A Novel Short Consensus Repeat-containing Molecule Is Related to Human Complement Factor H*

We have identified a novel factor H-related cDNA, which was isolated from a human liver cDNA library. The DOWN16 clone is 1269 base pairs in size and hybridized to a mRNA of 1.4 kilobases. Similar to the previously described factor H-related proteins, the predicted translation product of 331 amino acids con-tains a hydrophobic signal sequence followed by a stretch of five short consensus repeats (SCRs). These five SCRs display homology to SCRs of factor H: SCRsl-3 (DOWN16) are homologous to SCRs6-8 of factor H, while SCRs4 and -5 are related to SCRsl9 and -20. In vitro translation demonstrated that the DOWN16 cDNA encodes a primary translation product of an apparent molecular mass of 37,500 Da which is directed to the secretory pathway and is glycosylated. Thus, we propose that the protein will be present in human serum. The relatedness of structural elements between this novel gene and factor H may suggest common functions of these proteins not yet determined. The alternative corresponding cDNAs isolated cDNA The mRNA factor H is the organi-zation of factor H protein in 20 tandem repeating units approximately 60 amino acids, short consensus repeat (SCR). In SCRs several C3- and C4-binding proteins of the complement system; found in noncomplement proteins the b chain of the clotting factor XIII, the lymphocyte


A Novel Short Consensus Repeat-containing Molecule Is Related to Human Complement Factor H*
Federal Republic of Germany We have identified a novel factor H-related cDNA, which was isolated from a human liver cDNA library. The DOWN16 clone is 1269 base pairs in size and hybridized to a mRNA of 1.4 kilobases. Similar to the previously described factor H-related proteins, the predicted translation product of 331 amino acids contains a hydrophobic signal sequence followed by a stretch of five short consensus repeats (SCRs). These five SCRs display homology to SCRs of factor H: SCRsl-3 (DOWN16) are homologous to SCRs6-8 of factor H, while SCRs4 and -5 are related to SCRsl9 and -20. In vitro translation demonstrated that the DOWN16 cDNA encodes a primary translation product of an apparent molecular mass of 37,500 Da which is directed to the secretory pathway and is glycosylated. Thus, we propose that the protein will be present in human serum. The relatedness of structural elements between this novel gene and factor H may suggest common functions of these proteins not yet determined.
The complement glycoprotein factor H plays a regulatory role in the alternative pathway of complement activation, and the corresponding cDNAs have been isolated from human liver cDNA libraries (1)(2)(3). The mRNA encoding factor H is 4.4 kb' in size, and structural analysis indicates the organization of factor H protein in 20 tandem repeating units of approximately 60 amino acids, termed short consensus repeat (SCR). In addition to factor H, these SCRs are present in several C3-and C4-binding proteins of the complement system; they have also been found in noncomplement proteins such as the b chain of the clotting factor XIII, the lymphocyte homing receptor, the IL-2 receptor, and a vaccina virusencoded protein (4). Potentially, these molecules evolved by intragenic duplication of SCR elements, and some SCRcontaining elements are clustered in the human genome (5). Seven genes have been mapped to the RCA cluster (regulation of complement activation) on human chromosome 1 (6)(7)(8). These genes represent factor H, b chain of coagulation factor XIII, C4 binding protein, decay accelerating factor, membrane * This work was supported by Deutsche Forschungsgemeinschaft Grant Zi432/1.1. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequencefs) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number ( Several human factor H-related molecules have been isolated at the cDNA and at the protein level. So far, three factor H-related cDNAs have been isolated from human liver cDNA libraries. The 1.8-kb mRNA is supposedly derived from the human factor H gene by alternative splicing (9,10). Two distinct cDNAs, , both correspond to the 1.4-kb mRNA and seem transcribed from loci distinct from that of factor H. Two human serum proteins of 37,000 and 42,000 Da (h37, h42), encoded by the H36 cDNA, are both organized in five SCRs (11,14). The distinct molecular mass of these proteins shows that they represent differently glycosylated forms. Similarly, the human factor Hrelated serum proteins of 24,000 and 29,000 Da (h24, h29), derived from the DDESK59 cDNA also differ in their degree of glycosylation. These two proteins are organized in four SCRs (11,13). The use of cDNA fragments specific for factor H and for the factor H-related molecules as probes for Northern blot analyses indicated the existence of additional factor H-related transcripts in human liver (11). We describe the isolation of a novel factor H-related cDNA DOWN16. The protein encoded by this cDNA displays a signal peptide and is organized in five SCRs. These individual SCRs display homology to SCR6, -7, -8, -19, and -20 of factor H.

MATERIALS AND METHODS
Labeling of Oligonucleotide Probes and Screening-A human liver oligo(dT)-primed cDNA library in X zap was screened with a cDNA fragment of factor H, representing 1050 bp of the 3' end of the 1.8kb cDNA (kindly provided by T. Kristensen and B. Tack).
Sequence Analysis of cDNA Clones-The purified cDNA inserts were sequenced by the dideoxy chain termination method (15) using [cY-~'S]~ATP and Sequenase I1 (U. S. Biochemical Corp.). Various oligonucleotides were synthesized and used as primers to sequence the cDNA in both orientations.
RNA Isolation and Northern Blot Analysis-Total cellular RNA was extracted with guanidinium thiocyanate and isolated by centrifugation over CsCl (16). The RNA concentration was determined spectrophotometrically, and 8 pg of total cellular RNA was separated by electrophoresis in a formaldehyde agarose gel and subsequently transferred to a nylon membrane (PALL).
Southern Blot Analysis-Human DNA (10 pg) isolated from U937 cells was digested to completion with EcoRI or BamHI, separated in a 1.0% agarose gel, and transferred to a nylon membrane (PALL) (17).
Labeling and Hybridization-For library screening, a fragment representing 1050 bp of the 3' end of the factor H 1.8 cDNA was used. A specific TaqIIXhoI fragment (731-1264) of DOWN16 representing mainly SCR5 and the 3"untranslated region was used as a probe for Northern blotting. The cDNA inserts were excised, purified on low melt agarose gels, and, after labeling with 3zP by random priming (Amersham), used for hybridization at 42 "C (5 X Denhardt's, 5 X SSC, 0.1% SDS, 250 pg/ml denatured salmon sperm DNA, and 50% formamide). Following hybridization for 14-18 h, the filters were washed at a final stringency of 0.1% SSC at 55 "C. The filters were exposed at -70 "C using intensifying screens (Quanta 111, Du Pont-New England Nuclear).
In Vitro Transcription and Translation-For cell-free expression, the DOWN16 cDNA clone was linearized with Hind111 at the 3' end of the coding region, and in uitro transcription was carried out from the T3 promoter. The transcripts were translated in wheat germ lysate in the presence or absence of canine pancreatic microsomes as described previously (18). For the inhibition of high mannose-type N-glycosylation, an NH2-and COOH-terminally blocked tripeptide (benzoyl-Asn-Leu-Thr-N-methylamide) was included in the translation reaction, at a final concentration of 30 p M (19). Proteins were labeled by incorporation of ~-[~~S]methionine (Amersham), separated on 15% SDS-polyacrylamide gels, and visualized by autoradiography.

RESULTS
Isolation of DOWN16 cDNA Clones-In order to identify additional factor H-related genes, we screened an oligo(dT)primed human liver cDNA library with DNA fragments of factor H. The nucleotide sequence of one clone indicated that it was distinct from factor H. Consequently, this clone was used to isolate additional cDNA clones. The sequence of the longest clone (DOWN16) is shown in Fig. 1. This clone is 1269 nucleotides long, including the poly(A) tail. The motif TCT AAC ATG shows a good match (6 out of 9, including the ATG) with the consensus sequence of initiation sites GCC ACC ATG (20). There is a poly(A) signal sequence "AA-T A A A at position 1231-1236.
Deduced Protein Sequence-The nucleotide sequence of clone DOWN16 displays an open reading frame of 331 amino acids encoding a protein of 37,500 Da. Within the predicted amino acid sequence, four potential N-linked glycosylation sites of the type Asn-X-Ser/Thr were found at positions 108-110 (Asn-Ser-Thr), 186-188 (Asn-Ser-Ser), 206-208 (Asn-Ser-Ser), and 310-312 (Asn-Thr-Ser). The NHz-terminal amino acid residues are indicative of a signal peptide which indicates that the product of the DOWN16 gene may be directed to the secretory pathway (21). According to the criteria common for signal peptide cleavage sites, we suggest that the leader sequence is cleaved at position 19 (21). The molecular mass of the secreted, nonglycosylated product was calculated to 35,500 Da.
Structural Analysis and Homology-Structural alignment of the protein encoded by the DOWN16 cDNA indicated a protein composed of an NHz-terminal signal peptide followed by five SCRs (Fig. 2). Each SCR contained four of the characteristic Cys (C) residues (boxed with double lines in Fig.  21, and, additionally, a Gly (G), a Tyr (Y), 2 Gly (G), and a Pro (P) residue are conserved in all SCRs (Fig. 2). The signal peptide and the individual SCRs display homology to SCRs of factor H and to the products of the factor H-related cDNAs H36 (h37/h42) and DDESK59 (h24/h29), respectively ( Fig. 3 and Table I). SCRsl and -2 of DOWN16 are homologous to SCRs6 and -7 of factor H, having four and eight amino acids exchanged within a SCR. SCRs3-5 of DOWN16 are related to SCRs8, -19, and -20 of factor H, but display a lower degree of homology ( Fig. 3 and Table I).
Expression Analysis-In order to demonstrate the expression of the DOWN16 mRNA in human liver, we performed Northern blot analysis. A fragment specific for DOWN16 hybridized to a mRNA of approximately 1.4 kb, indicating that the cDNA is nearly full-length (Fig. 4).
Southern Blot Analyses-To determine whether the DOWN16 specific sequence is a member of a family of even more highly related genes, we probed human genomic DNA with a fragment specific for DOWN16. The data in Fig. 5 indicate a limited number of hybridizing fragments. Although DOWN16 is a member of a family of factor H-related genes, the DOWN16 specific fragment itself does not represent a large family of related sequences.
Cell-free Expression and Membrane Translocation of the Protein Encoded by DOWNI6"The DOWN16 cDNA is the first factor H-related cDNA for which no human serum protein has been identified so far. Since the other members of the factor H family are secreted, and since the predicted amino acid sequence of the DOWN16 product is indicative of a hydrophobic signal sequence, we tested whether the DOWN16 protein is directed to the secretory pathway. To this end, the cDNA was transcribed in vitro and expressed in a reconstituted heterologous system consisting of wheat germ lysate and canine pancreatic microsomes (22). In the absence of microsomes, the DOWN16 RNA was translated to a polypeptide of an apparent size of 37,500 Da (Fig. 6A, lane 1 ) . When microsomes were present in the translation reaction, the primary translation product was converted to three major products of larger sizes (Fig. 6A, lune 2 ) . These products were inside microsomal vesicles because they were protected from digestion with post-translationally added proteinase K (Fig.  6A, lane 3 ) , but they were susceptible to protease if the microsomes were lysed with Triton X-100 (Fig. 6A, lane 4 ) .
N-Linked glycosylation occurs in the lumen of the endoplasmic reticulum and results in an increase in molecular mass. In order to assess if the increase in molecular mass was due to N-linked glycosylation, translation and translocation experiments were carried out in the presence of an acceptor peptide which competitively inhibits N-linked glycosylation of newly synthesized polypeptides (Fig. 6B, lanes 2-4). In the presence of microsomal membranes, the primary translation product was converted to a slightly smaller form of approximately 35,500 Da, which in the absence of Triton X-100 was protected from proteinase K, but not in its presence (Fig. 6B,  lanes 1-4). The conversion of the primary translation product to a smaller form (Fig. 6B, lane 2 ) is consistent with the cleavage of an NHz-terminal signal sequence upon translocation into microsomes (23). These results confirm that the larger products generated in the absence of acceptor peptide have arisen from N-linked glycosylation (Fig. 6A, lunes 2 and  3 ) . Their different masses correspond to variant degrees of glycosylation.

DISCUSSION
We describe the isolation and characterization of a novel factor H-related cDNA clone, DOWN16. The predicted amino acid sequence displays a hydrophobic signal peptide. Similar to factor H and to the other factor H-related proteins, the DOWN16 protein is organized in SCRs. SCRsl and -2 of DOWN16 display a high degree of homology to SCRs6 and -7 of factor H ( Fig. 3 and Table I), while the homology of SCRs3-5 to SCRs8, -19, and -20 of human factor H is less pronounced (Table I). In vitro translation experiments demonstrate that the DOWN16 cDNA encodes a primary translation product of an apparent molecular mass of 37,500 Da which is directed to the secretory pathway.
Similar to factor H and to the other factor H-related molecules, the processed DOWN16 product consists exclusively of SCRs. Comparing the mobility of the DOWN16 mRNA to the 18 S ribosomal RNA indicates a size of about 1.4 kb. This size indicates that the DOWN16 cDNA is nearly full-length. The full-length cDNA clones of the other two factor H-related cDNAs (H36 and DDESK59) which are 1266 bp and 1051 bp in length, respectively, are also represented by 13).
The DOWN16 cDNA encodes a protein of 37,500 Da including an NHz-terminal signal sequence which directs the   translation product to the secretory pathway. This indicates that, similar to factor H and the two other factor H-related proteins (h24/h29 and h37/h42), the DOWN16 protein is secreted. So far, no factor H-related protein of a size similar to that of DOWN16 has been identified in human serum. However, the high relatedness between SCRsl and -2 (DOWN16) and SCRs6 and -7 (factor H) should allow crossreaction with anti-factor H antiserum. If all four N-linked glycosylation sites of the DOWN16 product are used in uiuo, we predict a molecular mass of the corresponding protein in the range of 57,000 Da. The presence of a functional cleaved signal peptide and the similarity of t,he DOWN16 protein to the human serum protein factor H and to the factor H-related proteins h37/h42 and h27/h29 suggest that the native protein must be secreted into human serum.
Additional factor H-related mRNAs have previously been described in human liver, indicating the existence of a family of factor H-related transcripts (11). So far, four human factor H-related cDNA clones have been isolated. One cDNA represents the alternatively spliced 1.8-kb mRNA (24) and the three distinct factor H-related cDNA clones (DOWN16, DDDESK59, and H36-1) are all encoded by mRNAs of 1. 4 kb (11-13). Despite this large number of factor H-related cDNAs, Southern blot analysis with a specific fragment in-     four distinct factor H-related clones have also been isolated from a liver cDNA library, but their corresponding genes do

Following translation and translocation, intact microsomes (lane 3)
or Triton X-100 (7'X)-solubilized microsomes ( l o n e 4 ) were digested with proteinase K ( P K ) . Translation products were labeled by incorporation of ["S]methionine, separated on 15% SDS-polyacrylamide gels, and visualized by autoradiography. Molecular weight standards are given x~O-'. not overlap with the mouse factor H gene (25,26).
The known biological functions of factor H (inactivation of the alternative pathway convertases and cofactor activity for the enzyme factor I, as well as polyanion/heparin binding) have been mapped to a 38,000-Da tryptic fragment representing SCRsl-5 and to SCR13,respectively (27,28). As the related SCRs of DOWN16 and of factor H exclude these functionally characterized SCRs, the function of the DOWN16 protein seems distinct from that known for factor H. A detailed functional characterization of the protein encoded by the DOWN16 cDNA requires the isolation of the native protein. Except for the alternatively spliced human 1.8-kb mRNA, factor H and all human and murine factor Hrelated cDNAs (H36, DDESK59, and DOWNl6; 3A4, 23L1, 9C4, and 13G1) isolated so far have at their COOH-terminal end a SCR which is homologous to SCR19 of factor H (11-13, 25, 26). The presence of SCR19 in that large number of molecules suggests a relevant and highly conserved function.