Title: AdipoQ Is a Novel Adipose-specific Gene Dysregulated in Obesity
Abstract: Adipose differentiation is accompanied by changes in cellular morphology, a dramatic accumulation of intracellular lipid and activation of a specific program of gene expression. Using an mRNA differential display technique, we have isolated a novel adipose cDNA, termed adipoQ. The adipoQ cDNA encodes a polypeptide of 247 amino acids with a secretory signal sequence at the amino terminus, a collagenous region (Gly-X-Y repeats), and a globular domain. The globular domain of adipoQ shares significant homology with subunits of complement factor C1q, collagen α1(X), and the brain-specific factor cerebellin. The expression of adipoQ is highly specific to adipose tissue in both mouse and rat. Expression of adipoQ is observed exclusively in mature fat cells as the stromal-vascular fraction of fat tissue does not contain adipoQ mRNA. In cultured 3T3-F442A and 3T3-L1 preadipocytes, hormone-induced differentiation dramatically increases the level of expression for adipoQ. Furthermore, the expression of adipoQ mRNA is significantly reduced in the adipose tissues from obese mice and humans. Whereas the biological function of this polypeptide is presently unknown, the tissue-specific expression of a putative secreted protein suggests that this factor may function as a novel signaling molecule for adipose tissue. Adipose differentiation is accompanied by changes in cellular morphology, a dramatic accumulation of intracellular lipid and activation of a specific program of gene expression. Using an mRNA differential display technique, we have isolated a novel adipose cDNA, termed adipoQ. The adipoQ cDNA encodes a polypeptide of 247 amino acids with a secretory signal sequence at the amino terminus, a collagenous region (Gly-X-Y repeats), and a globular domain. The globular domain of adipoQ shares significant homology with subunits of complement factor C1q, collagen α1(X), and the brain-specific factor cerebellin. The expression of adipoQ is highly specific to adipose tissue in both mouse and rat. Expression of adipoQ is observed exclusively in mature fat cells as the stromal-vascular fraction of fat tissue does not contain adipoQ mRNA. In cultured 3T3-F442A and 3T3-L1 preadipocytes, hormone-induced differentiation dramatically increases the level of expression for adipoQ. Furthermore, the expression of adipoQ mRNA is significantly reduced in the adipose tissues from obese mice and humans. Whereas the biological function of this polypeptide is presently unknown, the tissue-specific expression of a putative secreted protein suggests that this factor may function as a novel signaling molecule for adipose tissue. INTRODUCTIONAdipose tissue is highly specialized to play important roles in energy storage, fatty acid metabolism, and glucose homeostasis(1.Cornelius P. MacDougald O.A. Lane M.D. Annu. Rev. Nutr. 1994; 14: 99-129Crossref PubMed Scopus (572) Google Scholar, 2.Spiegelman B.M. Choy L. Hotamisligil G.S. Graves R.A. Tontonoz P. J. Biol. Chem. 1993; 268: 6823-6826Abstract Full Text PDF PubMed Google Scholar). Adipocytes synthesize and store triglyceride in periods of nutritional abundance and mobilize the lipids in response to fasting(2.Spiegelman B.M. Choy L. Hotamisligil G.S. Graves R.A. Tontonoz P. J. Biol. Chem. 1993; 268: 6823-6826Abstract Full Text PDF PubMed Google Scholar, 3.Spiegelman B.M. Hotamisligil G.S. Cell. 1993; 73: 625-627Abstract Full Text PDF PubMed Scopus (199) Google Scholar). Fat tissue is also involved in regulating blood glucose levels through the expression of the insulin responsive glucose transporter, Glu4(4.Kaestner K.H. Christy R.J. McLenithan J.C. Braiterman L.T. Cornelius P. Pekala P.H. Lane M.D. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 3150-3154Crossref PubMed Scopus (248) Google Scholar, 5.Kaestner K.H. Christy R.J. Lane M.D. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 251-255Crossref PubMed Scopus (226) Google Scholar). Fat and muscle, in fact, constitute the two major sites for insulin-regulated glucose uptake.At a molecular level, many genes involved in lipid metabolism and glucose homeostasis are prominently expressed in fat(1.Cornelius P. MacDougald O.A. Lane M.D. Annu. Rev. Nutr. 1994; 14: 99-129Crossref PubMed Scopus (572) Google Scholar). These include fatty acid synthase(6.Moustaid N. Sul H.S. J. Biol. Chem. 1991; 266: 18550-18554Abstract Full Text PDF PubMed Google Scholar), the fatty acid binding protein aP2(7.Spiegelman B.M. Frank M. Green H. J. Biol. Chem. 1983; 258: 10083-10089Abstract Full Text PDF PubMed Google Scholar, 8.Hunt C.R. Ro J.H.-S. Dobson D.E. Min H.Y. Spiegelman B.M. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 3786-3790Crossref PubMed Scopus (266) Google Scholar), lipoprotein lipase(9.Cornelius P. Enerback S. Bjursell G. Olivecrona T. Pekala P.H. Biochem. J. 1988; 249: 765-769Crossref PubMed Scopus (99) Google Scholar), phosphoenolpyruvate carboxykinase(10.Beale E.G. Tishler E.J. Biochem. Biophys. Res. Commun. 1992; 189: 925-930Crossref PubMed Scopus (13) Google Scholar), malic enzyme(11.Wise L.S. Sul H.S. Rubin C.S. J. Biol. Chem. 1984; 259: 4827-4832Abstract Full Text PDF PubMed Google Scholar), glyceraldehyde-3-phosphate dehydrogenase(12.Alexander M. Curtis G. Avruch J. Goodman H.M. J. Biol. Chem. 1985; 260: 11978-11985Abstract Full Text PDF PubMed Google Scholar), and Glut4 (4.Kaestner K.H. Christy R.J. McLenithan J.C. Braiterman L.T. Cornelius P. Pekala P.H. Lane M.D. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 3150-3154Crossref PubMed Scopus (248) Google Scholar). Receptors for lipogenic or lipolytic hormones such as insulin(13.Reed B.C. Lane M.D. Proc. Natl. Acad. Sci. U. S. A. 1980; 77: 285-289Crossref PubMed Scopus (161) Google Scholar, 14.Rubin C.S. Hirsch A. Fung C. Rosen O.M. J. Biol. Chem. 1978; 253: 7570-7578Abstract Full Text PDF PubMed Google Scholar), insulin-like growth factor 1(15.Campfield L.A. Smith F.J. Guisez Y. Devos R. Burn P. Science. 1995; 269: 546-549Crossref PubMed Scopus (3060) Google Scholar), and adrenergic compounds (16.Feve B. Emorine L.J. Briend-Sutren M.M. Lasnier F. Strosberg A.D. Pairault J. J. Biol. Chem. 1990; 265: 16343-16349Abstract Full Text PDF PubMed Google Scholar, 17.Guest S.J. Hadcock J.R. Watkins D.C. Malbon C.C. J. Biol. Chem. 1990; 265: 5370-5375Abstract Full Text PDF PubMed Google Scholar) are also expressed in adipocytes. In addition to these genes that clearly participate in the metabolic functions of adipose tissue, a group of genes that function in extracellular signaling have also been identified in fat. A prototype of these molecules is insulin-like growth factor 1, which is expressed in many tissues during development and plays an important role in cell proliferation(18.Froesch E.R. Schmid C. Schwander J. Zapf J. Annu. Rev. Physiol. 1985; 47: 443-467Crossref PubMed Google Scholar). In adipocytes, however, insulin-like growth factor 1 is found to stimulate cell differentiation(19.Smith P. Wise L.S. Berkowitz R. Wan C. Rubin C.S. J. Biol. Chem. 1988; 263: 9402-9408Abstract Full Text PDF PubMed Google Scholar). More interestingly, insulin-like growth factor 1 is synthesized by preadipocytes in response to growth hormone stimulation(20.Doglio A. Dani C. Fredrikson G. Grimaldi P. Ailhaud G. EMBO J. 1987; 6: 4011-4016Crossref PubMed Scopus (92) Google Scholar), thus potentially functioning in an autocrine or paracrine fashion to promote adipogenesis during development. Another signaling molecule from adipose tissue is TNF-α. 1The abbreviations used are: TNF-αtumor necrosis factor αDMEMDulbecco's modified Eagle's mediumPCRpolymerase chain reaction. TNF-α is secreted from fat, especially in obesity, and acts in an autocrine or paracrine manner to interfere with insulin action in fat and muscle(21.Hotamisligil G. Shargill N.S. Spiegelman B.M. Science. 1993; 259: 87-91Crossref PubMed Scopus (6031) Google Scholar, 22.Hotamisligil G.S. Arner P. Caro J.F. Atkinson R.L. Spiegelman B.M. J. Clin. Invest. 1995; 95: 2409-2415Crossref PubMed Scopus (2941) Google Scholar). The recent cloning and characterization of the ob gene product has further illustrated that adipose tissue secretes signaling molecules that function in an endocrine fashion(23.Zhang Y. Proenca R. Maffei M. Barone M. Leopold L. Friedman J.M. Nature. 1994; 372: 425-432Crossref PubMed Scopus (11632) Google Scholar). The ob gene product (leptin) is secreted from fat into the circulation and acts to regulate body weight, perhaps via a putative receptor in the cerebroventricular region of the brain(15.Campfield L.A. Smith F.J. Guisez Y. Devos R. Burn P. Science. 1995; 269: 546-549Crossref PubMed Scopus (3060) Google Scholar, 23.Zhang Y. Proenca R. Maffei M. Barone M. Leopold L. Friedman J.M. Nature. 1994; 372: 425-432Crossref PubMed Scopus (11632) Google Scholar, 24.Halaas J.L. Gajiwala K.S. Maffei M. Cohen S.L. Chait B.T. Rabinowitz D. Lallone R.L. Burley S.K. Friedman J.M. Science. 1995; 269: 543-546Crossref PubMed Scopus (4199) Google Scholar). Hence, molecules secreted from adipose tissue are capable of modulating diverse functions in fat and other tissues, thus representing a new facet of adipose tissue physiology.In this study, we have used mRNA differential display to clone a novel adipose gene termed adipoQ. Sequence analysis suggests that adipoQ is a secreted protein that shares significant homology to subunits of complement factor C1q and contains a collagenous structure at the NH2 terminus and a globular domain at the COOH terminus. The expression of this novel gene is highly regulated during the adipose differentiation process and is expressed predominantly in adipose tissue in vivo. Moreover, a significant down-regulation in adipoQ mRNA was observed in fat tissues from obese mice and humans. Our results provide a potentially valuable new molecular tool to explore the physiology of adipose tissue in normal and pathological states.EXPERIMENTAL PROCEDURESMaterialsDMEM, calf serum, and fortified calf serum were from Hyclone, Inc. Insulin, dexamethasone, and isobutylmethylxanthine were purchased from Sigma, Inc. Reverse transcriptase (Moloney murine leukemia virus), Klenow fragment and T4 kinase were obtained from BRL, Inc. Isotopes including [32P]dCTP (6000 Ci/mmol) and 35S-dATP (1300 Ci/mmol) were from DuPont NEN. GenAMP kit containing Taq DNA polymerase was purchased from Perkin-Elmer.Cell Lines and Cell CultureMurine fibroblastic 3T3-C2 cells and 3T3-F442A and 3T3-L1 preadipocytes were cultured as described (7.Spiegelman B.M. Frank M. Green H. J. Biol. Chem. 1983; 258: 10083-10089Abstract Full Text PDF PubMed Google Scholar, 25.Green H. Kehinde O. Cell. 1974; 1: 113-116Abstract Full Text PDF Scopus (737) Google Scholar). Induction of adipocyte differentiation was performed essentially as described(26.Tontonoz P. Hu E. Graves R.A. Budavari A.I. Spiegelman B.M. Genes & Dev. 1994; 8: 1224-1234Crossref PubMed Scopus (1983) Google Scholar). Briefly, differentiation was initiated by administration of insulin at 5 μg/ml at confluence for 3T3-F442A cells and dexamethasone (1 uM), isobutylmethylxanthine (0.25 mM), and insulin (5 μg/ml) for 3T3-L1 cells. For 3T3-L1 cells, cells were treated with dexamethasone/isobutylmethylxanthine/insulin mix for 48 h and then were refed by DMEM medium containing 10% fetal calf serum and 5 μg/ml insulin. Using this protocol, more than 90% of the cells in both cell lines acquire an adipocyte morphology 5-7 days after the initiation of differentiation. Culture medium was routinely changed every 2 days, and adipocyte differentiation was examined visually under the microscope.mRNA Differential DisplaymRNA differential display was performed essentially as described(27.Liang P. Pardee A.B. Science. 1992; 257: 967-971Crossref PubMed Scopus (4688) Google Scholar, 28.Liang P. Averboukh L. Pardee A.B. Nucleic Acids Res. 1993; 21: 3269-3275Crossref PubMed Scopus (883) Google Scholar). Briefly, total cellular RNA was isolated from 3T3-C2, 3T3-F442A preadipocytes, and differentiated 3T3-F442A adipocytes using the guanidine isothiocyanate extraction(29.Chirgwin J.M. Przybyla A.E. MacDonald R.J. Rutter W.J. Biochemistry. 1979; 18: 5294-5299Crossref PubMed Scopus (16620) Google Scholar). 50 μg of total RNA was then treated with 20 units of RNase free-DNase (BRL, Inc.). Subsequently 0.2 μg of treated RNA was used in a reverse transcription reaction using each of the four 1-base pair-anchored 3′ oligo(dT) primers (30.Liang P. Zhu W. Zhang X. Guo Z. O'Connell R.P. Averboukh L. Wang F. Pardee A.B. Nucleic Acids Res. 1994; 22: 5763-5764Crossref PubMed Scopus (307) Google Scholar) and 300 units of Mo-MLV reverse transcriptase (BRL) in 20 μl volume as recommended by the manufacturer. 2 μl of the reverse transcribed cDNA was used for each PCR reaction. PCR reaction was performed using the same 1-base pair-anchored 3′ oligo(dT) primer and 10 5′ arbitrary oligos of 10 nucleotides in length. The sequence of the 5′ arbitrary oligonucleotide that gave DD1 PCR product (see text) is 5′-AGTCATACAT-3′. The 50-μl PCR reaction contained 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.01% gelatin, 2 uM dNTP (except dATP), 1 μl of α-35S-dATP (1300 Ci/mmol), 2 uM of 5′ arbitrary oligo and 3′ anchored dT oligo, and 0.2 μl of Taq polymerase. Parameters for PCR were 30 cycles of denaturing at 95°C for 30 seconds, annealing at 40°C for 1 min, and extension at 72°C for 30 seconds. 5 μl of the PCR reaction mixture was loaded on a 8% sequence gel, and differentially amplified PCR fragments were visualized by exposing the dried sequencing gel to x-ray film. Candidate PCR products were excised from the sequencing gel, and the DNA was eluted from the gel slices by boiling the gel slice in TE (10 mM Tris, pH 7.5, 1 mM EDTA) buffer for 10 min. The eluted DNA fragment was re-amplified by using the same primer pair and subsequently cloned into the TA cloning vector (Invitrogen, Inc.).Library Screening, cDNA Cloning, and SequencingThe cDNA library screening, restriction fragment analysis, subcloning, and sequencing analysis were performed as described(31.Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). An adipocyte-specific λZAP II cDNA library was custom made by Stratagene Inc. as described(26.Tontonoz P. Hu E. Graves R.A. Budavari A.I. Spiegelman B.M. Genes & Dev. 1994; 8: 1224-1234Crossref PubMed Scopus (1983) Google Scholar). GenBank data base searches were performed using the Eugene program at the computer service in Dana-Farber Cancer Institute, and further homology searches were performed using Blast and Autosearch programs. Homology alignments were completed using the Pileup program from the Genetics Computer Group sequence analysis package (Madison, Wisconsin).In Vitro Translation of AdipoQ cDNAIn vitro transcription and translation was performed using a TNT in vitro translation system from Promega, Inc. according to the manufacturer's instructions. [35S]methionine (800 Ci/mmol, DuPont NEN) was used to label the translated protein and visualized on 12% SDS-polyacrylamide gel after fluorography. Molecular markers were from Amersham Corp. (rainbow markers).Expression of Flag-tagged AdipoQThe coding sequence of adipoQ was subcloned into pSV-sport eukaryotic expression vector (BRL) using oligonucleotides 5′-GAATTCGGGATGCTACTGTTGCAAGCT-3′ and 5′-CTCTTCCATGATACCAACGACTACAAGGACGACGATGACAAGTGAGAATTC-3′. The flag-epitope (DYKDDDDK, Kodak Scientific Imaging, Inc) was incorporated into the COOH terminus. The NIH-3T3 fibroblasts were transient transfected with pSV-sport-flag-adipoQ as described(31.Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). 24 h after transfection, cells were washed with phosphate-buffered saline, and DMEM medium with no serum was added. DMEM medium was collected after 24 h, and cells were lysed with RIPA buffer as described(31.Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Antibody against Flag-epitope (M2 monoclonal antibody, Kodak) was used to immunoblot the proteins separated on SDS-polyacrylamide gel electrophoresis.RNA Isolation and Northern Blot AnalysisTotal RNA was isolated from both 3T3-F442A and 3T3-L1 cell lines as well as from various mouse, rat, and human tissues as described(29.Chirgwin J.M. Przybyla A.E. MacDonald R.J. Rutter W.J. Biochemistry. 1979; 18: 5294-5299Crossref PubMed Scopus (16620) Google Scholar, 31.Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). 10 μg of RNA was denatured in formamide and formaldehyde at 55°C and separated in formaldehyde-containing gels as described(31.Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). RNA was blotted onto Hybond nylon membranes, and the nylon membranes were baked, hybridized, and washed as directed by the manufacturer. cDNA probes were radiolabeled to specific activities of at least 109 cpm/μg with [α-32P]dCTP (6000 Ci/mmol) using the random priming method(32.Fineberg A.P. Volgelstein B. Anal. Biochem. 1984; 137: 266-267Crossref PubMed Scopus (5171) Google Scholar).Fractionation of Rat Fat Pad into Stromal-vascular and Fat CellsFractionation of rat fat fads was performed as described (33.Gregoire F. Todoroff G. Hauser N. Remacle C. Biol. Cell. 1990; 69: 215-222Crossref PubMed Scopus (45) Google Scholar). Briefly, epididymal fat deposits were removed and transferred into Petri dishes containing the DMEM supplemented with 10% fetal calf serum and antibiotics (penicillin 100 μg/ml, streptomycin 10 μg/ml). The fat pads were minced by surgical scissors and digested with collagenase (5 mg/ml) for 45 min at 37°C under agitation. The resulting cell suspension was filtered through a 100-μm nylon filter and centrifuged at 400 × g for 5 min. The floating mature adipocytes were washed and centrifuged again. The pelleted fractions were collected and combined. Floating adipocytes and the pelleted stromal-vascular fraction were lysed with guanidine isothiocyanate, and total RNA was isolated as described(29.Chirgwin J.M. Przybyla A.E. MacDonald R.J. Rutter W.J. Biochemistry. 1979; 18: 5294-5299Crossref PubMed Scopus (16620) Google Scholar).RESULTSIdentification of AdipoQ cDNATo identify novel genes that are differentially expressed during adipose differentiation, we employed an mRNA differential display technique(27.Liang P. Pardee A.B. Science. 1992; 257: 967-971Crossref PubMed Scopus (4688) Google Scholar, 34.Liang P. Averboukh L. Pardee A.B. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 12515-12519Crossref PubMed Scopus (59) Google Scholar). RNA samples were prepared from 3T3-C2 cells (C), a fibroblastic cell line unable to differentiate into adipocytes, and a similarly derived preadipocyte cell line, 3T3-F442A cells before (P) and after (A) differentiation(25.Green H. Kehinde O. Cell. 1974; 1: 113-116Abstract Full Text PDF Scopus (737) Google Scholar, 35.Green H. Kehinde O. Cell. 1976; 7: 105-113Abstract Full Text PDF PubMed Scopus (612) Google Scholar). The RNA was used to synthesize the corresponding cDNA, which was subsequently used for differential display PCR reaction using 10 different arbitrary 5′ primers (see “Experimental Procedures”). A number of candidate α-35S-dATP-labeled PCR products visualized in sequencing gels were expressed preferentially in mature adipocytes (data not shown), and we focused on one such product, DD1 (Fig. 1A). A partial cDNA clone for DD1 was obtained by PCR re-amplification (see “Experimental Procedures”) and sequenced. No significant sequence homology with any other genes in GenBank was apparent from this 200-base pair fragment. However, putative polyadenylation signals were present in this short nucleotide sequence. Northern analysis using this cDNA fragment revealed a mRNA expressed predominantly in differentiated fat cells (Fig. 1B). These data suggested that this cDNA fragment reflected a genuine mRNA species that was differentially regulated. A full-length clone of DD1 was subsequently obtained by screening a λZAP II cDNA adipocyte library with the partial cDNA clone. Sequence analysis revealed a single open reading frame in the full-length cDNA clone (Fig. 2A).Figure 2:Nucleotide and deduced amino acid sequences of adipoQ: homology to components of C1q, collagen α1(X), and cerebellin. A, nucleotides are numbered from the 5′ end of the sequence. The amino acid sequence is derived from the longest open reading frame. The bold region indicates the collagen-like domain. Putative polyadenylation signals are underlined. The accession number for adipoQ in GenBank is U49915. B, homology of adipoQ with murine C1q A, B, and C chains. The bold region indicates identical amino acids in all four sequences. Two interruptions in collagenous region of C1q-B and C1q-C chains are indicated by an asterisk (see text). Symbols # and mark the conserved cysteines (see text) that are altered in adipoQ. C, comparisons of the globular region between C1q B chain, collagen α1(X), and cerebellin. Identical or conserved residues in all four sequences are indicated by bold letters. The two regions that are most homologous are underlined. Pair-wise comparisons between adipoQ and C1q-B chain, collagen α1(X), and cerebellin give 31, 38, and 25% identity, respectively. D, in vitro transcription and translation of adipoQ cDNA. TNT in vitro translation system (Promega) was used for translation (see “Experimental Procedures”). Varying amounts of purified cDNA plasmid (pBluescript) were added directly to the transcription-translation mixture with no RNA polymerase (lane 1), T3 RNA polymerase (0.5 μg DNA) (lane 2), T3 RNA polymerase (2 μg of DNA) (lane 3), T7 RNA polymerase (0.5 μg DNA) (lane 4) and T7 RNA polymerase (2 μg of DNA) (lane 5). [35S]Methionine-labeled products were separated on 10% SDS gel and visualized after fluorography. Molecular mass is indicated by kDa at the left side of the gel. E, Western blot of proteins from conditional medium and cell lysates of transiently transfected NIH-3T3 cells (see “Experimental Procedures”). Lane 1 and 2 were 1 μl of in vitro translated Flag-adipoQ and empty vector. Lanes 3 and 4 were 50 μl (10 μg of protein) of culture medium from Flag-adipoQ transfected (lane 3) and empty vector (lane 4). Lanes 5 and 6 were 10 μl (50 μg of protein) of cell lysates from Flag-adipoQ transfected (lane 5) and empty vector (lane 6). The proteins were separated on 12% SDS-polyacrylamide gel electrophoresis and immunoblotted with M2-Flag antibody from Kodak, Inc.View Large Image Figure ViewerDownload (PPT)Analysis of the putative protein sequence identified a hydrophobic leader from amino acid residues 2 to 17, presumably representing a signal peptide. A region of collagenous repeats (Gly-X-Y) was present from amino acids 45 to 110, with 22 individual Gly-X-Y repeats. Comparisons with genes in GenBank identified several regions of homology to the subunits (A, B, and C chains) of complement factor C1q (36.Petry F. Reid K.B.M. Loos M. J. Immunol. 1991; 147: 3988-3993PubMed Google Scholar, 37.Petry F. Reid K.B.M. Loos M. Eur. J. Biochem. 1992; 209: 129-134Crossref PubMed Scopus (21) Google Scholar), a tissue-specific collagen α1(X)(38.Elima K. Eerola I. Rosati R. Metsaranta M. Garofalo S. Perala M. De Crombrugghe B. Vuorio E. Biochem. J. 1993; 289: 247-253Crossref PubMed Scopus (64) Google Scholar), and a brain-specific protein cerebellin(39.Urade Y. Oberdick J. Molinar-Rode R. Morgan J.L. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 1069-1073Crossref PubMed Scopus (97) Google Scholar). The identity with the C1q chains is approximately 31% in the globular COOH-terminal region (Fig. 2B), with the homology localized primarily in two segments of uncharged, hydrophobic regions (Fig. 2C). In addition, adipoQ and C1q A, B, and C chains have a similar size of 240-250 amino acids (Fig. 2B). The number of Gly-X-Y repeats is similar as well, with 22 such repeats for adipoQ and 26-29 for the C1q chains. The similarity of this protein to collagen α1(X) and cerebellin is found mainly at the COOH-terminal globular domain (Fig. 2C) with 38 and 25% identity over a 130-amino acid region. Collagen α1(X), however, encodes a much larger protein (680 amino acids) with a long collagenous segment (154 Gly-X-Y repeats). Cerebellin, on the other hand, is a smaller polypeptide with 193 amino acid residues and does not contain a collagenous domain(39.Urade Y. Oberdick J. Molinar-Rode R. Morgan J.L. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 1069-1073Crossref PubMed Scopus (97) Google Scholar). Because of the similarity between this novel protein and all three components of C1q molecules in size, domain structure and overall homology, we termed this novel protein adipoQ.AdipoQ is clearly a distinct member of a proteins family characterized by a collagenous helical structure at the NH2 terminus, and a globular domain at the COOH terminus(40.Prockop D.J. Kivirikko K.I. Annu. Rev. Biochem. 1995; 64: 403-434Crossref PubMed Scopus (1368) Google Scholar). In addition to C1q A, B, and C chains (36.Petry F. Reid K.B.M. Loos M. J. Immunol. 1991; 147: 3988-3993PubMed Google Scholar, 37.Petry F. Reid K.B.M. Loos M. Eur. J. Biochem. 1992; 209: 129-134Crossref PubMed Scopus (21) Google Scholar) and collagen α1(X)(38.Elima K. Eerola I. Rosati R. Metsaranta M. Garofalo S. Perala M. De Crombrugghe B. Vuorio E. Biochem. J. 1993; 289: 247-253Crossref PubMed Scopus (64) Google Scholar), this protein family includes lung surfactant proteins SP-A and SP-D(41.White R.T. Damm D. Miller J. Sratt K. Schilling J. Hawgood S. Benson B. Cordell B. Nature. 1985; 317: 361-363Crossref PubMed Scopus (296) Google Scholar), mannan binding protein(42.Drickamer K. Dordal M.S. Reynolds L. J. Biol. Chem. 1986; 261: 6878-6887Abstract Full Text PDF PubMed Google Scholar), and the scavenger receptor and its homolog(43.Krieger M. Hertz J. Annu. Rev. Biochem. 1994; 63: 601-637Crossref PubMed Scopus (1057) Google Scholar, 44.Elomaa O. Kangas M. Sahlberg C. Tuukkanen J. Sormunen R. Liakka A. Thesleff I. Kraal G. Tryggvason K. Cell. 1995; 80: 603-609Abstract Full Text PDF PubMed Scopus (411) Google Scholar). These proteins often homo- or hetero-oligomerize via the collagenous structures. The presence of a collagenous domain in adipoQ suggests that this protein is likely to form oligomeric structures by itself or with other proteins.Although the COOH-terminal region of adipoQ shares significant similarity with the C1q chains, it also has some notable differences. For example, a cysteine (marked in Fig. 2B) in the globular region of C1q B (residue 194) and C (residue 198) chains is known to form disulfide bonds with activator molecules, e.g. IgG(45.Martin H. Kaul M. Loos M. Eur. J. Immunol. 1990; 20: 1641-1645Crossref PubMed Scopus (10) Google Scholar). In adipoQ sequence, this cysteine is not conserved and is replaced by an aspartate residue. Another conserved cysteine (residue 180 for C1q C chain, marked # in Fig. 2B) that is important for the formation of disulfide bonds and the stabilization of triplex strands in the collagenous domain (37.Petry F. Reid K.B.M. Loos M. Eur. J. Biochem. 1992; 209: 129-134Crossref PubMed Scopus (21) Google Scholar, 46.Sim R.B. Reid K.B.M. Immunol. Today. 1991; 12: 307-311Abstract Full Text PDF PubMed Scopus (78) Google Scholar) is also altered in adipoQ. In addition, an interruption (marked ∗ in Fig. 2B) in a collagenous motif found in C1q A (residue 61) and C (residue 65) chains is absent in adipoQ. These interruptions have been shown to be conserved between human and mice and result in a bend in the collagen triplex formation that can be observed under the electron microscope (47.Knobel H.R. Villinger W. Isliker H. Eur. J. Immunol. 1975; 5: 78-81Crossref PubMed Scopus (94) Google Scholar, 48.Brodsky-Doyle B. Leonard K.R. Reid K.B.M. Bochem. J. 1976; 159: 279-286Crossref PubMed Scopus (120) Google Scholar). These differences suggest that adipoQ may have structural and functional properties distinct from those of C1q.In an in vitro transcription and translation system, the adipoQ cDNA generates a protein of approximately 30 kDa in size (Fig. 2E). This is in agreement with the molecular mass predicted from the cDNA sequence.To test whether adipoQ is secreted, we constructed a flag-epitope tagged adipoQ and transiently transfected the DNA construct into NIH-3T3 cells. Western blot analysis (Fig. 2E) demonstrated that NIH-3T3 cells synthesized the adipoQ protein, and the protein is secreted into the medium.Differentiation-dependent Expression of AdipoQ mRNAWe next examined the expression of adipoQ mRNA during adipocyte differentiation. As was shown in Fig. 1B, a single mRNA species of approximately 1.3 kilobases was expressed in both 3T3-F442A and 3T3-L1 preadipocytes but not in fibroblastic 3T3-C2 cells. The expression of adipoQ mRNA was found to increase approximately 20-50-fold during adipocyte differentiation in both 3T3-F442A and 3T3-L1 cells (Fig. 3, A and B). Compared with the expression of early adipose differentiation markers such as lipoprotein lipase (9.Cornelius P. Enerback S. Bjursell G. Olivecrona T. Pekala P.H. Biochem. J. 1988; 249: 765-769Crossref PubMed Scopus (99) Google Scholar) and PPAR-λ2 (26.Tontonoz P. Hu E. Graves R.A. Budavari A.I. Spiegelman B.M. Genes & Dev. 1994; 8: 1224-1234Crossref PubMed Scopus (1983) Google Scholar), the expression of adipoQ mRNA is a late event in adipogenesis, first appearing at approximately day 4 after induction of differentiation. This kinetics is similar to