Upregulation of Endocrine Gland-Derived Vascular Endothelial Growth Factor in Papillary Thyroid Cancers Displaying Infiltrative Patterns, Lymph Node Metastases, and BRAF Mutation Daniela Pasquali, 1 Angela Santoro, 2 Pantaleo Bufo, 2 Giovanni Conzo, 1 William J Deery, 3 Andrea Renzullo, 1 Giacomo Accardo, 1 Valentina Sacco, 1 Antonio Bellastella, 1 and Giuseppe Pannone 2 Background: Endocrine gland-derived vascular endothelial growth factor (Prok1) and prokineticin 2 (Prok2) are involved in the organ-specific regulation of angiogenesis, which is a crucial step toward cancer progression in most tumors, including those of thyroid gland. The oncogene BRAF V600E mutation is associated with poor clinical outcome of papillary thyroid cancer (PTC) and can independently predict its recurrence. Design: Our hypothesis was that Prok1 and Prok2 expression levels associated with BRAF mutations can be prognostic factors for PTC outcome. Prok1 and Prok2 were examined in PTC, a cell line derived from a human PTC ( AU1 c FB-2 [TPC-1]), euthyroid multinodular goiter (MNG), Graves’ disease (GD), and contralateral normal thyroid (NT) tissues from PTC cases. We evaluated BRAF mutation and its relationship with Prok1 expression pattern in PTC. Methods: We studied Prok1 and Prok2 mRNAs by real-time polymerase chain reaction and BRAF mutation by mutant allele-specific polymerase chain reaction amplification. Formalin-fixed, paraffin-embedded blocks of PTC and NT were used for the immunohistochemical determination of Prok1 using anti-endocrine gland vas- cular endothelial growth factor primary antibody. Results: Prok1 and Prok2 transcripts were both present in thyroid tissues, and Prok1 was differentially expressed in PTC compared to MNG, GD, and NT. Prok1 mRNA levels were very low in NT and MNG and significantly higher in PTC, AU1 c FB-2, and GD ( p < 0.05). Prok1 protein was almost undetectable in NT but was highly expressed in all PTC samples having an infiltrative pattern of growth and lymph node metastases ( p < 0.05). Further, the expression of Prok1 in PTC was associated with 60% of the samples being positive for the BRAF mutation ( p < 0.05). Conclusions: We found that Prok1 is significantly increased in PTC, and its expression in PTC is related to BRAF mutation. These results suggest that Prok1 could be a new useful marker for thyroid cancer progression. Prok1 therefore could also be a potential target for novel therapeutic strategies, although the lack of functional data suggests caution against generalization of this assumption. Introduction P apillary thyroid cancer (PTC) represents *80% of all thyroid malignancies (1). The overall incidence of this par- ticular form of thyroid cancer is rising for reasons that remain unclear, but, in part, could be related to improved diagnostic procedures (2,3). Although differentiated thyroid cancer is typically an indolent disease, recurrence is common (15%–30% of patients) even in early stage disease (4–9). It is therefore crucial to identify patients at higher risk of recurrence, so that more aggressive therapy and monitoring can be considered. In vitro studies using benign thyroid cell models have dem- onstrated a particularly important role for BRAF as a central regulator of thyroid-specific protein expression (i.e., differenti- ation) and proliferative capacity (10). The discovery that mutations in BRAF resulting in constitutive kinase activation are common occurrences in solid tumors has led a number of groups to look for similar BRAF mutations in thyroid cancers (11). Overall, mutations that result in a V600E substitution in BRAF and consequent constitutive activation occur in *45% of adult PTCs, thus making BRAF mutations the most common defined genetic abnormality in thyroid cancers (12). In several 1 Department of Clinical and Experimental Medicine and Surgery, Endocrine Unit, Second University of Naples, Naples, Italy. 2 Department of Surgical Science, Pathology Section, University of Foggia, Foggia, Italy. 3 Department of Biochemistry and Cell Biology, Rice University, Houston, Texas. THYROID Volume 00, Number 00, 2011 ª Mary Ann Liebert, Inc. DOI: 10.1089/thy.2010.0168 1 THY-2010-0168-Pasquali_1P Type: research-article THY-2010-0168-Pasquali_1P.3d 01/31/11 1:15pm Page 1
Transcript
Upregulation of Endocrine Gland-Derived VascularEndothelial Growth Factor in Papillary Thyroid Cancers
Displaying Infiltrative Patterns, Lymph NodeMetastases, and BRAF Mutation
Daniela Pasquali,1 Angela Santoro,2 Pantaleo Bufo,2 Giovanni Conzo,1 William J Deery,3 Andrea Renzullo,1
Giacomo Accardo,1 Valentina Sacco,1 Antonio Bellastella,1 and Giuseppe Pannone2
Background: Endocrine gland-derived vascular endothelial growth factor (Prok1) and prokineticin 2 (Prok2) areinvolved in the organ-specific regulation of angiogenesis, which is a crucial step toward cancer progression inmost tumors, including those of thyroid gland. The oncogene BRAF V600E mutation is associated with poorclinical outcome of papillary thyroid cancer (PTC) and can independently predict its recurrence.Design: Our hypothesis was that Prok1 and Prok2 expression levels associated with BRAF mutations can beprognostic factors for PTC outcome. Prok1 and Prok2 were examined in PTC, a cell line derived from a humanPTC (AU1 c FB-2 [TPC-1]), euthyroid multinodular goiter (MNG), Graves’ disease (GD), and contralateral normalthyroid (NT) tissues from PTC cases. We evaluated BRAF mutation and its relationship with Prok1 expressionpattern in PTC.Methods: We studied Prok1 and Prok2 mRNAs by real-time polymerase chain reaction and BRAF mutation bymutant allele-specific polymerase chain reaction amplification. Formalin-fixed, paraffin-embedded blocks ofPTC and NT were used for the immunohistochemical determination of Prok1 using anti-endocrine gland vas-cular endothelial growth factor primary antibody.Results: Prok1 and Prok2 transcripts were both present in thyroid tissues, and Prok1 was differentially expressedin PTC compared to MNG, GD, and NT. Prok1 mRNA levels were very low in NT and MNG and significantlyhigher in PTC,AU1 c FB-2, and GD ( p< 0.05). Prok1 protein was almost undetectable in NT but was highly expressedin all PTC samples having an infiltrative pattern of growth and lymph node metastases ( p< 0.05). Further, theexpression of Prok1 in PTC was associated with 60% of the samples being positive for the BRAF mutation( p< 0.05).Conclusions: We found that Prok1 is significantly increased in PTC, and its expression in PTC is related to BRAFmutation. These results suggest that Prok1 could be a new useful marker for thyroid cancer progression. Prok1therefore could also be a potential target for novel therapeutic strategies, although the lack of functional datasuggests caution against generalization of this assumption.
Introduction
Papillary thyroid cancer (PTC) represents *80% of allthyroid malignancies (1). The overall incidence of this par-
ticular form of thyroid cancer is rising for reasons that remainunclear, but, in part, could be related to improved diagnosticprocedures (2,3). Although differentiated thyroid cancer istypically an indolent disease, recurrence is common (15%–30%of patients) even in early stage disease (4–9). It is thereforecrucial to identify patients at higher risk of recurrence, so thatmore aggressive therapy and monitoring can be considered.
In vitro studies using benign thyroid cell models have dem-onstrated a particularly important role for BRAF as a centralregulator of thyroid-specific protein expression (i.e., differenti-ation) and proliferative capacity (10). The discovery thatmutations in BRAF resulting in constitutive kinase activationare common occurrences in solid tumors has led a number ofgroups to look for similar BRAF mutations in thyroid cancers(11). Overall, mutations that result in a V600E substitution inBRAF and consequent constitutive activation occur in *45% ofadult PTCs, thus making BRAF mutations the most commondefined genetic abnormality in thyroid cancers (12). In several
1Department of Clinical and Experimental Medicine and Surgery, Endocrine Unit, Second University of Naples, Naples, Italy.2Department of Surgical Science, Pathology Section, University of Foggia, Foggia, Italy.3Department of Biochemistry and Cell Biology, Rice University, Houston, Texas.
THYROIDVolume 00, Number 00, 2011ª Mary Ann Liebert, Inc.DOI: 10.1089/thy.2010.0168
studies, the presence of a BRAF V600E mutation has been as-sociated with a more aggressive clinical course (13–16), al-though these results have not been confirmed in other studies(17–19). BRAF mutations alone may not be sufficient to inducededifferentiation, and the true predictive nature of BRAF mu-tations for aggressive behavior is still debatable.
Angiogenesis, the formation of new blood vessels from pre-existing vasculature, plays a key role in the development,growth, and metastases of carcinomas (20). Potential stimu-lators of angiogenesis, such as vascular endothelial growthfactor A (VEGF-A), have been identified (21). Overexpressionof VEGF has been documented in various malignancies, in-cluding PTC (22–26). The level of VEGF expression appears tobe closely correlated with tumor size, extra-thyroidal inva-sion, and stage, and is increased in PTC having BRAF muta-tion (26). Two new angiogenic factors that selectively act onthe endothelium of endocrine gland (EG) cells have been de-scribed and named EG-derived VEGF (also called prokineti-cin1 and prokineticin2) (27,28). Prok1 and Prok2 act byinducing proliferation, migration, and fenestration of endo-thelium derived from adrenal capillaries, but not of otherendothelium types, such as those derived from aorta, umbil-ical vein, and dermis (27). These peptides are structurallyunrelated to VEGF, regulate diverse biological functions, andare particularly involved in angiogenesis. Prok1 and Prok2exert their physiological functions through G-protein-coupledreceptors named PKAU2 c receptor type 1 and PK receptor type 2(29–31). Previously, we found that while prokineticins andtheir receptors, PK-R1 and PK-R2, are expressed at very lowlevel in normal prostate, their levels are increased with pros-tate malignancy (32). It is still unknown whether prokineticinsare present in normal thyroid (NT) and malignant thyroid orwhether their level of expression is correlated with tumormalignancy and BRAF mutation.
Thus, in the present study we examined Prok1 and Prok2expression in PTC using contralateral NT tissue as control,euthyroid goiter, Graves’ disease (GD) tissue, and also anAU2 c FB-2thyroid cancer cell line. We show in PTC tissue that there is asignificant relationship between BRAF mutation and Prok1expression levels, which has interesting implications for noveltreatments of the disease.
Materials and Methods
RNA isolation
Total RNA was extracted from 30 tissue samples of PTCand 30 contralateral NT tissues from PTC cases, 20 euthyroidmultinodular goiters (MNG), and 10 GD samples. All thepatients were treated at the Endocrine Unit and underwentthyroidectomy at Endocrine Surgery Unit of the Departmentof Clinical and Experimental Medicine and Surgery, SecondUniversity of Naples, Italy. Patients with PTC were subjectedto total thyroidectomy and prophylactic neck dissection.Tumor specimens were obtained in accordance with protocolsapproved by the institutional review board, and the infor-mative consent was achieved 1 day before surgery togetherwith the surgical one. Tissue samples were immediately fro-zen at �808C after surgery. RNA extracted from the FB-2(TPC-1), a cell line derived from a human papillary thyroidcarcinoma kindly supplied by Prof. Santoro M., Departmentof Endocrinology and Experimental Oncology, Naples, wasincluded for controls. Total RNA was recovered with TRIZOL
kit (Invitrogen). Residual DNA was removed by RNase-freeDNase I treatment (Promega). To evaluate the expression ofProk1 and Prok2, RT b AU2-polymerase chain reaction (PCR) wasperformed in which these genes were amplified with glycer-aldehyde-3-phosphate dehydrogenase as internal control(32,33). RNAs were reverse transcribed using 5mg total RNAas previously described (33). To obtain a negative control forthe amplification reactions, we carried out an RNA tran-scription without adding reverse transcriptase. cDNA(400 ng) obtained by RT of RNAs was amplified in a totalvolume of 50mL containing 10 mmol Tris-HCl, 1.5 mmolMgCl2, 50 mmol KCl (pH 8.3), and 100 ng of 50-30 end primers.
The oligonucleotide primer sequences were Prok1 sense(50CGC GAG TCT CAA TCA TGC TCC T-30) and antisense(50-GGC AAG GCG CTA AAA ATT GAT G-30), and Prok2sense (50-TTG GCC TGT TTA CGG ACT TC-30) and antisense(50-TGC AAGAGGAGGGAAGAGAA-30). Sequences usedfor glyceraldehyde-3-phosphate dehydrogenase are previ-ously reported (33).
PCR products were then separated on a 1.2% agarose gelcontaining ethidium bromide using a 100-bp DNA ladder(Life Technologies) as size marker.
Real-time quantitative RT-PCR
RNAs were reverse transcribed using 5mg total RNA aspreviously described (33). Real-time quantitative PCR wasused to determine the amounts of Prok1 and Prok2 mRNA inFB-2, PTC, euthyroid MNG, GD, and contralateral NT tissuesfrom PTC cases, as previously described (32). In these experi-ments, the amount of specific amplicon present was related tobeta2-microglobulin and subsequently to an internal control.Real-time PCR was repeated three times for each sample usingoligonucleotides Prok1 [50-CCACATGTATCCCTCGGTCT-30
(sense) and 50-ACCTGGGACTCTGAGCAATG- (antisense)];and Prok2 30 [50-CTTGCCTCTTCCACCTCAAA-30 (sense) and50-TGCAAGAGGAGGGAAGAGAA-30 (antisense)]. Real-time PCR was also repeated three times for a housekeepinggene, beta 2-microglobulin using the sense and antisenseprimers, 50-CCAGCAGAGAATGGAAAGTC-30, and 50-GATGCTGCTTACATGTCTCG-30, respectively. The iQ SYBRGreen Supermix kit (Bio-Rad Laboratories) was used in aniCycler iQ Real-Time PCR Detection System (Bio-Rad La-boratories). Data are expressed as the amount of specific PCRproducts from each gene after normalization based on thehousekeeping gene product beta2-microglobulin (whichshowed no significant difference).
Study population, clinic pathological data,and immunohistochemistry
Thirty-two patients affected by PTC were identified fromthe Italian University Hospital, ‘‘Ospedali Riuniti’’—Foggia,Italy. All PTC patients received surgical treatment only withcurative intention between 2003 and 2006. The histopatho-logical diagnosis, reports about histological variants, andstage identification of all PTCs were made and carefully re-viewed at the Section of Anatomic Pathology, Department ofSurgical Science, University of Foggia. Tumor extent, deter-mined from clinical records, was revised and classified ac-cording to the 2002 TNM classification (34). Demographicaland clinic pathological features of PTC patients are summa-rized in b T1Table 1.
Four-micrometer serial sections from formalin-fixed andparaffin-embedded blocks were cut and mounted on poly-L-lysine-coated glass slides. Immunostaining was performed bylinked streptavidin–biotin horseradish peroxidase technique.After sequential deparaffinization and rehydration, the slideswere treated with 0.3% H2O2 for 15 min to quench endoge-nous peroxidase. Antigen retrieval was performed by micro-wave heating a first time for 3 min at 650 W, and a second anda third time at 350 W of the slides immersed in 10 mM citratebuffer pH 6. After microwaving, the sections were blockedfor 60 min with 1.5% normal horse serum (Santa Cruz Bio-technology) diluted in phosphate-buffered saline buffer be-fore the reaction with primary antibody (Ab). Primarymonoclonal anti-EG-VEGF Ab (MAB1209; R&D Systems Inc.)was diluted 1:150 with 0.05 M Tris-HCl buffer pH 7.4 con-taining 1% bovine serum albumin and was incubated for 3 has previously described (32). The specificity of the anti-EG-VEGF Ab and its immunohistochemical utilization tech-nique has been previously described in the literature (32). Theproliferative index of neoplastic epithelial cells was identifiedin different sections of each sample using a mouse monoclonalanti-Ki67 Ab (mouse monoclonal anti-Ki67, clone MB67, andNovus Biologicals) (data not shown). After two washes withphosphate-buffered saline, the slides were treated with bio-tinylated species-specific secondary Ab and streptavidin–biotin enzyme reagent (DAKO), and the color was developedby 3,30-diaminobenzidine tetra hydrochloride chromogensolution. Sections were counterstained with Mayer’s hema-toxylin and mounted using the xylene-based mountingmedium. Negative control slides without primary Ab wereincluded for each staining. The results of the immuno-histochemical staining were evaluated separately by twoobservers.
Immunostained cells, analyzed at 40�with an opticalmicroscope (Olympus BX41), were counted in at least 10 high-power fields.
For each case, the cumulative percentage of positive cellsamong all sections examined was determined. Inter-rate re-liability between the two investigators blindly and indepen-dently examining the immunostained sections was assessed
by the Cohen’s K test, and yielded K values higher than 0.70 inalmost all instances.
Genomic DNA extraction
Paraffin-embedded tumor samples from 32 PTC patientsobserved at Foggia Hospital and 18 fresh PTC samples frompatients observed at Endocrine Units of the Second Universityof Naples were used to extract DNA for the exon 15 BRAFmutation analyses. Thus, the total number of PTC samplesanalyzed for BRAF mutation was from 50 patients. For nucleicacid extraction from paraffin-embedded tissues, 50 mm sec-tions were immersed in xylene for 30 min to remove paraffin,and washed in absolute then 70% ethanol. All samples weresubjected to digestion with 0.5% sodium dodecyl sulfate and0.5 mg/mL proteinase K at 378C overnight, and then ex-tracted with TRIZOL following manufacturer’s instructions.DNA concentrations were determined by absorbance at260 nm using a BioPhotometer (Eppendorf), and the resultinggenomic DNA (50–100 ng/sample) was used as a template.
FIG. 1. Prok1 and Prok2 mRNA expression inillustrative cases of the FBAU2 c -2 cell line, papillarythyroid carcinoma (PTC), Graves’ disease(GD), multinodular goiter (MNG), and normalcontralateral thyroid tissues (NT) from PTCcases. (A) Total RNA was reverse transcribedand amplified with GAPDH, Prok1 and Prok2primers, and the PCR products electro-phoresed on 1.2% agarose gels, and stainedwith ethidium bromide, as described in theMaterials and Methods section. Both Prok1andProk2 transcripts were detected in the repre-sentative thyroid tissues. Prok1 was expressedat high levels in the FB-2 cell line, PTC, andGD tissue, and at very low levels in NT andMNG. (B) Real-time PCR products of Prok1 and Prok2 were normalized to a housekeeping gene product (beta2-micro-globulin) in each experiment. Significantly higher levels of Prok1 were seen in FB-2 cells, PTC, and GD compared to MNGand NT. Data are the mean� standard deviation (n¼ 3; *p< 0.05). PCR, polymerase chain reaction; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.
Two primers (forward, 50-TCATAATGCTTGCTCTGATAGGA-30; reverse, 50-GGCCAAAATTTAATCAGTGGA-30)were used to amplify a 224 bp fragment of exon 15 of BRAFcontaining the site in which T1799A mutation occurs (35).PCR reactions were performed in 25mL of buffer containing1.5 mM MgCl2, 200 mM deoxynucleoside triphosphates, 50–100 ng genomic DNA, 0.5 mM of each primer, and 2.5 U Euro-Taq DNA polymerase (EuroClone). Thirty-five cycles withannealing temperatures optimized at 588C were used to ob-tain the PCR product. Amplification products were separatedon 1.2% agarose gel and observed by ethidium bromidestaining.
Mutant-allele-specific amplification analysis of BRAFmutation
Mutant-allele-specific amplification was used to identifyBRAF mutation as previously described (36). Two differentforward primers with substitution of a single base at the end ofthe primer (50-GTGATTTTGGTCTAGCTACAGT-30and 50-GTGATTTTGGTCTAGCTACAGA-30) were designed toamplify the wild-type allele or BRAF T1799A transversionmutation, respectively. The sequence of the reverse primerwas 50-GGCCAAAATTTAATCAGTGGA-30. PCRs were
performed at 948C for 2 min followed by 40 cycles of 948C for30 s, 528C for 45 s, and 728C for 45 s, and subsequent extensionat 728C for 8 min.
PCR products were analyzed in a 3% agarose gel stainedwith ethidium bromide. The presence of BRAF mutations wasdetermined by direct sequencing of the PCR products.
Statistical analysis
The data were analyzed by the Stanton Glantz statisticalsoftware (version 6, Mc Graw Hill, 2007) and GraphPad Prismsoftware version 4.00 for Windows (Graph Pad software,www.graphpad.com). Differences between the groups weredetermined using the one-way analysis of variance and theStudent-Newman-Keuls test. Only p-values <0.05 were con-sidered significant. Fisher exact test was used to test for anysignificant relation between two categorical variables. Forimmunohistochemistry, the cut-off value to segregate highversus low expressing cases was 50% of immunostained cellscalculated from at least 1000 tumor cells.
Results
Prok1 and Prok2 mRNA expression in thyroid tissues
We studied Prok1 and Prok2 mRNA expression by quanti-tative RT-PCR using thyroid gland tissue derived from 30 PTC
Table 2. Statistical Univariate Analysis of Prok1 IHC b AU1Expression and Associated
Clinico-Pathological Findings of 32 Papillary Thyroid Cancer
Variables n Mean SD SEMab AU5p value(Anovaa;
Student-Newman-Keuls b AU1b)
Prok1 in PTC versus Prok1 in normal thyroidProk1 in PTC 32 67.34 41.58 7.35 p¼ 0.000a; p< 0.05b
Prok1 in normal thyroid 13 0 0 0Prok1 according to age�40 years 8 85 32 11.34 p¼ 0.243a; p> 0.05b
>40 years 24 66 40.83 8.34Prok1 in PTC according to sex
Male 12 65.83 45.82 13.23 p¼ 0.849a; p> 0.05b
Female 20 68.06 38.39 9.05Prok1 in PTC according to surrounding thyroid
PTC from normal thyroid 13 63.46 41.51 11.5 p¼ 0.899a; p> 0.05b
PTC from Hashimoto’s thyroiditis 6 71.67 44.01 17.97PTC from goiter 13 70 42.43 11.77
Prok1 in PTC according to TNM stagingT1/T2 N0 M0 20 73.1 37.87 8.69 p¼ 0.315a; p> 0.05b
T3 N0 M0 4 77.5 45 22.5Any T with positive N 8 89.3 28.3 10.7M positive — — — —
Prok1 in PTC according to T sizeØ� 1.5cm 25 72.2 37.14 7.43 p¼ 0.426a; p> 0.05b
Ø> 1.5cm 7 29.5 47.2 23.6Prok1 in PTC according to histological PTC type
samples, 20 euthyroid MNG, and 10 GD tissues, with 30 con-tralateral NT tissues from PTC cases as controls; FB-2 thyroidcancer cell line samples were also analyzed (F1 c Fig. 1). We foundthat both Prok1 and Prok2 are present in NT and pathologicalthyroid tissues, as well as in the FB-2 cells (Fig. 1A). As shownin Figure 1B, Prok1 mRNA levels were differentially expressedin normal versus malignant tissues. Extremely low levels werefound in NT and MNG samples, whereas in PTC, FB-2, andGD, levels were significantly increased ( p< 0.05). In contrast,Prok2 transcripts were expressed at high levels both in controlas well as in pathological tissues (Fig. 1B).
Prok1 protein levels in NT and PTC thyroid tissues
We observed a strong upregulation of Prok1 protein inPTC. Twenty-nine of 32 PTC cases (90.6%) showed positiveimmunostaning for Prok1 (T2 c Table 2). All conventional typecases of PTC stained positively for Prok1 (F2 c Fig. 2), whereasPTC variants showed heterogeneity ranging from negative(75% of follicular variants) to strong immunostaining (one
sample of follicular variant shown in b F3Fig. 3, and one of diffusesclerosing variant). Statistical univariate analysis of Prok1levels stratified by clinico-pathological findings is summa-rized in Table 2. In Table 2, the growth patterns of analyzedtumors were classified as expansive in the presence of push-ing margins of invasion, and infiltrative in the presence ofneoplastic cells (or glands) having irregular and sharply de-fined borders that infiltrate the perilesional thyroidal tissue.Infiltrative neoplasias were valued cases with nodal metas-tases and tumors without lymph nodal metastases; no me-tastases were observed for the expansive PTCs. Theimmunohistochemical study (Fig. 2) showed that Prok1 wasalmost undetectable in NT tissues, whereas an overall higher
FIG. 2. Immunohistochemical analysis of Prok1 in NT fol-licles and in papillary thyroid carcinoma cells (classic vari-ant) at different pathological stages. Representative cases offormalin-fixed,
4C c
paraffin-embedded PTC tissues were im-mune-stained for Prok1 as described in the Materials andMethods section. Although specific cytoplasmic Prok1 ex-pression was not detected in NT follicles (A, right), robustexpression was detected in malignant PTC cells of micro-PTC(A), and in advanced PTC (B) with capsular invasion (B1).Linked streptavidin–biotin horseradish peroxidase techniquestaining is brown, and nuclear counterstaining with hema-toxylin is blue. Original magnification for (A) and (B1) is10�, and for (B) is 40�.
FIG. 3. Immunohistochemical analysis of Prok1 in a case offollicular variant of PTC from Has b 4Chimoto’s thyroiditis. (A)Neoplastic cells of a representative case of follicular PTCassociated to Hashimoto’s thyroiditis showed strong im-munostaining for Prok1. The tissue displayed the formationof lymphoid follicles with extensive lymphoplasmacytic in-filtrates and decreased colloid. (B) The characteristic strongcytoplasmic staining for Prok1 is observed at a higher mag-nification of PTC malignant follicles. Linked streptavidin–biotin horseradish peroxidase technique staining is brown,and nuclear counterstaining with hematoxylin is blue. Ori-ginal magnification for (A) is 4�, and for (B) is 63�.
degree of the protein compared to NT was seen in PTC( p< 0.05). We particularly found a trend in Prok1 proteinoverexpression relative to primary tumor growth pattern andlymph node involvement. The higher mean percentage ofProk1 expression was observed in PTCs with an infiltrativepattern of growth and lymph node metastases (Table 2). Also,a significant trend in Prok1 protein upregulation relative toTNM staging was observed, whereas no statistically signifi-cant correlations with sex, age, tumor maximum diameter,and associated pathology was noted. Mean levels of Prok1protein showed significant differences with respect to histo-logical type ( p< 0.05) (Table 2).
BRAF mutation in PTC samples
Paraffin-embedded tumor samples from 32 PTC patientsobserved at Foggia Hospital and 18 PTC samples from patientsobserved at Endocrine Units of the Second University of Na-ples were used to extract DNA for the exon 15 BRAF mutationanalyses. The total number of PTC samples analyzed for BRAFmutation was from 50 patients. The V600E mutation wasfound in 24 of 50 (48%) tumors, 4 examples of which are showninF4 c Figure 4. Correlation between BRAF mutation and Prok1expression was examined in 32 PTC cases described in Tables 1and 2. BRAF was mutated in 15 of 32 (47%) classical PTC, andin 1 of 4 (25%) follicular variant PTC. With respect to invasionpattern, we found V600E mutation in all seven samples (100%)of PTC infiltrative with metastatic lymph node involvementand in 11 of 16 (68%) infiltrative with negative lymph nodeinvolvement. Thus, the BRAF mutation showed a positivetrend of association with extra-thyroidal invasion.
Prok1 was significantly upregulated in BRAF V600E (þ)PTC, compared with BRAF V600E (�) PTC; expression ofProk1 was found in 60% of the PTC samples positive for theBRAF mutation ( p< 0.01). These data therefore showed asignificant trend in both Prok1 protein upregulation andBRAF mutation in tumors having an infiltrative pattern andlymph node involvement.
Discussion
Our data indicate that Prok1 and Prok2 are expressed inthyroid tissues, and show that Prok1 transcript is differentiallyexpressed in NT and pathological thyroid tissues. Besides thewell-known role for angiogenesis in cancer, it has becomeclear that it is also an integral component of a diverse range ofnon-neoplastic chronic inflammatory and autoimmune dis-eases, including GD (37). We find Prok1 mRNA highly ex-pressed in GD tissues, whereas it has recently been reportedthat VEGF is expressed at very low levels in these tissues (38).Thus, Prok1 that is specific of endothelial cells derived fromEGs could likely contribute to pathogenesis of structuralchanges, activation and proliferation of endothelial cells, aswell as to capillary and venule remodeling in GD (27,28). Wealso observe a significant increase of Prok1 levels in PTC dis-playing an infiltrative pattern. While Prok2 expression levelsare higher compared to Prok1, no significant differences areseen between NT and malignant thyroid tissues, suggesting arole for Prok1 in thyroid angiogenesis and tumor progression.Moreover, we observe a consistent relationship betweenBRAF mutation and Prok1 protein expression in PTC. Proki-neticins and their receptors are involved in a wide spectrum of
biological functions and pathologies of various tissue pa-thologies (39). The complexity and potential redundancy ofthis system remain unclear, and therefore progress of under-standing pathological roles for prokineticins and the devel-opment of new target-specific therapies is hindered. Thestrong angiogenic effects of prokineticins are known, andtheir aberrant signaling, which may cause hyperplasia andhyper-vascularity in various tissues, has been highly associ-ated with the development of polycystic ovarian syndrome,neuroblastoma, and testicular cancer (39,40). For example,Prok1 overexpression in a colorectal cancer line induces an-giogenesis and tumors when implanted into nude mice (41).In ovary carcinoma, Prok1 expression is detected in the earlystage and is reduced in advanced-stage of the disease (42).Our previous studies have demonstrated that prokineticinsand their receptors are expressed in human prostate, and thattheir levels increase with prostate malignancy (32). Differentstudies have also shown an increased expression of VEGF inPTC (26). VEGF is upregulated in BRAF V600E (þ) PTC, andthis may provoke an increase in tumor growth and vascula-ture. In differentiated thyroid cancer studies, preliminary trialresults using inhibitors of angiogenesis such as sorafenib,motesanib, axitinib, and vandetanib have shown promisingeffects (43). In particular, the immunoneutralizing antibodiesor antagonists of prokineticins may be useful tools for newanti-cancer treatments (44).
In the last five years, more than 200 publications have de-scribed the relationship between BRAF V600E and thyroidcancer (45,46). In PTC, the mutation is associated with risk ofrecurrence or decreased recurrence-free survival (47), and so-matic point mutations in the BRAF oncogene gene have beenidentified as the most common genetic event in PTC (*44% ofPTC cases) (12,48–51). Indeed, we also find BRAF mutation in
FIG. 4. Analysis of BRAF mutation in representative casesof PTC by mutant-allele-specific amplification. GenomicDNA was extracted from micro-dissected sections of repre-sentative PTC samples (PTC1-4), and analyzed for BRAFmutation by mutant-allele-specific amplification PCR usingmutant or wild-type-specific primers as described in theMaterials and Methods section. The 129-bp product wasobserved for all four samples using both primers. Lane (M) isthe base pair ladder.
47% of the PTC cases studied, but in particular, BRAF V600E isdetected in 84% of PTC with infiltrative pattern, and in 100% ofcases with infiltrative pattern and metastatic lymph nodes atdiagnosis. Since BRAF mutation is found in PTC samplesdisplaying significantly higher Prok1 protein expression lev-els, this relationship strongly suggests that Prok1 plays a rolein PTC tumor recurrence, and that it could be useful as amarker for thyroid cancer progression. Although Prok1 couldalso be a potential target for new therapeutic strategies intreating PTC cases with worse clinical outcomes, furtherstudies will be required to define the exact molecular signalingmechanisms at various stages of tumor progression.
Acknowledgments
The authors express their gratitude to the following grantagency: Progetti di Ricerca di Interesse Nazionale (PRIN) (toD.P.).
Disclosure Statement
The authors declare that no competing financial interestsexist.
References
1. Figge JJ 2006 Epidemiology of thyroid cancer. In: ThyroidCancer: A Comprehensive Guide to Clinical Management.Second edition. Humana press, New Jersey, pp 9–14AU3 c .
2. Cooper DS, Doherty GM, Haugen BR, Kloos RT, Lee SL,Mandel SJ, Ernest L, Mazzaferri EL, McIver B, Sherman SI,Tuttle RM 2006 Management guidelines for patients withthyroid nodules and differentiated thyroid cancer: theAmerican thyroid association guidelines taskforce. Thyroid12:109–142.
3. Enewold L, Zhu K, Elaine Ron E, Marrogi AJ, AlexanderStojadinovic A, George E, Peoples GE, Susan S, Devesa SS2009 Rising thyroid cancer incidence in the united states bydemographic and tumor characteristics, 1980–2005. CancerEpidemiol Biomarkers Prev 18:784–791.
4. Hay ID, Bergstralh EJ, Goellner JR, Ebersold JR, Grant CS1993 Predicting outcome in papillary thyroid carcinoma:development of a reliable prognostic scoring system in acohort of 1779 patients surgically treated at one institutionduring 1940 through 1989. Surgery 114:1050–1105.
5. Mazzaferri EL, Jhiang SM 1994 Long-term impact of initialsurgical and medical therapy on papillary and follicularthyroid cancer. Am J Med 97:418–428.
6. Fonseca E, Soares P, Rossi S, Sobrinho-Simoes M 1997Prognostic factors in thyroid carcinomas. Verh dtsch gespathol 1:82–96.
7. Gilliland FD, Hunt WC, Morris DM, Key CR 1997 Prog-nostic factors for thyroid carcinoma. A population-basedstudy of 15,698 cases from the Surveillance, Epidemiologyand End Results (SEER) program 1973–1991. Cancer 79:
564–573.8. LiVolsi VA, Fadda G, Baloch ZW 2000 Prognostic factors in
well-differentiated thyroid cancer. Rays 25:163–175.9. Sherman SI, Brierley JD, Sperling M, Ain KB, Bigos ST,
Cooper DS, Haugen BR, Ho M, Klein I, Ladenson PW,Robbins J, Ross DS, Specker B, Taylor T, Maxon HR 1998Prospective multicenter study of thyroid carcinoma treat-ment: initial analysis of staging and outcome. National
thyroid cancer treatment cooperative study registry group.Cancer 83:1012–1021.
10. Mitsutake N, Knauf JA, Mitsutake S, Mesa C Jr., Zhang L,Fagin JA 2005 Conditional BRAFV600E expression inducesDNA synthesis, apoptosis, dedifferentiation, and chromo-somal instability in thyroid PCCL3 cells. Cancer Res65:2465–2473.
11. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S,Teague J, Woffendin H, Garnett MJ, Bottomley W, Davis N,Dicks E, Ewing R, Floyd Y, Gray K, Hall S, Hawes R,Hughes J, Kosmidou V, Menzies A, Mould C, Parker A,Stevens C, Watt S, Hooper S, Wilson R, Jayatilake H, Gus-terson BA, Cooper C, Shipley J, Hargrave D, Pritchard-JonesK, Maitland N, Chenevix-Trench G, Riggins GJ, Bigner DD,Palmieri G, Cossu A, Flanagan A, Nicholson A, Ho JW,Leung SY, Yuen ST, Weber BL, Seigler HF, Darrow TL, Pa-terson H, Marais R, Marshall CJ, Wooster R, Stratton MR,Futreal PA 2002 Mutations of the BRAF gene in humancancer. Nature 417:949–954.
12. Xing M 2005 BRAF mutation in thyroid cancer. Endocr RelatCancer 12:245–262.
13. Namba H, Nakashima M, Hayashi T, Hayashida N, MaedaS, Rogounovitch TI, Ohtsuru A, Saenko VA, Kanematsu T,Yamashita S 2003 Clinical implication of hot spot BRAFmutation, V599E, in papillary thyroid cancers. J Clin En-docrinol Metab 88:4393–4397.
14. Nikiforova MN, Kimura ET, Gandhi M, Biddinger PW,Knauf JA, Basolo F, Zhu Z, Giannini R, Salvatore G, FuscoA, Santoro M, Fagin JA, Nikiforov YE 2003 BRAF muta-tions in thyroid tumors are restricted to papillary carci-nomas and anaplastic or poorly differentiated carcinomasarising from papillary carcinomas. J Clin Endocrinol Me-tab 88:5399–5404.
15. Xing M, Westra WH, Tufano RP, Cohen Y, Rosenbaum E,Rhoden KJ, Carson KA, Vasko V, Larin A, Tallini G, TolaneyS, Holt EH, Hui P, Umbricht CB, Basaria S, Ewertz M,Tufaro AP, Califano JA, Ringel MD, Zeiger MA, SidranskyD, Ladenson PW 2005 BRAF mutation predicts a poorerclinical prognosis for papillary thyroid cancer. J Clin En-docrinol Metab 90:6373–6379.
16. Riesco-Eizaguirre G, Gutierrez-Martinez P, Garcia-CabezasMA, Nistal M, Santisteban P 2006 The oncogene BRAFV600E is associated with a high risk of recurrence and lessdifferentiated papillary thyroid carcinoma due to the im-pairment of Naþ/I� targeting to the membrane. EndocrRelat Cancer 13:257–269.
17. Fugazzola L, Puxeddu E, Avenia N, Romei C, Cirello V,Cavaliere A, Faviana P, Mannavola D, Moretti S, Rossi S,Sculli M, Bottici V, Beck-Peccoz P, Pacini F, Pinchera A,Santeusanio F, Elisei R 2006 Correlation between B-RAFV600E mutation and clinico-pathologic parameters in pap-illary thyroid carcinoma: data from a multicentric Italianstudy and review of the literature. Endocr Relat Cancer13:455–464.
18. Kim TY, Kim WB, Song JY, Rhee YS, Gong G, Cho YM, KimSY, Kim SC, Hong SJ, Shong YK 2005 The BRAF mutation isnot associated with poor prognostic factors in Korean pa-tients with conventional papillary thyroid micro carcinoma.Clin Endocrinol 63:588–593.
19. Fugazzola L, Mannavola D, Cirello V, Vannucchi G, MuzzaM, Vicentini L, Beck-Peccoz P 2004 BRAF mutations in anItalian cohort of thyroid cancers. Clin Endocrinol 61:239–243.
20. Folkman J, Ingber KWD, Hanahan D 1989 Induction of an-giogenesis during the transition from hyperplasia to neo-plasia. Nature 339:58–61.
21. Yancopoulos GD, Davis S, Gale NW, Rudge JS, Wiegand SJ,Holash J 2000 Vascular-specific growth factors and bloodvessel formation. Nature 407:242–248.
22. Maeda K, Chung YS, Ogawa Y, Takatsuka S, Kang SM,Ogawa M, Sawada T, Sowa M 1996 Prognostic value ofvascular endothelial growth factor expression in gastriccarcinoma. Cancer 77:858–863.
23. Esser S, Lampugnani MG, Corada M, Dejana E, Risau W1998 Vascular endothelial growth factor induces VE-cadherin tyrosine phosphorylation in endothelial cells. J CellSci 111:1853–1865.
24. Akagi K, Ikeda Y, Miyazaki M, Abe T, Kinoshita J, MaeharaY, Sugimachi K 2000 Vascular endothelial growth factor-C(VEGF-C) expression in human colorectal cancer tissues. Br JCancer 83:87–891.
25. Ueda M, Terai Y, Yamashita Y, Kumagai K, Ueki K, Yama-guchi H, Akise D, Hung YC, Ueki M 2002 Correlation be-tween vascular endothelial growth factor-C expression andinvasion phenotype in cervical carcinomas. Int J Cancer98:335–343.
26. Jo YS, Li S, Song JH, Kwon KH, Lee JC, Rha SY, Lee HJ, SulJY, Kweon GR, Ro HK, Kim JM, Shong M 2006 Influence ofthe BRAF V600E mutation on expression of vascular endo-thelial growth factor in papillary thyroid cancer. J Clin En-docrinol Metab 91:3667–3670.
27. Schweitz H, Pacaud P, Diochot S, Moinier D, Lazdunski M1999 MIT (1), a black mamba toxin with a new and highlypotent activity on intestinal contraction. FEBS Letters461:183–188.
28. LeCouter J, Kowalski J, Foster J, Hass P, Zhang Z, Dillard-Telm L, Frantz G, Rangell L, DeGuzman L, Keller GA, PealeF, Gurney A, Hillan KJ, Ferrara N 2001 Identification of anangiogenic mitogen selective for endocrine gland endothe-lium. Nature 412:877–884.
29. Lin DC, Bullock CM, Eheler FJ, Chen JL, Tian H, Zhou QY2002 Identification and molecular characterization of twoclosely related G protein-coupled receptors activated byprokineticins/endocrine gland vascular endothelial growthfactor. J Biol Chem 277:19276–19280.
30. Masuda Y, Takatsu Y, Terao Y, Kumano S, Ishibashi Y,Suenaga M, Abe M, Fukusumi S, Watanabe T, ShintaniY, Yamada T, Hinuma S, Inatomi N, Ohtaki T, Onda H,Fujino M 2002 Isolation and identification of EG-VEGF/prokineticins as cognate ligands for two orphan G-protein-coupled receptors. Biochem Biophys Res Commun293:396–402.
31. Soga T, Matsumoto S, Oda T, Saito T, Hijama H, Takasaki J,Kamohara M, Ohishi T, Matsushime H, Furuichi K 2002Molecular cloning and characterization of prokineticin re-ceptors. Biochem Biophys Acta 1579:173–179.
32. Pasquali D, Rossi V, Staibano S, De Rosa G, Chieffi P, Pre-zioso D, Mirone V, Mascolo M, Tramontano D, Bellastella A,Sinisi AA 2006 The endocrine-gland-derived vascular en-dothelial growth factor (EG-VEGF)/prokineticin 1 and 2 andreceptor expression in human prostate: up-regulation of EG-VEGF/prokineticin 1 with malignancy. Endocrinology147:4245–4251.
33. Pasquali D, Rossi V, Prezioso D, Gentile V, Colantuoni V,Lotti T, Bellastella A, Sinisi AA 1999 Changes in tissuetransglutaminase activity and expression during retinoic
acid-induced growth arrest and apoptosis in primary cul-tures of human epithelial prostate cells. J Clin EndocrinolMetab 84:1463–1469.
34. Sobin LH, Wittekind C 2002 TNM: classification of malig-nant tumours. In: Greene FL (ed) Manual for Staging Cancer,Manual. Springer, American Joint Committee on Cancer.New York b AU4.
35. Quiros RM, Ding HG, Gattuso P, Prinz RA, Xu X 2005Evidence that one subset of anaplastic thyroid carcinomasare derived from papillary carcinomas due to braf and p53mutations. Cancer 103:2261–2268.
37. Folkman J 1995 Angiogenesis in cancer, vascular, rheuma-toid and other disease. Nat Med 1:27–31.
38. Itoh A, Iwase A, Jimbo S 2010 Expression of vascular en-dothelial growth factor and presence of angiovascular cellsin tissues from different thyroid disorders. World J Surg34:242–248.
39. Ngan ESW, Tam PKH 2008 Prokineticin-signaling pathway.Int J Biochem Cell Biol 40:1679–1684.
40. Maldonado-Perez D, Evans J, Denison F, Millar RP, JabbourHN 2007 Potential roles of the prokineticins in reproduction.Trends Endocrinol Metab 18:66–72.
41. Goi T, Fujioka M, Satoh Y, Tabata S, Koneri K, Nagano H,Hirono Y, Katayama K, Hirose K, Yamaguchi A 2004 An-giogenesis and tumor proliferation metastasis of humancolorectal cancer cell line sw620 transfected with endocrineglands-derived-vascular endothelial growth factor, as a newangiogenic factor. Cancer Res 64:1906–1910.
42. Zhang L, Yang N, Conejo-Garcia JR, Katsaros D, Mohamed-Hadley A, Fracchioli S, Schlienger K, Toll A, Levine B, RubinSC, Coukos G 2003 Expression of endocrine gland-derivedvascular endothelial growth factor in ovarian carcinoma.Clin Cancer Res 9:264–272.
43. Deshpande HA, Gettinger SN, Sosa JA 2008 Novel chemo-therapy options for advanced thyroid tumors: small mole-cules offer great hope. Curr Opin Oncol 20:19–24.
44. Sherman SI 2009 Advances in chemotherapy of differenti-ated epithelial and medullary thyroid cancers. J Clin En-docrinol Metab 94:1493–1499.
45. Nucera C, Goldfarb M, Hodin R, Parangi S 2009 Role ofB-Raf (V600E) in differentiated thyroid cancer and preclini-cal validation of compounds against B-Raf (V600E). BiochemBiophys Acta 1795:152–161.
46. Elisei R, Ugolini C, Viola D, Lupi C, Biagini A, Giannini R,Romei C, Miccoli P, Pinchera A, Basolo F 2008 BRAF(V600E)mutation and outcome of patients with papillary thyroidcarcinoma: a 15-year median follow-up study. J Clin En-docrinol Metab 93:3943–3949.
47. Xing M 2007 BRAF mutation in papillary thyroid cancer:pathogenic role, molecular bases, and clinical implications.Endocr Rev 28:742–762.
48. Fugazzola L, Puxeddu E, Avenia N, Romei C, Cirello V,Cavaliere A, Faviana P, Mannavola D, Moretti S, Rossi S,Sculli M, Bottici V, Beck-Peccoz P, Pacini F, Pinchera A,Santeusanio F, Elisei R 2006 Correlation betweenB-RAFV600E mutation and clinico-pathologic parameters inpapillary thyroid carcinoma: data from a multicentric Italianstudy and review of the literature. Endocr Relat Cancer13:455–464.
AU1: Please expand FB, IHC, and SEM.AU2: Please define PK, FB, and RT.AU3: Please provide editor names for Ref. 1.AU4: Please provide page range for Ref. 34.AU5: Please provide footnotes for ‘‘a’’ and ‘‘b’’ cited in Table 2.
