(1) Seminar on “Lymphocyte Development and Activation”
This U.S.-Japan Cancer Research Cooperative Meeting was held in Rockville, Maryland at the Holiday Inn Crowne Plaza, February 19-21, 1990. The meeting was organized by Drs. Hiromi Fujiwara and Richard Hodes and involved 11 U.S. and 8 Japanese scientists who met for 2 1/2 days to discuss issues relating to lymphocyte ontogeny, development, and activation.
I. Early hematopoietic development.
Dr. Gerald Spangrude (Rocky Mountain Laboratories) presented studies which demonstrated heterogeneity among multipotent hematopoietic stem cells. These cells were identified in bone marrow by their failure to express lineage-specific differentiation antigens, and could be divided into subpopulations by their expression or non-expression of Sca-1 (stem cell antigen), and by staining with the fluorescent vital dye rhodamine 123. The Sca-1+ cells had multipotent stem cell activity with as few as 30 cells sufficient to reconstitute lethally irradiated mice. The expression of Rh-123 allowed further subdivision of these stem cells as assayed by both in vitro and in vivo activities.
II. B Lymphocyte development and activation.
Dr. Paul Kincade (Oklahoma Medical Research Foundation) introduced the discussion of B lymphocyte differentiation by providing evidence for the importance of both humoral factors and cell-cell interactions in B cell development. Interleukin 7 (IL-7) was shown to be a potent growth factor for cells of the B lymphocyte lineage. Efficient development of mature B cells from their bone marrow precursors required in addition, however, the interaction with stromal cells. Dr. Kincade presented data indicating the importance of CD44 (Pgp-1) determinants on B cell precursors. CD44 appears to function in the interaction of B cell precursors with stromal cells, perhaps through interaction with hyaluronate. Dr. Shinichi Nishikawa (Kumamoto University) described the stepwise progression of B cell differentiation. In the studies which he summarized, an initial differentiation step was supported by cellular contact with stromal cells, followed by a second stage in which both stromal cell contact and IL-7 were required, and finally by a third stage in which differentiating cells were responsive to IL-7 alone.
Recent findings related to the mechanism of activation of B lymphocytes were presented by three speakers. Edward Clark (University of Washington) summarized the current state of understanding regarding signal transduction in B cell activation. Triggering through the surface Ig receptor on B cells appears to induce phospholipase c (PLC) activity with resultant induction of PI hydrolysis and phosphokinase c activity. These events may be mediated through a GTP-binding protein. In addition, new protein tyrosine phosphorylation is observed, and may represent an independent signalling pathway in B cells. Phorbol esters or antibodies to CD45 can differentially influence the PLC-mediated calcium response and phosphorylation events, indicating that two independent or at least sequential events mediate these components of the B cell activation response. Dr. John Monroe (University of Pennsylvania) emphasized studies of the expression of the Egr-1 gene during B cell activation. This appears to be a primary response gene which is rapidly induced following growth factor or mitogen stimulation of many cell types. Egr-1 is activated in response to those stimuli which induce a net positive activation signal to B cell proliferation. In contrast, Egr-1 is not induced under conditions of negative signalling to B cells. Toshiyuki Hamaoka described the role of IL-5 in B cell differentiation. In the monoclonal B cell system BCL1-CL-3, it was shown that growth is stimulated by either IL-4 or IL-5, whereas only IL-5 renders these cells responsive to IL-2. It was shown that IL-5 increases the expression of mRNA for the p75 chain of the IL-2 receptor, accounting for the induction of IL-2 responsiveness. In contrast, IL-4 strongly downregulates expression of the p55 chain of the IL-2 receptor. These studies thus allow a greater understanding of the integrated influence of multiple lymphokines on B cell growth in this monoclonal model.
III. Cell surface receptors and cell interaction molecules.
The role of IL-5 and its receptor in T and B cell differentiation were discussed by Kiyoshi Takatsu (Kumamoto University). IL-5, which is made by most T helper type 2 cells as well as by mast cells is both a growth factor and differentiation factor for B cells, T cells, and eosinophils. The receptors for this lymphokine are both high affinity (10-11) and low affinity (10-9). Both 60 kd and 120 kd chains have been associated with IL-5-binding activity. IL-5 transgenic mice have been made which demonstrate a striking eosinophilia in both blood and tissues as well as an increase in Ly1+ B cells, and increases in autoantibodies and overall serum Ig levels.
Kazuo Sugamura (Tohoku University) described the most recent progress in his characterization of IL-2 receptor subunits and functions. A third component of the high affinity IL-2 receptor was suggested by the observation that transfection with both!!
!and!!
!chains failed to reconstitute a functional high affinity receptor. A p64 chain, which is distinct by peptide mapping from the!!!and!!!chains, was identified by biochemical analysis of IL-2 binding protein. It was also demonstrated that the p75 chain of the IL-2 receptor is associated with tyrosine kinase activity which may play a role in signalling through the IL-2 receptor. Hideo Yagita (Juntendo University) characterized the adhesion molecules involved in murine lymphocyte interactions. Both CD2 and LFA1 were shown to have important function in T cell activation and/or development. In addition, a newly described monoclonal antibody toward a distinct cell interacrion molecule was described. This molecule, which appears to interact with fibronectin was also shown to be functionally important in T cell activation.
IV. T lymphocyte development.
The role of stromal cell interactions in T cell development was described by Sonoko Habu (Tokai University) and Hiromi Fujiwara (Osaka University). Dr. Habu described a thymic stromal cell clone TMC-R3, with which CD4+8+ thymocytes selectively associate in vitro After interaction with this stromal monolayer, immature CD4+8+ thymocytes differentiate into CD4-8+ cells in response to stimulation with anti-CD3 or SEA and IL-2. The response to SEA indicates that this is a clonally specific expansion. Dr. Fujiwara described another thymic stromal cell clone, MRL104.8a. CD4-8- thymocytes can be maintained in the presence of these stromal cells or their derived factor (TSTGF). More extended culture results in the appearance of CD4+8+, and then single positive CD4+ and CD8+ cells. Most of these single positive T cells are CD3+. The growth factor which is active under these conditions may consist, at least in part, of IL-7. Gene Shearer (NIH) described the effect of cyclosporin A (CsA) on T cell development. In a model of lethal irradiation and bone marrow reconstitution, it was shown that CSA administered in vivo leads to an arrest in the development of mature “single positive” T cells, with a dramatic decrease in T cell receptor (CD3) bright cells in thymus. The influence of CSA on T cell development was further studied in fetal thymus organ culture. In this system, it was again shown that there is an arrested development of mature single positive T cells. This effect seems to be upon the development of T cell receptor-expressing cells. Selective mechanisms acting in the development of the T cell receptor repertoire were described by Ada Kruisbeek (NIH) and Richard Hodes (NIH). Dr. Kruisbeek described mechanisms of tolerance induction and of positive selection. It was shown through the use of blocking with monoclonal antibodies in vivo that positive selection involves recognition by maturing T cells of specific MHC products. The CD8 molecule on T cells and class I or class II MHC products on thymic stromal cells are critical to this positive selection. The induction of tolerance can involve clonal deletion direct clonal inactivation, or indirect inactivation through regulatory cells. An example of clonal anergy was demonstrated in a model of neonatal thymectomy, in which cells expressing potentially self-reactive V receptor exist but are unresponsive to receptor-mediated stimuli. Dr. Hodes further characterized the process of clonal deletion as a mechanism of self tolerance. Deletions were characterized in 8 of the 16 commonly expressed T cell receptor V!!
! families. In each case, the tolerizing self ligand consisted of both MHC-encoded and non-MHC-encoded products. It was further demonstrated that efficient clonal deletion is thymus-dependent, since these deletions failed to occur in congenitally athymic nude mice. Expression of the!
!
!T cell receptor during thymocyte development was discussed by Dr. Jeffrey Bluestone (University of Chicago). The expression of specific!!!constant and variable genes was analyzed using both monoclonal antibodies and biochemical characterization. It was demonstrated that specific subsets of!!!expressing cells appear in the developing thymus at markedly different and characteristic sites and time points Thymus organ cultures recapitulated the sequential appearance of different subsets of!!!expressing T cells. Blocking of!!!T cell development in organ culture using monoclonal antibodies specific for these receptors also leads to a decrease in expression of!!!positive cells, indicating that!!!expressing T cells may in turn influence the!!!expressing population. Transgeneic mice expressing the!!!T cell receptor revealed a suppressed development of !!!expressing cells in some but not all mice. The expression of specified genes in different strains of mice has recently been analyzed, and preliminary results suggest that there may be strain-specific selection of the!!!receptor repertoire. V. T Cell Activation.
Junji Yodoi (Kyoto University) analyzed the expression and function of acute T cell leukemia-derived factor (ADF). This factor induces IL2 receptor expression on T cells and is produced by HTLV-1 positive human T cell lines. Dr. Yodoi has now indicated that ADF belongs to the thioredoxin family, a group of dithiol-related reducing co-enzymes. This reducing capacity may be related to the mechanism by which ADF functions. ADF-producing non-lymphoid cells in the thymus exist and may play a role in thymus differentiation.
Phenotypic and functional heterogeneity among CD4+ T cells was discussed by Stephen Shaw (NIH) and Kim Bottomly (Yale University). Dr. Shaw described the identification of naive and memory population of CD4+ T cells based on their respective expression of CD45 RA and CD45 RO. This, in combination with patterns of expression of other adhesion molecules, distinguishes these subsets of naive and previously activated or “memory” T cells. The role of adhesion molecules in T cell activation was further discussed. T cell activation mediated through the TCR/CD3 complex is modulated by co-existent interactions involving ICAM-1 and the LFA-1 receptor. This was discussed as a representative example of integrated signalling of T cells through multiple cell interaction molecules. Dr. Bottomly described the relation of CD45 “isoforms” to the distinction between naive and memory T cells in the mouse as well as rat and man. In addition the expression of different CD45 isoforms was shown to distinguish type I and type 2 T helper cell clones in the mice. Functional studies of co-capping indicated that the T cell receptor and CD4 are selectively associated with low molecular weight isoforms of the CD45 P molecules on activated T cells. Physical approximation of these cell surface molecules ma be involved in signal transduction during T cell activation y
Carl June (Naval Medical Research Institute) detailed recent studies of early signal transduction events in T cell activation. During triggering through the CD3/TCR complex, Jurkat cells or T cell blasts show specific tyrosine phosphorylations as the earliest, detectable changes following anti-CD3 stimulation. After a gap of 10-15 seconds this is followed by a rapid increase in IP3 and intra-cellular [Ca++]. Pretreatment of cells with a potent inhibitor of tryosine kinase completely blocked the anti-CD3 induced increase in intra-cellular [Ca++] without detectably altering the intact cell surface T cell receptor These findings were discussed in terms of alternative models for the roles of phospholipase c and protein tyrosine kinase in T cell activation.
In summary, it was felt by meeting participants that this was the most productive and stimulating immunology meeting to date in the US-Japan Cooperative Program. The past two years have seen an enormous expansion in the ability to analyze lymphoid cell development and activation at both cellular and biochemical levels. This meeting allowed leading investigators in the area of T and B cell development and biology to interact with scientist who are analyzing the biochemical basis for signal transduction and cell activation. The interaction of these two areas allowed extremely vigorous and productive discussions involving all participants and permitted a better focusing of biochemical approaches to some of the most topical questions of immunobiology.
(2) Seminar on “Transgenic Mouse Approach to Cancer Research”
A seminar entitled “Transgenic Mouse Approaches to Cancer Research” was conducted in the area of cancer diagnosis as part of the US-Japan Cooperative Cancer Research Program. The meeting was held on December 6-8, 1989 at the Silverado Country Club in Napa Valley, California and was co-organized by Professor Takehiko Sasazuki of Kyushu University and Dr. Stanley Korsmeyer of Washington University, St. Louis.
The first session “Models to Assess Gene Function” was initiated by Professor M. Katsuki of Tokai University with a discussion of somatically acquired mutations in human ras was further discussed. transgenes. Four independent founder lines bearing a normal genomic human H-ras construct with Gly12, Gln61 developed skin papillomas in 50% of the mice within 18 months. While the transgene was expressed at lower levels than the endogenous mouse Ras, site specific point mutations were confined to the human transgene. Papillomas frequently displayed Val12 while adenocarcinomas, hemangioendotheliomas and some papillomas demonstrated Leu61. Provocatively, the exposure of transgenics to carcinogens of NMU, DMBA or phorbol esters increased the incidence of carcinomas and also altered the amino acids substituted at codon 12. This outstanding model will contribute to our understanding of tissue specific mutagenesis.
Dr. Robert Tepper of Harvard School of Medicine presented evidence of a dramatic role for transgenic IL-4 in the alteration of lymphocyte development and carcinogenesis. Initial studies indicated that transfection of IL-4 producing constructs into tumor lines or even intra-tumor injection of IL-4 generated an eosinophil and mast cell infiltrate that resulted in tumor regression. Transgenics bearing a genomic IL-4 gene under the influence of the immunoglobulin enhancer element demonstrated severe runting and early death. Alternative promoter constructs bearing 1-3 copies of LacO displayed elevated CD8+ CD4- thymocytes while CD4+ CD8- and CD4+ CD8+ cells were decreased.
The topic then refocused on gene ablation with considerations of homologous gene recombination in embryonal stem cells by Drs. Elizabeth Robertson of Columbia University and Randall Johnson of Harvard Medical School. Dr. Robertson, who helped pioneer this field, discussed pivotal aspects in this technical approach. She then turned to a description of the first successful germline transmission of a C-Abl knockout utilizing a promoterless NeoR construct. The heterozygotes appear to be normal and have been mated to produce the homozygous state. Randall Johnson concentrated on a double drug selection protocol utilizing G418 and Gancyclovir co-selection to select for homologous recombinants that had inserted a NeoR gene but have not integrated the flanking thymidine kinase gene. He has successfully ablated Adipsin and a P2 which are not highly expressed in ES cells, yet no germline transmission has been noted as yet. The improvement in homologous integration with the addition of Gancyclovir selection is more frequently in the range of 10 rather than 1000 fold. Professor Katsuki concluded the consideration of gene interference with his myelin basic protein anti-sense mice. A marked decrease in CNS MBP is noted and mice develop the classic shiverer phenotype. However, the anti-sense RNA is approximately 10% of endogenous MBP RNA yet sense RNA is reduced in transgenics providing open avenues concerning the mechanism.
The evening Session II considered Class II major histocompatibility transgenic mice. Dr. Kikutani of Osaka University presented his effort of introducing missing Class II molecules into Non-obese Diabetic (NOD) mice. He found that the IE !
!!gene completely prevented insulinitis. However, the addition of the IAb gene also decreased the incidence of insulinitis. Thymectomy of NOD with transfer of IE+T cells also prevents insulinitis. d Professor Yamamura of Kumamoto University also demonstrated that E!!d would prevent diabetes yet backcross experiments suggested the role of a 2nd gene. NOD-Ak-Ak mice were generated that displayed a 75% protection from insulinitis that was parallel to their surface Class II. The implication of these workers is that sufficient levels of the appropriate Class II will enable deletion of autoreactive T cell clones. Professor Sasazuki concluded the session with fascinating data concerning a dominant antigen non-responsive model. Human DQW6A which is suspected of being a suppressor gene for streptococcal response was microinjected and transgenics proved tolerant to DQW6 and acquire responsiveness to strep antigen. Curiously introduction of DQ into high responder strains did not alter their response and these animals displayed DR restricted but not DQ restricted cells. Importantly, autologous but not allogeneic antigen presenting cells and antigen resulted in the generation of CD8 cells implying a need for a Class I gene in the development of suppression. This transgenic model provides a definitive system to identify such genes.
Dr. Fred Alt of Columbia University initiated Session III with a discussion of his work detailing the relationship of N-myc, L-myc, C-myc expression in early lymphoid development. Transgenic mice bearing a high copy number of Eµ-N myc mostly developed pre B tumors at 20 wks of age in which C-myc was down regulated, however V-Abl induced cell lines from Eµ-Nmyc animals do not down regulate C-myc. This strongly suggests a stage of differentiation dependence upon C-myc expression. In contrast Eµ-Lmyc mice demonstrate much slower tumor development. Dr. Stanley Korsmeyer followed with a presentation of the Bcl-2-Ig transgenic mouse model. The presence of the Ig enhancer insured widespread B + T cell lineage overexpression and overproduction of the transgenic Bcl-2. Despite this a follicular lymphoproliferation of small resting yet responsive IgM/IgD B cells occurred. This select B cell subset demonstrated an extended survival. Similarly introduction of Bcl-2 expression vectors results in death sparing for certain IL-3, GMCSF, IL-4 dependent cell lines. Over time the Bcl-2-Ig transgenic mice demonstrated tumor progression to life threatening monoclonal high grade lymphomas.
Dr. Yuichi Obata of Aichi Cancer Center opened the discussion of T cell neoplasia with his TL(thymus leukemia) transgenic model utilizing a 9.6 Kb Tla-3 genomic fragment that expressed within thymocytes of C3H CTL non-expressing mice. Other mice Tg Con.3 within an H-2 Kb promoter and T3b chimeric gene developed a 66% incidence of TL+, Thy1+, CD4-, CD8-, rearranged T cell lymphomas. These mice are an important model for examining the role of TL in pathway, T cell selection and T cell differentiation.
Dr. Glen Evans of the Salk Institute presented their approach to targeting genes to the T cell lineage. He has noted that the Ib-II intron of Thy-1 possesses a cis acting motif binding a trans-acting protein that appears to effect constitutive and brain specific expression while the 3’ UT has thymus and kidney specificity. They have been able to utilize the Thy-1 promoter plus the TCR enhancer to obtain expression in T cells in transgenics. Finally, he has used Thy-1 or CD4 promoters on a Ricin A chain construct whose expression has resulted in pre or peri-natal death.
Dr. Roger Perlmutter of the University of Washington concluded this session with a presentation concerning Lck, tyrosine kinase transgenic mice. CD4 and CD8 have conserved Cys containing motifs that interact with Lck Cys residues 23 and 25. A 5’ and 3’ promoter of Lck confers peripheral T cell or thymocyte activity respectively. Mice with a 3’ promoter with SV40 T Ag demonstrate decreased double positive cells and develop stromal dependent thymoblasts over time. Introduction of 5’ Lck promoter with substituted 505F construct results in high expression animals with a decrease of CD3+,CD4+, CD8+ and emergence of CD4-, CD8- tumors. In contrast low expressing lines result in peripheral T cells that don't express Lck and appear refractory to antigen-like activation signals. This model is providing crucial insights into the role of Lck in lymphocyte selection and activation.
The Final Session (IV) of the meeting dealt with transgenic models of viral disease. Dr. Kenichi Yamamura presented their model of human hepatitis B virus complete genome transgenic mice. Mice produced HBsAg and infectious Dane particles. These mice possess ' primed B cells to HBS and HBC Ag, yet remain tolerant perhaps reflecting T cell deletion. Despite this no pathogenic effects are noted. Dr. Frank Chisari of Scripps Clinic followed confirming that their complete HBV genome mice replicate virus that goes through the entire life cycle, but also don't develop hepatitis. However, he has expanded this model by producing HBsAg subgenomic transgenic mice. Deregulated envelope results in ER accumulation, injury, accumulation of classic “groundglass” hepatocytes, hyperplastic nodules and 100% incidence of multifocal hepatoma by 2 yrs of age. Curiously, these animals don’t display chronic active hepatitis or cirrhosis implicating a separate immune pathology for these events. This transgenic model indicates that deregulated HBsAg is the key to hepatoma, a remarkable insight that was only obtainable with a transgenic model.
Dr. Y. Iwakura of the University of Tokyo presented data on HTLV-I Px (Tax) transgenic mice. These mice demonstrate growth retardation, thymic atrophy and Tax expression that decreases over time in eye, spleen, heart and brain. A wide range of tumors are seen including B cell lymphomas, adenocarcinomas, schwannomas, and fibrosarcomas. Curiously 20-30% of mice develop an arthritis similar to that seen in some patients with ATL.
Dr. Gilbert Jay of the American Red Cross Laboratories concluded the meeting with a presentation of HIV LTR- Tat transgenic mice. Dr. Jay has discovered that the skin is the only site of expression and is inducible by 15 seconds of UV light. Within 15-18 mos. mice develop skin tumors of the dermis with endothelial cell involvement, RBC, and hyperplastic epidermis that resemble Kaposi’s sarcoma. Only males express tat in the skin to give hyperplasia and only males develop tumors; yet, tumors don’t express tat suggesting a cytokine component to neoplasia. Provocatively, the mice are demonstrating hepatomas as a secondary malignancy and AIDS patients are now showing an increase in hepatocellular carcmoma.
The US-Japan Seminar was of exceptionally high and uniform scientific quality in which an enormous amount of unpublished and instructive transgenic model data was presented. This resulted in wide ranging open discussions, stimulated the exchange of reagents and was widely regarded as a success.
(3) Seminaron “Cell Differentiatioin and Cancer”
A meeting on Cell Differentiation within the series of US-Japan Cancer Research conferences was held January 18-20, 1990 on Kauai. The meeting was organized by Tasuku Honjo, Kyoto University School of Medicine, and Igor B. Dawid, National Institute of Child Health and Human Development, NIH. In spite of its small size the workshop dealt with a broad range of topics, limited only by the decision to restrict the discussion to vertebrate systems. If a common theme were to be named it was the uniting influence of molecular concepts and techniques, and the growing application of the well-accepted view that developmental questions can be approached by focusing on cellular and molecular aspects in differentiation. Under this aegis topics as diverse as immunoglobulin gene rearrangement and induction in frog embryos could be viewed from a common perspective.
Cellular Adhesion.
Adhesive properties are critical to cell differentiation and tissue organization. M. Takeichi (Kyoto) described some of the extensive studies of his laboratory on the family of calcium-dependent cell adhesion proteins named cadherins. A major tool for functional studies is based on the fact that L cells do not express cadherins and consequently are poorly adhesive, but become adhesive upon expression of cadherins from suitable constructs. This approach allowed the demonstration of the homophillic nature of the interaction where cells expressing the same cadherin type adhere preferentially. More recently, chimeric cadherin molecules were studied with the aim to localize the region(s) responsible for recognition. P-cadherin-specific binding was still exhibited by cells expressing a hybrid molecule containing the amino-terminal 113 residues of P-cadherin attached to the E-cadherin body; two residues in this amino-terminal region were found most important in conferring specificity.
How do cadherins mechanically attach a cell to another? Takeichi suggests that the cytoplasmic domain of cadherins is linked to the cytoskeleton, specifically to actin filaments. The attachment does not appear to be a direct one, however. Cadherin antibodies coprecipitate two proteins, neither of which is actin; one of these, a 94 kD protein, is a strong candidate for an intracellular cadherin-binding factor.
U. Rutishauser (Cleveland) discussed cell adhesion mediated by N-CAM, an important adhesion molecule distinct from cadherins. Yet the main point of this presentation concerned the role of varying amounts of polysialic acid known to be attached to N-CAM. Convincing in vitro studies show that the presence of polysialic acid interferes with strong cell-cell and cell-substrate adhesion, whether mediated by N-CAM or by other molecules. This appears also to be true in vivo, as increases in polysialic acid correlate with loss of adhesion, for example in the initiation of cell migration. Rutishauser’s exciting if as yet hypothetical conclusion is that polysialic acid is a molecule regulating adhesion in its own right, by utilyzing is large hydrated volume to occupy available surface space. Its attachment to N-CAM may be incidental, simply using it as an abundant surface carrier on which to alight. If proven, this interpretation will significantly expand our understanding of cellular interactions.
Growth Factors and Embryonic Induction.
Moving from the outside of the cell inward we turn to a series of studies that deal with the role of growth factors in embryogenesis. The earliest interactions between cells that are critical in establishing the polarity of the embryo and initial tissue differentiation have attracted much attention for many years. It has been established in the amphibian embryo that growth factor-like substances are critical in the differentiation of the mesoderm. In addition to fibroblast growth factor (FGF), factors belonging to the large transforming growth factor-beta (TGF-!
!!) family were known to be involved. At this meeting D. Melton (Cambridge MA) presented evidence that implicates the distant TGF-!!!relative activin A as the major mesoderm inducer, a conclusion that has also been reached independently by Asashima (Yokohama) and J. C. Smith (London). As reported by Melton, activin A is a particularly effective inducer that can organize ectodermal explants into recognizable tissues, e.g. eyes. In addition, TGF-!!!3 acts as an effective mesoderm inducer while TGF-!!!1 and 5 are inactive and TGF-!!!2 is less active (I. Dawid, Bethesda). It appears that several growth factors may be involved in induction, but the important role of activin A is now clear. What is the immediate consequence of growth factor action in the embryo? The answer is, no doubt, complex, but it is becoming clear that a major consequence is the localized and graded activation of homeobox genes. This was discussed by Melton for Xhox3, Dawid for Mix.1 (as originally reported by F. Rosa), and by E. De Robertis (Los Angeles) for XlHbox1 and XlHbox6. Xhox3 is normally expressed in a posterior (high)-to-anterior (low) gradient, and is involved in the determination of polarity since artificial anterior expression suppresses head formation (Melton). In this context it is interesting that both Xhox3 and XlHbox6 genes are induced differentiatlly by different growth factors, yielding a high level in response to FGF but a low level with mesoderm inducing factor (MIF, now presumed to be activin A). In contrast, XlHbox1, an anterior gene, is strongly induced by MIF and weakly by FGF. XlHbox1 is not only involved in the development of axial polarity in the embryo as a whole but also during organogenesis, i.e., in the limb bud and in the chicken feather bud (De Robertis). The polarity-determining effect of inducing factors may thus be mediated by the differential and graded activation of homeobox genes. In a comparable though different way the Mix.1 homeobox gene is activated primarily in future endoderm, leading to the heretofore unconventional view that, as in the case of mesoderm endoderm determination may depend on induction (Dawid).
A fourth presentation that dealt with amphibian embryos was devoted to the analysis of the oncogene int-1, the vertebrate homolog of the Drosophila segmentation gene wingless. The gene product appears to be secreted and may act as a growth-factor-like substance. A. McMahon (Nutley, NJ) showed that overexpression of the int-1 product in Xenopus embryos leads to duplication of the neural plate. In the mouse, where int-1 was discovered, the question arose how this molecule could affect the early stage of neural plate formation even though it was not expressed until later in normal development. The apparent answer comes from McMahon’s observation that at least seven int-1-related genes occur in the mouse, some of which are expressed very early. Apparently these related genes function in early embryogenesis and the effect of int-1 overexpression may be a result of cross-reaction.
Intracellular Signal Processing.
How is a surface signal transmitted into the cell to affect its differentiation? In one of two contributions to this large subject M. Mikoshiba (Osaka) reported on the isolation and cloning of the receptor of inositol-tris phosphate (IP3). The receptor was isolated as a very large (313 kD) protein from Purkinje cells whose abundance is affected by certain neurological mutations (stagger, nervous). The production of antibodies and cloning of the gene showed that the molecule is expressed in different parts of the brain and in other tissues, and provided a model for the shape of this remarkably large protein that functions as an IP3-sensitive calcium channel.
In a different approach to signal processing M. Oishi (Tokyo) noted in his studies of two differentiation factors that differentiation in erythroid cells correlated with dephosphorylation of one of them. This suggested to Oishi the possibility that protein phosphatase activity may be a general, though not necessarily universal, correlate of differentiation in various cells. This idea was tested with the help of an antibiotic, Herbimycin A, which is known to reduce tyrosine phosphorylation and to counteract the effect of certain transforming genes (e.g., src) on cells. Oishi reported that Herbimycin A does in fact induce the differentiation of mouse erythroleukemia and F9 embryonal carcinoma cells. Herbimycin A, and other compounds with comparable activity, are likely to prove valuable tools in elucidating signal cascade pathways during certain differentiation events.
The Immune System and Viruses as Developmental Models.
A key question in analyzing the immune system is the mechanism of immunoglobulin gene rearrangement which involves recombination dependent on a heptamer/nonamer nucleotide sequence motif. M. Kawaichi (Kyoto) reported experiments in which a vector containing the recombinogenic sequences was constructed in such a way that the recombination event led to the forced expression of an easily detected (and sorted) surface marker. This assay allowed the recovery of cell lines that support active recombination. Using these lines Kawaichi succeeded in purifying and cloning a 63 kD protein that binds specifically to the recombinogenic sequences; this protein is distinct from the recombinase gene, Rag, recently reported by Baltimore and his colleagues in Cambridge, MA. Significantly, the 63 kD protein sequence shows homology to the integrase group of proteins, i.e., the Flp gene or the int gene of phage lambda.
Viewed from a different vantage point, the immune system stands out in that it utilizes pretty much every mechanism of expression and regulation that can be found in eukaryotes. This point was exemplified by the presentation of T. Honjo (Kyoto). In addition to a tantalizing mention of preliminary observations suggesting the occurance of trans-splicing in switch recombination, Honjo reported on elegant studies showing that the HTLV-1 rex gene product (p27) increases IL-2 expression by stabilizing IL-2 mRNA
The function of the HTLV-1 control region was the subject of the presentation of M. Hatanaka (Kyoto). The regulatory region named pX yields three products, tax, rex (p27), and p21; tax is the transforming gene of HTLV-1. Hatanaka and his colleagues have generated constructs that allow the separate express. ion of each of these products so that their functions could be studied independently. They showed that the rex protein is localized in the nucleolus; its function relates to splicing in that it may be required for the production of unspliced, genomic RNA. Detailed studies of the nucleolar targeting sequence has localized it to the amino terminal 19 residues of the rex product.
Transcription Factors and Gene Regulation.
Having arrived in the nucleus in our journey into the cell we come to a major topic of the meeting, the control of gene expression and the analysis of transcription factors. IC Ozato (Bethesda) discussed the control of expression of MHC class I genes; while most tissues express MHC antigens these genes are regulated developmentally and are modulated by certain cytokines, e.g., interferons. The MHC class I gene upstream control region contains a segment required for transcription and an overlapping segment required for interferon stimulation. Ozato and her colleagues have identified and cloned three transcription factors that bind specifically to these regions. One, named RIIBP, belongs to the hormone receptor class of zinc finger proteins, is expressed ubiquitously in the mouse, and is highly conserved in evolution. A second factor that has been cloned independently by several laboratories, is a C2H2 type zinc finger protein that binds to the control region of many genes; it may be a transcription factor of broad specificity. The third factor binds specifically to the interferon consensus sequence and is believed to be required in mediating the interferon effect on MHC transcription. Together with two related factors isolated by Taniguchi and colleagues (Osaka) it forms a new structural class of DNA binding proteins showing no homology to previously established motifs like homeodomains, zinc fingers or leucine zippers.
T. Watanabe (Fukuoka) reported on his studies of the human immunoglobulin heavy chain enhancer. Several transcription factor-binding motifs occur in the intron between VDJ and C regions. Watanabe’s studies identified a new enhancer region that is distinct from the octamer and other motifs known to be involved in heavy chain transcription. A binding sequence consensus was generated and shown to be functional. Using this sequence a binding protein has been isolated and the corresponding cDNA cloned. The sequence, will not doubt, be available soon.
M. Karin (San Diego) reported on his studies on growth hormone (GH) expression in the pituitary. Isolation of a specific factor in GH transcription (GHF-1, also called Pit-1) led to the cloning of the corresponding gene and its identification as a homeodomain protein. In addition to the homeodomain GHF-1 contains a highly conserved structural motif called the POU domain. Karin reported primarily on work that aims to understand the regulation of the GHF-1 gene itself - clearly, we are ready to look at what regulates the regulators. GHF-1 expression arises just after the pituitary rudiment is established. After several days GHF-1 protein appears and GH expression follows closely thereafter. The GHF-1 gene has been analyzed in some detail; interestingly, its upstream region contains a binding site for GHF-1 in addition to two CREB sites that mediate its responsiveness to cAMP, the second messenger for the hypothalamic hormone CH releasing factor.
Transgenic Mice.
Beyond molecular studies, i.e., analyses of transcription factors, gene regulation and function is best studied by genetics. Yet, classical genetics of vertebrate organisms remains limited in scope, leaving a void that transgenic animals must fill. Overexpression of certain genes creating dominant phenotypes is the approach discussed by H. Westphal (Bethesda). Westphal and his colleagues have focused on the specific expression of viral tumor antigens in the lens. With SV40 T antigen they note that the timing of oncogene expression influences tumor progression. If T antigen is made prior to lens fiber differentiation, rapid neoplastic growth of undifferentiated cells follows. In contrast, onset of T antigen expression after fiber differentiation results in slow tumors consisting of differentiated cells. The immortalizing T antigen of polyoma virus (Py) disrupts the normal process of lens differentiation but does not elicit neoplastic growth, possibly because differentiation factors can counteract Py T antigen mediated proliferation.
The other approach using genetically manipulated mice aims to eliminate (“knock-out”) specific genes by homologous recombination in ES cells, followed by the production of chimeric mice that carry ES cell derivatives in their germ line. This approach, which has only recently come to fruition in a limited number of examples, was discussed by M. Capecchi (Salt Lake City). An ingeneous selection procedure allows Capecchi to produce ES cells with the desired recombination events at very high efficiencies (between 5 and 80% of recovered cells in different experiments). The efficiency of homologous recombination in these cells did not depend on the length of the insert introduced to interrupt the resident gene, at least up to 4 kb. After introduction of ES into blastocysts a variable but substantial fraction (60-80%) of progeny mice were chimeric, and all chimeric mice that were recovered were germ line chimeras. After subjecting ES to drug selection (i.e., G418 and gancyclovir) the efficiency of generating germ line chimeras was reduced 2-10 fold. Using this approach Capecchi and his colleagues have generated mice that carry disruptions in several interesting genes, including many members of the Hox-1 and Hox-2 homeodomain gene clusters as well as the protooncogenes int-1 and int-2. Breeding and Southern blotting has established that the int-1 chimera is a germ line chimera. It will be most exciting to learn if the other chimeric animals are also germ line chimeras so that the phenotypes associated with homozygosing the mutant alleles can be evaluated.
SEMINAR AGENDA AND PARTICIPANTS
(1) SEMINAR ON “LYMPHOCYTE DEVELOPMENT AND ACTIVATION”
Rockville, Maryland, February 19-21, 1990
AGENDA