Tetracycline-controlled Gene Expression in Entamoeba histolytica

L. Hamann, H. Buß*), E. Tannich

To provide further tools for functional genetics of Entamoeba histolytica, we have tested the suitability of the bacterial TN10-encoded tet-repressor/tet-operator system for gene regulation in amoeba trophozoites. Expression of the tet-repressor within the amoeba was driven by the wild-type endogenous lectin gene promoter from episomal transfected plasmids. Tetracycline-inducible expression of a reporter gene driven by a modified tet-operator-bearing lectin gene promotor was monitored by transient or episomal transfections. Promotor activity was dependent on the position of the tet-operator insertion. Under appropriate conditions, expression of the reporter gene in tet-repressor expressing cells revealed only background levels but was inducible up to 240 fold by the addition of non-toxic amounts of tetracycline reaching full activity within 24 to 48 hours. Because of the tight and rapid control by tetracycline, the tet-repressor controlled lectin gene promotor should be a useful tool for reverse genetic approaches in E. histolytica as well as for recombinant protein expression within an anaerobic organism.
*) Parasitology Section
 
 

Molecular Modeling of Amoebapore and NK-lysin:
Common Structure for Cytolytic Effector Molecules of Most Distantly Related Organisms

T. Dandekar¹⁾, M. Leippe

Similar functional demands may lead to the creation of similar proteins, even in species separated by large evolutionary distances. This is also true for effector molecules of the innate defense systems, which are currently studied in a variety of organisms. We recently identified a significant sequence similarity in combination with a functional analogy between membrane-active polypeptides of organisms separated extremely early in evolution: the amoebapores of the protozoon Entamoeba histolytica  and NK-lysin of porcine cytotoxic lymphocytes. Both groups of the similarly sized peptides are cytolytically active toward bacteria and tumor cells, and the rather unspecific mode of action elucidated for the amoebic peptides, i.e. the formation of ion channels in target cell membranes, most likely favors such a broad target cell spectrum.
In anticipation of the soon-to-be-released tertiary structure of any member of these peptide groups, we predicted their tertiary structure by a computer-aided approach using the genetic algorithm. This method simulates evolution to find the „fittest fold“ of a given protein in that it selects according to basic protein structure principles. It has been shown previously that such simulations are sequence-specific and are able to predict the main-chain topology of known crystal structures of helical and of non-helical proteins. Molecular models were generated by exploiting available information such as the primary structures, secondary structure predictions and positions of disulfide bonds. Interestingly, for both types of peptides, a common structure of a four-a-helix bundle with three intramolecular crosslinks was obtained. The models generated strengthen the notion that amoebapore and NK-lysin comprise a distinct structural class of cytolytic and antibacterial polypeptides. The members of this class combine the typical conformation of lytic peptides dominated by amphipathic a-helices with a disulfide bonded compact structure known from the  -sheeted defensins and small toxins.
1) European Molecular Biology Laboratory (EMBL), Heidelberg
 
 

Potency of Amoebapores Compared to that of Other Membrane-permeating Peptides 

J. Andrä, O. Berninghausen, M. Leippe

Amoebapores are considered to represent an essential part of the cytotoxic machinery of Entamoeba histolytica. They are 77-residue peptides with an a-helical tertiary structure, which act by forming distinct ion-channels in target cell membranes. Three isoforms (amoebapore A, B, and C) have been isolated from cytoplasmic granules of the amoebae. Besides their extracellular cytolytic potential, they may act intracellulary as antibacterial agents to lyse ingested bacteria. Therefore, amoebapores may functionally resemble the defensins, antibacterial peptides from granules of mammalian neutrophils.
We compared the biological activities of the amoebapores with various other naturally occurring membrane-active peptides, which have been structurally and functionally characterized: i) melittin, the main lytic component from bee venom; ii) alamethicin, a pore-forming toxin from the fungi Trichoderma viride; iii) cecropins, antibacterial peptides from insect hemolymph; iv) magainins, potent defense peptides from Xenopus skin; v) dermaseptin, an antifungal peptide isolated from a tropical frog species; and vi) human defensins. All of these peptides appear to act by destroying the integrity of membranes.
We determined pore-forming activity and cytolytic activity of the aforementioned natural peptides by measuring the dissipation of a valinomycin-induced diffussion potential across the membranes of liposomes and by monitoring the release of a fluorescent dye from prelabeled Jurkat cells, respectively. It became apparent that amoebapores are considerably more active in the liposome assay than all other peptides tested. Among the latter, alamethicin was the most active, exerting approximately 20% of the activity of amoebapore A. As it has been reported for all the peptides that they share preferences for pH and phospholipid composition with the amoebapores, the results were not impaired because of particular assay conditions. The inferior pore-forming activity of most of the peptides tested compared to amoebapores is most likely due to the fact, that the amoebic peptides are capable of forming stable ion channels whereas the other, markedly smaller peptides are thought to disintegrate membranes by disrupting the lipid packing at a high peptide/phospholipid ratio. With regard to cytolytic activity, amoebapores were again highly active; particularly amoebapore C was virtually the most potent peptide tested and only melittin had similar potency. The other peptides displayed inferior activity only or were not active at all.
In conclusion, besides their proposed significance as pathogenicity factors, amoebapores are distinct among membrane-active peptides and are interesting model molecules for studying the (supra)molecular architecture of peptide-mediated channels.
 
 

Antimicrobial Activity of Synthetic Amoebapore Analogs

J. Andrä, B. Mannes¹⁾, M. Leippe
 
The emergence of microbial resistance to antibiotics has renewed the interest in natural antimicrobial agents. A widespread and crucial element of the first defense line of animals and plants against pathogens and also a constituent of human innate immunity are antimicrobial peptides that rapidly kill bacteria and fungi by permeating their membranes. Chemical synthesis or recombinant expression permits the enhancement of particular properties of these natural bioactive molecules by changing their amino acid sequences; these techniques also allow the production of a desired peptide in large quantities. Currently, the antimicrobial activity of many synthetic peptides are under investigation in order to employ them eventually against pathogens resistant to the classical antibiotics. The mode of action of these mostly cationic peptides, i.e. the physical disruption of microbial membranes, make them candidate molecules for new antibiotics since it may be suggested that their application will not create resistant strains. A few synthetic peptides have already entered clinical trials such as the synthetic magainin analog MSI-78, for which a phase III study for the topical treatment of diabetic foot ulcers has been completed successfully.
In a preceding study, we designed 15 synthetic peptides which represent short versions of amoebapore. The amino acid sequences of these peptides were based on that of the third helix of each amoebapore isoforms since, according to our studies of the structure-activity relationships of amoebapores, the third helix is the membrane-permeating element of the natural polypeptide. We determined the activity of the peptides towards artificial membranes, bacteria, and eukaryotic cells and found that some were even more antibacterial and/or cytotoxic than the parent molecules. The most promising model peptide for antibiotic application termed C1, although of poor solubility, displayed extraordinary pore-forming activity, exerted activity against some bacteria, but was of low cytotoxicity.
In the following, we used peptide C1 as a lead for sequence variations to enhance its solubility and antimicrobial potency. Stepwise exchange of positively charged lysine residues for negatively charged residues such as glutamic acid and aspartic acid led to synthetic peptides with increasing positive net charge. Additionally, we constructed a synthetic, extremely cationic peptide, which was based on a core region of NK-lysin of porcine lymphocytes that may be comparable to the third helix of amoebapores.
In the present study, we investigated the antifungal activity of all synthetic amoebapore analogs constructed so far by determining their minimal growth inhibiting and lethal concentrations against a clinical isolate of Candida albicans. The most active peptides were also brought into action against C. albicans strains from culture collections and were tested against a human keratinocyte cell line to estimate their cytotoxicity assuming a potential topical application. To complete the analysis of their activity spectrum, antibacterial and hemolytic activity of the novel peptides were also determined.
Among the most active peptides found are the newly constructed peptides, i.e. the C1 derivatives and the truncated NK-lysin. Although the pore-forming activity was lower than C1, the novel peptides displayed broad-spectrum activity: they were potent against Candida as well Gram-positive and Gram-negative bacteria. The enhanced efficacy of the substituted analogs compared to their lead peptide C1 may be due to their markedly enhanced solubility in biological buffers. Since the cytotoxic concentration of the C1-analogs and of the shortened NK-lysin were at least an order of magnitude higher than their antimicrobial concentration, these peptides are promising candidates for peptide antibiotics.
1) Medical Microbiology Section
 
 

Calcium-independent Cytolysis of Target Cells Induced by Entamoeba histolytica

O. Berninghausen, C. Benkert, M. Leippe

Early Ca²⁺ influx into target cells was interpreted to be a hallmark of the killing reaction of Entamoeba histolytica. It was our aim to determine the significance of calcium ions in target cell death induced by live trophozoites and amoebapores. Previous findings suggested that cytolysis induced by amoebae is depending on free Ca²⁺: when TMB-8, a Ca²⁺-transporter inhibitor, or EDTA were present during incubation with CHO cells, amoebae failed to kill these cells. We could confirm that observation with the human myeloic cell lines HL-60 and Jurkat; however, we found that in our hands the abolished killing of target cells was due to immobilization and disintegration of amoebae. Interestingly, the amoebae killed HL-60 cells in calcium-free buffer to the same extent as in the presence of 1 mM calcium, and irrespective of the presence of the intracellular calcium chelator BAPTA within target cells.
We have demonstrated previously the potent cytolytic activity of the three amoebapores A, B and C toward human tumor cell lines. In good agreement with our results with live trophozoites, we detected that Ca²⁺ influx into Jurkat cells loaded with the fluorescence probe fura-2 and incubated with a mixture of amoebapores occurred only when the target cells had already compromised membranes due to the action of the pore-forming peptides. The Ca²⁺ influx was detectable markedly later than that induced by equimolar amounts of the Ca²⁺ ionophore A23187. Jurkat cells preincubated with BAPTA prior to the experiments showed a substantially low intracellular Ca²⁺ concentration since the calcium ions have been chelated by the agent. Towards these target cells the degree of amoebapore-mediated killing was not diminished although their intracellular Ca²⁺ concentration did hardly increase. Moreover, amoebapores killed target cells irrespective of the presence or absence of calcium ions in the incubation buffer.
In conclusion, it appears that E. histolytica  and its amoebapores kill target cells, at least those we used here, independent of the induction of an irreversibly elevated Ca²⁺ level within the target cells.
 
 

Mechanism of Target Cell Death Induced by Entamoeba histolytica: Necrosis Versus Apoptosis.

O. Berninghausen, M. Leippe

The human pathogen Entamoeba histolytica is known to kill a variety of host cells including leukocytes. Using human myeloid cells as targets, we studied whether cytotoxicity of amoebic trophozoites in vitro is equivalent to the induction of apoptosis as has been proposed recently or whether these target cells die via necrosis. Upon morphological criteria, incubation of target cells with amoebae resulted in necrosis, since cell swelling, rupture of plasma membrane and release of cell contents including nucleic acids were detected using light and transmission electron microscopy. In contrast, characteristic features of apoptosis such as cell shrinking, surface blebbing, and chromatin condensation were not observed. Moreover, internucleosomal fragmentation of genomic DNA within target cells as a characteristic feature of apoptotic cell death did not occur as judged by the TdT-mediated dUTP Nick End labeling (TUNEL) technique in combination with automated fluorescent cell analysis. Consistently, cleavage of DNA was  detectable upon agarose gel electrophoresis only after a substantial part of the target cell population has already been lysed. We also analyzed the mechanism of cell death induced by amoebapores, pore forming peptides and primary candidate molecules for mediating the cytolytic activity of E. histolytica. At a time point, at which the majority of target cells showed membrane injury upon incubation with purified amoebapores, no DNA degradation was detectable in the victim cells. The data suggest that the target cells used in our study undergo necrosis rather than apoptosis when they are
killed by viable trophozoites as well as by isolated amoebapores.
 
 

Photoaffinity Labeling of the Adherence Lectin of Entamoeba histolytica

S. Hick, M. Leippe, T. Marti

Adherence of E. histolytica trophozoites to human colonic mucus, epithelium and host inflammatory cells is known to be a prerequisite for tissue invasion and destruction. This initial contact is primarily mediated by an amebic surface lectin which recognizes terminal galactose and N-acetylgalactosamine residues of glycoconjugates present on target cells. To identify the carbohydrate binding domain of the 220-kDa adherence lectin, a photoaffinity label was synthesized using a multi-step procedure. The disaccharide ligand N-acetyllactosamine was modified to contain a reactive amino group at its reducing end and subsequently coupled to a heterobifunctional crosslinker. Following purification and radioiodination the selectivity of the ligand was tested by photolabeling the lectin of Erythrina cristagalli, which displays a sugar specificity comparable to that of the E. histolytica adherence receptor. Crosslinking to the E. cristagalli lectin was displaceable at high concentrations of galactose and N-acetyllactosamine, whereas for the E. histolytica lectin a specific photoreaction could not be achieved. As multivalent glycoconjugates are known to display higher binding affinity to the E. histolytica lectin, a short glycopeptide containing a biantennary complex type carbohydrate chain was subsequently isolated from asialofetuin by enzymatic digestion. Following derivatization with a crosslinker and radioiodination, this glycopeptide ligand was used to photolabel the E. histolytica lectin. A specific photocrosslinking reaction was observed with both purified lectin and detergent-extracted membranes of trophozoites. The radiolabeled adherence lectin will now be subjected to enzymatic and chemical degradation. The radioactively labeled peptides will be isolated and the residues forming the carbohydrate binding region identified by amino acid sequence analysis.
 
 
 

A Putative ATPase Expressed in Complement-resistant Trophozoites of Entamoeba histolytica 

B. Urban, H. J. Sievertsen, R. D. Horstmann

Entamoeba histolytica is susceptible to complement attack in its lumen-dwelling state and develops complement resistance during pathogenic tissue invasion. As experimental evidence suggests that this change in phenotype is accompanied by a change in gene expres-sion, we constructed a subtractive cDNA library to identify genes involved. Poly(A)⁺RNA from complement-sensitive trophozoites was subtracted from single stranded cDNA derived from complement-resistant ones. Series of RNA hybridizations revealed three independent cDNA species which correspond to transcripts substantially enriched in complement-resistant trophozoites.
The subtractive cDNA sEhS2 in Northern-blot analyses was represented with an approximately tenfold abundance in both in vitro cultured and tissue-derived forms of complement-resistant E. histolytica. Southern-blot analysis suggested that sEHS2 corresponds to a single-copy gene. The cDNA-derived amino-acid sequence revealed similarity with the evolutionary conserved domains of ATPases of the Sec18/CDC48 family. These proteins mediate as different processes as membrane fusion, protein transport across membranes and transcription regulation.
A part of the amoeba protein was expressed recombinantly and used to raise an antiserum. Subsequent Western-blot analysis identified in amoebae lysates a single antigen of approximately 33 kDa molecular mass, which is in good accordance with 35 kDa calculated for the cDNA-derived polypeptide. The abundance of the protein was five- to tenfold in complement-resistant amoebae but decreased to the levels found in complement-sensitive ones if trophozoites were cultured in the absence of human serum for more than one week. The native protein was detected exclusively in the soluble fraction of E. histolytica lysates indicating its cytoplasmatic localization. It showed di- and trimerization, which could be inhibited by non-ionic detergents. Work is in progress to purify the native protein.
 
 

Hydrogen Peroxide Formation and Disulphide Reduction by an Entamoeba histolytica NADPH:flavin Oxidoreductase 

I. Bruchhaus, S. Richter, W. Stoltenberg, E. Tannich

Entamoeba histolytica normally resides and multiplies within the human gut under anaerobic or microaerophilic conditions. During tissue invasion, E. histolytica becomes confronted with an increased oxygen pressure. Although E. histolytica lacks a conventional respiratory electron transport chain terminating in the reduction of oxygen to water, previous studies indicate that the amoebae respire and tolerate up to 5% oxygen in the gas phase. We recently isolated an E. histolytica gene encoding a protein (Eh34) that shares substantial similarity with a class of disulphide oxidoreductases like the Escherichia coli thioredoxin reductase, the F52 subunit of the Salmonella typhimurium alkyl hydroperoxide reductase and the H₂O₂-forming NADH oxidase from Amphibacillus xylanus. Subsequent biochemical studies of recombinant as well as of native Eh34 revealed: (i) the enzyme catalyzes the reduction of oxygen to hydrogen peroxide as well as the reduction of disulphides like DTNB or cystine, (ii) both reactions are dependent on NADPH but not on NADH, and the enzyme requires flavins as cofactor; (iii) using reagents commonly used to inhibit sulphydryl-dependent reactions, the reduction of disulphides but not the H₂O₂-forming activity is inhibited; (iv) the enzyme is active in a monomeric as well as in a dimeric state, and is localized in the cytoplasm of the amoebae. Most likely, the E. histolytica NADPH:flavin oxidoreductase is involved in maintaining the disulphide redox balance of the cell. In addition, it may serve as scavenger to reduce toxic oxygen concentrations.
 
 

Removal of Hydrogen Peroxide by the Entamoeba histolytica 29-kDa Protein 

I. Bruchhaus, S. Richter, W. Stoltenberg, E. Tannich

During tissue invasion Entamoeba histolytica is exposed to elevated amounts of exogenous reactive oxygen species (ROS) such as superoxide radical anions or hydrogen peroxide, which are highly toxic and, therefore, have to be inactivated. In addition, E. histolytica has to inactivate ROS which are produced by endogenous enzymes. The activity of at least two E. histolytica enzymes, the iron-containing superoxide dismutase (EhFeSOD) and the NADPH:flavin oxidoreductase (Eh34), lead to the formation of hydrogen peroxide. The mechanism that enables E. histolytica to inactivate hydrogen peroxide remains obscure, since the parasite does not possess catalase or peroxidase activity. Recently we identified a gene encoding a 29-kDa protein of E. histolytica (Eh29). This polypeptide was found to have a substantial degree of homology to a new family of antioxidants identified in various prokaryotic and eukaryotic organisms like the AhpC subunit of the Salmonella typhimurium alkyl hydroperoxide reductase and the yeast thiol-specific antioxidant (TSA). The main difference between Eh29 and the other homologues is a cysteine-rich N-terminal extension only present in the E. histolytica molecule. Eh29 and a variant of the protein, which lacks the first 40 N-terminal amino acid residues, were expressed in Escherichia coli and purified to homogeneity. Both recombinant proteins were able to remove hydrogen peroxide although the mutated form was less active. Enzyme activity was dependent on the presence of DTT or amoebic NADPH:flavin oxidoreductase (Eh34), which both are required as hydrogen donors. The activity of Eh29 to inactivate hydrogen peroxide was further supported by results obtained by genetic complementation of the Escherichia coli strain TA4315. This bacterial strain lacks the entire ahp-locus and is highly sensitive to cumene hydroperoxide. Complementation by introducing the gene encoding Eh29 resulted in a substantial reduction of cumene hydroperoxide susceptibility.
 
 
 

Analysis of an Unusual Putative Actin-binding Protein of Entamoeba histolytica 

F. Ebert*), H. Buß*), N. Guillen**), E. Tannich, M. Leippe

In Entamoeba histolytica a variety of actin-binding proteins is involved in rearranging the cytoskeleton as well as in capping and shedding of the membrane. Therefore, this class of molecules is instrumental in important functions of the amoebae such as locomotion, phagocytosis or invasion.
We recently identified an E. histolytica gene containing an uninterrupted open reading frame of 4,698 nucleotides encoding a protein with an estimated molecular mass of 170 kDa. A data base search of the deduced amino acid sequence revealed striking similarities to known actin-binding proteins. However, the structure of the amoeba polypeptide is unusual, in that the N-terminal half revealed homology to coronin, an actin-binding protein of Dictostelium discoidem, whereas the C-terminal half revealed homology to gelsolin and villin which are found in higher eukaryotes, as well as to severin and fragmin found in lower organisms. In addition, substantial homology was found to the so-called head piece domain which is unique for villin and enables binding to F-actin in the absence of ionic calcium.
Fragments representing parts of the N-terminal or C-terminal half of the E. histolytica polypeptide were recombinantly expressed in E. coli and used to raise specific antibodies in rabbits. Immunofluorescence analyses revealed that at least the antiserum against the C-terminal half strongly reacts with caps induced by incubation of E. histolytica trophozoites with Con A. Double staining and confocal laser scanning microscopy revealed co-localization of fluorescent labeled antiserum with Con A-ligand complexes as well as with polymerized actin.
*) Parasitology Section
**) Institut Pasteur, Paris, France
 
 

Identification of an Epitope on the Entamoeba histolytica Galactose/N-acetylgalactosamine-inhibitable Lectin Conferring Antibody Mediated Protection against Invasive Amoebiasis 

H. Lotter, S. L. Stanley*), E. Tannich

The emergence of multi-drug resistant organisms and the failure to eradicate infection by a number of important pathogens has led to increased efforts to develop vaccines to prevent infectious diseases. However, the nature of the immune response to vaccination with a given antigen can be complex and unpredictable. An example is the galactose/N-acetylgalactos-amine-inhibitable lectin, a surface antigen of Entamoeba histolytica which has been identified as a major candidate in a vaccine to prevent amoebiasis. Vaccination with the lectin can induce protective immunity to amoebic liver abscess in some animals, but others of the same species exhibit exacerbations of disease after vaccination. To better understand this phenomenon, we used recombinant proteins corresponding to four distinct domains of the molecule, and synthetic peptides to localize both protective and exacerbative epitopes. Our results indicate that protective immunity after vaccination of gerbils can be correlated with the development of an antibody response to a region of 25 amino acid residues located on the heavy subunit of the lectin. The importance of the antibody response to this region was confirmed by passive immunization studies of scid mice. In addition, exacerbation of disease could be linked to the development of antibodies that bind to an N-terminal domain of the lectin. These findings are clinically relevant, as individuals who are colonized with E. histolytica but are resistant to invasive disease have a high prevalence of antibodies to the protective epitope(s) compared to individuals with a history of invasive amoebiasis. These studies should enable us to develop an improved vaccine for amoebiasis, and provide a model for the identification of protective and exacerbative epitopes of complex antigens.
*) Washington University, St. Louis, MO, USA
 
 

Prevalences of Entamoeba histolytica and E. dispar Infections in Two Villages in Southeast Anatolia 

G. D. Burchard, H. Lotter, B. Walderich*), V. Göral**), D. Britten***), J. Ackers***)

Limited information exists on the prevalence of E. histolytica with E. dispar being distinguished as a separate species. By doing so, Sargeaunt et al. found a prevalence of E. histolytica of 2.6% on the Seychelles and Acuna-Soto et al. one of 11.4% in Mexico. We studied the population of two villages in southeast Anatolia near Diyarbakir. Parasitological stool microscopy was carried out on a total of 238 subjects. In order to differentiate between E. histolytica and E. dispar, a polymerase chain reaction was performed directly with 210 faecal samples using the method of Britten et al. The primers were derived from sequences of repetitive elements in the 25 kDa rDNA episomes of E. histolytica. In addition, sera were analysed for reactive antibodies with ELISAs using crude antigens of E. histolytica, crude antigens of E. dispar, and a recombinant antigen of E. histolytica (rec-P1). The combined prevalence of E. histolytica / E. dispar using stool microscopy was 35% and 29%, in the two villages. The PCR yielded prevalences of 13% and 14%, respectively, for E. dispar, whereas E. histolytica was not found at all. The seroprevalence was 34% when using the
E. histolytica crude antigen, 27% with E. dispar crude antigen, and 0.6% when using the recombinant E. histolytica antigen. Although it cannot be excluded that the PCR may have yielded false-negative results in some cases, the results indicate that the prevalence of
E. histolytica can be very low compared to that of E. dispar. Seroprevalences using crude
E. histolytica antigens apparently reflect previous E. histolytica infections, infections with
E. dispar, or other antigenetically related organisms. They did not prove useful to indicate the actual prevalence of E. histolytica.
*) Institute for Tropical Medicine, Tübingen University
**) Dicle University, Diyarbakir, Turkey
***) London School of Hygiene and Tropical Medicine, United Kingdom
 
 

Cloning and Characterization of a HSP60-related Protein of the Human Filarial Parasite Onchocerca volvulus 

A. Koszarski, J. Bitter*), T. Marti, K. D. Erttmann, M. Gallin

Two-dimensional gel electrophoresis and immunoblot analysis of O. volvulus worm extracts with sera from exposed individuals indicated the presence of immunoreactive proteins related to heat-shock proteins of the 60 kDa group. Subsequent antibody screening of an expression cDNA library of O. volvulus led to the isolation of a full-length cDNA clone encoding for a protein of 550 amino acids (aa), termed O.v. HSP60. The deduced aa sequence shows the highest identity to rickettsial HSP60 (66%) but only of 43% to human HSP60 or to HSP60 of the nematode Caenorhabditis elegans. A phylogenetic tree based on HSP60 aa sequences constructed using the neighbor-joining distance method, indicates the closest relationship of O.v. HSP60 to rickettsial HSP60s. Immunoblot analysis of the recombinant O.v. HSP60 with sera from individuals exposed to infection with O. volvulus indicates that the protein is reactive with sera from all exposed individuals tested (onchocerciasis patients and putatively immune individuals). In contrast to the observed immunodominance of the O.v. HSP60 in exposed individuals, recognition of human HSP60 is restricted to few exposed individuals. The results indicate that we have cloned an HSP60-related protein of O. volvulus which represents the first of a parasitic worm. Reactivity to HSP60 proteins has been implicated in both the pathogenesis and protection against autoimmune and infectious diseases. Further studies with the recombinant O.v. HSP60 are aimed at investigating its role in immunity to O. volvulus.
*) formerly Bernhard Nocht Institute
 
 

Molecular Characterization of the Onchocerca volvulus Antigen S1/aS1 

C. Klosse, T. Kock, M. Gallin, K. D. Erttmann

We have previously cloned an O. volvulus antigen termed S1, based on its recognition by sera from patients with sowda. To further analyse the fine specificity of the immune response we subcloned fragments of the S1 cDNA. Three overlapping fragments encompassing the complete open reading frame of S1 were generated, expressed as non fusion proteins and purified by chromatography. These fragments were subjected to testing with sera and T-cells from individuals exposed to O. volvulus. As the S1 protein is recognized by antibodies from sowda patients but not by the PI tested, studies were begun using sowda sera. Preliminary results indicate that antibodies are most frequently directed against the third fragment which includes the C-terminal portion of the protein, but fragments I and II also are immunogenic to individual patients. Regarding the putative chemokine encoded in the S1 cDNA in antisense orientation, termed aS1, the cDNA encompassing the complete open reading frame of the predicted aS1 protein was cloned in a prokaryotic as well as in an eukaryotic vector expression system. Cloning into the pJC45 expression vector and induction resulted in low levels of expression of a protein of the predicted molecular mass. Cloning into the baculovirus expression system did not provide significant amounts of protein. Efforts to generate sufficient amounts of protein should enable further functional studies of this putative CC-chemokine.
 

The Putatively Protective Onchocerca volvulus Neuronal Protein E1 is a Member of the Death Domain Protein Family 

K. D. Erttmann, A. Domeyer, M. Gallin

Sequence alignment of E1, an ankyrin-related, potentially protective neuronal protein of the human filarial parasite Onchocerca volvulus, with members of the death domain protein family indicates that it contains a death domain (DD) motif. The DD in E1 is most similar to the DD of the human apoptotic molecule Mort1/FADD (39% identity), representing the highest degree of similarity between two members of the death domain protein family to date. Cloning of an E1 homologue from the rodent filarial parasite Litomosoides sigmodontis and sequence comparison of these filarial ankyrins to the ankyrin of the free-living nematode Caenorhabditis elegans defines in addition to the conserved DD motif two further conserved domains of 101 amino acids and 57 amino acids, respectively, whereby the intervening sequences are significantly less or not similar. One of these represents the C-terminal region of the spectrin-binding domain of ankyrins. The other represents a unique domain, most highly conserved between these nematodes, which may be functionally relevant. This domain does not show a significant overall identity to known proteins, but contains a short sequence motif of 21 amino acids present in the calcium-dependent protease calpain. The results suggest that E1 may be involved in apoptosis. This raises the possibility that protection against this parasitic helminth may be induced by apoptotic processes.
 
 

Analysis of Protein Interactions between the Neuronal Protein E1 of Onchocerca volvulus and Other Nematode Proteins Using the Two-hybrid System 

C. Klosse, K. D. Erttmann

The analysis of the amino acid sequence of E1 revealed the presence of three distinct protein domains. Sequence similarities to known proteins suggest that these domains may represent sites of protein-protein interaction. In order to identify those proteins which bind to the domains we are using the yeast two-hybrid system. Since domain III of E1 is highly conserved in O. volvulus and Caenorhabditis elegans we first generated a GAL4 AD fusion expression library from mRNA of C. elegans and cloned domain III of O. volvulus in the appropriate vector to be used as the bait. Clones identified in this fashion will be further characterized on the molecular level. Furthermore a corresponding library of hybrids will also be constructed from O. volvulus to allow screening with the other two domains of E1. These studies should help to elucidate the function of the E1 protein.
 
 

Analysis of E1 Protein Fragments Relevant to the Immune Response of Putatively Immune Individuals 

A. Domeyer, M. Gallin, K. D. Erttmann

To determine the epitopes of the O. volvulus E1 protein relevant to the immune response of putatively immune individuals (PI) we began to analyse fragments of the E1 protein. Three overlapping cDNA fragments encompassing the entire open reading frame of the E1 cDNA, each resulting in protein fragments of 155, 157 and 162 amino acids in length, were obtained by PCR using primers derived from the E1 cDNA sequence. In the same fashion, subfragments of these three were generated resulting in polypeptides of approximately 44 aa in length. The cDNA fragments and subfragments were cloned into the expression vector pJC45, expressed and purified by Ni-chelate affinity chromatography. The recombinant products will be subjected to testing for immunoreactivity using sera and T-cells from PI.
 
 

Host-parasite Interaction in Human Onchocerciasis: Cloning of Human Calgranulin C

T. Kock, K. D. Erttmann, T. Marti, M. Gallin

In order to examine the interaction between the novel human calgranulin C identified in extracts from adult O. volvulus and the parasite we used the primary amino acid sequence to generate synthetic peptides from a region of potential immunogenicity. These peptides were used to generate antibodies in a rabbit. Immunoblot analysis indicated weak reactivity to several bands in whole O. volvulus extract as well as in extracts of human neutrophils. In order to obtain the protein in larger quantities for antibody generation and functional analyses we used PCR with primers derived from the predicted nucleotide sequence to generate a cDNA probe from human cDNA. Screening of a human cDNA expression library led to the identification of a respective clone which is presently being further characterized. Cloning and expression of the human calgranulin C should enable studies aimed at the relevance of this protein of unknown function in the human immune response to O. volvulus.
 
 

Cloning and Immunologic Characterization of a Glyceraldehyde-3-Phosphate-Dehydrogenase (GAPDH) of the Human Filarial Parasite Onchocerca volvulus

E. Schneider, J. Bitter*), T. Marti, D. W. Büttner**), K. D. Erttmann, M. Gallin

To identify the antigenic targets of the antibody response in individuals putatively immune to Onchocerca volvulus (PI) we have analyzed the antibody recognition pattern by two-dimensional SDS-PAGE and immunoblot. One antigen preferentially recognized by PI and by few patients with hyperreactive onchocerciasis (sowda, SOW), but rarely by other onchocerciasis patients (GEN), was subjected to N-terminal amino acid sequencing. The obtained partial sequence showed similarity to mammalian and Caenorhabditis elegans GAPDHs. Glycolytic enzymes play an essential role during the helminthic life cycle and antibodies to GAPDH from the parasitic helminth Schistosoma mansoni have been associated with protection against schistosomiasis. Screening of an expression cDNA library of O. volvulus led to the isolation of a full-length cDNA encoding for 340 amino acids (aa), whereby the deduced aa sequence is most closely related to C. elegans and human GAPDH (81% and 73% identity, respectively). Expression of the open reading frame as a recombinant non-fusion protein with an approximate molecular mass of 38 kDa by SDS-PAGE, was used to test recognition by sera from exposed indviduals. The results show that recognition is indeed restricted to PI and few SOW. Simultaneous testing of the recognition of human GAPDH indicates that the widely distributed recognition of the O. volvulus GAPDH in the PI group is not based on crossreactivity with human GAPDH, reactivity to which was detected in only few sera from the other groups.
Immunohistology using human affinty-purified antibodies against the recombinant O. volvulus GAPDH indicates that the protein is abundant in infective larvae of O. volvulus. However, in O. volvulus adult worms no unambiguous localization was achieved in several samples studied. The results support the view that recognition of this antigen may be associated with protective immune responses in vivo and enable testing of the protective potential in respective animal models. Furthermore, the O. volvulus GAPDH may represent a useful serologic marker for the identification of putatively immune individuals.
*) formerly Bernhard Nocht Institute
**) Parasitology Section
 
 

Identification and Primary Structure of a Novel Human Calgranulin Present in Onchocerca volvulus  Extract

T. Marti, K. D. Erttmann, M.Y. Gallin

A novel calcium-binding protein of the S100 family, termed calgranulin-related protein (CGRP) was purified to homogeneity from O. volvulus adult worm extracts. Its complete primary structure was determined with 300 pmoles of protein, using microanalytical procedures for the generation, isolation and sequence analysis of proteolytic fragments. The primary structure of CGRP consists of 91 residues and displays identity with the previously reported amino-terminal sequence of an S100 protein present in human neutrophils. The human origin of CGRP is supported by the occurrence in O. volvulus extracts of additional human neutrophil proteins, including migration inhibitory factor-related protein 8 and defensins. The results suggest that these proteins interact with the worm surface following their release by activated neutrophils in the course of inflammatory reactions caused by O. volvulus infection. To date, the precise function of calgranulins in myeloid cells is unknown. Besides a possible role as calcium buffers, the proteins could be involved in calcium-dependent signal transduction. For example, a regulatory effect on cytoskeletal components may modulate neutrophil activities, such as migration, chemotaxis, degranulation or phagocytosis. Our current studies are aimed at elucidating the specific role of CGRP in the host-parasite interaction. Recently, CGRP was isolated by another research group from human neutrophils and its amino acid sequence determined, thereby proving the human origin of the protein.
 
 

The Onchocerca volvulus A3 Gene (OvA3) is Regulated by cis, trans and Alternative Splicing in Adult Female Worms

A. Bialonski, P. F. Zipfel.

In order to identify developmentally regulated genes, potentially involved in cell fate decision and cell-cell interaction, we have used conserved regions of the Caenorhabditis elegance glp-1 and lin-12 genes to isolate corresponding clones from Onchocerca volvulus. The isolated full length OvA3 cDNA clone is 1247 nucleotides in length and has an open reading frame of 367 amino acids, with a calculated molecular mass of 41.6 kDa. On the nucleotide level, the sequence displays homology to the lin12, and glp-1 genes of C. elegans, however a much higher degree of homology is identified to open reading frame CRC32 of C. elegans cosmid C26B2. The complete OvA3 gene was isolated and it is organized in seven exons, that cover about 3 kb of genomic DNA. Transcription of this gene is regulated by cis, trans and alternative splicing. Two different cDNA clones were isolated, that represent splice variants for exon 6. This exon is used in cDNA clone OvA8, but not in clone OvA3. Thus sequence analysis clearly demonstrates alternative use of a single exon. The alternative splicing affects the sequence of the protein as within this exon the open reading frame is terminated by a stop codon. Thus the two alternatively processed transcripts encode proteins with different C-terminal ends. Additionally the A3 transcript is processed by trans-splicing. The full length cDNA clone has a spliced leader sequence (SL1) at its 5’ end, a sequence motif present in a number of transcripts in nematodes. This is the first report showing complex proccessing of a single gene by cis, trans and alternative splicing in the nematode O. volvulus, thus demonstrating that this nematode is capable of complex processing of nuclear transcripts. Further work is in progress to elucidate the function of the encoded protein and to understand the regulation and processing of this transcript during development.
Supported in part by BMBF
 
 

Cytokine Response in T Cells Derived from Patients Infected with the Nematode Onchocerca volvulus

S. Schrum, A. Bialonski, B. Fleischer*), P. F. Zipfel

Different subsets of T lymphocytes are important for the outcome of a specific immune response. We are interested in characterizing the human immune response during infection with the nematode Onchocerca volvulus. To this end PBMCs were isolated from infected people from Guinea, West Africa, showing a generalized form of the disease. The proliferative response of these cells towards two differently prepared O. volvulus antigen extracts was determined. The cytokine profile, analyzed on the mRNA- and protein level, showed a response of type 0. On the mRNA level a weak induction of all the analyzed lymphokines and chemokines (IL-2, IFN-g, IL-4,IL-5, MIP-1a and RANTES) was observed, however the corresponding proteins were barely detectable. As PBMCs consist of a variety of cell types and due to the low number of Ag-specific cells in the periphery, we extended the analysis to local T cells, i.e. cells derived from the skin. T cells isolated from skin biopsies were propagated in the presence of PHA. All cloned T cells were CD4 positive and proliferate in response to lectin treatment. Again the cytokine profile of clones shows a type 0 profile, based on expression of IL-2, IFN-g, IL-4, IL-5, MIP-1a and RANTES. In addition, antigen specific T cell clones were generated using two differently prepared antigen extracts. Cells from six patients were cloned in the presence of antigen and were further expanded by PHA. A high fraction of the clones died within one week of culture and none of the surviving clones responded to antigen. However, as the cells were viable, we conclude that these cells reach a state of anergy. The reason(s) that lead to or cause this unresponsive phenotype are currently unclear. By using fractionated antigen extract as stimuli and by including additional costimulatory molecules we will analyze this molecular basis for this unresponsiveness in more detail.
Supported in part by BMBF
*) Medical Microbiology Section
 
 

The Human Factor H-related Protein 4 (FHR-4): A Novel Short Consensus Repeat-containing Protein is Associated with Human Triglyceriderich Lipoproteins

C. Skerka, J. Hellwage, W. Weber*), A.Tilkorn*), F. Buck**), T. Marti, E. Kampen, U. Beisiegel*), P. F. Zipfel

A novel apoprotein of an apparent molecular weight of 86 kDa in its unreduced form was identified in human triglyceride-rich lipoproteins. This protein was purified and the amino acid sequence of six proteolytic fragments was found to overlap with that of factor H-related proteins. In parallel we identified the cDNA encoding a new complement factor H-related protein, termed FHR-4. The sequences of the new apoprotein overlapped with that of the FHR-4 protein. Similar to the previously described factor H-related proteins, FHR-4 contains a hydrophobic signal sequence followed by a stretch of five repetitive elements termed short consensus repeats. Recombinant FHR-4 protein was expressed in the baculovirus system and has an apparent molecular mass of 42 kDa. In addition a 84 kDa dimeric form of the recombinant FHR-4 was detected. Using an immunoaffinity column with antibodies raised against the recombinant FHR-4 a 86 kDa protein was isolated from human plasma. The different molecular weight of the recombinant FHR-4 and the dimeric FHR-4 in plasma is due to different carbohydrate moieties. The 86 kDa plasma protein and the novel apolipoprotein had identical mobility on SDS-PAGE analysis and reacted with antisera raised against the reFHR-4 and the purified apoprotein. In conclusion, we have identified a novel factor H-related protein, FHR-4, and demonstrate that the dimeric form of this protein is present free in plasma and in triglyceride-rich lipoproteins. This observation provides an intriguing new aspect of possible function(s) of this novel protein and the other factor H-related proteins.
Supported in part by DFG
*) Medical Clinic, University Hospital Eppendorf, Hamburg
**) Institute for Cell Biology and Clinical Neurobiology, University of Hamburg, Hamburg
 
 

Binding Analyses of HUMAN Complement FACTOR H to C3b Attached to Activator and Non-activator Surfaces

J. Hellwage, T. Sakari Jokiranta*), S. Kühn, S. Meri*), P. F. Zipfel

Factor H serves as an important regulator of the alternative pathway of complement activation. This protein affects formation as well as dissociation of the alternative complement convertase C3b,5 and serves as a cofactor in factor I mediated degradation of C3b. The activity of the alternative pathway of complement to discriminate targets as either activator or non activators is mediated by different binding properties of factor H to surface associated C3b molecules. We are interested in investigating the recognition mechanism of the alternative pathway of complement activation and in characterizing the role of factor H in the discrimination of activator and non activator surfaces. To this end it is of interest to map the binding sites of factor H to C3b and cell surface structures. For five anti-factor H monoclonal antibodies we have mapped their binding sites within the factor H protein. Two mAb bound to SCR 1, two to SCR 5 and one to SCRs 8-15. The C3b binding sites of factor H have been mapped to SCRs 1-4. As one mAB (131X), that binds to SCRs 8-15 inhibits interaction of factor H with surface bound C3b, we conclude that an additional C3b binding epitope is present within this domain. Further experiments with recombinant fragments show that different domains of the protein are required for interaction and binding to zymosan-C3b and to C3b bound to sheep red blood cells (SRBC-C3b). Further experiments are in progress to elucidate the recognition mechanisms of factor H with activator and non activator surfaces and to identify the precise domains of factor H involved in cell surface interaction.
Supported in part by DFG and DAAD
*) Complement Research Unit, Dept. of Bacteriology and Immunology, Haartmann Institute, University of
 Helsinki, Helsinki, Finland
 
 

Identification and Localization of the C4b and C3b Binding Domains of the Recombinant Cofactor Protein of the Bony Fish Barred Sandbass (Paralabrax nebulifer)

C. Kemper, P. F. Zipfel, I. Gigli*)

In order to elucidate the evolution of regulatory proteins of activation products of the complement system studies were performed in a number of vertebrate species. The most primitive species displaying a complement regulatory system so far identififed is the teleost fish barred sandbass (Paralabrax nebulifer). A specific protease and a cofactor protein were purified in the plasma which cleave the a‘ chain of the human complement components C4 (C4b) and C3 (C3b). Recently we have isolated from a sand bass liver cDNA library a clone (SB1) of 3397 bp in length including a poly -A signal. The open reading frame encodes a translation product of 1053 amino acids with a calculated molecular weight of 110 kDa. The encoded protein (SBP1) has a hydrophobic N-terminal sequence, and the remaining part is organized in 17 repetitive domains called short consensus repeats (SCRs). A homology comparison indicates that individual SCRs of SBP1 display a high degree of homology to SCRs 1,2,3 of human C4bp and SCRs 2, 15, and 19 of human factor H. Given the structural and functional conservation of SBP1, hu C4bp and hu factor H, we were interested in localizing the C4b- and C3b-binding domains in the recombinant sandbass plasma protein. SBP1 and truncated mutants consisting of SCRs 1-5, 1-4, 1-3, and 1-2 were recombinantly expressed in the baculovirus system and analyzed for their ability to bind human C4b and C3b. C3b-binding to the full length protein (SBP1) was significantly weaker than C4b-binding. These results are in agreement with previous data obtained with the native cofactor protein purified from sand bass plasma.
Equimolar concentration of SBP1, SCRs 1-5 and SCRs 1-4 (0.2 mM) bind C4b at concentrations of 0.03 nM up to 0.2 mM. SCRs 1-3 showed less binding activity and SCRs 1-2 failed to bind huC4b. The localization of the C4b-binding domain to SCRs 1-4 of SBP1 is in good agreement with the homology comparison, as the binding domains of human C4bp have also been localized to the N-terminus of the protein. All truncated mutants failed to bind hu C3b. Since at least three C3b-binding sites have been identified in human factor H it is possible that SBP-1 binding to hu C3b requires additional C3-binding domains located in the COOH-terminal end of the protein. Alternatively a single C3b-binding domain within SBP1 may reside beyond SCRs 1-5.
The experiments described here indicate that the regulatory proteins of the complement system are well conserved in evolution. The reagents presently available will permit further investigations on the evolution of the complement system in even lower species.
*) Institute of Molecular Medicine for the Prevention of Human Diseases, The University of Texas, Houston
 Health Science Center, Houston, Tx., USA
 
 

The Zinc Finger Protein EGR-1 is an Important Activator of IL-2 Gene Induction

E. Decker, C. Skerka, P. F. Zipfel

Induction of interleukin 2 (IL-2) synthesis in response to antigen is a critical event for T cell proliferation and effector function. A 300 bp region of the human IL-2 gene promoter comprises binding sites for nuclear factors which confere responsiveness to signals generated by T-cell receptor interaction. Within this region several transcription factor binding sites and the corresponding nuclear factors have been identified in stimulated T cells. One important activator of cytokine gene expression is the nuclear factor of activated T cells (NFAT) which is considered to bind to the IL-2 promoter in combination with AP-1 proteins. We have previously identified a zinc finger protein binding region (termed ZIP) that is located upstream of the distal NFAT site, which serves as an overlapping binding site for the two transcription factors Sp1 and EGR-1. In stimulated Jurkat T cells the ZIP site functions as an activator of IL-2 gene expression. A strong functional cooperation of proteins binding the ZIP and NFAT sites was concluded from transient transfection assays in which a combination of the two sites show a superadditive effect on gene activation. In order to identifiy the factors mediating gene expression through these binding elements and to understand the interaction of these factors co-transfection studies were performed. Using a reporter construct with a threefold ZIP/NFAT region and expression vectors for the various ZIP and NFAT binding proteins, we can demonstrate a strong functional cooperation of the two trancription factors EGR-1 and NFAT. As the two proteins bind to adjacent promoter sites and are in close physical contact we are currently analyzing whether the two factors interact directly.
Supported in part by the Thyssen Stiftung
 
 

Induction of the EGR-3 Gene in T Cells Correlates with the Type of the Immune Response and the Pattern of Secreted Cytokines

P. F. Zipfel, S. Frosch*), E. Kampen, C. Skerka

Mitogenic stimulation of T cells leads to the coordinate induction of four immediate early genes, termed ‘early growth response genes’ (EGR-1 to EGR-4). Based on their homologous zinc finger domains, these genes represent a family of DNA binding proteins. In vitro binding experiments revealed, that the proteins bind sequence specifically to double stranded DNA and that all four factors recognize the same target sequence GCG T/GGG GCG. Corresponding and related motifs are found in the gene promoters of a number of T cell effector molecules, in particular of genes that are differently regulated during a type 1 and a type 2 immune response. We have previously identified a regulatory region in the human IL-2 gene promoter, which is a binding site for the zinc finger proteins EGR-1 and Sp1. As cytokines are important regulators of the immune response we are asking whether EGR proteins do participate in the control of a developing immune reaction. To this end we studied the induction and expression of EGR-genes in antigen specific human and murine T cell clones of the Th1, Th0 or Th2 phenotype. The various clones are characterized according to their pattern of secreted cytokines, and expression of EGR-genes was analyzed in mitogenic treated cells by RT-PCR. While RNA encoding the transcription factors EGR-1 and EGR-2 were detected in all clones analyzed, expression of EGR-3 was specifically regulated. EGR-3 coding mRNA was detected in clones of the Th1 and Th0 phenotype, but not in T cells of the Th2 subtype. This pattern of expression indicates an important regulatory role of the EGR-3 zinc finger protein during the regulation of a specific immune response.
*) Medical Microbiology Section
 
 

Two Human Gene Families Display Preference for Different Nucleotides and Have Distinct Codon Usage Patterns

W. O. Abel*), P. F. Zipfel

The availability of sequence information from a large panel of genes and proteins leads to the identification of common structural motifs and consequently to the grouping of gene families. This wealth of data allows to search for additional parameters useful for a comparison of gene families. Analysis of base composition among two human gene families showed a similar nucleotide distribution for all members of one gene family but revealed significant differences between the two families. The two groups selected for analysis are the human factor H-gene family, that represent six secreted human plasma proteins with functions in the immune defense; and a class of human zinc finger proteins, termed Early Growth Response Genes (EGRs) that represent four sequence specific DNA-binding proteins. The nucleotide distribution of each gene family is distinct. Members of the factor H gene family represent A+T-rich genes, displaying an overall A-T-nucleotide content of 62.8 % and a particular preference for A-nucleotides (33.9 %). In contrast the EGR-genes are G+C-rich (55.9 %) and C-nucleotides are used in 31.2 %. This nucleotide difference is of biological significance as it affects codon usage among synonymous codons. Both gene families select for codons which have the preferred nucleotide at position three. At this silent third position C-nucleotides are used by the EGR-family in 48.1 % of the 1876 analyzed codons, compared to 16 % of the 2503 codons analyzed for the factor H gene family. In contrast the factor H gene family has 36.3 % of the triplets ending with A-nucleotides compared to 10 % of the EGR-family. Thus nucleotide distribution and codon usage is not uniform within the human organism and the described differences most likely represent selection constraints between highly conserved genes with functions in cell cycle regulation and polymorphic genes with functions in the immune response.
*) Arbeitsbereich Genetik, Institut für Allgemeine Botanik, Universität Hamburg
 
 

Coordinate Induction of the Four EGR-Proteins in Activated Jurkat T-Cells and Specific Binding of EGR-1 to the EGR-consensus Motif

C. Skerka, E. L. Decker, P. F. Zipfel

The four EGR-genes (EGR-1 to EGR-4) are coordinately induced upon mitogenic stimulation of resting T-cells and their transcripts can be detected as early as 20 min after stimulation. In order to characterize the biological role of the four EGR-proteins, we studied their induction on the protein level and analyzed the binding of the newly induced cellular proteins to the target sequence. Expression of the four EGR-proteins was analyzed in unstimulated and in stimulated Jurkat T cells by Western-blotting. While the proteins were not detectable in unstimulated cells, all four proteins were found induced 2 h after treatment with PHA and PMA.
In vitro binding experiments with recombinant proteins revealed that each EGR-protein binds to the EGR-consensus motif (GCG GGG GCG). The cell extract from stimulated Jurkat cells, in which expression of all four EGR-proteins was demonstrated was used for DNA binding experiments. Only a single protein was identified in this extract that bound to the EGR-consensus site. Using specific antibodies for supershift experiments this band was shown to represent the EGR-1 protein. Binding of the three other related EGR-proteins to this EGR-consensus motif was excluded by the use of specific antisera. The reasons why in nuclear extracts only EGR-1 binds to the EGR-consensus site are currently under investigation. By using recombinant EGR-1 and AT133/EGR-4 protein expressed in baculovirus infected insect cells, we can demonstrate that these zinc finger proteins bind to the EGR-consensus motif with different affinities. In addition we detected specific binding of EGR-1, but not of EGR-4 to a related G-rich target sequence.
 
 

Assembly of Multihelical Membrane Proteins from Complementary Fragments

A. Jung, T. Marti

Unlike for soluble proteins, the pathway of in vitro folding of integral membrane proteins is poorly understood. This circumstance is primarily a consequence of the intrinsic hydrophobicity, which hampers the expression, purification and structural characterization of transmembrane proteins. We are investigating the process of folding and assembly of bacterio-rhodopsin, a prototypic member of receptors forming a seven-helical bundle structure. In order to test whether these polytopic membrane proteins are composed of autonomous folding domains, an efficient procedure was developed for the production and purification of transmembrane segments. This system uses the lambda cI-repressor/P_L-promoter for the temperature-inducible expression of membrane fragments in E. coli. The recombinant proteins are then extracted from E. coli membranes in a denatured state with an organic solvent mixture containing chloroform, methanol, and triethylamine. By repeated phase partitioning the membrane fragments are enriched to about 50% in the extract. Final purification is achieved by binding the proteins to an ion-exchange matrix via an endogenous, hydrophilic epitope that was introduced into each gene product. Using this methodology, eleven fragments comprising two to five of the transmembrane regions of bacteriorhodopsin have been produced in mg quantities. Upon reconstitution into mixed lipid-detergent micelles, different combinations of complementary fragments could be refolded to the native structure, as judged by the recovery of ligand binding and functional activity. These results suggest that multihelical membrane receptors can be assembled from several independent folding units.
 
 

The Influence of Marker and Disease Allele Frequencies on Genetic Association Studies

B. Müller-Myhsok, L. Abel*)

It has recently been claimed that the future of genetic studies in complex disorders belongs to association designs, such as the TDT, because of the inherently higher power of association studies (Risch and Merikangas, Science 273:1516, 1996). However, this power was computed in the setting that the allele was the disease allele itself. A more common situation is, and could well remain, the analysis of polymorphisms which have a low prior probability to be the disease allele even if they are within the actual disease gene. In this case, we show that the power of the TDT is highly dependant not only on the linkage disequilibrium between the disease allele and the allele analysed, but also on the relative frequencies of both these alleles. The power of association studies such as the TDT can be quite strong when there is a high probability that the allele studied is the causal allele, as shown by Risch and Merikangas. In other cases, scientists should be aware that the power of such association studies can be dramatically diminished as soon as the linkage disequilibrium becomes weaker and the ratio of the frequency of the disease allele and the allele analysed departs from unity.
*) INSERM U 436, Mathematical and Statistical Modelling in Biology and Medicine, Paris, France
 
 

Confirmation of a Locus on Human Chromosome 5q31-q33 Influencing the Intensity of Infection with Schistosoma mansoni

B. Müller-Myhsok, F. F. Stelma*), F. Guissé+), B. Muntau, T. Thye, G. D. Burchard, B. Gryseels+), R. D. Horstmann

In an area hyperendemic for Schistosoma mansoni transmission, others have shown that most individuals present with a low-susceptibility phenotype, that a major gene controls the intensity of infection, and that this gene is located on chromosome 5q31-q33 (Nature Genet. 14:181, 1996). Studying a population recently exposed to epidemic S. mansoni transmission, we found a more balanced distribution of phenotypes, no major gene being involved in controlling the intensity of infection and yet, using non-parametric methods, a significant contribution of a locus on 5q31-q33. The difference in the distribution of phenotypes in the two study groups may provide first evidence for a possible selective pressure of S. mansoni on the host population.
*) Department of Parasitology, University of Leiden, Leiden, The Netherlands
+) Institute for Tropical Medicine, Antwerp, Belgium
 

Hepatic Fibrosis Early in Schistosoma mansoni Infections: Ultrasound and Serological Findings

G. D. Burchard, F. Guissé Sow*), B. Gryseels*)

Due to environmental changes, an epidemic of Schistosomiasis mansoni started in the lower Senegal river basin in 1990. During a survey in 1993, no case of marked hepatosplenic schistosomiasis was found by ultrasound studies in Ndombo, a representative village of the region. Now, in the same village an ultrasound investigation has been carried out in order to (a) again determine the prevalence of a hepatic fibrosis, (b) evaluate various ultrasound classification systems, and (c) compare the findings with serological markers of hepatic fibrosis. A total of  470 village inhabitants were included. The prevalence of S. mansoni in stool samples was 81.5%, the geometric mean of the number of excreted eggs was 246 per g stool. The ultrasound study was carried out using portable equipment (Echo View 350, Shimadzu, Japan) with convex 3.75 Megahertz- and microconvex 3.5 Megahertz transducers. Procollagen-III-peptide, hyaluronic acid, and laminin were determined as serological markers. Twelve patients showed a distinct hepatic fibrosis, a degree III periportal fibrosis according to the Managil classification. Thus, in comparison to the first investigation in 1993, an increase in the prevalence of hepatic fibrosis was found. The Cairo classification did not prove to be satisfactory to classify mild forms of fibrosis as found in this study. The serum concentrations of the procollagen-III-peptide were shown to be elevated in 12.4% of adult inhabitants of the village, and those of hyaluronic acid in 26.8% of cases, but no correlation with ultrasound findings or with egg excretion rates were found (and neither with the occurrence of anti-HBc- or anti-HCV antibodies).
*) Prince Leopold Institute of Tropical Medicine, Antwerp, Belgium
 
 

Delineation of the DFNB1 Locus Linked to Non-syndromic Recessive Deafness Segregating in a Village with High Prevalence of the Disease in Ghana

B. Müller-Myhsok, G. Amedofu*), J. ter Meulen+), T. Thye, B. Muntau, G. K. W. Brobby*), R. D. Horstmann

Of several loci being linked to non-syndromic recessive deafness, DFNB1 on proximal human chromosome 13q has been the first to be localized. It appears to be related to one of the major causes for this disorder as it has been identified in various populations throughout the world. Recently, DFNB1 has been mapped centromeric to marker locus D13S175. In a Ghanaian village with a deafness frequency of approximately 20%, we identified DFNB1 as the gene linked to disease. Additionally we revealed in an affected individual a recombination event with marker D13S1236 located 3 cM centromeric of D13S175, which leaves only the interval between D13S1236 and D13S175 as the location of DFNB1.
*) Department of Ear, Nose and Throat, School of Medical Sciences, University of Science and Technology, Kumasi, Ghana
+) Department of Virology
 
 

Genetic Mapping of Trypanosusceptibility in the F2 Generation of Two Full-sibling N’Dama x Boran Families

A. Gelhaus, O. Hanotte*), R. D. Horstmann, A. J. Teale*)

Trypanosomiasis is considered a major constraint on livestock production in sub-Saharan Africa. The majority of Zebu cattle are susceptible to trypanosomiasis whereas some taurine breeds, mainly N’Dama and West African shorthorns, exhibit a certain degree of resistance termed trypanotolerance. Trypanotolerance is genetically determined but is also influenced by physiological, environmental and nutritional factors.
To study the genetic basis for trypanotolerance, trypanotolerant N’Dama bulls and trypanosusceptible Boran cows were mated at the International Livestock Research Institute (ILRI) in Nairobi to produce three-generation full-sibling families in which trypanotolerance genes were expected to segregate. Experiments using a mouse model had indicated the MHC as a candidate region. Accordingly, the bovine class II BoLA-DRB3 locus was chosen as a candidate gene. In two ILRI families DRB3 polymorphisms were found to be fully informative, and the respective F2 generations were typed using PCR-RFLP. In accordance with established criteria, anaemia was taken as a phenotypic marker for quantifying trypanotolerance, and the data of the packed-red-cell volume measured on day 150 post challenge (PCVd150) were used.
Altogether, 57 F2 animals were included in the analysis. DRB3-alleles were related to PCVd150 data using the computer programmes ‘Mapmaker/exp’ and ‘Mapmaker/qtl’. The analysis revealed linkage between the DRB3 locus and the PCVd150 phenotype indicating the presence of a gene in the MHC region which contributes to trypanotolerance or trypanosusceptibility. Further analysis using t-test statistics indicated that the PCVd150 values of animals carrying the DRB3*1301 allele were significantly lower than those of animals not carrying this allele (p<0.005) suggesting that DRB3*1301 itself is associated with trypanosusceptibility.
*) International Livestock Research Institute, Nairobi, Kenya
 
 

Bovine Analogue of a Human Trypanolytic Factor

A. Rink, B. Urban, M. Hess, B. Förster, D. Mehlitz*), R. D. Horstmann

Cattle are susceptible to Trypanosoma brucei brucei infection whereas humans are primarily resistant presumably because T. b. brucei is readily lysed in human serum. Several lines of evidence indicate that a product of the haptoglobin (Hp) gene family is a major trypanolytic factor of human serum.
We are interested in the mechanism of trypanotolerance expressed by certain cattle breeds and therefore studied the bovine counterpart of the human Hp gene family. We used as a probe a 289-bp PCR product of bovine genomic DNA obtained with a primer pair designed according to the ß chain sequence of the human haptoglobin-related protein. Hybridizations with bovine DNA and bovine liver RNA provided evidence for a single copy gene and a single transcript size, respectively. Sequence data from a full-length liver cDNA and from PCR-amplified introns indicate that the bovine gene corresponds to the human Hp2 gene. As phylogenetic studies have indicated that the Hp2 structure results from an unequal crossing over involving Hp1, it seems that the precursor gene disappeared from the bovine genome. So far, comparisons between a trypanotolerant and -susceptible animal revealed no differences in the Hp coding sequence or genomic structure.
*) Seminar für Tropenveterinärmedizin, Freie Universität Berlin