In this study, we have reported that type I IFNs induce a rapid differentiation of freshly isolated GM-CSF–treated human monocytes into short-lived DCs endowed with potent functional activities both in vitro and in vivo, using the hu-PBL-SCID mouse model. The comparison of DCs generated in the presence of IFN/GM-CSF with those obtained after the standard IL-4/GM-CSF treatment revealed that type I IFN was definitively superior in inducing a rapid and stable differentiation process and in conferring to DCs a full functional activity to trigger a potent primary human immune response both in vitro and in hu-PBL-SCID mice. In particular, IFN induced an early detachment of monocytes from culture plates, paralleled by rapid acquisition of high levels of CD40, CD54, CD80, CD86, and HLA-DR molecules within 3 d, whereas IL-4/GM-CSF–treated monocytes required at least 6 d to fully acquire the immature DC phenotype. Of interest, a remarkable percentage of CD83-expressing DCs was observed in IFN-treated cultures, but not in IL-4–stimulated cells. Notably, an upregulation of CD83 expression by type I IFN has been reported, under very different experimental conditions, by Luft et al. using DCs generated from CD34+
progenitor cells 17
and recently by Radvanyi et al. 19
. However, the novel finding concerning the CD83 induction by IFN reported in our study consists in the fact that the expression of this marker was acquired as early as 3 d after cytokine treatment, without any further treatment such as TNF 19
or LPS. We also found that CD83 expression was invariably associated with higher levels of HLA-DR and CD86. Notably, the fact that IFN/GM-CSF–treated cells had acquired a final commitment towards mature DC phenotype was suggested by the finding that upon cytokine removal, IFN/GM-CSF–treated cultures retained the DC phenotype, without adhering to the flask surface, whereas IL-4/GM-CSF–treated DCs reacquired the macrophage characteristics and readily readhered to culture flasks within 3 d, unless preventively stimulated to terminally differentiate. Of interest, the rapid differentiation observed in our IFN-treated cultures is reminiscent of an in vitro transendothelial trafficking model, in which immature DCs arose from monocytes within 2 d 4
In all of the experimental conditions reported in this study, the maturation effects induced by type I IFN proved to be superior to those achieved by the IL-4/GM-CSF combination, as evaluated by FACS® analysis of antigen expression as well as by both MLR assays with allogeneic PBLs, and lymphocyte proliferation responses to autologous antigen-pulsed DCs. Three different preparations of type I IFNs (CIFN, IFN-α2b, and IFN-β) exhibited comparable effects, emphasizing the general involvement of these prototypic cytokines in the process of DC maturation/activation from monocytes.
In this study, we have also reported that type I IFN induces IL-15 mRNA expression and IL-15 secretion in GM-CSF–treated DC cultures. To the best of our knowledge, this is the first report showing IL-15 induction by type I IFN in monocyte-derived DCs. Notably, Zhang et al. 33
had shown that type I IFN can induce IL-15 production in mouse macrophages. IL-15 is a recently identified pleiotropic cytokine produced by a wide variety of cells and tissues. This cytokine has been shown to act as a chemoattractant for T cells 34
, favoring APC–T cell interaction and enhancing the generation of cytotoxic T cells 35
. IL-15 increases IFN-γ production by NK and activated T cells 36
and strongly synergizes with IL-12 for IFN-γ production by upregulating IL-12Rβ1 expression on CD4+
T cells 37
. IL-15 mRNA has recently been shown to be restricted to the CD83+
fraction of monocyte-derived DCs at late stages of DC culture 38
. In fact, DC activation is generally needed to induce IL-15 mRNA expression in IL-4/GM-CSF–cultured DCs. At the protein level, IL-15 production by DCs has been reported to be low, unless triggered by phagocytic activity or cellular interactions 38
. In light of all this and of the recently considered involvement of IL-15 33
in the in vivo expansion and survival of CD8+
memory T cells in mice exposed to type I IFN 12
, we may assume that IL-15 secretion induced by type I IFN in DCs can represent an important event in the adjuvant activity of IFN for the proliferation and persistence of memory T cells in response to viral infections. Of interest, IFN-treated DCs induced a strong Th1-biased response, as evaluated by the overall cytokine production in allogeneic MLR assays as well as in proliferation assays with HIV-1–pulsed DCs. In particular, DCs generated in the presence of IFN/GM-CSF showed a potent ability to take up, process, and present inactivated HIV-1 to autologous T lymphocytes in vitro, which was clearly superior to that observed using DCs cultured with IL-4/GM-CSF. On the basis of these in vitro results, we decided to evaluate the capability of HIV-1–pulsed DCs generated in the presence of either IFN/GM-CSF or IL-4/GM-CSF to elicit a primary human immune response in vivo, by using SCID mice reconstituted with autologous PBLs. The hu-PBL-SCID mouse model 20
has been widely used for a variety of studies on pathogenesis 394041
and therapy 25
of viral infections, especially using HIV-1. However, the generation of a primary human immune response in the chimeras has only been reported in a very few studies 29303132
. We postulated, however, that an efficient immune response could be elicited in hu-PBL-SCID mice provided that appropriate antigen presentation could occur in vivo as a result of transfusion of autologous antigen–pulsed DCs at early times after reconstitution with hu-PBLs. Remarkably, we found that immunization of hu-PBL-SCID mice with IFN/GM-CSF–cultured autologous DCs pulsed with AT-2–inactivated HIV-1 resulted in the generation of potent primary immune response towards HIV-1, as evaluated by the detection of specific human antibodies against the whole spectrum of viral proteins (). At 7 d after immunization, human antibodies proved to be mostly IgM, whereas HIV-1–specific IgG1 antibodies were detected at 2 wk, suggesting a Th1-like response. This may reflect the DC capability of promoting both the generation of IgM response by naive B cells 42
and IL-10–independent IgG1 isotype switching 43
. Notably, the antibodies detected in the sera of mice injected with DCs generated in the presence of IFN had a potent neutralizing activity in vitro against HIV-1. The levels of human antibodies to HIV-1 were consistently higher in hu-PBL-SCID mice injected with DCs generated in the presence of type I IFN compared with those detected in the xenochimeras transplanted with the corresponding virus-pulsed DCs developed in the presence of IL-4.
Intriguingly, even though DCs generated in the presence of IFN showed an enhanced functional activity both in vitro (persisting for at least 6 d) and in vivo (in the hu-PBL-SCID mouse model), a considerable percentage of the IFN-treated cells underwent apoptosis at 5 d of culture. Comparative RT-PCR analysis revealed a marked induction of TRAIL expression in IFN/GM-CSF–treated DCs. Notably, both TRAIL and functionally active receptors TRAIL-R1 and R2 were upregulated in response to LPS treatment. Of interest, the TRAIL expression in DCs generated in the presence of IFN/GM-CSF proved to be functionally important in inducing apoptosis of target cells ( D). In addition to IL-15 induction and CD83 upregulation, TRAIL upregulation represents an additional intriguing similarity with the effects induced by LPS in DCs. We speculate that some common signal transduction pathways can be induced by both IFN and LPS, leading to DC maturation. Considerable amounts of IFN (500 U/ml) were secreted after LPS stimulation of IL-4/GM-CSF–cultured DCs. Further studies are needed to clarify whether the IFN secreted by DCs in response to LPS can play some role in the terminal maturation of DCs.
Our results suggest the following scenario of events, with potential new implications for vaccine research. As DCs are the key players in the induction of the immune response, a rapid availability of mature DCs at specific sites is of crucial importance for an effective immune control against infections. Large amounts of type I IFN are locally produced by specific cell types, such as natural IFN-producing cell (NIPC)/pDC2 89
, thus favoring DC development from monocytes. Locally produced type I IFN might protect T cells from antigen-induced apoptosis 1415
and induce IL-15 production by DCs, thus favoring the proliferation of certain T cell subsets 1233
. Likewise, type I IFN can also inhibit antigen-induced apoptosis of B cells 44
, thus promoting survival of antibody producing cells and enhancing humoral immune response. The IFN-induced upregulation of TRAIL expression may play a dual role: to induce a suicide of activated DCs, as suggested by apoptosis induction, and to induce apoptosis of virus-infected or tumor cells, which have been reported to be susceptible to TRAIL-induced apoptosis (23
; D). Both of these events might be beneficial for an efficient self-controlling host immune response against infections, and the IFN-mediated apoptosis of DCs may represent a natural feedback mechanism to limit Th1 response, thus avoiding subsequent tissue injury. However, the induction of TRAIL-mediated apoptosis of virus-infected or tumor cells by DC generated by IFN exposure may also enhance specific immunity. In fact, DCs have been shown to take up apoptotic bodies from dying cells and cross-prime cytotoxic CD8+
T cells 45
. Thus, the IFN/DC-mediated apoptosis of virus-infected or tumor cells may favor the direct priming of the CD8+
T cell response against viral or tumor antigens.
There are several important implications and perspectives stemming from the results described above. Viruses or other stimuli capable of inducing type I IFN (such as poly-I:C or LPS) and CD40L, which also induce IFN production by IL-4/GM-CSF monocyte-derived DCs (our unpublished observations), are the strongest inducers of maturation/activation of DCs. On the basis of the results described above, it is reasonable to assume that the activation of the IFN system can represent the early important mechanism involved in the maturation/induction of DCs in response to virus infection and possibly to other invading pathogens or tumors.
Type I IFNs are the most used cytokines in patients with viral and neoplastic diseases, even though the in vivo mechanisms underlying the considerable clinical response are still poorly understood. Notably, the current clinical use of these cytokines does not yet imply the rationale of their use as vaccine adjuvants. The results of this study, together with an ensemble of data recently obtained by several groups 1046
, indicate that type I IFNs can represent powerful natural adjuvants for the development of novel vaccination strategies against virus infections and tumors. Finally, in addition to these practical perspectives for vaccine development, the elucidation of the self-controlling mechanism underlying the “alliance” between DCs and type I IFN represents a crucial step in understanding the basis of the early immune response to infection and underlines the importance of these cytokines as the natural bridge system connecting innate and adaptive immunity.