Volume 382, Issue 4

The characterization of tumorassociated antigens recognized by human T lymphocytes in a major histocompatibility complex (MHC)restricted fashion has opened new possibilities for immunotherapeutic approaches to the treatment of human cancers. Dendritic cells (DC) are professional antigen presenting cells that are well suited to activate T cells toward various antigens, such as tumorassociated antigens, due to their potent costimulatory activity. The availability of large numbers of DC, generated either from hematopoietic progenitor cells or monocytes in vitro or isolated from peripheral blood, has profoundly changed preclinical research as well as the clinical evaluation of these cells. Accordingly, appropriately pulsed or transfected DC may be used for vaccination in the field of infectious diseases or tumor immunotherapy to induce antigenspecific T cell responses. These observations led to pilot clinical trials of DC vaccination for patients with cancer in order to investigate the feasibility, safety, as well as the immunologic and clinical effects of this approach. Initial clinical studies of human DC vaccines are generating encouraging preliminary results demonstrating induction of tumorspecific immune responses and tumor regression. Nevertheless, much work is still needed to address several variables that are critical for optimizing this approach and to determine the role of DCbased vaccines in tumor immunotherapy.

In this review we discuss intracellular bacteria as targets and carriers for vaccines. For clarity and ease of comprehension, we focus on three microbes, Mycobacterium tuberculosis, Listeria monocytogenes and Salmonella, with an emphasis on tuberculosis, one of the leading causes of death from infectious disease. Novel vaccination strategies against these pathogens are currently being considered. One approach favors the use of live attenuated vaccines and vaccine carrier strains thereof, either for heterologous antigen presentation or DNA vaccine delivery. This strategy includes both the improvement of attenuated vaccine strains as well as the de novo generation of attenuated variants of virulent pathogens. An alternative strategy relies on the application of subunit immunizations, either as nucleic acid vaccines or protein antigens of the pathogen. Finally, we present a short summary of the vaccination strategies against tuberculosis.

Invasive intracellular bacteria are able to transfer eukaryotic expression plasmids into mammalian host cells in vitro and in vivo. This can be used to induce immune responses toward protein antigens encoded by the plasmid or to complement genetic defects. Plasmid transfer takes place when the recombinant bacterium dies within the host cell, either due to metabolic attenuation or induction of autolysis. Alternatively, antibiotics can be used and spontaneous transfer has also been observed, indicating that this phenomenon might also occur under physiological conditions. Plasmid transfer has been reported for Shigella flexneri, Salmonella typhimurium and S. typhi, Listeria monocytogenes and recombinant Escherichia coli, but other invasive bacteria should also share this property. In vivo attempts were mainly directed toward vaccination using shigella and salmonella as carrier. So far a wide variety of antigens have been used succesfully in mice. Often this type of immunization was superior over direct application of antigen or using the same bacterium as a heterologous carrier expressing the antigen via a prokaryotic promoter. Characterization of the host cells revealed that macrophages and dendritic cells might be responsible for immune stimulation by either expressing the antigen or crosspresenting the antigen after uptake of apoptotic antigen expressing cells.

DNAbased vaccination is a novel technique to efficiently stimulate humoral (antibody) and cellular (T cell) immune responses to protein antigens. In DNAbased vaccination, immunogenic proteins are expressed in in vivo transfected cells of the vaccine recipients in their native conformation with correct posttranslational modifications from antigenencoding expression plasmid DNA. This ensures the integrity of antibodydefined epitopes and supports the generation of protective (neutralizing) antibody titers. Plasmid DNA vaccination is furthermore an exceptionally potent strategy to stimulate CD8[+] cytotoxic T lymphocyte (CTL) responses because antigenic peptides are efficiently generated by endogenous processing of intracellular protein antigens. These key features make DNAbased immunization an attractive strategy for prophylactic and therapeutic vaccination against extra and intracellular pathogens. In this brief review, we summarize the current state of expression vector design, DNA delivery strategies, priming immune responses to intracellular or secreted antigens by DNA vaccines and unique advantages of DNA versus recombinant proteinbased vaccines using the hepatitis B surface antigen (HBsAg) as a model antigen.

Malaria can be a very severe disease, particularly in young children, pregnant women (mostly in primipara), and malaria naïve adults, and currently ranks among the most prevalent infections in tropical and subtropical areas throughout the world. The widespread occurrence and the increased incidence of malaria in many countries, caused by drugresistant parasites (Plasmodium falciparum and P. vivax) and insecticideresistant vectors (Anopheles mosquitoes), indicate the need to develop new methods of controlling this disease. Experimental vaccination with irradiated sporozoites can protect animals and humans against the disease, demonstrating the feasibility of developing an effective malaria vaccine. However, developing a universally effective, long lasting vaccine against this parasitic disease has been a difficult task, due to several problems. One difficulty stems from the complexity of the parasites life cycle. During their life cycle, malaria parasites change their residence within the host, thus avoiding being reexposed to the same immunological environment. These parasites also possess some distinct antigens, present at different life stages of the parasite, the socalled stagespecific antigens. While some of the stagespecific antigens can induce protective immune responses in the host, these responses are usually genetically restricted, this being another reason for delaying the development of a universally effective vaccine. The stagespecific antigens must be used as immunogens and introduced into the host by using a delivery system that should efficiently induce protective responses against the respective stages. Here we review several research approaches aimed at inducing protective antimalaria immunity, overcoming the difficulties described above.

Synthetic immunogens, containing builtin adjuvanticity, B cell, T helper cell and CTL epitopes or mimotopes, are ideal and invaluable tools to study the immune response with respect to antigen processing and presentation. This serves as a basis for the development of complete and minimal vaccines which not need large carrier proteins, further adjuvants, liposome formulations or other delivery systems. Combinatorial peptide libraries, either completely random or characterized by one or several defined positions, are useful tools for the identification of the critical features of B cell epitopes and MHC class I and class II binding natural and synthetic epitopes. The complete activity pattern of O/X library with hundreds of peptide collections, each made up from billions of different peptides, represents the ranking of amino acid residues mediating contact to the target proteins of the immune system. Combinatorial libraries support the design of peptides applicable in vaccination against infectious agents as well as therapeutic tumour vaccines. Using the principle of lipopeptide vaccines, strong humoral and cellular immune responses could elicited. The lipopeptide vaccines are heatstable, nontoxic, fully biodegradable and can be prepared on the basis of minimized epitopes by modern methods of multiple peptide synthesis. The lipopeptides activate the antigenpresenting macrophages and cells and have been recently shown to stimulate innate immunity by specific interaction with receptors of the Toll family.

Although vaccines have proven very successful in preventing certain infectious diseases, progress in the field has been slowed by the tediousness of developing classical vaccines consisting of whole pathogens. Thus, there is great need for improvement in several areas: firstly, the range of diseases which can be treated has to be expanded. Secondly, antigens have to be defined to make the use of whole pathogens as antigen obsolete. And thirdly, new adjuvants have to be developed which show low toxicity, high potency and are also able to drive the immune response in the desired direction. Ideally, a vaccine would only consist of wellcharacterized, synthetic materials. This review summarizes the different approaches for the development of completely defined synthetic vaccines.

All organisms need protection against microorganisms, e. g. bacteria, viruses and fungi. For many years, attention has been focused on adaptive immunity as the main antimicrobial defense system. However, the adaptive immune system, with its network of humoral and cellular responses is only found in higher animals, while innate immunity is encountered in all living creatures. The turning point in the appreciation of the innate immunity was the discovery of antimicrobial peptides in the early eighties. In general these peptides act by disrupting the structural integrity of the microbial membranes. It has become clear that membraneactive peptides and proteins play a crucial role in both the innate and the adaptive immune system as antimicrobial agents. This review is focused on the functional and structural features of the naturally occurring antimicrobial peptides, and discusses their potential as therapeutics.

To be functional, nucleic acids need to adopt particular threedimensional structures. For a long time DNA was regarded as a rigid and passive molecule with the sole purpose to store genetic information, but experimental data has now accumulated that indicates the full dynamic repertoire of this macromolecule. During the last decade, fourstranded DNA structures known as Gquadruplexes, or DNA tetraplexes, have emerged as a threedimensional structure of special interest. Motifs for the formation of Gquadruplex DNA structures are widely dispersed in eukaryotic genomes, and are abundant in regions of biological significance, for example, at telomeres, in the promoters of many important genes, and at recombination hotspots, to name but a few in man. Here I explore the plethora of Gquadruplex DNA structures, and discuss their possible biological functions as well as the proteins that interact with them.

Heat shock proteins (HSPs) have been described as potent tumor vaccines in animal models and are currently studied in clinical trials. The underlying immune response relies on immunogenic peptides that the HSPs have acquired intracellularly by interfering with the classical antigen processing pathways. There have been numerous reports shedding light on how HSPs are able to gain this function and a number of important requirements for HSPmediated specific immunity have been described: first, the ability of HSPs to bind immunogenic peptides. Second, the acquisition of HSPs by specialized antigen presenting cells with efficient antigen processing pathways capable of inducing cellular immune responses. Third, the existence of specific receptors on the surfaces of antigen presenting cells, allowing efficient and rapid uptake of HSPpeptide complexes from the extracellular fluid. And fourth, the ability of heat shock proteins to activate antigen presenting cells, enabling the latter to prime cytotoxic T cell responses against the peptides associated to HSPs.

The KRAB domain of human Kox1, a member of the KRAB C2H2 zinc finger family, confers strong transcriptional repressor activities even to remote promoter positions. Here, HDAC inhibitors were used to demonstrate that histone deacetylation is not required for mediating transcriptional repression of KRAB zinc finger proteins. Two reporter systems with either stably integrated or transiently transfected templates, both under control of strong viral promoters, were analyzed. Under all circumstances, HDAC inhibition did not alter the repression potential of the KRAB domain. In case of the stably integrated luciferase reporter gene system, neither expression levels of the KRAB fusion protein nor complex formation with its putative corepressor TIF1β were significantly changed. Furthermore, the TIF1β/KRAB complex was devoid of mSin3A and HDAC1. In the transient transfection system, the transcriptional repression induced by TIF1β and HP1α was not diminished by HDAC inhibitors, whereas the repressory activity of TIF1α was significantly affected. Thus, KRAB, TIF1β and HP1α are likely to be functionally linked. In conclusion, HDAC activity is not essential for the strong transcriptional repressor activity mediated by the KRAB domain of Kox1 in particular and, presumably, by KRAB domains in general. This feature might be helpful in identifying and characterizing target genes under the control of KRAB zinc finger proteins.

The phenylpropanoid enzyme 4-coumarate:coenzyme A ligase (4CL) plays a key role in general phenylpropanoid metabolism. 4CL is related to a larger class of prokaryotic and eukaryotic adenylateforming enzymes and shares several conserved peptide motifs with these enzymes. In order to better characterize the nature of 4CL gene families in poplar, parsley, and tobacco, we used degenerate primers to amplify 4CL sequences from these species. In each species additional, divergent 4CL genes were found. Complete cDNA clones for the two new poplar 4CL genes were obtained, allowing examination of their expression patterns and determination of the substrate utilization profile of a xylemspecific isoform. Phylogenetic analysis of these genes and gene fragments confirmed previous results showing that 4CL proteins fall into two evolutionarily ancient subgroups . A comparative phylogenetic analysis of enzymes in the adenylate forming superfamily showed that 4CLs, luciferases, and acetate CoA ligases each form distinct clades within the superfamily. According to this analysis, four Arabidopsis 4CLlike genes identified from the Arabidopsis Genome Project are only distantly related to bona fide 4CLs or are more closely related to fatty acid CoA ligases, suggesting that the three Arabidopsis 4CL genes previously characterized represent the extent of the 4CL gene family in this species.

The deviant poly(A) signal of hepatitis B virus (HBV) not only controls the formation of the 3 end of all the viral RNA, but is also crucial for HBV replication. Hence, a cisreleasing hammerhead ribozyme (RzA) targeted to the poly(A) signal region of HBV subtype adr was investigated for its antiviral effects. In vitro, RzA cleaved HBV RNA at its target site up to 70%, while the disabled ribozyme (dRzA), which had a onebase mutation in the catalytic site, did not cleave the target RNA at all. When the ribozymes were cotransfected into HepG2 cells with the HBV genomecontaining plasmid p3.6II, the wildtype ribozyme RzA could effectively decrease HBV RNA levels and inhibit HBV replication, whereas its disabled form, dRzA, had much weaker effects, indicating that the active catalytic domain of the hammerhead ribozyme could markedly increase the extent of antisensemediated inhibition. In addition, there was a gradient of effectiveness: the higher the amount of released ribozyme, the more the reduction in target HBV RNA in cells as well as progeny DNA reduction. These results suggest the possibility of the hammerhead ribozyme RzA to be used for the gene therapy of HBV infection.

During protein synthesis the ribosome interacts with ligands such as mRNA, tRNA and translation factors. We have studied the effect of ribosomeligand interaction on the accessibility of 18S rRNA for single strandspecific modification in ribosomal complexes that have been assembled in vivo, i. e. native polysomes. A comparison of the modification patterns derived from programmed and nonprogrammed ribosomes showed that bases in the 630- and 1060- loops (530- and 790-loops in E. coli) together with two nucleotides in helices 33 and 34 were protected from chemical modification. The majority of the protected sites were homologous to sites previously suggested to be involved in mRNA and/or tRNA binding in prokaryotes and eukaryotes, implying that the interaction sites for these ligands are similar, if not identical, in naturally occurring programmed ribosomes and in in vitro assembled ribosomal complexes. Additional differences between programmed and nonprogrammed ribosomes were found in hairpin 8. The bases in helix 8 showed increased exposure to chemical modification in the programmed ribosomes. In addition, structural differences in helices 36 and 37 were observed between native 80S runoff ribosomes and 80S ribosomes assembled from isolated 40S and 60S subunits.

Interleukin-6 (IL-6) is a 185 amino acid residue helical cytokine with various biological activities (e. g. B cell development, acute phase reaction). We have investigated the role of the 168 185 Cterminal region of IL-6 in the induction of fibrinogen synthesis and expression of junB mRNA using synthetic peptides corresponding to this region. Circular dichroism spectroscopy data suggest that even truncated peptides have a strong tendency to adopt an ordered conformation. Peptides were tested alone or in combination with recombinant hIL-6 on an IL-6 responsive human hepatoma HepG2 cell line. The expression of the protooncogene junB monitored by competitive RTPCR represents an early, while the fibrinogen production detected by sandwich ELISA a late, marker of IL-6 initiated events. We found that peptides depending on their structure modulate spontaneous as well as IL-6 induced fibrinogen production and/or mRNA expression of junB by exhibiting inhibition (in the presence of IL-6) or stimulation (in the absence of IL-6).

We performed a threedimensional separation of pulsechase duallabelled rat liver cytosolic proteins using hydrophobic interaction chromatography, isoelectric focusing, and SDS gel electrophoresis. Due to very different expression rates but similar size and pI of rat liver cytosolic proteins, we demonstrate the impossibility of successful twodimensional separations of such complex protein mixtures. A prefractionation of proteins by hydrophobic interaction chromatography is therefore recommended prior to twodimensional gel electrophoresis. Our studies confirmed the correlation between protein turnover rates and surface hydrophobicity.

The conversion of the αhelical, protease sensitive and noninfectious form of the prion protein (PrP) into an insoluble, protease resistant, predominantly βsheeted and infectious form (PrP) is the fundamental event in prion formation. In the present work, two soluble and stable intermediate structural states are newly identified for recombinant Syrian hamster PrP(90 231) (recPrP), a dimeric αhelical state and a tetra or oligomeric, βsheet rich state. In 0.2% SDS at room temperature, recPrP is soluble and exhibits αhelical and random coil secondary structure as determined by circular dichroism. Reduction of the SDS concentration to 0.06% leads first to a small increase in αhelical content, whereas further dilution to 0.02% results in the aquisition of βsheet structure. The reversible transition curve is sigmoidal within a narrow range of SDS concentrations (0.04 to 0.02%). Size exclusion chromatography and chemical crosslinking revealed that the αhelical form is dimeric, while the βsheet rich form is tetra or oligomeric. Both the αhelical and βsheet rich intermediates are soluble and stable. Thus, they should be accessible to further structural and mechanistic studies. At 0.01% SDS, the oligomeric intermediates aggregated into large, insoluble structures as observed by fluorescence correlation spectroscopy. Our results are discussed with respect to the mechanism of PrP formation and the propagation of prions.

The pgktpi gene locus of Thermotoga maritima encodes both phosphoglycerate kinase (PGK) and a bienzyme complex consisting of a fusion protein of PGK with triosephosphate isomerase (TIM). No separate tpi gene for TIM is present in T. maritima. A frameshift at the end of the pgk gene has been previously proposed as a mechanism to regulate the expression of the two protein variants [Schurig et al., EMBO J. 14 (1995), 442 451]. Surprisingly, the complete T. maritima genome was found to contain a pgktpi sequence not requiring the proposed frameshift mechanism. To clarify the apparent discrepancy, a variety of DNA sequencing techniques were applied, disclosing an anomalous local variability in the pgktpi fusion region. The comparison of different DNA samples and the mass spectrometric analysis of the amino acid sequence of the natural fusion protein from T. maritima MSB8 confirmed the local variability of the DNA variants. Since not all peptide masses could be assigned, further variations are conceivable, suggesting an even higher heterogeneity of the T. maritima MSB8 strain.

We analysed the methylation patterns of CpG dinucleotides in a bidirectional promoter region (LRS, LMP 1 regulatory sequences) of latent EpsteinBarr virus (EBV) genomes using automated fluorescent genomic sequencing after bisulfiteinduced modification of DNA. Transcripts for two latent membrane proteins, LMP 1 (a transforming protein) and LMP 2B, are initiated in this region in opposite directions. We found that B cell lines and a clone expressing LMP 1 carried EBV genomes with unmethylated or hypomethylated LRS, while highly methylated CpG dinucleotides were present at each position or at discrete sites and within hypermethylated regions in LMP 1 negative cells. Comparison of high resolution methylation maps suggests that CpG methylationmediated direct interference with binding of nuclear factors LBF 2, 3, 7, AML1/LBF1, LBF5 and LBF6 or methylation of CpGs within an Ebox sequence (where activators as well as repressors can bind) is not the major mechanism in silencing of the LMP 1 promoter. Although a role for CpG methylation within binding sites of Sp1 and 3, ATF/CRE and a sisinducible factor (SIF) cannot be excluded, hypermethylation of LRS or regions within LRS in LMP 1 negative cells suggests a role for an indirect mechanism, via methylcytosine binding proteins, in silencing of the LMP 1 promoter.

Methylation of DNA occurs at the C[5] and N[4] positions of cytosine and N[6] of adenine. The chemistry of methylation is similar among methyltransferases specific for cytosineN[4] and adenineN[6]. Moreover these enzymes have similar structures and active sites. Previously it has been demonstrated that the DNA(adenineN[6])methyltransferases M.EcoRV, M.EcoRI, E. coli dam and both domains of M.FokI also modify cytosine residues at the N[4] position [Jeltsch et al., J. Biol. Chem. 274 (1999), 19538 19544]. Here we show that the cytosineN[4] methyltransferase M.PvuII, which modifies the second cytosine in CAGCTG sequences, also methylates adenine residues in CAGATG/ CAGCTG substrates in which the target cytosine is replaced by adenine in one strand of the recognition sequence. Therefore, adenineN[6] and cytosine N[4] methyltransferases have overlapping target base specificities. These results demonstrate that the target base recognition by Nspecific DNA methyltransferases is relaxed in many cases. Furthermore, it shows that the catalytic mechanisms of adenine N[6] and cytosineN[4] methyltransferases are very similar.

Menkes disease is an Xlinked disorder of copper metabolism that is usually fatal. The affected gene has recently been cloned and encodes one of the two human copper ATPases. If the Menkes ATPase is defective, copper is trapped in the intestinal mucosa, leading to systemic copper deficiency. In order to study copper transport by this ATPase and the effects of disease mutations on its function, we developed a Xenopus laevis oocyte expression system. Wildtype Menkes ATPase cDNA and a fusion of this gene with the green fluorescent protein (GFP) gene was transcribed in vitro and the mRNA injected into oocytes. Expression in oocytes was analyzed by Western blotting and fluorescence microscopy. The Menkes ATPase GFP chimera appeared to localize primarily to the plasma membrane as assessed by confocal microscopy. This system should thus provide an interesting new tool to study the function of the Menkes ATPase.