THE JOURNAL B 1991 by The American Society for Biochemistry and

OF BIOLOGICAL CHEMISTRY Molecular Biology, Inc

Vol. 266. No. 36. Issue of December 25, PP, 24627-24631,1991 Printed in U. S A .

A Novel Calcitonin Carboxyl-terminal Peptide Produced in Medullary Thyroid Carcinoma by Alternative RNA Processing of the Calcitonin/ Calcitonin Gene-related Peptide Gene*

(Received for publication, February 11, 1991)

Stephane MinvielleS, Segolene Giscard-Dartevelle, Regis Cohen, Jacqueline Taboulet, Frederic Labye, Annick Jullienne, Pierre Rivaille, Gerard Milhaud, Mohsen S. Moukhtar, and Francoise Lasmoles From the Znstitut National de la Santi et de la Recherche Medicale U113 and the Centre National de la Recherche Scientifique Unit6 de Recherche Associee 163, Centre Hospitals Universitaire Saint Antoine 27 rue Chaligny, 75012 Paris, France

The calcitonin/calcitonin gene-related peptide gene expresses two different mRNAs by tissue-specific al- ternative processing. The calcitonin mRNA is pro- duced in thyroid C cells by splicing of the first three exons to the fourth polyadenylated exon. It encodes a protein precursor containing an amino-terminal pep- tide, calcitonin, and a carboxyl-terminal peptide (CCP I). Calcitonin gene-related peptide (CGRP) mRNA is produced in neuronal cells by splicing of the three common exons to the fifth exon and the polyadenylated sixth exon, leading to the production of a CGRP pre- cursor. Our studies concerning the expression of the calcitonin/CGRP gene in human medullary thyroid carcinoma revealed the presence of a new RNA tran- script. Amplification by polymerase chain reaction and direct sequencing showed that the novel transcript is composed of exons 1,2, and 3, part of exon 4, exon 5 , and a polyadenylated exon 6. This transcript contains an open reading frame coding for the known amino- terminal and calcitonin peptides, as well as for a novel calcitonin carboxyl-terminal peptide, CCP 11. This third alternative pathway utilizes an internal donor site within the exon coding for calcitonin. The presence of CCP I1 was demonstrated in plasma and thyroidal tissues of medullary thyroid carcinoma patients, im- plying that this novel mRNA is actively translated in medullary thyroid carcinoma.

Alternative RNA processing is one of the multiple strategies used by cells to generate protein diversity (1, 2). The process- ing of the primary transcript of the calcitonin/CGRP’ gene can generate two distinct mRNAs (3). Calcitonin mRNA is mainly produced by the thyroid C cells and contains the first four exons with a polyadenylation site at the end of the exon 4. CGRP mRNA contains the sequence of the first three exons, exon 5 and the polyadenylated sixth exon. This latter

* This work was supported by Grant 6346 from the Association pour la Recherche sur le Cancer. 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 accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence($ reported in thispaper has been submitted

M64486. to the GenBankTM/EMBL Data Bank with accession number(s)

$ To whom reprint requests should be sent. The abbreviations used are: CGRP, calcitonin gene-related pep-

tide; CCP, calcitonin carboxyl-terminal peptide; kb, kilobase(s); bp, base pair(s); MTC, medullary thyroid carcinoma; PCR, polymerase chain reaction.

mRNA is predominantly expressed in the nervous system. The levels of CGRP mRNA represent less than 5% of the levels of calcitonin mRNA in the human medullary thyroid carcinoma (MTC) (4).

The molecular mechanisms involved in the regulation of the alternative processing of calcitonin/CGRP transcript are currently being investigated (5-8). Analyses of the processing pathways of the human calcitonin/CGRP primary transcript show that the last processing reactions leading to the calci- tonin mRNA are splicing of exon 3 to exon 4 and polyaden- ylation of exon 4. An intermediate RNA form of 3.3 kb observed in MTC represents the accumulation of products before the rate-limiting step in the processing pathway (4). Further studies from the same group suggest that the splicing of exon 3 to exon 4 is a slow process due to an unusual branch acceptor for calcitonin splicing. This acceptor could act as a cis-acting element involved in the regulation of the tissue- specific alternative splicing of the human calcitonin/CGRP transcript (5). Nevertheless, in vivo studies on the expression of rat calcitonin/CGRP gene suggest that the use of a poly(A) site is not a key regulatory event for the alternative RNA processing and that tissue-specific splicing is regulated by an alternative splice site selection. These results are compatible with a model in which cis-active sequences near the calcitonin acceptor site inhibit the production of calcitonin mRNA in a CGRP-producing cell line (6,7). In vitro studies using nuclear extracts from HeLa cells suggest that a specific binding factor interacts with regulatory sequences located in the first 45 bases of exon 4, thereby promoting recognition of the calci- tonin-specific 3”splice junction (8).

In this paper, we report our investigations on the expression of the calcitonin/CGRP gene in human MTC. Northern blot experiments revealed the presence of a new RNA transcript. In order to identify and characterize this transcript, the cDNA generated from RNA isolated from MTC was amplified by the polymerase chain reaction (PCR). Sequence analysis re- vealed that this transcript encodes a calcitonin precursor containing a new calcitonin carboxyl-terminal peptide, re- ferred to as CCP 11. It differs from CCP I by the last eight amino acids. We have synthesized this octapeptide, used it to produce specific antibodies, and detected its presence in thy- roid tissues and in plasma of MTC patients. Our results thus establish the existence of another splicing pathway that could result in the production of calcitonin and raises the question of its physiological significance.

24627

24628 A Novel Calcitonin Precursor

FIG. 1. Northern blot analysis of RNA extracted from hu- man MTC. Total RNA (15 pg) was hybridized to the exon 5 probe. Sizes were estimated using EcoRI and Hind111 digests of X DNA as molecular weight markers (Boehringer Mannheim).

A 1 2

Amplification of the first cDNA strand from poly(A)-rich RNAs extracted from human medullary carcinoma tissues using specific primers located in exons 1 and 6 demonstrated the existence of two DNA bands of 878 and 706 bp (Fig. 2A). The same two bands were observed after amplification of mRNA isolated from normal C cells (Fig. 2B). The 706-bp band corresponded to the expected size of CGRP mRNA amplified with these primers, whereas the 878-bp band ap- peared to be a novel transcript of the calcitonin/CGRP gene. DNA sequencing confirmed that the 706-bp product encodes the CGRP precursor already described (18). The sequence of the 878-bp band revealed the primary structure of a new calcitonin/CGRP gene transcript. This alternative mRNA encodes a second calcitonin precursor similar in size and in sequence to the known calcitonin precursor (19), with the exception of the last eight amino acids of the calcitonin carboxyl-terminal peptide (CCP). Cleavage of this precursor would liberate (in addition to the amino-terminal peptide and calcitonin) a novel calcitonin carboxyl-terminal peptide (CCP 11) (Fig. 3A). Both CCP I and I1 have the same length of 21 amino acids, which are identical for the first 13 amino acids but completely different in the eight carboxyl-terminal amino acids (Fig. 3B).

Comparison of the sequence of the 878-bp species with the DNA genomic sequence (20) demonstrated a new pathway of alternate splicing of the calcitonin/CGRP gene. In the novel transcript, only one part of exon 4, that we call exon 4', is joined to exon 5. The internal splice site that is used is located 172 nucleotides downstream from the beginning of exon 4. This donor site (CAT/GTTAGC) is in a good agreement with

B

FIG. 2. A, ethidium bromide stainingof PCR-amplified calcitonin/ CGRP transcripts electrophoresed on a 1% NuSieve, 1% SeaPlaque (FMC Corp.) gel. Estimated sizes are given in base pairs. Amplifica- tion was obtained as described under "Experimental Procedures." Lane I , 25 pl of the PCR reaction using reverse transcribed RNA extracted from MTC. Lane 2, molecular weight markers, BglI and HinfI digests of pBR328 (Boehringer Mannheim). B, autoradiogram of Southern blot performed on PCR products of human transcripts of calcitonin/CGRP gene hybridized with the exon 5 probe. Lane I , 25 pl of the PCR-amplified cDNA generated from total RNA ex- tracted from normal C cells. Lane 2, 2.5 p l of the PCR reaction using RNA extracted from MTC.

EXPERIMENTAL PROCEDURES~

RESULTS

Isolation and Direct Sequencing of the Amplified Products- Northern blot analysis of total RNA extracted from human medullary carcinoma tissues showed the presence of a 1.4-kb mRNA species which hybridized specifically with an exon 5 probe in addition to the 1.2 kb mature form of CGRP mRNA. An upper band that probably corresponds to the common precursor of the calcitonin and CGRP mRNAs was also observed (Fig. 1). Analysis indicated that there was approxi- mately twice as much of the 1.4-kb mRNA species than the 1.2-kb species.

* Portions of this paper (including "Experimental Procedures" and Fig. Sl) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

B

RMAN CCP I1 DWSSDLBRDERPANBCPEBLS

ccp 1 """"""_ VSIPQNAN

C

FIG. 3. A, sequence of human CCP I1 cDNA. The predicted amino acid sequence of the precursor is shown under the nucleotide se- quence. The cleavage sites are underlined. The sequence of the CCP I1 is boxed; the arrowhead indicates the point of divergence of the CCP I and CCP I1 precursors. B, predicted sequence of CCP I1 compared with CCP I. C, representation of both possible splice junctions in exon 5. In the first case, the arrangement leads to a CGRP precursor molecule, and in the second case, the exon 5 codes for the eight carboxyl-terminal amino acids of the CCP 11.

A Novel Calcitonin Precursor 24629

the consensus 5"splice site sequence (21,22). This new splic- ing changes the open reading frame of exon 5 (Fig. 3C).

CCP I I Radioimmunoassay-The carboxyl-terminal pep- tide fragment (eight amino acids) containing the predicted carboxyl-terminal sequence of CCP I1 (8-CCP 11) was synthe- sized and conjugated to tetanus toxoid to produce specific antibodies in mice. Antisera with titers greater than 1/1500 were obtained. The antibody did not cross-react with CCP I or calcitonin (see Fig. S1 in the Miniprint). We have thus established using these antibodies and '"I-labeled 8-CCP 11, a specific and sensitive radioimmunoassay.

Immunocytochemistry-The presence of CCP I1 in human MTC tissues was detected by specific immunostaining. No staining was observed if the antibodies were saturated with 8-CCP I1 (Fig. 4). Intensity of the specific staining was not affected if the antibodies were saturated with CCP I, calci- tonin, or tetanus toxoid (data not shown).

Measurement of CCP I I in Plasma and Thyroid Tissues from MTC Patients-Plasma and tissues CCP I1 immuno- reactive molecules were purified and concentrated using spe- cific affinity chromatography. We successfully measured cir- culating CCP I1 (0.25 ng/ml) in a MTC patient, whereas the peptide was at the limit of detection in the plasma of a normal subject. A major form of immunoreactive CCP I1 (-6000 Da) was detected in MTC extracts chromatographed on Sephadex G-50.

DISCUSSION

The 1.4-kb RNA species, demonstrated by Northern analy- sis of MTC RNA using exon 5 probe from the calcitonin/ CGRP gene, was 200 bases longer than the mature CGRP mRNA and could not be accounted for by the established splicing patterns of the primary transcript of the gene. We therefore amplified by PCR the single-stranded DNA comple- mentary to the mRNA species, using primers located in exon 1 and in exon 6, in order to avoid amplification of the known calcitonin mRNA (exons 1-4). Sequence analysis showed that this transcript contained, in addition to exons 1-3, an incom- plete exon 4 joined to exon 5, suggesting a third pathway of calcitonin/CGRP RNA processing, resulting in the produc- tion of a second calcitonin mRNA in thyroid tissues.

Fig. 5 schematically illustrates the complex alternative RNA processing pathways involved in the expression of the calcitonin/CGRP gene based upon the possible exon-intron arrangements of the gene and the primary structures of the calcitonin and CGRP precursor mRNAs. Splicing of the first three exons is invariant, but after this step, three pathways can be used. One involves splicing of exon 3 to exon 4 with a poly(A) tail, resulting in the calcitonin precursor previously

FIG. 4. A, CCP I1 immunoreactive C cells in human MTC (X500). B, specificity of the reaction was determined using antibodies satu- rated with 8-CCP I1 peptide (25 gg/ml).

Human calciloninlCGRP gene primary lranrcripl

Translalion and proteolytic ckava8e

FIG. 5. Schematic representation of alternative RNA proc- essing pathways in the expression of the human calcitonin/ CGRP gene. The numbered boxes represent exons; exon 4' indicates the part of the calcitonin coding exon 4 involved in alternative splicing, resulting in calcitonin mRNA 2. Black areas indicate non- coding regions.

described (19). In another one, exons 3,5, and 6 with a poly(A) tail are spliced (23), leading to the CGRP precursor. The third pathway, described here, demonstrates that splicing of exon 3, 4', 5, and 6 with a poly(A) tail results in a calcitonin precursor with a CCP diverging in the last eight carboxyl- terminal amino acids. The junction between exon 4' and exon 5 changes the open reading frame and introduces a nonsense codon, eight amino acids after the beginning of exon 5. This arrangement prevents translation of CGRP. Moreover a splice junction compatible with the production of this mRNA exists in the primary transcript of human calcitonin/CGRP gene, as there is a consensus sequence of donor site in exon 4 (see Fig. 3C).

This alternative pathway is used in normal and neoplastic C cells and seems to be restricted to humans. Comparison of the calcitonin mRNA sequences of rat (24), rabbit (25), chicken (26), and salmon (27) reveals that this alternative donor site does not exist in these species. Control studies using exon 1-6 PCR amplification of cDNA derived from poly(A)-rich RNA purified from rat thyroid tumors failed to detect such a phenomenon.

In order to determine if the new human RNA transcript is translated in a novel precursor protein, we produced antibod- ies specific for CCP I1 that do not cross-react with CCP I and probed MTC tissues for expression of this peptide. The pres- ence in MTC cells of molecules reacting with these CCP I1 antibodies is a direct proof of the expression and the trans- lation of this mRNA. Our preliminary data indicate that the precursor is processed as we detected by molecular sieving immunoreactive molecules of -6000 Da, which could corre- spond to the precursor having lost its amino-terminal peptide (28). Moreover, we detected significant levels of immunoreac- tive CCP I1 in the plasma of MTC patients, indicating that the peptide is secreted by the cells that synthesize it.

These results suggest a new model of calcitonin/CGRP alternative splicing regulation in which the use of a single polyadenylation site in exon 6 can produce calcitonin or CGRP. Therefore, cleavage and polyadenylation at the calci- tonin poly(A) site is not a prerequisite for splicing of exon 3 to exon 4 in human thyroid C cells, as was suggested for

24630 A Novel Calcitonin Precursor

expression of rat calcitonin mRNA (7). Recent in vitro splic- ing studies (29) show that deletion of the intron between exons 4 and 5 results in the utilization of a cryptic 5’-splice site in the exon 4, probably located at the same place as the internal donor site we described in vivo. Production of calci- tonin mRNA is achieved using either cleavage/polyadenyl- ation at the exon 4 poly(A) site, leading to CCP I, or the splicing of exon 4‘ to exon 5, leading to CCP 11. The regulation of this option could involve several mechanisms, such as specific poly(A) site selection or competition between the cleavage/polyadenylation reaction and the splicing reaction, as proposed for immunoglobulin gene (30, 31) or differences in splicing efficiency, as in the case of the rat kininogen genes (32).

In conclusion, we have demonstrated the existence of a second calcitonin mRNA produced by a splicing process avoiding the use of the internal cleavage and polyadenylation site, demonstrating that the exon 6 poly(A) site is not exclu- sively a CGRP mRNA poly(A) site. The use of this splicing pathway is not apparently due to C cell tumorization, as it also occurs in normal thyroid tissue. However, this CCP I1 transcript and CCP I1 peptide are much more abundant in MTC patients than in normal subjects. Thus, this peptide could be used as a diagnostic and perhaps a prognostic marker for MTC. Future studies will be necessary to establish its physiological role, its value as a new tumoral marker, and the processes involved in the regulation of the choice between the cleavage/polyadenylation reaction and the splicing reaction.

Acknowledgments-We are indebted to Claude Desplan for critical reading of the manuscript, to Claude Calmettes, supervisor of the French Medullary Study Group (Groupe #Etude des Tumeurs i Calcitonine), and to Robert Myrtil for his expert technical assistance.

REFERENCES 1. Leff, S. E., Rosenfeld, M. G., Evans, R. M. (1986) Annu. Reu.

Biochem. 55, 1091-1118 2. Breitbart, R. E., Andreadis, A., and Nadal-Ginard, B. (1987)

Annu. Reu. Biochem. 56,467-495 3. Amara, S. G., Jonas, V., Rosenfeld, M. G., Ong, E. S., and Evans,

R. M. (1982) Nature 298 , 240-244 4. Bovenberg, R. A. L., van de Meerendonk, W . P. M., Baas, P. D.,

Steenbergh, P. H., Lips, C. J. M., and Jansz, H. S. (1986) Nucleic Acids Res. 14, 8785-8803

5. Adema, G. J., van Hulst, K. L., and Baas, P. D. (1990) Nucleic Acids Res. 18, 5365-5373

6. Leff, S. E., Evans, R. M., and Rosenfeld, M. G. (1987) Cell 48 , 517-524

7. Emeson, R. B., Hedjran, F., Yeakley, J. M., Guise, J. W., and Rosenfeld, M. G. (1989) Nature 341 , 76-80

8. Cote, G. J., Nguyen, I. N., Berget, S. M., and Gage1 R. F. (1990) Mol. Endocrinol. 4, 1744-1749

9. Itoh, N., Nose, K., and Okamoto, H. (1979) Eur. J . Biochem. 97, 1-9

10. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, pp. 383-389, Cold Spring Har- bor Laboratory, Cold Spring Harbor, NY

11. Soudan, J., Caron de FrBmontel, C., Bernard, O., and Larsen, C. J. (1991) Leukemia, in press

12. Merrifield, R. B. (1963) J. Am. Chem. SOC. 85, 2149-2154 13. Rivaille, P., Gautron, J . P., Castro B., and Milhaud G. (1980)

14. Le-Nguyen D., Heitz, A., and Castro, B. (1987) J. Chem. SOC.

15. Tam, J. P., Heath,, W . H., Merrifield, R. B. (1983) J. Am. Chem.

16. Lane, L. C. (1978) Anal. Biochem. 86, 655-664 17. Hunter, W . M., and Greenwood, F. C. (1962) Nature 194,495-

496 18. Jonas, V., Lin, C. R., Kawashima, E., Semon, D., Swanson, L.

W., Mermod, J. J., Evans, R. M., and Rosenfeld, M. G. (1985) Proc. Natl. Acad. Sci. U. S. A. 8 2 , 1994-1998

19. Le Moullec, J. M., Jullienne, A., Chenais, J., Lasmoles, F., Gu- liana, J. M., Milhaud, G., and Moukhtar, M. s. (1984) FEBS Lett. 167,93-97

20. Broad, P. M., Symes, A. J., Thakker, R. V., and Craig, R. K. (1989) Nucleic Acids Res. 17, 6999-7011

21. Breathnach, R., and Chambon, P. (1981) Annu. Reu. Biochem.

22. Mount, S. (1982) Nucleic Acids Res. 10,459-472 23. Rosenfeld, M. G., Mermod, J. J., Amara, S. G., Swanson, L. W.,

Sawchenko, P. E., Rivier, J., Vale, W. W., and Evans, R. M. (1983) Nature 304, 129-135

24. Amara, S. G., Evans, R. M., and Rosenfeld, M. G. (1984) Mol. Cell. Bid. 4, 2151-2160

25. Martial, K., Minvielle, S., Taboulet, J., Jullienne, A., Milhaud, G., and Lasmoles. F. (1990) Biochem. Biophys. Res. Commun.

Tetrahedron 36,3413-3419

Perkin Trans. Z 1915-1919

SOC. 105,6442-6455

50,349-383

26.

27.

28.

171 , 1111-1114

and Moukhtar, M. S. (1985) EMBO J. 4,2603-2607

and Besemer, J. (1987) FEBS Lett. 226 , 96-100

, . ,

Lasmoles, F., Jullienne, A., Day, F., Minvielle, S., Milhaud, G.,

Poschl, E., Lindley, I., Hofer, E., Seifert, J.-M., Brunowsky, W.,

Birnbaum, R. S., Mahoney, W . C., and Roos, B. A. (1986) J. Biol. Chem. 261,699-703

(1989) Biochim. Biophys. Acta 1008 , 223-233

82,8658-8662

29. Bovenberg, R. A. L., Moen, T. C., Jansz, H. S., and Baas, P. D.

30. Danner, D., and Leder, P. (1985) Proc. Natl. Acad. Sci. U. S. A.

31. Peterson, M. L., and Perry, P. R. (1986) Proc. Natl. Acad. Sci.

32. Kakizuka, A., Kitamura, N., and Nakanishi, S. (1988) J. Biol. U. S. A . 83,8883-8887

Chem. 263,3884-3892

A Novel Calcitonin Precursor 24631

Supplemental Material to: A novel calcitonm C-terminal peptide (CCP) produced m medullary thyrold carcinoma by allemalive RNA pmesslng of the calcitoniKGRP gene

Sttphane Minvielle. Stgollne Giscard-Dartevelle, Regs Cohen. Jacqueline Tabulet. FddCric Labye. h i c k JuhMC, Pierre Rivallle, Gerard Milhaud. Mohsen S Moukhtar and Frangolse Lasmoles

by

EXPERIMENTAL PR0cU)URES

0.8

\ 0 C.lslon1" . CCPI

x W C P I I