| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,,,¶,
Divisions of
* Immunotherapy and
Gene Therapy, Departments of
Medical Oncology and
Pathology, Vrije Universiteit University Medical Center, Amsterdam, The Netherlands; and
¶ Division of Human Gene Therapy, Gene Therapy Center, Departments of Medicine, Pathology, and Surgery, University of Alabama, Birmingham, AL 35294
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DC infected by TAA-encoding adenoviral (Ad) vectors have been used successfully to generate antitumor T cell responses (2, 16). Although Ad vectors exhibit a broad tropism, retargeting of Ad vectors, both in vitro and in vivo, has led to the selective infection of specific cell types (17). This has successfully been achieved through the use of bispecific Ab conjugates, which block binding of Ad to its natural receptor (the coxsackie-adenovirus receptor (CAR)) and redirect Ad to alternative receptors (18, 19, 20). We previously reported an enhanced in vitro transduction efficiency of monocyte-derived DC (MoDC) by retargeting Ad, through a bispecific Ab conjugate, to CD40 on the cell surface (21). Moreover, a simultaneous CD40-dependent maturation of the transduced DC was observed, as evidenced by an increased expression of CD83, MHC, and costimulatory molecules, by the production of IL-12, and by an increased T cell stimulatory capacity in an allogeneic MLR (21). The DC-activating properties intrinsic to the employed CD40-targeted Ad vector, together with the reduced need to administer high titers of Ad to obtain a given expression level of the gene of interest in the DC, make this an attractive in vivo gene delivery system for immunotherapy. In the present study, we have determined the utility of this method for the in situ gene transfer to DC through intradermal (i.d.) injection of the CD40-targeted Ad vectors in a human skin explant model. Ex vivo skin explant cultures were previously used as a model system to study migrational patterns of skin-derived DC and shown to be representative of active migration of dermal and epidermal (i.e., Langerhans cells) DC through lymph vessels to the skin-draining LN (22, 23). Because vaccine administration through the skin affords excellent tumor protection (6, 24), we have selected this human model system to demonstrate targeted and enhanced transfer of Ad vectors to migrating DC, as well as their increased maturation and Ag-specific CTL-stimulatory ability, through a CD40-targeting Ab complex.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The culture medium used was IMDM (Life Technologies Laboratories, Paisley, Scotland) supplemented with 50 U/ml penicillin-streptomycin, 1.6 mM L-glutamine, 0.01 mM 2-ME, and 5% pooled, complement-inactivated, normal human pooled serum (HPS; Centraal Laboratorium van de Bloedtransfusiedienst, Amsterdam, The Netherlands). GM-CSF (Schering-Plough, Madison, NJ) and IL-4 (Centraal Laboratorium van de Bloedtransfusiedienst), diluted in serum-free IMDM without supplements (i.e., plain medium), were used for i.d. injection. Human skin specimens were obtained after informed consent from patients undergoing corrective breast or abdominal plastic surgery. Cytokines and/or Ad vectors were injected into the dermis with a MicroFine insulin syringe (0.33 (29G) x 12.7-mm needle; BD Biosciences, Franklin Lakes, NJ) in the indicated amounts and in a total volume of 20 µl. At the site of injection, a 5-mm urtica appeared and an exact punch biopsy of 6 mm was taken. The biopsy was lifted from the specimen with a forceps, and with scissors the dermis was cut at a depth of 2–3 mm to obtain skin explants. For immunohistochemical analysis, skin explants were cultured at air-medium interface with the epidermal side up in a six-well culture plate (Costar, Corning, NY) on sterilized stainless steel grids covered with a filter (Millipore, Bedford, MA; 45 µm), at 37°C in 5% CO2-containing humidified air. At the indicated time points, the explants were harvested, snap frozen, and stored in liquid nitrogen.
For the migration assay, the skin explants (12–20 samples/condition) were placed directly in 1 ml culture medium (floating with epidermal side up) in a 48-well culture plate (Costar). At the indicated time points, the explants were discarded, and the medium, containing migrated cells, was harvested and pooled per test condition. Absolute numbers of migrated DC were counted in hemocytometers using trypan blue exclusion, or with Flow-Count fluorospheres (Coulter, Miami, FL), following the manufacterer’s instructions, and recalculated to migrated DC number per skin explant. FACS analyses, MLR, or CTL activation experiments were subsequently performed.
Adenoviruses and conjugates
The used replication-defective recombinant type 5 adenoviruses
(with deletion of the E1 and E3 regions), encoding 
-galactosidase
(-gal) with a nuclear localization sequence (Ad-lacZ) or
the green fluorescent protein (Ad-GFP), both under the control of CMV
promoters, were produced, purified, and titered, as previously
described (20, 21). An E1-deleted Ad5 virus encoding the
Haeminfluenza matrix protein M1 (Ad-M1) was kindly
provided by C. Rickards and G. W. Wilkinson (University of Wales,
Cardiff, U.K.). The anti-CD40/anti-Ad (Fab anti-CD40)- and
anti-epidermal growth factor receptor (EGFR)/anti-Ad (Fab
anti-EGFR)-bispecific Ab conjugates, consisting of IgG1 mAbs targeting
CD40 or the EGFR and Fab binding the Ad fiber knob, were also prepared,
as previously described (21). To determine the amount of
Fab anti-CD40 needed per given number of viruses for the optimal
transduction of DC, the retargeting conjugate was titrated with Ad-GFP
at a multiplicity of infection of 100 and used to infect MoDC, as
described previously (21). Thus, 108
PFU of Ad was added to 833 ng Fab anti-CD40 and incubated at room
temperature for 30 min before injection into the skin explants or
addition to MoDC. Equivalent and effective amounts of the Fab
anti-EGFR conjugate were similarly determined with
EGFR+ target cells and used as a control for
Fc-mediated targeting (21).
Ad infection
Ad vectors were preincubated with appropriate amounts of the Fab anti-CD40 or Fab anti-EGFR for 30 min at room temperature in serum-free plain medium. Skin explants, previously injected with cytokines or plain medium and cultured with the epidermal side up on filter-covered grids at the air/medium interface over a period of 24 h, were subsequently injected with the indicated amounts of Ad or plain medium in a total volume of 10 µl using a MicroFine insulin syringe (0.33 (29G) x 12.7-mm needle; BD Biosciences). For immunohistochemical analysis, the explants were further cultured on filter-covered grids. Alternatively, the explants were placed in 1 ml culture medium (floating with epidermal side up) in a 48-well culture plate (Costar) and further cultured as indicated, before migrated DC were harvested.
Immunohistochemistry
Cryostat sections of 4 µm were cut and placed on poly(L-lysine)-coated slides, acetone fixed for 10 min, preincubated with normal rabbit serum (1:20; CLB, Amsterdam, The Netherlands) for 10 min, and incubated for 1 h with primary mAbs directed against CD83 (1:20; Immunotech, Marseille, France), CD1a (1:20; Immunotech), CD40 (1:100; Serotec, Oxford, U.K.), or with appropriate isotype control Abs. Subsequent incubation with rabbit anti-mouse biotin conjugate (1:150; DAKO, Glostrub, Denmark) for 30 min was followed by incubation with HRP-streptavidin complexes (1:500; DAKO). Staining was then visualized with 3-amino-9-ethyl-carbazol (ICN Biochemicals, Aurora, Ohio) in the presence of hydrogen peroxide. Slides were counterstained with hematoxylin and mounted.
For the enzymatic staining of cells transduced with the lacZ
gene encoding for -gal, slides were incubated with -gal staining
solution (Boehringer Mannheim, Mannheim, Germany) for 12–72 h
at 37°C after fixation with 2% formaldehyde and 0.2% glutaraldehyde
in PBS for 15 min. For phenotypic analysis of -gal-positive cells,
cryostat sections were incubated for 10 min with 10% HPS, followed by
CD1a-PE (Immunotech) or IgG1-PE isotype (BD Biosciences) for 60 min.
Subsequently, two investigators independently counted stained cells in
10 x400 magnification fields using a fluorescence microscope.
Flow cytometry
Cells were incubated on ice for 30 min in PBS with 0.1% BSA and 0.01% NaN3, in the presence of appropriate dilutions of FITC- or PE-labeled mouse mAbs to CD83, CD11c (Immunotech), CD1a, CD83, CD86, CD54, CD40, HLA-ABC (BD PharMingen, San Diego, CA), CD14, CD80, and HLA-DR (BD Biosciences), or with unconjugated mAbs to CMRF-44 (a kind gift of D. Hart, Brisbane, Australia), to Langerin (Immunotech), or to CCR7 (BD PharMingen). A second incubation step was performed for the unconjugated mAbs with FITC-labeled goat anti-mouse Abs (Centraal Laboratorium van de Bloedtransfusiedienst). The cells were subsequently analyzed, using a FACStarPlus and CellQuest FACS analysis software (BD Biosciences).
Mixed lymphocyte reaction
MLR was performed with the migrated skin DC. DC were added as stimulator cells to round-bottom 96-well tissue culture plates (Costar) at graded doses, reflecting the indicated responder-stimulator ratios. Allogeneic plastic nonadherent PBL, prepared as previously described (21), were used as a source of responder cells, and 1 x 105 lymphocytes/well were added to the migrated skin DC. Stimulations were performed in duplicate or triplicate. The cells were cultured for 5 days in medium containing 10% FCS (Life Technologies). During the last 18 h of culture, [3H]thymidine was added (0.4 µCi/well; Amersham, Aylesbury, U.K.), after which the cells were harvested onto fiberglass filters and [3H]thymidine incorporation was determined using a flatbed scintillation counter.
CTL activation and the IFN-
ELISPOT assay
Autologous heparinized blood (40 ml) was taken from skin donors after informed consent, and PBMC were isolated as previously described (21). From plastic-nonadherent PBMC (i.e., monocyte depleted), CD8+ CTL were isolated using MACS magnetic bead separation, according to the manufacturer’s protocol (CD8+ T cell isolation kit; Miltenyi Biotec, Bergisch Gladbach, Germany). Migrated cutaneous DC, either mock infected (through i.d. medium injection) or in situ Ad-M1 infected (untargeted or through CD40 targeting), were harvested and used as stimulator cells. A CTL culture was started, incubating the CD8+ T cells together with the migrated skin DC from different in situ Ad-targeting conditions at a ratio of 10:1 (5 ng/ml hIL-7 added at day 0 (Genzyme, Cambridge, MA)), for 1 wk at 37°C and 5% CO2.
To evaluate the activation of M1-specific CTL, an IFN- ELISPOT assay
was performed. Multiscreen 96-well filtration plates (Millipore) were
coated for 3 h at room temperature or overnight at 4°C with the
mAb 1-D1K (50 µl, 15 µg/ml in filtered PBS; Mabtech, Nacka,
Sweden). Plates were washed six times with serum-free medium and
subsequently blocked with filtered complete medium with 10% FCS for
30–60 min at room temperature. Autologous MoDC, cultured over the
course of 5 days from plastic-adherent monocytes in the presence of
GM-CSF and IL-4 as previously described (21), were
infected with either a wild-type vaccinia virus (i.e., control
wild-type vaccinia virus) or a vaccinia virus encoding the
Haeminfluenza matrix protein M1. One donor was found to be
HLA-A2 positive (by FACS analysis), and T2 cells (obtained from
American Type Culture Collection, Manassas, VA) were therefore loaded
overnight with either E7 peptide (i.e., control peptide, containing the
HPV-16 E7-derived HA-A2-binding epitope aa 11–20) (50 µg/ml) or M1
peptide (containing the M1-derived HLA-A2-binding epitope aa 58–66)
(50 µg/ml). The vaccinia virus-infected MoDC or peptide-loaded T2
cells were used as stimulators. The CTL effector cells (2.5 x
104 and 1.25 x 104 per well, in
duplicate) were incubated overnight with the different stimulator cells
(1 x 104 per well) at 37°C in an incubator with a 5%
CO2 humidified atmosphere. The cells were
discarded, and the plates were washed six times with filtered 0.05%
Tween 20 in PBS (PBS-T). A total of 50 µl of the mAb 7-B6-1 (1
µg/ml in filtered PBS; Mabtech) was added to each well, and plates
were left at room temperature for 2–4 h. After six washes with PBS-T,
50 µl streptavidin-alkaline phosphatase (1/1000 diluted in PBS;
Mabtech) was added to each well, and plates were left at room
temperature for 1–2 h. After six washes with PBS-T, 50 µl alkaline
phosphate reagent (AP conjugate substrate kit; Bio-Rad, Hercules, CA)
was added and left for 15 min to 1 h until spots had developed.
The reaction was stopped by washing with tap water. Spots were counted
using an automated ELISPOT plate counter (AID, Strassberg, Germany) and
recalculated to a standardized frequency/100,000 effector CTL.
Statistical analysis
Transduction efficiencies, transgene expression levels, and absolute numbers of (transduced) DC were compared after infection with untargeted or targeted Ad vectors, using the paired Student’s t test (two sided); differences were considered significant when p < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The usefulness of CD40 targeting for the selective transduction of
DC in the skin critically depends on the relative expression levels of
CD40 on the target DC as compared with other cell types with the
ability to express CD40 (25). Immunohistochemical
examination of fresh skin explants revealed a complete lack of CD40
expression both in the epidermis and the dermis (Fig. 1
). To achieve selective up-regulation of
CD40 on DC, GM-CSF was i.d. injected before excision and subsequent
culture of the skin explants. Injection of 100 ng GM-CSF led to optimal
activation of DC as judged on the basis of CD83 and CD40 expression
(range tested: 50–300 ng GM-CSF); up-regulation of both markers was
observed in the epidermal and particularly in the dermal layers with an
optimum at 24 h after injection (Fig. 1). Injection of plain
medium was also found to induce a certain amount of CD83 and CD40
expression after 24 h of culture, although to a lesser extent than
the injection of GM-CSF (Fig. 1). At 24 h of culture, abundant
CD1a expression, which (in contrast to CD40 and CD83) can also be found
on immature DC, was observed both in epidermis and dermis (Fig. 1). The
preferential localization of CD40 and CD83 in the dermis at this time
may reflect migration of these mature DC from the epidermis. In keeping
with our findings for MoDC (21), the cutaneous DC did not
express the primary Ad receptor CAR, but did express the integrin
v5, which acts as
coreceptor and mediates Ad internalization (data not shown). By
providing an alternative receptor to CAR (i.e., CD40) and with the
presence of the Ad coreceptor
(v5), all conditions
were thus met for transduction of DC by CD40-retargeted Ad.
|
Skin explants were injected i.d. with 100 ng GM-CSF in combination
with an unconjugated adenoviral vector encoding lacZ
(Ad-lacZ) or with Ad-lacZ complexed to a
chemically linked bispecific Ab conjugate, directed to the fiber knob
region of the Ad capsid and to CD40 (Fab anti-CD40). The explants
were harvested and snap frozen 48 h after i.d. injection, allowing
for an additional 24 h after optimal CD40 expression levels had
been reached to ensure expression of the gene of interest.
-Gal-expressing cells were visualized and scored on cryosections.
The optimal dose of Ad-lacZ was determined to be
108 PFU/injected biopsy (range tested
106-109 PFU). Injection of
Ad-lacZ resulted in the transduction of a high number of
cells within the dermis, but not in the epidermis (Fig. 2A). Similarly, the injection
of Fab anti-CD40-complexed Ad-lacZ only led to the
transduction of cells in the dermis. However, CD40 targeting of Ad
resulted in a drastic reduction in the number of transduced cells (Fig. 2A), indicating an effective block of the natural tropism of
the Ad vector.
|
-gal-positive cells were counted by microscopic
examination in 10 randomly chosen high power fields. Double staining
with a PE-labeled Ab demonstrated almost all (typically >99%) of the
untargeted Ad-transduced cells to be CD1a negative (Fig. 2
, B and C). In contrast, injection of Fab
anti-CD40-complexed Ad-lacZ resulted in a substantial
increase in the proportion of CD1a+ transduced
cells (at a mean of 44%; Fig. 2, B and C), thus
revealing a more selective targeting of the Ad vector to DC. In
parallel, Ad-lacZ complexed to Fab anti-EGFR (a
bispecific Ab conjugate (of the same isotype as the Fab
anti-CD40)-targeting Ad to the EGFR) was also injected into skin
explants. No -gal activity was detected using this conjugate (Fig. 2C), indicating effective blocking of the natural tropism of
Ad and excluding the possibility of Ad uptake by binding of Abs in the
conjugate to Fc receptors on the surface of DC. These data indicate
that the increased selectivity of cutaneous DC transduction is due to
specific binding to CD40. Phenotype of DC migrated from the skin explants
To further determine the effect of CD40-mediated Ad infection on
DC transduction efficiency and maturation, migrated cutaneous DC were
studied. Spontaneous migration of DC was observed from skin explants,
which were i.d. injected with plain medium and cultured for 48 h
while floating in 5% HPS-containing culture medium. Most of the cells
harvested from the explant cultures at 48 h had the typical
morphologic DC appearance of large cells, with a lobulated nucleus and
membrane invaginations and processes (Fig. 3A). This translated into
characteristic high side (SSC) and forward scatter (FSC) properties, as
measured by flow cytometry (Fig. 3B). Comigrating T cells
were also observed (see Fig. 3A). FACS analysis showed that
the migrated skin DC were positive for the DC markers CD1a and CD11c,
and expressed HLA-DR, -ABC, CD40, CD54, CD80, and CD86 (Fig. 3C). A majority of the DC showed weak expression of
Langerin, suggesting that most of the migrated cutaneous DC were
epidermally derived Langerhans cells (Fig. 3C). Only a small
subset of migrated cells within the SSC/FSC DC gate expressed the
monocyte marker CD14. All migrated DC expressed the early DC maturation
marker CMRF-44, but about one-half of these also expressed the late
maturation marker CD83, indicating the presence of a less and a more
mature migrated DC subset (Fig. 3C). CD83 expression always
coincided with the expression of CCR7 (Fig. 3C), indicating
the ability of fully matured cutaneous DC to home to the paracortical T
cell areas of LN.
|
GM-CSF induced CD40 expression, which was accompanied by DC
maturation (judged on the basis of CD83 expression; see Fig. 1).
Because DC maturation has previously been reported to interfere with Ad
infection (26), we studied the effect of cytokine-induced
maturation on the transduction efficiency of migrated cutaneous DC by
untargeted or CD40-targeted Ad. We compared the effects of i.d.
injection of plain medium, 100 ng GM-CSF, or 100 ng GM-CSF with 1000 U
IL-4, before the 48-h culture of the skin explants. Under all these
conditions, CMRF-44 expression was found on the migrated DC, indicative
of early maturation. Further maturation was induced after the i.d.
injection of GM-CSF and/or IL-4, evidenced by up-regulation of CD83 and
CD40 expression (Fig. 4A). The
most mature phenotype was consistently achieved through the
simultaneous injection of GM-CSF and IL-4 (Fig. 4A).
Accordingly, highest numbers of migrating DC were always found after
i.d. injection of GM-CSF in combination with IL-4 (significant in
comparison with injection of medium; see Table I). The ability of untargeted or
CD40-targeted Ad to infect cutaneous DC in situ in a mature state after
GM-CSF and IL-4 injection was subsequently compared with the more
immature control condition after injection of plain medium.
|
|
), at this
time CD40 expression on DC in situ was found to be optimal (data not
shown). The explants were subsequently placed directly into culture
medium and cultured for another 24 h, after which migrated DC were
harvested and analyzed by FACS. Significant and comparable increases in
transduction efficiency of migrated DC by CD40 targeting of Ad were
observed in both test conditions (Fig. 4B). Following medium
injection, Ad administration led to a mean transduction efficiency of
35.3% (range 14.9–47.7%), while CD40 targeting increased the mean
transduction efficiency to 59.0% (range 30.0–85.6%)
(n = 5, p = 0.005). Corresponding
transduction efficiencies after GM-CSF and IL-4 injection were 44.6%
(range 16.1–58.0%) and 73.4% (range 59.0–84.0%), respectively
(n = 7, p = 0.0003). Similarly,
expression levels of the transduced GFP gene were significantly
increased through CD40 targeting (Fig. 4B).
FACS analyses of migrated cutaneous DC revealed up-regulation of
maturation markers after CD40 targeting of Ad in situ. Although DC
migrating from medium- and GM-CSF/IL-4-injected skin explants already
displayed a (partially) mature phenotype on the basis of CD83
expression (Fig. 4C), this could be further up-regulated in
both cases by the injection of CD40-targeted, but not of untargeted Ad
vectors (Fig. 4C). A simultaneous up-regulation of CD80,
CD86, CD54, and HLA-ABC accompanying CD40 targeting was also observed
(Fig. 4C).
CD40 targeting of Ad consistently led to the migration of significantly
higher numbers of transduced, GFP-expressing DC, with the highest
absolute number of transduced DC migrating from the explants
preinjected with GM-CSF and IL-4 (see Table II).
|
, enhanced transduction efficiency and
transgene expression levels were specific to CD40 targeting and not
found after EGFR targeting. Moreover, Fig. 5A demonstrates
that all transduced cells uniformly expressed the tested DC markers,
but up-regulation of maturation and costimulatory markers was only
observed after CD40-targeted transduction.
|
The long-term effects of CD40 targeting of Ad on transgene
expression and DC activation state were examined in DC migrated from
explants, preinjected with plain medium without cytokines. Migrated
cutaneous DC were harvested as before, but cultured in cytokine-free
medium for an additional 5-day period. Transduction efficiencies and
transgene expression levels for the different targeting conditions were
comparable with those found immediately after migration, with similarly
increased levels of transduction after CD40 targeting (Fig. 6A). This demonstrates
persistent expression of the transgene, both for targeted and
untargeted conditions. However, CD83 expression, indicative of
persistent activation, by this time was only found on DC transduced by
CD40-targeted Ad (Fig. 6A). Moreover, this persistence of
maturation corresponded with the ability of DC to prime allogeneic T
cells in the MLR (Fig. 6B). To also demonstrate the ability
of these DC to activate specific CTL, we made use of an Ad vector
encoding the Haeminfluenza M1 matrix protein (Ad-M1). At
this late time point after migration, DC infected in situ by
CD40-targeted Ad-M1 had the ability to specifically activate autologous
CTL recognizing M1-derived epitopes (measured by IFN- release in an
ELISPOT assay at specific frequencies of about 100/100,000 CTL),
whereas DC infected by untargeted or EGFR-targeted Ad-M1 did not
(measured specific frequencies <10/100,000 CTL); see Fig. 6C.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Immunization by the use of replication-defective Ad vectors encoding specific TAA has resulted in CTL-mediated tumor rejection in murine models (2, 16). However, in vitro studies with MoDC have shown relatively low Ad transduction efficiencies (20–30%) (26, 27, 28). We previously showed that this is most likely due to the absence of the primary Ad receptor (i.e., CAR) on the DC surface (21). Only at extremely high virus titers (multiplicity of infection >1000) or in combination with liposomes can this relative resistance to Ad infection be overcome (26, 27, 28). However, the in vivo use of such high numbers of virus particles may be prohibitive due to considerable cytopathic effects on non-APC. Indeed, i.d. injection of Ad vectors and of DNA plasmids, or their delivery through the gene gun approach, has been reported to result in massive uptake of the vectors and expression of the Ag of interest by skin fibroblasts and keratinocytes (4, 7, 9, 10, 29). However, not these cells, but small numbers of DC, migrating out of the skin and ending up in the T cell areas of the draining LN, were shown to be responsible for the subsequent induction of both Th and CTL responses (7, 8). Theoretically, these DC could have been directly transfected or could have taken up apoptotic fragments from transfected keratinocytes or fibroblasts expressing the Ag of interest (a process known as cross-presentation or cross-priming) (7, 30). Both mechanisms have been shown to occur subsequent to genetic vaccination (4, 7, 8). However, Kurts et al. (11) reported that, in the absence of inflammation, uptake of exogenous Ags from non-APC by DC and subsequent presentation to CTL in draining LN could lead to the deletion of specific CTL clones. In addition, it has been proposed that the transfected keratinocytes, directly functioning as nonprofessional APC, might induce tolerance. In the context of these two latter scenarios, it is important to develop methods to selectively transfect and properly activate DC in vivo. Through our approach of CD40-targeted Ad transfer after cytokine-induced maturation of cutaneous DC in situ, both these conditions are met. At the same time, lower dosages of Ad can be used, thus eliminating cytopathic side effects while ensuring an optimal therapeutic efficacy of the vaccine.
In vivo targeting of Ag to DC provides an attractive immunotherapeutic alternative to the laborious ex vivo generation and Ag loading of autologous DC, followed by adoptive transfer. Immunotargeting of protein Ags to APC without a further need for adjuvants to induce humoral immune responses in vivo was previously described using such targeting molecules as MHC-II (31, 32, 33) and Fc receptors (33, 34). Targeting to more DC-restricted markers, such as CD11c, was shown to induce even stronger responses (35, 36, 37, 38).
A number of advantages adhere to our approach of combined GM-CSF/IL-4-induced maturation and CD40 targeting of DC in situ, as discussed below.
GM-CSF has been reported to enhance the number and maturation state of DC at the site of administration, leading to a more efficient Ag presentation and T cell-mediated tumor rejection in vivo (39, 40). The in vivo induction of a higher DC maturation state through the administration of IL-4 in addition to GM-CSF has also been reported (41). Indeed, several reports suggest the importance of IL-4 in the differentiation and maturation of DC (42, 43); without it they may become maturation resistant and acquire a T cell-tolerizing capacity (43). Although injection of medium alone already induced sufficient levels of CD40 expression to observe an improved transduction efficiency through CD40 targeting, the observed additional increase in maturation and migration certainly warrants the in vivo use of the GM-CSF/IL-4 cocktail.
GM-CSF- and CD40-mediated maturation does not impair endocytotic processes in DC, as evidenced by their reported maintained ability for cross-priming (44). This also suggests a continued ability to take up Ad-conjugate complexes. This is indeed confirmed by our data, revealing even higher transduction efficiencies and transgene expression levels after in situ maturation of the DC through GM-CSF and IL-4 administration before Ad injection. Moreover, the DC matured by CD40-targeted Ad infection appeared to preserve appropriate processing pathways for CTL epitope presentation, as evidenced by their ability to specifically activate CTL against HLA class I-binding epitopes contained within the Ad-encoded transgene M1.
GM-CSF and agonistic CD40 binding have been reported to enhance the survival of DC through increased apoptosis resistance (45, 46). This may protect them during migration and increase their chances to encounter and activate specific T cells in the draining LN.
Previous studies reported spontaneous migration of DC from skin explants after 48 h of culture (22, 23). We observed a significant increase in this migration by the i.d. injection of GM-CSF and IL-4. GM-CSF had previously been shown to stimulate the migration of DC from the skin to LN (39, 47). Although IL-4 was recently reported to hamper migration of Langerhans cells (48), we actually observed an increase in migration after the administration of IL-4 in combination with GM-CSF. Our results further demonstrate that the Ad-transduced DC retain their capacity to migrate. Nor did the binding of the Fab anti-CD40 conjugate impair migration, which is in keeping with previous murine studies (24). Indeed, an essential and regulatory role of CD40-CD40L interactions in the in vivo migration of cutaneous DC transporting Ag to the T cell areas in the LN was recently reported (49). In this context, our use of an agonistically binding Ab may ensure appropriate DC migration. That this may be the case is further supported by our recent finding of high transduction efficiencies (68–91%) of CD1a+ DC migrated from skin explants that were cultured for 48 h subsequent to injection with CD40-targeted Ad-GFP, contrasting with relatively low percentages (5–10%) of Ad-transduced cells among the CD1a+ nonmigrated DC derived from dermal single cell suspensions from the same explants (de Gruijl et al., unpublished data). These numbers suggest a selective migrational advantage of DC following their in situ transduction and activation by CD40-targeted Ad.
Our finding that CD40-targeted Ad transfer leads to preservation of the mature and T cell stimulatory phenotype of DC for up to 7 days after the start of migration suggests that it may be an effective vaccine delivery system for the generation of strong antitumor responses. In contrast, migrated DC infected by EGFR-targeted or untargeted Ad had lost their CD83 expression and no longer stimulated allogeneic T cells after 7 days. Over these 7 days, we also found the DC infected by CD40-targeted Ad to retain the ability to activate CTL specifically recognizing the influenza M1 transgene, whereas DC transduced by untargeted or EGFR-targeted Ad did not. This is in keeping with a recent study by Hawiger et al. (50), who showed that targeting of a model Ag to the DEC-205 receptor on murine DC led to specific T cell unresponsiveness within 7 days after immunization, but that this unresponsiveness could be overcome by the coinjection of a CD40 agonistic Ab.
The increased activation of DC effected through CD40 targeting of Ad, possibly coupled to enhanced survival and stability of the mature phenotype, may also account for the more effective CTL-dependent tumor rejection we previously observed in a mouse model, involving the adoptive transfer of DC, transduced in vitro by CD40-targeted Ad encoding the model TAA HPV-16 E7 (51). Other studies also revealed a key role for CD40-mediated DC activation in the generation of tumor-specific CTL and Th1 responses (52). This may in large part be due to CD40-induced IL-12 production, which we observed previously in MoDC after CD40-targeted Ad transduction (21). Similarly, Rea et al. (26) reported a synergistic enhancement of IL-12 production by simultaneous Ad infection and CD40 stimulation.
In conclusion, CD40-mediated DC targeting in situ increases the number of Ag-carrying and migrating DC and ensures a long-lived activated phenotype of the DC for optimal T cell stimulation. Based on our findings, we propose a vaccination approach whereby i.d. injection of GM-CSF (possibly in combination with IL-4) primes the DC at the vaccination site for subsequent infection by CD40-targeted Ad vectors, encoding a TAA of choice.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 T.D.d.G. and S.A.L.-d.B. contributed equally to this manuscript.
3 Address correspondence and reprint requests to Dr. Rik J. Scheper, Department of Pathology, VU University Medical Center, P.O. Box 7057, 1007 MB Amsterdam, The Netherlands. E-mail address: rj.scheper{at}vumc.nl
4 Abbreviations used in this paper: TAA, tumor-associated Ag; Ad, adenovirus; -gal, -galactosidase; CAR, coxsackie-adenovirus receptor; DC, dendritic cell; EGFR, epidermal growth factor receptor; FSC, forward scatter; GFP, green fluorescent protein; HPS, human pooled serum; i.d., intradermal; LN, lymph node; MoDC, monocyte-derived DC; RT, room temperature; SSC, side scatter.
Received for publication November 28, 2001. Accepted for publication August 20, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-producing T cells from tumor-infiltrating lymphocytes. J. Immunol. 164:6633.This article has been cited by other articles:
|
|
 |
|
  R. van de Ven, G. L. Scheffer, A. W. Reurs, J. J. Lindenberg, R. Oerlemans, G. Jansen, J.-P. Gillet, J. N. Glasgow, A. Pereboev, D. T. Curiel, et al. A role for multidrug resistance protein 4 (MRP4; ABCC4) in human dendritic cell migration Blood, September 15, 2008; 112(6): 2353 - 2359. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. D. de Gruijl, O. J. A. E. Ophorst, J. Goudsmit, S. Verhaagh, S. M. Lougheed, K. Radosevic, M. J. E. Havenga, and R. J. Scheper Intradermal Delivery of Adenoviral Type-35 Vectors Leads to High Efficiency Transduction of Mature, CD8+ T Cell-Stimulating Skin-Emigrated Dendritic Cells J. Immunol., August 15, 2006; 177(4): 2208 - 2215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. D. de Gruijl, C. C. Sombroek, S. M. Lougheed, D. Oosterhoff, J. Buter, A. J. M. van den Eertwegh, R. J. Scheper, and H. M. Pinedo A Postmigrational Switch among Skin-Derived Dendritic Cells to a Macrophage-Like Phenotype Is Predetermined by the Intracutaneous Cytokine Balance. J. Immunol., June 15, 2006; 176(12): 7232 - 7242. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Krebs, E. Scandella, B. Odermatt, and B. Ludewig Rapid Functional Exhaustion and Deletion of CTL following Immunization with Recombinant Adenovirus J. Immunol., April 15, 2005; 174(8): 4559 - 4566. [Abstract] [Full Text] [PDF]
|
|
|
|
