Artigo Acesso aberto Revisado por pares

Expression of an Olfactomedin-Related Gene in Rat Hair Follicular Papilla Cells

2005; Elsevier BV; Volume: 125; Issue: 1 Linguagem: Inglês

10.1111/j.0022-202x.2005.23746.x

ISSN

1523-1747

Autores

Qiong Cao, Dawen Yu, Andy Lee, Yuko Kasai, Birte Tychsen, Ralf Paus, Irwin M. Freedberg, Tung‐Tien Sun,

Tópico(s)

melanin and skin pigmentation

Resumo

Follicular papilla (FP) cells, but not their closely related dermal fibroblasts, can maintain hair growth suggesting cell type-specific molecular signals. To define the molecular differences between these two cell types, we generated a subtraction complementary DNA (cDNA) library highly enriched in FP-specific cDNA. Differential screening identified FP-1 as the most abundant cDNA sequence in this subtraction library. FP-1 message RNA is highly abundant in cultured rat vibrissa FP cells, can be detected at very low levels in the stomach and the ovary, and is undetectable in cultured dermal fibroblasts and in 16 rat non-follicular tissues. The full-length, 2.3 kb FP-1 cDNA encodes a protein of 549 amino acids harboring a signal peptide, collagen triple helix repeats, and an olfactomedin-like domain. Monospecific rabbit antibodies to FP-1 recognize in cultured FP cells a single ∼72 kDa glycoprotein with a ∼60 kDa protein core. FP-1 protein is expressed in vivo in a hair cycle-dependent manner, as it can be detected in FP during anagen, but not in catagen and telogen phases of the hair cycle. FP-1 is presumably a highly specific extracellular matrix protein synthesized by FP cells and may be involved in the organization of FP during certain phases of normal or pathological hair growth. Follicular papilla (FP) cells, but not their closely related dermal fibroblasts, can maintain hair growth suggesting cell type-specific molecular signals. To define the molecular differences between these two cell types, we generated a subtraction complementary DNA (cDNA) library highly enriched in FP-specific cDNA. Differential screening identified FP-1 as the most abundant cDNA sequence in this subtraction library. FP-1 message RNA is highly abundant in cultured rat vibrissa FP cells, can be detected at very low levels in the stomach and the ovary, and is undetectable in cultured dermal fibroblasts and in 16 rat non-follicular tissues. The full-length, 2.3 kb FP-1 cDNA encodes a protein of 549 amino acids harboring a signal peptide, collagen triple helix repeats, and an olfactomedin-like domain. Monospecific rabbit antibodies to FP-1 recognize in cultured FP cells a single ∼72 kDa glycoprotein with a ∼60 kDa protein core. FP-1 protein is expressed in vivo in a hair cycle-dependent manner, as it can be detected in FP during anagen, but not in catagen and telogen phases of the hair cycle. FP-1 is presumably a highly specific extracellular matrix protein synthesized by FP cells and may be involved in the organization of FP during certain phases of normal or pathological hair growth. complementary DNA expressed sequence tags follicular papilla glyceraldehyde-3-phosphate dehydrogenase messenger RNA polymerase chain reaction The hair follicle is a dynamic structure undergoing cyclic changes, including anagen (an active growing phase), catagen (a regressing or degenerating phase), and telogen (a resting phase) (Muller-Rover et al., 2001Muller-Rover S. Handjiski B. van der Veen C. et al.A comprehensive guide for the accurate classification of murine hair follicles in distinct hair cycle stages.J Invest Dermatol. 2001; 117: 3-15Crossref PubMed Google Scholar). The interactions between follicular epithelial cells and a group of specialized mesenchymal cells, known as follicular papilla (FP), are involved in the development and the maintenance of hair follicles. FP cells are derived from a group of fibroblast-like cells that form a condensation beneath the presumptive epidermis at the beginning of the embryonic follicle development (Hardy, 1992Hardy M.H. The secret life of the hair follicle.Trends Genet. 1992; 8: 55-61Abstract Full Text PDF PubMed Scopus (762) Google Scholar). Existing evidence indicates that FP plays a key role in controlling hair growth: (i) The size of FP is proportional to the size of the follicle epithelial matrix, and the diameter and the length of the hair fiber (Van Scott and Ekel, 1958Van Scott E.J. Ekel T.M. Geometric relationships between the matrix of the hair bulb and its dermal papilla in normal and alopecia scalp.J Invest Dermatol. 1958; 31: 281-287Abstract Full Text PDF PubMed Scopus (139) Google Scholar; Ibrahim and Wright, 1982Ibrahim L. Wright E.A. A quantitative study of hair growth using mouse and rat vibrissal follicles. I. Dermal papilla volume determines hair volume.J Embryol Exp Morphol. 1982; 72: 209-224PubMed Google Scholar; Elliott et al., 1999Elliott K. Stephenson T.J. Messenger A.G. Differences in hair follicle dermal papilla volume are due to extracellular matrix volume and cell number: Implications for the control of hair follicle size and androgen responses.J Invest Dermatol. 1999; 113: 873-877Crossref PubMed Scopus (164) Google Scholar). (ii) The upward movement of FP during the catagen phase appears to be crucially important for the follicle to enter into the next growing phase. In the hairless mutant mouse (hr/hr), failure of the FP to ascend upward at the end of the first hair cycle abolishes subsequent hair cycles and leads to complete hair loss in adult mice (Montagna et al., 1952Montagna W. Chase H.B. Melaragno H.P. Skin of hairless mice. I. Formation of cysts and the distribution of lipids.J Invest Dermatol. 1952; 19: 83-94Crossref PubMed Scopus (51) Google Scholar). (iii) Oliver et al have shown that if one surgically removes the lower one-third of the rat vibrissa follicle, the surrounding connective tissue sheath can give rise to a new, regenerated FP, and the follicle will survive. But if one removes more cells amounting to the loss of the entire lower one-half of the follicle, FP cannot be regenerated and the follicle die (Oliver, 1966Oliver R.F. Whisker growth after removal of the dermal papilla and lengths of follicle in the hooded rat.J Embryol Exp Morphol. 1966; 15: 331-347PubMed Google Scholar). Implantation of a new FP or a pellet of young passage cultured FP cells, but not regular skin fibroblasts, can rescue such a damaged follicle (Oliver, 1967Oliver R.F. Ectopic regeneration of whiskers in the hooded rat from implanted lengths of vibrissa follicle wall.J Embryol Exp Morphol. 1967; 17: 27-34PubMed Google Scholar; Jahoda and Oliver, 1984Jahoda C.A. Oliver R.F. Vibrissa dermal papilla cell aggregative behaviour in vivo and in vitro.J Embryol Exp Morphol. 1984; 79: 211-224PubMed Google Scholar). Like the vibrissa, the human follicle has also been shown to be able to regenerate an active bulb after follicular amputation (Kim and Choi, 1995Kim J.C. Choi Y.C. Regrowth of grafted human scalp hair after removal of the bulb.Dermatol Surg. 1995; 21: 312-313Crossref PubMed Scopus (38) Google Scholar). (iv) When cultured keratinocytes were combined with FP cells and grafted onto a nude (athymic) mouse, hair follicles were generated; however, no hair was formed when cultured keratinocytes were mixed with dermal fibroblasts (Kamimura et al., 1997Kamimura J. Lee D. Baden H.P. Brissette J. Dotto G.P. Primary mouse keratinocyte cultures contain hair follicle progenitor cells with multiple differentiation potential.J Invest Dermatol. 1997; 109: 534-540Crossref PubMed Scopus (80) Google Scholar). (v)Jahoda et al., 2001Jahoda C.A. Oliver R.F. Reynolds A.J. et al.Trans-species hair growth induction by human hair follicle dermal papillae.Exp Dermatol. 2001; 10: 229-237Crossref PubMed Scopus (63) Google Scholar showed trans-species hair induction by human scalp FP cells. Taken together, these results clearly demonstrate that FP cells, but not their closely related dermal fibroblasts, have at least two functions, i.e., inducing a new down-growth of follicular epithelial cells during anagen, and supporting the proliferation and/or viability of the epithelial matrix. Given the important roles of FP in regulating hair growth, we set out to identify FP-specific genes. There is abundant evidence that every specialized somatic cell in the body synthesizes a set of tissue-specific genes that are responsible for the specialized functions of that cell type. Examples of these include the specific keratins 3 and 12 of corneal epithelium (Sun et al., 1984Sun T.T. Eichner R. Schermer A. Cooper D. Nelson W.G. Weiss R.A. Classification, expression, and possible mechanism of evolution of mammalian epithelial keratins: A unifying model.Cancer Cell. 1984; 1: 169-176Google Scholar; Schermer et al., 1986Schermer A. Galvin S. Sun T.T. Differentiation-related expression of a major 64k corneal keratin in vivo and in culture suggests limbal location of corneal epithelial stem cells.J Cell Biol. 1986; 103: 49-62Crossref PubMed Scopus (1157) Google Scholar; Chaloin-Dufau et al., 1990Chaloin-Dufau C. Sun T.T. Dhouailly D. Appearance of the keratin pair k3/k12 during embryonic and adult corneal epithelial differentiation in the chick and in the rabbit.Cell Differ Dev. 1990; 32: 97-108Crossref PubMed Scopus (75) Google Scholar), uroplakins of urothelium (Wu and Sun, 1993Wu X.R. Sun T.T. Molecular cloning of a 47 kda tissue-specific and differentiation-dependent urothelial cell surface glycoprotein.J Cell Sci. 1993; 106: 31-43Crossref PubMed Google Scholar; Lin et al., 1994Lin J.H. Wu X.R. Kreibich G. Sun T.T. Precursor sequence, processing, and urothelium-specific expression of a major 15-kda protein subunit of asymmetric unit membrane.J Biol Chem. 1994; 269: 1775-1784Abstract Full Text PDF PubMed Google Scholar; Yu et al., 1994Yu J. Lin J.H. Wu X.R. Sun T.T. Uroplakins ia and ib, two major differentiation products of bladder epithelium, belong to a family of four transmembrane domain (4tm) proteins.J Cell Biol. 1994; 125: 171-182Crossref PubMed Scopus (159) Google Scholar; Sun et al., 1999Sun T.T. Liang F.X. Wu X.R. Uroplakins as markers of urothelial differentiation.Adv Exp Med Biol. 1999; 462: 7-18Crossref PubMed Scopus (61) Google Scholar), albumin, transthyretin, and glutamine synthetase of the liver (Fausto, 1990Fausto N. Hepatocyte differentiation and liver progenitor cells.Curr Opin Cell Biol. 1990; 2: 1036-1042Crossref PubMed Scopus (106) Google Scholar; Cereghini, 1996Cereghini S. Liver-enriched transcription factors and hepatocyte differentiation.Faseb J. 1996; 10: 267-282Crossref PubMed Scopus (461) Google Scholar; Susick et al., 2001Susick R. Moss N. Kubota H. et al.Hepatic progenitors and strategies for liver cell therapies.Ann N Y Acad Sci. 2001; 944: 398-419Crossref PubMed Scopus (53) Google Scholar; Zaret, 2001Zaret K.S. Hepatocyte differentiation: From the endoderm and beyond.Curr Opin Genet Dev. 2001; 11: 568-574Crossref PubMed Scopus (150) Google Scholar), and crystallins of the lens (Kodama and Eguchi, 1994Kodama R. Eguchi G. Gene regulation and differentiation in vertebrate ocular tissues.Curr Opin Genet Dev. 1994; 4: 703-708Crossref PubMed Scopus (16) Google Scholar; Cvekl and Piatigorsky, 1996Cvekl A. Piatigorsky J. Lens development and crystallin gene expression: Many roles for pax-6.Bioessays. 1996; 18: 621-630Crossref PubMed Scopus (241) Google Scholar; Bloemendal et al., 2004Bloemendal H. de Jong W. Jaenicke R. Lubsen N.H. Slingsby C. Tardieu A. Ageing and vision: Structure, stability and function of lens crystallins.Prog Biophys Mol Biol. 2004; 86: 407-485Crossref PubMed Scopus (654) Google Scholar). The characterization of these tissue-specific genes can lead to the identification of unique transcription factors, and can make available tissue-specific promoters allowing the study of tissue physiology and function by targeting transgenes to the tissue or by tissue-specific gene ablation. In our previous study, we have already found several genes that are preferentially expressed in FP cells versus dermal fibroblasts using differential display (Yu et al., 1995Yu D.W. Yang T. Sonoda T. et al.Message of nexin 1, a serine protease inhibitor, is accumulated in the follicular papilla during anagen of the hair cycle.J Cell Sci. 1995; 108: 3867-3874PubMed Google Scholar, Yu et al., 2001Yu D.W. Yang T. Sonoda T. et al.Osteopontin gene is expressed in the dermal papilla of pelage follicles in a hair-cycle-dependent manner.J Invest Dermatol. 2001; 117: 1554-1558Crossref PubMed Scopus (17) Google Scholar). None of these genes, however, are truly FP-specific because of the very nature of this approach. We have approached this problem using the subtraction library technique, which is particularly suitable for identifying novel genes that may not be represented in, e.g., the currently available complementary DNA (cDNA) microarrays. We found that, in a FP-specific subtraction cDNA library, FP-1 was the most abundant cDNA sequence. The FP-1 messenger RNA (mRNA) was highly expressed in cultured rat vibrissa FP cells, could be detected at very low levels in the stomach and the ovary, but was absent in cultured dermal fibroblasts and in 16 rat non-follicular tissues. FP-1 protein was expressed only in FP during the anagen phase of the hair cycle. Further studies of the structure and function of this tissue-specific gene and its product may open a new avenue to better understand hair growth regulation. To identify genes that are expressed preferentially in FP cells, we constructed an FP-specific subtraction cDNA library. Common messages were eliminated by hybridizing the cDNA of cultured rat vibrissa FP cells with those of fibroblasts that had been grown under identical culture conditions. For the fibroblasts, we decided to use a mixture of culture fibroblasts from the diaphragm, esophagus, and stomach, instead of skin dermal fibroblasts that are likely to be contaminated by FP cells. Such a contamination of the “driver” cDNA that was used in a quantity about 30 times higher than the “tester,” even if minor, could completely remove some of the FP-specific cDNA from the substation library; in this procedure, the differential expression of the isolated cDNA clones could be established later using authentic cultured dermal fibroblasts. To examine the subtraction efficiency, we performed a series of Southern blots using the following probes: (1) the total FP-specific cDNA (FP-probe); (2) the total fibroblasts-specific cDNA (F-probe); (3) a partial cDNA of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping gene; and (4) a partial cDNA of FP-1, a novel gene identified from the subtraction library. Our results showed a more than 10-fold enrichment of FP-1 in the subtracted FP library (Figure 1d), and a more than 20-fold reduction of GAPDH in both the subtracted FP- and fibroblast-specific libraries (Figure 1c). These data indicated that we had archived a more than 200-fold enrichment of the differentially expressed FP messages. Indeed, when we used the FP-specific cDNA as the probe (FP-probe), the signals of the subtracted FP-specific cDNA (Figure 1a, lane 1) were much stronger than those of the subtracted fibroblast cDNA (Figure 1a, lane 3). On the contrary, when we used the fibroblast-specific cDNA as the probe (F-probe), the signals of the subtracted fibroblast-specific cDNA (Figure 1b, lane 3) were much stronger than those of the subtracted FP cDNA (Figure 1b, lane 1). These data indicated that we had successfully enriched the FP-specific cDNA using the subtraction technique. To identify the FP-specific clones in the subtracted library, we performed a differential screening. Randomly picked clones from the FP-specific subtracted library were hybridized with the FP-probe and F-probe (Figure 2). We defined the FP-specific clones as those giving at least 5-fold stronger signals with the FP-probe than with the F-probe (Figure 2). By screening 465 randomly picked clones from the FP-specific subtracted library, we obtained 60 FP-specific clones representing nine expressed sequence tags (EST) and 25 known sequences (Table I). To verify that such clones were indeed overexpressed in FP cells versus cultured esophageal/diaphragm/stomach fibroblasts and, more importantly, dermal fibroblasts, we carried out virtual Northern blots by hybridizing polymerase chain reaction (PCR)-amplified cDNA from cultured cells with some of the identified clones, including FP-1, EST2, EST6, EST7, lysyl oxidase-like 2, serine protease, and tenascin c (Figure 3). Our results showed that all the identified cDNA clones were indeed expressed at significantly higher levels in FP cells than in the (non-dermal) fibroblast mixture (lane F in Figure 3), again proving the success of the subtraction. Six of seven clones were found to be also expressed at higher levels in FP cells than in dermal fibroblasts (lanes D and F in Figure 3); only one, tenascin c, showed about equal intensity in these two cell types (Figure 3g) suggesting that tenascin c expression was significantly higher in skin fibroblasts than those of the internal organs. These data proved the FP-specificity of many genes identified from the subtraction library, and indicated that the difference between FP and various types of fibroblasts was greater than the difference among the different types of fibroblasts.Table IFP-specific genesGene/ESTNo. of cloneScreening dot/colonyVirtual northernFP-1 (EST1-AI574756)850/5020EST2 (AI010449)115/1510EST3 (BE101484)130/20EST4 (BE553451)130/25EST5 (AI104295)120/15EST6 (BE543110)120/205EST7 (AV601118)120/205EST8 (BE303772)130/305EST9 (AA819288)110/05ß-actin510/05Vascular a-actin450/50Destrin (acting depolymerizing factor)350/5020OSF-2350/5015SH3-containing protein p4015315/10Tenascin C315/1015Fit-1240/40MUC18 glycoprotein220/20Serine protease220/105LOXL2215/055Calponin120/10Channel integral protein 28130/20Clp36 (homo. to reversion-induced LIM)115/15Collagen a1 (VI)115/10Growth potentiating factor110/05IGFBP2115/10IGFBP3110/05IGFBP7 (Follistatin-like protein, Mac25)120/1510a-meltrin (metalloproteinase/disintegrin)110/05Mitochondrial cytochrome c oxidase130/01P38IP110/05Rab3-GAP regulatory domain115/05Transgelin (SM22)130/305a-tropomyosin120/10WH syndrome candidate 2 protein110/05Total60FP, follicular papilla; EST, expressed sequence tags; OSF-2, osteoblast-specific factor-2; LOXL2, lysyl oxidase-like 2; IGFBP, IGF binding protein. Open table in a new tab Figure 3Confirmation of the cell type specificity of the complementary DNA (cDNA) clones identified by subtraction. The PCR-amplified double-stranded cDNA of follicular papilla (FP) cells, fibroblasts (diaphragm, esophagus, and stomach fibroblasts 1:1:1) (F), and dermal fibroblasts (DF) was separated electrophoretically and probed with (A) FP-1, (B) EST2, (C) EST6, (D) EST7, (E) lysyl oxidase-like 2 (LOXL2), (F) serine protease, (G) tenascin c, and (H) glyceraldehyde-3-phosphate dehydrogenase (GAPDH) partial cDNA. Note that although the three cDNA samples contained about equal amounts of GAPDH message, FP-1, EST2, EST6, EST7, LOXL2, serine protease, and tenascin c were preferentially expressed in FP cells versus the mixture of the three fibroblasts (F). Except tenascin c, the other six out of seven messenger RNA were also preferentially expressed in FP versus DF.View Large Image Figure ViewerDownload (PPT) FP, follicular papilla; EST, expressed sequence tags; OSF-2, osteoblast-specific factor-2; LOXL2, lysyl oxidase-like 2; IGFBP, IGF binding protein. Among the 25 known genes and nine EST sequences that had been identified from the FP-specific subtraction library, EST1 (Genebank access number AI574756) was the most abundant, represented by eight independent clones (Table I). The expression level of this EST in cultured rat vibrissa FP cells was more than 30-fold higher than that in cultured rat dermal fibroblasts (Figures 3a and 4a). Northern blot showed that this mRNA was detected at very low levels in the stomach and the ovary, and was undetectable in 16 other tissues including skin, diaphragm, esophagus, brain, lung, heart, liver, spleen, kidney, bladder, intestine, colon, uterus, prostate, testis, and skeletal muscle (Figure 4a). As these data indicated that this clone was preferentially expressed in FP, we named it “FP-1.” To obtain the full-length cDNA sequence of FP-1, we screened a cDNA phage library of cultured rat vibrissa FP cells, and performed a 5′ rapid amplification of cDNA ends (RACE). The full-length FP-1 cDNA was 2332 bp, which had a 1647 bp open reading frame encoding 549 amino acids (Figure 5). This FP-1 amino acid sequence was almost identical to a recently identified rat protein, gliomedin, except that the 461 residue of FP-1 is glutamine, whereas the corresponding residue of gliomedin is arginine (see Discussion). The predicted molecular weight of FP-1 protein is about 59.3 kDa. The N-terminal 33 amino acid residues of FP-1 may serve as a signal peptide (Figure 5). It has six potential glycosylation sites, and it has collagen triple helix repeats in amino acid positions 157–198 and olfactomedin-like domain in positions 317–542. To examine the protein expression pattern of FP-1, we generated five rabbit antisera against FP-1 (Figure 5a). Immunoblot showed that three of the FP-1 antisera (anti-epitopes 1, 2, and 3) recognized a single protein band of about 72 kDa in cultured rat FP cell lysate, with no detectable signals in cultured fibroblast cell lysate (Figure 6a). Digestion of FP-1 with endoglycosidase-H reduced the molecular weight of FP-1 to about 60 kDa, suggesting the presence of high-mannose sugars (Figure 6b). Immunofluorescent staining using the FP-1 antisera showed strong cytoplasmic signals in cultured FP cells, but was negative in fibroblasts (Figure 6c). These data verified that FP-1 was indeed preferentially expressed in cultured FP cells versus fibroblasts. We also found that the staining of COP I, a Golgi complex marker, overlapped partially with FP-1 staining, even though FP-1 staining covered a slightly broader area (Figure 6c), suggesting that FP-1 might be secreted by cultured FP cells. To investigate FP-1 expression in vivo, we performed indirect immunofluorescent staining of the depilated mouse back skin using the FP-1 antiserum. The FP-1 was strongly expressed in the FP during the anagen phase (Figure 7), but not in the catagen and telogen phases of the hair cycle (data not shown). This hair cycle-dependent expression pattern suggested that FP-1 might be involved in the control of hair growth. No staining was noted in the epidermis and other skin cells. To analyze FP-1 expression in the FP cells under the cultured condition, we performed immunofluorescent staining using primary cultures 4, 7, and 10 days after plating. Starting from day 4, all the cells of the whole colony derived from an FP were FP-1 positive, whereas it was barely detectable in the cultured fibroblasts (data not shown). As FP-1 was so abundantly expressed in the FP cells, we wanted to know its biological function. Based on the Jackson Laboratory database, there were a total of 337 mouse mutants that had a hair-related phenotype, and the chromosomal location of some of these mutations may be close to the FP-1 gene. It would therefore be interesting to determine whether some of these existing mouse hair mutants are due to mutations of the FP-1 gene. We have thus surveyed the chromosome localization of FP-1. BLAST search in the National Center for Biotechnology Information (NCBI) genome database revealed that rat FP-1 cDNA was 82% homologous to a Homo sapiens chromosome 15 genomic contig sequence (access number: NT 010194), which was located on human chromosome 15q15. No known hair-related disorders were located within this region. Fluorescent in situ hybridization (FISH) on mouse chromosomes using the rat FP-1 cDNA as a probe clearly showed that FP-1 was on mouse chromosome 9 B–C region (Figure 8). In this region, there were three hair-related mutants, including rough fur (ruf), rough coat (rc), and fur deficient (fd). To examine whether there were some gross changes of FP-1 in these three mouse mutants, we performed a genomic Southern blot. After digestion with seven different restriction enzymes, the genomic DNA of homozygous and heterozygous mutants and their background strains (considered as wild-type to the mutations) were compared by hybridizing with FP-1 cDNA probe. In theory, a size change greater than 1 kb because of insertion or deletion, which occurs frequently in the mouse genome, could be detectable by this approach. But no significant difference was found in all the three mutants suggesting that there was no deletion or insertion of a big DNA fragment (>1 kb) within the genomic region close to FP-1 in these mutants (data not shown). Using the subtraction cDNA library approach, we have identified several novel FP-specific genes. Among them, FP-1 is the most abundant. FP-1 protein sequence is almost identical to rat gliomedin, except one mismatch at their C-terminus (461 Q in FP-1 vs R in gliomedin). The rat gliomedin (also called olfactomedin-related protein) sequence was recently submitted to the NCBI database as a “myelinating glial cell-produced protein” with no additional information or publication (GeneBank accession # AAP22419). The rat FP-1 sequence is also homologous to a recently described mouse protein called CRG-L2 (or collomin; GeneBank accession # NP_852047) that is only known as an early liver cancer marker. We have surveyed various rat tissues using Northern blot and found that FP-1 is expressed at an extremely high level in cultured rat vibrissa FP cells, weakly detected in cultured rat skin fibroblasts and in the stomach and the ovary, but was not detectable in skin, diaphragm, esophagus, brain, lung, heart, liver, spleen, kidney, bladder, intestine, colon, uterus, prostate, testis, and skeletal muscle. Using representational difference analysis to compare normal liver and liver tumors obtained from diethylnitrosamine (DEN)-treated C3H/HeJ mice,Graveel et al., 2003Graveel C.R. Harkins-Perry S.R. Acevedo L.G. Farnham P.J. Identification and characterization of crg-l2, a new marker for liver tumor development.Oncogene. 2003; 22: 1730-1736Crossref PubMed Scopus (16) Google Scholar recently identified a novel gene, that they named CRG-L2 (cancer related gene-liver 2), which is upregulated in both mouse and human hepatocarcinoma. The tissue distribution pattern of this gene was surveyed by semi-quantitative PCR. The results indicated that in mouse the mRNA of this gene can be detected only after 34 cycles of PCR in the testis and still weakly in the skeletal muscle, but not in normal heart, brain, spleen, lung, liver, kidney, and embryos (7–17 d). In humans, this transcript was detected weakly in the placenta after 32 cycles of PCR, but not in normal spleen, thymus, prostate, testis, ovary, smooth muscle, leukocytes, heart, brain, lung, liver, skeletal muscle, and pancreas (Graveel et al., 2003Graveel C.R. Harkins-Perry S.R. Acevedo L.G. Farnham P.J. Identification and characterization of crg-l2, a new marker for liver tumor development.Oncogene. 2003; 22: 1730-1736Crossref PubMed Scopus (16) Google Scholar). Although it is difficult to compare our results with those of Graveel, the fact that they detected small amounts of FP-1 in mouse testis, which was negative in our Northern blot analysis, suggests that our Northern blot was less sensitive than their PCR assay. It is therefore highly significant that in our assay testis was negative, but FP cells, even at this low sensitivity setting, were found to express FP-1 at a level that is over 100 times higher than that of the testis. Taken together, these data strongly suggest that the expression of FP-1 gene is remarkably tissue-specific and is particularly highly expressed in FP cells. The FP-1/gliomedin/CRG-L2 protein sequence contains two conserved domains: amino acid positions 157–198 contain collagen triple helix repeats, and positions 317–542 are highly homologous to olfactomedin. Collagen proteins are involved in the formation of connective tissue structure. The triple helix repeat (G–X–Y) is common to all the family members. The first position of the repeat is G, the second and third positions can be any residue, but are frequently proline and hydroxyproline (Mayne and Brewton, 1993Mayne R. Brewton R.G. New members of the collagen superfamily.Curr Opin Cell Biol. 1993; 5: 883-890Crossref PubMed Scopus (111) Google Scholar). The olfactomedin-like domain is a conserved sequence shared by olfactomedin and its related proteins (such as gliomedin and neuronal olfactomedin-related endoplasmic reticulum (ER) localized protein) (Danielson et al., 1994Danielson P.E. Forss-Petter S. Battenberg E.L. deLecea L. Bloom F.E. Sutcliffe J.G. Four structurally distinct neuron-specific olfactomedin-related glycoproteins produced by differential promoter utilization and alternative mrna splicing from a single gene.J Neurosci Res. 1994; 38: 468-478Crossref PubMed Scopus (73) Google Scholar; Kulkarni et al., 2000Kulkarni N.H. Karavanich C.A. Atchley W.R. Anholt R.R. Characterization and differential expression of a human gene family of olfactomedin-related proteins.Genet Res. 2000; 76: 41-50Crossref PubMed Scopus (57) Google Scholar). Olfactomedin is, like collagens, an extracellular matrix glycoprotein, specifically expressed in olfactory neuroepithelium, forming homopolymers through disulfide bonds and carbohydrate interactions (Anholt et al., 1990Anholt R.R. Petro A.E. Rivers A.M. Identification of a group of novel membrane proteins unique to chemosensory cilia of olfactory receptor cells.Biochemistry. 1990; 29: 3366-3373Crossref PubMed Scopus (27) Google Scholar; Snyder et al., 1991Snyder D.A. Rivers A.M. Yokoe H. Menco B.P. Anholt R.R. Olfactomedin: Purification, characterization, and localization of a novel olfactory glycoprotein.Biochemistry. 1991; 30: 9143-9153Crossref PubMed Scopus (113) Google Scholar; Bal and Anholt, 1993Bal R.S. Anholt R.R. Formation of the extracellular mucous matrix of olfactory neuroepithelium: Identification of partially glycosylated and nonglycosylated precursors of

Referência(s)