Oncogenic RAS simultaneously protects against anti-EGFR ... - Nature

Oncogene (2013) 32, 2873–2881 & 2013 Macmillan Publishers Limited All rights reserved 0950-9232/13 www.nature.com/onc

ORIGINAL ARTICLE

Oncogenic RAS simultaneously protects against anti-EGFR antibody-dependent cellular cytotoxicity and EGFR signaling blockade S Kasper1, F Breitenbuecher1, H Reis2, S Brandau3, K Worm2, J Ko¨hler1, A Paul4, T Trarbach1, KW Schmid2 and M Schuler1 Monoclonal antibodies against the epidermal growth factor receptor (EGFR) are effective cancer therapeutics, but tumors harboring RAS mutations are resistant. To functionally dissect RAS-mediated resistance, we have studied clinically approved anti-EGFR antibodies, cetuximab and panitumumab, in cancer models. Both antibodies were equally cytotoxic in vitro. However, cetuximab, which also triggers antibody-dependent cellular cytotoxicity (ADCC), was more effective than panitumumab in vivo. Oncogenic RAS neutralized the activity of both antibodies in vivo. Mechanistically, RAS upregulated BCL-XL in cancer cell lines and in primary colorectal cancers. Suppression of BCL-XL by short hairpin RNA or treatment with a BH3 mimetic overcame RAS-mediated antibody resistance. In conclusion, RAS-mutant tumors escape anti-EGFR antibody-mediated receptor blockade as well as ADCC in vivo. Pharmacological targeting of RAS effectors can restore sensitivity to antibody therapy. Oncogene (2013) 32, 2873–2881; doi:10.1038/onc.2012.302; published online 16 July 2012 Keywords: cetuximab; colorectal cancer; RAS; ADCC; anti-EGFR antibodies; BCL-XL

INTRODUCTION Overexpression of the epidermal growth factor receptor (EGFR) is observed in several human cancers. It associates with adverse prognosis in patients with colorectal cancer (CRC) and head and neck cancer (HNSCC), thus defining the EGFR as an attractive therapeutic target.1 Combining anti-EGFR antibodies cetuximab or panitumumab with cytotoxic chemotherapy in patients with metastatic CRC increased remission rates, prolonged progressionfree survival and, in some clinical trials, increased overall survival. Importantly, more patients with isolated liver metastases were converted to a curatively resectable stage when chemotherapy was complemented by an anti-EGFR antibody. In addition, monotherapy with anti-EGFR antibodies has clinical activity in metastasized CRC patients relapsing after prior chemotherapies.2–5 Adding the anti-EGFR antibody cetuximab to standard chemotherapy increased response rates, progressionfree survival and overall survival in patients with recurrent or metastatic HNSCC.6 Also, locoregional control of advanced HNSCC was improved when cetuximab was applied simultaneously with radiotherapy.7 Recently, several retrospective analyses of studies conducted in CRC patients have suggested that the clinical benefit from both clinically approved anti-EGFR antibodies, cetuximab and panitumumab, is restricted to patients suffering from tumors devoid of somatic mutations of KRAS and additional RAS family oncogenes.4,8–11 Accordingly, more than 40% of all metastatic CRC patients are excluded from this therapeutic modality. Anti-EGFR antibodies block the receptor to interfere with ligand binding and possibly receptor dimerization, thus preventing signal transduction via the EGFR tyrosine kinase. The RAS family gene

products are involved in activating the mitogen-activated protein kinase pathway and additional pathways downstream of the EGFR. Hence, the current model of mutant RAS-mediated resistance to antibody therapy focuses on the ability of oncogenic RAS to compensate for upstream signals lacking owing to EGFR blockade by the antibody. In analogy, receptor blockade is regarded as the main effector mechanism of the two clinically approved anti-EGFR antibodies, cetuximab and panitumumab. Cetuximab is a chimeric IgG1 antibody capable to mediate antibody-dependent cellular cytotoxicity (ADCC) in vitro and in vivo.12,13 In contrast, panitumumab is unable to induce classical, natural killer (NK) cell-mediated ADCC owing to its IgG2 isotype.14 So far, no head-tohead comparison of these two antibodies in a clinical trial or in a preclinical in vivo model has been reported. Accordingly, it remains speculative whether the additional ADCC activity of cetuximab results in enhanced therapeutic activity over receptor blockade alone. Interestingly, Fc receptor-g polymorphisms, which may determine the efficacy of ADCC, have been correlated with outcome following cetuximab-based treatment in a retrospective analysis of CRC patients. This correlation was maintained in RASmutant and RAS-wild-type tumors.15 In vitro, RAS-mutant cancer cell lines were protected against direct antiproliferative actitivies of cetuximab and panitumumab, but not against ADCC enabled by cetuximab.16 These lines of evidence argue that CRC susceptibility to cetuximab-mediated ADCC is not determined by the mutation status of RAS proto-oncogenes. However, no apparent clinical benefit of cetuximab was derived in patients suffering from RAS-mutant CRC despite the theoretical possibility of preserved ADCC.

1 Department of Medical Oncology, West German Cancer Center, University Hospital Essen, University Duisburg-Essen, Essen, Germany; 2Department of Pathology and Neuropathology, West German Cancer Center, University Hospital Essen, University Duisburg-Essen, Essen, Germany; 3Department of Otorhinolaryngology, West German Cancer Center, University Hospital Essen, University Duisburg-Essen, Essen, Germany and 4Department of General, Visceral and Transplant Surgery, West German Cancer Center, University Hospital Essen, University Duisburg-Essen, Essen, Germany. Correspondence: Professor Dr M Schuler, Department of Medical Oncology, West German Cancer Center, University Hospital Essen, University Duisburg-Essen, Hufelandstrasse 55, 45147 Essen, Germany. E-mail: [email protected] Received 12 July 2011; revised 1 May 2012; accepted 31 May 2012; published online 16 July 2012

RAS-mediated antibody resistance S Kasper et al

2874 A full molecular understanding of relevant effector mechanisms of, as well as resistance mechanisms to, anti-EGFR antibody therapy will allow the development of strategies (i) for potentially synergistic combination treatments, and (ii) to overcome antibody resistance in RAS-mutant cancers. To this end, we have established preclinical models for the study of anti-EGFR therapy in vitro and in vivo. We find that ADCC indeed provides a therapeutic benefit in addition to EGFR blockade in vivo. However, oncogenic RAS completely abolishes both, receptor blockade and ADCC, by cetuximab in vivo. Dissecting RAS-mediated signaling alterations in cancer models and primary CRC, we identify anti-apoptotic BCL-XL as a clinically and functionally relevant target for therapeutic modulation of anti-EGFR antibody resistance. RESULTS Diverging efficacies of cetuximab and panitumumab in cancer models in vitro and in vivo Using a panel of EGFR-expressing cell lines, we confirmed that the cytotoxic activity of the anti-EGFR antibodies, cetuximab and panitumumab, in vitro strictly correlated with the extent of EGFR expression and KRAS mutation status (Supplementary Table 1). For further studies, we selected two cell lines, which exhibited a profound reduction in clonogenic survival upon antibody treatment in vitro (Figures 1a and b). In short-term assays, anti-EGFR antibodies effectively induced apoptosis in Difi cells, whereas A431 cells primarily responded by reduced proliferation (Figures 1c and d). Both anti-EGFR antibodies, cetuximab and panitumumab, were equally effective in vitro showing significant cytotoxicity as compared with the chimeric monoclonal anti-CD20 antibody rituximab (IgG1), which served as negative control (not shown). To study antibody therapy in a host context, tumors were established by subcutaneous injection of A431 or Difi cells in

Cetuximab

non-obese diabetic severe combined immunodeficiency (NOD/ SCID) mice. Injection of either cell line reproducibly led to outgrowth of flank tumors, which could be followed in vivo. Two models were established: (A) in a ‘therapeutic setting’, NOD/SCID mice bearing palpable flank tumors were treated twice weekly by intraperitoneal antibody injections (1 and 0.1 mg), and tumor growth was monitored. (B) In an ‘adjuvant setting’, intraperitoneal antibody injections (1 mg twice weekly) were initiated 1 day after the subcutaneous implantation of A431 cells in NOD/SCID mice, and tumor development and survival were monitored. Applying cetuximab, panitumumab and rituximab in therapeutic model A, we observed a dose-dependent activity of both anti-EGFR antibodies (Figure 2 and data not shown). In the ‘adjuvant’ model B, cetuximab prevented tumor development more effectively than panitumumab, resulting in superior survival of cetuximab-treated mice (Figures 3a–c). As both anti-EGFR antibodies were equally effective in vitro (Figure 1), where cytotoxic effects are solely mediated by EGFR blockade, we reasoned whether the superiority of cetuximab in vivo resulted from its ability to trigger ADCC. Canonical ADCC is executed by Fc receptor-bearing NK cells, which kill antibodymarked target cells by granule-mediated cytotoxicity.17 While NOD/SCID mice are lacking B and T cells, NK cell activity is at least partially preserved and thus can contribute to cetuximab activity in vivo.18 To formally prove this hypothesis, NOD/SCID mice were radiodepleted (1.5 Gy) of cellular ADCC effectors prior to anti-EGFR antibody treatment. Interestingly, the in vivo activity of cetuximab was markedly reduced in irradiated NOD/SCID mice as compared with untreated mice. In contrast, panitumumab was equally effective in irradiated and non-irradiated NOD/SCID mice (Figure 3d; Supplementary Table 2). These findings strongly suggest that ADCC mediated by radiosensitive cellular immune effectors at least in part explains the superior therapeutic activity of cetuximab as compared with panitumumab in vivo.

Medium

Cet 1µg/ml

Cet 10µg/ml

Cet 20µg/ml

Cet 50µg/ml

Cet 100µg/ml

Panitumumab

% of medium control

100 80 60 40 20 0 0

10

25

50

100

antibody concentration [ng/ml] Cetuximab

100

Panitumumab

100 % of medium control

sub G1 [%]

80 60 40 20 0

80 60 40 20 0

0

25 50 100 250 antibody concentration [ng/ml]

500

0

2

4

6

8

10

cetuximab concentration [µg/ml]

Figure 1. EGFR-positive cancer cells are sensitive to anti-EGFR antibody therapy in vitro. Difi (a, c) and A431 (b, d) cancer cell lines were grown in the presence of cetuximab or panitumumab at the indicated concentrations. (a) Percentage of clones (normalized to medium control) of Difi cells treated with cetuximab (black columns) or panitumumab (gray columns; mean±s.d. of three independent experiments). (b) Representative photomicrographs of clones of A431 cells grown in the presence of medium or cetuximab (Cet) at the indicated concentrations. (c) Fraction of apoptotic Difi cells with subgenomic DNA content (sub-G1) following treatment with cetuximab (black columns) or panitumumab (gray columns; mean±s.d. of three independent experiments). (d) Proliferation of A431 cells grown in the presence of increasing concentrations of cetuximab. Mean values (±s.d.) of three independent MTT assays. Oncogene (2013) 2873 – 2881

& 2013 Macmillan Publishers Limited


2875 140

Rituximab

Cetuximab

tumor size [mm2]

120 100 80 60 40 20 0 0

10

20

30 40 days post implantation

50

60

Figure 2. Effective treatment of established transplant tumors by an anti-EGFR antibody in vivo. Palpable flank tumors were established by subcutaneous injection of Difi cells in NOD/SCID mice. After 21 days (arrow), tumor-bearing mice were treated twice weekly by intraperitoneal injections of cetuximab (1 mg, closed triangles) or the control antibody rituximab (1 mg, closed circles). Tumor growth was monitored by palpation and measured using a caliper. Mean bidimensional tumor sizes þ s.d. of a representative experiment (five mice per treatment group) are given.

Oncogenic RAS confers resistance to ADCC in vivo and to EGFR blockade in vitro and in vivo To assess the impact of oncogenic RAS on antibody effector mechanisms in identical cellular backgrounds, we stably expressed clinically occurring RAS mutants, HRASG12V, KRASG12V and NRASG12V, in anti-EGFR antibody-sensitive A431 and Difi cells. Functional transgene expression was confirmed by immunoblot analyses of various phosphoepitopes indicating enhanced constitutive and ligand-induced mitogen-activated protein kinase activation in cells expressing mutant RAS (Figure 4a and not shown). As expected each of the RAS mutants, equally protected EGFR-positive cancer cells against cetuximab- and panitumumabinduced cytotoxicity in vitro (Figures 4b–d and Supplementary Figure 1). Accordingly, we selected the HRASG12V model for further studies. To rule out oncogene-induced antigen suppression, we confirmed that EGFR expression was not lost in HRASG12Vexpressing cancer cells (Supplementary Figure 2). This argued for RAS-mediated resistance mechanisms at the level of intracellular signal transduction. HRASG12V-expressing tumors established in NOD/SCID mice were significantly less responsive to anti-EGFR antibody therapy in vivo (Figure 5a). Surprisingly, HRASG12V-expressing tumors exhibited identical growth rates under ‘adjuvant’ treatment with cetuximab and panitumumab in vivo (Figure 5b). Moreover, radiodepletion of cellular ADCC effectors did not alter tumor outgrowth from HRASG12V-positive cancer cells in NOD/SCID mice treated with cetuximab (Figure 5c), which was in stark contrast to the results obtained with RAS-wild-type cancer models (Figures 3a and d). These results provide a strong argument for oncogenic RAS simultaneously conferring resistance against cetuximab-induced ADCC as well as cetuximab- and panitumumab-imposed EGFR blockade in vivo. Oncogenic RAS protects against anti-EGFR antibody therapy by upregulation of BCL-XL The above findings indicate that oncogenic RAS either triggers two distinct resistance mechanisms against the cytotoxic effects of EGFR signaling blockade and ADCC, or mutant RAS confers crossresistance via a unifying molecular pathway. Cetuximab-marked HRASG12V-expressing cancer cells effectively stimulated NK cell & 2013 Macmillan Publishers Limited

degranulation in vitro (Supplementary Figure 3a). Thus, ADCC resistance due to HRASG12V-imposed changes in NK cell activation seemed unlikely. As expected for an IgG2 antibody, panitumumabmarked EGFR-expressing HRASG12V or control cells had no effect on NK cell activation in vitro (Supplementary Figure 3b). Genetic mouse models have defined the mitochondrial pathway of apoptotic caspase activation, which is regulated by the BCL-2 protein family, as a main death effector mechanism following growth factor withdrawal in vivo.19–21 In keeping, pharmacological suppression of EGFR-mediated oncogenic signals by gefitinib triggered ‘mitochondrial’ apoptosis.22 Recently, we have shown that inhibition of this apoptotic pathway either by transgenic expression of BCL-XL or by enhanced mitogen-activated protein kinase and protein kinase B signaling protects cancer cells against tumor suppression by activated T cells and NK cells in vivo.23–25 Against this background, we immunohistochemically profiled the expression of antiapoptotic BCL-2, BCL-XL and MCL-1 in primary human CRC specimens. While at least one of these BCL-2 proteins was detectable in almost all cancer samples, there was a high prevalence of strong cytoplasmic expression of BCL-XL (Figures 6a and b). Interestingly, a significantly higher proportion of KRASmutant cancers stained positively for BCL-XL as compared with KRAS-wild-type tumors (median 97.5±6.25% vs 85.0±13.8%, P ¼ 0.02). No significant differences in expression of BCL-2 (33.3% positive vs 30% positive, P ¼ 0.63) or MCL-1 (median 95.5±7.62% vs 95.0±8.67%, P ¼ 0.731) were observed between KRAS-mutant and KRAS-wild-type cancers (Figures 6c and d and Supplementary Table 3). Based on these findings in primary CRC, we studied the impact of oncogenic RAS on the expression of BCL-2 family proteins in EGFR-positive cancer cell lines. A431 cells expressing HRASG12V, KRASG12V or NRASG12V exhibited markedly higher BCL-XL RNA and protein levels than control cells (Figures 6f and g and not shown). Blocking protein synthesis with cycloheximide revealed significantly enhanced stability of BCL-XL, but not BCL-2 or MCL-1, in the presence of HRASG12V, arguing for transcriptional activation as well as posttranslational mechanisms of action (Figure 6e). Taken together, upregulation of anti-apoptotic BCL-XL is frequently observed in KRAS-mutant human CRC. Moreover, induction and posttranslational stabilization of BCL-XL by oncogenic RAS might contribute to resistance against anti-EGFR antibody therapy. To support this hypothesis at a functional level, we stably expressed BCL-XL in Difi cells and A431 cells, both of which are devoid of endogenous mutations of HRAS, KRAS and NRAS (Figure 7a). BCL-XL effectively protected these RAS-wild-type cancer models against cetuximab-induced apoptosis in vitro (Figure 7b). In vivo, cetuximab and panitumumab were blunted in preventing the outgrowth of BCL-XL-expressing tumors in NOD/ SCID mice (Figure 7c and Supplementary Table 4). For example, median size of A431-BCL-XL tumors following 15 days of cetuximab treatment in vivo was 64.49 mm2 (±17.67 mm2, Figure 7c), as compared with 0 mm2 for A431 control tumors (Figure 3a) (P ¼ 0.0018, unpaired t-test). Panitumumab-treated A431-BCL-XL tumors grew to a median size of 124.1 mm2 (±13.06 mm2, Figure 7c) at 15 days, as compared with 16.54 mm2 (±14.18 mm2, Figure 3a) for A431 control tumors (Po0.0001, unpaired t-test). Interestingly, no significant difference in growth rates of cetuximab- and panitumumab-treated A431BCL-XL tumors was observed (Figure 7c), which was reminiscent of the results obtained by treating RAS-wild-type tumors growing in leukocyte-depleted mice (Figure 3d) or HRASG12V-expressing tumors (Figure 5b). These findings argue that BCL-XL alike oncogenic RAS equally protects against EGFR blockade and ADCC in vivo, and thus can substitute its upstream regulator in mediating resistance to anti-EGFR antibody therapy. To study the contribution of BCL-XL to RAS-mediated resistance at a functional level, we devised short hairpin RNA (shRNA) Oncogene (2013) 2873 – 2881


2876 Rituximab

300

Pantitumumab

Cetuximab

100 Cetuximab Panitumumab Rituximab

75

200

survival [%]

tumor size [mm2]

250

150 100

50 25

50 0

0

Group

5

10 15 20 25 30 35 days post implantation

Established Tumors

40

Median Survival [days]

cetuximab

0/10

n.r.

panitumumab

9/10

47

rituximab

8/10

25

45

0

tumor size [mm2]

0

500 450 400 350 300 250 200 150 100 50 0

10 20 30 40 50 60 70 80 Day Rituximab

0

5

Panitumumab

Cetuximab

10 15 20 25 days post implantation

30

35

Figure 3. Cetuximab is superior over panitumumab in prevention of tumor outgrowth in vivo. (a) NOD/SCID mice received subcutaneous injections of A431 cells on day 0. Starting on day 1 (arrow), mice were treated twice weekly by intraperitoneal injections of panitumumab (open boxes), cetuximab (closed triangles), or the control antibody rituximab (closed circles, 1 mg each). Tumor development was monitored by palpation and tumors were measured using a caliper. Mean bidimensional tumor sizes ( þ s.d.) of five mice per treatment group are given. (b) Kaplan–Meier plots of tumor-free survival of NOD/SCID mice following subcutaneous injection of A431 cells. Mice were treated as in (a) with cetuximab (solid line), panitumumab (dashed line), or the control antibody rituximab (dotted line). Cetuximab-treated mice exhibited a significantly prolonged survival as compared with panitumumab-treated (P ¼ 0.0027) or rituximab-treated mice (P ¼ 0.0031, log rank test). (c) Numeric representation of tumor development and median survival (days) in relation to antibody treatment (10 mice per treatment group; n.r. denotes not reached). (d) Following total body irradiation (1.5 Gy) for in vivo leukocyte depletion, NOD/SCID mice were subcutaneously injected with A431 cells on day 0. Starting on day 1 (arrow), mice were treated with panitumumab (open boxes), cetuximab (closed triangles), or the control antibody rituximab (closed circles) as in (a). Tumor development was monitored by palpation and tumors were measured using a caliper. Mean bidimensional tumor sizes ( þ s.d.) of five mice per treatment group are given. Note that in contrast to experiment (a), cetuximab was not superior over panitumumab in the prevention of tumor growth in irradiated NOD/SCID mice.

technology in HRASG12V-expressing Difi cells (Figure 7d). Suppression of endogenous BCL-XL restored sensitivity of HRASG12Vexpressing cancer cells to cetuximab-induced apoptosis (Figure 7e). Moreover, subtoxic concentrations of ABT-737, a pharmacologic BH3 mimetic antagonizing the anti-apoptotic activities of BCL-2, BCL-XL and BCL-W, but not MCL-1,26–28 effectively sensitized HRASG12V-expressing cancer cells to cetuximab-induced apoptosis (Figure 7f). This suggests that mutant RAS confers anti-EGFR antibody resistance by blocking apoptotic caspase activation through a pathway regulated by BCL-XL. DISCUSSION Somatic mutations of RAS proto-oncogenes are associated with clinical resistance to anti-EGFR antibody therapy in CRC.11 Accordingly, application of the clinically approved antibodies, cetuximab and panitumumab, is restricted to patients suffering from CRC, in which at least somatic KRAS ‘hot spot’ mutations have been ruled out. As most CRC express the EGFR antigen, patients harboring RAS-mutant CRC in theory could still benefit from antibody-mediated immunotherapy via complement-dependent cytotoxicity or ADCC. However, the lack of clinical activity of cetuximab, which is able to activate NK cells mediating ADCC (Supplementary Figure 3a) in KRAS-mutant CRC, argues against this hypothesis. Surprisingly, certain Fc receptor polymorphisms were reported to independently associate with an improved Oncogene (2013) 2873 – 2881

outcome following cetuximab-based therapy. This association was maintained when KRAS mutation status was taken into account.15 The availability of panitumumab, a fully human IgG2 antibody unable to trigger NK cell-mediated ADCC14 (Supplementary Figure 3b), allows to resolve this apparent discrepancy and to dissect the relative contribution of ADCC to the antitumor activity of cetuximab. So far, no head-to-head comparison of cetuximab and panitumumab has been reported in the clinic. Here we have devised EGFR-expressing cancer models to address the role of ADCC in anti-EGFR antibody therapy in relation to resistance mediated by oncogenic RAS. Transplanting anti-EGFR antibody-sensitive human cancer cells in NOD/SCID mice, we observed therapeutic superiority of cetuximab over panitumumab in vivo, which was abolished by depletion of cellular immune effectors. This is a strong indication of ADCC contributing to cancer control by cetuximab at least in the present model. Similar observations were reported from antibody treatment studies of leukemia xenografts in NOD/SCID mice, which proved sensitive to depletion or stimulation of NK cells in vivo.18 As NK cell activity of NOD/SCID mice is reportedly blunted,29 the relative contribution of ADCC to the cytotoxic effect of cetuximab might actually be underestimated by our experimental systems. Nevertheless, expression of oncogenic RAS interfered with both antibody effector mechanisms, EGFR signaling blockade (panitumumab and cetuximab) and ADCC (cetuximab), in that growth of RAS-mutant tumors was enhanced and the difference in efficacy between these antibodies was & 2013 Macmillan Publishers Limited


2877 HRAS G12V +

+

-

-

-

+

-

+

+

+

-

-

EGF Cetuximab

-

+

-

P-EGFR EGFR p-RAF-1 RAF-1 p-ERK

* 100 % of medium control

+

ERK

control

*

60

* *

*

*

40 * 20 0

KRAS G12V

HRAS G12V

*

80

Actin

* 100 10 50 0 25 cetuximab concentration [ng/ml] HRAS G12V

control

100

100 80

*

*

80

sub G1 [%]

% of medium control

control HRAS G12V KRAS G12V NRAS G12V

control

* 60

* *

40

* 20 0

*

60

*

40 20

*

#

0 0

10 25 50 100 500 panitumumab concentration [ng/ml]

50 100 500 0 cetuximab concentration [ng/ml]

Figure 4. Oncogenic RAS protects cancer cells against anti-EGFR antibody-mediated signaling blockade and cytotoxicity in vitro. A431 and Difi cells were retrovirally transduced to stably express the oncogenic RAS mutants HRASG12V, KRASG12V or NRASG12V. (a) Inhibition of constitutive and ligand-induced (EGF 10 ng/ml) phosphorylation of EGFR, and mitogen-activated protein kinase signal transducers RAF and ERK by cetuximab (1 mg/ml) in A431 cells expressing HRASG12V or controls. (b, c) Clonogenic survival of Difi cells expressing HRASG12V (gray columns), KRASG12V (checked bars), NRASG12V (striped bars) and controls (black bars) cultured in the presence of cetuximab (b) or panitumumab (c). Mean colony numbers ( þ s.d.) normalized to medium control from three independent experiments are given. (d) Induction of apoptosis by cetuximab in Difi cells expressing HRASG12V (gray columns) and controls (black bars). Mean percentages ( þ s.d.) of cells with subgenomic DNA content from three independent experiments are given. Asterisks (*) denote a statistically significant (Po0.05, t-test) difference in apoptosis observed in RAS-mutant cells as compared with control cells.

300

control

450

HRAS G12V

250

300

Panitumumab

150 100 50

tumor size [mm2]

200

300 250 200 150 100

10 20 30 days post implantation

40

200 150 100

0

0 0

Panitumumab

50

50

0

Cetuximab

250

350 tumor size [mm2]

tumor size [mm2]

Cetuximab

400

0

5

10

15 20 25 30 35 days post implantation

40

45

0

5

10 15 20 25 30 days post implantation

35

Figure 5. Oncogenic RAS protects tumors against anti-EGFR antibody-mediated direct cytotoxicity as well as ADCC in vivo. (a) Tumor growth following injection of Difi cells expressing HRASG12V (open boxes) or control cells (closed triangles) in NOD/SCID mice. After 21 days (arrow), mice were treated biweekly with intraperitoneal injections of cetuximab (0.1 mg). Mean bidimensional tumor sizes ( þ s.d.) of five mice per group are given. (b) NOD/SCID mice received subcutaneous injections of A431 cells expressing HRASG12V on day 0. Starting on day 1 (arrow), mice were treated with biweekly intraperitoneal injections of cetuximab (closed triangles) or panitumumab (open boxes, 1 mg each) and tumor growth was monitored. Mean bidimensional tumor sizes ( þ s.d.) of five mice per group are given. (c) Following total body irradiation (1.5 Gy) for in vivo leukocyte depletion, NOD/SCID mice were subcutaneously injected with A431 cells expressing HRASG12V on day 0. Starting on day 1 (arrow) mice were treated with biweekly intraperitoneal injections of cetuximab (closed triangles) or panitumumab (open boxes, 1 mg each) and tumor growth was monitored. Mean bidimensional tumor sizes ( þ s.d.) of five mice per group are given.

neutralized. Hence, it can be concluded that oncogenic RAS mediates protection against direct (inhibition of signal transduction) and indirect (immunological) cytotoxic activities of anti-EGFR antibodies in vivo. These findings are in line with recent observations from our group and others indicating that the responsiveness of cancer cells to cytotoxic cellular immune effectors is also determined by & 2013 Macmillan Publishers Limited

their ability to execute the ‘mitochondrial’ pathway of apoptotic caspase activation.23–25,30 Deregulated growth and survival factor signaling is a common phenomenon in human cancer.31 Frequently, this results in increased expression and function of anti-apoptotic BCL-2 family proteins, as observed for BCL-XL in our present model of RAS-mediated antibody resistance. As direct pharmacological targeting of oncogenic RAS so far has not been Oncogene (2013) 2873 – 2881


2878

BCL-2-

BCL-2+

BCL-XL +

BCL-XL ++

BCL-XL +++

MCL-1 +

MCL-1 ++

MCL-1 +++

57494/04 9902/03 22197/04 6282/02 22272/06 10427/02 20136/05 1577/04 39019/04 8677/01 12810/05 17532/04 16893/04 10199/05 20556/04 21578/04 11631/99 10543/04 62106/03 21442/02 3365/0B 18311/04 22197/04 4235/00

12810/05 6282/02 10427/02 39486/04 6282/03 22272/06 21442/02 22439/04 20556/04 1577/04 22197/04 9902/03 8677/01 39019/04 26955/06 13123/03 16893/04 11631/99 57494/04 8524/03 20136/05 17532/04 125/04 3365/0B 21442/02 62106/03 10543/04 22197/04 18311/04 21578/04 10199/05 25801/03 4235/00

BCL-XL

0

MCL-1

8 4 6 Immunoreactive Score

2

KRAS mutant

KRAS wild type 7%

18%

27%

9%

N=15

KRAS mutant

KRAS wild type

10% 10%

14%

40%

2V

ntr

ol

G1

co

AS HR

AS

G1

G1

KR

12

AS

10

2V

2V

N=7

NR

6 4 8 Immunoreactive Score

2

MCL-1 IRS 6 MCL-1 IRS 4

57%

N=10

MCL-1 IRS 12 MCL-1 IRS 9 MCL-1 IRS 8

14% 14% 40%

0

BCL-XL IRS 12

BCL-XL IRS 6 BCL-XL IRS 4

13% 13%

N=11

12

BCL-XL IRS 9 BCL-XL IRS 8

40%

64%

9%

10

BCL-XL 6h

2h

4h

0h

time with CHX

Actin

BCL-2

BCL-XL Actin +

-

+

-

+

-

+

-

HRAS G12V

relative expression

MCL-1

3

BCL-XL transcript level

*

2.5 2

*

1.5 1 0.5 0 control HRAS G12V HRAS G12V -Puro -Hygro

Figure 6. Upregulation of anti-apoptotic BCL-XL by oncogenic RAS. (a) Representative photomicrographs of immunohistochemically detectable BCL-2, BCL-XL and MCL-1 expression in tissue microarrays from 47 human CRC specimens. (b) Immunoreactive score (IRS) of BCLXL expression in CRC specimens from 33 patients in relation to KRAS and BRAF mutation status. Each bar represents the IRS of an individual sample. Black bars, KRASwild type/BRAFwild type tumors; blue bars, BRAFmutant tumors; red bars, KRASmutant tumors; gray bars, mutation analysis not informative. (c) IRS of MCL-1 expression in CRC specimens from 24 patients in relation to KRAS and BRAF mutation status. Each bar represents the IRS of an individual sample; color coding of bars as in (b). (d) Distrubution of immunhistochemical staining intensity (IRS) for BCL-XL (upper diagrams) and MCL-1 (lower diagrams) in KRASwild type and KRASmutant tumors. (e) Expression of anti-apoptotic BCL-2, MCL-1 and BCL-XL in A431 cells stably expressing HRASG12V ( þ ) and control cells ( ). To prevent new protein synthesis, cells were treated with cycloheximide (CHX, 10 mg/ml) for the indicated periods. (f ) Expression of BCL-XL in A431 cells stably expressing the oncogenic RAS mutants HRASG12V, KRASG12V or NRASG12V and A431 control cells. (g) Relative BCL-XL transcript level in Difi cells expressing HRASG12V from two different retroviral vectors (pBabe.puro and pBabe.hygro) as compared with control cells. Expression levels of BCL-XL were normalized to the housekeeping gene beta-ACTIN. Mean change ( þ s.d.) of BCL-XL transcript levels from three independent experiments are given. An asterisk (*) denotes a statistically significant (Po0.05, t-test) difference.

Oncogene (2013) 2873 – 2881



2879 100

BCL-XL

BCL-XL

sub G1 [%]

l on

tro

XL CL -

Di fi c

31

Di fi B

co

BC L-

nt ro

l

XL

80

A4

31 A4

control

60

*

*

* 40 * 20

Actin

0 0

400

cetuximab panitumumab

350 tumor size [mm2]

25 50 100 500 cetuximab concentration [ng/ml]

502

300

501

500

499

ctrl parental shRNA clone

250

BCL-XL

200 Actin

150 100 50 0 0

5

10

15

20

25

30

35

60 50

ctrl.

shRNA BCL-XL

*

80

40 30 20

60 40 20

10

0 Cetuximab

HRAS G12V

control

100

* sub G1 [%]

relative induction of apoptosis to ctrl. [%]

days post implantation

25ng/ml

50ng/ml

0 Cetuximab ABT-737

-

+ -

+

+ +

-

+ -

+

+ +

Figure 7. BCL-XL protects against cetuximab-mediated cytotoxicity in vitro and in vivo. (a) A431 and Difi cells were retrovirally transduced to stably express a human BCL-XL cDNA. (b) Induction of apoptosis by cetuximab treatment in Difi cells expressing BCL-XL (gray columns) or control cells (black columns). Mean percentages ( þ s.d.) of cells with subgenomic DNA content from three independent experiments. An asterisk (*) denotes a statistically significant (Po0.05, t-test) protection against apoptosis by BCL-XL. (c) NOD/SCID mice received subcutaneous injections of A431 cells expressing BCL-XL on day 0. Starting on day 1 (arrow), mice were treated with biweekly intraperitoneal injections of cetuximab (closed triangles) or panitumumab (open boxes, 1 mg each) and tumor growth was monitored. Mean bidimensional tumor sizes ( þ s.d.) of five mice per group are given. (d) BCL-XL protein levels in Difi-HRASG12V cells stably expressing four different lentiviral shRNA expression vectors targeting BCL-XL (499–502), a control shRNA vector (ctrl) and parental Difi-HRASG12V cells. (e) Difi-HRASG12V cells stably expressing shRNA vector 499 (black columns) or a control vector (gray columns) were treated with cetuximab (25 and 50 ng/ml). Mean relative percentages ( þ s.d.) of apoptotic cells with subgenomic DNA content from three independent experiments are given. An asterisk (*) denotes a statistically significant (Po0.05, t-test) difference. (f ) Difi-HRASG12V cells (black columns) or control cells (gray columns) were treated with cetuximab (50 ng/ml), the BH3-mimetic ABT-737 (10 nM), or combinations of both. Mean percentages ( þ s.d.) of apoptotic cells with subgenomic DNA content from three independent experiments are given.

successfully achieved, modulation of downstream effectors is an alternative therapeutic strategy. Using subtoxic concentrations of a BH3 mimetic, which antagonizes the activity of BCL-XL, we were indeed able to sensitize HRASG12V-expressing cancer cells to cetuximab-induced cytotoxicity. This provides proof-of-concept for a rational combination therapy approach to expand the population of cancer patients potentially benefitting from antiEGFR antibody therapy.

MATERIALS AND METHODS Cell lines and reagents The human EGFR-positive cancer cell lines A431 and Difi were obtained from DSMZ (Braunschweig, Germany) and Dr Robert Coffey (Nashville, TN, USA), respectively. All cells were cultured in DMEM medium supplemented with 10% fetal bovine serum (PAA, Coelbe, Germany), L-glutamine, penicillin and streptomycin (Invitrogen, Frankfurt, Germany). Stable expression of the HRASG12V, KRASG12V, NRASG12V and BCL-XL complementary DNAs (cDNAs) was achieved by retroviral transduction as described & 2013 Macmillan Publishers Limited

previously.32 Lentiviral vectors encoding short hairpin RNA (shRNA) from the MISSION TRC-Mm 1.0 library were purchased from Sigma-Aldrich (Munich, Germany); clones TRCN 0000033499-502 were used for BCL-XL suppression, and the MISSION pLKO.1-puro non-target shRNA plasmid (SHC016-1EA) served as negative control. Clinical grade cetuximab (Erbitux, Merck Serono, Darmstadt, Germany), panitumumab (Vectibix, Amgen, Thousand Oaks, CA, USA) and rituximab (Mabthera, Roche, GrenzachWyhlen, Germany) were purchased from the pharmacy of the University Hospital Essen; ABT-737 was provided by Abbott Laboratories (Abbott Park, IL, USA). For RNA expression analysis, total RNA was isolated (High Pure RNA Isolation Kit, Roche Diagnostics, Mannheim, Germany) and reversely transcribed into cDNA (Transcription High Fidelity cDNA Synthesis Kit, Roche Diagnostics) following the manufacturer’s instructions. Quantitative RT–PCR analysis was performed on a LC480 instrument (Roche Diagnostics) using SYBR Green 1 Master chemistry (Roche Diagnostics) and primers for human BCL-XL: 50 -GGCTGGGATACTTTTGTGGA-30 , 50 -TGTCTGGTCATTTCCGACTG-30 and human ACTIN: 50 -TCAGCTGTGGGGTCCTGT-30 , 50 -GAAGGGACAGGCAGTGAG-30 as previously described.33 The following primary antibodies were used for immunoblotting and immunolabeling following standard protocols: RAS, BCL-2 (C2), RAF-1, phospho-RAF-1S338 (all from Santa Cruz Biotechnology, Santa Cruz, CA, USA), phospho-ERK1/2, ERK1/2, Oncogene (2013) 2873 – 2881


2880 BCL-XL (54H6), EGFR, phopho-EGFRY1068 (all from Cell Signaling Technology, Danvers, MA, USA), actin (C4, ICN, Irvine, CA, USA), MCL-1 (Epitomics, Burlingame, CA, USA), EGFR-Phycoerythrin (R&D-Systems, Minneapolis, MN, USA).

Animal models All animal studies were conducted in compliance with institutional guidelines and German Animal Protection Law, and were approved by the responsible regulatory authority (Landesamt fu¨r Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen, Az. G969/08). NOD/SCID mice (Charles River Laboratories, France) received single subcutaneous flank injections of 2 106 A431 or 1 107 Difi cells suspended in 200 ml saline. Animals were monitored for tumor development twice weekly, and tumor growth was bidimensionally quantified using a caliper. Antibodies were dissolved in 200 ml saline and were administered as biweekly intraperitoneal injections. For in vivo leukocyte depletion, mice received a single fraction of total body irradiation (1.5 Gy) from a cobald source. Kaplan– Meier plots for tumor-free survival were analyzed using the log rank test. For statistical analysis of tumor size the unpaired t-test was used.

Cellular assays For clonogenic survival analysis, 5 103 A431 or Difi cells were seeded in six-well plates in the presence of cetuximab, panitumumab or rituximab. Following incubation for 7 to 14 days, colonies were fixed with ethanol (70% v/v) for 30 min, stained with Coomassie brilliant blue and automatically counted using an Infinity-100 System and the Vision Capt software (Vilber Lourmat, Eberhardzell, Germany). Proliferation was quantified by means of the MTT assay according to the manufacturer’s instruction (Roche, Mannheim, Germany). Apoptosis was quantified by flow cytometric determination of cells with subgenomic DNA content following hypotonic lysis and staining with propidium iodide as previously described.32 All results were obtained from at least three independent experiments. For statistical analysis, the unpaired t-test was used.

NK cell isolation and degranulation assay For NK cell isolation, lymphocytes from healthy donors were enriched by Ficoll centrifugation and purified using an NK cell isolation system including magnetic bead-coupled specific antibodies from Miltenyi Biotec (Bergisch Gladbach, Germany) according to the manufacturer’s instructions. Upregulation of surface CD107a was measured to quantify NK cell degranulation. Unstimulated or stimulated (200 U rhIL-2 for 24 h) NK cells were co-incubated with cetuximab- or panitumumab (1, 3 and 10 mg/ml)coated A431 cells at an effector:target cell ratio of 1:1 for 1 h in the presence of FITC-coupled anti-CD107. Following addition of monensin, cells were further incubated for 5 h and then stained with fluorochromeconjugated antibodies against CD56, CD16 or isotype control antibody (BD Biosciences, Heidelberg, Germany). Cells were washed, fixed and analyzed by flow cytometry following standard protocols.

Analysis of primary tumor samples Clinical data and surplus tumor specimens were retrieved from 47 patients with metastatic CRC. Tissue microarrays were prepared from formalin-fixed, paraffin-embedded tumor samples. Sections were subjected to hematoxilin & eosin staining and immunohistochemical analyses following diagnostic protocols of the Department of Pathology and Neuropathology, University Hospital Essen. Tumor-specific expression of each protein was quantified applying an immunoreactive score.34 For statistical analysis, the Mann–Whitney test (for BCL-XL and MCL-1) and the Fisher–Yates test (for BCL-2) was used. Tumor DNA was isolated from formalin-fixed, paraffinembedded tumor sections following microdissection, and KRAS and BRAF mutation status was determined by PCR amplification of the relevant exons followed by direct Sanger sequencing. All studies involving patient data or biosamples were approved by the Ethics Committee of the Medical Faculty of the University Duisburg-Essen.

CONFLICT OF INTEREST Martin Schuler has served as consultant to Amgen; Tanja Trarbach has received research funding, consulting and speaker honoraria from Merck Serono and Amgen. The other authors declare no conflict of interest.

Oncogene (2013) 2873 – 2881

ACKNOWLEDGEMENTS We thank Sandra Hoffarth, Sarah-Luise Stergar, Kirsten Bruderek, Jeannette Markowetz, Anna Even, Miriam Backs, Ali Sak, Sabine Harde and the staff of the Central Animal Facility, and the Pathology and Molecular Pathology Laboratories, University Hospital Essen, for their support. Robert Coffey, Roman Thomas, Hyatt Balke-Want, Scott W Lowe, Gary P Nolan and Abbott are acknowledged for providing reagents. This work was funded by grants from the Wilhelm Sander-Stiftung (2005.136.3, MS), the Deutsche Forschungsgemeinschaft (SCHU 1541/5-1, MS), Mercator Research Center Ruhr-MERCUR (An-2011-0031, SK) a CESAR Research Fellowship (SK), the Wiedenfeld-Stiftung (SK) and the IFORES program of the Medical Faculty of the University Duisburg-Essen (MS).

REFERENCES 1 Mayer A, Takimoto M, Fritz E, Schellander G, Kofler K, Ludwig H. The prognostic significance of proliferating cell nuclear antigen, epidermal growth factor receptor, and mdr gene expression in colorectal cancer. Cancer 1993; 71: 2454–2460. 2 Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, Santoro A et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004; 351: 337–345. 3 Jonker DJ, O’Callaghan CJ, Karapetis CS, Zalcberg JR, Tu D, Au HJ et al. Cetuximab for the treatment of colorectal cancer. N Engl J Med 2007; 357: 2040–2048. 4 Amado RG, Wolf M, Peeters M, Van CE, Siena S, Freeman DJ et al. Wild-type KRAS is required for panitumumab efficacy in patients with metastatic colorectal cancer. J Clin Oncol 2008; 26: 1626–1634. 5 Van CE, Peeters M, Siena S, Humblet Y, Hendlisz A, Neyns B et al. Open-label phase III trial of panitumumab plus best supportive care compared with best supportive care alone in patients with chemotherapy-refractory metastatic colorectal cancer. J Clin Oncol 2007; 25: 1658–1664. 6 Vermorken JB, Mesia R, Rivera F, Remenar E, Kawecki A, Rottey S et al. Platinumbased chemotherapy plus cetuximab in head and neck cancer. N Engl J Med 2008; 359: 1116–1127. 7 Bonner JA, Harari PM, Giralt J, Azarnia N, Shin DM, Cohen RB et al. Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med 2006; 354: 567–578. 8 Karapetis CS, Khambata-Ford S, Jonker DJ, O’Callaghan CJ, Tu D, Tebbutt NC et al. K-ras mutations and benefit from cetuximab in advanced colorectal cancer. N Engl J Med 2008; 359: 1757–1765. 9 Van CE, Kohne CH, Hitre E, Zaluski J, Chang CCR, Makhson A et al. Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer. N Engl J Med 2009; 360: 1408–1417. 10 Bokemeyer C, Bondarenko I, Makhson A, Hartmann JT, Aparicio J, de BF et al. Fluorouracil, leucovorin, and oxaliplatin with and without cetuximab in the firstline treatment of metastatic colorectal cancer. J Clin Oncol 2009; 27: 663–671. 11 De RW, Claes B, Bernasconi D, De SJ, Biesmans B, Fountzilas G et al. Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy of cetuximab plus chemotherapy in chemotherapy-refractory metastatic colorectal cancer: a retrospective consortium analysis. Lancet Oncol 2010; 11: 753–762. 12 Kurai J, Chikumi H, Hashimoto K, Yamaguchi K, Yamasaki A, Sako T et al. Antibodydependent cellular cytotoxicity mediated by cetuximab against lung cancer cell lines. Clin Cancer Res 2007; 13: 1552–1561. 13 Hara M, Nakanishi H, Tsujimura K, Matsui M, Yatabe Y, Manabe T et al. Interleukin2 potentiation of cetuximab antitumor activity for epidermal growth factor receptor-overexpressing gastric cancer xenografts through antibody-dependent cellular cytotoxicity. Cancer Sci 2008; 99: 1471–1478. 14 Schneider-Merck T, Lammerts van Bueren JJ, Berger S, Rossen K, van Berkel PH, Derer S et al. Human IgG2 antibodies against epidermal growth factor receptor effectively trigger antibody-dependent cellular cytotoxicity but, in contrast to IgG1, only by cells of myeloid lineage. J Immunol 2010; 184: 512–520. 15 Bibeau F, Lopez-Crapez E, Di FF, Thezenas S, Ychou M, Blanchard F et al. Impact of Fc{gamma}RIIa-Fc{gamma}RIIIa polymorphisms and KRAS mutations on the clinical outcome of patients with metastatic colorectal cancer treated with cetuximab plus irinotecan. J Clin Oncol 2009; 27: 1122–1129. 16 Barriere J, Fischel J, Formento P, Renee N, Francoual MFJ, Chefour M et al. Cetuximab-mediated antibody-dependent cellular cytotoxicity (ADCC) against tumor cell lines characterized for EGFR expression and K-ras mutation. J Clin Oncol 2010; 27: e14583. 17 Kagi D, Ledermann B, Burki K, Seiler P, Odermatt B, Olsen KJ et al. Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice. Nature 1994; 369: 31–37. 18 Piloto O, Nguyen B, Huso D, Kim KT, Li Y, Witte L et al. IMC-EB10, an anti-FLT3 monoclonal antibody, prolongs survival and reduces nonobese diabetic/severe combined immunodeficient engraftment of some acute lymphoblastic leukemia cell lines and primary leukemic samples. Cancer Res 2006; 66: 4843–4851.



2881 19 Strasser A, Harris AW, Vaux DL, Webb E, Bath ML, Adams JM et al. Abnormalities of the immune system induced by dysregulated bcl-2 expression in transgenic mice. Curr Top Microbiol Immunol 1990; 166: 175–181. 20 Knudson CM, Tung KS, Tourtellotte WG, Brown GA, Korsmeyer SJ. Bax-deficient mice with lymphoid hyperplasia and male germ cell death. Science 1995; 270: 96–99. 21 Yoshida H, Kong YY, Yoshida R, Elia AJ, Hakem A, Hakem R et al. Apaf1 is required for mitochondrial pathways of apoptosis and brain development. Cell 1998; 94: 739–750. 22 Paez JG, Janne PA, Lee JC, Tracy S, Greulich H, Gabriel S et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004; 304: 1497–1500. 23 Huber C, Bobek N, Kuball J, Thaler S, Hoffarth S, Huber C et al. Inhibitors of apoptosis confer resistance to tumour suppression by adoptively transplanted cytotoxic T-lymphocytes in vitro and in vivo. Cell Death Differ 2005; 12: 317–325. 24 Marques CA, Hahnel PS, Wolfel C, Thaler S, Huber C, Theobald M et al. An immune escape screen reveals Cdc42 as regulator of cancer susceptibility to lymphocytemediated tumor suppression. Blood 2008; 111: 1413–1419. 25 Hahnel PS, Thaler S, Antunes E, Huber C, Theobald M, Schuler M et al. signaling sensitizes cancer to cellular immunotherapy. Cancer Res 2008; 68: 3899–3906. 26 Oltersdorf T, Elmore SW, Shoemaker AR, Armstrong RC, Augeri DJ, Belli BA et al. An inhibitor of Bcl-2 family proteins induces regression of solid tumours. Nature 2005; 435: 677–681. 27 van Delft MF, Wei AH, Mason KD, Vandenberg CJ, Chen L, Czabotar PE et al. The BH3 mimetic ABT-737 targets selective Bcl-2 proteins and efficiently

28

29

30

31 32

33

34

induces apoptosis via Bak/Bax if Mcl-1 is neutralized. Cancer Cell 2006; 10: 389–399. Wesarg E, Hoffarth S, Wiewrodt R, Kroll M, Biesterfeld S, Huber C et al. Targeting BCL-2 family proteins to overcome drug resistance in non-small cell lung cancer. Int J Cancer 2007; 121: 2387–2394. Shultz LD, Schweitzer PA, Christianson SW, Gott B, Schweitzer IB, Tennent B et al. Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. J Immunol 1995; 154: 180–191. Ravi R, Fuchs EJ, Jain A, Pham V, Yoshimura K, Prouser T et al. Resistance of cancers to immunologic cytotoxicity and adoptive immunotherapy via X-linked inhibitor of apoptosis protein expression and coexisting defects in mitochondrial death signaling. Cancer Res 2006; 66: 1730–1739. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011; 144: 646–674. Schuler M, Maurer U, Goldstein JC, Breitenbucher F, Hoffarth S, Waterhouse NJ et al. p53 triggers apoptosis in oncogene-expressing fibroblasts by the induction of Noxa and mitochondrial Bax translocation. Cell Death Differ 2003; 10: 451–460. Kasper S, Breitenbuecher F, Heidel F, Hoffarth S, Markova B, Schuler M et al. Targeting MCL-1 sensitizes FLT3-ITD-positive leukemias to cytotoxic therapies. Blood Cancer J 2012; 2: e60. Remmele W, Stegner HE. [Recommendation for uniform definition of an immunoreactive score (IRS) for immunohistochemical estrogen receptor detection (ERICA) in breast cancer tissue]. Pathologe 1987; 8: 138–140.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)


Oncogene (2013) 2873 – 2881