DOI:10.1093/jnci/dju182

JNCI | Brief Communication Page 1 of 7jnci.oxfordjournals.org

© The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

Role of BRAFV600E in the First Preclinical Model of Multifocal Infiltrating Myopericytoma Development and MicroenvironmentPeter M. Sadow*, Carmen Priolo*, Simona Nanni, Florian A. Karreth, Mark Duquette, Roberta Martinelli, Amjad Husain, John Clohessy, Heinz Kutzner, Thomas Mentzel, Christopher V. Carman, Antonella Farsetti, Elizabeth Petri Henske, Emanuele Palescandolo, Laura E. Macconaill, Seum Chung, Guido Fadda, Celestino Pio Lombardi, Antonina M. De Angelis, Oreste Durante, John A. Parker, Alfredo Pontecorvi, Harold F. Dvorak, Christopher Fletcher, Pier Paolo Pandolfi, Jack Lawler, Carmelo Nucera

*Authors contributed equally to this work.

Manuscript received December 23, 2013; revised May 14, 2014; accepted May 20, 2014.

Correspondence to: Carmelo Nucera, MD, PhD, Harvard Medical School, Laboratory of Human Thyroid Cancers Preclinical and Translational Research, Division of Cancer Biology and Angiogenesis, Department of Pathology, Beth Israel Deaconess Medical Center, Harvard Medical School, 99 Brookline Avenue, Boston 02215, MA, USA. (e-mail: cnucera@bidmc.harvard.edu).

Myopericytoma (MPC) is a rare tumor with perivascular proliferation of pluripotent stem-cell–like pericytes. Although indolent, MPC may be locally aggressive with recurrent dis-ease. The pathogenesis and diagnostic biomarkers of MPC are poorly understood. We discovered that 15% of benign MPCs (thyroid, skin; 3 of 20 samples) harbored BRAFWT/V600E; 33.3% (1 of 3 samples) of BRAFWT/V600E-MPCs were multifocal/infiltrative/recurrent. Patient-MPC and primary MPC cells harbored BRAFWT/V600E, were clonal and expressed pericytic-differentiation biomarkers crucial for its microenvironment. BRAFWT/V600E-positive thyroid MPC primary cells triggered in vitro (8.8-fold increase) and in vivo (3.6-fold increase) angiogenesis. Anti-BRAFV600E therapy with vemurafenib disrupted angiogenic and metabolic properties (~3-fold decrease) with down-regulation (~2.2-fold decrease) of some extracellular-matrix (ECM) factors and ECM-associated long non-coding RNA (LincRNA) expression, with no effects in BRAFWT-pericytes. Vemurafenib also inhibited (~3-fold decrease) cell viability in vitro and in BRAFWT/V600E-positive thyroid MPC patient-derived xenograft (PDX) mice (n = 5 mice per group). We established the first BRAFWT/V600E-dependent thyroid MPC cell culture. Our findings identify BRAFWT/V600E as a novel genetic aberration in MPC pathogenesis and MPC-associated biomarkers and imply that anti-BRAFV600E agents may be useful adjuvant therapy in BRAFWT/V600E-MPC patients. Patients with BRAFWT/V600E-MPC should be closely followed because of the risk for multifocality/recurrence.

JNCI J Natl Cancer Inst (2014) 106(8): dju182

Myopericytoma (MPC) describes rare, nod-ular tumors characterized by a radial and multilayered perivascular growth of ovoid and spindled-shaped cells with mesenchy-mal stem-cell–like features, often with asso-ciated blood vessels arranged in an irregular, “staghorn” pattern (1–3). Most MPCs are benign (2), but some are malignant with metastatic potential and poor survival (4,5).

MPCs arise over a wide age range, primar-ily affecting subcutaneous tissues of the dis-tal extremities (2), and lesions occasionally present with skin ulceration, pain or tender-ness; additionally, symptomatic MPCs arise in the proximal extremities, head region and internal organs (1). Most MPC are treated surgically, and complete excision should prevent recurrent/persistent disease,

although negative surgical margins are diffi-cult to achieve outside of the extremities (2).

Biomarkers and oncogenic events driving MPC development are not well understood. In this study, we performed a comprehensive genomic and functional validation of associated MPC biomarkers with clinical implications. We have applied a high-throughput genotyping assay on 29 MPCs from available, formalin-fixed paraffin-embedded (FFPE), discarded/uni-dentified samples, using an Institutional Review Board-approved protocol (Beth Israel Deaconess Medical Center, Boston, MA) to unravel oncogenic events driving pathogenesis (Supplementary Methods and Supplementary Table  1, available online). The patient with thyroid MPC provided written informed consent for genetic analy-sis. For all other patients, we used discarded/unidentified tissue specimens and con-sent for genotyping test. Disease stage was assessed by radiologic imaging. We also used and immunohistochemistry, primary cell cultures, dynamic functional assays, shRNA, and developed an MPC-patient-derived xenograft (PDX) mouse model (for detailed methods, please see the Supplementary Methods, available online). All animal work was done in accordance with federal, local, and institutional guidelines at the Beth Israel Deaconess Medical Center (Boston, MA), and all experiments were performed with four-month-old Crl:NU(NCr)-Foxn1nu female, athymic, immunodeficient, nude mice (strain code: 490)  (Charles River, Wilmington, MA) (n = 5 per group).

Statistical analysis was carried out using GraphPad Prism 6 software (version Prism 6, GraphPad Software Inc., San Diego, CA). Mann-Whitney test was used to analyze the statistical significance of dif-ferences between two groups. For categori-cal data, Fisher’s exact test was used. All reported P values were two sided. Data are reported as the averaged value, and error bars represent the standard deviation of the average for each group in duplicate or trip-licate. Results with P values below .05 were considered statistically significant.

We developed the first translational model to date of multifocal and infiltrative thyroid MPC (Figure  1; Supplementary Figures 1 and 2, available online). The

BRIEF coMMunIcAtIon

at Beth Israel D

eaconess Medical C

enter on July 25, 2014http://jnci.oxfordjournals.org/

Dow

nloaded from

Page 2 of 7 Brief Communication | JNCI

Figure  1. A multimodality perioperative evaluation of patient with multifocal infiltrating thyroid myopericytoma harboring the heterozy-gous BRAFWT/V600E mutation. A1) Multinodular thyroid disease, 1.5 mm follicular variant papillary thyroid microcarcinoma (PTCFV) in the left thyroid lobe and a left thyroid mass (myopericytoma [MPC]) with medi-astinal extension in a 44-year old woman. A2) Neck Doppler ultrasound shows a 4.0 × 3.4 cm irregular, hypo-echoic, hyper-vascular mass in the left thyroid lobe with mediastinal extension. A3) Contrast enhanced T1-weighted magnetic resonance imaging before thyroid surgery shows a four-cm, hyper-intense mass (arrowhead) localized in the left thyroid lobe with mediastinal extension, impingement, rightward tracheal (asterisk) deviation and possible extension into soft tissue (arrow high-lights a right thyroid nodule). A4) Total body 18F-FDG (fluorodeoxyglu-cose) PET (Positron emission tomography)/CT (Computed Tomography) exam reveals high 18F-FDG uptake in the left persistent/recurrent mass (arrow). A5) Elastography analysis shows a non-deformable pattern (the arrow highlights a reduction of the elasticity, blue color) in the vast majority part of the left persistent/recurrent MPC. B1-B7) Histology of the thyroid and perithyroidal soft tissue with multifocal and infiltrating thyroid MPC (hematoxylin and eosin staining). (B1, scale bar = 500 µ) Macrofollicular thyroid tissue. (B2-B3, scale bar  =  200 µ) The thyroid MPC is present in several foci (asterisk) in the thyroid gland (B2-B3, arrowhead highlights thyroid follicles). (B4, scale bar = 500 µ) The MPC cells with apparent differentiation toward perivascular myoid cells and pericytes, best appreciated under oil immersion lens, show a bland, ovoid, and epithelioid to spindled morphology. (B5, scale bar = 200 µ) MPC shows a somewhat infiltrative appearance through adipose

tissue (arrowheads) with: (B6, scale bar  =  100 µ) radial and multilay-ered perivascular growth with a staghorn-like vascular growth pattern; and (B7, scale bar = 500 µ) destruction of thyroid tissue. (C1-C7, scale bar  =  100 µ) Immunohistochemistry staining in the thyroid MPC: C1) the Ki67 proliferation index (arrows highlight nuclear staining) is low in the MPC cells. C2) The MPC cells are clustered with a fascicular growth pattern around the vessels (clear, slit-like spaces, V) with intralesional, thin-walled branching vessels of small caliber, highlighted by CD31 (C2). C3-C4) MPC cells show cytoplasmic to membranous staining with antibodies against α-smooth muscle actin (αSMA) and platelet-derived growth factor receptor beta (PDGFRB). C5) MPC cells also show posi-tivity (punctate) for NG2 (chondroitin sulfate proteoglycan 4)  but are negative for desmin (C6). C7) Trichrome staining highlights the abun-dant amount of collagen deposition (blue staining, asterisk) surround-ing the MPC (purple staining) or the intratumoral deposition of collagen (arrows). D1-D2) DNA genotyping analysis by mass spectrometry (MS) traces of formalin-fixed paraffin embedded thyroid (n  =  1) (D1) and (D2) skin (eg, leg) (n = 28) MPC tissue samples reveals a heterozygous BRAFWT/V600E allele (long arrows) (A > T). The intensity of the signal vs mass of the analyte is plotted in the background, while the inset shows the cluster plot with all samples analyzed in the run (intensity of the sig-nals of the two alleles). The long arrow links the actual sample (circle in the inset) to the mass of the expected mutant V600E allele. Color coded by the software is automatic allele calls. Cyan squares for heterozygous samples, orange triangles for homozygous. Ambiguous calls are identi-fied by a red dot. Allele frequencies deviating from the expected values are assigned ambiguous or homozygous calls by the software.

initial pathologic diagnosis in October 2010, reported by the referring interna-tional institution, was undifferentiated

(anaplastic) thyroid carcinoma with two consequent cycles of chemotherapy given to the patient. By patient request, outside

review of the pathology was performed at the Massachusetts General Hospital (Boston) the following month, and the

at Beth Israel D

eaconess Medical C

enter on July 25, 2014http://jnci.oxfordjournals.org/

Dow

nloaded from

JNCI | Brief Communication Page 3 of 7jnci.oxfordjournals.org

diagnosis was amended to reflect a mul-tifocal and infiltrative left thyroid MPC (Figure 1; Supplementary Figures 1 and 2, available online), with no clinically suspi-cious lymph nodes or distant metastases. Completion thyroidectomy was performed in June 2011 (Supplementary Figure  1F, available online). Histologic evalua-tion confirmed residual thyroid MPC (Supplementary Figure  1, A-D, available online).

The MPC was about two-fold more metabolically active (standardized uptake value [SUV]  =  4.5) compared with the non-tumoral thyroid hyperplastic nodules (SUV  =  2.4) (Figure  1A4); but 18-FDG (Fluorodeoxyglucose) PET (positron emis-sion tomography)/CT (computed tomog-raphy) did not note any non-regional tissue involvement (Supplementary Figure  1D, available online). Twelve months post-operatively, imaging revealed persistent disease in the thyroid bed (Figure  1A5; Supplementary Figure  1, G-J, available online). MPC lacked features of malig-nancy (increased number of mitoses, necrosis, vascular invasion), showed appar-ent differentiation towards pericyte lineage (Figure  1, B2-B7) and was characterized by perivascular growth (Figure  1, B5 and B6, and 1C2). A  higher vascular density (CD31+) was found in the thyroid MPC (mean vessels/field  =  20.2, SD  =  0.4 ves-sels/field) compared with the adjacent normal thyroid (mean  =  5.2 vessels/field, SD  =  0.2, P  =  .007) (Figure  1C2). MPC cells are arranged circumferentially around smaller vessels (Figure  1, B5 and B6, and 1C2; Supplementary Figure 2B5, available online). The ratio of MPC cells to endothe-lial cells is about 3:1, quantified based on the number of platelet-derived growth factor receptor beta (PDGFRB)–positive MPC cells and CD31-positive endothelial cells (Supplementary Methods, available online). MPC biomarkers (alpha-smooth muscle actin [αSMA], PDGFRB, NG2 [neuron-glial antigen 2], and extracellular-matrix [ECM] molecules, eg, CollagenIA1 [COLIA1]), as well as desmin and p16/Ink4A immunoexpression are described in Figure 1, C3-C7; Supplementary Figure 2, B1-B19, and Supplementary Table 2, avail-able online.

As MPC is a poorly characterized “orphan” disease, we performed in-depth genotyping. We analyzed 29 FFPE tissues

(20 benign MPC, four benign intravas-cular MPC, and five malignant MPC) (Supplementary Table  1, available online) by Mass Spectrometry genomic technol-ogy, which interrogates about 1000 muta-tions in 112 validated oncogenes and tumor suppressors. Results were validated by per-forming Sequenom (6) (Figure  1, D1-D2; and Supplementary Figure 2, C1-C4, avail-able online), pyrosequencing and Sanger sequencing (data not shown). We found that 3/20 (15%) benign MPCs (ie, one thyroid and two cutaneous MPCs) from three different patients (Figure  1, D1-D2; Supplementary Figure  2, C1-C4, and Supplementary Table 1, available online) harbored the het-erozygous BRAFWT/V600E mutation in exon-15 hot-spot T1799A of the BRAF gene sequence. None of the four benign intra-vascular MPCs or five malignant MPCs harbored BRAFWT/V600E. Two out of three (66.6%) BRAFWT/V600E-MPCs (thyroid, skin) derived from two different patients’ infil-trated adjacent soft tissue; two of four (50%) multifocal MPCs harbored BRAFWT/V600E, compared with 1/25 (4%) unifocal MPCs with wild type (WT) BRAF (P = .04) (Supplementary Table  1, available online). One out of three (33.3%) MPCs (ie, thy-roid) with BRAFW/V600E showed recurrent/persistent disease after one year of follow-up (Figure  1A5;Supplementary Figure  1, G-J, available online). Furthermore, our analysis of X-chromosome inactivation and meth-ylation profile from the female patient with thyroid BRAFW/V600E-MPC (Figure 1D1 and Figure 2, A1 and A2) revealed that this tumor was monoclonal compared to the adjacent, uninvolved thyroid tissue (Figure 2A3) (7).

To provide a translational application for our study, we established early and late passages of primary cells cultured in vitro from human thyroid BRAFWT/V600E-MPC. These cells also harbored BRAFWT/V600E and expressed pericyte lineage-specific differentiation biomarkers (Figure  2A1). BRAFWT/V600E is the most frequently mutated oncogenic kinase. Vemurafenib is the first orally available selective inhibitor of BRAFV600E approved by the US Food and Drug Administration for the treatment of BRAFWT/V600E-melanoma (8–11). We tested the effects of vemurafenib on BRAFWT/V600E- MPC cells and BRAFWT pericytes (Figure 2, A-G). Vemurafenib substantially reduced phospho(p)ERK1/2 and pMEK1/2 pro-tein levels in BRAFWT/V600E-MPC cells as

compared with controls (Figure  2C). As a result, this treatment statistically signifi-cantly (P < .001) suppressed BRAFWT/V600E-MPC cell viability, with no effect on the growth of BRAFWT-pericytes (Figure  2B), suggesting its high specificity for the BRAFV600E vs BRAFWT. Furthermore, we hypothesized that BRAFWT/V600E plays a role in MPC angiogenic and metabolic proper-ties; we found that BRAFWT/V600E-MPC cells substantially grew as large cell aggregates on Matrigel (Supplementary Figure  3A, available online), and, when cocultured with human microvascular endothelial cells, stastistically significantly (8.8-fold, P = .002) triggered in vitro angiogenesis as compared with controls (Figure 2, D1-D2). Vemurafenib treatment statistically sig-nificantly (about 3-fold, P = .002) disrupted this effect (Figure  2, D1 and D2); addi-tionally, treatment with shRNA that tar-geted BRAFV600E statistically significantly reduced MPC cell adhesion and migration (Supplementary Figure 4C, available online) with no effect in BRAFWT-pericytes (data not shown). Remarkably, BRAFWT/V600E-MPC cells also statistically significantly (P = .002) increased vascular density in MPC-PDX mice (Supplementary Figure 5, available online), and vemurafenib therapy statistically significantly (3-fold, P  =  .007) suppressed both BRAFWT/V600E-MPC cell via-bility (Figure 2G; Supplementary Figure 5, available online) and vascular density/angio-genesis (3.6-fold, P = .002) (Supplementary Figure  5, available online) without any obvious toxicity. Subsequently, we found that COL1A1 (5-fold, P  =  .007), PDGFRB (1.3-fold increase, P  =  .02), integrin-β1 (ITGβ1) (1.7-fold increase, P  =  .02), ID2 (DNA-binding protein inhibitor) (6.5-fold increase, P = .002), and the long intergenic non-coding RNA (LincRNA) ID2 (2.8-fold increase, P  =  .02) were statistically signifi-cantly (with moderate or high copy num-ber) BRAFV600E-dependent in MPC cells as compared with control cells with BRAFWT (Figure  2F; Supplementary Table 3, available online). Their expression lev-els (ID2  =  2.2-fold; COL1A1  =  1.5-fold; PDGFRB  =  1.5-fold; ITGβ1  =  1.5-fold; and ID2 LincRNA  =  1.6) were statistically significantly reduced by vemurafenib treat-ment as compared with BRAFWT-pericytes (Figure 2F). Furthermore, MPC cell adhe-sion (P = .002, Supplementary Figure  3B, available online) and migration (P  =  .002,

at Beth Israel D

eaconess Medical C

enter on July 25, 2014http://jnci.oxfordjournals.org/

Dow

nloaded from

Page 4 of 7 Brief Communication | JNCI

MPC

Figure  2. In vitro and in vivo preclinical model of human primary thy-roid myopericytoma harboring the heterozygous BRAFWT/V600E muta-tion. A1) Immunocytochemistry of established non-immortalized primary human thyroid myopericytoma (MPC) cells with the heterozy-gous BRAFWT/V600E mutation. Phase image of MPC cells, scale bar = 100 µ. Immunohistochemistry staining (scale bar  =  400 µ) in the thyroid MPC cells show cytoplasmic to membranous staining with antibodies against α-smooth muscle actin (αSMA), platelet-derived growth factor receptor

beta (PDGFRB), and CD44 (stem cell and cell adhesion marker). Thyroid MPC cells also show positivity for NG2 (chondroitin sulfate proteoglycan 4), phospho(p)-ERK1/2, and vimentin. CD31, CD45, pan-keratin, desmin, and thyroglobulin (Tg) immunostains are negative. A2) DNA genotyping analysis by mass spectrometry (MS) traces of the human primary thyroid MPC cells (n = 1) reveals a heterozygous BRAFWT/V600E allele (long arrows) (A > T). The intensity of the signal vs mass of the analyte is plotted in the back-ground, while the inset shows the cluster plot with all samples analyzed

at Beth Israel D

eaconess Medical C

enter on July 25, 2014http://jnci.oxfordjournals.org/

Dow

nloaded from

JNCI | Brief Communication Page 5 of 7jnci.oxfordjournals.org

Figure  2E) were also statistically signifi-cantly reduced by vemurafenib treatment compared with BRAFWT-pericytes. MPCs continue to be under recognized, result-ing in inappropriate treatment and patient anxiety (12). Benign MPCs are generally indolent tumors (1); however, when they involve internal organs, they may be locally aggressive, with recurrent/persistent disease because of difficulty of complete excision. We discovered that BRAFV600E is a distinct genetic alteration seen in benign MPCs, often multifocal and infiltrative. Patients with multifocal MPCs present a treatment dilemma, often requiring multiple surgeries for lesions that may be quite painful, even if not malignant. BRAFV600E is a prognostic biomarker of tumor recurrence and aggres-siveness (13–24) and also facilitates tumori-genesis (25–29).

BRAFWT/V600E-induced MPC cell adhe-sion, migration, and angiogenesis associated with upregulation of molecules (eg, COL1A1, PDGFRB, ID2) (30–33) crucial for ECM

remodeling, angiogenesis, and for auto-crine and paracrine communication in the tumor microenvironment (Supplementary Figure 6, available online), which ultimately may lead to MPC aggressiveness. High doses of vemurafenib therapy were effective to inhibit BRAFV600E-MPC cell viability and angiogenesis, suggesting that this therapy blocked BRAFV600E-dependent pro-migra-tory pathways and so diminished pro-angio-genic capabilities of MPC cells.

Please see the Supplementary Results (available online) for additional findings that may be of interest.

Collectively, using anti-BRAFV600E therapy as a surgical adjuvant may pro-vide a novel advancement in the thera-peutic strategy and treatment of locally aggressive multifocal BRAFV600E-positive MPCs or possibly serve as a therapeutic alternative for cases in which surgical options are limited by location and extent of disease, or in medically poor surgical candidates.

Our study is limited by sample size (29 available MPCs), precluding optimal evalu-ation of MPC pathological features with/without BRAFV600E. However, our results were substantiated by our integrated in vitro and in vivo approaches. It is also possible that MPC heterogeneity reduced the sensitivity for detection of the BRAFV600E mutation in some cases; therefore, we cannot exclude the possibility that the percentage of BRAFV600E-positive MPC cases is higher.

In conclusion, our results demonstrate a subset of MPC harbor BRAFV600E that drives tumor development. We report the first MPC arising in the thyroid. Anti-BRAFV600E therapy effectively suppresses viability in the only currently available MPC short-term cell culture that harbors the BRAFV600E mutation. BRAFV600E plays a role in the MPC microenvironment (Supplementary Figure 6, available online) and might ultimately lead to aggressive behavior. Finally, we report a multiplex panel of diagnostic markers for MPCs

in the run (intensity of the signals of the two alleles). The long arrow links the actual sample (circle in the inset) to the mass of the expected mutant V600E allele. Color coded by the software is automatic allele calls. Cyan squares for heterozygous samples, orange triangles for homozygous. Ambiguous calls are identified by a red dot. Allele frequencies deviating from the expected values are assigned ambiguous or homozygous calls by the software. These findings were validated by two independent experi-ments. A3) Methylation-sensitive PCR analysis of the X-chromosome locus was assessed on DNA isolated from manually dissected female thy-roid MPC (n = 1) or non-malignant adjacent thyroid tissue (n = 1) formalin fixed paraffin embedded specimen. Screenshot of electropherogram dis-playing discordant patterns of X-chromosome inactivation (arrowheads) in the thyroid MPC vs the benign thyroid tissue. For each sample, the cor-responding plot is a quantitative representation of the size and amount of fluorescent PCR products amplified from undigested DNA or digested DNA with the methylation-sensitive enzyme HpaII when analyzed on an automated DNA sequencer. The peaks indicate the estimated allele size (in base pairs) that is the corresponding peak height (amount of PCR prod-uct), as quantified by DNA Genotyper software. These findings were vali-dated by triplicate sample assays. B) Thyroid MPC cells with heterozygous BRAFWT/V600E or normal pericytes with BRAFWild Type(WT) were treated with the indicated concentrations of vemurafenib for 48 hours, and viability was determined using the Cell Titer-Glo ATP-based luminescence assay, with Dimethyl sulfoxide, control (DMSO)–treated cells as the control. Cell growth curves were determined following two to three days of treatment with DMSO or vemurafenib. Arrows highlight change of cell shape in MPC cells vs control. All scale bars are =10 µ (MPC cells images) and 50 µ (pericytes images). These data represent the average ± standard deviation (error bars) of eight independent replicate measurements (*P < .05, **P < .01, ***P < .001, Mann-Whitney test, two-sided). C) A parallel plate simi-lar to (B) was set up and corresponding phospho(p)ERK1/2 and pMEK1/2 protein levels measured from BRAFWT/V600E-MPC cells or BRAFWT pericytes protein samples using western blotting assays. These data are representa-tive of three independent experiments. D1-D2) BRAFWT/V600E-thyroid MPC cells cocultured with human microvascular endothelial cells (HMVECs) induce tubule formation (asterisk, control) (magnifications in all images: 20×; scale bars = 200 µ) on growth factors reduced Matrigel in the pres-ence of very low serum concentration (0.2%) within six hours compared to cocultures of HMVECs and BRAFWT pericytes or HMVECs alone, in the presence of vehicle (control) or vemurafenib treatment. Phase-contrast

images were captured by optical microscopy. GFP (green fluorescent protein) and merge images were captured on a different microscopy field by confocal microscopy analysis: HMVECs alone are stained with ICAM-1 (human intercellular adhesion molecule) (green); merge= GFP-tagged MPC cells or Alexa Fluo 488-conjugated CD90 (green)–tagged pericytes, and Alexa Fluo 594-conjugated actin (red)–tagged HMVECs. These data represent the average ± standard deviation (error bars) of six independ-ent replicate measurements (**P = .002, Mann-Whitney test, two-sided). E) Cell migration assays within five hours in human primary thyroid MPC cells with BRAFWT/V600E or in pericytes with BRAFWT in the presence of vehi-cle (control) or vemurafenib treatment. These data represent the average ± standard deviation (error bars) of six independent replicate measure-ments (**P  =  .002, Mann-Whitney test, two-sided). F) Quantitation of mRNA expression levels (copy number) by multi-gene transcriptional real-time reverse transcriptase PCR (RT-PCR) analysis of genes fundamen-tal for extracellular matrix (ECM) remodeling and angiogenesis in human primary thyroid MPC cells with BRAFWT/V600E or in pericytes with BRAFWT in the presence of vehicle (control) or vemurafenib treatment for 24 hours (0.2% serum cell growth medium). Gene expression was classified as “low copy number” if it was below 1 mRNA copy/106 18S copies, “moderate copy number” if it was between 1 and 15 mRNA copies/106 18S copies, and “high copy number” if it was greater than 15 mRNA copies/106 18S copies. These data represent the average ± standard deviation (error bars) of four to six independent replicate measurements (*P < .05; **P < .01: ***P < .001, Mann-Whitney test, two-sided). Genes showing difference in values of comparisons (P values <.05) were considered differentially expressed and statistically significant. G) A  patient-derived xenograft (PDX) mouse model of BRAFWT/V600E-human primary thyroid MPC cells engineered to express GFP (green fluorescent protein, scale bar=100 µ) injected in the subcutaneous compartment of the left ear of nude mice (n = 10) to visually follow perivascular growth of MPC cells over time in vivo. Mice were treated starting four days after BRAFWT/V600E-MPC cells implantation (defined ‘baseline’) with vemurafenib (100 mg/kg twice daily by oral gavage) (n = 5) or vehicle (n = 5) for two consecutive weeks. MPC cell viability was determined by the CS5.1 histogram software calculating the absolute number of pixels in the GFP-positive areas (arrows) of the left ear divided by the total number of pixels (247,545) in the highlighted selec-tion area. These data represent the average ± standard deviation of two independent experiments (control-treated mice vs vemurafenib-treated mice, P = .007, Mann-Whitney test).

at Beth Israel D

eaconess Medical C

enter on July 25, 2014http://jnci.oxfordjournals.org/

Dow

nloaded from

Page 6 of 7 Brief Communication | JNCI

(eg, PDGFRB, NG2, αSMA, peritumoral-fibronectin, vimentin, CollagenIA1).

We propose that genetic testing for the BRAFWT/V600E-mutation is part of the patho-logic evaluation for multifocal and infiltrative MPCs, and patients with BRAFWT/V600E- positive MPCs should be periodically re-evaluated for recurrence, especially with available targeted drug therapies.

References 1. Fletcher CDM, Bridge JA, Hogendoorn P,

Mertens F. WHO Classification of Tumours of Soft Tissue and Bone. Fourth Edition. IARC Press, Lyon, France. 2013;5.

2. Mentzel T, Dei Tos AP, Sapi Z, Kutzner H. Myopericytoma of skin and soft tissues: clin-icopathologic and immunohistochemical study of 54 cases. AJSP. 2006;30(1):104–113.

3. Diaz-Flores L, Gutierrez R, Garcia MP, Diaz-Flores L, Jr, Valladares F, Madrid JF. Ultrastructure of myopericytoma: a contin-uum of transitional phenotypes of myoperi-cytes. Ultrastruct Pathol. 2012;36(3):189–194.

4. Fisher C. Unusual myoid, perivascular, and postradiation lesions, with emphasis on atypi-cal vascular lesion, postradiation cutaneous angiosarcoma, myoepithelial tumors, myo-pericytoma, and perivascular epithelioid cell tumor. Sem Diagn Pathol. 2013;30(1):73–84.

5. McMenamin ME, Fletcher CD. Malignant myopericytoma: expanding the spectrum of tumours with myopericytic differentiation. Histopathol. 2002;41(5):450–460.

6. Matulonis UA, Hirsch M, Palescandolo E, et  al. High throughput interrogation of somatic mutations in high grade serous cancer of the ovary. PloS One. 2011;6(9):e24433.

7. Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM, Belmont JW. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene corre-lates with X chromosome inactivation. AJHG. 1992;51(6):1229–1239.

8. Bollag G, Hirth P, Tsai J, et  al. Clinical effi-cacy of a RAF inhibitor needs broad target blockade in BRAF-mutant melanoma. Nature. 2010;467(7315):596–599.

9. Sosman JA, Kim KB, Schuchter L, et  al. Survival in BRAF V600-mutant advanced melanoma treated with vemurafenib. N Engl J Med. 2012;366(8):707–714.

10. Bollag G, Tsai J, Zhang J, et al. Vemurafenib: the first drug approved for BRAF-mutant can-cer. Nat Rev Drug Discov. 2012;11(11):873–886.

11. Sondergaard JN, Nazarian R, Wang Q, et al. Differential sensitivity of melanoma cell lines with BRAFV600E mutation to the spe-cific Raf inhibitor PLX4032. J Transl Med. 2010;8:39.

12. Dray MS, McCarthy SW, Palmer AA, et  al. Myopericytoma: a unifying term for a spec-trum of tumours that show overlapping fea-tures with myofibroma. A review of 14 cases. J Clin Pathol. 2006;59(1):67–73.

13. Xing M, Alzahrani AS, Carson KA, et  al. Association between BRAF V600E mutation and mortality in patients with papillary thy-roid cancer. JAMA. 2013;309(14):1493–1501.

14. Nucera C, Nehs MA, Nagarkatti SS, et  al. Targeting BRAFV600E with PLX4720 dis-plays potent antimigratory and anti-invasive activity in preclinical models of human thy-roid cancer. Oncologist. 2011;16(3):296–309.

15. Nucera C, Porrello A, Antonello ZA, et  al. B-Raf(V600E) and thrombospondin-1 pro-mote thyroid cancer progression. Proc Natl Acad Sci USA. 2010;107(23):10649–10654.

16. Xing M. Molecular pathogenesis and mechanisms of thyroid cancer. Nat Rev. 2013;13(3):184–199.

17. Montero-Conde C, Ruiz-Llorente S, Dominguez JM, et al. Relief of feedback inhi-bition of HER3 transcription by RAF and MEK inhibitors attenuates their antitumor effects in BRAF-mutant thyroid carcinomas. Cancer Discov. 2013;3(5):520–533.

18. Tiacci E, Trifonov V, Schiavoni G, et al. BRAF mutations in hairy-cell leukemia. N Engl J Med. 2011;364(24):2305–2315.

19. Chakravarty D, Santos E, Ryder M, et  al. Small-molecule MAPK inhibitors restore radioiodine incorporation in mouse thyroid cancers with conditional BRAF activation. J Clin Invest. 2011;121(12):4700–4711.

20. Peyrade F, Re D, Ginet C, et  al. Low-dose vemurafenib induces complete remission in a case of hairy-cell leukemia with a V600E mutation. Haematologica. 2013;98(2):e20–22.

21. Rosove MH, Peddi PF, Glaspy JA. BRAF V600E inhibition in anaplastic thyroid cancer. N Engl J Med. 2013;368(7):684–685.

22. Nehs MA, Nucera C, Nagarkatti SS, et al. Late intervention with anti-BRAF(V600E) therapy induces tumor regression in an orthotopic mouse model of human anaplastic thyroid cancer. Endocrinology. 2012;153(2):985–994.

23. Nehs MA, Nagarkatti S, Nucera C, Hodin RA, Parangi S. Thyroidectomy with neoadju-vant PLX4720 extends survival and decreases tumor burden in an orthotopic mouse model of anaplastic thyroid cancer. Surgery. 2010;148(6):1154–1162; discussion 62.

24. Shaik S, Nucera C, Inuzuka H, et al. SCF(beta-TRCP) suppresses angiogenesis and thyroid cancer cell migration by promoting ubiquit-ination and destruction of VEGF receptor 2. J Exp Med. 2012;209(7):1289–1307.

25. Choueiri TK, Cheville J, Palescandolo E, et al. BRAF mutations in metanephric adenoma of the kidney. Eur Urol. 2012;62(5):917–922.

26. Rosty C, Buchanan DD, Walters RJ, et  al. Hyperplastic polyp of the duodenum: a report of 9 cases with immunohistochemi-cal and molecular findings. Hum Pathol. 2011;42(12):1953–1959.

27. Michaloglou C, Vredeveld LC, Mooi WJ, Peeper DS. BRAF(E600) in benign and malignant human tumours. Oncogene. 2008;27(7):877–895.

28. Badalian-Very G, Vergilio JA, Degar BA, et al. Recurrent BRAF mutations in Langerhans cell histiocytosis. Blood. 2010;116(11):1919–1923.

29. Brastianos PK, Taylor-Weiner A, Manley PE, et al. Exome sequencing identifies BRAF mutations in papillary craniopharyngiomas. Nat Genet. 2014;46(2):161–165.

30. Franco M, Roswall P, Cortez E, Hanahan D, Pietras K. Pericytes promote endothelial cell survival through induction of autocrine VEGF-A signaling and Bcl-w expression. Blood. 2011;118(10):2906–2917.

31. Jain RK, Booth MF. What brings peri-cytes to tumor vessels? J Clin Invest. 2003;112(8):1134–1136.

32. Whelan MC, Senger DR. Collagen I initiates endothelial cell morphogenesis by inducing actin polymerization through suppression of cyclic AMP and protein kinase A. JBC. 2003;278(1):327–334.

33. Lasorella A, Rothschild G, Yokota Y, Russell RG, Iavarone A. Id2 mediates tumor initiation, proliferation, and angiogenesis in Rb mutant mice. MCB. 2005;25(9):3563–3574.

FundingCarmelo Nucera (Principal Investigator, Human Thyroid Cancers Preclinical and Translational Research at the Beth Israel Deaconess Medical Center/Harvard Medical School) was awarded grants by the National Cancer Institute/National Institutes of Health (1R21CA165039-01A1 and 1R01CA181183-01A1) the American Thyroid Association (ATA), and ThyCa:Thyroid Cancer Survivors Association Inc. for Thyroid Cancer Research. Carmelo Nucera was also recipient of the Guido Berlucchi “Young Investigator” research award 2013 (Brescia, Italy). Beth Israel Deaconess Medical Center was recipient of a research grant by Roche (2011). CP was awarded a grant by the Tuberous Sclerosis Complex Research Program (TSCRP) of the USA Department of Defense (DOD) (W81XWH-13-1-0262).

notesThe study funders had no role in the design of the study, the collection, analysis, or interpretation of the data, the writing of the manuscript, nor the deci-sion to submit the manuscript for publication.

We thank Drs. Yutaka Kawakami (Keio University, Tokyo, Japan) for providing HIV-U6 vectors. We also thank Drs. Shiva Gautam (statisti-cal consultant), Alexander Gimelbrant, Janice Nagy, Gideon Bollag, Riccardo Taulli and Francesca Ianni, as well as Mrs. Sudeepa Syamala, Mrs. Mei Zheng, Miss Nina Hu, Mr. Bhavik Padmani and Neal Smith for technical assistance. We thank Professor William Aird for critical suggestions.

Authors’ contributions: Peter Sadow: per-formed pathologic diagnosis, reviewed immu-nohistochemical data, and edited manuscript. Carmen Priolo: edited manuscript, performed in vitro experiments, analyzed data, and designed experiments. Simona Nanni: performed in vitro experiments. Florian A.  Karreth: edited manu-script, performed in vitro experiments, and ana-lyzed data. Mark Duquette: performed in vitro experiments. Roberta Martinelli: performed in vitro experiments. Amjad Hussain: performed

at Beth Israel D

eaconess Medical C

enter on July 25, 2014http://jnci.oxfordjournals.org/

Dow

nloaded from

JNCI | Brief Communication Page 7 of 7jnci.oxfordjournals.org

in vivo experiments. John Clohessy: performed in vivo experiments. Heinz Kutzner: per-formed clinical-pathological diagnosis. Thomas Mentzel: performed clinical-pathological diag-nosis. Christopher Carman: analyzed data. Antonella Farsetti: analyzed data. Elizabeth Petri Henske: edited manuscript and analyzed data. Emanuele Palescandolo: analyzed data. Laura E. Macconaill: analyzed data. Seum Chung: per-formed surgical therapy. Guido Fadda: performed pathological analysis. Celestino Pio Lombardi: performed surgical therapy. Antonella De Angelis: analyzed data. Oreste Durante: analyzed data. John A. Parker: analyzed data, edited manuscript and reviewed radiologic data. Alfredo Pontecorvi: contributed with inpatient clinical exams. Harold Dvorak: edited manuscript and designed experi-ments. Christopher Fletcher: edited manuscript, performed pathologic diagnosis, and reviewed histo-pathological data. Pier Paolo Pandolfi: ana-lyzed data and edited manuscript. Jack Lawler: analyzed data and edited manuscript. Carmelo Nucera: wrote and edited manuscript, reviewed

clinical and experimental data, analyzed data, col-lected data, analyzed clinical exams, performed in vitro and in vivo experiments, and designed experiments.

Affiliations of authors: Department of Pathology, Massachusetts General Hospital (PMS) and Department of Medicine, Brigham and Women’s Hospital (CP, EPH), Harvard Medical School, Boston, MA; Unit of Endocrinology, Department of Medicine, A.  Gemelli, Catholic University, Roma, Italy (SN, AP); Division of Cancer Genetics, Department of Medicine (FAK, JC, PPP) and Laboratory of Human Thyroid Cancers Preclinical and Translational Research, Division of Cancer Biology and Angiogenesis, Department of Pathology, Center for Vascular Biology Research (MD, CN), Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA; Department of Medicine (RM, CVC) and Department of Pathology (AH, HD, JL), Center for Vascular Biology Research, Beth Israel Deaconess Medical

Center, Boston, MA; Dermatopathologische Gemeinschaftspraxis, Siemensstrasse, Friedri-chshafen, Germany (HK, TM); National Research Council (CNR-IBCN) and Department of Experimental Oncology, Regina Elena Cancer Institute, Rome, Italy (AF); Department of Medical Oncology, Dana Farber Cancer Institute, Harvard Medical School, Boston, MA (EP, LEM); MIT/Broad Institute, Cambridge, MA (EP, LEM); Department of Plastic and Reconstructive Surgery, National Health Insurance Service Ilsan Hospital, Ilsan, Korea (SC); Department of Pathology (GF) and Department of Surgery (CPL), A. Gemelli, Catholic University, Roma, Italy; Department of Experimental and Clinical Medicine (AMDA) and Unit of Radiotherapy (OD), University of Catanzaro, Italy; Department of Nuclear Medicine, Beth Israel Deaconess Medical Center (JAP); Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA (CF).

at Beth Israel D

eaconess Medical C

enter on July 25, 2014http://jnci.oxfordjournals.org/

Dow

nloaded from

1

Supplementary Materials

Supplementary Materials and Methods

Histology and Immunohistochemistry

Histopathology evaluation of 29 myopericytomas (MPCs) was performed by pathologists (C.F.

and P.S.) on hematoxylin and eosin (HE)-stained formalin-fixed paraffin-embedded (FFPE)

tissues. For all patients we used discarded and unidentified tissues (Beth Israel Deaconess

Medical Center (BIDMC) IRB-approved exemption 4 protocol, Boston). Multifocal MPC was

defined as >1 discrete lesion (the presence of more than one morphologically distinct lesions

within the same patient in the same general region) (1). All tissue specimens were fixed with

10% buffered formalin phosphate and embedded in paraffin blocks. These were visualized with

an Olympus BX41 microscope and the Olympus Q COLOR 5 photo camera (Olympus Corp.,

Lake Success, NY, USA). Sections (4 µm thick) of FFPE human MPC (Supplementary Table

1) classified according to the World Health Organization (WHO) (1) or MPC-PDX (MPC

patient-derived xenograft) mouse tissue were used for IHC (Supplementary Table 2). After

baking overnight at +37°C, deparaffinization with xylene/ethanol and rehydration were

performed. IHC analysis was performed using primary antibodies (Supplementary Table 2).

The sections, treated with pressure cooker for antigen retrieval (Biocare Medical, Concord, CA),

were incubated at 123°C in citrate buffer (Dako Target Retrieval Solution, S1699; DAKO

Corp.), cooled and washed with PBS. Antigen retrieval was performed for 60 min at room

temperature. The primary antibody was detected using a biotin-free secondary antibody (K4011)

(Dako Envision system) and incubated for 30 min. All incubations were carried out in a humid

chamber at room temperature. Slides were rinsed with PBS between incubations. Sections were

2

developed using 3,3-diaminobenzidine (Sigma Chemical Co.) as a substrate and were

counterstained with Mayer’s hematoxylin (2, 3). The IHC markers (including Ki67

immunoexpression) were assessed semiquantitatively using the following scoring method: 0

(negative), 1 (1–10% positive cells, low expression), 2 (11–50% positive cells, moderate), and 3

(more than 50% positive cells, high expression). Microvascular density is defined by number of

vessels per microscope field showing CD31 staining. The ratio between MPC cells and

endothelial cells in MPC tissues was quantified based on the number of platelet-derived growth

factor receptor beta (PDGFRB)-positive MPC cells and CD31-positive endothelial cells.

Trichrome staining was performed on sections (4 µm thick) of FFPE MPC according to the

manufactures instructions (Ventana, USA) in order to assess the amount of collagen fibers in the

extracellular matrix (ECM) of MPCs.

Genotyping OncoMap Analysis and Mass Extend Sequenom Validations

Genomic DNA was extracted using our previous protocol (4) from 29 available FFPE

discarded/unidentified samples from 29 patients with MPC (Supplementary Table 1) (Beth

Israel Deaconess Medical Center (BIDMC) Institutional Review Board (IRB)-approved

exemption 4 protocol, Boston). Briefly, 30 µm paraffin sections were lysed in denaturing buffer

containing proteinase K (1 mg/ml) (Invitrogen, USA) during overnight incubation at +55°C.

DNA was purified using equal volumes of a phenol:chloroform mixture (Invitrogen, USA) and

eluted in 30 µL of distilled water. Genomic DNA was quantified using Quant-iT PicoGreen

dsDNA Assay Kit (Invitrogen, USA) per manufacturer’s protocol. 250 ng of genomic DNA was

used for mutation detection through Oncomap version 3 (high-throughput genomic technology),

which interrogates about 1000 known mutations in 112 validated oncogenes and tumor

3

suppressors (5). All genomic analyses were run in the Center for Cancer Genome Discovery

(CCGD, Dana Farber Cancer Institute (DFCI), Harvard Medical School, Boston, MA, USA) (5).

Primers and probes were designed using Sequenom MassARRAY Assay Design 3.0 software,

applying default multi-base extension parameters. Whole genome amplification (WGA) was

performed using the GenomePlex Complete WGA kit (Sigma, USA) based on chemical

fragmentation followed by adapter mediated PCR amplification. Samples were run on the mass

spectrometry-based genotyping platform (Sequenom) and analyzed according to current

standardized protocols (5). Sample identity and the possible introduction of artifacts by WGA

were evaluated using a 48 Single Nucelotide Polymorphisms (SNPs) panel comparing the pre-

WGA to the post-WGA DNA. If ≥3 SNP discrepancies were identified between SNPs found in

pre- and post-WGA samples, this sample was discarded. Validations were performed using

homogeneous and sensitive Mass Extend (hME) sequenom, a multi-base extension chemistry

performed on native unamplified genomic DNA (5); calls are based on an expected allelic frequency

of 50% as for germline SNPs and allele frequencies deviating from the expected values are assigned

ambiguous or homozygous calls by the software developed at the CCGD. The patient with thyroid

MPC provided written informed consent for genetic analysis (“A. Gemelli” Medical School,

Catholic University, Roma, Italy). For all other patients we used discarded/unidentified tissue

specimens and consent for genotyping test.

Pyrosequencing

DNA was amplified using the PyroMark Q24 BRAF kit (Qiagen, Valencia,

CA). PCR products were sequenced for BRAFV600E with 5-CACTCCATCGAGATTTC-3 as a

sequencing primer and CTGCATGCATGCA as the dispensation order using the PyroMark MD

4

System (Biotage AB and Biosystems, Uppsala, Sweden). Samples with mutant allele frequencies

below 4% were considered wild type; those with frequencies of 4% or greater were

considered mutant.

Cell cultures

Primary human thyroid MPC cells were established according to Fischer et al. (6) and Nucera et

al. (7) with some modifications (i.e. treatment with 100 µg/mL G418 (Life Technologies, USA)

for ~2 days), from the excess of thyroid MPC tissue, non-goiterous, and non-autoimmune thyroid

fresh tissue from patient undergoing thyroidectomy for a left sided neck mass in the “A.

Gemelli” Medical School (consent approved to AP). An exemption 4 protocol was approved by

the Beth Israel Deaconess Medical Center committee (Boston, MA, USA) approved through IRB

to CN for discarded and unidentified tissues. Human microvascular endothelial (HMVECs) cells

were obtained from Lonza (USA) and pericytes were obtained from Promo Cell (Heidelberg,

Germany). MPC and pericytes were grown in 1:1 DMEM:HAM’s F12 (CellGro, USA)

supplemented with 10% fetal bovine serum (FBS) and penicillin/streptomycin/amphotericin

(CellGro, USA). HMVECs were grown in EGM2 bullet kit (Lonza, USA) growth medium with

10% FBS. All in vitro assays were performed growing these cells with the specific growth

medium supplemented with 0.2% FBS. The 8505c human anaplastic thyroid cancer (ATC) cell

line harboring the BRAFV600E mutation was purchased from DSMZ (German collection of

microorganisms and cell culture, Braunschweig, Germany) (8, 9) and primary normal thyroid

(NT) cells were established and used according to Nucera et al. (7).

5

Vemurafenib preparation

Vemurafenib (PLX4032) (Roche/Genentech (NYC, USA) was dissolved in dimethyl sulfoxide

(DMSO) (Sigma, USA) to achieve a stock concentration of 10 mM for in vitro assays. For in

vivo studies, a drug suspension was prepared from micro-precipitated bulk powder (MBP) by

suspending the drug to a concentration of 25 mg active pharmaceutical ingredient (API)/mL in a

2% solution of hydroxypropylcellulose (Sigma, USA), according to the manufactures

instructions. Freshly prepared drug suspensions were stored at 4°C and used within 48 hours.

Mice were dosed twice daily (8 hours apart) with vehicle (control) or with vemurafenib

suspensions at 100 mg/kg by oral gavage using an 18G oral gavage needle for the time periods.

Cell viability and growth

Primary thyroid MPC cells or pericytes (1×104 cells/well) were cultured in growth medium

containing 0.2% fetal bovine serum (FBS) (CellGro, USA) in a 96-well sterile culture plate

(Thermo Fisher Scientific, USA). Cells were treated with or without various concentrations of

vemurafenib for 48 hours. Cell viability was measured using the CellTiter-Glo luminescent cell

viability assay kit (Promega, USA). The IC50 (50% maximal inhibitory concentration) for cell

viability was determined using the following doses: 0.01 µM, 0.1 µM, 1 µM, and 5 µM. The

calculation of the 90% maximal inhibitory concentration (IC90) of ERK1/2 phosphorylation was

performed according to Bollag et al. (10) comparing 1 µM vs. 10 µM vemurafenib.

6

Glucose uptake assay

One thousands primary human MPC cells with BRAFV600E or human normal pericytes with

BRAFWT were seeded in 24-well plates and grown in the presence of DMEM with 5 mM glucose

and 10% fetal bovine serum for at least 48 hours. Cells were then serum-starved and treated with

vemurafenib or vehicle (DMSO) for 5 hours. 1 µCi/ml of 14C[U]glucose was added to the media

for the last 2 hours of treatment. Following incubation with the drug or vehicle and

14C[U]glucose, cells were washed twice with PBS (Phosphate Buffered Saline) and trypsinized.

Cell suspension was then mixed with Ready-Safe Liquid Scintillation Fluid (Beckman Coulter),

and 14C counts (count per minutes, CPM) were read in a scintillation counter. CPM were

normalized to total protein content (cpm/microgram).

ELISA

MPC cells or pericytes were cultured in 6-well dishes with 0.2% serum medium in the presence

of vemurafenib or DMSO for 24 hours. The next day, the medium was collected and secreted

VEGFA or FGF2 protein levels were determined by ELISA (Enzyme-linked immunosorbent)

assay kit (R&D Systems, USA) according to the manufacturer’s instructions.

Cell cycle analysis

MPC cells or pericytes were seeded at 90% confluence in a 6-well dish and grown in

DMEM/Ham’s F12 supplemented with 0.2% FBS, treated with 5 µM vemurafenib or vehicle for

48 hours. After 48 hours, cells were trypsinized and fixed in pre-chilled (−20 °C) ethanol 75%,

pelleted at 400 × g at room temperature, resuspended in 0.5% PBS/BSA, incubated in 2M HCl

0.5% BSA for 20 min at room temperature, washed with 0.5% PBS/BSA, and centrifuged 5

7

minutes at 400 × g at room temperature. The cells were resuspended in 0.1 M sodium borate (pH

8.5) for 2 min at room temperature, washed with 0.5% PBS/BSA, and centrifuged 5 min at 400 ×

g at room temperature. Finally, the cells were pelleted, washed twice with 0.5% PBS/BSA, and

resuspended in 500 μL of this solution. Propidium iodide (Sigma) was added to a final

concentration of 10 μg/mL with RNase 10 mg/mL (Sigma). Cells were incubated at RT for 30

min and then analyzed by flow cytometry on a FACSCalibur (Becton Dickinson

Immunocytometry Systems, USA). Proliferating cells were calculated as percentage of cell in S-

phase (DNA synthesis).

Apoptosis assay

MPC cells or pericytes were seeded at 90% confluence in a 6-well dish and grown in

DMEM/Ham’s F12 supplemented with 0.2% FBS, treated with 5 µM vemurafenib or vehicle for

48 hours. After 48 hours, MPC cells or pericytes treated with vemurafenib or vehicle were

collected and fixed at 4 °C with 75% ethanol for propidium iodide staining and flow cytometry

analysis on a FACS Calibur (Becton Dickinson Immunocytometry Systems, USA) to evaluate

sub-G1 cell populations.

Western Blotting

Western blotting assays were performed according to a standard procedure, and the lysis buffer,

composed of 10 mM Hepes (pH 7.40), 150 mM NaCl, 5 mM EDTA, 1 mM EGTA, 1 mM

sodium vanadate, 5 mM sodium flouride, and 1% Triton X-100; protease and phosphatase

inhibitors (Pierce) were used for protein extractions (2). We used the following antibodies:

BRAF (clone F7, cat#5284, Santa Cruz Biotechnology, USA) phospho-ERK1/2 (cat#9101, Cell

8

Signaling, USA), total-ERK1/2, (cat#9102, Cell Signaling, USA), phospho-MEK1/2 (cat#9121,

Cell Signaling, USA), total-MEK1/2 (cat#4694, Cell Signaling, USA), mouse beta-actin

(1:2000) and tubulin (1:2000) (Sigma, USA).

In vitro angiogenesis assay

In vitro angiogenesis assays were performed as previously described (11). In brief, MPC cells or

pericytes (40×103) along with HMVECs (40×103) or HMVECs alone (40×103) in EGM medium

(Lonza, USA) containing 0.2% serum were seeded on growth factors reduced Matrigel

(cat#354230, BD Biosciences, USA) for about 2 hours and then treated with vehicle or with

vemurafenib for about 3.5 hours. After about 6 hours of incubation, cells were photographed.

The number of tubes was counted using a 10× or 20× objective and four fields were chosen per

well with two wells per each condition.

Silencing techniques (short-hairpin RNA)

Stable transduced (viral transductions) thyroid MPC cells engineered to express eGFP (Green

Fluorescent Protein) or silence BRAFV600E mRNA were established according to Nucera et al. (2)

Cell adhesion, migration, and invasion assays

Cell adhesion assays were performed as previously described (2) by coating 50 μg of type I

collagen, or 1% BSA as control, on 48-well plates for 2-3 hours at 37°C. Cells (5-9×103) were

added to the coated wells at 37 °C for 2 hours (treated with vehicle or vemurafenib) and then

fixed and washed before quantifying the number of adhesive cells. Invasion assays were

performed using 24-well BioCoat Matrigel invasion chambers according to manufacturer’s

9

instructions (BD Biosciences, USA). These chambers were used to study the effect of

vemurafenib or vehicle on the invasion of MPC cells or pericytes (8×103 cells/assay) for 12

hours in culture. The assay was performed in a serum free growth medium and we used 5% fetal

bovine serum (FBS) as the chemoattractant agent. The migration assay was performed using

8×103 MPC cells or pericytes loaded into the migration chamber (Corning Incorporated,

Corning, NY, USA) and grown for 3, 5 or 12 hours in culture. In all assays (adhesion, migration,

and invasion), cells were counted (number of cells/field) using a 20× objective, and four fields

were chosen each condition.

Immunofluorescence

Pericytes or eGPF stably transfected thyroid MPC cells and human microvascular endothelial

cells (HMVECs) were cocultured for about 6 hours to allow for tube formation (in vitro

angiogenesis) before performing confocal microscopy. Briefly cells were fixed in 3.7%

formaldehyde, permeabilized in PBS-0.1% TritonX-100 and stained for F-actin (phalloidin-

Alexa Fluo 594, red) (Invitrogen, USA) or CD90 (conjugated in house with Alexa Fluo 488,

green) (Invitrogen, USA) (12). ICAM-1 (human intra-cellular adhesion molecule 1) (clone

RR1/1) (13). Confocal imaging was conducted on a Zeiss LSM 510 (Zeiss, Heidelberg, DE)

using a 63× water immersion objective.

LincRNA and quantitative multigene profiling expression analysis

. We interrogated the long intergenic non-coding RNAs (lincRNA) database for association with

PDGFRB, COLIA1, ITGβ1, and ID2 genes. Accordingly to the LNCipedia 2.0, the latest version

of the lincRNA database that contains 32,183 human annotated lincRNAs (http://lncipedia.org/)

10

and UCSC genome database (http://genome.ucsc.edu/cgi-

bin/hgc?hgsid=340136521&c=chr2&o=8806772&t=8810423&g=lincRNAsTranscripts&i=TCO

NS_00004690), the lincRNA-ID2 (3,651 nucleotides, transcript: TCONS_00004690) is

composed by two exons and is localized on the chromosome 2 at the following genomic

position: chr2:8,806,773-8,810,423 (Supplementary Table 3).

The expression of lincRNA-ID2, ECM, or angiogenesis-associated genes was validated by multi-

gene transcriptional profiles analysis which provides a quantitative view of the expression of

many genes (14). RNA isolation was performed by Quiagen columns (Quiagen, USA) following

the manufacturer’s protocols. Quantitative multi-gene profiling was performed by absolute

quantification using real-time reverse transcriptase PCR (RT-PCR) according to Shih et al. (14).

Primer sequences used for the validation of about 90 genes are reported in Supplementary

Table 3. We classified gene expression as ‘low copy number’ if it was below 1 mRNA copy/106

18S copies, ‘moderate copy number’ if it was between 1 and 15 mRNA copies/106 18S copies,

and ‘high copy number’ if it was greater than 15 mRNA copies/106 18S copies. Genes showing

difference in values of comparisons (p values <0.05, two-sided) were considered differentially

expressed and statistically significant.

Senescence-associated β-galactosidase (SA-β-Gal) assay

Senescence-associated beta-galactosidase (SA-beta-Gal) represents the most widely used

biomarker for senescent and aging cells. SA-beta-Gal is a hydrolase enzyme that catalyzes the

hydrolysis of β-galactosides into monosaccharides only in senescent cells. BRAFV600E-thyroid

MPC cells or pericytes with BRAFWT were seeded at 90% confluence in a 24-well dish and

grown in DMEM/Ham’s F12 supplemented with 0.2% FBS, treated with 1 or 10 µM

11

vemurafenib or vehicle for 6 or 24 hours. After 6 or 24 hours, MPC cells or pericytes were fixed

in 2% formaldehyde/0.2% glutaraldehyde in PBS for 5 minutes and stained (staining solution:

citric acid/phosphate buffer pH 6.0, 150 mM NaCl, 2 mM MgCl, 5 mM potassium ferricyanide,

5 mM potassium ferrocyanide, and 1 mg ml−1 X-gal) overnight at 37 °C according to DeNicola et

al. (15). Fields of at least 50 cells were counted in triplicate for the SA-beta-Gal staining.

Colony formation assay

One hundred thousand primary MPC cells with BRAFV600E, or primary normal thyroid cells

(negative control), or 8505c anaplastic thyroid cancer cells with BRAFV600E (positive control)

were suspended in 0.4% agar and plated on a solidified bottom layer of 0.6% Agar Noble in 6-

well plates. One ml of medium was added to the plate weekly to prevent agar dehydration.

Following six weeks incubation at 37°C, images were taken at a bright field dissecting

microscope and analyzed by counting colonies of ≥ 30 cells.

X-chromosome inactivation assay and clonality analysis

Determination of nonrandom X chromosome inactivation patterns (16, 17) has been of interest in

studies of clonality of neoplastic cells in females (18). We isolated DNA from thyroid MPC

specimen (manually dissected) and thyroid tissues from same patient according to Nucera et al.

(2) For each DNA sample, three reactions were prepared: undigested DNA or digested DNA

with the restriction methylation-sensitive enzymes HpaII or HhaI (New England Biolabs, MA,

USA) according to Allen et al. (16) Methylation of the human androgen receptor (HUMARA)

gene on the X chromosome HpaII or HhaI restriction sites correlates with X chromosome

12

inactivation. Methylation-sensitive PCR reactions and primers were used according to Allen et

al. (16) The 5'-terminal forward primer was labeled with 6-FAM (6-carboxyfluorescein,

fluorescent dye used for labeling oligonucleotides, Life Technologies, USA). PCR reactions

were read by the 3730xl DNA analyzer (DNA sequencing-based genotyping) (Life

Technologies, USA) and analyzed using Genemapper software (Life Technologies, USA).

Patient-derived xenograft mouse model of human thyroid myopericytoma harboring

theheterozygous BRAFWT/V600E mutation

Primary human thyroid MPC cells with BRAFV600E engineered to express eGFP were cultured in

10-cm dishes and grown in DMEM:HAM’s F12 growth medium supplemented with 10% FBS,

penicillin, streptomycin, and amphotericin at 37°C with 5% CO2 atmosphere. On the day of

tumor implantation, the cells were trypsinized, gently centrifuged, and suspended in 100 µL

serum-free DMEM growth medium (Corning Cellgro, USA) to achieve a cell suspension

concentration of 10×104/10 µL. The cells were kept on ice until implantation. MPC cells were

injected subcutaneously, using a Hamilton syringe (Fischer, USA), in the left ear (ear injection

sites are advantageous to visually follow perivascular growth of MPC cells over time in mice) of

4 months-old Crl:NU(NCr)-Foxn1nu female athymic immunodeficient nude mice (strain code:

490) (Charles River, Wilmington, MA, USA) (Beth Israel Deaconess Medical Center,

Institutional Review Board-approved, Boston, MA, USA) (n=5 per group) according to Nagy et

al. (19). Ten mice were randomly divided into two equal groups of 5 mice for the purpose of

establishing a time course of MPC cell growth and response to the therapeutic intervention with

Vemurafenib. Vemurafenib treatment was started 4 days after MPC cell implantation and

performed for 2 consecutive weeks. MPC cell viability was determined by the Photoshop

13

histogram CS5.1 software calculating the absolute number of pixels in the selected eGFP-

positive areas of the left ear divided by the total number of pixels (250,000) in the entire image

of the left ear.

Imaging Techniques

Magnetic resonance imaging

Cross-sectional neck and chest evaluations were performed by means of gadolinium-enhanced

Magnetic resonance imaging (MRI) with the use of a clinical 1.5-Tesla system (General Electric

Medical Systems). Coronal and axial fast spin–echo T1- and T2-weighted sequences were

performed with and without fat suppression, and T1-weighted sequences after gadolinium

enhancement. MRI images were reviewed by experienced radiologist (AJP, BIDMC/Harvard

Medical School) without knowledge of clinical data.

Computed tomography

Computed tomography (CT) of the neck and chest was performed during full inspiration and

during full expiration, with the use of thin-section images. A low-dose non-contrast CT scan was

acquired first with 140 kV and 80 mA, pitch 1.5, and 4.25 mm of slice thickness. Images were

reconstructed using iterative algorithms (OSEM), with 28 subsets and 3 iterations, a 128×128

matrix, and attenuation correction based on CT images. CT images analyzed using a Xeleris 1

Workstation (GE Medical System). CT images were reviewed by experienced radiologist (AJP,

BIDMC/Harvard Medical School) without knowledge of clinical data.

14

Fluorine-18-fluorodeoxyglucose positron emission tomography/computed tomography

Patient, after fasting for at least 6 hours, underwent non-contrast whole-body computed

tomography (CT) scanning 1 hour after the intravenous administration of 198 MBq of Fluorine-

18-fluorodeoxyglucose (F18-FDG) for attenuation correction and for fusion with emission

positron emission tomography (PET) images using a GE Discovery LS scanner (General Electric

Healthcare, Chalfont St. Giles, UK). Patient rested comfortably in a reclined position between

injection and scanning. A series of overlapping emission PET 3D images was then obtained.

Patient had her fasting glucose level checked before injection of F18-FDG to ensure the level

was within the normal range. The area imaged spanned from the base of the skull to the chest.

In order to supplement visual interpretation in PET exam, a standardized uptake value (SUV)

was calculated (using the GEXeleris workstation (General Electric Healthcare, Chalfont St.

Giles, UK) for any F18-FDG-avid lesions from the equation SUV [Tact/Vmax]/[Dinj/B]; Tact is

the tumor activity (MBq), corrected for decay; Vmax is the volume of tumor (in grams); Dinj is

the injected dose (in mCi or MBq), and B is the body weight (in grams). Registered attenuation–

corrected PET, non-corrected PET, CT, and fused images, along with maximum intensity

projection images, were used for interpretation. 18-FDG-PET/CT images were reviewed by

experienced radiologist (AJP, BIDMC/Harvard Medical School) without knowledge of clinical

data.

15

Thyroid scintigraphy

99mTcO4 was administered intravenously in doses of 1 to 5 mCi. The usual dose was 2 mCi. After

a period of thirty to sixty minutes, both dot and photoscans were obtained with a 3-inch

rectilinear scanner. The pulse-height analyzer was set for a 126-to-154-keV window, the time

constant at 0.02 second, and scanning speed at 30 or 44 em per minute. A pinhole collimator was

employed. The correct setting of background suppression, which is of critical importance to the

success of this method, is discussed later. The suppression levels were set at 10 to 70 per cent,

allowing only a minimal imaging of non-target areas. Thyroid scan images were reviewed by

experienced nuclear medicine physician (John Antony Parker, BIDMC/Harvard Medical School)

without knowledge of clinical data.

Ultrasonography and Elastography

Thyroid ultrasound (US) with color doppler was performed by a single experienced radiologist

with a high-resolution ultrasound system (GE Logiq S8 Full Digital) equipped with a high-

energy linear probe of 14 MHz (PZT; Toshiba), allowing to work in fundamental B-mode

(lateral resolution: 0.17 mm; axial resolution: 0.11 mm) and in power Doppler mode (rate of 12

frames/sec, limit detection of 5 cm/sec with a pulse repetition frequency (PRF) of 17 KHz). US

examination included both central and lateral neck compartments. For each lymph node (LN)

detected with a short axis of 2 mm or more, eight US criteria were assessed: long axis, short axis

(perpendicular to the long axis), round shape (defined as a long to short axis ratio of less than 2),

loss of the fatty hyperechoic hilum, hypoechogenicity (by comparison with surrounding

muscles), cystic appearance, hyperechoic punctuations (due to colloid or to microcalcifications),

and peripheral vascularization either associated with central vascularization or not. Thyroid US

16

elastographic measurement was performed during the US examination, using the same real-time

instrument and the same probe. The probe was placed on the neck with light pressure, and a box

was highlighted by the operator that included the nodule to be evaluated. The principle of US

elastography is to acquire two ultrasonic images (before and after tissue compression by the

probe) and track tissue displacement by assessing the propagation of the imaging beam. A

dedicated software (GE Logiq S8 Full Digital) able to provide an accurate measurement of tissue

distortion was used. The US elastogram was displayed over the B-mode image in a color scale

that ranges from red, for components with greatest elastic strain (i.e. softest components), to blue

for those with no strain (i.e. hardest components). Thyroid and neck US images were reviewed

by experienced radiologist (AP, BIDMC/Harvard Medical School) without knowledge of clinical

data.

Laboratory Studies

To assess safety, at each visit the levels of FT3, FT4, TSH, Tg, AbTg, electrolytes, blood urea

nitrogen, creatinine, glucose, hepatic enzymes, bilirubin, serum lipids, and a complete blood

count were performed.

17

Supplementary Results

Multifocal infiltrating MPC of the thyroid gland: case presentation of a novel translational

model

A 44-year-old woman presented to her primary care physician in March of 2010 with a

palpable mass in the anterior triangle of the left side of the neck (Figure 1A1-A3) and difficulty

swallowing. Thyroid hormones (FT3 and FT4) and TSH serum levels were in the normal range.

Thyroid color-doppler ultrasound revealed a goiter and an irregular, hypo-echoic, hyper-vascular

mass in the left thyroid lobe (Figure 1A1-A2), with mediastinal extension. Magnetic Resonance

Imaging (MRI) performed in June 2010 showed a mass localized in the left lobe of the thyroid

with right-sided tracheal deviation and possible extension into soft tissue (Figure 1A3). Left

hemi-thyroidectomy was performed in October 2010. The initial pathologic diagnosis in October

2010, reported by the referring, international institution was undifferentiated (anaplastic) thyroid

carcinoma. By patient request, outside review of the pathology was performed at the

Massachusetts General Hospital (Boston) the following month, and the diagnosis was amended

to reflect a multifocal and infiltrative left thyroid MPC (Figure 1B2-B7), with no clinically

suspicious lymph-nodes or distant metastases. A persistent/recurrent irregular mass with

extension to the superior mediastinum was detected by Computed Tomography (CT) in October

2010 (Supplementary Figure 1A) after left hemi thyroidectomy, and two cycles of

chemotherapy with carboplatin AUC 5 (total 700 mg) and paclitaxel 200 mg/m2 (total 320 mg)

were given to the patient. At the request of the patient, outside review of the pathology was

performed at the Massachusetts General Hospital (Boston, USA) the following month, and the

diagnosis was revised from a malignant thyroid lesion (i.e. anaplastic thyroid carcinoma) to a left

thyroid MPC (Figure 1B2-B7) associated with a nodular goiter (Supplementary Figure 2A1)

18

and an incidental 1.5 mm papillary thyroid microcarcinoma (FVPTC) in the left lobe (Figure

1A1). The MPC extended into perithyroidal connective tissue (Figure 1B5-B7 and

Supplementary Figure 2A2-A3), but no clinically suspicious lymph nodes or distant metastases

were detected. The patient was re-examined in May 2011 and imaging revealed residual bilateral

disease with extension into the superior mediastinum (Supplementary Figure 1B-C).

Completion thyroidectomy was performed in June 2011 (Supplementary Figure 1F) and

revealed a left-sided thyroid mass with extension into the superior mediastinum, but, no

lymphadenopathy in the central or lateral neck (Supplementary Figure 1F). Pathology

evaluation confirmed a thyroid MPC associated with hyperplastic goiter (Supplementary

Figure 2A2-A3).

About one year post-near total thyroidectomy, imaging revealed bilateral masses in the

thyroid bed, consistent with recurrent/persistent disease (Supplementary Figure 1G-J). To our

knowledge, this is the first reported case of a poorly marginated fibrous-appearing benign thyroid

MPC, adherent to adjacent structures in the left paratracheal area (Figure 1A2-A3 and

Supplementary Figure 1A-C and 1F-J). The patient received therapy with levothyroxine (L-

T4) (125 µg/day) achieving TSH suppression, and Thyroglobulin (Tg) was below the normal

range according to the international guidelines; FT3 and FT4 were in the normal range.

Histopathology and immunohistochemical profile of MPC

We discovered that BRAFV600E is a distinct genetic alteration in MPC. It is, however, possible

that MPC heterogeneity reduced the sensitivity to detect the V600E mutation in some cases of

MPC. Thyroid MPC also expressed pAKT, S6, TSP-1 and TSP-5 (data not shown). Thyroid

MPC cells did not express (both in vitro and in the MPC specimen) thyroid-specific markers

19

(e.g. NIS, Tg, TSH-receptor, PAX8, and TTF1), PTH, calcitonin, and pan-cytokeratin (epithelial

marker). In addition, desmin, HMB-45, melan-A, chromogranin, S100 and E-cadherin were

negative. CD45/LCA (leukocyte common antigen) (Supplementary Figure 2B10) and CD68

(macrophage marker) (data not shown) immunohistochemical expression was negative in the

primary thyroid MPC cells in vitro and positive (i.e. tumor stromal microenvironment) in the

thyroid MPC specimen (CD45, Supplementary Figure 2B10). Importantly, both the thyroid

MPC described here and those arising in other organs (20) showed the same morphology and

immunohistochemical profile (αSMA positive and desmin negative). Benign MPC cells show

normochromasia and a normal nuclear-cytoplasmic ratio, somewhat stippled chromatin, and with

eosinophilic cytoplasm with a lack of a clearly demarcated cell borders (Figure 1B4).

Inhibition of BRAFWT/V600E impairs viability, adhesion, migration and glucose uptake of

human primary thyroid MPC cells harboring the heterozygous BRAFWT/V600E mutation

Five micromolar (µM) dose vemurafenib treatment was statistically significant (p<0.001) to

suppress BRAFWT/V600E-MPC cell viability but not BRAFWT pericytes viability (Figure 2B). Cell

growth curve showed that 5 µM dose vemurafenib statistically significant decreased the number

of BRAFWT/V600E-MPC cells following 2-3 days of treatment (p=0.02; Figure 2B). No substantial

effects are seen in pericytes with BRAFWT (Figure 2B). Ten µM dose vemurafenib substantially

affected BRAFWT/V600E-MPC cell morphology within 5 hours or 24 hour (from spindled cells in

the control to rounded shape upon vemurafenib treatment, arrows; Figure 2B). No substantial

morphological change or functional effects were observed in pericytes with BRAFWT (Figure

2B). Flow cytometric analysis showed an accumulation of vemurafenib-treated BRAFV600E-MPC

cells in the G1 phase and a statistically significant MPC cell population in sub-G1 (apoptosis)

20

compared to vehicle-treated MPC cells (p=0.02) (data not shown). Phospho(p)ERK1/2 and

pMEK1/2 protein expression levels were substantially decreased (about 90%) in BRAFWT/V600E-

thyroid MPC cells following treatment with 10 µM (IC90) vemurafenib compared to vehicle

(DMSO) (Figure 2C); 1 µM vemurafenib reduced for about 22% pERK1/2 and pMEK1/2 protein

expression levels. No changes in either pERK1/2 or pMEK1/2 protein expression levels were

found in pericytes with BRAFWT treated with vemurafenib compared to DMSO (control) (Figure

2C).

Since ki67 proliferative index was modest (~5% in either BRAFWT/V600E-MPC tissue or

primary BRAFWT/V600E-MPC cells), we assessed the protein levels of p16/INK4A, a crucial cell

cycle inhibitor and marker of quiescent cells. Sixty percent of BRAFWT/V600E-positive thyroid

MPC cells expressed nuclear p16/INK4A (Supplementary Figure 2B12) and 90% expressed

nuclear pERK1/2 (Supplementary Figure 2B7). All p16-positive nuclei also expressed

pERK1/2 (Supplementary Figure 2B13). Human lysosomal endogenous beta-galactosidase (β-

gal) protein, a marker of senescence, was expressed in less than 5% of MPC cells

(Supplementary Figure 2B14). Since p16 has also been proposed as a pro-senescence marker in

BRAFWT/V600E-positive benign nevi, we analyzed the percent of senescent BRAFWT/V600E-MPC

cells in vitro and found no statistically significant changes in the vemurafenib-treated (10 µM)

MPC cells (mean=0.9 β-gal positive cells/field, SD (Standard Deviation)=0.2) compared to

vehicle cells (mean=0.9 β-gal positive cells/field, SD=0.3). No statistically significant changes

were also observed in the normal BRAFWT-pericytes (from mean=14.3 β-gal positive cells/field

(SD=1.7) upon vemurafenib treatment to mean=14.7 cells/field (SD=2.5) in the vehicle control

condition). Conversely, stable BRAFV600E knockdown by short-hairpin RNA (shRNA) (which

showed a substantial decrease (~90%) in BRAFWT/V600E protein levels in MPC cells and

21

substantially reduced (~90%) endogenous levels of phosphorylated (p)ERK1/2 (Thr202/Tyr204)

protein levels, Supplementary Figure 4A-B), statistically significantly (p<0.001) induced MPC

cell senescence in vitro (Supplementary Figure 4D). SA-β-Gal assay showed a statistically

significant increased number of senescent cells (mean=9.2β-gal positive MPC cells/field

(SD=1.1)) in primary BRAFWT/V600E-MPC cells with BRAFV600E knock-down as compared to the

sh-control BRAFWT/V600E-MPC cells (mean=0.4 β-gal positive MPC cells/field (SD=0.5),

p<0.001; Supplementary Figure 4D). Overall, it may be possible that BRAFWT/V600E protected

MPC cells from senescence. The role of BRAFWT/V600E in the senescence process is still

controversial; our findings suggest that this oncogene might ‘antagonize’ cyclin p16 activity

and/or expression, which it is known to elicit inhibitory effects on cell cycle progression and

stimulatory effects on senescence.

Importantly, vemurafenib treatment statistically significant disrupted this effect (from

mean=17.6 tubules/field (SD=2.5) in the control to mean=7 tubules/field (SD=1) in the

vemurafenib-treated BRAFWT/V600E-MPC cells, p=0.002) (Figure 2D1-D2). Additionally,

vemurafenib (1 µM) impaired adhesion on type I collagen of BRAFWT/V600E-positive thyroid

MPC cell within 3 hours (from mean=11 cells/field (SD=0.89) upon vehicle (DMSO) treatment

to mean=4 cells/field (SD=0.8) after vemurafenib treatment, p=0.002; Supplementary Figure

3B). No statistically significant changes were found in adhesion of pericytes with BRAFWT

treated with vemurafenib (12±2 cells/field) as compared to vehicle (DMSO) treatment

(mean=11.6 cells/field (SD=0.5)) (Supplementary figure 3B). One µM vemurafenib treatment

statistically significant impaired BRAFWT/V600E-MPC cell migration within 5 hours (Figure 2E);

BRAFWT/V600E-MPC cells migrated more (19±1 cells/field) as compared to BRAFWT pericytes

(mean=9 cells/field (SD=1)) (p=0.002) (Figure 2E). BRAFWT/V600E-MPC cell migration was

22

inhibited ~2-fold by vemurafenib (mean=10.6 cells/field (SD= 0.5)) as compared to the vehicle

treated BRAFWT/V600E-MPC cells (p=0.002), but had no effect in BRAFWT-pericytes (Figure 2E).

Vemurafenib (10 µM) treatment statistically significant (p<0.01) almost completely blocked

MPC cell adhesion and migration (data not shown). Furthermore, treatment with specific short-

hairpin RNA (shRNA) that targets BRAFV600E statistically significant reduced adhesion (from

mean=13 thyroid MPC cells/field (SD=1.5) in the control to mean=7.6 MPC cells/field (SD=

0.5) when BRAFV600E was knocked-down, p=0.002) and migration (from mean=23 thyroid MPC

cells/field (SD=2.3) in the control to mean=6.3 MPC cells/field (SD=1.3) when BRAFV600E was

knocked-down, p=0.002) of MPC cells (Supplementary Figure 4C). Neither MPC cells nor

pericytes showed any substantial invasive capability through the Matrigel (data not shown).

BRAFWT/V600E-MPC cells compared to pericytes (data not shown) with BRAFWT were

substantially able to survive as single non-proliferating/quiescent cells in an anchorage-

independent microenvironment for 6 weeks (Supplementary Figure 3A). In fact, MPC cells

grew as single cells (4 cells/10× field) in the presence of 0.4% agar, or as cell aggregate in the

presence of Matrigel which reproduced a three-dimensional reconstituted ECM in vitro

(Supplementary Figure 3A).

Because the thyroid BRAFWT/V600E-MPC had high 18F-FDG uptake compared to normal

thyroid tissue (Figure 1A4 and Supplementary Figure 1D and 1H-I), we tested the effects of

vemurafenib on 14C[U]Glucose uptake in vitro. We found that treatment with vemurafenib

(either 1 or 10 µM) resulted in a 1.6-fold statistically significant decreased in glucose uptake

within 5 hours in MPC cells as compared to vehicle (p=0.02).

mRNA expression levels (copy number) of genes fundamental for extracellular matrix

(ECM) remodeling and angiogenesis were statistically significant up-regulated 1.6-fold (ITGβ1),

23

1.3-fold (PDGFRB), 5-fold (CollagenIA1 (COLIA1)), or 6.5-fold (DNA-binding protein

inhibitor (ID2)) in BRAFWT/V600E-positive thyroid MPC cells as compared to BRAFWT-pericytes

(Figure 2F). Cells were treated with 10 µM dose vemurafenib or DMSO (control) for 24 hours

in the presence of 0.2% serum cell growth medium. BRAFWT/V600E-MPC cells expressed the

following genes (absolute copies of mRNA): ITGβ1 (mean=476.8 copies (SD=19.3)/106 18S

copies compared to mean=349.5 copies (SD=8.2)/106 18S copies in the vemurafenib treated

cells, p=0.02), PDGFRB (mean=18.1 copies (SD=0.9)/106 18S copies compared to mean=12.6

copies (SD=0.5)/106 18S copies in the vemurafenib treated cells, p=0.02), COLIA1 (mean=1434

copies (SD=294.7)/106 18S copies in the control cells compared to mean=977.2 copies

(SD=76.5)/106 18S copies in the vemurafenib treated cells, p=0.007), and ID2 (mean=57.1

copies (SD=0.9)/106 18S copies compared to mean=26.5 copies (SD=2.1)/106 18S copies in the

vemurafenib treated cells, p=0.002). BRAFWT pericytes expressed the following mRNA

expression levels for these genes (absolute copies of mRNA): ITGβ1 (mean=285.3 copies

(SD=19.4)/106 18S copies in the control cells compared to mean=278.96 copies (SD=25)/106 18S

in the vemurafenib treated cells), PDGFRB (mean=14.7 copies (SD=1.5)/106 18S copies in the

control cells compared to mean=15.6 copies (SD=0.8)/106 18S copies in the vemurafenib treated

cells), COLIA1 (mean=282.6 copies (SD=47.8)/106 18S copies in the control cells compared to

mean=226.6.9 copies (SD=32)/106 18S copies in the vemurafenib treated cells), and ID2

(mean=8.7 copies (SD=0.3)/106 18S in the control cells compared to mean=7.7copies

(SD=1.2)/106 18S copies in the vemurafenib treated cells). Recently, it has been found that

microenvironment ECM genes may be associated with lincRNAs that may facilitate tumor

progression or, conversely, tumor suppressor-like function, and considered as prognostic

biomarkers. We sought to determine whether the ECM genes which were up-regulated in the

24

BRAFWT/V600E-MPC cells contain lincRNAs sequences. Our gene expression validation by RT-

PCR confirmed a lincRNA off the ID2 gene that is statistically significant (p=0.002) up-

regulated in the BRAFWT/V600E-MPC cells (Figure 2F). ID2-LincRNA is localized 12,000

nucleotides (nts) upstream from the first exon of the ID2 gene. According to the UCSC genome

database, the lincRNA-ID2 (3,651 nts, transcript: TCONS_00004690) is composed of two exons

and is localized on chromosome 2 at the following genomic position: chr2:8,806,773-8,810,423.

Vemurafenib treatment statistically significant (p=0.02) reduced (1.6-fold) ID2-lincRNA

expression levels as compared to vehicle treatment (data not shown).

Furthermore, BRAFWT/V600E-MPC cells expressed FN (mean=1426 copies (SD=62.1)/106

18S copies), TGFβ1 (mean=23.9 copies (SD=0.6)/106 18S copies), VEGFA (mean=4.4 copies

(SD=0.3)/106 18S copies), VEGFC (mean=10.7 copies (SD=0.5)/106 18S copies) and FGF2

(mean=18.6copies (SD=0.7)/106 18S copies) mRNA. Treatment with vemurafenib did not

change statistically the mRNA levels of these ECM/angiogenesis genes as compared to controls.

However, mRNA expression levels of FN (mean=103.7copies (SD=6.5)/106 18S copies;

p=0.002), VEGFC (mean=5.2 copies (SD=0.4)/106 18S copies; p=0.002), TGFβ1

(mean=7.8copies (SD=2.6)/106 18S copies; p=0.02) and FGF2 (mean=12.9 copies (SD=0.1)/106

18S copies; p=0.002) were statistically significantly lower in normal pericytes with

BRAFWT,whereas VEGFA (mean=3.6 copies (SD=0.4)/106 18S copies) mRNA levels did not

change statistically compared to BRAFWT/V600E-MPC cells.

BRAFWT/V600E-MPC cells grown in vitro secreted statistically comparable amounts of

VEGFA (62 pg/ml) and FGF2 (191 pg/ml) as normal pericytes with BRAFWT [VEGFA (65

pg/ml) and FGF2 (230 pg/ml)]. Furthermore, treatment with vemurafenib did not change

statistically the levels of these secreted factors compared to controls.

25

A PDX mouse model of heterozygous BRAFWT/V600E-positive human primary thyroid MPC

cells

Sixty percent of the injected mice at day 4 (baseline) after MPC cell implantation, showed an

intense eGFP protein expression with BRAFWT/V600E-MPC cell implantation in the left ear.

Vemurafenib (100 mg/kg twice daily by oral gavage) (n=5) or vehicle (n=5) treatment was

started 4 days after MPC cells implantation (defined ‘baseline’) and performed twice daily by

oral gavage for 2 consecutive weeks. MPC cell viability was determined by the Adobe

Photoshop CS5.1 histogram software calculating the absolute number of pixels in the selected

eGFP-positive areas of the left ear divided by the total number of pixels (247,545) in the

highlighted selection area. This analysis showed a statistically significant decrease (~3-fold) of

the eGFP-signal in the vemurafenib-treated mice at day 17 (number of pixels in the eGFP-

positive area/total pixels (%) = 2.8, SD=0.05), without any obvious toxicity, as compared to the

vemurafenib-group mice at the baseline (number of pixels in the eGFP-positive area/total pixels

(%) = 8.2, SD=0.1) (p=0.007). Whereas, a statistically significant increase (~1.3-fold) of the

eGFP-signal was observed in the vehicle-treated mice at the day 17 (number of pixels in the

eGFP-positive area/total pixels (%) = 9.2, SD=0.4) as compared to the baseline (number of

pixels in the eGFP-positive area/total pixels (%) = 7.5, SD=1.1) (p=0.007).

Importantly, eGFP immunostain highlighted MPC cells in both control and treated

animals, with a statistically significant protein expression in the control-treated mice (mean=57.6

eGFP+ MPC cells/field (SD=2.2)) as compared to the vemurafenib-treated mice (mean=15.6

eGFP+ MPC cells/field, SD=3.6) (p=0.002), likely reflective of a reduced tumor burden

26

(Supplementary Figure 5). CD31 immunostaining highlights small blood vessels in both

control mice and treated mice (Supplementary Figure 5); the number of vessels (CD31+)

(mean=11 vessels/field, SD=0.8) is statistically significantly higher in the control-treated mice

compared to the vemurafenib-treated mice (mean=3 vessels/field, SD=0.8) (p=0.002)

(Supplementary Figure 5). Immunohistochemical staining also showed that ki67 (proliferation

marker) was focally expressed (score: 1+) in the MPC cells of the control-treated mice and

substantially absent in the MPC cells of the vemurafenib-treated mice (score: 0)

(Supplementary Figure 5).

References 1 Fletcher CDM, Bridge, J.A., Hogendoorn, P., Mertens, F. WHO Classification of Tumours of Soft Tissue and Bone. Fourth Edition. IARC Press, Lyon, France. 2013;5. 2 Nucera C, Porrello A, Antonello ZA, et al. B-Raf(V600E) and thrombospondin-1 promote thyroid cancer progression. Proc Natl Acad Sci USA. 2010;107(23):10649-54. 3 Nucera C, Mazzon E, Caillou B, et al. Human galectin-3 immunoexpression in thyroid follicular adenomas with cell atypia. J Endocrinol Invest. 2005 Feb;28(2):106-12. 4 Beroukhim R, Mermel CH, Porter D, et al. The landscape of somatic copy-number alteration across human cancers. Nature. 2010 Feb 18;463(7283):899-905. 5 Matulonis UA, Hirsch M, Palescandolo E, et al. High throughput interrogation of somatic mutations in high grade serous cancer of the ovary. PloS one. 2011;6(9):e24433. 6 Fischer AH, Taysavang P, Jhiang SM. Nuclear envelope irregularity is induced by RET/PTC during interphase. AJP. 2003 Sep;163(3):1091-100. 7 Nucera C, Nehs MA, Nagarkatti SS, et al. Targeting BRAFV600E with PLX4720 displays potent antimigratory and anti-invasive activity in preclinical models of human thyroid cancer. Oncologist. 2011;16(3):296-309. 8 Nehs MA, Nucera C, Nagarkatti SS, et al. Late intervention with anti-BRAF(V600E) therapy induces tumor regression in an orthotopic mouse model of human anaplastic thyroid cancer. Endocrinology. 2012 Feb;153(2):985-94. 9 Shaik S, Nucera C, Inuzuka H, et al. SCF(beta-TRCP) suppresses angiogenesis and thyroid cancer cell migration by promoting ubiquitination and destruction of VEGF receptor 2. J Exp Med. 2012 Jul 2;209(7):1289-307. 10 Bollag G, Hirth P, Tsai J, et al. Clinical efficacy of a RAF inhibitor needs broad target blockade in BRAF-mutant melanoma. Nature. 2010 Sep 30;467(7315):596-9. 11 Ponce ML. In vitro matrigel angiogenesis assays. Methods Mol Med. 2001;46:205-9.

27

12 Martinelli R, Newton G, Carman CV, Greenwood J, Luscinskas FW. Novel role of CD47 in rat microvascular endothelium: signaling and regulation of T-cell transendothelial migration. ATVB. 2013 Nov;33(11):2566-76. 13 Rothlein R, Dustin ML, Marlin SD, Springer TA. A human intercellular adhesion molecule (ICAM-1) distinct from LFA-1. J. Immunol. 1986. 137: 1270-1274. J Immunol. 2011 May 1;186(9):5034-8. 14 Shih SC, Smith LE. Quantitative multi-gene transcriptional profiling using real-time PCR with a master template. Exp Mol Pathol. 2005 Aug;79(1):14-22. 15 DeNicola GM, Karreth FA, Humpton TJ, et al. Oncogene-induced Nrf2 transcription promotes ROS detoxification and tumorigenesis. Nature. 2011 Jul 7;475(7354):106-9. 16 Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM, Belmont JW. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X chromosome inactivation. Am J Hum Genet. 1992 Dec;51(6):1229-39. 17 Shattuck TM, Westra WH, Ladenson PW, Arnold A. Independent clonal origins of distinct tumor foci in multifocal papillary thyroid carcinoma. N Engl J Med. 2005 Jun 9;352(23):2406-12. 18 Vogelstein B, Fearon ER, Hamilton SR, Feinberg AP. Use of restriction fragment length polymorphisms to determine the clonal origin of human tumors. Science. 1985 Feb 8;227(4687):642-5. 19 Nagy JA, Shih SC, Wong WH, Dvorak AM, Dvorak HF. Chapter 3. The adenoviral vector angiogenesis/lymphangiogenesis assay. Methods Enzymol. 2008;444:43-64. 20 Mentzel T, Dei Tos AP, Sapi Z, Kutzner H. Myopericytoma of skin and soft tissues: clinicopathologic and immunohistochemical study of 54 cases. Am J Surg Pathol. 2006 Jan;30(1):104-13.

28

Supplementary Table 1. Clinical-pathologic data and genotyping of 29 patients with human myopericytoma (MPC).

Histology (case)

Site Gender Age Focality Infiltrative appearance

(tumor margins)

Genotyping

Benign myopericytoma

Thyroid F 44 Multifocal Yes BRAFV600E

Benign myopericytoma

Finger (left hand)

F 42 Single nodule

Yes BRAFV600E

Benign myopericytoma

Left leg M 37 Multifocal No BRAFV600E

Benign myopericytoma

Parotid F 40 Multifocal No BRAFWT

Benign myopericytoma

Left thumb M 53 Single nodule

No BRAFWT

Benign myopericytoma

Left lower leg M 43 Single nodule

No BRAFWT

Benign myopericytoma

Lower leg F 35 Single nodule

No BRAFWT

Benign myopericytoma

Forearm F 35 Single Nodule

No BRAFWT

Benign myopericytoma

Trunk F 48 Single nodule

No BRAFWT

Benign myopericytoma

Left cheek M 74 Single nodule

No BRAFWT

Benign myopericytoma

Left ankle, joint M 66 Single nodule

No BRAFWT

Benign myopericytoma

Left plantar M 83 Single nodule

No BRAFWT

Benign myopericytoma

Trunk F 35 Single nodule

No BRAFWT

Benign myopericytoma

Left upper thigh F 22 Multifocal No BRAFWT

Benign intravascular myopericytoma

Left temple M 67 Single nodule

No BRAFWT

Benign intravascular myopericytoma

Left knee M 48 Single nodule

No BRAFWT

Benign intravascular myopericytoma

Right artery ulnaris

M 67 Single nodule

No BRAFWT

Benign intravascular myopericytoma

Right forearm distal

F 64 Single nodule

No BRAFWT

Benign myopericytoma

Left thigh M 35 Single nodule

No BRAFWT

Benign myopericytoma

Right calf F 64 Single nodule

No BRAFWT

Benign myopericytoma

Left leg F 40 Single nodule

No BRAFWT

Benign myopericytoma

Submental F 3 Single nodule

No BRAFWT

Benign myopericytoma

Left leg M 10 Single nodule

No BRAFWT

Benign myopericytoma

Left shoulder F 16 Single nodule

Yes BRAFWT

29

Malignant myopericytoma

Left arm F 81 Single nodule

No BRAFWT

^Malignant myopericytoma

Right heel M 19 Single nodule

Yes BRAFWT

Malignant myopericytoma

Left thigh M 46 Single nodule

No BRAFWT

Malignant myopericytoma

Left neck F 80 Single nodule

No BRAFWT

Malignant myopericytoma

Left buttock F 52 Single nodule

No BRAFWT

M=male; F=female. ^Malignant MPC arose from a focus of a multifocal benign MPC. Multifocal MPC= the presence of more than one morphologically distinct lesions within the same patient in the same general region.

30

Supplementary Table 2. Antibodies list and experimental conditions

Primary Antibody Company Dilution and time of incubation

(All antibodies were incubated for 45 minutes

at room temperature except for some ones

with overnight incubation at +4 degree)

Antigen retrieval (time: 30-45 minutes)

NG2 Abcam Cat# 83508

1:1000, overnight 10 minutes proteinase digestion

PDGFRB Epitomics 1469-1

1:500, overnight Citrate buffer and pressure cooker

CD44 Abcam Ab16728

1:500, 45 minutes Citrate buffer and pressure cooker

αSMA Sigma A2547

1:20,000 none

Vimentin Dako M7020

1:400 20 minutes proteinase digestion

Desmin Sigma D1033

1:5000 Citrate buffer and pressure cooker

D2-40 Covence Sig-3730

1:100 none

CD31 Dako M0823

1:30 15 minutes proteinase digestion

CD34 Dako M7165

1:400 none

Chromogranin Neomarker MS-324

1:4000 Citrate buffer and pressure cooker

Synaptophysin Invitrogen 180130

1:50 none

Calponin Dako M3556

1:2000 Citrate buffer and pressure cooker

P63 Biocare CM163B

1:250 Citrate buffer and pressure cooker

pan-keratin (AE1, AE3) Dako M3515

1:200 10 minutes proteinase digestion

CAM-5.2 BD Pharmingen 349205

1:50 10 minutes proteinase digestion

E-cadherin Dako M3612

1:75 Citrate buffer and pressure cooker

KI67 Dako M7240

1:200 Citrate buffer and pressure cooker

Phospho-ERK1/2 Cell Signaling 4376

1:150, overnight Citrate buffer and pressure cooker

Phospho-AKT Cell signaling 4060

1:50, overnight Citrate buffer and pressure cooker

P-S6 Cell signaling 4858

1:75, overnight Citrate buffer and pressure cooker

TSP-1 Abcam Ab1823

1:25, 45 minutes Citrate buffer and pressure cooker

TSP-5 Kamiya biomedical 1:300, overnight Citrate buffer and pressure cooker

CD99 Covance Sig 3620

1:150 Citrate buffer and pressure cooker

LCA (CD45) Dako M0701

1:600 Citrate buffer and pressure cooker

31

HMB45 Dako M0634

1:400 none

Melan-A Dako M7196

1:200 Citrate buffer and pressure cooker

CD68 Dako M0814

1:1000 10 minutes proteinase digestion

P16 Ventana 9517

1:2 Citrate buffer and pressure cooker

P53 Immunotech 1767

1:1200 Citrate buffer and pressure cooker

NIS (sodium-iodide symporter)

Provided from dr. Sabine Costagliola (Free University of

Brussels)

1:10, overnight Citrate buffer and pressure cooker

Tg (thyroglobulin) Novus NB100-65184

1:250 No treatment

TPO (thyroid peroxidase) Sigma A007987

1:25, 45 minutes room temperature

Citrate buffer and pressure cooker

TSH (thyroid stimulating hormone) receptor

Sigma A926689

1:50, overnight Citrate buffer and pressure cooker

PAX8 Proteintech 10036-1-AP

1:800 or 1:1200 Citrate buffer and pressure cooker

TTF1 Dako M3575

1:300 Citrate buffer and pressure cooker

S100 Dako Z0311

1:1000 None

VEGF Abcam Ab27620

1:2 Citrate buffer and pressure cooker

VEGFR2 Cell Signaling 2479

1:300 EDTA and pressure cooker

Beta-Galactosidase Abcam Ab616

1:750 Citrate buffer and pressure cooker

Collagen1A1 Sigma HPA008405

1:250 Citrate buffer and pressure cooker

Fibronectin Dako, A0245 1:10,000 10 minutes proteinase digestion

Integrin Beta 1(ITGB1) Provided by dr. Richard Hynes (MIT,

Cambridge, MA, USA)

1:500 Citrate buffer and pressure cooker

GFP (Green fluorescent protein)

Santa Cruz Biotechnology

sc-8334

1:800 overnight Citrate buffer and pressure cooker

32

Supplementary Table 3. Primers sequences list F= forward primer R= reverse primer Hu = human Genes Primers sequences Hu SYK.F ACGCCAAGATCAGTGATTTCG Hu SYK.R GTACCACTTGACAGGCCACTTTC Hu CD36.F AAGTCACTGCGACATGATTAATGG Hu CD36.R CTGCAATACCTGGCTTTTCTCA Hu VEGFR2.F GGCAAATGTGTCAGCTTTGTACA Hu VEGFR2.R GGTCACGTGGAAGGAGATCAC Hu THBS1 (TSP1).F GGACCAGAGAGACACTGATATGGAT Hu THBS1 (TSP1).R CTCCAATGCGGTCTGAGTCA Hu CD29 (ITGB1).F CAACGAGGTCATGGTTCATGTT Hu CD29 (ITGB1).R CCAGCTACAATTGGAATGATGTCT Hu BRAF.F CCTTCCGACCAGCAGATGA Hu BRAF.R ATGCACATTGGGAGCTGATG Hu MMP2.F TTGATGGCATCGCTCAGATC Hu MMP2.R CTTGTCACGTGGCGTCACA Hu MMP9.F CCTGGAGACCTGAGAACCAATC Hu MMP9.R GATTTCGACTCTCCACGCATCT Hu MMP7.F GAATGTTAAACTCCCGCGTCAT Hu MMP7.R TCCTGTAGGTGACCACTTTGGA Hu ADAMTS9.F GAAGCGGAACTCAAGCACTGT Hu ADAMTS9.R ATTCCTGAGCAGAGGCTGATG Hu CTSB.F AAGCTTCGATGCACGGGAACAATG Hu CTSB.R TGGATGCAGATCCGGTCAGAGAT Hu CTSS.F TCACCACTGGCATCTCTGGAAGAA Hu CTSS.R AGGTGGTTCATGCCCAGATCGTAT Hu CD49C (ITGA3).F ACAGGCTCTGGAGAACCACACT Hu CD49C (ITGA3).R CTGCAAGTTGCTCTCACACTTGT Hu CD49F (ITGA6).F AATATCAGGCTGCCAAATGCA Hu CD49F (ITGA6).R GGTACTCCCGAATACTGAGCTACAG Hu VEGFA.F GCGGAGAAAGCATTTGTTTGT Hu VEGFA.R CGGCTTGTCACATCTGCAA Hu FN1.F TGTTCGTGCAGCTGTTTACCA Hu FN1.R TCTGTGACACAGTGGCCATAGG Hu TGFB1.F AGGGCTACCATGCCAACTTCT Hu TGFB1.R CCCGGGTTATGCTGGTTGT 18S.F TGTCTCAAAGATTAAGCCATGCA 18S.R GCGACCAAAGGAACCATAACTG Hu VEGFC.F CCCCACCAATTACATGTGGAAT Hu VEGFC.R ATCTCCAGCATCCGAGGAAA Hu CD36.F AAGTCACTGCGACATGATTAATGG Hu CD36.R CTGCAATACCTGGCTTTTCTCA Hu HMGB1.F GCCTCCTTCGGCCTTCTT Hu HMGB1.R ACAGGCCAGGATGTTCTCCTTT Hu PTK2.F AGCCCAATGTGGAGGACTCTAC Hu PTK2.R GGATTAGACGCTCTTCCATCAGA Hu HGFR (MET; C-MET).F TGGGCTGCAGACATTTCCA Hu HGFR (MET; C-MET).R CTCCGATCGCACACATTTGT Hu CD44.F CGGAGAGGCCAGCAAGTC Hu CD44.R ACTGGTCTGGAGTTTCTGACGAC Hu TGFBR1.F AGTCATCACCTGGCCTTGGT Hu TGFBR1.R AACATGAGTGAGATGCAGACGAA Hu CD66C (CEACAM6).F TCCTATATGTGCCAAGCCCATA Hu CD66C (CEACAM6).R CCAGAGACTGTGATCATCGTGACT Hu MCAM.F AGGAACACAGTGGGCGCTAT Hu MCAM.R CTCACTCGGACGTCAGACACA Hu CD331 (FGFR1).F CGTGGAGTTCATGTGTAAGGTGTAC Hu CD331 (FGFR1).R CCATTCACCTCGATGTGCTTT Hu CD45.F GTGGAGAAAGGACGCATGCT Hu CD45.R TGCCAAGAGTTTAAGCCACAAA

33

Hu CD56 (NCAM1).F GGGCATCCTCATCGTCATCT Hu CD56 (NCAM1).R GCACATGAACAGGCCACACT Hu CD73 (NT5E).F TATCCGGTCGCCCATTGAT Hu CD73 (NT5E).R GGCAATACAGCAGCCAGGTT Hu EGFR.F TGAAGGAGCTGCCCATGAG Hu EGFR.R GGATGCTCTCCACGTTGCA Hu COL1A1.F CCAGTGTGGCCCAGAAGAAC Hu COL1A1.R CGCCATACTCGAACTGGAATC Hu LAMC1.F GGTTGATTGAAATCGCATCCA Hu LAMC1.R TGGCTGAGTGACTGACACATTG Hu FGF1 #1.F CAGGAGCGACCAGCACATT Hu FGF1 #1.R CGGTGTCCATGGCCAAGTA Hu NOTCH1.F CGGGTCCACCAGTTTGAATG Hu NOTCH1.R CCGCAGAGGGTTGTATTGGTT Hu CSPG4 (NG2).F TGCTCTTCTACCTCCGAAAACG Hu CSPG4 (NG2).R GCGAAAGGTCTCGGTGTCA Hu ACTB (B-Actin).F CTGGAACGGTGAAGGTGACA Hu ACTB (B-Actin).R AGTCCTCGGCCACATTGTG Hu CD34.F CCTCTGTAATCAGCACAGTGTTCA Hu CD34.R CAGGTGACAGGCTAGGCTTCA Hu ENG (CD105).F GACGCCATGACCCTGGTACT Hu ENG (CD105).R AGAAGGTCAGGCCCGTGAT Hu CD90 (Thy1).F GCCTAACGGCCTGCCTAGT Hu CD90 (Thy1).R GGGTGAACTGCTGGTATTCTCAT Hu AKT1 #1.F ACCATGAGCGACGTGGCTAT Hu AKT1 #1.R GCCGCCAGGTCTTGATGTAC Hu VEGFR1.F GCTGACAGCAACATGGGAAA Hu VEGFR1.R ACCAAGGTGCTAGCCATCTTATTA Hu FLT4(VEGFR3).F GTAACAGGCATTCGCATCGA Hu FLT4(VEGFR3).R GTGGCGCTTGGACTTTGTTC Hu TGFA.F AGGAGGACAAGCCAGCATGT Hu TGFA.R CGGTGATGGCCTGCTTCTT Hu FGF2.F GTATGTGGCACTGAAACGAACTG Hu FGF2.R AATCAGCTCTTAGCAGACATTGGA Hu PDGFB.F CTCGATCCGCTCCTTTGATG Hu PDGFB.R GGGTCATGTTCAGGTCCAACTC Hu CD140b (PDGFRB).F GGTGGGCACACTACAATTTGC Hu CD140b (PDGFRB).R GGTGGGTAGGCCTCGAACA Hu CDH5 (CD144; VE-Cadherin).F GAACCCAAGATGTGGCCTTTAG Hu CDH5 (CD144; VE-Cadherin).R GATGTGACAACAGCGAGGTGTAA Hu TP53.F AACAGCTTTGAGGTGCGTGTT Hu TP53.R GCGGAGATTCTCTTCCTCTGTG Hu NRP1.F GGCTAGAAATCTGGGATGGATTC Hu NRP1.R GATCGGATTCGACCTGGTGTT Hu HK2#2.F CGTGGTGACCCTGCTGAAG Hu HK2#2.R GCAACCACATCCAGGTCAAAC Hu SLC2A1 (GLUT1).F TGCTCATGGGCTTCTCGAA Hu SLC2A1 (GLUT1).R TGGTCAGGCCGCAGTACAC Hu PTEN #2.F GGATTATAGACCAGTGGCACTGTTG Hu PTEN #2.R GCTGGCAGACCACAAACTGA Hu HIF1A.F TGCTGACCCTGCACTCAATC Hu HIF1A.R AGTGCTTCCATCGGAAGGACTA Hu XLKD1.F GGTGGCATCCCCTTACTCTACA Hu XLKD1.R TTCTCCGTGGAATAGAAGTGGAA Hu CD332 (FGFR2).F GGAACCCAATGCCAACCAT Hu CD332 (FGFR2).R GTAGCCTCCAATGCGATGCT Hu SP1.F TGCTGGCACAGTCACTGTGA Hu SP1.R GTACCCAATGCACTGAGGTTAATG Hu FOXM1.F GGAGCAGCGACAGGTTAAGGT Hu FOXM1.R GGCCGCTCAGACACAGAGTT Hu NRP2.F CGCGGAGGAGACAGTATCACT Hu NRP2.R TCCTTGCCGTTTAGGCTGTAG Hu RAB5A.F AGGCAAGCAAGTCCTAACATTGT Hu RAB5A.R GACTGTGCTTCCTGGAAATCTACTG Hu ID2.F ATGAAAGCCTTCAGTCCCGT Hu ID2.R TTCCATCTTGCTCACCTTCTT Hu FOXA1#2.F GTTCTCCATCAACAACCTCATGTC Hu FOXA1#2.R GGCAACGTAGAGCCGTAAGG

34

Hu Raf1#2.F GGCACGGAGATGTTGCAGTA Hu Raf1#2.R AACAGCCACCTCATTCCTGAA Hu FIGF#2(VEGFD).F TTGCTGGAACAGAAGACCACTCT Hu FIGF#2(VEGFD).R GCAACGATCTTCGTCAAACATC Hu COL1A2.F CAATGCTGGCAGCCAGTTT Hu COL1A2.R AGGTGATGTTCTGAGAGGCATAGTT Hu COL3A1.F TGGTCAGTCCTATGCGGATAGAG Hu COL3A1.R CGGATCCTGAGTCACAGACACA Hu COL4A1.F CTGGCAGAGATGGTGTTGCA Hu COL4A1.R CTTTGAGCCGCAAGTCGAA Hu COL4A2.F AGCAAGGCAACAGAGGACTTG Hu COL4A2.R GGTGGAGGGTGTCTGATGGA Hu COL4A3BP.F GCGTCGACATGGCTCAATG Hu COL4A3BP.R GAATGAAGAGGTGGATGTTGCA Hu COL5A1.F CCATGCCTGAAACCAGTGAAG Hu COL5A1.R CTCGCCATAGGTGAGGTCATC Hu COL5A2.F CACTCAGAATGGCCAGATGTACTT Hu COL5A2.R TCCTGGCATTCTATCTTGTCACA Hu COL12A1.F GTGAACGGCACGTGTTCATT Hu COL12A1.R AGAGGACAACTTGAAGTGGCAGTT Hu COL18A1#1.F GCAGCAGCACGGAAGATTC Hu COL18A1#1.R GGAAGTGTCTGAGCTCCTAACGA Hu LAMC2.F AAGGCTGTCCAACGAAATGG Hu LAMC2.R GTCTAGTCGTTGGGCTGAGCTAA Hu CD31 (PECAM1).F CACCTGGCCCAGGAGTTTC Hu CD31 (PECAM1).R AGTACACAGCCTTGTTGCCATGT Hu CAPZA1.F GTCGTTGGAGATCAGAGTGGAAGT Hu CAPZA1.R TAACCAACTGAACATTGCCATCTT Hu CD62E (SELE).F CTGCCAAGTGGTAAAATGTTCAAG Hu CD62E (SELE).R CTTGCACACAGTGCCAAACAC Hu VCL.F TCATTCAGTGTGCCAAGGACAT Hu VCL.R GTGCACTGCTTGGCAACCT Hu XLKD1(LYVE1).F GGTGGCATCCCCTTACTCTACA Hu XLKD1(LYVE1).R TTCTCCGTGGAATAGAAGTGGAA Hu DLL4.F CTGGAGCTCAGCGAGTGTGA Hu DLL4.R AGCCATCCTCCTGGTCCTTAC Hu ANG1.F CGCTGCCATTCTGACTCACATA Hu ANG1.R CCGGTTATATCTTCTCCCACTGTTT Hu ANG2.F GCCGCTCGAATACGATGACT Hu ANG2.R GCTTCATTAGCCACTGAGTGTTGT Hu TIE2.F TGCCCAGATATTGGTGTCCTTAA Hu TIE2.R CATAAAGCGTGGTATTCACGTAGGT Hu CD51 (ITGAV).F GCGTCTTCCCGGATGTTTC Hu CD51 (ITGAV).R GGGTAGAAGACCAGTCACATTTGAG Hu SRC#1.F GCACTACAAGATCCGCAAGCT Hu SRC#1.R AGCTGCTGCAGGCTGTTGA Hu CD49B (ITGA2).F ACTCTGTGGCTGCAATTTCTACTG Hu CD49B (ITGA2).R GCACTATCTGGCCGGTATAATTTG Hu CD49D (ITGA4).F AGGTGTCCAGCAGAGAAGCTAACT Hu CD49D (ITGA4).R GGTGGTAAGCAGCTTCAATCTGA Hu CD11b (Mac1, ITGAM).F TCGAGTACGTGCCACACCAA Hu CD11b (Mac1, ITGAM).R TTCCGAAGCTCAGCCAGAA Hu IGF1.F CTTCCGGAGCTGTGATCTAAGG Hu IGF1.R CGGACAGAGCGAGCTGACTT Hu IL1b.F CGAATCTCCGACCACCACTAC Hu IL1b.R TGTCCATGGCCACAACAACT Hu IL6.F GGCAGAAAACAACCTGAACCTT Hu IL6.R GGCAAGTCTCCTCATTGAATCC Hu IL8.F CTCTGTGGTATCCAAGAATCAGTGA Hu IL8.R TATTGCATCTGGCAACCCTACA Hu TNFA.F TCAGAGGGCCTGTACCTCATCT Hu TNFA.R GGTGTGGGTGAGGAGCACAT Hu ID2-lincRNA.F CTTTGCATTCCAGTGTGGTG Hu ID2-lincRNA.R GGTGAGTCCAGGAAGGGAAT

35

Supplementary Figures Supplementary Figure 1

Supplementary Figure 1. A multimodality perioperative evaluation of patient with multifocal, infiltrating and recurrent thyroid myopericytoma harboring the heterozygous BRAFWT/V600E mutation. (A) Computed Tomography (CT) with contrast performed 1 month after left hemi thyroidectomy showed a ~3 cm hypo dense, irregular, and heterogeneous mass (clinically "wooden") localized in the left thyroid bed (left para tracheal region, arrow) with extension into the superior mediastinum, adherent to adjacent structures in the left paratracheal area with infiltration of neck muscles and the recurrent laryngeal nerve (arrowhead highlights a right thyroid nodule). (B) CT with contrast performed 6 months after left hemithyroidectomy shows ~3.3 cm hypo dense, irregular, and heterogeneous mass localized in the left thyroid bed (left para-tracheal region, arrow) with extension into the superior mediastinum. Arrowhead highlights a right thyroid nodule with mediastinal extension. (C) Thyroid Doppler ultrasound shows the left thyroid remnant and a persistent irregular and hypo-echoic 53×23 mm mass with intra-lesional vascularity and extension into the superior mediastinum after hemithyroidectomy. (D) Total body 18F-FDG (Fluorodeoxyglucose) PET (Positron emission tomography)/CT exam after hemi thyroidectomy revealed high 18F-FDG uptake in the left persistent/recurrent mass (arrow) (standardized uptake value (SUV) = 4.5) and a SUV=2.4 in the right nodule. (E) Thyroid scan with 99mTcO4

- shows a non-functional

36

left thyroid lobe after hemithyroidectomy. (F) Completion thyroidectomy: presence of poorly marginated fibrous-appearing lesion, adherent to adjacent structures in the left para-tracheal area. (G) Conventional thyroid Doppler ultrasound 1 year after near-total thyroidectomy shows a 12 mm irregular hypo-echoic mass with intra-lesional vascularity in the left thyroid bed. (H-I) Total body 18F-FDG PET/CT exam 1 year after near-total thyroidectomy revealed high 18F-FDG metabolic uptake in the left persistent/recurrent mass (SUV=5.6) and a SUV=6 in the upper part of the cricoid cartilage in the right para-tracheal region. (J) Contrast enhanced T1-weighted magnetic resonance imaging (MRI) 1 year after near-total thyroidectomy shows a recurrent/persistent 21 mm hypo-intense left para-tracheal lesion.

37

Supplementary Figure 2

Supplementary Figure 2. Immunohistochemistry of the multifocal and infiltrative thyroid myopericytoma harboring the heterozygous BRAFWT/V600E mutation. (A1-A3, hematoxylin and eosin staining): (A1, scale bar= 500 µ) Thyroid hyperplastic nodule, (A2-A3, scale bar= 500 µ) Myopericytoma (MPC) cells are clustered with a fascicular growth pattern around these vessels (clear, slit-like spaces, V); the tumor shows a somewhat infiltrative appearance through peri-thyroidal connective tissue (A2) and adipose tissue (arrowheads, A3). MPC cells show cytoplasmic to membranous staining with antibodies against platelet-derived growth factor receptor beta (PDGFRB) (C1, scale bar= 500 µ) and α-smooth muscle actin (αSMA) (C2, scale bar= 500 µ and C3, scale bar= 100 µ).

38

Intralesional thin-walled branching vessels of small caliber are highlighted by both αSMA (C4, scale bar= 500 µ) and CD31 (C5, scale bar= 500 µ). MPC cells also show positivity for NG2 (chondroitin sulfate proteoglycan 4) (punctate, C6, scale bar= 500 µ), nuclear phospho (p)-ERK1/2 (C7, scale bar= 100 µ), vimentin (C8, scale bar= 100 µ), but are negative for E-cadherin (C9, scale bar= 100 µ). Lymphoid infiltration in the MPC is highlighted by the leukocyte common antigen (CD45/LCA) staining, a membrane glycoprotein restricted to leukocytes (C10, scale bar= 100 µ). (C11, scale bar= 100 µ) Weak p53 expression (arrows), (C12, scale bar= 100 µ) focal p16/INK4A nuclear expression (arrows) and (C13, scale bar= 100 µ; inset, magnification = 1000×, scale bar= 500 µ) nuclear co-localization pERK1/2 and p16 (arrows) in in MPC cells. (C14, scale bar= 500 µ) Human lysosomal endogenous beta-galactosidase (β-gal) protein, a marker of senescence, was expressed in scattered MPC cells (arrows). (C15, scale bar= 100 µ) Double immunohistochemical staining shows protein co-localization between αSMA and CD44 (cell surface glycoprotein, marker of cell adhesion and stem cell), or (C16, scale bar= 100 µ) between pERK1/2 and CD44 (arrows) in MPC cells. Immunohistochemical expression of: (C17, scale bar= 100 µ) CollagenIA1 (COLIA1) (asterisk), (C18, scale bar= 100 µ) Fibronectin (FN) (asterisk) in the peritumoral (MPC) connective tissue (arrow) and (C19, scale bar= 100 µ) cytosolic integrin β1 (ITGβ1, asterisk highlights MPC area and arrow marks infiltrative MPC cells into adjacent adipose tissue. (C1-C4) DNA genotyping analysis by mass spectrometry (MS) traces of formalin-fixed paraffin embedded tissue samples. (C1) MS traces of the follicular variant of papillary thyroid microcarcinoma (FVPTC, n=1) and (C2) of the normal thyroid tissue (n=1) show no BRAFV600E mutation. (C3) MS analysis of a skin multifocal MPC reveals the mutant copy of heterozygous BRAFWT/V600E (A>T) (n=28), whereas the normal skin (n=1) in C4 is homozygous wild type (WT) BRAF. The intensity of the signal versus mass of the analyte is plotted in the background, while the inset shows the cluster plot with all samples analyzed in the run (intensity of the signals of the two alleles). The long arrow links the actual sample (circle in the inset) to the mass of the expected mutant V600E allele peak or of the absent V600E allele peak. Color coded by the software is automatic allele calls. Cyan squares for heterozygous samples, orange triangles for homozygous. Ambiguous calls are identified by a red dot. Allele frequencies deviating from the expected values are assigned ambiguous or homozygous calls by the software.

39

Supplementary Figure 3

Supplementary Figure 3. Anchorage independent and collagen-dependent assays in primary human thyroid myopericytoma cells harboring the heterozygous BRAFWT/V600E mutation. (A) Anchorage-independent cell growth assay (0.4% agar noble; scale bar= 50 µ; inset with a scale bar= 50 µ) and anchorage-dependent cell growth assay (in vitro reconstituted basement membrane using growth factors reduced Matrigel matrix; scale bar= 100 µ) of BRAFWT/V600E-positive human primary thyroid myopericytoma (MPC) cells seeded in a 6-well plate for 6 weeks. Arrows highlight MPC cells which grew as single cells. These findings were validated by two independent experiments. (B) Cell adhesion on type I collagen of BRAFWT/V600E-positive thyroid MPC cell and pericytes within 3 hours in the presence of vehicle (control) or vemurafenib treatment. These data represent the average ± standard deviation (error bars) of six independent replicate measurements (**p=0.002, Mann-Whitney test, two-sided).

40

Supplementary Figure 4

Supplementary Figure 4. BRAFV600E silencing in primary human thyroid myopericytoma cells harboring the heterozygous BRAFWT/V600E mutation. (A) Western blot protein analysis in human primary thyroid myopericytoma (MPC) cells harboring BRAFWT/V600E and transduced with short-hairpin RNA (shRNAs) (sh) #1 and #10 targeting BRAFV600E. (B) MPC cells (magnification in all images= 20×; scale bar in all images= 100 µ) transduced with either sh-Luciferase (Luc, control) or sh-BRAFWT/V600E expressing eGFP (green fluorescent protein). These findings were validated by two independent experiments. (C) Cell adhesion and migration assays in human primary thyroid MPC cells targeting BRAFV600E with specific short-hairpin RNA (shRNA).

41

Magnification in all images: 20×; scale bar= 50 µ in the cell adhesion assays images and 100 µ in the cell migration assays images. These data represent the average ± standard deviation (error bars) of six independent replicate measurements (**p=0.002, Mann-Whitney test, two-sided). (D) Senescence associated β-galactosidase (SA-β-Gal) assay in sh-control MPC cells or sh-BRAFWT/V600E MPC cells. Arrows indicate the β-gal positive MPC cells. Magnification in all images: 20×; scale bars= 20 µ. These data represent the average ± standard deviation (error bars) of six independent replicate measurements (***p<0.001, Mann-Whitney test, two-sided).

42

Supplementary Figure 5

Supplementary Figure 5. Immunohistochemical analysis in a patient-derived xenograft (PDX) mouse model of human primary thyroid myopericytoma cells harboring the heterozygous BRAFWT/V600E mutation. Human thyroid primary myopericytoma (MPC) cells engineered to express eGFP (green fluorescent protein) injected in the subcutaneous compartment of the left ear of nude mice (n=10). Vemurafenib treatment was started 4 days after MPC cells implantation and performed twice daily by oral gavage for 2 consecutive weeks. Mice were treated with vemurafenib (100 mg/kg twice daily by oral gavage) (n=5) or vehicle (n=5) for 2 weeks. Control-treated BRAFWT/V600E-MPC cells can be seen infiltrating the left ear dermis (arrows) or vemurafenib-treated BRAFWT/V600E-MPC cells with substantial minimal foci of residual tumor (arrows) in the left ear dermis, using hematoxylin/eosin (H&E) staining (scale bar= 200 µ).

43

These findings were validated by two independent experiments. eGFP immunostain highlights MPC cells in both control-treated and vemurafenib-treated nude mice. Microvascular density is defined by number of vessels per field showing CD31 staining which highlights small blood vessels in the control-treated mice and vemurafenib-treated mice (arrowheads, scale bar= 200 µ). Immunohistochemical staining shows that ki67 (proliferation marker) is substantially focally expressed (arrowheads, score: 1+) in the MPC cells of control-treated nude mice and absent in the MPC cells of vemurafenib-treated nude mice (score: 0) (scale bar= 100 µ). The IHC markers were assessed semiquantitatively using the following scoring method: 0 (negative), 1 (1–10% positive cells, low expression), 2 (11–50% positive cells, moderate), and 3 (more than 50% positive cells, high expression).

44

Supplementary Figure 6

Supplementary Figure 6. Experimental model of BRAFWT/V600E-positive human myopericytoma microenvironment. Heterozygous BRAFWT/V600E regulates autocrine and paracrine signaling in the myopericytoma (MPC) microenvironment (e.g. thyroid) by up-regulating the expression of COLIA1 (collagen IA1), ID2 (inhibitor of DNA binding 2), ITGβ1 (integrin beta 1), and PDGFRB (platelet-derived growth factor receptor beta) genes. Hematoxylin and eosin (H&E) staining of thyroid MPC image with a magnification= 400× and scale bar= 500 µ. Inset (CD31+) with a magnification= 1000× and scale bar= 500 µ.