dasatinib inhibits mammary tumour development in a genetically engineered mouse model

Journal of PathologyJ Pathol 2013; 230: 430–440Published online in Wiley Online Library(wileyonlinelibrary.com) DOI: 10.1002/path.4202

ORIGINAL PAPER

Dasatinib inhibits mammary tumour development in a geneticallyengineered mouse modelSaadia A Karim,1,# Helen Creedon,2,# Hitesh Patel,2 Neil O Carragher,2 Jennifer P Morton,1 William J Muller,3Thomas RJ Evans,1 Barry Gusterson,4 Owen J Sansom1 and Valerie G Brunton2,*

1 Beatson Institute for Cancer Research, Garscube Estate, Switchback Road, Glasgow G61 1BD, UK2 Edinburgh Cancer Research UK Centre, University of Edinburgh, Crewe Road South, Edinburgh EH4 2XR, UK3 Goodman Cancer Research Center, McGill University, Montreal, Canada H3A 1A34 Institute of Cancer Sciences, Glasgow University, Glasgow G12 8QQ, UK

*Correspondence to: Dr Valerie Brunton, Edinburgh Cancer Research UK Centre, University of Edinburgh, Crewe Road South, Edinburgh EH4 2XR,UK. e-mail: [email protected]

#These authors contributed equally to this study.

AbstractSrc family kinase activity is elevated in a number of human cancers including breast cancer. This increasedactivity has been associated with aggressive disease and poor prognosis. Src inhibitors are currently in clinicaldevelopment with a number of trials currently assessing their activity in breast cancer. However, the results todate have been disappointing and a further evaluation of the preclinical effects of Src inhibitors is required tohelp establish whether these agents will be useful in the treatment of breast cancer. In this study we investigatethe effects of dasatinib, which is a potent inhibitor of Src family kinases, on the initiation and development ofbreast cancer in a genetically engineered model of the disease. The mouse model utilized is driven by expressionof activated ErbB-2 under the transcriptional control of its endogenous promoter coupled with conditionalloss of Pten under the control of Cre recombinase expressed by the BLG promoter. We show that daily oraladministration of dasatinib delays tumour onset and increases overall survival but does not inhibit the proliferationof established tumours. The striking difference between the dasatinib-treated group of tumours and the vehiclecontrols was the prominent squamous metaplasia that was seen in six out of 11 dasatinib-treated tumours. Thiswas accompanied by a dramatic up-regulation of both E-cadherin and β-catenin and down-regulation of ErbB-2in the dasatinib-treated tumours. Dasatinib also inhibited both the migration and the invasion of tumour-derivedcell lines in vitro. Together these data support the argument that benefits of Src inhibitors may predominate inearly or even pre-invasive disease.Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Keywords: breast cancer; Src; mouse model; dasatinib

Received 1 October 2012; Revised 2 April 2013; Accepted 4 April 2013

No conflicts of interest were declared.

Introduction

The non-receptor tyrosine kinase Src plays a centralrole in a number of signalling pathways, integratingcues from both growth factor receptors and integrins,and therefore controls numerous cellular processesinvolved with the malignant phenotype, includingproliferation, migration, invasion, and angiogenesis[1,2]. In addition, it has been known for many yearsthat Src activity and expression are increased in anumber of tumour types [3], which has led to thedevelopment of small molecule Src kinase inhibitors[4], with numerous phase I/II clinical trials nowcompleted or underway [5–7].

In breast cancer, increased Src activity has beenlinked to reduced survival [8,9] and Src is known toplay a key role in many of the signalling pathways

that drive breast cancer development. For example,Src is activated following activation of the oestrogenreceptor (ER), while Src can in turn phosphorylate theER and regulate its activity [10]. Several preclinicalstudies have reported that combining Src inhibitorswith anti-hormonal therapies leads to more effectiveinhibition of breast cancer growth than seen with thesingle agents, and in addition can delay the onset ofresistance [11–15]. Src is also involved in signallingdownstream of the epidermal growth factor receptor(ErbB/HER) family of receptors [16]. ErbB-2 is over-expressed or amplified in around 20% of breast cancersand is associated with poor prognosis and reducedpatient survival [17]. Trastuzumab is a monoclonalantibody which targets HER2 and it improves survivalin both early and advanced breast cancer, althoughacquired resistance is emerging as a significant cause

Copyright 2013 Pathological Society of Great Britain and Ireland. J Pathol 2013; 230: 430–440Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk www.thejournalofpathology.com

Dasatinib inhibits mammary tumour formation 431

of treatment failure. Recently, Src has been identifiedas a common signalling axis in trastuzumab resistance,indicating that Src inhibitors may provide an alternativetreatment for trastuzumab-resistant tumours [18].

Dasatinib is a small molecule inhibitor of Src familykinases which inhibits a number of additional kinasesincluding BCR-ABL, c-Kit, and the platelet-derivedgrowth factor receptor (PDGFR) [19]. Dasatinibinhibits the growth of only a small proportion oftumour cell lines, including those from breast [20–23],while it can effectively inhibit tumour cell migrationand invasion in a number of different tumour types[22,24–27]. In addition, dasatinib can inhibit the devel-opment of metastases in mouse models [25,28,29].A number of phase II trials of dasatinib monotherapyin patients with advanced breast cancer have beencompleted or are underway and to date, there havebeen some reports of partial responses and diseasestabilization [30,31]. However, there is still debateabout how best to utilize dasatinib (and other Srcinhibitors) in the clinical setting as it is unclear whichpatients will benefit from treatment and at what stageof their disease. It is therefore important to assessthe effects of Src kinase inhibitors in more complexin vivo mouse models where we can follow thedevelopment of disease progression.

In this study, we have investigated the effects ofdasatinib on the initiation and development of breastcancer in a genetically engineered mouse model inwhich tumours are driven by activated ErbB-2 andloss of the tumour suppressor PTEN. Previously, it hasbeen shown that in this model, transgenic mice carryingan MMTV-Cre-inducible activated Erbb2 under thetranscriptional control of the endogenous Erbb2 pro-moter (Flneo NeuNT mice [32]; hereafter referred to asErbb2KI) coupled with conditional loss of Pten resultsdevelop mammary tumours [33]. Here we have usedmice where Cre recombinase is under the control of theβ-lactoglobulin enhancer (BLG-Cre), which we havepreviously shown to provide efficient Cre-mediatedrecombination in mammary epithelial cells [34]. Weshow that dasatinib significantly delays tumour onsetand increases overall survival in the resulting BLG-Cre,Erbb2KI/+, Ptenfl/+ mice. Dasatinib does not inhibitproliferation of the primary tumours but does inducedifferentiation of the tumours. This was accompa-nied by increased expression of β-catenin and E-cadherin and down-regulation of ErbB-2 expression inthe dasatinib-treated tumours.

Materials and methods

Animals and genotypingBLG-Cre [STOCK Tg(LGB-cre)74Acl/J, Jackson\hbox{Laboratory,} C57BL/6 J F3 segregating for theFVB and S129 genomes] [35], Erbb2KI (FVB/n) [32],and floxed Pten (strain C;129S4-Ptentm1Hwu/J, JacksonLaboratory, C57BL/6 J F3 segregating for the FVB andS129 genomes) mice were bred to generate BLG-Cre,

Erbb2KI/+, Ptenfl/+ mice on a mixed background.Mice were genotyped by polymerase chain reactionanalysis as described previously [32,36]. For dasatinibtreatment, female virgin mice were randomly placedinto two cohorts after genotyping and where possible,littermates of the same genotype were assigned to dif-ferent cohorts: dasatinib (n = 9) and vehicle (n = 11).Experiments were performed in compliance with UKHome Office guidelines. Mice were dosed daily byoral gavage with 10 mg/kg dasatinib (Bristol-MyersSquibb, Princeton, NJ, USA) in 80 mmol/l citratebuffer as vehicle from 8 weeks of age. Mice weremonitored weekly for tumour formation by palpation(tumour onset was defined as the presence of apalpable tumour). Animals were sacrificed once theirtumour burden had reached 1.7 cm (in any direction).Tumours and tissues were removed and fixed in10% buffered formalin at sacrifice. For histologicaldetection of lung metastases, non-serial sections fromthroughout the lung were analysed. No metastaticdeposits were seen in other tissues.

ImmunohistochemistryImmunohistochemistry of formalin-fixed, paraffin-embedded tissues was performed as describedpreviously [22]. The primary antibodies used werephospho-Src Y418 (Cell Signaling, Danvers, MA,USA; 1 : 200), Ki67 (Thermo Fisher ScientificWaltham, MA; 1 : 200), ErbB-2 (Invitrogen, Carlsbad,CA, USA; 1 : 100), cytokeratin 5 (CK5) (CambridgeBioscience, Cambridge, UK; 1 : 250), E-cadherin (BDBiosciences, San Jose, CA, USA; 1:300), and β-catenin(BD Biosciences; 1 : 1000). For quantification ofKi67-positive nuclei, three random high-power fieldsper tumour were analysed. Areas of keratinizationin the dasatinib-treated tumours were excluded fromthe analysis. Expression levels of phospho-Src Y418staining were scored based on staining intensity andarea of tumour cells using a weighted histoscore calcu-lated from the sum of (1 × % weak staining) + (2 × %moderate staining) + (3 × % strong staining).

Western blottingWestern blot analysis was performed as describedpreviously [22]. The primary antibodies used werephospho-Src Y418 (Cell Signaling, 1 : 1000), Src(Cancer Research UK, London, UK; 1 : 1000),phospho-AKT S473 (Cell Signaling; 1 : 1000), AKT(Cell Signaling; 1 : 1000), E-cadherin (BD Bio-sciences; 1 : 1000), PTEN (Cell Signaling; 1:1000),β-actin (Sigma, St Louis, MO, USA; 1 : 5000), andβ-tubulin (Sigma; 1 : 5000).

Cell culture and dasatinib treatmentMammary tumour cell lines (BLG-5957 and BLG-6222) were generated from primary mammary tumourstaken from BLG-Cre, Erbb2KI/+, Ptenfl/+ mice andthen passaged in growth media (Dulbecco’s modified


432 SA Karim et al

Eagle medium containing 10% fetal bovine serum and2 mmol/l L-glutamine). Dasatinib was prepared as a100 mmol/l stock in dimethyl sulphoxide and dilutedin growth media.

Migration assayConfluent monolayers of BLG5957 or BLG6222 cellswere scored with a fine pipette tip to produce a wound.Migration into the wound was monitored by time-lapsevideo microscopy over 18 h in the presence or absenceof dasatinib using 20× magnification using an auto-mated Olympus ScanR microscope. Four representativeareas were scored for each treatment and the cell veloc-ity was calculated from the distance advanced into thewound over time using Cell R software (Olympus,Hamburg, Germany).

Invasion assayCells were seeded on the bottom of Corning Transwellinserts (Thermo Scientific, Waltham, MA) containingMatrigel (Becton Dickinson, Franklin, NJ), which wereplaced into serum-free medium. Medium supplementedwith 10% fetal calf serum was placed on top of thematrix in the absence or presence of dasatinib. Fivedays after seeding, invading cells were stained withCalcein-AM (Invitrogen) and visualized by confocalmicroscopy (Olympus FV1000). Serial optical sectionswere captured at 10 µm intervals and quantified usingImageJ software using the area analysis module. Inva-sion was calculated as cells that had moved more than20 µm into the collagen.

Reverse-phase protein array (RPPA) analysisThe fourth mammary fat pad of 7-week-old CD-1 nudemice was inoculated with 1 × 106 BLG-6222 cells.Once tumours had reached a volume of 0.2 cm3, micewere randomized to receive either a single dose ofdasatinib, 10 mg/kg (n = 3), or vehicle (citrate buffer,pH 3.1, n = 3). Mice were sacrificed 2 h later. Follow-ing harvest, the tumours were washed with ice-coldred blood cell lysis solution (PBS 7.5 ml, ammoniumchloride 2.5 ml; Stem Cell Technologies, Grenoble,France), complete protease inhibitor cocktail tablet(Roche, Basel, Switzerland), and phosSTOP proteaseinhibitor tablet (Roche), and subsequently snap-frozenin liquid nitrogen. Protein extracts were preparedin CLB1 lysis buffer (Zeptosens-Bayer TechnologyServices, Leverkusen, Germany) and samples wereanalysed by Zeptosens RPPA as previously described[37]. Data are presented as relative fluorescence inten-sity (RFI) values for each sample replicate followingglobal normalization. Further details are provided inthe Supplementary methods.

Results

Dasatinib reduces mammary tumour developmentBLG-Cre, Erbb2KI/+, Ptenfl/+ mice were treated from 8weeks of age with dasatinib (n = 9) or vehicle (n = 11)

by daily oral gavage. There was a significant delay intumour onset in the dasatinib-treated cohort of animalscompared with the control animals. The mediantumour onset was 467 days in the vehicle-treated ani-mals compared with 585 days in the dasatinib-treatedanimals (Figure 1a, p = 0.046, Wilcoxon signed ranktest). There was also an increase in the survival of thedasatinib-treated animals compared with the controlanimals (Figure 1b, p = 0.06, Wilcoxon signed ranktest), which reflects the delay in tumour onset, as therewas no difference in the time from tumour onset to thetime at which the animals had to be sacrificed whentheir tumour burden reached the maximum value asallowed by the Home Office (Figure 1c, p = 0.284,Wilcoxon signed rank test; median time from palpabletumour to death for vehicle 42 days and for dasatinib-treated animals 39 days). The majority of the mice(85%) developed only one tumour, while the remain-ing mice developed a smaller second tumour in anadditional gland (for further details see SupplementaryTable 2). The incidence of lung metastases was unal-tered in the two cohorts of animals, with 18% (2/11)of vehicle-treated animals and 22% (2/9) of dasatinib-treated animals developing lung metastases. However,whereas in the control animals there were multiplelesions within the lungs (23 and 35), there were only2–4 metastatic lesions in the dasatinib-treated animalsthat developed lung metastases (Figure 1d).

Dasatinib alters tumour differentiationThe vector control animals had moderately to poorlydifferentiated adenocarcinomas with a variable growthpattern and little to no CK5 expression (Figures 2aand 2b). The striking difference between the dasatinib-treated group of tumours and the vehicle controls wasthe prominent squamous metaplasia that was seen insix out of nine dasatinib-treated tumours. In five of thetumours, the squamous elements occupied over 80%of the section and one tumour had both epidermal(keratin whorls and stratified squamous epithelium) andsebaceous elements (Figure 2b). The glandular elementin all of these tumours was moderately to poorlydifferentiated adenocarcinoma. The CK5 staining alsoclearly showed the squamous metaplasia arising fromthe basal layer of the glandular elements (Figure 2b).In all cases, the metastases reflected the growth patternand cytology of the primary tumour. For example,Supplementary Figure 1 shows a poorly differentiatedmetastatic lesion with the same nuclear and cytologicaldetails as seen in the primary tumour. This was alsotrue for the metastases that developed in the dasatinib-treated animals (Supplementary Figure 1). Mammary-specific activation of Wnt signalling, and in particularstabilization of β-catenin, results in the formation ofsquamous metaplasias [38,39]. Immunohistochemicalanalysis showed that vehicle-treated tumours had verylow levels of β-catenin, while increased staining bothin the nucleus and at cell–cell junctions was seen indasatinib-treated tumours, which was associated with



(a)

(c)

(d)

(b)

Figure 1. Dasatinib inhibits tumour formation in BLG-Cre, Erbb2KI/+, Ptenfl/+ mice. (a) Tumour onset (detection of measurable tumour bypalpation) in vehicle- (n = 11) and dasatinib-treated (n = 9) mice. p = 0.046, Wilcoxon signed rank test. (b) Survival curves (time of death)for vehicle- and dasatinib-treated mice. p = 0.06, Wilcoxon signed rank test. (c) Survival curves plotting time from tumour onset to deathfor vehicle- and dasatinib-treated mice. p = 0.284, Wilcoxon signed rank test. (d) Incidence of lung metastases expressed as number ofmice with metastases as a percentage of the total number of mice in each cohort and number of metastatic lesions in the two vehicle-(mouse IDs: 88138 and 93020) and two dasatinib-treated animals (mouse IDs: 95295 and 98852).

areas of squamous metaplasia (Figure 2c). Specifically,nuclear β-catenin was seen in regions surroundingthe areas of keratinization (Figure 2c, arrowheads),indicating that increased β-catenin activation in thedasatinib-treated tumours may drive the squamousdifferentiation.

Dasatinib does not inhibit mammary tumourproliferationAs there was no delay in tumour progression in thedasatinib-treated mice once the tumours had estab-lished (Figure 1c), we then looked at proliferationof the tumours in the two cohorts using Ki67 as amarker (Figure 3a). Quantification of the percentage ofKi67-positive cells in the vehicle- and dasatinib-treatedtumours showed that there was no significant differencein the proliferative capacity between the two cohorts(Figures 3a and 3b). In addition, we saw no evidenceof apoptosis, as measured by the presence of cleavedcaspase 3, in tumours from vehicle- or dasatinib-treatedanimals (results not shown). Inhibition of Src kinaseactivity was confirmed in the dasatinib-treated tumoursby assessing the phosphorylation of Src on tyrosine418 (pTyr418 Src), widely used as a biomarker of Srcactivity, by both immunohistochemistry and westernblotting (Figures 3c and 3d) [22,40].

Dasatinib inhibits migration and invasionWe then looked at whether dasatinib could inhibitmigration and invasion. To facilitate these studies,tumour-derived cell lines from BLG-Cre, Erbb2KI/+,Ptenfl/+ mice were established (BLG-5957 and BLG-6222). Using a wound-healing assay, we found thatdasatinib treatment inhibited the migration of bothcell lines, with a significant reduction in the speedof migration in the dasatinib-treated cells (Figure 4a).Dasatinib also inhibited invasion of the BLG-6222cells into Matrigel (the BLG-5957 cells did not invadein this assay) (Figure 4b). No effect on proliferationof the cells was seen at these concentrations of dasa-tinib (results not shown), although Src activity wasinhibited, as shown by a reduction in Src autophos-phorylation (Figure 4c). Interestingly, the reduction inautophosphorylated Src was seen despite an increasein total Src expression following dasatinib treatmentin both cell lines (Figure 4c). Previously, we haveshown that dasatinib can inhibit tumour cell invasionvia stabilization of E-cadherin-mediated cell–celljunctions [41,42]. E-cadherin was expressed in alltumours and there was evidence of both cytoplasmicand membrane-associated E-cadherin, although in thecontrol tumours, there was very little localized atthe cell membrane (Figure 4d). Dasatinib treatment


434 SA Karim et al

(a)

(b)

(c)

Figure 2. Dasatinib induces tumour differentiation. (a) Representative haematoxylin and eosin (H&E)-stained section of vehicle- anddasatinib-treated tumours (top panels). In the vehicle tumour, the asterisk shows poorly differentiated elements and the arrowheadglandular elements, while the metastatic lesion shows the same poorly differentiated elements. The dasatinib-treated tumour showsglandular elements (arrowheads), sebaceous elements (red asterisk), and keratin whorls (arrow). Scale bars = 100 µm. Lower panelsshow higher-magnification images. Scale bars = 50 µm. (b) Immunohistochemical analysis of CK5 in vehicle- (left-hand panel) anddasatinib-treated tumours (right-hand panel). (c) Immunohistochemical analysis of β-catenin in vehicle- and dasatinib-treated tumours.β-catenin was expressed at cell–cell junctions (arrows) and within the nucleus (arrowhead). Scale bars = 100 µm.

resulted in a dramatic increase in E-cadherin expres-sion at the cell membrane and within the cytoplasm,throughout the tumours (Figure 4d). This was con-firmed by western blot analysis of the tumours, whichshowed a significant increase in E-cadherin expressionin the dasatinib-treated tumours (Figure 4e).

Dasatinib inhibits downstream signallingWe next looked at the effect of dasatinib on ErbB-2 anddownstream signalling pathways. Using an antibody

that recognizes one of the main autophosphorylationsites in ErbB-2 (Tyr 1248), we were unable to detectphosphorylated ErbB-2 in the tumours by eitherwestern blot or immunohistochemistry. However,immunohistochemical analysis showed that ErbB-2expression was down-regulated in the dasatinib-treatedtumours (Figure 5a). In all tumours analysed, ErbB-2expression was almost undetectable and this wasirrespective of the tumour phenotype. Unfortunately,we were not able to detect ErbB-2 expression by



(a) (b)

(c)

(d)

Figure 3. Dasatinib does not inhibit tumour proliferation. (a) Immunohistochemical analysis of Ki67 in vehicle- and dasatinib-treatedtumours. (b) Box-whisker plot analysis of Ki67-positive nuclei in tumours from vehicle- (n = 4) and dasatinib-treated (n = 6) animals.p = 0.7832, Mann–Whitney U-test. (c) Immunohistochemical analysis of pTyr418 Src in vehicle- and dasatinib-treated tumours. Scalebars = 100 µm. Histoscore analysis is shown in the right-hand panel. p = 0.1099, Mann–Whitney U-test. (d) Western blot analysis ofpTyr418 Src and Src in vehicle- (n = 4) and dasatinib-treated (n = 4) tumours.

western blot in the tumour lysates. Looking down-stream of ErbB-2, we found that dasatinib treatmentresulted in reduced phosphorylation of AKT, withtotal levels of AKT being unaltered (Figure 5b).Interestingly, the reduction in AKT phosphorylation inthe dasatinib-treated cells was seen despite a reductionin PTEN expression in these tumours (Figure 5b). Aswe were limited by the amount of available materialfor analysis of signalling in the tumours from thegenetically engineered mice, tumours within the mam-mary fat pad were established following inoculationof the BLG-6222 cells. We then carried out RPPAanalysis on vehicle- and dasatinib-treated tumours,which allowed us to look at changes in a numberof signalling molecules including Src, ErbB-2, AKT,MAPK, Stat5, and mTOR (Supplementary Figure 2).As expected, treatment with dasatinib resulted in a sig-nificant reduction in Src phosphorylation (Figure 5c).Autophosphorylation of ErbB-2 on Tyr1248 was alsoreduced but this did not reach significance (p = 0.057).Although the antibody used also detects EGFRphosphorylated on Tyr1173, there was no decreasein signal in the dasatinib-treated tumours when anantibody specific to EGFR pTyr1173 was used. Thus,

short-term treatment (2 h) with dasatinib can inhibitboth Src and ErbB-2 phosphorylation. Further analysisof the vehicle- and dasatinib-treated tumours showedthat there was also a significant reduction in AKTphosphorylation on both Thr308 and Ser473, as wehad seen in the genetically engineered model. The onlyother statistically significant difference between thevehicle- and dasatinib-treated tumours was a reductionin phosphorylation of PLCγ on Tyr783 (Figure 5c).

Discussion

There has been much interest in the development of Srcinhibitors for the treatment of solid tumours, includingbreast cancer. Although a number of phase II trials ofdasatinib in breast cancer have now been completed,the results have been disappointing. There is a need tobetter understand the activity of Src inhibitors such asdasatinib and understand in which patients and at whatstage of disease these inhibitors should be evaluated.

We have used a mouse model in which activatedErbB-2 under the control of its endogenous promoter(Erbb2KI) [32], combined with heterozygous loss ofPten , drives mammary tumourigenesis by using a


436 SA Karim et al

(a)

(b)

(d)(e)

(c)

Figure 4. Dasatinib inhibits migration and invasion. (a) Migration of BLG-5957 and BLG-6222 cells into a wound was followed over18 h and the velocity calculated from the distance advanced into the wound over time. Values are mean ± SD of five experiments.Representative images of the wound closure in both cell lines from which the quantification was made are shown in the right-hand panels(b) Invasion of BLG-6222 cells into Matrigel in the presence or absence of dasatinib. Values are mean ± SD from triplicate wells. (c)Western blot analysis of pTyr418 Src and Src in cells treated with a range of dasatinib concentrations for 18 h. (d) Immunohistochemicalanalysis of E-cadherin in vehicle- and dasatinib-treated tumours (arrows indicate staining at cell–cell junctions).

mammary-specific BLG-driven Cre recombinase. Wehad previously shown that the BLG-Cre strain providedefficient Cre-mediated recombination in virgin mam-mary epithelial cells [34]; however, tumours developedin our model with a very long latency (median tumouronset was 467 days in the vehicle-treated tumours).Previous use of this model in which Cre recombi-nase was under transcriptional control of the MMTVpromoter enhancer had reported a tumour latency of6.5 months [33]. There are a number of possible rea-sons for these differences. Firstly, the BLG-Cre maynot be as efficient as the MMTV-Cre strain used; and

secondly, in our model we used mice on a mixed back-ground, which may not be as susceptible to mammarytumour formation. The longer latency may be in partattributed to the retention of PTEN within the tumoursthat develop in our model. Loss of heterozygosity ofPten was reported in 50% of tumours that developedin the MMTV-driven model [33]. However, we foundthat all tumours retained PTEN expression. In addition,we saw very low levels of ErbB-2 expression comparedwith those found in tumours from other ErbB-2-drivenmammary tumour models (our unpublished data).



(a)

(b)

(c)

Figure 5. Dasatinib induced alterations in signalling pathways. (a) Immunohistochemical analysis of ErbB-2 expression in vehicle- anddasatinib-treated tumours. Scale bars = 50 µm. (b) Western blot analysis of pSer473 AKT, AKT, and PTEN in tumours from four vehicle- andfour dasatinib-treated animals. (c) Reverse-phase protein array analysis of vehicle- (n = 3) and dasatinib-treated (n = 3) BLG-6222 tumours(2 h, 10 mg/kg). The graph represents globally normalized relative fluorescence intensity (RFI) values, where the ratio of phosphorylated tototal protein levels have been plotted. p values generated by ANOVA of each data set.

We found a significant delay in the onset of tumourformation, suggesting that Src plays an importantrole in the early stages of tumour development.Further support for this comes from studies inwhich mammary-specific expression of activated Srcleads to hyperplasia but rarely to the formation oftumours [43], while mammary-specific deletion ofSrc in the MMTV-polyomavirus middle T (PyMT)model of mammary tumourigenesis leads to reducedproliferation of hyperproliferative lesions and aninability to form tumours [44]. Treatment with theSrc inhibitor SKI-606 also reduced the growth ofhyperplastic lesions in the MMTV-PyMT modeland delayed the onset of tumour formation [45].

In addition, in a mouse skin carcinogenesis model,Src inhibition reduces the formation of papillomasbut not their conversion to squamous cell carcinomasand high Src activity is associated with human actinickeratoses, which are pre-malignant hyperproliferativeskin lesions [46]. The delayed tumour onset may bedue to the development of resistance to dasatinib. Weknow that this is not due to lack of target inhibitionas Src activity is still inhibited in the dasatinib-treatedtumours. However, there may be compensatory up-or down-regulation of other pathways that contributesto the delayed onset and indeed there was signifi-cant down-regulation of ErbB-2 in dasatinib-treatedtumours.


438 SA Karim et al

Many studies have shown that interference with Srcactivity inhibits the migration and invasion of epithelialtumor cells, including breast cancer-derived lines,which may contribute to the potential anti-metastaticactivity of inhibitors such as dasatinib [25,28,47–50].Although we did not see a reduction in the number ofmice with metastases following dasatinib treatment,we did find that the number of metastatic foci wasdramatically reduced in the dasatinib-treated animals.However, due to the low incidence of metastasis inthis model, it was not possible to carry out statisticalanalysis on these data. The findings are, though, in linewith previous reports where dasatinib treatment hasbeen shown to specifically reduce metastatic burdenin mouse models of pancreatic and prostate cancer[25,28]. Our in vitro data confirmed that dasatinibcould inhibit both the migration and the invasion ofthe BLG-Cre, Erbb2KI/+, Ptenfl/+-derived tumour cells.The marked increase in E-cadherin expression in thedasatinib-treated tumours supports a role for Src incontrolling E-cadherin endocytosis and protein stabilityin vivo which contributes to tumour spread [41,51].

Src activity is increased in ErbB-2-driven mousemammary tumours and this activation is linked tothe formation of a complex between the SH2 domainof Src and the activated ErbB-2 receptor [52–54].Furthermore, the formation of a Src–ErbB-2 complexis associated with changes in breast cancer cell growth,survival, and polarity [55–57]. However, dasatinibtreatment did not alter the proliferation of tumoursformed in the BLG-Cre, Erbb2KI/+, Ptenfl/+ mice,although Src activity was effectively inhibited. Theinability of dasatinib to inhibit the proliferation ofthe tumours in the BLG-Cre, Erbb2KI/+, Ptenfl/+ micemost likely reflects the fact that other genetic changeshave occurred in these tumours that are driving theirproliferation. Further support for this comes froma study in which mammary-specific expression ofthe negative regulator of Src, CSK, did not alter thegrowth of ErbB-2-dependent mammary tumours inmice, although Src kinase activity was drasticallyreduced [58]. This is also in line with studies in humanbreast cancer cell lines where there was no correlationwith ErbB-2 activation and sensitivity to dasatinib[20,21]. In addition, although some cases of diseasestabilization have been reported in the on-going clin-ical trials of dasatinib in patients with breast cancer,this has not been linked to patients whose tumoursare expressing ErbB-2 [7]. In other tumour types, ithas also been shown that dasatinib treatment does notinhibit tumour proliferation. For example, in a geneti-cally engineered model of pancreatic cancer, dasatinibinhibits the formation of metastasis but has no effecton the proliferation of the primary tumours [25]. Thishighlights the need to identify predictive biomarkerswith which to select patients who would potentiallybenefit from dasatinib treatment where Src activitycontributes to the proliferation of the tumour [59].

The most distinctive difference between the tumoursin the vehicle- and dasatinib-treated groups was the

presence of squamous elements in the latter. Squamousmetaplasia is associated with mammary-specific acti-vation of the Wnt pathway in transgenic mice [38],and we saw increases in β-catenin expression sur-rounding the areas of keratinization in the dasatinib-treated tumours indicative of increased activation ofsignalling via β-catenin driving the squamous differen-tiation in the dasatinib-treated tumours. Hebbard et alalso reported that the Src inhibitor SKI-606 inducedtumour differentiation in the MMTV-PyMT mousemodel, which had similarities to the differentiation pat-tern observed in some types of Wnt-induced mammarytumours [45]. Thus, the co-ordinated activities of Srcand Wnt can regulate mammary differentiation. Thishas been previously reported during osteoblast differen-tiation, where dasatinib induces activation of β-cateninand drives osteoblast differentiation [60,61].

Importantly, these data serve as further evidence thattransgenic models can recapitulate the pathophysiologyof human disease and may therefore play an importantrole in further therapeutic developments. In humans,elevated levels of activated Src are frequently seen inDCIS [62], where it is associated with factors indica-tive of high risk of recurrence, including HER2 posi-tivity [62]. Previous data on the relationship betweenactivated Src and tumour stage have been conflicting[63–65]. However, Ito et al reported that elevated acti-vated Src levels were more frequently associated withearly invasive disease than with advanced disease [64].Together with the recognized anti-migratory propertiesof Src inhibitors, this further fuels the argument that thebenefits of Src inhibitors may predominate in early oreven pre-invasive disease. Whilst DCIS is not associ-ated with mortality, it continues to result in significantmorbidity. Current treatment strategies include mastec-tomy or breast-conserving surgery (BCS) with/withoutradiotherapy. Recurrence rates following BCS can beas high as 40% in HER2-positive subgroups [66].Furthermore, at the time of recurrence, approximately50% of patients have developed invasive disease [67],prompting aggressive surgical management of initialhigh-risk DCIS. Clinically, Src inhibitors may prove auseful risk reduction strategy permitting BCS, despitean initial high-risk presentation.

Acknowledgments

This work was supported by a Cancer ResearchUK (CR-UK) programme grant (C157/A9148; MCF,VGB), CR-UK Beatson Institute core funding (SAK,JPM, TRJE, OJS), and a CR-UK Clinical Fellowship(C6088/A12063; HC).

Author contribution statement

SAK, HC, HP, JPM, and NOC carried out the exper-iments. TRJE, WJM, VGB, and OJS designed theexperiments. VGB, OJS, and BG analysed the data



and wrote the manuscript. All authors approved thesubmitted manuscript.

References1. Frame MC. Newest findings on the oldest oncogene; how activated

src does it. J Cell Sci 2004; 117: 989–998.2. Parsons SJ, Parsons JT. Src family kinases, key regulators of signal

transduction. Oncogene 2004; 23: 7906–7909.3. Summy JM, Gallick GE. Src family kinases in tumor progression

and metastasis. Cancer Metastasis Rev 2003; 22: 337–358.4. Brunton VG, Frame MC. Src and focal adhesion kinase as

therapeutic targets in cancer. Curr Opin Pharmacol 2008; 8:427–432.

5. Aleshin A, Finn RS. SRC: a century of science brought to the clinic.Neoplasia 2010; 12: 599–607.

6. Creedon H, Brunton VG. Src kinase inhibitors: promising cancertherapeutics? Crit Rev Oncog 2012; 17: 145–159.

7. Mayer EL, Krop IE. Advances in targeting SRC in the treatment ofbreast cancer and other solid malignancies. Clin Cancer Res 2010;16: 3526–3532.

8. Elsberger B, Tan BA, Mitchell TJ, et al . Is expression or activationof Src kinase associated with cancer-specific survival in ER-, PR-and HER2-negative breast cancer patients? Am J Pathol 2009; 175:1389–1397.

9. Kanomata N, Kurebayashi J, Kozuka Y, et al . Clinicopathologicalsignificance of Y416Src and Y527Src expression in breast cancer.J Clin Pathol 2011; 64: 578–586.

10. Shupnik MA. Crosstalk between steroid receptors and the c-Src-receptor tyrosine kinase pathways: implications for cell prolifera-tion. Oncogene 2004; 23: 7979–7989.

11. Chen Y, Alvarez EA, Azzam D, et al . Combined Src and ERblockade impairs human breast cancer proliferation in vitro andin vivo. Breast Cancer Res Treat 2011; 128: 69–78.

12. Chen Y, Guggisberg N, Jorda M, et al . Combined Src and aromataseinhibition impairs human breast cancer growth in vivo and bypasspathways are activated in AZD0530-resistant tumors. Clin Cancer

Res 2009; 15: 3396–3405.13. Herynk MH, Beyer AR, Cui Y, et al . Cooperative action of

tamoxifen and c-Src inhibition in preventing the growth of estrogenreceptor-positive human breast cancer cells. Mol Cancer Ther 2006;5: 3023–3031.

14. Hiscox S, Morgan L, Green T, et al . Src as a therapeutic targetin anti-hormone/anti-growth factor-resistant breast cancer. Endocr

Relat Cancer 2006; 13 (Suppl 1): S53-S59.15. Vallabhaneni S, Nair BC, Cortez V, et al . Significance of ER–Src

axis in hormonal therapy resistance. Breast Cancer Res Treat 2011;130: 377–385.

16. Ishizawar R, Parsons SJ. c-Src and cooperating partners in humancancer. Cancer Cell 2004; 6: 209–214.

17. Slamon DJ, Clark GM, Wong SG, et al . Human breast cancer:correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987; 235: 177–182.

18. Zhang S, Huang WC, Li P, et al . Combating trastuzumab resis-tance by targeting SRC, a common node downstream of multipleresistance pathways. Nature Med 2011; 17: 461–469.

19. Lombardo LJ, Lee FY, Chen P, et al . Discovery ofN -(2-chloro-6-methyl-phenyl)-2-(6-(4-(2-hydroxyethyl)-piperazin-1-yl)-2-methylpyrimidin-4-ylamino)thiazole-5-carboxamide(BMS-354825), a dual Src/Abl kinase inhibitor with potentantitumor activity in preclinical assays. J Med Chem 2004; 47:6658–6661.

20. Finn RS, Dering J, Ginther C, et al . Dasatinib, an orally active smallmolecule inhibitor of both the src and abl kinases, selectively inhi-

bits growth of basal-type/‘triple-negative’ breast cancer cell linesgrowing in vitro. Breast Cancer Res Treat 2007; 105: 319–326.

21. Huang F, Reeves K, Han X, et al . Identification of candidate molec-ular markers predicting sensitivity in solid tumors to dasatinib:rationale for patient selection. Cancer Res 2007; 67: 2226–2238.

22. Serrels A, Macpherson IR, Evans TR, et al . Identification ofpotential biomarkers for measuring inhibition of Src kinase activityin colon cancer cells following treatment with dasatinib. Mol Cancer

Ther 2006; 5: 3014–3022.23. Song L, Morris M, Bagui T, et al . Dasatinib (BMS-354825)

selectively induces apoptosis in lung cancer cells dependent onepidermal growth factor receptor signaling for survival. Cancer Res

2006; 66: 5542–5548.24. Buettner R, Mesa T, Vultur A, et al . Inhibition of Src family kinases

with dasatinib blocks migration and invasion of human melanomacells. Mol Cancer Res 2008; 6: 1766–1774.

25. Morton JP, Karim SA, Graham K, et al . Dasatinib inhibits thedevelopment of metastases in a mouse model of pancreatic ductaladenocarcinoma. Gastroenterology 2010; 139: 292–303.

26. Nam S, Kim D, Cheng JQ, et al . Action of the Src family kinaseinhibitor, dasatinib (BMS-354825), on human prostate cancer cells.Cancer Res 2005; 65: 9185–9189.

27. Tatarov O, Mitchell TJ, Seywright M, et al . SRC family kinaseactivity is up-regulated in hormone-refractory prostate cancer. Clin

Cancer Res 2009; 15: 3540–3549.28. Park SI, Zhang J, Phillips KA, et al . Targeting SRC family kinases

inhibits growth and lymph node metastases of prostate cancer in anorthotopic nude mouse model. Cancer Res 2008; 68: 3323–3333.

29. Trevino JG, Summy JM, Lesslie DP, et al . Inhibition of SRCexpression and activity inhibits tumor progression and metastasisof human pancreatic adenocarcinoma cells in an orthotopic nudemouse model. Am J Pathol 2006; 168: 962–972.

30. Finn RS, Bengala C, Ibrahim N, et al . Dasatinib as a single agentin triple-negative breast cancer: results of an open-label phase 2study. Clin Cancer Res 2011; 17: 6905–6913.

31. Mayer EL, Baurain JF, Sparano J, et al . A phase 2 trial of dasatinibin patients with advanced HER2-positive and/or hormone receptor-positive breast cancer. Clin Cancer Res 2011; 17: 6897–6904.

32. Andrechek ER, Hardy WR, Siegel PM, et al . Amplificationof the neu/erbB-2 oncogene in a mouse model of mammarytumorigenesis. Proc Natl Acad Sci U S A 2000; 97: 3444–3449.

33. Dourdin N, Schade B, Lesurf R, et al . Phosphatase and tensinhomologue deleted on chromosome 10 deficiency accelerates tumorinduction in a mouse model of ErbB-2 mammary tumorigenesis.Cancer Res 2008; 68: 2122–2131.

34. Gallagher RC, Hay T, Meniel V, et al . Inactivation of Apc perturbsmammary development, but only directly results in acanthoma inthe context of Tcf-1 deficiency. Oncogene 2002; 21: 6446–6457.

35. Selbert S, Bentley DJ, Melton DW, et al . Efficient BLG-Cremediated gene deletion in the mammary gland. Transgenic Res

1998; 7: 387–396.36. Li G, Robinson GW, Lesche R, et al . Conditional loss of PTEN

leads to precocious development and neoplasia in the mammarygland. Development 2002; 129: 4159–4170.

37. van Oostrum J, Calonder C, Rechsteiner D, et al . Tracing pathwayactivities with kinase inhibitors and reverse phase protein arrays.Proteomics Clin Appl 2009; 3: 412–422.

38. Miyoshi K, Rosner A, Nozawa M, et al . Activation of differentWnt/beta-catenin signaling components in mammary epitheliuminduces transdifferentiation and the formation of pilar tumors.Oncogene 2002; 21: 5548–5556.

39. Miyoshi K, Shillingford JM, Le Provost F, et al . Activation of beta-catenin signaling in differentiated mammary secretory cells inducestransdifferentiation into epidermis and squamous metaplasias. Proc

Natl Acad Sci U S A 2002; 99: 219–224.


440 SA Karim et al

40. Luo FR, Barrett YC, Yang Z, et al . Identification and validation ofphospho-SRC, a novel and potential pharmacodynamic biomarkerfor dasatinib (SPRYCEL), a multi-targeted kinase inhibitor. Cancer

Chemother Pharmacol 2008; 62: 1065–1074.41. Canel M, Serrels A, Miller D, et al . Quantitative in vivo imaging

of the effects of inhibiting integrin signaling via Src and FAK oncancer cell movement: effects on E-cadherin dynamics. Cancer Res

2010; 70: 9413–9422.42. Serrels A, Timpson P, Canel M, et al . Real-time study of E-cadherin

and membrane dynamics in living animals: implications for diseasemodeling and drug development. Cancer Res 2009; 69: 2714–2719.

43. Webster MA, Cardiff RD, Muller WJ. Induction of mammaryepithelial hyperplasias and mammary tumors in transgenic miceexpressing a murine mammary tumor virus/activated c-src fusiongene. Proc Natl Acad Sci U S A 1995; 92: 7849–7853.

44. Marcotte R, Smith HW, Sanguin-Gendreau V, et al . Breast CancerSpecial Feature: Mammary epithelial-specific disruption of c-Srcimpairs cell cycle progression and tumorigenesis. Proc Natl Acad

Sci U S A 2012; 109: 2808–2813.45. Hebbard L, Cecena G, Golas J, et al . Control of mammary tumor di-

fferentiation by SKI-606 (bosutinib). Oncogene 2011; 30: 301–312.46. Serrels B, Serrels A, Mason SM, et al . A novel Src kinase

inhibitor reduces tumour formation in a skin carcinogenesis model.Carcinogenesis 2009; 30: 249–257.

47. Jallal H, Valentino ML, Chen G, et al . A Src/Abl kinase inhibitor,SKI-606, blocks breast cancer invasion, growth, and metastasis in

vitro and in vivo. Cancer Res 2007; 67: 1580–1588.48. Johnson FM, Saigal B, Talpaz M, et al . Dasatinib (BMS-354825)

tyrosine kinase inhibitor suppresses invasion and induces cell cyclearrest and apoptosis of head and neck squamous cell carcinoma andnon-small cell lung cancer cells. Clin Cancer Res 2005; 11:6924–6932.

49. Pichot CS, Hartig SM, Xia L, et al . Dasatinib synergizes withdoxorubicin to block growth, migration, and invasion of breastcancer cells. Br J Cancer 2009; 101: 38–47.

50. Shor AC, Keschman EA, Lee FY, et al . Dasatinib inhibits migrationand invasion in diverse human sarcoma cell lines and inducesapoptosis in bone sarcoma cells dependent on SRC kinase forsurvival. Cancer Res 2007; 67: 2800–2808.

51. Coluccia AM, Benati D, Dekhil H, et al . SKI-606 decreases growthand motility of colorectal cancer cells by preventing pp60(c-Src)-dependent tyrosine phosphorylation of beta-catenin and its nuclearsignaling. Cancer Res 2006; 66: 2279–2286.

52. Luttrell DK, Lee A, Lansing TJ, et al . Involvement of pp60c-srcwith two major signaling pathways in human breast cancer. Proc

Natl Acad Sci U S A 1994; 91: 83–87.53. Muthuswamy SK, Muller WJ. Direct and specific interaction of

c-Src with Neu is involved in signaling by the epidermal growthfactor receptor. Oncogene 1995; 11: 271–279.

54. Muthuswamy SK, Siegel PM, Dankort DL, et al . Mammary tumorsexpressing the neu proto-oncogene possess elevated c-Src tyrosinekinase activity. Mol Cell Biol 1994; 14: 735–743.

55. Belsches-Jablonski AP, Biscardi JS, Peavy DR, et al . Src familykinases and HER2 interactions in human breast cancer cell growthand survival. Oncogene 2001; 20: 1465–1475.

56. Kim H, Chan R, Dankort DL, et al . The c-Src tyrosine kinaseassociates with the catalytic domain of ErbB-2: implications forErbB-2 mediated signaling and transformation. Oncogene 2005; 24:7599–7607.

57. Marcotte R, Zhou L, Kim H, et al . c-Src associates withErbB2 through an interaction between catalytic domains and con-fers enhanced transforming potential. Mol Cell Biol 2009; 29:5858–5871.

58. Kaminski R, Zagozdzon R, Fu Y, et al . Role of SRC kinasesin Neu-induced tumorigenesis: challenging the paradigm usingCsk homologous kinase transgenic mice. Cancer Res 2006; 66:5757–5762.

59. Moulder S, Yan K, Huang F, et al . Development of candidategenomic markers to select breast cancer patients for dasatinibtherapy. Mol Cancer Ther 2010; 9: 1120–1127.

60. Garcia-Gomez A, Ocio EM, Crusoe E, et al . Dasatinib as a bone-modifying agent: anabolic and anti-resorptive effects. PLoS One

2012; 7: e34914.61. Lee YC, Huang CF, Murshed M, et al . Src family kinase/abl

inhibitor dasatinib suppresses proliferation and enhances differenti-ation of osteoblasts. Oncogene 2010; 29: 3196–3207.

62. Wilson GR, Cramer A, Welman A, et al . Activated c-SRCin ductal carcinoma in situ correlates with high tumour grade,high proliferation and HER2 positivity. Br J Cancer 2006; 95:1410–1414.

63. Diaz N, Minton S, Cox C, et al . Activation of stat3 in primarytumors from high-risk breast cancer patients is associated withelevated levels of activated SRC and survivin expression. ClinCancer Res 2006; 12: 20–28.

64. Ito Y, Kawakatsu H, Takeda T, et al . Activation of c-Src is inverselycorrelated with biological aggressiveness of breast carcinoma.Breast Cancer Res Treat 2002; 76: 261–267.

65. Madan R, Smolkin MB, Cocker R, et al . Focal adhesion proteinsas markers of malignant transformation and prognostic indicatorsin breast carcinoma. Hum Pathol 2006; 37: 9–15.

66. Han K, Nofech-Mozes S, Narod S, et al . Expression of HER2neuin ductal carcinoma in situ is associated with local recurrence. ClinOncol (R Coll Radiol) 2012; 24: 183–189.

67. Fisher B, Dignam J, Wolmark N, et al . Tamoxifen in treatment ofintraductal breast cancer: National Surgical Adjuvant Breast andBowel Project B-24 randomised controlled trial. Lancet 1999; 353:1993–2000.

SUPPORTING INFORMATION ON THE INTERNETThe following supporting information may be found in the online version of this article.

Supplementary methods.

Figure S1. Representative haematoxylin and eosin (H&E)-stained section of vehicle- and dasatinib-treated tumours (top panels) and thecorresponding pulmonary metastases (bottom panels).

Figure S2. Heat map displaying the relative abundance of protein levels and post-translational modifications across dasatinib- and vehicle-treatedin vivo tumour samples.

Table S1. Antibodies for reverse-phase protein array analysis.

Table S2. Location and phenotype of tumours in BLG-Cre, Erbb2KI/+, Ptenfl/+ mice.


dasatinib inhibits mammary tumour development in a genetically engineered mouse model

Documents