ACREDITACIÓN DE CALIDAD DE LAS INSTITUCIONES DE

As described earlier a vast and diverse range of genes are regulated by cAMP and in m ost cases the critical prom oter element has been identified as a CRE containing the core motif CGTCA (see section 1.9). This represents a half site for CREB binding (Lee et al., 1987; Lin and Green, 1988; Yamamoto et al., 1988; Nichols et al., 1992) and in m any cases CREs contain the palindrom e TGACGTCA, that binds a CREB dimer symmetrically.

The diversity of genes regulated th ro u g h CREs is m irrored by the variety of biological functions in which CREB (or a CREB relative) is thought to be involved. These include gluconeogenesis (Boshart et al., 1990), p itu itary proliferation (Struthers et al., 1991), opiate tolerance (G uitart et al., 1992), neuronal signalling (Comb et al., 1987), sperm atogenesis (Foulkes et al., 1992; Foulkes et al., 1993; Delmas et al., 1993), setting of circadian rhythm s (Ginty et al., 1993) and the establishm ent of long term m em ory (Bourtchuladze et al., 1994; Frank and Greenberg, 1994; Yin et al., 1994). Accordingly these diverse genes are not coordinately expressed in response to cAMP rather they are finely reg u lated in a cell and p ro m o ter specific m anner. In d eed elevation of

Chapter One

intracellular cAMP m ay result in either stim ulation or repression of specific genes (Lalli and Sassone-corsi, 1994 and references therein).

Detailed studies of the cAMP signal transduction pathw ay indicates that this integral non-coordinate regulation is generated at all levels from the cell surface dow n to the transcriptional m achinary itself (for review see Borelli et al., 1992). The regulatory role of nuclear factors in specifying responses subsequent to activation of PKA is considered next.

a) Structure of CREB b in d in g sites

In addition to the core m otif CGTCA, flanking bases can significantly influence DNA binding affinity providing a m eans by which CRE activity can be varied. Firstly depending on the nature of the CREB binding site (CBS), CREB DNA b inding activity can be increased by phosphorylation. W hile symmetric sites such as the som atostatin CRE (TCACCTCA) have high affinity for CREB and DN A-binding is not strongly stim ulated by PKA, asym m etric sites such as th at in the tyrosine amino transferase gene (TCACCCAC), have low affinity for unphosphorylated CREB and high affinity for phosphorylated CREB (Nichols et al., 1992). These findings are consistent w ith the observation th at basal transcription (ie, w hen CREB is unphosphorylated) directed by sym m etric sites can be higher than for asymmetric sites b u t that the induced level (ie, w hen CREB is phosphorylated) is similar for both types (M uchardt et al., 1990). Secondly a particular CBS may bind alternative members of the CREB family w ith different stability. D epending upon the particular CBS the binding of ATFl for example is highly unstable compared to CREB (Hurst et al., 1990).

b) C ontributions from additional prom oter elem ents

The relative responsiveness of different CREs to cA M P-induced transcriptional activation has been show n to vary. For exam ple w hilst the som atostatin and gonadotrophin a-su b u n it CREs w ere found to be highly

responsive to cAMP w hen transfected into JEG3 hum an choriocarcinoma cells, the glucagon and parathyroid CREs w ere m uch less active (Deutsch et al., 1988). These four CREs however share the palindrom e TGACGTCA which lead to the su g g estio n th at flanking bases account for th e differences in transcriptional activity. This has subsequently been confirmed w ith tw o distinct m echanism s accounting for the effects. Firstly, flanking nucleotides affect the DNA binding stability of CRE binding proteins (Ryseck and Bravo, 1991) similar to the effect of nucleotides w ithin the CRE (see section 1.14a). Secondly, flanking nucleotides m ay represent binding sites for additional transcription factors that m odulate the function of CREB (for reviews see M asson and Lee, 1993; Schmid et al., 1993). For example tyrosine am inotransferase is expressed exclusively in liver parenchym al cells and can be induced by cAMP. Induction by cAMP is m ediated by an enhancer region (at -3.6 kb) consisting of two essential elements. One is a CRE that binds CREB, the other interacts w ith the liver enriched transcription factor HNF4. Both elements are essential for cell specific and cAMP dependent expression (Schmid et al., 1993 and references therein). M eanwhile the ability of CREB to activate the rat glucagon gene in response to cAMP is decreased by additional proteins w hich bind adjacent to the CRE (Miller et al., 1993). W hilst in m ost cases the precise m echanism s by w hich these additional transcription factors m odulate CREB transcriptional activity are not clear, direct protein-protein interactions m ay be involved. Such interactions m ight influence DNA binding stability or else m o d u late the interaction of CREB transactivation dom ains (see section 1.11) w ith protein kinases or phosphatases or w ith com ponents of the basal transcriptional ap p aratu s. A lternatively binding of additional transcription factors to the prom oter m ay alter DNA conformation in a m anner that affects the function of the CREB binding site. G row ing evidence suggests that transcription factors can im part significant DNA conformational changes upon binding (Gustafson et al., 1989; Schrech et al., 1990; Kerppola and Curran, 1991).

Chapter One

c) C ontributions from other ATF fam ily m em bers

As m entioned earlier CBSs m ay be targeted by different member(s) of the CREB/ATF family w ith different regulatory a n d /o r functional properties. Such considerations are m ade all the m ore im portant given that w ithin the family alternative splicing and heterodim erisation create the potential for even greater diversity (see section 1.9).

Multiple isoforms of CREB were in fact amongst the first to be identified (see D iagram 6). A ctivator forms term ed CREB a and CREB A (the latter differing from CREB a by the exclusion of the 14 amino acid a -p e p tid e (see section 1.11)), appear to be ubiquitously expressed in somatic cells w ith CREB A 3-fold higher than CREB a (Berkowitz and Gilman 1990; R uppert et al., 1992). O ther CREB isoforms (CREB ay, CREB y, CREB D, CREBW and CREB aW ) are highly expressed only in adult testis however their function is not obvious since they lack the DN A b in d in g do m ain an d nu clear tran slo catio n signal (Yamamoto et al., 1990; W aeber et al., 1991; R uppert et al., 1992; W aeber and Habener, 1992).

Of particular relevence are the factors ATFl and CREM (and various isoforms -see later) which comprise the im m ediate CREB family (see section 1.9). These share extensive hom ology w ith CREB, especially in the bZIP and KID regions (see Diagrams 3 and 7), form selective heterodim ers w ith CREB (see section 1.9) and can be regulated directly by PKA (see sections 1.15 and 1.16). Transgenic experim ents using dom inant negative CREB m utants and hom ozygous deletions of the CREB gene indicate that som e CREB functions (Struthers et al., 1991; H um m ler et al., 1994), although not all (Barton et al., 1996), can be effectively perform ed by the various isoforms of CREM a n d /o r ATFl (see sections 1.15 and 1.16) suggesting th at a degree of functional redundancy exists.

Structural similarities in the organisation of the CREB (R uppert et al., 1992) and CREM genes (Laoide et al., 1993) suggest a common ancestral origin.

Significantly how ever each has acquired unique features and results in the generation of proteins w ith some distinct properties w hich are believed to reflect specialised roles. In line w ith this m any also exhibit complex patterns of gene expression. Both CREM and ATFl are discussed in tu rn below.