Modeling Drug Information for a Prescription – Oriented Knowledge Base on Drugs

There exists little theoretical analysis how to represent knowledge on drugs required for computerized drug-prescription applications. A work package drug information modeling is described which was part of the European GPADE project. We describe the content and structure of a Drug Knowledge Base (DKB) designed to meet the requirements of decision-support systems in the domain of drug therapy, and to facilitate data transfer from various information sources. The definition of the DKB content is derived from the analysis of information requirements at the various stages of the process of the clinical usage of drugs (prescribing, administration, and follow-up). The DKB structure results from the classification of the various data items along two dimensions: (1) entities in the pharmaco-therapeutic domain for which information must be defined (the Pharmaco-Therapeutic Group, the Component, the Manufactured Preparation, and the Presentation), and (2) the validity score of the pharmaco-therapeutic information (international, national, or local).

A General Model of Drug Prescription

Currently, there is no widely accepted structured representation of drug prescription. Nevertheless, a structured representation is required for entering and storing drug prescriptions avoiding free text in computerized systems, and for drug prescription reviews. Derived from part of the work of the European OPADE project, we describe an object-oriented model of drug prescription which incorporates important concepts such as the phase and triggering event concepts. This model can be used to record all drug prescriptions, including infusions, in a structured way avoiding free text. The phase concept allows the storage of sequentially ordered dosage regimens for a drug within the same prescription. The prescription triggering event concept allows recording of the administration of a drug conditional to dates, symptoms and clinical signs, medical procedures, and everyday life events. This model has been implemented within the OPADE project; the corresponding aspects of the user interface are presented to show how this model can be used in practice. Even if other new attributes may be added to the described objects, the structure of this model is suitable for general use in software which requires the entry, storage and processing of drug prescriptions.

Indicators for the Automated Analysis of Drug Prescribing Quality

Irrational and inconsistent drug prescription has considerable impact on morbidity, mortality, health service utilization, and community burden. However, few studies have addressed the methodology of processing the information contained in these drug orders used to study the quality of drug prescriptions and prescriber behavior. We present a comprehensive set of quantitative indicators for the quality of drug prescriptions which can be derived from a drug order. These indicators were constructed using explicit a priori criteria which were previously validated on the basis of scientific data. Automatic computation is straightforward, using a relational database system, such that large sets of prescriptions can be processed with minimal human effort. We illustrate the feasibility and value of this approach by using a large set of 23,000 prescriptions for several diseases, selected from a nationally representative prescriptions database. Our study may result in direct and wide applications in the epidemiology of medical practice and in quality control procedures.

Object-Oriented Modeling and Terminologies for Drug Contraindications

There is no available standardized representation of contraindications that can be used in drug knowledge bases. In this paper, an objectoriented representation of contraindications is proposed that would allow computerized checking of drug prescription safety. It distinguishes four types of contraindications: pathological state, physiological state, findings of investigation procedures, and diagnostic or therapeutic procedures. The vocabulary usable for expressing the contraindications appropriate for automated prescription checking was also investigated. ICDlO (International Classification of Diseases, tenth revision), SNOMED III (Systematized Nomenclature of Medicine), ICPC (International Classification of Primary Care), and ATC (Anatomic Therapeutic and Chemical Classification) were studied as potential sources of standardized vocabulary. A system to support entering data on contraindications into a drug knowledge base, compatible with the proposed structure and using the identified vocabulary sources, is also presented.

Switch junction sequences in PMS2-deficient mice reveal a microhomology-mediated mechanism of Ig class switch recombination

Isotype switching involves a region-specific, nonhomologous recombinational deletion that has been suggested to occur by nonhomologous joining of broken DNA ends. Here, we find increased donor/acceptor homology at switch junctions from PMS2-deficient mice and propose that class switching can occur by microhomology-mediated end-joining. Interestingly, although isotype switching and somatic hypermutation show many parallels, we confirm that PMS2 deficiency has no major effect on the pattern of nucleotide substitutions generated during somatic hypermutation. This finding is in contrast to MSH2 deficiency. With MSH2, the altered pattern of switch recombination and hypermutation suggests parallels in the mechanics of the two processes, whereas the fact that PMS2 deficiency affects only switch recombination may reflect differences in the pathways of break resolution.

The intrinsic hypermutability of antibody heavy and light chain genes decays exponentially

Somatic hypermutation, essential for the affinity maturation of antibodies, is restricted to a small segment of DNA. The upstream boundary is sharp and is probably related to transcription initiation. However, for reasons unknown, the hypermutation domain does not encompass the whole transcription unit, notably the C-region exon. Since analysis of the downstream decay of hypermutation is obscured by sequence-dependent hot and cold spots, we describe a strategy to minimize these fluctuations by computing mutations of different sequences located at similar distances from the promoter. We pool large databases of mutated heavy and light chains and analyse the decay of mutation frequencies. We define an intrinsic decay of probability of mutation that is remarkably similar for heavy and light chains, faster than anticipated and consistent with an exponential fit. Indeed, quite apart from hot spots, the intrinsic probability of mutation at CDR1 can be almost twice that of CDR3. The analysis has mechanistic implications for current and future models of hypermutation.

Human CD1d-glycolipid tetramers generated by in vitro oxidative refolding chromatography

CD1 molecules are specialized in presenting lipids to T lymphocytes, but identification and isolation of CD1-restricted lipid specific T cells has been hampered by the lack of reliable and sensitive techniques. We here report the construction of CD1d-glycolipid tetramers from fully denatured human CD1d molecules by using the technique of oxidative refolding chromatography. We demonstrate that chaperone- and foldase-assisted refolding of denatured CD1d molecules and beta(2)-microglobulin in the presence of synthetic lipids is a rapid method for the generation of functional and specific CD1d tetramers, which unlike previously published protocols ensures isolation of CD1d tetramers loaded with a single lipid species. The use of human CD1d-alpha-galactosylceramide tetramers for ex vivo staining of peripheral blood lymphocytes and intrahepatic T cells from patients with viral liver cirrhosis allowed for the first time simultaneous analysis of frequency and specificity of natural killer T cells in human clinical samples. Application of this protocol to other members of the CD1 family will provide powerful tools to investigate lipid-specific T cell immune responses in health and in disease.

Ligand-independent assembly of recombinant human CD1 by using oxidative refolding chromatography

CD1 is an MHC class I-like antigen-presenting molecule consisting of a heavy chain and beta(2)-microglobulin light chain. The in vitro refolding of synthetic MHC class I molecules has always required the presence of ligand. We report here the use of a folding method using an immobilized chaperone fragment, a protein disulphide isomerase, and a peptidyl-prolyl cis-trans isomerase (oxidative refolding chromatography) for the fast and efficient assembly of ligand-free and ligand-associated CD1a and CD1b, starting with material synthesized in Escherichia coli. The results suggest that "empty" MHC class I-like molecules can assemble and remain stable at physiological temperatures in the absence of ligand. The use of oxidative refolding chromatography thus is extended to encompass complex multisubunit proteins and specifically to members of the extensive, functionally diverse and important immunoglobulin supergene family of proteins, including those for which a ligand has yet to be identified.

In vivo and in vitro studies of immunoglobulin gene somatic hypermutation

Following antigen encounter, two distinct processes modify immunoglobulin genes. The variable region is diversified by somatic hypermutation while the constant region may be changed by class-switch recombination. Although both genetic events can occur concurrently within germinal centre B cells, there are examples of each occurring independently of the other. Here we compare the contributions of class-switch recombination and somatic hypermutation to the diversification of the serum immunoglobulin repertoire and review evidence that suggests that, despite clear differences, the two processes may share some aspects of their mechanism in common.

The molecular biology of CD1

CD1 were the first human differentiation antigens to be identified by monoclonal antibodies. In this review, we summarize some key results from the molecular study of CD1, with particular reference to their relationship to MHC antigens, and to the existence of two distinct groups of CD1 molecules.

Bacterial and yeast chaperones reduce both aggregate formation and cell death in mammalian cell models of Huntington's disease

Huntington's disease (HD) is an autosomal dominant neurodegenerative condition caused by expansions of more than 35 uninterrupted CAG repeats in exon 1 of the huntingtin gene. The CAG repeats in HD and the other seven known diseases caused by CAG codon expansions are translated into long polyglutamine tracts that confer a deleterious gain of function on the mutant proteins. Intraneuronal inclusions comprising aggregates of the relevant mutant proteins are found in the brains of patients with HD and related diseases. It is crucial to determine whether the formation of inclusions is directly pathogenic, because a number of studies have suggested that aggregates may be epiphenomena or even protective. Here, we show that fragments of the bacterial chaperone GroEL and the full-length yeast heat shock protein Hsp104 reduce both aggregate formation and cell death in mammalian cell models of HD, consistent with a causal link between aggregation and pathology.

With the benefit of hindsight

The development of a method to produce monoclonal antibodies against a defined antigen has had profound impact in many branches of research and signalled a shift in the analysis of biological problems. This article describes the background to these developments and their important practical applications that impinge on the relationship between the world of basic science and commerce.

Memory in the B-cell compartment: antibody affinity maturation

In the humoral arm of the immune system, the memory response is not only more quickly elicited and of greater magnitude than the primary response, but it is also different in quality. In the recall response to antigen, the antibodies produced are of higher affinity and of different isotype (typically immunoglobulin G rather than immunoglobulin M). This maturation rests on the antigen dependence of B-cell maturation and is effected by programmed genetic modifications of the immunoglobulin gene loci. Here we consider how the B-cell response to antigen depends on the affinity of the antigen receptor interaction. We also compare and draw parallels between the two processes, which underpin the generation of secondary-response antibodies: V gene somatic hypermutation and immunoglobulin heavy-chain class switching.

The hybridoma revolution: an offshoot of basic research

In this narrative, I describe how my interest in the nature and origin of antibody diversity led me to tackle the problem by using somatic cell genetic techniques. The first hybridoma (an immortal antibody-secreting cell line derived by fusion of a short-lived lymphocyte and a myeloma cell line) was an offshoot of this approach. Although not intended for such purposes, it soon became obvious that this invention had widespread potential in basic research and industry. Indeed, the technique opened new inroads into the study of complex biological substances and became the method of choice to define new differentiation markers. Hybridomas also allowed us to dissect the immune response to a simple antigen and to demonstrate the critical role of somatic mutations in the generation of high affinity antibodies. Now, monoclonal antibodies can be derived and manipulated in vitro, leading to important new developments in therapeutic applications. BioEssays 1999;21:966-973.

Inhibition of apoptosis of a PARP(-)/(-)cell line transfected with PARP DNA-binding domain mutants

We have investigated the possibility of the involvement of PARP in apoptosis, independently of its enzymatic activity. We thus transfected PARP(-)/(-)A11 cells with a DNA construct encoding the PARP DNA-binding domain (DBD) fragment or mutants DBDbd(-), defective in DNA binding to DNA strand breaks, and DBDcl(-), resistant to caspase-3 cleavage. We found that in the absence of PARP, while expression of DBD has only a marginal effect, expression of the mutants strongly inhibits the apoptosis induced by staurosporine, as measured by the binding of annexin V. Moreover, the mutants, but not DBD, inhibit the cleavage of DNA PKcs, suggesting inhibition of activation of caspase-3. In addition, the mutant transfectants are fractionally less susceptible to low doses of an alkylating agent than the DBD transfectant or the original A11 line. The results suggest that the DBD fragment of PARP, apart from its classical role of nick detection and DNA binding, participates in complexes involved in upstream events leading to activation of the caspase cascade.

Both DNA strands of antibody genes are hypermutation targets

During the maturation of the immune response, antibody genes are subjected to localized hypermutation. Mutations are not evenly distributed along the V gene; intrinsic hot spots exist that are correlated with primary sequence motifs. Although the mechanism of hypermutation remains unknown, it has been proposed to exhibit DNA strand polarity because purine residues on the coding strand are more frequently targeted for mutation than pyrimidines. However, this polarity may not be an intrinsic property of the hypermutation mechanism but a consequence of evolutionary-selected peculiarities of V gene sequences. Furthermore, the possibility that both strands are hypermutation targets has received little attention. To discriminate between these possibilities, we have analyzed the average frequency of mutations of each of the three bases of all nucleotide triplets by using large databases taken from both V and non-V mutation targets. We also have reassessed the sequence motifs associated with hot spots. We find that even in non-Ig sequences, A mutates more than T, consistent with a strand-dependent component to targeting. However, the mutation biases of triplets and of their inverted complements are correlated, demonstrating that there is a sequence-specific but strand-independent component to mutational targeting. Thus, there are two aspects of the hypermutation process that are sensitive to local DNA sequences, one that is DNA strand-dependent and the other that is not.

Hot spot focusing of somatic hypermutation in MSH2-deficient mice suggests two stages of mutational targeting

Likely creation of mismatches during somatic hypermutation has stimulated interest in the effect of mismatch repair deficiency on the process. Analysis of unselected mutations in the 3' flank of VH rearrangements in germinal center B cells revealed that MSH2 deficiency caused a 5-fold reduced mutation accumulation. This might reflect ectopic effects of the Msh2 disruption; indeed, the mice exhibit other perturbations within the B cell compartment. However, that MSH2 (or factors dependent upon it) plays a role in the mechanism of mutation fixation is indicated by a strikingly increased focusing of the mutations on intrinsic hot spots. We propose two phases to hypermutation targeting. The first is hot spot focused and MSH2 independent; the second, MSH2-dependent phase yields a more even spread of mutation fixation.

Monitoring and interpreting the intrinsic features of somatic hypermutation

We have used both normal and transgenic mice to analyse the recruitment and targeting of somatic hypermutation to the immunoglobulin loci. We compare methods for analysing hypermutation and discuss how large databases of mutations can be assembled by PCR amplification of the rearranged V-gene flanks from the germinal centre B cells of normal mice as well as by transgene-specific amplification from transgenic B cells. Such studies confirm that hypermutation is preferentially targeted to the immunoglobulin V gene with the bcl6 gene, for example, escaping this intense mutational targeting in germinal centre B cells. We review our data concerning the nature of the hypermutation domain and the targeting of hotspots within that domain. We consider how enhancer-mediated recruitment of hypermutation to the immunoglobulin loci operates in a clonally maintained fashion and illustrate how both the degree of expression and demethylation of the transgene broadly correlate with its mutability.

Low antigen dose favours selection of somatic mutants with hallmarks of antibody affinity maturation

The immunization schedule is critical for the derivation of high-affinity antibodies, low antigen dose being particularly favourable for the development of a more efficient memory response. To analyse the molecular events underpinning this preference, we analysed the early maturation of the response to the hapten 2-phenyloxazolone (phOx) using low and high doses of immunogen. The phOx response is initially dominated by antibodies expressing the VkOx1-Jk5 light chain and the hallmark of the early stages of maturation is the substitution of His 34 by Asn or Gln increasing affinity 10- or eightfold, respectively, and of Tyr 36 by Phe. High-affinity antibodies express mutations at both sites. We cloned and sequenced VkOx1-Jk5 light chains from antigen-specific B cells taken 14 and 21 days after immunization with high and low antigen doses. We found that overall, the derived sequences were more mutated both at longer times and at higher dose. At day 14, His 34 was more frequently mutated at the higher than at the lower dose, while at day 21 the reverse was true. On the other hand, the His 34/Tyr 36 mutation pair was more frequent at low than high doses at both 14 and 21 days. Furthermore, at both times, the low immunization protocol yielded double mutants in cells with a lower mutation background. It appears therefore that while the higher dose may favour the acquisition of individual critical mutations, low-dose immunization favours the selection of a more focused mutational pattern, whereby advantageous mutations are associated with a low mutational background.

Indicators for the automated analysis of drug prescribing quality

Irrational and inconsistent drug prescription has considerable impact on morbidity, mortality, health service utilization, and community burden. However, few studies have addressed the methodology of processing the information contained in these drug orders used to study the quality of drug prescriptions and prescriber behavior. We present a comprehensive set of quantitative indicators for the quality of drug prescriptions which can be derived from a drug order. These indicators were constructed using explicit a priori criteria which were previously validated on the basis of scientific data. Automatic computation is straightforward, using a relational database system, such that large sets of prescriptions can be processed with minimal human effort. We illustrate the feasibility and value of this approach by using a large set of 23,000 prescriptions for several diseases, selected from a nationally representative prescriptions database. Our study may result in direct and wide applications in the epidemiology of medical practice and in quality control procedures.

Object-oriented modeling and terminologies for drug contraindications

There is no available standardized representation of contraindications that can be used in drug knowledge bases. In this paper, an object-oriented representation of contraindications is proposed that would allow computerized checking of drug prescription safety. It distinguishes four types of contraindications: pathological state, physiological state, findings of investigation procedures, and diagnostic or therapeutic procedures. The vocabulary usable for expressing the contraindications appropriate for automated prescription checking was also investigated. ICD10 (International Classification of Diseases, tenth revision), SNOMED III (Systematized Nomenclature of Medicine), ICPC (International Classification of Primary Care), and ATC (Anatomic Therapeutic and Chemical Classification) were studied as potential sources of standardized vocabulary. A system to support entering data on contraindications into a drug knowledge base, compatible with the proposed structure and using the identified vocabulary sources, is also presented.

The 5' hypermutation boundary of kappa chains is independent of local and neighbouring sequences and related to the distance from the initiation of transcription

The hypermutation of antibody genes targets 1-2 kb of DNA which includes the rearranged V(D)J gene segments. The precise nature, location and limits of the targeted region are of considerable interest in terms of the mechanism of hypermutation. We have analyzed the frequency and distribution of mutations in the 5' region of immunoglobulins using several modified kappa transgenes. We found that the position of the boundary, relative to the transcription initiation site, is not affected by the sequence of the V segment or by substituting the kappa chain promoter for a beta-globin promoter. Furthermore, the deletion of the leader intron (containing the hypermutation boundary) does not affect hypermutation per se, but shifts the boundary from the leader intron to the V region such that the distance between the boundary and the site of initiation of transcription remains constant. These results show that the position of the hypermutation boundary (about 185 bases downstream of the site of initiation of transcription) is not defined by the nucleotide sequence but rather by the distance to a fixed upstream position. Although mutations are also observed in the region upstream of the boundary, the frequency at which they occur is one order of magnitude lower relative to the frequency observed in the V segment. Nonetheless this upstream mutation rate remains more than two orders of magnitude higher than that of somatic genes. We discuss possible mechanisms explaining the nature and position of the boundary in the context of an error-prone DNA repair model.

Cost-related information to be provided by computerised drug-prescription systems to promote cost-effective prescribing

Drug prescriptions are a source of steadily increasing healthcare expenditure in most developed countries. As the use of Computerised Drug Prescription Systems (CDPS) increase both in hospital and community-care settings, the potential of such systems to promote cost-effective prescribing and help contain prescription costs should be considered. This article describes the cost-related information that could be provided by a CDPS, namely decision-support information to be provided on-line during the prescription entry, and retrospective information made available by processing stored prescription records. The review also describes design and specification requirements for building a cost-information module that can be used in various health-delivery systems. These are: (i) adequate and well-organised data; (ii) pertinent background knowledge of the domain; and (iii) algorithms that allow adaptation to site-specific features. The propositions expounded in this article result from a part of the work performed during the Optimisation of Drug Prescription using Advanced Informatics (OPADE) European project.

Cells strongly expressing Ig(kappa) transgenes show clonal recruitment of hypermutation: a role for both MAR and the enhancers

The V regions of immunoglobulin kappa transgenes are targets for hypermutation in germinal centre B cells. We show by use of modified transgenes that the recruitment of hypermutation is substantially impaired by deletion of the nuclear matrix attachment region (MAR) which flanks the intron-enhancer (Ei). Decreased mutation is also obtained if Ei, the core region of the kappa3'-enhancer (E3') or the E3'-flank are removed individually. A broad correlation between expression and mutation is indicated not only by the fact that the deletions affecting mutation also give reduced transgene expression, but especially by the finding that, within a single mouse, transgene mutation was considerably reduced in germinal centre B cells that poorly expressed the transgene as compared with strongly expressing cells. We also observed that the diminished mutation in transgenes carrying regulatory element deletions was manifested by an increased proportion of B cells in which the transgene had not been targeted at all for mutation rather than in the extent of mutation accumulation once targeted. Since mutations appear to be incorporated stepwise, the results point to a connection between transcription initiation and the clonal recruitment of hypermutation, with hypermutation being more fastidious than transcription in requiring the presence of a full complement of regulatory elements.

Modifying the sequence of an immunoglobulin V-gene alters the resulting pattern of hypermutation

Affinity maturation of antibodies requires localized hypermutation and antigen selection. Hypermutation is particularly active in certain regions (notably the CDRs of light and heavy chains) due to the local accumulation of hot spots. We have now analyzed the role of individual nucleotides in the origin of hot spots and show that mutability is largely defined by the nucleotide sequence. We compared the mutability profile of wild-type and modified kappa transgenes that contain silent mutations in the CDR1 segment. We found a new hot spot created at the third base of Ser-31 when its wild-type AGT codon was substituted by AGC. Two major hot spots associated with this AGC vanished when Ser-31 was encoded by the synonymous TCA. In addition to these, which were the most prominent changes, there were compensatory alterations in mutability of residues not directly related to the introduced silent mutations, so that the average hypermutation remained constant. Thus, mutations arising early in the immune response, even silent ones, could affect the mutability of critical residues and alter the pattern of affinity maturation. When analyzing hybridomas, we detected such alterations, but they seemed to better correlate with changes in average rather than local mutation rates. Overall, this paper shows how evolution could have optimized the mutability of individual residues to minimize deleterious mutations. Thus, the optimal strategy for affinity maturation may involve the incorporation of multiple point mutations before antigen selection of the relevant cells.

The targeting of somatic hypermutation

Somatic hypermutation does not occur randomly within immunoglobulin V genes but, rather, is preferentially targeted to certain nucleotide positions (hot spots) and away from others (cold spots). Cold spots often coincide with residues essential for V gene folding. Hotspots, which appear to be strategically located to favour affinity maturation, are most frequently located in the CDRs (particularly CDR1) though conserved hotspots are also found at the base of FR3. Hotspots are in part created by local DNA sequence and the strong biases of codon usage in V genes indicate that the genes have evolved such that somatic hypermutation is targeted to those parts of the V where it is likely to prove most useful. These features of mutational hotspots and biased codon usage are also evident in V genes of lower animals suggesting that diversification by strategic targeting of non-templated mutation may have evolved early in antigen receptor evolution.

Nonsense mutations inhibit RNA splicing in a cell-free system: recognition of mutant codon is independent of protein synthesis

Mutations resulting in premature termination codons reduce the corresponding mRNA levels. We describe a cell-free system in which depletion of the mutant immunoglobulin kappa mRNA pool correlates with inefficient splicing and not with RNA decay. Splicing deficiency does not depend on the sequence surrounding the in-frame nonsense codon and can be partially corrected by mutating the methionine initiation codon. Despite the apparent link between translation and low mutant mRNA levels, inefficient splicing is not dependent on protein synthesis. Abnormal splicing of mutant immunoglobulin RNA is observed with B-cell but not with HeLa or T-cell extracts. A nonsense mutant beta-globin RNA is normally spliced by B-cell extract. We propose that the phenomenon exhibits tissue and gene specificity.

Modeling drug information for a prescription-oriented knowledge base on drugs

There exists little theoretical analysis how to represent knowledge on drugs required for computerized drug-prescription applications. A work package drug information modeling is described which was part of the European OPADE project. We describe the content and structure of a Drug Knowledge Base (DKB) designed to meet the requirements of decision-support systems in the domain of drug therapy, and to facilitate data transfer from various information sources. The definition of the DKB content is derived from the analysis of information requirements at the various stages of the process of the clinical usage of drugs (prescribing, administration, and follow-up). The DKB structure results from the classification of the various data items along two dimensions: (1) entities in the pharmaco-therapeutic domain for which information must be defined (the PharmacoTherapeutic Group, the Component, the ManufacturedPreparation, and the Presentation), and (2) the validity score of the pharmaco-therapeutic information (international, national, or local).

A general model of drug prescription

Currently, there is no widely accepted structured representation of drug prescription. Nevertheless, a structured representation is required for entering and storing drug prescriptions avoiding free text in computerized systems, and for drug prescription reviews. Derived from part of the work of the European OPADE project, we describe an object-oriented model of drug prescription which incorporates important concepts such as the phase and triggering event concepts. This model can be used to record all drug prescriptions, including infusions, in a structured way avoiding free text. The phase concept allows the storage of sequentially ordered dosage regimens for a drug within the same prescription. The prescription triggering event concept allows recording of the administration of a drug conditional to dates, symptoms and clinical signs, medical procedures, and everyday life events. This model has been implemented within the OPADE project; the corresponding aspects of the user interface are presented to show how this model can be used in practice. Even if other new attributes may be added to the described objects, the structure of this model is suitable for general use in software which requires the entry, storage and processing of drug prescriptions.

Targeting of non-Ig sequences in place of the V segment by somatic hypermutation

Affinity maturation of antibodies is characterized by localized hypermutation of the DNA around the V segment. Here we show, using mice containing single or multiple transgene constructs, that an immunoglobulin V kappa segment can be replaced by human beta-globin or prokaryotic neo or gpt genes without affecting the rate of hypermutation; the V gene itself is not necessary for recruiting hypermutation. The ability to target hypermutation to heterologous genes in vivo could find more general applications in biology.

A reporter gene to analyse the hypermutation of immunoglobulin genes

The affinity maturation of antibodies is driven by somatic hypermutation which is localized to specific segments of the coding genes. The information available on this process derives from studies in vivo. With the intention of developing new approaches, we have constructed a fusion gene between a kappa chain and a selectable neomycin resistance gene, neor. The neor gene, which includes the SV40 small t intron and polyadenylation site, but not the upstream elements nor its first 12 amino acids, is an in-frame substitution of the FR2-CDR3 fragment of a rearranged V kappa OX1-J kappa 5 gene. Expression of neor activity is therefore dependent on the upstream immunoglobulin sequence. A stop codon was placed in the CDR1 region so that only mutants survive treatment with geneticin sulphate (G418). The effectiveness of the system was tested by transfecting the NS0 myeloma cell line and isolating spontaneous mutants. Neomycin-resistant clones arose at an estimated rate of 1 x 10(-8)/cell division, and over 90% were authentic structural mutants. Unlike the somatic hypermutations, the majority arose by in-frame deletions including the stop codon, although up to 30% involved a point mutation. The reporter gene was then modified by substituting all the sequences downstream of the J kappa 5 with others known to be required for full hypermutation in vivo. Different cell lines were transfected and G418-resistant clones analyzed. No significant increase in the rate of reversion or in the generation of point mutations versus deletions was detected, even using conditioned culture medium. In the presence of azacytidine however, a mutant involving multiple events (single base addition and deletion plus two point mutations) was detected. The reporter gene system therefore seems suitable to test culture conditions and modifications of the host cells aimed at the derivation of an in vitro assay of somatic hypermutation.

Somatic hypermutation

For the generation of secondary response antibodies, immunoglobulin genes are subjected to hypermutation. Cells expressing antibodies with higher affinity are then selected by antigen. Recent clues to the mechanism of hypermutation come from experiments using transgenic mice enabling analysis of the controlling cis-acting elements and the intrinsic features of the hypermutation, dissociated from the effects of antigenic selection.

Universal cloning and direct sequencing of rearranged antibody V genes using C region primers, biotin-captured cDNA and one-side PCR

Polymerase chain reaction (PCR) cloning has greatly facilitated the cloning of heavy and light chain genes from B cells and hybridomas and has been critical for the generation of natural antibody gene libraries for expression in bacteria and on filamentous phages. There remain difficulties, however, in cloning VH and VL genes from a number of mouse and rat hybridoma lines and from B cells from several other species due to insufficient sequence information. Here we describe a rapid and 'universal' strategy for cloning rearranged antibody genes from any species for which the sequence of the C segment(s) are known. First strand synthesis is primed with a biotinylated C region primer and full length cDNA is captured on streptavidin-coated magnetic beads for tailing with dGTP and terminal deoxynucleotidyl transferase. After tailing, the cDNA is captured again, amplified using polyC primers and used for direct sequencing or cloning. The use of C region primers and cDNA capture ensures that this one-side PCR procedure is efficient and rapid as well as being entirely independent of the sequence of the V segment. We demonstrate its application to the direct sequencing or cloning of the H and L chain genes from six mouse and rat hybridomas and propose that the method described will find applications in three areas: (i) cloning rearranged antibody genes in all cases in which cloning with V-J primers is not possible; (ii) repertoire studies in which an unbiased cloning procedure is required for accurate estimate of gene usage; and (iii) generation of VH and VL gene libraries from immunised animals.

Somatic mutation of immunoglobulin lambda chains: a segment of the major intron hypermutates as much as the complementarity-determining regions

The rate and nature of hypermutation of immunoglobulin genes are of prime importance in the affinity maturation of antibodies. Although a considerable body of information has been gathered for kappa light chains, there is much less data for lambda chains. We have derived a large data base of somatic mutants of mouse lambda 1 light chains from Peyer's patches germinal center B cells. The endogenous lambda 1 genes mutate at a rate comparable to that previously found for a kappa transgene (V kappa ox1). There are intrinsic hot spots of mutation common to both in-frame and out-of-frame rearrangements; these hot spots cluster in hypermutating domains. In contrast to the pattern seen for V kappa Ox1, the hot spot clusters are found not only in complementarity-determining region (CDR)1 but also in CDR2 and CDR3; mutations also cluster in the joining/constant region intron. The differences between the pattern of mutations in V kappa Ox1 and lambda 1 light chains are discussed.

Prescription guidelines in OPADE: what are they, how are they used?

Many computerised drug prescription systems have been developed, but they are rarely used in clinical practice; among the reasons are their lack of integration with the functioning of medical institutions and the lack of consideration of general and local clinical practice rules. We present in this paper how OPADE, a computerised drug prescription system does answer this shortcoming by introducing prescription guidelines called Prescribing Principles. We argue that introduction of these Prescribing Principles will not only allow for integration of the computer in medical practice but will also introduce a positive feed back loop in the prescribing process.

Age-related decrease in the proportion of germinal center B cells from mouse Peyer's patches is accompanied by an accumulation of somatic mutations in their immunoglobulin genes

Somatic hypermutation of immunoglobulin genes and the generation of memory B cells seems to take place in germinal centers, which are chronically present in Peyer's patches. Age-associated changes in the germinal center B cell compartment of Peyer's patches and in the mutations of a kappa light chain transgene were analyzed in unimmunized mice. Somatic mutations accumulate in germinal center B cells slowly and continuously to reach an apparent plateau when the animals are around 5 months old. In contrast, the proportion of germinal center B cells reaches a maximum in very young mice (about 2 months old) and decreases progressively thereafter. These results suggest that the highly mutated B cells in older mice arise by the successive accumulation of mutations in memory cells. The data also show that the optimum time for the analysis of hypermutation of transgenes in Peyer's patches is when the mice are about 5 months old.

Conformational isomerism and the diversity of antibodies

The fact that one cell encodes a single antibody sequence does not necessarily mean that the resulting antibody folds into a single structure, although this is a common assumption. Here we challenge this view and suggest that many antibodies do not have a single conformation at the combining site. The basis for this proposal comes from the kinetic analysis of a set of murine hybridomas derived from defined stages of the immune response to 2-phenyl-5-oxazolone (Ox). Among them we have identified three antibodies that exhibit complex hapten-binding kinetics. We observed biphasic or triphasic reactions in stopped-flow fluorescence experiments, indicating that ligand binding involved isomerization, as well as associative steps. The existence of an equilibrium between at least two antibody conformations, with ligands binding preferentially to one form, was deduced from the variation with hapten concentration of the apparent rate of each phase.

Low cytoplasmic mRNA levels of immunoglobulin kappa light chain genes containing nonsense codons correlate with inefficient splicing

We have previously reported down-regulation of mRNA expression of some of the kappa light chain transgenes in a hybridoma derived from a secondary immune response. Of the five heavily mutated transgene copies present in that hybridoma, three included premature stop codons and were poorly represented at the mRNA level. Here we show that the nonsense mutations are the cause of the low mRNA levels. While we found no evidence that the reduction in mRNA abundance was attributable to an increased rate of cytoplasmic mRNA decay, the amount of cytoplasmic mRNA correlated with the accumulation of unspliced transcripts in the nucleus. Similar results were obtained with a chimeric immunoglobulin gene containing a premature chain termination codon in the variable gene segment. We suggest that inhibition of splicing induced by in-frame premature stop codons is an important mechanism for down-regulation of undesirable immunoglobulin transcripts.

OPADE: optimization of drug prescription using advanced informatics

The computerized drug prescription system (CDPS) is an old, almost obsolete, dream, and numerous researchers have developed prototypes or even full fledged packages. But few systems are currently used in clinical practice. The OPADE project, presented in this paper, aims at developing a computerized system which allows for optimization of drug prescription from a medical, patient compliance and economical point of view; the system will be adapted to different European countries, integrated and customized to local medical practice so as to insure user acceptance.

Alternative splicing generates secretory isoforms of human CD1

Human CD1 genes are a family of five non-polymorphic genes that, although homologous to both class I and II major histocompatibility complex genes, map to chromosome 1. Only three of the antigens, CD1a, -b, and -c, have been clustered with monoclonal antibodies. They are noncovalently associated with beta 2-microglobulin and may function as nonclassical antigen-presenting molecules. Here we analyze their expression in mouse myeloma transfectants and human thymocytes and find mRNA splicing complexity. This manifests itself as incomplete splicing, alternative splicing, utilization of cryptic splice sites, and the generation of alternative reading frames. In the case of CD1A transfectants, we demonstrate that the major protein product is secreted and show by amino acid sequence analysis that this is derived from an unspliced transcript. A second major CD1a component appears to be retained intracellularly. The production of alternatively spliced transcripts in the thymus is not a feature of all CD1 genes. Although in the case of CD1A only the transcript encoding the cell surface CD1a isoform is found, CD1C and -E produce complex intrathymic splicing patterns. The CD1C transcripts predict the expression of a secreted CD1c isoform in the human thymus, which we detect in CD1C transfectant culture supernatants. CD1 gene expression is thus characterized by considerable splicing complexity, and the difference between the splicing patterns found in different environments suggests that this is tissue specific.

The 5' boundary of somatic hypermutation in a V kappa gene is in the leader intron

The maturation of the immune response involves the hypermutation of antibody genes and the selection of B cells expressing receptors with improved antigen binding properties. Somatic hypermutation of antibody genes is targeted to a small region approximately 1 kb surrounding the rearranged V gene. The precise definition of the 5' limit is not yet clear since the data base of somatic mutations upstream of the V region is very restricted. The available data suggest that it lies close to the promoter region and this has been used to implicate transcription in the mechanism leading to hypermutation. Here we present an extensive analysis of mutations in the 5' region of a single kappa light chain gene. A large data base from highly mutated sequences was obtained from anti-oxazolone hybridomas expressing the V kappa Ox1-J kappa 5 light chain and from polymerase chain reaction-derived clones from splenic and Peyer's patches of transgenic mice expressing the same V kappa Ox1-J kappa 5 gene combination. Although mutations were found in the 5'-flanking segment, the rate of mutation in the V-J segment was about 20-fold higher. A sharp decline between those two mutation rates is evident but the boundary was found in the leader intron of the V kappa Ox1 gene, about 150 bases downstream of the initiation of transcription site.

Elements regulating somatic hypermutation of an immunoglobulin kappa gene: critical role for the intron enhancer/matrix attachment region

Following encounter with antigen, the immunoglobulin genes in B lymphocytes undergo somatic hypermutation. Most nucleotide substitutions are introduced into a region flanked by the V gene promoter and intron enhancer. Experiments described here using transgenic mice revealed that the V kappa promoter does not contain specific signals since hypermutation was retained on substituting it by a beta-globin promoter. However, both the kappa intron and kappa 3' enhancer regions were found to be essential for full hypermutation. This dependence of hypermutation on both enhancers contrasts with transgene expression in hybridomas in which only the 3' enhancer (and not the intron enhancer) is necessary to achieve high mRNA levels. The results show that full hypermutation depends on multiple elements, removal of some of which may drastically impair but not totally abolish the process.

Transcriptional regulation of HLA-A and -B: differential binding of members of the Rel and IRF families of transcription factors

HLA-A and -B transplantation antigens can be expressed differentially at the basal level and in response to interferons (IFNs). To determine which DNA control elements and nuclear factors are responsible for these differences, HLA-A and -B upstream regulatory regions were used in expression and mobility-shift analyses. The HLA-A enhancer was found to contain two Rel (KBF/NF-kappa B) binding motifs, while the HLA-B enhancer has only one and is transactivated less well by overexpression of the NF-kappa B p65 subunit. On the other hand, the HLA-B IFN response element mediates a much stronger induction by IFNs and has a higher affinity for IRF-1 and -2, which are transcription factors implicated in the regulation of major histocompatibility complex class I genes. These results suggest a molecular basis for the way in which HLA-A and -B loci have adapted to be differentially expressed and to respond to different sets of cytokine signals.

Analysis of somatic hypermutation in mouse Peyer's patches using immunoglobulin kappa light-chain transgenes

We have exploited mice transgenic for an immunoglobulin kappa light chain in order to show that immunoglobulin genes in the B cells of Peyer's patches in unimmunized mice carry a high level of somatic mutations. Most of the mutations are found in the subpopulation of B cells which, based on peanut agglutinin binding, derive from the germinal centers. The number of mutations per clone and their distribution along the variable gene segment (indicative of untemplated point mutations) are very similar to those found in antigen-specific splenic B cells of normal mice after secondary immunization. The mutations accumulate mainly in complementarity-determining region 1, in particular in some specific codons (Ser-26, Ser-31, and Ser-77) which have been previously recognized as intrinsic hypermutational hotspots. These results suggest that, as in the spleen, somatic mutation occurs in B cells which have migrated to the germinal centers, probably as a consequence of stimulation by antigens present in the gut environment. Transgenic animals are increasingly being used to define the signals involved in hypermutation. However, their subsequent study is very time-consuming because it is based on immunization and analysis of hybridomas or antigen-selected cells. We propose that the use of Peyer's patches of unimmunized adult mice offers a reliable and simple approach to analyze hypermutation of transgenes.