B21. Regulation of the B"/PR72/PR130 subunits of protein phosphatase 2A by Ca2+ and m-calpain-mediated limited proteolysis V. JANSSENS, I. STEVENS, E. MARTENS, C. VAN HOOF, W. MERLEVEDE and J. GORIS (Afdeling Biochemie, Faculteit Geneeskunde, Katholieke Universiteit Leuven, Herestraat 49, B-3000 Leuven, Belgium) Type 2A protein phosphatases (PP2A) comprise a major phosphoserine/threonine-specific phosphatase family implicated in many cellular processes (Janssens & Goris, 2001). PP2A holoenzymes consist of a catalytic C subunit and a scaffolding A subunit to which a regulatory B subunit can be bound. The B subunits specify the catalytic activity, substrate selectivity and/or subcellular localization of the trimeric enzymes, and therefore their physiological function. To date, four classes of B subunits have been described, denoted B, B', B" and B"' (Janssens & Goris, 2001). The B" family members PR72 and PR130 are splice variants of a single gene and have identical C-termini (Hendrix et al., 1993). This common C-terminal part contains two well-conserved EF-hand Ca2+ binding motifs and a PEST-domain. Both EF-hand domains bind 45Ca2+ in an overlay assay, but with different affinities. Addition of Ca2+ to a purified PR72-containing PP2A trimer (PP2AT72) stimulates the catalytic activity of PP2A, suggesting that Ca2+ could regulate this holoenzyme through binding to these EF-hands. Moreover, PP2AT72 co-purifies with a Ca2+-dependent cysteine protease, that was identified as m-calpain. PR72, as well as PR130, are m-calpain substrates in vitro, whereas the A, C and B/PR55 subunits are resistant to m-calpain-mediated proteolysis. Proteolysis of PR72 and PR130 is limited and eventually results in the formation of a common m-calpain-resistant fragment of 45 kDa, denoted PR45. The N- and C-terminal borders of the truncations could be determined by specific anti-peptide antibodies and by degradation of several in vitro translated N- and C-truncated forms of PR72/PR130 of different lengths. PR45 still interacts with the A subunit in a yeast two-hybrid assay, and can therefore still be considered as a PP2A subunit. Cleavage of PP2AT72 by m-calpain results in a trimeric enzyme with distinct biochemical in vitro characteristics, of which the high sensitivity to protamine stimulation is the most striking. Moreover, PR45 fused to EGFP has a different subcellular localization, compared to PR72-EGFP and PR130-EGFP fusion proteins. Together, the data suggest regulation of PR72/PR130-containing PP2A trimers by a Ca2+-signal, either directly through binding of Ca2+ to PR72/PR130, or indirectly through activation of a Ca2+-dependent protease that cleaves PR72/PR130 into a smaller fragment, resulting in a PP2A holoenzyme with altered biochemical characteristics and different subcellular localization. This work was supported by the E.U. Commission (QLK 3-2000-01038). V. J. and C. V. H. are post-doctoral fellows of the F.W.O.-Vlaanderen. I. S. is a post-doctoral fellow of the K.U.Leuven Onderzoeksfonds. References HENDRIX, P., MAYER-JAEKEL, R.E., CRON, P., GORIS, J., HOFSTEENGE, J., MERLEVEDE, W. & HEMMINGS, B.A. (1993) J. Biol. Chem. 268, 15267-15276. JANSSENS, V. & GORIS, J. (2001) Biochem. J. 353, 417-439.

B21-B22. Genetic identification of distinct loci controlling mammary tumor multiplicity, latency and aggressiveness J.-F. LAES1, D. STIEBER1, X. QUAN1, J. RUSSO2, D. WEDEKIND3, W. COPPIETERS4, M. GEORGES4, J. SZPIRER1 and C. SZPIRER1 (1Université Libre de Bruxelles, IBMM, B-6041 Gosselies, Belgium; 2Fox Chase Cancer Center, Breast Cancer Research Laboratory, Philadelphia, PA 19111, USA; 3Medizinische Hochschule Hannover, MHH Zentrales Tierlabor, D-30623 Hannover, Germany; 4Université de Liège, Département de Génétique, B-4000 Liège, Belgium) The rat is considered to provide an excellent model of human breast cancer and thus appears to be appropriate for dissecting the genetic basis of susceptibility to mammary cancer. Previous studies based on crosses involving the susceptible strain WF (crossed with the resistant strains COP or WKY), and focused on tumor multiplicity as the susceptibility phenotype led to the genetic identification of eight loci controlling this trait. The present study aims at determining whether other loci can be identified, on the one hand, by analyzing crosses derived from another susceptible strain and, on the other hand, by including other phenotypes. A backcross was generated between the susceptible strain SPRD/Cu3 and the resistant strain WKY. The female progeny was genotyped using microsatellite markers covering all rat autosomes, treated with a single dose of DMBA and phenotyped with respect to tumor latency, tumor multiplicity and tumor aggressiveness. Five loci controlling mammary tumor development were detected, only one of which colocalizes with a locus previously identified. New loci have thus been uncovered by analyzing a new pair of susceptible and resistant strains. Two distinct, newly identified loci control tumor latency and tumor aggressiveness. This study demonstrates that different genetic determinants control the tumor multiplicity, tumor latency and tumor aggressiveness phenotypes.

B22. Elongation Factor Like-1, a cytoplasmic GTPase homologous to the ribosomal translocases EF-G/EF-2, is required for ribosomes synthesis D.L.J. LAFONTAINE (F.N.R.S., Université Libre de Bruxelles, Institut de Biologie et de Médecine Moléculaires, Gosselies, Belgium) Ribosome biogenesis involves the synthesis, maturation, and assembly of ribosomal RNAs (rRNAs) and ribosomal proteins (r-proteins). In eukaryotes, three out of the four rRNAs (5.8S, 18S, and 25S rRNAs) are interspersed with noncoding sequences in a single, large precursor from which they are released following a complex processing pathway involving both endo- and exonucleolytic digestions. Concomitantly, pre-rRNAs are extensively modified by base and ribose methylation and formation of pseudouridine residues and are bound by the r-proteins. R-proteins are synthesized in the cytoplasm and need to be specifically targeted to the nucleolus where most steps of ribosomal assembly occur. In the course of assembly, pre-ribosomes are exported to the cytoplasm and undergo late and ill-defined steps of maturation before being committed to translation. Altogether, this represents a major metabolic activity for the cells. Deletion of Elongation Factor Like-1 (Efl1p), a cytoplasmic GTPase homologous to the ribosomal translocases EF-G/EF-2, results in nucle(ol)ar pre-rRNA processing and pre-60S subunits export defects. Efl1p interacts genetically with Tif6p, a nucle(ol)ar protein stably associated with pre-60S subunits and required for their synthesis and nuclear exit. In the absence of Efl1p, 50% of Tif6p is relocated to the cytoplasm. In vitro, the GTPase activity of Efl1p is stimulated by 60S, and Efl1p promotes the dissociation of Tif6p-60S complexes. We propose that Tif6p binds to the pre-60S subunits in the nucle(ol)us and escorts them to the cytoplasm where the GTPase activity of Efl1p triggers a late structural rearrangement which facilitates the release of Tif6p and its recycling to the nucle(ol)us. Passage of large ribonucleoprotein particles (RNPs), such as pre-ribosomes, through the nuclear pore complexes is likely to be preceded and followed by major structural rearrangements and the concomitant release of RNP-associated proteins, including processing and assembly factors. This presumably requires a considerable input of energy. In mature ribosomes, translocases act as force-generating motor proteins utilizing the energy from GTP hydrolysis to promote a series of synchronized molecular movements. By analogy, we propose that the energy provided by Efl1p-mediated GTP-hydrolysis is used for a late cytoplasmic structural rearrangement in the pre-ribosomes. The overall homology of Efl1p to ribosomal translocases further suggests that their binding sites on the ribosome are conserved. If so, the association of Efl1p with cytoplasmic pre-ribosomes could provide a quality control step to check that the binding sites for elongation factors are properly shaped on mature subunits before they engage in translation. D.L.J.L. is supported by the F.N.R.S., Fonds Émile Defay and Banque Nationale de Belgique Reference LAFONTAINE, D.L.J. & TOLLERVEY, D. (2001) Nature Reviews Mol. Cell. Biol. 2, 514-520.

B22-B23. XNAP, a conserved ankyrin repeat-containing protein with a role in the Notch pathway during Xenopus primary neurogenesis K. LAHAYE, S. KRICHA and E. J. BELLEFROID (Université libre de Bruxelles, IBMM, rue des professeurs Jeener et Brachet 12, B-6041 Gosselies, Belgium) The Notch signaling pathway is an evolutionarily highly conserved mechanism for cell-cell communication that is important for cellular differentiation in various developmental processes. Notch signal transduction is mediated by proteolysis of the receptor and translocation of the intracellular domain into the nucleus, where it function as a positive regulator of HES gene expression after binding to the DNA-binding protein Su(H). In Xenopus, during primary neurogenesis, Notch inhibits neuronal differentiation induced by proneural gene expression and directs cells to an undifferentiated fate. Here we investigate the regulation and function of the conserved gene XNAP, which is a member of the Delta-Notch synexpression group in Xenopus (Gawantka et al., 1998). XNAP encodes a small protein with two C-terminal tandem ankyrin repeats which is expressed in the neurectoderm and in the presomitic mesoderm in a pattern that resembles that of other component of the Notch pathway. When a myc-tag form of XNAP is overexpressed in Xenopus or Hela cells, XNAP protein is detected both in the nucleus and the cytoplasm. In embryos and in animal cap assays, XNAP expression is activated, perhaps directly, by the Notch pathway and this activation appears to be Su(H) dependent. Overexpression of XNAP in embryos decreases HES gene expression, which leads to an increase in the number of primary neurons that form within the domains of the neural plate where neurogenesis normally occurs. In contrast, overexpression of a XNAP deletion mutant that lacks the region N-terminal to the ankyrin repeats decreases the number of N-tubulin positive cells and thus behave as a dominant negative mutant, supporting the idea that XNAP interacts with other proteins and that different regions of the protein may be involved in these interactions. In cultured cells, XNAP overexpression interferes with ICD activation of a Notch regulated reporter gene, suggesting that XNAP is acting on downstream events of Notch signaling involving the association of ICD with Su(H) and activation of target genes. In conclusion, our data indicate that XNAP is a novel target of the Notch pathway that may, in a feed-back loop, modulate its activity. Recent data have indicated that XNAP directly binds to ICD and Su(H) and that it promotes the loss of ICD (Lamar et al., 2001). We are currently analysing how XNAP promotes the degradation of Notch ICD and whether XNAP binding to ICD and Su(H) affects the binding of other proteins that modulate Notch mediated transcription. This work was supported by a grant from the FRSM (contract n°3.455.01). L. Lahaye is recipient of a FRIA fellowship. References GAWANTKA, V., POLLET, N., DELIUS, H., VINGRON, M., PFISTER, R., NITSCH, R., BLUMENSTOCK, C. & NIEHRS, C. (1998) Mech. Dev. 77, 95-141. LAMAR, E., DEBLANDRE, G., WETTSTEIN, D., GAWANTKA, V., POLLET, N., NIEHRS, C. & KINTNER, C. (2001) Genes Dev. 15, 1888-1899.

B23. The hoxa-1/hoxa-2 1,25 kb enhancer contains an exonic Hox/Pbx responsive element that is critical for its function X. LAMPE, C. MATIS, J. PICARD and R. REZSOHAZY (Unité de Génétique du Développement, Université Catholique de Louvain, B-1200 Bruxelles, Belgium. E-mail : rene.rezsohazy@gede.ucl.ac.be) The hox genes encode transcription factors that play key roles for the developmental program, such as patterning the embryonic antero-posterior axis (reviewed in Capecchi, 1997). In mammals, they are organized in four chromosomic complexes. The position of the hox genes along the chromosomal clusters correlates with their expression pattern: genes located at the 3' end of a cluster are expressed earlier and more anteriorly in the embryo than those found at the 5' end. This property is called colinearity (reviewed in Duboule, 1998). The control of colinearity is poorly understood but it may be partly explained by self and cross regulations among -at least- the 3' hox genes (ManzanareS, 2001). Previous works have shown that a 1,25 kb region overlapping with the hoxa2 coding region controls the expression of hoxa-1 and/or hoxa-2 in the rhombomere 4 (r4) territory (Frash et al., 1995). However, until now, nothing was known about the cis- regulatory elements this region contains and the trans activating factors acting through it. In r4, hoxb-1 is positively regulated by Hoxa-1 and hoxb-2 expression is controlled by Hoxb-1(reviewed in Nonchev, 1997). Furthermore, both hoxb-1 and hoxb-2 show positive self regulations. To further unravel the regulatory relationships between the hox genes involved in governing the r4 fate, we analysed the hoxa-1/hoxa-2 1,25 kb regulatory region for the presence of possible hox responsive elements. Here we show that a luciferase reporter gene placed under the control the 1,25 kb region is strongly upregulated upon cotransfection of Hoxa-2, Pbx1a and Prep1 expression vectors in COS7 cells. By deletion analysis and examination of the 1.25 kb sequence, we pointed out a Hox/Pbx responsive element (HPRE-1 : TGATTGATGA) that is crucial for this upregulation. HPRE-1 lies in the first exon of hoxa-2, 33 pb downstream of a potential binding site for Prep1/Meis1 proteins. Cotransfection experiments performed in teratocarcinoma P19 cells further emphazised the importance of this HPRE 1 for the activity of the 1,25 kb regulatory region. Preliminary gel retardation experiments show that Hox proteins specifically bind double stranded oligonucleotides bearing the HPRE1 site. This work was supported by the Belgian National Foundation for Scientific Research, and the "Fonds Spéciaux de Recherche" of the Université Catholique de Louvain. XL and CM hold a FRIA fellowship. References CAPECCHI, M. R. (1997) Cold Spring Harbor Symposia on Quantitative Biology.,Vol LXII. DUBOULE, D. (1998) Curr. Opin. Genet. Dev. 8, 514-518. FRASH, M. et al. (1995) Development 121, 957-974. MANZANARES, M. (2001) Development 128, 3595-3607. NONCHEV, S. (1997) Cold Spring Harbor Symposia on Quantitative Biology, Vol LXII.

B23-B24. Trypanosoma brucei: analysis of chromatin structure by immunoprecipitation with anti-H4 antibodies L. LECORDIER1, M. NAVARRO2, J.A. GARCIA-SALCEDO1, E. PAYS1 and L. VANHAMME1 (1Laboratoire de Parasitologie Moléculaire, ULB-IBMM Gosselies, Belgium ; 2Manchester University, Manchester, UK) The protozoan parasite Trypanosoma brucei is the causative agent of sleeping sickness in humans and nagana in cattle. During its life cycle, the trypanosome has to adapt to very different environments, the tse tse fly and the bloodstream of its mammalian host. This implies activation/inactivation of subsets of specific genes, in particular the genes encoding the major surface proteins, the Procyclin in the fly and the VSG in bloodstream. In its mammalian host, the parasite escapes the immune response by antigenic variation, by the periodic replacement of the VSG coat. This VSG gene is picked up in a repertoire of a thousand and expressed from one out of several dozens of "expression sites". In addition, the trypanosome possesses several unusual features of transcription. First, the transcription is polycistronic and appears to be regulated at the elongation level rather than at initiation step. Second, no polII promoters have been identified. Third, the expression sites (ES) containing the VSG genes are transcribed by a highly processive polI type polymerase, which starts in all ES but reaches the telomeric VSG gene in only one of them. The role of chromatin structure in transcriptional control has been widely demonstrated in other eukaryotes. In order to investigate the involvement of this structure in the control of gene expression in trypanosomes, we compared chromatin of trypanosomes bearing tagged active and inactive expression sites, and polI and polII transcribed genes. Micrococcal nuclease digested chromatin was immunoprecipited using anti-histone H4 and anti-acetylated histone H4 antibodies. Immunoprecipitated DNA was hybridised with probes for the inactive and active ES and several pollII transcribed genes. No obvious correlation could be drawn between H4 acetylation and ES activity. The inactive ES were shown to be enriched in acetylated H4 and the active ES were found to be hypersensitive to micrococcal nuclease digestion, from the promoter region to the telomeric VSG gene. The nucleosomal structure of all ES telomeres appeared to be very loose. Interestingly, with the exception of Actin encoding genes, the polII transcribed genes seemed to be largely deficient in acetylated H4. These results will be discussed in the context of polycistronic organization of transcription units in trypanosomes.

B24. Purification and analysis of factors which bind to a silencer element of the alpha-foetoprotein promoter P. LIENARD, C. SZPIRER and J. SZPIRER (Laboratoire de biologie du développement, Institut de biologie et médecine moléculaires, Université Libre de Bruxelles) Alpha-foetoprotein (AFP) is a serum protein, the expression of which is developmentally regulated. AFP is produced at high level in the fetal liver and is repressed after birth. Its re-expression in the adult is associated with liver pathology and especially hepatomas. A region of the promoter, located between &endash;80pb and &endash;38pb is involved in the control of the AFP gene expression during development (it acts as an enhancer in expressing cells and as a negative regulatory region in non-expressing cells). We report that the repressive activity of this region can be correlated with the presence of i) a DNAse hypersensitive site at &endash;40pb in footprinting experiments and ii) a band shift generated by non-expressing cell nuclear extracts in EMSA experiments made with a probe centred on position &endash;40pb. In an attempt to purify the protein responsible for this shift, we identified a protein co-migrating with it: CaBP1 (calcium binding protein 1), a protein of the Protein Disulfide Isomerase family which may be implicated in the redox regulation of the negative factor. We showed indeed that this negative factor has to be reduced to bind DNA. Moreover, we showed that an oxidative stress prior to the protein extraction induces an intensification of the shift; this dual role of in vitro and in vivo oxidation is similar to that observed with NF-KB. We verified that in vitro translated CaBP1 is able to inhibit DNA binding of the negative factor which is in the process of being purified and microsequenced. Another protein binding to the AFP promoter region has been purified: the heterodimer Ku. Though known to bind non-specifically to DNA ends, Ku can also bind specific sequences in promoters and play a role in transcriptional regulation. In our case, Ku binds to a site included in a peculiar sequence; indeed, we showed that this sequence can adopt a cruciform structure. This observation may be correlated to the helicase function of the heterodimer Ku. Since the existence of cruciform structures in promoter sequences has already been described as a potential regulatory element; the precise role of the heterodimer Ku in the case of the AFP gene regulation remains to be evaluated.

B24-B25. Morphological and functional study of an in vitro reconstructed fully differentiated human epidermis S. MARCOUX, Y. POUMAY and M. HERIN (Département Histologie-Embryologie, Facultés Universitaires Notre-Dame de la Paix, B-5000 Namur, Belgium) We here report a simple serum-free tissue culture method for keratinocytes aiming at the in vitro reconstruction of a fully differentiated human epidermis. Morphological and functional properties of the in vitro epidermis have been studied and data are discussed. Keratinocytes were isolated by enzymatic dermis/epidermis separation of human adult normal skin obtained by surgery at abdominoplasty. Keratinocytes were first expanded in KGM-2 growth medium (Biowhittaker) as previously described (Poumay et al., 1999), then second-passage proliferating keratinocytes were seeded at high cellular density on a polycarbonate filter substrate (diameter size: 10 mm, pore size: 0.4 µm) in 1.5 mM calcium-supplemented KGM-2 growth medium (six-well plates, Falcon). Increasing the calcium concentration allowed stratification and differentiation of the keratinocytes (Hennings et al., 1980). After one week, cultures were raised to the air-medium interface and cultured for one more week, allowing terminal differentiation (keratinization) of the multilayered epidermis which results in formation of a stratum corneum (Prunieras et al., 1983). Polycarbonate filter substrate allows the separated co-culture of keratinocytes (attached to the filter) and mitomycin-treated 3T3 feeder cells (attached to the bottom of the culture well). The feeder layer proved to be essential for epidermis thickness, although it was not required for multilayering and differentiation. Histological studies of the in vitro epidermis revealed the presence of basal, spinous, granular and cornified layers, while electron microscopy analysis illustrated typical keratohyalin granules, lamellar bodies, desmosomes, hemidesmosomes and basement membrane, in accordance with the in vivo epidermis. The presence of basement membrane components was demonstrated by immunostaining of laminin-1 and collagen IV. Expression of keratinocyte markers was assessed by immunohistochemistry (involucrin, keratins 10 and 14) and perfectly illustrated a normal epidermal differentiation process. This human in vitro reconstructed epidermis, obtained by an easy and straightforward culture method, should prove to be useful in various biological approaches such as toxicology skin assays, differentiation and cellular interactions (e.g. relationship between keratinocytes and lymphocytes in immune-based skin pathologies) studies. S. Marcoux holds a fellowship from the "Fonds pour la Formation à la Recherche dans l'Industrie et l'Agriculture" and the work is supported in part by Salus Sanguinis Foundation. References HENNINGS, H., MICHAEL, D., CHENG, C., STEINERT, P., HOLBROOK, K. & YUSPA, S.H. (1980) Cell 19, 245-254. POUMAY, Y., HERPHELIN, F., SMITS, P., DE POTTER, I.Y. & PITTELKOW, M.R. (1999) Mol. Cell Biol. Res. Commun. 2, 138-144. PRUNIERAS, M., REGNIER, M. & WOODLEY, D. (1983) J. Invest. Dermatol. 81, 28s-33s.

B25. Suppressor of cytokine signaling (SOCS)-7 N. MARTENS1, M. WÉRY3, S. DEVOS1, E. QUARTIER2, R. HOOGHE1,4 and E.L. HOOGHE-PETERS1 (1Pharmacology Dept and 2Biochemistry Dept, Medical School, Free University of Brussels (VUB), 1090 Brussels; 3Laboratoire de Génétique Moléculaire (Prof. Jean Vandenhaute, Head), Unité de Recherche en Biologie Moléculaire (URBM), Facultés Universitaires Notre-Dame de la Paix (FUNDP), 5000 Namur and 4Flemish Institute for Technological Research, 2400 Mol, Belgium) Several members of the suppressor of cytokine signaling (SOCS) family modulate signal transduction by cytokines and other agonists (Yasukawa et al., 2000). For some SOCS factors such as SOCS-7, very limited information is so far available. By using the double-hybrid test in yeast (Y-2H), SOCS-7 was first cloned from a fetal brain library as a partner interacting with the adapter molecule Nck-1 (Matuoka et al., 1997). This interaction was confirmed by GST-pull-down. In murine fibroblasts, SOCS-7 also interacted with the EGF-R, phospholipase Cg and with another adapter molecule, Ash/Grb2. We have identified several additional partners using SOCS-7 as bait and a human leukocyte cDNA bank as prey, in the Gal4 Y-2H assay. Among them are members of the src-kinase family, adapter proteins and nuclear proteins. The relevance of these interactions is being further investigated by biochemical methods. We have detected constitutive SOCS-7 expression using PCR in several epithelial and leukemic cell lines and in normal leukocytes. SOCS-7 expression in leukocytes was up-regulated by mitogens. Using co-immunoprecipitation, we have shown that SOCS-7 interacts with Ash and Nck in Jurkat and in Raji cells and with Lck in Jurkat cells. It can be hypothesized that kinases (such as the src-kinases and the EGF-R) are targets for a (suppressive?) action of SOCS-7. SOCS-7 could also, like other SOCS factors, target substrates for ubiquitination and degradation by the proteasome pathway. Whereas the initial study identified partners for SOCS-7 in fibroblasts, our data also support a role for SOCS-7 in leukocytes. Supported by GOA 97-02-04, the VUB and the Region of Brussels-Capital. Many thanks to Dr Koozi Matuoka (Rudgers University, Piscataway, NJ) and Dr Jean Vandenhaute (FUNDP) for their continuous help and interest during this work. References MATUOKA, K., MIKI, H., TAKAHASH, K. & TAKENAWA, T. (1997) Biochem. Biophys. Res. Commun. 239, 488-492. YASUKAWA, H., SASAKI, A. & YOSHIMURA, A. (2000) Annu. Rev. Immunol. 18, 143-164.

B25-B26. Analysis of gene expression in human thyroid tumors by microarray technology H. MIRCESCU, F. PÉCASSE, S. DELEU, D. VENET, A. BURNIAT, J.E. DUMONT and C. MAENHAUT (IRIBHN, Université Libre de Bruxelles, Belgium) The identification of genes involved in cellular proliferation and the changes occurring in cancer cells is essential for the understanding of molecular mechanisms that govern cell development. The behavior and biology of any cell type is defined and conditioned by the genes that this cell expresses as RNA and as proteins. Recent technological advances have made possible the simultaneous study of the quantitative expression for thousands of genes. In the present study we have used the newly developed DNA microarray technology as a means of identifying genes involved in proliferation and tumorigenesis. We analyzed the expression profiles of 2400 genes in solitary thyroid autonomous adenomas and their corresponding quiescent tissue using the MICROMAX‰ cDNA microarray system (NEN‚ Life Science Products) that allows to perform the experiments with very small quantities of starting material due to their efficient system of signal amplification (TSA: tyramide signal amplification). Comparison of gene expression in the quiescent tissue and the adenomas demonstrates a close correspondence for the great majority of expressed genes. Approximately 25 different genes were found to be systematically overexpressed (> 2 fold) in mRNA from autonomous adenoma tissue compared with similar amounts of mRNA from normal tissue. These overexpressed genes included receptors, small G proteins, and structural proteins. 29 genes were downregulated in all our adenomas. They are mainly genes involved in apoptosis, thyroid function genes and growth factors. Studies are currently ongoing to confirm these regulations by Northern blotting or RT-PCR and to extend these results to a larger number of adenomas. Thyroid carcinoma, despite its low mortality rate, remains a significant medical problem because of the cost of establishing an accurate diagnosis, the need for prolonged follow-up and a relatively important rate of recurrence. The identification and characterization of genes responsible for abnormal cellular proliferation and aggressiveness of malignancy, as well as the proteins they code for, is one of the primordial goals of fundamental research in cancer. The results of such research could potentially be used in the diagnostic interventions (by providing new histochemical markers), allow a better understanding of the pathophysiology of the tumors and provide ways by which the targeted therapeutic interventions can be improved.

B26. Identification and characterization of a new protein of the thyroid TSH/cAMP pathway H. MIRCESCU, I.PIRSON, J.E. DUMONT and C.MAENHAUT (IRIBHN, Université Libre de Bruxelles, Belgium) Thyroid tumors are the most prevalent endocrine tumors and are an example of multistep transformation events that lead to neoplasia. Of the three thyroid cascades of mitogenic signal transduction, TSH/adenylyl cyclase/cAMP, EGF-HGF/ras/MAPK and phorbol esters/phospholipase C/protein kinase C, only the cAMP pathway promotes both proliferation and differentiation within the thyroid cell. Despite its effects on differentiation, there have been reports of neoplastic transformation or acquisition of malignant potential when the TSH pathway is improperly activated. Our aim is to identify putative protooncogenes and antioncogenes of the thyroid cell. To address this issue, we have chosen to study differentially expressed genes in response to the mitogenic factors. New genes were identified by differential screening of a cDNA library from a dog thyroid chronically stimulated in vivo by thyrotropin following antithyroid drug administration. This study presents the characterization of one of the isolated clones (clone 45) that was preferentially expressed in the thyroid tissue and was modulated by TSH. It codes for a protein of 686 amino acids with potential phosphorylation sites on tyrosine and serine in the central part. There is 87% amino acids identity between the human and canine sequences. The regulation of its mRNA was investigated by Northern blotting using a dog thyrocyte primary cell culture that is a good reflection of normal human thyrocyte physiology. On day 4 of culture, when the cells are quiescent, stimulation was performed with agents activating the mitogenic cascades: TSH 1 mU/ml, forskolin 10-5M, EGF 25 ng/ml, HGF 50 ng/ml, and PMA 10 ng/ml. In response to TSH stimulation, mRNA levels increased after 2-4 hours, reached a plateau at 6-8 hours and declined thereafter. A second peak was observed with longer stimulation times at 72h. Experiments performed with forskolin confirmed the role of the adenylyl cyclase/cAMP pathway in this regulation. EGF, HGF and PMA, which induce proliferation and dedifferentiation, slightly decreased the levels of clone 45 mRNA as early as 2-4h. There was an escape from this inhibition with control levels being reached after 24h, despite the continued presence of the modulating agent. Polyclonal antibodies were synthesized in order to fully characterize the role of this protein in the TSH-cAMP signaling cascade and experiments are currently in progress. In conclusion, we have identified a new clone which is preferentially expressed in the thyroid tissue and whose modulation by the mitogenic pathways is divergent.

B26. Role of calcium and ERK in the hypoxia-induced activation of HIF-1   D. MOTTET, G. MICHEL, P. RENARD, N. NINANE, M. RAES and C. MICHIELS (Laboratory of Biochemistry and cellular Biology, University of Namur, 61 rue de Bruxelles, 5000 Namur) HIF-1 (hypoxia-inducible factor-1) is the main transcription factor responsible for increased gene expression in hypoxia: VEGF, erythropoietin and glycolytic enzymes are such target genes and all participate in the adaptative response of cells to hypoxia. The oxygen-dependent regulation of HIF-1 activity occurs at multiple levels in vivo. The mechanisms regulating HIF-1a protein expression have been most extensively analyzed but the ones modulating HIF-1 transcriptional activity remain unclear. Changes in the phosphorylation and/or redox status of HIF-1a certainly play a role. Here, we show that ionomycin could activate HIF-1 transcriptional activity in a way which was additive to the effect of hypoxia without affecting HIF-1a protein level and HIF-1 DNA binding capacity. In addition, a calmodulin dominant negative mutant as well as BAPTA, an intracellular calcium chelator, inhibited the hypoxia-induced HIF-1 activation. These results indicate that elevated calcium in hypoxia could participate in HIF-1 activation. Furthermore, ERK but not JNK phosphorylation was evidenced in both conditions, ionomycin and hypoxia. PD98059, an inhibitor of the ERK pathway, as well as a ERK1 dominant negative mutant also blocked HIF-1 activation by hypoxia and by ionomycin. A MEKK1 (a kinase upstream of JNK) dominant negative mutant had no effect. In addition, BAPTA, calmidazolium, a calmodulin antagonist and PD98059 inhibited VEGF secretion by hypoxic HepG2. All together, these results suggest that calcium and calmodulin would act upstream of ERK in the hypoxia signal transduction pathway leading to enhanced HIF-1 transcriptional activity.

B26-B27. Synergistic activation of the BLV promoter transcriptional activity by TaxBLV and inhibitors of deacetylases   T.L.-A. NGUYEN1, E. ADAM1, C. CALOMME1, S. NIZET1, C. BRENNER3, A. BURNY1, Y. DE LAUNOIT2 and C. VAN LINT1 (1ULB, IBMM, Service de Chimie Biologique, Laboratoire de Virologie Moléculaire, Rue des Profs Jeener et Brachet 12, 6041 Gosselies, Belgium ; 2ULB, Faculté de Médecine, Laboratoire de Microbiologie, 808 Route de Lennik, 1070 Bruxelles, Belgium) Efficient transcription and replication of the Bovine Leukemia Virus (BLV) genome require both the viral long terminal repeat (LTR) and the virus-encoded transcriptional activator TaxBLV, which functions through three 21pb enhancers (called TxREs for TaxBLV responsive elements). There is no evidence for direct binding of the TaxBLV protein to the 21pb enhancer DNA. However, it is clearly established that the binding of CREB/ATF transcription factors to the CRE-like elements in the center of each 21pb enhancer mediates TaxBLV trans-activation. Moreover, the CREB/ATF transcription factors are known to recruit the co-activator proteins p300/CBP and P/CAF, which possess intrinsic acetyltransferase activity. It is now well established that reversible acetylation of histone and non-histone proteins is a post-translational modification playing a critical role in transcriptional regulation. To examine a potential regulatory link between protein acetylation and TaxBLV trans-activation, we tested the effect of two deacetylase inhibitors, Trichostatin A (TSA) and Sodium Butyrate (NaBut), on basal and TaxBLV-induced LTR activity. We performed transient co-transfection experiments of B-lymphoid and non B-lymphoid cell lines with an LTR-luciferase reporter construct and a TaxBLV expression vector. Remarkably, when cells were both exposed to increasing amounts of TaxBLV and treated with TSA (or NaBut), a strong transcriptional synergism was observed between the two activators TaxBLV and TSA (NaBut). This synergism required intact TxREs in the BLV LTR, since point mutations in these cis-elements abrogated the synergistic effect. Moreover, we confirmed by RNase protection analysis that the TaxBLV/TSA synergism occurred at the level of transcription. Three lysine residues in TaxBLV (K149, K288 and K296) are potential acetylation sites. To study further the mechanism of transcriptional synergism between TaxBLV and TSA, we performed similar co-transfection experiments using TaxBLV expression vectors either wild-type or mutated in the three lysines. Western blot analysis showed that the mutations did not affect nuclear localization and level of expression of TaxBLV. Mutations of the lysine residues resulted in a marked decrease in synergism compared to the wild-type TaxBLV protein, establishing a correlation between reduced TaxBLV/TSA synergism and loss of potential acetyl-acceptor lysines. These results demonstrated that TSA (or NaBut) synergistically enhanced TaxBLV-dependent transcriptional activation of the BLV promoter, and suggested that TaxBLV trans-activation could be regulated by direct acetylation on internal lysine residues. To test in vivo TaxBLV acetylation, Cos-7 cells were transfected with a TaxBLV expression vector and labeled with tritiated sodium acetate. Cellular extracts were then immunoprecipitated using a Tax-specific antiserum. We found that intracellularly-expressed TaxBLV was acetylated. TaxBLV acetylation thus appears to play a critical role in transcriptional activation of the BLV promoter.

B27. Characterization of human peroxiredoxin 5 promoter, and identification of multiple transcription and translation start sites N. NGUYEN-NHU1, J. BERCK1, A. CLIPPE1,2, A. BERNARD2 and B. KNOOPS1 (1Laboratory of Cell Biology, ISV and 2Unit of Industrial Toxicology and Occupational Medicine, Catholic University of Louvain, B-1348 Louvain-la-Neuve, Belgium) Peroxiredoxins (PRDXs) are a novel family of peroxidases that have been identified in prokaryotes and eukaryotes. In mammals, PRDXs reduce peroxides with the use of thioredoxin, glutathione or cyclophilin A. PRDXs cooperate with other antioxidant enzymes and non-enzymatic antioxidants to regulate the level of reactive oxygen species (ROS) in cells and tissues. Recently, we have cloned and characterized human, rat and mouse PRDX5. We have shown that PRDX5 may be addressed intracellularly to mitochondria, peroxisomes and the cytosol (Knoops et al., 1999; Declercq et al., 2001). Cloning and sequencing the human gene revealed that PRDX5 gene consisted in six exons and five introns spanning 4 kb on chromosome 11q13. To characterize human PRDX5 promoter, we subcloned the 3 kb genomic fragment upstream exon 1. Transfection experiments were performed in COS-7 and HepG2 cells with the 3 kb genomic fragment cloned upstream of luciferase reporter gene in the pGL3 plasmid. Reporter gene expression revealed that the 3 kb fragment was sufficient to promote transcription in both cell types with higher expression of luciferase in HepG2. Moreover, PRDX5 promoter displayed features of a housekeeping gene promoter containing GC-rich regions and devoid of TATA box. Previous studies have shown that mitochondrial, cytosolic and peroxisomal forms of human PRDX5 are 161 amino acids long although a full-length PRDX5 cDNA encoding a protein of 214 amino acids was cloned. Moreover, the sequence composed of additional 53 amino acids at the N-terminus presented the features of a cleavable mitochondrial targeting sequence. Interestingly, two AUGs were present upstream and downstream the sequence encoding the mitochondrial targeting sequence suggesting the existence of two different translation start sites. To gain more insight into transcription and/or translation mechanisms involved in the expression of the different subcellular PRDX5 proteins, 5'RACE-PCR was used to map 5'-ends of PRDX5 mRNAs from human liver. Three major transcription start sites were identified at -55, -11 and +27 from the first AUG. The cDNAs corresponding to the three transcripts were then cloned into pcDNA3.1 and transfected into HepG2 cells. A mitochondrial protein was expressed in cells transfected with the long transcript whereas a mitochondrial and a cytosolic/peroxisomal PRDX5 were identified in cells transfected with the intermediate transcript. Translation of the shortest transcript gave a cytosolic/peroxisomal protein. These results show that the use of alternative transcription and translation start sites determines the subcellular localization of PRDX5. This work was supported by the FRIA, the FNRS and the EU key action « Environment and Health » (CT99 QLK4-1308). References DECLERCQ, J.-P., EVRARD, C., CLIPPE, A., VANDER STRICHT, D., BERNARD, A. & KNOOPS, B. (2001) J. Mol. Biol. 311, 751-759. KNOOPS, B., CLIPPE, A., BOGARD, C., ARSALANE, K., WATTIEZ, R., HERMANS, C., DUCONSEILLE, E., FALMAGNE, P. & BERNARD, A. (1999) J. Biol. Chem. 274, 30451-30458.

B28. Analysis of the post-transcriptional regulation of the human interferon b mRNA M. PASTÉ, G. HUEZ and V. KRUYS (Laboratoire de Chimie Biologique, Institut de Biologie et de Médecine Moléculaires, Université Libre de Bruxelles, 12 rue des Professeurs Jeener et Brachet, 6041 Gosselies, Belgium) Human interferon b (hIFNb) gene expression is inducible by viruses and by double-stranded RNA. After induction, the gene of hIFNb is transcriptionally activated and IFNb mRNA is rapidly accumulated. The human IFNb mRNA contains AU-rich elements (ARE) in its 3'untranslated region (3'UTR). This motif was shown to confer instability and to reduce translational efficiency. In this work we followed the induction of IFNb mRNA in human epithelial cells (Hec1B) induced by Sendaï virus. Eight hours after induction, the level of IFNb mRNA decreases and 12 h after induction, no more IFNb mRNA transcripts are detected by Northern Blot analysis. Here we show that 8 h after induction by Sendaï virus, IFNb mRNA is deadenylated. The deadenylation leads to the accumulation of a short IFNb mRNA, which is rapidly degraded. In order to study the role of the ARE in deadenylation, we generated constructs containing a tagged hIFNb gene containing or not the ARE. Plasmidic DNAs or in vitro transcribed mRNAs from these constructs were transfected in Hec1B cells; the in vitro transcribed mRNAs were not deadenylated following induction whereas the mRNAs transcribed from the DNA constructs followed the same deadenylation pattern as the endogenous IFNb mRNA, suggesting an important role of the nucleo-cytoplamic transit in the deadenylation of IFNb mRNA. Several proteins, which specifically bind the ARE, were identified in eukaryotic cells and were suggested to play a role in the control of mRNA stability and/or translational efficiency. We used the electrophoretic mobility shift assay (EMSA) to identify proteins that specifically bind to the hIFNb mRNA ARE in cells extracts derived from Hec1B infected by the Sendaï virus. By inducing the cells in presence of cycloheximide, we showed that ongoing translation is not required for the protein complex binding to the IFNb 3'UTR. These data suggest that pre-existing proteins are modified after induction and become able to bind to the hIFNb ARE.

B28. Cloning and characterization of bc8, a hairy and Enhancer of split related transcription factor B. PICHON1, V. TAELMAN2, G. DOUMONT1, E. BELLEFROID2 and D. CHRISTOPHE1 (1IRIBHN and 2Laboratoire d'Embryologie Moléculaire, ULB-IBMM, Gosselies, Belgium) Members of the Hairy/E(spl) family are basic helix-loop-helix (bHLH) transcription factors that are induced by the Notch pathway and play an important role in cell fate determination. These factors are transcriptional repressors that act by interacting with the Groucho/TLE co repressors via their WRPW carboxy-terminal motif. By using a functional screening procedure to identify nuclear proteins (Pichon et al., 2000), we have cloned a cDNA coding for a bHLH factor belonging to a new family of proteins related to the Hairy/E(spl) family, the HRT family (Nakagawa et al., 1999). As compared with the Hairy/E(spl), HRTs lack the WRPW motif but exhibit a new conserved carboxy-terminal motif TEI/VGAF. Moreover, a proline residue characteristic of the Hairy/E(spl) is replaced by a glycine in the basic domain of the HRTs. In this study, we have analysed the DNA-binding properties and the transcriptional activity of bc8. The DNA-binding properties of bc8 were first investigated by EMSA by testing the interaction of the protein with different E-boxes. We found that bc8 binds preferentially to the class B E-box (CACGTG) whereas the hairy/E(spl) member, HES-1, preferred the class C E-box (CACGCG). As nucleotides surrounding the E-boxes are known to be important for the binding affinity of the bHLH factors we decided to search the consensus binding site of bc8 by using the "selex" method with a double stranded oligonucleotide containing ten variable nucleotides. After three rounds of selection, 40 sequences were analysed. 70% of the E-boxes found corresponded to the class B E-box and 30% to the class C E-box; in more than 60% of these sequences, the E-boxes were preceded by two Gs and followed by two Cs. We can thus propose the sequence "ggCACGT/CGcc" as the consensus binding site for bc8, the sequence "ggCACGTGcc" constituting the highest affinity site as confirmed by EMSA. The transcriptional activity of bc8 was investigated in co-transfection experiments. We found that the terminal 186 amino acids of bc8, fused to the DNA-binding domain of Gal4, were able to exert a 50% repression on a reporter gene under the control of five UASs. Similarly, we showed that bc8 with its own DNA-binding domain repressed a reporter gene controlled by five high affinity bc8 sites, to the same extent. bc8, thus presents the characteristics of a transcriptional repressor. Its repressing activity is lower than that exerted by the ErbA and Hes-1 repressor domains used as controls. By introducing several deletions in bc8, we showed that the transcriptional repressing activity was contained in a 43 amino acid region located behind the bHLH domain. This region called the "orange" domain, in the hairy/E(spl) family, is also conserved in the HRT family. It has been suggested that the "orange" domain could also participate in the repressing activity of the members of the hairy/E(spl) family in addition to the WRPW motif but the molecular mechanisms involved remain to be elucidated. References NAKAGAWA, O., NAKAGAWA, N., RICHARDSON, J.A., OLSON, E.N. & SRIVASTAVA, D. (1999) Dev. Biol. 216, 72-84. PICHON, B., MERCAN, D., POUILLON, V., CHRISTOPHE-HOBERTUS, C. & CHRISTOPHE, D. (2000) Anal. Biochem. 284, 231-239.

B29. Synergistic activation of HIV-1 transcriptional promoter activity by NF-kappaB and inhibitors of deacetylases V. QUIVY1, E. ADAM1, Y. COLLETTE2, D. DEMONTE1, C. VANHULLE1, B. BERKHOUT3, R. CASTELLANO2, Y. DE LAUNOIT4, A. BURNY1, J. PIETTE5, V. BOURS6 and C. VAN LINT1 (1Université Libre de Bruxelles, Institut de Biologie et de Médecine Moléculaires (IBMM), Service de Chimie Biologique, Laboratoire de Virologie Moléculaire, Rue des Profs Jeener et Brachet 12, 6041 Gosselies, Belgium; 2INSERM U119, 27 Boulevard Lei Roure, 13009 Marseille, France ; 3Department of Human Retrovirology, Academic Medical Center, University of Amsterdam, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands; 4Université Libre de Bruxelles, Faculté de Médecine, Laboratoire de Microbiologie, 808 Route de Lennik, 1070 Bruxelles, Belgium ; 5Laboratoire de Virologie and 6Laboratoire de Chimie et d'Oncologie Médicales, Institut de Pathologie, Université de Liège, Sart-Tilman, 4000 Liège, Belgium) The inducible transcription factor NF-kB plays a central role in the human immunodeficiency virus type 1 (HIV-1) activation pathway. HIV-1 transcription is also regulated by acetylation of histones and non-histone proteins. We and others have previously shown that treatment with deacetylase inhibitors such as trichostatin A (TSA) or sodium butyrate (NaBut) markedly induces HIV-1 transcriptional activity of the Long Terminal Repeat (LTR) promoter and virus production. In this study, we demonstrate that TSA or NaBut synergized with both ectopically expressed p50/p65 and Tumor Necrosis Factor (TNF)-induced NF-kB to activate the LTR promoter. This synergism required intact kB sites in the HIV-1 enhancer and was observed with LTRs from subtypes A through G of the HIV-1 group M (Major) with a positive correlation between the number of kB sites present in the respective LTRs and the amplitude of the TNF/TSA synergism. TSA and NaBut drugs prolonged TNF-induced NF-kB binding to DNA, likely by prolonging the intranuclear presence of p65. Remarkably, a marked delay in the recovery of the inhibitory protein IkB-alpha correlated temporally with the sustained NF-kB binding activity. Importantly, the physiological relevance of the TNF/TSA(NaBut) synergy was shown on HIV-1 replication in both acutely and latently HIV-infected cell lines. Our results could therefore open new therapeutic strategies aimed at forcing viral expression and at contributing, in the presence of an efficient antiviral treatment, to a reduction of the pool of latently HIV-infected cellular reservoirs.

B29. Control of Bacteriophage Mu Lysogenic Repression C. RANQUET1,2, G. MAENHAUT-MICHEL2, J. GEISELMANN1 and A. TOUSSAINT3 (1Laboratoire du Contrôle de l'Expression Génique, Université J. Fourier, BP 53, F-38041 Grenoble cedex 9, France; 2Laboratoire de Génétique des Procaryotes, Université Libre de Bruxelles, IBMM,12 rue de Professeurs R. Jeneer et J. Brachet, B6401 Gosselies, Belgium; 3 Service de Conformation des Macromolécules Biologiques et Bioinformatique, Université Libre de Bruxelles, 50 av F.D. Roosevelt - CP 160/16, B-1050, Bruxelles, Belgium) When residing as a latent prophage, the transposable phage Mu is integrated at a random location in the host genome. No chemical or physical treatment triggers the massive induction of repressed Mu prophages, which nevertheless switch to lysis spontaneously at low frequency. In the laboratory, Mu lysogens can be induced using mutants of the virus, which express mutant forms (cts or vir) of the phage lysogenic repressor Repc. In both cases inactivation of Repc is the cause of Mu induction. The cts and vir mutants exemplify two mechanisms of relieving repression by Repc, lowering its affinity for the operators or decreasing its intracellular concentration through rapid proteolytic degradation by host proteases. Mu prophages are derepressed in lysogens grown to stationary phase (Shapiro and Higgins, 1989, Lamrani et al., 1999). This "S" derepression depends on host proteases, the stationary phase specific sigma factor RpoS (Lamrani et al., 1999) and the SsrA RNA (Ranquet et al., 2001) and its associated protein SmpB (Karzai et al., 1999). We have measured the amount and the stability of wild type and cts62 Repc in wild type, clpX, clpP and lon mutants of E.coli in various physiological states. Repc was degraded by the Lon and ClpP proteases at high temperature and in stationary phase. Repressor accumulated in cells where the Mu lytic pE promoter was fully derepressed. clpX and clpP host mutations although generating a similar phenotypic behavior (e.g. absence of S derepression) had a different incidence on Repc expression (accumulation in the clpP host vs. unchanged in clpX). These and former data suggest that Mu Repc is flexible and undergoes conformational changes between active/stable and non-active/unstable forms. Some forms are dominant and stimulate the conversion of others into one with properties similar to their own. Other forms are resistant to that induced change of conformation. We propose that Repc activity is controlled by the regulated expression and stability of the various forms of the protein and their interactions via phage and host functions. This generates a complex array of transduction signals through which the phage senses its host physiology and is induced for instance when the host cell starves. References KARZAI, A.W., SUSSKIND, M.M. & SAUER, R.T. (1999) EMBO J. 18, 3793-3799. LAMRANI, S., RANQUET, C., GAMA, M.J., NAKAI, H., SHAPIRO, J.A., TOUSSAINT, A. & MAENHAUT-MICHEL, G. (1999) Mol. Microbiol. 32, 327-343. RANQUET, C., GEISELMANN, J. & TOUSSAINT, A. (2001) Proc. Natl. Acad. Sci.USA 98, 10220-10225. SHAPIRO, J.A. & HIGGINS, N.P. (1989) J. Bacteriol. 171, 5975-5986.

B30. Identification of pre-B-cell colony-enhancing factor (PBEF) as the mammalian nicotinamide phosphoribosyltransferase (NAmPRTase), a cytosolic enzyme involved in NAD biosynthesis A. RONGVAUX1, R. J. SHEA2, M. H. MULKS2, D. GIGOT3, J. URBAIN1, O. LEO1 and F. ANDRIS1 (1Laboratoire de Physiologie Animale, Université Libre de Bruxelles, 6041 Gosselies, Belgium; 2Department of Microbiology, Michigan State University, East Lansing, Michigan 48824-1101, U.S.A and 3Laboratoire de Microbiologie, Université Libre de Bruxelles, 1070 Bruxelles, Belgium) Nicotinamide phosphoribosyltransferase (NAmPRTase, EC 2.4.2.12) is an essential enzyme catalyzing a key step in the salvage pathway of NAD biosynthesis from nicotinamide. However, very little is known about the structure, function and regulation of this enzyme in multicellular organisms. A bacterial NAmPRTase has been recently identified, and shown to display significant homology with a mammalian gene, PBEF (pre-B-cell colony enhancing factor), thought to encode an hematopoietic growth factor. We demonstrate in this work that the murine PBEF is a cytosolic protein with NAmPRTase activity, catalyzing the condensation of nicotinamide with 5-phosphoribosyl-1-pyrophosphate, to yield nicotinamide mononucleotide, an intermediate in the biosynthesis of NAD. The mouse gene was functional beyond the phylogenic boundary, as it confers to a NAmPRTase-defective A. pleuropneumoniae bacterial strain the ability to grow in the absence of NAD, with nicotinamide as a unique source of pyridine nucleotide. Collectively, these results demonstrate that PBEF represents a key enzyme in NAD metabolism, and cast some doubts on its potential role as a secreted cytokine.

B30. Responses to cadmium in different populations of Thlaspi caerulescens N. ROOSENS 1, A. SMITH2 , P. MEERTS3 and N. VERBRUGGEN1 (1Laboratoire de Physiologie et de Génétique Moléculaire des Plantes; 3Laboratoire de Génétique et Ecologie Végétale, Free University of Brussels, Belgium; 2Department of Plant Sciences,Oxford University, England) On soils enriched in metals, a rare class of plants named hyperaccumulators is able to survive and accumulate exceptionally high concentrations of metals in their shoots (above 100 ppm cadmium (Cd) or 1000 ppm Zinc (Zn) in their dry weight). Thlaspi caerulescens (T.c.) possesses several characteristics conferring the ability to hyperaccumulate and tolerate Zn and Cd (Escarre et al., 2000). However, the significance of the Cd and Zn hyperaccumulation and the basis of tolerance and sequestration of these metals are poorly understood. Zn is an essential metal, which is toxic at supra-optimal concentrations (above 300-500ppm in the shoot) while Cd is not generally believed to have a biological function and is highly toxic even at low concentration (above 1ppm) (Pence et al., 2000). Seven ecotypes of T.c from different European metalliferous sites were analysed in hydroponic culture upon 100µM ZnSO4 and 3 or 30 µM CdSO4 treatment. Growth, Cd/Zn in the plant and the xylem sap were analysed. Measurements of the biomass showed that CdSO4 exerts an adverse effect on most ecotypes. In marked contrast, Cd addition in the growth medium had a stimulator effect on 2 ecotypes especially at low concentration (3 µM CdSO4) but also at higher doses. This lead to the hypothesis of a biological role for Cd in higher plants like it was demonstrated in marine diatoms (Lane and Morel, 2000). Comparison of the capacity to accumulate Cd and Zn in the ecotypes showed that they have not the same ability to hyperaccumulate these heavy metals (5000 to 10000 µg /g DW for the Zn and 3000 to 10000 µg/g DW for the Cd). However, for all them the concentration of Zn and Cd in the shoot was always greater than in the root. Moreover, Cd and Zn concentrations in the shoot are always positively correlated but the 2 populations that exhibit a stimulation of the growth by Cd, present a higher Cd/Zn ratio. This is confirmed by a higher concentration of Cd in the xylem sap. Further analysis of the data suggested that tolerance and Cd hyperaccumulation in the shoot are inversely correlated and that the ratio root-shoot is a major determinant of the Cd and Zn hyperaccumulation. Future work will focus on stimulatory mechanisms that stimulate the growth of T.c. This work was supported by a grant from the Fondation Wiener-Anspach (ULB). Nancy Roosens is " Chargé de Recherche " FNRS References ESCARRÉ, J., LEFÈBVRE, C., GRUBER, W., LEBLANC, M., LEPART, J., RIVIÈRE, Y. & DELAY, B. (2000) New Phytol. 145, 429-437. LANE, T. & MOREL, F. (2000) Proc. Natl. Acad. Sci.U SA 97 , 4627-4631. PENCE, N., LARSEN, P., EBBS, S., LETHAM, D., LASAT, M., GARVIN, D., EIDE, D. & KOCHIAN, L. (2000). Proc. Natl. Acad. Sci.U SA 97, 4956-4960.