[CANCER RESEARCH 54, 4744-4749, September 1. 1994]

Expression of Thyroid-specific Transcription Factors TTF-1 and PAX-8 inHuman Thyroid Neoplasms1

Dora Fabbro, Carla Di Loreto, Carlo Alberto Beltrami, Antonino Belfiore, Roberto Di Lauro, andGiuseppe Damante2

Dipartimento di Scienze e Tecnologie Biomediche [D. F., R. D. L, G. D.¡and Dipartimento di Scienze Mediche e Morfologiche ¡C.D. L, C. A. BJ, Università di Udine,Ospedale Gervasutta. via Cervasutta 48 Udine, and Cattedra di Endocrinologia, Università di Catania, Catania ¡A.BJ, Italy

ABSTRACT

TTF-1 and PAX-8 are tissue-specific transcription factors expressed in

the thyroid follicular cells, contributing to the maintenance of the differentiated phenotype. In fact, it has been demonstrated that TTF-1 andPAX-S are able to activate transcription from thyroglobulin and thyroper-

oxidase (TPO) promoters, the transcriptional activity of which is in vivorestricted only to the thyroid follicular cell. In order to gain insight intohow these transcription factors control in vivo the differentiation of thethyroid cell and to have a better molecular characterization of humanthyroid tumors, TTF-1, PAX-8, thyroglobulin, and TPO mRNA levels

were measured in nonmalignant and malignant human thyroid tissues.Results indicate that the expression of TTF-1 and PAX-8 is not sufficientper se for the expression of the thyroid-differentiated phenotype. Furthermore, in follicular adenomas, PAX-8 mRNA levels are strictly related toTPO mRNA levels, suggesting that the amount of PAX-8 could play a rolein the modulation of TPO gene expression. TTF-1 mRNA is always well

detectable in papillary carcinomas and, in contrast, always absent inanaplastic carcinomas. Identical results were obtained when the expression of TTF-1 protein was investigated using immunohistochemistry.Thus, TTF-1 gene expression could be a molecular marker in order to

distinguish these two types of thyroid neoplasms.

INTRODUCTION

Transcription factors play a pivotal role in the determination andmaintenance of cellular phenotype. The activity of transcription factors is in fact considered as the main switch to regulate gene expression (1). Transcription factors may be classified according to thelocalization of their expression. In this manner, these factors can bedivided into ubiquitous and tissue-specific groups. Members of the

latter group are present in a few cell types only and participate in thetranscriptional regulation of genes expressed only in these cells (2-6).Hence, tissue-specific transcription factors can be very important for

expression of the differentiated phenotype of each cell.Two transcription factors, expression of which is restricted to the

thyroid follicular cell, have been cloned thus far: TTF-1 (7) andPAX-8 (8). In addition to the follicular thyroid cell, in the adultanimal, TTF-1 and PAX-8 are also expressed in the lung and kidney,

respectively. However, the two factors are present together only in thethyroid follicular cell, suggesting that this unique combination couldplay a role in the expression of the thyroid-specific phenotype. Thisview is strengthened by molecular data; both TTF-1 and PAX-8 bindto sequences of Tg3 and TPO promoters and, albeit to a different

extent, they are able to activate transcription of Tg and TPO genes(9-10). These data have been obtained by cotransfection of cell lineswith both expression vectors of TTF-1 or PAX-8 and reporter genestranscriptionally controlled by Tg and TPO promoters. A main draw-

Received 3/16/94; accepted 7/6/94.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1This work is funded by the Progetto Finalizzato Oncologia of the Consiglio Nazio

nale delle Ricerche and by the Associazione Italiana Ricerca Cancro.2 To whom requests for reprints should be addressed.3 The abbreviations used are: Tg, thyroglobulin; TPO, thyroperoxidase; GAPDH,

glyceraldehyde-3-phosphate dehydrogenase.

back of this experimental approach is the difficulty to quantitativelyreproduce the real concentration of the expressed transcription factors.This problem is very important because, for several of these proteins,it has been clearly demonstrated that relatively small changes of theirexpression can have enormous biological effects (11-12). For this

reason, the specificity of the phenotype observed using systems basedon cell transfection is difficult to be clearly defined.

For these motivations, in order to better understand the role ofTTF-1 and PAX-8 in the maintenance of the thyroid-differentiated

phenotype, it would be relevant to use approaches able to complementfindings obtained by in vitro studies. Thus, we decided to study theexpression of TTF-1 and PAX-8 genes, together with the expression

of their putative target genes Tg and TPO, both in normal andneoplastic human thyroid tissues. Thyroid tumors originating from thefollicular cells comprise lesions with a broad spectrum of cell differentiation which include benign adenomas, differentiated (papillaryand follicular) carcinomas, poorly differentiated carcinomas, and thehighly undifferentiated anaplastic carcinomas (13). The degree ofdifferentiation of these classes of neoplasms is heterogeneous. Often,heterogeneity in the expression of the differentiated phenotype isencountered for tumors of the same class (14).

The studies presented in this paper could also provide clinicallyrelevant information. In fact, differentiated thyroid carcinomas (whichrepresent over 90% of the total malignancies in thyroid) are a heterogeneous disease: some of them are a mortal disease while others donot affect life expectancy (15-17). Thus far, clinical and histomor-

phological characteristics are often inadequate to predict their clinicalbehavior. Therefore, new parameters, such as molecular characterization, could help to better classify them and to more accurately predicttheir prognosis. Because in tumor cells differentiation markers andmalignant phenotype are very often inversely related, the expressionof tissue-specific transcription factors controlling the differentiated

phenotype could be an additional tool in the evaluation of canceraggressiveness.

MATERIALS AND METHODS

Plasmids. The DNA fragments used as probes in Northern analysis were:for Pax-8, a HindlWEcoRl fragment of human Pax-8 complementary DNAcontained in plasmid H26P/S3; for Tg, plasmid pRTG2 containing a 5' coding

fragment of Tg complementary DNA (18); for TPO, the Sall/EcoRl codingfragment of pTPO 5'c (18); for TTF-1, the 0.7-kilobase Soci fragment of

plasmid prTTF-1/4 (7); for GAPDH, a 1.3-kilobase Pstl fragment of plasmidpGAPDHl containing the coding region of GAPDH (from H. Francis-Lang,

Imperial Cancer Research Fund, Oxford, United Kingdom).Tissue Collection, RNA Extraction, Northern Blots, and Values Nor

malization. Thyroid tissue was obtained from patients undergoing surgery forclinical indications. After surgery, thyroid pieces were cut into fragments ofabout 1 cm3 and quickly frozen. Histological diagnosis was obtained for all

tissues. Total RNA from frozen tissues and FRTL-5 cells was prepared by theacid guanidinium thiocyanate-phenol chloroform procedure (19). Northern

blots were performed using standard procedures (formaldehyde/agarose geland filter hybridization using the protocol described in Ref. 20), and at the end,filters were exposed at -80C° for autoradiography. The intensity of the signals

was quantitated by scanning densitometry of the autoradiograms. Each value

4744

Research. on October 6, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

TTF-1 AND PAX-8 EXPRESSION IN THYROID NEOPLASMS

Tg

TPO

TTF-1

PAX-8

GAPDH

„

further normalized for Tg, TPO. TTF-1. and PAX-8 gene expression of

FRTL-5 cells. To perform this normalization, in each gel a same amount of

FRTL-5 RNA was run. In this manner, the values of the thyroid-specific gene

expression in human tissue do not indicate the absolute amounts hut the ratio

between the gene expression of human tissues and the gene expression ofFRTL-5 cells.

Immunohistnchemistry. Immunohistochemical detection of TTF-1 wasperformed using a polyclonal antibody to TTF-1 produced in rabbit. Specimens

were fixed in 4% buffered formaldehyde and were routinely embedded inparaffin. The streptavidin-biotinylated alkaline phosphatase technique was

applied. Briefly, after rehydration, the sections were incubated with normal

swine serum. Subsequent incubations were as follows (all at room temperature): antibody to TTF-1 (1:200 dilution) for 2 h; biotin-labeled swine anti-rabbit for 30 min; streptavidin-biotinylated alkaline phosphatase complex

(Dakopatts A/S, Gloustrup, Denmark) for 30 min; and incubation in newfuchsin substrate solution. The sections were counterstained with hematoxylinand mounted with Glycergel. Negative controls were carried out. Cells wereconsidered positive when red staining of the nucleus was identified. The extentof TTF-1 positivity in each case was evaluated by determining the percentage

of positivity in at least 5000 cells.

Table 1 Tg. TPO. TTF-1. und PAX-X mRNA lei-els in thyroid follicular adenomas

mRNA levels were obtained by densitometric scanning of autoradiograms and normalized as described in "Materials and Methods."

kilobases; TTF-1, 2.4 kilobases; PAX-8, 3.1 kilobases.

GPV oP^ Case no. TgTPO4

1.120.17ion

in normal and nonmalignantwere performed as described incative cases only are shown. For " 'h of: Tg, 8.5 kilobases: TPO, 3.2 ^ £63**J46

1.638.0511.3610.1560.582.44iRNA.

For this gene, mRNA gj Q^262•sstudied (data not shown). 66 0.040.06different

gels, each value was 72 0.415.4^ll!,,,,,!,

1C...3.5

-3

-2.5

-Tg

2-1.5

-1

-0.5

-•

R:0.43••^*~**~•----

— •^^— -4•^

^"~£

•1

4-1

2-1

0-TPO

8-"4

-2

-n

.TTF-11.8924.1

37.419.439.035.741.649.910.611.33.113.5PAX-834327

74430620

34636956717833121753•

* R: 0.33

•<*

•

mm0

10 20 30 40 50 0 10TTF-120

30TTF-140

5

Fig. 2. Relationship belwecn the mRNA levels ofthe thyroid-specific transcription factors (TTF-1 andPAX-8) and target genes (Tg and TPO) in thyroid

follicular adenomas. Cases and mRNA levels arethose numerically presented in Table 2.

3.5-3

•2.5

-Tg

2-1.5

-1

-0.5

-0

-•

R:0.12•*—

— — — "•*""••*

•

100 200 300 400 500 600 700 800PAX-8

4745

100 200 300 400 500 600 700 800PAX-8

Research. on October 6, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

TTF-1 AND PAX-8 EXPRESSION IN THYROID NEOPLASMS

Tg

TPO

TTF-1 . . •

PAX-8

GAPDH

100L

41 23

J L94

J

Papillarycarcinoma

Follicularcarcinoma

Fig. 3. Tg, TPO, TTF-1, and PAX-8 gene expression in malignant thyroid tissues. RNAextraction and Northern blot were performed as described in "Materials and Methods."

Autoradiograms of exemplicativc cases only are shown.

RESULTSTg, TPO, TTF-1, and PAX-8 Gene Expression in Normal and

Nonmalignant Thyroid Tissues. We determined whether the variability of our experimental approach (described in "Materials andMethods") allowed a reliable comparison between values obtained

from samples run in different gels and those independently processed.Several samples were run in two different gels that were indepen

dently blotted and further processed. Tg, TPO, TTF-1, and PAX-8

mRNA levels were detected and normalized as described in Methods.Successively, the differences of values obtained for each sample in thetwo different blots were calculated. For each gene, the mean difference was not over 10% of the mean value. Hence, Tg, TPO, TTF-1,and PAX-8 gene expression was evaluated in normal thyroids (n = 3)

and in differentiated, nonneoplastic pathological tissues such as goiters (;¡= 13) and follicular adenomas (n = 12). In all of these tissues,Tg, TPO, TTF-1, and PAX-8 gene expression was detectable (Fig. 1),

although at variable levels. The levels of expression of the mRNAsstudied were especially variable in thyroid adenomas. (Values areshown in Table 1.) An attempt was made to correlate the expressionof Tg and TPO with that of their regulators PAX-8 and TTF-1 (Fig.2). Weak correlations were found between Tg and TTF-1 mRNAlevels and TPO and TTF-1 mRNA levels. Conversely, a highlysignificant correlation was found between TPO and PAX-8 mRNAlevels, suggesting that the amount of PAX-8 in the thyroid follicular

cells could be a limiting factor for TPO gene expression. No correlation was found between levels of PAX-8 and the expression of Tg.These data indicate that PAX-8 mostly acts on TPO gene expression.The weak correlations between TTF-1 and target genes suggest thatthe amount of TTF-1 plays only a marginal role for the modulation of

Tg and TPO gene expression.Tg, TPO, TTF-1, and PAX-8 Gene Expression in Malignant

Thyroid Tissues. The expression of thyroid-specific genes was in

vestigated in 21 differentiated carcinomas (20 papillary and 1 follicular), in 1 poorly differentiated carcinoma (insular carcinoma), and in3 undifferentiated (anaplastic) carcinomas. In differentiated carcinomas, Tg, TPO, TTF-1, and PAX-8 mRNA gene expression was highly

heterogeneous. In Fig. 3, a representative sample of the results obtained is shown. In some cancers, all four genes were expressed atlevels similar to normal tissues, while in some others, the expression

a) b)

Gena expression(\ of normal -issues)

140-120-100->n

uea)80

-SO

-40

-20-*1«XOriff"X••joatXjri••••a0•••••o

Gene expressionof nomai tiiiues)

TPO PAX-8 TTF-1 TB TPO PMC-8

Fig. 4. a, Tg, TPO. TTF-1. and PAX-8 mRNA levels of differentiated thyroid carcinomas. Values were obtained as described in "Materials and Methods" and for each gene

normalized for respective mean values of normal tissues, b, relationship between TPO and PAX-8 values of papillary carcinomas, a and h: •,mRNA levels of papillary carcinomas;O, gene expression of a papillary cancer where an anaplastic focus was evident, a, the insular carcinoma (crosses) and the follicular carcinoma (EH)mRNA levels are shown.

4746

Research. on October 6, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

TTF-1 AND PAX-8 EXPRESSION IN THYROID NEOPLASMS

of one or more thyroid-specific gene(s) was undetectable or very low.In order to better analyze the results, Tg, TPO, TTF-1, and PAX-8

gene expression levels were quantitated as described previously. Results of differentiated carcinomas are shown in Table 2 and Fig. 4.TTF-1 and PAX-8 gene expression was detected in all papillarycarcinomas. TTF-1 mRNA was always well detectable, except in one

case (case 18C, see Table 2 and Fig. 4) where anaplastic foci weregrowing inside a papillary carcinoma. PAX-8 gene expression was

slightly reduced or slightly increased compared to normal tissues in 6of 20 cases, whereas in the remaining cases, it was markedly reduced,reaching levels as low as 5% of the normal levels. Tg mRNA wasundetectable in 5 of 20 (25%) cases, and TPO mRNA was undetectable in 7 of 20 (35%) cases. The levels of expression of these twogenes were not correlated. These data suggest that Tg and TPO geneexpression is controlled by, at least in part, independent mechanisms.In general, a good correlation was present between TPO and PAX-8

mRNA levels in papillary carcinomas (Fig. 4b). However, in threecases (nos. 41, 57, and 94 in Table 2), there was a dramatic reductionin TPO mRNA levels, in spite of only an about 50% reduction ofPAX-8 mRNA levels. TTF-1 and PAX-8 mRNA levels were not

correlated to each other, suggesting that these two genes are, at leastpartially, independently controlled.

As shown in Table 2, in order to investigate whether the level ofTTF-1 and PAX-8 mRNAs was related to the different aggressiveness

in papillary carcinomas, these 20 neoplasms were subdivided into ahigh risk group (8 cases with a maximal diameter 5:1.5 cm, extrathy-

roidal tumor extension, and nodal métastases)and a low risk group(12 cases without extrathyroidal tumor extension and nodal métastases). Values of Tg, TPO, and PAX-8 but not TTF-1 were lower in

neoplasms of the high risk group. Only two cases of the high riskgroup (cases 21C and 11C) showed overlapping values with the lowrisk group; in case 21C, however, TPO was undetectable, and PAX-8

was low. The only carcinoma classified at histology as poorly differentiated (insular carcinoma) unexpectedly showed a good expressionof all differentiation markers except Tg that was quite reduced compared to control values (Fig. 4). Analysis of anaplastic carcinomasshowed a complete shut-off of all thyroid-specific genes in the three

cases studied.

Table 2 Thyroid-specific gene expression in hitman papillary cancers

For each gene, data are expressed as % of values observed in normal tissues. Caseswere grouped into low and high risk papillary carcinomas according to the criteriadescribed in the text.

Table 3 Inimitnohislochemislry detection of TTF-1 in malignati! timi nonmalignant

hitman thyroid tissuesResults are expressed as % of nuclei specifically stained by a specific TTF-1 antibody.

Caseno.Low-risk

papillarycarcinoma62333415794100120122133139161High-risk

papillarycarcinoma40163ne12C14C15CIXC21CTg10815131127922106255210413339001120IIo842TPO558102050870158752021070I)II100TTF-11158412684639411S8463525284737394635231HI84PAX-8145210758664613216141272513515120713113820

No.Diagnosis44°Goiter50"53555816419

Adenoma24"3551"16516627

Papillarycarcinoma33"40°41"527612116716816982"

Follicularcarcinoma17017117217317417517617717889"

Anaplasticcarcinoma113179180TTF-1

protein (% of

positivenuclei)68556162696873847677838784928073768884SO858972777682887575SISII820II00

' Tissues studied by Northern blots.

TTF-1 Protein Detection by Immunohistochemistry. Data obtained using Northern blots indicate that TTF-1 could be used as a

molecular marker to distinguish differentiated and undifferentiatedthyroid carcinomas. In order to further support this finding, theimmunohistochemistry detection of TTF-1 in paraffin-embedded sec

tions was performed. Malignant and nonmalignant thyroid tissueswere studied. Results are expressed in Table 3, and exemplicativecases are shown in Fig. 5. All papillary and follicular carcinomasshowed a high percentage of nuclear positivity, superimposable to thatobserved in goiters and follicular adenomas. In contrast, in anaplasticcancers, no staining for TTF-1 was observed.

DISCUSSION

The data obtained in this study strongly support the notion thatTTF-1 and PAX-8 play a major role in controlling the tissue-specific

gene expression of the follicular thyroid cells. In fact, in none of thetissues studied could the presence of Tg or TPO mRNAs be observedin the absence of TTF-1 or PAX-8 mRNAs.

TTF-1 is a very strong activator of Tg promoter transcriptionalactivity, as demonstrated by cotransfection experiments.4 Moreover,

during development, TTF-1 mRNA appears before Tg mRNA (21).These data suggest that TTF-1 is an important determinant for Tg

gene expression in follicular thyroid cells. Accordingly, in the presentstudy, we found that, in anaplastic thyroid carcinomas, the absence ofTTF-1 is associated to the absence of Tg gene expression. This finding

1M. De Felice. G. Damante. R. Di Lauro, unpublished data.

4747

Research. on October 6, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

TTF-1 AND PAX-K I XI'RI SSION IN THYROID NEOPLASMS

» « ' -J>T' fr.'.

cFig. 5. TTF-1 expression in thyroid carcinomas. In papillary carcinoma (a) and

follicular carcinoma (b), nuclei of a large percentage of tumor cells show positivity forTTF-1. Anaplastic carcinoma (c) is negative (immunoalkaline phosphatase-hematoxylin;X 4(X)).

is in agreement with a study performed on anaplastia thyroid carcinoma cell lines (22). However, some papillary carcinomas expressTTF-1 but not Tg. This finding is in agreement with two other in vivo

conditions: (a) in the rat, between days 10 and 15 of development,TTF-1 is present in the thyroid, but Tg mRNA is not yet transcribed;and (b) in the lung, TTF-1 is present (7), but Tg is not transcribed. Inour opinion, all these findings indicate that TTF-1 is a factor necessarybut not sufficient for the full expression of the thyroid-differentiated

phenotype. In fact, several studies indicate that the two essentialmolecular functions of TTF-1, DNA binding and transcriptional activation, can be regulated. Using "/';; vitro" systems, situations have

been described: (a) where TTF-1, although present, is not able to bindDNA (23); and (b) where TTF-1, although capable of binding to

DNA, is not able to activate transcription (18). Our present dataindicate that similar situations could occur in human papillary thyroidcarcinomas.

By cotransfection experiments, it has been demonstrated that PAX-8 isa dose-dependent activator of the TPO promoter and, but to a much lesser

extent, of the Tg promoter (10). Present data show a tight correlationbetween TPO and PAX-8 mRNA levels in all thyroid tissues examined,

especially in follicular adenomas. Lack of correlation between TPO andPAX-8 mRNA levels was observed in three papillary carcinomas andcould indicate that the regulatory effect of PAX-8 on TPO gene expres

sion is coupled to at least another independent regulatory event. It isimportant to note that, in FRTL-5 cells, TPO mRNA levels are controlledby thyroid-stimulating hormone, at least partially, through nontranscrip-

tional mechanisms (24). Nonetheless, our data suggest that the amount ofPAX-8 present in the cell is a major determinant for the levels of TPOgene expression; therefore, variations in the amount of PAX-8 levels

could play a role not only in the determination and maintenance of thedifferentiated phenotype but also in controlling the functional behavior ofthe thyroid follicular cell (10).

From the clinical point of view, the markers studied could certainlydifferentiate between differentiated and anaplastic thyroid carcinomas. The diagnosis of anaplastic thyroid carcinomas usually can beachieved by clinical and histopathological criteria. However, the differentiation markers can be useful, in addition to morphologicalparameters, in very early stages of anaplastic changes, like in case18C of our series.

When we grouped papillary carcinomas into high risk (diameter^1.5 cm, extrathyroidal invasion, and métastases)and low risk (absence of these characteristics), we found significantly lower values inthe expression of PAX-8, Tg, and TPO in the high risk group, with

value overlap present only in two cases.Due to the indolent evolution of most papillary carcinomas, the

duration of follow-up did not allow us to correlate differentiation

markers to tumor relapse and mortality rates. Insular carcinomas areclassified at histology in the class of less differentiated cancers and areusually aggressive. There is increasing evidence, however, that theyare a heterogeneous group (like papillary carcinomas are), and someof them may concentrate iodine (25-26). This characteristic is clinically important because it allows us to treat métastaseswith radioio-

dine therapy. Accordingly, the case described here showed a goodcorrelation of all differentiation markers.

It is likely that these markers, in addition to a panel of othermolecular markers (e.g.. oncogene and tumor suppressor gene mutations, expression of growth factors, and their receptors; Refs. 27-36)

will, in the future, help more accurately in classifying both differentiated and less differentiated thyroid carcinomas.

ACKNOWLEDGMENTS

We thank Dr. Alfonso Colombatti for the use of the radioisotope facilitiesat the Centro di Riferimento Oncologico. Aviano, Italy. We thank Carlo LoCascio for assistance during the artwork in the preparation of the figures.

REFERENCES

1. Mitchell. P. J.. and Tjian, R. Transcriptional regulations in mammalian cells bysequence-specific DNA binding proteins. Science (Washington DC), 245: 371-378,

1989.2. Umek, R. M, Friedman A. D., and McKnight. S. L. CAAT-enhanccr binding protein:

a component of differentiation switch. Science (Washington DC), 251: 288-292,1991.

3. Voss, J. W., and Roscnfeld, M. G. Anterior pituitary development: short tales fromdwarf mice. Cell, 70: 527-530, 1992.

4. Gehring, W. J. The homcobox in perspective. Trends Biochem. Sci., 17: 277-280,1992.

5. Verrijzer, C. P.. and Van der Vliet. P. C. POU domain transcription factors. Biochem.Biophys. Acta, 1173: 1-21, 1993.

6. Weintraub, H., Davis, R., Tapscott, S., Thayer, M., Krause. M., Benezza, R., Black-well, D., Turner, D.. Rupp, R.. Hollemberg, S., Zhuang, Y., and Lassar, A. The tmo

4748

Research. on October 6, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

TTF-1 AND PAX-8 EXPRESSION IN THYROID NEOPLASMS

D gene family: nodal point during specification of the muscle cell lineage. Science(Washington DC), 251: 761-766, 1991.

7. Guazzi, S„Price, M., De Felice, M., Damante, G., Matici, M. G., and Di Lauro, R.Thyroid transcription factor 1 (TTF-1) contains a homeodomain and displays a novelDNA-binding specificity. EMBO J., 9: 3631-3639, 1990.

8. Plachov, D., Chowdhury, K., Walther, C, Simon, D., Guenet, J. L., and Grass, P.Pax8, a murine paired box expressed in the developing excretory system and thyroidgland. Development (Camb.), 110: 643-651, 1990.

9. Francis-Lang, H., Price, M., Policarpou-Schwarz, M., and Di Lauro, R. The promoterof thyroperoxidase gene indicates common mechanisms for thyroid-specific geneexpression. Mol. Cell. Biol., 12: 576-588, 1992.

10. Zaniiini, M., Francis-Lang, H., Plachow, D., and Di Lauro, R. Pax-8, a paireddomain-containing protein, binds to a sequence overlapping the recognition site of ahomeodomain and activates transcription from two thyroid-specific promoters. Mol.Cell. Biol., 12: 4230-4241, 1992.

11. Strahl, G., Strahl, K., and Macdonald, P. M. The gradient morphogen bicoid is aconcentration-dependent transcriptional activator. Cell, 57: 1259-1273, 1989.

12. Strahl, G., Johnston, P., and Lawrence, P. A. Control of drosophila body pattern bythe hunchback morphogen gradient. Cell, 69: 237-249, 1992.

13. Williams, E. D. The aetiology of thyroid tumors. Clin. Endocrinol. Metab., 8:193-207, 1979.

14. Brabant, G., Maenhaut, C., Korle, J., Scheumann, G„Dralle, H., Hoang-Vu, C.,

Hesch, R. D., von zur Mühlen,A., Vassart, G., and Dumont, J. E. Human thyrotropinreceptor gene: expression in thyroid tumors and correlation to markers of thyroiddifferentiation and dedifferentiation. Mol. Cell. Endocrinol., 82: R7-R12, 1991.

15. Mazzaferri, E. L. Treatment of carcinoma of follicular epithelium. In: S. H. Ingbarand L. E. Bravermann (eds.), The Thyroid. A Fundamental and Clinical Text, Ed. 5,pp. 1329-1348. Philadelphia: J. B. Lippincott Company, 1986.

16. Cady, B., Sedgwick, C. F., Meissner, W. A., Bookwalter, J. R., Romagosa, V., andWeber, J. Changing clinical, pathologic, therapeutic, and survival patterns in differentiated thyroid carcinoma. Ann. Surg., 184: 541-543, 1976.

17. Hedinger, C. E. Histológica! typing of thyroid tumors. In: C. E. Hedinger (ed.).International Histológica! Classification of Tumors, Vol. 11. Berlin: Springer-Verlag,185-190, 1988.

18. Francis-Lang, H., Zannini, M., De Felice, M., Berlingieri, M. T., Fusco, A., and DiLauro, R. Multiple mechanisms of interference between transformation and differentiation in thyroid cells. Mol. Cell. Biol., 12: 5793-5800, 1992.

19. Chomczynski, P., and Sacchi, N. A single step method of RNA isolation by acidguanidinium thiocyanate-phenol chloroform extraction. Anal. Biochem., 762:156-159, 1987.

20. Church, G. M., and Gilbert, W. Genomic sequencing. Proc. Nati. Acad. Sci. USA, 81:1991-1995, 1984.

21. Lazzaro, D., Price, M., De Felice, M., and Di Lauro, R. The transcription factor

TTF-1 is expressed at the onset of thyroid and lung morphogenesis and in restrictedregions of the foetal brain. Development (Camb.), 113: 1093-1104, 1991.

22. Heldin, N. E., and Westermark, B. The molecular biology of the human anaplasticcarcinoma cell. Thyroidol. Clin. Exp., 3: 127-131, 1991.

23. Avvedimento, V. E., Musti, A. M., Ueffing, M., Obici, S., Gallo, G., Sanchez, M.,DeBrasi, D., and Gottesman, M. E. Réversibleinhibition of a thyroid-specific transacting factor by raj. Genes Dev., 5.- 22-28, 1991.

24. Damante, G., Chazenbalk, G., Russo, D., Rapoport, B., Foli, D., and Filetti, S.Thyrotropin regulation of thyroid peroxidase mRNA levels in cultured rat thyroidcells: evidence for the involvement of a non-transcriptional mechanism. Endocrinology, 124: 2889-2894, 1989.

25. Justin, E. R., Seabold, J. E., Robinson, R. A., Walker, W. P., Gurll, N. J., and Hawes,D. R. Insular carcinoma: a distinct thyroid carcinoma with associated ¡odine-131localization. J. Nucí.Med., 32: 1358-1363, 1991.

26. Papotti, M., Botto Micca, F., Pavero, A., Palestini, N., and Bussolari, G. Poorlydifferentiated thyroid carcinoma with primordial cell component. Am. J. Surg.Pathol., 17: 291-301, 1993.

27. Lemoine, N. R., Mayall, E. S., Wyllie, F. S., el al. High frequency of ras oncogeneactivation in all stages of human thyroid tumorigenesis. Oncogene, 4: 159-164,1988.

28. Suarez, H. G., du Villard, J. A., Severino, M., Schlumberger, M., and Monier, R.Presence of mutations in all three ras genes in human thyroid tumors. Oncogene, 5:565-570, 1990.

29. Suarez, H. G., du Villard J. A., Caillou, B., Schlumberger, M., Parmentier, C., andMonier, R. Gsp mutations in human thyroid tumors. Oncogene, 6: 677-679, 1991.

30. Dobashi, Y., Sakamoto, A., Sugimura, H., and Machinami, R. Overexpression of p53as a possible prognostic factor in human thyroid carcinoma. Am. J. Surg. Pathol., 17:375-381, 1993.

31. Fagin, J., Matsuo, K., Karmakar, A., et al. High prevalence of mutations of the p53gene in poorly differentiated human carcinoma. J. Clin. Invest., 91: 179-184, 1993.

32. Di Renzo, M. F., Olivero, M., Ferro, S., Prat, M., Borganzone, I., Pilotti, S., Belfiore,A., Costantino, A., Vigneri, R., Pierotti, M. A., and Comoglio, P. M. Overexpressionof c-Met/HGF receptor gene in human thyroid carcinomas. Oncogene, 7: 2549-2553,

1992.33. Haugen, D. R. F., Akslen, L. A., Varhug, J. E., and Lillehaug, J. R. Expression of

c-erb-2 protein in papillary thyroid carcinomas. Br. J. Cancer, 65: 852-857, 1992.

34. Santoro, M., Carlomagno, F., Hay, I. D., et al. Rei oncogene activation in humanthyroid neoplasms is restricted to the papillary cancer subtype. J. Clin. Invest., 89:1517-1522, 1992.

35. Vannelli, G. B., Barni, T., Modigliani, U., el al. Insulin-like growth factor-I receptorsin nonfunctioning thyroid nodules. J. Clin. Endocrinol. Metab., 71: 1175-1182, 1990.

36. Schimitzu, T., Usuda, N., Yamanda, T., and lida, F. Proliferative activity of humanthyroid tumors evaluated by proliferating cell nuclear antigen/cyclin immunohisto-chemical studies. Cancer (Phila.), 71: 2807-2812, 1993.

4749

Research. on October 6, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

1994;54:4744-4749. Cancer Res   Dora Fabbro, Carla Di Loreto, Carlo Alberto Beltrami, et al.   PAX-8 in Human Thyroid NeoplasmsExpression of Thyroid-specific Transcription Factors TTF-1 and

  Updated version

  http://cancerres.aacrjournals.org/content/54/17/4744

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/54/17/4744To request permission to re-use all or part of this article, use this link

Research. on October 6, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from