A frequent alteration of p53 gene in carcinoma in adenoma of colon.

In general, colorectal carcinoma is thought to originate mainly from adenoma, and this pathway is called the adenoma-carcinoma sequence. Carcinoma in adenoma is an appropriate model for analysis of this mechanism, because adenoma and carcinoma tissues coexist in the same polyp and the carcinoma is thought to have originated from the surrounding adenoma. Expression of the p53 protein was analyzed in 36 cases of carcinoma in adenoma in the colon by immunohistochemistry using an anti-human p53 monoclonal antibody (PAb1801). Alterations of the p53 gene were analyzed by the polymerase chain reaction for microanalysis of normal mucosa, adenoma, and carcinoma from histological slides. Mutations were assessed by the polymerase chain reaction-single strand conformation polymorphism analysis and identified by DNA sequencing in some cases. Loss of heterozygosity was studied by polymerase chain reaction-restriction fragment length polymorphism analysis. Positive staining for p53 was detected in three (8%) of 37 adenomas and 20 (53%) of 38 focal carcinomas. One (7%) of 15 adenomas with mild dysplasia, three (14%) of 22 adenomas with moderate dysplasia, and 16 (42%) of 38 focal carcinomas had a mutation in exon 5 through exon 8 of the p53 gene. As for allelic loss in the p53 gene locus, only one adenoma with moderate dysplasia had loss of heterozygosity, whereas six (40%) of 15 focal carcinomas had loss of heterozygosity. Of those tumors (3 of 37 adenomas and 20 of 38 focal carcinomas) that reacted with PAb1801, 78% (18 of 23) showed genetic alterations. Among 52 tumors which showed negative staining, five tumors had a p53 mutation and four of them were nonsense mutations. Putting all of these results together, 71% (24 of 34) of the cases underwent p53 gene and protein alterations during the conversion from adenoma to focal carcinoma. These data clearly indicate that genetic alterations of p53 are involved mainly in the malignant transformation from adenoma to focal carcinoma in colon carcinogenesis. In addition, some cases show heterogeneity of the p53 gene in carcinoma in adenoma of the colon. There may be other pathways than p53 responsible for malignant change in the colon.

Identification of a novel gene marker specific for epithelial cells by utilizing a 3'-directed cDNA library.

To achieve reliability of molecular diagnosis using reverse transcription-PCR (RT-PCR), we established a unique method to search for a novel gene marker specific for colonic epithelial cells. Of eight candidate genes selected from a 3'-directed cDNA library in colonic mucosa, two genes were expressed in normal mucosa and cancer of the colon but not in either normal lymph node or normal liver tissue. Known sequences of these genes were reported to be located in the 3' noncoding region, and an additional sequence just upstream to gs04094 (one of the candidate genes) was determined. According to the newly identified sequence, we designed a new set of primers so that we could distinguish the DNA fragment amplified in RT-PCR from that in genomic PCR. RT-PCR using these primers demonstrated that gs04094 was expressed in all of 10 primary colon cancers and 4 liver metastases from colon cancer but in none of 5 normal lymph nodes, 10 peripheral blood samples, and 2 normal liver tissues. Sensitivity of this method was so high as to detect gs04094 mRNA in 10(-6) microg of colon cancer RNA per 1 microg of normal lymph node RNA. Thus, our strategy to search for a novel gene marker using 3'-directed cDNA library proved to be highly efficient.

Localization and expression of p27KIP1 in multistage colorectal carcinogenesis.

The cyclin-dependent kinase inhibitor p27Kip1 can inhibit the G1 to S transition of the cell cycle and is a putative tumor suppressor. However, our laboratory found that a variety of human cancer cell lines express relatively high levels of this protein and that this is often associated with increased expression of cyclin D1 or cyclin E. Therefore, in the present study we analyzed by immunohistochemistry the expression of p27Kip1 in a series of human tissue samples representing various stages of colon carcinogenesis, using 20 samples of normal colon mucosa, 20 hyperplastic polyps, 19 samples of adenomatous polyps, and 40 samples of various types of colorectal carcinomas. Parallel immunostaining was done for cyclin D1 and also for Ki67 to evaluate cell proliferation. An additional 17 human colon carcinoma samples, together with paired adjacent normal mucosa samples, were analyzed for levels of expression of the p27Kip1 protein by Western blot analysis, and 7 of these pairs of samples were examined by Northern blot analysis for levels of p27Kip1 mRNA. We did not find a positive or negative correlation between p27Kip1 expression and cell proliferation in the normal mucosa and tumor samples. There was, however, an inverse correlation between p27Kip1 and Ki67 expression in the lymphoid follicles present in the colonic mucosa. There was no evidence for a consistent increase or decrease in p27Kip1 expression in the mucosal cells during colon carcinogenesis, because the mean values for percentage p27Kip1-positive cells were similar in the normal mucosa, adenomatous polyps, and carcinoma samples. This is in contrast to Ki67 and cyclin D1 expression, which did show significant increases in mean values with tumor development. A subset (35%) of the carcinomas displayed diffuse cytoplasmic staining, in addition to nuclear staining, for p27Kip1, and in these cases the percentage of cells that were positive for p27Kip1 was higher than in cases that had only nuclear staining. There was a significant correlation between p27Kip1 expression and tumor grade; ie., well and moderately differentiated carcinomas had high p27Kip1 expression, whereas poorly differentiated carcinomas had lower expression. The Western blot analysis data on p27Kip1 expression confirmed this correlation. Comparisons of Northern and Western blots did not show a correlation between the level of p27Kip1 mRNA and the corresponding protein, a finding consistent with evidence that the p27Kip1 protein is regulated mainly via a posttranscriptional mechanism. The immunostaining studies revealed a significant correlation between high p27Kip1 protein expression and high cyclin D1 expression in the adenomatous polyps and in the subset of carcinomas that had only nuclear p27Kip1 expression. This may reflect the existence of a homeostatic feedback mechanism that is lost in the high-grade carcinomas that express low levels of p27Kip1.

Paradoxical increase in retinoblastoma protein in colorectal carcinomas may protect cells from apoptosis.

The retinoblastoma (Rb) gene is inactivated in a variety of human cancers, but in colorectal carcinomas there is frequently increased expression of this gene. This is paradoxical in view of the known role of Rb as a tumor suppressor gene. In the present study, we compared the levels of expression of the Rb protein (pRb) in normal human colorectal mucosa, adenomatous polyps, and carcinomas by immunohistochemistry. In vitro studies were also done to examine the phenotypic effects of an antisense oligodeoxynucleotide (AS-Rb) targeted to Rb mRNA in the HCT116 colon carcinoma cell line that expresses a relatively high level of pRb. The incidence of pRb-positive cells was increased during multistage colorectal carcinogenesis. In vitro treatment of HCT116 cells with AS-Rb decreased the level of pRb by about 70% and also decreased the levels of the cyclin D1 protein and cyclin D1-associated kinase activity. AS-Rb inhibited growth of HCT116 cells and induced apoptosis. Reporter assays indicated about a 17-fold increase in E2F activity. These findings suggest that the increased expression of pRb in colorectal carcinoma cells may provide a homeostatic mechanism that protects them from growth inhibition and apoptosis, perhaps by counterbalancing potentially toxic effects of excessive E2F activity.

Laparoscopic management for local recurrence of early gastric cancer after endoscopic mucosal resection.

Endoscopic mucosal resection (EMR) is a widely accepted technique for early gastric cancer because it is minimally invasive; however, incomplete resection with subsequent cancer recurrence in the remnant remains a difficult problem. Generally, the margins of the local recurrence lesions are unclear, and second EMR is difficult to perform because of scar formation after the first EMR. We performed a laparoscopic treatment on six patients with residual lesions after EMR and reviewed the safety and efficacy of this management. Laparoscopic management consisted of two techniques: laparoscopic wedge resection with a lesion-lifting method and laparoscopic-assisted distal gastrectomy with mini-laparotomy. Cancerous lesions were completely resected with sufficient surgical margins circumferentially. Mean operative time was 171 min, mean estimated blood loss was 16.5 g, time to first walking was 1 day, duration of epidural analgesia was 2.2 days, and mean length of hospital stay was 13.5 days. There were no intra- and postoperative complications, no conversion to open surgery, and no recurrence after surgery. No patients died of gastric cancer during a median follow-up of 60.3 months (range, 38-84). Laparoscopic management for residual lesions of early gastric cancer after EMR is a safe, effective, and minimally invasive procedure by which curative resection can be expected.

The human thyrotropin-releasing hormone gene is regulated by thyroid hormone through two distinct classes of negative thyroid hormone response elements.

TRH is the principal positive regulator of TSH synthesis and secretion in man. T3 is able to control TRH synthesis through feedback inhibition at the transcriptional level, presumably by binding to its receptor which interacts with one or more negative thyroid hormone response elements (TREs) present within the human TRH promoter. In the present study we have identified the specific negative TREs within the TRH promoter and characterized their ability to interact with thyroid hormone receptors (TRs), and the retinoid X receptor (RXR). Our analysis demonstrates that ligand-independent and dependent regulation of the human TRH promoter is restricted to the TR beta 1 isoform. Deletional analysis of the TRH promoter identified two discrete regions that are responsible for mediating ligand-dependent negative regulation of the TRH promoter. Mutagenesis of potential TR binding half-sites within these regions identified three separate half-sites (site 4 from -55 to -60 base pairs (bp); site 5, +14 to +19 bp; and site 6, +37 to +42 bp) which act in combination to allow for negative regulation. Mutation and/or deletion of each of these sites leads to a loss of negative regulation of the TRH promoter by T3. Gel-mobility shift assays of site 4 and its surrounding nucleotides revealed that this region of the promoter is capable of binding TR monomers, homodimers, and TR-RXR heterodimers. Mutagenesis of site 4 leads to a loss of all binding to this region. The region encompassing sites 5 and 6 binds only TR monomer, and the addition of RXR to the binding reaction leads to a loss of specific monomeric binding. To assess the functional importance of site 4 and its surrounding nucleotides we cotransfected RXR isoforms along with TR beta with TRH promoter constructs containing either site 4 or its mutant. In the presence of wild type site 4 sequence, cotransfected RXR enhanced negative regulation of the TRH promoter. Mutation and or deletion of site 4 leads to a loss of this enhancement. These data demonstrate that two structurally different negative TREs cooperate to allow for negative regulation of the human TRH promoter and that negative regulation is TR isoform-specific and modulated by the RXR-signaling pathway through a novel negative TRE.

Requirement of thyrotropin-releasing hormone for the postnatal functions of pituitary thyrotrophs: ontogeny study of congenital tertiary hypothyroidism in mice.

We recently reported that TRH-deficient mice showed characteristic tertiary hypothyroidism. In the present study, we investigated how this tertiary hypothyroidism occurred particularly in pre- and postnatal stages. Immunohistochemical analysis revealed a number of TSH-immunopositive cells in the TRH-/- pituitary on embryonic day 17.5 and at birth. The mutant pituitary at birth in pups born from TRH-deficient dams also showed no apparent morphological changes, indicating no requirement of either maternal or embryonic TRH for the development of pituitary thyrotrophs. In contrast, apparent decreases in number and level of staining of TSH-immunopositive cells were observed after postnatal day 10 in mutant pituitary. Similar decreases were observed in the 8-week-old mutant pituitary, while no apparent changes were observed in other pituitary hormone-producing cells, and prolonged TRH administration completely reversed this effect. Consistent with these morphological results, TRH-/- mice showed normal thyroid hormone levels at birth, but the subsequent postnatal increase was depressed, resulting in hypothyroidism. As expected, TSH content in the TRH-/- pituitary showed a marked reduction to only 40% of that in the wild type. Despite hypothyroidism in the mutant mice, both the pituitary TSHbeta and alpha mRNA levels were lower than those of the wild-type pituitary. These phenotypic changes were specific to the pituitary thyrotrophs. These findings indicated that 1) TRH is essential only for the postnatal maintenance of the normal function of pituitary thyrotrophs, including the normal feedback regulation of the TSH gene by thyroid hormone; 2) neither maternal nor embryonic TRH is required for normal development of the fetal pituitary thyrotroph; and 3) TRH-deficient mice do not exhibit hypothyroidism at birth. Moreover, reflecting its name, TRH has more critical effects on the pituitary thyrotrophs than on other pituitary hormone-producing cells.

Human immunodeficiency virus type 1 Tat binding protein-1 is a transcriptional coactivator specific for TR.

The DNA-binding domain of nuclear hormone receptors functions as an interaction interface for other transcription factors. Using the DNA-binding domain of TRbeta1 as bait in the yeast two-hybrid system, we cloned the Tat binding protein-1 that was originally isolated as a protein binding to the human immunodeficiency virus type 1 Tat transactivator. Tat binding protein-1 has subsequently been identified as a member of the ATPase family and a component of the 26S proteasome. Tat binding protein-1 interacted with the DNA-binding domain but not with the ligand binding domain of TR in vivo and in vitro. TR bound to the amino-terminal portion of Tat binding protein-1 that contains a leucine zipper-like structure. In mammalian cells, Tat binding protein-1 potentiated the ligand-dependent transactivation by TRbeta1 and TRalpha1 via thyroid hormone response elements. Both the intact DNA-binding domain and activation function-2 of the TR were required for the transcriptional enhancement in the presence of Tat binding protein-1. Tat binding protein-1 did not augment the transactivation function of the RAR, RXR, PPARgamma, or ER. The intrinsic activation domain in Tat binding protein-1 resided within the carboxyl-terminal conserved ATPase domain, and a mutation of a putative ATP binding motif but not a helicase motif in the carboxyl-terminal conserved ATPase domain abolished the activation function. Tat binding protein-1 synergistically activated the TR-mediated transcription with the steroid receptor coactivator 1, p120, and cAMP response element-binding protein, although Tat binding protein-1 did not directly interact with these coactivators in vitro. In contrast, the N-terminal portion of Tat binding protein-1 directly interacted in vitro and in vivo with the TR-interacting protein 1 possessing an ATPase activity that interacts with the activation function-2 of liganded TR. Collectively, Tat binding protein-1 might function as a novel DNA-binding domain-binding transcriptional coactivator specific for the TR probably in cooperation with other activation function-2-interacting cofactors such as TR-interacting protein 1.

Aberrant dynamics of histone deacetylation at the thyrotropin-releasing hormone gene in resistance to thyroid hormone.

Histone acetylation status influences transcriptional activity, and the mechanism of negative gene regulation by thyroid hormone remains unclear, although its impairment by a mutant thyroid hormone receptor (TR) is critical for resistance to thyroid hormone (RTH). We found a novel RTH mutant, F455S, that exhibited impaired repression of the TRH gene and had a strong dominant-negative effect on the gene. F455S strongly interacted with nuclear receptor corepressor (NCoR) and was hard to dissociate from it. To analyze the dynamics of histone acetylation status in vivo, we established cell lines stably expressing the TRH promoter and wild-type or F455S TR. Treatment with a histone deacetylase (HDAC) inhibitor completely abolished the repression of the gene by T3. The histones H3 and H4 at the TRH promoter were acetylated, and addition of T3 caused recruitment of HDACs 2 and 3 within 15 min, resulting in a transient deacetylation of the histone tails. TR and NCoR were located on the promoter, and T3 caused NCoR dissociation and steroid receptor coactivator-1 recruitment. In the presence of F455S, the histones were hyperacetylated, and HDAC recruitment and histone deacetylation were significantly impaired. This is the first report demonstrating the direct involvement of aberrant dynamics of chromatin modification in RTH.

Isoform variable action among thyroid hormone receptor mutants provides insight into pituitary resistance to thyroid hormone.

Resistance to thyroid hormone (RTH) is due to mutations in the beta-isoform of the thyroid hormone receptor (TR-beta). The mutant TR interferes with the action of normal TR to cause the clinical syndrome. Selective pituitary resistance to thyroid hormone (PRTH) results in inappropriate TSH secretion and peripheral sensitivity to elevated thyroid hormone levels. Association of the PRTH phenotype with in vitro behavior of the mutant TR has proved elusive. Alternative exon utilization results in two TR-beta isoforms, TR-beta 1 and TR-beta 2, which differ only in their amino termini. Although the TR-beta 1 isoform is ubiquitous, the TR-beta 2 isoform is found predominantly in the anterior pituitary and brain. To date, in vitro evaluation of RTH mutations has focused on the TR-beta 1 isoform. Site-directed mutagenesis was used to create several PRTH (R338L, R338W, V349M, R429Q, I431T) and generalized RTH (delta 337T, P453H) mutations in both TR-beta isoforms. The ability of mutant TRs to act as dominant negative inhibitors of wild type TR-beta function on positive and negative thyroid hormone response elements (pTREs and nTREs, respectively) was evaluated in transient transfection assays. PRTH mutants had no significant dominant negative activity as TR-beta 1 isoforms on pTREs found in peripheral tissues or on nTREs found on genes regulating TSH synthesis. PRTH mutants, in contrast, had strong dominant negative activity on these same nTREs as TR-beta 2 isoforms. Cotransfected retinoid X receptor-alpha was required for negative T3 regulation via the TR-beta 1 isoform but was not necessary for negative regulation via the TR-beta 2 isoform in CV-1 cells. The differing need for retinoid X receptor cotransfection demonstrates two distinct negative T3-regulatory pathways, one mediated by the TR-beta 1 and the other mediated by TR-beta 2. The selective effect of PRTH mutations on the TR-beta 2 isoform found in the hypothalamus and pituitary vs. the TR-beta 1 isoform found in peripheral tissues suggests a molecular mechanism for the PRTH disorder.

p120 acts as a specific coactivator for 9-cis-retinoic acid receptor (RXR) on peroxisome proliferator-activated receptor-gamma/RXR heterodimers.

p120 was originally isolated as a novel nuclear co-activator for thyroid hormone receptor. In this study, we characterized its interaction and transactivation of peroxisome proliferator-activated receptor-gamma (PPARgamma) and 9-cis-retinoic acid receptor (RXR) heterodimers. Transient transfection study revealed that p120 enhanced the transcriptional activation of PPARgamma/RXR induced by PPARgamma- or RXR-specific ligands. In the glutathione-S-transferase pull-down assay, while steroid receptor coactivator-1 showed apparent interactions with both RXR and PPARgamma, p120 bound only to RXR in a 9-cis-retinoic acid (RA)-dependent manner and also did not bind to PPARgamma even in the presence of thiazolidinediones. The yeast two-hybrid analysis showed no interaction of p120 with PPARgamma under any conditions, and electophoretic mobility shift assay showed apparent DNA-PPARgamma/RXR/p120 complex formation only in the presence of 9-cis-RA. Furthermore, the yeast three-hybrid assay clearly revealed a significant interaction between p120 and PPARgamma via RXR of PPARgamma/RXR heterodimer only in the presence of 9-cis-RA. These findings indicate that p120 acts as a specific co-activator for the RXR of PPARgamma/RXR heterodimer in a 9-cis-RA-dependent manner.

Abundance of cyclo (His-Pro)-like immunoreactivity in the brain of TRH-deficient mice.

Cyclo(His-Pro) or CHP was initially discovered as a metabolite of thyrotropin-releasing hormone (TRH) resulting from the action of the enzyme Pyroglutamyl aminopeptidase. Physiologic and pharmacologic studies that followed this initial discovery provided indirect evidence that all CHP may not be derived from TRH. However, the recent availability of a TRH-deficient mouse has made it possible to reinvestigate whether CHP is derived from TRH. In the present study, we examined distribution of CHP and TRH in TRH-deficient mice. Northern blot analysis confirmed the absence of preproTRH mRNA in both the hypothalamus and the cortex of TRH-deficient mice. Brains from the wild-type and TRH-deficient mice were dissected into 7 regions, and TRH and CHP concentrations were determined by specific radioimmunoassay (RIA) in each region. Whereas TRH was identified in all regions of the wild-type brain, with the highest concentration in the hypothalamus, no detectable TRH was observed in any region in the TRH-deficient mice. While CHP-like immunoreactivity (CHP-LI) was present in all regions in the wild-type brain, its concentration was reduced by approximately 50% in the hypothalamus and cerebral cortex of TRH-deficient mice, with no change in other brain regions. Furthermore, the CHP-LI present in the brain of TRH-deficient mice was immunologically and chromatographically identical to synthetic CHP. These findings strongly suggest that a portion of the CHP in the brain is derived from sources other than TRH.

Regulation of the mouse preprothyrotropin-releasing hormone gene by retinoic acid receptor.

Retinoic acid (RA) has been reported to inhibit the secretion and synthesis of the pituitary TSH in vivo and in vitro. However, little is known about the influence of RA on the expression of the prepro-TRH gene. We therefore investigated whether the promoter activity of the mouse TRH gene is directly regulated by RA using a transient transfection assay into CV-1 cells. In the absence of cotransfected RA receptor (RAR), all-trans-RA did not affect the promoter activity. In contrast, the cotransfected RARalpha significantly stimulated promoter activity in the absence of ligand, and all-trans-RA reversed basal promoter activation. The cotransfected thyroid hormone receptor-beta (TRbeta), but not 9-cis-RA receptor (RXR), had an additive effect on the RAR-dependent stimulation. TR and RAR can similarly interact with the corepressor proteins, and the cotransfected nuclear receptor corepressor (N-CoR) has been demonstrated to augment the transcriptional stimulation of the TRH gene by unliganded TR. As observed with TR, the coexpression of a N-CoR variant significantly enhanced the ligand-independent stimulation by RAR. A mutant RAR (RAR403) lacking the C-terminal activation function-2 (AF-2) activation domain that was essential for ligand-induced corepressor release constitutively stimulated the promoter activity. The constitutive stimulation by RAR403 was augmented by the cotransfected N-CoR variant. A deletion analysis of the 5'-flanking region of the TRH gene revealed that the minimal promoter region for the regulation by RAR was -83 to +53, with a consensus half-site motif for the thyroid hormone response element at -57. In contrast to the strong binding of TR to the thyroid hormone response element half-site in gel retardation assays, no binding of RAR homodimer, RAR/ RXR heterodimer, or RAR/TR heterodimer was observed to the minimal promoter region. These results collectively suggest that RAR without heterodimerization with RXR and TR regulates transcription of the mouse TRH gene in cooperation with the corepressor, and that the DNA binding of RAR appeared to be unnecessary for regulation of the TRH gene promoter.

Genomic organization and promoter function of the mouse uncoupling protein 2 (UCP2) gene.

We cloned and characterized the mouse uncoupling protein 2 (UCP2) gene and its promoter region. The gene spans approximately 6.3 kb and contains eight exons and seven introns. Two short exons are located in the 5' untranslated region, and each of the remaining exons encodes one of the transmembrane domains. 3'-RACE analysis showed that a polyadenylation signal 257 bp downstream from the stop codon was functional. Primer extension analysis indicated a single transcriptional start site 369 bp upstream from the translational start site. The promoter region lacks both TATA and CAAT boxes but is GC-rich. A construct containing 1250 bp of the promoter region showed significant activity in all 6 cell lines examined, and the region between -160 and -678 bp exhibited strong positive regulatory activity. These features of the UCP2 gene are different from those of the UCP1 gene and may contribute to its ubiquitous expression.

Transcriptional down-regulation by epidermal growth factor of TRH receptor mRNA in rat pituitary cells.

To gain insight into the mechanism underlying the epidermal growth factor (EGF)-induced changes in responsiveness to TRH and in the numbers of TRH receptors (TRH-Rs) in the pituitary, we investigated the transcriptional regulation by EGF of the TRH-R gene in GH4C1 cells. Northern blot analyses and binding studies revealed that EGF reduced both TRH binding and TRH-R mRNA levels in a dose- and time-dependent manner, while no significant changes were observed in beta-actin mRNA levels. Addition of actinomycin D caused an acute increase in the basal TRH-R mRNA level, and the rate of decrease of the TRH-R mRNA was identical in control and EGF-treated groups, suggesting that the stability of the TRH-R mRNA was not significantly affected in EGF-treated cells. Incubation with cycloheximide also induced an increase in the basal TRH-R mRNA level and completely reversed the EGF-induced reduction of TRH-R mRNA levels. Furthermore, a nuclear run-on assay demonstrated that the rate of transcription of the TRH-R gene was significantly inhibited in cells treated with EGF. We conclude that (1) EGF decreases the expression of the TRH-R mRNA largely by reducing its rate of transcription, and this action requires the synthesis of new proteins, and (2) inhibitors of protein and RNA synthesis cause a significant increase in the basal TRH-R mRNA level, suggesting that there may be a short-lived protein suppressing the TRH-R mRNA level in the pituitary.

Extraction and analysis of diagnostically useful proteins from formalin-fixed, paraffin-embedded tissue sections.

We describe and discuss a method of protein extraction for Western blot analysis from formalin-fixed, paraffin-embedded tissue sections. From 5-mm2 50-micron-thick tissue sections, an abundance of proteins could be extracted by incubating the sections in lysis buffer containing 2% sodium dodecyl sulfate (SDS) at 100C for 20 min followed by incubation at 60C for 2 hr. Extracts yielded discernible protein bands ranging from 10 kD to 120 kD as identified by SDS-polyacrylamide gel electrophoresis (PAGE). Western blot analysis successfully detected membrane-bound protein such as E-cadherin, cytosolic protein such as beta-catenin, and nuclear proteins including proliferating cell nuclear antigen (PCNA), mutant-type p53, cyclin D1, cyclin E, and cyclin-dependent kinases (CDKs). With this technique, we could examine cyclin D1 and CDK2 expression in small adenomas compared with cancer tissues and normal mucosa. The simple method of protein extraction described here should make it possible to use large-scale archives of formalin-fixed, paraffin-embedded samples for Western blot analysis, and its application could lead to detailed analysis of protein expression. This new technique should yield valuable information for molecular biology.

Expression of truncated midkine in human colorectal cancers.

Midkine (MK) is a growth differentiation factor originally found as the product of a retinoic acid-responsive gene. The expression of MK was examined in 35 surgically resected specimens of primary colorectal cancer using the reverse transcription-polymerase chain reaction (RT-PCR). All of the cancerous tissues expressed MK. In 5/25 cancerous tissues a truncated form of MK, which was recently found in various human tumor cell lines, was detected in addition to the full-size MK. In contrast, the truncated from of MK could not be detected in non-cancerous tissues, whereas the wild-type form was detected in 8/10 non-cancerous tissues. These results suggest that the expression of the truncated form of MK may be associated with tumorigenesis.

PyNPase expression in human colon cancer.

Although colorectal cancer tissue is rich in pyrimidine nucleoside phosphorylase (PyNPase), there is no consensus as to whether cancer cells or stromal cells predominately express PyNPase. We micro-dissected OCT compound embedded frozen tissue sections into epithelial and stromal components and then analyzed the extracted samples separately. The PyNPase expression level was higher in stromal cells than in cancer cells and the difference increased with inflammation induced by the immunostimulator OK432. These results suggest that stromal cells are the major PyNPase source in colorectal cancer.

DNA binding and interaction with the nuclear receptor corepressor of thyroid hormone receptor are required for ligand-independent stimulation of the mouse preprothyrotropin-releasing hormone gene.

A negative thyroid hormone response element (TRE) in the mouse preprothyrotropin-releasing hormone (TRH) gene was previously mapped within the proximal promoter element between -83 and +53 that contained a TRE half-site motif at -57 (-57TGACCT-51). In transfection experiments, the promoter activity is stimulated by unliganded thyroid hormone receptor (TR) and T3 reverses the basal promoter stimulation. In this study, we determined whether the direct binding of TR to the TRE half-site in the mouse TRH gene is required for the ligand-independent stimulation using a transient transfection assay into CV-1 cells and electrophoretic mobility shift assays (EMSA). In addition, the role of a corepressor protein for the ligand-independent stimulation was examined using a putative splicing variant of the nuclear receptor corepressor (N-CoRI). Point mutations introduced into the TRE half-site at -57 eliminated the binding of TR and the stimulatory effect of unliganded TR. Two mutant TRs lacking DNA-binding activity and two CoR box mutant TRs showed no stimulation in the wild-type TRH promoter. The cotransfected N-CoRI potentiated the ligand-independent stimulation by the wild-type TR, but did not compensate for the impaired function of the CoR box mutant TR. In EMSA, TR strongly bound as homodimers and weakly as heterodimers with retinoid X receptor (RXR) to the element containing the TRE half-site at -57. Binding of TR to the TRE half-site was essential to form homo- and heterodimers, and the RXR binding site appeared to be located downstream of the TRE half-site. In vitro translated N-CoRI preferentially bound TR homodimers over TR/RXR heterodimers. These results collectively suggest that the DNA-bound TR/corepressor complex might be directly involved in the ligand-independent stimulation of the mouse TRH gene promoter.

Immunohistochemical study of retinoblastoma gene-expression in colorectal carcinomas.

The expression of the retinoblastoma gene (Rb-1) in colorectal carcinomas was studied by Western blotting and immunohistochemistry. Western blot analysis using monoclonal antibodies raised against synthetic peptides of the Rb-1 gene product (pRB) revealed that colorectal carcinomas overexpressed pRB with a molecular weight of around 110 kD when compared to normal mucosa. Immunohistochemical staining of formalin-fixed, paraffin-embedded tissue sections from 48 colorectal carcinomas demonstrated that pRB expression was exclusively localized to the nucleus. More than 50% of the carcinoma cells expressed nuclear pRB in 14 tumors (29.2%), while 10-50% of the carcinoma cells did so in 21 tumors (43.8%), and less than 10% of cells did so in 13 tumors (27.1%). Although there was no clear correlation between pRB expression and clinico-pathologic parameters such as tumor stage, tumor size, depth of invasion, and lymph node metastasis, a higher incidence of pRB expression was observed in well to moderately differentiated adenocarcinoma than in the signet ring cell carcinoma. Thus the present study demonstrated for the first time that the oncosuppressor gene, Rb-1, is overexpressed at the protein level in most colorectal carcinomas.