The potential benefits of immunotherapy as an adjunct to conventional forms of cancer treatment are under active investigation.1, 2, 3, 4 Activated cytotoxic T lymphocytes (CTLs) capable of recognizing and destroying cancer cells are generated in immunized mice and patients. The immunity is directed toward unique MHC class I restricted TAAs expressed by the malignant cells.5, 6, 7, 8
Although experimental immunotherapy protocols in mice are revealing the potential of this form of treatment, effective vaccination strategies in cancer patients are wanting. One possible explanation is that even though the immune system can adversely affect diffuse and smaller tumors, it cannot effectively destroy large, established neoplasms. An immunotherapeutic strategy that would allow treatment at an early stage of the disease could have significant benefits.
Tumor cells are the richest source of tumor antigens. Immunization with malignant cells modified to secrete immune-augmenting cytokines such as IL-2,9, 10, 11 GM-CSF,12 IL-4,13 IL-614 and IL-1215, 16 resulted in rejection of the cytokine-secreting cells and the induction of T cell mediated immunity toward the neoplastic cells. In some instances, the induced immunity was sufficient to prolong the lives of mice with established neoplasms. However, the direct modification of cancer cells from a primary neoplasm is technically challenging. It requires the establishment of a tumor cell line, which cannot always be accomplished. This is especially the case for breast cancer in patients. It is notoriously difficult to establish breast cancer cell lines from primary neoplasms.
In prior reports,17, 18, 19 we described the results of studies in mice with breast cancer treated by immunization with a vaccine prepared by transfer of sheared genomic DNA fragments from various murine neoplasms, including adenocarcinoma of the breast, into a highly immunogenic, mouse fibroblast cell line. The rationale was that genes specifying TAA would be expressed in a highly immunogenic form by the transfected cells. As the transferred DNA is integrated into the genome of the recipient cells, and replicated as the cells divide, the vaccine could be prepared from DNA derived from relatively small numbers of cancer cells. Sufficient DNA could be recovered from as few as 10 million cancer cells. (A tumor of 4 mm contains an equivalent number of cells.) The vaccine was readily prepared from primary neoplasms. The establishment of a tumor cell line was not required. Furthermore, as the DNA was not fractionated before transfer, it was likely that multiple mutant/dysregulated genes in the breast cancer cells specifying an array of unidentified weakly immunogenic TAAs were expressed by the transfected cells.
Like many other vaccination strategies, however, the DNA-based vaccine was not effective against larger and more aggressive tumors. The survival of mice with well-established breast neoplasms treated solely by immunization with the vaccine was not significantly different than that of untreated mice with breast cancer.
In an attempt to improve the therapeutic outcome, we combined treatment with the vaccine with paclitaxel, a drug commonly used in breast cancer therapy.20, 21 C3H/He mice bearing an established, highly aggressive breast cancer (generation time = 18.3 hr) derived from a neoplasm that arose spontaneously in a C3H/HeJ mouse were treated with noncurative amounts of paclitaxel, followed by immunization with the DNA-based cell vaccine. The results indicated that the survival of mice with established breast cancer receiving the combined therapy exceeded that of tumor-bearing mice receiving either form of treatment alone. To our knowledge, this report is the first indicating that combination therapy with paclitaxel along with immunization with a unique vaccine prepared from microgram amounts of tumor tissue was capable of prolonging the survival of mice with breast cancer.
Material and methods
Experimental animals and tumor cell lines
Eight- to 10-week-old pathogen-free C3H/HeJ female mice were from the Jackson Laboratory (Bar Harbor, ME). The animals, between 10 and 14 weeks old when used in the experiments, were maintained according to NIH Guidelines for the Care and Use of Laboratory Animals. SB5b cells were a short-term passage adenocarcinoma of the breast cell line derived from a breast neoplasm that arose spontaneously in a C3H/He mouse in our animal colony. B16 cells, a melanoma cell line of C57BL/6 origin, were obtained originally from I. Fidler (M.D. Anderson, Houston, TX). The cells were maintained by serial passage in histocompatible C3H/HeJ or C57BL6J mice, respectively, or at 37°C in a humidified 7% CO2/air atmosphere in Dulbecco's modified Eagle's medium (DMEM) (Gibco BRL, Grand Island, NY) supplemented with 10% heat inactivated fetal bovine serum (FBS) (Sigma, St. Louis, MO) and antibiotics (Gibco BRL) (growth medium). LM cells were from the American Type Culture Collection (ATCC number CCL-1.3).
Isolation of DNA from the breast cancer cells
A DNeasy isolation kit (Qiagen, Valencia, CA) was used to obtain genomic DNA from the breast cancer cells, according to the manufacturer's instructions. In brief, 1 × 107 actively proliferating plastic adherent breast cancer cells from in vitro culture were disassociated from the plastic by treatment with EDTA (10−4 M). The cell suspension was centrifuged at 300 × g for 5 min. Afterward, the cell pellet was resuspended in 200 μl PBS, followed by the addition of 400 μg RNase A and incubation at RT for 2 min. After incubation, 12 mAU proteinase K and 200 μl lysis buffer containing guanidine HCl were added, followed by further incubation at 70°C for 10 min. Afterward, 200 μl of 100% ethanol was added. The extracted DNA was loaded onto the DNeasy spin column, and eluted after 2 washes with buffer. The A260/A280 ratio of the isolated DNA was greater than 1.8 in each instance. The molecular size of the extracted DNA was ˜25 kb, as determined by agarose gel electrophoresis. The same procedure was followed to isolate DNA from B16 cells, a melanoma cell line.
Modification of LM fibroblasts to secrete IL-2
To augment their nonspecific immunogenic properties, the fibroblasts were modified to secrete IL-2 before they were transfected with DNA from the breast cancer cells. A vector (pZipNeoSV-IL-2; from M.K.L. Collins, University College, London, England) specifying human IL-2 was used for this purpose. The IL-2-specifying vector was packaged in GP+env AM (2) cells (from A. Bank, Columbia University, New York, NY). The vector also included a neo r gene under control of the Moloney leukemia virus long terminal repeat. The neor gene conferred resistance to the aminoglycoside antibiotic neomycin-derivative, G418 (Gibco BRL), used for selection.
Virus-containing supernatants of GP+env AM12 cells transduced with pZipNeoSV-IL-2 were added to the fibroblasts, followed by overnight incubation at 37°C in growth medium containing polybrene (Sigma; 5 μg/ml, final concentration). The cells were maintained for 14 days in growth medium containing 400 μg/ml G418 (Gibco BRL). One hundred percent of nontransduced cells died in medium supplemented with G418 during this period. Colonies of cells proliferating in the G418-containing growth medium were pooled and maintained as modified cell lines for later use in the experiments. An ELISA (BioSource, Camarillo, CA) was used to determine the quantity of IL-2 secreted by the transduced fibroblasts (LM-IL-2 cells).
Modification of the cytokine-secreting fibroblasts to express H-2Kb-class I-determinants
Allogeneic class I-determinants are strong immune adjuvants.21, 22, 23, 24 To further augment their immunogenic properties, the fibroblasts were modified to express MHC H-2Kb-determinants, allogeneic in C3H/He mice. A plasmid (pBR327H-2Kb) (Biogen Research, Cambridge, MA) encoding H-2Kb-determinants was used. Ten microgram of pBR327H-2Kb and 1 μg of pBabePuro (from M. K. L. Collins), a plasmid specifying a gene that confers resistance to puromycin, were mixed with Lipofectin (Gibco BRL), and added to 1 × 106 cytokine-secreting fibroblasts in 10 ml of DMEM, without FBS. (A 10:1 ratio of tumor-DNA to plasmid DNA was used to increase the likelihood that cells converted to puromycin-resistance took up tumor-DNA as well.) The IL-2-secreting fibroblasts were incubated under standard cell culture conditions for 18 hr at 37°C in growth medium. After incubation, the cell cultures were divided and replated in complete growth medium supplemented with 3.0 μg/ml puromycin (Sigma, St. Louis, MO), followed by incubation at 37°C for 7 additional days. The surviving colonies were pooled and maintained as a cell line for later use (LM-IL-2Kb cells). One hundred percent of nontransduced cells maintained in growth medium containing equivalent amounts of puromycin died during the 7-day period of incubation.
Immunofluorescent staining and cytofluorometric measurements
Quantitative immunofluorescence measurements were used to detect the expression of MHC class I determinants and costimulatory/adhesion molecules by the fibroblasts used as recipients of DNA from the breast cancer cells. The measurements were performed in an Epic V flow cytofluorograph (Coulter Electronics, Hialeah, FL) equipped with a multiparameter data acquisition and display system (MDADS). For the analysis, 0.1 mM EDTA in PBS was used to disassociate the monolayer cultures from plastic cell culture flasks. The cell suspensions were washed with PBS containing 0.2% sodium azide and 0.5% FBS. Afterward, FITC/PE-conjugated H-2Kk, H-2Kb, I-Ak, B7.1, B7.2 or ICAM-1 mAbs (Pharmingen, San Diego, CA) were added to the cell cultures, followed by incubation at 4°C for 1 hr. The cells were then washed with PBS containing 0.5% FBS and 0.2% sodium azide. Background staining was determined by substituting FITC-conjugated IgG2a isotype serum (DAKO, Carpenteria, CA) for the specific mAbs. One-parameter fluorescence histograms were generated by analyzing at least 1 × 104 cells in each instance.
Preparation of the vaccine
The vaccine was prepared by transfer of sheared, unfractionated DNA fragments from SB5b breast cancer cells into LM fibroblasts, which had been modified to secrete IL-2 and to express H-2Kb-determinants (LM-IL-2Kb/SB5b cells). The method described by Wigler et al.25 was used, as modified. In brief, 50 μg of sheared (25 kb) genomic DNA derived from ˜1 × 107 breast cancer cells was mixed with 5 μg pHyg (from L. Lau, University of Illinois, Chicago), a plasmid that encoded the E. Coli enzyme hygromycin B phosphotransferase gene, conferring resistance to hygromycin B, used for selection. The sheared DNA and pHyg were mixed with Lipofectin, according to the manufacturer's instructions (Gibco BRL) and added to 1 × 107 modified fibroblasts divided 24 hr previously into 10 100 mm plastic cell culture plates. Eighteen hours after the addition of the DNA/Lipofectin mixture to the cells, the growth medium was replaced with fresh growth medium containing sufficient quantities of hygromycin (500 μg/ml; Boehringer Mannheim, Indianapolis, IN) to kill 100% of the nontransfected cells. For use as a control, 5 μg of pHyg alone mixed with Lipofectin was added to an equivalent number of the modified fibroblasts. As an additional control, the same procedure was followed except that DNA from B16 melanoma cells was substituted for DNA from SB5b cells. In each instance, the cells were maintained for 14 days in growth medium containing 500 μg/ml hygromycin B. None of the nontransfected cells maintained in the selection medium were viable by the end of this period. The remaining colonies (at least 2 × 104) were pooled and maintained as cell lines for use in the experiments (LM-IL-2Kb/SB5b cells and LM-IL-2Kb/B16 cells, respectively).
ELISPOT IFN-γ assays were used to detect the presence of spleen cells responsive to the breast cancer cells in mice immunized with the modified, transfected fibroblasts. Responder (R) T cells from the spleens of C3H/HeJ mice immunized with the transfected cells were added into individual wells (1 × 106 cells/well in 0.2 ml growth medium) of 96-well ELISPOT IFN-γ plates (B-D Pharmingen, ELISPOT Mouse IFN-gamma Set (Cat no. 551083)) coated with 100 μl of the capture Ab (5 μg/ml in PBS). Stimulator (S) SB5b breast cancer cells were then added at an R:S ratio of 10:1. After incubation for 18 hr at 37°C, the cells were removed by washing with PBS-Tween (0.05%). Detection antibodies (2 μg/ml) were then added to each well. The plates were incubated for 2 hr at RT and the washing steps were repeated. Afterward, streptavidin-peroxidase (Streptavidin-HRP, 5 μg/ml) was added to the individual wells and the plates were washed 4 times with PBS-Tween and twice with PBS. One hundred microliter of aminoethylcarbazole staining solution was added to each well to develop the spots. The reaction was stopped after 4–6 min with deionized water. The spots were counted by computer-assisted image analysis (ImmunoSpot Series 2 analyzer: Cellular Technology, Cleveland, OH).
Detection of CTLs reactive with the breast cancer cells by 51Cr-release cytotoxicity assays
A standard 51Cr-release assay was used to detect the presence of spleen cells with cytotoxic activity toward the breast cancer cells in mice immunized with the transfected fibroblasts. Spleen cell suspensions were prepared as described. After washing, the cells were coincubated under standard cell culture conditions for 5 days with (mitomycin C-treated) SB5b cells. The incubation medium consisted of RPMI-1640 medium (Gibco BRL) supplemented with 100 U/ml human IL-2, 10% FBS, 5 × 10−2 mmol 2-mercaptoethanol, 15 mmol HEPES, 0.5 mmol sodium pyruvate and penicillin/streptomycin (Gibco). The ratio of spleen cells to mitomycin-C-treated breast cancer cells during the coincubation was 30:1. At the end of the 5 day incubation period, the population that failed to adhere to the plastic cell culture flasks was collected and used as the source of effector cells for the cytotoxicity determinations.
For the cytotoxicity assay, 5 × 106 SB5b cells were labeled with 51Cr during a 1 hr incubation period at 37°C in growth medium containing 100 μCi Na251CrO4 (Amersham, Arlington Heights, IL). After 3 washes with DMEM, 1 × 10451Cr-labeled cells were incubated for 4 hr at 37°C with the effector cell-population. The quantity of isotope released was measured in a gamma counter (Beckman, Palo Alto, CA).
The percent specific cytolysis was calculated as
The spontaneous release of 51Cr was less than 15% of the total release in each instance.
Kaplan-Meier log rank analyses were used to determine the statistical differences between the survival of mice in the various experimental and control groups. A p value less than 0.05 was considered significant. Student t test one-way Anova was used to determine statistical difference between experimental and control groups in the in vitro experiments.
Paclitaxel inhibited the growth of breast cancer cells in C3H/He mice
Paclitaxel is a potent inhibitor of cell division.26, 27, 28 It blocks cells in the G2/M phase of replication through its effect on the formation and function of microtubules within the cell. To determine whether paclitaxel affected the growth of the breast cancer cells used in the experiments described here, SB5b cells were injected into the mammary fat pad of naïve C3H/He mice. Six days after injection of the cancer cells, the mice received a single i.p. injection of varying amounts of paclitaxel (range = 0.5–2.25 mg/kg). The effect of paclitaxel on the growth of SB5b cells was determined by measurements of tumor volume 6 days later. The results (Fig. 1) indicated that although tumor growth occurred at the injection site in each instance, the greatest inhibitory effect was in mice treated with the highest dose of paclitaxel tested. In addition to its direct toxic effect on the rapidly proliferating breast cancer cells, it is likely that paclitaxel increased the space for cytotoxic T cells generated in mice immunized with the vaccine. Mice with breast cancer receiving the combined therapy were treated with a single i.p. injection of 2.25 mg/kg of paclitaxel (˜90 mg/m2) before the first immunization.
The effect of paclitaxel on the white blood cell count in C3H/He mice
Paclitaxel is highly toxic. As mounting an effective immune response requires robust cell proliferation following antigen administration, peripheral white blood counts were measured at varying times after an injection of paclitaxel. The objective was to administer the vaccine when the white blood count returned at least to its preinjection value. The results indicated that 6 days after a single injection of 2.25 mg/kg paclitaxel, the white blood count had returned to preinjection levels, consistent with a full recovery from the toxic effects of the drug (Table I).
Table I. WBC Counts in C3H/HE Mice following an Injection of Paclitaxel
WBC count (106 cells/mm3)
C3H/He mice received a single i.p. injection of paclitaxel (2.25 mg/kg/mouse). At varying times afterwards, the peripheral WBC count was determined. There were 2 mice in each group.
4.8 ± 1.6
1.6 ± 0.3
1.1 ± 0.4
1.2 ± 0.4
4.7 ± 0.3
5.5 ± 0.1
6.2 ± 0.1
4.9 ± 0.4
Cytokine-secretion by LM mouse fibroblasts transduced with pZipNeoSVIL-2, a retroviral vector specifying IL-2
Among other advantages, the use of a fibroblast cell line as the recipient of DNA from the breast cancer cells enables the recipient cells to be modified in advance of DNA-transfer to augment their nonspecific immunogenic properties. In this instance, the fibroblasts were modified to secrete IL-2 and to express allogeneic MHC class I-determinants. IL-2 is a growth and maturation factor for CTLs. Allogeneic MHC class I-determinants are strong immune adjuvants.22, 23, 24
A replication-defective retroviral vector (pZipNeoSVIL-2) was used to modify the fibroblasts used as DNA-recipients to secrete IL-2. The vector specified the gene for human IL-2, along with a gene (neor), which conferred resistance to the neomycin analog, G418. (Like mouse IL-2, human IL-2 stimulates the proliferation and maturation of mouse T cells.) After selection in growth medium containing sufficient quantities of G418 to kill 100 percent of nontransduced cells, the surviving colonies were pooled and maintained as a cell line (LM-IL-2 cells). An analysis by ELISA of the culture supernatants of LM-IL-2 cells indicated that 106 retrovirally transduced cells formed 196 pg IL-2/ml/48 hr. The culture supernatants of LM fibroblasts transduced with the IL-2 negative vector pZipNeoSV(X), like that of nontransduced LM cells, failed to form detectable quantities of IL-2. Every third passage, the transduced cells were placed in medium containing 600 μg/ml G418. Under these circumstances, equivalent quantities of IL-2 were detected in the culture supernatants of cells transduced with pZipNeoSVIL-2 for more than 6 months of continuous culture. The generation time of transduced and nontransduced fibroblasts, approximately 24 hr in each instance, were equivalent. The introduction of DNA from the breast cancer cells into the IL-2-secreting cells did not affect the quantity of IL-2-secreted (these data are not presented). Toxic effects of IL-2 were not observed. Paracrine secretion of IL-2 by the vaccine avoids the toxicity of systemic cytokine administration. As the effect is in the local microenvironment of the injected vaccine, the amount secreted is likely to be in excess of that required for a maximal effect.
Modification of LM fibroblasts to express allogeneic MHC class I (H-2Kb)-determinants
H-2Kb-determinants are allogeneic in C3H/He mice (H-2k). To further augment their immunogenic properties, the cytokine-secreting fibroblasts were modified to express H-2Kb-determinants. A plasmid, pBR327H-2Kb, specifying H-2Kb-determinants was used for this purpose. LM-IL-2 cells were cotransfected with pBR327H-2Kb DNA along with pBabePuro DNA, used for selection. A 10:1 ratio of pBR327H-2Kb to pBabePuro DNA was used to ensure that the cells that incorporated pBabePuro DNA took up pBR327H-2Kb DNA as well. After selection in medium containing sufficient quantities of puromycin to kill 100 percent of nontransduced cells, the surviving colonies were pooled and maintained as cell line (LM-IL-2Kb cells).
Quantitative immunofluorescence staining with FITC-labeled mAbs for mouse H-2Kb determinants was used to measure expression of the class I-determinants. As a control, aliquots of the puromycin-resistant cell suspension were incubated with FITC-conjugated IgG2a isotype Ig. The results (Fig. 2) indicated that more than 99 percent of the transduced fibroblasts stained positively (mean fluorescence index (MFI = 6.88 ± 0.77) significantly (p < 0.01) higher than cells stained with FITC-conjugated isotype Ig (MFI = 0.37 ± 0.01), taken as background. Under similar conditions, nontransduced fibroblasts or fibroblasts incubated with FITC-conjugated isotype Ig failed to stain. The expression of H-2Kb-determinants by the transduced cells was a stable property. The staining intensity was essentially unchanged after 3 months of continuous culture (these data are not presented).
An analogous procedure was used to further characterize the cells used as DNA-recipients. The modified fibroblasts were stained with FITC-labeled mAbs for H-2Kk class I-determinants, FITC-labeled I-Ak or with FITC-labeled mAbs for the costimulatory/cell adhesion molecules B7.1, B7.2 and ICAM-1. The results indicated that 99 percent of the fibroblasts, derived from C3H/He mice, expressed H-2Kk and B7.1 determinants constitutively (mean fluorescence index 5-fold greater than the control (substitution of FITC-labeled isotype Ig for the mAbs, taken as background). Fibroblasts incubated with FITC-labeled ICAM-1, B7.2 or I-Ak mAbs failed to stain above background (substitution of FITC-labeled isotype Ig for the mAbs). The expression of MHC class I-determinants and the costimulatory molecule by LM cells was consistent with various reports indicating that fibroblasts, such as dendritic cells, are efficient antigen presenting cells.29, 30, 31, 32
A similar approach was used to determine whether the breast cancer cells (C3H/He mouse origin (H-2Kk)) used in the study expressed MHC class I-determinants. As indicated (Fig. 2), the cells stained positively with mAbs for H-2Kk-determinants. Under similar conditions, T cells from the spleens of naïve C3H/He mice expressed H-2Kk determinants. Both cell-types failed to express H-2Kb or I-A determinants.
Immunity to breast cancer in mice immunized with LM-IL-2Kb/SB5b cells
C3H/He mice are highly susceptible to the growth of SB5b cells. The injection of 1 × 105 SB5b cells into the mammary fat pad of C3H/He mice resulted in progressive tumor growth and death of the animals in 20–30 days.
To determine whether the vaccine induced immunity to breast cancer in tumor-free C3H/He mice, (inhibition of tumor growth and survival), naïve mice received a single s.c. injection of 5 × 106 LM-IL-2Kb/SB5b cells. Fourteen days later, 1 × 105 SB5b cells were injected s.c. into the left hind mammary fat pad, using a syringe equipped with a number 27-gauge needle. As a control, an equivalent number of SB5b cells were injected into the mammary fat pad of naïve C3H/He mice. To determine whether paclitaxel augmented the vaccine's therapeutic effect, the same protocol was followed except that the mice were injected with the drug 2 days before the injection of the vaccine. The results (Fig. 3a) indicated that 100 percent of the animals in the control groups injected with SB5b cells alone or with SB5b cells and paciltaxel alone died within 27 days. In contrast, mice injected with the vaccine, followed by the injection of SB5b cells survived for significantly longer periods than naïve mice in the control groups (p < 0.001). Two of 10 mice immunized with LM-IL-2Kb/SB5b cells, followed by the challenging injection of SB5b cells, appeared to have rejected the breast cancer cells. They survived indefinitely (more than 62 days). Paclitaxel alone had no significant effect. The survival of mice injected with paclitaxel, followed by immunization with LM-IL-2Kb/SB5b cells before the injection of SB5b cells, was essentially the same as that of mice injected with the vaccine alone (Fig. 3a).
Measurements of tumor growth in the preimmunized mice injected with the breast cancer cells were consistent with the vaccine's immunoprotective properties (Fig. 3b). Tumor growth was inhibited both in mice immunized with the vaccine and in mice injected with paclitaxel and the vaccine before the injection of the breast cancer cells.
To further investigate the vaccine's immunogenic properties, spleen cells from C3H/He mice immunized with LM-IL-2Kb/SB5b cells were tested in 51Cr-release cytotoxicity assays. The spleen cells were obtained from mice that had received a single s.c. injection of 5 × 106 LM-IL-2Kb/SB5b cells 14 days previously. Cells from the immunized mice were coincubated for 5 days with (mitomycin-C-treated; 50 μg/ml; 45 min.) SB5b cells. After incubation, the surviving cells were tested in a standard 51Cr-release assay for cytotoxic activity toward 51Cr-labeled SB5b cells. The results (Fig. 4a) indicated that the specific release of isotope from the labeled breast cancer cells was significantly increased, relative to that of the control, untreated group (p < 0.005). Analogous results were obtained if the spleen cells were tested in ELISPOT IFN-γ assays (Fig. 4b). The number of spots developing in spleen cell cultures from immunized mice coincubated with SB5b cells was significantly higher than that of spleen cell cultures from control nonimmunized mice (p < 0.001). With the exception of the cytotoxic response at a low effector/target ratio, an injection of paclitaxel before immunization had no significant effect upon the cytotoxicity assay. The number of spots developing in spleen cell cultures from mice injected with the vaccine were not significantly different than that of mice receiving the combined therapy. Antibody inhibition studies indicated that prior treatment of the spleen cell suspensions with CD8+ mAbs and complement (C) but not CD4+ mAbs and C significantly (p < 0.005) inhibited the antibreast cancer cytotoxicity responses (Fig. 4c). Coadministration of paclitaxel had no significant effect upon either the cytotoxicity response or the number of IFN-γ spots in the ELISPOT assays.
Thus, immunity to the breast cancer was generated in C3H/HeJ mice immunized with a vaccine prepared by transfection of modified fibroblasts with genomic DNA-fragments from the breast cancer cells. Our prior experience19, 33, 34, 35 indicated that tumor immunity failed to develop in mice immunized with nontransfected fibroblasts, or mice immunized with fibroblasts transfected with DNA from a heterologous tumor. This point is investigated further, later.
Tumor growth and survival of C3H/He mice with established breast cancer treated with a combination of paclitaxel and immunotherapy with modified fibroblasts transfected with DNA-fragments from breast cancer cells (LM-IL-2Kb/SB5b)
The therapeutic effects of paclitaxel administered in combination with the DNA-based vaccine were investigated in mice with breast neoplasms derived from SB5b cells. Cancers were first established in the mammary glands of C3H/He mice following a single injection of 1 × 105 SB5b cells. Six days later, when the average tumor was approximately 3 mm, the mice received a single s.c. injection of 2.25 mg/kg paclitaxel. Six days later, the mice received the first of 3 s.c. injections at weekly intervals of 5 × 106 LM-IL-2Kb/SB5b cells. As a control, an equivalent number of SB5b cells were injected into the mammary fat pad of C3H/He mice. As additional controls, SB5b cells were injected into the mammary fat pad and a single injection of paclitaxel was administered i.p., or paclitaxel was injected i.p., followed by the vaccine. As indicated (Fig. 5a), mice with established breast neoplasms that received the combination of paclitaxel, followed by immunization with the transfected fibroblasts survived significantly (p < 0.01) longer than mice in any of the other groups. The survival of mice with breast cancer treated with paclitaxel alone, or by immunotherapy alone, was not significantly different than that of untreated mice with breast cancer. To determine whether the therapeutic effects of the vaccine were specific, the experiment was repeated to include treatment of mice with breast cancer with a vaccine prepared by transfer of DNA-fragments from B16 melanoma cells into the modified fibroblasts (LM-IL-2Kb/B16 cells). As indicated (Fig. 5c), the survival of mice with breast cancer treated with a combination of paclitaxel and LM-IL-2Kb/B16 cells was not significantly different than that of untreated mice or mice treated with paclitaxel alone. Immunization with nontransfected modified fibroblasts (LM-IL-2Kb cells) had no significant therapeutic effect. Measurements of tumor growth in mice with breast cancer treated by the combined therapy were consistent with therapeutic outcome. Tumor growth was delayed in mice receiving the combined therapy, relative to that of mice in any of the other groups (Figs. 5b and 5d). The data are consistent with a reduction, but not elimination of the entire tumor cell population.
The results of two independent spleen cell assays designed to detect the presence of CTLs reactive with SB5b cells were consistent with the enhanced survival of mice receiving the combined therapy. Mice with breast cancer were treated according to the same protocol with paclitaxel, followed by immunization with LM-IL-2Kb/SB5b cells. Seven days after the last injection of the vaccine, spleen cells from the immunized, tumor-bearing mice were tested in 51Cr-release cytotoxicity assays. As indicated (Fig. 6a), the percent specific lysis from the group of mice receiving the combined therapy was significantly (p < 0.001) higher than that of any of the other groups including spleen cells from mice immunized with the vaccine alone or mice immunized with paclitaxel and LM-IL-2Kb/B16 cells (p < 0.001). Analogous results were obtained in ELISPOT IFN-γ assays. The highest number of spots was obtained if the spleen cells were from mice receiving the combined therapy (Fig. 6b). To determine the classes of cells mediating resistance to the tumor, mAbs for NK1.1, CD8+ or CD4+ determinants were added to the spleen cell suspensions before the 51Cr-release cytotoxicity assays were performed. As indicated (Fig. 6c), the greatest inhibitory responses were in the groups treated with NK1.1, CD8+ mAbs.
Thus, combining the administration of paclitaxel and immunotherapy with a unique DNA-based cell vaccine successfully prolonged the survival of mice with breast cancer.
In our study, the vaccine was prepared by transfer of sheared genomic DNA-fragments derived from an aggressive breast cancer into a highly immunogenic mouse fibroblast cell line. This unique approach was an application of classic studies, indicating that the introduction of high molecular weight genomic DNA from one cell type into another results in stable integration of the transferred DNA and alteration of both the genotype and the phenotype of the cells that incorporate the exogenous DNA.38, 39 Oncogenes were first discovered by this approach.40, 41 We reasoned that mutant/dysregulated genes specifying an array of undefined weakly immunogenic breast cancer antigens would be expressed in a highly immunogenic form by the transfected cells. As the transferred DNA is integrated and replicated as the recipient cells divide, the number of vaccine cells could be expanded as desired for multiple rounds of immunization. Thus, the vaccine could be prepared by transfer of microgram amounts of DNA derived from small quantities of tumor tissue, enabling treatment at an early stage of the disease. In addition, transfer of tumor-DNA into a highly immunogenic cell line enabled the recipient cells to be modified to secrete a Th1 cytokine, such as IL-2. Paracrine secretion of IL-2 by the vaccine itself avoided the toxicity often associated with systemic cytokine administration. As the effect is in the local microenvironment at the site of injection of the vaccine, it is likely that the amount secreted was in excess of that required to obtain a maximal effect.
The validity of this approach was indicated by the finding that immunization with a vaccine prepared by transfer of genomic tumor-DNA-fragments into a mouse fibroblast cell line resulted in the induction of robust T cell-mediated immunity directed toward the malignant cells. Our prior experience17 indicated that the immune response was specific and long term. The immunity was sufficient to prolong survival, although in most instances, like other immunotherapeutic approaches, tumor growth eventually recurred. Analogous findings were obtained ex vivo in patients with squamous cell carcinoma of the head and neck.42
Mouse fibroblasts, rather than dendritic cells, were chosen as the recipients of DNA from the cancer cells for several compelling reasons. Like dendritic cells, fibroblasts are efficient antigen presenting cells.29, 30, 31 As reported here, the fibroblasts expressed MHC class I-determinants and costimulatory molecules required for T cell activation. Among other advantages, the use of a fibroblast cell line enabled the cells to be modified in advance of DNA transfer to augment their immunogenic properties. In our study, in addition to modifying the cells to secrete IL-2, the fibroblasts were also modified to express a foreign (allogeneic) MHC-determinant. We made no attempt to compare the immunogenic properties of transfected fibroblasts modified to secrete various other immune augmenting cytokines, either singly or in combination or different foreign determinants. In addition to their important adjuvant properties, the presence of allogeneic MHC determinants ensured that the vaccine would be rejected. In no instance were toxic effects such as a tumor derived from the vaccine itself or the appearance of an autoimmune disease observed.
In spite of these steps, immunization of tumor-bearing mice with the DNA-based vaccine alone was unable to successfully control the growth of the highly aggressive breast cancer. The survival of mice with established breast cancer treated solely by immunization with the DNA-based vaccine was not significantly different than that of untreated mice with breast cancer. This might be expected as the first immunization was administered when the breast cancer reached a size of approximately 3 mm, the result of an injection of 1 × 105 cells 6 days previously. The animals died from pernicious tumor growth within 20–30 days, in spite of treatment with the vaccine. However, the survival of tumor-bearing mice was significantly enhanced if the animals were treated with by a combination of paclitaxel and immunization with the vaccine. Without reduction in the tumor cell population by drug therapy, the rate of tumor growth may exceed the capacity of the immune system to respond. Mice with established breast cancer receiving the combined therapy survived significantly longer than mice treated with paclitaxel alone, or by immunization alone. The immunity was specific as immunization with a vaccine prepared by transfer of DNA-fragments from a heterologous neoplasm (B16 melanoma) failed to induce immunity to breast cancer.
We considered several possible explanations. One possibility is that the injection of paclitaxel resulted in a reduction in the size of the tumor “burden,” allowing time after immunization until a sufficient number of tumor-reactive T cells could be generated. However, the survival of mice with breast cancer treated solely with paclitaxel was not significantly different than that of untreated mice with breast cancer. Another possibility is that the vaccine itself was insufficiently immunogenic, and failed to induce tumor-reactive T cells. However, the prolonged survival of mice immunized with the vaccine before tumor challenge was consistent with the vaccine's highly immunogenic properties. Another possibility is that T cell proliferation increased during the recovery phase from the toxic effects of paclitaxel. The space created by the drug may be filled with vaccination-induced proliferating T cells. As a result, greater numbers of tumor-reactive T cells were present in the immunized mice. The results of the ELISPOT and cytotoxicity assays for tumor-reactive T ells were consistent with this possibility.
Transfer of DNA from breast cancer cells into a well-characterized fibroblast cell line has important practical advantages. One advantage is that leukophoresis to isolate and culture autologous dendritic cells for each immunization is not required. DNA from the patient's neoplasm, obtained during routine care, is conveniently transfected into the fibroblasts. As the transferred DNA is integrated, and replicated as the recipient cells divide, the number of vaccine cells can be readily expanded for multiple rounds of immunization. The fact that immunization with tumor-DNA-transfected cells resulted in a robust antitumor immune response indicated that a sufficient number of transfected cells expressed relevant genes specifying therapeutic breast cancer antigens to induce the immune response. Prior reports indicate that fibroblasts, like dendritic cells, are antigen presenting cells. Kundig et al.,29 reported that fibroblast are efficient antigen presenting cells in lymphoid organs. Our own experience is consistent with this result.33 Both direct antigen presentation and indirect antigen presentation (cross priming) are involved in the induction of the antitumor immune response.
C3H/He mice were used as a source of the breast cancer cells. They may have been infected with exogenous MMTV. Conceivably, products of the virus contributed to the efficacy of the vaccine. However, our previous experience indicates that mice with melanoma and squamous carcinoma immunized tumor-DNA-transfected cells, like mice immunized with fibroblasts transfected from breast cancer cells arising in C3H mice, developed specific cell-mediated immune responses.17, 18
The data presented in this work raise the possibility that human fibroblasts ideally derived from donors that share identity with the cancer patient at one or more MHC class I alleles may be readily modified to provide immunologic specificity for breast cancer antigens expressed by the patient's own neoplasm. The technique allows the vaccine to be prepared from quite small amounts of tumor tissue, providing an opportunity to treat patients at an early stage of the disease. Immunization at an appropriate interval following chemotherapy may result in an enhanced antitumor immune response.