Molecular identification of ADP-ribosylation factor mRNAs and their expression in mammalian cells

ADP-ribosylation factors (ARFs) are approximately 20-kDa guanine nucleotide-binding proteins that serve as GTP-dependent allosteric activators of cholera toxin ADP-ribosyltransferase activity. Four species of mammalian ARF, termed ARF 1-4, have been identified by cloning. Hybridization of a bovine ARF 2 cDNA under low stringency with mammalian poly(A)+ RNA resulted in multiple bands that were subsequently assigned to the known ARF genes using ARF-specific oligonucleotide probes. The relative signal intensities of some bands (e.g. the 3.8- and 1.3-kilobase (kb) mRNAs) that hybridized with the cDNA were not, however, consistent with the intensities observed with the individual ARF-specific oligonucleotide probes. These inconsistencies suggested that other ARF-like mRNAs were comigrating with known ARF mRNAs. To explore this possibility, a cyclic AMP-differentiated HL-60 Lambda ZAP library was screened using the bovine ARF 2 cDNA. Clones corresponding to known ARF genes (1, 3, and 4) were identified by hybridization of positive clones with oligonucleotide probes specific for each ARF species; ARF 2 cDNA-positive, oligonucleotide-negative clones were sequenced. Two new ARF-like genes, ARF 5 and 6, encoding proteins of 180 and 175 amino acids, respectively, were identified. Both proteins contain consensus sequences believed to be involved in guanine nucleotide binding and GTP hydrolysis. ARF 5 was most similar in deduced amino acid sequence to ARF 4, which also has 180 amino acids. ARF 6, whose deduced amino acid sequence is identical with that of a putative chicken pseudogene (CPS1) except for a serine/threonine substitution, was different from other ARF species in size and deduced amino acid sequence. With mammalian poly(A)+ RNA from a variety of tissues and cultured cells, ARF 5 preferentially hybridized with a 1.3-kb mRNA, whereas ARF 6 hybridized with 1.8- and 4.2-kb mRNAs. The fact that the sizes of these mRNAs are similar to those of other ARFs (ARF 1, 1.9 kb; ARF 2, 2.6 kb; ARF 3, approximately 3.8 and 1.3 kb; ARF 4, 1.8 kb) explain the previously observed inconsistencies between the cDNA and ARF-specific oligonucleotide hybridization patterns. All six ARF cDNAs are more similar to each other than to other approximately 20-kDa guanine nucleotide-binding proteins.     

Differential expression during development of ADP-ribosylation factors, 20-kDa guanine nucleotide-binding protein activators of cholera toxin

Cholera toxin exerts its effects on cells in large part through the ADP-ribosylation of guanine nucleotide-binding proteins. Toxin-catalyzed ADP-ribosylation is enhanced by approximately 20-kDa guanine nucleotide-binding proteins termed ADP-ribosylation factors (ARFs), which are allosteric activators of the toxin catalytic unit. Rabbit antiserum against a purified bovine brain ARF (sARF II) reacted on immunoblots with two approximately 20-kDa ARF-like proteins (sARF I and II) in tissue extracts from bovine, rat, frog, and chicken. Levels of ARF were higher in brain than in non-neural tissues. In rat brain, on the second postnatal day, amounts of sARF I and II were similar. By the 10th postnatal day and thereafter, sARF II predominated. Relative levels of ARF determined by immunoreactivity were in agreement with levels assessed in functional assays of cholera toxin-catalyzed ADP-ribosylation. Based on nucleotide and deduced amino acid sequences of human and bovine cDNAs, there appear to be at least six different ARF-like genes. Northern blots of rat brain poly(A)+ RNA were hybridized with cDNA and oligonucleotide probes specific for each of the human and bovine ARF genes. From the second to the 27th postnatal day, ARF 3 mRNA increased, whereas mRNAs for ARFs 2 and 4 decreased; and those for ARFs 1, 5, and 6 were apparently unchanged. Partial amino acid sequence of sARF II is consistent with it being either the ARF 1 or 3 gene product. The developmental changes in rat brain ARF parallel neuronal maturation and synapse formation.     

Multiple forms of Go alpha mRNA: analysis of the 3'-untranslated regions

Go, a guanine nucleotide binding protein found predominantly in neural tissues, interacts in vitro with rhodopsin, muscarinic, and other receptors and has been implicated in the regulation of ion channels. Despite the virtual identity of reported cDNA sequences for the alpha subunit of Go (Go alpha), multiple molecular weight forms of mRNA have been identified in tissues from all species examined. To investigate the molecular basis for the size heterogeneity of Go alpha mRNAs, four cDNA clones were isolated from the same retinal lambda gt10 cDNA library that was used earlier to isolate lambda GO9, a clone encompassing the complete coding region of Go alpha. These clones were identified as Go alpha clones based on nucleotide sequence identity with lambda GO9 in the coding region; they diverge, however, from lambda GO9 in the 3'-untranslated region 28 nucleotides past the stop codon. An oligonucleotide probe complementary to a portion of the 3'-untranslated region of lambda GO9 that differs from the newly isolated clones hybridized with 3.0- and 4.0-kb mRNAs present in bovine brain and retina whereas a similar probe for the unique region of the new clones hybridized with a 4.0-kb mRNA in both tissues and with a 2.0-kb mRNA found predominantly in retina. A similar hybridization pattern was observed when brain poly(A+) RNA from other species was hybridized with the different 3'-untranslated region probes. It appears that differences in the 3'-untranslated regions could, in part, be the basis for the observed heterogeneity in Go alpha mRNAs.     

Guanine nucleotide dependent formation of a complex between choleragen (cholera toxin) a subunit and bovine brain ADP-ribosylation factor

Cholera toxin activates adenylyl cyclase by catalyzing the ADP-ribosylation of Gs alpha, the stimulatory guanine nucleotide binding protein of the cyclase system. This toxin-catalyzed reaction, as well as the ADP-ribosylation of guanidino compounds and auto-ADP-ribosylation of the toxin A1 protein (CTA1), is stimulated, in the presence of GTP (or GTP analogue), by 19-21-kDa proteins, termed ADP-ribosylation factors or ARFs. These proteins directly activate CTA1 in a reaction enhanced by sodium dodecyl sulfate (SDS) or dimyristoylphosphatidylcholine (DMPC)/cholate. To determine whether ARF stimulation of ADP-ribosylation is associated with formation of a toxin-ARF complex, these proteins were incubated with guanine nucleotides and/or detergents and then subjected to gel permeation chromatography. An active ARF-toxin complex was observed in the presence of SDS and GTP gamma S [guanosine 5'-O-(3-thiotriphosphate)] but not GDP beta S [guanosine 5'-O-(2-thiodiphosphate)]. Only a fraction of the ARF was capable of complex formation. The substrate specificities of complexed and noncomplexed CTA differed; complexed CTA exhibited markedly enhanced auto-ADP-ribosylation. In the presence of GTP gamma S and DMPC/cholate, an ARF-CTA complex was not detected. A GTP gamma S-dependent ARF aggregate was observed, however, exhibiting a different substrate specificity from monomeric ARF. These studies support the hypothesis that in the presence of guanine nucleotide and either SDS or DMPC/cholate, ARF and toxin exist as multiple species which exhibit different substrate specificities.     

Complete murine cDNA sequence, genomic structure, and tissue expression of the high mobility group protein HMG-I(Y)

A cDNA coding for the non-histone chromosomal protein HMG-I, or its isoform HMG-Y, was isolated from a murine Friend cell library using synthetic oligonucleotide hybridization probes. Sequence analysis showed that the 1670-base pair full length cDNA insert consists of a 201-base pair, G/C-rich (74%), 5'-untranslated region, a 288-base pair amino acid coding sequence, and an unusually long 1182-base pair 3'-untranslated region. The deduced 96-residue amino acid coding sequence of the murine HMG-I(Y) cDNA is very similar to the reported amino acid sequence of human HMG-I, except that it lacks 11 internal amino acids reported in the human protein. Based on Southern blot hybridization analysis of genomic DNA, there appear to be fewer than five copies of HMG-I(Y) genes in the haploid murine genome. These murine HMG-I(Y) genes contain a large (at least 890 base pairs) exon that includes most, or all, of the 3'-untranslated region; whereas the much shorter 5'-untranslated region and amino acid coding sequences are interrupted by at least one intron. A single size class (approximately 1700 nucleotides in murine cells and 2000 nucleotides in human cells) of HMG-I(Y) mRNAs was detected at high levels in total RNA extracts from rapidly dividing, transformed cells, but to a lesser extent, or not at all, in extracts from slowly or non-dividing cells.     

A single gene codes for two forms of rat nucleolar protein B23 mRNA

Protein B23 (38 kDa, pI = 5.1) is an abundant RNA-associated nucleolar phosphoprotein and putative ribosome assembly factor. A full length cDNA clone (lambda JH1) encoding a major expressed form of rat protein B23, now designated B23.1, was reported recently (Chang, J. H., Dumbar, T. S., and Olson, M. O. J. (1988) J. Biol. Chem. 263, 12824-12837). In this paper the isolation from a rat brain library and sequence of a cDNA clone (lambda JH2) coding for a second form (B23.2) of protein B23 is reported. Isoforms B23.1 and B23.2 are polypeptides of 292 and 257 amino acids, respectively. The 5'-untranslated regions of the two cDNAs and the amino-terminal 255 amino acids of the proteins are identical in the two isoforms. However, the 3'-untranslated regions of the mRNAs are completely different, and the dipeptide Gly-Gly in B23.1 (residues 256 and 257) is replaced by Ala-His in B23.2 indicating that the former is not a precursor of the latter. The finding of AGGT sequences in the 3' regions of lambda JH1 suggest the presence of intron-exon boundaries at the point where the two cDNAs begin to differ. To investigate the origin of the two isoforms, two rat genomic libraries were screened with oligonucleotide probes based on sequences from the unique regions of the two cDNAs. One of the genomic clones isolated (lambda JH125) contained a 6.5-kilobase fragment encoding the 3' end of both cDNAs. lambda JH125 contains four exons designated W, X, Y, and Z in the order indicated. Exons W and X encode 36 amino acids at the carboxyl terminus of B23.2, whereas exons W, Y, and Z encode the carboxyl-terminal 71 amino acid residues of B23.1. Exons X and Z each contain distinct 3'-untranslated sequences in which are found polyadenylation signals. These data suggest that two different mRNAs are formed by alternative splicing of separate 3' segments onto a common 5' region.     

Interspecies relationships among ADP-ribosylation factors (ARFs): Evidence of evolutionary pressure to maintain individual identities

ADP-ribosylation factors (ARFs) are ～20-kDa guanine nucleotide-binding proteins that are allosteric activators of the NAD:arginine ADP-ribosyltransferase activity of cholera toxin and appear to play a role in intracellular vesicular trafficking. Although the physiological roles of these proteins have not been defined, it has been presumed that each has a specific intracellular function. To obtain genetic evidence that each ARF is under evolutionary pressure to maintain its structure, and presumably function, rat ARF cDNA clones were isolated and their nucleotide and deduced amino acid sequences were compared to those of other mammalian ARFs. Deduced amino acid sequences for rat ARFs 1, 2, 3, 5 and 6 were identical to those of the known cognate human and bovine ARFs; rat ARF4 was 96% identical to human ARF4. Nucleotide sequences of both the untranslated as well as the coding regions were highly conserved. These results indicate that the ARF proteins are, as a family, extraordinarily well conserved across mammalian species. The unusually high degree of conservation of the untranslated regions is consistent with these regions having important regulatory roles and that individual ARFs contain structurally unique elements required for specific functions.     

Mechanism of activation of cholera toxin by ADP-ribosylation factor (ARF): both low- and high-affinity interactions of ARF with guanine nucleotides promote toxin activation

Activation of adenylyl cyclase by cholera toxin A subunit (CT-A) results from the ADP-ribosylation of the stimulatory guanine nucleotide binding protein (GS alpha). This process requires GTP and an endogenous guanine nucleotide binding protein known as ADP-ribosylation factor (ARF). One membrane (mARF) and two soluble forms (sARF I and sARF II) of ARF have been purified from bovine brain. Because the conditions reported to enhance the binding of guanine nucleotides by ARF differ from those observed to promote optimal activity, we sought to characterize the determinants influencing the functional interaction of guanine nucleotides with ARF. High-affinity GTP binding by sARF II (apparent KD of approximately 70 nM) required Mg2+, DMPC, and sodium cholate. sARF II, in DMPC/cholate, also enhanced CT-A ADP-ribosyltransferase activity (apparent EC50 for GTP of approximately 50 nM), although there was a delay before achievement of a maximal rate of sARF II stimulated toxin activity. The delay was abolished by incubation of sARF II with GTP at 30 degrees C before initiation of the assay. In contrast, a maximal rate of activation of toxin by sARF II, in 0.003% SDS, occurred without delay (apparent EC50 for GTP of approximately 5 microM). High-affinity GTP binding by sARF II was not detectable in SDS. Enhancement of CT-A ADP-ribosyltransferase activity by sARF II, therefore, can occur under conditions in which sARF II exhibits either a relatively low affinity or a relatively high affinity for GTP. The interaction of GTP with ARF under these conditions may reflect ways in which intracellular membrane and cytosolic environments modulate GTP-mediated activation of ARF.     

Demonstration of an unusual allelic variation of mouse factor H by the complete cDNA sequence of the H.2 allotype

Three allotypes of murine factor H have been identified serologically in the previous study (denoted H.1, H.2, and H.3). A cDNA clone coding for the entire length of murine factor H was isolated from a library constructed from the livers of STR/N mice which have H.2 allotype and was fully sequenced. The insert of this clone (STR309) contained 4184 nucleotides and consisted of a 47-bp 5' noncoding region, a 54-bp coding for leader peptide, a 3648 bp for the mature factor H protein, and a 435-bp 3' noncoding region. Compared with the previously reported sequence of the cDNA clone (MH8) isolated from B10.WR mice that have H.1 allotype, the size of the protein coding region was exactly the same, but 21 nucleotide substitutions resulting in 15 amino acid replacements were observed. The amino acid replacement/nucleotide substitution ratio (0.71) is far higher than those observed in the allotypic variations of other proteins. Four 15-base oligonucleotide probes specific for either STR309 or MH8 were synthesized and used in Northern blot analysis. The probes specific for STR309 hybridized with mRNA isolated from the livers of STR/N mice but not with mRNA from the livers of BALB/c mice that have H.1 allotype, whereas the reverse pattern was observed with the oligonucleotide probes specific for MH8. These results strongly suggest that the nucleotide sequence of STR309 represents H.2 allotype of factor H protein, providing an example of an unusual allotype with high ratio of amino acid replacements to nucleotide substitutions.     

Stimulation of choleragen enzymatic activities by GTP and two soluble proteins purified from bovine brain

Choleragen (cholera toxin) activates adenylate cyclase by catalyzing ADP-ribosylation of Gs alpha, the stimulatory guanine nucleotide-binding protein. It was recently found (Tsai, S.-C., Noda, M., Adamik, R., Moss, J., and Vaughan, M. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 5139-5142) that a bovine brain membrane protein known as ADP-ribosylation factor or ARF, which enhances ADP-ribosylation of Gs alpha, also increases the GTP-dependent NAD:arginine and NAD:protein ADP-ribosyltransferase, NAD glycohydrolase, and auto-ADP-ribosylation activities of choleragen. We report here the purification and characterization of two soluble proteins from bovine brain that similarly enhance the Gs alpha-dependent and independent ADP-ribose transfer reactions catalyzed by toxin. Like membrane ARF, both soluble factors are 19-kDA proteins dependent on GTP or GTP analogues for activity. Maximal ARF effects were observed at a molar ratio of less than 2:1, ARF/toxin A subunit. Dimyristoyl phosphatidylcholine was necessary for optimal ADP-ribosylation of Gs alpha but inhibited auto-ADP-ribosylation of the choleragen A1 subunit and NAD:agmatine ADP-ribosyltransferase activity. It appears that the soluble factors directly activate choleragen in a GTP-dependent fashion. The relationships of the ARF proteins to the ras oncogene products and to the family of guanine nucleotide-binding regulatory proteins that includes Gs alpha remains to be determined.     

Multiple mRNA species code for the catalytic subunit of the cAMP-dependent protein kinase from LLC-PK1 cells. Evidence for two forms of the catalytic subunit

We present evidence for the existence of two forms of the catalytic (C) subunit of the cAMP-dependent protein kinase. A lambda gt-11 cDNA library constructed from poly(A)-rich RNA from the porcine kidney cell line, LLC-PK1, was screened using a 1.5-kb EcoRI fragment from a bovine cDNA for the C subunit. Two independent classes of cDNAs were identified on the basis of partial restriction map and sequence data. These two cDNAs, lambda CAT4 and lambda CAT3, apparently encode two forms of C subunit designated C alpha and C beta, respectively. The nucleotide sequence of the C alpha and C beta cDNAs revealed differences in the coding region and particularly in the 3' untranslated region. However, the deducted amino acid sequences of C alpha and C beta subunits were 96% homologous to the sequences so far determined. Specific probes from the 3' coding region of the two cDNA species were used to investigate C subunit mRNA expression in LLC-PK1 cells. Northern analysis showed a major mRNA species of 2.8 kb with the C alpha probe while the C beta probe detected two mRNA species of 5.0 kb and 3.8 kb. These data were supported by genomic blot analysis which showed distinct hybridization patterns with either the C alpha or C beta probes. All the available evidence suggests that at least two distinct genes encode the C subunit which are expressed in LLC-PK1 cells.