Symmetric division of cancer stem cells--a key mechanism in tumor growth that should be targeted in future therapeutic approaches.

As cancer stem cells (SCs) drive tumor growth, it is only through the elimination of those cancer SCs that a pharmacologic cure can be attained. To study ways to develop drugs that target cancer SC, we investigated changes in cellular mechanisms and kinetics that occur in SC populations during colorectal cancer (CRC) development. We used computer modeling to determine which changes could give rise to exponential increases in both SC and non-SC populations in CRC. Our results show that the only mechanism that can explain how these subpopulations increase exponentially in CRC development involves an increase in symmetric SC cell division. This finding suggests that any systemic therapies designed to effectively treat CRC and other cancers must act to control or eliminate symmetrical cancer SC division in tumors, while minimally affecting normal SC division in non-tumor tissues.

Increased expression of the retinoblastoma gene in human colorectal carcinomas relative to normal colonic mucosa.

We report the first evidence of increased levels of the retinoblastoma (Rb) message in a majority of colorectal cancers when compared with normal mucosa. Southern blot analysis showed an increase in Rb gene copy number in at least 28% of colorectal carcinomas relative to normal mucosa. These results plus previous reports of nonrandom chromosome 13 gains in approximately 50% of colorectal cancers suggest that an increase in Rb gene copy number occurs frequently in these tumors. Possible mechanisms pertaining to overexpression of the Rb gene are discussed in relation to its role as a recessive cancer gene.

Computer modeling implicates stem cell overproduction in colon cancer initiation.

On the basis of our investigation of the premalignant crypt phenotype in familial adenomatous polyposis patients, the hypothesis is developed that tumor initiation in the colon is caused by crypt stem cell overproduction. A novel kinetic model for the colonic crypt was used to investigate how the earliest tissue abnormality (altered crypt labeling index) arises in these patients who have a mutant APC genotype. Only an increase in crypt stem cell number, not changes in the rate of cell cycle proliferation, differentiation, or apoptosis of the non-stem cell population, simulated this abnormality. This suggests that APC regulates the number of stem cells in the colonic crypt and when the cells become mutant, an expansion of the crypt stem cell population results.

Evidence that APC regulates survivin expression: a possible mechanism contributing to the stem cell origin of colon cancer.

Because colorectal cancers (CRCs) frequently display APC mutation, inhibition of apoptosis, and increased expression of the antiapoptotic protein survivin, we hypothesized that APC mutation inhibits apoptosis by allowing constitutive survivin expression. Using HT-29 CRC cell lines having inducible wild-type APC (wt-APC) or transfected dominant-negative TCF-4, we show that wt-APC down-regulates survivin expression via APC/beta-catenin/TCF-4 signaling. Using normal colonic epithelium, we found survivin by immunostaining/reverse transcription-PCR to be preferentially expressed in the lower crypt (which inversely correlates with wt-APC's expression pattern). Thus, wt-APC, by progressively decreasing survivin and increasing apoptosis from crypt bottom to top, may limit the population size of stem cells and other proliferative cells in the lower crypt; mutant APC may allow expansion of these populations, thereby initiating tumorigenesis.

Cellular growth response to epidermal growth factor in colon carcinoma cells with an amplified epidermal growth factor receptor derived from a familial adenomatous polyposis patient.

The receptor binding and cellular growth responses to exogenous epidermal growth factor (EGF) were studied using the DiFi cell line established from a familial adenomatous polyposis patient. The number of cell membrane EGF receptors on DiFi cells, as measured by competitive radioligand binding assays and Scatchard analysis of 125I-EGF binding isotherms, was calculated to be 4.8 x 10(6) receptors/cell. An acid prewash step performed prior to ligand binding assays did not reveal additional receptor numbers. A single, low-affinity receptor population was identified by Scatchard analysis, with an apparent Kd of 4.6 nM. This result was confirmed by radioligand binding studies performed in the presence and absence of the receptor-antagonist monoclonal antibody 528 IgG that binds predominantly to the low-affinity form of the EGF receptor. DiFi cells at 50-60% confluence, when exposed to 50 nM exogenous EGF, exhibited a rapid but partial (30%) reduction in their cell membrane-associated receptor, characteristic of sequestration. Exposure of DiFi cells to 50 nM EGF for longer periods of time (4 h) did not result in any further reduction in EGF-receptor number. The cellular growth response of DiFi cells to exogenous EGF was studied in monolayer cultures as well as in a soft agarose assay. Inhibition of soft agar colony formation was observed at exogenous EGF concentrations greater than 1.7 nM, and inhibition of monolayer growth occurred at EGF concentrations greater than 1 nM. In immune complex kinase assays, the DiFi receptor showed similar specific activity to that from the well-characterized A431 cell line. Additionally, phosphorylation of the receptor on tyrosine was qualitatively similar to that of A431 cells, further suggesting that the DiFi receptors identified by EGF-binding studies were biologically functional.

Ifosfamide and mesna: marginally active in patients with advanced carcinoma of the pancreas.

Ifosfamide, an analogue of cyclophosphamide, has therapeutic activity against a wide variety of human malignancies. In a phase II trial in carcinoma of the pancreas, we treated 31 patients who had not received prior chemotherapy with a median ifosfamide dose of 2 g/m2/d (range, 1.5 to 2 g/m2/d) administered intravenously (IV) over one hour for five consecutive days every 3 weeks. 2-mercaptoethane sulphonate (mesna), an acrolein antagonist with known uroendothelial protective properties, was administered IV at a dose of 400 mg/m2 over 15 minutes before the daily dose of ifosfamide and repeated every four hours for two additional doses. Among 30 evaluable patients, one patient achieved a complete remission (26+ months) and another patient had a partial remission (4 months). The median duration of survival of all patients from the start of ifosfamide therapy was only 3 months (range, 1 to 26+ months). Treatments were generally well tolerated. The most common toxic effects included granulocytopenia, nausea and vomiting, malaise, anorexia, and mild hematuria. Mesna offers an adequate protection against uroendothelial injury caused by ifosfamide. Despite the previously reported response rate of greater than 20% at the same or lower doses of ifosfamide in other studies, our data suggest that ifosfamide is only marginally active against cancer of the pancreas and appears to be of minimal value in the treatment of patients with this tumor.

Manganese and magnesium dependent properties and inner plasma membrane surface localization of guanylate cyclase from murine plasmocytoma cells.

The particulate fraction from murine plasmocytoma cells contained 90 per cent of the total guanylate cyclase activity. Triton X-100 produced a 6 fold stimulation of guanylate cyclase activity in plasma membrane enriched fractions obtained by zonal centrifugation. Isolated inside out (10) vesicles contained 9 times more activity than rightside out (RSO) vesicles. This difference was abolished by Triton X-100 treatment of the vesicles indicating that the catalytic site of guanylate cyclase is located on the inner face of the plasma membrane. Kinetic studies of membranous guanylate cyclase showed that optimal activity was found with manganese. Only 20 per cent of this activity was obtained with magnesium. The Km for GTP with magnesium (1.4 mM) was about 7 fold greater than with manganese (0.2 mM). Positive cooperativity was obtained in both cases and the Hill coefficients were 1.8 for manganese and 1.6 for magnesium. Physiological concentrations of ATP were found to inhibit both manganese and magnesium supported activities indicating a possible regulatory mechanism for this nucleotide in vivo.

Phase II study of adriamycin with sequential methotrexate and 5-fluorouracil (AMF) in gastric carcinoma.

The purpose of this study was to evaluate the response rate, methotrexate plasma levels, and toxicity of a three-drug regimen in patients with gastric carcinoma. A total of 37 patients with advanced measurable adenocarcinoma of the stomach were treated with Adriamycin, methotrexate, and 5-fluorouracil (AMF). Adriamycin and methotrexate were given as i.v. infusions on day 1; 24 h following methotrexate administration, patients received an i.v. infusion of 5-fluorouracil concomitantly with oral leukovorin factor (given over 48 h). Methotrexate levels were monitored regularly in all patients, and courses were repeated every 3 weeks. The median dose levels per course were 50 mg/m2 (range, 40-60 mg/m2) for Adriamycin, 1,000 mg/m2 (range, 650-1,250 mg/m2) for 5-fluorouracil, and 500 mg/m2 (range, 160-625 mg/m2) for methotrexate. Of 36 evaluable patients, 8 (22%) achieved an objective response, including 1 complete remission. Stable disease was noted in 11 patients and a minor tumor regression occurred in 1. The median survival duration of all patients was 6 months (range, 2-31+ months). AMF was well tolerated; toxicities were mild to moderate, most frequently involving nausea and vomiting, mucositis, and neutropenia with or without fever. There was no death directly attributable to chemotherapy. Although the AMF regimen used a well-documented preclinical concept of synergism between methotrexate and 5-fluorouracil, response and survival results suggest a modest activity of this combination in patients with gastric cancer. Better preclinical models are necessary for the development of effective combination chemotherapy.

Growth effects of tamoxifen on Lovo colon carcinoma cells and cultured cells from normal colonic mucosa.

The growth effects of tamoxifen (TAM) were studied using normal and malignant colonic epithelial cells. Addition of TAM (0.8 to 10 microM) to cultured normal human colon epithelial cells and Lovo colon adenocarcinoma cells produced a dose response decrease in growth of 42 to 76%. Histamine (1 to 10 microM) did not affect cell growth and did not negate the TAM effect. Calmodulin (2 micrograms/ml) totally blocked growth inhibitory effects of a calmodulin antagonist, trifluoperazine (10 microM), but did not block the TAM inhibitory effect. These data suggest that antiestrogen binding sites (AEBS) may play an important role in the growth-inhibitory effects of TAM on colon cells. Competition with estrogen and antagonism with histamine or calmodulin do not appear to be significant in this regard.

The effects of steroid hormones on a human colon cancer cell line in vitro.

Estrogen analogues, moxestrol (10(-8)-10(-5) M) and ethinyl estradiol (10(-8)-10(-6) M), produced a 30% and 15% inhibition of LoVo cell growth, respectively, in serum-free Ham's F-10 medium. Under the same conditions, no growth effects were observed on these cells following the addition of progesterone or testosterone (10(-8)-10(-6) M); however, metribolone (10(-8)-10(-6) M), a synthetic androgen with glucocorticoid receptor-binding properties, moderately stimulated cell growth (18%). The synthetic antiandrogen, RU 23908 (10(-6) M), did not reduce metribolone effects, and hydrocortisone (10(-9)-10(-7) M) stimulated LoVo cell growth by 31% in serum-free medium. In medium containing 10% charcoal-treated fetal bovine serum, the inhibitory effects of estrogens were not observed, and the lower concentrations (10(-11) M) of moxestrol and ethinyl estradiol facilitated cell growth (10 to 15%). The other steroid hormones produced the same results as observed with serum-free medium. These data suggest that estrogen and glucocorticoid hormones may play an important role in the growth of colon carcinoma cells. Androgen and progesterone hormones appear to be less significant in this regard. Serum factors alter the effects of estrogen, but not of glucocorticoids.

Expression of the gene coding for the light chain of calpactin I (annexin II) in cell lines DiFi, HT-29, and WI-38.

Calpactin I (annexin II) light chain gene messages were expressed in the DiFi and HT-29 human colon cancer cell lines, as well as in the diploid lung fibroblast cell line WI-38. However, expression of an approximately 1.0 kb transcript was stronger in DiFi and HT-29 cells than in WI-38 cells. The moderate to strong expression of such transcripts in DiFi and HT-29 cells indicates that the calcium binding protein, calpactin I, may be abundant in colon carcinoma cells. Calpactin I is the major substrate of pp60v-src, a tyrosine-specific protein kinase encoded by v-src, whose cellular homologue c-src also codes for a tyrosine kinase (pp60c-src), known to be activated in colon carcinomas and in cell lines derived from them (including HT-29). Abundance of calpactin I in such cells is consistent with the possibility that activation of the pp60c-src tyrosine kinase contributes to the origin of human colon cancers.

A study of chromosome 6 allele loss in human colorectal carcinomas.

Matched normal/tumor DNA pairs from patients with sporadic and hereditary (FAP = familial adenomatous polyposis) colorectal carcinoma were examined for tumor-specific allele loss on chromosome 6 using cDNA probes for the avian myeloblastosis viral oncogene homologue (MYB on 6q22-q23), the estrogen receptor (ESR on 6q24-q27), and for the alpha polypeptide of human chorionic gonadotropin (CGA on 6q14-q21). No chromosome 6 allele loss was observed at these gene loci among 22 colorectal carcinomas examined, although such losses were relatively frequent (37.5% of informative individuals) at the D17S28 locus on chromosome 17. These results are consistent with karyological studies and indicate that chromosome 6 allele loss from colorectal carcinomas may occur less frequently than previously reported.

Co-transfection of MDR1 and MRP antisense RNAs abolishes the drug resistance in multidrug-resistant human lung cancer cells.

The resistance of lung cancer cells to the therapeutic actions of anticancer drugs is a serious clinical problem often encountered during cancer chemotherapy. It is very important, therefore, to investigate how to prevent and/or reverse this drug resistance. To this end, we took advantage of the fact that the overexpression of MDR1 and MRP genes, two genes known to be associated with the development of drug resistance, is very common in lung cancer. We used antisense RNA in an attempt to prevent expression of the protein products of these genes. Using a retrovirus, we introduced the antisense RNAs of MDR1 and MRP genes into doxorubicin-selected, multidrug-resistant GAOK cells, a cell which overexpresses both MDR1 and MRP genes. The expression levels of the products of the MDR1 gene (Pgp) and MRP gene (Mrp) in the transfected cells were analyzed using flow cytometry, and the drug resistances of the transfected cells were detected by a cell viability (MTT) assay. The expression of Pgp and Mrp in the transfected cells was almost completely inhibited by the antisense RNAs: expression levels decreased 64% and 93%, respectively. In parallel, the drug resistance of these cells decreased about 99% to doxorubicin, 98% to vinblastine, and 97% to colchicine. These results show that: a) antisense RNAs can attenuate drug resistance, an inhibition that might lead to new treatments for patients who are, or become, refractory to conventional chemotherapy; b) MDR1 and MRP appear to be cooperating to confer drug resistance in GAOK cells.

The role of vitamin D3 in the proliferation of a human colon cancer cell line in vitro.

LoVo, a cultured colon cancer cell line, is shown to possess a receptor for 1,25-dihydroxy vitamin D3 (1 alpha,25(OH)2D3) with a low capacity (28 fmol/mg protein) and high affinity (Kd: 1.9 x 10(-21)0M). When these cells were grown in monolayer culture in a chemically defined serum-free medium, a significant inhibition of proliferation was seen in the presence of 10 nM to 1 microM of 1 alpha,25(OH)2D3 (p less than 0.005. Furthermore, 1 alpha,25(OH)2D3 delayed early attachment of cells. After 8 days of treatment, aggregated cuboidal cells showed a marked change to an apparently spindle like morphology. The 1 alpha,25(OH)2D3 growth-inhibitory effect was modulated by verapamil (1 microM), a calcium channel blocker, hydrocortisone (1 microM), and moxestrol (1 mM), an estrogen analogue, and 2% charcoal-treated fetal bovine serum. This study represents the first demonstration of 1 alpha,25(OH)2D3 modulation of growth of human colon cells.

Immunodetection of the presence or absence of full-length APC gene product in human colonic tissues.

In order to detect the presence or absence of wild-type adenomatous polyposis coli (APC) gene protein (APC) in human colonic tissues, we immunoaffinity purified two polyclonal rabbit antibodies (APC-1 and APC-2) directed against defined epitopes in the middle and carboxyl regions of APC. Such antibodies proved useful in western blot analysis of matched colonic mucosa and tumor sample pairs. A 300 kDa band corresponding to APC was detected in samples from normal colonic mucosa using both antibodies. No tumor samples (n = 14) showed a detectable 300 kDa band. SW480 colon carcinoma cells, known to express truncated APC lacking the carboxyl half of the protein, were also negative. These results indicate that our antibodies bind to full-length but not truncated APC. Thus, western blot analysis employing APC-1 and APC-2 antibodies may be used to evaluate the absence or presence of wild-type APC. The value of this methodology in detecting APC mutations, which mainly involve protein truncation or allelic loss, is based on its ability to demonstrate negative or reduced level of immunoreactivity toward full-length APC in tissues that contain such mutations.

Characterization of beta-adrenergic receptors in DiFi and HT-29 cells.

The beta-adrenergic receptors present in two human colorectal adenocarcinoma cell lines were characterized by measuring specific binding of [125I]-cyanopindolol (CYP). Whole, cultured, DiFi (derived from a familial adenomatous polyposis [FAP] patient) and HT-29 cells were used in radioligand binding assays. Scatchard analysis of specific 125I-CYP binding gave KDS of 38.6 +/- 5.7 pM in DiFi cells and 54 +/- 9.1 pM in HT-29 cells. However, binding site density (Bmax) in the DiFi cells was greater than that in HT-29 cells. In DiFi cells, the kinetically determined KD was similar to that calculated from Scatchard analysis. Studies in DiFi cells of the displacement of specific 125I-CYP binding by nonselective (propranolol), beta 1-selective (metoprolol and atenolol), and beta 2-selective (ICI 118-551) antagonists revealed only a single class of beta 2-adrenergic receptors. This provides the first evidence that colorectal adenocarcinoma cell lines contain beta-adrenergic receptors and shows that only beta 2-adrenergic receptors are present in DiFi cells. Mechanisms possibly affecting beta-adrenergic-receptor expression in such cells are discussed in relation to colon carcinogenesis.

Development of cross-resistance to tamoxifen in raloxifene-treated breast carcinoma cells.

Selective estrogen receptor modifiers (SERMs) are used chronically in the treatment of breast cancer and osteoporosis but some patients become resistant, at which point second-line SERMs are considered as options. Because the use of SERMs is increasing and breast cancer is so common, we tested the hypothesis that treatment with SERMs can induce cross-resistance to other SERMs. We used three cultured breast carcinoma cell lines (MCF-7, ZR-75-1, and T47D) which are estrogen-receptor-positive (ER+) and are prone to developing resistance to hormonal treatment. Cell lines were exposed to increasing doses of raloxifene. Raloxifene-resistant clones were selected and tested for cross-resistance to tamoxifen. Compared to untreated cells, raloxifene-resistant clones showed an increased IC50 (reduced potency) of about 15,000-fold with no apparent change in maximal inhibition of cell growth. These same raloxifene-resistant clones were also about 15-fold more resistant to the growth-inhibiting effects of tamoxifen. While the resistance to tamoxifen is considerably less marked (1000-fold less), it is large enough to raise the question as to whether patients who become resistant to raloxifene will benefit by switching to tamoxifen or vice versa.

Genetic changes at the beta-2-adrenergic receptor locus on chromosome 5 in human colorectal carcinomas.

Matched normal/tumor DNA pairs from 44 colorectal carcinoma patients were examined for tumor-specific genetic changes using a probe for the beta-2-adrenergic receptor (ADRB2R) gene on chromosome 5. This locus (5q31-q32) maps close to the site of chromosomal deletions recently reported to occur in colorectal carcinomas and distal to the chromosomal location of the familial adenomatous polyposis (FAP) gene (5q21-q22). Our investigation shows tumor-specific allele loss or allelic rearrangement of at least 29% at the AdRb2R locus on chromosome 5 in informative cases. These results suggest that the mechanism by which colorectal carcinomas lose genetic material on chromosome 5 can affect this functional gene located distally to the FAP gene. The possible functional significance that ADRB2R gene changes may have in neoplastic progression is discussed.

Colorectal cancer staging and adjuvant chemotherapy.

Colorectal cancer is a significant cause of morbidity and mortality in Western populations. The standard of care for staging patients with colorectal cancer to determine prognosis and identify patients who will receive adjuvant therapy continues to be histopathology of regional lymph nodes. However, the significant variability in survival within each staging category likely reflects the heterogeneity of detecting micrometastatic disease employing this technique. Novel molecular markers of micrometastases currently in development will permit more accurate staging of patients with colorectal cancer. These advances in staging will distinguish patients who will maximally benefit from adjuvant therapy from those who have an especially good prognosis in whom chemotherapy can be avoided. In addition, new adjuvant chemotherapeutic agents, novel combinations of those agents and creative dosing schedules currently being investigated will offer considerable advantages with respect to ease of administration, safety and tolerability, quality of life and efficacy. Ultimately, it is anticipated that advances in molecular diagnostics will define unique biochemical characteristics of patients' tumours, permitting individualization of chemotherapeutic regimens employing novel agents that specifically exploit those characteristics.

A randomized study of two schedules of copovithane in patients with advanced colorectal carcinoma.

Copovithane (BAY i 7433), a copolymer of 1,3-bis(methylaminocarboxyl)-2-methylenepropanecarbamate and N-vinyl pyrrolidone, has antitumor activity in vitro and in vivo. The mechanism of this polymer is unknown. Thirty patients with advanced colorectal carcinoma with measurable tumors were treated in an open randomized study with two schedules of copovithane (schedule A: 6 g/m2 i.v. over 30 min daily for 5 consecutive days and repeated every 3 weeks; schedule B: 10 g/m2 i.v. over 30 min once a week until progression, unacceptable toxicity, or up to 18 months). Fifteen patients received copovithane on a weekly schedule, and 15 patients received it on an intermittent schedule. None of the patients on either schedule achieved a complete or partial remission. Copovithane was ineffective against colorectal carcinoma by both schedules selected in this study.