Background
Cancer vaccines targeting well-established tumor antigens have demonstrated modest activity in clinical trials performed in the era predating effective immune checkpoint blockade. Even with more potent vaccine strategies, tumor escape may occur due to downregulation or loss of targeted antigens, as such antigens, not critical for tumor survival and proliferation, may be subject to immune editing without affecting the malignant phenotype [
1]. In contrast, targeting “driver” antigens that are critical components of cellular proliferation, survival, or resistance mechanisms is an attractive strategy, as these “driver” antigens cannot be downregulated or lost due to their requirement for maintenance of the malignant phenotype. Nonetheless, the adaptive immune response against chronically overexpressed tumor antigens is often minimized or diminished due to immune tolerance and/or immunoregulation [
2]. We hypothesize that a novel therapeutic strategy would be to target proteins associated with the malignant phenotype or acquired therapeutic resistance that are initially sequestered from the immune system but may become upregulated upon the initiation of therapy or tumor progression. One such upregulated mediator of therapeutic resistance is the human epidermal growth factor receptor (HER) family member HER3, associated with poor prognosis in several epithelial malignancies including breast cancer.
Although having reduced catalytic kinase activity [
1‐
4], HER3 is thought to function as a signaling substrate for other HER proteins with which it heterodimerizes [
5] thus promoting tumor proliferation and survival [
6]. Importantly, it is a co-receptor for epidermal growth factor receptor (EGFR) and HER2 with which it is synergistically co-transforming [
7] and rate-limiting for transformed growth [
8]. Treatment of HER2-amplified breast cancers with HER2-targeting tyrosine kinase inhibitors (TKIs) promotes an increase in HER3 plasma membrane localization and downstream signaling, which can lead to resistance to the HER2-targeted therapies [
9‐
11]. HER3 expression has been associated with poor clinical outcomes including central nervous system (CNS) metastasis in both the triple negative (TNBC) and HER2 subtypes of breast cancer [
12,
13].
The pivotal role of HER3 as a hub for HER family signaling has made it an attractive therapeutic target, but its reduced kinase activity has limited the development of small molecule inhibitors. One proven method has been to disrupt the HER2-HER3 heterodimer formation. The HER2-specific monoclonal antibody pertuzumab effectively disrupts heregulin-induced HER2-HER3 dimerization and signaling [
14] and has proven clinical benefit. Nonetheless, it is less effective at disrupting the elevated basal state of ligand-independent HER2-HER3 interaction and signaling in HER2-overexpressing tumor cells [
15]. Alternatively, HER3 may be targeted directly, specifically with antibodies having diverse functional consequences depending on their binding site [
16]. For example, HER3-specific monoclonal antibodies inhibit ligand-induced activation of the receptor [
17], inhibiting downstream signaling; however, none are currently commercially available.
As an alternative to monoclonal antibodies, we and others have demonstrated that polyclonal antibodies induced by vaccination against receptors such as HER2 can recognize the cell-expressed receptor, suppress its phosphorylation, mediate profound receptor internalization and degradation, and retard the growth of established receptor-dependent tumor xenografts [
18,
19]. Further, rather than repeated administration of antibodies, vaccination induces long-term anti-tumor immune responses that can be periodically boosted. Therefore, we sought to generate a vaccine capable of inducing potent anti-HER3 antibody responses.
We recently reported generation of a recombinant adenoviral vector expressing human HER3 (Ad-HER3) and demonstrated that it induced HER3-specific T cell responses and had antitumor activity [
20]. However, we are aware tumor antigen-specific T cell responses may be ineffective in some patients with advanced malignancies expressing the target tumor antigen and sought evidence of alternative antitumor mechanisms. Therefore, we examined whether Ad-HER3 vaccine-induced multifunctional antibody responses, including complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC), may mediate antitumor responses. In addition to the expected immune-mediated functions, we also wished to demonstrate whether serum containing anti-HER3 antibody (subsequently referred to as HER3-VIA) could have a direct effect on HER3 biology, specifically mediating HER3 internalization and degradation, as well as inhibiting the downstream signaling of HER3 heterodimers. Finally, we sought to demonstrate in vivo that HER3-specific polyclonal anti-HER3 serum alone, when transferred to tumor-bearing animals, retards growth of both HER2 therapy-resistant tumors and TNBC.
Methods
Cell lines and cell culture
The human breast cancer cell lines BT474, MCF-7, MDA-MB-231, MDA-MB-468, SKBR3, and T47D (obtained from the American Type Culture Collection (ATCC), Manassas, VA, USA) were grown in the recommended medium. The BT474M1 human breast tumor cell line (kind gift from Dr. Mien-Chie Hung at The University of Texas M. D. Anderson Cancer Center, Houston, TX, USA) was grown in DMEM/F12 with 10% FBS. Lapatinib-resistant BT474 (rBT474) was generated as previously described [
21]. Frozen stocks of these cell lines were made at earlier passages, and after thawing, cells were cultured no longer than 8 weeks for the experiments. Cells were authenticated by morphology and growth curve analysis and were routinely tested for the absence of mycoplasma by PCR. All mycoplasma tests performed during this study were negative. For tumor challenge, cell lines were tested for rodent pathogens (IMPACT Profile III) and proven to be negative before the injection to mice.
Reagents
Trastuzumab (Herceptin™, Genentech, San Francisco, CA, USA) and cetuximab (Erbitux®, Bristol-Myers Squibb, New York, NY, USA) were purchased from the Duke University Medical Center Pharmacy. Lapatinib was purchased from Sigma-Aldrich (CDS022971, St. Louis, MO, USA). Heregulin (377-HB/CF), heregulin with a C-terminal 6-His tag (5898-NR) and allophycocyanin (APC)-conjugated anti-His Tag antibody (IC050A) were purchased from R&D Systems (Minneapolis, MN, USA).
Adenovirus vector preparation
The human HER3 complementary DNA (cDNA) was excised from a pCMVSport6-HER3-HsIMAGE6147464 plasmid (cDNA clone MGC:88033/IMAGE:6147464 obtained from ATCC). Construction of a first-generation (E1-, E3-) Ad vector containing human full length HER3 under control of human cytomegalovirus (CMV) promoter/enhancer elements was performed using the pAdEasy system (Agilent technologies, Santa Clara, CA, USA) as previously described [
22‐
24]. Similar Ad-vectors containing the green fluorescence protein (GFP) or lacZ rather than HER3 was similarly generated to serve as controls.
Mice
BALB/c and NOD.CB17-Prkdcscid/J mice were purchased from Jackson Labs (Bar Harbor, ME, USA). All mice were maintained under specific pathogen-free conditions, and all work was conducted in accordance with Duke Institutional Animal Care and Use Committee (IACUC)-approved protocols.
Production of vaccine-induced antibodies (VIA)
BALB/c mice were vaccinated on day 0 and day 14 by footpad injection of Ad-GFP (control), or Ad-HER3 vectors (2.6 × 1010 particles/ mouse). At 14 days after the second vaccination, mice were euthanized and serum was collected, pooled from every 20 vaccinated mice, and 1-mL aliquots were made and stored at − 80 °C until use. Approximately 80 mice for HER3-VIA and 80 mice for control VIA were vaccinated to collect and pool serum (24 mL for both VIAs) for this study.
Cell-based ELISA
The 4 T1 cells were transduced with the HER3 gene by lentiviral vectors (4 T1-HER3 cell): 4 T1 and 4 T1-HER3 cells were incubated overnight at 37 °C in 96-well plates (3 × 104 cells/well). Mouse serum (HER3-VIA, LacZ-VIA, GFP-VIA) was diluted (final titrations 1:50 ~ 1:6400), added to the wells (50 μL/well), and incubated for 1 h on ice. The plates were washed with PBS twice, and then cells were fixed with diluted formalin (1:10 dilution).Then, near infrared (nIR) dye-conjugated anti-mouse IgG (IRDye 800CW, LI-COR Biosciences, Lincoln, NE, USA) was added (1:2000 dilution, 30 min, room temperature). After washing with PBS, the nIR signal was detected by a LI-COR Odyssey Imager (LI-COR) at 800 nm channel.
Analysis of anti-HER3 antibody binding by flow cytometry
HER3 vaccine-induced antibodies in vaccinated mouse serum were measured by flow cytometry as reported [
25]. Briefly, 3 × 10
5 human breast cancer cells were incubated with diluted (1:100 to 1:51200) mouse, post-vaccine serum (HER3-VIA or GFP-VIA) for 1 h at 4 °C and then washed with 1% BSA-PBS. The cells were further stained with phycoerythrin (PE)-conjugated anti-mouse IgG (Dako, catalog number R0480) for 30 min at 4 °C and washed again. Samples were analyzed on a BD LSRII flow cytometer (Becton Dickenson, San Jose, CA, USA) and mean fluorescence intensity (MFI) reported. For the analysis of HER family expression on tumor cells, PE-conjugated anti-EGFR, anti-HER2 (BD Biosciences), and anti-HER3 antibody (BioLegend) were used as in the manufacturer’s instructions.
Detection of HER3 epitopes bound by vaccine-induced antibody
Epitopes were mapped using spotted peptide arrays of 15-mer peptides overlapping by four amino acids representing the full length of the human HER3 protein. HER3 peptides were coated onto cellulose membranes using a Spot Robot ASP 222 (AbiMed) and HER3-VIA (1:100 dilution in saline) epitopes were mapped as described [
26].
Heregulin binding assay
BT474 cells (HER3+) were incubated with medium containing no serum at 37 °C for 24 h. At 30 min before the assay, the culture plates were placed at 4 °C to avoid internalization of HER3 receptor. The following procedures were performed on ice. Cells were pre-incubated with heregulin (final concentrations: 0, 10, 100 nM) or GFP-VIA/HER3-VIA (final dilution: 1:100) for 10 min, then heregulin-His Tag (final concentration: 100 nM) was added and further incubated for 10 min. After washing three times with cold PBS, cells were incubated with APC-conjugated anti-His Tag antibody for 30 min. Cells were washed with cold PBS three times, harvested from the flask, and analyzed using the LSRII flow cytometer.
MTT assay to detect cell proliferation
The effect of HER3-VIA on the proliferation of human breast cancer cell lines was measured as previously described [
27]. Briefly, 5000 cells per well were cultured in a 96-well plate with HER3-VIA (1:33 dilution), GFP-VIA (1:33 dilution) or trastuzumab 20 μg/ml for 3 days and proliferation was assessed by a 3-(4,5-dimethylthiazol-2-yl)-2.5-diphenyltetrazolium bromide (MTT) assay.
Assessment of HER3 internalization
Human HER3+ breast cancer cells (SKBR3, BT474M1 and MDA-MB-468) were incubated with 1:100 HER3-VIA or GFP-VIA at 37 °C for 60 min (SKBR3, BT474M1) or 3 h (MDA-MB-468). After washing, fixation with 4% paraformaldehyde (PFA), and application of permeabilizing solution 2 (Becton Dickenson), nonspecific binding was blocked with 2.5% goat serum at 37 °C for 30 min. Cells were incubated with 1:100 Red™-conjugated anti-mouse IgG (H + L) (Jackson ImmunoResearch Laboratories Inc. West Grove, PA, USA) in a dark chamber for 1 h at room temperature and washed with PBS. For the detection of EGFR, MDA-MB-468 cells were incubated with HER3-VIA for 3 h, then fixed and permeabilized. Cells were then labeled with cetuximab (40 μg/mL) for 10 min, followed by staining with Cy2-conjugated anti-human IgG antibody (ab97169, Abcam, Cambridge, MA, USA). For the detection of HER2, SKBR3 cells were labeled with trastuzumab (40 μg/mL) for 10 min, washed with medium, and then incubated with HER3-VIA for 3 h. Then, cells were fixed and permeabilized, and stained with Cy2-conjugated anti-human IgG antibody. Slides were mounted in VectaShield containing 4′,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA, USA) and images were acquired using a Zeiss Axio Observer wide-field fluorescence microscope (Carl Zeiss, München-Hallbergmoos, Germany).
Complement-dependent cytotoxicity assay
We performed complement-dependent cytotoxicity assays using our previously published protocol [
27]. Briefly, target cells were incubated with rabbit serum (1:100) as a source of complement and the HER3-VIA or GFP-VIA in serum from mice immunized as above diluted (1:100), or trastuzumab (20 μg/mL) at 37 °C for 2 h. After incubation, cytotoxicity was measured using the CytoTox 96 Nonradioactive Cytotoxicity Assay (Promega; per manufacturer’s instructions) to measure lactate dehyrdrogenase (LDH) release in the culture medim as evidence of cytotoxicity.
Antibody-dependent cell-mediated cytotoxicity assay
Antibody-dependent cell-mediated cytotoxicity was measured against HER3-expressing JC-HER3 cells and parental JC cells (HER3-negative) using mFcγRIV ADCC Reporter Bioassay (Promega, catalog number M1211). Target cells (25,000 or 10,000 cells/well) were seeded into 96-well plates, incubated overnight, and pre-incubated with 1:10 dilution of HER3-VIA or GFP-VIA for 30 min at room temperature. Then, effector cells were applied per manufacturer’s instructions. Two different effector-target ratios (3:1, 7.5:1) were tested. After 6 h of co-incubation, Bio-Glo™ Reagent was added, and luminescence was measured. Luminescence of JC cells was subtracted from luminescence of JC-HER3 cells for each VIA.
Treatment of established HER3+ human breast tumor xenografts by passive transfer of vaccine-induced antibodies
BT474M1 cells, lapatinib-resistant rBT474 cells, or MDA-MB-468 cells (5 × 10
6, 1 × 10
6, 1 × 10
6 cells/mouse, respectively) were implanted in the mammary fat pads of 8–10-week-old NOD.CB17-
Prkdcscid/J mice. At 2 days prior to tumor implantation, 17-beta-estradiol pellets (0.72 mg 60-day continuous release pellets; Innovative Research of American, Sarasota, FL, USA) were subcutaneously implanted in the backs of the mice, except for the experiment with MDA-MB-468 cells. Tumors were allowed to develop for 14 days (BT474M1), 2 months (rBT474), or 12 days (MDA-MB-468), to reach the volume of approximately 50–100 mm
3, and then mice were randomized to receive intravenous injection of either GFP-VIA or HER3-VIA. VIA (100–150 μL) was injected at 2–3-day intervals for a total of 10 administrations. Tumor growth was measured in two dimensions using calipers and tumor volume was determined using the formula:
$$ \mathrm{Volume}=\frac{1}{2}\ \left[{\left(\mathrm{Width}\right)}^2\mathrm{x}\ \left(\mathrm{Length}\right)\right]. $$
Western blotting to analyze pathway inhibition
Tumors were isolated from euthanized, VIA-treated mice and immediately flash frozen. Tissue extracts were prepared as previously described [
27]. Equal amounts of proteins (50 μg) were resolved by 4–15% gradient SDS PAGE. After transfer, membranes were probed with specific antibodies recognizing target proteins: pTyr (Sigma), ErbB2, ErbB3, Akt, pAkt473, Erk 1/2, pErk1/2 (Cell Signaling, Beverly, MA, USA), survivin, actin (Sigma, St. Louis, MO, USA), 4EBP-1, p4EBP-1, s6, ps6 (Santa Cruz Biotech, Santa Cruz, CA, USA), and then with IRDye 800 conjugated anti-rabbit or mouse IgG or Alexa Fluor 680 anti-rabbit IgG, and were visualized using the Odyssey Infrared Imaging System (LI-COR, Lincoln, NE, USA) as previously described [
27].
Immunohistochemical analysis of HER3 expression in tumor tissue
BT474M1 or rBT474 tumors were collected when mice were sacrificed, fixed with 10% neutral-buffered formalin, and embedded to paraffin. Tissue sections of 4 μm thick were deparaffinized, and heat-induced antigen retrieval was performed in sodium citrate buffer for 20 min. After blocking endogenous peroxidase activity with 3% H2O2, 10% normal horse serum was applied for the blocking of nonspecific binding sites. Anti-HER3 antibody (Santa Cruz) was put on the sections and incubated overnight at 4 °C. After washing with PBS, biotinylated secondary antibody (Bio-Rad) was applied for 30 min, followed by an VECTASTAIN ABC kit (Vector Lab) and then the color was developed using the DAB Peroxidase substrate kit (Vector Lab). Counterstaining was performed with hematoxylin.
Statistical analysis
Tumor volume over time was standardized by the baseline tumor volume. Area under the tumor growth curve was calculated under spline interpolation [
28] and adaptive quadrature. Groups were compared based on the Kruskal-Wallis test [
29] followed by multiple comparisons performed by the non-parametric Tukey test [
30]. If only two groups were compared then the Mann-Whitney test [
31] was applied. Normality assumption was verified by the Shapiro-Francia test [
32] and homogeneity of variances by the Levene test [
33]. Difference in luminescence intensity in the ADCC assay was analyzed by Fisher’s exact test. All tests of hypotheses were two-sided, with a significance level of 0.05. Calculations were performed using R, version 3.2.5 [
34].
Discussion
Although EGFR and HER2-targeted therapy has substantial activity in EGFR and HER2-overexpressing malignancies, respectively, therapeutic resistance and progression eventually develops in responders with metastatic disease who remain on therapy. Significant data implicate HER3 and more specifically, EGFR:HER3 and HER2:HER3 heterodimers as mediators of this resistance [
36]. Monoclonal antibodies would be one mechanism for inhibiting HER family heterodimerization, but trastuzumab is ineffective against HER2:HER3 heterodimers [
37], resistance develops to trastuzumab/pertuzumab combinations, there are no pertuzumab equivalents for EGFR, and anti-HER3 monoclonal antibodies are thus far not commercially available [
38]. Our overall goal has been to develop an immunologic strategy to target HER3-expressing malignancies, reasoning that immune responses will be adaptive, pleiotropic, and persistent. Previously, we generated an Ad-HER3 vaccine and demonstrated that it induced potent HER3-specific T cell responses with anti-tumor activity [
20]. In that study, we also observed preliminary evidence of the induction of HER3-VIA polyclonal serum in Ad-HER3-vaccinated mice, which suggested that further characterization of the antibody response was warranted thus leading to the experiments presented herein. We believe that while the T cell response is often regarded as the primary anti-tumor effector mechanism, antibody responses are also important as major histocompatibility complex (MHC) downregulation and epitope loss, observed in some tumors, would render them resistant to T cell-mediated cell death [
39].
Vaccine-induced antibodies by virtue of their polyclonality should have the capacity to bind to multiple different epitopes and have multiple different functions, including commonly ascribed immune activities such as ADCC and CDC. Indeed, in the current study we observed immune-mediated anti-tumor activity mediated by Ad-HER3-induced antibodies (CDC and ADCC against HER3-expressing HER2+ and TNBC cells).
Additional anti-tumor activities that may be mediated by antibodies include internalization and signaling inhibition of the target-surface-expressed molecule [
40‐
42]. Indeed, we observed that the anti-proliferative effects of HER3-VIA were likely due to internalization and signaling inhibition rather than inhibition of heregulin binding of HER3. These findings are consistent with our previously reported ability to generate polyclonal antibodies against the HER family member HER2, where we observed that these antibodies mediated HER2 receptor internalization and degradation in both mouse and human studies [
19].
An important implication of HER3 downregulation is inhibition of signaling through HER3 heterodimers. Interestingly, treatment of lapatinib-sensitive and lapatinib-resistant BT474 cells with HER3-VIA led to decreased HER3, pHER3 and pERK1/2 as expected, but only treatment of the resistant BT474 cell led to a decrease in HER2, pAkt(S473), pS6, p4EPB1, and survivin expression. The decrease in the protein survivin, an inhibitor of apoptosis, suggests that there is also an increase in apoptotic cells after Ad-HER3 treatment. In a previous study [
43], when the partially trastuzumab-resistant 4 T1-HER2-expressing tumors were treated with lapatinib or HER2-VIA alone, we observed no change in survivin expression, but when these tumors were treated with a combination of lapatinib and HER2-VIA, we observed a decrease in survivin expression [
27] implying that complete HER2 signaling blockade decreased survivin expression [
43]. In an analogous fashion, our current findings suggest that complete blockade of HER2:HER3 signaling in lapatinib-refractory tumors is accomplished by treatment with HER3-VIA, resulting in the decreased expression of survivin.
An important observation was that the growth in vitro and in vivo of more aggressive breast cancer subtypes such as HER2+ breast cancer and TNBC, and of lapatinib-refractory (HER2 small molecule inhibitor-refractory) tumors was slowed by HER3-VIA. We believe our findings lay a framework for an immune-mediated strategy for treating aggressive breast cancer subtypes (TNBC and HER2+ breast cancer), by co-administration of Ad-HER3 with HER2-targeted or EGFR-targeted therapies. Also, HER3 may play a role in therapeutic resistance to anti-estrogen therapies in estrogen receptor (ER)-positive breast cancers [
44‐
47], and the development of castration resistance in prostate cancers [
48]. Therefore, our findings also suggest a role for immunization with Ad-HER3 to induce HER3-VIA prior to the development of therapeutic resistance. Clinical trials of these approaches are in development.
Conclusions
In this study, we revealed that polyclonal antibodies induced by Ad-HER3 vaccine (HER3-VIAs) are multifunctional, including induction of CDC, ADCC, anti-proliferative effect, HER3 internalization, and interruption of HER3 heterodimer-driven tumor signaling pathways. In addition to the T cell anti-tumor response induced by Ad-HER3, the HER3-VIAs provide additional functions to eliminate tumors in which HER3 signaling mediates aggressive behavior or acquired resistance to HER2-targeted therapy. These data support clinical studies of vaccination against HER3 prior to or concomitantly with other therapies to prevent outgrowth of therapy-resistant HER2+ and triple negative clones.