Abstract
Clustering of Complement Receptor 1 (CR1) in the erythrocyte membrane is important for immune-complex transfer and clearance. CR1 contains the Knops blood group antigens, including the antithetical pairs Swain-Langley 1 and 2 (Sl1 and Sl2) and McCoy a and b (McCa and McCb), whose functional effects are unknown. We tested the hypothesis that the Sl and McC polymorphisms might influence CR1 clustering on erythrocyte membranes. Blood samples from 125 healthy Kenyan children were analysed by immunofluorescence and confocal microscopy to determine CR1 cluster number and volume. In agreement with previous reports, CR1 cluster number and volume were positively associated with CR1 copy number (mean number of CR1 molecules per erythrocyte). Individuals with the McC b /McC b genotype had more clusters per cell than McC a/McC a individuals. However, this association was lost when the strong effect of CR1 copy number was included in the model. No association was observed between Sl genotype, sickle cell genotype, α+thalassaemia genotype, gender or age and CR1 cluster number or volume. Therefore, after correction for CR1 copy number, the Sl and McCoy polymorphisms did not influence erythrocyte CR1 clustering, and the effects of the Knops polymorphisms on CR1 function remains unknown.
Similar content being viewed by others
Introduction
Complement receptor 1 (CR1) is a polymorphic transmembrane glycoprotein expressed in humans on erythrocytes, most leucocytes and glomerular podocytes1. It is encoded by the CR1 gene at the 1q32 locus, spanning approximately 133 kb and consisting of 39 exons2. Four size polymorphisms of CR1 have been identified, the most common form being CR1*13. This classical form has a molecular weight of 190 kDa under non-reducing conditions, with an ectodomain comprising 30 short consensus repeats (SCRs), the first 28 of which are arranged into four long homologous repeats (LHRs) A-D (Fig. 1)4. Insertion or deletion of LHRs accounts for the variation seen in the molecule size4. Variation also occurs in erythrocyte CR1 copy number, with the mean number of CR1 molecules per cell varying between individuals from 50 to 12005. A HindIII restriction fragment length polymorphism in intron 27 of the CR1 gene correlates with high (H allele) or low (L allele) CR1 expression on erythrocytes6. However, this polymorphism is not associated with CR1 copy number in African populations7,8.
CR1 is a key regulator of complement cascade activation, namely through accelerating the decay of C3 and C5 convertases thus reducing production of anaphylotoxins C3a and C5a9, and acting as a cofactor with Factor I to inactivate C3b and C4b to iC3b and iC4b10. These functions map to specific sites within the CR1 molecule, with site 1 (SCRs 1–3 in LHR-A) displaying high decay accelerating activity11 and binding C4b12,13, and site 2 (SCRs 8–10 in LHR B and SCRs 15–17 in LHR C) being responsible for cofactor activity with Factor I and the binding of C3b and C4b (Fig. 1)12,13.
Diagram of the most common CR1 size variant (CR1*1). Adapted from Schmidt et al.70. The extracellular domain of CR1 is composed of 30 Complement Control Protein (CCP) domains organized into four “Long Homologous Repeats” (LHR). Two major functional sites are found within the CR1 molecule. Site 1 is located in LHR-A (CCP 1–3) and is the binding site for C4b and the Plasmodium falciparum invasion ligand PfRh4, and has decay accelerating activity. Site 2 is located in LHR-B (CCP 8–10) and LHR-C (CCP 15–17), and binds C3b and C4b, and interacts with P. falciparum infected erythrocytes to form rosettes. In addition, site 2 has Factor I-cofactor activity. SNPs determining the Sl and McC polymorphisms are localized in LHR-D (CCP 25, shown in red). CCP regions within Sites 1 and 2 that share high sequence identity (between 1 to 3 amino acid changes) are represented with identical shading. TM, transmembrane region; CYT, cytoplasmic tail.
The function of the LHR-D region of CR1 is unclear, but may include binding sites for C1q14 and Mannose Binding Lectin (MBL)15. SCR 25 in LHR-D also contains the Knops blood group antigens, which include the Swain-Langley (antithetical pairs Sl1 and Sl2) and McCoy (McCa and McCb) antigens (Fig. 1)16. The Sl2 allele results from a single nucleotide polymorphism (SNP) (rs17047661), encoding an amino acid change from arginine to glycine (R1601G), while the McC b allele results from a SNP 33 base pairs away (rs17047660), encoding an amino acid change from lysine to glutamic acid (K1590E) (Fig. 1)16. The functional effects of the Sl2 and McC b polymorphisms remain unknown. Recent work investigating short recombinant proteins bearing the Sl and McC antigens found no difference between the variant forms in their affinities for immobilised C3b, C4b or C1q, nor in their co-factor activity for cleavage of C3b and C4b in conjunction with Factor I17.
A marked geographical difference is seen in the frequency of Sl2 and McC b alleles, with the alleles being common in African populations and rare in Caucasians18. This has led to the suggestion that these polymorphisms arose due to selection pressure from malaria19, however, studies to date have been conflicting. An observational study from Western Kenya reported an association between the Sl2/Sl2 genotype and a reduced odds ratio of cerebral malaria (OR 0.17, p = 0.02) when compared to the Sl1/Sl1 genotype20. Recently, analysis of a large case-control study from Kilifi (eastern Kenya) identified an association between the Sl2/Sl2 genotype and protection against cerebral malaria and death from malaria, but conversely found addition of the McC b allele to be associated with increased odds of cerebral malaria (Opi, Swann et al., submitted). However, studies from The Gambia and Ghana did not find significant associations between the polymorphisms and severe malaria21,22.
In addition to complement regulation, CR1 on erythrocytes is responsible for immune adherence, the binding and transfer of opsonised immune complexes (ICs) from the circulation to resident macrophages in the liver and spleen for phagocytosis23,24. The clustering of CR1 in the erythrocyte membrane is thought to play a key role in its immune adherence function. On the resting erythrocyte membrane, CR1 is widely dispersed in small clusters, which coalesce into larger clusters when cross-linked by ligands or antibodies25,26,27,28,29,30,31. CR1 has also been demonstrated to co-immunoprecipitate with Fas-associated binding protein 1 (FAP-1) in the cytoskeleton fraction, which acts as a scaffolding protein for cross-linked CR1 and cluster formation28. It has been hypothesized that CR1 clustering after cross-linking by ligands could prevent ingestion of the erythrocyte during IC transfer to macrophages by concentrating opsonic stimuli in a few areas of the membrane28. Ligation of CR1 has also been demonstrated to increase erythrocyte deformability, which could allow IC-laden erythrocytes to traverse the hepatic and splenic capillaries and transfer their cargo to the resident macrophages more easily29.
The number of CR1 clusters per erythrocyte increases with CR1 copy number25,32,33, although CR1 copy number alone does not completely explain inter-individual variation in clustering, implying involvement of other factors32. The possibility that the Sl2 polymorphism might affect CR1 clustering has been raised34, but never investigated. If the Knops blood group is indeed associated with malaria, there are a number of mechanistic possibilities that could include a role for variation in CR1 clustering. If the Sl2 or McCb mutations influenced the size or number of CR1 clusters formed, this might influence the process of invasion of Plasmodium falciparum into the erythrocyte via CR1 “capping” of the merozoite35. Alternatively, CR1 clustering could be important in the pathogenesis of cerebral malaria. Children with cerebral malaria may have higher circulating levels of immune complexes (ICs) than children with other malaria disease syndromes36. ICs deposited on the endothelial surface can activate complement and stimulate the release of local pro-inflammatory mediators37. ICs can also cross-link Fc-receptors on monocyte and macrophages, stimulating production of tumour necrosis factor38, a cytokine whose relationship with cerebral malaria has been previously reported39. As avidity of IC binding to erythrocytes depends on CR1 clustering, clearance of ICs could potentially vary with different CR1 clustering patterns.
We therefore set up a study using blood samples from healthy children in Kilifi, Kenya to examine the hypothesis that the Sl2 and McC b polymorphisms might influence erythrocyte CR1 clustering.
Results
Erythrocyte CR1 clustering was assessed in blood samples from 125 Kenyan children with the homozygous Knops genotypes Sl1/Sl1 McC a /McC a, Sl2/Sl2 McC a /McC a or Sl2/Sl2 McC b /McC b. These genotypes are hereafter referred to as 1a, 2a and 2b respectively. Only homozygous genotypes were studied in order to examine the associations of the “pure forms” of the polymorphisms on CR1 clustering. The characteristics of the children are summarised in Table 1; no significant differences were seen between the 1a, 2a or 2b genotype groups in terms of gender, age, ABO blood group, α+thalassaemia genotype or sickle cell genotype.
In order to examine CR1 clusters, we carried out immunofluorescent staining of unfixed erythrocytes with a monoclonal antibody to CR1 (Fig. 2A), and analysed confocal microscopy images with an automated protocol developed using Volocity Image Analysis software (version 6.3, Perkin Elmer) to determine CR1 erythrocyte cluster number and volume (Fig. 2B–H and Supplementary Videos 1 and 2).
Immunofluorescent staining of erythrocyte CR1 clusters and image analysis. (A) Erythrocyte CR1 clusters were visualised by staining with 5 μg/ml of CR1 monoclonal antibody J3D3, followed by 4 μg/ml of Alexa Fluor 488 goat anti-mouse IgG. See methods for full details. (B) For each donor, five confocal microscopy images were taken from the central area of the microscope slide as indicated by the grey shaded areas. Each image was a Z-stack comprising 8–10 steps. (C–F) To illustrate the image analysis process, a section containing 4 erythrocytes from one of the 5 images from a single donor is shown. Overall, at least 200 erythrocytes were examined per donor. (C) For each image, the bright-field view was used to identify the positions of erythrocytes. (D) Non-touching cells were then stamped. (E) The stamped region was then applied to the fluorescent image. (F) Cluster numbers and volumes within the stamped regions were assessed by an automated protocol using Volocity software as described in the methods. (G) An example of the clusters identified in a single erythrocyte. (H) The same erythrocyte as shown in G, with distinct clusters being shown in different colours. (I) A different erythrocyte with each of the 8 steps that comprise the Z-stack shown separately (1 to 8). These images can also be viewed as a video (Supplementary Videos 1 and 2).
Increased CR1 copy number in individuals with the McCb/McCb genotype
CR1 copy number (mean number of CR1 molecules per erythrocyte) positively correlates with number of CR1 clusters25,32,33 and was therefore measured in all individuals. We first examined the association of CR1 copy number and donor variables in univariate analyses. We observed that individuals with the 2b genotype showed a trend towards higher CR1 copy number than individuals with 1a and 2a genotypes (Table 1 and Fig. 3A, p = 0.098 by ANOVA). This was driven by the McC component of the genotype, as individuals with the McC b /McC b genotype had a significantly higher CR1 copy number than McC a /McC a (mean 658 copies per cell versus 483 copies per cell, p = 0.033, t-test, Fig. 3B). The Sl component of the genotype was not significantly associated with CR1 copy number (p = 0.978, t-test, Fig. 3C). Age, gender, ABO blood group and sickle cell trait showed no significant association with CR1 copy number in this study. Individuals who were homozygotes for α +thalassaemia had a lower CR1 copy number (441) than those who were heterozygotes (495) or who had normal α globin (520), however this was not significant (p = 0.366, ANOVA, Figure S1).
The McC b genotype is associated with increased CR1 copy number and cluster number. Preliminary exploration of the association of Swain Langley (Sl) and McCoy (McC) genotypes with CR1 copy number (mean number of CR1 molecules per erythrocyte), median CR1 cluster number per cell and mean cluster volume. Panels (A–C) Relationship between CR1 copy number and (A) combined Sl and McC genotype (1a = Sl1/Sl1 McC a /McC a, 2a = Sl2/Sl2 McC a /McC a and 2b = Sl2/Sl2 McC b /McC b), (B) McC genotype and (C) Sl genotype. Panels (D–F) Relationship between median number of CR1 clusters per cell and (D) combined Sl and McC genotype, (E) McC genoype and (F) Sl genotype. Panels (G–I) Relationship between mean CR1 cluster volume per cell and (G) combined Sl and McC genotype, (H) McC genoype and (I) Sl genotype. Red bars indicate mean and standard deviation, with statistical testing by t test or ANOVA as indicated. Individuals with the McC b genotype had significantly higher CR1 copy numbers than McC a individuals, and showed a trend towards higher CR1 cluster numbers, although this was not statistically significant.
Increased CR1 cluster number in individuals with the McCb/McCb genotype
As previously reported26,29,32, the number of CR1 clusters per cell varied widely within each individual. Overall, the number of clusters per cell ranged from 0 to 43, with the maximum number of clusters per cell ranging from 15 to 43 amongst different individuals (Figure S2). The volume of CR1 clusters also varied within individuals, with the largest clusters being > 3μm3 (Figure S3).
Preliminary analysis of the relationship between Knops genotype and erythrocyte CR1 cluster number was conducted by ANOVA, with data summarised at the level of the individual donor (i.e. using a single data point per donor, representing the median cluster number for that donor). These data were suggestive of a higher cluster number in 2b individuals, driven by the McC b genotype, although this was not statistically significant (Fig. 3D–F). Similar analysis by ANOVA of the relationship between Knops genotype and mean CR1 cluster volume revealed no significant associations (Fig. 3G–I). There were also no significant associations between CR1 cluster number (Figure S4) or cluster volume (Figure S5) and age, gender, ABO blood group, sickle cell genotype and α +thalassaemia genotype.
In agreement with previous findings32,33, a strong positive linear correlation was observed between CR1 cluster number and CR1 copy number (R2 = 0.530, p < 0.001, Fig. 4A), and a positive correlation was also observed between CR1 cluster volume and CR1 copy number (R2 = 0.398, p < 0.001, Fig. 4B). Cluster number and cluster volume were also strongly positively correlated (R2 = 0.573, p < 0.001, Fig. 4C). Because of the potential confounding effect of CR1 copy number on the relationship between Knops genotype and CR1 cluster number and volume, more detailed statistical analyses were carried out.
Positive correlation between CR1 copy number, cluster number and cluster volume. (A) Scatter plot of median CR1 cluster number per cell by CR1 copy number (mean number of CR1 molecules per cell) with linear regression. Each dot represents an individual donor. (B) Scatter plot of mean CR1 cluster volume by CR1 copy number with linear regression. Each dot represents an individual donor. (C) Scatter plot of mean CR1 cluster volume by median CR1 cluster number with linear regression. Each dot represents an individual donor. There were significant positive correlations between all three variables studied.
No evidence of an association between Knops genotype and CR1 cluster number after adjustment for CR1 copy number
A hierarchical mixed effects Poisson regression analysis was constructed to include all available data at the level of the cell (using at least 200 data points per individual donor, representing the CR1 cluster number for each cell imaged) and to allow for adjustment for CR1 copy number. The analysis examined each donor variable (combined Sl and McC genotype, Sl genotype alone, McC genotype alone, gender, age, ABO blood group, α+thalassaemia genotype or sickle cell genotype) in turn. Each donor variable was specified as a fixed effect and donor identifier and date of assay were specified as random effects. The relationship between each donor variable and the number of CR1 clusters per cell was reviewed both before and after adjustment for CR1 copy number.
Prior to inclusion of CR1 copy number in the model, the 2b genotype was associated with a higher number of clusters per cell than the 1a genotype (multiplicative increase of 1.200 (95% CI 0.997–1.443, p = 0.053, Table 2)). When broken down by component genotypes, this association was driven by the McC component rather that the Sl component, with the McC b /McC b genotype associated with a significantly higher number of clusters per cell than the McC a /McC a genotype (multiplicative increase = 1.216, 95% CI 1.027–1.439, p = 0.023, Table 2). No significant associations were seen for gender, age, ABO blood group, α+thalassaemia genotype or sickle cell genotype.
The correlation between erythrocyte CR1 copy number and number of clusters per cell was strong, with each increase in CR1 copy number of 100 being associated with a multiplicative increase in CR1 cluster number of 1.084 (95% CI 1.067−1.101, p < 0.001). Adjustment for CR1 copy number obscured any association between cluster number and McC genotype (Table 2). This suggests that the original association was likely to be driven by a difference in CR1 copy number, rather than any unique effect of the McC genotype itself.
No evidence of an association between Knops genotype and erythrocyte CR1 cluster volume after adjustment for CR1 copy number
Similarly, a more detailed hierarchical mixed effects linear regression analysis was constructed to include all available data at the level of the individual clusters, and to allow for adjustment for CR1 copy number. Donor variables (combined Sl and McC genotype, Sl genotype alone, McC genotype alone, gender, age, ABO blood group, α+thalassaemia genotype or sickle cell genotype) were examined individually as fixed effects, while cell identifier and donor identifier were specified as random effects (Table 3). The relationship between each donor variable and CR1 cluster volume was reviewed both before and after adjustment for CR1 copy number.
No donor variables were found to have a significant association with cluster volume before adjustment for CR1 copy number (i.e. t values were all <1.96 or >−1.96). The correlation between CR1 copy number and cluster volume was extremely strong, with each increase in CR1 copy number of 100 being associated with an absolute increase in CR1 cluster volume of 0.039 µm3 (t value 9.446, Table 3. Bootstrapped p value <0.001). Adjustment for CR1 copy number had no impact on the non-significant relationships between cluster volume and combined Sl and McC genotype, Sl genotype alone, McC genotype alone, gender, age, α+thalassaemia genotype or sickle cell genotype (Table 3). However, in models adjusting for CR1 copy number, the B blood group was associated with an absolute increase in cluster volume of 0.057 µm3 when compared to the reference O blood group (t value = 2.305, Table 3. Bootstrapped p value = 0.028).
Discussion
In agreement with existing literature, we report that both the number of CR1 clusters per erythrocyte and CR1 cluster volume are highly correlated with an individual’s erythrocyte CR1 copy number25,32,33. The strength of association between CR1 copy number and cluster number was such that the inclusion of CR1 copy number in the statistical model obscured any other associations. However, on review of individual explanatory variables (without CR1 copy number in the model), the McC b /McC b genotype was associated with significantly higher cluster number than the McC a /McC a genotype. Individuals of the McC b /McC b genotype were found to have a significantly higher CR1 copy number than McC a /McC a, which is in agreement with a larger analysis of Kenyan children40. As such, it is likely that the association of the McC b /McC b genotype with increased CR1 cluster number was due to the higher CR1 copy numbers in McC b /McC b individuals, rather than being a distinct biological effect of the variant itself. Exactly how the McC b polymorphism influences erythrocyte CR1 copy number is unknown, but could be due to resistance to cleavage by tryptic proteases resulting from the lysine to glutamic acid (K1590E) mutation, as discussed previously40.
In contrast to existing literature40,41, we found no evidence of an association between donor age and CR1 copy number. However, the previous studies included children between birth and 13 years40 or birth and 32 years41, with the major changes occurring during the first four years of life. The narrower age range of our study (5.7 years –12.7 years) may explain why no association between age and CR1 copy number was detected. We also found no significant effect of α +thalassaemia genotype on CR1 copy number, unlike other studies which showed lower CR1 copy number in α +thalassaemia homozygotes compared to normal individuals in Papua New Guinea42 and Kenya40. Whilst we also observed a lower CR1 copy number in α +thalassaemia homozygote individuals, the small sample size in our study may have prevented this from reaching statistical significance.
One unexpected finding from this study was that blood group B was associated with significantly larger CR1 clusters than the reference blood group O. There is no known link between ABO blood group and CR1 function, and further work will be required to determine if this is a reproducible finding, and if so, to determine its biological significance.
Owing to the high level of within and between donor variation, this was a challenging assay to develop and interpret. The sample size of this study was small, in particular for the 2b genotype, which is a limitation of this study. However, given the logistics of sampling (see methods), it was not possible to recruit more individuals prospectively. Another consideration is that although all children positive for malaria on a thick film were excluded, information on other infective or inflammatory conditions that might influence CR1 copy number was not available. Therefore, it is possible that the CR1 copy number of some children may have been transiently altered by concurrent conditions. However, all children were sampled in the community and were not inpatients in health centres, therefore were in broadly good health.
The use of field samples also conferred other limitations on our study. When carrying out the confocal imaging in Edinburgh, we used conservative imaging settings to avoid photo-bleaching the slides during image acquisition. This was done in case repeated imaging was needed, as the slides themselves could not be replaced. Therefore, it is possible that we may have under-sampled the immunofluorescent slides, resulting in missing very small clusters. As temperature has been reported to affect cluster size in some studies43, the use of physiological temperature during the assay (rather than 4 °C as used here) may have increased the relevance of the findings, as may the use of a physiological ligand to induce clustering (e.g. C3b covered beads or immune complexes). Furthermore, the J3D3 monoclonal antibody used here recognises multiple epitopes in LHRs A, B and C44, and results could therefore be influenced by CR1 size polymorphisms (allotype), which was not accounted for here. An antibody with a single epitope on CR1 that would not be affected by CR1 allotype43 could be used in future studies. Other technical aspects of the assay may have affected the results shown here. For example, the mounting medium used after staining and preparing the thin blood films can cause lysis of the erythrocytes, potentially bringing the top and bottom membranes closer together. This could cause overlap of some clusters making them appear fewer or larger than in vivo. Despite these limitations, our methods were consistent throughout all samples, thus allowing for direct comparison between the different genotypes.
The functional effects of the Sl and McC polymorphisms remain elusive. Laboratory evidence suggests that Sl2 may affect rosetting (the binding of Plasmodium falciparum infected erythrocytes to uninfected erythrocytes)19, a parasite virulence factor associated with severe malaria in African children45. Rosetting occurs when P. falciparum Erythrocyte Membrane Protein 1 (PfEMP1) expressed on the surface of infected erythrocytes binds to uninfected erythrocyte receptors, and is thought to involve site 2 on CR1 (the C3b binding site) in LHR-B and LHR-C46. Erythrocytes from Sl2/Sl2 donors rosette less well than erythrocytes from Sl1/Sl1 donors when PfEMP1 is expressed heterologously on the surface of COS-7 cells19. However, a study using short recombinant proteins of the Sl and McC variants found no difference in their abilities to disrupt rosettes17. Further investigation of the effects of the Knops polymorphisms on P. falciparum rosetting is required.
In addition to its role in rosetting, CR1 has also been identified as a sialic acid-independent invasion receptor for P. falciparum 47, via the reticulocyte-binding–like homolog 4 (PfRh4)48. The binding of PfRh4 has been mapped to site 1 in LHR-A49,50,51. Immuno-fluorescent staining of freshly released invasive merozoites showed CR1 on erythrocytes to be more intense around the merozoite, referred to by the authors as “capping,” suggesting CR1 clustering may have a role in invasion47. However, the previously cited study using short recombinant proteins of the Sl and McC variants also found no difference in their abilities to inhibit erythrocyte invasion by P. falciparum 17.
Interestingly, the binding site for mannose binding lectin (MBL) has been mapped to CCP 24–25, in close proximity to the Sl and McC polymorphisms52. MBL may influence malaria in Ghanaian children, with alleles responsible for low levels of MBL associated with increased parasite load, lower blood glucose levels and increased odds of severe malarial anaemia53,54. It has also been suggested that MBL may opsonize P. falciparum infected erythrocytes53 or merozoites55. However, whether the Sl or McC polymorphisms have functional effects on MBL interactions with CR1 has not yet been explored.
In summary, after accounting for the effect of CR1 copy number, there was no significant association between Sl or McC genotypes (separately or in combination) and erythrocyte CR1 cluster number or cluster volume in Kenyan children. The effects of the Sl and McC Knops blood group polymorphisms on CR1 function remain unknown.
Materials and Methods
Study Population and Ethical Approval
The Immunology Cohort Study Group in Kilifi, Kenya comprises 1041 children resident in the Junju or Ngerenya regions of the Kilifi Demographic Surveillance System (KDSS)56,57. The Kilifi district borders the Indian Ocean and the local economy depends on subsistence farming57. It is a malaria endemic area with seasonal transmission after the long (April -June) and short (October-November) rainy seasons, although recently malaria transmission has fallen in the area58. Ethical approval for this study was obtained from the KEMRI Ethical Review Committee and all methods were performed in accordance with the relevant guidelines and regulations. Following informed consent from a parent or guardian, blood samples were obtained by venepuncture between 2nd and 30th April 2013 (inclusive).
Donor selection and genotyping
Donors had previously been genotyped for Sl and McC polymorphisms using the SEQUENOM iPLEX® Gold platform59, for sickle cell (HbS) and α+thalassaemia genotype (–α3.7 deletion) by PCR60,61 and typed for ABO blood group by standard haemagglutination methods62. Our investigation was limited to the Sl and McC homozygote genotypes, 1a (Sl1/Sl1 McC a /McC a), 2a (Sl2/Sl2 McC a /McC a) and 2b (Sl2/Sl2 McC b /McC b). The 1b genotype does not occur because the McC b polymorphism is only found on the background of the Sl2 variant16,18. 148 children in the cohort with 1a, 2a or 2b genotypes, and with complete data on α+thalassaemia and sickle cell status (both reported to influence CR1 copy number40,42,63) were eligible for inclusion. Nine children did not attend for venepuncture, 12 were excluded because they were positive for P. falciparum on thick blood smear, and two were accidently discarded, giving 125 samples for analysis (29 with genotype 1a, 88 with genotype 2a and eight with genotype 2b).
Sample collection and processing
4.4 mls of venous blood was collected into a lithium heparin tube, centrifuged at 400 g for 10 minutes and plasma removed. The sample was then made up to 5 mls with R2 medium (500 ml RPMI (catalogue number R0883, Sigma-Aldrich, Missouri, USA) with final concentration of 0.02 M Hepes (Gibco, New York, USA), 0.1 mM L-glutamine (Gibco, New York, USA), 1% vol/vol Penicillin–Streptomycin Solution Hybri-Max (Sigma-Aldrich, Missouri, USA) and 2% pooled human AB serum). The cell suspension was layered onto 3 ml of Lymphoprep (PROGEN Biotechnik, Heidelberg, Germany) and centrifuged at 1000 g for 20 minutes at 20 °C. The white cell layer was removed and the resulting erythrocytes washed twice in R2 medium without serum by centrifugation at 400 g for 5 minutes. They were then adjusted to 4% haematocrit by adding R2 medium without serum. The clustering assay was subsequently performed on these unfixed cells. As storage of blood samples can affect CR1 copy number and function43,64,65, all clustering assays were performed within eight hours of venepuncture.
CR1 copy number determination
An aliquot of washed erythrocytes from above was fixed in 5% formaldehyde (Sigma-Aldrich, Missouri, USA) on the day of venepuncture and CR1 copy number was determined within 8 weeks as previously described in detail65. Briefly, this method uses flow cytometry to quantify CR1 copy number on fixed erythrocytes using a standard curve derived from donors with known CR1 copy numbers that have been fixed at the same time as the samples. Erythrocytes were stained with 0.5 μg/ml J3D3 anti-CR1 monoclonal antibody (GTX44217, GeneTex, California, USA), followed by 5 μg/ml of Alexa Fluor 488-conjugated goat anti-mouse IgG (A-11001, ThermoFisher Scientific, Paisley, UK) and the mean fluorescence intensity determined on a FC500 flow cytometer (Beckman-Coulter Inc.). Examples of the gating strategy and staining of low, intermediate and high CR1 standard donors are shown in Figure S6.
Immunofluorescent staining of CR1 clusters
The IgG1 mouse anti-CR1 monoclonal antibody J3D3 was used to stain CR1 clusters, as it does not recognise epitopes in the LHR-D region44, so was considered unlikely to influence any effect the Sl or McC polymorphisms might have on clustering. Two aliquots from each sample of washed, unfixed erythrocytes at 4% haematocrit in PBS/1% BSA were incubated with 5 μg/ml of either J3D3 or IgG1 mouse anti-human isotype control (MCA928, AbD Serotec, Kidlington, UK) to give a final sample volume of 50 μl. Samples were incubated at 4 °C for one hour (resuspended every ten mins) and washed three times in 1 ml cold PBS (Oxoid, Basingstoke, UK). 50 μl of 4 μg/ml Alexa Fluor 488-conjugated goat anti-mouse secondary antibody was added and the sample was incubated at 4 °C in the dark for one hour (resuspended every ten mins) and subsequently washed three times in 1 ml cold PBS. The sample was resuspended at 30–40% haematocrit, smeared on a glass slide, air-dried, mounted under a 22 mm × 22 mm glass coverslip with 10 μl of DABCO-glycerol (2.5 mg DABCO (Sigma-Aldrich, Poole, UK), 0.5 ml glycerol and 0.5 ml PBS) and sealed with nail varnish. All steps to this point were undertaken at the KEMRI-Wellcome Trust Research Unit, Kilifi and the resultant slides were shipped to Edinburgh for confocal microscopy.
Confocal microscopy
Slides were imaged with a Leica SP5 confocal microscope using the glycerol × 63 objective, argon laser and differential interference contrast settings (now referred to as “bright field”). Five images per slide (Fig. 2A shows a representative image) were taken from predefined areas (Fig. 2B). To prevent bias for stained cells, each slide was navigated using the bright field image, ensuring that only erythrocyte outlines, rather than CR1 clusters, were visible. To minimise photo-bleaching, a Z-stack of images was taken at 0.8 μm intervals and between eight to ten steps were acquired per stack (an example of the 8 steps comprising a Z-stack are shown in Fig. 2I and Supplementary Videos 1 and 2). The images in theZ-stack were then merged into one 3D image using Volocity 3D Image Analysis Software version 6.3 (PerkinElmer, Massachusetts, USA). Slides were imaged blind to donor information.
Image Analysis
Using the bright field image (to obscure CR1 clusters and increase objectivity) (Fig. 2C), a region of interest stamp 9 µm in diameter (area 63.6 µm2, slightly larger than an erythrocyte) was used to identify individual erythrocytes that were not touching other cells (Fig. 2D). Approximately 40 erythrocytes per image were stamped. Image analysis was then carried out within the stamped areas in the fluorescent channel (Fig. 2E,F). An automated software protocol was developed using the “Find objects” task in the “Measurements” module in Volocity version 6.3 (Perkin Elmer), and was used to determine the number of CR1 clusters per erythrocyte (Fig. 2G,H), and to calculate the volume of each cluster. The protocol identified objects within the stamped area based on a minimum size threshold and subsequently separated touching objects using an object size guide. To establish the protocol settings, five cells from five separate donors (total cell number = 25) were reviewed. Settings over a range from 0.01 to 2.56 μm3 were examined for both minimum size and object size guide. A minimum size threshold of 0.06 μm3 and object size guide of 0.06 μm3 gave the optimal correlation with number of clusters identified by eye (Pearson’s correlation coefficient of 0.927 and r2 of 0.859), therefore these settings were used throughout. To account for background fluorescence, 30 maximum fluorescent intensity (MFI) readings were taken from areas of cells without visible CR1 clusters for each donor. The median of these readings was calculated for each individual to give the background MFI. The median of all donor medians was then calculated, giving a universal background correction factor (median = 554, IQR = 554–693) that was subtracted from all samples before running the Volocity protocol.
In order to determine the optimal number of cells to image from each donor, 400 erythrocytes were imaged from a single donor. The standard error of the mean (SEM) of the number of CR1 clusters per erythrocyte was plotted against the number of erythrocytes imaged (Figure S7). It was decided to image 200 erythrocytes for each donor, as gains in SEM beyond this point were small.
The reproducibility of the assay was assessed by imaging a blood sample taken from a single donor at the same time each day for three consecutive days. Neither the median number of CR1 clusters per cell, nor the mean CR1 cluster volume was significantly different across the three days (Kruskal-Wallis test p = 0.337 and 0.054 respectively) (Figure S8).
Statistical analyses
Our experiments generated data on individual CR1 clusters from at least 200 erythrocytes per donor. These data were then summarised at three levels: (1) at the level of each individual cluster (to allow for detailed analysis of cluster volume); (2) at the level of each individual cell (to allow for detailed analyses of cluster number per cell) and (3) at the level of each individual donor (to provide a single data point for each participant which represented their median CR1 cluster number or mean cluster volume).
Initial exploratory analyses used the data at the level of the individual donor to examine the influence of donor variables on CR1 copy number, CR1 cluster number and cluster volume. The donor variables explored in our analyses were basic demographic data (age and gender), other variables reported to influence CR1 copy number (α+thalassaemia genotype and sickle cell genotype)40,42,63, ABO blood group (repeatedly associated with malaria in this population)59,62 and Knops genotype (1a, 2a and 2b). Correlation between continuous variables was analysed by linear regression, and differences between means were analysed by one-way ANOVA or unpaired t tests.
To examine factors potentially influencing CR1 cluster number, the data were then analysed at the level of the individual cells. To permit inclusion of all the data collected and account for the intra-donor individual variation in the number of clusters per cell, a two-level hierarchical mixed effects Poisson regression model was used. Donor identifier and date the experiment was performed were included as random effects. Each donor variable (as above) was then examined in turn as a fixed effect. In view of the well documented relationship between CR1 cluster number and CR1 copy number25,32,33, each variable was examined both with and without CR1 copy number in the model.
Finally, to examine factors potentially influencing CR1 cluster volume and account for both intra-individual and intra-cell variation in volume of clusters, a three-level hierarchical mixed effects linear regression model was used with both donor identifier and cell identifier included as random effects. Each donor variable was then examined as a fixed effect with and without CR1 copy number in the model. Where t values were significant, bootstrapping with 1000 iterations was performed to provide an equivalent p value for ease of interpretation.
Analyses were performed using R statistical software (R Foundation for Statistical Computing, Vienna, Austria)66 using the packages “plyr”, “ggplot2”, “boot” and “lme4”67,68,69. Graphs were drawn using GraphPad Prism 7 (GraphPad Software, Inc., La Jolla, CA, USA).
Data availability
Data at the level of individual donors are in Supplementary dataset 1, data at the level of individual cells are in Supplementary dataset 2, and data at the level of individual clusters are in Supplementary dataset 3.
References
Fearon, D. T. Human Complement Receptors for C3b (CR1) and C3d (CR2). J. Invest. Dermatol. 85, 53s–57s, https://doi.org/10.1111/1523-1747.ep12275473 (1985).
Vik, D. P. & Wong, W. W. Structure of the gene for the F allele of complement receptor type 1 and sequence of the coding region unique to the S allele. J. Immunol. 151, 6214–6224 (1993).
Dykman, T. R., Hatch, J. A., Aqua, M. S. & Atkinson, J. P. Polymorphism of the C3b/C4b receptor (CR1): characterization of a fourth allele. J. Immunol. 134, 1787–1789 (1985).
Wong, W. W. Structural and Functional Correlation of the Human Complement Receptor Type 1. J. Invest. Dermatol. 94, 64s–67s, https://doi.org/10.1111/1523-1747.ep12875150 (1990).
Moulds, J. M., Moulds, J. J., Brown, M. & Atkinson, J. P. Antiglobulin testing for CR1-related (Knops/McCoy/Swain-Langley/York) blood group antigens: negative and weak reactions are caused by variable expression of CR1. Vox Sang 62, 230–235 (1992).
Wilson, J. G. et al. Identification of a restriction fragment length polymorphism by a CR1 cDNA that correlates with the number of CR1 on erythrocytes. J. Exp. Med. 164, 50–59 (1986).
Herrera, A. H., Xiang, L., Martin, S. G., Lewis, J. & Wilson, J. G. Analysis of complement receptor type 1 (CR1) expression on erythrocytes and of CR1 allelic markers in Caucasian and African American populations. Clin. Immunol. Immunopathol. 87, 176–183 (1998).
Rowe, J. A. et al. Erythrocyte CR1 expression level does not correlate with a HindIII restriction fragment length polymorphism in Africans; implications for studies on malaria susceptibility. Genes Immun. 3, 497–500, https://doi.org/10.1038/sj.gene.6363899 (2002).
Iida, K. & Nussenzweig, V. Complement receptor is an inhibitor of the complement cascade. J. Exp. Med. 153, 1138–1150, https://doi.org/10.1084/jem.153.5.1138 (1981).
Medof, M. E. & Nussenzweig, V. Control of the function of substrate-bound C4b-C3b by the complement receptor Cr1. J. Exp. Med. 159, 1669–1685 (1984).
Krych-Goldberg, M. et al. Decay accelerating activity of complement receptor type 1 (CD35). Two active sites are required for dissociating C5 convertases. J. Biol. Chem. 274, 31160–31168 (1999).
Klickstein, L. B. et al. Identification of distinct C3b and C4b recognition sites in the human C3b/C4b receptor (CR1, CD35) by deletion mutagenesis. J. Exp. Med. 168, 1699–1717 (1988).
Krych, M., Hourcade, D. & Atkinson, J. P. Sites within the complement C3b/C4b receptor important for the specificity of ligand binding. Proc. Nat. Acad. Sci. USA 88, 4353–4357 (1991).
Klickstein, L. B., Barbashov, S. F., Liu, T., Jack, R. M. & Nicholson-Weller, A. Complement receptor type 1 (CR1, CD35) is a receptor for C1q. Immunity 7, 345–355 (1997).
Ghiran, I. et al. Complement Receptor 1/Cd35 Is a Receptor for Mannan-Binding Lectin. J. Exp. Med. 192, 1797–1808, https://doi.org/10.1084/jem.192.12.1797 (2000).
Moulds, J. M. et al. Molecular identification of Knops blood group polymorphisms found in long homologous region D of complement receptor 1. Blood 97, 2879–2885, https://doi.org/10.1182/blood.V97.9.2879 (2001).
Tetteh-Quarcoo, P. B. et al. Lack of Evidence from Studies of Soluble Protein Fragments that Knops Blood Group Polymorphisms in Complement Receptor-Type 1 Are Driven by Malaria. PLoS ONE 7, e34820, https://doi.org/10.1371/journal.pone.0034820 (2012).
Moulds, J. M. The Knops blood-group system: a review. Immunohematol. 26, 2–7 (2010).
Rowe, J. A., Moulds, J. M., Newbold, C. I. & Miller, L. H. P. falciparum rosetting mediated by a parasite-variant erythrocyte membrane protein and complement-receptor 1. Nature 388, 292–295, https://doi.org/10.1038/40888 (1997).
Thathy, V., Moulds, J., Guyah, B., Otieno, W. & Stoute, J. Complement receptor 1 polymorphisms associated with resistance to severe malaria in Kenya. Malaria J. 4, 54, https://doi.org/10.1186/1475-2875-4-54 (2005).
Zimmerman, P. A. et al. CR1 Knops blood group alleles are not associated with severe malaria in the Gambia. Genes Immun 4, 368–373, https://doi.org/10.1038/sj.gene.6363980 (2003).
Hansson, H. H. et al. Human genetic polymorphisms in the Knops blood group are not associated with a protective advantage against Plasmodium falciparum malaria in Southern Ghana. Malaria J. 12, 400, https://doi.org/10.1186/1475-2875-12-400 (2013).
Schifferli, J. A. & Taylor, R. P. Physiological and pathological aspects of circulating immune complexes. Kidney Int. 35, 993–1003 (1989).
Emlen, W., Carl, V. & Burdick, G. Mechanism of transfer of immune complexes from red blood cell CR1 to monocytes. Clin. Exp. Immunol. 89, 8–17 (1992).
Paccaud, J. P., Carpentier, J. L. & Schifferli, J. A. Direct evidence for the clustered nature of complement receptors type 1 on the erythrocyte membrane. J. Immunol 141, 3889–3894 (1988).
Chevalier, J. & Kazatchkine, M. D. Distribution in clusters of complement receptor type one (CR1) on human erythrocytes. J. Immunol. 142, 2031–2036 (1989).
Paccaud, J. P., Carpentier, J. L. & Schifferli, J. A. Difference in the clustering of complement receptor type 1 (CR1) on polymorphonuclear leukocytes and erythrocytes: effect on immune adherence. Euro. J. Immunol. 20, 283–289, https://doi.org/10.1002/eji.1830200209 (1990).
Ghiran, I., Glodek, A. M., Weaver, G., Klickstein, L. B. & Nicholson-Weller, A. Ligation of erythrocyte CR1 induces its clustering in complex with scaffolding protein FAP-1. Blood 112, 3465–3473, https://doi.org/10.1182/blood-2008-04-151845 (2008).
Glodek, A. M. et al. Ligation of complement receptor 1 increases erythrocyte membrane deformability. Blood 116, 6063–6071, https://doi.org/10.1182/blood-2010-04-273904 (2010).
Lapin, Z. J., Höppener, C., Gelbard, H. A. & Novotny, L. Near-field quantification of complement receptor 1 (CR1/CD35) protein clustering in human erythrocytes. J Neuroimmune Pharmacol. 7, 539–543, https://doi.org/10.1007/s11481-012-9346-3 (2012).
Melhorn, M. I. et al. CR1-mediated ATP Release by Human Red Blood Cells Promotes CR1 Clustering and Modulates the Immune Transfer Process. J. Biol. Chem. 288, 31139–31153, https://doi.org/10.1074/jbc.M113.486035 (2013).
Cosio, F. G., Shen, X. P. & Hebert, L. A. Immune complexes bind preferentially to specific subpopulations of human erythrocytes. Clin. Immunol. Immunopathol. 55, 337–354 (1990).
Taylor, R. P., Pocanic, F., Reist, C. & Wright, E. L. Complement-opsonized IgG antibody/dsDNA immune complexes bind to CR1 clusters on isolated human erythrocytes. Clin. Immunol. Immunopathol. 61, 143–160 (1991).
Krych-Goldberg, M., Moulds, J. M. & Atkinson, J. P. Human complement receptor type 1 (CR1) binds to a major malarial adhesin. Trends Mol. Med. 8, 531–537 (2002).
Biryukov, S. et al. Complement and Antibody-mediated Enhancement of Red Blood Cell Invasion and Growth of Malaria Parasites. EBioMedicine 9, 207–216, https://doi.org/10.1016/j.ebiom.2016.05.015 (2016).
Mibei, E. K., Otieno, W. O., Orago, A. S. & Stoute, J. A. Distinct pattern of class and subclass antibodies in immune complexes of children with cerebral malaria and severe malarial anaemia. Parasite Immunol. 30, 334–341, https://doi.org/10.1111/j.1365-3024.2008.01030.x (2008).
Jancar, S. & Sanchez Crespo, M. Immune complex-mediated tissue injury: a multistep paradigm. Trends Immunol. 26, 48–55, https://doi.org/10.1016/j.it.2004.11.007 (2005).
Debets, J. M. et al. Fc-receptor cross-linking induces rapid secretion of tumor necrosis factor (cachectin) by human peripheral blood monocytes. J. Immunol. 141, 1197–1201 (1988).
Kwiatkowski, D. et al. TNF concentration in fatal cerebral, non-fatal cerebral, and uncomplicated Plasmodium falciparum malaria. Lancet 336, 1201–1204 (1990).
Opi, D. H., Uyoga, S., Orori, E. N., Williams, T. N. & Rowe, J. A. Red blood cell complement receptor one level varies with Knops blood group, α(+)thalassaemia and age among Kenyan children. Genes Immun. 17, 171–178, https://doi.org/10.1038/gene.2016.2 (2016).
Waitumbi, J. N., Donvito, B., Kisserli, A., Cohen, J. H. M. & Stoute, J. A. Age-related changes in red blood cell complement regulatory proteins and susceptibility to severe malaria. J. Infect. Dis. 190, 1183–1191, https://doi.org/10.1086/423140 (2004).
Cockburn, I. A. et al. A human complement receptor 1 polymorphism that reduces Plasmodium falciparum rosetting confers protection against severe malaria. Proc. Nat. Acad. Sci. USA 101, 272–277, https://doi.org/10.1073/pnas.0305306101 (2004).
Chen, C. H. et al. Antibody CR1-2B11 recognizes a non-polymorphic epitope of human CR1 (CD35). Clin. Exp. Immunol 148, 546–554, https://doi.org/10.1111/j.1365-2249.2007.03355.x (2007).
Nickells, M. et al. Mapping epitopes for 20 monoclonal antibodies to CR1. Clin. Exp. Immunol. 112, 27–33 (1998).
Rowe, A., Obeiro, J., Newbold, C. I. & Marsh, K. Plasmodium falciparum rosetting is associated with malaria severity in Kenya. Infect. Immun. 63, 2323–2326 (1995).
Rowe, J. A. et al. Mapping of the region of complement receptor (CR) 1 required for Plasmodium falciparum rosetting and demonstration of the importance of CR1 in rosetting in field isolates. J. Immunol. 165, 6341–6346 (2000).
Spadafora, C. et al. Complement receptor 1 is a sialic acid-independent erythrocyte receptor of Plasmodium falciparum. PLoS Pathog. 6, e1000968, https://doi.org/10.1371/journal.ppat.1000968 (2010).
Tham, W.-H. et al. Complement receptor 1 is the host erythrocyte receptor for Plasmodium falciparum PfRh4 invasion ligand. Proc. Nat. Acad. Sci. USA 107, 17327–17332, https://doi.org/10.1073/pnas.1008151107 (2010).
Tham, W.-H. et al. Plasmodium falciparum uses a key functional site in complement receptor type-1 for invasion of human erythrocytes. Blood 118, 1923–1933, https://doi.org/10.1182/blood-2011-03-341305 (2011).
Park, H. J. et al. Using mutagenesis and structural biology to map the binding site for the Plasmodium falciparum merozoite protein PfRh4 on the human immune adherence receptor. J. Biol. Chem., https://doi.org/10.1074/jbc.M113.520346 (2013).
Lim, N. T. et al. Characterization of Inhibitors and Monoclonal Antibodies That Modulate the Interaction between Plasmodium falciparum Adhesin PfRh4 with Its Erythrocyte Receptor Complement Receptor 1. J. Biol. Chem. 290, 25307–25321, https://doi.org/10.1074/jbc.M115.657171 (2015).
Jacquet, M. et al. Deciphering complement receptor type 1 interactions with recognition proteins of the lectin complement pathway. J. Immunol. 190, 3721–3731, https://doi.org/10.4049/jimmunol.1202451 (2013).
Garred, P. et al. Mannose-binding lectin is a disease modifier in clinical malaria and may function as opsonin for Plasmodium falciparum-infected erythrocytes. Infect. Immun. 71, 5245–5253 (2003).
Holmberg, V. et al. Mannose-binding lectin variant associated with severe malaria in young African children. Microbes Infect. 10, 342–348, https://doi.org/10.1016/j.micinf.2007.12.008 (2008).
Korir, J. C., Nyakoe, N. K., Awinda, G. & Waitumbi, J. N. Complement Activation by Merozoite Antigens of Plasmodium falciparum. PLoS ONE 9, https://doi.org/10.1371/journal.pone.0105093 (2014).
Williams, T. N. et al. Sickle cell trait and the risk of Plasmodium falciparum malaria and other childhood diseases. J. Infect. Dis. 192, 178–186, https://doi.org/10.1086/430744 (2005).
Scott, J. A. G. et al. Profile: The Kilifi Health and Demographic Surveillance System (KHDSS). Int. J. Epidemiol. 41, 650–657, https://doi.org/10.1093/ije/dys062 (2012).
O’Meara, W. P. et al. Effect of a fall in malaria transmission on morbidity and mortality in Kilifi, Kenya. Lancet 372, 1555–1562, https://doi.org/10.1016/S0140-6736(08)61655-4 (2008).
Fry, A. E. et al. Common variation in the ABO glycosyltransferase is associated with susceptibility to severe Plasmodium falciparum malaria. Hum. Mol. Gene.t 17, 567–576, https://doi.org/10.1093/hmg/ddm331 (2008).
Waterfall, C. M. & Cobb, B. D. Single tube genotyping of sickle cell anaemia using PCR-based SNP analysis. Nucleic Acids Res. 29, E119 (2001).
Chong, S. S., Boehm, C. D., Higgs, D. R. & Cutting, G. R. Single-tube multiplex-PCR screen for common deletional determinants of alpha-thalassemia. Blood 95, 360–362 (2000).
Rowe, J. A. et al. Blood group O protects against severe Plasmodium falciparum malaria through the mechanism of reduced rosetting. Proc. Nat. Acad. Sci. USA 104, 17471–17476, https://doi.org/10.1073/pnas.0705390104 (2007).
Otenio, W. et al. Sickle Cell Trait (HbAS) is Associated with Increased Expression of Erythrocyte Complement Regulatory Proteins CR1 and CD55 Levels in Children. Int. J. Trop. Dis. 3, 133–147 (2013).
Pascual, M., Lutz, H. U., Steiger, G., Stammler, P. & Schifferli, J. A. Release of vesicles enriched in complement receptor 1 from human erythrocytes. J. Immunol. 151, 397–404 (1993).
Cockburn, I. A., Donvito, B., Cohen, J. H. M. & Rowe, J. A. A simple method for accurate quantification of complement receptor 1 on erythrocytes preserved by fixing or freezing. J. Immunol. Methods 271, 59–64 (2002).
R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, http://www.R-project.org (2008).
Wickham, H. The Split-Apply-Combine Strategy for Data Analysis. J. Stat. Software 40 (2011).
Wickham, H. Ggplot2: elegant graphics for data analysis. (Springer, 2009).
lme4: Linear mixed-effects models using Eigen and S4 v. R package version 1.0-4 (2013).
Schmidt, C. Q., Kennedy, A. T. & Tham W. -H. More than just immune evasion: Hijacking complement by Plasmodium falciparum. Molecular Immunology 67(1), 71–84 (2015).
Acknowledgements
This paper is published with the permission of the Director of KEMRI. We would like to thank Jennifer Musyoki and Metrine Tendwa for their excellent laboratory support. Our thanks also to the staff of Kilifi County Hospital and the KEMRI-Wellcome Trust Research Programme, Kilifi for their help with data and sample collection. We are also grateful to the volunteers who donated blood used as controls in erythrocyte CR1 level determination. Lastly, we would like to acknowledge the generosity of the Immunology Cohort Study Group participants and their guardians, without whom this research would not be possible. This work was funded by the Wellcome Trust through an Edinburgh Clinical Academic training fellowship to OVS (grant number 101910) and through Senior Research Fellowships awarded to JAR (grant number 084226) and TNW (grant number 091758).
Author information
Authors and Affiliations
Contributions
Study design: O.V.S., T.N.W., J.A.R. Experimental work: O.V., D.H.O., E.N., A.M., S.U. Data analysis: O.V.S., E.M.H., J.A.R. Wrote paper: O.V.S., J.A.R. Reviewed and edited paper: all authors.
Corresponding author
Ethics declarations
Competing Interests
The authors declare that they have no competing interests.
Additional information
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Swann, O.V., Harrison, E.M., Opi, D.H. et al. No Evidence that Knops Blood Group Polymorphisms Affect Complement Receptor 1 Clustering on Erythrocytes. Sci Rep 7, 17825 (2017). https://doi.org/10.1038/s41598-017-17664-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-017-17664-9
- Springer Nature Limited