Human co-infection is difficult to replicate in vitro. However, human malaria parasites can readily be cultured in vitro, as can freshly isolated human peripheral blood mononuclear cells naturally infected with HIV. This provides an excellent model for studying early immune responses to malaria parasites in the context of HIV co-infection.
Malaria and HIV co-infection is a growing health priority. However, most research on malaria or HIV currently focuses on each infection individually. Although understanding the disease dynamics for each of these pathogens independently is vital, it is also important that the interactions between these pathogens are investigated and understood.
We have developed a versatile in vitro model of HIV-malaria co-infection to study host immune responses to malaria in the context of HIV infection. Our model allows the study of secreted factors in cellular supernatants, cell surface and intracellular protein markers, as well as RNA expression levels. The experimental design and methods used limit variability and promote data reliability and reproducibility.
All pathogens used in this model are natural human pathogens (Plasmodium falciparum and HIV-1), and all infected cells are naturally infected and used fresh. We use human erythrocytes parasitized with P. falciparum and maintained in continuous in vitro culture. We obtain freshly isolated peripheral blood mononuclear cells from chronically HIV-infected volunteers. Every condition used has an appropriate control (P. falciparum parasitized vs. normal erythrocytes), and every HIV-infected donor has an HIV uninfected control, from which cells are harvested on the same day. This model provides a realistic environment to study the interactions between malaria parasites and human immune cells in the context of HIV infection.
Co-infection, infection with multiple concurrent infections, is the norm in natural environments. Co-infection can have a major impact on disease pathology and on the clinical management of each infection. In the context of co-infection, vaccine and drug efficacy, as well as diagnostic testing, can be negatively impacted (reviewed in 1). However, despite its importance, the majority of pathogen research considers only single infections.
Malaria and HIV-1 (HIV) are leading causes of morbidity and mortality globally. Areas of malaria and HIV endemicity share a wide geographic overlap, putting millions of people at risk of co-infection and consequently at risk for more severe clinical disease 2–10. The two diseases negatively interact. In HIV-infected individuals, higher HIV viral loads and temporary decreases in CD4+ T-cell counts can be seen during a malaria infection, while malaria parasite burdens and risk of clinical and severe malaria are higher in co-infected individuals 2,3,5,7,8,10. The mechanisms by which HIV increases malaria severity are not fully understood and warrant further investigation.
Here we describe a method by which malaria and HIV co-infection can be studied in vitro. Specifically, this method allows for the examination of malaria-specific immune responses in the context of HIV infection. Our protocol describes a versatile co-culture system using freshly isolated peripheral blood mononuclear cells (PBMCs) isolated from chronically HIV infected donors and in vitro cultured P. falciparum parasitized erythrocytes (PfRBC). The impact of HIV antiretroviral therapy on these responses can also be examined using prospectively collected PBMCs from HIV(+) donors pre- and post-therapy.
We have used this system to investigate the impact of HIV infection on malaria-specific innate immune responses 11,12, and were able to determine that malaria-specific IFNγ and TNF responses are impaired in NK cells, NKT cells, γδ T cells from HIV(+) donors pre- and post-HIV antiretroviral therapy. Additionally, we were able to use this system to determine that monocytic functions are also impaired in HIV(+) donors, but recover post-HIV antiretroviral therapy.
This protocol requires the recruitment of donors for serum and RBC to be used for parasite culture, and HIV(+) and uninfected donors for PBMC isolation. Institutional Review Boards must approve all studies and all donors must provide informed consent prior to blood draw.
CAUTION: Working with human blood samples and human malaria parasites requires precautionary measures. Always wear a lab coat, gloves, and work in a level 2 Biosafety cabinet. In the event of accidental percutaneous exposure to human malaria, report to Health and Safety for prophylactic treatment. Additional safety consideration should also be put in place for working with HIV(+) blood. Wear a back closing lab coat. Double glove (top glove should be latex). Perform all handling in a level 2 Biosafety cabinet. Do not use any glass or sharp objects. Do not use aspirator. Place all contaminated items in virox solution or bleach for a minimum of 1 hr prior to discarding. Wash all surfaces with virox and UV for 1 hr following use. Note that each institution will have their own specific biosafety regulations that need to be followed. Report all accidental exposures to HIV-infected blood to Health and Safety for evaluation and consideration of possible post-exposure prophylaxis.
Note: Different strains of P. falciparum malaria parasites exist. ITG was used for these experiments, but different strains can be used. Excellent instructions on freezing and thawing Plasmodium falciparum parasites are available at the MR4 website 13.
1. Making RPMI-A for Malaria Parasite Culture
2. Preparing Human Red Blood Cells for Parasite Culture
Note: Blood donors should be type O.
3. Maintaining Parasite Cultures
4. Parasite Synchronization
Note: The day before the experiment, synchronize the parasite culture by treating with alanine. Only ring stage parasites and uninfected RBC will survive this treatment. Alanine synchronization will give you a pure trophozoite culture the next day that can be used in the co-culture experiments. Make sure to start with a parasite culture that contains a majority of ring stage parasites.
5. Parasite and RBC Preparation for Co-culture Experiments
6. Isolation of Human Peripheral Blood Mononuclear Cells (PBMC)
7. Malaria/HIV Co-infection Culture
8. Detection of Malaria Immune Responses
Note: Perform co-culture experiments for as long as 4 days. No medium change is required during this time. The optimal time point will depend on the cell type of interest and the question asked. An incubation of 12-48 hr is optimal for monocytic responses to PfRBCs, while lymphocyte responses were best observed at 72-96 hr. The time points will need to be optimized based on the experimental question. Shorter periods (2-4 hr) may be used if interaction between intact PfRBC and PBMCs is of interest.
9. Intracellular Flow Cytometry for Cell-specific Ccytokine Responses Using PBMC Co-cultured with P. falciparum Infected RBC
Note: As mentioned above the length of co-culture will depend on the cell type of interest. If interested in monocytic responses to PfRBC a shorter incubation period is needed. Times will be longer for innate lymphocyte responses γδ T cells, NK cells, NKT cells), and longer still for CD4 and CD8 T cells. Optimization will be required.
The graphs depict the levels of IFNγ production from NKT cells (Figure 2), using CD56+CD3+γδ- gates to obtain the NKT cells population (data not shown). The cells were cultured for 72 hr prior to staining. Once stained, 100,000 CD3+ cells were acquired on the flow cytometer to obtain large enough populations of NK, NKT and γδ cells (cells of interest). A minimum of 5,600 NKT cells are displayed on each graph. TNF production is obtained in the same manner (data not shown). The graphs clearly demonstrate that IFNγ production is lower in cells from HIV(+) individuals compared to HIV(-) individuals exposed to PfRBC.
Flow cytometry analysis is very subjective. It is therefore essential to have all the appropriate controls for each experiment (see Figure 2). Background levels are calculated using the FMO samples, which allows for a true representation of the cytokine staining. Cells stimulated with PMA/Ionomycin were used as a positive control (data not shown). PMA and Ionomycin are strong stimulators of T cell cytokine production. A lack of IFNγ production in these samples would most likely indicate a problem with the staining protocol. However, other variables such as cell viability or inactive reagents may also be at fault.
Figure 1. Depiction of Ficoll gradient post spin demonstrating the position of the PBMCs.
Figure 2. IFNγ production by Natural Killer T cells. The flow diagrams were obtained by gating on CD56+CD3+γδ- cells (minimum of 5,600 events). IFNγ production is detectable in the HIV(-) sample stimulated with P. falciparum infected red blood cells. This cytokine response is no longer apparent in the context of a chronic HIV infection. Please click here to view a larger version of this figure.
Our protocol has been optimized in order to most realistically study HIV-malaria co-infection in vitro. First, fresh human RBCs and serum are required for malaria parasite culture. This is vital to obtain a healthy population of malaria parasites. Parasite lysates cannot be substituted for live parasites as cytokine production is much more rapid and intense when using live P. falciparum infected RBC (PfRBC) 17,18. In addition, activation of cell types like NK cells, requires whole PfRBCs, and does not work efficiently with parasite lysates. This may be due to the need for direct contact between the PfRBC and the leukocyte or may be due to the unstable nature of the parasite-derived ligands that interact with cell surface receptors 18. Malaria PfRBC cultures must also be well synchronized (Protocol 4) prior to the experiment. Synchronized PfRBCs improve the reproducibility of the experimental data. Although this protocol describes the use of trophozoite stage PfRBC for co-culture, it can easily be modified for the study of ring stage PfRBC.
Human leukocytes can be artificially infected with HIV, and this method has been used in studies of malaria-HIV co-infection research 20. However, this does not model the immune dysregulation that results from chronic HIV infection. In this co-culture system we use PBMCs isolated from human participants that are chronically infected with HIV-1. When participants are required for a study, it is important to carefully select this population. Inclusion and exclusion criteria are vital to ensure minimum variability and maximum reproducibility. Our criteria included a clear definition of chronic HIV infection (HIV-infected for >1 year, with CD4+ T-cell count decline of >50 cells/mm3/year) and the exclusion of anyone with a concurrent infection. This ensured that the results obtained could be attributed solely to the effect of chronic HIV infection, and not another infection. Additionally, since we were interested in innate immune responses to malaria we excluded donors that had previous malaria infection.
HIV-uninfected controls are also used for each HIV(+) donor, at every sampling time point, to allow for data normalization. An attempt should be made to match controls to their respective HIV(+) donor for at least age but preferably also sex (although in our previous study 12 we did not observe a significant difference in cell subsets or cytokine responses between female and male HIV-uninfected participants). If prospective sampling is planned maintaining the same HIV-uninfected control for each sampling time point is beneficial. Responses can vary from experiment to experiment due to multiple factors including the health of the PfRBCs, their degree of synchronization, parasitemia level and stage of maturity of PfRBCs, and hematocrit level. Care must be taken to keep as many of these consistent between experiments. Normalizing to an HIV-uninfected control allows for some accounting for these variables.
Using fresh cells is paramount to avoid artificial results. Freezing and thawing samples can have a significant impact on cell viability 21,22, cytokine production 23-26 and cell surface phenotypic markers 27.
If monocytes are of particular interest, it is important that glass is not used during the protocol. Monocytes adhere to glass and many other plastics 28. We use polypropylene plastic pipettes, transfer pipettes, and tubes throughout to minimize monocyte adherence and ultimate removal from our study cell populations.
When setting up the flow cytometry assay, any combination of antibodies can be used depending on the cells of interest. We were particularly interested in IFNγ and TNF production from NK cells, NKT cells and γδ T cells. This defined our antibody panel. However, if using a different combination, it is important to optimize the amounts of antibody for each panel. This is vital to avoid erroneous data. Also, to avoid variability and ensure validity, FMOs are required to assess background cytokine levels for each condition (RBC, PfRBC, medium, and PMA/Ionomycin) and participant population (HIV(-) and HIV(+), Figure 2). Measuring in vitro intracellular cytokine production usually results in high background levels, especially when cells have been cultured prior to staining. Since the culture conditions affect background levels, appropriate FMOs ensure knowledge of the background level for each sample. Our supply of cells was limited, therefore we designed our FMOs to contain all surface stains but no intracellular stains. This allows for the specific comparison of cytokine background levels on all the different cell types studied. We also ran one single stain per experiment for each of the surface markers to confirm their background levels.
The protocol described is a versatile one, which can be used to examine responses within hours or days depending on the cells of interest. For optimal PfRBC-induced cytokine production from innate lymphocytes, we measured flow cytometry output after 2 or 3 days. When looking at monocytes, earlier time points (1 or 2 days) are recommended. This system allows for multiple cell types and cell responses to be distinguished by flow cytometry, secretory responses to be measured in cell supernatants, and expression profiles to be assessed in extracted RNA. Further, use of antibodies to block specific receptors or neutralize cytokines can be utilized in this system to further dissect mechanisms. We have successfully used IL-18 receptor blockade to implicate IL-18 receptor in PfRBC-induced IFNγ responses 12. This system represents a realistic method by which to assess a number of innate immune responses to malaria in the context of HIV infection.
The authors have nothing to disclose.
C.A.M.F. and L.S. participated in protocol design, acquisition and analysis of data, and drafting of the article.
The authors wish to thank Dr. Kain, Dr. Loutfy, Dr. Wasmuth, and Dr. Ayi for their contributions.
C.F. was supported by a CTN/Ontario HIV Treatment Network (OHTN) postdoctoral fellowship. L.S. is supported by an OHTN Junior Investigator Development Award. The present work was supported by a Canadian Institutes of Health Research (CIHR) operating grant (MOP-13721 and 115160), and a CIHR New Investigator Catalyst grant.
alanine | Sigma | A7377 | |
antibodies (see other table) | |||
BD Cytofix/Cytoperm with BD GolgiPlug | BD | 555028 | includes brefeldin A, cytofix/cytoperm buffer and perm/wash buffer |
BD Vacutainer ACD Solution A | BD | 364506 | |
BD Vacutainer Sodium Heparin | BD | 17-1440-02 | |
DPBS (no calcium, no magnesium) | Corning | 21-031-CV | |
Fetal Bovine Serum | Sigma | F1051 | heat inactivate before use |
Ficoll-Paque PLUS | GE Healthcare | 17-1440-02 | |
gentamycin (10mg/ml) | Gibco | 15710-064 | |
Hema3 Staining Set | Fisher | 122-911 | |
HEPES | Fisher | BP310-500 | |
hypoxanthine | Sigma | H9636 | |
Ionomycin | Sigma | I3909 | |
MEM non-essential amino acids (10mM) | Gibco | 11140 | |
PMA | Sigma | P8139 | |
RPMI-1640 powder | Life-Technologies | 31800-022 | |
RPMI-1640 with L-glutamine and HEPES | Thermo Scientific | SH30255.01 | |
sodium bicarbonate (powder, cell culture) | Sigma | S5761 | |
Sodium Pyruvate (100mM) | Gibco | 11360 | |
Tris (Trizma base) | Sigma | T6066 | |
Trizol | Ambion | 15596018 | |
Trypan Blue (0.4%) | Gibco | 15250-061 | |
BD CompBead | BD | 552843, 552845 | depends on antibodies used |
Parasite Gas Mixture | By special order | 3% CO2, 1% O2, balance N2 |