for the CHMI-SIKA Study Team
(2023)
The Dantu blood group prevents parasite growth in vivo: Evidence from a controlled human malaria infection study
eLife 12:e83874.
https://doi.org/10.7554/eLife.83874

Categories and tags

Research organisms

Further reading

    1. Epidemiology and Global Health
    2. Genetics and Genomics

    Within-host diversity improves phylogenetic and transmission reconstruction of SARS-CoV-2 outbreaks

    Arturo Torres Ortiz, Michelle Kendall ... Louis Grandjean
    Research Article

    Accurate inference of who infected whom in an infectious disease outbreak is critical for the delivery of effective infection prevention and control. The increased resolution of pathogen whole-genome sequencing has significantly improved our ability to infer transmission events. Despite this, transmission inference often remains limited by the lack of genomic variation between the source case and infected contacts. Although within-host genetic diversity is common among a wide variety of pathogens, conventional whole-genome sequencing phylogenetic approaches exclusively use consensus sequences, which consider only the most prevalent nucleotide at each position and therefore fail to capture low frequency variation within samples. We hypothesized that including within-sample variation in a phylogenetic model would help to identify who infected whom in instances in which this was previously impossible. Using whole-genome sequences from SARS-CoV-2 multi-institutional outbreaks as an example, we show how within-sample diversity is partially maintained among repeated serial samples from the same host, it can transmitted between those cases with known epidemiological links, and how this improves phylogenetic inference and our understanding of who infected whom. Our technique is applicable to other infectious diseases and has immediate clinical utility in infection prevention and control.

    1. Cancer Biology
    2. Genetics and Genomics

    Complementary CRISPR screen highlights the contrasting role of membrane-bound and soluble ICAM-1 in regulating antigen specific tumor cell killing by cytotoxic T cells

    Ann-Kathrin Herzfeldt, Marta Puig Gamez ... Lee Kim Swee
    Research Article

    Cytotoxic CD8+ T lymphocytes (CTLs) are key players of adaptive anti-tumor immunity based on their ability to specifically recognize and destroy tumor cells. Many cancer immunotherapies rely on unleashing CTL function. However, tumors can evade killing through strategies which are not yet fully elucidated. To provide deeper insight into tumor evasion mechanisms in an antigen-dependent manner, we established a human co-culture system composed of tumor and primary immune cells. Using this system, we systematically investigated intrinsic regulators of tumor resistance by conducting a complementary CRISPR screen approach. By harnessing CRISPR activation (CRISPRa) and CRISPR knockout (KO) technology in parallel, we investigated gene gain-of-function as well as loss-of-function across genes with annotated function in a colon carcinoma cell line. CRISPRa and CRISPR KO screens uncovered 187 and 704 hits respectively, with 60 gene hits overlapping between both. These data confirmed the role of interferon‑γ (IFN-γ), tumor necrosis factor α (TNF-α) and autophagy pathways and uncovered novel genes implicated in tumor resistance to killing. Notably, we discovered that ILKAP encoding the integrin-linked kinase-associated serine/threonine phosphatase 2C, a gene previously unknown to play a role in antigen specific CTL-mediated killing, mediate tumor resistance independently from regulating antigen presentation, IFN-γ or TNF-α responsiveness. Moreover, our work describes the contrasting role of soluble and membrane-bound ICAM-1 in regulating tumor cell killing. The deficiency of membrane-bound ICAM-1 (mICAM-1) or the overexpression of soluble ICAM-1 (sICAM-1) induced resistance to CTL killing, whereas PD-L1 overexpression had no impact. These results highlight the essential role of ICAM-1 at the immunological synapse between tumor and CTL and the antagonist function of sICAM-1.

    1. Genetics and Genomics
    2. Neuroscience

    Methylglyoxal-derived hydroimidazolone, MG-H1, increases food intake by altering tyramine signaling via the GATA transcription factor ELT-3 in Caenorhabditis elegans

    Muniesh Muthaiyan Shanmugam, Jyotiska Chaudhuri ... Pankaj Kapahi
    Research Article

    The Maillard reaction, a chemical reaction between amino acids and sugars, is exploited to produce flavorful food ubiquitously, from the baking industry to our everyday lives. However, the Maillard reaction also occurs in all cells, from prokaryotes to eukaryotes, forming Advanced Glycation End-products (AGEs). AGEs are a heterogeneous group of compounds resulting from the irreversible reaction between biomolecules and α-dicarbonyls (α-DCs), including methylglyoxal (MGO), an unavoidable byproduct of anaerobic glycolysis and lipid peroxidation. We previously demonstrated that Caenorhabditis elegans mutants lacking the glod-4 glyoxalase enzyme displayed enhanced accumulation of α-DCs, reduced lifespan, increased neuronal damage, and touch hypersensitivity. Here, we demonstrate that glod-4 mutation increased food intake and identify that MGO-derived hydroimidazolone, MG-H1, is a mediator of the observed increase in food intake. RNAseq analysis in glod-4 knockdown worms identified upregulation of several neurotransmitters and feeding genes. Suppressor screening of the overfeeding phenotype identified the tdc-1-tyramine-tyra-2/ser-2 signaling as an essential pathway mediating AGEs (MG-H1) induced feeding in glod-4 mutants. We also identified the elt-3 GATA transcription factor as an essential upstream regulator for increased feeding upon accumulation of AGEs by partially controlling the expression of tdc-1 gene. Further, the lack of either tdc-1 or tyra-2/ser-2 receptors suppresses the reduced lifespan and rescues neuronal damage observed in glod-4 mutants. Thus, in C. elegans, we identified an elt-3 regulated tyramine-dependent pathway mediating the toxic effects of MG-H1 AGE. Understanding this signaling pathway may help understand hedonistic overfeeding behavior observed due to modern AGEs-rich diets.