Journal of Infectious Diseases Advance Access published May 9, 2015
SUPPLEMENT ARTICLE
Immune Response to Marburg Virus Angola Infection in Nonhuman Primates Lisa Fernando,1 Xiangguo Qiu,1,2 P. Leno Melito,1 Kinola J. N. Williams,1,3 Friederike Feldmann,1,a Heinz Feldmann,1,2,a Steven M. Jones,1,a and Judie B. Alimonti1 1
Special Pathogens Program, National Microbiology Laboratory, Public Health Agency of Canada, 2Department of Medical Microbiology, and 3Department of Immunology, University of Manitoba, Winnipeg, Canada
Background. The 2005 outbreak of Marburg virus (MARV) infection in Angola was the most lethal MARV infection outbreak in history, with a case-fatality rate (90%) similar to that for Zaire ebolavirus (EBOV) infection. However, very little is known about the pathogenicity of MARV Angola, as few studies have been conducted to date. Therefore, the immune response was examined in MARV Angola–infected nonhuman primates. Methods. Cynomolgus macaques were infected with MARV Angola and monitored for survival. The effect of MARV Angola on the immune system was examined by immunophenotyping whole-blood and by analyzing cytokine and chemokine levels in plasma and spleen specimens, using flow cytometry. Results. The prominent clinical findings were rapid onset of disease and death (mean time after infection, 6.7 days), fever, depression, anorexia, petechial rash, and lymphopenia. Specifically, T, B, and natural killer cells were severely depleted in the blood by day 6. The typical cytokine storm was present, with levels of interferon γ, tumor necrosis factor, interleukin 6, and CCL2 rising in the blood early during infection. Conclusions. MARV Angola displayed the same virulence and disease pathology as EBOV. MARV Angola appears to cause a more rapid onset and severe outcome of infection than other MARV strains. Keywords. Filovirus; Marburgvirus; lymphocytes; natural killer cells; nonhuman primates; Marburg Angola; cytokines; Ebolavirus; cynomolgus macaques. The Marburgvirus genus belongs to the Filoviridae family. Unlike the other Filoviridae genus, Ebolavirus, the Marburgvirus genus consists of only 1 species, Marburg marburgvirus, which contains 2 viruses, Marburg virus (MARV) and Ravn virus [1]. MARV was first identified in 1967, when laboratory workers became ill after processing tissues from African green monkeys [2]. Since then, there have been isolated episodes of Marburg hemorrhagic fever involving a small number of individuals in Africa [3–5]. In 1998, the first large-scale community outbreak occurred in Durba, Democratic Republic of the Congo, where the case-fatality rate was 83%; this was the first outbreak to resemble outbreaks involving
a Present affiliations: Virology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rocky Mountain Laboratories, Hamilton, Montana (F. F. and H. F.); Cognoveritas Consulting, Winnipeg, Canada (S. M. J.). Correspondence: Lisa Fernando, MSc, Special Pathogens Program, National Microbiology Laboratory, Public Health Agency of Canada, 1015 Arlington St, Winnipeg, Manitoba, R3E 3R2 Canada ([email protected]).
The Journal of Infectious Diseases® © Crown copyright 2015. DOI: 10.1093/infdis/jiv095
the most pathogenic Ebolavirus, Zaire ebolavirus (EBOV), in terms of case-fatality rate [6, 7]. The largest outbreak of Marburg hemorrhagic fever, with the same 90% casefatality rate as EBOV hemorrhagic fever, occurred in 2005, in Uige, Angola [1, 8–11], where the original case started in the towns’ hospital before dispersing among several villages [8, 11]. Transmission was primarily through contact with an infected body and bodily fluids, usually in family members involved in the burying rituals of infected relatives [10]. The initial disease symptoms began on about day 5, with the majority of deaths occurring during the second week [12]. Several characteristics make the 2005 Angola outbreak very unique. First, there were numerous cases among young children, something that had never been seen before [10]. Additionally, there was a shorter incubation period and more-virulent disease associated with this new strain of MARV. The rapid onset and high death rate suggested that MARV Angola was a severely virulent pathogen on par with EBOV. While EBOV has been well studied in nonhuman primates (NHPs), including extensive immunological analysis of EBOV pathogenesis, very Marburg Virus Angola Immune Response
•
JID
•
S1
little is known about the immune response to MARV Angola infection. NHPs are the preferred filovirus disease model, and in cynomolgus and rhesus macaques MARV infections are 100% lethal when using the most common Ci67 and Musoke strains. So far, only 5 NHP studies have been published on MARVAngola infection [13–17]. Typically, infection with any of the MARV strains results in a syndrome closely resembling Marburg hemorrhagic fever in humans, including fever, cutaneous rashes, and elevated enzyme levels associated with organ dysfunction [11, 15]. The immune response is not as well characterized in NHPs for MARV as it is for EBOV. Lymphopenia is a hallmark of Ebola hemorrhagic fever. When cynomolgus macaques were infected with EBOV by any inoculation route, there was a decline in T-lymphocyte counts, specifically in CD4+ and CD8+ T-cell subsets, and, in some studies, also the natural killer (NK) population [18–20]. Similarly, in tissue culture, when human peripheral blood mononuclear cells were infected with EBOV, both subsets of T lymphocytes declined by 60% on day 8 after infection [21]. In contrast, lymphopenia is not a consistent feature in NHPs infected with MARV Ci67, Angola, and Musoke [13, 18, 22, 23]. However, depletion of lymphocytes with both EBOV and MARV has been attributed to apoptosis and was most prominent in EBOV infections [24]. Because macrophages and dendritic cells (DCs) are the primary targets of filovirus infections, the mechanism of apoptosis is hypothesized to be due to incorrect T-cell activation signals by DCs and macrophages [25]. Since the immune system is unable to control the virus, the infection worsens, leading to the hallmark cytokine storm. Although this is well documented for EBOV, there is limited evidence for this with MARV. MARV Ci67 is associated with a rapid increase in proinflammatory cytokines at the time of death [22, 26]. As for MARV Angola, one study in cynomolgus macaques demonstrated elevated levels of CCL2, CCL4, interleukin 6 (IL-6), and CXCL8 by days 7–9 after an aerosol challenge [13]. Additionally, in rhesus macaques, IL-6 and monocyte chemotactic protein 1 (MCP-1) levels increased on day 7 then decreased on day 8 when infected with MARV Angola [15]. Clearly, more information is needed to understand the immune response in MARV infections in general and in MARV Angola in particular. Since the MARV Angola strain displayed unique disease qualities as compared to other MARV isolates, it was of interest to investigate the immune pathogenesis in cynomolgus macaques. While others have described clinical, virological, and pathological findings for NHPs infected with MARV Angola [14–16, 23], we are the first to report a detailed immunological analysis in NHPs during MARV Angola infection. Our analysis allows us to determine whether lymphopenia and the cytokine response play an important role during MARV Angola infection. Furthermore, by studying circulatory immune cells in the blood, we are able to describe how this latest MARV strain affects the immune system. S2
•
JID
•
Fernando et al
MATERIALS AND METHODS Animals
Five healthy 16–22-year-old female cynomolgus macaques (Macaca fascicularis) from a NHP colony at Health Canada Animal Resources Division (Ottawa, Ontario) were housed individually in cages and acclimatized 14 days prior to infection. Each 4-cage bank was inside a negative pressure isolator with high-efficiency particulate arrestance filtration. Animals were monitored twice daily and fed commercial monkey chow, treats, and fruit; enrichment consisted of commercial toys and treats. All animal procedures were conducted in a containment level 4 laboratory according to the Canadian Council on Animal Care and with approval by the Animal Care Committee of the Canadian Science Centre for Human and Animal Health. Challenge and Examination
MARV Angola was obtained from an infected patient during the 2005 outbreak in Uige and passaged once in Vero E6 cells. Three NHPs were challenged with 1000 50% tissue culture infective doses (TCID50) of MARV Angola intramuscularly. Two control animals received Roswell Park Memorial Institute 1640 medium. Routine examination on days 0, 2, 4, and 6 after challenge included monitoring of temperature (rectal), respiration rate, weight, hydration status, and discharges. Animals were euthanized once they reached the end point score, in accordance with the approved Animal Care Committee animal protocol. Homogenization of Spleens
Upon euthanasia (days 11 and 12 for the 2 control NHPs and days 6 and 7 for the 3 NHPs challenged with MARV Angola), the spleen was harvested, and a portion was weighed and homogenized in 1.5 mL of Roswell Park Memorial Institute 1640 medium, using the BD Medimachine. The spleen homogenate was centrifuged at 1300g for 10 minutes, and the supernatant was used in the cytometric bead array (CBA) assay. The remaining pellet was washed twice in phosphate-buffered saline and then underwent immunophenotyping, using fluorescenceactivated cell-sorting (FACS) analysis. Immunophenotyping of Whole-Blood Specimens and Splenocytes
Whole-blood specimens (100 µL) or 1 × 106 splenocytes were blocked with 10 µg of human γ-globulin then stained with a 100 µL antibody master mix. Two antibody panels were used, one T-cell panel (CD45, CD3, CD4, CD8, CD25, and CD69) and one NK cell panel (CD45, CD3, CD8, CD16, CD20, and CD69). Splenocytes were also stained with a macrophage panel (CD45, CD11b, CD14, CD15, CD69, and HLA-DR). After incubation for 30 minutes at 4°C, red blood cells and splenocytes were lysed with 1.5 mL and 0.5 mL of 1 × FACS lysing solution (Becton Dickinson), respectively; vortexed; and incubated at room temperature for 15 minutes. Pellets were washed
with 1 mL of wash buffer (Dulbecco’s phosphate-buffered saline without calcium or magnesium [Invitrogen], with 1% heat inactivated fetal bovine serum). Cells were pelleted and resuspended in 900 µL of phosphate-buffered saline, with 4% paraformaldehyde. Samples were evaluated on a BD LSRII flow cytometer, and data were analyzed using FACS Diva software v5.0.2 (BD Biosciences). Prior to acquisition, 100 µL of Flow-Count Fluorospheres (Beckman Coulter) were added to determine absolute counts. Cytometric Bead Array (CBA) Analysis of Plasma Samples and Spleen Homogenate Supernatant
A flow cytometric BD Biosciences cytokine/chemokine CBA Flex Set system was used to assess the levels of cytokines and chemokines in plasma samples and spleen homogenate supernatant that were exposed to 5 MRads of γ-irradiation on dry ice upon removal from the containment level 4 laboratory. The samples were evaluated in duplicate, using the CBA Flex Set system according to the manufacturer’s instructions, and included beads for interleukin 2 (IL-2), interleukin 4 (IL-4), IL-6, CXCL8 (interleukin 8 [IL-8]), interferon γ (IFN-γ), CCL2 (MCP-1), CCL5 (regulated on activation, normal T-cell expressed and secreted [RANTES]), and tumor necrosis factor (TNF). Samples were run on the LSRII (BD Biosciences), with FACSDiva v5.0.2 software and analyzed with BD’s FCAP v1.0.1 software. Statistical Analyses
The t test was used to determine significant responses for the CBA and flow cytometry data. A significant result was defined as a P value of
SUPPLEMENT ARTICLE
Immune Response to Marburg Virus Angola Infection in Nonhuman Primates Lisa Fernando,1 Xiangguo Qiu,1,2 P. Leno Melito,1 Kinola J. N. Williams,1,3 Friederike Feldmann,1,a Heinz Feldmann,1,2,a Steven M. Jones,1,a and Judie B. Alimonti1 1
Special Pathogens Program, National Microbiology Laboratory, Public Health Agency of Canada, 2Department of Medical Microbiology, and 3Department of Immunology, University of Manitoba, Winnipeg, Canada
Background. The 2005 outbreak of Marburg virus (MARV) infection in Angola was the most lethal MARV infection outbreak in history, with a case-fatality rate (90%) similar to that for Zaire ebolavirus (EBOV) infection. However, very little is known about the pathogenicity of MARV Angola, as few studies have been conducted to date. Therefore, the immune response was examined in MARV Angola–infected nonhuman primates. Methods. Cynomolgus macaques were infected with MARV Angola and monitored for survival. The effect of MARV Angola on the immune system was examined by immunophenotyping whole-blood and by analyzing cytokine and chemokine levels in plasma and spleen specimens, using flow cytometry. Results. The prominent clinical findings were rapid onset of disease and death (mean time after infection, 6.7 days), fever, depression, anorexia, petechial rash, and lymphopenia. Specifically, T, B, and natural killer cells were severely depleted in the blood by day 6. The typical cytokine storm was present, with levels of interferon γ, tumor necrosis factor, interleukin 6, and CCL2 rising in the blood early during infection. Conclusions. MARV Angola displayed the same virulence and disease pathology as EBOV. MARV Angola appears to cause a more rapid onset and severe outcome of infection than other MARV strains. Keywords. Filovirus; Marburgvirus; lymphocytes; natural killer cells; nonhuman primates; Marburg Angola; cytokines; Ebolavirus; cynomolgus macaques. The Marburgvirus genus belongs to the Filoviridae family. Unlike the other Filoviridae genus, Ebolavirus, the Marburgvirus genus consists of only 1 species, Marburg marburgvirus, which contains 2 viruses, Marburg virus (MARV) and Ravn virus [1]. MARV was first identified in 1967, when laboratory workers became ill after processing tissues from African green monkeys [2]. Since then, there have been isolated episodes of Marburg hemorrhagic fever involving a small number of individuals in Africa [3–5]. In 1998, the first large-scale community outbreak occurred in Durba, Democratic Republic of the Congo, where the case-fatality rate was 83%; this was the first outbreak to resemble outbreaks involving
a Present affiliations: Virology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rocky Mountain Laboratories, Hamilton, Montana (F. F. and H. F.); Cognoveritas Consulting, Winnipeg, Canada (S. M. J.). Correspondence: Lisa Fernando, MSc, Special Pathogens Program, National Microbiology Laboratory, Public Health Agency of Canada, 1015 Arlington St, Winnipeg, Manitoba, R3E 3R2 Canada ([email protected]).
The Journal of Infectious Diseases® © Crown copyright 2015. DOI: 10.1093/infdis/jiv095
the most pathogenic Ebolavirus, Zaire ebolavirus (EBOV), in terms of case-fatality rate [6, 7]. The largest outbreak of Marburg hemorrhagic fever, with the same 90% casefatality rate as EBOV hemorrhagic fever, occurred in 2005, in Uige, Angola [1, 8–11], where the original case started in the towns’ hospital before dispersing among several villages [8, 11]. Transmission was primarily through contact with an infected body and bodily fluids, usually in family members involved in the burying rituals of infected relatives [10]. The initial disease symptoms began on about day 5, with the majority of deaths occurring during the second week [12]. Several characteristics make the 2005 Angola outbreak very unique. First, there were numerous cases among young children, something that had never been seen before [10]. Additionally, there was a shorter incubation period and more-virulent disease associated with this new strain of MARV. The rapid onset and high death rate suggested that MARV Angola was a severely virulent pathogen on par with EBOV. While EBOV has been well studied in nonhuman primates (NHPs), including extensive immunological analysis of EBOV pathogenesis, very Marburg Virus Angola Immune Response
•
JID
•
S1
little is known about the immune response to MARV Angola infection. NHPs are the preferred filovirus disease model, and in cynomolgus and rhesus macaques MARV infections are 100% lethal when using the most common Ci67 and Musoke strains. So far, only 5 NHP studies have been published on MARVAngola infection [13–17]. Typically, infection with any of the MARV strains results in a syndrome closely resembling Marburg hemorrhagic fever in humans, including fever, cutaneous rashes, and elevated enzyme levels associated with organ dysfunction [11, 15]. The immune response is not as well characterized in NHPs for MARV as it is for EBOV. Lymphopenia is a hallmark of Ebola hemorrhagic fever. When cynomolgus macaques were infected with EBOV by any inoculation route, there was a decline in T-lymphocyte counts, specifically in CD4+ and CD8+ T-cell subsets, and, in some studies, also the natural killer (NK) population [18–20]. Similarly, in tissue culture, when human peripheral blood mononuclear cells were infected with EBOV, both subsets of T lymphocytes declined by 60% on day 8 after infection [21]. In contrast, lymphopenia is not a consistent feature in NHPs infected with MARV Ci67, Angola, and Musoke [13, 18, 22, 23]. However, depletion of lymphocytes with both EBOV and MARV has been attributed to apoptosis and was most prominent in EBOV infections [24]. Because macrophages and dendritic cells (DCs) are the primary targets of filovirus infections, the mechanism of apoptosis is hypothesized to be due to incorrect T-cell activation signals by DCs and macrophages [25]. Since the immune system is unable to control the virus, the infection worsens, leading to the hallmark cytokine storm. Although this is well documented for EBOV, there is limited evidence for this with MARV. MARV Ci67 is associated with a rapid increase in proinflammatory cytokines at the time of death [22, 26]. As for MARV Angola, one study in cynomolgus macaques demonstrated elevated levels of CCL2, CCL4, interleukin 6 (IL-6), and CXCL8 by days 7–9 after an aerosol challenge [13]. Additionally, in rhesus macaques, IL-6 and monocyte chemotactic protein 1 (MCP-1) levels increased on day 7 then decreased on day 8 when infected with MARV Angola [15]. Clearly, more information is needed to understand the immune response in MARV infections in general and in MARV Angola in particular. Since the MARV Angola strain displayed unique disease qualities as compared to other MARV isolates, it was of interest to investigate the immune pathogenesis in cynomolgus macaques. While others have described clinical, virological, and pathological findings for NHPs infected with MARV Angola [14–16, 23], we are the first to report a detailed immunological analysis in NHPs during MARV Angola infection. Our analysis allows us to determine whether lymphopenia and the cytokine response play an important role during MARV Angola infection. Furthermore, by studying circulatory immune cells in the blood, we are able to describe how this latest MARV strain affects the immune system. S2
•
JID
•
Fernando et al
MATERIALS AND METHODS Animals
Five healthy 16–22-year-old female cynomolgus macaques (Macaca fascicularis) from a NHP colony at Health Canada Animal Resources Division (Ottawa, Ontario) were housed individually in cages and acclimatized 14 days prior to infection. Each 4-cage bank was inside a negative pressure isolator with high-efficiency particulate arrestance filtration. Animals were monitored twice daily and fed commercial monkey chow, treats, and fruit; enrichment consisted of commercial toys and treats. All animal procedures were conducted in a containment level 4 laboratory according to the Canadian Council on Animal Care and with approval by the Animal Care Committee of the Canadian Science Centre for Human and Animal Health. Challenge and Examination
MARV Angola was obtained from an infected patient during the 2005 outbreak in Uige and passaged once in Vero E6 cells. Three NHPs were challenged with 1000 50% tissue culture infective doses (TCID50) of MARV Angola intramuscularly. Two control animals received Roswell Park Memorial Institute 1640 medium. Routine examination on days 0, 2, 4, and 6 after challenge included monitoring of temperature (rectal), respiration rate, weight, hydration status, and discharges. Animals were euthanized once they reached the end point score, in accordance with the approved Animal Care Committee animal protocol. Homogenization of Spleens
Upon euthanasia (days 11 and 12 for the 2 control NHPs and days 6 and 7 for the 3 NHPs challenged with MARV Angola), the spleen was harvested, and a portion was weighed and homogenized in 1.5 mL of Roswell Park Memorial Institute 1640 medium, using the BD Medimachine. The spleen homogenate was centrifuged at 1300g for 10 minutes, and the supernatant was used in the cytometric bead array (CBA) assay. The remaining pellet was washed twice in phosphate-buffered saline and then underwent immunophenotyping, using fluorescenceactivated cell-sorting (FACS) analysis. Immunophenotyping of Whole-Blood Specimens and Splenocytes
Whole-blood specimens (100 µL) or 1 × 106 splenocytes were blocked with 10 µg of human γ-globulin then stained with a 100 µL antibody master mix. Two antibody panels were used, one T-cell panel (CD45, CD3, CD4, CD8, CD25, and CD69) and one NK cell panel (CD45, CD3, CD8, CD16, CD20, and CD69). Splenocytes were also stained with a macrophage panel (CD45, CD11b, CD14, CD15, CD69, and HLA-DR). After incubation for 30 minutes at 4°C, red blood cells and splenocytes were lysed with 1.5 mL and 0.5 mL of 1 × FACS lysing solution (Becton Dickinson), respectively; vortexed; and incubated at room temperature for 15 minutes. Pellets were washed
with 1 mL of wash buffer (Dulbecco’s phosphate-buffered saline without calcium or magnesium [Invitrogen], with 1% heat inactivated fetal bovine serum). Cells were pelleted and resuspended in 900 µL of phosphate-buffered saline, with 4% paraformaldehyde. Samples were evaluated on a BD LSRII flow cytometer, and data were analyzed using FACS Diva software v5.0.2 (BD Biosciences). Prior to acquisition, 100 µL of Flow-Count Fluorospheres (Beckman Coulter) were added to determine absolute counts. Cytometric Bead Array (CBA) Analysis of Plasma Samples and Spleen Homogenate Supernatant
A flow cytometric BD Biosciences cytokine/chemokine CBA Flex Set system was used to assess the levels of cytokines and chemokines in plasma samples and spleen homogenate supernatant that were exposed to 5 MRads of γ-irradiation on dry ice upon removal from the containment level 4 laboratory. The samples were evaluated in duplicate, using the CBA Flex Set system according to the manufacturer’s instructions, and included beads for interleukin 2 (IL-2), interleukin 4 (IL-4), IL-6, CXCL8 (interleukin 8 [IL-8]), interferon γ (IFN-γ), CCL2 (MCP-1), CCL5 (regulated on activation, normal T-cell expressed and secreted [RANTES]), and tumor necrosis factor (TNF). Samples were run on the LSRII (BD Biosciences), with FACSDiva v5.0.2 software and analyzed with BD’s FCAP v1.0.1 software. Statistical Analyses
The t test was used to determine significant responses for the CBA and flow cytometry data. A significant result was defined as a P value of
