Various isozyme gene expression patterns among human colorectal adenocarcinoma cell lines and tissues.

Seven tissue-specific isozymes--lactate dehydrogenases A and B (LDHA and LDHB), esterases B2 and A4 (ESB2 and ESA4), creatine kinase B (CKB), and red blood cell and tissue forms of adenosine deaminases (ADA)--were examined in the human adenocarcinoma cell line LS174T and its subclones grown under a wide variety of conditions. Strong expressions of ESB2, ESA4 and CKB were stable in all material studied, as was a greater expression of LDHA than of LDHB. ADA isozyme expression varied with culture conditions. Identical expression of the five stable isozymes was observed in six surgical specimens of normal bowel mucosa, while cell lines derived from normal mucosa varied in the expression of these isozymes. An additional 23 colorectal cell lines, derived from primary or metastatic adenocarcinomas in different stages of invasiveness and differentiation from 18 individuals, were verified as to their origin by genetic signature analysis and similarly studied for isozyme expression. These lines had different, yet stable, gene expression patterns similar to differences observed among six adenocarcinoma surgical specimens. These differences could not be related to the number of passages that the cell lines were in culture nor to the stages of the tumors at the time that the cell lines were initiated or surgical samples tested. Gene expression patterns in adenocarcinomas differed, these differences might have been due to the origin of colorectal adenocarcinomas from different cell types within a heterogeneous mucosa, and the colorectal cell lines studied represented appropriate models for colon cancer.

Human colon adenocarcinoma cells. III. In vitro organoid expression and carcinoembryonic antigen kinetics in hollow fiber culture.

Cell line LS174T was established in our laboratory from a primary human colon adenocarcinoma and was serially cultured for 4 years. Following 1 month of culture in perfused hollow fiber matrices, organoid growth reminiscent of the patient's original tumor was observed. The cellular organization was predominantly glandular with mucin located within gland lumina and cells. Glands had irregular brush borders, well-developed junctional complexes, and intracytoplasmic lumina lined with microvilli, features found in adenocarcinoma tissue. Alternate-day determinations of carcinoembryonic antigen (CEA) revealed biphasic kinetics in the extracapillary fluids of the hollow fiber system but not in identically treated monolayer cultures. The initial rate of CEA release was similar to that of monolayer cultures. A significantly accelerated secondary CEA release phase was observed in 10,000- and 50,000-molecular weight (MW)-exclusion fiber cultures (P less than or equal to 0.001 and P = 0.05, respectively). Totals of 43.8 micrograms and 112.6 micrograms CEA were released into the extracapillary fluids of 10,000- and 50,000-MW-exclusion fiber cultures, respectively, which were onefold and 2.3-fold increases, respectively, over monolayer supernatant yields. Most CEA was released during the final days of the secondary phase. In monolayer cultures, maximum CEA release occurred during the stationary phase of growth.

Human colon adenocarcinoma cells. II. Tumorigenic and organoid expression in vivo and in vitro.

A human colon epithelial tumor cell line (LS174T) recultured in vitro following passage through hamsters displayed differences in its cell doubling time and synthesis of carcinoembryonic antigen when compared with the cells grown solely in vitro. These animal-passaged cells more closely resembled the parent tumor cell line (LS180) derived from the primary tumor than LS174T, the trypsinized variant of LS180. Analysis of lactate dehydrogenase isoenzymes indicated that the tumor cells recovered from the hamsters were free of xenogeneic host tissue. Furthermore, LS174T grafted to athymic (nude) mice grew as a mucinous adenocarcinoma microscopically resembling the original tumor. The altered growth potential of LS174T was also demonstrated on confluent feeder monolayers of normal cells and by uninhibited multiplication in vitro. These results suggest that, at least in this one case, short-term passage of long-term cultured cells into xenogeneic hosts may effect a phenotypic reversion such that the cells regain properties observed in the primary tumor and the initial in vitro explant.

Cell surface changes associated with malignant transformation of bladder epithelium in vitro.

Three properties of cell surface membranes of normal bladder epithelium and of two malignant urothelial lines transformed in vitro by Dr. Y. Hashimoto and Dr. H. S. Kitagawa were analyzed by means of morphological and chemical tools. Under scan electron microscopy, normal bladder epithelium displayed an appearance devoid of the pleomorphic, abundantly distributed, microvillous structures seen on the neoplastic cells. The molecular weight profiles of proteins dispersed from purified plasma membrane fractions demonstrated quantitative differences in the content of three molecular-weight classes between the native and the transformed cells. More striking differences were seen upon two-dimensional analysis of proteins solubilized from these two cell types, using 3 M KCl. These findings suggest that further investigations of the chemical moieties appearing on the cell surface early after transformation may enhance our understanding of proteins amenable for chemical and/or immunological attack to achieve control of the progression of bladder neoplasia.

Longitudinal karyotype and genetic signature analysis of cultured human colon adenocarcinoma cell lines LS180 and LS174T.

Giemsa-banded chromosomes were analyzed at intervals during either 34 or 70 serial subcultivations of two cell lines, LS180 and LS174T, established from one primary human colon adenocarcinoma, and at passage 14 of autochthonous normal bowel cells, NB(LS174T). The cell lines were established and subcultured by either scraping or trypsin treatment of primary cultures; the scraped cell line was designated LS180, and the trypsin-dispersed cell line was named LS174T. Early passages of LS174T cells were composed mainly of 46,XX (38%) and 45,X (34%) karyotypes; LS180 cultures possessed cells with 46,XX (54%), 45,X (7.5%), and 47,XX+D (19.5%) chromosome modes. In both cell lines, the 45,X karyotype predominated in later subcultivations. After the fifth passage, all LS180 cells examined exhibited a translocation from the long arm of the X chromosome to the long arm of the No. 5 chromosome. Cultures from the patient's normal bowel mucosa and peripheral blood leukocytes had normal 46,XX karyotypes. Genetic signature analysis sustantiated the common genetic origin of the cell lines, and we concluded that differences observed between LS180 and LS174T were not due to contamination with other cell lines. LS180 and LS174T represent closely related cell lines differing cytogenetically in a translocation.

In vitro correlates of in vivo therapy with cyclosporine to immunosuppress rejection of heterotopic rat cardiac allografts across strong (RT-1) plus weak (non-RT-1) histocompatibility differences.

This study correlated different oral cyclosporine doses with in vivo graft survival, blood and tissue drug levels, and in vitro immune performances. Wistar-Furth (WFu, RT-1u) hosts engrafted with heterotopic cardiac transplants from strongly histoincompatible Buffalo (BUF, RT-1b) rats were treated postoperatively with 14-day courses of different doses of CsA delivered per gavage. There was a graded prolongation of graft survival--namely, no effect at the 1.5 mg/kg dose; a modest effect at 3 mg/kg; a therapeutic effect at 5 mg/kg; and long-term unresponsiveness at 10 mg/kg. Whole blood, serum, and tissue CsA concentrations correlated with drug dose. On day 7 posttransplantation--that is, during the peak of the immune response of untreated recipients and midway during the period of daily CsA therapy--in vitro immune performances were examined in each experimental group. On the one hand, the mixed lymphocyte reaction of WFu host splenic T cells toward donor-type BUF stimulators poorly reflected the administered CsA dose. On the other hand, there was a good correlation between drug dose and both impaired cell-mediated lympholysis and reduced frequency of alloantigen-specific T cytotoxic cell precursors f(CTL)p. Animals treated with therapeutic doses of CsA showed different patterns of T cell-mediated lympholysis: 3 mg/kg did not prevent anti-BUF Tc cell sensitization; 5 mg/kg maintained f(CTL)p levels similar to the normal controls; and 10 mg/kg significantly reduced Tc clones against donor but not third-party targets. These data demonstrate that the fate of alloantigen-specific Tc clones activated in vivo depends upon the local drug concentration. Furthermore, the present studies suggest that CML and f(CTL)p afford useful in vitro indices of in vivo immunosuppression with CsA in rat cardiac allograft recipients.

Variable oral absorption of cyclosporine. A biopharmaceutical risk factor for chronic renal allograft rejection.

The inter- and intrapatient variability in cyclosporine (CsA) pharmacokinetics obfuscates the relationship between therapeutic outcome and administered dose, thereby impeding the development of secure algorithms for CsA therapy. In an attempt to understand these variabilities, we previously performed serial pharmacokinetic profiles on 160 renal transplant recipients during the first 3 posttransplant months. Drug exposure was estimated by the average CsA concentration (Cav), which was defined as a time-corrected (tau, hours) expression of the area under the concentration-time curve (AUC), i.e., Cav = (AUC/tau). Low Cav values correlated with an increased occurrence of acute rejection episodes and 1-year rate of renal transplant loss. The present study examines the results of serial pharmacokinetic profiling of a cohort of 204 patients treated for up to 5 years with CsA doses selected to achieve target Cav values. Multivariate analyses correlated demographic factors, laboratory values, clinical parameters, and CsA pharmacokinetic parameters with the occurrence of chronic rejection. The factors that predisposed to chronic rejection included a previous acute rejection episode, initial acute tubular necrosis, diastolic blood pressure above 85 mmHg, and African-American race. Once regression models were adjusted to account for the impact of these factors, we examined the association between the incidence of chronic rejection and individual pharmacokinetic parameters, including the mean values of the absolute and dose-corrected trough, peak, and Cav concentrations, as well as the percent coefficient of variation of each of these values. Receiver operating characteristic curves documented that 27% of the total risk for the occurrence of chronic rejection was attributable to a greater than 20% coefficient of variation of the dose-corrected Cav, namely, AUC/(tau.mg). This study suggests that variable oral bioavailability of CsA represents a biopharmaceutical risk factor for the occurrence of chronic rejection.

Influence of demographic factors on cyclosporine pharmacokinetics in adult uremic patients.

The causes of variability in cyclosporine (CS) clearance (CL) are mostly unknown. The pharmacokinetics of CS were studied in 30 adult uremic patients after single intravenous and oral doses by analyzing serial concentrations in serum by radioimmunoassay (SR) and in whole blood by radioimmunoassay (WR) and high pressure liquid chromatography (WH). Bioavailability (F) and CL were calculated by noncompartmental models and were significantly different depending upon the assay method except for FSR = FWR: FSR = 43.2 +/- 21.7%; FWR = 43.5 +/- 18.5%; FWH = 36.4 +/- 17.3%; CLSR = 849 +/- 363 ml/min; CLWR = 380 +/- 156 ml/min; CLWH = 559 +/- 174 ml/min. The age of the patients and parameters describing body size such as weight, surface area and percent of ideal weight were not correlated with CL. The height of the patients correlated with CLWH but not CLSR or CLWR. Parameters responsible for CS binding in blood such as cholesterol, triglyceride, hemoglobin concentration or hematocrit did not explain variability in CL. Of the factors indicative of liver function alanine transaminase activity but not aspartate transaminase, lactate dehydrogenase, alkaline phosphatase activity nor total bilirubin concentration in serum was correlated with CL. F was not correlated with any of the demographic factors except for alanine transaminase. None of the significant correlations explained enough of the variability to afford a reliable prediction of CL or F.

Challenges in cyclosporine therapy: the role of therapeutic monitoring by area under the curve monitoring.

Cyclosporine has revolutionized the practice of transplantation, but its clinical application has been beclouded by a narrow therapeutic window between immunosuppressive and nephrotoxic concentrations. Marked intra- and interindividual pharmacokinetic differences preclude the use of routine dosing regimens. For example, at the intraindividual level, cyclosporine absorption improves during the first 90 days after institution of therapy. A wide range of demographic factors, namely, age, race, and concomitant drug therapy, as well as individual-specific factors produce unique pharmacokinetic behaviors in any given patient. We introduced a pharmacokinetic strategy for cyclosporine administration almost 10 years ago based on the observation that the best estimate of drug exposure was the area under the concentration-time kinetic curve (AUC) not the trough level. Early studies documented the relation between AUC and the incidence of acute rejection. Subsequent studies revealed that not only is the AUC an important predictor, but so is the consistency of drug absorption over time; namely, patients with variations > 25% among AUC determinations display an increased risk of chronic rejection episodes. Therefore therapeutic drug monitoring plays an important role in the optimal care of patients under cyclosporine therapy.

NMR relaxation times of water protons in human colon cancer cell lines and clones.

Established lines of human colon cancer cells from several sources (LS180, LS174T, HT29, SW480, SW1345) had water proton nuclear magnetic resonance (NMR) spin-lattice relaxation times (T1) of 460 +/- 45 msec to 982 +/- 9 msec and spin-spin relaxation times (T2) of 83 +/- 6 msec to 176 +/- 6 msec. Two clones derived from single cells of line LS174T were similar in T1 and T2 to the parent line. Differences among the cell lines were not totally a function of cellular hydration. Normal adult and fetal human primary colon cells were wetter and had higher T1 and T2 values than established cell lines. Relaxation times in this study substantiate variations seen for human colon tumors in earlier studies. Established cell lines maintained water relaxation times similar to tumor tissue values. Along with other morphological and biochemical criteria, the relaxation times suggest that these established human colon cancer cell lines may serve as a good experimental model for the study of human colon cancer.

A new approach for the immunogenic presentation of membrane-bound human colon tumor antigens.

Liposomes bearing human tumor membrane vesicles were effective immunogenic complexes for inducing antibodies in rabbits. Multilamellar liposomes (MLV) at 7:4:1 molar ratios of phosphatidylcholine, cholesterol and phosphatidic acid were prepared with sonicated membrane (MN) isolated from LS174T colon tumor cells. MLV liposomes prepared together with MN (i.e., MN-MLV antigens) or MN added to preformed MLV (i.e., MN+MLV antigens), and MN antigens alone were used as immunogens. Rabbits were immunized i.v. with 100 g protein of each antigen group. Boosters (i.v.) were at days 13 and 29. Binding assays were performed by indirect radioimmunoassay on viable tumor cell targets. The MN+MLV groups were distinguished by earlier reactivity and greater specificity to the colon tumor antigens.

Active specific immunotherapy potential for the treatment of large bowel cancer.

Active specific immunotherapy, harnessing the strength and specificity of the host immune response to destroy neoplastic cells, may offer an ideal surgical adjuvant treatment modality for human colon cancer. Unfortunately, achievement of this goal has been obscured by 1) the effect of excess residual disease to interfere with the host's destructive response, 2) the weak nature of tumor resistance, 3) the potential adverse effect of concomitant treatments such as chemotherapy, and 4) the present limitation of poorly defined immunogens to induce, as well as insensitive assay systems to detect, host sensitizaion. Recent immunologic and chemical research revealing distinctive surface membrane structures on colon cancer cells suggests that a controlled trial of irradiated, autochtonous cell vaccines (without mycobacterial adjuvants) may provide a new therapeutic tool for Dukes B2 and C stages of human colon cancer.

Characterization of a human colonic adenocarcinoma cell line, LS123.

An epithelial cell line, LS123, was established in 1974 from the second in a series of three primary colonic tumors resected from a Caucasian female. The cell line is aneuploid, releases low concentrations of carcinoembryonic antigen (CEA), fails to grow progressively in nude mice, and forms colonies only in enriched semisolid medium developed for tumor stem cells. LS123 cells grow on confluent cell monolayers and in either low serum or serum-free medium. In the chick embryonic skin assay, LS123 cells grew as a well-differentiated abnormal colonic epithelium with little mitotic activity but with some indication of invasion. On floating collagen gels LS123 cells formed a one to three-cell-layer-thick undifferentiated epithelial sheet. The apparent low invasiveness of the cells of this line is supported by the patient's history of three primary colon tumors without systemic metastases during the past 30 yr. Therefore, although LS123 cells possess several properties associated with neoplasia, they have little invasive potential. Thus, LS123 cells may represent an important model for the study of human colon cancer.

Purification of tumor-specific antigens. An overview of the relevance to human colon carcinoma.

Methods which dissociate intramolecular noncovalent bonds have been used to prepare soluble derivatives of cell-surface antigens. Applications of these techniques to human colon carcinoma are underway. Continuous tissue-culture strains derived from primary lesions were developed and shown to be composed of malignant epithelial elements. Parallel data on the preparation and activity of soluble materials in a murine model methylcholanthrene system reveal that although cultured cells are a satisfactory source for antigen extraction, they are poor targets of the immune response. The development of methods to quantitate the biologic activity of colon-specific, soluble materials may provide indicator systems to define the antigenic determinants, to permit purification, and to serve as assays of the efficacy of immunotherapy.

Human colonic adenocarcinoma cells. I. Establishment and description of a new line.

A series of human colonic epithelial cell lines have been cultured from a single patient: LS-180 the original adenocarcinoma, LS-174T a trypsinized variant, and normal colonic tissue. The malignant cells, 20 to 40, mum in diameter and oval to polygonal, exhibited characteristics of normal colonic mucosal cells, namely, abundant microvilli prominent in secretory cells, and the presence of intracytoplasmic mucin vacuoles. The cultured adenocarcinoma cells, but not normal, demonstrated neoplastic properties by producing high levels of carcinoembryonic antigen (CEA) and by the ability to be propagated in hamster cheek pouches and in immunodeprived mice. The CEA production by the newly established line LS-180 released 900 times more CEA per cell into the culture medium and bore 30 times more cell-associated material than the established line, HT-29. These cell lines may permit detection of distinctive chemical, physiological, pharmacologic, and immunologic characteristics of neoplastic colonic cells.