A novel E1A domain mediates skeletal-muscle-specific enhancer repression independently of pRB and p300 binding.

The adenovirus E1A oncoprotein completely blocks muscle differentiation and specifically inhibits the transactivating function of myogenic basic helix-loop-helix (bHLH) transcription factors. This inhibition is dependent on the conserved region CR1 of E1A, which also constitutes part of the binding sites for the pocket proteins pRB, p107, and p130 and the transcriptional coactivators p300 and CBP. Here we report a detailed mutational analysis of E1A and the identification of a muscle inhibition motif within CR1. This motif encompasses amino acids 38 to 62 and inhibits Myf-5- or MyoD-mediated activation of myogenin and the muscle creatine kinase gene. Overexpression of this E1A region also inhibits the conversion of 10T1/2 fibroblasts to the myogenic lineage. The sequence motif EPDNEE (amino acids 55 to 60) within CR1 appears to be particularly important, because point mutations of this sequence diminish the E1A inhibitory activity. Interactions of E1A with pRB and with p300 do not seem to be necessary for the muscle-specific enhancer repression, because E1A mutants which lack these interactions still inhibit Myf-5- and MyoD-mediated transactivation. Moreover, overexpression of p300 fails to overcome muscle-specific inhibition by wild-type E1A and mutant E1A protein which lacks pRB binding. Since we have no evidence for direct E1A interaction with bHLH proteins, we propose that E1A may target a necessary cofactor of the muscle-specific bHLH transcription complex.

The uteroglobin promoter targets expression of the SV40 T antigen to a variety of secretory epithelial cells in transgenic mice.

Adenocarcinomas derived from the lining epithelia of various organs are the most common malignant tumors in human pathology and about 50% are hormone dependent. The tissue-specific and hormally regulated expression of the rabbit uteroglobin gene is secretory epithelial cells could provide a means of establishing in vivo models for a variety of human tumors originating from such tissues. We have generated trangenic mice inheriting a hybrid gene containing 4.7 kb of the rabbit uteroglobin 5'-flanking sequences fused to the SV40 T antigen encoding region. All transgenic founders examined exhibited bronchio-alveolar adenocarcinomas, probably due to expression of the transgene in Clara cells. Most founders also developed tumors of the submandibular salivary gland, and adenocarcinomas of the stomach. Adenocarcinomas and dysplasias in epithelial cells of the male and female genital tract were found in single founders. Immunohistochemical analysis showed that T antigen expression interfered with stable maintenance of the differentiated phenotype as documented by expression of the endogenous uteroglobin gene. One founder gave rise to a mouse line, UT7.1. Transgenic descendants of UT7.1 developed lung adenocarcinomas and, depending on the genetic background, exhibited tumors of the stomach, the salivary gland and the pancreas. Sporadically male descendants developed prostatic adenocarcinoma whereas females developed dysplasias and adenocarcinomas of the uterus and the oviduct. Thus, the UT7.1 mouse line could be a useful model for several epithelial neoplasias.

Cell-specific, developmentally and hormonally regulated expression of the rabbit uteroglobin transgene and the endogenous mouse uteroglobin gene in transgenic mice.

We have generated a transgenic mouse line by introducing the rabbit uteroglobin gene with 4 kb of 5'-flanking DNA and 1 kb of 3'-flanking DNA into the mouse germ line via microinjection into fertilized oocytes. Expression of the rabbit uteroglobin transgene was examined and compared with the endogenous mouse uteroglobin gene. Both genes are expressed in the lung, male genital tract and uterus. In the lung, mRNA expression is enhanced by glucocorticoids and restricted to the Clara cells that line terminal and respiratory bronchioli. During embryonic lung development, transcripts are first detected at day 17. Expression in the uterus is restricted to the glandular epithelium and can be induced by sequential treatment with estrogens and progesterone. In the uterus of these pseudopregnant mice the level of rabbit uteroglobin transcripts is higher than that of the mouse endogenous uteroglobin transcripts. In the male genital tract, expression of both genes is restricted to the epithelial layers of the vesicular gland, vas deferens and epididymis. Our results indicate that the rabbit uteroglobin gene together with 4 kb of 5'-flanking DNA and 1 kb of 3'-flanking DNA contains the information required for cell type-specific, developmentally, and hormonally regulated expression.

A comparison of mouse and rabbit embryos for the production of transgenic animals by pronuclear microinjection.

Procedures for the production of transgenic animals have low overall efficiency. To evaluate factors responsible for low efficiency, zygotes of two species, varying intensities of microscope light, different qualities of injection pipettes, and six different genes were tested for their influence on the efficiency of pronuclear gene injection for the production of transgenic rabbits and mice. Rabbit zygotes were less sensitive to mechanical manipulation during injection than mouse zygotes. Exposing zygotes to a microscope light intensity of 5550 lx significantly reduced their cleavage rate, while a lower intensity (2280 lx) did not. Using pipettes with a filament for pronuclear gene injection of mouse zygotes resulted in a higher cleavage rate of zygotes after injection than when pipettes were used without filament (30.3 vs 20.6%). Implantation rates varied between 2.9% (HB72CAT) and 23.1% (ts 58-2) depending on the gene used. No transgenic animals were obtained after injection of uteroglobin-CAT-hybrid genes (B2B3UGCAT, HB72CAT), while all other genes used (UG 11.8, UGTAg, RSV lacZ, ts 58-2) resulted in transgenic embryos, fetuses, and newborn animals.

A transgenic mouse model for lung adenocarcinoma.

Lung cancer is a leading cause of tumor-related deaths in humans but its origin and development are poorly understood. To study the biology of these tumors, appropriate animal and cell culture models will be of eminent importance. Uteroglobin is a marker protein for the nonciliated epithelial Clara cells lining the respiratory and terminal bronchioli of the lung. We have used the promoter and 5'-flanking sequences of the rabbit uteroglobin gene to target expression of the SV40 T antigen to the lung of transgenic mice. All transgenic founders as well as the descendants from an established line, UT7.1, developed multifocal bronchioloalveolar adenocarcinomas originating from Clara cells. At least three different stages in tumor development with progressive loss of the differentiated phenotype can be distinguished by immunohistochemical data and in situ hybridization. Only in the initial stage did bronchiolar cells express both uteroglobin and SV40 T antigen, whereas at later stages, only SV40 T antigen was detected, and the most advanced tumors were negative for both proteins. Starting from the lungs of UT7.1 mice, a bronchiolar cell line was established that maintains the features of differentiated Clara cells. This system provides a useful model for further studying the development and progression of lung adenocarcinomas in vivo and in vitro.

Transcriptional activation of the myogenin gene by MEF2-mediated recruitment of myf5 is inhibited by adenovirus E1A protein.

The basic helix-loop-helix (bHLH) transcription factor myogenin plays a crucial role in terminal differentiation of committed myoblasts into mature myocytes. Transcriptional activation of the myogenin gene requires coordinate action of myocyte enhancer factor 2 (MEF2) proteins and the myogenic bHLH regulators, MyoD or Myf5. Here we show that transcription of the myogenin gene in differentiated cells correlates with MEF2 and NF1 binding to their cognate sites in the proximal myogenin promoter but not with binding of Myf5 or MyoD to the E-box. The importance of MEF2 activity was further demonstrated by expression of antisense MEF2 RNA which repressed MEF2 and Myf5-mediated MEF2 site-dependent reporter gene activation and the synergistic transactivation of a myogenin CAT reporter by Myf5 and MEF2. Adenovirus E1A which has previously been shown to specifically interfere with myogenin gene transcription also inhibited the cooperative transactivation by Myf5/MEF2 and MEF2. Consistently, coimmunoprecipitation studies revealed impaired MEF2/Myf5 protein-protein interactions. These results support a model of transcriptional activation and stabilization of myogenin expression in which DNA-bound MEF2 recruits myogenic bHLH factors into an active but E1A-sensitive transcription factor complex.