My laboratory studies the structure-function relationships of the mammalian B-ZIP class of sequence-specific DNA-binding dimeric proteins. More than 50 B-ZIP genes have been identified in the mammalian genome. In the most general terms, B-ZIP proteins both activate and repress gene expression in response to physiological changes, be it growth factors (FOS), stress (ATF2), neuronal signaling (CREB), or metabolic changes (CEBP). We are studying B-ZIP transcriptional function using dominant-negatives (DNs) we have designed that inhibit B-ZIP DNA binding. A problem with the design of such reagents is that B-ZIP proteins become stabilized by binding DNA. We have overcome this problem by extending the dimerization domain into the basic region to produce A-ZIPs.
Click to enlarge image.
The A represents an N-terminal acidic amphipathic extension of the leucine zipper that replaces the basic region critical for sequence-specific DNA binding of the B-ZIP dimer. The acidic extension forms an a-helical coiled coil structure with the basic region, extending the leucine zipper. The stabilization that occurs through the interaction of the acidic extension with the basic region of the B-ZIP domain prevents DNA binding of B-ZIP proteins. The pathology of activated stress pathways caused by B-ZIP proteins can be examined using these A-ZIPs. Ultimately, we hope to use these gene-based A-ZIPs as adjuvants with other medical approaches to cure human disease, particularly chemotherapy-resistant cancers. The hypothesis driving this work is that direct transcriptional targets of a B-ZIP protein can be identified by expression of the corresponding A-ZIP protein.
Our laboratory has systematically analyzed the contribution of individual amino acids to coiled coil stability and dimerization specificity. These studies have helped us generate designs for dominant negatives to B-ZIP proteins by replacing the B-ZIP basic region with an acidic amphipathic protein sequence N-terminus of the leucine zipper. The acidic extension of the leucine zipper heterodimerizes with the B-ZIP basic region to produce a coiled coil extension of the leucine zipper. This zippering up of the leucine zipper into the basic region by the acidic extension stabilizes the heterodimer 2.5 to 5 kcal mol-1. In a competition assay containing an equal concentration of the B-ZIP and A-ZIP protein, DNA binding of the B-ZIP protein is abolished. The acidic extension interacts with all the B-ZIP basic regions we have examined.
Thus, we have developed a general strategy for generating A-ZIPs that inhibit B-ZIP DNA binding in a leucine zipper-dependent manner. We have demonstrated this for the mammalian B-ZIP proteins CEBP, JUN, TFE, CREB, and ATF2. A similar strategy of replacing the basic region with an acidic protein sequence to produce A-HLH-ZIP dominant negatives works for the B-HLH-ZIP proteins examined, MYC/MAX, USF, and MITF.
Click to enlarge image.
The utility of these reagents to alter living systems was examined by producing transgenic mice expressing an A-ZIP dominant negative in only fat cells. The transgenic mice have 20 copies of a construct consisting of 7.6 kb of the fat-specific aP2 promoter driving expression of an A-ZIP dominant negative that inhibits the DNA binding of both CEBP and FOS/JUN B-ZIP domains. The resultant mouse has essentially no fat. The metabolic consequences are profound, mimicking the human disease lipodystrophy. The mouse eats 1.7-fold more food and the liver is 2.3-fold larger than normal. The diabetic nature of the metabolic confusion is seen in insulin levels that are 100-fold higher than normal and glucose levels that are 3-fold elevated. The mice die prematurely, starting at 5 months. Troglitazone, a thiazolidinedione agonist of the transcription factor PPRg, reverses the diabetes.