| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Immunology |
1 School of Biochemistry and Molecular Biology, Faculty of Science, 2 Division of Immunology and Genetics, John Curtin School of Medical Research, The Australian National University, Canberra, Australian Capital Territory, Australia; 3 Institut Pasteur, Unitè de Gènètique Molèculaire Bactèrienne, Paris, France; and 4 The Centenary Institute of Cancer Medicine and Cell Biology and 5 Discipline of Medicine, University of Sydney, Sydney, New South Wales, Australia
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
or lipopolysaccharide). Administration of the DC-targeting vaccine to mice challenged with B16-OVA cells induced a dramatic immunotherapeutic effect and prolonged disease-free survival. The results show that the targeting of antigen to DCs in this way is highly effective at inducing immunity and protection against the tumor, with protection being at least partially dependent on the eosinophil chemokine eotaxin. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
DCs originate from progenitors in the bone marrow and migrate as immature cells to peripheral tissues where they internalize antigen and undergo a complex maturation process. Antigen is internalized via a number of surface receptors, including the complement receptor CD11c/CD18 (8, 9, 10)
and the endocytic receptor DEC-205 (11
, 12)
. During antigen acquisition, immature DCs also may receive "danger" signals in the form of pathogen-related molecules such as bacterial cell wall lipopolysaccharide (LPS), or maturation and/or inflammatory stimuli via cytokines such as IFN-. DCs then migrate to the secondary lymphoid organs, maturing to become competent antigen presenting cells (13)
. The receptors CD11c/CD18 and DEC-205 are believed to play a crucial role in the process of antigen capture and presentation and are expressed almost exclusively on DCs. It is conceivable, therefore, that both receptors also could be used for targeting antigen directly to DCs in vivo. Consistent with this notion, a fusion protein consisting of antigen fused with antibodies (Ab) to DEC-205, and a DEC-205 monoclonal antibody (mAb) chemically conjugated with antigen have been shown to target DCs in vivo, inducing T-cell activation when coadministered with inflammatory stimulators such as anti-CD40 Ab (14
, 15)
.
Synthetic liposomes have the potential to deliver large quantities of antigen to DCs (16, 17, 18) , but to date their targeting to specific surface molecules on DCs has been difficult to achieve in practice (18, 19, 20, 21, 22, 23) . Clearly, an effective strategy that combines the antigen carrying capacity of liposomes and the specificity of molecular recognition to target multiple antigens and maturation and/or "danger" signals directly to DCs in vivo would have enormous potential in simplifying DC immunotherapies. Here we examine whether a novel chelator-lipid, 3(nitrilotriacetic acid)-ditetradecylamine (NTA3-DTDA) can be used to anchor histidine-tagged forms of single chain full-length variable Ab fragments (ScFv), which target DCs, onto either tumor-derived plasma membrane vesicles (PMV) or onto antigen-containing stealth liposomes. The results show that the targeting of antigen directly to DC in this way can elicit strong antitumor responses and provide an effective alternative to the ex vivo manipulation of DCs for use in cancer immunotherapy.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
-palmitoyl-ß-oleoyl-phosphatidylcholine (POPC) were obtained from Avanti Polar Lipids Inc. (Alabaster). 2-(4,4-Difluoro-5octyl-4-bora-3a,4a-diaza-s-indacene-3-pentanoyl)-L-hexadecanoyl-sn-glycero 3-phosphocholine (PC-BODIPY) and 5-(and-6)-carboxyfluorescein diacetate, succinimidyl ester, mixed isomers were purchased from Molecular Probes (Eugene, OR). The chelator-lipid NTA3-DTDA, consisting of three nitrilotriacetic acid (NTA) head groups covalently linked to two ditetradecylamine (DTDA) chains was synthesized essentially as described (24)
, but with additional steps to covalently couple a NTA group onto each COOH group of the NTA-DTDA, to produce NTA3-DTDA. NiSO4 was used for all of the additions of Ni2+ to buffers.
Monoclonal Abs and Proteins
Murine CD56 (clone 42.18; rat IgG2a) mAb was from the 6th Human NK Cell Workshop and the murine CD3 mAb (clone 145–2C11; Armenian hamster IgG) was purchased from PharMingen (San Diego, CA). Recombinant murine IFN- and granulocyte macrophage colony-stimulating factor (GM-CSF) were supplied by PeproTech Inc. (Rocky Hill, NJ). Recombinant ScFv Abs N418 (anti-CD11c) and NLDC145 (anti-DEC-205; Refs. 25
, 26
) were developed using the VH and VL regions amplified from cDNA from the respective hybridomas and shown to bind to murine DCs.6
A hexahistidine (6H) tag was incorporated at the COOH terminal of the Abs, CD11c-ScFv and DEC-205-ScFv, respectively, and these were expressed in the baculovirus expression system and purified as described previously (27)
. Peptides were synthesized by the Biomolecular Resource Facility, John Curtin School of Medical Research, Australian National University, Canberra. The L2 peptide (GHHPHGHHPH), a sequence of 10 amino acids found in the plasma protein histidine-rich glycoprotein, was used routinely to engraft control PMV and stealth liposome (SL) because it binds to Ni-NTA3-DTDA with high avidity and can block its nonspecific binding to cells. The peptide SIINFEKL-6H, representing the immunodominant CTL epitope of OVA in H-2b mice (OVA residues 257–264), with hexahistidine tag attached was used for peptide antigen delivery to DCs.
Mice and Cell Lines
Female or male C57BL/6 mice (H-2b) 6–8 weeks of age were supplied by the Animal Breeding Establishment, and C57BL/6 eotaxin knockout mice (H-2b, eotaxin–/–) were obtained from Dr. Paul Foster, Division of Molecular Bioscience, John Curtin School of Medical Research (Australian National University). All of the animal experimentation protocols were approved by the Australian National University Animal Experimentation Ethics Committee. The highly metastatic murine B16-OVA melanoma [C57BL/6 (H-2b)], an OVA-secreting tumor cell line, was cultured at 37°C in an atmosphere of 5% CO2 in RPMI 1640 (Life Technologies, Inc., Invitrogen, Melbourne, Australia) containing 10% FCS (Trace Biosciences, Noble Park, Victoria, Australia) and 0.5 mg/ml Geneticin (Invitrogen). Murine fetal skin DCs (FSDC) [C57BL/6-DBA/2J F1 (H-2b/d)] were cultured in the same medium but without geneticin. Murine long-term culture DCs (LTC-DC) [B10.A(2R; H-2k/b)], isolated and cultured as described (28)
, were a gift from Dr. Helen O’Neill (School of Biochemistry and Molecular Biology, Australian National University).
Isolation of T Cells
Murine T cells were isolated from the spleens of C57BL/6 mice. The spleens were dissociated into single cell suspensions, and after removing red cells by hypotonic lysis, the T cells were isolated using a nylon wool column (29)
. Although this procedure leads to an enrichment in the number of T cells, it does not substantially deplete the number of DCs in the purified cell population. Hence, it can be expected that antigen presentation and cross-presentation could still occur in lymphocyte cultures with these cells.
Engrafted PMVs and SLs
PMVs.
PMV from cultured cells were prepared by sucrose gradient centrifugation as described previously (30
, 31)
. Briefly, cultured B16-OVA cells (1 x 108) were washed twice with PBS to remove proteins from the culture medium. The cells were suspended in homogenization buffer [10 mM Na2HPO4/NaH2PO4 (pH 7.4) containing 30 mM NaCl, 1 mM MgCl2, and 0.02% NaN3] and homogenized by brief sonication at 4°C. The cell lysate was then layered over a 41% sucrose gradient and centrifuged (95,000 x g) for 1 h at 4°C. The PMVs were collected from the interfacial band and washed twice in homogenization buffer by centrifugation (95,000 x g; 20 min; 4°C). Different batches of PMVs were always prepared from a known starting number of cells after the same procedure; each batch of PMVs was standardized for total protein and number of "cell equivalents," with typically 1 x 108 cell equivalents of PMVs being suspended in 100 µl of PBS. Stock PMV suspensions were stored at –20°C and were briefly resonicated before use in each experiment.
Modification of PMVs.
A procedure similar to that used for the incorporation of NTA-DTDA into PMV (31)
was used also for the incorporation of NTA3-DTDA into PMV, because this method was found to be optimal when assessed in studies (data not shown) of the binding of engrafted PMVs to cells in flow cytometry experiments analogous to those shown in Fig. 2A
. For the present work the liposomes used to modify PMVs were prepared as follows: ethanolic solutions of POPC, NTA3-DTDA, LPS, and PC-BODIPY (molar ratio 94:2:2:2); or POPC, NTA3-DTDA, and PC-BODIPY (molar ratio 96:2:2), were mixed, dried under a stream of N2, then rehydrated in 100 µl of PBS containing 60 µM Ni2+. Where indicated, as an alternative to LPS, either IFN- or GM-CSF (50 ng) was included in the rehydration buffer. Hydrated mixtures were sonicated (three times; 15 s bursts on ice) using a TOSCO 100W ultrasonic disintegrator (Measuring and Scientific Ltd., London, United Kingdom) at maximum amplitude. In initial studies the efficiency of LPS/cytokine incorporation in liposomes was assessed after purification by gel filtration and analysis of the "free" and liposome-associated LPS/cytokine by SDS-PAGE; in all of the instances >75% incorporation was obtained (data not shown). Furthermore, consistent with previous reports, our studies showed that liposomes with incorporated LPS bound to FSDC, and liposomes with incorporated IFN- and GM-CSF were functionally active, with GM-CSF-containing liposomes stimulating proliferation when added to cultures of FSDC in serum-free medium and IFN--containing liposomes inhibiting FSDC proliferation in serum-containing medium (data not shown).
|
SLs.
SLs were prepared as follows: POPC, NTA3-DTDA, PE-PEG2000, LPS, and PC-BODIPY (molar ratio 96:1:1:1:1); or POPC, NTA3-DTDA, PE-PEG2000, and PC-BODIPY (molar ratio 97:1:1:1) dissolved in ethanol were dried under a stream of N2 then rehydrated in 100 µl PBS containing 30 µM Ni2+ (total lipid 1 mM). Preliminary studies showed that the inclusion of 1% PE-PEG2000 in the SL lipid mixture had no significant effect on the efficiency of engraftment of either of the two hexahistidine-tagged ScFvs to be used for the targeting of the engrafted SL to LTC-DC in vitro (data not shown), when assessed in binding studies by flow cytometry. For mixtures lacking LPS, IFN- or GM-CSF (50 ng) was included in the PBS (as described for modification of PMV; see above). Lipid mixtures were sonicated and SL purified (as above). For functional studies the PC-BODIPY was omitted from all of the lipid mixtures.
Incorporation of OVA and SIINFEKL Peptide in SLs.
Encapsulation of the immunodominant epitope of the OVA protein, SIINFEKL, into SL was attempted but proved difficult because this peptide has low solubility at the pH used to produce the SL and to engraft histidine-tagged ScFv (pH 7.4). However, a hexahistidine-tagged form of the peptide, SIINFEKL-6H, permitted efficient encapsulation and/or engraftment of the peptide onto NTA3-DTDA-containing SL. Binding studies using fluorescence-activated cell sorter analysis showed that CD11c-ScFv- or DEC-205-ScFv-engrafted SL containing SIINFEKL-6H could effectively target receptors on DCs in vitro, provided that the amount of SIINFEKL used did not exceed 2 µM (data not shown). Thus, where indicated, SIINFEKL-6H (2 µM) was included to simultaneously engraft with ScFv. The efficient encapsulation of OVA into SL containing POPC, NTA3-DTDA, and PE-PEG2000, was achieved by rehydrating the desiccated lipid mixture in PBS containing 0.1 mg of OVA (1 mg/ml) followed by brief sonication. In preliminary studies the total amount of OVA that was encapsulated was checked by SDS-PAGE analysis of the total amount of OVA in the "free" and SL-associated fractions after separation by gel filtration. Under the conditions used the efficiency of OVA encapsulation was >85%. OVA- or SIINFEKL-containing SL were always prepared from known amounts of lipids, OVA, and SIINFEKL peptide, and unassociated material was removed from the SL or modified PMV preparation by gel filtration before engraftment of ScFv for targeting and use in experiments.
Engraftment of ScFv onto Modified PMV and SL.
The engraftment of targeting proteins onto the modified PMV and SL was carried out by a procedure similar to the engraftment of T-cell costimulatory molecules (31)
. Thus, hexahistidine-tagged ScFv to murine CD11c or DEC-205, dissolved in PBS, was incubated with a suspension of the SL or PMV (that had been modified to contain the NTA3-DTDA lipid and the appropriate incorporated LPS, cytokine, and/or the OVA antigen) for 1 h at room temperature. The amount of ScFv used in the engraftment was in the same molar ratio as the effective concentration of NTA3-DTDA used for preparing the SL or modified PMV preparations. This concentration was found to give optimal binding of the ScFv-engrafted SLs and PMV to LTC-DC and FS-DC in preliminary experiments using flow cytometry (data not shown), in studies analogous to those presented in Fig. 2A
.
|
Targeting of DC in Vivo |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
|
Cytotoxicity Assays |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
-irradiated (5000 rad) B16-OVA cells. After 5 days of coculture, the cytolytic activity of the T cells was assessed in a standard 51Cr release assay, as described (31)
. |
Immunization of Animals and Tumor Challenge in Vivo |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
0.16 µg total lipid) bearing associated antigen (0.2 µg of OVA or 0.8 ng of SIINFEKL-6H) and cytokine (0.1 ng), each suspended in a 200 µl volume of PBS. It should be noted that these effective amounts of OVA and IFN- administered per vaccination are small (1%) relative to the amounts typically administered as soluble proteins in vivo. Two weeks after the last injection, the mice were challenged by the i.v. injection of 3 x 105 B16-OVA cells. At day 16 the lungs were removed, and the number of tumor foci was counted visually under a dissection microscope. Alternatively, mice were immunized with ScFv-engrafted B16-OVA PMV at 3, 6, and 9 days after i.v. injection of 1.5 x 105 B16-OVA cells. |
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
. The first entailed the use of a crude preparation of tumor cell-derived PMV modified by engraftment of ScFv targeting DCs, and the second was a preparation of antigen-containing SLs also engrafted with DC targeting ScFv. SLs, also known as sterically stabilized liposomes, are synthetic lipid structures, which have been sterically stabilized by the inclusion of lipids such as PE-PEG2000, and, by virtue of their ability to escape nonspecific elimination by the reticuloendothelial system, can remain in the blood circulation for days after their i.v. administration (18
, 20
, 33)
. The use of the chelator lipid NTA-DTDA to modify tumor cells and tumor cell-derived PMV for engraftment of T-cell costimulatory molecules has been described (31
, 34)
. We have produced recently a novel lipid, NTA3-DTDA (Fig. 1A)
, which is related to NTA-DTDA, but by achieving a higher local density of NTA head groups, can permit a more stable anchoring (effective Kd 5–10-fold lower for NTA3-DTDA than for NTA-DTDA) of histidine-tagged proteins onto NTA3-DTDA-containing SLs and modified PMV.7
Thus, liposome attachment, via NTA3-DTDA, of histidine-tagged ScFv against DC markers such as CD11c and DEC-205 should allow effective targeting of the SLs (Fig. 1B)
and modified PMV (Fig. 1C)
to DCs.
|
|
). A summary of the different liposome and modified PMV preparations used in this study is shown in Table 1
. Because the modified membranes also were made to contain PC-BODIPY as a fluorescent tracer, their targeting to DCs could be assessed by flow cytometry.
LTC-DCs exhibited little binding of control-modified PMV (PMV-L2; 2-fold increase in fluorescence above background), but when incubated with PMV engrafted with either CD11c-ScFv (PMV-CD11c) or DEC-205-ScFv (PMV-DEC-205), there was a 4–8-fold increase in binding above control cells (Fig. 2A)
. A similar result was obtained when LTC-DCs were incubated with SL engrafted with CD11c-ScFv (SL-CD11c) and DEC-205-ScFv (SL-DEC-205; data not shown). Similarly, the incubation of FSDC expressing CD11c, with PMV or SLs engrafted with CD11c-ScFv, resulted in a fluorescence increase substantially above that of control cells (data not shown). Importantly, preincubation of LTC-DCs with either the anti-CD11c mAb N418 or the anti-DEC-205 mAb NLDC145, but not an isotype-matched control mAb, inhibited binding of the respective ScFv-engrafted SL or PMV by 90% (see Table 2
), demonstrating binding specificity for the engrafted ScFv.
|
0.1%), whereas a considerable number of fluorescein-positive cells was seen in mice injected with ScFv-engrafted PMV (3.8%; see Fig. 2B
, panels i-iii), with the vast majority of these cells being CD11c+. We also carried out analogous studies in which sections of the draining lymph node were examined by confocal scanning laser microscopy, after incubation with biotinylated CD11c mAb and streptavidin-rhodamine. The results show that compared with the control (Fig. 2C
, panels i and ii), virtually all of the CD11c+ cells in the draining lymph node from mice injected with ScFv-engrafted PMV exhibit fluorescein fluorescence (Fig. 2B
, panels iii–vi). The findings show that ScFv-engrafted PMV can target DCs in vivo.
Liposome-Mediated Targeting of Antigens to DCs Induces Potent Tumor-Specific Immunity Both in Vitro and in Vivo.
To determine whether antigen-bearing PMV and SL targeted to DCs can be used to induce functional responses in vivo, we initially examined the ability of ScFv-engrafted PMV and SL to stimulate antigen-specific CTL responses. Recent studies have demonstrated the importance of danger/inflammatory signals during antigen exposure and DC maturation (14
, 15
, 36)
in determining the type of immune response initiated by DCs. Thus, although the studies presented above showed that engrafted liposomes and PMV can target antigen to DCs in vitro, previous studies suggest that for this approach to induce an immune response in vivo, the codelivery of a maturation and/or danger signal to DCs also is required. To deliver both antigen and a danger signal to DCs simultaneously, we produced antigen-bearing modified PMV and SL that contained incorporated LPS, IFN-, or GM-CSF.
It is well known that LPS, a bacterial cell-wall component, can incorporate into lipid membranes. In preliminary studies we showed that the incorporation of LPS (1%) into SL can increase their binding to FSDC, suggesting that at least some of the liposome-associated LPS can interact with DC surface receptors, consistent with previous observations (37)
. In addition, evidence suggests that liposome-associated IFN- is biologically active and that IFN- can be encapsulated within liposomes and as well can bind to the outer surface of liposomes (38
, 39)
. Similar findings have been reported with GM-CSF (40)
. These attributes permit liposome-associated LPS, GM-CSF, and IFN- to interact with DC surface receptors and also provide the possibility that upon the binding, disruption, and/or fusion of the targeted SL or PMV, with the membrane of DCs, the SL-encapsulated agent that is released is then able to interact with receptors on DCs. Experiments also indicated that up to 1% of LPS could be included in the lipid mixture and that PMV and SL could be made to incorporate GM-CSF and IFN- with high efficiency, without significantly interfering with the ability of ScFv-engrafted SL to target DCs in vitro, as assessed by binding studies using flow cytometry (data not shown). Moreover, similar to what has been reported for the soluble cytokines (41)
, proliferation studies with FSDC showed that SL-incorporated GM-CSF induces the proliferation of FSDC in serum-free medium, whereas SL-incorporated IFN- inhibits proliferation in complete medium (data not shown). The results thus show that these liposome-associated cytokines are functional. Importantly, FSDC proliferation assays also could be used to monitor cytokine incorporation in the SL, which correlated well with the amounts of "free" versus SL-associated cytokine fractions as determined by SDS-PAGE analysis of fractions separated by gel filtration. The results showed that typically >85% of the GM-CSF and >75% of IFN- became associated with the SL under the conditions used (data not shown).
To determine whether DC-targeted PMV or antigen-containing SL could generate CTL responses in vivo, we immunized C57BL/6 mice i.v. with preparations that either lacked or contained danger/maturation signals such as LPS, IFN-, or GM-CSF. We then isolated splenic T cells, restimulated the cells in vitro with -irradiated B16-OVA tumor cells, and assessed their cytolytic activity toward B16-OVA cells in a standard 51Cr release assay. The T-cell preparations used in the CTL assays against the B16-OVA were not depleted of DCs, and, hence, it would be expected that cross-presentation of B16-OVA-derived antigens can occur in vitro via these cells. Representative lytic curves are shown in Fig. 3A
for animals that were immunized with various PMV preparations engrafted with the DEC-205-ScFv. The data show that little CTL activity can be detected when mice were preimmunized with PMV engrafted with the L2 peptide or with DEC-205-ScFv in the absence of a danger/maturation signal (Fig. 3A)
. The incorporation of either LPS or IFN- in the DEC-205-ScFv-engrafted PMV, however, resulted in the induction of high levels of cytolytic activity, with 50% specific lysis of target cells still occurring at a 1:1 E:T ratio (Fig. 3A)
. In contrast, GM-CSF was a much less effective inducer of CTL activity.
|
. Maximum CTL activity was observed with splenic T cells from mice immunized with PMV or SL (SIINFEKL or OVA bearing) containing IFN- or LPS as the danger molecule. CD11c-ScFv-engrafted PMV and SL were somewhat less immunogenic, with GM-CSF being generally a less effective maturation or danger signal than IFN- or LPS but, nevertheless, inducing significant CTL activity when associated with PMV and OVA containing SL. Interestingly, cultures containing T cells from animals injected with ScFv-engrafted PMV or SL lacking an associated maturation or danger signal, gave near background levels of lysis (Fig. 3, A and B)
. In parallel experiments we tested B16, EL-4, and EL-4-OVA cells, instead of B16-OVA cells, as targets. Lytic activity against EL4-OVA cells was higher than against B16 or EL4 cells, but significantly lower (3–4-fold) than against B16-OVA cells (data not shown). The cytolytic response generated by ScFv-engrafted PMV and SL, therefore, was specific for OVA and possibly antigens unique to the B16-OVA cell line.
Liposome-Based Vaccines That Target DCs Induce Protective Immunity Against Tumors.
Mice immunized with the various B16-OVA preparations were examined for their ability to resist an i.v. challenge of B16-OVA tumor cells, with lung metastases being quantified 16 days after tumor cell injection. Compared with control mice, a much lower number of metastases was observed in mice immunized with PMV or OVA-bearing SL engrafted with ScFv and containing either LPS or IFN- (Fig. 4A)
. Fig. 4B
shows representative images of the lungs from mice vaccinated with PMV containing IFN- and engrafted with either L2 peptide (Fig. 4B
, panel i), or with DEC-205 ScFv (Fig. 4B
, panel ii), from which clear differences in the appearance of the lungs can be seen. In experiments where the PMV or OVA-bearing SL were not engrafted with a ScFv and/or did not contain LPS or IFN- little protection to tumor cell challenge was detected. In stark contrast, SIINFEKL containing SL was unable to protect mice against tumor challenge (Fig. 4A)
, despite some of the vaccine constructs inducing potent CTL activity (Fig. 3)
. These data are consistent with the B16-OVA being resistant to clearance by CD8+ CTLs to this epitope (35)
.
|
at day 3 and also at day 6 and day 9, after challenge with 1.5 x 105 B16-OVA tumor cells. Interestingly, vaccinated mice subsequently did not show any signs of tumor development up to 8 months after tumor challenge, whereas control animals had to be euthanized at day 22 due to an increasing tumor burden in the lungs, which contained an average of 250 ± 37 tumor foci.
The inhibition of tumor growth observed by targeting ScFv-engrafted PMV or SL containing antigen (and encapsulated with IFN- or LPS) to DCs in vivo raised the question of the mechanism by which the antitumor effect is induced. In vitro studies in which splenic DCs isolated from C57BL/6 mice were pulsed separately with different PMV and OVA-SL preparations indicated that in addition to antigen, engrafted ScFv (either CD11c or DEC-205) was essential to induce functional antigen presentation, which led to the proliferation of both CD4+ and CD8+ T cells, almost to the same extent (data not shown). Also, in these studies SIINFEKL-SL stimulated proliferation primarily of CD8+ T cells, as expected. Because CD4+ T cells have been implicated recently in the clearance of B16-OVA melanoma lung metastases through a mechanism involving the eosinophil chemokine eotaxin (35)
, the possibility that eotaxin also plays a role in the antitumor effects seen by targeting antigen directly to DCs in vivo was tested. For these studies eotaxin knockout mice were immunized with different vaccine preparations (either controls or PMV-DEC-205) before assessing the CTL activity of their splenic T cells toward B16-OVA cells as targets and assessing tumor growth in mice inoculated with the B16-OVA tumor. The results show that whereas the CTL activity of T cells from normal and eotaxin knockout mice are essentially identical (Fig. 5A)
, eotaxin knockout mice immunized with PMV-DEC-205 exhibit a marked deficiency in their ability to inhibit tumor growth and metastasis (Fig. 5B)
. This supports a role for eotaxin in inhibiting tumor growth in this system.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
, when incorporated into either SLs or into tumor cell-derived PMV (B16-OVA), enables the stable engraftment of hexa-histidine-tagged ScFv that target surface molecules on DCs (Fig. 1, B and C)
. The present work shows that PMV and SLs engrafted with ScFv specific for the DC markers CD11c and DEC-205 bind specifically to DCs in vitro (Table 2)
and, based on flow cytometry and confocal microscopy studies, can target associated antigens directly to DCs in vivo (Fig. 2)
. Whereas hybrid Abs, or mAb conjugated with antigens, have been used to target DCs (9
, 10)
, to our knowledge the present study is the first to demonstrate that DCs can be targeted using ScFv engrafted onto a membrane system.
Evidence suggests that maturation and/or "danger" signals are important in the maturation and migration of DCs after antigen exposure, and can avoid induction of tolerance to the presented antigen (14
, 15
, 36)
. Such signals are generally not required in in vitro antigen presentation assays, presumably because the DCs are "perturbed" or activated during their isolation. Because LPS and cytokines like GM-CSF and IFN- are known to influence the ability of DCs to take up antigen and to mature (13
, 41, 42, 43)
, for animal studies, therefore, we incorporated LPS, IFN-, or GM-CSF, within PMV and SL, thereby providing the means to simultaneously deliver both antigen and a danger signal to DCs.
An examination of the ability of ScFv-engrafted PMV and SL containing antigen to induce DCs to initiate CTL responses revealed that, compared with control cells, T cells from animals immunized with ScFv-engrafted PMV or antigen-bearing SL exhibit an increased ability, after in vitro restimulation, to lyse B16-OVA target cells in vitro (Fig. 3)
. Importantly, the results show that in vivo priming for cytolytic activity is dependent on the presence of a maturation or "danger" signal, with LPS and IFN- stimulating the greatest response (Fig. 3)
. Both the xenogeneic OVA protein and a hexahistidine-tagged form of SIINFEKL could be associated with SLs for targeting via the engrafted ScFv. CTL assays thus demonstrate that targeting ScFv-engrafted PMV and antigen-bearing SLs to DCs in this way can be effective in stimulating antitumor responses, and highlights the importance of maturation and/or danger signals in the induction of these responses (Fig. 3)
. Moreover, the finding that ScFv-engrafted SL containing SIINFEKL-6H can induce a significant cytotoxic response (Fig. 3)
shows that the approach using NTA3-DTDA-containing SLs may be an effective strategy for targeting any histidine-tagged peptide antigen to DCs in vivo.
A finding of paramount importance in this work was our observation that syngeneic animals immunized with CD11c-ScFv- and DEC-205-ScFv-engrafted PMV had much lower numbers of tumor metastases in their lungs compared with controls, after challenge with the B16-OVA melanoma. Similarly, syngeneic animals immunized with ScFv-engrafted SL containing OVA and either LPS or IFN- had substantially lower numbers of metastases (Fig. 4)
. The results additionally show that tumor immunity was completely dependent on the presence of the maturation or "danger" signals, LPS and IFN- (Figs. 3
and 4
). The immunization of mice with CD11c-ScFv- and DEC-205-ScFv-engrafted PMV and antigen-bearing SL, therefore, target the associated antigen(s) to DCs, which then process and present the antigens to T cells inducing antigen-specific T-cell activation and elicit a strong inhibition in the growth and metastasis of the B16-OVA tumor in vivo. An additionally significant finding was the fact that, unlike control mice, which all developed severe lung metastases, mice that had been vaccinated with DEC-205-ScFv-engrafted PMV containing IFN- after challenge with B16-OVA tumor cells subsequently did not show any signs of tumor development, indicating that the DC targeting vaccine has therapeutic activity. This therapeutic effect has been maintained for at least 8 months, and there is no evidence of tumors in the surviving mice, demonstrating a clear therapeutic potential for the use of the novel DC-targeted vaccine to elicit in vivo antitumor responses.
A particularly intriguing aspect of this study is that the apparent generation of CTL activity against the B16-OVA melanoma was not associated with tumor protection. This point is particularly evident with the SIINFEKL-SL vaccine that would be expected to generate only a CD8+ CTL response against OVA produced by the B16-OVA tumor cells. Despite the vaccine inducing a strong in vitro recall CTL response against B16-OVA tumor cells, no in vivo protection against the tumor was afforded by the immunization. One possibility is that this apparent discrepancy reflects an involvement of other tumor cell-derived antigenic peptides, in the case of the PMV vaccine, affording protection against the tumor. In contrast, the fact that the OVA-SL vaccine also induced a comparable recall CTL response to the tumor (compared with the SIINFEKL-SL preparation) implies that a CD4+ T-cell response may be mediating tumor protection (as OVA also contains CD4+ T-cell epitopes).
It is known that the B16-OVA melanoma line expresses very low levels of MHC class I and, consequently, is resistant to CTL lysis unless high avidity CTLs are used (35) . The fact that T cells from mice immunized with DC targeting preparations of PMV or SL could lyse B16-OVA tumor cells after restimulation with tumor cells in vitro implies that high avidity CTLs can be generated against this tumor cell line. An example of in vivo priming leading to the generation of high frequency and high avidity CTLs that recognize target cells expressing only low levels of class I MHC peptide has already been reported in the literature (44) . However, the present work suggests that such CTLs are either not generated or are not particularly effective in vivo in this system. Consistent with this interpretation, previous studies in fact indicate that CD4+ rather than CD8+ T cells are effective against B16-OVA metastases, with CD4+ T cells with a cytokine profile characteristic of T-helper 2 cells being particularly effective (35) .
To explore a possible role of CD4+ T cell-mediated eosinophil recruitment in the antitumor effects induced by targeting antigen directly to DC in vivo, eotaxin knockout mice were immunized with ScFv-engrafted PMV before inoculating them with the B16-OVA tumor. The results show that compared with controls, eotaxin knockout mice exhibit a markedly reduced ability to inhibit the growth and metastasis of the B16-OVA tumor (Fig. 5)
. Eotaxin is a potent eosinophil chemokine and, therefore, the findings are consistent with the recruitment of eosinophils into the tumor constituting an important component of the antitumor response. Whereas previous studies showed that eotaxin plays a role in the elimination of B16-OVA lung metastases after the adoptive transfer of CD4+ T cells (35)
, the present study shows clearly that the antitumor effect elicited by vaccination with the novel DC-targeted vaccine is at least partially dependent on eotaxin, thereby providing some insight into the mechanism by which the DC-targeted vaccine elicits its antitumor effects. The fact that vaccination offers partial protection in eotaxin knockout mice (Fig. 5)
suggests that other mechanisms, presumably dependent on CTLs, also are involved.
In summary, the modified PMV and SL system described herein offers a number of advantages over current strategies using DCs for tumor immunotherapy. Firstly, the system can deliver antigens directly to DCs in vivo, thus eliminating the need to isolate DCs from patients and to manipulate the cells ex vivo for use in immunotherapies. Secondly, a targeted or active liposome-mediated delivery of antigen to DCs has the potential to deliver more antigen and/or several different antigens, simultaneously, potentially stimulating a more effective immune response. The same approach could potentially deliver to DCs any antigen or immunostimulatory agent, such as "danger" signals, RNA, DNA, and cytokines, or combinations thereof, which cannot be easily achieved using antigens fused to DC targeting proteins (14 , 15) . Thirdly, the approach is versatile and would be convenient to use clinically, because potentially any DC targeting protein(s) possessing a histidine tag can be engrafted onto the modified PMV or SL to deliver specific tumor antigens or other agents to enhance tumor immunity in patients.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Joseph G. Altin, School of Biochemistry and Molecular Biology, Faculty of Science, The Australian National University, Canberra Australian Capital Territory 0200, Australia. Phone: 61-2-6125-4495; Fax: 61-2-6125-0313; E-mail: Joseph.Altin{at}anu.edu.au
6 C. Demangel, J. Zhou, A. B. H. Choo, G. Shoebridge, G. M. Halliday, and W. J. Britton. Selective targeting of protein antigens to dendritic cells using single chain fragments binding DEC-205 and CD11c, manuscript in preparation. 
7 C. L. van Broekhoven and J. G. Altin. Highly stable binding of histidine-tagged proteins to membranes containing the novel chelator lipid 3 (nitrilotriacetic acid)-ditetradecylamine (NTA3-DTDA): an improved strategy for simultaneous delivery of antigen, cytokine and T cell costimulatory signals for development of tumor vaccines and immunotherapies, manuscript in preparation.
Received 1/15/04. Revised 3/ 8/04. Accepted 4/ 2/04.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSTargeting of DC in...Cytotoxicity AssaysImmunization of Animals and...RESULTSDISCUSSIONREFERENCES |
|---|
show improved pharmacokinetics and biological acivity and reduced toxicity in mice. J Immunother, 20: 180-93, 1997.
This article has been cited by other articles:
|
|
 |
|
  S. Gehring, S. H. Gregory, P. Wintermeyer, C. Aloman, and J. R. Wands Generation of Immune Responses against Hepatitis C Virus by Dendritic Cells Containing NS5 Protein-Coated Microparticles Clin. Vaccine Immunol., February 1, 2009; 16(2): 163 - 171. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Tacken, B. Joosten, A. Reddy, D. Wu, A. Eek, P. Laverman, A. Kretz-Rommel, G. J. Adema, R. Torensma, and C. G. Figdor No Advantage of Cell-Penetrating Peptides over Receptor-Specific Antibodies in Targeting Antigen to Human Dendritic Cells for Cross-Presentation J. Immunol., June 1, 2008; 180(11): 7687 - 7696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Adotevi, B. Vingert, L. Freyburger, P. Shrikant, Y.-C. Lone, F. Quintin-Colonna, N. Haicheur, M. Amessou, A. Herbelin, P. Langlade-Demoyen, et al. B Subunit of Shiga Toxin-Based Vaccines Synergize with {alpha}-Galactosylceramide to Break Tolerance against Self Antigen and Elicit Antiviral Immunity J. Immunol., September 1, 2007; 179(5): 3371 - 3379. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Zaks, M. Jordan, A. Guth, K. Sellins, R. Kedl, A. Izzo, C. Bosio, and S. Dow Efficient Immunization and Cross-Priming by Vaccine Adjuvants Containing TLR3 or TLR9 Agonists Complexed to Cationic Liposomes. J. Immunol., June 15, 2006; 176(12): 7335 - 7345. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Rosen, J. Bettadapura, M. Alsharifi, P. A. Mathew, H. S. Warren, and L. L. Lanier Cutting Edge: Lectin-Like Transcript-1 Is a Ligand for the Inhibitory Human NKR-P1A Receptor J. Immunol., December 15, 2005; 175(12): 7796 - 7799. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Warren, A. L. Jones, C. Freeman, J. Bettadapura, and C. R. Parish Evidence That the Cellular Ligand for the Human NK Cell Activation Receptor NKp30 Is Not a Heparan Sulfate Glycosaminoglycan J. Immunol., July 1, 2005; 175(1): 207 - 212. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
