Guidelines for the use of cell lines in biomedical research.

Cell-line misidentification and contamination with microorganisms, such as mycoplasma, together with instability, both genetic and phenotypic, are among the problems that continue to affect cell culture. Many of these problems are avoidable with the necessary foresight, and these Guidelines have been prepared to provide those new to the field and others engaged in teaching and instruction with the information necessary to increase their awareness of the problems and to enable them to deal with them effectively. The Guidelines cover areas such as development, acquisition, authentication, cryopreservation, transfer of cell lines between laboratories, microbial contamination, characterisation, instability and misidentification. Advice is also given on complying with current legal and ethical requirements when deriving cell lines from human and animal tissues, the selection and maintenance of equipment and how to deal with problems that may arise.

Hypomethylation of WNT5A, CRIP1 and S100P in prostate cancer.

Oligoarray analysis of a matched pair of prostate cancer and normal cell lines derived from the same radical prostatectomy specimen identified 113 candidate hypomethylated genes that were overexpressed in the cancer cells and contained CpG islands. Hypomethylation of wingless-related MMTV integration site 5A (WNT5A), S100 calcium-binding protein P (S100P) and cysteine-rich protein 1(CRIP1) was confirmed in the cancer cells by bisulfite sequencing. Treatment of the corresponding normal prostate epithelial cells 1542-NPTX with the DNA methyltransferase inhibitor 5-Aza-2'-deoxycytidine (5-aza-CdR) induced higher levels of mRNA expression and partial loss of methylation on these genes. Primary prostate cancers were tested using methylation-specific polymerase chain reaction. WNT5A was hypomethylated in 11/17 (65%) tumors, S100P in 8/16 (50%) and CRIP1 in 13/20 (65%). Bisulfite sequencing of a section of the 5' untranslated region (UTR) of WNT5A revealed that three CpG sites (15, 24 and 35) were consistently methylated (93%) in the normal cell line and normal tissues, but not in the prostate cancer cell line and eight primary prostate cancers. Multiple putative binding sites for the transcription factors SP1 and AP-2 were found adjacent to CpG sites 15 and 24. A putative c-Myb binding site was located within the CpG site 35. Anti-c-Myb antibody co-precipitation with WNT5A was methylation-sensitive in 1542-NPTX cells. It is likely that an epigenetic mechanism regulates WNT5A expression in prostate cancer.

Defective repair of cisplatin-induced DNA damage caused by reduced XPA protein in testicular germ cell tumours.

Metastatic cancer in adults usually has a fatal outcome. In contrast, advanced testicular germ cell tumours are cured in over 80% of patients using cisplatin-based combination chemotherapy [1]. An understanding of why these cells are sensitive to chemotherapeutic drugs is likely to have implications for the treatment of other types of cancer. Earlier measurements indicate that testis tumour cells are hypersensitive to cisplatin and have a low capacity to remove cisplatin-induced DNA damage from the genome [2] [3]. We have investigated the nucleotide excision repair (NER) capacity of extracts from the well-defined 833K and GCT27 human testis tumour cell lines. Both had a reduced ability to carry out the incision steps of NER in comparison with extracts from known repair-proficient cells. Immunoblotting revealed that the testis tumour cells had normal amounts of most NER proteins, but low levels of the xeroderma pigmentosum group A protein (XPA) and the ERCC1-XPF endonuclease complex. Addition of XPA specifically conferred full NER capacity on the testis tumour extracts. These results show that a low XPA level in the testis tumour cell lines is sufficient to explain their poor ability to remove cisplatin adducts from DNA and might be a major reason for the high cisplatin sensitivity of testis tumours. Targeted inhibition of XPA could sensitise other types of cells and tumours to cisplatin and broaden the usefulness of this chemotherapeutic agent.

Allelic imbalance and biochemical outcome after radical prostatectomy.

OBJECTIVE: To compare the incidence of allelic imbalance (AI) in men with rapid disease progression with those who remained disease free after radical prostatectomy, with the aim of identifying genetic markers to predict prognosis and guide further treatment. PATIENTS AND METHODS: Tumour and normal DNA were extracted from two matched groups of 31 men with extracapsular node-negative (pT3N0) prostate cancer who had undergone radical prostatectomy. One group comprised men who developed biochemical recurrence within 2 years of surgery and one group were prostate-specific antigen (PSA) free for at least 3 years. Men were matched for Gleason grade, preoperative PSA and pathological stage. Analysis was performed by genotyping. RESULTS: Allelic imbalance was analysed using 30 markers, and was seen in at least one marker in 57 (92%) of the cases. Deletion at marker D10S211 (10p12.1) was significantly more common in the relapse group than the non-relapse group (35 vs 5%, P=0.03). CONCLUSIONS: This study demonstrates significant association between AI on chromosome 10 and biochemical progression after radical prostatectomy.

Differential sensitivities of human testicular and bladder tumor cell lines to chemotherapeutic drugs.

The in vitro drug sensitivities of 5 human testicular tumor cell lines (Tera II, SuSa, NEC-8, 833K, T3B1) and 5 human bladder carcinoma cell lines (RT4, RT112, T24, HT1197, HT1376) were compared. Cytotoxicities of cisplatin and doxorubicin were assessed by inhibition of colony-forming ability during continuous exposure to a range of drug concentrations. The ranges of the drug concentrations required to kill 70% of clonogenic cells obtained against the testicular cell lines were 1-7 ng/ml and 21-161 ng/ml for doxorubicin and cisplatin, respectively, compared with 4-19 ng/ml and 112-431 ng/ml for the bladder cell lines. This study shows that continuous cell lines retain the relative clinical chemosensitivities of their tumors of origin. The results also indicate that testicular tumor cells are inherently more sensitive to the cytotoxic effects of chemotherapeutic drugs than are bladder cancer cells.

Cyclic Variation of DNA synthesis in human breast epithelium.

DNA synthesis in normal breast epithelium from premenopausal women was assessed by use of autoradiography. In parous women the labeling indexes decreased during the follicular phase of the menstrual cycle and increased to a significantly higher level during the luteal phase.

Hypersensitivity of human testicular tumors to etoposide-induced apoptosis is associated with functional p53 and a high Bax:Bcl-2 ratio.

Metastatic testicular cancers are curable, whereas bladder cancers and most other solid tumors are not. Cell lines derived from human testicular (GH, GCT27, and 833K) and bladder (RT4, RT112, and HT1376) tumors retain this differential chemosensitivity in vitro. We have investigated the hypothesis that differential sensitivity to chemotherapy is related to differences in the threshold of susceptibility to undergoing apoptosis. Sensitivity to etoposide was not directly related to the frequency of DNA strand breaks. DNA damage was on average 2-fold greater in the testicular than the bladder tumor cell lines; in contrast, the testicular tumor lines were 15-fold more sensitive to etoposide cytotoxicity than the bladder tumor lines (IC90 values of 19 +/- 6 versus 293 +/- 180 microM, respectively). Using equidamaging (550 rad equivalents) etoposide treatments, the percentage of cells that underwent drug-induced apoptosis was on average higher in the testicular tumor cell lines than the bladder tumor cell lines. The testicular tumor lines have two characteristics that could confer sensitivity to drug-induced apoptosis. First, they have functional p53: the product of the p53-dependent gene waf-1 was increased after etoposide treatment. Second, the testicular tumor lines expressed relatively high levels of the apoptosis-promoting protein Bax, but there was no expression of the suppressor of apoptosis Bcl-2. In contrast, only one of the three bladder cell lines (RT4) had functional p53, and all of the bladder lines had readily detectable levels of Bcl-2 and low levels of Bax. In the testicular cell lines, increases in p53 and p53-transactivated genes were associated with apoptosis but not arrest in G1. In contrast, in the bladder cell line (RT4), increases in p53 and Waf-1 were associated with both arrest in G1 and apoptosis. The differences in the ratio of Bax:Bcl-2 could contribute to the differential sensitivity of the two tumor types. However, in contrast to earlier reports, the ratio of Bax and Bel-2 was not perturbed by DNA damage.

Inherent sensitivity and induced resistance to chemotherapeutic drugs and irradiation in human cancer cell lines: relationship to mutation frequencies.

Metastatic nonseminomatous testicular germ cell tumors are curable using combination chemotherapy in approximately 80% of patients. In contrast, most other patients with other types of cancer either present with or acquire drug-resistant disease following chemotherapy. Cell lines derived from testis tumors retain hypersensitivity to both drugs and radiation in vitro, thus providing a model system with which to investigate the genetic basis of hypersensitivity to these agents. This study compared the spontaneous and both ethyl methanesulfonate- and cisplatin-induced frequencies of mutation of 6-thioguanine resistance in 3 human bladder and 3 testis tumor cell lines and a bladder and a testis cell line with cisplatin resistance induced in vitro. The two tumor types showed similar frequencies of both spontaneous and induced mutation frequencies at this locus. Therefore, we failed to provide evidence for the hypothesis that the curability of testis tumors is associated with a low frequency of mutation to drug resistance.

Sensitivities of monolayers and spheroids of the human bladder cancer cell line MGH-U1 to the drugs used for intravesical chemotherapy.

The in vitro cytotoxicities of the four drugs most frequently used for intravesical chemotherapy (Adriamycin, epodyl, mitomycin C, Thiotepa) and epirubicin were compared using monolayers and multicellular tumor spheroids of the human bladder cancer cell line, MGH-U1. Adriamycin and epirubicin were most cytotoxic against monolayer cultures, whereas mitomycin C killed more cells in spheroids. Epodyl was least cytotoxic against both two- and three-dimensional cultures. Thiotepa was the only drug more cytotoxic to three- than two-dimensional cultures. Topographic analysis of bromodeoxyuridine-stained nuclei using image analysis indicated that Adriamycin selectively removed or killed superficial cells in multicellular tumor spheroids, but had little effect on DNA synthesis within the spheroids. In contrast Thiotepa killed cells throughout the spheroids. These in vitro data appear to reflect clinical experience using intravesical chemotherapy to treat superficial bladder cancer.

Relationship between topoisomerase II level and chemosensitivity in human tumor cell lines.

Patients with metastatic testis tumors are generally curable using chemotherapy, whereas those with disseminated bladder carcinomas are not. We have compared levels of the nuclear enzyme topoisomerase II in three testis (SuSa, 833K, and GH) and three bladder (RT4, RT112, and HT1376) cancer cell lines which differ in their sensitivity to chemotherapeutic agents. The testis cell lines were more sensitive than the bladder lines to three drugs whose cytotoxicity is mediated in part by inhibiting topoisomerase II: amsacrine; Adriamycin; and etoposide (VP16). The frequency of DNA strand breaks induced by amsacrine was higher (1.5- to 13-fold) in the testis cells than in the bladder cells. The level of topoisomerase II-mediated DNA strand breakage in vitro, measured by filter trapping of amsacrine-induced protein:DNA cross-links, was similarly higher in nuclear extracts from the testis than the bladder cells. Western blot analysis showed a generally higher level of topoisomerase II protein in testis than in bladder cell nuclear extracts. Topoisomerase II protein expression broadly correlated with drug-induced strand breakage in both protein extracts and whole cells, but not with population doubling time. However, despite a 2- to 20-fold increased sensitivity to the different topoisomerase II inhibitors, the testis line 833K had a less than 2-fold higher level of topoisomerase II protein than that of the bladder line RT4. These results indicate that the level of expression of topoisomerase II is an important determinant of the relative chemosensitivity of testis and bladder tumor cell lines, but that additional factors must contribute to the extreme chemosensitivity of testis cells.

Differential repair of platinum-DNA adducts in human bladder and testicular tumor continuous cell lines.

The formation and removal of four platinum-DNA adducts were immunochemically quantitated in cultured cells derived from a human bladder carcinoma cell line (RT112) and from two lines derived from germ cell tumors of the testis (833K and SUSA), following exposure in vitro to 16.7 microM (5 micrograms/ml) cisplatin. RT112 cells were least sensitive to the drug and were proficient in the repair of all four adducts, whereas SUSA cells, which were 5-fold more sensitive, were deficient in the repair of DNA-DNA intrastrand cross-links in the sequences pApG and pGpG. Despite expressing a similar sensitivity to SUSA cells, 833K cells were proficient in the repair of all four adducts, although less so than the RT112 bladder tumor cells. In addition, SUSA cells were unable to repair DNA-DNA interstrand cross-links whereas 50-85% of these lesions were removed in RT112 and 833K cells 24 h following drug exposure. It is possible that the inability of SuSa cells to repair platinated DNA may account for their hypersensitivity to cisplatin.

Tissue culture model of transitional cell carcinoma: characterization of twenty-two human urothelial cell lines.

Twenty-two continuous cell lines derived from normal and neoplastic urothelium, maintained under identical culture conditions, were characterized in terms of isozyme phenotype, tumorigenicity, and xenograft morphology following xenotransplantation to nude mice, cytological appearance, in vitro growth rate, labelling index, and colony-forming efficiency, in parallel with separate studies of in vitro drug sensitivities and monoclonal antibody reactivities. Three groups were identified: (a) distinct lines with differing isozyme patterns, a broad spectrum of growth characteristics, and xenograft morphologies similar to the histopathology of the parent tumors after periods of up to 17 yr following establishment in vitro; (b) cross-contaminated sublines (maintained separately in different laboratories for periods of up to 10 yr), with identical isozyme patterns and similar growth characteristics, but differing markedly in tumorigenicity and xenograft morphology; and (c) lines derived from normal urothelium which were nontumorigenic and had an isozyme pattern usually only encountered in untransformed cells. These data indicate that cell lines representative of human transitional cell carcinomas can be selected on the basis of xenograft morphology and isozyme patterns, and that a panel of lines derived from normal and neoplastic urothelium could provide a model system to study the biology and treatment of this disease.

Plexin-B1 signalling promotes androgen receptor translocation to the nucleus.

Semaphorins and their receptors plexins have diverse roles in many cancers affecting tumour growth, metastasis and angiogenesis. Plexin-B1, the receptor for semaphorin4D (Sema4D), has been implicated in prostate cancer where mutation of the gene and overexpression of the protein occur. It is not clear, however, as to which of the several Sema4D-activated signalling pathways downstream of plexin-B1 function in prostate cancer progression. We show here that Sema4D/plexin-B1 increases the expression of androgen-responsive genes and activates the transcriptional activity of the androgen receptor (AR). Activation of plexin-B1 results in phosphorylation of AR at Serine 81, a site that is phosphorylated by nuclear kinases. Cell fractionation and immunocytochemistry studies demonstrated that the proportion of cells with AR in the nucleus increases significantly upon Sema4D treatment. The N-terminal (AF-1) domain of AR, which contains binding sites for transcription regulators, is not required for this response. Depletion of AR suppressed Sema4D-induced anchorage-independent growth of LNCaP and LNCaP-LN3 cells, demonstrating the functional significance of these findings. These results show that Sema4D/plexin-B1 signalling promotes the translocation of AR to the nucleus and thereby enhances AR transcriptional activity. Plexin-B1 is therefore a promising target for cancer therapy, especially in low androgen situations such as those imposed by androgen deprivation therapy.

Genome-wide screening for genetic changes in a matched pair of benign and prostate cancer cell lines using array CGH.

Copy number alterations in a matched pair of benign epithelial and prostate cancer cell lines derived from the same patient were assessed using array-based comparative genomic hybridisation (aCGH). The cancer cell line showed a gain of chromosome 7, deletion of chromosome 8, gains (including high level) and losses on chromosome 11, loss of 18p and gain of 20q. Deletions on chromosome 8 were confirmed with microsatellite markers. The aCGH results were compared to gene expression data obtained using DNA microarrays and suggested the involvement of caspases and ICEBERG on 11q and E2F1 on chromosome 20q.

5 alpha-reductase expression by prostate cancer cell lines and benign prostatic hyperplasia in vitro.

5 alpha-reductase (5 alpha R) activity in two human prostate cancer cell lines was compared to that in benign prostatic hyperplasia (BPH) tissue and COS cells transfected with and expressing the human genes for 5 alpha-reductase type 1 (5 alpha R1) and type 2 (5 alpha R2). Comparisons were based on pH profiles and sensitivities to selective inhibitors of 5 alpha-reductase In the cancer lines, activity was greatest over the pH range 7-8, compared to a sharp peak of activity between pH 5-5.5 in BPH tissue and COS cells expressing 5 alpha R2. Finasteride and SKF105,657 were potent inhibitors of 5 alpha-reductase activity in BPH tissue and COS cells expressing 5 alpha R2, but weak inhibitors in the cancer lines and in COS cells expressing 5 alpha R1. In contrast, UK117,026 was a more potent inhibitor of 5 alpha-reductase activity in the prostate cancer cell lines and in COS cells expressing 5 alpha R1. These data indicate that human prostate cancer cell lines express 5 alpha-reductase activity similar to that in COS cells transfected with 5 alpha R1, but different from that in BPH tissue. This may be a consequence of in vitro culture. Alternatively, it may reflect a change occurring as a result of neoplastic transformation in which case it will be important to select appropriate inhibitors in the clinic.

Screening of new anticancer agents in vitro using panels of human cell lines derived from non-seminomatous germ cell tumours and transitional cell carcinomas of the bladder.

Metastatic testis tumours are cured in over 80% of patients using combination chemotherapy, and this hypersensitivity is retained by the cells in vitro. To determine whether differential toxicity to testis tumour cells is useful in the screening of novel anticancer agents, we compared the toxicities of 12 compounds against panels of human bladder and testis tumour cell lines using a clonogenic assay. The compounds had screened negative against P388 in vivo, and had been retested using the human tumour colony forming assay (HTCFA) and in selected cases against human tumour xenografts. NSC 339004, chloroquinoxaline sulphonamide, was 7-fold more toxic to testis tumour than bladder cancer cells, comparing the mean of the concentrations reducing colony-forming ability by 70%. This was the only one of the compounds selected by the HTCFA shown to have clinical activity. Compound R was selectively toxic to the bladder cancer cells, and might be of value as an intravesical agent. These data indicate that panels of testis and bladder cancer cell lines might be a useful addition to the disease-oriented screening programme.

Development and characterization of cisplatin-resistant human testicular and bladder tumour cell lines.

Cisplatin-resistant cells were derived in vitro from a human bladder carcinoma cell line (RT112) and a testicular tumour cell line (SuSa) by continuous exposure to increasing concentrations of cisplatin for 14 and 11 months, respectively. Both resistant cell lines had a four-fold level of resistance relative to their parental cell lines, comparing the cisplatin concentration to inhibit colony forming ability by 70%. These levels of resistance were retained in the absence of cisplatin for at least 3 months. In each case, four-fold fewer micronuclei were produced in the resistant lines by the same cisplatin concentrations. Cross-resistance to carboplatin and methotrexate was observed in both resistant cell lines, but neither line was resistant to doxorubicin. Isozyme and DNA analysis with hypervariable probes confirmed the origin of each resistant cell line from its parental line. Population doubling times and intermitotic times were similar in each of the pairs of cell lines. Karyotyping showed that the resistant cell lines had gained and lost marker chromosomes, but there were no changes common to both resistant cell lines.

Fatty acid composition of normal and malignant cells and cytotoxicity of stearic, oleic and sterculic acids in vitro.

The aim of this study was to investigate the hypothesis that saturated fatty acids are differentially cytotoxic to cancer cells. Three studies were undertaken to: (1) measure the toxicities of stearic and oleic acids to normal and malignant cells in vitro, (2) assess if there is any relationship between toxicity and relative fatty acid composition and (3) determine whether the relative fatty acid composition of a cancer cell line could be modified by sterculic acid, an inhibitor of delta-9-desaturase. Stearic (18:0) and oleic (18:1) acids inhibited the colony-forming abilities of five human cancer cell lines and two non-neoplastic cell lines in a dose-dependent fashion. The concentration of oleic acid required to reduce colony formation ability by 50% was 2.5-6.0-fold greater than that of stearic acid. Addition of sterculic acid to a cancer cell line resulted in steady-state levels of stearic acid and increasing percentage of oleic acid.

Sensitivity of testis tumour cells to chemotherapeutic drugs: role of detoxifying pathways.

In contrast to most other types of cancer, metastatic testicular germ cell tumours (TGCT) are cured in most patients using cisplatin-based combination chemotherapy. The biochemical mechanisms underlying this sensitivity have not been defined. Drug detoxification can modulate response to chemotherapy in vivo and in vitro, and therefore we measured levels of glutathione (GSH), glutathione-S-transferase (GST) and both constitutive and cisplatin- and dexamethasone-induced levels of metallothionein (MT) in five human testis tumour cell lines. The levels were compared with those in five human bladder cancer cell lines and two cell lines with cisplatin resistance acquired in vitro. GSH levels were relatively low in the testis tumour cell lines, as might be expected in drug-sensitive cells, and there was a 77-fold increase in GSH levels in the cisplatin-resistant testis tumour cell line. GST levels were similar in the two cell types, while metallothionein levels were relatively high in the testis tumour cell lines. These data indicate that GSH may contribute to the sensitivity of TGCT to chemotherapy, and that GSH expression may be involved in the acquisition of cisplatin resistance in these tumours.

The differential induction of collateral resistance to 62.5 MeV (p-->Be+) neutrons and 4 MeV photons by exposure to cis-platinum.

PURPOSE: To determine the relative sensitivity to cis-platinum, 4 MeV photons and 62.5 MeV (p-->Be+) neutrons in five human tumor cell lines, and their cis-platinum resistant variants. METHODS AND MATERIALS: The degree of cross-resistance of five human in-vitro cell lines to photons or fast neutrons was analysed for both cisplatinum-sensitive and resistant variants. RESULTS: The development of acquired cis-platinum resistance conferred collateral resistance to 62.5 MeV (p--Be+) neutrons in all five cell lines, but did not consistently decrease the photon sensitivity of these same cells. CONCLUSION: The reduction in photon and neutron sensitivity following the development of acquired cis-platinum resistance may possibly be regulated by different mechanisms. The reduction in neutron sensitivity was primarily due to a 1.3-1.7 fold reduction in the magnitude of the initial slope (alpha), which was independent of the degree of platinum resistance induced, suggesting a non-stochiometric relationship between the mechanisms responsible for acquired cis-platinum, and 62.5 MeV (p-->Be+) neutron resistance.