|
|
||||||||
CUTTING EDGE |
,
,
,¶,||
,
,
,||
Departments of
*
Urology and
Microbiology,
Prostate Cancer Research Group,
Immunology Program,
¶ Veterans Administration Medical Center, and
||
Cancer Center, University of Iowa, Iowa City, IA 52242
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
A number of investigators have reported effective CTL generation with recombinant viral vectors that effectively controlled tumor growth (8, 9, 10, 11). Although viruses are effective in generating CTL-mediated antitumor activity in naive mice, the use of viruses in settings where the host has been previously exposed to the virus significantly reduces transgene expression (12, 13). Studies by Yang and associates formally demonstrated that T cell responses, specifically CTL, to viral proteins were responsible for destruction of cells expressing the transgene (12). Furthermore, Abs in the serum diminish the ability of viruses to deliver the transgene (13), and the resulting decrease in gene expression reduces subsequent CTL activation (14, 15, 16, 17). The diminished ability to generate CTL, when Abs to the viral vector are present, has been suggested to be an important reason for the lack of CTL after adenovirus delivery of melanoma Ags in clinical trials (18).
Numerous attempts to circumvent the adverse effects of viral immunity on gene expression and immune activation have been reported, including boosting regimens with a different viral vector (19) and inhibiting the host anti-viral immune response with agents such as soluble TNF receptor (20), CTLA4Ig (21), or cyclophosphamide (22). The collagen-based matrix, Gelfoam (Pharmacia and Upjohn, Kalamazoo, MI) is primarily used as an intraoperative hemostatic agent, but also has been used to deliver a number of different compounds, including insulin (23), various cytokines, and growth factors (24, 25), to improve and sustain delivery. We have recently reported the use of Gelfoam as a solid-state delivery vehicle that significantly enhanced subsequent gene expression and improved tumor inhibition in an in situ immunotherapy protocol (26). Here, we demonstrate the ability of Gelfoam to enhance the generation of CTL, both in the presence of neutralizing Abs to the delivery virus as well as in a self-Ag model.
| Materials and Methods |
|---|
|
|
|---|
The BALB/c murine prostate cancer model, RM-11, was obtained
from Dr. Timothy C. Thompson (Baylor College of Medicine, Houston, TX)
(27). A RM-11psa clone was obtained by transfecting the
full-length prostate-specific Ag
(PSA)2 cDNA (provided
by Donald Tindall, Mayo Clinic, Rochester, MN) into pH
Apr-1-neo
(28). RM-11neo was generated in like manner with the
pH
Apr-1-neo vector alone (RM-11neo). A second PSA-expressing line,
the DBA/2 (H-2d) mastocytoma cell line P815, was
transfected with pH
Apr-1-neo-PSA (P815psa). The murine T cell
lymphoma cell line, EL-4 and the OVA-expressing EL-4 cell line, EG.7,
were used for CTL assays in the self-Ag model, rat insulin promoter
(RIP)-mOVA-transgenic mice (29). Mice were either bred in
our animal facility (RIP-mOVA) or obtained through the National Cancer
Institute. All animal studies were approved by the University of Iowa
Animal Review Board and were performed in accordance with institutional
guidelines.
Gene transfer vectors
A replication-deficient type 5 adenovirus recombinant for PSA (Ad5-PSA) expressed from the CMV promoter was generated using standard methods by the University of Iowa Gene Transfer Vector Core (Iowa City, IA) (30). Briefly, the entire coding sequence of PSA cloned into pAd5CMVK-NpA. The resultant plasmid and adenovirus backbone sequences that had the E1 (E1A and E1B) genes deleted were transfected into human embryonic kidney (HEK) 293 cells, and viral particles were isolated and amplified for analysis of PSA expression. Type 5 adenovirus recombinant for OVA (Ad5-mOVA) used for experiments in the self-Ag model, as well as the control vector Ad5-lacZ (31), were prepared as described for Ad5-PSA.
Delivery system and immunizations
Gelfoam is a collagen matrix prepared from purified pork skin collagen (23, 24, 25, 26). Preparation of the viral vectors in the Gelfoam for injections was performed as previously described (26). The matrix/virion mixture was injected s.c. into the flanks of mice. CTL activity induced by matrix-delivered vectors was compared with s.c. injection of the same number of viral particles without matrix (fluid phase).
To investigate possible benefit of Ad5-PSA immunization with Gelfoam in the presence of anti-adenovirus Abs, mice were injected i.p. with Ad5-lacZ (109 PFU) 2 wk before immunization with Ad5-PSA. Control mice were injected i.p. with PBS. Serum samples were obtained before and after immunization and subsequently assayed for adenovirus-specific Ab titers.
Isolation and detection of lytic T cells
Fourteen days after virus injection, splenocytes were isolated and tested for CTL activity as previously described (17). Briefly, for PSA-specific CTL, splenocytes were incubated in 24-well plates at 1 x 107 cells/well together with the mitomycin C-treated P815psa stimulator cells (4 x 105) in DMEM supplemented with 10% FCS, 5 x 10-5 M 2-ME, and 10 IU/ml recombinant human (rh)IL-2. The CTL assays in the RIP-mOVA self-Ag model were performed in a similar fashion, except that in vitro stimulation was performed as a 5-day coculture with Ad5-mOVA-infected dendritic cells or SIINFEKL-pulsed 1500 cGy irradiated syngeneic spleen cells as stimulators (17, 32, 33). CD8-depleted effector cells were obtained by depleting CD8+ T cells after the 5-day culture with the Vario Macs magnetic bead system (Miltenyi Biotec, Auburn, CA) as described by the manufacturer. CD8-depleted effector cells contained <2% CD8+ T cells as determined by flow cytometry.
| Results |
|---|
|
|
|---|
Previous studies demonstrated that Gelfoam enhanced the
distribution and expression of genes delivered by viral vectors in vivo
(26). To determine whether Gelfoam also enhanced CTL
activation, Ad5-PSA (1 x 1071 x
109 PFU) was injected into BALB/c mice with and
without Gelfoam. Injection of mice with Ad5-lacZ (1 x
109 PFU) served as a negative control. Detection
of CTL against PSA-expressing targets was observed without Gelfoam only
at the highest virion concentration (1 x
109 PFU; Fig. 1
A). Immunization with Gelfoam
consistently demonstrated significant specific lysis at lower virion
concentrations (1 x 1071 x
109 PFU, Fig. 1
B) in at least three
separate experiments. Further studies in which either
CD8+ T cells were depleted with anti-CD8 mAb
(clone 2.43) or anti-CD8 was added to the lytic assay resulted in
abrogation of lysis, demonstrating the effector cell population to be
CD8+ T cells (data not shown).
|
Previous studies showed diminished CTL activity to transgene
products if delivered by a viral vector to which the animal was immune
(14, 15, 16, 17). Because the impact of viral immunity on CTL
activation has been implicated as a negative factor in clinical studies
and Gelfoam showed the capacity to enhance CTL activity, we tested the
effect of Ad5-PSA in Gelfoam on CTL activation in the presence of
anti-adenoviral immunity (18). Mice were first
injected i.p. with 1 x 109 PFU Ad5-lacZ or
a similar volume of PBS. Serum anti-adenoviral Ab titers of
Ad5-lacZ-injected mice from both Ad5-PSA and Ad5-PSA with Gelfoam
groups showed high levels of anti-adenovirus Ab at the time of
Ad5-PSA injection (52.4 and 62.1 µg/ml, respectively). Subsequently,
Ad5-lacZ-primed and control mice were immunized with Ad5-PSA with and
without Gelfoam. As expected, prior infection with Ad5-lacZ
significantly inhibited the ability of Ad5-PSA to induce a PSA-specific
CTL response (Fig. 2
). However,
immunization with the same Ad5-PSA concentration (1 x
109 PFU) delivered with Gelfoam restored
PSA-specific CTL lytic activity to the levels observed in
adenovirus-naive mice.
|
Significant improvement of immunization was demonstrated using
Gelfoam as a delivery vehicle in the foreign Ag model using human PSA
in a murine system, which lacks a PSA homolog. However, limited success
in activating CTL has been reported in nonmutated self-Ag models
(9). To determine whether viral delivery with Gelfoam
would improve immunization to self-Ags, we used RIP-mOVA mice, which
develop tolerance to OVA-specific peptides (29, 32).
RIP-mOVA mice were immunized with Ad5-mOVA (1 x
108 PFU) with and without Gelfoam. Ad5-mOVA
(1 x 108 PFU) in the fluid phase also was
injected into both RIP-mOVA mice and, as a control, into normal C57BL/6
mice. Immunization of normal C57BL/6 mice demonstrated OVA-specific
lysis of EG.7 target cells, whereas Ad5-mOVA immunization (i.p. or
s.c.) of RIP-mOVA mice showed no OVA-specific lysis (Fig. 3
). In contrast, Ad5-mOVA immunization in
Gelfoam stimulated significant OVA-specific CTL. Lysis by cells
activated with Ad5-mOVA in Gelfoam was mediated by
CD8+ T cells (Fig. 3
, inset).
Moreover, histological evaluation of the pancreas from
Ad5-mOVA-plus-Gelfoam-primed mice showed
-islet-associated
inflammation but no obvious
-islet destruction (data not shown).
|
| Discussion |
|---|
|
|
|---|
Several strategies have been adopted to overcome or bypass this neutralizing immune response to optimize the use of these viral vectors. Tumor Ags can be delivered as defined peptides (35), proteins designed to access the MHC class I pathway of APCs in vivo (36), or by adoptive transfer of Ag-loaded APCs such as dendritic cells (37). DNA vaccines encoding tumor Ags can be delivered as naked DNA or encapsulated by liposomes (38). Xiang et al. have recently described an autologous oral DNA vaccine delivered with an attenuated strain of Salmonella typhimurium to successfully immunize mice in a self-Ag model of melanoma (39). In addition, less immunogenic vectors have been used including adeno-associated virus, lentivirus, and gutless adenovirus. Also, avoidance of the antiviral immune response of the more immunogenic viral vectors may be possible by varying the vectors that are used for boosting the immune response (19). Finally, treatment with suppressive agents such as soluble TNF receptor (20), CTLA4Ig (21), cyclophosphamide (22), and anti-CD40 (40) has been shown to inhibit the host anti-viral immune response, improve transgene expression, and enhance the ability to use vectors for multiple injections. Although the latter suppressive approaches are effective in blocking the development of immunity to viral proteins, their use in the context of vector-induced CTL activation is controversial. Clearly, alternative approaches for gene delivery and CTL activation are needed.
Our previous data demonstrated that intratumoral injection of viral
vectors in Gelfoam enhanced both the distribution and the expression
level of reporter transgenes transferred by viruses (26).
Furthermore, the enhanced gene delivery with the Gelfoam enhanced
biologic activity of an in situ immunotherapy protocol
(26). As a result of the enhanced gene delivery and
augmented antitumor effects induced by coinjection of Gelfoam and
recombinant viruses, we tested the effects of Gelfoam on
Ad5-PSA-mediated CTL activation. Ad5-PSA was previously observed to
induce CTL activity in a murine tumor model system, and to activate
CTL-mediated PSA-specific antitumor activity.3
Here we show that Gelfoam enhances CD8+ CTL
activation in both foreign and autologous Ag settings. Consistent with
previous reports in the RIP-mOVA model where OT-1 adoptive transfer
frequency was low, inflammation was observed in the pancreas but
-islet destruction was not apparent (Ref. 40 , and data
not shown). Importantly, we show that administration of Ad5-PSA with
Gelfoam abrogated the inhibitory effects on CTL activation of
preexisting immunity to adenovirus. These data suggest that coinjection
of recombinant viruses with Gelfoam may provide an approach for
augmenting CTL activation in the clinical setting and also may provide
a means of using viral vectors for multiple injections to boost
immunity.
The mechanism(s) by which Gelfoam enhances CTL priming is not known. Previous studies demonstrated that CTL priming either with pox or polio virus required MHC class I-matched bone marrow-derived cells. Interestingly, either recombinant Ag produced in the periphery or Ag produced by bone marrow-derived cells was sufficient to activate CTL (41). In this regard, distribution studies, which were performed using PCR amplification of adenoviral DNA, showed an absence of detectable DNA in the liver, lungs, and other tested organs, but adenoviral DNA was detected in the draining lymph nodes for both fluid phase and Gelfoam-delivered virus. These initial biodistribution studies were not quantitative but clearly show the presence of virus in draining lymph nodes. Perhaps Gelfoam enhances delivery to the draining nodes where higher Ag expression could enhance T cell priming (42). It is clear that the delivery of genes in Gelfoam results in a 5- to 10-fold enhancement of gene expression and an extension of the time course of expression (26). Likewise, it is known that Ag levels play an important role in CTL priming (33, 43). In this regard, Xiang and associates recently showed that by directing proteins to the proteosome, which enhanced MHC class I Ag presentation and limited Ab development, CTL activation was enhanced sufficiently to mediate antitumor activity (39). Thus, the simplest explanation for the observed ability of Gelfoam to augment CTL activation is elevated production of the transgene product, which delivers a higher Ag load to bone marrow-derived APCs. Determination of the functional role of Gelfoam is currently under investigation.
| Footnotes |
|---|
2 Abbreviations used in this paper: PSA, prostate-specific Ag; Ad5-PSA, type 5 adenovirus recombinant for PSA; Ad5-mOVA, type 5 adenovirus recombinant for OVA; RIP, rat insulin promoter. ![]()
3 B. D. Elzey, D. R. Siemens, T. L. Ratliff, and D. M. Lubaroff. Immunization with type 5 adenovirus recombinant for PSA in combination with ALVAC cytokine gene delivery induces destruction of established prostate tumors. Submitted for publication. ![]()
Received for publication June 14, 2000. Accepted for publication November 21, 2000.
| References |
|---|
|
|
|---|
decreases inflammation and prolongs adenovirus gene expression in lung and liver. Hum. Gene Ther. 9:1875.[Medline]
-actin expression vector system directs high-level accumulation of antisense transcripts. Proc. Natl. Acad. Sci. USA 84:4831.This article has been cited by other articles:
![]() |
B. D. Elzey, N. W. Schmidt, S. A. Crist, T. P. Kresowik, J. T. Harty, B. Nieswandt, and T. L. Ratliff Platelet-derived CD154 enables T-cell priming and protection against Listeria monocytogenes challenge Blood, April 1, 2008; 111(7): 3684 - 3691. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. L. VanOosten and T. S. Griffith Activation of Tumor-Specific CD8+ T Cells after Intratumoral Ad5-TRAIL/CpG Oligodeoxynucleotide Combination Therapy Cancer Res., December 15, 2007; 67(24): 11980 - 11990. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Liu, D. P. Evanoff, X. Chen, and Y. Luo Urinary Bladder Epithelium Antigen Induces CD8+ T Cell Tolerance, Activation, and Autoimmune Response J. Immunol., January 1, 2007; 178(1): 539 - 546. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |