key: cord-010278-loey5xq9
authors: Huh, Changgoo; Nagle, James W.; Kozak, Christine A.; Abrahamson, Magnus; Karlsson, Stefan
title: Structural organization, expression and chromosomal mapping of the mouse cystatin-C-encoding gene (Cst3)
date: 1995-01-23
journal: Gene
DOI: 10.1016/0378-1119(94)00728-b
sha: 
doc_id: 10278
cord_uid: loey5xq9

Cystatin C (CstC) is a potent cysteine-proteinase inhibitor. The structure of the mouse CstC-encoding gene (Cst3) was examined by sequencing a 6.1-kb genomic DNA containing the entire gene, as well as 0.9 kb of 5′ flanking and 1.7 kb of its 3′ flanking region. The sequence revealed that the overall organization of the gene is very similar to those of the genes encoding human CstC and other type-2 Cst, with two introns at positions identical to those in the human gene. The promoter area does not contain typical TATA or CAAT ☐es. Two copies of a Spl-binding motif, GGGCGG, are present in the 5′ flanking region within 300 bp upstream from the initiation codon. A hexa-nucleotide, TGTTCT, which is a core sequence of the androgen-responsive element (ARE), is found in the promoter region. This region also contains a 21-nucleotide sequence, 5′-AGACTAGCAGCTGACTGAAGC, which contains two potential binding sites for the transcription factor, AP-1. The mouse Cst3 mRNA was detected in all of thirteen tissues examined by Northern blot analysis. Cst3 was mapped in the mouse to a position on distal chromosome 2.

The cystatins (Cst) are a group of potent cysteineproteinase inhibitors. There are at least five distinct types in the Cst superfamily, each type consisting of several proteins (Rawlings and Barrett, 1990; Devos et al., 1993) . CstC belongs to the family of type-2 Cst and consists of 120 aa, with two intrachain disulfide bonds (Barrett et al., 1986) .

Although the proteinase-inhibiting function of the CstC has been thoroughly investigated, less is known about its broader biological role. Recent reports indicate that CstC may play a role in cancer progression (Sloane, 1990) , bone resorption (Lerner and Grubb, 1992) , modulation of neutrophil chemotactic activity and inflammation (Leung-Tack et al., 1990a,b) , and resistance to viral infection (Collins and Grubb, 1991) . Furthermore, a point mutation in the CST3 gene, resulting in a Leu~Gln substitution, is the primary cause of autosomal dominant hereditary disorder, hereditary CstC amyloid angiopathy (HCCAA) (Grubb et al., 1984) . As young adults, carriers of this mutation suffer from repeated and massive brain hemorrhages due to deposition of the mutant protein in the walls of the cerebral arteries.

The gene structures for several type-2 Cst have been determined. The human CST1 and CST2 genes (Saitoh et al., 1987) , and the human CST3 (Abrahamson et al., 1990) and CST4 (Freije et al., 1991) genes show very similar structural organization with respect to the number and position of the introns. Structural analysis of the genes for these proteins will be necessary to understand the function and evolution of the members in the cystatin multigene superfamily. In this paper, we report the structural organization and expression of the mouse Cst3 gene, compare its regulatory elements with that of other Cst genes and map the Cst3 gene in the mouse.

Using two primers mcyc3: (5'-ATG GCC AGC CCG CTG CGC TCC TTG-3') and mcyc4: (5'-GGC ATT TTT GCA GCT GAA TTT TGT CAG-3'), a 420-bp DNA fragment within the coding region was generated from mouse Cst3 cDNA, labeled with 32p by random primer extension, and used as a probe to screen a ~,FixlI genomic DNA library from 129/Sv mice (purchased from Stratagene, La Jolla, CA, USA). Among several hybridizing clones, one clone, ~,cygl3, was chosen for characterization. Southern blot and PCR analysis indicated the presence of the mouse Cst3 gene in a 15-kb DNA insert. Southern blot analysis of the genomic DNA did not show any evidence for the presence of a Cst3 pseudogene. A 6.1-kb genomic DNA fragment containing the entire gene, as well as 0.9 kb of 5' flanking and 1.7 kb of 3' flanking region, was subjected to nt sequencing.

The 6125-nt sequence covering the entire mouse Cst3 gene is shown in Fig. 1 . Comparison of the mouse Cst3 gene sequence with that of the corresponding cDNA (Solem et al., 1990) revealed that the gene contains two intron sequences located between the nt triplets encoding aa 55-56 and 93-94 of the proposed mature polypeptide chain, exactly as in the human CST3 gene. The presence of two introns, at homologous positions, has also been reported in the other type-2 Cst genes fully characterized to date: the human CSTI, CST2 and CST4 genes. The intron-exon junctions in the mouse Cst3 gene are all close matches to the consensus sequences for the donor and acceptor splice sites of introns (Mount, 1982) .

Some differences were observed between the exons of the genomic DNA sequence and the published cDNA, The five differences are summarized in Table 1 . Two of these positions, 994 and 3421, are in the coding region. The nt 994 in exon 1 results in a GCC codon (coding for Ala) towards the C-terminal part of the leader sequence. However, a GGC codon (coding for Gly) was reported at this position in the published cDNA. This GCC codon found in the genomic DNA exists in the corresponding site of the rat and human Cst3 cDNAs. The nt 342~ in exon 2 forms a TTG codon for Leu (TTT coding for Phe in cDNA). The differences between genomic and cDNA sequence may be due to an error during cDNA synthesis or due to polymorphism between the mouse strains 129/Sv and BALB/c.

The sequence of the 0.9-kb segment flanking exon 1 of the mouse Cst3 gene at the 5' end, did not reveal a typical TATA or CAAT box in the suggested promoter area. However, a TATA-box-like TAAAA sequence is present at 78-82 nt upstream from the start codon. A similar slightly atypical TATA-box is found at the homologous position in the human CST3 gene (ATAAAA), the human CST4 gene (ATAAAT), the human CSTI and CST2 genes (ATAAA). The TATA-box is preceded by a Spl-binding GC-box sequence (Pugh and Tjian, 1990) with the core consensus sequence, GGGCGG, ending 23 nt upstream from the AT-sequence. A corresponding sequence in the human CST3 gene is located slightly closer to the TATAbox (distance 16 nt). In the human CST4 gene, a GC-box is also found in the immediate 5'-flanking region (upstream distance from the AT-sequence 41 nt). By contrast, in the human CSTI and CST2 genes, a segment similar to the CAAT consensus is found instead of the GC box.

A core sequence of the ARE, TGTTCT, is found 22 nt downstream from the TAAAA sequence. This hexanucleotide is located in a partial palindromic setting. Some naturally occurring sequences and synthetic constructs containing this core sequence in a partial palindromic structure were shown to be inducible with androgens (Ham et al., 1988) . This ARE is not found in the promoters of any other type-2 Cst genes published. However, it is present in Cst-related protein-encoding genes whose expression is regulated by androgen in the ventral prostate and lachrymal gland (Chamberlain et al., 1983 ).

An exact match of nine nt with the pituitary transcription factor (Pit-l) recognition element is centered around nt -795 from the start codon, but is probably of low significance for the expression of the gene because multiple recognition elements have been shown to be needed for markedly increased expression of the rat prolactin gene by Pit-1 (Ingraham et al., 1988 recognized by the leader binding protein (LBP-1), 5'-WCTGG-3' or its inverse, that is present in several copies in the HIV-1 promoter and contribute to its basal function (Jones et al., 1988) , is strikingly abundant in the 5'-flanking region of the mouse Cst3 gene. Another five LBP-I motifs are found within a 300-bp segment in the first part of the first intron.

Transcription factor AP-l-binding sites that bind the Jun-Fos protooncogene complexes contain the consensus sequence 5'-TGACTCAGC. The mouse Cst3 promoter contains two AP-l-like binding sequences within the sequence 5'-AGACTAGCAGCTGACTGAAGC, immediately upstream from the first Spl-binding site. This 21-mer sequence contains direct repeats of two adjacent potential AP-l-binding sites, each slightly deviating from the consensus sequence. It has been shown that two adjacent AP-l-like binding sites act synergistically to confer inducibility beyond that observed for a single AP-1 consensus sequence (Friling et al., 1992) . The presence of the two AP-l-like binding sites in the promoter indicates that Differences between the mouse Cst3 gene sequence and that of the published eDNA (Solem et al., 1990) Position"

Genomic DNA b cDNA c aa a (genomic/cDNA) 994 C (11) G Ala/Gly 3421 G(5) T Leu/Phe 4393 C 15) T 4564-5 GC (3) AT a The nt positions refer to Fig. 1 . b The nt refers to the genomic DNA sequence. Sequence was determined on both strands (number of independent sequencing runs in parentheses). c The nt refers to the cDNA sequence. d The aa encoded by nt in columns Genomic DNA and cDNA.

transcription factor AP-1 may play a role in the Cst3 gene expression. There is evidence that induction of gene expression by TGF-13 is mediated by transcription factor AP-1. Autoregulation of TGF-131 expression is mediated by the binding of AP-1 to a loose consensus binding site, TGAGACA, in the TGF-131 promoter (Kim et al., 1990) . A strong positive regulation of the Cst3 gene by TGF-13 in serum free mouse embryo cells has also been reported (Solem et al., 1990) . The presence of AP-l-like binding sites in the mouse Cst3 promoter suggests that cystatin C induction by TGF-[3 may be mediated by the AP-I complex,

The Y-flanking region of the human CST3 gene has a notably high G+C content, with >70% G+C in the 400 bp sequence upstream from the start codon. The G+C-rich region also includes the coding part of exon 1 and the 5' part of the first intron, which together represents a 900-bp segment with a G+C content of 73%, and contains CpG/GpC dinucleotides in a ratio close to unity (Abrahamson et al., 1990) . The immediate 5'-flanking region of the mouse Cst3 gene does not have such a strikingly high G + C content, but is more similar to the human CST1, CST2 and CST4 genes in having a GC content of approx. 60%. However, the CpG/GpC ratio is 1/1.3 in the 300 bp region upstream from the start codon (as compared to 1/4.1 over the entire 6. l-kb sequenced region), differing markedly from ratios of 1/6, 1/9 and 1/16 for the human CST4, CSTI and CST2 genes, respectively. Thus, the mouse Cst3 gene Y-flanking region is not a typical housekeeping gene promoter having extremely high G + C content. Rather it is similar to these promoters and the human CST3 promoter because it displays several Spl-binding sites and contains a high number of CpG dinucleotides. This may indicate a low degree of methylation due to constant transcription of the gene (Bird, 1986) . Proposed promoter regions of several type-2 Cst are compared in Fig. 2 (Saitoh et al., 1987 : Abrahamson et al., 1990 : Freije et al., 1991 .

Sequence determination of the mouse Cst3 gene 1.7-kb 3'-flanking segment revealed no alternative polyadenylation signals in addition to the one present in the corresponding eDNA, 213bp downstream from the stop codon. Analysis of short tandem repeats within the entire gene sequence revealed the presence of(GT)zl and (GA)26 in the region immediately 3'-flanking the polyadenylation signal and three stretches of perfect CT repeats, (CT)13, (CT)14 and (CT)25, 1300 bp further downstream.

Analysis of two multilocus crosses .was used to define the Chr location for the mouse Cst3 gene: (NFS/N or C58/J × M. m. musculus) × M. m. musculus ) and (NFS/N ×M. spretus)×M, spretus or C58/J (Adamson et al., 1991) . DNAs extracted from parental mice and progeny of the crosses were typed by Southern blotting analysis for RFLPs of Cst3 using the mouse Cst3 cDNA as probe. SstI digestion produced fragments of 8.6 and 6.0 kb in M. spretus and 6.0 and 2.8 kb in NFS/N and C58/J mice, and BamHI fragments of 22.0 and 18.8 kb were detected in M. m. musculus and NFS/N, respectively.

Inheritance of the parental fragments was followed in both crosses and compared with inheritance of almost 650 markers previously typed and mapped to all 19 autosomes and the X chromosome. As shown in Fig. 3 The numbers given between adjacent loci represent percent recombination±the standard error calculated according to Green (1981) . (B) Chr 2 linkage maps. The map on the right was generated from the two crosses described here and indicates the position of Cst3 relative to the other markers typed in this cross. Distances between adjacent markers (in centiMorgans) are indicated to the immediate left of the map. The map on the left is an abbreviated version of the composite genetic map (Siracusa and Abbott, 1993) . Numbers to the left of the map are centiMorgan distances from the centromere. Human map locations for homologs of the underlined mouse genes are indicated to the far left of this map.

and distal to Snap (encoding synaptosomal associated protein), markers which were typed in these crosses as previously described (Joseph et al., 1990; Grimaldi et al., 1992) . It has recently been shown that the human homolog of this gene maps to 20pll (Schnittger et al., 1993) . This is consistent with our results, since the distal end of mouse Chr 2 contains a substantial region of linkage conservation with human Chr 20 (Siracusa and Abbot, 1993; Fig. 3b ). The human CST3 is part of a cluster which includes up to eight members of the CST gene family (Schnittger et al., 1993) further suggesting that the mouse homologs of these genes are likely to map to the same site on Chr 2.

We examined the expression of the mouse Cst3 gene in 13 different tissues by Northern blot analysis using the mouse Cst3 cDNA probe. As expected, Cst3 mRNA was detected in all tissues examined, including stomach, brain, intestines, liver, muscle, spleen, heart, kidney, lung, pancreas, testis, uterus and ovary (data not shown). The pattern of the mouse Cst3 gene expression is similar to that of its human counterpart. Both species show expression of the gene in all tissues examined with high level of Cst3 messenger RNA in brain and testis, and lowest level in pancreas. This overall similarity between the two species indicates that mouse may be suitable for generating an animal model for the human genetic disease HCCAA.

Structure and expression of the human cystatin C gene

The mouse homolog of the gibbon ape leukemia virus receptor: genetic mapping and a possible receptor function in rodents

Nomenclature and classification of the proteins homologous with the cysteine-proteinase inhibitor chicken cystatin

CpG-rich islands and the function of DNA methylation

Isolation, properties, and androgen regulation of a 20-kilodalton protein from rat ventral prostate

Inhibitory effects of recombinant human cystafin C on human coronaviruses

Structure of rat genes encoding androgen-regulated cystatin-related proteins (CRPs): a new member of the cystatin superfamily

Structure and expression of the gene encoding cystatin D, a novel human cysteine proteinase inhibitor

Two adjacent AP-l-like binding sites form the electrophile-responsive element of the murine glutathione S-transferase Ya subunit gene

Genetics and Probability in Animal Breeding Experiments

Genomic structure and chromosomal mapping of the murine CD40 gene

Abnormal metabolism of 7-trace alkaline microprotein

Characterization of response elements for androgens, glucocorticolds and progestins in mouse mammary tumour virus

A tissue-specific transcription factor containing a homeodomain specifies a pituitary phenotype

Structural arrangements of transcription control domains within the 5'-untranslated leader regions of the HIV-I and HIV-2 promoters

Characterization and expression of the complementary DNA encoding rat histidine decarboxylase

Autoinduction of transforming growth factor 131 is mediated by the AP-1 complex

Molecular genetic markers spanning mouse chromosome 10

Neutrophil chemotactic activity is modulated by human cystatin C, an inhibitor of cysteine proteases

Modulation of phagocytosis-associated respiratory burst by human cystatin C: role of the N-terminal tetrapeptide Lys-Pro-Pro-Arg

Human cystatin C, a cysteine proteinase inhibitor, inhibits bone resorption in vitro stimulated by parathyroid hormone and parathyroid hormone-related peptide of malignancy

A catalogue of splice junction sequences

Mechanism of transcriptional activation by Sp 1; evidence for coactivators

Evolution of proteins of the cystatin superfamily

Human cysteineproteinase inhibitors: nucleotide sequence analysis of three members of the cystatin gene family

Cystatin C (CST3), the candidate gene for heretary cystatin C amyloid angiopathy (HCCAA), and other members of the cystatin gene family are clustered on chromosome 20pl 1.2

Mouse chromosome 2. Mammal

Cathepsin B and cystatins: evidence for a role in cancer progression

Transforming growth factor beta regulates cystatin C in serum-free mouse embryo (SFME) cells

We thank Dr. Jakob Reiser for critical reading of the manuscript.