Previous Article | Next Article ![]()
Journal of Virology, May 2005, p. 5507-5515, Vol. 79, No. 9
0022-538X/05/$08.00+0 doi:10.1128/JVI.79.9.5507-5515.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Veit Hornung,2,
Stefan Finke,1
Margit Günthner-Biller,2
Sabrina Marozin,1
Krzysztof Brzózka,1
Sharareh Moghim,1
Stefan Endres,2
Gunther Hartmann,2 and
Karl-Klaus Conzelmann1*
Max von Pettenkofer Institute and Gene Center,1 Division of Clinical Pharmacology, Department of Internal Medicine, Ludwig-Maximilians-Universität München, Munich, Germany2
Received 2 November 2004/ Accepted 27 December 2004
|
|
|---|
|
|
|---|
isotypes and IFN-ß (21). Although most types of cells can produce IFN through recognition of cytosolic double-stranded RNA (1a, 36, 44), or upon stimulation of TLR3 and TLR4 through double-stranded RNA or lipopolysaccharide, respectively (1), the vast amount of IFN upon entry of bacterial and viral pathogens is produced by a specialized cell population, plasmacytoid dendritic cells (PDC) (2, 6). Transcriptional induction of IFN genes is controlled by interferon regulatory factors (IRFs). IRF-3 mainly regulates IFN-ß induction, whereas IRF-7 has the ability to activate IFN-
promoters (22, 25, 45). In contrast to other cell types, PDC constitutively express high levels of IRF-7 such that expression of IFN-
by PDC is independent of the IFN-
receptor-mediated positive feedback via IFN-ß (3, 13, 16, 18), explaining in part the promptness of high-capacity IFN-
production.
The TLR repertoire of human PDC is composed of TLR7 and TLR9, both located in the endosomal membrane. As shown recently, TLR7 and TLR8 recognize viral single-stranded RNA (8, 12) as well as imidazoquinolines such as imiquimod and resiquimod (R848) and guanosine analogs (reviewed in references 1 and 42). In contrast, TLR9 recognizes bacterial or viral DNA (1), including synthetic CpG oligodeoxynucleotides (ODN) (11). Indeed, recent work revealed IFN-
production in PDC after incubation with a variety of inactivated or live DNA and RNA viruses, including herpes simplex virus types 1 and 2 (16, 19, 23), murine cytomegalovirus (7), human immunodeficiency virus (46), influenza A virus (8, 24), Sendai virus (14, 16), and vesicular stomatitis virus (3, 24). For herpes simplex virus (19, 23), Influenza A virus (8, 24), and vesicular stomatitis virus (24), the critical involvement of MyD88 adaptor-dependent TLR9 and TLR7 signaling has been demonstrated.
In addition to perceiving external virus components through TLR7 and TLR9, human PDC have the means to sense cytosolic replicating RNA viruses. As we could show recently, respiratory syncytial virus (RSV) escapes from recognition by PDC TLRs (14). Nevertheless, infection with a particular laboratory strain of RSV (subtype A, strain Long), or cytosolic delivery of double-stranded RNA but not of poly(I:C) led to potent IFN-
induction in PDC in a TLR- and protein kinase R-independent manner (14).
The considerable repertoire of tools for sensing pathogens combined with a tremendous capacity to produce IFN make human PDC the key sentinels for exciting a generalized host alert upon infection. Indeed, activation of PDC by a variety of pathogens and synthetic TLR agonists profoundly shapes the host immune system by promoting Th1 immune responses and suppressing Th2 immune responses (6, 11, 15, 42). This effect is exploited in the therapeutic use of TLR7 and TLR9 ligands, such as R848 and CpG ODN, which are promising as immunoprotective agents, vaccine adjuvants, and antiallergens (1, 11, 42).
It may therefore appear from the current literature that PDC are impeccable in sensing intruders. However, in view of the key role of human PDC in innate and adaptive immunity, natural viruses must have evolved tools to counteract IFN production by PDC. In order to identify such viruses, we scrutinized two negative-stranded RNA viruses of the Paramyxoviridae family, measles virus (MeV) and RSV, which were considered successful candidates because of their immune-suppressive features and Th2-biased immune response (10, 27, 35). Since RSV strain Long was previously identified as an efficient inducer of IFN in PDC (14), infection of different human cell types, including PDC, other immune cells, and fibroblast cell lines, was done as a control. Infection of PDC with RSV A2 and four primary RSV isolates from hospitalized children and the vaccine MeV Schwarz effectively and similarly repressed IFN expression. Most importantly, TLR7- and TLR9-dependent IFN-inducing pathways stimulated by R848 and CpG ODN were abolished by infection with MeV and RSV A2. Thus, these viruses can make PDC blind to other pathogens and therapeutic IFN stimuli.
|
|
|---|
Primary human RSV isolates from hospitalized children suffering from bronchiolitis or obstructive bronchitis were provided by H. Werchau, Bochum, Germany, and were first amplified by two passages on HEp2 cells (ATCC CCL-23) before preparation of IFN-free virus stocks on Vero cells as described above.
MeV strain Schwarz was obtained from a commercial batch (Ch.-B: W5663-2) of measles vaccine Meriéux (Pasteur Mérieux MSD, Leimen, Germany); 500 50% tissue culture infective doses were incubated with 2.5 x 106 Vero cells for 1 h in a 15-ml Falcon tube with 2.5 ml of DMEM without FCS. Cells were seeded in DMEM supplemented with 2.5% FCS at 37°C in a 5% CO2 atmosphere until development of an extensive cytopathic effect. A virus stock for infection experiments was prepared from this first passage on Vero cells as described above for RSV.
For infection of epithelial cell lines HEp2, 293, and A549, 2.5 x 104 cells were seeded in DMEM with 10% FCS in 48-well plates and were grown for 16 h at 37°C and 5% CO2. Cell numbers were then determined from three extra wells. FCS-containing medium was replaced by 200 µl of DMEM without FCS and containing the indicated viruses adjusted to a multiplicity of infection of 3. After 1 h of incubation virus-containing medium was removed and cells were washed and further incubated with DMEM plus 10% FCS. As a negative control for infection, cells were incubated in parallel with mock supernatant obtained from freeze-thaw cell lysates of noninfected Vero cells.
Isolation of human hematopoietic cells. Human peripheral blood mononuclear cells were prepared from whole blood of young healthy donors by Ficoll-Hypaque density gradient centrifugation (Biochrom, Berlin, Germany). PDC were isolated by magnetic activated cell sorting (MACS) using the BDCA-4 dendritic cell isolation kit from Miltenyi Biotec as described (14). Briefly, PDC were labeled with anti-BDCA-4 antibody coupled to colloidal paramagnetic microbeads and passed through a magnetic separation column once (LS column; Miltenyi Biotec). The purity of isolated PDC (lineage-negative, major histocompatibility complex class II-positive, CD123-positive) was between 75% and 100%. PDC from individual donors were used separately in all experiments and were not pooled. Contaminating cells represented mainly T cells.
Prior to isolation of monocytes and B cells, PDC were depleted by MACS (CS column; Miltenyi Biotec). B cells were isolated by MACS using the BDCA-1 dendritic cell isolation kit and B-cell isolation kit, respectively (both Miltenyi Biotec). Monocytes and T cells were isolated from buffy coats by Ficoll (Lymphoflot, Biotech, Dreieich) gradient centrifugation (33) at 1,500 rpm in a Heraeus Varifuge. T cells were separated from monocytes and B cells by plastic adherence for 1 h. A fraction of strongly adherent monocytes was separated from B cells by overnight incubation and rigorous washing three times. Remaining adherent monocytes were trypsinized and primary cultures of monocytes or T cells were grown in suspension in RPMI supplemented with 10% FCS at 37°C and 5% CO2 atmosphere. The purity of primary cell cultures was determined by fluorescence-activated cell scanning (FACS) analysis using phycoerythrin-labeled monoclonal CD3 (Serotec), CD14 (Serotec), and CD19 (Serotec) antibodies. The purity of B cells, monocytes, and T cells was greater than 95%, 95%, and 80%, respectively. Viability was >95% as determined by trypan blue exclusion.
Prior to virus infection, T cells were activated by incubation with phytohemagglutinin (5 µg/ml, Sigma), while B cells and monocytes were activated by incubation with lipopolysaccharide (100 ng/ml, Sigma) in RPMI with 10% FCS at 37°C and 5% CO2 for 16 h. For infection of 5 x 104 PDC or 10 x 104 B cells, T cells, or monocytes, cells were seeded in 100 µl of RPMI (Gibco) plus 10% FCS into 96-cluster-well plates; 100 µl of RPMI plus 10% FCS containing infectious virus adjusted to yield a multiplicity of infection of 3 was added, and cells were incubated at 37°C in a 5% CO2 atmosphere.
TLR agonists. R848 was purchased from Invivogen (Toulouse, France). CpG ODN 2216 (5'-ggGGGACGATCGTCgggggG-3'; lowercase letters are phosphorothioate linkage; capital letters are phosphodiester linkage 3' of the base; italics are CpG dinucleotides) was kindly provided by Coley Pharmaceutical Group (Wellesley, Mass.). R848 or CpG ODN 2216 was added to PDC at the indicated time points at final concentrations of 2 µg/ml and 6 µg/ml, respectively.
Flow cytometry. Infection of epithelial cell lines and of primary hematopoietic cells was determined by cell surface staining of viral envelope proteins using 2.5 x 105 cells (except for PDC, where 2 x 104 cells were used), fixed for 5 min with 3% paraformaldehyde at room temperature. After washing with FACS buffer (phosphate-buffered saline containing 0.4% FCS and 0.02% NaN3), staining for 30 min on ice was performed using the mouse monoclonal antibodies RSV-F (Serotec) and K83, recognizing MeV H (kindly provided by S. Schneider-Schaulies, Würzburg, Germany). As a negative control, mock-treated cells were stained in parallel. After washing, cells were incubated for 30 min with fluorescein isothiocyanate-labeled anti-mouse immunoglobulin antibody (Dianova) followed by washing and FACS analysis on a BD Biosciences (Heidelberg, Germany) FACSCalibur (excitation at 488 and 635 nm).
Detection of cytokines.
For enzyme-linked immunosorbent assay (ELISA) detection of human interleukin-8 and IFN-
, the interleukin-8 kit and the IFN-
ELISA module set (detection range, 8 to 500 pg/ml) of Bender MedSystems (Graz, Austria) was used. The IFN-
module set ELISA detects most IFN-
isoforms but not IFN-ß. ELISA for an individual experiment was performed in parallel with 20 µl of cell-free supernatant, each collected at indicated time points and stored at 20°C. Assays were performed according to the manufacturer's recommendations.
|
|
|---|
induction in epithelial cells. Human A549, 293, and HEp-2 cells were equally well infected with RSV A2 and Long, as demonstrated by expression of the fusion (F) protein on the surface of infected cells (Fig. 1B). However, IFN-
production was significantly lower in RSV A2-infected cells at 12, 24, and 36 h postinfection (Fig. 1A). This is consistent with previous observations indicating that nonstructural proteins (NS) 1 and 2 of RSV A2 are able to counteract IFN induction in A549 cells (39). Among other IFN antagonistic activities (4, 31), the RSV nonstructural (NS) proteins interfere with activation of the essential IFN transcription factor IRF-3 (5). In contrast to RSV A2, however, RSV Long induced approximately five times more IFN-
, suggesting that this virus has lost the ability to effectively counteract IRF-3 activation in virus-infected cells. Similar to RSV A2, infection by MeV Schwarz did not lead to IFN-
induction considerably exceeding that of mock-infected cells, confirming potent IFN-antagonistic activity described for measles virus (29).
![]() View larger version (31K): [in a new window] |
FIG. 1. MeV Schwarz and RSV strain A2 but not RSV strain Long are able to inhibit production of IFN- in epithelial cell lines. (A) Human A549, 293, and HEp2 cells were infected with RSV A2, RSV Long, or the MeV vaccine strain Schwarz. After 12, 24, and 36 h (white, grey, and black bars, respectively) IFN- secreted into the culture supernatants was determined by ELISA. Results are shown as induction compared to supernatants of mock-infected cells. (B) Cell surface expression of RSV fusion protein (F) and MeV hemagglutinin (H), determined 36 h postinfection by FACS with monoclonal antibodies, was used to verify productive virus infection. Error bars indicate the standard deviation in three independent experiments.
|
production in T and B cells and in monocytes, whereas RSV A2 and MeV completely suppressed IFN production (Fig. 2A). Notably, cells infected with the latter viruses produced even less IFN-
than mock-infected and lipopolysaccharide- or phytohemagglutinin-activated cells, suggesting an inhibition of TLR4-mediated IFN production.
![]() View larger version (29K): [in a new window] |
FIG. 2. MeV and RSV A2 but not RSV Long inhibit production of IFN- in primary human hematopoietic cells. (A) Freshly isolated T cells, B cells, and monocytes were either incubated with medium or activated by addition of lipopolysaccharide (LPS) or phytohemagglutinin (PHA) for 16 h, followed by virus infection or incubation with mock supernatant. After 24, 36, and 48 h (white, grey, and black bars, respectively), the amount of IFN- in supernatants was determined by ELISA. (B) Cell surface expression of RSV F protein and MeV H protein determined by FACS at 48 h postinfection as described for Fig. 1. Error bars indicate the standard deviation of mean in experiments using primary cells from three (T cells), six (B cells), or seven (monocytes) individual donors.
|
![]() View larger version (28K): [in a new window] |
FIG. 3. MeV and RSV A2 but not RSV Long inhibit production of IFN- in infected PDC. Freshly isolated PDC were infected with the indicated viruses and treated with mock supernatant or supernatant containing 4 µg/ml of ODN2216. After 12, 24, and 36 h (white, grey, and black bars, respectively), secreted IFN- (A) and interleukin-8 (B) were determined by ELISA. Error bars in indicate the standard deviation of mean of four experiments using PDC from six individual donors (A) or three donors (B). Infection of PDC was assayed by FACS monitoring of cell surface expression of RSV F and MeV H protein at 36 h (thin line) and 72 h (bold line) postinfection. For a negative control (dotted line), mock-infected cells were incubated for 72 h.
|
production, PDC were treated with 4 µg of CpG ODN 2216, which strongly activates PDC through TLR9 binding (20)/ml. Supernatants from virus-infected, mock-infected, and ODN-treated PDC were assayed 12, 24, and 36 h later for IFN-
production by ELISA. Both CpG ODN 2216-treated and RSV Long-infected PDC produced huge and comparable amounts of IFN-
, with concentrations reaching 13.8 ng/ml and 11.6 ng/ml at 36 h, respectively (Fig. 3A). However, in PDC infected with RSV A2 or measles virus, IFN-
production was suppressed. In particular, MeV infection allowed not much more IFN to be produced than in mock-infected cells. Importantly, infection with the different viruses had no obvious effect on production of interleukin-8, which is NF-
B dependent (18, 28). Treatment of PDC with supernatants from either mock-infected or virus-infected cell cultures led to a similar and approximately 10-fold increase of interleukin-8 compared to PDC treated with fresh medium (Fig. 3B). These data confirm the viability of virus-infected PDC and indicate that RSV NS proteins block IRF-3 and IRF-7 activation in PDC, whereas the activity of the NF-
B and AP-1 transcription factors is not affected, as was demonstrated previously for epithelial cells (5). Similarly, the major target of MeV appears to be the IRF rather than the NF-
B activation pathway.
To confirm that natural RSV shares the ability of A2 to prevent IFN-
induction in PDC, four clinical RSV isolates from children hospitalized with bronchiolitis or obstructive bronchitis were included in the assay. All four isolates were able to infect PDC similarly to RSV A2 (Fig. 4C) and suppressed IFN-
production equally well (Fig. 4A), while they did not suppress production of interleukin-8 (Fig. 4B). Thus, infection of human PDC and potent inhibition of IFN-
induction are intrinsic features of clinical human RSV isolates. RSV A2 but not RSV Long has fully retained this important wild-type phenotype.
![]() View larger version (30K): [in a new window] |
FIG. 4. Clinical RSV isolates from hospitalized children counteract IFN- production in infected PDC. PDC were infected with the indicated clinical RSV isolates or RSV Long or incubated with mock supernatant. After 12, 24, and 36 h (white, light grey, and dark grey bars, respectively), secreted IFN- (A) and interleukin-8 (B) were determined by ELISA. Error bars show the standard deviation of the mean of four experiments using PDC from individual donors. (C) Expression of RSV F on the surface of PDC as determined by FACS (mean fluorescence) after 36 h (grey) and 72 h (black) of incubation.
|
expression. Since RSV A2 and MeV infection of PDC suppresses IFN-
production, it is suggested that these viruses interfere with the pathway leading from recognition of cytosolic viral double-stranded RNA to activation of IRF-3.
To assess whether RSV A2 and MeV are also able to interfere with the TLR-dependent IFN-
-inducing pathways, we measured the effect of virus infection on TLR7 and TLR9 signaling upon stimulation with known TLR agonists. First, the response to the TRL7 ligand R848 (resiquimod) was investigated. Stimulation of mock-infected PDC with 2 µg of R848 per ml led to production of up to 9 ng/ml IFN-
(Fig. 5). Even higher amounts, up to 15 ng/ml, of IFN were produced in PDC infected with RSV Long. Since similarly high IFN-
levels were previously observed in the absence of R848 in RSV Long-infected cells (see Fig. 3), a synergistic effect of the IFN-inducing virus and R848 stimulation was not evident. In striking contrast to the situation with RSV Long, R848 treatment of PDC infected with RSV A2 and MeV did not result in production of considerable amounts of IFN-
. Thus, RSV A2 and MeV are able to antagonize TLR7-mediated IFN induction pathways.
![]() View larger version (12K): [in a new window] |
FIG. 5. RSV A2 and MeV infection precludes IFN production mediated by the TLR7 agonist R848. Human PDC were mock infected or infected with the indicated viruses for 16 h, followed by addition of fresh medium or medium containing 2 µg/ml of R848. IFN- production was determined at 12, 24, and 36 h post-R848 treatment (white, grey, and black bars, respectively). Error bars indicate the standard deviation of the mean in three experiments using PDC from five individual donors.
|
, reaching more than 15 ng/ml at 36 h (Fig. 6A). In cells previously infected with RSV Long, similar amounts of IFN were produced. An additive activation of IFN-
by ODN 2216 and the IFN-inducing RSV Long was indicated by even higher IFN production at 12 h postinfection. In striking contrast, IFN secretion was almost completely abolished in cells infected with RSV A2 or measles virus (Fig. 6A). Thus, RSV A2 and measles virus are able to prevent CpG-mediated activation of IFN through TLR9.
![]() View larger version (19K): [in a new window] |
FIG. 6. MeV and RSV A2 abolishes ODN-induced IFN- production of PDC. (A) PDC were infected with the indicated viruses for 16 h, followed by stimulation with 6 µg/ml ODN2216. IFN production after 12, 24, and 36 h (white, grey, and black bars, respectively) was measured by ELISA. Error bars indicate the standard deviation of the mean in three experiments using PDC from five individual donors. (B) Human PDC were first stimulated by addition of ODN 2216 for 6 h (upper panel) or 24 h (lower panel). Then, the medium was removed and the cells were infected with the indicated viruses at a multiplicity of infection of 3 or mock infected. The effect of the 6-h ODN 2216 stimulus is almost completely abolished by MeV and RSV A2, and infection after the long 24-h stimulus is still sufficient to diminish IFN production by approximately 40% and 50%, respectively. Error bars indicate the standard deviation in experiments using PDC from four individual donors.
|
, infection with RSV A2 and measles virus strongly diminished IFN-
secretion turned on during the 6-h ODN incubation period. Indeed, values were only slightly greater than previously observed for nonstimulated and mock-infected PDC (e.g., Fig. 6A). Moreover, virus infection was able to considerably reduce IFN production even after a 24-h ODN 2216 stimulation. Compared to mock- and RSV Long-infected PDC, a reduction of IFN-
levels to less than 50% was observed in RSV A2-infected PDC at all time points monitored (Fig. 6B, lower panel). These results show that previously activated operational TLR9 signaling can be shut down by both RSV A2 and measles virus. |
|
|---|
Intriguingly, deaths from measles are largely due to an increased susceptibility to secondary bacterial and viral infections, attributed to a prolonged state of immune suppression (27, 35). One mechanism contributing to immune suppression is contact-mediated inhibition of T-cell proliferation by measles virus surface proteins (32). The present findings that human PDC are highly permissive for infection with measles virus in vitro and fail to produce IFN upon different TLR stimuli may further explain how measles virus facilitates bacterial superinfections in vivo and encumbers the development of an adequate immune response. Of note in this respect is the typical strong Th2 bias in measles (27). Since IFN-
drives the immune response towards Th1, shutdown of the major human IFN-producing cells by measles virus may considerably contribute to this feature of measles pathology.
Unlike measles virus, which causes a generalized infection by targeting mainly cells of hematopoietic origin and therefore may reach a considerable portion of the PDC population, RSV infection remains restricted mainly to epithelial cells of the upper and lower respiratory tract. The finding that clinical RSV isolates have the capacity to prevent PDC IFN responses, however, strongly suggests that this feature is also vital for local infection of the respiratory tract by RSV. Notably, further immune-relevant features are shared by measles virus and RSV, including a pronounced in vitro contact inhibition of T cells by the viral surface proteins (33) and a Th2-biased immune response (10). Although bacterial superinfections in RSV infection appear to contribute little to the disease in humans, they are often described for infections of calves with bovine RSV (see the references in reference 33), a relevant animal model for human RSV.
In RSV-infected non-PDC cell types, induction of IFN-
and IFN-ß is prevented by the activity of two unique proteins, the viral NS1 and NS2 proteins (5, 31, 39, 43). As we have shown recently, the NS1 and -2 proteins interfere with the activation of the essential IFN transcription factor IRF-3, while NF-
B and AP-1 activity is not affected (5). Both cytosolic double-stranded RNA-triggered and TLR3- and TLR4-dependent IFN induction pathways appear to merge in the phosphorylation of IRF-3 (and IRF-7, if present) by two homologous kinases, tank binding protein 1 (TBK-1), and the inducible inhibitor of kappa kinase, IKK-i (1, 9, 26, 38). IRF-3 activation through TLR3 and TLR4 involves the adapter TRIF, which has been shown to associate with TBK-1 and IKK-i (reviewed in reference 1). In contrast, TLR7 and TLR9 IFN signaling depends on MyD88 (8, 12, 17), which directly binds IRF-7 and TRAF6 but not IRF-3 (17).
The observation that RSV Long has lost the ability to counteract both TLR7- and TLR9-dependent and cytoplasmic IFN-inducing pathways while RSV A2 and all clinical RSV isolates have retained the full inhibitory set suggests molecular targets common to TLR-dependent and -independent IFN inducing pathways, namely, phosphorylation of the IRFs by the different IRF kinases. This is supported by our experiments with virus-infected B cells and monocytes stimulated by lipopolysaccharide (in order to render these cells permissive for virus infection), which indicate that RSV and measles virus not only abolish the TLR7/9 MyD88-dependent signaling to IFN activation but might also counteract TRIF-dependent TLR4 signaling (see Fig. 2). As the NS proteins of RSV are critically involved in inhibiting IFN induction in non-PDC and the bovine RSV NS has been shown to prevent phosphorylation of IRF-3 (5), it will be interesting to determine whether these proteins are also involved in counteracting TLR-mediated IFN induction e.g., by preventing activation of IRF-7.
The situation with the IFN antagonists responsible for preventing IFN induction of wild-type measles virus (29) is less clear. As for other Paramyxovirinae, the V protein of measles virus interferes with IFN signaling through STAT (30), and a corresponding activity has been reported for the measles virus C protein (37). Since PDC are capable of producing IFN-
independently of the IFN-
receptor-mediated positive feedback via IFN-ß (3, 13, 16, 18), these viral activities on IFN signaling are not sufficient to account for the observed effective IFN-
shutdown in PDC. It therefore remains to be clarified which measles virus factors interfere with IFN induction.
From the present experiment, it appears that measles virus and RSV have similar targets. Interestingly, measles virus strain Edmonston, which is closely related to strain Schwarz, has been shown to activate IRF-3 and to induce IFN in 293 cells (41). Comparison of the sequences of the strains might help to reveal the genetic basis for interference with IFN induction in fibroblasts and probably in PDC. Further analysis of the mechanisms used by RSV and measles virus to prevent IFN production in PDC will not only lead to recombinant vaccine viruses with abolished or reduced capacity to undermine the host immune response but may also help to develop tools to modulate the activities of therapeutic TLR ligands.
This work was supported by the Deutsche Forschungsgemeinschaft through SFB 455 and SPP1089 Co260/1-1 and by the European Commission (5thFP-PRIOVAX-QLK2-CT-2002-81399). G.H. was supported by grants from the Deutsche Forschungsgemeinschaft, HA 2780/4-1 and SFB 571, the Mildred Scheel Stiftung 10-2074, the Friedrich-Baur-Stiftung, and the Human Science Foundation of Japan.
J.S. and V.H. contributed equally to this work. ![]()
|
|
|---|
induction in human plasmacytoid dendritic cells by a single-stranded RNA virus. J. Immunol. 173:5935-5943.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»