S2
Si INOSITOL LIPIDS AND CELL PROLIFERATION M.J. Berridge, Dept of Zoology, University of Cambridge, Downing Street, Cambridge CB2 3EJ. Many of the stimuli responsible for controlling cell proliferation are known to induce changes in phosphoinositide metabolism. There are two main effects. Firstly, growth factors stimulate the hydrolysis of phosphatidylinositol 4,5-bisphosphate to generate two second messenger pathways. Inositol 1,4,5-trisphosphate (Ins(1,4,5)P3) is released to the cytosol to mobilize internal carlcium and to promote an influx of external calcium whereas the diacylglycerol which remains within the plane of the membrane activates protein kinase C. The calcium signalling pathway displays complex spatial and temporal patterns which may be particularly re1evant to cell proliferation since they have been observed in eg s undergoing fertilization and in lymphocytes and fibroblasts stimulated with growth factors. This bifurcating signalling pathway is activated by growth factors operating either through G proteins or through tyrosine phosphorylation. In the latter case, receptors for growth factors such as PDGF and EGF activate PLC- by tyrosine phosphorylation which means that the generation of calcium siqnals and the activation of protein kinase C are almost always associated with the onset of cell proliferation. What remains a subject of debate is whether or not these signals have any direct role to play in stimulating DNA synthesis. The second major effect of growth factors on metabolism is to stimulate an inositol yhosphoinositide ipid 3-kinase which leads to the formation of a novel class of inositol lipids carrying a phosphate on the 3-position. Together with PLC Ra 1 and GAP this 3-kinase is activated by tyrosine phosphorylation. Since the appearance of these novel inositol lipids is also apparent following transformation by certain oncogenes such as src, it seems they may play an important role in cell proitTeration perhaps related to changes in cell morphology. While it is clear that the phosphoinositides undergo profound changes when cells are induced to grow, there is still considerable uncertainty as to their precise role in controlling events during the cell cycle.
Inosltol phospholiplds as precursors of lntracellular
second messengers. C.M. Macpheel, A.N. Carter2, F. Ruiz-Larrea3 and C. Peter Downes2, ICellular Pharmacology, Smith Kline Beecham, The Frythe, Welwyn, Herts. AL6 9AR,2Department of Biochemistry. University of Dundee, Dundee DDI 4HN, 3Ludwig Institute for Cancer Research, Coutauld Building, London WIP 8BT. The inositol-containing phospholipids have a critical role in coordinating cellular activity because they furnish cells with a number of intracellular signal molecules in response to a wide variety of hormones, neurotransmitters and growth factors. It is now generally accepted that, upon agonist activation, many cell surface receptors stimulate a phospholipase C which cleaves phosphatidylinositol (4,5)-bisphosphate to give diacylglycerol, which remains associated with the membrane, and inositol (1,4,5)-trisphosphate (InsP3). which diffuses into the cytosol. Diacylglycerol is the second messenger which activates protein kinase C and InsP3 stimulates the release of Ca2+ from intracellular storage sites. These second messengers control responses as widespread as fertilisation, early embryonic development, smooth muscle contraction, cell proliferation, secretion and neuronal excitability. Recently, a novel group of inositol phospholipids has been identified in which an additional monoester phosphate moiety is inserted in the D-3 position of the inositol ring by phosphatidylinositol 3-kinase (Ptdlns 3-kinase). Several lines of investigation have implicated one of the products of this novel pathway, phosphatidylinositol (3,4,5)-trisphosphate, as a second messenger in mitogenic and inflammatory cell signalling. Inositol phospholipid kinases, therefore, occupy pivotal positions in second messenger pathways by directing the synthesis of phospholipid isomers with apparently distinctive signalling functions. We have investigated the substrate specificity of 3 types of phosphatidylinositol kinase using a variety of ATP site ligands and synthetic enantiomers of phoshatidylinositol. Both types of compound reveal substantial differences in the substrate recognition sites of Ptdlns 3-kinase and Ptdlns 4-kinases, respectively, which could be exploited in the developement of specific inhibitors of these enzyms:s.
S4
PROTEIN KINASE C, P.J. Parker, S.Cazaubon, N. Goode, A. Olivier, C. Pears, J. Ridden. Protein Phosphorylation Laboratory, Imperial Cancer Research Fund, 44 Lincoln's Inn Fields, London WC2A 3PX.
S3
ONCOGENES, GROWTH FACTORS AND PHOSPHATIDYLINOSITOL 3-KINASE L. Cantley, Dept. of Physiology, Tufts University School of Medicine, Boston, MA 02111 USA.
Phosphatidylinositol 3-kinase (Ptdlns 3-kinase) was discovered because of its specific association with oncogenic protein-tyrosine kinases (Whitman et al., 1985 Nature 315, 239; Whitman et al., 1988 Nature 332, 644). This enzyme phosphorylates the D-3 position of the inositol ring of either phosphatidylinositol (Ptdlns), Ptdlns-4-P or Ptdlns-4,5-P2 to generate the novel lipids PtdIns-3-P, Ptdlns-3,4-P2 and PtdIns-3,4,5-P3 respectively. These lipids are not substrates for Ptdlns-specific phospholipases C and they are not in the canonical Ptdlns turnover pathway. The correlation between the transforming ability of mutant proteintyrosine kinases and their association with Ptdlns 3-kinase argues that this enzyme is in a critical pathway for regulation of cell growth and transformation (Cantley et al., 1991 Cell 64, 281). Notably, Ptdlns-3,4-P2 and Ptdlns-3,4,5-P3 are absent in quiescent cells but appear within minutes of addition of growth factors or upon transformation by oncogenes of the protein-tyrosine kinase family. The Ptdlns 3-kinase is a heterodimer of 85 kd and 110 kd subunits (Carpenter et al., 1990 J. Biol. Chem. 265, 19704). The 85 kd subunit is phosphorylated on tyrosine in vivo in polyoma middle t-transformed cells or platelet-derived growth factorstimulated cells (Kaplan et al., 1987 Cell 50, 1021) and can be phosphorylated in vitro by baculovirus-expressed polyoma middle t/pp6Oc-src complex. The gene for the 85 kd subunit has been cloned by several laboratories and the predicted protein has regions of homology to the non-catalytic domain of pp6Oc-src (SH-2 and SH-3 domains). The involvement of the SH-2 domain in association with tyrosinephosphorylated proteins will be discussed. The potential role of the lipid produlcts of the Ptdlns 3-kinase in cell signaling will also be discussed.
Protein kinase C is made up of a family of structurally related
polypeptides. The consistent feature of this family, is the
presence of a kinase domain (defincd both functionally and through sequence alignment) and a cysteine rich region conferring phorbol ester (and by inference diacylglycerol) binding. Where studied it is evident that the binding of phorbol esters/diacylglycerol in a suitable phospholipid context leads to activation of the kinase. Thus these proteins have been considered to operate immediately downstream of agonist evoked diacylglycerol production following activation of phosphatidylinositol 4,5 bisphosphate directed phospholipase C. The coincident production of inositol 1,4,5 trisphosphate and consequent Ca2+ mobilisation will synergise with diacylglycerol in the activation of the Ca2+-dependent forms of protein kinase C. By contrast the more recently described Ca2 + -independent protein kinase Cs have the potential to respond to diacylglycerol produced both in the presence or absence of a rise in cytoplasmic Ca2+; the latter being exemplified by the breakdown of phosphatidylcholine. This suggests that the patterns of expression of these two branches of the PKC family (Ca2 +-dependent and -independent) permit cells to respond to different effector systems. The manner of the response is governed by the substrate specificity of the individual PKCs; there is ample evidence that there are significant differences in specificity. Thus this growing family of signal transducing proteins appear to operate to diversify the circumstances in which the pathway is triggered and the nature of the output.
3
S6
GROWTH FACTOR RECEPTORS IN CELL SIGNALLING, W.J. Gullick, ICRF Oncology Group, Hammersmith Hospital, Du Cane Road, London W12 OHS Since the discovery of the epidermal growth factor receptor, a large number of both structurally and functionally related molecules have been identified which are either known to be or likely to be involved in transmitting mitogenic stimuli across cell membranes. This family of growth factor receptor molecules is defined by possessing an extracellular domain responsible for ligand binding and an intracellular domain with protein tyrosine kinase activity. In all cases so far examined, ligand binding promotes tyrosine kinase activity. In most cases this leads to stimulation of mitogenesis although there are some ligands such as amphiregulin which may also have inhibitory properties on some cell types. Several tyrosine kinase receptors can be activated by mutation or overexpression to behave as dominant in oncogenes vitro and in transgenic animals. Examination of human tumour biopsies has identified alterations in the c-erbB-1, c-erbB-2, trk, met and Ksam genes. The c-erbB-1 (EGFR) gene is amplified or overexpressed in many squamous carcinomas and the cerbB-2 gene in several types of adenocarcinoma. Overexpression of either gene has in some cases been shown to indicate a short relapse free interval and survival from the disease. Such cell surface molecules provide an attractive target for new forms of therapies. Our own work has focused on designing compounds which prevent receptor dimerisation, an event necessary for receptor activation. Other strategies for the production of GFR inhibitors include ligand antagonists and antisense oligonucleotides. If such drugs can be produced they have potential for treatment of many common solid tumours.
REGULATION OF THE ACTIVITY OF p2lras IN T-LYMPHOCYTES. J.Downward, Imperial Cancer Research Fund, Lincoln's Inn Fields, London WC2A 3PX.
S5
The ras proteins bind and hydrolyse GTP: p21.GTP is the active from of the molecule while p21.GDP is inactive. The physiological signals responsible for controlling the activation state of p2lras have long been sought. We have recently demonstrated that in intact T lymphocytes the activation state of p2lraS, as determined by the amount of GTP bound to it relative to GDP, is under control of the T-cell antigen receptor. Protein kinase C mediates this stimulation of p2lraS which appears to occur through the suppression of GAP activity. An inhibitor of GAP activity is stimulated in Tcells upon treatment with protein kinase C activators. A permeabilised cell system has been used to study this activation of p2lras further: this shows that the rate of guanine nucleotide exchange on p21 in both lymphoblasts and fibroblasts is greatly stimulated relative to on pure p21 in vitro. The activation state of p21raS in both cell types appears to be controlled by a dynamic balance between two highly active and opposing proteins, GAP/NF-1 and a guanine nucleotide exchange factor.
S8
S7
SIGNAL TRANSDUCTION MECHANISMS IN ANTICANCER DRUG ACTION AND RESISTANCE. Thomas R. Tritton, Department of Pharmacology, University of Vermont Medical School, Burlington, VT 05405 USA. One of the hallmarks of cancer is loss of cellular growth regulation. Now that it is established that many oncogenes code for proteins involved in signalling and growth control, there has developed an increased recognition that an understanding of these basic processes offers new prospects for controlling neoplasia. In this presentation we will discuss two ideas stemming from our laboratory's work: (1) the ability of cytotoxic anticancer drugs like adriamycin to kill susceptible cells appears to involve a series of events initiated at the plasma membrane and proceeding through the protein kinase C signal transduction pathway to ultimate damage to DNA in the nucleus; and (2) resistance to antineoplastics, particularly multidrug resistance, may be influenced both by modulations of protein kinase C biology, and by the action of oncogenes like c-fos and c-jun, which bind to the promoter region of genes involved in drug resistance and regulate their transcription.
CELL MEMBRANE/SIGNAL TARGETS FOR ANTICANCER CHEMOTlERAPY
D Crowther, CRC Dept Medical Oncology, Christie Hospital, Manchester, M20 9BX, U.K.
Our understanding of the way in which cells interact with their environment is improving and its complexity is already appreciated. The mechanisms whereby the cell surface interacts with other cell membranes, the cell matrix, growth factors and hormones, and the consequent cascade of biochemical events leading to changes in cell growth and function suggest a substantial number of potential targets for systemic anticancer therapy. The fact that several elements in this system have been identified as proto oncogene products suggest that mutant products could be targetted with a degree of specificity for the tumour cell compared with its normal counterpart. Already clinical studies investigating the use of growth factors, their analogues, receptors or antibodies directed against them have been started. Other approaches involve the use ofagents designed to influence cell signal mechanisms. These include modulators of protein kinase C (Bryostatin- 1, Staurosporins) and the ether lipids such as SRI 62-834 and hexadecylcholine phosphate. Some are currently in Phase I trials under the auspices of the CRC. Families of enzymes such as the phospholipases, tyrosine kinases and phosphates and the GTPase superfamily are also targets against which new agents are being developed. Other agents being developed have structural similarities to heparan sulphate (HS) with the capacity to interfere with growth factors/receptor interaction and currently one of these (Suramin) is in clinical phase I/l trials in the USA and Europe. A better understanding of HS chemistry and an appreciation ofthe diversity of molecular structure and potential specificity for mediating biological events could well lead to the development of an important new group of agents with anticancer activity. Some agents active on the cell membrane or on the cell signalling cascade 'now being developed are highly active in biological systems and starting doses in clinical trial may be at the 5jg/ml level with expected serum levels below the detection limit for current assays. These and drug scheduling issues pose special problems for clinical study.
4
MOLECULAR EPIDEMIOLOGY OF FAMILIAL S12 CANCERS, F. Li, Clinical Epidemiology Branch, National Cancer Institute and Dana-Farber Cancer Institute, Boston, Mass., 02115 U.S.A.
Sil
SMALL CELL LUNG CANCER (SCLC) AND NEUROBLAS1 ~~~~~~~~2 TOMA (NB), J. Pritchard' and R.L. Souhami 1Dept. of Haematology & Oncology, Institute of Child Health, London WC1, 2Dept. of Oncology, University College & Middlesex School of Medicine, London Wl.
Clinical studies, epidemiological research and laboratory investigations have contributed to understanding of human cancer biology. My colleagues and I have attempted to integrate these 3 approaches to facilitate accrual of new knowledge. We study cancerof families at exceptionally high risk as human models
Both SCLC and NB express markers of neuroendocrine (NE) differentiation, such as neural cell adhesion molecule, neurosecreting granules and NSE. Poor prognosis in NB appears to be associated with amplification of n-myc which is also amplified in SCLC lines although the relationship to prognosis is less clear. Chromosome deletion at lp and 17p are features of NB and 3p and 17p in SCLC. Both tumours may undergo a change in morphology after chemotherapy perhaps reflecting differentiation, but this is less marked in SCLC and is more difficult to achieve in tissue culture. The metastatic patterns of both tumours are very similar with frequent spread to local nodes, liver and bone marrow. NB and SCLC are sensitive to chemotherapy, the same drugs being active in both diseases. Although remission rates are high, relapse occurs in the majority of patients with NB who are over 1 year of age and who have stage 3 and 4 disease. In SCLC cure is extremely uncommon (approximately 3%). In both tumours control of systemic disease is the main problem. In both NB and SCLC it appears that short intensive remission induction procedures are probably superior in producing remission. This is now being tested formally in both diseases. In NB there is evidence that 'consolidation' with high dose chemotherapy prolongs remission. This remains a possibility, and may improve survival, though less prectical, in SCLC. The definition of non-cross resistance and the design of high dose strategies presents similar problems in both diseases. There are reasonable grounds for believing that chemotherapy lessons learned from the treatment of one disease may have implications for the other. This is also true for other approaches to therapy such as modulation of autocrine growth and antibody-guided therapy
cancer susceptibility. These patients represent rare "experiments of nature" that can reveal new insights into carcinogenic processes. The rationale for investigating rare family cancers rests in Knudson's
hypothesis, which is supported by experimental data on retinoblastoma. This model indicates that the multiple
steps in carcinogenesis can be dissected into 2 components in cancer families for ease of study. Specially, germ line of familial cases can reveal the first mutation, and subsequent mutations are detected by comparing tumor DNA with the corresponding germ line DNA. Our interdisciplinary research employs clinical observations to find susceptible patients, epidemiological studies to quantitate the excess risk, and laboratory investigations to examine the biological basis of susceptibility. The studies have uncovered several new patterns of familial cancers. This presentation describes our investigations of three disorders: 1) the syndrome of sarcomas, breast cancer, and other neoplasms; 2) co-inheritance of renal cell carcinoma and a constitutional chromosome translocation in a kindred; and 3) familial Wilms' tumor that indicates genetic heterogeneity of the neoplasm.
FATHER'S EXPOSURE TO RADIATION AT SELLAFIELD AND CHILDHOOD CkNCER, Martin J. Gardner, Medical Research Council Environmental Epidemiology Unit, (University of Southampton), Southampton General Hospital, Southampton S09 4XY. The original suggestion that the level of childhood leukaemia was raised near the Sellafield nuclear site in Cumbria, particularly in the nearby village of
S14
ENDOCRINE SEQUELAE FOLLOWING THE TREATMENT OF CHILDHOOD CANCER, S. M. Shalet, Dept of Endocrinology, Christie Hospital & Manchester H20 8BX. Holt Radium Institute,
S 15
hypothalamic, impair directly may Radiation pituitary, thyroid and gonadal function or alternatively it may induce the development of hyperparathyroidism, Cytotoxic chemotherapy thyroid adenomas or carcinomas.
Seascale, was made during 1983 by a television programme. The Black inquiry, through routinely available cancer registration and mortality data, was able to substantiate this suggestion by the availability of relevant geographical analyses - although the additional suggestion that the raised level of leukaemia was a consequence of radioactive discharges from the site to the environment was determined to be most unlikely. An epidemiological cohort study of Seascale children indicated that the excess of leukaemia was concentrated among children born there rather than among children moving in after birth, which suggested that if there was a locally specific cause then it appeared to act before birth or early in life. A subsequent epidemiological case-control study in West Cumbria has shown that cases of childhood leukaemia and non-Hodgkin's lymphoma, but not Hodgkin's disease, are particularly associated with their fathers having raised recorded levels of exposure to external ionising radiation during employment at Sellafield before the children's conception. This association can explain statistically the observed excess of leukaemia in Seascale. Other issues and studies relating to the interpretation of this finding will be discussed.
may damage the gonad and both irradiation and cytotoxic chemotherapy may interfere with the normal growth of bone. These complications of treatment may lead to various clinical presentations including infertility, gynaecomastia, impaired growth leading to short stature, development, pubertal undergo normal failure to precocious puberty, hyperparathyroidism, hypothyroidism, thyroid tumours and varying degrees of hypopituitarism. of practice the commonest defect In clinical hypothalamic-pituitary function following irradiation is There is now isolated growth hormone (GH) deficiency. unequivocal evidence that GH therapy will improve the height prognosis of children with radiation-induced GH The consequences of GH deficiency in adult deficiency. life and the need for GH replacement therapy however, are only being explored now. The ovarian damage following cytotoxic drugs or Therefore, in such irradiation is often irreversible. females, donor oocyte and in vitro fertilisation may represent their only chance to carry a pregnancy. Studies of uterine function will determine which females might benefit from such an approach.
5
Abstracts of the following invited talks were not received
si0
WHY ARE THERE DIFFERENT GROUPS OF TUMOUR SUPPRESSOR GENE INACTIVATIONS IN VARIOUS PAEDIATRIC AND ADULT TUMOURS? S.H. Friend (MGH Cancer Center, Charlestown, USA)
CHILDHOOD CANCER IN ADULTS: BIOLOGIC BASIS AND CLINICAL OVERLAP T.J. Triche (Children's Hospital of Los Angeles, USA)
ENVIRONMENTAL EXPOSURE TO RADIATION IN THE AETIOLOGY OF CANCER V. Beral (ICRF, Oxford)
WAYS OF IMPROVING PSYCHOLOGICAL CARE OF CANCER PATIENTS P. Maguire (Manchester)
S9
S13
S16
NEW APPROACHES TO THE PREVENTION AND TREATMENT OF BREAST CANCER T. Howell (Christie Hospital, Manchester) S17
S18
RESULTS OF THE OVERVIEW ANALYSIS OF CHEMOTHERAPY FOR ADVANCED OVARIAN CANCER
C. Williams (Southampton)
ADVANCES IN THE MANAGEMENT OF GESTATIONAL TROPHOBLASTIC TUMOURS E. Newlands (Charing Cross Hospital, London)
S20
S19
ENHANCING THE SELECTIVITY OF ANTICANCER DRUGS
M. D'Incalci (Milan)
GENES AND MOLECULES IMPORTANT FOR RESISTANCE TO ANTICANCER DRUGS: IMPLICATIONS FOR CANCER CHEMOTHERAPY V. Ling (Toronto)
S21
6
Si INOSITOL LIPIDS AND CELL PROLIFERATION M.J. Berridge, Dept of Zoology, University of Cambridge, Downing Street, Cambridge CB2 3EJ. Many of the stimuli responsible for controlling cell proliferation are known to induce changes in phosphoinositide metabolism. There are two main effects. Firstly, growth factors stimulate the hydrolysis of phosphatidylinositol 4,5-bisphosphate to generate two second messenger pathways. Inositol 1,4,5-trisphosphate (Ins(1,4,5)P3) is released to the cytosol to mobilize internal carlcium and to promote an influx of external calcium whereas the diacylglycerol which remains within the plane of the membrane activates protein kinase C. The calcium signalling pathway displays complex spatial and temporal patterns which may be particularly re1evant to cell proliferation since they have been observed in eg s undergoing fertilization and in lymphocytes and fibroblasts stimulated with growth factors. This bifurcating signalling pathway is activated by growth factors operating either through G proteins or through tyrosine phosphorylation. In the latter case, receptors for growth factors such as PDGF and EGF activate PLC- by tyrosine phosphorylation which means that the generation of calcium siqnals and the activation of protein kinase C are almost always associated with the onset of cell proliferation. What remains a subject of debate is whether or not these signals have any direct role to play in stimulating DNA synthesis. The second major effect of growth factors on metabolism is to stimulate an inositol yhosphoinositide ipid 3-kinase which leads to the formation of a novel class of inositol lipids carrying a phosphate on the 3-position. Together with PLC Ra 1 and GAP this 3-kinase is activated by tyrosine phosphorylation. Since the appearance of these novel inositol lipids is also apparent following transformation by certain oncogenes such as src, it seems they may play an important role in cell proitTeration perhaps related to changes in cell morphology. While it is clear that the phosphoinositides undergo profound changes when cells are induced to grow, there is still considerable uncertainty as to their precise role in controlling events during the cell cycle.
Inosltol phospholiplds as precursors of lntracellular
second messengers. C.M. Macpheel, A.N. Carter2, F. Ruiz-Larrea3 and C. Peter Downes2, ICellular Pharmacology, Smith Kline Beecham, The Frythe, Welwyn, Herts. AL6 9AR,2Department of Biochemistry. University of Dundee, Dundee DDI 4HN, 3Ludwig Institute for Cancer Research, Coutauld Building, London WIP 8BT. The inositol-containing phospholipids have a critical role in coordinating cellular activity because they furnish cells with a number of intracellular signal molecules in response to a wide variety of hormones, neurotransmitters and growth factors. It is now generally accepted that, upon agonist activation, many cell surface receptors stimulate a phospholipase C which cleaves phosphatidylinositol (4,5)-bisphosphate to give diacylglycerol, which remains associated with the membrane, and inositol (1,4,5)-trisphosphate (InsP3). which diffuses into the cytosol. Diacylglycerol is the second messenger which activates protein kinase C and InsP3 stimulates the release of Ca2+ from intracellular storage sites. These second messengers control responses as widespread as fertilisation, early embryonic development, smooth muscle contraction, cell proliferation, secretion and neuronal excitability. Recently, a novel group of inositol phospholipids has been identified in which an additional monoester phosphate moiety is inserted in the D-3 position of the inositol ring by phosphatidylinositol 3-kinase (Ptdlns 3-kinase). Several lines of investigation have implicated one of the products of this novel pathway, phosphatidylinositol (3,4,5)-trisphosphate, as a second messenger in mitogenic and inflammatory cell signalling. Inositol phospholipid kinases, therefore, occupy pivotal positions in second messenger pathways by directing the synthesis of phospholipid isomers with apparently distinctive signalling functions. We have investigated the substrate specificity of 3 types of phosphatidylinositol kinase using a variety of ATP site ligands and synthetic enantiomers of phoshatidylinositol. Both types of compound reveal substantial differences in the substrate recognition sites of Ptdlns 3-kinase and Ptdlns 4-kinases, respectively, which could be exploited in the developement of specific inhibitors of these enzyms:s.
S4
PROTEIN KINASE C, P.J. Parker, S.Cazaubon, N. Goode, A. Olivier, C. Pears, J. Ridden. Protein Phosphorylation Laboratory, Imperial Cancer Research Fund, 44 Lincoln's Inn Fields, London WC2A 3PX.
S3
ONCOGENES, GROWTH FACTORS AND PHOSPHATIDYLINOSITOL 3-KINASE L. Cantley, Dept. of Physiology, Tufts University School of Medicine, Boston, MA 02111 USA.
Phosphatidylinositol 3-kinase (Ptdlns 3-kinase) was discovered because of its specific association with oncogenic protein-tyrosine kinases (Whitman et al., 1985 Nature 315, 239; Whitman et al., 1988 Nature 332, 644). This enzyme phosphorylates the D-3 position of the inositol ring of either phosphatidylinositol (Ptdlns), Ptdlns-4-P or Ptdlns-4,5-P2 to generate the novel lipids PtdIns-3-P, Ptdlns-3,4-P2 and PtdIns-3,4,5-P3 respectively. These lipids are not substrates for Ptdlns-specific phospholipases C and they are not in the canonical Ptdlns turnover pathway. The correlation between the transforming ability of mutant proteintyrosine kinases and their association with Ptdlns 3-kinase argues that this enzyme is in a critical pathway for regulation of cell growth and transformation (Cantley et al., 1991 Cell 64, 281). Notably, Ptdlns-3,4-P2 and Ptdlns-3,4,5-P3 are absent in quiescent cells but appear within minutes of addition of growth factors or upon transformation by oncogenes of the protein-tyrosine kinase family. The Ptdlns 3-kinase is a heterodimer of 85 kd and 110 kd subunits (Carpenter et al., 1990 J. Biol. Chem. 265, 19704). The 85 kd subunit is phosphorylated on tyrosine in vivo in polyoma middle t-transformed cells or platelet-derived growth factorstimulated cells (Kaplan et al., 1987 Cell 50, 1021) and can be phosphorylated in vitro by baculovirus-expressed polyoma middle t/pp6Oc-src complex. The gene for the 85 kd subunit has been cloned by several laboratories and the predicted protein has regions of homology to the non-catalytic domain of pp6Oc-src (SH-2 and SH-3 domains). The involvement of the SH-2 domain in association with tyrosinephosphorylated proteins will be discussed. The potential role of the lipid produlcts of the Ptdlns 3-kinase in cell signaling will also be discussed.
Protein kinase C is made up of a family of structurally related
polypeptides. The consistent feature of this family, is the
presence of a kinase domain (defincd both functionally and through sequence alignment) and a cysteine rich region conferring phorbol ester (and by inference diacylglycerol) binding. Where studied it is evident that the binding of phorbol esters/diacylglycerol in a suitable phospholipid context leads to activation of the kinase. Thus these proteins have been considered to operate immediately downstream of agonist evoked diacylglycerol production following activation of phosphatidylinositol 4,5 bisphosphate directed phospholipase C. The coincident production of inositol 1,4,5 trisphosphate and consequent Ca2+ mobilisation will synergise with diacylglycerol in the activation of the Ca2+-dependent forms of protein kinase C. By contrast the more recently described Ca2 + -independent protein kinase Cs have the potential to respond to diacylglycerol produced both in the presence or absence of a rise in cytoplasmic Ca2+; the latter being exemplified by the breakdown of phosphatidylcholine. This suggests that the patterns of expression of these two branches of the PKC family (Ca2 +-dependent and -independent) permit cells to respond to different effector systems. The manner of the response is governed by the substrate specificity of the individual PKCs; there is ample evidence that there are significant differences in specificity. Thus this growing family of signal transducing proteins appear to operate to diversify the circumstances in which the pathway is triggered and the nature of the output.
3
S6
GROWTH FACTOR RECEPTORS IN CELL SIGNALLING, W.J. Gullick, ICRF Oncology Group, Hammersmith Hospital, Du Cane Road, London W12 OHS Since the discovery of the epidermal growth factor receptor, a large number of both structurally and functionally related molecules have been identified which are either known to be or likely to be involved in transmitting mitogenic stimuli across cell membranes. This family of growth factor receptor molecules is defined by possessing an extracellular domain responsible for ligand binding and an intracellular domain with protein tyrosine kinase activity. In all cases so far examined, ligand binding promotes tyrosine kinase activity. In most cases this leads to stimulation of mitogenesis although there are some ligands such as amphiregulin which may also have inhibitory properties on some cell types. Several tyrosine kinase receptors can be activated by mutation or overexpression to behave as dominant in oncogenes vitro and in transgenic animals. Examination of human tumour biopsies has identified alterations in the c-erbB-1, c-erbB-2, trk, met and Ksam genes. The c-erbB-1 (EGFR) gene is amplified or overexpressed in many squamous carcinomas and the cerbB-2 gene in several types of adenocarcinoma. Overexpression of either gene has in some cases been shown to indicate a short relapse free interval and survival from the disease. Such cell surface molecules provide an attractive target for new forms of therapies. Our own work has focused on designing compounds which prevent receptor dimerisation, an event necessary for receptor activation. Other strategies for the production of GFR inhibitors include ligand antagonists and antisense oligonucleotides. If such drugs can be produced they have potential for treatment of many common solid tumours.
REGULATION OF THE ACTIVITY OF p2lras IN T-LYMPHOCYTES. J.Downward, Imperial Cancer Research Fund, Lincoln's Inn Fields, London WC2A 3PX.
S5
The ras proteins bind and hydrolyse GTP: p21.GTP is the active from of the molecule while p21.GDP is inactive. The physiological signals responsible for controlling the activation state of p2lras have long been sought. We have recently demonstrated that in intact T lymphocytes the activation state of p2lraS, as determined by the amount of GTP bound to it relative to GDP, is under control of the T-cell antigen receptor. Protein kinase C mediates this stimulation of p2lraS which appears to occur through the suppression of GAP activity. An inhibitor of GAP activity is stimulated in Tcells upon treatment with protein kinase C activators. A permeabilised cell system has been used to study this activation of p2lras further: this shows that the rate of guanine nucleotide exchange on p21 in both lymphoblasts and fibroblasts is greatly stimulated relative to on pure p21 in vitro. The activation state of p21raS in both cell types appears to be controlled by a dynamic balance between two highly active and opposing proteins, GAP/NF-1 and a guanine nucleotide exchange factor.
S8
S7
SIGNAL TRANSDUCTION MECHANISMS IN ANTICANCER DRUG ACTION AND RESISTANCE. Thomas R. Tritton, Department of Pharmacology, University of Vermont Medical School, Burlington, VT 05405 USA. One of the hallmarks of cancer is loss of cellular growth regulation. Now that it is established that many oncogenes code for proteins involved in signalling and growth control, there has developed an increased recognition that an understanding of these basic processes offers new prospects for controlling neoplasia. In this presentation we will discuss two ideas stemming from our laboratory's work: (1) the ability of cytotoxic anticancer drugs like adriamycin to kill susceptible cells appears to involve a series of events initiated at the plasma membrane and proceeding through the protein kinase C signal transduction pathway to ultimate damage to DNA in the nucleus; and (2) resistance to antineoplastics, particularly multidrug resistance, may be influenced both by modulations of protein kinase C biology, and by the action of oncogenes like c-fos and c-jun, which bind to the promoter region of genes involved in drug resistance and regulate their transcription.
CELL MEMBRANE/SIGNAL TARGETS FOR ANTICANCER CHEMOTlERAPY
D Crowther, CRC Dept Medical Oncology, Christie Hospital, Manchester, M20 9BX, U.K.
Our understanding of the way in which cells interact with their environment is improving and its complexity is already appreciated. The mechanisms whereby the cell surface interacts with other cell membranes, the cell matrix, growth factors and hormones, and the consequent cascade of biochemical events leading to changes in cell growth and function suggest a substantial number of potential targets for systemic anticancer therapy. The fact that several elements in this system have been identified as proto oncogene products suggest that mutant products could be targetted with a degree of specificity for the tumour cell compared with its normal counterpart. Already clinical studies investigating the use of growth factors, their analogues, receptors or antibodies directed against them have been started. Other approaches involve the use ofagents designed to influence cell signal mechanisms. These include modulators of protein kinase C (Bryostatin- 1, Staurosporins) and the ether lipids such as SRI 62-834 and hexadecylcholine phosphate. Some are currently in Phase I trials under the auspices of the CRC. Families of enzymes such as the phospholipases, tyrosine kinases and phosphates and the GTPase superfamily are also targets against which new agents are being developed. Other agents being developed have structural similarities to heparan sulphate (HS) with the capacity to interfere with growth factors/receptor interaction and currently one of these (Suramin) is in clinical phase I/l trials in the USA and Europe. A better understanding of HS chemistry and an appreciation ofthe diversity of molecular structure and potential specificity for mediating biological events could well lead to the development of an important new group of agents with anticancer activity. Some agents active on the cell membrane or on the cell signalling cascade 'now being developed are highly active in biological systems and starting doses in clinical trial may be at the 5jg/ml level with expected serum levels below the detection limit for current assays. These and drug scheduling issues pose special problems for clinical study.
4
MOLECULAR EPIDEMIOLOGY OF FAMILIAL S12 CANCERS, F. Li, Clinical Epidemiology Branch, National Cancer Institute and Dana-Farber Cancer Institute, Boston, Mass., 02115 U.S.A.
Sil
SMALL CELL LUNG CANCER (SCLC) AND NEUROBLAS1 ~~~~~~~~2 TOMA (NB), J. Pritchard' and R.L. Souhami 1Dept. of Haematology & Oncology, Institute of Child Health, London WC1, 2Dept. of Oncology, University College & Middlesex School of Medicine, London Wl.
Clinical studies, epidemiological research and laboratory investigations have contributed to understanding of human cancer biology. My colleagues and I have attempted to integrate these 3 approaches to facilitate accrual of new knowledge. We study cancerof families at exceptionally high risk as human models
Both SCLC and NB express markers of neuroendocrine (NE) differentiation, such as neural cell adhesion molecule, neurosecreting granules and NSE. Poor prognosis in NB appears to be associated with amplification of n-myc which is also amplified in SCLC lines although the relationship to prognosis is less clear. Chromosome deletion at lp and 17p are features of NB and 3p and 17p in SCLC. Both tumours may undergo a change in morphology after chemotherapy perhaps reflecting differentiation, but this is less marked in SCLC and is more difficult to achieve in tissue culture. The metastatic patterns of both tumours are very similar with frequent spread to local nodes, liver and bone marrow. NB and SCLC are sensitive to chemotherapy, the same drugs being active in both diseases. Although remission rates are high, relapse occurs in the majority of patients with NB who are over 1 year of age and who have stage 3 and 4 disease. In SCLC cure is extremely uncommon (approximately 3%). In both tumours control of systemic disease is the main problem. In both NB and SCLC it appears that short intensive remission induction procedures are probably superior in producing remission. This is now being tested formally in both diseases. In NB there is evidence that 'consolidation' with high dose chemotherapy prolongs remission. This remains a possibility, and may improve survival, though less prectical, in SCLC. The definition of non-cross resistance and the design of high dose strategies presents similar problems in both diseases. There are reasonable grounds for believing that chemotherapy lessons learned from the treatment of one disease may have implications for the other. This is also true for other approaches to therapy such as modulation of autocrine growth and antibody-guided therapy
cancer susceptibility. These patients represent rare "experiments of nature" that can reveal new insights into carcinogenic processes. The rationale for investigating rare family cancers rests in Knudson's
hypothesis, which is supported by experimental data on retinoblastoma. This model indicates that the multiple
steps in carcinogenesis can be dissected into 2 components in cancer families for ease of study. Specially, germ line of familial cases can reveal the first mutation, and subsequent mutations are detected by comparing tumor DNA with the corresponding germ line DNA. Our interdisciplinary research employs clinical observations to find susceptible patients, epidemiological studies to quantitate the excess risk, and laboratory investigations to examine the biological basis of susceptibility. The studies have uncovered several new patterns of familial cancers. This presentation describes our investigations of three disorders: 1) the syndrome of sarcomas, breast cancer, and other neoplasms; 2) co-inheritance of renal cell carcinoma and a constitutional chromosome translocation in a kindred; and 3) familial Wilms' tumor that indicates genetic heterogeneity of the neoplasm.
FATHER'S EXPOSURE TO RADIATION AT SELLAFIELD AND CHILDHOOD CkNCER, Martin J. Gardner, Medical Research Council Environmental Epidemiology Unit, (University of Southampton), Southampton General Hospital, Southampton S09 4XY. The original suggestion that the level of childhood leukaemia was raised near the Sellafield nuclear site in Cumbria, particularly in the nearby village of
S14
ENDOCRINE SEQUELAE FOLLOWING THE TREATMENT OF CHILDHOOD CANCER, S. M. Shalet, Dept of Endocrinology, Christie Hospital & Manchester H20 8BX. Holt Radium Institute,
S 15
hypothalamic, impair directly may Radiation pituitary, thyroid and gonadal function or alternatively it may induce the development of hyperparathyroidism, Cytotoxic chemotherapy thyroid adenomas or carcinomas.
Seascale, was made during 1983 by a television programme. The Black inquiry, through routinely available cancer registration and mortality data, was able to substantiate this suggestion by the availability of relevant geographical analyses - although the additional suggestion that the raised level of leukaemia was a consequence of radioactive discharges from the site to the environment was determined to be most unlikely. An epidemiological cohort study of Seascale children indicated that the excess of leukaemia was concentrated among children born there rather than among children moving in after birth, which suggested that if there was a locally specific cause then it appeared to act before birth or early in life. A subsequent epidemiological case-control study in West Cumbria has shown that cases of childhood leukaemia and non-Hodgkin's lymphoma, but not Hodgkin's disease, are particularly associated with their fathers having raised recorded levels of exposure to external ionising radiation during employment at Sellafield before the children's conception. This association can explain statistically the observed excess of leukaemia in Seascale. Other issues and studies relating to the interpretation of this finding will be discussed.
may damage the gonad and both irradiation and cytotoxic chemotherapy may interfere with the normal growth of bone. These complications of treatment may lead to various clinical presentations including infertility, gynaecomastia, impaired growth leading to short stature, development, pubertal undergo normal failure to precocious puberty, hyperparathyroidism, hypothyroidism, thyroid tumours and varying degrees of hypopituitarism. of practice the commonest defect In clinical hypothalamic-pituitary function following irradiation is There is now isolated growth hormone (GH) deficiency. unequivocal evidence that GH therapy will improve the height prognosis of children with radiation-induced GH The consequences of GH deficiency in adult deficiency. life and the need for GH replacement therapy however, are only being explored now. The ovarian damage following cytotoxic drugs or Therefore, in such irradiation is often irreversible. females, donor oocyte and in vitro fertilisation may represent their only chance to carry a pregnancy. Studies of uterine function will determine which females might benefit from such an approach.
5
Abstracts of the following invited talks were not received
si0
WHY ARE THERE DIFFERENT GROUPS OF TUMOUR SUPPRESSOR GENE INACTIVATIONS IN VARIOUS PAEDIATRIC AND ADULT TUMOURS? S.H. Friend (MGH Cancer Center, Charlestown, USA)
CHILDHOOD CANCER IN ADULTS: BIOLOGIC BASIS AND CLINICAL OVERLAP T.J. Triche (Children's Hospital of Los Angeles, USA)
ENVIRONMENTAL EXPOSURE TO RADIATION IN THE AETIOLOGY OF CANCER V. Beral (ICRF, Oxford)
WAYS OF IMPROVING PSYCHOLOGICAL CARE OF CANCER PATIENTS P. Maguire (Manchester)
S9
S13
S16
NEW APPROACHES TO THE PREVENTION AND TREATMENT OF BREAST CANCER T. Howell (Christie Hospital, Manchester) S17
S18
RESULTS OF THE OVERVIEW ANALYSIS OF CHEMOTHERAPY FOR ADVANCED OVARIAN CANCER
C. Williams (Southampton)
ADVANCES IN THE MANAGEMENT OF GESTATIONAL TROPHOBLASTIC TUMOURS E. Newlands (Charing Cross Hospital, London)
S20
S19
ENHANCING THE SELECTIVITY OF ANTICANCER DRUGS
M. D'Incalci (Milan)
GENES AND MOLECULES IMPORTANT FOR RESISTANCE TO ANTICANCER DRUGS: IMPLICATIONS FOR CANCER CHEMOTHERAPY V. Ling (Toronto)
S21
6
