Caspase-mediated cleavage of the U snRNP-associated Sm-F protein during apoptosis

Recent studies have implicated the dying cell as a potential reservoir of modified autoantigens that might initiate and drive systemic autoimmunity in susceptible hosts. The spliceosomal Sm proteins are recognized by the so-called anti-Sm autoantibodies, an antibody population found exclusively in patients suffering from systemic lupus erythematosus. We have studied the effects of apoptosis on the Sm proteins and demonstrate that one of the Sm proteins, the Sm-F protein, is proteolytically cleaved in apoptotic cells. Cleavage of the Sm-F protein generates a 9-kDa apoptotic fragment, which remains associated with the U snRNP complexes in apoptotic cells. Sm-F cleavage is dependent on caspase activation and the cleavage site has been located near the C-terminus, EEED(81) downward arrow G. Use of different caspase inhibitors suggests that besides caspase-8 other caspases are implicated in Sm-F cleavage. A C-terminally truncated mutant of the Sm-F protein, representing the modified form of the protein, is capable of forming an Sm E-F-G complex in vitro that is recognized by many anti-Sm patient sera.     

A comparison of snRNP-associated Sm-autoantigens: human N, rat N and human B/B''.

N is a tissue-specific, Sm-epitope bearing, snRNP-associated protein found predominantly in brain. The cDNA sequence encoding human N is compared to those for rat N and human B/B'. The amino acid sequences of human and rat N are 100% conserved. Although the amino acid sequences of N and B/B' are very similar to each other, B/B' contains 50 amino acids which are not present in N. On Northern blots the cDNAs encoding N and B/B' recognize two different RNA species. A comparison of the codon usage, as specified by the open reading frames of N and B/B' as well as results from Southern blots, show that N and B/B' are derived from different genes.     

The [U4/U6.U5] tri-snRNP-specific 27K protein is a novel SR protein that can be phosphorylated by the snRNP-associated protein kinase.

SR proteins play important roles in the recognition and selection of the 3' and 5' splice site of a given intron and contribute to the phosphorylation/dephosphorylation-mediated regulation of pre-mRNA splicing. Recent studies have demonstrated that the U1 snRNP is recruited to the 5' splice site by protein/protein interactions involving the SR domains of the U1-70K protein and SF2/ASF. Recently, it was suggested that SR proteins might also contribute to the binding of the [U4/U6.U5] tri-snRNP to the pre-spliceosome (Roscigno RF, Garcia-Blanco MA, 1995, RNA 1:692-706), although it remains unclear whether these SR proteins interact with proteins of the tri-snRNP complex. As a first step toward the identification of proteins that could potentially mediate the integration of the [U4/U6.U5] tri-snRNP complex into the spliceosome, we investigated whether purified [U4/U6.U5] tri-snRNP complexes contain SR proteins. Three proteins in the tri-snRNP complex with approximate molecular weights of 27, 60, and 100 kDa were phosphorylated by purified snRNP-associated protein kinase, which has been shown previously to phosphorylate the serine/ arginine-rich domains of U1-70K and SF2/ASF (Woppmann A et al., 1993, Nucleic Acids Res 21:2815-2822). These proteins are thus prime candidates for novel tri-snRNP SR proteins. Here, we describe the biochemical and molecular characterization of the 27K protein. Analysis of a cDNA encoding the 27K protein revealed an N-terminal SR domain strongly homologous (54% identity) to the SR domain of the U1 snRNP-specific 70K protein. In contrast to many other SR proteins, the 27K protein does not contain an RNA-binding domain. The 27K protein can be phosphorylated in vitro by the snRNP-associated protein kinase and exhibits several isoelectric variants upon 2D gel electrophoresis. Thus, the tri-snRNP-specific 27K protein could potentially be involved in SR protein-mediated protein/protein interactions and, additionally, its phosphorylation state could modulate pre-mRNA splicing.     

Filaggrin, an intermediate filament-associated protein: structural and functional implications from the sequence of a cDNA from rat.

Filaggrin is an intermediate filament-associated protein that is involved in aggregation of keratin filaments in fully cornified cells of the mammalian epidermis, and is an important marker for epidermal differentiation. In this report, the sequence of a rat cDNA clone coding for a portion of the polymeric precursor, profilaggrin, is presented. The cDNA is 2,314 bp long with 1,875 bp of coding region ending with an A-T-rich 3' noncoding region. Genomic analysis indicates that the profilaggrin gene consists of 20 +/- 2 repeats of 1,218 bp of sequence coding for 406 amino acids, making the mRNA at least 25-27 kb in length. Each repeat consists of a filaggrin domain and a linker sequence with an estimated size of 380 and 26 amino acids, respectively. High levels of profilaggrin mRNA are found only in keratinizing epithelia. Comparison of the rat filaggrin sequence with that of mouse and human filaggrin and with the sequence of phosphorylated peptides from mouse profilaggrin indicates that the proteins share extensive amino acid sequence similarities, especially in the two phosphorylated regions. Proteolytic processing sites are also quite similar in rat and mouse. The three species show blocks of sequence that are similar in length and composition which alternate with sequences that are variable in length. This analysis suggests that the evolution of the present-day filaggrins has been constrained by maintenance of phosphorylation sites and overall amino acid composition. The cDNAs for the profilaggrins are similar in structure, reflecting genes that have simple repeating structures and lack introns within their coding regions. Mouse and rat profilaggrin terminate with a nonpolar sequence atypical of the rest of the coding region, and have similar 3' noncoding regions. To explain these observations, a novel evolutionary model is proposed.     

Nucleotide sequence of cloned cDNA specific for rat ribosomal protein L35a

A cDNA clone specific for rat ribosomal protein L35a, which is known to be a tRNA-binding protein, was isolated by hybrid-selected translation from a cDNA library made for 8-9-S poly(A)-rich RNA from regenerating rat liver. The nucleotide sequence of the cDNA was determined. It consists of one base pair from the 5' leading sequence, the entire coding sequence of 333 base pairs and 14 base pairs from the 3' trailing sequence. The primary structure of protein L35a was deduced from the nucleotide sequence. It consists of 109 amino acids with a molecular mass of 12422. The calculated amino acid composition is consistent with that reported for the hydrolysate of L35a. The amino acid sequence showed marked homology with the reported partial sequence of Xenopus leavis ribosomal protein L32, but not significant homology with Escherichia coli ribosomal proteins that bind to tRNA.     

Amino acid sequence deduced from a rat kidney cDNA suggests it encodes the Zn-peptidase aminopeptidase N

We have isolated and characterized rat kidney cDNA clones encoding a 140-kDa glycoprotein that exhibits characteristics of a cell surface Zn-peptidase. Structural features predicted for this putative kidney Zn-peptidase (KZP) are most consistent with properties previously determined for the Zn-peptidase aminopeptidase N. The deduced amino acid sequence of rat KZP is almost identical to the NH2-terminal sequence of aminopeptidase N purified from rabbit. The overall amino acid composition predicted for rat KZP is remarkably similar to that previously determined for rabbit and pig aminopeptidase N. The predicted Mr of rat kidney KZP approximates the Mr of the unglycosylated form of aminopeptidase N. The topology predicted for KZP is identical to that observed for aminopeptidase N: a short cytoplasmic domain at the NH2 terminus immediately precedes an uncleaved signal/anchor domain; a stalk region connects this membrane anchor to the extracellular, hydrophilic bulk of the protein containing catalytic sites required for Zn-peptidase activity. In addition, mRNA encoding KZP is present in tissues known to exhibit aminopeptidase N activity. Taken together with the observation that only a single gene homologous to KZP DNA is present in the rat and human genomes, these results suggest that we have established the primary structure of rat kidney aminopeptidase N.