Published online before print January 31, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Viral Immunology Group, School of Biotechnology, Dublin City University, Dublin, Ireland; and
National Institute of Biological Standards and Controls, Herts, UK
1 Correspondence: Viral Immunology Group, School of Biotechnology, Dublin City University, Dublin 9, Ireland. E-mail: patricia.johnson{at}dcu.ie
 TOP ABSTRACT REFERENCES |
|---|
Key Words: monocytes • T cells • viral infections
Circulating monocytes continuously exit the bloodstream and enter body tissues, where they can give rise to tissue resident macrophages, as well as specialized dendritic cells (DCs) [1
]. During respiratory viral infection, virus replication occurs in the epithelial layer of the respiratory tract [2
], ultimately resulting in maturation and migration of DCs to secondary lymphoid organs, where they can activate naïve T cells [3
]. This highlights the importance of the microenvironment at the site of infection and its influence on DC generation. Conventional DCs, representing the monocyte-derived DCs (IL-4 DCs), which are generated in the presence of GM-CSF and IL-4, have played center stage for many years in standard in vitro assays [4
, 5
]; nevertheless, their relevance in vivo remains unclear, especially as this DC generation pathway, involving IL-4, is unlikely to represent the cytokine milieu present at a site of intercellular infection [6
]. High levels of Type 1 IFN are physiologically produced in vivo in response to viral infections or inflammatory stimuli [7
], and this has led to the more recent use of the Type 1 IFN-DC [6
]. However, in view of the subversive effects of several viruses on the Type 1 IFN pathway, these artificial methods of DC generation may inappropriately skew immune responses to viral infections [8
]. In the light of the critical role these cells play in defining immunological outcomes, not only from infection but inflammation and neoplasia, it is imperative that the most appropriate cell type be generated and examined in a given disease state. Therefore, we comprehensively compared the maturation profiles of monocytes, IL-4 DCs, IFN-DCs, and DCs, generated from infected epithelial supernatants (A549 DCs), in response to viral infections, TNF-
treatment, or untreated controls. In this report, we show that DCs, generated from direct infection of monocytes with human parainfluenza virus Type 3 (HPIV3) or influenza, mimic those generated from infected epithelial supernatants. Virally infected IFN-DCs induce distinct DC subsets from the virally infected monocytes and IL-4 DCs. We also show that this difference in DC generation is translated into diverse T cell responses to each distinct DC subset. Thus, the pathway used for generating DCs is crucial to the outcome of viral infections, as different DC subsets can respond differently to the same virus, thereby generating conflicting ex vivo immune data.
CD14+ monocytes were purified by positive selection from PBMC preparations to >90% using an antibody/magnetic cell sorter (MACS) magnetic bead cell separation protocol (Miltenyi Biotec, Auburn, CA, USA). Unstimulated, purified monocytes displayed an immature/inactivated phenotype (hiCD14, hi CD11c, loCD83, loMHC lo/mod CD80 CD86; Fig. 1D
). IL-4 DCs and IFN-DCs were generated as described previously [4
, 6
]. Monocytes, IL-4 DCs, and IFN-DCs were infected with HPIV3 (kindly provided by Marius Loetscher, Berna Biotech AG, Switzerland) at a tissue culture infectious dose (TCID), 50/ml, of six or influenza virus (kindly provided by Robert Newman, National Institute of Biological Standards and Controls, UK) at a TCID, 50/ml, of seven for 2 h at 37°C and then subsequently washed following incubation [5
] and cultured for the indicated time-points. As a positive control, monocytes, IL-4 DCs, and IFN-DCs were also cultured with 25 ng/ml TNF-. A549 epithelial cells were also infected directly with HPIV3 or influenza, as described above, and supernatants were harvested at 24 h postinfection and UV-inactivated or cultured directly with monocytes (A549 DC) at a 50% (v:v) ratio for the indicated time-points.
 View larger version (24K): [in a new window] |
Figure 1. Phenotypic pattern from direct viral infection of monocytes and A549 DCs versus IL-4 DCs and IFN-DCs. Multicolor flow cytometry was performed using standard protocols [8
]. Briefly, cells were stained with human FITC-conjugated anti-CD80, CD11c; PE-conjugated anti-CD14, -CD86, -CD83, and -CD123; and allophycocyanin (APC)-conjugated HLA-DR (eBioscience, San Diego, CA, USA). Isotype controls included FITC-, PE-, and APC-conjugated mouse IgG1 and FITC and APC-conjugated mouse IgG2b (eBioscience). All FACS was done on the BD FACSCalibur using CellQuest software (Becton Dickinson, San Jose, CA, USA). Data represent the mean fluorescence intensity (MFI) of (A) IL-4 DCs, (B) IFN-DCs, (C) A549 DCs, and (D) monocytes following 48 h culture. Results are a culmination of two to four independent experiments.
|
chain (CD123), following infection with HPIV3 (Fig. 1C
and 1D)
. Consistently, cells infected with live virus were positive for viral (HPIV3 and influenza) nucleoprotein by RT-PCR. UV-treated samples were negative (data not shown). HPIV3-infected IFN-DCs showed a similar expression profile of costimulatory molecules, MHC Class II, and CD123 to monocytes (Fig. 1B)
. However, infection of the IL-4 DCs with HPIV3 generated a DC subset totally distinct from the infected monocyte DC (Fig. 1A)
. Although there was strong up-regulation of the costimulatory marker CD86 compared with its positive control TNF-, the lack of expression of CD123 and poor up-regulation of MHC Class II indicate the emergence of two distinct DCs. Similarly, influenza virus generated A549 DCs, which mimic direct infection of the monocytes but with less potency than the HPIV3-generated DCs (Fig. 1C
and 1D)
. We also observed up-regulation of costimulatory markers and MHC Class II from influenza-infected monocytes, IFN-DCs, and IL-4 DCs (Fig. 1A
1B
1C
, and data not shown). However, influenza-infected monocytes were poorly activated compared with the influenza-infected IL-4 DCs and IFN-DCs. These results suggest that artificial pretreatment of these cells may over-ride immunomodulation strategies used by the different viruses.
Next, we investigated the cytokine secretion profile from these infected DCs. Significant levels of the antiviral, proinflammatory cytokine IFN- were secreted from all DC subsets infected with HPIV3 (Fig. 2A
2B
2C
2D
). It is interesting that IL-4 DCs were the only DC subset capable of secreting large amounts of TNF- in response to HPIV3 infection (Fig. 2A)
. IL-4 and IFN-DCs secreted a significant amount of IL-12p40 in response to TNF- compared with directly infected monocytes or A549 DCs (Fig. 2A
2B
2C
2D)
. High levels of the regulatory cytokine IL-10 were only observed in supernatants from the HPIV3-infected monocytes compared with the HPIV3-infected IL-4 DCs, IFN-DCs, and A549 DCs (Fig. 2A
2B
2C
2D)
. In contrast, little or no cytokines were secreted from the influenza-infected DCs (Fig. 2A
2B
2C
2D)
. Thus, we demonstrate further the importance of different DC generation pathways in influencing the production of distinct cytokine profiles from virally infected cells.
 View larger version (26K): [in a new window] |
Figure 2. Cytokine secretion profile of virally infected IL-4 DCs, IFN-DCs, monocytes, and A549 DCs. Cells were infected with HPIV3 or influenza and cultured for 48 h as described above. Supernatants were harvested, and IFN-, IL-10, IL-12p40, and TNF- secretion was determined by ELISA (R&D Systems, Minneapolis, MN, USA). Data reflect the mean concentration ± SD for infected (A) IL-4 DCs, (B) IFN-DCs, (C) A549 DCs, and (D) monocytes. Statistical differences between HPIV3-infected cells and the various other treatments for IFN- were assessed using two-sample Student’s t-tests, and P values were as follows: (A) P  0.02; (B) P 0.04; (C) P 0.02; (D) P 0.05. Results are representative of two to three individual experiments.
|
, which is a potent inducer of DC maturation [4
], greatly improved the survival of the IL-4 DCs, and 80% of the cells still survived following the 2-day culture (Fig. 3B)
. Also, TNF- increased the survival of the IFN-DCs slightly but had little or no effect on the monocytes (Fig. 3B)
. It is interesting that HPIV3 induced extremely high levels of apoptosis (>80%) in IL-4 DCs (Fig. 3C)
and moderate levels in monocyte and IFN-DCs (Fig. 3C)
. Again, influenza-infected IL-4 DCs exhibited the higher level of apoptosis (
50% of cells) compared with influenza-infected IFN-DCs and monocytes (Fig. 3D)
. Thus, IL-4 DCs appear to be more sensitive to viral-induced apoptosis than the IFN-DCs or monocytes. Similar apoptosis profiles were also observed at 24 h poststimulation (data not shown).
 View larger version (67K): [in a new window] |
Figure 3. Viral-induced apoptosis of IL-4 DCs, IFN-DCs, and monocytes. Cells were infected with HPIV3 or influenza and cultured for 48 h as above. Apoptosis of cells was determined using the Annexin V Detection Kit I (BD PharMingen, San Diego, CA, USA) samples. (A) Control, (B) TNF--treated, (C) HPIV3, and (D) influenza-infected cells were analyzed by flow cytometry. Quadrant markers were set based on isotype controls, and values reflect the percentage of cells in each quadrant. Lower left, Living cells; lower right, early apoptotic cells; upper right, late apoptotic cells. Data are representative of two to three separate experiments.
|
and influenza-treated cocultures (Fig. 4A
and 4B
). However, it is most interesting that HPIV3-infected monocytes and A549 DCs (UV and live HPIV3) did not inhibit proliferation of purified CD3+ T cells (Fig. 4D)
. Thus, preprimed DCs exhibit different T cell stimulatory capacities to the infected monocytes and A549 DCs.
 View larger version (24K): [in a new window] |
Figure 4. T cell proliferation profiles in response to virally infected or treated DC subsets. Cells were infected with HPIV3 or influenza virus and were cultured for 24 h as above. Cocultures of DCs with allogeneic CD3+ T cells or in MLR (CD14-depleted fraction) were set up at a 1:10 ratio for 5 days. Proliferation was evaluated by adding 0.5 µCi/well 3H-thymidine to the cocultures for the last 5 h of incubation. Results reflect the mean scintillation cpm ± SD for (A) IL-4 DC, (B) IFN-DC, (C) A549 DC, and (D) monocyte cocultures and are representative of two to four independent experiments.
|
-induced maturation treatment, and TNF- alone induced little or no effect on the surface marker expression of the monocytes. It is interesting that distinct DC subsets were generated by virally infected IL-4 DCs and IFN-DCs, emphasizing the critical nature of different DC generation pathways in shaping or directing the overall immune response.
Each virally infected cell population also induced their own distinct cytokine secretion profile. HPIV3-infected IL-4 DCs secreted large amounts of TNF- compared with the other subsets, and higher levels of IL-10 were secreted from the HPIV3-infected IL-4 DCs and monocytes compared with the HPIV3-infected IFN-DCs. As IL-10 is a potent, immunoregulatory cytokine, it can have important implications on polarizing T cells, shifting the protective TH1 bias toward a regulatory T cell, such as Tr1, as demonstrated by McGuirk et al. [10
]. Little or no cytokines were secreted from the influenza-infected DCs, probably as a result of cytokine suppression by the nonstructural protein 1 of this virus [8
]. Thus, differing routes to DC generation can result in disparate cytokine responses to viral infection. In addition, we observed varied levels of apoptosis induced from the virally infected cell populations. It is surprising that the IL-4 DCs showed the greatest sensitivity to viral-induced apoptosis compared with the infected IFN-DCs and monocytes. IL-4 DCs are the most commonly used DCs in in vitro assays [4
, 5
]. Our findings that these DCs are hypersensitive to viral-induced apoptosis could have vast implications on viral studies, resulting in inaccurate or misleading conclusions from results.
We investigated the immunostimulatory capacity of these virally infected DCs to examine their functional relevance in vitro. Similar T cell proliferation profiles for the MLR and purified T cells were observed from the virally infected, preprimed DC cocultures. However, MLR proliferation profiles were significantly different to the purified T cells for HPIV3-infected monocytes and A549 DC MLR cocultures. Previous reports have demonstrated that HPIV3-infected, preprimed DCs can inhibit T cell proliferation [5 ]. Thus, discovering that HPIV3-infected monocytes and A549 DCs appear to restore proliferation of purified T cells compared with HPIV3-infected, preprimed cocultures emphasizes the discrepancies that may occur when using virally infected, preprimed DCs. These data strongly suggest that preprimed DCs may not be the most appropriate cell to use in in vitro assays when studying viral infections, as they appear to skew immune responses, and that direct viral infection of monocytes and A549 DCs generated from monocytes cultured with supernatants from infected epithelial cells appears to be the most natural pathway of DC generation. We propose that this DC generation pathway could aid therapeutic advances in viral studies by representing a more accurate model for viral in vitro studies.
Received October 6, 2006; revised November 7, 2006; accepted December 21, 2006.
|
TOPABSTRACTREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. M. Noone, E. Paget, E. A. Lewis, M. R. Loetscher, R. W. Newman, and P. A. Johnson Natural Killer Cells Regulate T-Cell Proliferation during Human Parainfluenza Virus Type 3 Infection J. Virol., September 15, 2008; 82(18): 9299 - 9302. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
