Endocrine-Related Cancer (2008) 15 439–449

Patterns and changes in expression following neo-adjuvant anti-estrogen treatment in estrogen receptor-positive breast cancer

Vera Cappelletti1*, Manuela Gariboldi1,2*, Loris De Cecco1,2, Sara Toffanin1, James F Reid1,2, Lara Lusa1,2, Emilio Bajetta3, Luigi Celio3, Marco Greco4, Alessandra Fabbri5, Marco A Pierotti 2,6 and Maria Grazia Daidone1

1Department of Experimental Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori, Via Venezian 1, 20133 Milano, Italy 2Molecular Cancer Genetics, Fondazione Istituto FIRC di Oncologia Molecolare (IFOM), Milan, Italy 3Departments of Medical Oncology 4Surgery, 5Pathology and 6Scientific Direction, Fondazione IRCCS Istituto Nazionale dei Tumori, Via Venezian 1, 20133 Milano, Italy (Correspondence should be addressed to M G Daidone, Department of Experimental Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori, Via Venezian 1, 20133 Milano, Italy; Email: [email protected]) *V Cappelletti and M Gariboldi contributed equally to this work

Abstract This study aimed to define a gene expression profile associated with response to anti-estrogen treatment in estrogen receptor a (ERa)-positive breast cancer from elderly patients and to identify possible candidate associated with resistance by detecting those modulated by treatment. Using cDNA microarrays containing 16 702 unique clones, 21 pre-treatment and 11 paired post- treatment samples collected in a neo-adjuvant toremifene trial on elderly patients with operable and locally advanced ERa-positive breast cancer were profiled. Gene expression profiles generated from pre-treatment samples were correlated with treatment-induced tumor shrinkage and compared with those obtained from post-treatment paired samples to define genes differentially modulated following anti-estrogen treatment. Correlation analysis on 21 pre-treatment samples highlighted 53 genes significantly related to treatment response (P!0.001). Genes involved in cell cycle and proliferation were more frequently upregulated in responders compared with non- responders. Class comparison analysis identified 101 genes significantly modulated independently of treatment response; 82 genes were modulated in non-responders, whereas only 8 genes were differently expressed after treatment in responders. Gene expression profiles appear to be more frequently modulated by anti-estrogen treatment in non-responding patients and may harbor interesting genes possibly involved in anti-estrogen resistance, including clusterin, MAPK6, and MMP2. This concept was corroborated by in vitro studies showing that silencing of CLU restored toremifene sensitivity in the ER anti-estrogen-resistant breast cancer cell line T47D. Integration between neo-adjuvant therapy and transcriptional profiling has therefore the potential to identify therapeutic targets to be challenged for overcoming treatment resistance. Endocrine-Related Cancer (2008) 15 439–449

Introduction therapy in this tumor type. However, ER transcrip- The use of hormonal therapy has significantly contri- tional effects are not simply determined by the ligand, buted to the general reduction in breast cancer and ER activity is regulated through complex mortality achieved in the last decade (Peto et al. interactions with multiple cell signaling pathways 2000). Classical anti-estrogen therapies act by bloc- (Razandi et al. 2003, Stoica et al. 2003). king or downregulating the estrogen receptor (ER) and Neo-adjuvant trials represent a tool for investigating represent the first example of biologically targeted biological markers associated with treatment response

Endocrine-Related Cancer (2008) 15 439–449 Downloaded from Bioscientifica.comDOI: 10.1677/ERC-07-0274 at 09/30/2021 07:03:58PM 1351–0088/08/015–439 q 2008 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.orgvia free access V Cappelletti, M Gariboldi et al.: Gene expression and anti-estrogen treatment as well as those modulated by the treatment, thus needle. Post-treatment tissue was sampled by the giving clues for understanding mechanisms involved in pathologist within the residual tumor at surgery. clinical resistance. The combination of this approach For molecular analyses, core biopsies and surgical with genomic expression profiling could provide an specimens were snap-frozen in liquid nitrogen within opportunity to identify molecular signatures associated 30 min of removal, and stored at K80 8C; gene profile and/or predictive of treatment response. was carried out on samples containing more than 70% Preliminary data eliciting molecular signatures neoplastic cells, as assessed by hematoxylin and eosin- associated with clinical outcome following endocrine stained sections from frozen tissue specimens. treatment are already available in the metastatic and Clinical response was evaluated as percentage of adjuvant settings (Ma et al. 2004, Paik et al. 2004, residual tumor size from baseline after 3 months of Jansen et al. 2005). treatment (Cappelletti et al. 2004). Patients were We exploited a neo-adjuvant trial recently con- assigned to responder or non-responder groups cluded in our institution, originally planned to according to WHO criteria except that patients with evaluate the predictive role of ER-b on response residual tumor !75 and R50% were classified as to the anti-estrogen toremifene (Cappelletti et al. achieving minor changes and considered as respon- 2004), to investigate the molecular signatures ders (Geisler et al. 2001), according to previously associated with tumor shrinkage on pre-treatment reported criteria to evaluate clinical response to neo- samples and to identify changes in gene expression adjuvant endocrine treatment. We also investigated pattern in post-treatment samples compared with pre- changes of gene expression pattern in the absence of treatment one. any intercurrent systemic treatment on a series of We report here a gene signature associated with seven paired bioptic and surgical specimens from response to treatment and a distinct set of genes patients subjected to diagnostic biopsy followed modulated by treatment, which might be involved in within 3–4 weeks by radical surgery without anti-estrogen resistance mechanisms. The latter intervening therapies. include genes already known to be involved in The study has been approved by the IRB and Ethics mechanisms of resistance to anti-estrogens, such as Committee, and patients gave written informed MPK6 and MMP2, and a new candidate CLU consent to donate the leftover tissue after diagnosis (clusterin). A validation in an experimental model is to the Istituto Nazionale Tumori of Milan for the reported for one of the candidate genes, CLU. We present and future research. demonstrated that toremifene treatment upregulated CLU expression and that silencing of CLU restored RNA isolation and expression profiling sensitivity to the anti-estrogen in an ERCanti- Total RNA extraction from core biopsies and surgical estrogen-resistant breast cancer cell line. samples, probe labeling and hybridization were This suggests that the integration between neo- performed as described previously (De Cecco et al. adjuvant therapy and transcriptional profiling has the 2004). Two types of slides were used: type 7 star slides potential to identify therapeutic targets to be chal- (Amersham Biosciences) for the toremifene-treated lenged to overcome treatment resistance. cases and UltraGAPS (Corning, Lowell, MA, USA) for the control dataset. The samples and a reference RNA (Universal Human Reference RNA, Stratagene, La Patients and methods Jolla, CA, USA) were labeled directly with Cy3-dCTP (reference RNA) or Cy5-dCTP (sample RNA; Patients, tissue samples, and study design Amersham Biosciences) and indirectly with 3DNA This study employed leftover samples obtained from a Submicro Expression Array Detection kit (Genisphere, prospective study (Cappelletti et al. 2004) in which Montvala, NJ, USA). Hybridizations were carried out women aged 65 years or older with operable or locally in a hybridization station (Genomic Solutions, Ann advanced ER-positive (ERC) breast cancer but Arbor, MI, USA), slides were scanned using the previously untreated were subjected at Istituto Nazio- GenePix 4000B microarray scanner and quantified nale Tumori of Milan to pre-operative toremifene using GenePix Pro 5.0.1.24 (Axon Instruments, (Fareston, Shering Plough, Bulkham, Australia) at a Molecular Devices, Sunnyvale, CA, USA). The dose of 60 mg once daily for 3 months and then RNAs were hybridized on two different cDNA underwent surgery. Pre-treatment tumor material was microarrays containing a total of 16 702 unique clones obtained by three to four core biopsies with a 14 gauge selected from the Human sequence verified IMAGE

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access 440 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 439–449 clone collection (Research Genetics/Invitrogen) and buffered saline (DPBS) 1!, 0.1% Tween-20 with 5% spotted in triplicate. non-fat dry milk and incubated with primary against CLU (1:500, sc-6419; Santa Cruz Biotechnology, Quantitative real-time PCR Santa Cruz, CA, USA) and against b-actin (1:1000, 20– 33; Sigma–Aldrich). The membranes were washed and Total RNA isolated for the microarray analysis was incubated for 1 h at room temperature with peroxidase- used to verify the quantity of specific messengers by conjugated anti-goat (Santa Cruz Biotechnology) and real-time PCR for 11 biologically or clinically relevant anti-rabbit (Amersham Biosciences Europe) secondary differentially expressed genes. RNA was reverse (1:2000). transcribed using the High-Capacity cDNA Archive Signals were detected by chemiluminescence (ECL, Kit (Applied Biosystems, Foster City, CA, USA). The Amersham Biosciences Europe) according to the samples were amplified in multiplex PCRs using one manufacturer’s instructions. The relative amounts of of the assays of interest labeled with FAM and the CLU were quantified by densitometric analysis housekeeping gene GADPH labeled with VIC. Data and normalized with respect to b-actin signal. analysis was done using the Sequence Detector version 1.9 software (Applied Biosystems, Foster City, CA, USA) and statistical analyses were performed using Data analysis and statistics the R-statistical computing programing language Raw microarray images and quantifications were (R development Core Team 2006). The same statistical stored and processed in BioArray Software Environ- tests employed in the gene expression experiment were ment (BASE, Lund, Sweden; Saal et al. 2002). Poor used to confirm the results with quantitative real-time signal quality of background-corrected Cy3 and Cy5 K PCR. Relative log expression of the genes ( DCt) was intensities were flagged and a lowess normalization obtained subtracting the number of cycle threshold (Yang et al. 2002) applied to each slide. Replicated observed for the GADPH gene from that observed for spots were averaged and their log (base2) expression the gene of interest. ratios (tumor/reference) were downloaded from BASE and imported into BRB-ArrayTools version 3.2.2 siRNA transfection (Bethesda, MD, USA) for further analyses (Simon T47D cells were pre-incubated with Lipofectamine et al.2007). All IMAGE clone annotations were 2000 (Invitrogen-Life Technologies Inc.) in serum-free updated with the latest release of NCBI Unigene Opti-MEM for 20 min before adding siRNA oligo- (build no. 194) using SOURCE (Diehn et al. 2003). nucleotides (25 nM), diluted in the same medium Inthepresentstudy, we analyzed two different datasets. (3 mg/ml lipofectamine final concentration). The first one contained 32 hybridizations and was used siRNAs, CLU-V-siRNA (selective for the cytoplas- to evaluate: 1) gene expression patterns of pre-treatment mic form of CLU), 50-AUG AUG AAG ACU CUG CUG core biopsies correlated with tumor shrinkage (as a C-30 and 30-UAC UAC UUC UGA GAC GAC G-50; categorical dichotomous or a continuous variable) after control siRNA (scrambled of BIRC6 gene), 50-GCA toremifene treatment (21 pre-treatment samples) and 2) 0 0 gene expression changes in residual tumor specimens to GUA CAU GGU AUG AUU AdTdT-3 ,and3-dTd- TCGU CAU GUA CCA UAC UAA U-50 were custom- identify molecular patterns after toremifene exposure made (Dharmacon, Research Inc., Lafayette, CO, USA). (11 paired pre- and post-treatment samples). The second After 24 h, medium containing the RNA duplex and one containing 14 hybridizations, from 7 paired lipofectamine was replaced with Dulbecco’s modified biopsy–surgical specimens, was used to evaluate gene Eagle’s medium/Ham’s F12 (DMEM/F12) supple- expression changes in residual tumor specimens without mented with 5% stripped FBS and cells were treated any intercurrent local or systemic treatment. with 10K7 M 4-OH–TOR. Starting from both chips containing 8303 and 8399 The cell number was determined after 3 days by clones spotted in triplicate, low-quality spots (filtering direct counting of viable cells in a Burker chamber. criteria are described in the web Supplementary Experiments were repeated thrice. Material, which can be viewed online at http://erc. endocrinology-journals.org/supplemental/) from both datasets were flagged (set to missing value) and Western blotting subsequently filtered to contain no more than 20% Cytosolic (30 mg) were separated by SDS- missing values for each clone, which reduced the PAGE and transferred onto nitrocellulose membranes first dataset to 12 784 clones and the second one to that were blocked for 1 h in Dulbecco’s phosphate 4205 clones.

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access www.endocrinology-journals.org 441 V Cappelletti, M Gariboldi et al.: Gene expression and anti-estrogen treatment

Genes whose expression was significantly correlated Results with treatment response where identified by computing Pre-treatment gene profiles a statistical significance level for each gene by testing the hypothesis that Spearman’s correlation between gene Individual clinicopathological and biological tumor expression and tumor shrinkage was zero. Class features of the 21 patients included in this study are comparisons were performed using random-variance reported in Table 1. Tumors were clinically classified t-tests (Wright & Simon 2003;pairedwhenapplicable). as T2 (90%), node-negative (62%), and PgR-positive In all analyses, genes were considered statistically (71%), and the clinical response (i.e., tumor shrinkage significant if their P value was !0.001. The proba- R25%) was achieved in 67% of patients. Patients were bility of finding a given number of genes below this comparable for clinicopathological and biological threshold by chance was tested by repeated analysis on findings to the original subset of 38 toremifene-treated permutated class labels (Simon et al. 2003). patients entered a prospective study adequately Prior to the functional annotation of the lists using statistically sized to evaluate the predictive role of SOURCE (Diehn et al. 2003), DAVID (Dennis et al. ER-b expression (Cappelletti et al. 2004) whose 2003), and data from PubMed, clones with no gene leftover biological samples were used for the present symbols assigned to them by means of Unigene were study. Initial class comparison results of clinical mapped to the UCSC using BLAT response considered as a binary variable (i.e., (Kent 2002, Karolchik et al. 2003). responders versus non-responders) identified only 15 Hierarchical clustering of samples and genes was genes at the 0.001 level with a high probability of done using the 1-Pearson correlation as distance and finding such a number by chance (PZ0.153). There- average linkage method. after, when clinical response was considered as a All protocols, expression data, sample description, continuous variable defined as percent tumor shrinkage the list of annotated genes, and the microarray data in a (Table 1), 53 genes (Fig. 1), either positively (32) or format that conforms to the Minimum Information negatively (21) correlated with tumor reduction About a Microarray Gene Experiment (MIAME) (P!0.001), were identified (PZ0.0024 of finding 53 guidelines of the Microarray Gene Expression Data genes at 0.001 level by chance). Functional analysis Society (MGED) are available at http://pierotti.group. (excluding unknown reporters) showed that they were ifom-ieo-campus.it/suppl/br_tor.html. Microarray data mainly involved in cell cycle and proliferation have been deposited to the EBI ArrayExpress. (19.4%), signal transduction (16.7%), nucleic acid

Table 1 Clinicopathological and biological tumor characteristics

Initial (pre-treatment) Clinical axillary % Tumor Post-treatment Patient ID tumor size (cm) positive nodes PgR variation specimen availability

09 3!3NoCK100 No 10 4!4 Yes CK75 Yes 13 2.5!2.5 Yes CK100 Yes 18 6!5 Yes CK80 No 20 2.5!2NoKK25 No 21 3!3NoCK30 No 22 4.5!4 Yes CK32 Yes 24 3.5!3.5 No CK60 No 25 3.5!2.5 No CK74 No 26 3!3 Yes KK17 No 29 5!3.5 Yes KK14 Yes 30 3.5!4 Yes KK14 Yes 32 3!3NoCK17 Yes 33 4!4NoC 0 Yes 34 3!2NoCC50 Yes 37 2.5!2 Yes CK55 Yes 38 2.5!2.5 No CK64 No 45 4!3.5 No CK25 No 51 3!2.5 No K 0 Yes 57 2.1!2.1 No KK49 Yes 79 2.5!2.5 No CK52 No

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access 442 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 439–449

Genes involved in cell proliferation were more frequently upregulated in responding patients than in resistant ones (26.3% vs 11.8%) and included MOBK1B, DYRK2, RUNX3, BMI1, and GSPT1. So were also genes involved in membrane solute transport (including members of the solute carrier family as SLC11A2, SLCO3A1, SLC27A1, and ABCA5, belonging to ATP-binding cassette transporter family). Conversely, genes involved in nucleic acid processing, including zinc finger proteins (LOC90379, ZNF558, and PIAIS1) and helicases (SBNO1), were more frequently upregulated in non-responders.

Treatment-modulated gene profiles Before comparing the gene expression pattern in matched pre- and post-treatment specimens, we investigated expression profile modulation in serial paired biopsy and surgical specimens in the absence of intercurrent treatments. Unsupervised hierarchical clustering demonstrated that five out of the seven pairs had closely related transcriptional profiles. The class comparison based on paired t-test identified, at the 0.001 significance level, only two differentially expressed genes: an expressed sequence tag and TFDP1.

Unsupervised analysis of paired pre- and post-treatment specimens For 11 patients (five responders and six non-res- ponders), paired samples before and after toremifene treatment were available. Unsupervised hierarchical cluster analysis using all available genes showed that pre- and post-treatment specimens clustered together in Figure 1 Heatmap of genes correlated with response to toremifene (percentage of tumor reduction). For each gene, four out of five responding patients and in only two out expression levels above the mean are shown in red and those of the six matched specimens from non-responding below the mean are shown in blue. The color scale ranges from patients (Fig. 2). K1 (darkest blue) to C1 (darkest red). IMAGE clone identifiers and gene symbols are reported in the rows whereas the samples and their percentage of tumor reduction are reported in Supervised analysis of paired pre- and the columns. The gene expression matrix was mean centered post-treatment specimens and the gene expression profiles were clustered using the average linkage hierarchical clustering method and 1-Pearson Class comparison results considering all paired pre- and correlation as the distance method. All samples (14 responders post-treatment samples identified 101 differentially C 7 non-responders) were ordered according to the tumor expressed genes (P!0.001) following toremifene shrinkage. *Annotated using BLAT. treatment (PZ0.009 of getting at least 101 genes significant by chance), with a negligible overlap with processing/transcription (16.7%), and membrane the 53 genes associated with tumor shrinkage. Only solute transport (13.9%). We also identified some four genes were common to the two lists: MNT, SYAP1, genes involved in protein transport/metabolism and FLJ90757 and an expressed sequence tag, which (11.1%), immune response (5.5%), and cell adhesion/ were all downregulated by the treatment and negatively cytoskeleton (2.8%), and 17 genes (32.1%) had associated with tumor reduction. unknown function. In Table 2, the list of known Since hierarchical clustering result suggested that genes associated with response is reported as sub- gene expression before and after treatment seemed dividing genes according to biological function. more similar in responders than in non-responders,

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access www.endocrinology-journals.org 443 V Cappelletti, M Gariboldi et al.: Gene expression and anti-estrogen treatment

Table 2 Genes associated with treatment response

Biological function Gene symbol P value

Genes positively associated with response Adhesion/cytoskeleton C1QTNF7 0.0006799 Protein transport/metabolism VPS39 0.0003019 UBA52 0.0004614 Signal transduction FRS2 0.0002194 CLEC2D 0.0002404 INPP4B 0.0008168 Nucleic acid processing and transcription ZNF558 0.0008585 PIAS1 0.0008585 SBNO1 0.0008853 LOC90379 0.0009408 Cell cycle/proliferation MNT 0.0001893 MAPK8 0.0007346 Immunology BPHL 1.29!10K5 Membrane solute transport SYAP1 0.0004961 Other CLPTM1 0.0003625 ADCK4 0.000575 TCTE3 0.0007346 Genes negatively associated with response Protein transport/metabolism GALNT13 0.0009408 EXOC1 0.0007878 Signal transduction PIP5K1B 0.0004431 CAB39 0.0005004 GRBP 0.0008775 Nucleic acid processing and transcription SNAPC3 0.0003496 INTS12 0.0009488 K Cell cycle/proliferation MOBK1B !1!10 7 BMI1 0.0002498 GSPT1 0.0005487 DYRK2 0.0007878 RUNX3 0.0001817 Steroid metabolism OSBPL8 9.5!10K6 Immunology COL4A3BP !1!10K7 Membrane solute transport SLCO3A1 0.0005407 ABCA5 0.0005487 SLC27A1 0.0005842 SLC11A2 0.0007395 Other C1QC 0.0003251 we speculated that many of the differentially expressed which were absent among the 53 genes associated with genes after treatment were in part due to the hetero- response, were also represented (8.3%). geneity observed in the latter subset. As expected, the Among genes modulated in non-responding patients, class comparison of treatment-induced expression cell adhesion/cytoskeleton, nucleic acid processing/ among the responders (five patients) identified 8 transcription, and signal transduction emerged as the differentially expressed genes (P!0.001, PZ0.3125 most represented functional categories of observing such a number), while 82 differentially Among genes modulated in all patients upregulation expressed genes (P!0.001, PZ0.03125 of observing predominated in the cell cycle proliferation category such a number) were identified among non-responders (JUNB, S100A10, and PPARG) and slightly in the cell (six patients). adhesion and cytoskeleton category (CTTN, KAL1, Within genes modulated independently of treatment and MXRA8). On the contrary, in the category of outcome protein transport/metabolism, cell signaling, protein transport/metabolism genes, downregulation and nucleic acid processing/transcription were more (18.9%) slightly overcame upregulation (14.3%). The represented, while the cell adhesion class was less former included RAB27A involved in protein transport represented (Table 3). Genes coding for proteins and also in small GTPase-mediated signal transduc- related to cell stress response and apoptosis pathways, tion, SYT11, RPE, TTC7L1, and ANKIB1 involved in

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access 444 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 439–449

Among genes exclusively modulated in non- responders upregulation of cell stress/apoptosis was more frequent and included both antiapoptotic (NCL and CLU) and pro-apoptotic (BCL2L13) genes (Table 4). Genes related to signal transduction and cell cycle/proliferation were more frequently down- regulated and included only two upregulated genes, DKK3, a negative regulator of WNT signaling and MAPK6. Furthermore, a striking upregulation (41.2%) of genes associated with cell adhesion/cytoskeleton (ITGB1, SMOC2, FBLN2, FMOD, SERPIN1, TIMP3, and MMP2) was observed.

Table 4 Genes modulated by treatment in non-responder a Figure 2 Hierarchical clustering of the gene expression patients only matrix containing no missing values of 11 paired samples (five responders and six non-responders) before and after Biological function Gene symbol P value treatment. The solid lines indicate paired samples with similar gene expression profiles of responding patients and Upregulated genes the dashed lines indicate pairing of non-responding Cell stress/apoptosis CLU 0.0001582 patients. The percentage of tumor reduction is reported BCL2L13 0.0006041 under each pre-treatment sample. NCL 0.0008061 Adhesion/cytoskeleton FMOD 0.000246 ubiquitin-dependent protein catabolism, the mito- TIMP3 0.0004362 SERPIN1 0.0005444 chondrial carrier protein MCART6 and CRLS1, a FBLN2 0.0005989 synthase involved in phospholipid biosynthesis, while SMOC2 0.0008047 upregulated genes included PDZRN3 involved in MMP2 0.000805 protein ubiquitination, PPIG, a peptidyl-prolyl isomerase ITGB1 0.0008549 Protein transport/ TOMM40 0.0009673 accelerating protein folding and also involved in metabolism regulation of pre-mRNA splicing, and the translation Signal transduction DKK3 0.0009238 initiation factor EIF3S2, besides a mannosidase Nucleic acid processing ZNF71 0.0004078 (MAN1C1) involved in glycosylation. and transcription GABPB2 0.0005375 K Cell cycle/proliferation MAPK6 8.56!10 5 Membrane solute SLC39A13 0.0006409 Table 3 Classification into functional categories of genesa transport modulated by the treatment independently from treatment Other PPT1R11 0.0006333 response Downregulated genes Functional categories of Adhesion/cytoskeleton C1QTNF7 0.708 genes modulated by In all In non-responder CLDN18 0.0002255 treatment patientsb patientsb ARPC4 0.0004617 Signal transduction ADRBK2 0.0004755 Stress/apoptosis 8.3 9.1 DKFZP686A10121 0.0005211 Cell adhesion/cytoskeleton 6.9 20.0 PTCH 0.0006665 Protein transport/ 16.7 10.9 TXK 0.0009083 metabolism Nucleic acid processing ZNF677 0.0002018 Signal transduction 13.9 14.5 and transcription Nucleic acid processing 13.9 20.0 SBNO1 0.0004757 Cell cycle proliferation 8.3 7.3 TAF1B 0.0006169 ER signaling – 1.8 ZNF331 0.0007066 Steroid metabolism 2.8 – SRPK2 0.0007598 Immunology 1.4 – Cell cycle/proliferation ORC1L 0.0001482 Membrane solute transport 9.7 7.3 HSMPP8 0.0007817 Other classes 18.1 9.1 ER signalling NCOA5 0.0008897 Other BACE 0.0004043 aThe percentage of unknown reporters was 28.7% on the DLD 0.0005991 overall series and 49.1% exclusively in specimens from non- responder patients. aOnly those genes that are modulated in non-responders are bRelative percentages on known genes. reported, excluding genes common to other gene lists.

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access www.endocrinology-journals.org 445 V Cappelletti, M Gariboldi et al.: Gene expression and anti-estrogen treatment

Confirmation of gene expression data by silenced treating the cells with an siRNA specific for real-time PCR cytoplasmic CLU, the form responsible for the We validated cDNA microarray data by real-time PCR cytoprotective effects. CLU appeared on blots as the for eight genes differentially expressed in pre- and post- 60 kDa precursor and an about 40 kDa protein smear treated specimens from non-responding patients comprising the two a and b dimers and their glyco- (TIMP3, MMP2, FBLN2, NLC, BCL2L13, CLU, sylated forms. Treatment with siRNA effectively MAPK6, and DKK3) and three genes whose expression suppressed the expression of the precursor and the K K correlated with tumor shrinkage (RUNX3, GSPT1, and active forms ( 75 and 95% respectively; Fig. 3B), while no effects on protein expression were observed DYRK2). Paired t-test and Spearman’s rank correlation K7 analysis confirmed all but one (DYRK2) of the tested with the control siRNA. Treatment with 10 M genes (see Supplementary Table 1). toremifene for 3 days did not affect cell proliferation (percentage of untreated control: 102G4% vs 100G 5%). Addition of unrelated control siRNA, which did Biological validation of CLU not affect CLU expression, resulted in a slight but not a CLU, one of the genes upregulated following treatment statistically significant reduction (83G14%), while exclusively in the subset of non-responders (Supple- treatment with the siRNA CLU-V significantly mentary data Table 1), was identified as possible can- inhibited cell proliferation down to 56G0.6% didate for attempting to modulate the sensitivity to (PZ0.032 with respect to siRNA ctrl). anti-estrogens in a toremifene-resistant cell line: T47D. CLU was chosen as candidate gene due to its Discussion association with survival and apoptosis (Shannon et al. 2006). Treatment with 10K7 M 4-OH toremifene The combination of primary endocrine therapy with or with the similar drug tamoxifen slightly upregulated global gene profiling provides the opportunity to collect CLU expression (Fig. 3A). CLU expression was information to predict response to SERM and to identify genes involved in anti-estrogen resistance. The clinical response of the tumor can in fact be regarded as a surrogate of response to systemic treatment, and genes modulated in non-responding patients might help to identify pathways associated with resistance. In the present study, few genes were differentially expressed between responders and non-responders confirming the difficulties in finding a reliable molecular predictor of response using a top-down approach. Limited duration of clinical treatment, which resulted in few patients achieving pCR, scanty cases available for microarray analysis, and their heterogeneity could account for this failure. We did, however, identify a gene set significantly correlated with toremifene-induced tumor shrinkage evaluated on a continuous scale in pre-operative samples, without overlapping with previously identified signatures of endocrine treatment response in the adjuvant setting (Ma et al. 2004) or in women with advanced disease (Paik et al. 2004, Jansen et al. 2005). Conversely, a novel contribution of our study to the relation between gene expression profile and treatment effect resides in Figure 3 (A) Western blotting of CLU active forms in T47D cells K K the analysis of transcriptome changes after anti- treated for 6 days with 10 7 M 4-OH toremifene and 10 7 M 4-OH tamoxifene. (B) Western blotting of CLU precursor and estrogen treatment. To exclude the possibility that active forms, in controls and in siRNA-treated samples. T47D genes differentially expressed before and after treat- cells were treated with 25 nM control siRNA and siRNA against ment could be due to surgical procedure, elapsed time the secreted form of CLU for 24 h in serum-free medium. The medium was replaced and after 3 days cells were harvested or simply to some random effects, rather that to and lysed. The figure reports representative blots of three treatment itself, we verified in advance that a negligible separate experiments. fraction of genes (2/4205) were differentially

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access 446 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 439–449 expressed in matched biopsies and surgical specimens linked to resistance. This latter concept was further collected within a time lapse of 3–4 weeks and not supported by the fact that our data indirectly confirm in a submitted to intercurrent treatment. Investigations on clinical setting some experimental studies aimed at gene modulation in the absence of treatment are understanding the action mechanism of SERM stimu- generally underrepresented in the literature, and even lation and the role of membrane ER (Shou et al. 2004), studies dealing with post-treatment changes in gene which associates with the scaffolding protein caveolin expression are scanty, with only one paper reporting and activates proximal signaling molecules like differentially expressing genes following anastrozole G-proteins (Razandi et al. 2003). Such a signaling in locally advance breast cancer (Nedelcheva et al. results into stimulation of phospholipase C and ErbB2/ 2005), and few other papers investigating transcrip- EGFR heterodimers, which lead to activation of PIrK tional profiling following docetaxel alone (Chang et al. and AKT pathway (Stoica et al. 2003), and finally of 2003, 2005) or in combination with doxorubicin and MAPK. Specific G-proteins allow the activation cyclophosphamide (Hannemann et al. 2005), or erlo- of MMP able to transactivate heparin-bound tinib (Yang et al. 2005). EGF (Razandi et al. 2003). In our study, toremifene In our study, treatment-modulated genes appear to affected different genes involved in G-protein-mediated differ from treatment outcome-associated genes, a signaling, including GEM, which was upregulated finding also reported with a slightly different type of independently of treatment response and could play a analysis in the only other study investigating post- role as a regulatory protein in receptor-mediated signal treatment changes as a function of clinical response to transduction. Exclusively in the non-responder subset, anastrozole (Nedelcheva et al. 2005). These authors toremifene upregulated MMP2 and MAPK6, which observed that most pre- and post-treatment tumor were shown to be activated by estradiol following specimens clustered together, indirectly suggesting the binding to the membrane receptor, a mechanism existence of a set of treatment outcome-associated involved in anti-estrogen resistance. A validation of genes unchanged after treatment. This finding gene signatures obtained from this study on independent indirectly confirms and provides support to our patients set was not possible due to the lack of patients observation of a negligible overlap between treatment treated with pre-operative anti-estrogens. We, however, response-associated and response-modulated genes. In attempted a biological validation of the concept that addition, in our study treatment outcome appears to genes modulated by the treatment exclusively in the affect gene modulation, since transcriptome variations non-responding patients, may include genes involved in were minimal in responding patients in agreement with hormone resistance. We chose an ERC anti-estrogen- Chang et al. (2005) but in contrast with Hannemann resistant cell line, T47D. Treatment with anti-estrogens et al. (2005). However, disagreements on gene upregulated CLU expression in agreement with that of modulations according to clinical response might be Warri et al. (1993), although to a lesser extent because due to differences in treatment type and criteria used to the anti-estrogen-resistant cell line expressed much assess clinical/pathological response, and to the more CLU compared with the sensitive MCF7 cells possibility that in non-responding patients a differential used by Warri et al. Silencing of clusterin using an gene expression reflects biological differences between siRNA specific for the cytoplasmic variant restored the epithelial and stromal cells (the latter frequently sensitivity of T47D cells toward growth inhibitory increased after primary treatment). Despite such effects of toremifene. Such results do not demonstrate a problems and potential bias due to investigating direct involvement of CLU in anti-estrogen-resistant treatment-induced changes mainly in partially or non- mechanisms, but strongly support the concept that CLU responding patients, we reasoned that focusing on may modulate response to anti-estrogens and represents genes exclusively modulated in non-responders could therefore an interesting pharmaceutical target. provide information on anti-estrogen resistance. Overall, this study suggests that knowledge of Indeed, many among the genes modulated only in molecular networks modulated by treatment is parti- non-responders are involved in the balance between cularly important in view of the abundance of pro-apoptotic and anti-apoptotic signals as well as in promising molecularly targeted inhibitors currently cell signaling from membrane receptors, which are available as treatment options. known to play a role in the establishment of hormone resistance. This observation represents a proof of our hypothesis that genes upregulated after treatment in Acknowledgements resistant patients may include potential therapeutic This work was supported by the Italian Association for targets as well as genes representative of pathways Cancer Research (AIRC) and Special Grant ‘Ricerca

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access www.endocrinology-journals.org 447 V Cappelletti, M Gariboldi et al.: Gene expression and anti-estrogen treatment

Finalizzata’ from the Italian Health Ministry. The Karolchik D, Baertsch R, Diekhans M, Furey TS, Hinrichs A, authors declare that there is no conflict of interest that Lu YT, Roskin KM, Schwartz M, Sugnet CW, Thomas DJ would prejudice the impartiality of this scientific work. et al. 2003 The UCSC genome browser database. Nucleic Acids Research 31 51–54. Kent WJ 2002 BLAT – the BLAST-Like alignment tool. Genome Research 12 656–664. References Ma XJ, Wang Z, Ryan PD, Isakoff SJ, Barmettler A, Fuller A, Muir B, Mohapatra G, Salunga R, Tuggle JT et al. 2004 A Cappelletti V, Celio L, Bajetta E, Allevi A, Longarini R, two-gene expression ratio predicts clinical outcome in Miodini P, Villa R, Fabbri A, Mariani L, Giovanazzi R breast cancer patients treated with tamoxifen. Cancer Cell 5 et al. 2004 Prospective evaluation of estrogen receptor-b 607–616. in predicting response to neoadjuvant antiestrogen Nedelcheva Kristensen V, Sorlie T, Geisler J, Yoshimura N, therapy in elderly breast cancer patients. Endocrine- Linegjaerde OC, Glad I, Frigessi A, Harada N, Lonning Related Cancer 11 761–770. PE & Borresen-Dale AL 2005 Effects of anastrazole on De Cecco L, Marchionni L, Gariboldi M, Reid JF, the intratumoral gene expression in locally advanced Lagonigro MS, Caramuta S, Ferrario C, Bussani E, Mezzanzanica D, Turatti F et al. 2004 Gene breast cancer. Journal of Steroid Biochemistry and expression profiling of advanced ovarian cancer: Molecular Biology 95 105–111. characterization of a molecular signature involving Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M, Baehner fibroblast growth factor 2. Oncogene 23 8171–8183. FL, Walker MG, Watson D, Park T et al. 2004 A Chang JC, Wooten EC, Tsimelzon A, Hilsenbeck SG, multigene assay to predict recurrence of tamoxifen- Gutierrez MC, Elledge R, Mohsin S, Osborne CK, treated, node-negative breast cancer patients. New Chamness GC, Allred DC et al. 2003 Gene profiling for England Journal of Medicine 351 2817–2826. the prediction of therapeutic response to docetaxel in Peto R, Boreham J, Clarke M, Davies C & Beral V 2000 UK patients with breast cancer. Lancet 362 362–369. and USA breast cancer deaths down 25% in year 2000 at Chang JC, Wooten EC, Tsimelzon A, Hilsenbeck SG, ages 20–69 years. Lancet 355 1822. Gutierrez MC, Tham YL, Kalidas M, Elledge R, Razandi M, Pedram A, Park ST & Levin ER 2003 Proximal Mohsin S, Osborne CK et al. 2005 Patterns of resistance events in signalling by plasma membrane estrogen and incomplete response to docetaxel by gene receptors. Journal of Biological Chemistry 278 expression profiling in breast cancer patients. Journal of 2701–2712. Clinical Oncology 23 1169–1177. R Development Core Team 2006 R: a language and Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane environment for statistical computing. R Foundation for HC & Lempicki RA 2003 DAVID: database for Statistical Computing, Vienna, Austria. ISBN 3-900051- annotation, visualization, and integrated discovery. 07-0, URL http://www.R-project.org. Genome Biology 4 P3. Saal LH, Troein C, Vallon-Christersson J, Gruvberger S, Diehn M, Sherlock G, Binkley G, Jin H, Metese JC, Borg A & Peterson C 2002 BioArray software environ- Hernandez-Boussard T, Reees CA, Cherry JM, Botstein D, ment (BASE): a platform for comprehensive management Brown PO et al. 2003 SOURCE: a unified genomic and analysis of microarray data. Genome Biology 3 resource of functional annotations, ontologies, and gene software0003.1–software0003.6. expression data. Nucleic Acids Research 31 219–223. Shannon B, Seifert M, Leskov K, Willis J, Boothman D, Geisler J, Detre S, Berntsen H, Ottestad L, Lindtjorn B, Tilgen W & Reichrath J 2006 Challenge and promise: Dowsett M & Eistein Lonning P 2001 Influence of roles for CLU in pathogenesis, progression and neoadjuvant anastrazole (Arimidex) on intratumoral therapy of cancer. Cell Death and Differentiation 13 estrogen levels and proliferation markers in patients with 12–19. locally advanced breast cancer. Clinical Cancer Research Shou J, Massarweh S, Osborne CK, Wakeling Wakeling AE, 7 1230–1236. Ali S, Weiss H & Schiff R 2004 Mechanisms of tamoxifen Hannemann J, Oosterkamp HM, Bosch CA, Velds A, resistance: increased estrogen receptor-HER2/neu cross- Wessels LF, Loo C, Rutgers EJ, Rodenhuis S & van de talk in ER/HER2-positive breast cancer. Journal of the Vijver MJ 2005 Changes in gene expression associated National Cancer Institute 96 926–935. with response to neoadjuvant chemotherapy in breast Simon R, Korn E, McShane L, Radmacher MD, Wright GW & cancer. Journal of Clinical Oncology 23 3331–3342. Zhao Y 2003 In Design and Analysis of DNA Microarray Jansen MP, Foekens JA, van Staveren IL, Dirkzwager-Kiel Investigations, ch 7. New York: Springer-Verlag. MM, Ritstier K, Look MP, Meijervan Gelder ME, Simon R, Lam A, Li MC, Ngan M, Menenzes S & Zhao Y Sieuwerts AM, Portengen H, Dorssers LC et al. 2005 2007 Analysis of gene expression data using BRB-Array Molecular classification of tamoxifen-resistant breast tools. Cancer Informatics 2 11–17. carcinomas by gene expression profiling. Journal of Stoica GE, Franke TF, Moroni M, Mueller S, Morgan E, Iann Clinical Oncology 23 732–740. MC, Winder AD, Reiter R, Wellstein A & Martin MB

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access 448 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 439–449

2003 Effect of estradiol on estrogen receptor-a gene Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J & expression and activity can be modulated by the ErbB2/PI Speed TP 2002 Normalization for cDNA microarray 3-K/Akt pathway. Oncogene 22 7998–8011. data: a robust composite method addressing single and Wa¨rri AM, Huovinen RL, Laine AM, Martikainen PM & multiple slide systematic variation. Nucleic Acids Harkonen PL 1993 Apoptosis in toremifene-induced growth Research 3 e15. inhibition of human breast cancer cells in vivo and in vitro. Yang SX, Simon RM, Tan AR, Nguyen D & Swain SM Journal of the National Cancer Institute 85 1412–1418. 2005 Gene expression patterns and profile changes Wright GW & Simon R 2003 A random variance model for pre-and post-erlotinib treatment in patients with detection of differential gene expression in small metastatic breast cancer. Clinical Cancer Research 11 microarray experiments. Bioinformatics 19 2448–2455. 6226–6232.

Downloaded from Bioscientifica.com at 09/30/2021 07:03:58PM via free access www.endocrinology-journals.org 449