| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
*
Experimental Transplantation and Immunology Branch, National Cancer Institute, Bethesda, MD 20892; and
Science Application International Corp. and
Laboratory of Molecular Immunoregulation, National Cancer Institute, Frederick, MD 21702
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-defensins 2 and 3 bind murine
CCR6, similarly to inflammatory chemokine macrophage-inflammatory
protein 3
, and they chemoattract bone marrow-derived immature, but
not mature DC. Using various chemokines or defensins fused with
nonimmunogenic tumor Ags, we studied their capacity to delivery Ags to
subsets of immune cells to elicit antitumor immunity. We demonstrate
that DNA immunizations with fusion constructs with -defensin 2 or
inflammatory chemokines that target immature DC, but not homeostatic
chemokines secondary lymphoid tissue chemokine, CCL21, or stromal
cell-derived factor 1, CXCL12, which chemoattract mature DC,
elicit humoral, protective, and therapeutic immunity against two
different syngeneic lymphomas. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(MIP-3; EBV-induced molecule 1
ligand chemokine, CCL19 (ELC)) and secondary lymphoid tissue chemokine,
CCL21 (SLC), to recruit mature DC in lymph node (16, 17). Recently, it has been demonstrated that human -defensin
2, an antimicrobial peptide involved in innate immunity against
infection, binds to the chemokine CCR6 (18). Defensins are
proposed to play a role in inflammation, wound repair, and regulation
of the specific immunity by inducing the expression of cytokines and
chemokines, the production of histamine (19), and
enhancing Ab responses (20).
Perhaps a general strategy for induction of effective adaptive immunity
against weakly immunogenic tumor Ags may be to target the delivery of
such Ags to receptors on professional APC by linkage to their
-defensins or proinflammatory chemokine ligands. Recently, we
reported that lymphoma Id or sFv could be rendered immunogenic in mice
by fusion to a chemokine moiety such as monocyte chemoattractant
protein 3 (MCP-3) (21), suggesting that chemokines
could be an efficient carrier for tumor and other Ags by targeting APC
via chemokine receptors. Herein, using various chemokines and
defensins, we demonstrate that protective antitumor immunity depends on
the ability of the vaccine to target immature, but not mature, DC.
Thus, MIP-3 and -defensins render otherwise nonimmunogenic tumor
Ags immunogenic and induce protective and therapeutic antitumor
immunity. In contrast, immunizations with homeostatic chemokines SLC or
SDF1 do not elicit antitumor immunity. Although both humoral and
cellular immune responses are required for rejection of the more
aggressive 38C13 tumor that expresses IgM primarily on its surface
(17, 21), only cellular, but not humoral, immunity, is
protective for slower growing A20 lymphoma, which largely secretes its
idiotypic Ag. Thus, the breadth of this approach as a general strategy
for vaccine development was also made apparent by its ability to elicit
for the first time eradication of established A20 lymphoma.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cloning strategies for lymphoma-specific
VH and VL fragments from
38C13 (22) and A20 (23) cells as sFv38 and
sFv20, respectively, with MCP-3 and IFN-
-inducible protein 10
(IP-10) were described previously (21). Genes for mature
murine -defensins were cloned from LPS (10 ng/ml)-treated BALB/c
mouse skin in-frame to the 5' end of sFv by RT-PCR from total RNA using
specific primers as described previously (21).
The following pairs of primers were used: for -defensin
2 (GenBank accession no. AJ011800), PRmDF2-5'
(ACCATGGAACTTGACCACTGCCACACC) and PRmDF2-3'
(TGAATTCAAGATCTTTCATGTACTTGCAACAGGGGTTGTT); and for -defensin 3
(GenBank accession no. AF092929), PRmDF3-5'
(ACCATGGAAAAAATCAACAATCAGTAAGTTGTTTGAGG) and PRmDF3-3'
(CTCGAGCTAGAATTCTTTTCTCTTGCAGCATTTGAGGAAA). The -pro-defensin 2
gene was cloned for eukaryotic expression using PRproDF2L-5'
(AAAGCTTCCACCATGAGGA CTCTCTGCTCT) and PRmDF2-3', which contained
native secretion signal sequence. SDF-1 (GenBank accession no.
HSU16752) was cloned from 10 ng/ml LPS-treated human monocytes using
PRhSDF1-5' (CTCTAGACACCATGAACGCCAAGGTCGTGGTCGTGCTG) and
PRhSDF1-3' (TGAATTCCATCTTGAACCTCTTGTTTAAAGCTTT). Murine MIP-3
(GenBank accession no. AJ222694) was cloned from a mixture of thymus
and kidney cDNA using PRmMIP3-5'
(ACCATGGCAAGCAACTACGACTGTTGCCTC) and PRmMIP3-3'
(ATAGAATTCCATCTTCTTGACTCTT AGGCTGA). Murine SLC (GenBank accession no.
U88322) was recloned from the plasmid (gift from Dr. Shakhov, Science
Application International Corp.-Frederick) using PRmSLC-5'
(ACCATGGATGGAGGGGGACAGGACTGCT) and PRmSLC-3'
(ATAGAATTCTCCTCTTGAGGGCTGTGTCTGT). All constructs were verified by
the DNA dideoxy sequencing method (Amersham, Arlington Heights, IL) and
purified using a plasmid purification kit (Qiagen, Valencia,
CA).
Recombinant fusion proteins purified as inclusion bodies after 8 h
of induction in SuperBroth (Digene Diagnostics, Beltsville, MD) with
0.8 mM isopropyl -D-thiogalactoside as described
previously (21) and refolded according to Buchner et al.
(24). The refolded fusion proteins were purified by
heparin-Sepharose chromatography (Pharmacia Biotech, Uppsala, Sweden).
The integrity and purity (>90%) of recombinant proteins were tested
by PAGE and Western blot hybridization with 9E10 anti c-myc
mAb (Sigma, St. Louis, MO). Correct folding of purified sFv38 proteins
was determined by the ability to bind to anti-idiotype mAb S1C5 in
ELISA (21). Briefly, serially diluted sFv were added to
microtiter plates coated with 10 µg/ml anti-c-myc mAb
9E10. After washing, plates were incubated with 1/300 diluted
biotin-labeled S1C5, followed by streptavidin-HRP (1/5000; The Jackson
Laboratory, Bar Harbor, ME), and developed with ABTS peroxidase
substrate (Kirkegaard & Perry, Gaithersburg, MD). Proteins were
biotinylated using EZ-Link sulfo-NHS-LC-biotin following the
manufacturer’s protocol (Pierce, Rockford, IL).
Isolation of murine bone marrow-derived DC was described previously
(25). Briefly, bone marrow was collected from tibias and
femurs of 4- to 6-mo-old BALB/c mice. Erythrocytes were lysed with ACK
lysis buffer (BioWhittaker, Walkersville, MD).
CD8+, CD4+,
B220+, and I-Ab cells were
depleted using a mixture of mAbs and rabbit complement. The mAbs were
TIB-146 (anti-B220), TIB-150 (anti-CD8), TIB-207
(anti-CD4), and TIB-229 (anti- I-Ab)
obtained from American Type Culture Collection (Manassas, VA). Cells
were cultured in DC medium (RPMI 1640 containing 5% heat-inactivated
FBS, 1% penicillin, streptomycin, 1% L-glutamine, and
5 x 10-5 2-ME) containing 10 ng/ml each of
murine IL-4 and GM-CSF (PeproTech, Rocky Hill, NJ). Adherent cells were
harvested on days 4 and 7 and used in subsequent experiments. DC were
matured by TNF- (10 ng/ml; BD PharMingen, San Diego, CA) overnight
in DC medium. Immature DC on days 4–5 of culture were in general
CD11+ (69%), B7.2+ and
I-Ab+ (21%), B7.2- and
I-Ab+ (18%), and CD40+
(27%). Upon maturation, the DC were CD11c+
(87%), B7.2+ and I-Ab+
(62%), B7.2- and I-Ab+
(3%), and CD40+ (87%). The following mAb were
used for FACS: CD-11c-APC, MAC3-PE, Gr-1-FITC, B220-PE, Thy 1.2-FITC,
I-Ab-FITC, B7.2-PE, and CD40-PE (BD
PharMingen).
In vitro chemotaxis assay
The migration of DC (50 µl, 106
cells/ml) was assessed using a 48-well microchemotaxis chamber
(NeuroProbe, Cabin John, MA) with a 5-µm polycarbonate filter
(Osmonics, Livermore, CA) as previously described (26, 27). Cells were incubated at 37°C in humidified air with 5%
CO2 for 1.5 h. DC migrating across the
filter were counted using a Bioquant semiautomatic counting
system (Bioquant, Nashville, TN). The results (mean ± SE
of triplicate samples) are presented as the chemotactic index, defined
as the fold increase in the number of migrating cells in the presence
of test factors over the spontaneous cell migration (in the absence of
test factors). Human MIP-3 and MIP-3 were obtained from
PeproTech.
Cell lines and mice
The carcinogen-induced, C3H 38C-13 B cell lymphoma
(22) was a gift from Dr. R. Levy (Stanford, CA). The
38C-13 tumor secretes and expresses IgM
on the cell surface. The
BALB/c A20 lymphoma (23) was obtained from the American
Type Culture Collection and expresses IgG. Murine CCR6-expressing
HEK293 cells (HEK293/CCR6) were a gift from Dr. J. Farber (National
Institute of Allergy and Infectious Disease, National Institutes of
Health, Bethesda, MD).
Flow cytometric analysis
Trypsinized HEK293/CCR6 or HEK293 cells (2.5 x
106 cells/ml) were incubated with 20 µg/ml
biotinylated mMIP3sFv38 for 30 min on ice in PBS with 2% BSA
(PBS/BSA) and 20% mouse serum. Cells were stained on ice for 30 min
with 0.2 µg/ml streptavidin-PE (BD PharMingen) and fixed with 1%
paraformaldehyde.
In vivo immunizations and tumor protection experiment
Animal care was provided in accordance with the procedures outlined in a Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication 86-23, 1985). Six- to 12-wk-old female C3H/HeNCrlBR or BALB/c mice (Charles River Laboratories, Frederick, MD) were used. Mice (10/group) were immunized with the Helios Gene Gun System (Bio-Rad, Hercules, CA) with 1–2 µg of plasmid DNA three times per 2 wk as previously described (21). Two weeks after the last immunization, mice were challenged i.p. with 2 x 103 or 2.5 x 105 38C13 or A20 lymphoma cells, respectively, and followed for survival. Differences in survival between groups were determined by nonparametric log-rank test (BMDP Statistical software, Los Angeles, CA). The p values refer to comparisons with the group immunized with DNA expressing the same chemokine or defensin fused with an irrelevant sFv or sFv fused with mutant chemokine, unless otherwise specified.
Therapy of established tumor with DNA vaccine
Six- to 12-wk-old female BALB/c mice (10/group) were challenged
with 2.5 x 105 syngeneic A20 tumor cells.
On days 1, 4, 8, and 18, these mice were gene gun immunized with DNA
plasmid (containing 
1–2 µg of DNA/immunization), and mice were
followed for tumor progression.
Adoptive transfer experiments
BALB/c mice were gene gun immunized with 1–2 µg of pMCP3sFv20 twice biweekly, and splenocytes and sera were removed 10 days after the last immunization. Ten BALB/c mice per group were i.p. injected in saline with 2.5 x 105 A20 tumor cells/mouse mixed with 2 x 107 splenocytes or sera from immune or mock-treated mice, and mice were followed for tumor progression.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-defensins and chemokines retain their functional
integrity as fusion proteins with sFv and chemoattract immature, but
not mature, DC
First, we cloned and purified a variety of chemokine
and -defensin fusion proteins with sFv, a lymphoma Ig-derived
nonimmunogenic Fv from the two different B cell lymphomas, 38C-13 and
A20 (Table I
). For example, sFv fusion
proteins with murine inflammatory chemokine MIP-3, murine
-defensin 2, and -defensin 3 were designated mMIP3sFv38,
mDF2sFv38, and mDF3sFv38 or mMIP3sFv20, mDF2sFv20, and
mDF3sFv20, respectively. Similarly, sFv fusion proteins with
homeostatic chemokines SLC and SDF1 were designated
mSLCsFv38 and SDF1sFv38, respectively. Control proteins contained
either sFv fusions with mutant chemokine, generated by replacing the
first Cys residue with Ser or by truncation of the amino termini to
abrogate a respective receptor-binding, or an inactive form
pro-defensin 2, -defensin 2, with its pro-sequence (mproDF2sFv38;
Table I). All fusion proteins had comparable idiotype folding, as
tested by inhibition ELISA with anti-Id mAbs, which bind only
properly folded parental lymphoma Id (data not shown).
|
fusion protein (Fig. 1A). Furthermore, as expected,
chemokine fusion proteins, but not control mutant fusion proteins,
induced chemotaxis of THP-1 cells (data not shown) or murine DC (Fig. 1, B–D). Since human -defensin 2 was reported to
interact with CCR6, we tested murine -defensin fusion proteins for
their ability to induce chemotaxis of murine CCR6-transfected cells,
HEK293/CCR6. These cells were chemoattracted by murine MIP-3 (a
ligand of CCR6) fusion protein (mMIP3sFv38; Fig. 1B).
Both fusion proteins -defensins 2 and 3, but not fusion proteins
-defensin 2 containing pro-sequence (mproDF22sFv38) or control
viral chemokine (vMIP1 MsFv38), chemoattracted murine CCR6-expressing
cells in a dose-dependent manner (Fig. 1B). A control
parental cell line, HEK293, which does not express CCR6 was not
attracted to these proteins (data not shown). Next, we tested the
ability of these proteins to attract murine bone marrow-derived DC,
which are known to express CCR6 (13, 14). The purity of
murine bone marrow-derived DC cells and their immature phenotype on day
4 of cultivation was confirmed by expressions of
CD11c+ and low levels of I-A, B7.2, and CD40,
respectively (data not shown, see Materials and
Methods). Similarly to MIP-3 (data not shown), both
-defensin fusion proteins induced chemotaxis of immature DC in a
dose-dependent manner, with peak activity at 10 and 100 ng/ml for
mDF2sFv38 (Fig. 1C) and mDF3sFv38 (data not shown),
respectively. The predominant immature phenotype of these DC was also
supported by their ability to migrate to human MIP-3, a
chemoattractant specific for CCR6+ immature DC
(16, 28), and their limited ability to react to human
MIP-3 (13), a chemoattractant specific for
CCR7+ mature DC (Fig. 1C). In
contrast, none of the defensin and MIP-3 fusion proteins stimulated
chemotaxis of TNF-induced mature DC, which migrated to MIP-3 (Fig. 1D). Furthermore, control fusion protein with mutant
MIP-3 (data not shown) or mproDF2sFv38 did not induce chemotaxis
of any DC (Fig. 1, B–D). Therefore, murine -defensin 2
and 3 fusion proteins can specifically target iDC and induce their
CCR6-mediated chemotaxis, similarly to MIP-3.
|
-defensin or inflammatory chemokine fusion
constructs on capacity of nonimmunogenic tumor Ags to induce humoral
immunity
The lymphoma idiotype alone is nonimmunogenic in syngeneic mice.
Similarly, DNA immunizations with lymphoma-derived Fv or sFv alone,
particularly from 38C13 and A20 lymphomas, do not induce
immunity in syngeneic mice (21). Therefore, we tested
whether a fusion construct of inflammatory peptides, such as MIP-3
or -defensins, with these sFv Ags would induce specific immunity
when administered as a DNA vaccine in mice. Ten mice per group were
immunized by gene gun with plasmids encoding fusion proteins consisting
of mature forms of -defensins (pmDF2sFv38 and pmDF3sFv38,
respectively) or MIP-3 (pmMIP3sFv38). Control mice were immunized
with DNA constructs encoding sFv fused with inactive pro-defensin
(pproDF2sFv38) or mutated and inactive chemokines (pvMC148MsFv38).
Furthermore, as will be discussed below, no Ab was
generated when mice were immunized with 2 µg DNA expressing a mixture
of plasmids containing unlinked sFv and murine -defensin
(pmDF3Muc1 and sFv38; Fig. 2A) or chemokine (data not
shown). In contrast, mice immunized with plasmids encoding sFv fusion
proteins with murine -defensins (Fig. 2A), MIP-3, or
SLC (Fig. 2B) produced idiotype-specific Ab levels
comparable to those induced by vaccination with tumor-derived intact Ig
protein conjugated to keyhole limpet hemocyanin (KLH; Fig. 2). However,
control mice immunized with an inactive pro--defensin
(pproDef2sFv38) or mutant chemokine sFv fusion constructs
(pvMC148MsFv38; Fig. 2B) or with sFv38 alone (data not
shown) did not produce any anti-Id Ab responses. Interestingly, the
two types of -defensins differed in their capacity to elicit Ab
responses; although both produced predominantly specific IgG1 Abs,
-defensin 3 was superior to -defensin 2 for induction of specific
Abs to every Ag tested (Fig. 2).
|
-Defensin fusion vaccines elicit protective and therapeutic
antitumor immunity
The vaccine protocol used to elicit protective antitumor immunity
was as follows. First, mice were immunized with 2 µg of DNA
constructs three times at biweekly intervals, then 2 wk after the last
immunization mice were challenged with a 20-fold lethal dose of
syngeneic tumor. No survival was observed in control groups immunized
with PBS or plasmids encoding sFv38 fused with inactive
pro--defensin-2 (pproDef2sFv38; Fig. 3A), with irrelevant chemokine
plasmid vaccines pMDC-EGFP (Fig. 3A), or with mutant
constructs pMC148MsFv38 (Fig. 3B). In contrast, significant
protective immunity was elicited in mice immunized with
pmDF2sFv38 (Fig. 3A) or pmDF3sFv38 (Fig. 3B; by log-rank test, p < 0.001 and 0.004
compared with pproDF2sFv38 and pMC148MsFv38, respectively). The
protection elicited with both -defensin constructs was comparable to
that induced by Ig-KLH protein vaccine (Fig. 3A), a
prototype vaccine that consists of lymphoma-derived IgM cross-linked
with KLH that is being successfully tested in a Phase III clinical
trial (1). Similarly, as shown in Fig. 3B, DNA
immunizations with MIP-3 fusions elicited potent tumor protection
(pmMIP3sFv38, by log-rank test, p < 0.0001 compared
with pMC148MsFv38). In contrast, none of the mice immunized with
constructs encoding sFv fusion with homeostatic chemokine murine SLC
were protected (Fig. 3B; pmSLCsFv38; by log-rank test,
p < 0.02 compared with pmMIP3sFv38) despite the
fact that this SLC fusion construct generated anti-Id-specific Abs
comparable to the pmMIP3sFv38-vaccinated group (Fig. 2
B).
Moreover, no immunity was detected in mice immunized with constructs
expressing human SDF1, which binds to murine CXCR4 (data not shown).
Therefore, these data suggest that fusion constructs with inflammatory
mediators, such as -defensins or MIP-3, which target immature DC,
can render a nonimmunogenic tumor Ag (sFv) immunogenic and elicit
protective antitumor immunity even for a very aggressive lymphoma,
38C13, which kills all control mice within 20 days postchallenge.
|
, A and B), IP-10
(pmIP10sFv20, Fig. 4A), and MIP-3 (pmMIP3sFv20; Fig. 4B). The mice were immunized with three booster vaccinations
at 4, 8, and 18 days posttumor challenge. Tumor-bearing mice randomized
to control treatment with the same chemokine, but fused with sFv
derived from the 38C-13 lymphoma (pMCP3sFv38, Fig. 4, A and
B), sFv20 alone (data not shown), or A20 lymphoma-derived
Ig-KLH protein (data not shown) did not survive. In contrast,
significant survival was observed in pMCP3sFv20 (by log-rank test,
p < 0.01 compared with pMCP3sFv38; Fig. 4, A and B) or pmMIP3sFv20 (by log-rank test,
p < 0.0003 compared with the mock group; Fig. 4B)-treated mice. Furthermore, angiostatic or other
nonspecific effects of functional chemokines, as potentially induced by
IP-10, were apparently not sufficient for tumor eradication (Fig. 4A). No mice survived in the pmIP10sFv20 immunization group
(Fig. 4A), which is in concordance with lack of CXCR3
expression on skin iDC. Therefore, these data demonstrate that
nonimmunogenic tumor Ag fused with inflammatory chemokines MIP-3 and
MCP-3, but not IP-10, which bind receptors expressed on iDC (13, 14), induce potent therapeutic antitumor immunity.
|
-defensin vaccines.
Tumor-bearing mice (10 mice/group) were immunized with 2 µg of DNA
vaccine expressing -defensin 2 and 3 fusions with sFv20
(pmMIP3sFv20, pmDF2sFv20, and pmDF3sFv20) starting 1 day after
challenge with a lethal dose of A20 tumor, followed by three booster
vaccinations. There were no survivors among tumor-bearing mice
randomized to control treatment with the same -defensin but fused
with sFv derived from the 38C-13 lymphoma (pmDF2sFv38; Fig. 4B), suggesting that nonspecific effects of -defensins
were apparently not sufficient for tumor eradication. In contrast, a
significant number of surviving mice were observed in the pmDF2sFv20
treatment group, as shown by the representative experiment in Fig. 4B (by log-rank test, p < 0.002 compared
with pmDF2sFv38). It is notable that in this model, although the
-defensin 3 fusion vaccine also induced a superior specific Ab
production (data not shown), the vaccine was not able to elicit
antitumor immunity (Fig. 4B). Superior Ab production also
was generated by vaccinating with A20 lymphoma derived Ig-KLH protein
(21), although it was not sufficient to elicit both
protective (21) and therapeutic immunity in A20 model.
Thus, these data suggest that induction of humoral immunity is not
sufficient to eradicate A20 B cell lymphoma, and that fusions with
inflammatory chemokines MCP-3 and MIP-3 or -defensin 2 induced
specific cellular antitumor responses.
Previously we reported that antitumor immunity in the 38C13 lymphoma
model, induced with MCP-3 fusion constructs, required effector
CD8+ T cells (21). To further
elucidate the basis for the antitumor effects, 10 naive BALB/c
mice/group were subjected to adoptive transfer of a lethal dose of A20
tumor mixed with sera or splenocytes from immune mice. No protection
was observed in mice injected with A20 tumor mixed with immune serum
(data not shown). However, the mice that received splenocytes from mice
immunized with pMCP3sFv20, but not from mock-treated mice, exhibited
significant prolongation of survival (by log-rank test,
p < 0.01; Fig. 4C). Although we and others
have not yet been able to demonstrate Id-specific cellular immunity in
A20 or 38C13 tumor models, these data suggest that DNA vaccinations
with lymphoma Id-derived sFv chemokine or defensin induce cellular
immunity that can be transferred with immune cells.
Requirement for chemokine receptor targeting with fusion constructs
Next we tested whether -defensin needs to be physically linked
with sFv by immunizing with a mixture of unlinked -defensin and sFv.
Ten mice per group immunized with a mixture of separate plasmids
encoding -defensin 3 (pmDF3Muc1T) and sFv Ag (sFv38) failed to
elicit a specific humoral response, thus demonstrating a requirement
for sFv to be physically linked to -defensin (Fig. 2A).
Furthermore, these mice challenged with a lethal dose of syngeneic
38C13 tumor exhibited no protection (data not shown). These data also
suggest that chemokine receptor engagement with chemokine- or
defensin-sFv fusion is essential for the induction of
immunity. It was not sufficient to simply attract APC or induce
inflammation at the site of production of sFv Ag; direct APC targeting
with Ag fused to -defensin or chemokine was required, presumably via
the involvement of chemotactic receptors. To further test
this idea, we tried to inhibit this receptor-mediated immunity by
injection of the competing ligand. C3H mice (10/group) were immunized
with either pmDF3sFv38 alone or mixed with DNA encoding -defensin
fused with irrelevant Ag (pmDF3Muc1T; Fig. 5A). Sera from mice immunized
with a plasmid encoding sFv protein fused with murine -defensin 3 by
itself or in the presence of an irrelevant plasmid contained 300
µg/ml idiotype-specific Abs, on the average, which was 2- to 3-fold
higher than the levels induced by vaccination with tumor-derived intact
Ig protein conjugated to KLH. However, much lower levels (5–15
µg/ml) of specific Abs were detected in sera of mice coimmunized with
pmDF3sFv38 mixed with competing pmDF3Muc1T. Two weeks after the
last immunization all mice were challenged with a 20-fold lethal
dose of syngeneic 38C13 tumor. No survival was observed in control
groups immunized with PBS or plasmid encoding -defensin 3 fused with
an irrelevant Ag (pmDF3Muc1T; Fig. 5A). Similarly, no
protection was detected in mice coimmunized with pmDF3sFv38 and
competing pmDF3Muc1T (pmDF3Muc1T/pmDF3sFv38; Fig. 5A). In contrast, 40% of mice immunized with pmDF3sFv38
were protected (by log-rank test, p < 0.001 compared
with pmDF3Muc1T; Fig. 5A). Therefore, these data support
the view that immunity to nonimmunogenic tumor Ags fused with defensins
or chemokines depended on their ability to engage chemokine
receptor(s).
|
B). In addition, no
protection was detected in mice immunized with DNA expressing a
secretable, but mutated, MCP-3 fusion protein, which could not bind the
respective receptor (pMCP3MsFv38; Fig. 5B).
The inability of the MCP3MsFv38 construct to bind chemokine receptor
was supported by its inability to chemoattract THP-1 cells in vitro
(Fig. 5C). These data are consistent with the hypothesis
that immunity was induced primarily by APC which took up (via chemokine
receptor) a functionally active chemokine fusion protein that was
produced and secreted by transduced bystander or APC cells rather than
by APC directly being transduced by gene gun immunization. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-defensins 2 and 3, or
inflammatory chemokines MIP-3 and MCP-3, which target receptors
preferentially expressed on iDC (16, 30). This immunity is
correlated with the ability of murine -defensins 2 and 3 or
inflammatory chemokine MIP-3 to induce chemotaxis of murine bone
marrow-derived immature, but not mature, DC. Human -defensin 2 binds
to CCR6 preferentially expressed on iDC and resting memory T cells
(18, 28), but the murine -defensin receptor(s) has not
been identified to date. Herein we demonstrate for the first time that
both murine -defensins 2 and 3 also bind to CCR6 (Fig. 1), although
the possibility that -defensins bind to other receptors is not
excluded. The involvement of additional receptors may explain the
differences in immune responses induced by -defensins 2 and 3. For
example, -defensin 3 fusion constructs generated higher Ab titers in
both tumor models, but, unlike -defensin 2, did not elicit antitumor
immunity in the A20 model. The importance of targeting immature DC
residing in the skin (a site of Ag production) is also supported by
data obtained from mice immunized with another inflammatory chemokine,
IP-10, which binds to CXCR3. This receptor is not detected in skin
cells, such as epidermal DC and keratinocytes, although it is expressed
in normal human lymphoid organs by certain subsets of DC, specifically
myeloid-derived CD11c-positive cells, and by peripheral B cells
(32, 33). We previously reported that substantial
infiltration of mononuclear cells and polymorphonuclear leukocytes was
detected at the site of injection of IP-10 or MCP-3 fusion proteins
(21). However, mice immunized with IP-10 fusion constructs
also failed to elicit efficient antitumor immunity in the A20 tumor
model despite the generation of humoral immunity. Furthermore, no
antitumor immunity was elicited when we used the homeostatic chemokines
SLC and SDF1, which bind to receptors expressed on mature DC. The
lack of antitumor immunity in mice immunized with homeostatic chemokine
fusion constructs is not due to inadequate expression of the fusion
genes, since both constructs were comparably expressed by transient
transfection assay in HEK293 cells (data not shown). The integrity of
the chemokine was also not impaired, since SDF fusion protein competed
specifically with native labeled SDF1 for binding to CXCR4 (data not
shown). Moreover, mice immunized with the SLC fusion construct
generated Id-specific Ab comparable with MIP-3 fusion vaccine (Fig. 2B). This indicates that CCR7-expressing APC have the
capacity to take, process, and present nonimmunogenic Ag sufficiently
well to induce a humoral immune response. Even though SDF1 fusion
proteins can also interact with CXCR4 on immature DC, whether their
availability might be reduced because CXCR4 is expressed by many
different cells or by the presence of competing native SDF1 needs
further studies.
Taken together, the appeal of using fusion constructs with defensin or
chemokine ligands was based not only on the ability of these mediators
of innate and adaptive immunity to target surface receptors on APC,
particularly on iDC (18), presumably resulting in
increased uptake of Ag, but possibly also to induce the expression of
costimulatory molecules and, in turn, the production of other
proinflammatory cytokines and factor. The importance of coexpression of
genes for costimulatory molecules B-7.1 and CD-40 for induction of
immune responses against weak immunogens has been repeatedly reported
by others (34, 35, 36, 37). Human -defensins induce the
expression of costimulatory molecules on lymphocytes (20)
and the production of proinflammatory cytokines (38). The
potential effects of inflammatory chemokine or -defensin fusions on
the induction of local inflammation in vivo need to be elucidated. It
is tempting to speculate that they may target not only Ag to a
receptor(s) on APC, but also preferentially prime for Th1 or Th2
cytokine responses in vivo. Moreover, Th1 or Th2 cells could be
differentially attracted by chemokines, thus modulating immunity. For
example, MCP-1 stimulates IL-4 production (39) and thereby
induces control Th2 polarization (40). Moreover,
macrophage-derived chemokine, CCL22 (MDC), has been reported to
selectively chemoattract Th2 cells toward APC (41). In
support of this view, we also observed in A20 tumor models that
nonimmunogenic tumor Ags fused with MDC induced inefficient therapeutic
antitumor immunity despite inducing the production of high levels of
specific Ab (data not shown). Therefore, these data suggest that
nonimmunogenic tumor Ags are rendered immunogenic by linking them as
fusion constructs to proinflammatory ligands to target immature DC. The
identity of chemokine or defensin moieties determines whether humoral
and/or cellular immunity are induced.
The fusion constructs were delivered by cutaneous DNA bombardment and
therefore could have been expressed directly in LC and other APC
residing in the skin (34, 42). However, experiments with
constructs designed to either secrete or express fusion proteins in
intracellular compartments clearly demonstrated that secretion was
required, although we cannot completely rule out a role of directly
transduced APC. Moreover, the mechanism of targeting of APC by secreted
fusion proteins was supported by evidence for a receptor-dependent
process, since idiotype-specific Ab and protective antitumor immunity
were elicited only in mice immunized with DNA encoding fusion
constructs with functionally active murine -defensins or
proinflammatory chemokines. Immunizations with fusion constructs
encoding mutant chemokine or an inactive -pro-defensin moiety, which
could not bind to the respective chemokine receptor and induce
chemotaxis, failed to elicit any immune response. In addition,
induction of chemotaxis alone to the site of Ag production by unfused
mixtures of ligands and Ag was not sufficient to induce
nonresponsiveness to tumor Ag, probably due to inefficient uptake of Ag
by infiltrating cells. Specifically, no immune responses were elicited
in mice immunized with a combination of DNA expressing free,
unlinked murine -defensin or chemokine and either DNA encoding sFv
Ag (Fig. 2) or sFv protein alone (data not shown). Only sFv physically
linked with the defensin/chemokine moiety elicited efficient antitumor
immunity.
The ability to efficiently induce both types of adaptive immune
responses, humoral and cellular, may explain the effect on established
lymphoma in mice immunized by DNA encoding inflammatory chemokine or
-defensin 2 fusion proteins (Figs. 3 and 4). It is noteworthy that
the use of these fusion constructs yielded superior protective immunity
than intact Ig protein conjugated to carrier (KLH), the prototype
vaccine currently in Phase III clinical testing against human lymphoma
(1). In particular, immunizations with MIP-3 and MCP-3
or -defensin 2 fusion constructs elicited protective and therapeutic
antitumor immunity against the A20 lymphoma, a model in which Ig-KLH
and other idiotypic vaccine formulations previously failed
(21). In contrast with the A20 model, generation of
Id-specific Ab contributes to the protection against 38C13 lymphoma
(3) (21). As we reported previously, tumor
protection was abrogated completely by depletion of
CD8+ effector cells in mice vaccinated with
pMCP3sFv38 (21). Although we and others have not yet been
able to demonstrate Id-specific cellular immunity in A20 or 38C13 tumor
models using a conventional CTL assay, the Id-specific cellular
immunity can be transferred by adoptive transfer of immune splenocytes
(Fig. 4C), but not sera (data not shown). In agreement with
this, we recently demonstrated that mice immunized with MCP-3 or
-defensin 2, but not -defensin 3 or MDC, fusion constructs
induced significant CD8+ CTL (both systemic and
mucosal) against Env protein of HIV-1, in addition to high levels of
Abs (A. Biragyn, I. M. Belyakov, Y. H. Chow, T. C. VanCott, D.
C. Dimitrov, J. A. Berzofsky, and L. W. Kwak, manuscript in
preparation). These results are also consistent with our previous
observation that protective antilymphoma immunity elicited by MCP-3 sFv
fusion vaccines required effector CD8+ T cells
(21). These data suggest that DNA vaccinations with
lymphoma Id-derived sFv (nonimmunogenic tumor Ag) fused with
inflammatory chemokines MIP-3 and MCP-3 or -defensin 2, which
target immature DC, induce both humoral and cellular immunities.
Overall, our data suggest that the use of these ligands linked to tumor
Ags could provide a general strategy for the development of simpler and
more effective vaccine formulations for human use.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 On leave of absence from the Divisione di Oncologia Medica Falck, Ospedale Niguarda Ca Granda, Milan, Italy.
3 Abbreviations used in this paper: sFv, single-chain Fv fragment; DC, dendritic cell; iDC, immature dendritic cell; IP-10, IFN--inducible protein 10; KLH, keyhole limpet hemocyanin; LC, Langerhans cell; MCP, monocyte chemoattractant protein; MIP, macrophage-inflammatory protein; ELC, EBV-induced molecule 1 ligand chemokine, CCL19; SLC, secondary lymphoid tissue chemokine, CCL21; MDC, macrophage-derived chemokine, CCL22.
Received for publication August 2, 2001. Accepted for publication September 19, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
augments antitumor immune responses induced by protein and DNA vaccines. J. Immunol. 157:5503.[Abstract]
is expressed at inflamed epithelial surfaces and is the most potent chemokine known in attracting Langerhans cell precursors. J. Exp. Med. 192:705.-Defensins: linking innate and adaptive immunity through dendritic and T cell CCR6. Science 286:525.1 express functional C-C chemokine receptor 6. J. Immunol. 163:1737.This article has been cited by other articles:
|
|
 |
|
  D. J. Betting, K. Kafi, A. Abdollahi-Fard, S. A. Hurvitz, and J. M. Timmerman Sulfhydryl-Based Tumor Antigen-Carrier Protein Conjugates Stimulate Superior Antitumo |
