appear to possess different DNA-binding properties, transcriptional activities, 
stability, etc. (Morgan and Curran 1991; Ryseck and Bravo 1991). Such 
formation of heterodimers has significant consequences as it dramatically 
increases the number of functional transcriptional complexes that can be 
formed from a finite number of factors. In addition, transcription factors 
contain an activation domain, which is thought to be the region of the proteins 
responsible for regulation of transcription (Ptashne and Gann 1990). Each 
class of transcription factor has been shown to bind to a particular DNA 
response element. However, these DNA-binding sites, such as those shown 
in table 1 , represent only “consensus sequences” in that the actual sites to 
which the factors bind in different genes can vary. For example, functional 
cyclic AMP response elements (CREs) in various neural genes are not all 
“TGACGTCA”; in fact, only the last five nucleotides appear to be constant 
among all known CREs (Goodman 1990; Montminy et al. 1990). In addition, 
it can be seen from table 1 that certain response elements share considerable 
homology; for example, the CRE and activator protein-1 (AP-1) sites. These 
observations raise the possibility that the specificity of the cyclic AMP (cAMP) 
response element binding protein (CREB) and Fos/Jun families for CREs and 
AP-1 sites, respectively, is not absolute, with Fos/Jun potentially able to bind 
CRE sites and vice versa. It has even been reported that members of the 
Fos/Jun and CREB families, both of which contain the LZ motif, can cross- 
dimerize and thereby form heterodimers with unique DNA-binding and 
regulatory properties (Hai and Curran 1991). These examples of cross- 
reactivity between different classes of transcription factors illustrate the 
enormous complexity of possible mechanisms underlying the regulation of 
neural gene expression. 
Most forms of transcription factors are highly regulated, and it is through such 
regulatory processes that important aspects of neural plasticity (including drug 
addiction) would appear to be achieved. Four mechanisms appear to be most? 
prominent. First, the total amounts of certain transcription factors can be 
dramatically altered via changes in their own transcription. This is the case b 
for the Fos/Jun family; the expression of these proteins is regulated in the 
adult CNS in response to diverse types of stimuli (Morgan and Curran 1991). 
Second, the ability of transcription factors to regulate gene expression can 
be altered via their phosphorylation by diverse types of protein kinases. The 
prototypical example for this is CREB, whose ability to regulate transcription 
is controlled primarily through its phosphorylation by cAMP-dependent or 
calcium-dependent protein kinases (Goodman 1990; Montminy et al. 1990; 
Sheng et al. 1990, 1991; Van Nguyen et al. 1990). It appears that CREB 
phosphorylation does not dramatically alter its DNA-binding activity but instead 
increases the ability of CREB, bound to DNA, to activate the RNA polymerase II 
transcriptional complex. Third, the ability of transcription factors to bind DNA 
94 
