Cancer Research Has Developed Several Models Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Cancer research has developed several models to investigate and evaluate targeted therapies. Especially the introduction of SCID mice, allowing xenografts of human tumors, contributed significantly to progress in oncological drug development. As SCID mice display different degrees of immunodeficiency, such models are only of limited use in immunotherapeutical research. To investigate the in vivo efficacy of active immunotherapies, currently only a few models are available, which are based on TAA-transgene expression under organ-specific promotors, or on transplantation of transgenic cell lines overexpressing the desired target. These models may lack the functionality of the transgene product leading to failure to achieve the proof-of-concept.

We thus present in this study a novel model using the human HER-2 overexpressing breast cancer cell line BT474 (derived from solid, invasive ductal carcinoma) for xenografts in immunocompetent BALB/c mice. Growth and tumor formation of this established and reliable model cell line is feasible because of its described MHC-I downregulation and consecutive immune evasion.

Upon subcutaneous grafting of 2x106 cells, first tumors could be detected in naïve mice after 6 days post transplantation and reached the end point volume of 300mm3 between day 12 and 18. Upon immunization against HER-2, significant benefit could be demonstrated, as mice showed slower tumor growth resulting in significantly longer survival. As BT474 cells also express EGFR, this new xenotransplantation model could serve as a highly interesting tool for future translational oncological trials. Upon application in immunotherapy settings, cancer research will be able to gain important insights into the EGFR/HER-2 biology as well as MHC-1 specific tumor escape strategies, with possible benefits for the development of novel immunotherapies for oncological patients.


The human breast cancer cell line BT474 is commonly used in cancer research to model mammary carcinoma [1-6]. BT474 or HTB-20ƒ”, as the cell line is listed by ATCC (American Type Culture Collection - ATCC, Rockville, Maryland, USA), were first isolated by E. Lasfargues and W.G. Coutinho from a solid, invasive ductal carcinoma of a 60-year old caucasian female [7]. Its overexpression of the human epidermal growth factor receptor-2 [8], one of the most attractive targets in oncology, established this cell line as an important tool in cancer immunotherapeutic research. The human epidermal growth factor receptor-2 (HER-2, ErbB-2) is a receptor tyrosine kinase, mediating key features of malignant growth, such as invasiveness, vascularization and development of distant metastases [9-10]. The overexpression of HER-2 on cancer cells is clinically evaluated via the immunhistochemical HercepTest® (DAKO, Glostrup, Denmark) [11]. This test classifies tumors for HER-2 expression on a scale from „0" to „+++", where 0 indicates little (less than 10% of all cells) to no membrane staining. If more than 10% of cells are stained, tumors are classified as „+"; (+) equals incomplete membrane staining, (++) corresponds to weak to moderate complete membrane staining and (+++) indicates strong and complete membrane staining in >10% of all tumor cells [12-13]. BT474 cells are classified „+++" in the HercepTest® score, making them an ideal model to study novel therapeutic agents targeting carcinomas highly overexpressing HER-2 [14-15]. Currently, two HER-2 targeting drugs are in clinical use, trastuzumab (Herceptin®, Roche), a monoclonal antibody and lapatinib (Tykerb®, GlaxoSmithKline), a small molecule tyrosine kinase inhibitor, both approved for the therapy of metastatic breast cancer overexpressing HER-2 [14].

Furthermore, BT474 cells also express the human epidermal growth factor receptor-1 (EGFR, ErbB-1) [8], another member of the ErbB-family that is currently targeted in clinical immunotherapy of cancer [16]. EGFR is amongst others overexpressed in human colon [17-18] and head and neck cancer [19], where it regulates genes needed for cell proliferation [20]. Silencing of these growth signals, either with the monoclonal antibody cetuximab (Erbitux®, Merck) or with the small molecule tyrosine kinase inhibitor erlotinib (Tarceva®, Genentech), is part of several therapeutic regimens of metastatic colorectal cancer [21]. Although BT474 cells display considerable amounts of EGFR on their surface [8], this cell line is not regarded as a classical model for EGFR positive cancers such as e.g. A431 (listed as CRL-1555â„¢ by ATCC), an epidermoid carcinoma cell line, isolated by Giard and Aaronson back in the early 1970ies [22]. Nonetheless, several recent studies also employed BT474 cells in this regard, focussing on the interaction between EGFR and HER-2 [23-25]. BT474 cells were also successfully applied in multiple xenograft trials in SCID mice, allowing evaluation of targeted therapies with xeno-antibodies such as humanized or chimeric immunoglobulins. As our group has been involved in the development of mimotope vaccines for EGFR, HER-2 and other important TAAs [26-37], we attempted to develop an immunocompetent animal model for proof-of-concept (PoC) studies. To investigate the in vivo efficacy of active immunotherapies, currently only a few models are available, which are based on TAA-transgene expression under organ-specific promotors, or on transplantation of transgenic cell lines overexpressing the desired target. Notably, the action of anti-tumor immunoglobulins such as trastuzumab is as much conferred by interference with signaling as with immunologically mediated ADCC (antibody-dependent cell-mediated cytotoxicity) [38-39]. We demonstrate in this study that transplant models may sometimes lack the functionality of the transgene product leading to failure to achieve the proof-of-concept. For instance, targeting the murine mammary carcinoma cells D2F2, transfected with human HER-2 [40] failed in our studies, although the relevant target molecules were expressed at a high density. The present study thus intended to investigate whether the fully functional human cancer cells BT474 could be used for xenografts in an immunocompetent BALB/c mouse model. Against our own expectations, our data propose that this is feasible due to their low expression of MHC-I molecules. This new model opens up the possibility for more consistent proof-of-concept studies for the human EGFR and HER-2 targets in a less artificial mouse organism than SCID, harboring all needed effector cells. Our model is a contribution to facilitate developments of future immunological anti-cancer therapeutics, especially anti-cancer vaccines.

Materials and Methods

Cell lines, recombinant proteins and monoclonal antibodies

BT474 cells (listed as HTB-20ƒ” in the American Type Culture Collection - ATCC, Rockville, Maryland, USA) were a kind gift of Prof. Dr. Thomas Grunt (Institute of Cancer Research, Medical University of Vienna) and cultivated in ¡-MEM medium supplemented with 10% fetal calf serum (FCS), 2 mM l-glutamine, penicillin (100 U/mL) and streptomycin (100 µg/mL).

The parental cell line D2F2, a murine mammary carcinoma cell line of BALBc background, and D2F2/E2 cells, transfected with human HER-2, were provided by Wei-Zen Wei (Karmanos Cancer Institute, Wayne State University School of Medicine, Detroit, Michigan). D2F2, as well as D2F2/E2 cells were grown in Iscove's modification of Dulbecco's medium (IMDM), supplemented with 5% FCS, penicillin (100 U/mL) and streptomycin (100 µg/mL).

A431 cells (CRL-1555â„¢, ATCC) cells were allowed to grow in DMEM medium augmented with 10% FCS, penicillin (100 U/ml) and streptomycin (100µg/ml).

All cells were grown in a humidified atmosphere of 5% CO2 at 37°C.

Peripheral blood mononuclear cells (PBMC), collected from healthy human donors were obtained from STEMCELL Technologies (STEMCELL Technologies SARL, Grenoble, France).

Cetuximab (Erbitux®), a chimeric IgG1 anti-ErbB-1 (EGFR) monoclonal antibody, was obtained from Merck KGaA (Darmstadt, Germany), and trastuzumab (Herceptin®), a humanized IgG1 monoclonal anti-ErbB-2 (HER-2) antibody, from Roche (Hertfordshire, United Kingdom). Rituximab (MabThera®), a chimeric IgG1 anti-CD20 monoclonal antibody, employed as isotype control, was also purchased from Roche (Hertfordshire, United Kingdom).

The α-asialo GM1-antibody was obtained from Wako (Wako Pure Chemical Industries, Ltd., Osaka, Japan).

The expression system for producing recombinant HER-2 (rHER-2) was used as described by Cho et al. [41]. In short, recombinant soluble HER2 (residues 1-631) was expressed in Lec1-cells, kindly provided by Prof. Daniel J. Leahy (The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA) and purified via its His-Tag using a Ni-NTA column. Lec-1 cells were kept in a humidified atmosphere of 5% CO2 at 37°C in DMEM/F12, supplemented with 5% FCS, 100nM methotrexate and 10 µg/mL gentamicin sulfate.

Flow cytometric analysis

For flow cytometric assessment of HER-2 surface-receptor expression on human BT474, as well as murine D2F2 and D2F2/E2 cells, cells were incubated with 200 µl of either 10 µg/mL cetuximab, trastuzumab or rituximab, respectively, for 30 min at 4°C, followed by two washes in ice cold FACS-buffer (phosphate buffered saline (PBS), 5% normal goat serum) and incubation with anti-human IgG AF488 antibodies (200 µl of 20 µg/mL, Invitrogen, Life Technologies, Grand Island, New York, USA) for 30 min at 4°C. After washing with ice cold FACS-buffer, analysis was performed on a dual laser FACSCaliburâ„¢ (BD Biosciences, Franklin Lakes, New Jersey, USA).

Staining against HLA-molecules was performed analogous on BT474, A431 and human peripheral blood mononuclear cells (PBMC) with an anti-human HLA-ABC antibody (200 µl of 1 µg/mL DAKO, Glostrup, Denmark), followed by detection with an anti-mouse IgG FITC (200 µl of 20 µg/mL, BioLegend, Inc., San Diego, California, USA). Purified mouse IgG2a (Southern Biotech, Birmingham, Alabama, USA) served as isotype control. PBMC were analysed in total, but HLA-expression of T-cells was employed for further analysis. Therefore, T-cells were gated via staining with anti-CD-3 AF647 (eBioscience, Frankfurt, Germany).

Analysis was again performed on a dual laser FACSCaliburâ„¢.

Cell viability assays

The dependency of tumor cell growth on signaling via the HER-2 receptor was measured by means of the EZ4U® proliferation assay (EZ4U® the 4th Generation non radioactive cell proliferation & cytotoxicity Assay kit, Biomedica, Vienna, Austria), which displays the viability of cells, based on its ability to reduce tetrazolium salts. BT474, D2F2 and D2F2/E2 cells were seeded in 96-well plates at a density of 3 x 104 cells per well (in 100µl complete medium) and allowed to adhere overnight under standard culture conditions prior to assays. Cells were exposed to 10 µg/ml trastuzumab or 10 µg/ml cetuximab antibodies over a period of 24, 48 and 72 hours. Furthermore, combined treatment with 5 µg/ml trastuzumab plus 5 µg/ml cetuximab was investigated. Control groups received media alone or 10 µg/ml human IgG1 isotype control (rituximab) Following treatments, tetrazolium solution, prepared according to the manufacturer's instructions, was added at 20 µl per well and cells were incubated for a further 1 h prior to recording absorbance at 450 nm with 620 nm as a reference using a 96-well plate reader (Infinite M200 PRO, Tecan Group AG, Maennedorf, Switzerland). The quantity of the formazan product as measured by the amount of 450 nm absorbance directly correlates to the number of living cells in culture.

Tumor graft experiments

For evaluating the trastuzumab sensitivity of BT474 cells in vivo, human mammary carcinoma cells BT474 were grafted into severe combined immunodeficient mice. Six to eight weeks old C.B-17 scid/scid (SCID) mice were employed for all experiments (Harlan Winkelman, Borchen, Germany). All mice were kept under pathogen-free conditions in filter cages (IVC racks, EHRET Labor- & Pharmatechnik GmbH & Co.KG, Emmendingen, Germany) on the basis of authorization of the Animal Ethics Committee of the Medical University according to the Austrian, European Union and FELASA guidelines for animal care and protection (BWMF-66.009/0047-II/10b/2010).

The tumor therapy experiment was performed in male animals, n=6 per group. The respective groups received at day 0 of the experiment 25μg α-asialo GM1-antibody (Wako Pure Chemical Industries, Ltd., Osaka, Japan) intraperitoneally to eliminate the NK-cell mediated resistancy of mice against the human cells. At day 1 all mice received 1x106 BT474 cells, grafted subcutaneously into the right flank and 3x106 peripheral blood mononuclear cells (PBMC) intravenously, purified of healthy human donors. The immunotherapeutically treated group received 100μg trastuzumab antibody injected intravenously (in 100μl of 0.09% NaCl saline) directly after reconstitution of the mice with immune cells.

Tumors were measured by caliper measurement and tumor volume was calculated employing the formula V (mm3) = d2 (mm2) x D (mm)/2, where d stands for the smallest and D for the largest tumor diameter. Additionally mice were monitored for signs of tumor related side effects such as stopping of food ingestion or weight loss. When tumors reached a volume of 300mm3, mice were sacrificed.

To determine the functionality of a HER-2 overexpressing tumor graft model in immunocompetent mice, six to eight weeks old female BALB/c mice were employed (Institute for Laboratory Animal Science and Genetics, Medical University of Vienna) on the basis of authorization of the Animal Ethics Committee of the Medical University according to the Austrian, European Union and FELASA guidelines for animal care and protection (GZ: BWMF -66.009/0003-II/3b/2011). Mice were immunized 4 times subcutaneously with 50μg of recombinant, extracellular domain HER-2 (rHER-2) adsorbed to aluminum hydroxide solution (Al(OH)3, Alum) as an adjuvant at intervalls of 2 weeks. After confirming the humoral immune response via ELISA, mice received HER-2 transfected D2F2/E2 cells (2x106 cells in 100 µl IMDM, subcutaneously, day 1).

To determine the effect of anti-HER-2 vaccination treatment on naturally HER-2 expressing tumors, again six to eight weeks old female BALB/c mice were employed (Institute for Laboratory Animal Science and Genetics, Medical University of Vienna, GZ: BWMF -66.009/0003-II/3b/2011). Mice were immunized 3 times according to the immunization scheme described above and grafted 2x106 BT474 cells in 100 µl α-MEM subcutaneously at day 1. Tumor growth was monitored daily by caliper measurement and mice were sacrificed when tumors reached a volume of 300mm3.

Enzyme Linked Immunosorbent Assay (ELISA)

To determine the immune response of immunocompetent BALB/c mice towards vaccination with rHER-2, HER-2 specific antibody levels were measured in ELISA. Recombinant, extracellular domain HER-2 (rHER-2) was coated on microtiter plates (Immuno Maxi-Sorpƒ”, Nunc, Roskild, Denmark) at a concentration of 1­g/ml. After a blocking step with 1% dried milk powder (DMP) in TRIS-buffered saline, 0.05% Tween20 (TBST), mouse sera diluted 1:100 in TBST/0.1% DMP was allowed to adhere overnight on 4°C. Bound immunoglobulins were detected with rat anti-mouse IgG1-antibodies (BD Biosciences, Franklin Lakes, New Jersey; diluted 1:1000), followed by a horseradish-peroxidase labeled goat anti-rat IgG antibody (Amersham, Buckinghamshire, UK; diluted 1:1000). For detection, 3,3',5,5' tetramethylbenzidine (TMB) was added and the optical density (OD) was measured at 450nm with 620nm as reference wavelength.


Mammary tumor samples were fixed in buffered 10% formaline and paraffin embedded. Serial 4-µm sections were cut and immunohistochemistry was performed using standard protocols for Hematoxylin and Eosin staining. Furthermore, HER-2 specific staining was conducted, using the FDA-approved in vitro diagnostic HercepTest® (DAKO, Glostrup, Denmark) kit according to the manufacturer's instructions. Evaluation was performed according to the HercepTest® classification.

Data handling and statistics

Flow cytometric experiments of receptor expression were at least repeated three times, and the histograms depicted in this publication show one representative example. Cell viability assays were performed with n=6 for every treatment condition at 24 and 48 hours-timepoints and n=4 for the 72 hours incubation. Statistical analysis of this assay was performed by means of one-way-ANOVA and Bonferroni's correction was used as post-test correction.

For survival experiments, Kaplan-Meier curves of treated animal groups were plotted and analysed applying Log-rank (Mantel-Cox) Test.

In all statistical calculations of this study, significance was accepted at p < 0.05 (*), p < 0.01 (**), and p< 0.001 (***).


Flow cytometric analysis of HER-2 expression on BT474, D2F2 and D2F2/E2 cells

In order to determine the practical value of BT474 cells as a model for HER-2 targeting, the surface expression of the receptor tyrosine kinase was assessed via flow cytometry. Fig. 1a depicts the respective histograms, demonstrating considerable HER-2 staining of BT474 cells with trastuzumab resulting in median fluorescence intensity (MFI) of 1498.93, compared to MFI=4.26 for the secondary, detecting antibody control and again MFI=9.91 for the anti-CD20 antibody rituximab, serving as isotype control. This is in line with previous observations, including our own findings of moderate, but homogenous EGFR expression (MFI=30.78), detected with cetuximab [8, 24]. At the surface of D2F2 cells, no HER-2 specific staining could be observed (MFI=3.16, Fig. 1b), whereas their transfected derivatives, D2F2/E2 cells, expressed high levels of HER-2 (MFI=99.1, Fig. 1c).

Tumor Xenograft Experiment with BT474C.B-17 scid/scid (SCID) mice

To evaluate the tumoricidic effects of trastuzumab in vivo, we employed the commonly used xenograft method of transplanting human cancer cells in immunocompromised mice. For BT474 cells, the most widespread regimen contains the use of subcutaneous implantation of estrogen-releasing pellets prior to tumor graft [42]. As two recently published papers described adverse effects of estrogen pellet implantation on the urinary tract, involving dilated urinary bladders, hydronephrosis [43] as well as urinary retention and cystitis [42], this procedure was excluded from the protocol. In fact, 100­l of BT474 cell suspension were grafted subcutaneously at a total number of 1x106cells in the right flanks of male C.B-17 scid/scid (SCID) mice.

For all studies, mice were pretreated with α-asialo GM1-antibodies intraperitoneally to eliminate the NK-cell activity against the human cells, which is further nourished by low- HLA-1 expression [44] To reconstitute the mice with effector cells, 3x106 PBMC of healthy human donors were infused intravenously and the therapy group received additionally 100­g trastuzumab as a single treatment. As Figure 2 clearly demonstrates, not a single animal of the trastuzumab treated group developed a tumor until day 152. Actually, at this time point the maximal volume was not yet achieved, but the experiment had to be terminated due to aggressive behaviour of the mice. In clear contrast, mice of both control groups developed tumors. Thus as expected, trastuzumab clearly showed a protective effect in C.B-17 scid/scid (SCID) mice.

Tumor graft of D2F2/E2 cells in BALB/c mice

To investigate whether a similar tumor-protective effect like that observed with trastuzumab could be achieved via active immunization with rHER-2, we employed a tumor graft model in immunocompetent BALB/c mice. Mice were immunized with 50­g rHER-2 adsorbed to Al(OH)3 as adjuvant. After 4 rounds of immunzation, when mice displayed high levels of antibodies against HER-2 in ELISA (Fig. 3a), D2F2/E2 cells were grafted. D2F2/E2 cells were transplanted subcutaneously and from day 3 on tumors could be detected in almost all animals. Although one could observe delayed tumor growth in HER-2 immunized mice between day 5 and 10 of the study, this small and non-significant effect was completely abolished towards the end of the study (Fig.3b).

In order to exclude loss of HER-2 expression as a possible explanation for the failure of immunoprotection, histological staining against HER-2 was performed. All tumors of both naïve and treated groups showed strong and homogenous surface expression of HER-2 (Supplementary Fig. 1, right panels), proving the specificity of the chosen model cell line.

Cell viability assays

Consequently, our next aim was to evaluate the susceptibility of the tested mammary carcinoma cells towards growth factor silencing. Thus D2F2 versus D2F2/E2, and BT474 cells were treated with cetuximab or trastuzumab, and cell viability was measured.

Fig. 4a shows for BT474 cells, that until 24 hours of single treatment with either cetuximab or trastuzumab, none showed any growth inhibitory effect, whereas combination treatment with both antibodies indicated a trend towards diminished growth compared to untreated and isotype control treated groups. After 48 hours of treatment with trastuzumab or combination treatment of both antibodies, growth could be significantly reduced compared to untreated cells (p<0.001). This observation was even more pronounced after 72 hours, when trastuzumab treated cells displayed in average only 58% growth of untreated ones (p<0.001) and growth of cells treated with combination treatment was almost diminished by 50% (p<0.001).

In contrast to this finding, neither D2F2 nor D2F2/E2 cells displayed any growth inhibition upon treatment with trastuzumab at any timepoint (Fig. 4b,c), irrespective of their HER-2 status. This indicated that although D2F2/E2 strongly expressed the target molecules, they were non-functional. Therefore, the immune-mediated effect via the Fc-domains of the tested targeting antibodies is obviously not sufficient to combat the cancer cells.

Tumor xenograft of BT474 cells in immunocompetent BALB/c mice

Thus, to refine the tumor transplant mouse model, BT474 cells, being highly susceptible to HER-2 targeting (Fig. 3a), were grafted in immunocompetent BALB/c mice. Prior to tumor transplantation, mice were immunized three times with rHER-2 adsorbed to Al(OH)3 and formed antibody titres, measured in ELISA (Fig. 4a). In this improved model clear survival benefits of the HER-2 immunized group could be observed. In contrast, in the naïve mice the experiment had to be terminated before day 20 of the trial, when more than 50% of the treated mice were still alive (Fig. 4b).

HLA-expression of BT474 cells

In order to elucidate how human cells, which are foreign to the immune system of BALBc mice can adhere and grow in vivo, HLA-receptor expression of BT474 cells was determined. Thus, flow cytometric analysis of BT474, A431 as well as T-cells was performed. As can be seen in Fig. 6, BT474 cells display homogenous, but diminished HLA-molecules on their surfaces (MFI=34.29) compared to T-cells (MFI=609.76) or A431 cells (MFI=679.25) which did not undergo MHC-downregulation. Isotype controls show only background staining in all tested conditions (dashed lines).


Targeted therapy of malignancies has gained great importance in clinical oncology. To develop and evaluate targeted therapies, reliable in vitro and in vivo models are needed. Cell lines generated from isolated cancer samples of human patients have enabled cancer researchers to measure surrogate parameters in order to assess tumoricidic efficacy as well as toxicity of novel compounds. Currently those tumoricidic effects, that could be observed in vitro, are confirmed in vivo in xenograft models in mice with different degrees of immunodeficiency. Although the concrete predicitive effect of xenograft models is discussed controversely and recent strategies try to refine the screening procedure, for example via establishing xenografts from primary solid-tumor isolates, the basic principle still serves as the 'workhorse' of the pharmaceutical industry [45]. Successful targeted therapies often comprise immunotherapy with monoclonal antibodies due to their high specificity [46]. But one of the major drawbacks of early immunotherapy against cancer was the reaction of the patient's immune system against the first monoclonal antibodies, which were of murine origin [47]. These so-called HAMAs (human anti-murine antibodies) [48] or HAGAs (human anti-globulin antibodies) [49] recognized the therapeutically applied antibodies as foreign proteins, leading on the one hand to rapid clearance of the therapeutic agent from the circulation, resulting in low efficacy [50] and on the other hand to symptoms such as serum sickness or immediate type hyperreactivity [51], as well as to unexplained tumor marker changes [52].

To overcome this situation, chimeric, humanized or finally fully human antibodies were generated [53]. The clinically applied cetuximab is a chimeric and trastuzumab a humanized antibody, respectively. These recombinant proteins are well tolerated by patients, and proved to be successful in various clinical studies [54-55]. Whereas, if these compounds are applied to mice in preclinical proof-of-concept studies, these mice will react similarly as humans and produce antibodies against the foreign protein. This leads to the same risks concerning serum sickness and immediate type hyperreactivity. Therefore and mainly because of the lack of tumor graft rejection, SCID mice were chosen as a widespread model for testing of anti-cancer compounds. Severe combined immunodeficient (SCID) mice lack B- and T-cells due to the lack of V(D)J recombination [56-57]. Since they are not able to elicit specific immune responses, they cannot produce antibodies or cytotoxic T-cells against foreign proteins. These mice enabled cancer researchers to graft human tumors into mice, without triggering transplant rejection [45]. As of course other factors also contribute to immune reactions against transplants or foreign proteins, like e.g. NK-cells, these mouse models have been refined, e.g. crossing SCID mice onto a nonobese diabetic (NOD/Lt) background or adding defects of the interleukin-2 receptor gamma chain (IL-2Rg) resulting in mice also lacking NK-cell activity (NOD-scid IL2Rnull) [58].

Also in this context, BT474 cells are employed to mimic mammary carcinoma [1-6]. In this respect, BT474 tumor grafts are regarded as HER-2 positive, estrogen-responsive cancer based on responsive effects of estrogen modulation in vitro [59-61] and in vivo [62]. The standard protocol of BT474 tumor grafting includes the implantation of estrogen pellets [42], a method, which is also used for grafting of MCF-7 cells, another human mammary carcinoma cell line [63-65]. Also in other study protocols investigating the effect of HER-2 targeting, female mice are usually implanted with estrogen-releasing pellets prior to tumor graft [2, 4, 6].

While we were designing a xenograft model to investigate novel immunotherapeutic compounds against HER-2, two closely published papers described adverse effects after estrogen pellet implantation on the urinary tract. Gakhar et al. observed dilated urinary bladders and hydronephrosis [43]; Pearse et al. report urinary retention and cystitis and investigated the expression pattern of the estrogen receptor alpha (ER-¡) in the urinary tract [42]. As this study could only determine that the ER-¡ seems to be not involved in the pathophysiology of the disease and the literature is unclear about these findings, we decided to circumvent these problems and to establish our model in male C.B-17 scid/scid (SCID) mice. As Fig. 2 demonstrates, this model of grafting BT474 cells in male mice plus reconstituting their immune system with human PBMCs turned out to be a reliable xenograft model to investigate effects of HER-2 targeting treatment without the risk of urinary tract side effects. A single dose treatment of 100µg trastuzumab completely protected mice from developing tumors. This xenograft model is HER-2 sensitive, working in the absence of foreign hormone treatment and incorporates the effect of human immune cells targeting the malignancy.

But this model is still quite artificial considering the severe immunodeficiency of the mice. Possible side effects, like serum sickness or immediate type hyperreactivity, which would constrain the clinical use of novel substances, are still not perceivable. Furthermore modern immunotherapy of cancer focuses on generating cancer vaccines, due to their possible advantages concerning easier applicability, longer-lasting effects and better cost effectiveness [66]. To display the efficiency of immune responses elicited by vaccines, a tumor graft model in immunocompetent mice would be needed. Thus, for translational studies aiming to investigate and develop active immunotherapies, models have evolved, which exploit transgene expression under organ-specific promotors, or transplantation of transgenic cell lines which often lack the intrinsic biological functions of the transgene product. Fig. 3 illustrates that such models, like e.g. the use of murine mammary carcinoma cells D2F2 of BALB/c background, transfected with human HER-2, fail to display the protective effect of anti-HER-2 vaccine approaches. Although the immunocompetent BALB/c mice developed high levels of antibodies against HER-2 upon vaccination with rHER-2 (Fig. 3a), these antibodies did not act tumor growth-inhibitory (Fig. 3b). One of the possible explanations for this immune escape of the grafted tumors could be shedding of HER-2 [67], leading to the expression of a truncated version of HER2 that lacks the receptor's extracellular domain, (named ''p95 HER2'' due to its molecular weight) [68]. In clear contrast to that hypothesis, all tumors of the naïve as well as the immunized group displayed, strong, clear and homogenous membrane expression of HER-2 in situ (Supplementary Fig 1). As one of trastuzumab's closest investigated mechanism of action is silencing of vital growth signals mediated via HER-2, we aimed to investigate the susceptibility of D2F2/E2 cells in this respect. Although the transfected D2F2/E2 cells displayed high levels of HER-2 on their surface (Fig. 1c), incubation with trastuzumab did not result in any change of cell viability, even after 72 hours of treatment (Fig. 4c). D2F2/E2 cells were as resistant to anti-HER-2 treatment as the parental cell line D2F2, regardless of their high HER-2 expression. On the contrary, naturally overexpressing BT474 cells (Fig. 1a) turned out to be highly susceptible, resulting in significant growth inhibition, even after 48 hours (Fig. 4a). Based on this encouraging result and the fact that BT474 tumor grafts already had proven to be highly reliable in SCID models, we considered to graft human BT474 cells in immunocomptetent BALB/c mice. Prior to grafting, mice were again vaccinated three times with rHER-2 and proved to form high levels of HER-2 specific antibodies (Fig. 6a). When mice were grafted 2*106 BT474 cells subcutaneously, the human tumor cells started to homogenously grow and form tumors in the mouse flanks. When tumors reached a total volume of 300mm3, mice were sacrificed and survival curves were plotted. In contrast to the SCID mouse trial, in which a single application of the anti-HER-2 antibody trastuzumab protected the mice from tumor growth, treated immunocompetent BALB/c mice still developed tumors. Nonetheless vaccinated mice had a clear and significant survival benefit (Fig. 6), indicating the value of this novel xenograft model.

Xenograft rejection is normally caused by incompatibility of major histocompatibility complex (MHC) antigens [69]. MHC-I downregulation as a mechanism of immune escape is a well-observed [70] and heavily discussed topic in viral and tumor immunology [71]. When we investigated the MHC-I expression of the explanted BT474 tumors histologically, we found a profound downregulation, which is in line with a recent report [72]. This observation was not only found in vitro, as Maruyama et al. also described an inverse correlation between HER-2 and MHC-I expression on oesophageal squamous cell carcinoma in situ [72].

Our findings underline the importance of refinement of xenograft models in cancer research. Our described model of BT474 xenotransplantation in immunocompetent BALB/c mice resembles human carcinoma according to expression and functionality of HER-2 as well as a functional immune system of the host. Furthermore it contains an important immune escape strategy of tumors, MHC-I downregulation which could be a promising target of its own for novel therapies.

Therefore we propose this new xenotransplantation model a highly interesting tool for future translational oncological trials. Upon application in immunotherapy settings, cancer research will be able to gain important insights into the EGFR/HER-2 biology as well as MHC-1 specific tumor escape strategies, with possible benefits for the development of novel immunotherapies for oncological patients.


The authors would like to express their gratitude to all members of the Jensen-Jarolim and Walter Berger labs for fruitful discussions and their support and are grateful to Sanjula Jain and Dr. Franziska Roth-Walter for their critical review of the manuscript.

This work was supported by the CCHD PhD program of the Austrian Science Fund FWF, project ##APW01205FW (JS, EJJ), APP23398FW (EJJ), a Foerderstipendium of the Medical University of Vienna, and Biomed Int. R&D, Vienna.Figure Legends

Fig. 1: Flow cytometric assessment of surface expression of human EGFR/HER-2 on cancer cells.

EGFR was detected with cetuximab (dotted grey line), HER-2 with trastuzumab (black line), anti-CD20 antibody rituximab served as isotype control (dashed dark grey line). Cells stained with AF488-labeled secondary detection antibody are depicted as filled histograms in grey.

BT474 cells display homogenously HER-2 overexpression (median fluorescene intensity, MFI=1498.93) as well as EGFR expression (MFI=30.78), whereas only background fluorescence could be detected with the humanized IgG antibody isotype control Rituximab (MFI=9.91). Parental D2F2cells neither express HER-2 (MFI=3.16), nor EGFR (MFI=3.37), whereas HER-2 transfected D2F2/E2 cells display considerable amounts of HER-2 (MFI=99.1).

Fig. 2: Kaplan-Meier survival curve of BT474 tumor grafts in immunodeficient C.B-17 scid/scid mice.

In both control groups, first grafted tumors reached a total volume of 300 mm3 between day 40 and 50, leading to the sacrifice of the respective mice. Consecutively, further animals of the untreated as well as PBMC grafted groups had to be sacrificed. In clear contrast, not a single mouse of the trastuzumab treated group developed any sign of a tumor or a symptom of malignant disease. Although treatment was only once with a single shot of 100µg trastuzumab, mice were protected from HER-2 overexpressing tumors until day 154, when the trial had to be terminated due to aggressive behavior of mice.

Fig. 3: Tumor graft of D2F2/E2 cells in immunocompetent BALB/c mice.

Fig. 3a: ELISA measuring rHER-2 specific IgG1 antibody levels (OD-values) in sera of BALB/c mice.

PIS=Pre-Immune Serum, 1MIS=1st Mouse Immune Serum, after 1 round of immunization with rHER-2; 2MIS=2nd Mouse Immune Serum, 2 immunizations; 3MIS=3rd Mouse Immune Serum, 3 immunizations; 4MIS=4th Mouse Immune Serum after 4 immunization rounds.

Immunized mice display a strong increase in HER-2 specific antibody levels after each round of immunization. Mice of the control group, immunized with PBS display only background immunoglobulin levels.

Fig. 3b: Tumor growth curve of grafted D2F2/E2 tumors.

Tumors appeared at day 3 after transplantation and reached a volume of 300mm3 at day 12. Tumors of HER-2 immunized mice display similar growth as those of untreated animals.

Fig. 4: Cell viability assay measuring the impact of HER-2 silencing on the proliferation of cancer cells.

a: BT474 cells display a trend towards slower growth upon combination treatment with trastuzumab and cetuximab even after 24 hours, whereas after 48 hours single treatment with trastuzumab as well as combination treatment leads to highly significant growth inhibition. This effect is even increased after 72 hours. In clear contrast to that neither HER-2 negative D2F2 (b) nor HER-2 positive D2F2/E2 (c) cells are affected by any tested treatment in their growth properties.

Fig. 5: Tumor graft of BT474 cells in immunocompetent BALB/c mice.

Fig. 5a: ELISA measuring rHER-2 specific IgG1 antibody levels in sera of BALB/c mice.

PIS=Pre-Immune Serum, 1MIS=1st Mouse Immune Serum, after 1 round of immunization with rHER-2; 2MIS=2nd Mouse Immune Serum, 2 immunizations; 3MIS=3rd Mouse Immune Serum, 3 immunizations;

Fig. 5b: Kaplan-Meier survival curve of BT474 tumor grafts in immunocompetent BALB/c mice.

Fig. 6: Surface staining of HLA-molecules on human cancer cell lines BT474 and A431 compared to human CD3+ T-cells.

BT474 cells show reduced surface expression of HLA-molecules (MFI=34.29) compared to CD3+ Tcells (MFI=609.76) and A431 cells (MFI=685.39). Isotype controls (dotted lines) show only background staining on both cell lines.