Evaluation of CD44 transcription variants in human digestive tract carcinomas and normal tissuesкод для вставкиСкачать
Znt. J. Cancer: 66,l-5 (1996) 0 1996 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de I’Union InternationaleContre le Cancer REPORT ON THE 5TH JAPANESE-GERMAN WORKSHOP ON MOLECULAR AND CELLULAR ASPECTS OF CARCINOGENESIS Meeting held in Essen, Germany, July 11-13, 1995 Toshio KUROKIand Manfred F. RAJEWSKY At this 5th Japanese-German Workshop, 12 Japanese and 12 German scientists presented r studies, ranging from the molecular basis of carcinogenesis and tumor progression to future pos innovatory approaches to the early diagnosis and therapy of cancer based on current rapid advances in molecular genetics and cell biology. Understanding the molecular basis of carcinogenesis in epithelial cell systems is of prime importance, since epithelial tumors are by far the most frequent cancers in humans. Dr. T. Kuroki (Tokyo) reported on a possible signal-transduction pathway mediating squamous differentiation. Two clones of Ca2+-independentisoforms of protein kinase C (PKC), i.e., the 7 and 8 isoforms, were isolated from a cDNA library of mouse skin. The isoform was highly expressed at the mRNA and protein levels in the epithelial tissues in close association with differentiation. Cholesterol sulphate was found to act as a second messenger activating the q isoform. Activation of the q isoform leads to transcription of transglutaminase I, which in turn results in the cross-linking of structural proteins of keratinocytes. Cholesterol sulphate was further shown to be a potent inhibitor of tumor promotion and of the growth of keratinocytes and their malignant counterparts. This study elucidated several steps of the signal-transduction pathway leading to squamous differentiation, which may act as a negative regulator of carcinogenesis. To characterize sequential genetic and henotypic changes associated with different stages of the process of skin carcinogenesis, Dr. N.E. Fusenig (Heidelberg7 has been using an in vitro system (human skin-keratinocyte cell line HaCaT and its c-Ha-rus-I-transfected derivative clones, which, in nude mice, give rise to benign or malignant tumors respectively). Focusing on growth regulation and epithelial-mesenchymal interactions, his analyses have shown that both tumorigenic phenotypes had acquired an altered response to growth factors and autonomous growth potential in vitro. Malignant (invasive) cells could be discriminated in vivo only where they gave rise to squamous-cell carcinomas, while benign cells formed stationary cysts, and immortal cells were growth-inhibited followed by terminal differentiation. These observations suggested decreasing sensitivity to, or escape from, growth-control mechanisms mediated by the mesenchyme. This could be visualized more clearly in a surface-transplantation assay, in which tumor and host cells are initially separated by an acellular collagen matrix and early tumor-host interactions can be studied. In these conditions, HaCaT cells and benign sub-clones formed organized and differentiating epithelia, while malignant cells rapidly elicited the activation of mesenchyme and directed angiogenesis before onset of invasive growth. Dr. N. Huh (Tokyo) discussed the morphogenesis of mouse embryonic skin in a floating membrane culture. Embryonic mouse skin undergoes drastic morphological change from the 13th to 16th gestational day, i.e., formation of rudiments of hair follicles, and stratification and cornification of inter-follicular epidermis. To reveal the underlying molecular mechanisms of the morphogenesis, a culture system on floating membrane was established to allow the skin tissues isolated from 12- or 13-day embryos to develop in a histologically similar way, as with the process in vivo. Expression of differentiation markers of epidermal keratinocytes, including cholesterol sulphotransferase and cytokeratin KI, were induced during the cultivation period, as also occurs in the corresponding period in vivo. EGF, TGFa and KGF specifically inhibited hair-follicle formation with marginal effects on interfollicular epidermis. The inhibitory action by EGF was reversible and stage-specific, i.e., at an early stage of the development of hair rudiments. An adenovirus-based expression vector was used to introduce exogenous genes, which were expressed in almost all the epidermal cells in the histological architecture. Thus, the present culture method, in cqmbination with appropriate expression vectors, provides a useful experimental system for studying the molecular processes of morphogenesis in mice. Dr. R. Muller (Marburg) presented evidence indicating that interference with transcriptional activation by a novel promoter element (CDE) is a common mechanism in cell-cycle-regulated transcription. This was shown for the TATA-less cdc25C gene, which reaches maximum activity in late S/G2. CDE-mediated repression also plays a crucial role in the cell-cycle regulation of other genes, such as cdc2 and cyclin A. The CDE is located close to one of the major transcription initiation sites in all 3 genes. Genomic dimethyl-sulphate footprinting experiments, either with synchronized cells or with cycling cells sorted by FACS, showed the formation of CDE-protein complexes both in Go and in GI cells, and their dissociation in G2. Mutation of CDE severely impairs cell-cycle regulation of the cdc25C, cdc2 and cyclin-A promoters and leads to high expression in Go/Gl. Apparently, the CDE is a cell-cycle-regulated cis-acting repressor element, blocking the activation of transcription by the glutamine-rich activators Spl and NF-Y (which bind to the cdc-25C, cdc2 and cyclin-A enhancers). Correspondingly, cell-cycle regulation is lost upon removal of the enhancer region located immediately upstream of the CDE, but is largely restored when an enhancerless minimal promoter fragment is linked to the constitutive SV40 early promoter/enhancer. The CDE repressor element may represent a candidate molecule in the context of gene therapeutic approaches towards proliferative disorders. The SeriThr-specific kinase Raf is part of a Ras-dependent signalling pathway, with Ras lying upstream of Raf; the functional role of Ras being to move its effector Raf to the plasma membrane where it becomGs activated by an as-yet-unknown mechanism. Dr. N. Nassar (Dortmund) reported on the X-ray crystal structure, at 1.9 A resolution, of the complex between residues 1-167 of the Ras-related (identical effector region) GTP-binding human protein RaplA (capable of reverting transformation induced by the K-rus oncogene) and the Ras-binding domain (RBD) expressed from human c-Rafl (residues 55-131). The topology of RBD strongly resembles that of ubiquitin (“ubiquitin fold”). The RaplAiRBD complex was purified in the presence of the GTP analog GppNHp prior to crystallization. The structure of Received: December 7,1995 2 KUROKI AND RAJEWSKY complexed RaplA is very similar to uncomplexed Ras-GppNHp. Tight interactions-via an anti-parallel P-sheet formed by strands p2-P3 from RaplA and Bl-B2 from RBD-provide for the inhibitory effect of RBD on guanine-nucleotide dissociation. The structure of the RaplAiRBD complex accounts for the interest in whether the observed mode of interaction is also used by other GTP-binding proteins (e.g., Rho, Ran) and in the case of other effector molecules (eg., GAP). Moreover, the small size of the RaplA (Ras)/RBD interface may permit the design of inhibitory “anti-Ras drugs”. Genes that are physiologically expressed, but down-regulated, in the presence of a mutationally activated rus gene, represent potential suppressors of neoplastic phenotypes. Dr. R. Schafer (Zurich) has identified a number of candidate genes (H-rev genes) with these properties in primary rat fibroblasts, in rat 208F and mouse 3T3 cells (non-tumorigenic) and in cells of the rodent cell lines REF-52 and EK-3 (transformation-resistant). Phenotypic revertants of rus-transformed cells exhibited at least partially restored expression levels of genes encoding extracellular matrix components (e.g., collagen type I) and the matrix-modifying enzyme lysyl oxidase. Coordinate over-expression of several H-rev genes may thus be involved in the suppression of neoplastic phenotypes in rus transformants. Interestingly, expression of the angiogenesis inhibitor thrombospondin was also down-regulated in rus-transformed cells (probably contributing to the rapid neovascularization of tumors originating from these cells); in the revertants, however, up-regulation was not consistently seen. Two novel genes were isolated by subtractive hybridization: n’l (H-rev18), a member of the heterogeneous class of LIM genes, and H-revl07, encoding an 18-kDa polypeptide without similarity to known proteins. While proliferating 208F cells contained low levels of the cytoplasmic form of the H-rev107 protein, the membraneassociated form accumulated in density-arrested cells. Withdrawal of serum did not result in elevated H-rev107 levels in cultured fibroblasts. Preliminary results indicate the absence of H-rev107 expression in neoplastic cell lines and experimental tumors. The heterogeneity of tumor histology is considered to reflect alterations in morphogenetic gene function. Dr. S. Hirohashi (Tokyo) presented evidence that the cadherin system, which mediates Ca++-dependent homophilic cell-cell adhesion, is inactivated by multiple mechanisms in carcinoma cells which are poorly differentiated and highly invasive. Mutations have been found in the genes for E-cadherin, a major cadherin in epithelial cells, and its undercoat proteins, a and P catenins, which connect E-cadherin to actin filaments and establish firm cell-cell adhesion. Transcriptional inactivation of E-cadherin expression caused by CpG-methylation of the E-cadherin promoter was also shown to play a significant role. In addition, tyrosine-phosphorylation of the E-cadherin-catenin complex was hypothesized to inactivate cell adhesion, and association of c-erbB-2 with p-catenin was demonstrated. Dr. Hirohashi proposed that the mechanisms for inactivation of a particular function are multiple, and include genetic and non-genetic alterations. Tumor heterogeneity, irreversible and reversible, represents the sum of these multiple alterations in key systems regulating cell growth and morphogenesis. The adenomatous polyposis coli (APC) gene was cloned as a gene responsible for familial adenomatous polyposis coli (FAP). Dr. T. Noda (Tokyo) established a mouse model for FAP, by introducing a frame-shift mutation at codon 1309 of the APC gene, since this mutation is most frequently observed in FAP families. While introducing a mutation, a carboxyl terminus of mutant APC protein was tagged with HA epitope and the analysis with anti-HA antibody clearly showed that the mutant APC protein is dominantly localized in cytoplasm, and also forms a heterodimer with wild-type APC protein. Multiple tumor formation was observed along the intestinal tracts of all the heterozygous mutant mice. These tumors were adenomas or adenocarcinomas showing loss of heterozygosity of the APC gene. This result reconfirmed the idea that mutation of the APC gene leads to the transformation of intestinal epithelium through a second hit of the wild-type APC gene. The effect of p53 mutation on the process of intestinal tumorigenesis was also analyzed by breeding these mutants with p.53-mutant mice. The results showed that p53 mutation did not affect the process of progression of tumors, but clearly increased the incidence of tumor formation in the colon of the APC mutants. Dr. A. von Deimling (Bonn) discussed the results of a systematic study of 300 human brain tumors based on recent advances in molecular genetics and aimed at establishing new grounds for the clinical management of nervous-system malignancies. These analyses have permitted more precise molecular classification of astrocytic gliomas (pilocytic astrocytomas WHO grade I; astrocytomas WHO grade 11; anaplastic astrocytomas WHO grade Ill; glioblastoma multiforme WHO grade IV), and oligoastrocytomas (WHO grade II), and of meningiomas. Thus, pilocytic astrocytomas appear to represent a genetically distinct entity that does not share molecular features with other astrocytic tumors. Progression from astrocytoma (predominant lesion: LOH 17p) to anaplastic astrocytoma seems to involve a gene localized on chromosome 19q, and progression to glioblastoma multiforme may be influenced by gene(s) on chromosome 10; however, the majority of the latter tumors (like oligoastrocytomas and oligodendrogliomas) do not involve LOH 17p and may therefore arise de novo. Both oligoastrocytomas and oligodendrogliomas frequently exhibit LOH 19q and l p (i.e., lesions not seen in astrocytomas). LOH 19q and l p were also seen both in the astrocytic and in the oligodendroglial portions of oligoastrocytomas, suggesting a monoclonal origin of the latter tumors. A highly significant association of LOH22 and mutations of the NF2 gene was found in different sub-types and WHO grades of meningiomas, supporting the notion that NF2 is the tumor-suppressor gene commonly involved in meningioma formation. However, distinct histopathological meningioma sub-types showed significant differences in NF2 mutation frequencies (fibroblastic/ transitional meningiomas, 80%; meningotheliomatous meningiomas, 25%), suggesting genetically different molecular pathways in the genesis of these tumors. To specify genes involved in the multi-step process of T-cell Iymphomagenesis in mice, Dr. T. Moroy (Marburg) studied Ep-L-myc and Ep-pim-1 “single-gene transgenics,” both of which are predisposed to developing T-cell lymphomas, as well as Ep-L-myclpim-1 double transgenics. In the latter, efficient cooperation of both oncogenes results in a dramatic acceleration of tumorigenesis, which is further accentuated in animals infected with MoMuLV neonatally (proviral tagging; latency periods, approx. 45 days vs. approx. 100 days). A region of 60 kb on chromosome 5 was found to be a major hot-spot for viral integration. This region contains several genes, one of which (gfi-1) encodes a zinc finger protein and was hit by proviral integration in 10% of tumors in double transgenics and in 25 to 30% of tumors in Ep-L-myc or Ep-pim-1 single transgenics. T-cell lymphomas with MoMuLV integration into this gene exhibited high level gfi-1 expression. Transfection of complete murine gfi-l cDNA under the control of the lck promoter into the 11-2-dependent mouse T-cell line CTLL resulted in increased survival upon IL-2 depletion. As the emergence of growth-factor MOLECULAR AND CELLULAR ASPECTS OF CARCINOGENESIS 3 independence is often associated with a more malignant stage of the disease, gfi-1 may play an important role in the progression of lymphoid tumors. The enzyme 06-methylguanine-DNA methyltransferase repairs pre-mutagenic DNA lesions induced by alkylating agents. Making use of gene targeting, Dr. M. Sekiguchi (Fukuoka) established mouse lines deficient in the methyltransferase gene; tissues in these mice contained no methyltransferase activity. Administration of methylnitrosourea (MNU; 50 pgig body weight) to these gene-targeted mice led to early death, yet normal mice treated in the same manner showed no untoward effects. Following MNU treatment, the bone marrow became hypocellular and there was a drastic decrease in the number of leukocytes and platelets, indicating an impaired reproductive capacity of hematopoietic stem cells. These results clearly indicate that methyltransferase protects animals from pancytopenia caused by alkylating agents. Dr. K. Tanaka (Osaka) described how the XPA protein, which is missing or altered in xeroderma-pigmentosumgroup-A patients, preferentially bound to DNA damaged by UV, cisplatin or osmium teroxide, suggesting that the XPA protein is involved in the damage-recognition step of nucleotide-excision-repair (NER) processes. The damaged DNA-binding domain of the XPA protein is contained within the 122 amino acids with a C4-type zinc-finger motif. The XPA protein also bound to the ERCCl repair protein and RPA (replication protein A). Interactions between the XPA and ERCCl proteins or RPA enhanced the damaged DNA-binding activity of the XPA protein, suggesting that a specific interaction between these proteins is required for the early steps of NER processes. To elucidate the in vivo function and molecular basis of the XP-group-A phenotype, XPA-deficient mice were established by the ES-cell techniques. The XPA-deficient mice were defective in NER processes, and developed squamous-cell carcinoma on the shaved back skin at high frequency when irradiated with low doses of UVB, while normal or heterozygous mice did not. The mice provide a useful model, therefore, for studying UV-induced skin carcinogenesis in XP-group-A patients. 8-oxo-7,8-dihydro-2’-deoxyguanosine triphosphate (8-0x0-dGTP) is produced by active oxygen species in the nucleotide pool of the cell and can be incorporated into cellular DNA. Mammalian cells contain an enzyme activity that hydrolyzes 8-0x0-dGTP to 8-0x0-dGMP, thereby preventing the occurrence of mutations caused by misincorporation. Dr. H. Hayakawa (Fukuoka) discussed evidence that 8-0x0-dGTP can be generated not only by direct oxidation of dGTP but also by phosphorylation of 8-0x0-dGDP by nucleoside diphosphate kinase. 8-oxo-dGMP, derived from 8-0x0-dGTP, is further degraded to 8-0x0-deoxyguanosineby a nucleotidase. The nucleotidase was partially purified from an extract of human Jurkat cells and the mode of action was elucidated. Dr. K. Sakumi (Fukuoka) reported on the cloning of cDNA for human and mouse 8-0x0-dGTPase and on the elucidation of their genomic sequences. Mice deficient in 8-0x0-dGTPase activity were constructed. Dr. Kasai (Fukuoka) reported on the increase of S-hydroxyguanine(8-0H-Gua)-repairactivity as an indicator of cellular exposure to oxygen radicals. 8-OH-Gua is one of the major DNA modifications caused by reactive oxygen species. 8-OH-Gua-repair activity was induced after cells were exposed to oxidative stress such as ionizing radiation. The measurement of 8-OH-Gua-repair activity is therefore useful in assessing cellular oxidative stress. When renal carcinogen Fe-NTA was administered to rats, an increase of 8-OH-Gua-repair activity was observed in the kidney within 6 to 24 hr. An increase of 8-OH-Gua-repair activity was observed in human leukocytes after physical exercise, with varying increased levels for each individual. A higher level of 8-OH-Gua-repair activity was also detected in the leukocytes of smokers as compared with those of non-smokers. A 7-fold inter-individual difference was observed in the 8-OH-Gua-repair activity of smokers. On the basis of the 8-OH-Gua-repair assay of each organ after chemicals were administered to rats, it was concluded that the carcinogenicity of the chemicals and target organs can be predicted, and that the 8-OH-Gua-repair assay of human leukocytes provides a basis for estimating the cancer risk due to the oxidative stress of each individual. Dr. P. Herrlich (Karlsruhe) described gene-regulatory pathways triggered by chemical and physical carcinogens. DNA-damaging agents greatly influence the expression of genes. Altered programs of gene expression result both in beneficial (DNA repair, survival) and in detrimental consequences (mutagenesis, apoptosis), probably depending on the net outcome of induced signal flow and of reactions to the stimuli received by a cell. Adverse agents, such as DNA-damaging chemicals or radiation, can induce a cellular response only if their primary interaction with cellular material becomes translated into “physiological language”. The language that cells understand involves signal transduction and post-translational modification of regulatory proteins (e.g., transcription factors). Two pathways induced by UVC irradiation have been identified. One of these originates from DNA lesions; it is not yet clear how transcription factors (e.g.,p53) are reached and modified. The other pathway involves the activation of growth-factor receptors at the cell surface as the primary event. Here, the molecular mechanism appears to be down-regulation of dephosphorylation of receptor protein tyrosine kinases. The extensive changes in gene expression induced by DNA-damaging agents are explained by the activation of many cell-surface receptors and their respective pathways and of other pathways elicited by UVC or alkylating agents. Activation of the transcription factor CREB is mediated by a UVC-induced p108 CREB kinase, which is not apparently part of the aforementioned 2 pathways. MMS or MNNG activate Jun and ATF-2 through Jun kinases; however, kinetics and other features suggest the existence of another, undefined pathway triggered by alkylation. The expression of the tumor suppressor genep53 during cell proliferation is tightly regulated. As shown by Dr. W. Deppert (Hamburg), only a moderate level ofp53 mRNA and hardly any p53 protein are found in Go lymphocytes. When these cells are triggered into the cell-cycle, the level ofp53 mRNA rapidly rises during GI, but p53 protein remains low. With the onset of DNA synthesis, the p53 mRNA level begins to decline, while there is a pronounced increase in p53 biosynthesis, pointing to an auxiliary role ofp53 connected with the repair of DNA damage. Interestingly, highly purified p53 protein revealed 3’-5‘ exonuclease activity. The negative-feedback autoregulation ofp53 synthesis provides a means to rapidly increase the level ofp53 upon DNA damage, either by halting cells in G I for pre-replicative DNA repair, or by driving cells into apoptosis. When Go lymphocytes were gamma-irradiated with a lethal dose, no increase in p53 protein was found, despite the presence of p53 mRNA. A G1-specific factor may thus be required for abrogation of the negative-feedback control of p53 synthesis. Irradiated Go lymphocytes underwent apoptosis rapidly and quantitatively, indicating that the induction of apoptosis was p53-independent. When the irradiated cells were triggered into proliferation by concanavalin A, the level of p53 rose immediately; apoptosis was drastically delayed and affected only a fraction of cells. In Golymphocytes, therefore, p53-protein expression is intricately regulated at the translational level, and 4 KUROKI AND RAJEWSKY the response to DNA damage induced by gamma irradiation is primarily directed towards enhanced DNA repair rather than apoptosis. It has been known for some time that activation of some oncogenesis induces cellular apoptosis which is thought to be one host defense mechanism against neoplastic development. In malignant cells, therefore, the cell-death mechanism must be inhibited. Thus, molecular analyses of cell death should provide a useful insight into cancer prevention and treatment. Dr. Y. Tsujimoto summarized recent studies on the molecular basis of cellular apoptosis, focusing especially on the bcl-2 and ICEICed-3 family genes, which function against cellular apoptotis and drive the death machinery, respectively. A widely accepted mechanism involves bcl-2 activity on reactive oxygen species (ROS). However, the data obtained from a system of cell death in hypoxia showed that bcl-2, or a related gene with a similar death-sparing activity, bcl-xL, exerts an anti-cell-death function by a mechanism other than by regulation of ROS activity, and also that ROS are not common mediators of apoptosis, but rather a trigger inducing apoptosis. Several novel ICE-related genes (rics) have recently been identified. One of them, ric-2 (the same as T X ) ,is able to induce apoptosis when over-expressed. Apoptosis induced by ric-2 is inhibited by crmA, an ICE inhibitor, but not by YVAD-CHO, a tetrapeptide ICE inhibitor, which indicates that ric-2 has a similar but distinct substrate specificity from that of ICE. Tumor angiogenesis is a requirement for growth of the primary tumor as well as its metastatic dissemination. An important element in the development of novel strategies of cancer therapy is, therefore, the targeting of cellular molecules involved in this process. With the ultimate aim of identifying molecules that are selectively expressed by angiogenic endothelial cells, Dr. H.G. Augustin (Gottingen) has performed comparative analyses of gene expression in migrating and resting endothelial cells, using differential RNA display as well as monoclonal and recombinant antibody technology. A number of novel endothelial-cell-specific monoclonal antibodies were generated. RNA display showed the gene encoding follistatin, an activin-binding protein, to be exclusively expressed by migrating (as opposed to quiescent) endothelial cells, suggesting involvement of the follistatin/activin system in the autocrine regulation of endothelial cells during angiogenesis, i.e., the existence in endothelial cells of a growth-regulatory pathway that modulates the growth-inhibitory functions of members of the TGF-P family of growth factors. While there is active angiogenesis during nervous-system development, it is down-regulated in the mature brain under physiological conditions. However, angiogenesis may be reactivated under pathological conditions, such as tumor growth. To elucidate the underlying molecular mechanisms, Dr. K.H. Plate (Freiburg i. Br.) has studied the expression of vascular endothelial growth factor (VEGF) and its cognate tyrosine-kinase receptors (flt-1/VEGF receptor 1 and flk-1/KDR/ VEGF receptor-2) during brain development and glioma-induced angiogenesis. The results suggest a paracrine mode of control of endothelial-cell proliferation and angiogenesis. This control is tightly regulated and transient during brain development, switched-off in mature brain, and turned on in glioma cells (VEGF) and in the host vasculature (flt-1 and flk-1IKDR) during tumor growth. This pattern was indistinguishable in human-glioblastoma and rat-cerebral transplants of C6 or GS-9L glioma cells. The growth of gliomas originating from tumor-cell grafts in nude mice or syngeneic rats was inhibited significantly by transfer of a signalling-defectiveflk-1 gene into endothelial cells in situ, thus identifying VEGF as a tumor-angiogenesis factor in human and rodent glial tumors, and the VEGF/flk-1 system as a possible therapeutic target. Dr. S. Nishimura (Tsukuba), who has participated in all the Workshops of this series, presented recent results on NB-506, a novel indolocarbazole anti-cancer agent targeting topoisomerase I. Contrary to other known topoisomerase-I inhibitors, NB-506 also inhibited DNA polymerase (Y and RNA polymerase 11. A unique feature of NB-506 was its differential cytotoxicity in various cell lines, with good correlation between cytotoxic activity and the cellular content of NB-506, indicating that the selective cytotoxicity of NB-506 is due at least partly to its specific accumulation in the cells. NB-506 showed no cross-resistance in multidrug-resistant cell lines, and was quite effective in the treatment of various human cancers transplanted to nude mice. The compound not only inhibited tumor growth, but also caused regression of tumor nodules. In addition, NB-506 strongly inhibited the growth of metastasized tumors in various metastasis models of mice. Interestingly, NB-506 has very low cumulative toxicity when measured by mouse mortality, and has quite a wide therapeutic window. These results indicate that NB-506 is a promising candidate as a future anti-cancer agent. Dr. T. Oki (Toyama) reported that novel antifungal and anti-HIV antibiotics, pradimicins, possessing a core structure of glycosylated dihydrobenzo[a]napthacenequinone substituted with a D-amino acid at the C-15 position (MW 870 kDa) specificallybound to terminal mannose residues on the surface of yeasts, fungi and HIV in the presence of calcium, but not to healthy mammalian cells. Once the sugar composition of the cell membrane of rat basophilia leukemia cells was modified by treatment with 1-deoxymannojirimycin, pradimicins then efficiently bound to the cell membrane and killed the leukemia cells. It is assumed that the differential absorption of dihydrobenzo[a]na hthacenequinone antibiotics to micro-organisms and mammalian cells may induce apoptosis by blocking a Ca++-depen ent signal-transduction pathway under specific physiological conditions. Amplification and/or over-expression of type-I/EGF-receptor-related tyrosine kinases (ErbBIEGFR, ErbB-2, ErbB-3, ErbB-4) is observed in various types of human tumors, particularly mammary and ovarian carcinomas, and correlates with an unfavorable prognosis. High expression on the surface of tumor cells, as opposed to low expression in normal e ithelial tissues, makes these growth-factor receptors candidate target-molecules for tumor-directed therapy, i e . , for the esign of cytotoxic agents preferentially attacking tumor cells. Uptake of therapeutic molecules bound to the extracellular domains of cell-surface receptors can be effected either through receptor turnover or by ligand-induced internalization. Dr. W. Wels (Freiburg i. Br.) described the cloning of the variable domains of monoclonal antibodies FRP5 and 225 which bind to the extracellular domains of ErbB-2 and EGF, respectively. Fusion genes coding for single-chain antibody molecules (scFv) were produced by joining light- and heavy-chain variable domains with a synthetic nucleotide linker. Recombinant immunotoxin genes were constructed by adding scFv encoding DNA to sequences encoding truncated Pseudomonas exotoxin A (ETA) devoid of its own cell-binding domain. A recombinant immunotoxin specific for ErbB-3 and ErbB-4 was generated by PCR amplification of human cDNA encoding the EGF-like domain of the growth factor heregulin p l (HRGPl) and fusion with the modified ETA gene. The bacterially expressed immunotoxins scFv(FRP5)-ETA, scFv(225)-ETA, and HRGP-ETA bind specifically to the respective receptors and exhibit potent cytocidal activity in human tumor cells in vitro. In nude-mouse models, low-dose treatment with these immunotoxins specifically inhibited the growth of tumor xenografts expressing the appropriate receptors. 1 B MOLECULAR AND CELLULAR ASPECTS OF CARCINOGENESIS 5 This Workshop-like the previous Workshops of this series-assembled outstanding cancer researchers from both countries in a friendly, informal atmosphere permitting extensive scientific exchange and the initiation of new collaborative projects. Presentations focused on topical issues in molecular genetics and cell biology relevant to carcinogenesis and clinical oncology. In particular, the discussions centred on the molecular control of cell-cycle progression and differentiation, (tumor) angiogenesis, and programmed cell death (apoptosis), emphasizing epithelial and neural cell systems; on the significance of specific DNA-repair genes and proteins as determinants of mutation/ transformation risk, hereditary cancer susceptibility, and therapy resistance; on gene-regulatory pathways triggered by DNA-reactive chemicals and radiation; and on the development of novel approaches to the early molecular diagnosis and therapy of human cancer. Transgene and gene “knock-out’’ analyses in vivo play an ever-greater role in defining the function of specific genes in processes (eg., signal transduction, DNA repair, apoptosis) that may critically influence carcinogenic risk in particular types of cells, or may be exploited to increase the efficiency of tumor-selective cancer therapy. It was recognized, however, that while advances are being made at an unprecedented pace with respect to the characterization of individual genes and proteins and some of their interactions and athways, the definition of their positions and functions within the extremely complex regulatory networks in mamma ian cells is an imperative, albeit formidable, task. Concomitantly, cancer research faces an increasing multiplicity of network elements, the alterations of which may result in the carcinogenic subversion of normal cellular phenotypes. P ACKNOWLEDGEMENTS This Workshop continued the Essen series of biennial Japanese-German Workshops on “Molecular and Cellular Aspects of Carcinogenesis,” organized traditionally at the Institute of Cell Biology (Cancer Research), University of Essen Medical School, Germany, under the auspices of the Japanese-German Cooperative Program in Cancer Research (see Reports by Kuroki and Rajewsky, 1990, 1992 and 1994). It was supported by the Federal Ministry of Education, Science, Research and Technology, the Deutsche Forschungsgemeinschaft, Germany, and by the International Cooperation Program for the 2nd-Term Comprehensive 10-Year Strategy for Cancer Control of the Ministry of Health and Welfare, and the International Scientific Research Program of the Ministry of Education, Science and Culture, Japan. The Workshop was held in conjunction with the 6th Charles Heidelberger Conference on “Control of Cell Proliferation and Differentiation: Molecular Targets in Carcinogenesis and Cancer Therapy”. REFERENCES KUROKI,T. and RAJEWSKY, M.F., Second Japanese-German Workshop on molecular and cellular aspects of carcinogenesis. Int. J. Cancer, 46,155-158 (1990). KUROKI,T. and RAJEWSKY, M.F., Report on the third Japanese- German Workshop on molecular and cellular aspects of carcinogenesis. Int. J. Cancer, 51,330-342 (1992). KUROKI,T. and RAJEWSKY, M.F., Report on the fourth JapaneseGerman Workshop on molecular and cellular aspects of carcinogenesis. Int. J. Cancer, 57,451457 (1994). LIST OF PARTICIPANTS AUGUSTIN, H.G., Cell Biology Ldboratoly, Department of Gynecology and Obstetrics, Georg-August-Universitat Gottingen, Robert-KochStrasse, 40,37075 Gottingen, Germany. DEPPERT, W., Abteilung fur Tumorimmunologie, Heinrich-PetteInstitut, Universitat Hamburg, Martinistrasse 52, 20251 Hamburg, Germany. FUSENIG, N.E., Research Program Tumor-Cell Regulation, Deutsches Krebsforschungszentrum, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. HAYAKAWA, H., Department of Biochemistry, Kyushu University, School of Medicine, Maidashi 3-1-1, Higashi-ku, Fukuoka 812, Japan. HERRLICH,P., Institut fur Genetik, Forschungszentrum Karlsruhe GmbH, Technik und Umwelt, Postfach 3640, 76021 Karlsruhe, Germany. S., Pathology Division, National Cancer Center Research HIROHASHI, Institute, Tjukiji 5-chome, Chuoku, Tokyo 104, Japan. HUH,N.-H., Department of Cancer-Cell Research, Institute of Medical Science, University of Tokyo, Shirokdnedai, Minato-ku, Tokyo 108, Japan. KASAI,H., University of Occupational and Environmental Health, Kita-Kyushu, Fukuoka 807, Japan. KUROKI, T., Department of Cancer-Cell Research, Institute of Medical Science, University of Tokyo, Shirokanedai, Minato-ku, Tokyo 108, Japan. MOROY,T., Institut fur Molekularbiologie und Tumorforschung (IMT), Philipps-Universitat Marburg, Emil-Mannkopff-Strasse 2,35037 Marburg, Germany. MULLER,R., Institut fur Molekularbiologie und Tumorforschung (IMT), Philipps-Universitat Marburg, Emil-Mannkopff-Strasse 2,35037 Marburg, Germany. NASSAR, N., Abteilung Strukturelle Biologie, Max-Planck-Institut fur Molekulare Physiologie, Rheinlanddamm 201, 44139 Dortmund, Germany. S., Merck Research Laboratories, Banyu Tsukuba ReNISHIMURA, search Institute, Okubo 3, Tsukuba 300-33, Japan. NODA, T., Department of Cell Biology, Cancer Institute, KamiIkebukuro 1-37-1,Toshima-ku, Tokyo 170, Japan. OKI,T., Biotechnology Research Center, Toyama Prefectural University, 5180, Kosugi-machi, Toyama 939-03, Japan. PLATE, K.H., Institut fur Neuropathologie, Universitat Marburg, Baldinger Strasse, 35033 Marburg, Germany. RAJEWSKY, M.F., Institut fur Zellbiologie (Tumorforschung) (IFZ), Westdeutsches Tumorzentrum Essen, Universitatsklinikum Essen, Hufelandstrasse 55,45122 Essen, Germany. SAKUMI, K., Medical Institute of Bioregulation, Kyushu University, Maidashi 3-1-1, Higashi-ku, Fukuoka 812, Japan. SCHAFER,R., Division of Cancer Research, Department of Pathology, University of Zurich, Schrnelzbergstrasse 12,8091 Zurich, Switzerland. SEKIGUCHI, M., Department of Biochemistry, Medical Institute of Bioregulation, Kyushu University, Maidashi 3-1-1,Higashi-ku, Fukuoka 812, Japan. TANAKA,K., Division of Cell Genetics, Institute for Molecular and Cellular Biology, Osaka University, 1-3 Yamada-oka, Suita, Osaka 565, Japan. TSUJIMOTO, Y., Laboratory of Molecular Genetics, Department of Molecular Genetics, Biomedical Research Center, Osaka University Medical School, 2-2 Yamada-oka, Suita, Osaka 565, Japan. A,, Institut fur Neuropathologie, Universitatskliniken VONDEIMLING, Bonn, Sigmund-Freud-Strasse 25,53105 Bonn, Germany. WELS,W., Klinik fur Tumorbiologie an der Albert-Ludwigs-Universitat Freiburg, Postfach 1220,79011 Freiburg i. Br., Germany.