Phenobarbital (PB), the more classic tumor promoter of rodent hepatocarcinogenesis, appears to have a somewhat more complex mechanism of signal transduction than either TPA or TCDD. TPA inhibited the PB induction of phase I genes (Steele and Virgo, 1988), while PB inhibited the translocation of protein kinase C from the soluble to the particulate or membrane fraction of the cell (Brockenbrough et al., 1991), which is the normal pathway by which TPA activates pro- tein kinase C to initiate signal transduction pathways. Like that of TPA, chronic administration of PB caused a decrease in epidermal growth factor receptor protein levels (Orton et al., 1996). PB also increased expression of the transcription factor c-fos, which is important in the enhance- ment of gene expression by TPA through the AP-1 regulatory site on DNA (see above) as well as in enhancing the expression of glutathione S-transferase in preneoplastic foci in rodent liver (Bitsch et al., 1999). However, the exact signal transduction pathway of phenobarbital has not been as readily elucidated as those of TPA and TCDD. In 1991, He and Fulco reported on a sequence in DNA which had some of the characteristics of a PB-responsive element. Later stud- ies from this laboratory (Liang et al., 1995) defined the wild-type sequence, termed a Barbie box, as AT(A/C)AAAAGCTGGTG. Such sequences were found in the regulatory regions of several PB-responsive genes (Fournier et al., 1994). More recently, Negishi and associates de- fined another regulatory DNA element involved in the phenobarbital induction of cytochrome P450 CYP2B genes (cf. Honkakoski and Negishi, 1998). This sequence, termed the PB-respon- sive enhancer module (PBREM), did not contain any sequences of the Barbie box but usually occurred distal in the 5′ region of genes containing Barbie boxes. The sequence of the PBREM 51 base-pair enhancer element is seen in Figure 7.12.
These same workers have now demonstrated that several regulatory factors including a constitutively active receptor (CAR) are capable of transactivating the PBREM (Kawamoto et
Figure 7.11 Mechanisms involved in the induction of cytochrome P450 1A1 gene transcription by TCDD. The figure demonstrates the various steps occurring from interaction of TCDD with the Ah receptor (Ah-R), the complex being an association with heat shock proteins and chaperones, Hsp90 and p50. This complex dissociates with an interaction of the ligand, TCDD, with the Ah-R, the complex then interacting with the Ah receptor nuclear translocator, Arnt, through helix- loop-helix (HLH) domains in the two proteins. This complex is translocated into the nucleus, where it may be phosphory- lated by kinases such as protein kinase C, then interacting with specific regions of DNA and other proteins of the RNA polymerase complex. This complex is shown associating with adapter proteins and the TATA-box binding protein (TBP), ultimately resulting in transcription of the specific gene. (Modified from Whitlock, 1993, and DeVito and Birnbaum, 1994.)
Figure 7.12 The 51-bp enhancer element (PBREM) as delineated by transfection in mammalian cells. The two putative nuclear receptor binding motifs are within the boxes as well as an NFI binding sequence. Variations in other cytochrome P450IIB genes are noted below the primary sequence with the differences indicated. (Adapted from Honkakoski et al., 1998, with permission of the authors and publisher.)
al., 1999). Phenobarbital appears to have an active function in this process, since it is capable of altering protein binding to this unit in native chromatin (Kim and Kemper, 1997). Thus, the phe- nobarbital signal transduction pathway appears to be quite analogous to that seen with TCDD in relation to the CAR receptor, but this does not explain the function of the Barbie box. Other studies (Stolz et al., 1998) indicate that phenobarbital induction of the PB-responsive cyto- chrome P450s requires interactions among multiple regulatory proteins and multiple cis-acting elements in the regulatory regions of these genes.
The mechanisms described for the action of these three promoting agents—TPA, TCDD, and PB—are responsible for the phenotypes noted in preneoplastic lesions, which differ signifi- cantly from those of their normal cellular counterparts. However, based on studies up to the present time, these mechanisms involved in both preneoplastic and normal cells appear to be the same. How these mechanisms result in selective cell replication and gene expression within pre- neoplastic tissues, giving them a growth advantage over their normal counterparts but being to- tally dependent on a continued presence of the promoting agent, is not clear at the present time.