- Top of page
- Materials and methods
Four different isoforms of the catalytic subunit of cAMP-dependent protein kinase, termed Cα, Cβ, Cγ and PrKX have been identified. Here we demonstrate that the human Cβ gene encodes six splice variants, designated Cβ1, Cβ2, Cβ3, Cβ4, Cβ4ab and Cβ4abc. The Cβ splice variants differ in their N-terminal ends due to differential splicing of four different forms of exon 1 designated exon 1-1, 1-2, 1-3, 1-4 and three exons designated a, b and c. All these exons are located upstream of exon 2 in the Cβ gene. The previously identified human Cβ variant has been termed Cβ1, and is similar to the Cβ isoform identified in the mouse, ox, pig and several other mammals. Human Cβ2, which is the homologue of bovine Cβ2, has no homologue in the mouse. Human Cβ3 and Cβ4 are homologous to the murine Cβ3 and Cβ2 splice variants, whereas human Cβ4ab and Cβ4abc represent novel isofoms previously not identified in any other species. At the mRNA level, the Cβ splice variants reveal tissue specific expression. Cβ1 was most abundantly expressed in the brain, with low-level expression in several other tissues. The Cβ3 and Cβ4 splice variants were uniquely expressed in human brain in contrast to Cβ2, which was most abundantly expressed in tissues of the immune system, with no detectable expression in brain.
We suggest that the various Cβ splice variants when complexed with regulatory subunits may give rise to novel holoenzymes of protein kinase A that may be important for mediating specific effects of cAMP.
Cyclic 3′,5′-adenosine monophosphate (cAMP) is a key intracellular signaling molecule, the main function of which is to activate the cAMP-dependent protein kinases . Protein kinase A is a heterotetrameric enzyme containing a regulatory (R) subunit dimer and two catalytic (C) subunits. The holoenzyme is activated when four molecules of cAMP bind to the R subunit dimer, two to each R subunit, releasing two free active C subunits . In human enzyme, four different R subunits (RIα, RIβ RIIα, RIIβ), and four different C subunits (Cα, Cβ, Cγ and PrKX) have been identified . The Cα and Cβ subunits are expressed in most tissues, while the Cγ subunit, which is transcribed from an intron-less gene and represents a retroposon derived from the Cα subunit , is only expressed in human testis . PrKX is an X chromosome-encoded protein kinase, and was recently identified as a protein kinase A C subunit because it is inhibited by both protein kinase inhibitor (PKI) and regulatory subunit Iα (RIα), and because the RIα/PrKX complex is activated by cAMP .
Splice variants of both Cα and Cβ have been identified. The splice variants encoded by the Cα gene have been termed Cα1 (previously named Cα), Cα2 and Cα-s. Originally, Cα2 was isolated from interferon-treated cells as a C-terminally truncated and enzymatically inactive Cα subunit, probably representing a retroposon. Recently a novel Cα2 variant was reported , this Cα splice variant is homologous to the Cα-s, which has been previously been identified and characterized in ovine  and human sperm . Furthermore, Cα-s, now designated Cα2 is encoded with a truncated and nonmyristylated N-terminal end when compared to Cα1[9–11]. The variable parts of Cα1 and Cα2 are located upstream of exon 2 in the murine Cα gene and are encoded by alternative use of different first exons . In bovine, two splice variants of Cβ have been identified, termed bovine Cβ1 and bovine Cβ2. Bovine Cβ1 is ubiquitously expressed whereas Cβ2 is expressed at low levels in most tissues with the highest expression in the spleen, thymus, and kidney . The bovine splice variants contain variable N-terminal ends in which the nonidentical sequences are probably encoded by different forms of exon 1, as is the case with Cβ splice variants identified in the mouse, where three splice variants designated Cβ1, Cβ2 and Cβ3 have been identified . Whereas mouse Cβ1 is ubiquitously expressed, mouse Cβ2 and mouse Cβ3 have so far only been identified in the brain. The mouse and bovine Cβ1 are similar along the entire sequence, demonstrating that they represent orthologous protein sequences. In contrast, mouse and bovine Cβ2 are not similar in the N-terminal region, indicating that their N-termini are encoded by unrelated exons. Thus, mouse and bovine Cβ2 are not orthologous proteins. Previous to this study, only a single splice variant of human Cβ with high homology (more than 98%) to mouse and bovine Cβ1 has been identified, demonstrating this isoform is the human Cβ1 splice variant. In the present study, we demonstrate that the Cβ gene encodes at least six different gene products, designated Cβ1, Cβ2, Cβ3, Cβ4, Cβ4ab and Cβ4abc. As is the case with the murine and bovine splice variants, all the human Cβ splice variants vary in the N-terminal region preceding that encoded by exon 2. All Cβ splice variants identified in mouse and bovine were identified in human (Cβ1, Cβ2, Cβ3 and Cβ4) in addition to two novel Cβ splice variants (Cβ4ab and Cβ4abc), that have previously not been identified in any other species.
- Top of page
- Materials and methods
Here we demonstrate that the human Cβ gene encodes five novel Cβ splice variants, designated Cβ2, Cβ3, Cβ4, Cβ4ab and Cβ4abc, in addition to the previously identified splice variant Cβ1. All the Cβ splice variants contained a unique N-terminus, and showed tissue specific expression. As we found no evidence of an additional exon upstream of exon 1-1 and all the cDNA characterized had unique 5′ ends, it is reasonable to assume that the exons 1-1, 1-2, 1-3 and 1-4 each contain a separate promoter, and that the resulting mRNA products are due to use of alternative promoters. Despite this, we cannot rule out the possibility that two or more of these splice variants share a common promoter used to alternatively splice the different exons. Furthermore, we found two Cβ variants, Cβ4ab and Cβ4abc, which were the results of alternative splicing of either exon a and b, or exon a, b and c, between exon 1-4 and exon 2. The presence of the corresponding mRNA was confirmed by hybridizing a Northern blot with a probe complementary to the sequences found in exons a and b. This probe and the probe specific for Cβ4 bound to an RNA with the same apparent length located in human brain. The location of the exons a, b and c may suggest that they generate splice variants of Cβ in addition to those demonstrated here. Indeed, a short cDNA from human infant brain has been sequenced and demonstrated to contain a combination of exons 1-3, a, b and 2 (GenBank accession no. AA351487; see Fig. 2C). We were unable to produce such a cDNA, which could be due to low-level expression of Cβ3 in adult brain.
In all species examined, the various Cβ splice variants contain the same conserved catalytic domain encoded by the same sets of exons in the Cβ gene, which may be indicative of similar enzymatic features. This is in contrast to the exons upstream of exon 2, which encode N-terminal sequences with low or no homology. In fact, the N-terminus of Cβ1 shows higher homology to the Cα1 N-terminus than to any of the N-termini found in the other Cβ splice variants. The Cβ1 and Cα1 N-terminals are 98% homologous, form an α helix, and contain two sites for post-translational modification, an N-terminal G which is myristylated [16,20] and a conserved autophosphorylation site . It should, however, be noted that both human and mouse Cβ3, possess an N-terminal G. Despite this, the G in mouse Cβ3 is not myristylated in vivo[14,20], because the G is not followed by the required amino acid . Thus, because the mouse and human Cβ3 are 100% identical at the N-terminus, we suggest that human Cβ3 when expressed in vivo is not myristylated.
The N-terminal consensus autophosphorylation motif, KKGS(7–10), is identified in both Cα1 and Cβ1[7,12], but not in any of the other Cβ splice variants. Instead, we identified a potential autophosphorylation site, RKSS(3–6), in Cβ4ab and Cβ4abc that is encoded by exon a. To what extent this site represents a true autophosphorylation site, or if it is phosphorylated by other kinases, remains to be investigated.
The functional role(s) of the N-terminal domains and the post-translational modifications are elusive. However, expression of the N-terminally truncated and nonmyristylated C subunits such as the mouse Cβ2 and Cβ3 revealed that they are enzymatically active in vivo, suggesting that the α helix, the myristyl group and autophosphorylation are not crucial for catalytic activity [16,19,22]. This may also imply that the human Cβ3, Cβ4, Cβ4ab and Cβ4ab are active when expressed in vivo.
The increasing number of reports of C subunits with variable N-terminal ends lacking the ability to be myristylated and autophosphorylated in vivo may suggest differential features associated with the N-terminal domain. The myristyl group fills a hydrophobic pocket in the large lobe and was first suggested to be important for C subunit structure stability [16,23]. However, recent reports have shown that the absence or altered location of myristyl in Cα2 and Cα1, respectively, may influence hydrophobic properties, and thus, may be implicated in subcellular localization of the C subunits [9,11,24,25]. Based on this, we suggest that the N-terminally truncated and presumably nonmyristylated human Cβ3 and Cβ4 splice variants may display altered hydrophobic properties in vivo compared to Cβ1.
The human Cβ2 splice variant was similar to the previously identified bovine Cβ2 splice variant, but showed a somewhat different tissue distribution. Whereas the human Cβ2 subunit was absent from the brain, and expressed at high levels in immune tissues as examined by Northern blotting, the bovine Cβ2 variant was shown using the same method to be expressed in the brain and several other tissues, including immune tissues . The diverging results may be caused by the specificity of the probes used. Furthermore, despite Thullner and colleagues’ identification  of bovine Cβ2 using a Cβ2 polyclonal antisera in rodents, we were unable to identify a similar splice variant in mice and rats as demonstrated both by Northern- and Zoo-blot analysis. Our results indicate that rodents lack the Cβ2 specific sequence in the genome and thus, do not express mRNA and protein for this subunit. The suggestion that rodents may lack the Cβ2 homologue is strengthened as mice ablated for Cβ1 do not express any Cβ homologue in peripheral tissues, including spleen and thymus that is recognized by the mouse Cβ cDNA . Despite this, we can not rule out that a homologue not recognized by the probes used in this paper and in previous studies , but that is recognized by the bovine antiserum may exist. Finally, Cβ2 is expressed with an unusually long N-terminus. Thus, Cβ2 appears to be the most atypical Cβ splice variant and may have specific and unique features that will require further studies for a full characterization.
We suggest that tissue-specific expression of various Cβ splice variants when complexed with R subunits suggests the presence of novel protein kinase A holoenzymes with specific functional features that may be important as mediators of cAMP effects.