Martha A Clark1, Morgan M Goheen1, Nicholas A Spidale1, Raj S Kasthuri2, Anthony Fulford3, Carla Cerami4. 1. Department of Microbiology & Immunology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America. 2. Division of Hematology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America. 3. Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America. 4. Medical Research Council International Nutrition Group, London School of Hygiene and Tropical Medicine, London, United Kingdom & Medical Research Council Keneba, The Gambia.
Abstract
Plasmodium falciparum invasion of host erythrocytes is essential for the propagation of the blood stage of malaria infection. Additionally, the brief extracellular merozoite stage of P. falciparum represents one of the rare windows during which the parasite is directly exposed to the host immune response. Therefore, efficient invasion of the host erythrocyte is necessary not only for productive host erythrocyte infection, but also for evasion of the immune response. Host traits, such as hemoglobinopathies and differential expression of erythrocyte invasion ligands, can protect individuals from malaria by impeding parasite erythrocyte invasion. Here we combine RBC barcoding with flow cytometry to study P. falciparum invasion. This novel high-throughput method allows for the (i) direct comparison of P. falciparum invasion into different erythrocyte populations and (ii) assessment of the impact of changing erythrocyte population dynamics on P. falciparum invasion.
Plasmodium falciparum invasion of host erythrocytes is essential for the propagation of the blood stage of malaria infection. Additionally, the brief extracellular merozoite stage of P. falciparum represents one of the rare windows during which the parasite is directly exposed to the host immune response. Therefore, efficient invasion of the host erythrocyte is necessary not only for productive host erythrocyte infection, but also for evasion of the immune response. Host traits, such as hemoglobinopathies and differential expression of erythrocyte invasion ligands, can protect individuals from malaria by impeding parasite erythrocyte invasion. Here we combine RBC barcoding with flow cytometry to study P. falciparum invasion. This novel high-throughput method allows for the (i) direct comparison of P. falciparum invasion into different erythrocyte populations and (ii) assessment of the impact of changing erythrocyte population dynamics on P. falciparum invasion.
Malaria is responsible for significant morbidity and mortality in the developing world, causing an estimated 250 million infections and 1 million deaths annually[1]. Malaria can be caused by any of several different plasmodia species capable of infecting humans; of these, Plasmodium falciparum is the most malignant. Upon transmission to the human host by an Anopheles mosquito, plasmodial sporozoites migrate to and infect liver hepatocytes. Following the asymptomatic liver stage, merozoites are released into the bloodstream where they infect host red blood cells (RBCs). It is the RBC stage of infection that is responsible for all symptoms of disease. Upon entry into the host RBC the malaria parasite progresses through a 48 hour lifecycle. At the end of the 48 hour intra-erythrocytic lifecycle the host RBC is ruptured and newly formed merozoites are released into the serum to invade new host RBCs [2]. The extracellular merozoite stage of the malaria parasite is very brief. Invasion occurs within 10 minutes of RBCrupture and the 48 hour intra-erythrocyte cycle begins again [3]. Successful invasion of the host RBC is essential for propagation of the parasite. Additionally the merozoite is one of the only parasite stages directly exposed to the host immune system. As a result, merozoite invasion represents a promising point of attack for antimalarial drugs and vaccines [4].The RBC population of a given individual is heterogeneous; RBC age, anemia, acidosis, infection, and Glucose-6-phosphate dehydrogenase (G6PD) deficiency are examples of a few conditions which can contribute to heterogeneity in an individual's RBC population [5]–[9]. Therefore, the most relevant approach for studying how RBC heterogeneity influences P. falciparum invasion would be to combine different RBC populations in a single culture condition and study subsequent P. falciparum infection. In addition, direct comparison of P. falciparum invasion into different “target” RBC populations would allow for increased sensitivity to small differences that may exist between parasite invasion of different RBCs. The diversity and redundancy that exists in both host RBC ligands and merozoite invasion receptors makes accuracy in the measurement of invasion important [4], [10], [11].Flow cytometry has revolutionized the study of malaria; increasing the throughput, sensitivity, and accuracy of many malaria assays [12]. Pattanapanyasat et al. was the first to use the power of flow cytometry to distinguish two discrete RBC populations in a single culture condition [13], [14]. Since only one of the two RBC populations was labeled, the approach taken by Pattanapanyasat et al. does not account for “contaminant” RBCs present in the inoculum. Moreover, Pattanapanyasat et al. examined overall growth of P. falciparum and not invasion. Previous groups have dealt with “contaminant” RBCs by enzymatically treating the parasitized RBC (pRBC) inoculum to remove parasite invasion ligands from uninfected RBCs present in the inoculum [15], [16]. Alternatively, a two-color flow cytometry based invasion assay developed by Theron et al. addresses the issue of “contaminant” RBCs by labeling “target” RBCs with CellTrace dyes, making “contaminant” and “target” RBCs clearly distinguishable from one another [17]. Neither the enzyme treatment of inoculum RBCs nor the two-color approach allow for the direct comparison of P. falciparum invasion into different “target” RBC populations contained in the same well.Here we utilize fluorescent RBC barcoding to build upon the methods described by Theron et al.[17] and Pattanapanyasat et al.[14]. RBC barcoding allows direct comparison of P. falciparum invasion into two distinct “target” RBC populations in the same well. This high-throughput approach provides increased sensitivity for detecting differences in merozoite invasion as well as a method for studying how different RBC populations interact to determine malaria infection.
Materials and Methods
P. falciparum culture
P. falciparum parasite lines 3D7 (MR4, MRA-102), Dd2 (MR4, MRA-156), and FCR3-FMG/Gambia (MR4, MRA-736) were routinely cultured in O positive (O+) RBCs within two weeks of being obtained from healthy individuals. RBCs were collected at the Clinical and Translational Research Center at the University of North Carolina, Chapel Hill and their use for this study was approved by the Institutional Review Board at the University of North Carolina at Chapel Hill (IRB# 09-0559). Written consent was obtained from all donors using a consent form specifically approved by the IRB. Cultures were maintained at 2% hematocrit in complete media containing RPMI 1640 with 10% AlbuMAX II, 1 mM hypoxanthine, 20 mM L-glutamine, 0.45% glucose, and 10 µg/L gentamicin (ACM). Cultures were shaken at 37°C in 5% O2, 5% CO2 and 90% N. Parasite density was maintained between 0.5% and 10% P. falciparum parasitized RBCs (pRBCs). pRBC cultures were synchronized using 5% D-sorbitol to select for ring stage parasites, followed by Magnetic Activated Cell Sorting (MACS) (Miltenyi Biotec) isolation of hemozoin containing trophozoite and schizont stage pRBCs 24 hours later [18].
Barcoded RBC invasion assay
Briefly, O+ RBCs were labeled at 2% hematocrit in RPMI with 5 µM of either CellTrace Violet (RBCViolet) or DDAO (RBCDDAO) (Invitrogen) for two hours with shaking at 37°C, washed twice with ACM, incubated 30 minutes in ACM with shaking at 37°C, and finally washed twice with RPMI and stored in RPMI at 4°C. DDAO and Violet fluorescent intensity varied between experiments. To confirm distinct separation of the fluorescent signals of RBCunlabeled, RBCDDAO, and RBCViolet, each was analyzed by flow cytometry alone and in combination. Invasion assays were performed with fluorescently labeled RBCs within two days of labeling. For barcoded RBC invasion assays, RBCDDAO and RBCViolet were counted using a Nexcelom Bioscience Cellometer Auto 2000 and 2×107 total RBCs were delivered in triplicate into a 96 well plate. MACS purified pRBCs from routine cultures were counted using a Nexcelom Bioscience Cellometer Auto 2000 and 2×105 pRBCs were added to each well. Experiments were maintained for 18–24 hours under standard culture conditions to allow for schizont rupture and subsequent invasion of RBCDDAO and RBCViolet. Following merozoite invasion, cells were stained with 1× DNA dye SYBR Green I (Invitrogen), fixed with 1% paraformaldehyde and 0.0075% glutaraldehyde in Alsever's Solution (Sigma) as described previously[19], and analyzed by flow cytometry or microscopy. A small (0.16%) double positive Violet and DDAORBC population – expected to be RBC doublets that escaped singlet gating – was occasionally observed and was excluded from the analysis. We have confirmed that dye is not transferred between RBCs during the course of barcoded RBC invasion experiments (data not shown).
Enzyme treatment of RBCs
RBCs were treated with trypsin, chymotrypsin, or neuraminidase in accordance with the Sanger Institute flow cytometry-based invasion phenotyping protocol (http://www.sanger.ac.uk/research/projects/malariaprogramme-rayner/sop-flow-cytometry-invasion-assay.pdf). Briefly, RBCs labeled with either CellTrace Violet or DDAO were treated with 0.02 U/mL neuraminidase (Sigma), 50 µg/mL trypsin (Sigma), or 0.91 mg/mL chymotrypsin (Sigma) at 2% hematocrit in RPMI for 1 hour with shaking at 37°C, washed twice with ACM, incubated 30 minutes in ACM with shaking at 37°C, and finally washed twice with RPMI. Enzyme treated RBCs were stored in RPMI at 4°C for up to 1 day before being used in experiments. Untreated RBCDDAO and RBCViolet were combined and included as a control for all experiments.
Microscopy
pRBCDDAO and pRBCViolet were stained with 1× DNA dye SYBR Green I, fixed with 1% paraformaldehyde and 0.0075% glutaraldehyde, and then viewed by oil immersion with a 63X/1.4 numerical aperture Oil Plan Apo objective lens on a Zeiss CLSM 710 Spectral Confocal Laser Scanning Microscope. Images were captured with Zeiss ZEN 2011 software.
Flow cytometry analysis
Flow cytometry was performed at the UNC Flow Cytometry Core Facility on a Beckman-Coulter (Dako) CyAn ADP cytometer. Channels and probes used on the CyAn ADP Cytometer included: CellTrace Violet (405 nM excitation, 450/50 bandpass), SYBR Green I (488 nm excitation, 530/40 bandpass), and CellTrace DDAO (635 nm excitation, 665/20 bandpass). Detector gain settings varied between experiments to optimize signal but were kept constant within individual experiments and spectral overlap compensation was not necessary with this configuration. P. falciparumpRBCs were gated on SYBR Green I signal. Dako Cyan data was collected and analyzed with Summit v4.3.01. Linear amplification of forward scatter was used to set event threshold in order to exclude cell debris, microparticles and doublets. For all experiments samples were diluted to 0.001–0.002% haematocrit and 100,000–500,000 total events were collected.
Data analysis and statistical methods
All experiments were performed in triplicate. Results are from either one representative experiment of at least three independent experiments or the combined results of at least three independent experiments. To compare the relative susceptibility of two different RBC populations to P. falciparum invasion, we determined the susceptibility index (SI) with an unadjusted Odds Ratio (performed with Stata/IC, v10, Stata Corp, College Station, TX) as shown in the equation:The SI represents the relative risk of the two RBC populations (represented by X and Y in the equation) to P. falciparum invasion. For experiments comparingP. falciparum invasion of RBCDDAO and RBCViolet, X = RBCDDAO and Y = RBCViolet. For experiments comparing P. falciparum invasion of untreated (RBCØ) and enzyme treated (RBCN, RBCT, or RBCC), X = enzyme treated RBC and Y = RBCØ. An SI of 1.0 indicates no difference in P. falciparum invasion of the two RBC populations. P. falciparum invasion rate is the number of RBC invasions per inoculated pRBC. Linear regression was employed to investigate the associations between RBC enzyme treatment and parasite invasion in vitro using the number of each RBC population as the independent variable and total invasions/1×105 RBCs as the dependent variable. Analysis of covariance (ANCOVA) was conducted to determine whether the invasion into the two groups were the same. An alpha of 0.05 was set a priori to determine statistically significant differences.
Results
Barcoding RBCs with CellTrace dyes DDAO and Violet allows for the direct comparison of P. falciparum invasion into different RBC populations
To develop a P. falciparumRBC invasion assay that would allow for direct comparison of invasion of P. falciparum into two different RBC populations, we utilized a fluorescent RBC staining approach that permits the definitive detection of two distinct RBC populations by flow cytometry. DDAO was used to label the first RBC population (RBCDDAO) (
) and CellTrace Violet was used to label the second RBC population (RBCViolet) (
). As has been reported with CellTrace DDAO [17], CellTrace Violet does not affect P. falciparumRBC invasion (). We next examined whether RBCDDAO could be combined with RBCViolet in the barcoded RBC invasion assay. An equal number of RBCDDAO and RBCViolet were combined, inoculated with Magnetic Activated Cell Sorting (MACS) purified trophozoite stage pRBCs and incubated for 18–24 hours to allow for schizont rupture and subsequent merozoite invasion of RBCDDAO and RBCViolet. RBCDDAO is readily distinguished from RBCViolet by microscopy (
) and flow cytometry (
). SYBR Green I staining allows for the identification of parasitized RBCDDAO (pRBCDDAO) and pRBCViolet by both microscopy (
) and flow cytometry (
). No spectral interference of the three fluorescent dyes was detected by flow cytometry or microscopy. Additionally, we confirmed that dye is not transferred between RBCs during the course of barcoded RBC invasion experiments (data not shown). The relative risk of two different target RBC populations to P. falciparum invasion was determined with an unadjusted Odds Ratio. We have termed this measurement the Susceptibility Index (SI). An SI of 1.0 indicates no difference in the risk of invasion of two target RBC populations. The SI of RBCDDAO and RBCViolet for 3D7, Dd2, and FCR3-FMG parasite invasion was 0.97 (confidence interval [CI] 0.96–0.99), 0.90 (CI 0.89–0.91), and 1.08 (CI 1.06–1.09) respectively (
).
Figure 1
RBCs barcoded with CellTrace DDAO and Violet can be combined to directly compare P. falciparum invasion in an invasion assay.
RBCs were labeled with 5 µM of either DDAO (A) or CellTrace Violet (B). Cells were then combined and infected with MACS purified unlabeled pRBCs. Experiments were incubated 18–24 hours. Cells were then stained with DNA dye SYBR Green I, fixed, and examined by brightfield (C) and fluorescence microscopy (D, E) and by flow cytometry (F–H). (A) shows red channel only, (B) shows violet channel only, (C) shows brightfield, (D) shows green channel only and (E) shows merge of red, violet, and green channels. (F) Flow cytometry plot of RBCs stained with CellTrace DDAO (R1) and CellTrace Violet (R4) and non-stained pRBC (R3). (G) Flow cytometry plot shows DDAO negative pRBCs (R5), DDAO negative uninfected RBCs (R7), DDAO positive pRBCs (R6) and DDAO positive uninfected RBCs (R8). (H) Flow cytometry plot shows CellTrace Violet negative pRBCs (R9), CellTrace Violet negative uninfected RBCs (R11), CellTrace Violet positive pRBCs (R10) and CellTrace Violet positive uninfected RBCs (R12).
Figure 2
Direct comparison of P. falciparum strains 3D7, Dd2 and FCR3-FMG invasion into RBCs barcoded with DDAO and CellTrace Violet.
Equal numbers (1×107) of RBCs stained with either 5 µM CellTrace DDAO (RBCDDAO) or CellTrace Violet (RBCViolet) were combined (total of 2×107 RBCs per well) and inoculated with 2×105 MACS purified trophozoite stage P. falciparum strains 3D7, Dd2 or FCR3-FMG. Invasion experiments were incubated 18–24 hours to allow for rupture of schizonts and subsequent invasion of merozoites into labeled RBCs, then stained with DNA dye SYBR Green I (to identify pRBCs), fixed and analyzed by flow cytometry. The Susceptibility Index (SI), an unadjusted Odds Ratio assessing the relative risk of RBCDDAO and RBCViolet to 3D7, Dd2, and FCR3-FMG invasion. The marker represents the SI point estimate and the bar represents the 95% confidence interval (CI). A SI of 1.0 indicates no difference in parasite invasion of two RBC populations. Data is the combination of four, seven, and three independent experiments performed in triplicate with 3D7, Dd2, and FCR3-FMG respectively.
RBCs barcoded with CellTrace DDAO and Violet can be combined to directly compare P. falciparum invasion in an invasion assay.
RBCs were labeled with 5 µM of either DDAO (A) or CellTrace Violet (B). Cells were then combined and infected with MACS purified unlabeled pRBCs. Experiments were incubated 18–24 hours. Cells were then stained with DNA dye SYBR Green I, fixed, and examined by brightfield (C) and fluorescence microscopy (D, E) and by flow cytometry (F–H). (A) shows red channel only, (B) shows violet channel only, (C) shows brightfield, (D) shows green channel only and (E) shows merge of red, violet, and green channels. (F) Flow cytometry plot of RBCs stained with CellTrace DDAO (R1) and CellTrace Violet (R4) and non-stained pRBC (R3). (G) Flow cytometry plot shows DDAO negative pRBCs (R5), DDAO negative uninfected RBCs (R7), DDAO positive pRBCs (R6) and DDAO positive uninfected RBCs (R8). (H) Flow cytometry plot shows CellTrace Violet negative pRBCs (R9), CellTrace Violet negative uninfected RBCs (R11), CellTrace Violet positive pRBCs (R10) and CellTrace Violet positive uninfected RBCs (R12).
Direct comparison of P. falciparum strains 3D7, Dd2 and FCR3-FMG invasion into RBCs barcoded with DDAO and CellTrace Violet.
Equal numbers (1×107) of RBCs stained with either 5 µM CellTrace DDAO (RBCDDAO) or CellTrace Violet (RBCViolet) were combined (total of 2×107 RBCs per well) and inoculated with 2×105 MACS purified trophozoite stage P. falciparum strains 3D7, Dd2 or FCR3-FMG. Invasion experiments were incubated 18–24 hours to allow for rupture of schizonts and subsequent invasion of merozoites into labeled RBCs, then stained with DNA dye SYBR Green I (to identify pRBCs), fixed and analyzed by flow cytometry. The Susceptibility Index (SI), an unadjusted Odds Ratio assessing the relative risk of RBCDDAO and RBCViolet to 3D7, Dd2, and FCR3-FMG invasion. The marker represents the SI point estimate and the bar represents the 95% confidence interval (CI). A SI of 1.0 indicates no difference in parasite invasion of two RBC populations. Data is the combination of four, seven, and three independent experiments performed in triplicate with 3D7, Dd2, and FCR3-FMG respectively.
The barcoded RBC invasion assay allows for the direct comparison of P. falciparum invasion into two different RBC populations
Neuraminidase (N), trypsin (T), and chymotrypsin (C) are commonly used to study P. falciparum merozoite RBC invasion. Treatment of RBCs with any of these enzymes reduces merozoite RBC invasion. Different P. falciparum laboratory strains as well as clinical isolates exhibit different enzyme sensitivities[2]. To demonstrate the functionality of the barcoded RBC invasion assay, we have utilized RBCs treated with these enzymes in combination with P. falciparum strains 3D7 and Dd2, which have well characterized invasion phenotypes. Specifically, 3D7 invasion is less sensitive to neuraminidase than Dd2, 3D7 and Dd2 invasion is similarly sensitive to trypsin, and Dd2 invasion is less sensitive to chymotrypsin than 3D7 [17]. Equal numbers of untreated (RBCØ) and either neuraminidase (RBCN), trypsin (RBCT), or chymotrypsin (RBCC) treated RBCs were combined in the barcoded RBC invasion assay and the SI of RBCØ and enzyme treated RBCs to 3D7 and Dd2 parasite invasion was determined. The SI of RBCØ and RBCN, RBCT, and RBCC to 3D7 parasite invasion was 0.56 (CI 0.54–0.58), 0.33 (CI 0.32–0.35), and 0.04 (CI 0.04–0.04) respectively. The SI of RBCØ and RBCN, RBCT, and RBCC to Dd2 invasion was 0.03 (CI 0.02–0.03), 0.33 (CI 0.31–0.34), and 0.15 (CI 0.14–0.15) respectively. The SI of all RBCØ and enzyme treated RBC combinations were significantly different from the control conditions (combination of RBCDDAO and RBCViolet) (
). These results are consistent with previously reported enzyme sensitivities of strains 3D7 and Dd2.
Figure 3
Direct comparison of P. falciparum invasion into untreated and enzyme treated RBCs.
RBCs were labeled with either CellTrace DDAO or Violet before enzyme treatment (neuraminidase, trypsin, or chymotrypsin). For barcoded RBC invasion assays, 1×107 RBCs of differentially labeled and enzyme treated RBC populations were combined for a total of 2×107 RBCs per well. For two-color invasion assays 2×107 of each RBC population – untreated (RBCØ), neuraminidase (RBCN), trypsin (RBCT), and chymotrypsin (RBCC) – were inoculated into separate wells. Two-color and barcoded RBC invasion assays were inoculated with 2×105 of MACS enriched trophozoite stage 3D7 or Dd2 P. falciparum parasites and invasion assays were performed as previously described. (A) SI for 3D7 and Dd2 invasion into barcoded RBC invasion assays containing RBCØ and either RBCN, RBCT, or RBCC. The marker represents the SI point estimate and the bar represents the 95% CI. (B) 3D7 and Dd2 invasion of RBCN, RBCT and RBCC normalized to invasion of RBCØ for both two-color and barcoded RBC invasion assays. Bars represent the mean and error bars represent the SD. Student's t-tests were used to calculate differences in invasion between two-color and barcoded RBC invasion assays, *p<0.005, **p<0.0002, and ***p<7E-6. Data is from a representative experiment of three independent experiments, each performed in triplicate.
Direct comparison of P. falciparum invasion into untreated and enzyme treated RBCs.
RBCs were labeled with either CellTrace DDAO or Violet before enzyme treatment (neuraminidase, trypsin, or chymotrypsin). For barcoded RBC invasion assays, 1×107 RBCs of differentially labeled and enzyme treated RBC populations were combined for a total of 2×107 RBCs per well. For two-color invasion assays 2×107 of each RBC population – untreated (RBCØ), neuraminidase (RBCN), trypsin (RBCT), and chymotrypsin (RBCC) – were inoculated into separate wells. Two-color and barcoded RBC invasion assays were inoculated with 2×105 of MACS enriched trophozoite stage 3D7 or Dd2P. falciparum parasites and invasion assays were performed as previously described. (A) SI for 3D7 and Dd2 invasion into barcoded RBC invasion assays containing RBCØ and either RBCN, RBCT, or RBCC. The marker represents the SI point estimate and the bar represents the 95% CI. (B) 3D7 and Dd2 invasion of RBCN, RBCT and RBCC normalized to invasion of RBCØ for both two-color and barcoded RBC invasion assays. Bars represent the mean and error bars represent the SD. Student's t-tests were used to calculate differences in invasion between two-color and barcoded RBC invasion assays, *p<0.005, **p<0.0002, and ***p<7E-6. Data is from a representative experiment of three independent experiments, each performed in triplicate.We next compared the two-color [17] and barcoded RBC invasion assays. The two-color invasion assay measures P. falciparum invasion of different RBC populations in adjacent but separate culture wells, while the barcoded RBC invasion assay directly compares P. falciparum invasion of two different RBC populations in the same culture well. In both assays, the invasion of each parasite strain into RBCN, RBCT and RBCC was normalized to invasion into RBCØ. 3D7 and Dd2 parasite sensitivity to all three enzymes was significantly greater in the barcoded RBC than the two-color invasion assay. Specifically, 3D7 and Dd2 invasion into RBCN was 67.7% (SD±1.8) and 4.5% (SD±0.2) in the two-color assay and 56.1% (SD±1.9) and 2.6% (SD±0.7) in the barcoded RBC assay. 3D7 and Dd2 invasion into RBCT was 65.7% (SD±1.8) and 69.6% (SD±1.0) in the two-color assay and 33.1% (SD±0.45) and 32.5% (SD±0.9) in the barcoded RBC assay. Finally, 3D7 and Dd2 invasion into RBCC was 5.1% (SD±0.4) and 32.9% (SD±0.9) in the two-color assay and 3.9% (SD±0.1) and 15.6% (SD±1.1) in the barcoded RBC assay (
). Together, these results show that the two-color and barcoded RBC assays reveal the same trends but that the barcoded RBC assay is better able to detect small differences in P. falciparum invasion.
The barcoded RBC invasion assay can be used to study the effect of RBC population dynamics on P. falciparum infection
To study the effect of RBC population dynamics on P. falciparum invasion, we replaced 10%, 50%, and 90% of RBCØ with RBCN, RBCT, or RBCC and assessed P. falciparum invasion. The SI reflects the relative invasion of P. falciparum into two different RBC populations. However, the SI does not reflect the total invasion capacity of P. falciparum. Therefore for RBCØ replacement experiments, we examined the invasion rate of P. falciparum strains 3D7 and Dd2 into wells containing different combination of RBCØ and enzyme treated RBC. The invasion rate of both 3D7 and Dd2 parasites decreased significantly as RBCØ decreased from 90% to 10% and the enzyme treated RBC population (RBCN, RBCT, or RBCC) increased from 10% to 90% of the total RBC population (
and
). Parasite invasion did not significantly change for control conditions in which RBCViolet were replaced with RBCDDAO ().
Figure 4
Invasion of P. falciparum strain 3D7 decreases as untreated RBCs are replaced with enzyme treated RBCs.
RBCs were labeled with either CellTrace DDAO or Violet before enzyme treatment (neuraminidase, trypsin, or chymotrypsin). 1.8×107, 1×107, and 2×106 non-enzymatically treated RBCs (RBCØ) were combined with 2×106, 1×107, and 1.8×107 of enzyme treated RBCs (RBCN, RBCT, or RBCC) respectively to achieve 10∶1, 1∶1, and 1∶10 combinations of RBCØ to either RBCN, RBCT, or RBCC in the barcoded RBC invasion assay. Invasion assays were inoculated with 2×105 of MACS purified trophozoite stage P. falciparum strain 3D7 parasites and invasion assays were performed as previously described. Data is from a single representative experiment of three independent experiments performed in triplicate. (A) Rate of 3D7 invasion into 1∶10, 1∶1, and 10∶1 combinations of RBCØ and either RBCN, RBCT, or RBCC. Bars represent the mean invasion rate and error bars represent the SD. Elongated triangles below the X-axis represent the replacement of RBCØ (white triangle) with either RBCN, RBCT, or RBCC (gray triangle) in the total RBC population. Student's t-tests were used to compare invasion rates, *p<0.02 and **p<0.0002. (B, C, and D) Data shows the number of invasion events into RBCØ (triangles) and either RBCN, RBCT, or RBCC (circles) as the frequency of each RBC type increases from 10% to 90% of the total RBC population. Linear regression was used to determine the best fit line for P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. The null hypothesis was no difference between RBCØ and either RBCN, RBCT, or RBCC (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.0001. (E) RBCØ datum from panels B–D were superimposed to compare invasion into RBCØ when RBCØ was combined with RBCN (circles) RBCT (triangles) or RBCC (diamonds). Linear regression was used to determine the best fit line for RBCØ invasion data. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ. The null hypothesis was that there would be no difference in the invasion of RBCØ in the presence of the other RBC populations (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.2599.
Figure 5
Invasion of P. falciparum strain Dd2 decreases as untreated RBCs are replaced with enzyme treated RBCs.
RBCs were labeled with either CellTrace DDAO or Violet before enzyme treatment (neuraminidase, trypsin, or chymotrypsin). 1.8×107, 1×107, and 2×106 non-enzymatically treated RBCs (RBCØ) were combined with 2×106, 1×107, and 1.8×107 of enzyme treated RBCs (RBCN, RBCT, or RBCC) respectively to achieve 10∶1, 1∶1, and 1∶10 combinations of RBCØ to either RBCN, RBCT, or RBCC in the barcoded RBC invasion assay. Invasion assays were inoculated with 2×105 of MACS purified trophozoite stage P. falciparum strain Dd2 parasites and invasion assays were performed as previously described. Data is from a single representative experiment of three independent experiments performed in triplicate. (A) Rate of Dd2 invasion into 1∶10, 1∶1, and 10∶1 combinations of RBCØ and either RBCN, RBCT, or RBCC. Bars represent the mean invasion rate and error bars represent the SD. Elongated triangles below the X-axis represent the replacement of RBCØ (white triangle) with either RBCN, RBCT, or RBCC (gray triangle) in the total RBC population. Student's t-tests were used to compare invasion rates, *p<0.04, **p<0.0003, and ***p<5E-5. (B, C, and D) Data shows the number of invasion events into RBCØ (triangles) and either RBCN, RBCT, or RBCC (circles) as the frequency of each RBC type increases from 10% to 90% of the total RBC population. Linear regression was used to determine the best fit line for P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. The null hypothesis was no difference between RBCØ and either RBCN, RBCT, or RBCC (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.0002. (E) RBCØ data from panels B–D were superimposed to compare invasion into RBCØ when RBCØ was combined with RBCN (circles) RBCT (triangles) or RBCC (diamonds). Linear regression was used to determine the best fit line for RBCØ invasion data. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ. The null hypothesis was that there would be no difference in the invasion of RBCØ in the presence of the other RBC populations (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.8335.
Invasion of P. falciparum strain 3D7 decreases as untreated RBCs are replaced with enzyme treated RBCs.
RBCs were labeled with either CellTrace DDAO or Violet before enzyme treatment (neuraminidase, trypsin, or chymotrypsin). 1.8×107, 1×107, and 2×106 non-enzymatically treated RBCs (RBCØ) were combined with 2×106, 1×107, and 1.8×107 of enzyme treated RBCs (RBCN, RBCT, or RBCC) respectively to achieve 10∶1, 1∶1, and 1∶10 combinations of RBCØ to either RBCN, RBCT, or RBCC in the barcoded RBC invasion assay. Invasion assays were inoculated with 2×105 of MACS purified trophozoite stage P. falciparum strain 3D7 parasites and invasion assays were performed as previously described. Data is from a single representative experiment of three independent experiments performed in triplicate. (A) Rate of 3D7 invasion into 1∶10, 1∶1, and 10∶1 combinations of RBCØ and either RBCN, RBCT, or RBCC. Bars represent the mean invasion rate and error bars represent the SD. Elongated triangles below the X-axis represent the replacement of RBCØ (white triangle) with either RBCN, RBCT, or RBCC (gray triangle) in the total RBC population. Student's t-tests were used to compare invasion rates, *p<0.02 and **p<0.0002. (B, C, and D) Data shows the number of invasion events into RBCØ (triangles) and either RBCN, RBCT, or RBCC (circles) as the frequency of each RBC type increases from 10% to 90% of the total RBC population. Linear regression was used to determine the best fit line for P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. The null hypothesis was no difference between RBCØ and either RBCN, RBCT, or RBCC (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.0001. (E) RBCØ datum from panels B–D were superimposed to compare invasion into RBCØ when RBCØ was combined with RBCN (circles) RBCT (triangles) or RBCC (diamonds). Linear regression was used to determine the best fit line for RBCØ invasion data. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ. The null hypothesis was that there would be no difference in the invasion of RBCØ in the presence of the other RBC populations (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.2599.
Invasion of P. falciparum strain Dd2 decreases as untreated RBCs are replaced with enzyme treated RBCs.
RBCs were labeled with either CellTrace DDAO or Violet before enzyme treatment (neuraminidase, trypsin, or chymotrypsin). 1.8×107, 1×107, and 2×106 non-enzymatically treated RBCs (RBCØ) were combined with 2×106, 1×107, and 1.8×107 of enzyme treated RBCs (RBCN, RBCT, or RBCC) respectively to achieve 10∶1, 1∶1, and 1∶10 combinations of RBCØ to either RBCN, RBCT, or RBCC in the barcoded RBC invasion assay. Invasion assays were inoculated with 2×105 of MACS purified trophozoite stage P. falciparum strain Dd2 parasites and invasion assays were performed as previously described. Data is from a single representative experiment of three independent experiments performed in triplicate. (A) Rate of Dd2 invasion into 1∶10, 1∶1, and 10∶1 combinations of RBCØ and either RBCN, RBCT, or RBCC. Bars represent the mean invasion rate and error bars represent the SD. Elongated triangles below the X-axis represent the replacement of RBCØ (white triangle) with either RBCN, RBCT, or RBCC (gray triangle) in the total RBC population. Student's t-tests were used to compare invasion rates, *p<0.04, **p<0.0003, and ***p<5E-5. (B, C, and D) Data shows the number of invasion events into RBCØ (triangles) and either RBCN, RBCT, or RBCC (circles) as the frequency of each RBC type increases from 10% to 90% of the total RBC population. Linear regression was used to determine the best fit line for P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ, RBCN, RBCT, and RBCC. The null hypothesis was no difference between RBCØ and either RBCN, RBCT, or RBCC (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.0002. (E) RBCØ data from panels B–D were superimposed to compare invasion into RBCØ when RBCØ was combined with RBCN (circles) RBCT (triangles) or RBCC (diamonds). Linear regression was used to determine the best fit line for RBCØ invasion data. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCØ. The null hypothesis was that there would be no difference in the invasion of RBCØ in the presence of the other RBC populations (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.8335.Utilizing the ability to distinguish RBCØ and enzyme treated RBCs provided by the barcoded RBC invasion assay, we next characterized the invasion of P. falciparum strains 3D7 and DD2 into RBCØ and RBCN, RBCT, and RBCC as the number of each increased in the total RBC population. We observed that the number of 3D7 invasions into all RBC populations increased linearly as each increased in frequency from 10% to 90% of the total RBC population, and that 3D7 invasion as a function of either RBCN, RBCT, or RBCC number was significantly less than that of RBCØ, p<0.0001 (
). Similar analysis of Dd2 invasion of RBCØ, RBCN RBCT, and RBCC also increased linearly as each increased from 10% to 90% of the total RBC population. As with 3D7 invasion, Dd2 invasion as a function of either RBCN, RBCT, or RBCC number was significantly less than that of RBCØ, p<0.0002 (
). Together this data demonstrates the capacity of the barcoded RBC invasion assay to study changing RBC population dynamics and P. falciparum invasion.Finally, we investigated the hypothesis that the presence of one RBC population may impact the susceptibility of another to P. falciparum invasion. To determine whether the presence of RBCN, RBCT, or RBCC affects 3D7 or Dd2 invasion into RBCØ, we compared 3D7 and Dd2 invasion of RBCØ when RBCØ were replaced with either RBCN, RBCT, or RBCC and the total number of RBCs remained constant. We observed that the number of 3D7 and Dd2 invasions into RBCØ was not affected by the presence of RBCN, RBCT, or RBCC (
and
).
Discussion
In the present study we have coupled RBC barcoding and flow cytometry to develop a P. falciparum invasion assay in which parasite invasion of two RBC populations is directly compared. The use of CellTrace dyes to label or “barcode” cell populations was developed to perform high-throughput drug screening and signal profiling [20]. Here, we demonstrate the first application of this approach to “barcode” RBC populations with the aim of studying P. falciparum invasion within a heterogeneous population.We've demonstrated that RBCs labeled with CellTrace dyes DDAO and Violet and subsequently combined can be definitively identified by both microscopy and flow cytometry. Furthermore, we demonstrate that parasitized DDAO and Violet RBCs can be identified with SYBR Green I staining (
). With the introduction of the susceptibly index (SI) as a robust tool for analyzing the relative risk of two RBC populations to P. falciparum invasion, we proceeded to demonstrate that DDAO and Violet RBCs are similarly susceptible to P. falciparum strains 3D7, Dd2, and FCR3-FMG invasion (
).Having established that P. falciparum invasion of two RBC populations may be compared directly by barcoding RBCs with CellTrace dyes, we proceeded to pursue the power of the barcoded RBC invasion assay by directly comparing P. falciparum invasion into untreated and enzyme treated RBCs. The effect of enzymatically treating RBCs with neuraminidase, trypsin, and chymotrypsin on P. falciparum invasion has been well studied [21] and therefore served as an ideal system for validating the barcoded RBC invasion assay. Direct comparison of P. falciparum (strains 3D7 and Dd2) invasion of RBCØ and either RBCN, RBCT, or RBCC with the barcoded RBC invasion assay resulted in the same invasion trends as previously described [17]. Moreover, we observed 3D7 and Dd2 sensitivity to all three enzymes to be significantly greater in the barcoded RBC invasion assay (direct measure of invasion) as compared to the two-color invasion assay (indirect measure of invasion) (
). This is consistent with Pattanapanyasat et al. 's conclusion that direct comparison of parasite growth allows for greater sensitivity in detecting small differences in growth rates between physiologically different RBCs than independent growth rate assessment [13]. Assay sensitivity is very important to consider as there is diversity in both host RBC and parasite factors which interact to shape malaria pathogenesis [4], [22]. In the case of merozoite RBC invasion, the redundancy in parasite invasion pathways increases the necessity of sensitive methods [23]–[27].The RBC is essential for the erythrocytic stage of malaria infection; consequentially humanRBC traits, such as hemoglobinopathies [22], G6PD deficiency [28] and differential expression of host RBC invasion ligands [29], have arisen in the human population providing protection from malaria. In addition, RBC age as well as other potential factors such as iron deficiency, oxidative stress, and heterozygous G6PD deficiency result in heterogeneity in a single individual's RBC population [30]. As RBC age, oxidative damage, and G6PD enzyme levels are all known to impact erythrocyte stage P. falciparum infection [15]–[18], it is likely that the changing dynamics of an individual's RBC population has the potential to dramatically impact the course of the disease [31]. The barcoded RBC invasion assay introduces the first method for studying the effect of changing RBC populations on P. falciparum infection. In the present study, we demonstrate that replacing a susceptible RBC population (RBCØ) with a less susceptible RBC population (enzyme treated RBCs) reduces the invasion rate of P. falciparum (
and
). Furthermore, barcoding allowed for the quantitation of merozoite invasion events into RBCØ and enzyme treated RBC populations as the number of each increased in the total RBC population. We show the number of P. falciparum invasion events into RBCØ and enzyme treated RBCs is linearly related to the RBC number. As would be expected, P. falciparum invasion of RBCØ and enzyme treated RBCs as a function of RBC number is significantly different (
and
). Finally, we show that P. falciparum invasion of untreated RBCs is not competitive as it is not affected by the presence of enzyme treated RBCs (
and
).In conclusion, we have developed a sensitive, high-throughput experimental assay for directly comparing the susceptibility of different RBC populations to P. falciparum merozoite invasion. We additionally demonstrate how the barcoded RBC invasion assay may be utilized to study the impact of changing RBC population dynamics on overall P. falciparum invasion. Furthermore, by using different Celltrace dyes and/or different concentrations of a given dye, the RBC barcoding approach is easily modifiable and could be expanded to study multiple RBC populations simultaneously. Future experimentation using this approach will provide invaluable insight into the relationship between P. falciparum and its human host.Barcoding RBCs with CellTrace Violet does not impact
invasion. An equal number of RBCViolet and RBCunlabeled (1×107) were combined for a total of 2×107 uninfected RBCs per well, and inoculated with 2×105 MACS purified trophozoite stage P. falciparum strains 3D7, Dd2 or FCR3-FMG. Invasion experiments were incubated 18–24 hours to allow for rupture of schizonts and subsequent invasion of merozoites into labeled RBCs, then stained with DNA dye SYBR Green I (to identify pRBCs), fixed and analyzed by flow cytometry. The SI, an unadjusted odds ratio assessing the relative risk of RBCViolet and RBCunlabeled to 3D7 and Dd2 invasion, was determined. The marker represents the SI point estimate and the bar represents the 95% confidence interval (CI). Data is from a representative experiment of three independent experiments, each performed in triplicate.(TIF)Click here for additional data file.Effect of replacing CellTrace Violet barcoded RBCs with DDAO barcoded RBCs on
strain 3D7 and Dd2 invasion. RBCs were labeled with either CellTrace DDAO or Violet and 1.8×107, 1×107, and 2×106 Violet labeled RBCs (RBCViolet) were combined with 2×106, 1×107, and 1.8×107 of DDAO labeled RBCs (RBCDDAO) to achieve 10∶1, 1∶1, and 1∶10 combinations of RBCViolet to RBCDDAO in the barcoded RBC invasion assay. Invasion assays were inoculated with 2×105 of MACS purified trophozoite stage P. falciparum strain 3D7 or Dd2 parasites and invasion assays were performed as previously described. Data is from a single representative experiment of three independent experiments performed in triplicate. (A) Rate of 3D7 and Dd2 invasion into 1∶10, 1∶1, and 10∶1 combinations of RBCViolet and RBCDDAO. Bars represent the mean invasion rate and error bars represent the SD. Elongated triangles below the X-axis represent the replacement of RBCViolet (white triangle) with RBCDDAO (gray triangle) in the total RBC population. (B and C) Data shows the number of 3D7 (B) and Dd2 (C) invasion events into RBCViolet (triangles) and RBCDDAO (circles) as the frequency of each RBC type increases from 10% to 90% of the total RBC population. Linear regression was used to determine the best fit line for P. falciparum invasion of RBCViolet and RBCDDAO. ANCOVA was performed to compare the slopes of the lines fit to P. falciparum invasion of RBCViolet and RBCDDAO. The null hypothesis was no difference between RBCØ and either RBCN, RBCT, or RBCC (H0: βØ = βenzyme, α = 0.05). ANCOVA performed with GraphPad, Prism, v. 5.04, La Jolla, CA calculated a p<0.0001. and p<0.06 for 3D7 and Dd2 respectively.(TIF)Click here for additional data file.
Authors: Sean C Gifford; Jure Derganc; Sergey S Shevkoplyas; Tatsuro Yoshida; Mark W Bitensky Journal: Br J Haematol Date: 2006-09-21 Impact factor: 6.998
Authors: P M Lantos; A D Ahouidi; A K Bei; C V Jennings; O Sarr; O Ndir; D F Wirth; S Mboup; M T Duraisingh Journal: Parasitology Date: 2009-01 Impact factor: 3.234
Authors: Andrea Cossarizza; Hyun-Dong Chang; Andreas Radbruch; Mübeccel Akdis; Immanuel Andrä; Francesco Annunziato; Petra Bacher; Vincenzo Barnaba; Luca Battistini; Wolfgang M Bauer; Sabine Baumgart; Burkhard Becher; Wolfgang Beisker; Claudia Berek; Alfonso Blanco; Giovanna Borsellino; Philip E Boulais; Ryan R Brinkman; Martin Büscher; Dirk H Busch; Timothy P Bushnell; Xuetao Cao; Andrea Cavani; Pratip K Chattopadhyay; Qingyu Cheng; Sue Chow; Mario Clerici; Anne Cooke; Antonio Cosma; Lorenzo Cosmi; Ana Cumano; Van Duc Dang; Derek Davies; Sara De Biasi; Genny Del Zotto; Silvia Della Bella; Paolo Dellabona; Günnur Deniz; Mark Dessing; Andreas Diefenbach; James Di Santo; Francesco Dieli; Andreas Dolf; Vera S Donnenberg; Thomas Dörner; Götz R A Ehrhardt; Elmar Endl; Pablo Engel; Britta Engelhardt; Charlotte Esser; Bart Everts; Anita Dreher; Christine S Falk; Todd A Fehniger; Andrew Filby; Simon Fillatreau; Marie Follo; Irmgard Förster; John Foster; Gemma A Foulds; Paul S Frenette; David Galbraith; Natalio Garbi; Maria Dolores García-Godoy; Jens Geginat; Kamran Ghoreschi; Lara Gibellini; Christoph Goettlinger; Carl S Goodyear; Andrea Gori; Jane Grogan; Mor Gross; Andreas Grützkau; Daryl Grummitt; Jonas Hahn; Quirin Hammer; Anja E Hauser; David L Haviland; David Hedley; Guadalupe Herrera; Martin Herrmann; Falk Hiepe; Tristan Holland; Pleun Hombrink; Jessica P Houston; Bimba F Hoyer; Bo Huang; Christopher A Hunter; Anna Iannone; Hans-Martin Jäck; Beatriz Jávega; Stipan Jonjic; Kerstin Juelke; Steffen Jung; Toralf Kaiser; Tomas Kalina; Baerbel Keller; Srijit Khan; Deborah Kienhöfer; Thomas Kroneis; Désirée Kunkel; Christian Kurts; Pia Kvistborg; Joanne Lannigan; Olivier Lantz; Anis Larbi; Salome LeibundGut-Landmann; Michael D Leipold; Megan K Levings; Virginia Litwin; Yanling Liu; Michael Lohoff; Giovanna Lombardi; Lilly Lopez; Amy Lovett-Racke; Erik Lubberts; Burkhard Ludewig; Enrico Lugli; Holden T Maecker; Glòria Martrus; Giuseppe Matarese; Christian Maueröder; Mairi McGrath; Iain McInnes; Henrik E Mei; Fritz Melchers; Susanne Melzer; Dirk Mielenz; Kingston Mills; David Mirrer; Jenny Mjösberg; Jonni Moore; Barry Moran; Alessandro Moretta; Lorenzo Moretta; Tim R Mosmann; Susann Müller; Werner Müller; Christian Münz; Gabriele Multhoff; Luis Enrique Munoz; Kenneth M Murphy; Toshinori Nakayama; Milena Nasi; Christine Neudörfl; John Nolan; Sussan Nourshargh; José-Enrique O'Connor; Wenjun Ouyang; Annette Oxenius; Raghav Palankar; Isabel Panse; Pärt Peterson; Christian Peth; Jordi Petriz; Daisy Philips; Winfried Pickl; Silvia Piconese; Marcello Pinti; A Graham Pockley; Malgorzata Justyna Podolska; Carlo Pucillo; Sally A Quataert; Timothy R D J Radstake; Bartek Rajwa; Jonathan A Rebhahn; Diether Recktenwald; Ester B M Remmerswaal; Katy Rezvani; Laura G Rico; J Paul Robinson; Chiara Romagnani; Anna Rubartelli; Beate Ruckert; Jürgen Ruland; Shimon Sakaguchi; Francisco Sala-de-Oyanguren; Yvonne Samstag; Sharon Sanderson; Birgit Sawitzki; Alexander Scheffold; Matthias Schiemann; Frank Schildberg; Esther Schimisky; Stephan A Schmid; Steffen Schmitt; Kilian Schober; Thomas Schüler; Axel Ronald Schulz; Ton Schumacher; Cristiano Scotta; T Vincent Shankey; Anat Shemer; Anna-Katharina Simon; Josef Spidlen; Alan M Stall; Regina Stark; Christina Stehle; Merle Stein; Tobit Steinmetz; Hannes Stockinger; Yousuke Takahama; Attila Tarnok; ZhiGang Tian; Gergely Toldi; Julia Tornack; Elisabetta Traggiai; Joe Trotter; Henning Ulrich; Marlous van der Braber; René A W van Lier; Marc Veldhoen; Salvador Vento-Asturias; Paulo Vieira; David Voehringer; Hans-Dieter Volk; Konrad von Volkmann; Ari Waisman; Rachael Walker; Michael D Ward; Klaus Warnatz; Sarah Warth; James V Watson; Carsten Watzl; Leonie Wegener; Annika Wiedemann; Jürgen Wienands; Gerald Willimsky; James Wing; Peter Wurst; Liping Yu; Alice Yue; Qianjun Zhang; Yi Zhao; Susanne Ziegler; Jakob Zimmermann Journal: Eur J Immunol Date: 2017-10 Impact factor: 6.688
Authors: M M Goheen; R Wegmüller; A Bah; B Darboe; E Danso; M Affara; D Gardner; J C Patel; A M Prentice; C Cerami Journal: EBioMedicine Date: 2016-11-09 Impact factor: 8.143
Authors: Martha A Clark; Morgan M Goheen; Anthony Fulford; Andrew M Prentice; Marwa A Elnagheeb; Jaymin Patel; Nancy Fisher; Steve M Taylor; Raj S Kasthuri; Carla Cerami Journal: Nat Commun Date: 2014-07-25 Impact factor: 14.919
Authors: Andrea Cossarizza; Hyun-Dong Chang; Andreas Radbruch; Andreas Acs; Dieter Adam; Sabine Adam-Klages; William W Agace; Nima Aghaeepour; Mübeccel Akdis; Matthieu Allez; Larissa Nogueira Almeida; Giorgia Alvisi; Graham Anderson; Immanuel Andrä; Francesco Annunziato; Achille Anselmo; Petra Bacher; Cosima T Baldari; Sudipto Bari; Vincenzo Barnaba; Joana Barros-Martins; Luca Battistini; Wolfgang Bauer; Sabine Baumgart; Nicole Baumgarth; Dirk Baumjohann; Bianka Baying; Mary Bebawy; Burkhard Becher; Wolfgang Beisker; Vladimir Benes; Rudi Beyaert; Alfonso Blanco; Dominic A Boardman; Christian Bogdan; Jessica G Borger; Giovanna Borsellino; Philip E Boulais; Jolene A Bradford; Dirk Brenner; Ryan R Brinkman; Anna E S Brooks; Dirk H Busch; Martin Büscher; Timothy P Bushnell; Federica Calzetti; Garth Cameron; Ilenia Cammarata; Xuetao Cao; Susanna L Cardell; Stefano Casola; Marco A Cassatella; Andrea Cavani; Antonio Celada; Lucienne Chatenoud; Pratip K Chattopadhyay; Sue Chow; Eleni Christakou; Luka Čičin-Šain; Mario Clerici; Federico S Colombo; Laura Cook; Anne Cooke; Andrea M Cooper; Alexandra J Corbett; Antonio Cosma; Lorenzo Cosmi; Pierre G Coulie; Ana Cumano; Ljiljana Cvetkovic; Van Duc Dang; Chantip Dang-Heine; Martin S Davey; Derek Davies; Sara De Biasi; Genny Del Zotto; Gelo Victoriano Dela Cruz; Michael Delacher; Silvia Della Bella; Paolo Dellabona; Günnur Deniz; Mark Dessing; James P Di Santo; Andreas Diefenbach; Francesco Dieli; Andreas Dolf; Thomas Dörner; Regine J Dress; Diana Dudziak; Michael Dustin; Charles-Antoine Dutertre; Friederike Ebner; Sidonia B G Eckle; Matthias Edinger; Pascale Eede; Götz R A Ehrhardt; Marcus Eich; Pablo Engel; Britta Engelhardt; Anna Erdei; Charlotte Esser; Bart Everts; Maximilien Evrard; Christine S Falk; Todd A Fehniger; Mar Felipo-Benavent; Helen Ferry; Markus Feuerer; Andrew Filby; Kata Filkor; Simon Fillatreau; Marie Follo; Irmgard Förster; John Foster; Gemma A Foulds; Britta Frehse; Paul S Frenette; Stefan Frischbutter; Wolfgang Fritzsche; David W Galbraith; Anastasia Gangaev; Natalio Garbi; Brice Gaudilliere; Ricardo T Gazzinelli; Jens Geginat; Wilhelm Gerner; Nicholas A Gherardin; Kamran Ghoreschi; Lara Gibellini; Florent Ginhoux; Keisuke Goda; Dale I Godfrey; Christoph Goettlinger; Jose M González-Navajas; Carl S Goodyear; Andrea Gori; Jane L Grogan; Daryl Grummitt; Andreas Grützkau; Claudia Haftmann; Jonas Hahn; Hamida Hammad; Günter Hämmerling; Leo Hansmann; Goran Hansson; Christopher M Harpur; Susanne Hartmann; Andrea Hauser; Anja E Hauser; David L Haviland; David Hedley; Daniela C Hernández; Guadalupe Herrera; Martin Herrmann; Christoph Hess; Thomas Höfer; Petra Hoffmann; Kristin Hogquist; Tristan Holland; Thomas Höllt; Rikard Holmdahl; Pleun Hombrink; Jessica P Houston; Bimba F Hoyer; Bo Huang; Fang-Ping Huang; Johanna E Huber; Jochen Huehn; Michael Hundemer; Christopher A Hunter; William Y K Hwang; Anna Iannone; Florian Ingelfinger; Sabine M Ivison; Hans-Martin Jäck; Peter K Jani; Beatriz Jávega; Stipan Jonjic; Toralf Kaiser; Tomas Kalina; Thomas Kamradt; Stefan H E Kaufmann; Baerbel Keller; Steven L C Ketelaars; Ahad Khalilnezhad; Srijit Khan; Jan Kisielow; Paul Klenerman; Jasmin Knopf; Hui-Fern Koay; Katja Kobow; Jay K Kolls; Wan Ting Kong; Manfred Kopf; Thomas Korn; Katharina Kriegsmann; Hendy Kristyanto; Thomas Kroneis; Andreas Krueger; Jenny Kühne; Christian Kukat; Désirée Kunkel; Heike Kunze-Schumacher; Tomohiro Kurosaki; Christian Kurts; Pia Kvistborg; Immanuel Kwok; Jonathan Landry; Olivier Lantz; Paola Lanuti; Francesca LaRosa; Agnès Lehuen; Salomé LeibundGut-Landmann; Michael D Leipold; Leslie Y T Leung; Megan K Levings; Andreia C Lino; Francesco Liotta; Virginia Litwin; Yanling Liu; Hans-Gustaf Ljunggren; Michael Lohoff; Giovanna Lombardi; Lilly Lopez; Miguel López-Botet; Amy E Lovett-Racke; Erik Lubberts; Herve Luche; Burkhard Ludewig; Enrico Lugli; Sebastian Lunemann; Holden T Maecker; Laura Maggi; Orla Maguire; Florian Mair; Kerstin H Mair; Alberto Mantovani; Rudolf A Manz; Aaron J Marshall; Alicia Martínez-Romero; Glòria Martrus; Ivana Marventano; Wlodzimierz Maslinski; Giuseppe Matarese; Anna Vittoria Mattioli; Christian Maueröder; Alessio Mazzoni; James McCluskey; Mairi McGrath; Helen M McGuire; Iain B McInnes; Henrik E Mei; Fritz Melchers; Susanne Melzer; Dirk Mielenz; Stephen D Miller; Kingston H G Mills; Hans Minderman; Jenny Mjösberg; Jonni Moore; Barry Moran; Lorenzo Moretta; Tim R Mosmann; Susann Müller; Gabriele Multhoff; Luis Enrique Muñoz; Christian Münz; Toshinori Nakayama; Milena Nasi; Katrin Neumann; Lai Guan Ng; Antonia Niedobitek; Sussan Nourshargh; Gabriel Núñez; José-Enrique O'Connor; Aaron Ochel; Anna Oja; Diana Ordonez; Alberto Orfao; Eva Orlowski-Oliver; Wenjun Ouyang; Annette Oxenius; Raghavendra Palankar; Isabel Panse; Kovit Pattanapanyasat; Malte Paulsen; Dinko Pavlinic; Livius Penter; Pärt Peterson; Christian Peth; Jordi Petriz; Federica Piancone; Winfried F Pickl; Silvia Piconese; Marcello Pinti; A Graham Pockley; Malgorzata Justyna Podolska; Zhiyong Poon; Katharina Pracht; Immo Prinz; Carlo E M Pucillo; Sally A Quataert; Linda Quatrini; Kylie M Quinn; Helena Radbruch; Tim R D J Radstake; Susann Rahmig; Hans-Peter Rahn; Bartek Rajwa; Gevitha Ravichandran; Yotam Raz; Jonathan A Rebhahn; Diether Recktenwald; Dorothea Reimer; Caetano Reis e Sousa; Ester B M Remmerswaal; Lisa Richter; Laura G Rico; Andy Riddell; Aja M Rieger; J Paul Robinson; Chiara Romagnani; Anna Rubartelli; Jürgen Ruland; Armin Saalmüller; Yvan Saeys; Takashi Saito; Shimon Sakaguchi; Francisco Sala-de-Oyanguren; Yvonne Samstag; Sharon Sanderson; Inga Sandrock; Angela Santoni; Ramon Bellmàs Sanz; Marina Saresella; Catherine Sautes-Fridman; Birgit Sawitzki; Linda Schadt; Alexander Scheffold; Hans U Scherer; Matthias Schiemann; Frank A Schildberg; Esther Schimisky; Andreas Schlitzer; Josephine Schlosser; Stephan Schmid; Steffen Schmitt; Kilian Schober; Daniel Schraivogel; Wolfgang Schuh; Thomas Schüler; Reiner Schulte; Axel Ronald Schulz; Sebastian R Schulz; Cristiano Scottá; Daniel Scott-Algara; David P Sester; T Vincent Shankey; Bruno Silva-Santos; Anna Katharina Simon; Katarzyna M Sitnik; Silvano Sozzani; Daniel E Speiser; Josef Spidlen; Anders Stahlberg; Alan M Stall; Natalie Stanley; Regina Stark; Christina Stehle; Tobit Steinmetz; Hannes Stockinger; Yousuke Takahama; Kiyoshi Takeda; Leonard Tan; Attila Tárnok; Gisa Tiegs; Gergely Toldi; Julia Tornack; Elisabetta Traggiai; Mohamed Trebak; Timothy I M Tree; Joe Trotter; John Trowsdale; Maria Tsoumakidou; Henning Ulrich; Sophia Urbanczyk; Willem van de Veen; Maries van den Broek; Edwin van der Pol; Sofie Van Gassen; Gert Van Isterdael; René A W van Lier; Marc Veldhoen; Salvador Vento-Asturias; Paulo Vieira; David Voehringer; Hans-Dieter Volk; Anouk von Borstel; Konrad von Volkmann; Ari Waisman; Rachael V Walker; Paul K Wallace; Sa A Wang; Xin M Wang; Michael D Ward; Kirsten A Ward-Hartstonge; Klaus Warnatz; Gary Warnes; Sarah Warth; Claudia Waskow; James V Watson; Carsten Watzl; Leonie Wegener; Thomas Weisenburger; Annika Wiedemann; Jürgen Wienands; Anneke Wilharm; Robert John Wilkinson; Gerald Willimsky; James B Wing; Rieke Winkelmann; Thomas H Winkler; Oliver F Wirz; Alicia Wong; Peter Wurst; Jennie H M Yang; Juhao Yang; Maria Yazdanbakhsh; Liping Yu; Alice Yue; Hanlin Zhang; Yi Zhao; Susanne Maria Ziegler; Christina Zielinski; Jakob Zimmermann; Arturo Zychlinsky Journal: Eur J Immunol Date: 2019-10 Impact factor: 6.688
Authors: Morgan M Goheen; Amat Bah; Rita Wegmüller; Hans Verhoef; Bakary Darboe; Ebrima Danso; Andrew M Prentice; Carla Cerami Journal: Sci Rep Date: 2017-12-15 Impact factor: 4.379
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5