In vitro peptide synthesis
Expression vectors
Construction of a genomic library
Progress in malaria
Progress in Anaplasmosis
Progress in theileriosis: 1. Construction of Theileria parva genomic libraries and identification of sporozoite and macroschizont-specific genes
Progress in Theileriosis: 2. Identification of genes encoding important protective antigens of sporozoites of Theileria parva
Chairperson: A. F. Barbet
H. H. Geysen
Rapporteur's summary
A technique has been developed for the simultaneous synthesis of many peptides on solid supports in the form of polyethylene rods coated with a polymer of acrylic acid. These peptides can be used to determine those epitopes recognized by antibodies raised in a classical manner against native antigens. The advantage of this technique is that it analyses the immunodominant regions of an antigen directly.
Identified epitopes can be analyzed further to indicate those residues directly involved with antibody binding. The residues which are freely replaceable by nonsimilar amino acids are identified by determining the effect of all possible single amino-acid substitutions on antibody recognition.
Starting with a monoclonal antibody which has been shown both to neutralize foot-and-mouth disease virus strongly and to bind to a discontinuous epitope on that virus, peptides simulating the epitope were determined. An iterative procedure was used based on the progressive definition of amino acids in a random octapeptide sequence. Optimization of the antibody-binding sequence was carried out at several intermediate stages. It was found that W-Q-M-G-H-S was the optimal hexapeptide made from the common L-amino acids. Longer sequences gave no significant increase in binding activity. Further optimization of this hexapeptide was carried out either by the incorporation of flexible spacers (beta-alanine) between residues, or substitution of the L-isomer by the D-isomer. Strongly binding peptides were found to consist of the two elements W-Q-M and H-S separated by a flexible spacer. It was also shown that the element W-Q-M was best incorporated as the D-isomers of each of the residues and the element H-S as the L-isomers.
This general procedure will allow the determination of an antibody-binding peptide for a monoclonal antibody for which the sequence of the immunogen, or even its identity, is completely unknown. It is anticipated that the extensive 'mapping' of the antigen-combining site of antibodies (using this technique) will answer questions about the nature of discontinuous epitopes and the stereochemical requirements for antigen-antibody interactions. How-ever, its use in defining potentially useful peptide immunogens is possibly more important. Using this technique to map substrate-binding sites in enzymes may allow rational design of selective inhibitors for use as drugs.
Points arising from the discussion
As different species can recognize different epitopes on the same antigen, it is hoped to in-corporate degeneracy into the immune response by using incompletely specified peptide immunogens. In this manner a greater cross-reactivity for related antigens may be induced and at the same time this approach would overcome 'holes' in the repertoire and minimize interspecies differences.
W. Rutter
Rapporteur's summary
There are many factors involved in the successful cloning of desired genetic sequences into suitable expression vectors and in subsequently obtaining high levels of expression in various host cells. Obtaining expression involves the inclusion of initiation and termination signals as well as sequences that ensure vector replication and methods for selecting transformed cells. Sometimes product synthesis must be regulated to prevent toxic effects on host cells.
In addition to transcription of gene sequences, effective expression requires proper protein folding and may involve modifications such as glycosylation and intermolecular interactions between amino-acid chains following the translation process. Processing of translation products must be carried out in the proper order and compartment of the host cell. An example can be given of a host secretory cell which contains the protein disulphide isomerase compartmentalized within the endoplasmic reticulum which catalyzes formation of disulphide linkages in proteins just before secretion. The enzyme has two thioredoxin domains at either end with a highly charged connecting sequence which wobbles to allow the appropriate interaction of the sulph-hydryl residues in the, natural configuration. This illustrates the difficulties in preparing such a protein by recombinant DNA technologies.
Recombinant DNA technology has been used to develop a vaccine against the hepatitis B virus. This virus has an outer surface antigen coat and a eve which contains the DNA polymerase and DNA. The vaccine for hepatitis B virus was produced by genetic engineering using the gene for the surface protein antigen which makes the outer coat. Cloning and expression of this gene were carried out in yeast, rather than in bacterial, expression systems as they can produce large quantities of gene product more stably. The yeast vector employed contained the yeast 2u plasmid and the Leu 2 gene (for selection), a yeast promoter and an Escherichia coli origin of replication (to allow plasmid production in bacteria). An additional expression method employed an SV40 cloning vector transformed into COS cells. Unlike cloning in yeast, this method of cloning produces glycosylation of the expressed protein in a normal way. However, glycosylation is not required for immunogenicity.
In characterizing the surface antigen gene it was discovered that multiple initiation sites were present within the viral gene which produced 45,000-, 31,000- and 25,000-dalton molecules whose antigenicity increased with increasing size. This enhanced antigenicity is correlated with increasing protrusion of the peptide from the surface of the viral particle. The P31 surface antigen produced in yeast also assembled into hepatitis-shaped particles which lacked the core antigen and viral DNA. At present scientists are engineering modified gene sequences that carry other important immunogenic gene sequences to enhance the response. For example, sequences from the herpes viral genome have been fused to the hepatitis surface antigen sequence and effectively triggered an anti-herpes humoral response.
Points arising from the discussion
It was not known if cell-mediated immunity was triggered by the viral particle, nor if cell-mediated immunity was necessary in immunity to hepatitis. Antibodies alone can provide passive protection. Although difficult, it is also feasible to employ transfection of genes into cells using shuttle vectors to test for cell-mediated immune responses.
J. Dame
Rapporteur's summary
A method has been developed to generate genomic expression libraries. It entails the generation of gene-sized fragments using digestion with mung bean nuclease in the presence of varying concentrations of formamide. This approach is particularly useful for identifying genes from small parasite genomes when the use of cDNA technology may prove difficult due to low representation of particular mRNA species.
Digestion of nuclear DNA with mung bean nuclease in the presence of formamide releases gene-sized fragments in a number of instances including plasmodial genes for tubulin, actin the small rRNA subunit, the histidine-rich protein and the circumsporozoite (CS) protein. This procedure can also be used for Drosophila ADH genes or for genes in schistosomes and trypanosomes and retroviruses in mammalian DNA.
The most extensive analysis was performed with the CS antigen gene of Plasmodium falciparum, which was isolated from a lambda gill expression library. Clones selected from this library with a range of monoclonal antibodies demonstrated that the whole gene was cloned and, in addition, in most eases expression was not in the form of a B-galactosidase fusion protein. Expression was observed with the gene in either orientation in the vector but most often in an orientation opposite to that of the B-galactosidase gene. It was suggested that transcription was initiated in these clones from a lambda promotor and that the polycystronic transcript was processed in some way to produce the mature CS protein. One clone was in the orientation of the B-galactosidase gene, but was out of phase with the B-gal gene and was probably produced by translational slippage. Another clone was detected with the monoclonal antibodies which produced a fusion protein with B-gal but was unrelated to the CS protein gene. In some eases screening expression libraries with monoclonal antibodies can lead to the selection of clones which are apparently unrelated to the antigen being sought.
Points arising from the discussion
The nature of the cleavage sites with mung bean nuclease are at present unknown and are apparently not related to the AT content of the sequences. It was suggested that since formamide is necessary, local structural transition states between coding and noncoding sequences may be involved. Apparently the enzyme does not cleave at introns.
Optimal formamide concentration was a characteristic of the genome as opposed to particular genes. The particular optimum was widely different for different genomes, noncoding DNA sequences also occurred in libraries.
G.V. Brown
Rapporteur's summary
So far most progress in malaria has been made with the circumsporozoite protein studies, and vaccination trials in humans are planned for later this year. Another important area is the transmission blocking studies of the sexual stages of the parasite. In these studies monoclonal antibodies identifying antigens of 250,000, 60,000 and 55,000 daltons have been shown to block extracellular gametes of Plasmodium falciparum and inhibit fertilization and ookinete stages in the mosquito. A synergistic effect of two monoclonal antibodies on fertilization blockage was noted, whereas neither alone had an effect. Human immune responses naturally appear not to be directed against sexual stages. The monoclonal antibody binding to sexual stages of P falciparum does not, so far, appear to be isolate specific and antigens associated with the sexual stage may be used as part of a combined vaccine.
A large body of evidence demonstrates that immunity is directed at asexual stages of the parasite. Candidate molecules for vaccine development have been identified. Some of these have apparent biological functions while the biological roles of others are unknown. The following molecules are of potential use as vaccines:
(1) the 195, 000-dalton schizont surface antigen; this a protein which is precursor to the 83,000-dalton processed product on the merozoite surface. The molecule appears to be altered following the release of the schizont from the host cell. The 83, 000-dalton merozoite surface protein is shed during the process of erythrocyte entry. Immunization against this molecule appears to confer partial protection.(2) cyto-adherent molecules; these are thought to mediate adherence of infected host erythrocytes to capillary endothelial cells. Although of parasite origin, cyto-adherent molecules appear as knobs on the surface of infected erythrocyte membranes and can be detected by surface labeling techniques. These molecules vary from isolate to isolate. Monoclonal antibodies raised against them inhibit cyto-adherence. The histidine-rich protein is thought to be one of the molecules present in infected erythrocyte knobs. Cyto-adherent molecules are lost during in vitro passage.
(3) ring-infected erythrocyte surface antigen (RESA)
(4) glycophorin binding protein which occurs on the merozoite surface
(5) rhoptry proteins (41,000 daltons) show no variation from parasite to parasite and monoclonal antibodies have been shown to confer some protection.
(6) s-antigen; genes encoding this antigen have been shown to be localized on different chromosomes by pulsed gradient field gel electrophoresis. Some Plasmodium chromosomes appear haploid.
Several of the malaria antigens consist of short tandemly repeated units. Analysis of these repeats within a given antigen and comparison of repeats of different antigens in one stage or at different stages of the lifecycle have revealed a great deal of variation. The biological significance of this variation is unknown.
The new technique of pulsed gradient field gel electrophoresis demonstrated that field isolates had different chromosome-size DNA profiles. It also showed the presence of at least six or seven resolvable chromosomes ranging in size from 800 kb to 2 Mb.
Points arising from the discussion
Immunization with sexual forms resulted in the failure of ookinetes to develop. It was indicated that gametocyte neutralization would require large amounts of antibody. Experiments are in progress on the effect of trypsinization on merozoites.
It was indicated that polydisperse repeats exist in the human genome and one of these has been identified at the 5' end of the human insulin gene. This element has a repeating unit of 15 to 20 bases which is repeated many times. Studies in family members show rapid sequence variation. The role of this insulin polymorphic region is unknown.
A.F. Barbet
Rapporteur's summary
The iodinated antibody screening method of Young and Davis is used to isolate genomic clones containing the gene for the 105, 000-dalton protein which appears to be protective against Anaplasma marginale. Monoclonal antibodies against the 105,000- and a 36,000-dalton A marginale antigen were tested in extensive controlled experiments with infected and uninfected erythrocytes and the purified target proteins. The later threshold of the sensitivity of the assay used was 1 ng of the 105,000-dalton protein.
Partial Sau 3A fragments of A marginale DNA were cloned into the Bam H1 site of pBR 322, and 8,000 recombinants were screened with a polyvalent antibody prepared against the 105,000-dalton protein immunopurified on an anti-105,000-dalton monoclonal antibody absorbent column. Thirteen positive colonies yielded 10 stably expressing clones. Four genomic fragments of different sizes were identified in these clones, three of these appeared, by mapping, to be derived from the same genomic sequence. 35S labeling of transformed bacteria containing any of the four cloned fragments revealed a 105,000-dalton protein that was absent in control bacteria and was specifically precipitated by the antibody used in screening, but did not react with a monoclonal antibody recognizing the authentic 105,000-dalton protein. Two possible explanations for the failure of recognition by the monoclonal antibody were considered. The native 105, 000-dalton molecule might be modified post-transcriptionally. Alternatively, since Southern blots with the cloned gene suggested the presence of three genes, it was possible that another member of a homologous gene family had been cloned. The mature 105,000-dalton protein appeared as a closely spaced doublet in SDS gels and the recombinant protein migrated between the two components.
Points arising from the discussion
It was considered essential that the recombinant protein should be used, either as whole bacteria or after immunopurification, to raise an antiserum that could be tested in reaction with the mature 105, 000-dalton protein. Neutralization of the initial bodies of A. marginale by the anti-recombinant protein antibody could also test the potential of the cloned gene for production of a vaccine.
K. Iams
Rapporteur's summary
Theileria parva DNA has been obtained from piroplasms of infected bovine blood. Lambda gt11 clone libraries were constructed from the purified DNAs with 200- to 500-base pair sonicated fragments. The lambda libraries have been screened with antisera and monoclonal antibodies to identify stage-specific gene sequences. Several sporozoite surface antigen gene sequences have been identified by Dr R. Hall (ILRAD) and two macroschizont-specific sequences have also been identified. Dr P. Conrad (ILRAD) is now using the macroschizont sequences in experiments aimed at differentiating T parva strains. These macroschizont clones and one sporozoite clone have been used to characterize digestion fragments produced by mung bean nuclease treatment of T parva DNAs. The results indicated that T parva DNAs are reduced specifically to gene-size fragments on treatment with mung bean nuclease in the presence of 45 to 50% formamide.
Mung bean nuclease libraries are now being constructed to allow identification of complete gene sequences for sporozoite and macroschizont genes. Some difficulty has been encountered in maintaining the recombinant clones stably in pUC9.
Future work will involve production of T parva genomic libraries from mung bean nuclease treatment of DNAs. These libraries will be used to identify macroschizont-specific genes which will be tested for their ability to trigger in vitro cytotoxic responses on transfection into bovine lymphocytes.
Points arising from the discussion
Instability of recombinants expressing fusion protein in pUC vectors, probably due to leaky repression, has been encountered by others using pUC. Lambda gt11 and other non-pUC expression vectors were recommended.
Purification of macroschizonts from Theileria-infected lymphocytes is being developed using aureolysin and centrifugation techniques. This should allow preparation of pure Theileria DNA from cells grown in vitro. Meanwhile Hoescht dye 3325 which appears to bind preferentially to AT-rich DNA could be used to separate parasite DNA from lymphocyte DNA by isopycnic centrifugation.
It was suggested that in screening for Theileria antigens involved in cytotoxic responses an inhibition assay with standard target cells might be easier to set up. However, other workers have found such inhibition assays difficult. PH-15 cells may be the best to use for transfection to make target cells.
R. Hall
Rapporteur's summary
A series of monoclonal and polyclonal antibodies has been used to screen genomic expression libraries constructed in lambda gt11. Several clones have been identified and characterized. One clone studied encodes the sequence of part of a protein antigen of 18,500 daltons. The antiserum to this protein blocks infection of target cells by sporozoites in vitro. Three other clones encode peptide sequences for a protein of 25,500 daltons, which may be a breakdown product of a 67,000-dalton antigen. The serum to this antigen is also capable of inhibiting infection of target cells in vitro. A number of problems have been encountered in the course of these studies including:
(l) lysogen instability
(2) insert instabilities
(3) reduction of recombination frequencies upon amplification
(4) selection of bovine DNA sequences
(5) no positive clones detected with monoclonal antibodies.
Other studies have been carried out on the translation of mRNA from infected tick salivary glands. These indicate that the level of sporozoite mRNA is very low compared to tick mRNA. Immunoprecipitation of in vitro translated products with one serum revealed a sporozoite mRNA-encoded antigen of 105,000 daltons. However, several other antigens precipitated by this serum are of tick origin. A cDNA library was constructed and screened with this serum and seven clones were isolated. Problems encountered with this approach were low overall levels of sporozoite mRNA and anti-tick antibodies in the sera.
Future research will be directed towards purifying fusion proteins. These will be used for immunization experiments and in vitro blocking experiments to detect genes encoding sporozoite receptors.
Points arising from the discussion
Discussion focused on the validity of the clones identified using anti-sporozoite antibodies. Some examples of antibodies recognizing spurious clones were given. A suggestion was made that these sequences be used to probe mung bean nuclease digests of Theileria DNA for the presence of gene-sized fragments. These experiments are already in progress. A novel approach to identifying the reading frame of some of the genes expressed in the bacteria was suggested using the micropeptide synthesis method described by Geysen.
