Open Full Page

Published OnlineFirst December 15, 2016; DOI: 10.1158/1535-7163.MCT-16-0569

Small Molecule Therapeutics Molecular Cancer Therapeutics Folate-Hapten–Mediated Immunotherapy Synergizes with Vascular Endothelial Growth Factor Receptor Inhibitors in Treating Murine Models of Cancer N. Achini Bandara1, Cody D. Bates1, Yingjuan Lu2, Emily K. Hoylman1, and Philip S. Low1

Abstract

The overexpression of folate receptors (FR) on many human immune suppression) synergize with the folate-hapten–tar- cancers has led to the development of folate-linked drugs for geted immunotherapy to reduce tumor growth in three differ- the imaging and therapy of FR-expressing cancers. In a recent ent syngeneic murine tumor models. We further demonstrate phase I clinical trial of late-stage renal cell carcinoma patients, that the combination therapy not only enhances tumor infil- þ þ folate was exploited to deliver an immunogenic hapten, fluo- tration of CD4 and CD8 effector cells, but surprisingly þ rescein, to FR tumor cells in an effort to render the cancer cells reduces tumor neovasculogenesis more than predicted. Subse- more immunogenic. Although >50% of the patients showed quent investigation of the mechanism for this unexpected prolonged stable disease, all patients eventually progressed, suppression of neovasculogenesis revealed that it is indepen- suggesting that the folate-hapten immunotherapy was insuffi- dent of elimination of any tumor cells, but instead likely þ cient by itself to treat the cancer. In an effort to identify a derives from a reduction in the numbers of FR tumor-asso- companion therapy that might augment the folate-hapten ciated macrophages and myeloid-derived suppressor cells, that immunotherapy, we explored coadministration of two is, immunosuppressive cells that release significant quantities approved cancer drugs that had been previously shown to also of VEGF. These data suggest that a reduction in stromal cells of stimulate the immune system. We report that sunitinib and myeloid origin can inhibit tumor growth by suppressing neo- axitinib (VEGF receptor inhibitors that simultaneously mitigate vasculogenesis. Mol Cancer Ther; 16(3); 461–8. 2016 AACR.

Introduction The first folate-targeted therapy to enter human clinical trials involved the delivery of a highly immunogenic hapten (fluo- Folic acid, a vitamin required for the synthesis of nucleotide rescein) to FR-overexpressing tumor cells. In this strategy, the bases, methylation of DNA, and the posttranslational modifi- patient was first vaccinated against fluorescein by immunization cation of G proteins, enters cells via either the reduced folate with keyhole limpet hemocyanin (KLH) linked to fluorescein. carrier, the proton-coupled folate transporter, or a folate recep- After establishing the presence of a high anti-fluorescein anti- tor (FR; refs. 1–4). Although the reduced folate carrier and body titer, the patient was injected with folate-fluorescein to proton-coupled folate transporter are expressed in virtually all decorate the surfaces of FR-expressing cancer cells with the cells, the FR is present and accessible primarily on activated immunogenic hapten, thereby marking the malignant cells for macrophages, the proximal tubule cells of the kidney, and elimination by the immune system (15–19). Although the certain cancers of epithelial origin, including malignancies of phase I human clinical data in advanced renal cell carcinoma the ovary, endometrium, kidney, lung, and breast (5–8). patients (>75% stage 4) with bulky disease revealed no com- Because folate-linked imaging and therapeutic agents can only plete responses and few partial responses, >50% of the patients enter cells via the FR, their uptake in normal tissues is highly showed prolonged stable disease (20, 21), suggesting some limited, allowing folate to be used as a targeting ligand to deliver degree of tumor suppression. Unfortunately, when more potent attached drugs to cancers and sites of inflammation (9–14). immune stimulants, IFNa and IL2, were later added to the immunotherapy to further stimulate immune activity, influen- za-like symptoms triggered by the cytokine treatments emerged, 1 2 Department of Chemistry, Purdue University, West Lafayette, Indiana. Endo- discouraging further clinical testing of the combination therapy cyte, Inc., West Lafayette, Indiana. (20). Nevertheless, the prolonged stable disease in the majority Note: Supplementary data for this article are available at Molecular Cancer of patients motivated further exploration of less toxic compan- Therapeutics Online (http://mct.aacrjournals.org/). ion therapies that might augment the ability of folate-fluores- þ Corresponding Author: Philip S. Low, Purdue University, 720 Clinic Drive, West cein to label and stimulate removal of FR cancers. Lafayette, IN 47907-2084. Phone: 765-494-5273; Fax: 765-494-5272; E-mail: In this paper, we examine the efficacy of combining the [email protected] aforementioned folate-fluorescein immunotherapy with inhi- doi: 10.1158/1535-7163.MCT-16-0569 bitors of vascular endothelial growth factor receptor (VEGFR). 2016 American Association for Cancer Research. Our reasons for exploring this combination lies not only in the

www.aacrjournals.org 461

Bandara et al.

observation that VEGFR inhibitors suppress unregulated neo- mice (The Jackson Laboratory). For M109 tumor implantation, vascularization (22–24), but they also enhance tumor immu- frozen cells at early passage were thawed and allowed to grow to nogenicity (25, 26). Thus, sunitinib, a VEGFR inhibitor, has not confluence. On the day of tumor injection, M109 or Renca cells only been shown to reduce tumor neovascularization, but also were collected and suspended in folate-deficient RPMI1640 medi- to decrease immunosuppressive Treg and myeloid-derived sup- um containing 1% syngeneic female Balb/c serum. Mice were pressor cell (MDSC) populations in tumor-bearing animals injected with 1 106 cells subcutaneously, and therapy was (26, 27). Because sunitinib is approved for treatment of mul- usually initiated between day 10 and 14, when tumors had tiple malignancies, we elected to examine whether a combina- reached 50–75 mm3. Similar procedures were followed for tion of a VEGFR inhibitor and the folate-hapten therapy might L1210A tumor growth, only cells were suspended in PBS imme- exceed the efficacies of either therapy alone. In this paper, we diately prior to implantation and treatment began in 7–10 days report significant synergy between the VEGFR inhibitors and when tumors had reached 50–75 mm3. Tumor growth was the folate-targeted immunotherapy, not only in their abilities monitored at 48 hours intervals by measurement of size with to suppress tumor growth, but also in their capacities to reduce laboratory calipers and tumor volume was calculated using a angiogenesis. standard two-dimensional tumor volume equation; that is, the length of the longest axis of each tumor (L) and the length of the axis perpendicular to it (W) were recorded in millimeters and the Materials and Methods corresponding tumor volume in mm3 was determined by solving Antibodies and reagents 0.5(L W2). Folic acid, KLH, carboxymethylcellulose (CMC), Tween 20, and female Balb/c serum were purchased from Sigma Aldrich. BSA Immunization with KLH-FITC conjugated to fluorescein (BSA-FITC), folate-EDA-fluorescein Animals were vaccinated with a KLH labeled with fluorescein in (Folate-FITC), and the GPI-0100 adjuvant was kindly provided order to elicit an anti-fluorescein antibody response. The by Endocyte, Inc. The structure of folate-FITC has previously been KLH–FITC conjugate was synthesized by reaction of KLH protein published (28). Sunitinib malate and axitinib free base were with excess FITC according to previously published procedures purchased from LC Laboratories. Disposable PD-10 desalting (15). The resulting KLH–FITC conjugate was purified on a PD-10 columns were obtained from GE Healthcare Bio-Sciences. The desalting column followed by dialysis in PBS at 4C. KLH-FITC biotin-conjugated goat anti-mouse IgG2a antibody and strepta- stock solutions were stored in the dark at 20C until use in vidin–HRP conjugate for ELISA antibody titer analysis was man- immunization cocktails. ufactured by Caltag Laboratories. The Shandon Cryomatrix resin Female Balb/c and DBA/2 mice were vaccinated with 35 mg was purchased from Thermo Scientific. All fluorescently labeled KLH-FITC in 50 mg GPI-0100 adjuvant formulated in sterile saline antibodies for flow cytometry and confocal microscopy were every 2 weeks for 6 weeks by subcutaneous injection alternating obtained from either BioLegend or eBioscience, Inc. The special between the base of tail and back of the neck. Blood samples from folate-deficient diet on which animals in treatment studies were immunized mice were collected by submandibular puncture one maintained was purchased from Harlan Laboratories. week after the second and third vaccinations and analyzed for anti-fluorescein titer by ELISA, as described previously (15). Cell lines and culture L1210A cells were a generous gift from Dr. Gerrit Jansen, Formulation of drugs for in vivo administration Department of Rheumatology at the University Hospital Vrije Human cancer patients generally take sunitinib or axitinib Universiteid (Amsterdam, the Netherlands) in 2012. M109 cells orally (29, 30). Therefore, in order to mimic human treatment were kindly provided by Dr. Alberto Gabizon, Hadassah-Hebrew procedures, sunitinib malate was formulated in a CMC suspen- University Medical Center (Jerusalem, Israel) in 2012. Both sion (0.5% carboxymethylcellulose, 1.8% sodium chloride, 0.4% L1210A (lymphocytic leukemia) and M109 (lung cancer) cells Tween 80, and 0.9% benzyl alcohol, pH 6) and administered to were selected for high FR expression and have been shown to mice by oral gavage. Axitinib was dosed similarly, but was maintain these elevated FR levels through multiple passages. formulated in a different CMC suspension (0.5% CMC in deio- M109 cells were harvested by digestion of solid tumor xenografts nized water acidified to pH 2–3). Both drug formulations were grown in Balb/c mice and cryopreserved at passage 0 or 1. When prepared freshly every week and stored in the dark at 4 C. The desired, cell lines were cultured in folate-deficient RPMI1640 suspensions were shaken vigorously to ensure even drug distri- medium (Invitrogen) supplemented with 10% heat-inactivated bution before dosing. Folate-FITC stock solutions supplied by FBS (Sigma Aldrich), penicillin (100 units/mL), and streptomycin Endocyte, Inc. were diluted to the desired concentration in sterile (100 mg/mL) at 37C in a humidified atmosphere containing 5% PBS, aliquoted, and stored in the dark at 20 C. Aliquots were CO2. Adherent M109 cells were passaged in a monolayer, while thawed completely on the day of treatment and mice in the L1210A cells were grown in suspension before implantation into appropriate groups were injected subcutaneously with 100 mLof syngeneic mice. Renca (renal cell carcinoma) cells were used as a the diluted conjugate. FR-negative control. Combination therapy studies Animals and tumor models The day of tumor cell implantation was designated as day 0 All procedures conducted on animals were carried out in for all experiments. Once tumors had reached approximately accordance with protocols approved by the Purdue Animal Care 50–75 mm3, mice were randomized into different treatment and Use Committee. M109 and Renca tumors were grown in 5- to groups: PBS control, folate-FITC immunotherapy, sunitinib or 7-week-old female Balb/c mice (Harlan Laboratories), whereas axitinib therapy alone, or the combination of folate-FITC plus L1210A tumors were implanted in 5- to 7-week-old female DBA/2 sunitinib or axitinib. Mice in the PBS control group were

462 Mol Cancer Ther; 16(3) March 2017 Molecular Cancer Therapeutics

Folate-Hapten Immunotherapy and VEGFR Inhibitors

injected daily with 100 mL of sterile PBS (subcutaneously) and manner as labeled slides. Confocal microscopy data were mice treated with folate-FITC alone were injected with analyzed on FLUOVIEW Viewer software. The confocal images 500 nmol/kg on a 5 days on 2 days off schedule (subcutane- of tumor sections stained with CD31 were also analyzed for ously). Mice in the sunitinib alone or axitinib alone groups vascular density by calculating the percentage of each image's were gavaged daily with 20 mg/kg sunitinib or 15 mg/kg surface that was occupied by a colored pixel. axitinib, respectively. All mice treated with the combination of folate-FITC plus sunitinib or axitinib were dosed as described Statistical analysis above for each of the individual components. Treatments were Graphing was performed using GraphPad Prism software. administered continually until the PBS control tumors reached Statistical analyses were performed using the PROC MIXED 1,000 to 1,500 mm3, at which point mice were euthanized and procedure of SAS statistical software version 9.3. Data were tumors resected for further analysis. To reduce serum folate analyzed by two-tailed ANOVA and differences were considered levels to concentrations comparable with folate levels in statistically signiﬁcant at P < 0.05. Pairwise comparisons were humans, all mice were placed on a special folate-deﬁcient diet performed using a Tukey–Kramer adjustment. 1 week following the second immunization.

Flow cytometric analysis of isolated spleen cells Results At the conclusion of each study, all animals were euthanized by Induction of anti-FITC antibody titer in mice CO2 asphyxiation, and their tumors and spleens were harvested, Mice were immunized three times at 2-week intervals with a washed in PBS, and weighed. Each tumor was then snap-frozen for KLH–FITC conjugate formulated in GPI-0100 adjuvant. Blood immunofluorescence staining. For analysis of immune cells in the was collected following both the second and third immunization, spleens, spleens were gently mashed and pressed through a 40-mm and antibody titers were assessed to ensure development of a cell strainer, after which the strainer was carefully washed with robust anti-FITC immune response. As shown in Supplementary PBS to collect any residual cells. Red blood cells were then Fig. S1, a strong anti-FITC antibody titer was induced in all removed using lysis buffer (Sigma Aldrich) and the remaining immunized groups, and subsequent titer analysis at completion splenocytes were suspended in labeling buffer (PBS containing of each study further established maintenance of the titer during 1% BSA). Fc receptors were blocked using a commercially avail- the treatment period. able Fc blocker (BD Biosciences) to reduce nonspecific labeling. Blocked cells were then incubated for 1 hour at 4C with the Sunitinib synergizes with folate-hapten–mediated following antibody combinations: APC or PE-conjugated anti- immunotherapy in L1210A tumor-bearing mice mouse F4/80 for macrophages; PE-conjugated anti-mouse CD3 VEGFR inhibitors have been observed to not only down- and FITC-conjugated anti-mouse CD4 or FITC-conjugated anti- þ þ regulate angiogenesis, but also inhibit the immunosuppressive mouse CD8 for CD4 and CD8 T cells, respectively; and APC- activities of both MDSCs and regulatory T cells (T ; ref. 26). conjugated anti-mouse CD11b and PE-conjugated anti-mouse regs To test whether these immunomodulatory properties might GR-1 for MDSCs. Following labeling, cells were washed with cold augment the potency of our receptor-targeted hapten immu- labeling buffer and analyzed on a Becton Dickinson FACSCalibur notherapy, it was necessary to grow FR-expressing tumors in instrument using CellQuest software. At least 1 105 cells were immunocompetent mice. For this purpose, L1210A tumors counted from each sample. All flow cytometry data were analyzed were implanted subcutaneously in DBA/2 mice and the mice on FlowJo software. were treated with either sunitinib, folate-hapten immunotherapy, or a combination of the two. As seen in Fig. 1, the PBS- Cryopreservation and confocal imaging of tumors treatedDBA/2micegrewtumorsthatexpandedto1,000mm3 Each resected tumor was embedded in Shandon Cryomatrix in <3 weeks. Surprisingly, neither the moderate doses of suni- resin and snap frozen by partial submersion in liquid nitrogen. tinib nor folate-targeted immunotherapy used in this study The frozen tumor tissues were then sectioned (10 mm) using a exerted a significant effect on this rapid tumor expansion. In Shandon SME Cryotome cryostat (GMI, Inc.) and mounted on contrast, a combination of the same doses of sunitinib- and to polylysine-coated Superfrost microscopy slides (Thermo folate-targeted immunotherapy caused a decrease in tumor Fisher Scientific). For confocal imaging, the slides were allowed volume (Fig. 1A). These data imply that some type of synergy to equilibrate at room temperature for approximately 1 hour must exist between the tumor-targeted hapten therapy and before fixation and permeabilization in ice-cold acetone VEGFR inhibitor therapy. (10 minutes). After washing in PBS, the tissue sections were Because L1210A cells derive from a murine lymphocytic leu- incubated in PBS-Tween containing 1% BSA to block any kemia that commonly metastasizes to the spleen, it was also nonspecific binding followed by washing in PBS (2 2min). instructive to examine whether the spleen weights of each treat- Individual slides were then incubated with a 1:50 dilution of ment group might differ as an indication of tumor cell metastasis AlexaFluor 647 conjugated rat anti-mouse CD31 antibody (31). As shown in Fig. 1B, spleens became similarly enlarged in the (BioLegend) in a dark humidified chamber for 1 hour at room PBS-, sunitinib-, and folate-FITC–treated animals, but were rela- temperature. The stained slides were washed (3 5 minutes in tively small in the combination therapy animals. Isolation of PBS), dried, and mounted with cover slips before examination extractable cells from these spleens demonstrated that their under an Olympus FV1000 inverted confocal laser scanning increase in size derived exclusively from accumulation of meta- microscope using a Plan Apo 10 objective. All fluorescence static L1210A cells. These data suggest that metastasis of tumor images were obtained under identical conditions and unla- cells to the spleen is suppressed by the combination therapy but beledtissuesectionsusedascontrolsweretreatedinthesame not by either individual therapy.

www.aacrjournals.org Mol Cancer Ther; 16(3) March 2017 463

Bandara et al.

A L1210A with axitinib plus folate-FITC, however, displayed signiﬁcantly 1,750 P <

) PBS Control smaller tumor volumes (23% of controls, 0.05). 3 1,500 Folate-FITC 1,250 Sunitinib Immune effector and suppressor cell levels are altered in 1,000 Folate-FITC + Sunitinib combination therapy–treated mice 750 Studies reported in the literature propose several mechanisms 500 for the known abilities of angiogenesis inhibitors to downregulate 250 immunosuppressive MDSCs and Tregs (22, 25, 26). To determine Tumor volume (mm Tumor 0 whether the observed augmentation of folate-FITC immunother- 7 9 11 13 15 17 19 21 apy might arise from a reduction in either of these immune cell Days post tumor implantation types, spleen cells were isolated from all mice in the above studies B 0.55 and analyzed for different immune cell populations by ﬂow 0.50 cytometry. Although no differences in Treg numbers could be 0.45 þ 0.40 detected among treatment groups (data not shown), CD4 and þ 0.35 CD8 T cells were found to be elevated in the combination 0.30 therapy groups in both L1210A and M109 tumor models 0.25 0.20 (Fig. 4A and B). Moreover, MDSC numbers were reduced in the 0.15 spleens of M109 tumor-bearing mice treated with the combina- Spleen weight (g) 0.10 tion therapy (Fig. 4C), although no difference could be observed 0.05 0.00 in the L1210A model. Because immunosuppressive T cells con- PBS Control Folate-FITC Sunitinib Combination stitute only a fraction of the total T-cell population (usually <3%; þ þ ref. 32), the signiﬁcant increase in both CD4 and CD8 T cells Figure 1. A, Effect of folate-FITC (500 nmol/L/kg, 5 days/week) and/or sunitinib (20 mg/kg daily) therapy on the growth of L1210A tumors in DBA/2 mice. Treatments started on day 7 when tumors reached 50–75 mm3. B, Average spleen weight. Data are mean SEM (n ¼ 5–9).

Sunitinib synergizes with folate-hapten–mediated immunotherapy in the M109 tumor-bearing mice To determine whether the above hypothesized synergy between the folate-hapten therapy and anti-VEGFR therapy þ mightalsoapplytoothertumortypes,asecondFR tumor was similarly examined in syngeneic mice. As seen in Fig. 2, M109 tumor-bearing mice treated with either sunitinib or folate-FITC alone responded only mildly to the individual therapy, resulting in moderately retarded tumor growth. How- ever, mice treated with both folate-FITC and sunitinib again maintained a signiﬁcantly slower tumor growth rate than either monotherapy group alone. Combination therapy resulted in a 77% reduction in tumor volume and a 55% reduction in tumor weight compared with PBS-treated controls (P < 0.05), which supports the hypothesis that a combination of folate-hapten plus anti-VEGFR therapy results in augmentation of each drug's individual efﬁcacy.

Axitinib synergizes with folate-hapten–mediated immunotherapy in L1210A tumor-bearing mice To determine whether other angiogenesis inhibitors might similarly synergize with the folate-hapten immunotherapy, we explored the use of a more selective VEGFR inhibitor (axitinib) than sunitinib in combination with our FR-targeted immunotherapy. For this purpose, L1210A tumors were again implanted in DBA/2 mice and the affected mice were treated with axitinib (15 mg/kg), folate-targeted immunotherapy, or the combination of the two. As shown in Fig. 3, axitinib, like sunitinib, combined with folate-FITC immunotherapy to Figure 2. A, Effect of folate-FITC (500 nmol/kg, 5 days/week) and/or sunitinib (20 mg/kg signiﬁcantly slow tumor growth and prolong survival of all daily) therapy on the growth of M109 tumors in Balb/c mice. B, Average weights treated mice. The average tumor volume in the axitinib alone of excised tumors from M109 tumor-bearing Balb/c mice dosed with the treated group at the end of the study (day 20) was 53% of the indicated therapies. Data are mean SEM (n ¼ 5–7). Means that do not average tumor volume of the PBS-treated mice. Mice treated share a letter are signiﬁcantly different, P < 0.05.

464 Mol Cancer Ther; 16(3) March 2017 Molecular Cancer Therapeutics

Folate-Hapten Immunotherapy and VEGFR Inhibitors

argues that the combination therapy augments the cellular immune response against the tumor.

Synergistic inhibition of tumor vascular growth is observed in combination therapy–treated mice Because a proposed mechanism of antitumor activity of angiogenesis inhibitors derives from their retardation of neovascularization, we anticipated a reduction in tumor blood vessel density in sunitinib- and axitinib-treated animals, but no signiﬁcant diminution in other treatment groups. To ascertain the extent of inhibition of neovasculogenesis, we stained tumor sections from mice in each of the treatment groups with an antibody to CD31 (i.e., an established endothelial cell marker), and examined the vascular density/distribution by confocal microscopy. In contrast Figure 3. to expectations, sunitinib alone had little effect on blood vessel Effect of folate-FITC and/or axitinib therapy on the growth of L1210A tumors density in either of the tumor models, whereas the combination in DBA/2 mice. Data are mean SEM (n ¼ 6). Means that do not share therapy showed a dramatic reduction in CD31 staining in both a letter are signiﬁcantly different, P < 0.05. L1210A and M109 tumors (Fig. 5, top and center). Indeed,

Figure 4. A, CD4þ and CD8þ T-cell populations in the spleens of L1210A tumor-bearing DBA/2 mice and (B) M109 tumor-bearing Balb/c mice. Isolated splenocytes stained with florescent antibodies against CD3, CD4, and CD8, were analyzed by flow cytometry. C, Percent of MDSC found in the spleens isolated from M109 tumor- bearing mice dosed with the indicated therapies. Isolated splenocytes labeled with fluorescent antibodies against CD11b and GR-1 were analyzed by flow cytometry. A total of 1 105 cells were counted from each sample. Data are mean SEM (n ¼ 5–7). Means that do not share a letter are significantly different, P < 0.05 (#P ¼ 0.06).

www.aacrjournals.org Mol Cancer Ther; 16(3) March 2017 465

Bandara et al.

Figure 5. Fluorescent staining and imaging of the vasculature within tumors harvested from treated mice. Top row shows images obtained from solid L1210A tumors treated with PBS, folate-FITC, sunitinib, or a combination of folate-FITC and sunitinib. The center row represents images obtained from solid M109 tumors treated with PBS, folate-FITC, sunitinib, or a combination of folate-FITC and sunitinib. The bottom row shows images collected from stained L1210A tumors collected from mice treated with PBS, folate-FITC, axitinib, or a combination of folate-FITC and axitinib. Frozen tumors were sectioned and stained with an anti-CD31 antibody labeled with AlexaFluor 647 and imaged under an Olympus FV1000 confocal microscope using a 10 objective. Error bars, SEM. Means that do not share a letter are signiﬁcantly different, P < 0.05.

quantitative analysis of CD31 staining in each image showed the because all Renca tumor cells are FR-negative. As shown in Fig. 6A, vascular density of combination treated tumors to be approxi- inhibition of tumor growth was minimal in both sunitinib and mately half that of tumors treated with either folate-FITC immu- folate-hapten monotherapies. In contrast, the combination ther- notherapy or sunitinib alone. These data indicate that folate- apy caused a dramatic decrease in tumor growth, suggesting that hapten therapy synergizes with VEGFR inhibitors to augment the mechanism of synergy does not rely on any direct effect of downregulation of the tumor vasculature. However, as initially folate-FITC on the cancer cells. Importantly, anti-CD31-stained anticipated, tumors treated solely with axitinib showed a signif- sections of the same Renca tumors revealed that their vascular icant reduction in tumor vasculature staining (57% of control), density was lower in the combination than monotherapy-treated and the combination of axitinib plus immunotherapy further animals (Fig. 6B). These data further argue that suppression of decreased this vascular density to 37% of the PBS-treated control. neovascularization does not depend on prior destruction of the Representative images of anti-CD31-stained tumors derived from cancer cells, but rather that a reduction in the vasculature can each treatment group as well as the corresponding graphs of a independently inhibit tumor growth. Based on all of the above computer analysis of the vascular density are shown in Fig. 5. data, it is not unreasonable to posit that downregulation of The unanticipated observation that folate-hapten immuno- neovasculogenesis may constitute the primary mechanism of therapy can augment the ability of VEGFR inhibitors to suppress tumor growth inhibition in the combination therapy group. tumor neovascularization raised the question of whether inhibition of neovasculogenesis might constitute the predominant mechanism accounting for their synergistic effect on tumor Discussion growth. To test this hypothesis, FR-negative Renca tumors were As noted above, the folate-hapten–targeted immunotherapy implanted in Balb/c mice and treated with folate-hapten immu- showed signiﬁcant promise in human clinical trials, but failed to notherapy, sunitinib, or a combination of the two. In this tumor yield complete responses probably due to inadequate activation model, any impact of the folate-targeted immunotherapy would of the immune system (21). Attempts to augment the immu- þ have to be exerted on the FR tumor-inﬁltrating macrophages, notherapy's potency by coadministration of cytokines also failed

466 Mol Cancer Ther; 16(3) March 2017 Molecular Cancer Therapeutics

Folate-Hapten Immunotherapy and VEGFR Inhibitors

Figure 6. A, Effect of folate-FITC and/or sunitinib therapy on the weight of FR Renca tumors in Balb/c mice (n ¼ 2–3). Individual data points and means (horizontal lines) are shown. B, Quantification of CD31 fluorescent staining and imaging of the vasculature within Renca tumors harvested from mice treated with PBS, folate-FITC, sunitinib, or a combination of folate-FITC and sunitinib. Data are from five tumor sections. Individual data points and means (horizontal lines) are shown. C, Representative images of CD31 staining for above tumors. Frozen tumors were sectioned and stained with an anti-CD31 antibody labeled with AlexaFluor 647 and imaged under an Olympus FV1000 confocal microscope using a 10 objective.

due to unacceptable toxicities (20). In an effort to identify a less inhibitors. Not only can the targeted therapy lead to antibody þ toxic and more effective companion therapy, we explored coad- opsonization and immune cell–mediated destruction of FR þ ministration of VEGFR inhibitors which are known to activate cancer cells (16–18), but the concurrent elimination of FR immune effector cells (22, 33, 34). We found a strong synergy tumor-associated macrophages and MDSCs could reduce both between the VEGFR inhibitors and the folate-hapten immuno- the immunosuppressive microenvironment and growth of new therapy, where the combination therapy significantly augmented blood vessels within the tumor. Because many human cancer suppression of tumor vascularization, reduction of MDSCs, and cells do not express FR, but virtually all human tumors contain þ þ þ recruitment of CD4 and CD8 T cells. large numbers of FR myeloid cells (35, 40, 41), the combi- One of the more interesting results from this study was the nation therapy described here could conceivably prove useful observed synergy in suppression of tumor neovasculogenesis. in controlling many human tumor types. Although reduction in blood vessel formation by the two VEGFR inhibitors (i.e., sunitinib and axitinib) was anticipated, fl the strong augmentation of this reduction by the FR-targeted Disclosure of Potential Con icts of Interest fi fi immunotherapy was not. Contemplation of plausible mechan- P.S. Low is a Chief Scienti cOf cer in Endocyte, Inc. and also reports receiving a commercial research grant from Endocyte, Inc. No potential isms raised the possibility that the immunotherapy-mediated conflicts of interest were disclosed by the other authors. elimination of tumor-associated myeloid cells [i.e., tumor- associated macrophages (TAM) and MDSCs] will simultaneously reduce the levels of growth factors that they produce. Authors' Contributions Thus, recent examination of FR beta expression in a large Conception and design: N.A. Bandara, Y. Lu, P.S. Low number of human tumor samples reveals that virtually all Development of methodology: N.A. Bandara, Y. Lu þ tumor types are characterized by significant infiltration of FR Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): N.A. Bandara, C.D. Bates, E.K. Hoylman TAMs and MDSCs (35). Because these tumor-associated mye- Analysis and interpretation of data (e.g., statistical analysis, biostatistics, loid cells are known to release VEGF, FGF, and many other computational analysis): N.A. Bandara, C.D. Bates angiogenic factors (36–39), immune-mediated elimination of Writing, review, and/or revision of the manuscript: N.A. Bandara, P.S. Low these TAMs/MDSCs should significantly reduce the tumor's Administrative, technical, or material support (i.e., reporting or organizing supply of cytokines that impact tumor angiogenesis. Together data, constructing databases): E.K. Hoylman with the blockade of VEGF receptor signaling by sunitinib and Study supervision: P.S. Low axitinib (22, 24, 29), the observed synergy of the combination therapy in suppressing vasculogenesis should probably have Acknowledgments been predicted. If the same reduction in vascular density can We thank Endocyte, Inc. for providing folate-FITC (EC17) and lower the probability of tumor metastasis, then the mechanism GPI-0100. We are grateful to Dr. Leroy Wheeler for advice during design underpinning the observed synergy in the combination ther- of flow cytometry experiments and to Dr. Ahmad Gheith for assistance apy's ability to eliminate L1210A cell metastasis to the spleen in interpretation of confocal images of CD31-stained tumor tissues. fl might also be explained. The authors gratefully acknowledge the ow cytometry resources from the Purdue University Center for Cancer Research, NIH grant P30 Taken together, our data suggest that multiple mechanisms CA023168, and the imaging facilities of the Bindley Bioscience Center, may contribute to the efficacy of the combination therapy a core facility of the NIH-funded Indiana Clinical and Translational comprised of the folate-hapten immunotherapy plus VEGFR Sciences Institute.

www.aacrjournals.org Mol Cancer Ther; 16(3) March 2017 467

Bandara et al.

Grant Support advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate This work was supported in part by a grant from Endocyte, Inc. (awarded to this fact. P.S. Low) and NIH grant P30 CA023168. The costs of publication of this article were defrayed in part by the Received August 25, 2016; revised December 9, 2016; accepted December 9, payment of page charges. This article must therefore be hereby marked 2016; published OnlineFirst December 15, 2016.

References 1. Antony A. The biological chemistry of folate receptors. Blood 1992; 23. Goodman VL, Rock EP, Dagher R, Ramchandani RP, Abraham S, Gobburu 79:2807–20. JV, et al. Approval summary: sunitinib for the treatment of imatinib 2. Matherly LH, Goldman ID. Membrane transport of folates. Vitam Horm refractory or intolerant gastrointestinal stromal tumors and advanced renal 2003;66:403–56. cell carcinoma. Clin Cancer Res 2007;13:1367–73. 3. Lucock M. Folic acid: nutritional biochemistry, molecular biology, and role 24. Hu-Lowe DD, Zou HY, Grazzini ML, Hallin ME, Wickman GR, Amundson in disease processes. Mol Genet Metab 2000;71:121–38. K, et al. Nonclinical antiangiogenesis and antitumor activities of axitinib 4. Kim Y-I. Nutritional epigenetics: impact of folate deficiency on DNA (AG-013736), an oral, potent, and selective inhibitor of vascular endo- methylation and colon cancer susceptibility. J Nutr 2005;135:2703–9. thelial growth factor receptor tyrosine kinases 1, 2, 3. Clin Cancer Res 5. Elnakat H, Ratnam M. Distribution, functionality and gene regulation of 2008;14:7272–83. folate receptor isoforms: implications in targeted therapy. Adv Drug Deliv 25. Ozao-Choy J, Ma G, Kao J, Wang GX, Meseck M, Sung M, et al. The novel Rev 2004;56:1067–84. role of tyrosine kinase inhibitor in the reversal of immune suppression and 6. Parker N, Turk MJ, Westrick E, Lewis JD, Low PS, Leamon CP. Folate modulation of tumor microenvironment for immune-based cancer ther- receptor expression in carcinomas and normal tissues determined by a apies. Cancer Res 2009;69:2514–22. quantitative radioligand binding assay. Anal Biochem 2005;338:284–93. 26. Xin H, Zhang C, Herrmann A, Du Y, Figlin R, Yu H. Sunitinib inhibition of 7. Weitman SD, Lark RH, Coney LR, Fort DW, Frasca V, Zurawski VR, et al. Stat3 induces renal cell carcinoma tumor cell apoptosis and reduces Distribution of the folate receptor GP38 in normal and malignant cell lines immunosuppressive cells. Cancer Res 2009;69:2506–13. and tissues. Cancer Res 1992;52:3396–401. 27. Draghiciu O, Nijman HW, Hoogeboom BN, Meijerhof T, Daemen T. 8. Weitman SD, Weinberg AG, Coney LR, Zurawski VR, Jennings DS, Kamen Sunitinib depletes myeloid-derived suppressor cells and synergizes with BA. Cellular localization of the folate receptor: potential role in drug a cancer vaccine to enhance antigen-specific immune responses and tumor toxicity and folate homeostasis. Cancer Res 1992;52:6708–11. eradication. Oncoimmunology 2015;4:e989764. 9. Ayala-Lopez W, Xia W, Varghese B, Low PS. Imaging of atherosclerosis in 28. De Jesus E, Keating JJ, Kularatne SA, Jiang J, Judy R, Predina J, et al. apoliprotein e knockout mice: targeting of a folate-conjugated radiophar- Comparison of folate receptor targeted optical contrast agents for intrao- maceutical to activated macrophages. J Nucl Med 2010;51:768–74. perative molecular imaging. Int J Mol Imaging 2015;2015:469047. 10. Turk MJ, Breur GJ, Widmer WR, Paulos CM, Xu LC, Grote LA, et al. 29. Motzer RJ, Rini BI, Bukowski RM, Curti BD, George DJ, Hudes GR, et al. Folatetargeted imaging of activated macrophages in rats with adju- Sunitinib in patients with metastatic renal cell carcinoma. JAMA 2006; vantinduced arthritis. Arthritis Rheum 2002;46:1947–55. 295:2516–24. 11. Wang S, Luo J, Lantrip DA, Waters DJ, Mathias CJ, Green MA, et al. Design 30. Rini BI, Wilding G, Hudes G, Stadler WM, Kim S, Tarazi J, et al. Phase II and synthesis of [111In] DTPA-folate for use as a tumor-targeted radio- study of axitinib in sorafenib-refractory metastatic renal cell carcinoma. pharmaceutical. Bioconjug Chem 1997;8:673–9. J Clin Oncol 2009;27:4462–8. 12. Reddy JA, Dorton R, Dawson A, Vetzel M, Parker N, Nicoson JS, et al. In vivo 31. Hofer KG, Prensky W, Hughes WL. Death and metastatic distribution of structural activity and optimization studies of folate tubulysin conju- tumor cells in mice monitored with 125I-iododeoxyuridine. J Natl Cancer gates. Mol Pharm 2009;6:1518–25. Inst 1969;43:763–73. 13. Reddy JA, Low PS. Folate-mediated targeting of therapeutic and imaging 32. Caramalho I, Lopes-Carvalho T, Ostler D, Zelenay S, Haury M, Demengeot agents to cancers. Crit Rev Ther Drug Carrier Syst 1998;15:587–627. J. Regulatory T cells selectively express toll-like receptors and are activated 14. Low PS, Henne WA, Doorneweerd DD. Discovery and development of by lipopolysaccharide. J Exp Med 2003;197:403–11. folic-acid-based receptor targeting for imaging and therapy of cancer and 33. Bose A, Taylor JL, Alber S, Watkins SC, Garcia JA, Rini BI, et al. Sunitinib inflammatory diseases. Acc Chem Res 2007;41:120–9. facilitates the activation and recruitment of therapeutic antitumor immu- 15. Lu Y, Low PS. Folate targeting of haptens to cancer cell surfaces mediates nity in concert with specific vaccination. Int J Cancer 2011;129:2158–70. immunotherapy of syngeneic murine tumors. Cancer Immunol Immun- 34. Huang Y, Yuan J, Righi E, Kamoun WS, Ancukiewicz M, Nezivar J, et al. other 2002;51:153–62. Vascular normalizing doses of antiangiogenic treatment reprogram the 16. Lu Y, Low PS. Immunotherapy of folate receptor-expressing tumors: review immunosuppressive tumor microenvironment and enhance immunother- of recent advances and future prospects. J Control Release 2003;91:17–29. apy. Proc Natl Acad Sci 2012;109:17561–6. 17. Lu Y, Sega E, Leamon CP, Low PS. Folate receptor-targeted immunotherapy 35. Shen J, Putt KS, Visscher DW, Murphy L, Cohen C, Singhal S, et al. of cancer: mechanism and therapeutic potential. Adv Drug Deliv Rev Assessment of folate receptor-b expression in human neoplastic tissues. 2004;56:1161–76. Oncotarget 2015;6:14700–09. 18. Lu Y, Sega E, Low PS. Folate receptortargeted immunotherapy: induction of 36. Lewis C, Leek R, Harris A, McGee J. Cytokine regulation of angiogenesis in humoral and cellular immunity against haptendecorated cancer cells. Int J breast cancer: the role of tumor-associated macrophages. J Leukoc Biol Cancer 2005;116:710–9. 1995;57:747–51. 19. Lu Y, Xu L-C, Parker N, Westrick E, Reddy JA, Vetzel M, et al. Preclinical 37. Murdoch C, Muthana M, Coffelt SB, Lewis CE. The role of myeloid cells in pharmacokinetics, tissue distribution, and antitumor activity of a folate- the promotion of tumour angiogenesis. Nat Rev Cancer 2008;8:618–31. hapten conjugate–targeted immunotherapy in hapten-immunized mice. 38. Pepper MS, Mandriota SJ, Montesano R. Angiogenesis-regulating cyto- Mol Cancer Ther 2006;5:3258–67. kines. In: Fan T-PD, Kohn EC, editors. The new angiotherapy. Totowa, NJ: 20. Amato RJ, Shetty A, Lu Y, Ellis PR, Mohlere V, Carnahan N, et al. A phase I/Ib Humana Press; 2002. p. 7–40. study of folate immune (EC90 vaccine administered with GPI-0100 39. Yang L, DeBusk LM, Fukuda K, Fingleton B, Green-Jarvis B, Shyr Y, et al. adjuvant followed by EC17) with interferon-a and interleukin-2 in patients Expansion of myeloid immune suppressor GrþCD11bþ cells in tumor- with renal cell carcinoma. J Immunother 2014;37:237–44. bearing host directly promotes tumor angiogenesis. Cancer Cell 2004; 21. Amato RJ, Shetty A, Lu Y, Ellis R, Low PS. A phase I study of folate 6:409–21. immune therapy (EC90 vaccine administered with GPI-0100 adjuvant 40. Cortez-Retamozo V, Etzrodt M, Newton A, Rauch PJ, Chudnovskiy A, followed by EC17) in patients with renal cell carcinoma. J Immunother Berger C, et al. Origins of tumor-associated macrophages and neutrophils. 2013;36:268–75. Proc Natl Acad Sci 2012;109:2491–6. 22. Ellis LM, Hicklin DJ. VEGF-targeted therapy: mechanisms of anti-tumour 41. Noy R, Pollard JW. Tumor-associated macrophages: from mechanisms to activity. Nat Rev Cancer 2008;8:579–91. therapy. Immunity 2014;41:49–61.

468 Mol Cancer Ther; 16(3) March 2017 Molecular Cancer Therapeutics

Folate-Hapten−Mediated Immunotherapy Synergizes with Vascular Endothelial Growth Factor Receptor Inhibitors in Treating Murine Models of Cancer

N. Achini Bandara, Cody D. Bates, Yingjuan Lu, et al.

Mol Cancer Ther 2017;16:461-468. Published OnlineFirst December 15, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-16-0569

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2016/12/14/1535-7163.MCT-16-0569.DC1

Cited articles This article cites 40 articles, 14 of which you can access for free at: http://mct.aacrjournals.org/content/16/3/461.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/16/3/461.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/16/3/461. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.