Profiling trefoil factor family (TFF) expression in the mouse: identification of an antisense TFF1-related transcript in the kidney and liver

Peptides 25 (2004) 755–762

Profiling trefoil factor family (TFF) expression in the mouse: identification of an antisense TFF1-related transcript in the kidney and liver Silvia C. Hertel a , Caroline E. Chwieralski a , Margitta Hinz a , Marie-Christine Rio b , Catherine Tomasetto b , Werner Hoffmann a,∗ a b

Institut für Molekularbiologie und Medizinische Chemie, Otto-von-Guericke-Universität, D-39120 Magdeburg, Germany Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), 67404 Illkirch Cedex, C.U. de Strasbourg, France Received 3 September 2003; accepted 14 November 2003 Available online 13 May 2004

Abstract The expression of the trefoil factor family (TFF) genes (TFF1, TFF2, and TFF3) was systematically analyzed in 18 different organs from male or female mice using RT-PCR analysis. The expression patterns showed some gender-specific differences, e.g., TFF3 transcripts in the urinary bladder and liver. Furthermore, the murine expression profile differed from that in human, e.g., in the respiratory tract and uterine cervix. As a hallmark, an aberrant TFF1-related transcript was detected specifically in the kidney and liver of several mouse strains. Molecular characterization of this rare 1.8 kb long transcript from the kidney clearly revealed that its 3 region originated from the antisense strand of the TFF1 locus containing particularly large parts of the antisense strands of introns 1 and 2. Homology searches using various databases revealed that this antisense TFF1-related transcript is subject of intense alternative splicing and no protein product encoded by this antisense TFF1-related transcript could be identified. Although the function of this transcript is not known currently, we can speculate that this antisense TFF1-related transcript might have a gene silencing effect particularly on TFF1 expression in the murine kidney and liver. © 2004 Elsevier Inc. All rights reserved. Keywords: Antisense RNA; TFF-domain; Trefoil factor; Gene expression; Gene silencing; TFF1; Kidney; Liver

1. Introduction Trefoil factor family (TFF) peptides (TFF1, TFF2, and TFF3) are characteristic secretory products of the mucous epithelia and are integral constituents of the mucus layer (for reviews, see [11,12,26]. They are essential in maintaining the surface integrity of these delicate epithelia. TFF peptides were shown to interact directly with mucins [22,33] influencing the rheologic properties of the viscous biopolymers [9,30]. TFF1-deficient mice have an altered mucus layer in their antral and pyloric stomach [16,34] and developed antropyloric adenoma with full penetrance suggesting that TFF1 is a tumor supressor gene [16]. Additionally, it has been shown that TFF1 deficiency permanently activates the unfolded protein response in this part of the stomach [34]. Thus, TFF1 could have an additional function within the secretory pathway, e.g., for correct folding or secretion ∗

Corresponding author. E-mail address: [email protected] (W. Hoffmann). 0196-9781/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.peptides.2003.11.021

of mucins. Furthermore, TFF1 was shown to reduce cell proliferation by delaying the G1–S phase transition [2,4]. TFF peptides are capable of triggering intracellular signaling cascades which can cause motogenic, cell scattering, anti-apoptotic, and pro-angiogenic effects as well as modulate inflammatory processes (for reviews, see [11,12]). Currently, CRP-ductin is one candidate for a putative TFF receptor [31]. This protein is now recognized as a product of the DMBT1 gene [14]. TFF peptides are highly expressed in the mouse gastrointestinal tract, TFF1 and TFF2 are secreted by the stomach and TFF3 by the intestine [17,21]. No TFF1 transcripts were detectable in the kidney or liver by Northern blot analysis [17]. The three TFF genes are clustered on mouse chromosome 17 and promotor analysis showed a perfect correlation between the demethylation of their promoter and their expression [25]. In the murine stomach, the expression of the clustered TFF genes seems to depend on each other. For example, 70% of TFF1-deficient mice failed to express gastric TFF2 [16] while TFF3-deficient mice showed reduced expression of both gastric TFF1 and TFF2 [29].

756

S.C. Hertel et al. / Peptides 25 (2004) 755–762

TFF peptides, besides their prominent expression in the gastrointestinal tract, are also synthesized in other mucous epithelia, e.g., the respiratory tract, the uterus, the conjunctiva, and the salivary glands as well as in the central nervous system (for reviews, see [11,12]). In this study, the expression patterns of all three TFFs were systematically investigated in different organs of the adult mouse using RT-PCR analysis. Surprisingly, an antisense TFF1-related transcript was detected specifically within the kidney and liver.

2. Materials and methods 2.1. Animals, RNA extraction and reverse transcription-polymerase chain reaction (RT-PCR) analysis Adult C57BL/6JOIaHsd or BALB/cOlaHsd mice were obtained from Harlan Winkelmann GmbH (Borchen, Germany). Transgenic TFF1(−/−) mice and their wild type control strain 129S2/Sv were described previously [14]. RNA extraction and semi-quantitative RT-PCR analysis (35 cycles) for TFF1, TFF2, or TFF3 expression including the sequences of the synthetic oligonucleotides MD7/MD8, MD5/MD6, and MD3/MD4 has been described in detail [10]. As a control for the integrity of the cDNA preparations, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcripts were amplified in parallel (primer pair: MD1/MD2; [10]). Additional oligonucleotides used were: MB197 d(CATGGAGCACAAGGTGATC), MB198 d(GAATGACGAACACAGAATAAC), MB206 d(CTGCTCTGATGGACTCATC), MB212 d(ACAGAATAGTTGTGGGAAG), MB220 d(TCACCCACCAGGATGAATAG), MB223 d(CCAGTCCGTATCTCTGGTTG), MB267 d(AATGAGCAAACAGTACATCAC), and MB135 d(CCCTCGAGGATCCGAATTCT18 VN). The position of the oligonucleotides within the TFF1 locus is displayed in Fig. 2. 2.2. DNA sequence analysis of RT-PCR products, homology searches RT-PCR products were subcloned using the pGEMR -T Easy Vector System (Promega GmbH, Mannheim, Germany). Plasmid DNA was prepared from transformed Escherichia coli JM109 using a Plasmid Midi Kit (Qiagen GmbH, Hilden, Germany). Commercial DNA sequencing was performed by Invitek (Berlin, Germany) using the PUC/M13 forward or reverse primer, respectively. Homology searches were performed in the “National Center for Biotechnology and Information” (NCBI) and the FANTOM2.00.seq [23] databases using BLASTN [1]. 2.3. Northern blot analysis Total RNA was extracted from murine stomach or kidney as described [10]. Poly(A)+ RNA was isolated using

the Micro-Fast TrackTM 2.0 Kit (Invitrogen, Karlsruhe, Germany). RNA blot analysis on HybondTM -N membranes (Amersham Biosciences Europe GmbH, Freiburg, Germany) was similar as described [6]. Typically, 25 ␮g poly(A)+ RNA per lane were loaded onto the 1% denaturing agarose gel (20 cm × 19 cm). Electrophoresis was for 5 h at 80 V. As a marker, a RNA standard ladder was used (Invitrogen). Hybridization probes were generated by RT-PCR, purified with the QIAquick PCR Purification Kit (Qiagen), and labeled with [␣-32 P]dCTP (Amersham Biosciences Europe GmbH) using a Nick Translation Kit (Roche Diagnostics GmbH, Mannheim, Germany). After prehybridization, hybridization, and washing steps, the membranes were exposed to Kodak Biomax MS film at −80 ◦ C. 2.4. Protein extraction, Western blot analysis Protein extraction of murine tissues and Western blot analysis were performed under reducing conditions as reported [13]. Protein concentrations of the samples loaded onto the gel were monitored by the Bio-Rad Protein Assay (20 ␮g per lane). The following polyclonal antisera were used: Ab502 (in a 1:1000 dilution) against the 16 C-terminal amino acid residues of mouse TFF1 [16], antiserum against the 16 C-terminal amino acid residues of mouse TFF2 (1:1000 dilution) [32], and affinity-purified anti-hTFF3-2 (1:500 dilution) [37]. 3. Results 3.1. TFF expression pattern in the adult mouse Organs from adult male or female C57BL/6 mice were dissected and systematically analyzed for their TFF1, TFF2, and TFF3 mRNA contents by RT-PCR analysis (Fig. 1). The 18 organs dissected include various brain regions, heart, lung, spleen, various regions of the gastrointestinal tract, urinary tract, and reproductive tract. The size of the RT-PCR products was as expected in all samples, except the kidney and liver. Surprisingly, an aberrant band of an unexpected size (about 0.8 kb) was detectable with the TFF1-specific primer pair MD7/MD8 as the major RT-PCR product in these two organs. The expression of the three TFF genes in the mouse is regulated differently. Unlike using Northern blot analysis, RT-PCR analysis revealed that TFF1 transcripts are present in nearly all tissues investigated. TFF2 expression showed a more restricted pattern with the highest level in the stomach. The highest TFF3 transcript levels were detected in the entire gastrointestinal tract starting in the esophagus. Interestingly, gender-specific differences in the TFF3 transcript levels were monitored in the urinary bladder (compare Figs. 1A and B) as well as the liver (Fig. 1C; representative experiment of three single male or three single female animals, respectively).

S.C. Hertel et al. / Peptides 25 (2004) 755–762

757

3.2. Partial molecular characterization of the aberrant TFF1-related transcript from the murine kidney

Fig. 1. RT-PCR analysis (35 cycles). TFF1 (primers: MD7/MD8), TFF2 (primers: MD5/MD6), TFF3 (primers: MD3/MD4), or antisense TFF1 expression (primers: MB267/MB206) was monitored in different organs from a single adult female (A) or male C57BL/6 mouse (B). Furthermore, the liver, spleen, or kidney was analyzed from a single male (m) or female (f) C57BL/6 mouse (C). Total RNA from murine stomach or duodenum was analyzed as positive controls for TFF1, TFF2, or TFF3 transcripts, respectively (lane c). The integrity of the oligo(dT)12–18 -primed cDNAs was tested by monitoring the GAPDH transcripts. The molecular size standard (bp) is shown left. B, brain lacking cerebellum and cerebral cortex; C, colon; Cb, cerebellum; Co, cerebral cortex; Cv, uterine cervix; D, duodenum; E, esophagus; En, endometrium; H, heart; K, kidney; L, lung; Li, liver; R, rectum; S, stomach; Sp, spleen; T, testicles including ejaculatory ducts; U, urinary bladder; and Ur, ureter.

Molecular cloning and sequence analysis of the aberrant 0.8 kb RT-PCR product (TFF1-specific primers: MD7/MD8) from a C57BL/6 kidney revealed clearly a relation to the TFF1 gene (Figs. 2 and 3). By alignment with the TFF1 genomic sequence we found that the transcript starts as expected within exon 2 and terminates correctly within exon 3. Surprisingly, the RT-PCR product contains also the 5 part of intron 2, but it is lacking the 3 part of intron 2 as well as the 5 part of exon 3. Thus, the kidney-specific RT-PCR product represents a novel transcript related to the TFF1 gene locus and lacking 581 bp of genomic sequence. The existence of similar TFF1-related transcripts was confirmed by finding of several ESTs originating from the kidney or liver (Fig. 2). Surprisingly, these ESTs did not contain the complete 3 end of exon 3 (match only until position 5818; Fig. 3) in spite of this, they continued downstream of exon 3. Furthermore, these regions did not match the genomic sequences following exon 3 (see Fig. 2). A similar RT-PCR product was also generated from murine kidney cDNA using the oligonucleotides MB220/MB135 as primers (Fig. 2). The aberrant RT-PCR product (primers: MD7/MD8) from the kidney is lacking 581 bp when compared with the genomic sequence (Figs. 2 and 3). Such a transcript cannot be generated by splicing of a TFF1 pre-mRNA because the intron consensus sequences [3] are lacking. However, the antisense strand would contain the splice consensus sequences necessary to allow correct splicing of this 581 bp long region. Thus, RT-PCR analysis was performed with different mouse kidney cDNAs either primed with oligonucleotides complementary to the TFF1 sense strand (e.g., MB206 or MD8; Fig. 2) or the TFF1 antisense strand (e.g., MB267 or MB223; Fig. 2) in order to determine the orientation of this aberrant TFF1-related transcript (amplification was with the primer combinations MD7/MD8 or MB220/MD8 or MD7/MB212). In every case only the MB267- or MB223-primed cDNAs generated the expected RT-PCR products (data not illustrated). This indicated clearly that the transcription of this TFF1-related gene in the kidney occurs on the reverse strand than the regular TFF1 pre-mRNA. Thus, this transcript is now termed “antisense TFF1-related”. Furthermore, RT-PCR analysis of oligo(dT)12–18 -primed cDNAs from 18 different murine organs with MB267 (located within intron 1; see Fig. 2) and MB206 (see Fig. 2) generated a 1.045 kb band only in the kidney and liver and very scarcely also in the stomach (termed as “TFF1” in Fig. 1). This indicates that the antisense TFF1-related transcript in the murine kidney and probably the liver also contains sequences from intron 1. This result is in full agreement with ESTs BB667664 and AK050258 (Fig. 2). In another set of RT-PCR experiments it was tested whether the antisense TFF1-related transcript in the kidney extends into exon 1 of the TFF1 gene. The primer combination MB197

758

S.C. Hertel et al. / Peptides 25 (2004) 755–762

Fig. 2. Schematic comparison of the mouse TFF1 genomic region (as deposited to the EMBL database with accession number AJ271002) and various murine kidney cDNA clones. The three TFF1 exons as determined previously [22] as well as the exons encoding the antisense TFF1-related transcript are hatched. Transcribed regions were identified from RT-PCR products obtained with MD7/MD8 or MB220/PCR3 as well as from mouse ESTs (GenBank accession numbers: BG969460 from kidney, BB502343 from neonate kidney, BB667664 from a liver tumor, and AK085535 from neonate kidney) and the FANTOM2 cDNA annotation system (DDBJ accession number AK050258 from liver). Also indicated are the positions of the different primers used for RT-PCR analysis.

(located at the 5 end of exon 1; see Fig. 2) and MD8 did not result in an enlarged RT-PCR product using kidney cDNA as a template (Fig. 4). This indicates clearly, that the antisense TFF1-related transcript in the kidney does not contain sequences from the 5 end of exon 1. In a control experiment, this combination gave the expected products with stomach cDNA. Interestingly, a stomach-specific RT-PCR product could be generated also with MB197/MB198 (Fig. 4; for position of these primers, see Fig. 2) indicating that the 3 end of the TFF1 transcript is longer than expected probably using alternative polyadenylation sites. A Northern blot analysis of poly(A)+ -RNA from the stomach or kidney of C57BL/6 mice revealed a band at about 1.8 kb for the antisense TFF1 transcript in the murine kidney Fig. 5). However, this is certainly a rare transcript when compared with the TFF1 mRNA level in the stomach (Fig. 5A). This transcript is enriched in the kidney, but is also present at very low levels in the stomach Figs. 1 and 5B), the latter being confirmed by cDNA cloning. RT-PCR analysis (TFF1-specific primers: MD7/MD8) of kidney cDNAs from different mouse strains clearly revealed that the antisense TFF1 transcript is present in all three of the wild type strains investigated (Fig. 4). The kidneys of these strains also contained variable minor amounts of the regular TFF1 transcript (individual differences between different animals of the same strain). In contrast, an enlarged antisense TFF1-related transcript was hardly detectable in kidneys of TFF1-deficient mice (insertion mutants; [16]);

whereas an enlarged TFF-related transcript was easily detectable in the stomach. 3.3. Western blot analysis of murine kidney and stomach extracts TFF1, TFF2, and TFF3 peptides are clearly not detectable in the kidney of three different mouse strains (Fig. 6). There is also a lack of TFF1–3 synthesis in the kidney of the TFF1-deficient mice. However, the stomachs of BALB/c, C57BL/6 and 129S2/Sv mice secrete all three TFF peptides, whereas in the stomach of TFF1(−/−) knock out mice not only TFF1 is missing, but also TFF2 and TFF3 (Fig. 6). 4. Discussion 4.1. TFF expression pattern in the adult mouse Taken together, the TFF expression profile in the mouse differs in many respects from the situation in human. For example, the human respiratory tract expresses mainly TFF3 and little TFF1 and is lacking TFF2 transcripts [37]. This is in contrast to the murine lung where TFF2 seems to be the predominantly expressed TFF gene. A similar discrepancy was observed for the female reproductive tract (human uterine cervix: mainly TFF3 [36]; murine uterine cervix: mainly TFF1). The reasons for these species-specific differ-

S.C. Hertel et al. / Peptides 25 (2004) 755–762

759

Fig. 3. Nucleotide sequence of part of the murine TFF1 gene (as deposited to the EMBL database with accession number AJ271002). Exons 2 and 3 as determined previously [22] are enclosed in boxes. The sequences, which are missing in the RT-PCR products obtained with the primer combinations MD7/MD8 or MB220/PCR3, are shown in lower case letters (precisely corresponding to the introns of the antisense TFF1-related transcript in Fig. 2). Potential polyadenylation sites are underlined. The positions of the different primers used for RT-PCR analysis are shown in bold.

ences are not yet known. However, these results are in line with previous reports on dramatic differences in the TFF2 synthesis in human and murine pancreas [7,16]. Generally, the species differences observed might indicate that the TFF peptides differ in their biologic properties and each species has adapted the optimal TFF peptide for its own specific physiologic needs. Furthermore, TFF3 expression was clearly demonstrated in the stomach by RT-PCR analysis (Fig. 1). This result was confirmed also on protein level for three different mouse strains (Fig. 6). This fact was not recognized in the past but is in line with a previous report on TFF3 expression in the rat stomach [5] as well as very recent findings in the human stomach, where TFF3 appeared to be typically secreted in the antrum and pylorus [15]. Interestingly, the

TFF1-deficient mice were lacking gastric TFF3 synthesis (Fig. 6) which agrees with reports that expression of the clustered TFF genes is coordinately regulated in the murine stomach [16,29]. However, the striking finding was the detection of an aberrant RT-PCR product in the kidney and liver when using TFF1-specific primers (Fig. 1). 4.2. A TFF1-related antisense RNA is generated in the murine kidney and liver Molecular characterization of the aberrant RT-PCR product from murine kidney clearly suggested that this transcript originates from the minus strand of the TFF1 locus. Indeed, the plus strand did not contain splice consensus sequences [3] that could lead to this transcript; whereas all the cor-

760

S.C. Hertel et al. / Peptides 25 (2004) 755–762

Fig. 4. RT-PCR analysis. TFF1 and antisense TFF1 expression was monitored in the stomach (S) or the kidney (K) of the following mouse strains: BALB/c, C57BL/6, 129S2/Sv, and TFF1(−/−). The primer combinations used were: MD7/MD8, MD267/MD206 (30 cycles: 30 s at 55 ◦ C, 1 min at 72 ◦ C, 30 s at 94 ◦ C), MB197/MD8 (25 cycles: 45 s at 55 ◦ C, 30 s at 72 ◦ C, 45 s at 94 ◦ C), or MB197/MB198 (30 cycles: 1 min at 58 ◦ C, 45 s at 72 ◦ C, 1 min at 94 ◦ C). Additionally, a somewhat extended amplification cycle was used for the amplification of the antisense TFF1-related transcript with MD7/MD8 (lane K∗ ; 35 cycles: 30 s at 60 ◦ C, 2 min at 72 ◦ C, 30 s at 94 ◦ C). The integrity of the oligo(dT)12–18 -primed cDNAs was tested by monitoring the GAPDH transcripts (MD1/MD2). The molecular size standard (bp) is shown left.

rect splice sites were present on the minus strand (three splice junctions between positions 5169/5170, 5750/5751, and 5818/5819, respectively; Fig. 3). Such a transcript was detectable in the kidneys of five different mouse strains (BALB/c, C57BL/6, and 129S2/Sv: Fig. 4; NMRI and TFF3-deficient [21] mice: data not illustrated) as well as the murine liver. Homozygous TFF1-deficient mice, which had the TFF1 gene inactivated by insertion of the neomycin resistance gene into exon 2 [16], expressed an equivalent TFF1-related sequence in the kidney only scarcely (lane K∗ in Fig. 4); whereas an enlarged TFF1 transcript resulting from the insertion of the neomycin resistance gene casette within exon 2 [16] was detectable in the stomach of these animals as expected. Thus, insertion of the neomycin resistance gene within exon 2 of the TFF1 gene reduced expression of the antisense TFF1-related transcript in the kidney. As displayed in Fig. 2 the sequence of this novel transcript overlaps with various mouse ESTs generated from the kidney (GenBank accession number BG969460, BB502343, AK085535) or a liver tumor (GenBank accession number BB667664). Furthermore, a FANTOM2 cDNA from the

liver (DDBJ accession number AK050258; [23]) is partially congruent with our sequence. A comparison of all these sequences revealed that they differ mainly by relatively short deletions/insertions at the 5 portion which are expected to be the result of intense alternative splicing. However, a precise analysis of the exon–intron structure is currently not possible because the corresponding genomic sequences are not available. The length of the antisense TFF1-related transcript is about 1.8 kb as determined by Northern blot analysis (Fig. 5). Thus, the portion known spans the region between the 5 end of EST AK085535 and the 3 end of EST BB667664, i.e., about 1.7 kb not including a poly(A)-tail. This suggests that the known portion of the antisense TFF1 transcript represents about the full length sequence. The sequence AGUAAA might function as a polyadenylation signal which is located 13 bases from the 3 end and represents a variant of the canonical hexameric signal [18]. Thus, so far no protein encoded by the antisense TFF1 transcript could be identified unambiguously. There is a relatively short open reading frame starting at position 126 from EST BB667664 which could encode a peptide of 112

S.C. Hertel et al. / Peptides 25 (2004) 755–762

761

of the 2431 pairs of sense–antisense transcripts identified recently by analysis of the mouse transcriptome [23,28]. The function of the antisense TFF1 transcript is not known currently. However, there is increasing evidence that antisense transcripts are important regulators of gene expression not only in vitro but also in vivo [8,19,27]. For example, gene silencing by antisense RNA has been reported recently to cause a human genetic disease [35]. The observation that the TFF1 mRNA level varies in single animals of the same strain (Fig. 4) might be a first indication for regulation via antisense TFF1. Thus, it will be challenging to study the regulation of antisense TFF1 expression in the murine kidney and liver. Furthermore, 44% of gastric carcinomas showed a loss of TFF1 expression [20,24]. Thus, it would be interesting to test the possibility that TFF1 gene silencing is due to the presence of a similar antisense TFF1-related transcript in these cancers.

Acknowledgments

Fig. 5. Northern blot analysis of poly(A)+ RNA from pooled stomachs (S) or kidneys (K) of five male and five female C57BL/6 mice. (A) Hybridization with the 32 P-labeled RT-PCR product generated with the primer combination MB223/MB206 (see Fig. 2; template: oligo(dT)12–18 -primed murine kidney cDNA). (B) Hybridization with the 32 P-labeled RT-PCR product generated with the primer combination MB223/MB212 (see Fig. 2; template: oligo(dT)12–18 -primed murine kidney cDNA). The RNA size standard (kb) is shown left.

amino acid residues. However, this deduced sequence did not show similarity with any known protein. It is tempting to speculate that this transcript possibly belongs to the class of spliced functional non-coding RNAs and is probably one

Fig. 6. Western blot analysis. Detection of TFF1, TFF2, or TFF3 in extracts of the stomach (S) or the kidney (K) of the following mouse strains: BALB/c, C57BL/6, 129S2/Sv, TFF1(−/−). The polyclonal antisera used were: anti-mouse TFF1, anti-mouse TFF2, and affinity-purified anti-hTFF3-2. The molecular size standard is shown left.

The authors thank Prof. E. Wahle (Halle) for helpful discussions and Dr. J. Lindquist for comments on the manuscript. This work was supported by the Land Sachsen-Anhalt (1918A/0025H to W. H.) and the “Fonds der Chemischen Industrie” (0163615 and 0500058 to W.H.). References [1] Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997;25:3389–402. [2] Bossenmeyer-Pourié C, Kannan R, Ribieras S, Wendling C, Stoll I, Thim L, et al. The trefoil factor 1 participates in gastrointestinal cell differentiation by delaying G1–S phase transition and reducing apoptosis. J Cell Biol 2002;157:761–70. [3] Breathnach R, Chambon P. Organization and expression of eucaryotic split genes coding for proteins. Ann Rev Biochem 1981;50:349–83. [4] Calnan DP, Westley BR, May FEB, Floyd DN, Marchbank T, Playford RJ. The trefoil peptide TFF1 inhibits the growth of the human gastric adenocarcinoma cell line AGS. J Pathol 1999;188:312– 7. [5] Chinery R, Poulsom R, Rogers LA, Jeffery RE, Longcroft JM, Hanby A, et al. Localization of intestinal trefoil-factor mRNA in rat stomach and intestine by hybridization in situ. Biochem J 1992;285:5–8. [6] Debailleul V, Laine A, Huet G, Mathon P, d’Hooghe MC, Aubert JP, et al. Human mucin genes MUC2, MUC3, MUC4, MUC5AC, MUC5B, and MUC6 express stable and extremely large mRNAs and exhibit a variable length polymorphism: an improved method to analyze large mRNAs. J Biol Chem 1998;273:881–90. [7] Ebert MPA, Hoffmann J, Haeckel C, Rutkowski K, Schmid RM, Wagner M, et al. Induction of TFF1 gene expression in pancreas overexpressing transforming growth factor α. Gut 1999;45:105–11. [8] Haddad F, Bodell PW, Qin AX, Giger JM, Baldwin KM. Role of antisense RNA in coordinating cardiac myosin heavy chain gene switching. J Biol Chem 2003;278:37132–8. [9] Hauser F, Poulsom R, Chinery R, Rogers LA, Hanby AM, Wright NA, et al. hP1.B, a human P-domain peptide homologous with rat intestinal trefoil factor, is expressed also in the ulcer-associated cell lineage and the uterus. Proc Natl Acad Sci USA 1993;90:6961–5.

762

S.C. Hertel et al. / Peptides 25 (2004) 755–762

[10] Hinz M, Schwegler H, Chwieralski CE, Laube G, Linke R, Pohle W, Hoffmann W. Trefoil factor family (TFF) expression in the mouse brain and pituitary: changes in the developing cerebellum. Peptides 2004, this issue. [11] Hoffmann W, Jagla W. Cell type specific expression of secretory TFF peptides: colocalization with mucins and synthesis in the brain. Int Rev Cytol 2002;213:147–81. [12] Hoffmann W, Jagla W, Wiede A. Molecular medicine of TFF-peptides: from gut to brain. Histol Histopathol 2001;16:319–34. [13] Jagla W, Wiede A, Kolle S, Hoffmann W. Differential expression of the TFF-peptides xP1 and xP4 in the gastrointestinal tract of Xenopus laevis. Cell Tissue Res 1998;291:13–8. [14] Kang W, Reid KBM. DMBT1, a regulator of mucosal homeostasis through the linking of mucosal defense and regeneration? FEBS Lett 2003;540:21–5. [15] Kouznetsova I, Peitz U, Vieth M, Meyer F, Vestergaard EM, Malfertheiner P. A gradient of TFF3 (trefoil factor family 3) peptide synthesis within the normal human gastric mucosa. Cell Tissue Res 2004;316:155–65. [16] Lefebvre O, Chenard M-P, Masson R, Linares J, Dierich A, LeMeur M, et al. Gastric mucosa abnormalities and tumorigenesis in mice lacking the pS2 trefoil protein. Science 1996;274:259–62. [17] Lefebvre O, Wolf C, Kédinger M, Chenard M-P, Tomasetto C, Chambon P, et al. The mouse one P-domain (pS2) and two P-domain (mSP) genes exhibit distinct patters of expression. J Cell Biol 1993;122:191–8. [18] Legendre M, Gautheret D. Sequence determinants in human polyadenylation site selection. BMC Genomics 2003;4:7. [19] Lehner B, Williams G, Campbell RD, Sanderson CM. Antisense transcripts in the human genome. Trends Genet 2002;18:63–5. [20] Luqmani Y, Bennett C, Paterson I, Corbishley CM, Rio M-C, Chambon P, Ryall G. Expression of the pS2 gene in normal, benign and neoplastic human stomach. Int J Cancer 1989;44:806–12. [21] Mashimo H, Wu DC, Podolsky DK, Fishman MC. Impaired defense of intestinal mucosa in mice lacking intestinal trefoil factor. Science 1996;274:262–5. [22] Newton JL, Allen A, Westley BR, May FEB. The human trefoil peptide, TFF1, is present in different molecular forms that are intimately associated with mucus in normal stomach. Gut 2000;46:312–20. [23] Okazaki Y, et al. Analysis of the mouse transcriptome based on functional annotation of 60,770 full-length cDNAs. Nature 2002;420:563–73.

[24] Park WS, Oh RR, Park JY, Lee JH, Shin MS, Kim HS, et al. Somatic mutations of the trefoil factor family 1 gene in gastric cancer. Gastroenterology 2000;119:691–8. [25] Ribieras S, Lefebvre O, Tomasetto C, Rio MC. Mouse trefoil factor genes: genomic organization. Gene 2001;266:67–75. [26] Ribieras S, Tomasetto C, Rio C. The pS2/TFF1 trefoil factor, from basic research to clinical applications. Biochim Biophys Acta 1998;1378:F61–77. [27] Shendure J, Church GM. Computational discovery of sense–antisense transcription in the human and mouse genomes. Genome Biol 2003;3:0044.10–4. [28] Runkel F, Michels M, Franz T. Fxyd3 and Lgi4 expression in the adult mouse: a case of endogenous antisense expression. Mamm Genome 2003;14:665–72. [29] Taupin D, Wu DC, Jeon WK, Devaney K, Wang TC, Podolsky DK. The trefoil gene family are coordinately expressed immediate-early genes: EGF receptor-and MAP kinase-dependent interregulation. J Clin Invest 1999;103:R31–8. [30] Thim L, Madsen F, Poulsen SS. Effect of trefoil factors on the viscoelastic properties of mucus gels. Eur J Clin Invest 2002;32:519– 27. [31] Thim L, Mørtz E. Isolation and characterization of putative trefoil peptide receptors. Regul Peptides 2000;90:61–8. [32] Tomasetto C, Karam SM, Ribieras S, Masson R, Lefebvre O, Staub A, et al. Identification and characterization of a novel gastric peptide hormone: the motilin-related peptide. Gastroenterology 2000;119:395–405. [33] Tomasetto C, Masson R, Linares J-L, Wendling C, Lefebvre O, Chenard M-P, et al. pS2/TFF1 interacts directly with the VWFC cysteine-rich domains of mucins. Gastroenterology 2000;118:70–80. [34] Torres L-F, Karam SM, Wendling C, Chenard M-P, Kershenobich D, Tomasetto C, et al. Trefoil factor 1 (TFF1/pS2) deficiency activates the unfolded protein response. Mol Med 2002;8:273–82. [35] Tufarelli C, Stanely JAS, Garrick D, Sharpe JA, Ayyub H, Wood WG, et al. Transcription of antisense RNA leading to gene silencing and methylation as a novel cause of human genetic disease. Nat Genet 2003;34:157–65. [36] Wiede A, Hinz M, Canzler E, Franke K, Quednow C, Hoffmann W. Synthesis and localization of the mucin-associated TFF-peptides in the human uterus. Cell Tissue Res 2001;303:109–15. [37] Wiede A, Jagla W, Welte T, Köhnlein T, Busk H, Hoffmann W. Localization of TFF3, a new mucus-associated peptide of the human respiratory tract. Am J Respir Crit Care Med 1999;159:1330–5.