An immunohistochemical evaluation of the proteins Wnt1and glycogen synthase kinase (GSK)-3b in invasive breastcarcinomas

Eleni Mylona,1,* Ioannis Vamvakaris,2,* Ioanna Giannopoulou,2 Irene Theohari,2

Christos Papadimitriou,3 Antonios Keramopoulos3 & Lydia Nakopoulou215th Department of Internal Medicine, Evagelismos Hospital, 21st Department of Pathology, Medical School, University

of Athens, and 3Department of Clinical Therapeutics, Alexandra General Hospital, Medical School, University of Athens,

Athens, Greece

Date of submission 18 July 2012Accepted for publication 9 January 2013Published online Article Accepted 13 January 2013

Mylona E, Vamvakaris I, Giannopoulou I, Theohari I, Papadimitriou C, Keramopoulos A & Nakopoulou L

(2013) Histopathology 62, 899–907

An immunohistochemical evaluation of the proteins Wnt1 and glycogen synthase kinase(GSK)-3b in invasive breast carcinomas

Aims: Our purpose was to investigate, in breast car-cinomas, the prognostic importance of the proteinsWnt1 and glycogen synthasekinase (GSK)-3b, andtheir associations with classic clinicopathological indi-ces.Methods and results: Immunohistochemistry was per-formed on paraffin-embedded tissue specimens from288 invasive breast carcinomas to detect the expres-sion of the proteins Wnt1, GSK3b, oestrogen receptor(ER), progesterone receptor (PR), erbB2, p53, Ki67,caspase-3 and b-catenin. Both Wnt1 and GSK3bwere detected predominantly in the cytoplasm of theinvasive tumour cells and the in-situ component,while GSK3b was also detected in the stromal fibro-blasts. Wnt1 immunoreactivity in the invasive

tumour cells showed an inverse association with his-tological grade (P = 0.002), Ki67 (P = 0.008) andp53 (P = 0.031), while its relation with ER, erbB2and caspase-3 was found to be positive (P = 0.007,P = 0.018 and P = 0.03, respectively). CytoplasmicWnt1 expression was related to a favourable progno-sis within the subgroup of patients with stage IIdisease (P = 0.032).Conclusions: Wnt1 expression in the invasive tumourcells seems to promote differentiation and apoptosis,while being related inversely to proliferation. There-fore, this suggests its participation in the primarystages of breast carcinogenesis. The latter is supportedfurther by the immunodetection of Wnt1 in in-situcarcinomas.

Keywords: apoptosis, breast cancer, GSK3b, immunohistochemistry, proliferation, Wnt1

Introduction

The Wnt proteins constitute a family of cysteine-richsecreted glycoproteins implicated in the activation ofmultiple intracellular signalling cascades.1–3 The

intracellular signalling pathway activated by Wntswas identified originally as a b-catenin-dependentpathway (also referred to as the canonical Wnt sig-nalling pathway). In this, the binding of Wnts to acell–surface receptor complex inhibits the serine/thre-onine kinase glycogen synthase kinase-3b (GSK3b).This results in dephosphorylation, stabilization andthe accumulation of b-catenin in the cell cytoplasmand its entry into the nucleus, where it activates thetranscription of target genes.1,2 In the absence of theWnt signal, active GSK3b phosphorylates b-catenin

Address for correspondence: Professor L Nakopoulou, 1st

Department of Pathology, MedicalSchool, University of Athens,

75 Mikras Asias Street, Goudi, GR-11527, Athens, Greece.

e-mail: lnako@med.uoa.gr

*Both authors contributed equally to this work.

© 2013 Blackwell Publishing Limited.

Histopathology 2013, 62, 899–907. DOI: 10.1111/his.12095

and targets b-catenin for ubiquitination and degrada-tion by the proteasome.1,2

Wnts have also been implicated in the activation ofb-catenin independent pathways (also referred to asnon-canonical Wnt pathways), such as the planar cellpolarity (PCP) pathway and the Ca2+pathway.3 Althoughthese pathways are involved in the regulation of tissuepolarity and cell migration in Drosophila,4 their exactrole in mammals is not well understood. What hasbeen reported, however, is that non-canonical Wntsantagonize canonical Wnt signalling.5

The already complex pathway of Wnt signalling iscomplicated further by the role of GSK3b which,apart from participating in Wnt/b-catenin signalling,is a crossroads at which diverse pathways (such asRas/Raf/MAPK, EGFR/PI3K/Akt and IP3/PKC) con-verge.6

Aberrant activation of the canonical Wnt/b-cateninpathway is one of the most frequent abnormalitiesknown in human cancers.7 In the mammary gland,in particular, Wnt signalling is implicated in severalstages of growth and differentiation, as well as inmammary neoplasia.8,9 Increased Wnt signalling inhuman mammary epithelial cells has been reportedto elicit a DNA damage response, which is consideredto be an early event in human carcinogenesis.10 Itsoncogenic potential has been confirmed further bythe generation of transgenic mice expressing Wnt1.These animals rapidly develop lobulo-alveolar hyper-plasia, and subsequently focal carcinoma, after a longlatency period.11

In recent years it has become apparent that a widerange of human cancers, with colorectal cancer as theprime example, carry mutations in at least one compo-nent of the canonical Wnt/b-catenin pathway.7 How-ever, no such mutations have been reported in breastcarcinomas,9,12 while there is evidence that b-cateninis stabilized in about half of them. Among the explana-tions proposed for the pathway activation in breastcancer is overexpression of the Wnt ligands.8,9

The purpose of the present study was to investigatethe subcellular localization of Wnt1 and GSK3b pro-teins in invasive breast carcinomas to determine theirprognostic importance, as well as their relation towell-known clinicopathological indices, proliferationand apoptosis.

Materials and methods

P A T I E N T S A N D S A M P L E S S T U D I E D

A consecutive series of 290 paraffin blocks withtumour samples were available from patients with

resectable breast cancer who had undergone radicalsurgery in the Gynecology Department of AlexandraUniversity Hospital, Greece, between 1992 and 1996.We selected only women with histologically proven,clearly invasive ductal or lobular breast carcinomas,regardless of their initial stage, in whom axillarylymph node dissection had been performed, who hadhad all their resected materials studied histologically,who had had a follow-up period of more than 5 yearsand who had signed informed consent permitting sci-entific investigation of their tumour tissue. Thepatients were aged from 25 to 87 years (mean58 years). None of them had received radiation orchemotherapy pre-operatively.Routine histological examination was performed

with haematoxylin and eosin staining. All carcinomaswere classified in accordance with World Health Orga-nization criteria13 and were recorded as invasive duc-tal or invasive lobular. The combined histologicalgrade (1,2 or 3) of infiltrating ductal and lobular car-cinomas was obtained according to a modified Scarff–-Bloom–Richardson histological system with guidelinesas suggested by Nottingham City Hospital patholo-gists.14 Nuclear grading was based on nuclear pleo-morphism. Staging at the time of diagnosis was basedon the TNM system.15 The clinicopathological charac-teristics of the series are shown in Tables 1 and 2.Follow-up was available for 288 patients, of whom

66 died of breast cancer and 81 had a recurrence.Patient outcome was defined as overall and recur-rence-free survival. Mean overall survival time was81.59 months (range 6–318 months) and medianrecurrence-free survival time was 75.5 months. Allpatients received conventional postoperative treat-ment depending on the extent of the disease, includ-ing adjuvant chemotherapy, radiation therapy andanti-oestrogen therapy, when indicated, according tothe consensus recommendations at the time.

I M M U N O H I S T O C H E M I S T R Y

Immunohistochemical staining for Wnt1 protein wasperformed on 4-lm-thick formalin-fixed paraffin sec-tions after being heated overnight at 37°C as well assubsequent deparaffinization in xylene and rehydra-tion through graded alcohols. After the quenching ofendogenous peroxidase activity with 0.3% hydrogenperoxide in Tris-buffered saline (TBS)for 30 min, weproceeded to microwave-mediated antigen retrieval in0.01M citrate buffer, pH 6.0, at 750 W for 10 min.After rinsing with TBS, normal horse serum wasapplied for 30 min to block non-specific antibodybinding. For GSK3b, deparaffinization, rehydration

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

900 E Mylona et al.

and antigen retrieval was performed in a one-stepprocedure by treating the slides in a microwave withthe Trilogy reagent (pH 8.0) (Cell Marque,Hotsprings,AR, USA) for 18 min. Endogenous peroxidase activitywas quenched with 0.3% hydrogen peroxide in TBS.Subsequently, sections were incubated overnight at

4°C with the primary antibodies. After additional rins-ing in TBS, sections were incubated with biotinylatedhorse anti-goat (Wnt1) and anti-rabbit (GSK3b) anti-bodies (Vector Laboratories, Burlingame, CA, USA) for30 min at room temperature, and then incubatedwith avidin–biotin–peroxidase complex (VectastainElite ABC kit; Vector Laboratories) for 30 min. Theperoxidase reaction was developed with a 0.5 mg/mlsolution of 3,3′-diaminobenzidine tetrahydrochloride(Sigma Chemical Co., St Louis, MO, USA) supple-mented with 0.01% H2O2. Finally, sections werecounterstained with haematoxylin and mounted.A goat polyclonal antibody (G19) against Wnt1 (sc

6280;Santa Cruz Biotechnology, Santa Cruz, CA,USA) was used at a dilution of 1:50; and a rabbitpolyclonal antibody against GSK3b (clone Η76; SantaCruz Biotechnology) at a dilution of 1:70. For thedetection of oestrogen receptor (ER), progesteronereceptor (PR), p53, erbB2, Ki67, caspase-3 and b-catenin,we used the following antibodies overnight, in aprocess similar to the aforementioned: anti-ER, clone1D5 (1:450; Dako, Glostrup, Denmark); anti-PR,clone 1A6 (1:150; Dako); anti-p53, clone BP 53.12.1(1:50; Oncogene, Cambridge, MA, USA); anti-cerbB2,

Table 1. Association of cytoplasmic Wnt1 protein overexpres-sion with clinicopathological parameters and biological indices

Total

Cytoplasmic Wnt1

n % P value

Menopausal statusBefore 81 28 34.5 NS

After 207 73 35.2

Histological typeDuctal 231 80 34.6 NS

Lobular 57 24 42.1

Histological grade1 28 16 57.1 0.002

2 174 70 40.2

3 86 18 21.9

Nuclear grade1 113 46 40.7 NS

2 101 33 32.6

3 74 22 29.7

Tumor size<2cm 80 32 40.0 NS

>2cm 208 70 33.6

Lymph nodesNot infiltrated 115 39 33.9 NS

Infiltrated 173 62 35.8

Stage1 60 20 33.3 NS

2 186 67 36.0

3 42 15 35.7

ERNegative 104 26 25.0 0.007

Positive 184 77 41.8

PRNegative 102 32 31.3 NS

Positive 186 71 38.1

c-erbB2Negative 235 71 30.2 0.018

Positive 53 26 49.0

p53Negative 227 111 48.8 0.031

Positive 61 19 31.1

Table 1. (Continued)

Total

Cytoplasmic Wnt1

n % P value

Ki67Negative 151 70 46.3 0.008

Positive 137 38 27.7

caspase-3Negative 110 57 51.8 0.030

Positive 178 134 75.2

GSKbNegative 114 38 33.3 NS

Positive 174 65 37.3

b-cateninNegative 144 49 34.0 NS

Positive 144 51 35.4

Bold denotes statistically significant.

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

Wnt1 and GSK3b in breast cancer 901

clone CB11 (1:150; Biogenex, San Ramon, CA, USA);rabbit polyclonal anti-Ki67 (1:50; Dako); goat poly-clonal anti-caspase-3 (1:80; Santa Cruz Biotechnol-ogy); and anti-phospho-b-catenin (#9561, 1:50; CellSignaling Technology, Beverly, MA, USA). The resultsof these immunomarkers were retrieved from ourarchival database.16,17

I M M U N O H I S T O C H E M I S T R Y E V A L U A T I O N

Following the above process, 288 sections were suit-able for evaluation. Two of the available sampleswere considered to be unsuitable because the tissuehad been destroyed during the immunohistochemicalstaining process.Evaluation of immunohistochemical staining was

performed independently by two pathologists throughlight microscopic observation and without knowledgeof the clinical data for each patient. Cases of disagree-ment were reviewed jointly to reach a consensusscore. As positive controls, we used breast cancer tis-sue sections known previously to be Wnt1 or GSK3bimmunoreactive. Negative controls had the primaryantibody omitted and replaced by non-immune normal

Table 2. Association of cytoplasmic and stromal GSK3bprotein overexpression with clinicopathological parametersand biological indices

Total

CytoplasmicGSK3b Stromal GSK3b

n %Pvalue n %

Pvalue

Menopausal statusBefore 83 52 62.6 NS 69 83.3 NS

after 205 127 61.9 158 77.0

Histological typeDuctal 230 144 62.6 NS 183 79.5 NS

Lobular 58 32 55.1 41 70.6

Histological grade1 29 13 44.8 NS 24 82.7 NS

2 174 107 61.4 136 78.1

3 85 55 64.7 64 75.2

Nuclear grade1 120 71 59.1 NS 93 77.5 NS

2 91 55 60.7 74 72.7

3 77 50 64.9 56 70.5

Tumor size<2cm 79 45 56.9 NS 68 86.0 NS

>2cm 209 131 62.6 156 74.6

Lymph NodesNegative 114 68 59.6 NS 88 77.1 NS

Positive 174 109 62.6 135 77.5

Stage1 59 33 55.9 NS 48 81.3 NS

2 188 119 63.2 145 75.6

3 41 25 61.9 31 75.9

ERNegative 105 69 65.7 NS 86 81.9 NS

Positive 183 105 57.3 134 73.2

PRNegative 100 68 68.0 NS 79 79.0 NS

Positive 188 106 56.3 141 75.0

c-erbB2Negative 224 137 61.1 NS 180 80.3 NS

Positive 54 32 59.2 45 83.3

Table 2. (Continued)

Total

CytoplasmicGSK3b Stromal GSK3b

n %Pvalue n %

Pvalue

p53Negative 223 140 62.7 NS 178 79.8 NS

Positive 65 39 60.0 45 69.2

Ki67Negative 149 94 63.0 NS 101 67.7 0.05

Positive 139 80 57.5 78 56.1

Caspase-3Negative 111 74 66.7 NS 59 79.0 0.009

Positive 177 108 61.0 140 77.0

Wnt1Negative 182 108 59.3 NS 143 75.4 NS

Positive 106 67 63.2 80 78.5

b-cateninNegative 143 86 60.1 NS 107 74.8 NS

Positive 145 96 66.2 116 80.0

Bold denotes statistically significant.

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

902 E Mylona et al.

serum from the same species as the primary antibody.For each sample, the fraction of positively stainedcells was scored by examining 10 high-power fields(9400) of one section, and calculating the average ofthe percentage of cytoplasmic Wnt1 or GSK3b posi-tive cells in each of the 10 fields. Staining intensitywas also scored, on a scale of 0–4: 0= no staining,1= weak staining in fewer than 10% of tumour cells,2= weak to moderate staining in 11–50% of the cells,3= moderate to strong staining in 51–70% and 4=strong staining in more than 70% of tumour cells.After several statistical trials to determine the

threshold which made the best discrimination possiblebetween the clinicopathological parameters, biologicalindices and survival curves, cut-off points were deter-mined as follows: cases with Wnt1 expression in�70% of tumour cells examined were classified asnegative, and cases with expression in >70% oftumour cellsas positive; and for GSK3b, cases withcytoplasmic expression in �50% of tumour cells wereconsidered to be negative, and those with expressionin >50% of tumour cellsas positive.Stromal GSK3b im-munodetection was evaluated as present or absent.Staining for ER and PR was evaluated according to

CAP/ASCO recommendations, i.e. ER and PR assaysare considered positive if there are at least 1% posi-tive tumour nuclei in the sample in the presence ofthe expected reactivity of internal and external con-trols.18 The evaluation of p53, Ki67 and caspase-3proteins was performed as described previously, using10% as the cut-off value for all proteins.16,17 Thefraction of erbB2-positive stained cells was scoredfrom 0 to 3 according to CAP/ASCO guidelines.19

S T A T I S T I C S

The significance of the relationship between theexpression of Wnt1 and GSK3b and clinicopathologi-cal parameters was evaluated by univariate analysisusingthe v2 test and Fisher’s exact probability test.The effect of Wnt1 and expression on postoperativesurvival rates was assessed by both univariate (log-rank test) and multivariate (stepwise forward Coxproportional-hazard regression model) analysis.P�0.05 was considered statistically significant.

Results

In the present study, both Wnt1 and GSK3b weredetected predominantly in the cytoplasm of the inva-sive tumour cells in 36.5% and 60.6% of the cases,respectively (Figures 1A and 2A). With regard to

Wnt1, there was no stromal immunoreactivity, whileGSK3b was also detected in the peritumoral stroma(76.2%) (Figure 2A, arrow). No expression of eitherprotein was observed in the nuclei of invasive tumourcells. The in-situ component, where present, was im-munostained intensively for Wnt1, while adjacentnormal cells were stained faintly (Figure 1C,D).Regarding GSK3b, both in-situ carcinoma and normalbreast epithelium, where they existed adjacent to theinvasive tumour, were immunoreactive (Figure 2B,C).Wnt1 immunoreactivity in cancer cells showed no

relation to the classic clinicopathological markerssuch as menopausal status, histological type, nucleargrade, tumour size, lymph nodes, stage or PR, exceptfor histological grade, to which it was associatedinversely (P = 0.002) (Figure 1B), as well as ER anderbB2, to which its relation was positive (P = 0.007and P = 0.018, respectively). An inverse relation wasfound between Wnt1 expression and Ki67 as well asp53 (P = 0.008 and P = 0.031, respectively), whilethere was apositive association with caspase-3(P = 0.03) (Table 1). No relation was detectedbetween cytoplasmic Wnt1 and b-catenin or GSK3b.Regarding GSK3b, cytoplasmic expression in the

tumour cells was associated with none of the classicclinicopathological markers, b-catenin or Wnt1(Table 2). Stromal localization of GSK3bwas found tobe related inversely to Ki67 (P = 0.05) and positivelyto caspase-3 (P = 0.009) of the tumour cells. No rela-tion was found between stromal GSK3bexpressionand the traditional clinicopathological indices of theinvasive tumour cells (Table 2).Regarding overall survival, cytoplasmic Wnt1

expression was related to a favourable prognosis forpatients with stage II disease (P = 0.032) (Figure 3).Neither cytoplasmic Wnt1 nor cytoplasmic or stromalGSK3b exerted a prognostic impact on the disease-free or overall survival of the unselected cohort ofour patients (data not shown).

Discussion

In the present study, Wnt1 protein was immunode-tected in the cytoplasm of epithelial and stromal cells,in agreement with the literature.20,21 Normal epithe-lium, where it existed adjacent to tumour, showedweak staining, in accordance with data showing that,while several members of the Wnt gene family havebeen found to be expressed in the developing mammarygland, Wnt1 itself has not.22,23 In contrast, in-situtumour was immunostained intensively, an observationwhich suggests that Wnt1 is activated in premalignant

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

Wnt1 and GSK3b in breast cancer 903

stages of breast tumorigenesis, and which is in agree-ment with the findings from in vitro and in vivo experi-ments. Thus, Wnt1 expression was found to beinsufficient for tumorigenesis in mouse mammaryepithelial cell lines24 and, furthermore, geneticallyaltered mice expressing Wnt1 were reported todevelop lobulo-alveolar hyperplasia rapidly, and subse-quently focal carcinoma, after a long latency period.11

Wnts are secreted glycoproteins which act afterbinding membrane receptors.1 Interestingly, however,in the present and several other studies concerningbreast20,21 or other25,26 carcinomas Wnt1 stainingwas found to be exclusively cytoplasmic, while nostromal localization was detected. Cytoplasmic im-munodetection of Wnt1 protein may be due to endo-cytosis of the Wnt/ligand complex, a mechanismwhich has been suggested to regulate the specificityof activation downstream of Wnts.27 The lack of

stromal immunodetection indicates the existence ofan autocrine function of Wnt signalling in malignantcells. This has been confirmed by experiments con-cerning Wnt2: Wnt2 was reported to be presentexclusively within the fibroblasts of normal breast,while it was expressed in the epithelium of infiltratingductal tumours,28 suggesting a switch from a para-crine to an autocrine mechanism during the processof carcinogenesis.29

As far as GSK3b is concerned, this is the first immu-nohistochemical study to investigate its subcellularlocalizations and, contrary to Wnt1, it was found inthe peritumoral stroma. No staining was detected inthe nuclei of invasive tumour cells, despite the sugges-tion that as some of the substrates of GSK3b arenuclear, the protein is also expected to be located innuclei.30 The absence of Wnt1 expression in stromalfibroblasts where GSK3b was detected indicates that

A B

C D

Figure 1. Intense Wnt1 immunoexpression in the cytoplasm of a low-grade invasive ductal breast carcinoma (A). Faint Wnt1

immunoexpression in the cytoplasm of a high-grade breast carcinoma (B). Wnt1 expression in in-situ carcinoma (C) and in normal

epithelium (faint) (D) (immunoperoxidase).

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

904 E Mylona et al.

stromal GSK3b may be activated by pathways otherthan that involved in Wnt/b-catenin signalling.6

The positive association of tumour Wnt1 with ER andcaspase-3 and its inverse correlation with histological

grade, Ki67 and mutated p53 suggests an associationwith well-differentiated tumours of lower proliferativeand high apoptotic potential; in other words, withtumours of favourable phenotype. In addition, the posi-tive relation between Wnt1 and favourable prognosiswithin the subgroup of patients with stage II of the dis-ease further confirms its propitious role and the sugges-tion that overexpression of Wnt1 seems to be an earlyevent in breast tumorigenesis. This finding is unex-pected given that Wnt signalling is considered to leadto stabilization of b-catenin with the subsequent acti-vation of genes implicated in tumorigenesis.1,2 More-over, up-regulation of Wnt1 protein reported invarious human carcinomas such as those of lung,oesophagus and stomach has, in fact, been related tounfavourable phenotypes and worse prognoses.25,26,31

However, in breast cancer, Wong et al., in accordancewith our findings, reported elevation of Wnt1 expres-sion in grade 1 tumours, yet with a subsequentdecrease in higher tumour grades.21 This unexpectedfinding can be attributed to the involvement of Wnts inboth canonical and non-canonical pathways depend-ing on cell types and conditions.27 In the present study,the lack of association between Wnt1 and b-catenin orGSK3b expression implies that Wnt/b-catenin (canoni-cal) signalling is in an inactive state, and suggests theactivation of non-canonical pathways. Interestingly,non-canonical signalling has been reported to antagonize

A

B

C

Figure 2. GSK3b protein expression in (A) the cytoplasm and

the stromal fibroblasts (arrow) of an invasive ductal breast

carcinoma, (B) in-situ carcinoma and (C) normal epithelium

(immunoperoxidase).

Stage II disease1.0

0.8

0.8

Cum

Sur

viva

l

0.4

0.2

0.1

0 50 100 150 200Overall survival time (months)

250

p=0.032

Wnt1 positive

Wnt1 negative

Figure 3. Effect of Wnt1 overexpression on the overall survival of

patients with stage II disease (Kaplan–Meier curve).

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

Wnt1 and GSK3b in breast cancer 905

the canonical pathway which is implicated mainly intumorigenesis5 and, moreover, it has been shown thatnon-canonical Wnts are unable to transform mousemammary epithelial cells.32 In other words, there maybe circumstances under which the activation of non-canonical pathways is selected, and through inhibitionof the canonical signalling, Wnts are likely to confer atumour-suppressive role in tumours that rely uponcanonical signalling for their survival. In line with theabove data, Wnt5a, in addition to activatingnon-canonical signalling,33 has also been reported toinhibit the activation of the canonical signalling path-way34 as well as to exert tumour-suppressing effects onsome cancer types.35

With regard to GSK3b, its immunoexpression ininvasive tumour cells had no association with theclassic clinicopathological markers, due perhaps to itsparticipation in various signalling pathways,6 withvarious and probably contradictory effects on tumori-genesis. Its stromal expression, however, was foundfor the first time to be related positively to apoptosisand associated inversely with proliferation. Thisfinding shows that stromal GSK3b, in a paracrinemanner, may facilitate apoptosis induction, whilecontributing to proliferation suppression. This is inagreement with data suggesting that stroma may actas a local modulator to impede tumorigenesis at anearly stage and that the desmoplastic response is ahost defence reaction designed to confine the develop-ing tumour.36

In conclusion, in the present series of invasivebreast carcinomas Wnt1 expression in the invasivetumour cells and GSK3b expression in the stromal fi-broblasts, both through alternative and not thecanonical Wnt/b-catenin signalling pathway, seem topromote differentiation and apoptosis, while beingrelated inversely to proliferation, clearly suggestingtheir participation in the primary stages of breast car-cinogenesis. Further investigation is required in orderto elucidate the exact role of these two molecules inbreast tumorigenesis.

Acknowledgements

We are grateful to S Koulyra for editing the manu-script.

References

1. MacDonald BT, Tamai K, He X. Wnt/b-catenin signaling: com-

ponents, mechanisms and diseases. Dev. Cell 2009; 17; 9–26.2. Clevers H. Wnt/b-catenin signaling in development and

disease. Cell 2006; 127; 469–480.

3. Widelitz R. Wnt signaling through canonical and non-canoni-

cal pathways: recent progress. Growth Factors 2005; 23;

111–116.4. Miller RK, McCrea PD. Wnt to build a tube: contributions of

Wnt signaling in epithelial tubulogenesis. Dev. Dyn. 2010;

239; 77–93.5. Van Amerogen R, Mikels A, Nusse R. Alternative Wnt signal-

ing is initiated by distinct receptors. Sci. Signal 2008; 1; re9.

6. Doble B, Woodgett JR. GSΚ-3: tricks of the trade for a multi-

tasking kinase. J. Cell Sci. 2003; 116; 1175–1186.7. Paul S, Dey A. Wnt signaling and cancer development: thera-

peutic implication. Neoplasma 2008; 55; 165–176.8. Brennan KR, Brown AMC. Wnt proteins in mammary develop-

ment and cancer. J. Mammary Gland Biol. Neoplasia 2004; 9;

119–131.9. Howe LR, Brown AMC. Wnt signaling and breast cancer. Can-

cer Biol. Ther. 2004; 3; 36–41.10. Ayyanan A, Givenni G, Giarloni L et al. Increased Wnt signal-

ing triggers oncogenic conversion of human breast epithelial

cells by a Notch-dependent mechanisms. Proc. Natl Acad. Sci.

USA 2006; 103; 3799–3804.11. Tsukamoto AS, Grosschedl R, Guzman RC, Parslow T, Varmus

HE. Expression of the int-1 gene in transgenic mice is associ-

ated with mammary gland hyperplasia and adenocarcinomas

in male and female mice. Cell 1988; 55; 619–625.12. Candidus S, Bischoff P, Becker KF, Hofler H. No evidence

for mutations in the alpha- and beta-catenin genes in

human gastric and breast carcinomas. Cancer Res. 1996; 65;

49–52.13. Tavassoli FA, Devilee P eds. Pathology and genetics:tumors of the

breast and female genital organs. IARC WHO Classification of

Tumours, no. 4. Lyon: WHO Press, 2003; 25–27.14. Robins P, Pinder S, de Klerk N. Histological grading of breast

carcinomas: a study of interobserver agreement. Hum. Pathol.

1995; 6; 873–879.15. Hermanek P, Sabin H eds. TNM Classification of international

union against cancer. In TNM atlas,3rd rev, 4th edn. Berlin:

Springer-Verlag, 1992; 15–25.16. Nakopoulou L, Alexandrou P, Stefanaki K, Panayotopoulou E,

Lazaris AC, Davaris PS. Immunohistochemical expression of

caspase-3 as an adverse indicator of the clinical outcome in

human breast cancer. Pathobiology 2001; 69; 266–273.17. Nakopoulou L, Mylona E, Papadaki I et al. Study of the phos-

phor-b-catenin subcellular distribution in invasive breast carci-

nomas in relation to their phenotype and the clinical outcome.

Mod. Pathol. 2006; 19; 556–563.18. Hammond MEH, Hayes DF, Dowsett M et al. American Society

of Clinical Oncology/College of American Pathologists guideline

recommendations for immunohistochemical testing of estrogen

and progesterone receptors in breast cancer. J. Clin. Oncol.

2010; 28; 2786–2795.19. Wolff AC, Hammond MEH, Schwartz JN et al. American Soci-

ety of Clinical Oncology/College of American Pathologists

guideline recommendations for human epidermal growth factor

receptor 2 testing in breast cancer. J. Clin. Oncol. 2007; 25;

118–145.20. Wong SCC, Lo SFE, Lee KC, Yam JWP, Chan JKC, Hsiao WLW.

Expression of frizzled-related protein and Wnt-signaling mole-

cules in invasive breast tumours. J. Pathol. 2002; 196;

145–153.21. Watanabe O, Imamura H, Shimizu T et al. Expression of twist

and Wnt in human breast cancer. Anticancer Res. 2004; 24;

3851–3856.

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

906 E Mylona et al.

22. Weber-Hall SJ, Phippard D, Niemeyer C, Dale TC. Developmen-

tal and hormonal regulation of Wnt gene expression in the

mouse mammary gland. Differentiation 1994; 57; 205–214.23. Gavin BJ, McMahon AP. Differential regulation of the Wnt

gene family during pregnancy and lactation suggests a role in

postnatal development of the mammary gland. Mol. Cell. Biol.

1992; 12; 2418–2423.24. Hollmann CA, Kittrell FS, Medina D, Butel JS. Wnt1 and int-2

mammary oncogene effects on the b-catenin pathway in

immortalized mouse mammary epithelial cells are not sufficient

for tumorigenesis. Oncogene 2001; 20; 7645–7657.25. Nakashima T, Liu D, Nakano J et al. Wnt1 overexpression asso-

ciated with tumor proliferation and a poor prognosis in non-

small cell lung cancer patients. Oncol. Rep. 2008; 19; 203–209.26. Ly J, Cao XF, Ji L, Zhu B, Tao L, Wang DD. Association of

Wnt1/beta-catenin with clinical pathological characteristics

and progression of esophageal squamous cell carcinomas.

Genet. Test Mol. Biomarkers 2010; 14; 363–369.27. Kikuchi A, Yamamoto M, Sato A. Selective activation mecha-

nisms of Wnt signaling pathways. Trends Cell Biol. 2009; 19;

119–129.28. Dale TC, Weber-Hall SJ, Smith K et al. Compartment switching

of WNT2 expression in human breast tumors. Cancer Res.

1996; 56; 4320–43232.

29. Brown AMC. Wnt signaling in breast cancer: have we come

full circle? Breast Cancer Res. 2001; 3; 351–355.30. Jope RS, Johnson GVM. The glamour and gloom of glycogen

synthase kinase-3. Trends Biochem. Sci. 2004; 29; 95–102.31. Zhang H, Xue Y. Wnt pathway is involved in advanced gastric

carcinoma. Hepatogastroenterology 2008; 55; 1126–1130.32. Shimizu H, Julius MA, Giarr�e M, Zheng Z, Brown AMC, Kiitajew-

ski J. Transformation by Wnt family proteins correlates with reg-

ulation of b-catenin. Cell Growth Differ. 1997; 8; 1349–1358.33. Nomachi A, Nishita M, Inaba D, Enomoto M, Hamasaki M, Mi-

nami Y. Receptor tyrosine kinase Ror2 mediates Wnt5a-

induced polarized cell migration by activating c-Jun N-terminal

kinase via actin-binding protein filamin A. J. Biol. Chem. 2008;

283; 27973–27981.34. Torres MA, Yang-Snyder JA, Purcell SM, DeMarais AA,

McGrew LL, Moon RT. Activities of the Wnt-1 class of secreted

signaling factors are antagonized by the Wnt-5a class and by a

dominant negative cadherin in early Xenopus development.

J. Cell Biol. 1996; 133; 1123–1137.35. McDonald SL, Silver A. The opposing roles of Wnt-5a in can-

cer. Br. J. Cancer 2009; 101; 209–214.36. Angeli F, Koumakis G, ChenM-C , Kumar S, Delinassios JG.

Role of stromal fibroblasts in cancer: promoting or impeding?

Tumor Biol. 2009; 30; 109–120.

© 2013 Blackwell Publishing Ltd, Histopathology, 62, 899–907.

Wnt1 and GSK3b in breast cancer 907