Characterization of newly established colorectal cancer cell lines: correlation between cytogenetic abnormalities and allelic deletions associated with multistep tumorigenesis

We have established a series of 20 colorectal cancer cell lines and performed cytogenetic and RFLP analyses to show that the recurrent genetic abnormalities of chromosomes 1, 5, 17 and 18 associated with multistep tumorigenesis in colorectal cancer, and frequently detected as recurrent abnormalities in primary tumours, are also retained in long-term established cell lines. Earlier studies by us and other investigators showed that allelic losses of chromosomes 1 and 17 in primary colorectal cancers predicted poorer survival for the patients (P = 0.03). We utilized the cell lines to identify specific chromosomal sites or gene(s) on chromosomes 1 and 17 which confer more aggressive phenotype. Cytogenetic deletions of chromosome 1p were detected in 14 out of the 20 (70%) cell lines, whereas allelic deletions for 1p using polymorphic markers were detected in 13 out of 18 (72%) informative cell lines for at least one polymorphic marker. We have performed Northern blotting, immunohistochemical staining (p53 mRNA, protein) and RFLP analysis using several probes including p53 and nm23. RFLP analysis using a total of seven polymorphic markers located on 17p and 17q arms showed allelic losses aroundthe p53 locus in 16 out of the 20 cell lines (80%), four of which were losses of thep53 locus itself. In addition, seven cell lines (out of nine informative cases) also showed losses of thenm23 gene, four with concurrent losses of thep53 locus, while the remaining three were homozygous. In addition, five out of seven cell lines withnm23 deletions were derived from hepatic metastatic tumours, and one cell line was obtained from recurrent tumour. A comparison between allelic deletions of 1p and functional loss ofnm23 gene revealed a close association between these two events in cell lines derived from hepatic metastasis. Following immunohistochemical staining, nine out of the twenty cell lines showed high levels (25–80%) of mutant p53, four showed intermediate levels (>20%), and seven had undetectable levels of the protein. Of these seven, four showed complete absence of mRNA. Of the remaining three cell lines one showed aberrant mRNA due to germline rearrangement of thep53 gene, whereas in two cell lines normal levels of mRNA were present. Nineteen of the 20 cell lines had normal germline configurations for thep53 gene, while one showed a rearrangement. These data suggest that functional loss ofp53 andnm23 genes accomplished by a variety of mechanisms may be associated with poor prognosis and survival. In addition, concurrent deletions of chromosome regions 17p, 17q and 1p were closely associated with high-stage hepatic metastatic disease. These cell lines with well-characterized genetic alterations and known clinical history provide an invaluable source of material for various biological and clinical studies relating to multistep colorectal tumorigenesis.     

The primary structure of human chromogranin A and pancreastatin

A full-length clone encoding human chromogranin A has been isolated from a lambda gt10 cDNA library of a human pheochromocytoma. The nucleotide sequence reveals that human chromogranin A is a 439-residue protein preceded by an 18-residue signal peptide. Comparison of the protein sequence of human chromogranin A with that of bovine chromogranin A shows high conservation of the NH2-terminal and COOH-terminal domains as well as the potential dibasic cleavage sites, whereas the middle portion shows remarkable sequence variation (36%). This part of human chromogranin A contains a sequence homologous to porcine pancreastatin at residues 250-301. The sequence variation in this part of human chromogranin A compared to porcine pancreastatin is 32% and thus of the same magnitude as that between human and bovine chromogranin A. Therefore, the difference between porcine pancreastatin and the corresponding portions of bovine or human chromogranin A can be explained by species variation, suggesting that pancreastatin is derived from chromogranin A itself rather than a protein that is only similar to chromogranin A. Moreover, the pancreastatin sequence contained in human chromogranin A is flanked by sites for proteolytic processing. Together, these observations suggest that human chromogranin A may be the precursor for a human pancreastatin molecule and possibly for other, as yet unidentified, biologically active peptides.     

Gene response upon illumination in forming mRNA encoding peroxisomal glycollate oxidase

Glycollate oxidase is a constituent of leaf peroxisomes. Its biosynthesis is, like the biosynthesis of many chloroplastic proteins, controlled by light, via phytochrome. The level of mRNA coding for glycollate oxidase was determined at different stages of greening of etiolated plant cells. The appearance of glycollate oxidase mRNA in the cytoplasm was measured by hybridization with cDNA containing part of the coding sequence for glycollate oxidase. cDNA was prepared from enriched mRNA, inserted into the Pst I site of pBR 322, and cloned in Escherichia coli DH-1. By differential colony hybridization and hybrid selection, a clone containing a 670 bp sequence complementary to mRNA encoding glycollate oxidase was selected and identified. Northern blot hybridization was used to investigate mRNA levels induced by light. It was found that continuous light affected the formation of glycollate oxidase mRNA. When a large population of microbodies was present in the cells being induced, the immediate mRNA increase was very pronounced, and was detectable as little as 20 min after the beginning of the light treatment. In contrast, a lag period in the mRNA increase was observed when the induction was performed with etiolated leaves which are characterized by the occurrence of a rather small population of microbodies. For comparison, we measured the time-course of formation of mRNA coding for a light-induced chloroplastic protein, i.e., a protein of the light-harvesting complex. The time-courses of levels of the two mRNAs indicate that the program of gene expression differs between the two particular proteins destined either for chloroplasts or for peroxisomes. The formation of glycollate oxidase mRNA could also be stimulated by a short pulse of light, a treatment of 15 s being a sufficient trigger.     

Mechanism of postsegregational killing by the hok gene product of the parB system of plasmid R1 and its homology with the relF gene product of the E. coli relB operon.

The parB region of plasmid R1 encodes two genes, hok and sok, which are required for the plasmid-stabilizing activity exerted by parB. The hok gene encodes a potent cell-killing factor, and it is regulated by the sok gene product such that cells losing a parB-carrying plasmid during cell division are rapidly killed. Coinciding with death of the host cell, a characteristic change in morphology is observed. Here we show that the killing factor encoded by the hok gene is a membrane-associated polypeptide of 52 amino acids. A gene located in the Escherichia coli relB operon, designated relF, is shown to be homologous to the hok gene. The relF gene codes for a polypeptide of 51 amino acids, which is 40% homologous to the hok gene product. Induced overexpression of the hok and relF gene products results in the same phenomena: loss of cell membrane potential, arrest of respiration, death of the host cell and change in cell morphology. The parB region and the relB genes were cloned into unstably inherited oriC minichromosomes. Whereas the parB region also conferred a high degree of genetic stability to an oriC minichromosome, the relB operon (with relF) did not; therefore the latter does not appear to 'stabilize' its replicon (the chromosome). The function of the relF gene is not known.     

The Ca2+ permeability of sarcoplasmic reticulum vesicles II. Ca2+ efflux in the energized state of the calcium pump

Ca²⁺ efflux from sarcoplasmic reticulum vesicles was studied by measurements of net Ca²⁺ uptake, ⁴⁵Ca²⁺ flux and hydrolysis of energy-rich phosphate. The maximal Ca²⁺ uptake capacity (150–200 nmol/mg protein at pH 6.7, 10 mM MgCl₂ and μ=0.26) was independent of the nature and concentration of the energy-donating substrate (ATP or carbamyl phosphate) and of temperature (15–35°C), suggesting coupling between influx and efflux of Ca²⁺. In the presence of high concentrations of ATP, this efflux of Ca²⁺ was much higher than the passive Ca²⁺ permeation, measured after ATP or Ca²⁺ depletion of the reaction medium. Ca²⁺ efflux was imperceptible at vesicle filling levels below 35–40 nmol Ca²⁺/mg protein, and uncorrelated to the inhibition of the Ca²⁺-ATPase by high intravesicular Ca²⁺ concentrations. Analysis of the data indicated that Ca²⁺ efflux under our conditions probably is associated with one of the Ca²⁺-ATPase partial reactions occurring after dephosphorylation, rather than with a reversal of the Ca²⁺ translocation step in the phosphorylated state of the enzyme. Furthermore, passive Ca²⁺ permeation may be concurrently reduced during the enzymatically active state. It is proposed that both Ca²⁺ efflux and passive Ca²⁺ permeation (Ca²⁺ outflow) proceed via the same channels which are closed (occluded) during part of the Ca²⁺-ATPase reaction cycle.     

The primary structure of human secretogranin II, a widespread tyrosine-sulfated secretory granule protein that exhibits low pH- and calcium-induced aggregation

Secretogranin II (previously also called chromogranin C) is a tyrosine-sulfated secretory protein found in secretory granules in a wide variety of endocrine cells and neurons. Here, we have determined the primary structure of human secretogranin II from a full length cDNA clone and have investigated its properties, predicted from the sequence, by studying the behavior of purified secretogranin II under conditions characteristic of the milieu of secretory granules. Analysis of a 2.35-kilobase cDNA clone isolated from a human pituitary library and identified as secretogranin II by various criteria showed that human presecretogranin II is a 617-residue polypeptide containing an NH2-terminal located signal peptide. Secretogranin II lacks the disulfide-bonded loop structure near the NH2 terminus which is conserved in chromogranin A and chromogranin B (secretogranin I), two other widespread constituents of neuroendocrine secretory granules, but like the latter two proteins contains (i) an -E-N/S-L-X-A/D-X-D/E-X-E-L- motif and (ii) multiple potential dibasic cleavage sites for the generation of smaller, perhaps biologically active peptides. Another structural feature that secretogranin II shares with chromogranin A and chromogranin B (secretogranin I) is the abundance of acidic residues all along the polypeptide chain whose negative charge must somehow be neutralized to allow condensation and packaging of the protein into secretory granules. Experiments with purified secretogranin II showed that in the presence of 10 mM calcium at pH 5.2, conditions characteristic of the milieu of neuroendocrine secretory granules, this protein formed aggregates. Immunoglobulin G, a secretory protein that in vivo is not packaged into secretory granules, did not form aggregates under these in vitro conditions and was excluded from the secretogranin II aggregates. Very little aggregation of secretogranin II was observed in the absence of calcium at pH 5.2 or in the presence of calcium at neutral pH. In vivo, ammonium chloride, which is known to neutralize the pH of acidic intracellular compartments, inhibited the packaging of newly synthesized secretogranin II into secretory granules. Our results suggest that the low pH- and calcium-induced aggregation of secretogranin II may be important for the organization of the secretory granule matrix and raise the possibility that aggregation of secretogranin II may be involved in its sorting to secretory granules.     

Multipoint linkage analysis in Menkes disease.

Linkage analyses were performed in 11 families with X-linked Menkes disease. In each family more than one affected patient had been diagnosed. Forty informative meioses were tested using 11 polymorphic DNA markers. From two-point linkage analyses high lod scores are seen for DXS146 (pTAK-8; maximal lod score 3.16 at recombination fraction [theta] = .0), for DXS1 (p-8; maximal lod score 3.44 at theta = .0), for PGK1 (maximal lod score 2.48 at theta = .0), and for DXS3 (p19-2; maximal lod score 2.90 at theta = .0). This indicates linkage to the pericentromeric region. Multilocus linkage analyses of the same data revealed a peak for the location score between DXS146(pTAK-8) and DXYS1X(pDP34). The most likely location is between DXS159 (cpX289) and DXYS1X(pDP34). Odds for this location relative to the second-best-supported region, between DXS146(pTAK-8) and DXS159 (cpX289), are better than 74:1. Visualization of individual recombinant X chromosomes in two of the Menkes families showed the Menkes locus to be situated between DXS159(cpX289) and DXS94(pXG-12). Combination of the present results with the reported absence of Menkes symptoms in male patients with deletions in Xq21 leads to the conclusion that the Menkes locus is proximal to DXSY1X(pDP34) and located in the region Xq12 to Xq13.3.     

Detection of urinary TNF, IL 1, and IL 2 after local BCG immunotherapy for bladder carcinoma

Intravesical application of bacillus Calmette-Guérin (BCG) is highly active against recurrences of superficial urothelial bladder carcinoma. In an attempt to monitor the immunological effects of this therapy, we analyzed the urine of patients following the sixth intravessical instillation, to show the presence of the monokines TNF and IL 1 and the lymphokine IL 2. Within 24 hours following the instillation, all (n = 10) patients exhibited a strong increase in urinary cytokine secretion, which was significantly different from the control group (n = 10), with respect to TNF L929 biological assay (P less than 0.01), TNF sandwich-ELISA (P less than 0.01), IL 2 CTL 6 biological assay (P less than 0.05), IL 2 sandwich-ELISA (P less than 0.005), and IL 1 sandwich-ELISA (P less than 0.05), but not to the IL 1 fibroblast biological assay. The maximum urinary secretion varied between individual patients and different cytokines, but was generally found within 2 to 8 hr after the instillation. A persistent high urinary activity was demonstrated in BCG-treated patients for IL 2 in sandwich-ELISA. These results reflect the local inflammatory response to BCG and suggest an immunomodulatory mode of action against urothelial carcinoma cells. Elucidation of the possible role of each urinary cytokine against this cancer warrants further investigations.