Modifications to the INSM1 promoter to preserve specificity and activity for use in adenoviral gene therapy of neuroendocrine carcinomas.

The INSM1 gene encodes a transcriptional repressor that is exclusively expressed in neuronal and neuroendocrine tissue during embryonic development that is re-activated in neuroendocrine tumors. Using the 1.7 kbp INSM1 promoter, an adenoviral HSV thymidine kinase gene therapy was tested for the treatment of neuroendocrine tumors. An unforeseen interference on the INSM1 promoter specificity from the adenoviral genome was observed. Attempts were made to protect the INSM1 promoter from the influence of essential adenoviral sequences and to further enhance the tissue specificity of the INSM1 promoter region. Using the chicken ß-globin HS4 insulator sequence, we eliminated off-target tissue expression from the Ad-INSM1 promoter-luciferase2 constructs in vivo. In addition, inclusion of two copies of the mouse nicotinic acetylcholine receptor (n(AchR)) neuronal-restrictive silencer element (NRSE) reduced nonspecific activation of the INSM1 promoter both in vitro and in vivo. Further, inclusion of both the HS4 insulator with the n(AchR) 2 × NRSE modification showed a two log increase in luciferase activity measured from the NCI-H1155 xenograft tumors compared with the original adenovirus construct. The alterations increase the therapeutic potential of adenoviral INSM1 promoter-driven suicide gene therapy for the treatment of a variety of neuroendocrine tumors.

The insulinoma-associated 1: a novel promoter for targeted cancer gene therapy for small-cell lung cancer.

The insulinoma-associated 1 (INSM1) gene is expressed exclusively during early embryonal development, but has been found re-expressed at high levels in neuroendocrine tumors. The regulatory region of the INSM1 gene is therefore a potential candidate for regulating expression of a therapeutic gene in transcriptionally targeted cancer gene therapy against neuroendocrine tumors. We analyzed expression of a reporter gene from a 1.7 kb region of the INSM1 promoter in a large number of small-cell lung cancer (SCLC) cell lines. This INSM1 promoter region showed very high levels of expression in most of the SCLC cell lines and expression was absent in cell lines of non-neuroendocrine origin. Inclusion of the general transcriptional enhancer from SV40 compromised the specificity of the promoter and did not enhance transcription in most of the SCLC cell lines. For comparison, the region of the gastrin releasing peptide (GRP) previously suggested for SCLC gene therapy was analyzed in a similar manner. High expression was observed for a number of cell lines, but unlike for the INSM1 promoter, reporter gene expression from the GRP promoter did not correlate to the relative GRP mRNA levels, demonstrating that this region may not contain all necessary regulatory elements. Expression of the suicide gene herpes simplex virus thymidine kinase (HSV-TK) from the INSM1 promoter in combination with treatment with the prodrug ganciclovir (GCV) caused a significant increase in GCV sensitivity specifically in INSM1-expressing cell lines. The INSM1 promoter is therefore a potential novel tool for transcriptionally targeted gene therapy for neuroendocrine tumors.

Multiple promoters exist in the human GR gene, one of which is activated by glucocorticoids.

A new human GR gene sequence (hGR 1Ap/e), which is distinct from the previously identified human GR promoter and coding sequences, has been isolated and characterized. The hGR 1Ap/e sequence is approximately 31 kbp upstream of the human GR coding sequence. This sequence (2,056 bp) contains a novel promoter (the hGR 1A promoter; 1,075 bp) and untranslated exon sequence (hGR exon 1A sequence; 981 bp). Alternative splicing produces three different hGR 1A-containing transcripts, 1A1, 1A2, and 1A3. GR transcripts containing exon 1A1, 1A2, 1B, and 1C are expressed at various levels in many cancer cell lines, while the exon 1A3-containing GR transcript is expressed most abundantly in blood cell cancer cell lines. Glucocorticoid hormone treatment causes an up-regulation of exon 1A3-containing GR transcripts in CEM-C7 T-lymphoblast cells and a down-regulation of exon 1A3-containing transcripts in IM-9 B-lymphoma cells. Deoxyribonuclease I footprinting using CEM-C7 cell nuclear extract reveals four footprints in the promoter region and two intraexonic footprints. Much of the basal promoter-activating function is found in the +41/+269 sequence, which contains two deoxyribonuclease I footprints (FP5 and FP6). When this sequence is cloned into the pXP-1 luciferase reporter gene, hormone treatment causes a significant increase in luciferase activity in Jurkat T cells that are cotransfected with a GR expression vector. FP5 is an interferon regulatory factor-binding element, and it contributes significantly to basal transcription rate, but it is not activated by steroid. FP6 resembles a glucocorticoid response element and can bind GRbeta. This novel hGR 1Ap/e sequence may have future applications for the diagnosis, prognosis, and treatment of T-cell leukemia and lymphoma.

The human glucocorticoid receptor promoter upstream sequences contain binding sites for the ubiquitous transcription factor, Yin Yang 1.

Studies on the human glucocorticoid receptor (GR) promoter were carried out so as to understand the regulation of GR expression. A -2738 to +19 fragment of the human GR promoter was used to identify important regulatory elements involved in the control in GR expression in NIH 3T3 cells (mouse fibroblasts) and HeLa cells (human cervical carcinoma cells). Important regulatory domains in the distal region of the human GR promoter were identified by sequential 5' end deletion analysis. A region between -2490 and -2025 contributed 50% of the measured transcriptional activity to the promoter. Using DNase I footprint analysis, four sites in this region were identified: -2362 to -2339 (mouse footprint, mFP); -2301 to -2293 (distal YY1, dYY1); -2130 to -2122 (middle YY1, mYY1); and, -2086 to -2078 (proximal YY1, pYY1). Three sites contained an identical core sequence, CCAAGATGG and were identified as Yin Yang 1 (YY1) binding sites. The site located at -2362 to -2339 was footprinted in NIH 3T3 cells only. The sequence of this site is a direct repeat with a 2-nucleotide spacer region, and it does not share homology with any known transcription factor binding sites. Computer analysis of the entire promoter sequence revealed an additional YY1 site located at -260 to -249 (initiator YY1, iYY1) with the sequence CTCCTCCATTTTG. Electrophoretic mobility supershift assays, with an anti-YY1 antibody, were used to confirm YY1 binding to these four putative YY1 binding sites. Site-directed deletion of all three upstream YY1 sites but not the iYY1 site, or the iYY1 site alone, showed a approximately 60% decrease in transcriptional activity of the hGR promoter in HeLa cells but had no effect in NIH 3T3 cells. A similar (50%) decrease in the expression of a full-length hGR/luciferase reporter gene was obtained when HeLa cells were cotransfected with a full-length antisense YY1 expression plasmid. Additionally, a region between -1841 and -1689 contributed to hGR promoter activity in both cell types tested. An Sp1 binding site was identified in this region (-1748 to -1733) by DNase I footprint and mobility supershift analyses. The presence of four YY1 binding sites in the human GR promoter suggests that these sites play a critical role in GR gene regulation.

The glucocorticoid receptor and c-jun promoters contain AP-1 sites that bind different AP-1 transcription factors.

The glucocorticoid receptor (GR) promoter contains several potential transcription factor recognition sites, including a putative AP-1 site. The GR promoter AP-1 site differs from the consensus AP-1 site by a nucleotide substitution, while the c-jun promoter contains a functionally characterized AP-1 site that varies from the consensus AP-1 site by a nucleotide insertion. Electrophoretic mobility shift assays were performed using nuclear extracts from a mouse pituitary tumor cell line (AtT-20) to test the binding capability of the AP-1 proteins, Jun and Fos, to the putative glucocorticoid receptor and the c-jun AP-1 sites. In addition, a comparison of the complexes formed at the GR AP-1 and c-jun AP-1 sites was done using antibodies specific for the Jun and Fos family members. The complexes formed with the GR AP-1 and c-jun AP-1 sites revealed striking differences. The GR AP-1 site formed complexes with both Jun and Fos family members. JunD was the most abundant Jun family member present, followed by JunB. cJun was absent from the complex. The amount of Fra-2 was greater than FosB in GR promoter AP-1 site complexes while Fra-1 was absent. A small amount of cFos may bind to the GR AP-1 site. In contrast to the GR promoter AP-1 site, only Jun family members were involved with complex formation on the c-jun promoter using AtT-20-cell nuclear extract, with JunD binding exceeding that of cJun. These results confirm previous studies suggesting that the c-jun promoter is stimulated solely by Jun family members. They also show preferential binding of Jun family members to different AP-1 sites present in different promoters. Finally, this study supports the hypothesis that the coordinate regulation of GR and c-jun gene regulation is mediated by crosstalk involving a Jun protein.

Cellular distributions of the prohormone processing enzymes PC1 and PC2.

The prohormone convertases PC1 (also known as sPC3) and PC2 are known to mediate the proteolytic conversion of inactive neuropeptide and hormone precursors to bioactive peptide products. In this study we have used sucrose density centrifugation to determine the subcellular distributions of the various forms of PC1 and PC2 in three different cell types, AtT-20, beta TC3, and PC12 cells. The former two cell lines naturally express PC enzymes, while PC12 cell clones expressing PCs were obtained by stable transfection. Our data show considerable cell-line specific variation in PC processing, with PC12 cells exhibiting the most complete processing of both enzyme precursors. While in all cell lines mature forms of both enzymes were stored within particles having the same buoyant density as secretory granule markers, in some cell lines substantial amounts of mature PC1 and PC2 were also associated with the Golgi marker. Processing of the two PC precursors was not interdependent since PC12 cells expressing only one of the two PCs were fully capable of enzyme maturation. Interestingly, analysis of intracellular processing of an endogenous peptide precursor, proneurotensin, revealed that transfected PC1, but not PC2, showed enzymatic activity against this precursor.

Differential processing of proenkephalin by prohormone convertases 1(3) and 2 and furin.

Recombinant vaccinia virus vectors were used to coexpress mouse prohormone convertase 1 (mPC1), mPC2, or human furin together with human proenkephalin in GH4C1 cells (rat pituitary somatomammotrophs) to examine the proteolytic processing of proenkephalin by these enzymes. Radioimmunoassays performed on high pressure gel permeation size-fractionated extracts obtained from GH4C1 cells and corresponding conditioned media revealed distinct profiles of immunoreactivity for products generated by each enzyme. PC1 produced intermediate sized processing products (3-10 kDa); the major immunoreactive enkephalin-containing species observed eluted at the positions of peptide B, the 5.3-kDa fragment, and free Leu5-enkephalin. PC2 exhibited a more complete processing profile. The major immunoreactive enkephalins produced were free Met5-enkephalin-Arg-Phe, free Met5-enkephalin-Arg-Gly-Leu, free Leu5-enkephalin, and free Met5-enkephalin. Thus PC2 appears to be more capable of generating active opioid units from proenkephalin than is PC1. Finally, furin cleaved proenkephalin to generate peptide B, an unidentified peak between the 18- and 5.3-kDa fragments, and a small amount of the 5.3-kDa fragment. Radiosequencing data verified that the production of the 5.3-kDa fragment by PC1 occurred as a result of a Lys-Lys cleavage. The ability of PC1 to cleave proenkephalin (but not proopiomelanocortin) at a Lys-Lys site implies that the structural context of the paired basic cleavage site may be more important in the determination of cleavage specificity than the particular pair of basic residues at the site.