Tissue transglutaminase selectively modifies proteins associated with truncated mutant huntingtin in intact cells.

The cause of Huntington's disease (HD) is a pathological expansion of the polyglutamine domain within the N-terminal region of huntingtin. Neuronal intranuclear inclusions and cytoplasmic aggregates composed of the mutant huntingtin within certain neuronal populations are a characteristic hallmark of HD. However, how the expanded polyglutamine repeats of mutant huntingtin cause HD is not known. Because in vitro expanded polyglutamine repeats are excellent glutaminyl-donor substrates of tissue transglutaminase (tTG), it has been hypothesized that tTG may contribute to the formation of these aggregates in HD. However, an association between huntingtin and tTG or modification of huntingtin by tTG has not been demonstrated in cells. To examine the interactions between tTG and huntingtin human neuroblastoma SH-SY5Y cells were stably transfected with full-length huntingtin containing 23 (FL-Q23) (wild type) or 82 (FL-Q82) (mutant) glutamine repeats or a truncated N-terminal huntingtin construct containing 23 (Q23) (wild type) or 62 (Q62) (mutant) glutamine repeats. Aggregates were rarely observed in the cells expressing full-length mutant huntingtin, and no specific colocalization of full-length huntingtin and tTG was observed. In contrast, in cells expressing truncated mutant huntingtin (Q62) there were numerous complexes of truncated mutant huntingtin and many of these complexes co-localized with tTG. However, the complexes were not insoluble structures. Further, truncated huntingtin coimmunoprecipitated with tTG, and this association increased when tTG was activated. Activation of tTG did not result in the modification of either truncated or full-length huntingtin, however proteins that were associated with truncated mutant huntingtin were selectively modified by tTG. This study is the first to demonstrate that tTG specifically interacts with a truncated form of huntingtin, and that activated tTG selectively modifies mutant huntingtin-associated proteins. These data suggest that proteolysis of full-length mutant huntingtin likely precedes its interaction with tTG and this process may facilitate the modification of huntingtin-associated proteins and thus contribute to the etiology of HD.

Huntingtin's WW domain partners in Huntington's disease post-mortem brain fulfill genetic criteria for direct involvement in Huntington's disease pathogenesis.

An elongated glutamine tract in mutant huntingtin initiates Huntington's disease (HD) pathogenesis via a novel structural property that displays neuronal selectivity, glutamine progressivity and dominance over the normal protein based on genetic criteria. As this mechanism is likely to involve a deleterious protein interaction, we have assessed the major class of huntingtin interactors comprising three WW domain proteins. These are revealed to be related spliceosome proteins (HYPA/FBP-11 and HYPC) and a transcription factor (HYPB) that implicate huntingtin in mRNA biogenesis. In HD post-mortem brain, specific antibody reagents detect each partner in HD target neurons, in association with disease-related N-terminal morphologic deposits but not with filter trapped insoluble-aggregate. Glutathione S:-transferase partner 'pull-down' assays reveal soluble, aberrantly migrating, forms of full-length mutant huntingtin specific to HD target tissue. Importantly, these novel mutant species exhibit exaggerated WW domain binding that abrogates partner association with other huntingtin isoforms. Thus, each WW domain partner's association with huntingtin fulfills HD genetic criteria, supporting a direct role in pathogenesis. Our findings indicate that modification of mutant huntingtin in target neurons may promote an abnormal interaction with one, or all, of huntingtin's WW domain partners, perhaps altering ribonucleoprotein function with toxic consequences.

Polyglutamine-mediated dysfunction and apoptotic death of a Caenorhabditis elegans sensory neuron.

The effect of expressing human huntingtin fragments containing polyglutamine (polyQ) tracts of varying lengths was assessed in Caenorhabditis elegans ASH sensory neurons in young and old animals. Expression of a huntingtin fragment containing a polyQ tract of 150 residues (Htn-Q150) led to progressive ASH neurodegeneration but did not cause cell death. Progressive cell death and enhanced neurodegeneration were observed in ASH neurons that coexpressed Htn-Q150 and a subthreshold dose of a toxic OSM-10::green fluorescent protein (OSM-10::GFP) fusion protein. Htn-Q150 huntingtin protein fragments formed protein aggregates in ASH neurons, and the number of ASH neurons containing aggregates increased as animals aged. ASH neuronal cell death required ced-3 caspase function, indicating that the observed cell death is apoptotic. Of interest, ced-3 played a critical role in Htn-Q150-mediated neurodegeneration but not in OSM10::GFP-mediated ASH neurodegeneration. ced-3 function was important but not essential for the formation of protein aggregates. Finally, behavioral assays indicated that ASH neurons, coexpressing Htn-Q150 and OSM10::GFP, were functionally impaired at 3 days before the detection of neurodegeneration, cell death, and protein aggregates.

Polyglutamine-expanded human huntingtin transgenes induce degeneration of Drosophila photoreceptor neurons.

Huntington's disease (HD) is an autosomal dominant neurodegenerative disorder. Disease alleles contain a trinucleotide repeat expansion of variable length, which encodes polyglutamine tracts near the amino terminus of the HD protein, huntingtin. Polyglutamine-expanded huntingtin, but not normal huntingtin, forms nuclear inclusions. We describe a Drosophila model for HD. Amino-terminal fragments of human huntingtin containing tracts of 2, 75, and 120 glutamine residues were expressed in photoreceptor neurons in the compound eye. As in human neurons, polyglutamine-expanded huntingtin induced neuronal degeneration. The age of onset and severity of neuronal degeneration correlated with repeat length, and nuclear localization of huntingtin presaged neuronal degeneration. In contrast to other cell death paradigms in Drosophila, coexpression of the viral antiapoptotic protein, P35, did not rescue the cell death phenotype induced by polyglutamine-expanded huntingtin.

Huntingtin interacts with a family of WW domain proteins.

The hallmark neuropathology of Huntington's disease (HD) is due to elongation of a polyglutamine segment in huntingtin, a novel approximately 350 kDa protein of unknown function. We used a yeast two-hybrid interactor screen to identify proteins whose association with huntingtin might be altered in the pathogenic process. Surprisingly, no interactors were found with internal and C-terminal segments of huntingtin. In contrast, huntingtin's N-terminus detected 13 distinct proteins, seven novel and six reported previously. Among these, we identified a major interactor class, comprising three distinct WW domain proteins, HYPA, HYPB and HYPC, that bind normal and mutant huntingtin in extracts of HD lymphoblastoid cells. This interaction is mediated by huntingtin's proline-rich region and is enhanced by lengthening the adjacent glutamine tract. Although HYPB and HYPC are novel, HYPA is human FBP-11, a protein implicated in spliceosome function. The emergence of this class of proteins as huntingtin partners argues that a WW domain-mediated process, such as non-receptor signaling, protein degradation or pre-mRNA splicing, may participate in HD pathogenesis.

Amyloid formation by mutant huntingtin: threshold, progressivity and recruitment of normal polyglutamine proteins.

Huntington's disease (HD) is caused by an expanded CAG trinucleotide repeat encoding a tract of consecutive glutamines near the amino terminus of huntingtin, a large protein of unknown function. It has been proposed that the expanded polyglutamine stretch confers a new property on huntingtin and thereby causes cell and region-specific neurodegeneration. Genotype-phenotype correlations predict that this novel property appears above a threshold length (approximately 38 glutamines), becomes progressively more evident with increasing polyglutamine length, is completely dominant over normal huntingtin and is not appreciably worsened by a double genetic dose in HD homozygotes. Recently, an amino terminal fragment of mutant huntingtin has been found to form self-initiated fibrillar aggregates in vitro. We have tested the capacity for aggregation to assess whether this property matches the criteria expected for a fundamental role in HD pathogenesis. We find that that in vitro aggregation displays a threshold and progressivity for polyglutamine length remarkably similar to the HD disease process. Moreover, the mutant huntingtin amino terminus is capable of recruiting into aggregates normal glutamine tract proteins, such as the amino terminal segments of both normal huntingtin and of TATA-binding protein (TBP). Our examination of in vivo aggregates from HD post-mortem brains indicates that they contain an amino terminal segment of huntingtin of between 179 and 595 residues. They also contain non-huntingtin protein, as evidenced by immunostaining for TBP. Interestingly, like the in vitro aggregates, aggregates from HD brain display Congo red staining with green birefringence characteristic of amyloid. Our data support the view that the expanded polyglutamine segment confers on huntingtin a new property that plays a determining role in HD pathogenesis and could be a target for treatment. Moreover, the new property might have its toxic consequences by interaction with one or more normal polyglutamine-containing proteins essential for the survival of target neurons.

Frataxin gene of Friedreich's ataxia is targeted to mitochondria.

Friedreich's ataxia is caused by a triplet repeat expansion in intron 1, a noncoding region of the frataxin gene (X25). We have generated a chimeric gene composed of the frataxin gene fused with the green fluorescent protein (GFP) gene as a reporter. Transfection of the fusion construct into living COS cells revealed that the frataxin-GFP construct localizes to organelles that double-label with 8-(4'-chloromethyl) phenyl-2,3,5,6,11,12,14,15-octahydro-1H,4H,10H-13H-diquinolizin o-8H-xanthylium chloride (CMXRos), a novel mitochondrial dye. Thus, frataxin appears to be a nuclear-encoded mitochondrial protein.

Functional in vivo interaction between the amino-terminal, transactivation domain and the ligand binding domain of the androgen receptor.

The ligand binding domain (LBD) and the amino-terminal, transactivation domain (TAD) of the androgen receptor (AR) were separately linked to the GAL4 DNA binding domain (DBD) and to the GAL4(TAD). Resulting constructs were tested in the yeast two-hybrid system for protein-protein interactions. In the presence of androgen [methyltrienolone (R1881) or dihydrotestosterone (DHT)] a transcriptionally active complex was formed, reflecting an association between the AR(LBD) and the AR(TAD). No interactions were found in the presence of low-affinity ligands like estradiol (E2), promegestone (R5020), or progesterone (Pg). Use of the Thr-868-Ala mutated AR(LBD) in the assay resulted not only in a clear AR TAD-LBD interaction in the presence of R1881 and DHT but also in the presence of E2, Pg, and R5020, corresponding to the alteration in ligand specificity induced by the mutation. Coexpression of the fusion protein Gal4(DBD)AR(LBD) and the separate AR(TAD) also gave rise to the formation of a transcriptionally active complex. No interactions were found between two AR LBDs at the low-expression level of the two components. However, LBD-LBD interaction was detectable by application of a high-expression vector for GAL4(TAD)AR(LBD), albeit at high ligand concentrations. To substantiate the observation of the AR LBD-TAD interaction, CHO cells were cotransfected with expression plasmids for a truncated AR, which lacks the TAD [AR(DBD)(LBD)], and for the separate AR(TAD). This resulted in stimulation of a MMTV-LUC reporter gene in the presence of R1881 but not in the absence of hormone. This finding indicates that, like in the yeast system, in mammalian cells, TAD-LBD interactions are of importance for AR activation. In the mammalian system, a maximal AR TAD-LBD interaction was obtained at approximately 10-fold higher ligand concentrations than required for full-length AR activation. In the presence of low-affinity ligands, the AR TAD-LBD interaction as measured by transcriptional activation was considerably weaker than the activity of the full-length AR. From the present results a concept of hormone-dependent AR activation is proposed, which requires a functional, direct or indirect intramolecular interaction between the TAD and the LBD.

An androgen response element in a far upstream enhancer region is essential for high, androgen-regulated activity of the prostate-specific antigen promoter.

Prostate-specific antigen (PSA) is expressed at a high level in the luminal epithelial cells of the prostate and is absent or expressed at very low levels in other tissues. PSA expression can be regulated by androgens. Previously, two functional androgen-response elements were identified in the proximal promoter of the PSA gene. To detect additional, more distal control elements, DNasel-hypersensitive sites (DHSs) upstream of the PSA gene were mapped in chromatin from the prostate-derived cell line LNCaP grown in the presence and absence of the synthetic androgen R1881. In a region 4.8 to 3.8 kb upstream of the transcription start site of the PSA gene, a cluster of three DHSs was detected. The middle DNAseI-hypersensitive site (DHSII, at approximately -4.2 kb) showed strong androgen responsiveness in LNCaP cells and was absent in chromatin from HeLa cells. Further analysis of the region encompassing DHSII provided evidence for the presence of a complex, androgen-responsive and cell-specific enhancer. In transient transfected LNCaP cells, PSA promoter constructs containing this upstream enhancer region showed approximately 3000-fold higher activity in the presence than in the absence of R1881. The core region of the enhancer could be mapped within a 440-bp fragment. The enhancer showed synergistic cooperation with the proximal PSA promoter and was found to be composed of at least three separate regulatory regions. In the center, a functionally active, high-affinity androgen receptor binding site (GGAACATATTGTATC) could be identified. Mutation of this element almost completely abolished PSA promoter activity. Transfection experiments in prostate and nonprostate cell lines showed largely LNCaP cell specificity of the upstream enhancer region, although some activity was found in the T47D mammary tumor cell line.

Differential expression of normal and mutant Huntington's disease gene alleles.

Huntingtin expression was examined by Western blot and immunoprecipitation studies of lymphoblastoid cell lines from Huntington's disease (HD) homozygotes, heterozygotes, and a phenotypically normal individual with a t(4p16.3;12p13.3) breakpoint in the HD gene. The latter produced a reduced level of normal huntingtin without evidence of an altered protein, indicating that simple loss of huntingtin activity does not cause HD. In juvenile onset HD heterozygotes, NH2- and COOH-terminal antisera revealed reduced relative expression from the mutant allele. Pulse-chase studies indicated that huntingtin is a stable protein whose differential allelic expression is not due to destabilization of the mutant isoform. No stable breakdown products specific to mutant huntingtin were detected in either HD homozygotes or heterozygotes. These data are consistent with HD involving either a gain of function or a dominant negative loss of function that operates within severe constraints and suggest that in either case the pathogenic process is usually saturated by the amount of abnormal huntingtin produced from a single mutant allele.

Tissue specific and androgen-regulated expression of human prostate-specific transglutaminase.

Transglutaminases (TGases) are calcium-dependent enzymes catalysing the post-translational cross-linking of proteins. In the prostate at least two TGases are present, the ubiquitously expressed tissue-type TGase (TGC), and a prostate-restricted TGase (TGP). This paper deals with the molecular cloning and characterization of the cDNA encoding the human prostate TGase (hTGP). For this purpose we have screened a human prostate cDNA library with a probe from the active-site region of TGC. The largest isolated cDNA contained an open reading frame encoding a protein of 684 amino acids with a predicted molecular mass of 77 kDa as confirmed by in vitro transcription-translation and subsequent SDS/PAGE. The hTGP gene was tissue-specifically expressed in the prostate, yielding an mRNA of approx. 3.5 kb. Furthermore, a 3-fold androgen-induced upregulation of hTGP mRNA expression has been demonstrated in the recently developed human prostate cancer cell line, PC346C. Other well established human prostate cancer cell lines, LNCaP and PC-3, showed no detectable hTGP mRNA expression on a Northern bolt. The gene coding for prostate TGase was assigned to chromosome 3.

Androgen regulation of the rat keratinocyte growth factor (KGF/FGF7) promoter.

Keratinocyte Growth Factor (KGF/FGF7) is a candidate andromedin in normal embryonic development of male accessory sex glands, such as the prostate and seminal vesicles. The expression of KGF mRNA and protein is androgen-responsive. To elucidate the regulation of expression of the KGF gene, we isolated the first two exons of the KGF gene and approximately 15 kb upstream sequences. The major transcription start site was mapped. It is preceded by a CAAT-box and a TATA-box. Transient transfections in LNCaP cells revealed that, upon treatment with the synthetic androgen R1881, KGF promoter activity is upregulated 6 to 11 fold, indicating androgen regulation of the KGF promoter in the region from position - 900 to -1200. The longest construct (BH-pLuc: -4700 to +901) has a much higher basal activity than the shorter constructs, indicating that in the region -4700 to -2700 additional activating sequences are present.

Loss of heterozygosity of chromosome 8 microsatellite loci implicates a candidate tumor suppressor gene between the loci D8S87 and D8S133 in human prostate cancer.

To search for specific chromosome 8 aberrations in human prostate cancer, DNA was isolated from 44 human prostate tumor samples. Twenty six tumor samples were obtained from locally progressive tumors by transurethral resection, 12 were from radical prostatectomy specimens, and 6 were from lymph node metastases. Tumor DNAs were screened for allelic losses using 16 highly polymorphic microsatellite loci (14 covering the p arm, 2 on the q arm). In general, the detected deletions were large. In 59% of the tumor DNAs, allelic loss of 3 or more 8p loci was observed. Loss of 8p loci occurred in between 36 and 69% of the informative cases; for the two 8q markers, the percentages of loss were 11 and 25%, respectively, indicating preferential loss of (part of) 8p. In one tumor, two separate 8p deletions were found. The percentage of loss of heterozygosity was considerably higher in transurethral resection (65%) and lymph node metastases (83%) than in radical prostatectomy specimens (33%), suggesting that 8p deletion is a relatively late step in tumor progression. The maximal overlapping deleted region in all tumor DNAs is between the distal locus D8S133 and the proximal locus D8S87, indicating the localization of a candidate tumor suppressor gene within this region.

Structure, chromosome localization, and regulation of expression of the interferon-regulated mouse Ifi54/Ifi56 gene family.

The interferon-alpha (IFN-alpha) regulated mouse Ifi54/Ifi56 gene family, which is composed of at least four members (Ifi54, Ifi56, Ifi56-ps1, and Ifi56-ps2), was isolated and characterized. In addition, the chromosomal localization of the four genes was determined. The Ifi54 and Ifi56 genes show an identical organization. Both are composed of a very small first exon and a second exon, which contains the complete open reading frame, except for the ATG start codon and the first two nucleotides of the second codon. In both genes, the two exons are separated by a small intron (5 and 2.5 kb, respectively). Expression of both genes is rapidly induced by IFN-alpha (within 2 h). The Ifi54 promoter region contains two sequences, which are closely related to the interferon stimulated response element (ISRE) consensus sequence (ISRE 1, GGTTTCAATTTCT, and ISRE 2, AGTGTTACTTTCT). The two elements are located directly adjacent to each other. A similar organization was recently established for the hamster Ifi54 promoter (Bluyssen et al., 1994). However, the mouse promoter is 70% less active than the hamster promoter. It turned out that ISRE 2 is hardly active, due to the G at position 4, which is a T in the hamster Ifi54 ISRE 2 and in the ISRE consensus sequence. The Ifi56 promoter region contains at a similar position two functional ISREs of identical strength (ISRE 1, AGTTTCAGTTTCT, and ISRE2, AGTTTCACTTTCC). In the Ifi56 promoter, the two ISRE motifs are separated by 6 bp. In addition to the Ifi56 gene, parts of two closely related genes (Ifi56-ps1 and Ifi56-ps2) were isolated. Both fragments contain an Ifi56-related open reading frame. However, we were unable to isolate the presumed first exon of Ifi56-ps1 and Ifi56-ps2, nor could we show expression of the genes. The Ifi54, Ifi56, Ifi56-ps1, and Ifi56-ps2 genes could all be assigned to mouse chromosome 19D1, suggesting a tight clustering.

Two different, overlapping pathways of transcription initiation are active on the TATA-less human androgen receptor promoter. The role of Sp1.

In this study, the minimal promoter requirements of the TATA-less human androgen receptor (hAR) gene promoter are described. The hAR promoter is characterized by a short GC-box (-59/-32) and a long homopurine stretch (-117/-60). Two major transcription initiation sites, AR transcription initiation site I (AR-TIS I, (+1/2/3)) and AR transcription initiation site II (AR-TIS II, (+12/13)) are located in a 13-base pair region (Faber, P. W., van Rooij, H. C. J., van der Korput, J. A. G. M., Baarends, W. M., Brinkmann, A. O., Grootegoed, J. A., and Trapman, J. (1991) J. Biol. Chem. 266, 10743-10749). Transient transfection of COS cells with hAR promoter deletion and mutant constructs, followed by RNA isolation and S1 nuclease protection analysis showed that the process of transcription initiation through AR-TIS I and AR-TIS II is regulated by different promoter sequences. The GC-box directed initiation from AR-TIS II but did not affect AR-TIS I utilization, which is dependent upon sequences between positions -5 and +57. Band shift analysis identified the transcription factor Sp1 as the protein interacting with the GC-box. A single Sp1 binding sequence was found to be present in the GC-box. Footprint analysis confirmed the interaction of Sp1 with this sequence. The differential initiation through AR-TIS I and AR-TIS II was substantiated by the introduction of point mutations in the Sp1 binding sequence: only mutations that specifically abolished Sp1 binding interfered with AR-TIS II utilization, but all mutations left AR-TIS I initiation intact.

Regulation of androgen receptor expression in the human heterotransplantable prostate carcinoma PC-82.

In vivo effects of androgen withdrawal and substitution on human androgen receptor (hAR) expression were evaluated in the androgen-dependent human prostatic carcinoma tumor line PC-82. By application of several antibodies reactive with different epitopes of the hAR molecule, hAR protein expression was studied in tumor transplants by immunohistochemistry and immunoblotting. hAR messenger RNA (mRNA) levels were quantitated in PC-82 tumor tissue with a S1-nuclease protection assay. Most PC-82 tumor cells (> 97%) from testosterone-supplemented mice displayed nuclear hAR protein expression immunohistochemically. The almost complete reduction of nuclear hAR immunoreactivity within 5 days after androgen withdrawal (< 10%) was restored after androgen substitution within 1 day. The immunochemical data were confirmed by Western blot analysis. In contrast, no significant changes were observed in hAR mRNA content of PC-82 cells after 5 days of androgen withdrawal. Correlating hAR expression with proliferative activity of PC-82 tumor tissue during endocrine manipulation, a rapid, castration-induced decline of the percentage of bromodeoxyuridine-labeled cells accompanied the loss of hAR. Androgen substitution in castrated male mice restored the proliferative activity. However, this increase of proliferative activity lagged at least 24 h behind the normalization of the hAR protein level. In contrast to the steroid receptor down-regulation by homologous ligands observed in other experimental models, our data support the concept of hAR up-regulation by androgen. Since the hAR mRNA content of PC-82 tumor tissue was hardly affected by castration, expression of the hAR in PC-82 is thought to be modulated by translational and/or posttranslational mechanisms.

The mouse androgen receptor. Functional analysis of the protein and characterization of the gene.

Screening a mouse genomic DNA library with human androgen-receptor (hAR) cDNA probes resulted in the isolation and characterization of eight genomic fragments that contain the eight exons of the mouse androgen-receptor (mAR) gene. On the basis of similarity to the hAR gene, the nucleotide sequences of the protein-coding parts of the exons as well as the sequences of the intron/exon boundaries were determined. An open reading frame (ORF) of 2697 nucleotides, which can encode an 899-amino-acid protein, could be predicted. The structure of the mAR ORF was confirmed by sequence analysis of mAR cDNA fragments, which were obtained by PCR amplification of mouse testis cDNA, using mAR specific primers. A eukaryotic mAR expression vector was constructed and mAR was transiently expressed in COS-1 cells. The expressed protein was shown by Western blotting to be identical in size with the native mAR. Co-transfection of HeLa cells with the mAR expression plasmid and an androgen-responsive chloramphenicol acetyltransferase (CAT) reporter-gene construct showed mAR to be able to trans-activate the androgen-responsive promoter in a ligand-dependent manner. Transcription-initiation sites of the mAR gene were identified by S1-nuclease protection experiments, and the functional activity of the promoter region was determined by transient expression of mAR promoter-CAT-reporter-gene constructs in HeLa cells. Structural analysis revealed the promoter of the mAR gene to be devoid of TATA/CCAAT elements. In addition, the promoter region is not remarkably (G + C)-rich. Potential promoter elements consist of a consensus Sp1 binding sequence and a homopurine stretch. The polyadenylation sites of mAR mRNA were identified by sequence similarity to the corresponding sites in the hAR mRNA.

Characterization of the human androgen receptor transcription unit.

A full length human androgen receptor (hAR) cDNA was constructed from cDNA and genomic clones. Structurally the 10.6-kilobase (kb) hAR cDNA consists of a long 5'-untranslated region (5'-UTR, 1.1 kb), a previously described open reading frame (ORF, 2.7 kb) (Trapman, J., Klaassen, P., Kuiper, G. G. J. M., van der Korput, J. A. G. M., Faber, P. W., van Rooij, H. C. J., Geurts van Kessel, A., Voorhorst, M. M., Mulder, E., and Brinkmann, A. O. (1988) Biochem. Biophys. Res. Commun. 153, 241-248; Faber, P. W., Kuiper, G. G. J. M., van Rooij, H. C. J., van der Korput, J. A. G. M., Brinkmann, A. O., and Trapman, J. (1989) Mol. Cell. Endocrinol. 61, 257-262), and a very long 3'-untranslated region (3'-UTR, 6.8 kb). The complete 5'- and 3'-UTRs were found to be encoded by the previously reported first and eight protein coding exons of the hAR gene, respectively (Kuiper, G. G. J. M., Faber, P. W., van Rooij, H. C. J., van der Korput, J. A. G. M., Ris-Stalpers, C., Klaassen, P., Trapman, J., and Brinkmann, A. O. (1989) J. Mol. Endocrinol. 2, R1-R4). Two major sites of transcription initiation were identified in a 13-base pair region. DNA fragments spanning these transcription initiation sites conferred promoter activity upon a promoterless chloramphenicol acetyltransferase reporter gene construct. Two equally effective, functional polyadenylation signals (ATTAAA and CATAAA) at a mutual distance of 221 base pairs were detected. The ATTAAA hexamer sequence gave rise to multiple sites of poly(A) addition, whereas only one position was used following the CATAAA hexamer. In LNCaP prostatic carcinoma cells an alternatively spliced hAR mRNA species was identified which lacks 3 kb of the 3'-UTR.

The androgen receptor: functional structure and expression in transplanted human prostate tumors and prostate tumor cell lines.

The growth of the majority of prostate tumors is androgen-dependent, for which the presence of a functional androgen receptor is a prerequisite. Tumor growth can be inhibited by blockade of androgen receptor action. However, this inhibition is transient. To study the role of the androgen receptor in androgen-dependent and androgen-independent prostate tumor cell growth, androgen receptor mRNA expression was monitored in six different human prostate tumor cell lines and tumors, which were grown either in vitro or by transplantation on (male) nude mice. Androgen receptor mRNA was clearly detectable in three androgen-dependent (sensitive) tumors and absent or low in three androgen-independent tumors. Growth of the LNCaP prostate tumor cell line can be stimulated both by androgens and by fetal calf serum. In the former situation androgen receptor mRNA expression is downregulated, whereas in the latter no effect on androgen receptor mRNA levels can be demonstrated. Sequence analysis showed that the androgen receptor gene from LNCaP cells contains a point mutation in the region encoding the steroid-binding domain, which confers an ACT codon encoding a threonine residue to GCT, encoding alanine.

The rat androgen receptor gene promoter.

The androgen receptor (AR) is activated upon binding of testosterone or dihydrotestosterone and exerts regulatory effects on gene expression in androgen target cells. To study transcriptional regulation of the rat AR gene itself, the 5' genomic region of this gene was cloned from a genomic library and the promoter was identified. S1-nuclease protection analysis showed two major transcription start sites, located between 1010 and 1023 bp upstream from the translation initiation codon. The area surrounding these start sites was cloned in both orientations in a CAT reporter plasmid. Upon transfection of the constructs into COS cells, part of the promoter stimulated transcription in an orientation-independent manner, but the full promoter showed a higher and unidirectional activity. In the promoter/reporter gene constructs, transcription initiated from the same positions as in the native gene. Sequence analysis showed that the promoter of the rat AR gene lacks typical TATA and CCAAT box elements, but one SP1 site is located at about 60 bp upstream from the major start site of transcription. Other possible promoter elements are TGTYCT sequences at positions -174 to -179, -434 to -439., -466 to -471, and -500 to -505, resembling half-sites of the glucocorticoid-responsive element (GRE). Furthermore, a homopurine stretch containing a total of 8 GGGGA elements and similar to sequences that are present in several other GC-rich promoters, is located between -89 and -146 bp upstream from the major start site of transcription.