|Home | About | Journals | Submit | Contact Us | Français|
Early results have recognized that influenza virus infects the innate and adaptive immune cells. The data presented in this paper demonstrated that influenza virus labeled with fluorescent dye not only retained the ability to infect and replicate in host cells, but also stimulated a similar human immune response as did unlabeled virus. Influenza virus largely infected the innate and activated adaptive immune cells. Influenza B type virus was different from that of A type virus. B type virus was able to infect the immature lymphocytes, but in lower amounts when compared to activated lymphocytes. Protection from influenza is tightly associated with cellular immunity. Traditional methods of cellular immunity assay had limitations to imitate the natural human cell-mediated responses to influenza virus. Labeled viruses could be used in the assay of virus-specific cytotoxicity, which might reflect the natural process more closely. Furthermore, human immune cells activated by one influenza subtype virus could kill the cells infected by other subtype virus. These results implied the human immune cells could directly handle and remove free virus using similar mechanism that was used to remove virus-infected non immune cells, which might help to simplify the design and production of influenza vaccine, thereby reduce the cost.
Influenza virus is controlled by both humoral and cell-mediated immune responses (Doherty et al., 2006; Thomas et al., 2006). The traditional detection of human cell-mediated responses used 51Cr-release assays. Enzyme-linked immunospot, intracellular cytokine staining, CD107 staining, and specific T cell receptor staining by the use of fluorescent major histocompatibility class (MHC) I peptide complexes are alternative methods for the detection of CTL responses (Altman et al., 1996; Rubio et al., 2003; Sun et al., 2003; Tomaru et al., 2003). These methods are aimed at the detection of virus-specific CTLs in lymphocyte populations (Yang, 2003; Lieberman, 2004). As non-radioactive alternatives for the 51Cr assay, several assays used fluorescent dyes to discriminate between target and effector cells and to assess target-cell elimination (Lecoeur et al., 2001; Lee-MacAry et al., 2001; Liu et al., 2002). In order to measure the outcome of the cytolytic cascade, the detection methods of cytolytic activity have been developed; they include sets of overlapping peptides pulsing and directly transfecting target cells with DNA vector encoding antigen-green fluorescent protein (GFP) fusion protein (Maecker et al., 2001; van Baalen et al., 2005). A disadvantage to this approach is that the infection of cells may differ from that identified by exogenously added peptides and DNA vector (Addo et al., 2001). It has been reported that conjugates of HA protein of influenza virus are suitable for obtaining a strong immune response (Huebner, 1997). Fluorescent isothiocyanate (FITC) labeled influenza virus was able to infect both innate and adaptive immune cells, and kept the ability of infection and replication as unlabeled virus (Nichols et al., 1993). However, whether there is a difference between native and activated lymphocytes in the infection of influenza virus and whether influenza virus labeled with fluorescent dye stimulates the same immune response as unlabeled virus remains unclear. The data presented in this paper demonstrated that influenza virus labeled with fluorescent dyes kept their abilities of infection, replication and stimulation to immune response. Labeled influenza viruses had a selective infection for innate and activated adaptive immune cells. The target cells, which were infected with fluorescent dye labeled influenza virus, were co-cultured with CTL effector cells, and their killing by CTLs was monitored by flow cytometry.
Human peripheral blood was collected in heparinized tubes from healthy donors. Influenza virus strains: A/Puerto Rico/8/34/[H1N1], A/Wyoming/03/2003 [H3N2], A/New Caledonia/20/99 [H1N1], A/Panama/2007/99 [H3N2], and B/Hong Kong/330/2201 were purchased from the Centers for Diseases Control and Prevention (CDC), USA. This work was approved by the University of Connecticut Health Center.
Human PBMC were prepared with Histopaque-1077 (Sigma-Aldrich, St. Louis, MO), washed with PBS once, treated with RBC Lysis Solution (Bio-Rad, Hercules, CA) for 5 minutes, washed with PBS once again, then resuspended in AIM-V media (Gibco Laboratories, Grand Island, NY), then adjusted into 2.0 × 106/ml. 1 ml of suspended cell was added into each well of a 48-well multiplate (Nalge Nunc, Rochester, NY), and stimulated with influenza virus (TCID50=4 × 10 6/ml, MOI=2) at 37°C in 5% CO2. The concentration of phytohemagglutinin (PHA, Sigma) was 1μg/ml AIM-V medium.
The following antibodies used for flow cytometry were purchased from BD Pharmingen: anti-CD8-PerCP, anti-Perforin (Perf) -PE and anti-CD107a-PE antibodies. In addition, anti-CD3-PE-Cy7 antibody was purchased from eBioscience. Fv17 single chain anti-human granzyme B antibody (scFv GrB), which was a gift from C. R. Bleackley and K. P. Kane (Departments of Biochemistry and Immunology, University of Alberta, Edmonton, Alberta, Canada), was labeled with reactive fluorescent isothiocyanate (FITC, Sigma) (Rong et al., 2004). Cells for flow cytometry were prepared as previously described (McElhaney et al., 2006). In brief, cells (0.5-1×106) were incubated with surface antibodies, washed with cold 0.2%BSA/PBS before and after fixing with 2% paraformaldehyde, then resuspended in a cold permeabilization buffer (0.3% saponin, 5% normal human serum PBS). Following scFv GrB intracellular staining, cells were washed with 0.1% saponin and 0.2%BSA/PBS, resuspended in 0.2%BSA/PBS, and transferred to FACS tubes for data acquisition. Data were acquired on BD LSR II, while 30,000 events were counted for each sample and analyzed using Flow Jo software (Tree Star, Ashland, OR).
Fresh PBMC (0.7 × 106) were labeled with 0.05mCi 51 Cr (chromium, supplied as Na2CrO4; ICN) for 1 hour at 37°C. PBMC were washed with AIM-V three times, infected with A/Wyoming/03/2003 [H3N2] at MOI=2 as the target. 100 μl (5000cells) of labeled target cells in AIM-V were added to various PBMC, as indicated by the effector-to-target ratios in final 200μl. Target cells with 5% Tween 20 or AIM-V media were used to determine the maximum release and spontaneous release. The mixtures were cultured in u-bottomed 96-well plates for 4 h. Supernatants (100 μl) were harvested, mixed with scintillation fluid, and counted in a 1450 MicroBeta Trilux liquid scintillation counter (Wallac Inc. Fort Myers, Florida). The percentage of specific lysis was calculated using the formula: % specific lysis = [(experimental release spontaneous release)/(maximum release spontaneous release)] × 100 (Powers and Belshe, 1993).
Fitc-Annexin-V was purchased from BD Bioscience.
Human PBMC (1.5 × 106 cells/ml in AIM-V media) were stimulated with virus (MOI=2) for 5 days and PBMC lysates were prepared and analyzed by the GrB assay using the Acetyl-Ile-Glu-Pro-Asp-paranitroanilide (Ac-IEPD-pNA, Bachem) substrate according to previously described methods (Ewen et al., 2003; McElhaney et al., 2006). GrB activity was measured in PBMC lysates by cleavage of the substrate. GrB activity was calculated by A405 U/mg protein in the PBMC lysate. The lysate of YT cells was used as the standard. The supernatants were collected into new tubes as samples for the detection of IFN-γ. IFN-γ was measured with Human IFN-γ ELISA Ready-SET-Go kits (eBiosciece, San Diego, CA), following the company provided protocol.
2.0μl of 1M sodium bicarbonate was added into 10μl (100–500 μg) of purified virus and mixed. Fluorescent dyes including Cyanine (Cy3, Amersham), Cy5 (Amersham), Phycoerythrin (PE, Dojindo Molecular Technologies, Inc. Gaithersburg, MD) and Allophycocyanin (APC, Dojindo Molecular Technologies, Inc. Gaithersburg, MD) were dissolved in 3μl of DMSO. 3μl of fluorescent dye was added to the virus and mixed gently but thoroughly. The mixture of virus and dye was incubated in the dark for one hour at room temperature. Added 6μl of 4M hydroxylamine to the mixture, mixed thoroughly and incubated for 15 minutes in the dark at room temperature. The mixture was then loaded on the top of a 1ml G-50 column (Ambion Inc. Austin, TX), and centrifuged at 750g for two minutes. The dye labeled virus ran through to the 1.5ml tube. Free dye (Cy3 and Cy5) was retained in the column matrix. Virus separated from free PE and APC was centrifuged by Beckman L8–55 Ultracentrafuge at 100,000rpm for 30 minutes. The dye optical density (OD) values of the labeled virus were measured by DU® Series 650 UV/Vis Scanning Spectrophotometer. Using TCID50 as particle number, each virus particle could couple 2000–6500 dye molecules in all experiments.
Madin-Darby canine kidney (MDCK) cells were digested with 0.25% trypsin, washed with PBS, adjusted to 1 × 105/ml in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco Laboratories, Grand Island, NY). Viruses were diluted sequentially in 0.002% trypsin DMEM. 100μl of the virus and 100μl MDCK cells were mixed into each well of a 96-well plate and incubated at 37°C in 5% CO2 for 48 hours. 100μl of supernatant and 100μl of washed chicken (0.5%) red blood cells were then mixed in a v-bottomed 96-well plate. The plates were kept at 4°C and read at 60 minutes. The aggregation of chicken red blood cells was as positive.
Data were expressed as mean and standard error. Comparisons were analyzed using Student’s paired t- tests. The significance level was set at 0.05 (p<0.05).
Early studies showed that fluorescent dye labeled influenza virus had the capacity of infection and replication in host cell (Nichols et al., 1993). It remained unclear, however, whether there are differences between labeled and unlabeled virus in the stimulation of the human immune response. In this study, influenza viruses were labeled with Cy3, Cy5, PE or APC. Labeled viruses were titrated with MDCK cell and chicken red blood cells. These results indicated that labeled influenza viruses kept the ability of infection and replication in the host cell. Labeled and unlabeled viruses were used to stimulate human PBMC for 5 days. PBMC were probed with anti human CD8, CD3, GrB and Perf antibodies. The activity of GrB of cell lysates and IFN-γ in culture media were measured. The infection with Cy5-labeled virus A/Panama/2007/99 [H3N2] enhanced the expression of GrB in CD8 T cells as did unlabeled virus did. All of the GrB+ CD8 T cells were Perf+ (Fig. 1A). Both labeled and unlabeled influenza virus stimulated a similar increase in the activity of GrB in human PBMC (Fig. 1B). The labeled and unlabeled influenza virus (Cy3-virus: A/Puerto Rico/8/34, [H1N1]) also stimulated an increase in production of IFN-γ (Fig. 1C). The viruses labeled with various fluorescent dyes showed the same results (data not shown), which suggested that labeled influenza virus not only kept the capacity of infection and replication in the host cell, but also had the ability to stimulate a similar human immune system responses as unlabeled virus.
Fresh human PBMC were infected with labeled virus. In the forward and side scatter dot plot, the inactivated lymphocytes were defined as gate 1 (G1), monocytes and activated lymphocytes as gate 2 (G2). The results showed that Cy3 labeled influenza virus, A/Panama/2007/99 [H3N2], infected almost the entire G2 population, but not the inactivated lymphocytes (G1) (Fig. 2A). The viral infection was dynamically monitored by flow cytometry (Fig. 2B, 2C). Cy5 labeled influenza A strain virus, A/Puerto Rico/8/34 [H1N1], only infected the G2 population. Labeled influenza B type virus strain B/Shanghai/361/2002 infected both G1 and G2 populations, but lower amounts of the B virus infected G1 as compared to the G2 population. Other fluorescent dyes labeled viruses showed the same results (data not shown). Furthermore, human PBMC were infected with PE labeled virus A/Wyoming/03/2003 [H3N2] and probed with anti-CD3, CD4, CD8 and GrB antibodies. Virus A/Wyoming/03/2003 [H3N2] was largely taken by activated (GrB+) lymphocytes virus (Fig. 2D).
The traditional 51Cr release assay was used to test the killing caused by influenza virus A/Wyoming/03/2003 [A/H3N2]. PBMC were stimulated with virus for five days as effectors (E). PBMC that were not stimulated with virus served as a controls. Autologous PBMC, stimulated with PHA for five days followed by infection with virus, were used as the target (T). A mixture of various ratios of effector/target (E:T) were incubated for 4 hours at 37°C, in 5% CO2. The percentage of specific lysis 51Cr release increased with E:T ratio. The results suggested that at least a 5:1 E:T ratio could be used to setup a killing test (Fig. 3A).
The labeled virus killing experiments were carried out as follows: PBMC stimulated with virus A/New Caledonia/20/99 [H1N1] for 5 days were added to autologous PBMC labeled influenza virus A/Puerto Rico/8/34 [H1N1] at an E:T ratio of 6:1. The results of flow cytometric analysis showed that the peak of the cells infected with labeled virus dropped in the mixture with PBMC stimulated with virus, but still existed in the control mixture (Fig. 3B). Most cells infected with Cy3-labeled virus were killed by PBMC stimulated with the virus.
The target cells infected with Cy3 labeled virus A/Panama/2007/99 [H3N2] were mixed with the PBMC (effector) that were stimulated with virus A/Puerto Rico/8/34, [H1N1] for five days. Fluorescence of labeled virus decreased with time after targets and effectors were mixed (Fig. 4A). Tracked with the early apoptotic marker, annexin-V, and labeled virus, the results showed that apoptosis occurred in virus infected cells (Fig. 4B).
The cumulative exposure of granular membrane proteins (CD107a and b) on the cell surface of responding antigen-specific T cells provides a positive marker for degranulation. Significant expression of cell surface CD107a and b can be observed as early as 30 minutes following stimulation of primary CD8 T cells, and reaches a maximum by 4 hours (Betts et al., 2003; Rubio et al., 2003). Previous intracellular staining showed that the expression of GrB and CD107a and b were identical, GrB and/or CD107a and/or b could be a marker of virus specific CTL (Xie and McElhaney, 2007). The killing process caused by influenza virus infection was probed with anti CD3, CD8, CD107a and GrB antibodies. The results showed that the killing process was accompanied by the increase of CD107a on GrB+ CD8 T cell surfaces (Fig. 5A). The number of GrB+ CD8 T cells also increased. The killing process was dynamically monitored with Annexin V and anti-CD107a antibody. Fig. 5B and 5C showed that apoptosis occurred rapidly (within 5 minutes) and almost simultaneously with degranulation, reaching a peak at about 30 minute. This result strengthened the conclusion that the degranulation of CD8 T cells was involved in the apoptosis occurring in virus infected PBMC.
Although it is well-known that influenza virus can agglutinate erythrocytes by attachment to specific sialic glycoprotein receptors (Fields, 1996), the interaction between influenza virus and human immune system remains unclear. Since PBMC are mainly involved in immune system responses, these cytotoxicity assays indicated that influenza virus could infect these immune cells and the infected cells could be killed by effector cells. Current knowledge states that a virus-infected cell must be producing viral peptides and displaying them on the surface of the cell in order to be recognized by cytotoxic immune cells. However, this simplistic understanding of the interaction between the immune system and the influenza virus hardly explains how the immune system clears itself of infection. Does the interaction involve binding of virus to specific immune cells or does virus bind all immune system components? Can the human immune system directly remove free virus? Therefore, it is necessary to develop a better understanding of the immune mechanisms that are linked to influenza virus infection. Early studies showed that fluorescence labeled virus infected a majority of monocytes/macrophages and lymphocytes (Nichols et al., 1993). This study further proved that influenza A type virus largely infected the innate and mature adaptive immune cells. Influenza B type virus was different from that of A type virus. Lower amounts of B type virus infected the immature (inactivated) lymphocytes when compared to activated lymphocytes.
The results presented in this paper suggested that the human immune cells activated by one influenza subtype virus could kill the cells infected by other type virus (Fig. 3B and Fig. 4A). Even B type virus-activated immune cells killed A type virus infected cells and vice versa (data not shown). These suggest that CTL activated by on one type virus can handle other type virus. In other words, activated CTL has lower target specificity. Therefore, producing enough CTLs might be the most important issue in combating virus infection. These results were different from traditional opinion.
Dynamic analysis of the killing process showed that apoptosis and GrB release (degranulation) occurred within 5 minutes after effectors and targets were mixed. Cytotoxicity assays using labeled virus might reflect the natural process more closely, and therefore offer an advantage over methods utilizing radioactivity and viral peptides. Unlike the traditional 51Cr release, which is toxic to cells, this assay used no radioactivity. Combining with other surface and intracellular markers, such as CD4, CD8, CD3, CD62L, etc., this labeled virus also supports to track the fate of cells that have been infected at low MOI.
Finally, these results suggested that human immune cells could directly handle and remove free virus using a similar mechanism that was used to remove virus-infected non immune cells. These might give more options for the vaccine design and the treatment of infection diseases.
The work was supported by grants of NIH RO1 AG20634.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.