AU9439298A - Detection of measles virus-specific antibodies - Google Patents

Description

WO 99/12038 PCT/EP98/05546 DETECTION OF MEASLES VIRUS-SPECIFIC ANTIBODIES The present invention relates to a method for the identification of measles virus specific antibody in a sample, comprising contacting a sample suspected of containing measles virus specific antibody with a measles virus specific glycoprotein recombinantly produces in mammalian cells using a high expression system; and detecting the presence or absence of said measles virus specific antibody in said sample. Preferably, the expression system is based on a togavirus expression system, more preferred on an alphavirus expression system, and most preferred on a Semliki Forest virus expression system. Additionally, the present invention relates to a kit comprising said recombinantly produced glycoproteins. The method of the present invention allows an easy and reliable assay of the immune status of the human with respect to the present or past infection with measles virus. Measles is a major word-wide health problem. In developing countries 60 million cases are reported each year, with a mortality of approximately 1 million children. In the industrialised countries eradication of measles remains elusive and despite vaccination programmes out-breaks continue to occur. Considering the failure to control measles in the developed countries and the continuing high measles related mortality in the developing countries there is a continuing need for improved methods for monitoring immunity and for diagnosis. This need becomes even more urgent, since rare and isolated cases tend to become more difficult to diagnose clinically. Also, diseases with similar skin involvement such as allergies are becoming more frequent while measles tends to become less frequent. With the frequency of measles patients decreasing, medical personnel tends to loose the experience for diagnosing this disease. Moreover, incomplete protection after vaccination can result WO 99/12038 PCT/EP98/05546 2 in clinical measles with less characteristic symptoms (Aaby et al. 1986 J. Infect. Dis. 154:858-863. Edmonson et al. 1990 JAMA 263:2467-2471). All these problems of diagnosis are compounded by the difficulty to diagnose measles in dark pigmented skin. The measles eradication programme of the Word Health Organisation requires an intensive surveillance of measles immunity world-wide. Therefore, surveillance of measles epidemiology and immunity will rely increasingly on serological parameters. It is well known that after contact with measles virus the immune system develops antibodies against different antigenic components of the virus. Measles immunity consists of cellular and humoral effector mechanisms. Cellular immunity is normally not measured directly and the relevance of serum immunoglubulins to susceptibility and immunity is an open issue. Different serological assays measure different subsets of antibodies, but it is not clear which subset reflects the overall immune status. For an antigenic component of the measles virus to be useful for monitoring measles or measles immunity it is critical that the immunoglobulins generated against the measles virus component accurately reflect the measles immune status. In addition, the antigenicity of the antigen used in the immune assay for detecting said immunoglobulins must sufficiently resemble the virus and unspecific binding to the antigen preparation must be low. After contact with measles antibodies are generated against different antigenic components. These include the nucleoprotein and surface glycoproteins. It has been shown that immunoglobulins specific for the nucleoprotein correlate with total measles virus antibodies (Hummel et al. J.Clin Microbiol 30:2874-2880, 1992) and with functional antibodies (Erdman et al. J. Med. Virol. 41:44-48, 1993). This has not been shown for antibodies against surface glycoproteins of the measles virus and it is not clear how antibodies against glycoproteins correlate with measles immunity. To monitor measles immunity after vaccination different serological assays including hemagglutination inhibition, neutralization, complement fixation and immuno assays WO 99/12038 PCT/EP98/05546 3 have been used (Norrby, E. 1962.; Boteler et al. 1983; Erdman et al. 1991; Warnes et al. 1994, 1995; Weigle et al. 1984, Neumann et al. 1985; Cremer et al. 1985, Albrecht et al. 1981). Currently, the most simple assays for measuring measles immunity are immunoassays based on whole measles virus (MV) or virus-infected cells as antigen (Weigle et al. 1984, Cremer et al 1985). (Boteler et al.1983 J. Clin. Microbiol. 17:814-818; Erdman et al. 1991 J. Clin. Microbiol. 29:1466-1471). Commercial diagnostics tests such as Measles ENZYGNOST, supplied by Behringwerke, Mannheim, Germany or "MEASELISA" and "MEASLESTAT M," supplied by Whittaker Bioproducts, Inc., are available for the detection of IgG and IgM antibodies, respectively, and employ simian kidney cells infected with measles virus or the Edmonston vaccine strain of measles virus grown in MK-2 cells as whole virus antigen. (Boteler et al.1983 J. Clin. Microbiol. 17:814-818.) Detection of specific IgM by EIA can be performed in different ways. The serum may be pretreated to remove interfering IgG antibodies. Alternatively, in the capture ELISA, serum IgM is immobilised to the solid support using an anti-human IgM antibody (Erdman et al. 1991 J. Clin. Microbiol. 29: 1466-1471). Diagnosis of measles may be confirmed by virus isolation, by the demonstration of a significant increase in specific immunoglobulin G (lgG) titers, or by the detection of anti-measles virus (MV) IgM antibodies using radioimmunoassays (Arstila et al. 1977 J. Gen. Virol. 34:167-176; Jankowski et al. 1982 Acta Virol. 26:481-487; Vuorimaa et al. 1978 J. Virol. 2:271-278), enzyme-linked immunosorbent assays (ELISA) (Pedersen et al. 1982 Acta Path. Microbiol. Immunol. Scand. Sect. B. 90:153-160; Tuokko and Salmi 1983 Med. Microbiol. Immunol. 171:187-198), direct or indirect fluorescence-antibody techniques (Kleiman et al. 1986 J. Clin. Microbiol. 18:652-657; Minnich and Ray 1980 J. Clin. Microbiol. 12:2865-2867). MV-lgM appears at the time of rash (Forghani et al. 1983 J. Clin. Microbiol. 18:652-657; Lievens and Brunell 1986 J. Clin. Microbiol. 24:391-394; Tuokko and Salmi 1983 Med. Microbiol. Immunol. 171:187-198) and can be detected three days after onset of rash in most individuals (Perry et al. 1993 J Med Virol. 40:235-240; Rossier et al. 1991 J. Clin. Microbiol. 29:1069-1071). IgM peaks on day 7 to 10 and wanes within weeks (Pederson et al. 1986 Vaccine 4:173-178). Since IgM is transient, the demonstration of specific IgM corresponds to a recent primary measles infection (Helfand et al.

WO 99/12038 PCT/EP98/05546 4 1997 J. Infect. Dis. 175:195-199; Lievens and Brunell 1986 J. Clin. Microbiol. 24:391 394). A single serum specimen collected at the appropriate time (Helfand et al. 1997 J. Infect. Dis. 175:195-199) is now generally accepted to be sufficient to diagnose measles (Erdman et al. 1991 J. Clin. Microbiol. 29:1466-1471; Forghani et al. 1983 J. Clin. Microbiol. 18:652-657; Helfand et al. 1997 J. Infect. Dis. 175:195-199. James 1990 Microbiol. Rev. 3:132-152; Ozanne and d'Halewyn 1992 J. Clin. Microbiol. 30:564-569; Rossier et al. 1991 J. Clin. Microbiol. 29:1069-1071; Tuokko and Salmi 1983 Med. Microbiol. Immunol. 171:187-198; Vuorimaa et al. 1978 J. Virol. 2:271 278). Thus, most commercially available enzyme immunoassays (EIAs) rely on whole virus from cell culture. Measles virus grown in culture generally yields a low and variable titers, which adds to the variability of the antigen. Keeping these variables under control is both difficult and costly. Moreover, the stability of whole virus is limited under normal environmental conditions. This restricts the application of assays based on whole virus to the laboratory setting. Antigens with a higher inherent stability could be used to develop assays which could potentially be used under field conditions. The possibility to measure measles-specific antibodies directly in the field would represent a major advantage over current commercially available immuno assays for measles. Thus, there is a need for a more rational and economical method of producing large quantities of measles virus antigens, such as proteins, which are potentially more stable under field conditions as part of an simpler test for monitoring measles disease and immunity. In this respect, recombinant proteins of the measles virus produced by gene technology in heterologous host cells are an alternative to whole virus as antigen. Such a system provides a virtually unlimited supply of measles antigens suitable as antigen for detecting measles-specific antibodies. Moreover expressing proteins by recombinant means would also allow for lower batch to batch variation and greater ease in calibrating and quantifying assays. Among the difficulties of using recombinant glycoproteins is the choice of the suitable expression system. This is essential since proteins expressed in prokaryotic and WO 99/12038 PCT/EP98/05546 5 simple eukaryotic systems, such as yeast, often display immunological characteristics different from their natural counter parts. Such differences can limit the utility of recombinant proteins in assays that require recognition of serum antibodies that bind to native protein. Recombinant baculoviruses were employed for the high-level expression of genes in insect cells. However, the post-translational modifications, such as glycosylation, are different in these cells from those in mammalian cells. This is not necessarily important for proteins which are not glycosylated such as the measles virus nucleoprotein. However, the measles virus fusion and hemagglutinin protein proper glycosylation is impaired when these glycoproteins are expressed in the baculovirus system. Vialard et al. (1990) J. Virol. 64: 37-50. Further, Bellini et al., WO 93/22683 have expressed the measles hemagglutinin and fusion protein encoding genes in a baculovirus expression system. The expressed proteins proved unsatisfactory for the detection of serum antibodies when tested in conventional immunoassays. Mammalian expression systems may, however, also have some disadvantages: Usually the expression is to low to efficiently produce large amounts of the recombinant antigen; due to close contacts between humans and other mammals antibodies against mammalian tissues may cause elevated background levels against contaminating cell debris. In addition, complicated steps may be required to isolate the antigen with sufficient purity. The technical problem underlying the present invention therefore was to overcome the above recited problems and to provide a conveniently usable and reliable detection system for measles virus glycoprotein induced antibody formulation. The solution to said technical problem is achieved by providing the embodiments characterized in the claims. Accordingly, the present invention relates to a method for the identification of measles virus specific antibody in a sample comprising (a) contacting a sample suspected of containing measles virus specific antibody with a measles virus specific glycoprotein recombinantly produces in mammalian cells using a high expression system; and WO 99/12038 PCT/EP98/05546 6 (b) detecting the presence or absence of said measles virus specific antibody in said sample. The term "identification" as used in accordance with the present invention in its broadest aspect is intended to mean that it is determined whether such antibodies are at all present in said sample. Said term further comprises the quantitative and/or qualitative measurement of such antibodies such as the isotype determination. The person skilled in the art, of course, will readily appreciate that "isotype determination" also comprises isotype subclass determination. The term "measles virus specific antibody", in accordance with the present invention, is intended to mean any antibody that essentially only reacts with epitopes specific for a measles virus antigen. Accordingly, said antibody does not or does not significantly cross-react with any other antigen such as antigens of viruses other than measles viruses that belong to the family Paramyxoviridae. The term "measles virus specific glycoprotein" is intended to mean any full-length glycoprotein that corresponds to the naturally corresponding glycoproteins. In addition, said term comprises fragments and derivatives of said glycoproteins that retain the or at least part of the epitope specificity of said naturally occurring measles virus antigens. In this connection, it will be appreciated that the antigens can be the native antigens or can be modified versions thereof. Well known techniques of molecular biology can be used to alter the amino acid sequence of a measles antigen to produce modified versions of the antigen that may be used in accordance with the invention. These techniques are useful to alter patterns of post-translational modification. For instance, changes in the amino acid sequence of a protein can alter its glycosylation or phosphorylation pattern. Such techniques are also useful to provide specific functional moieties that aid efficient expression or purification of recombinantly expressed proteins, inter alia. It will be understood, therefore, that, in accordance with the invention, the entire panoply of recombinant DNA techniques can be employed to provide antigens useful to detecting anti-measles immunoglobulins.

WO 99/12038 PCT/EP98/05546 7 The term "recombinantly produced" comprises any production method that employs recombinant DNA technology. It will be readily evident for the person skilled in the art that "recombinant DNA technology" is not restricted to DNA only but comprises any recombinant nucleic acid technology like, e.g., also recombinant RNA technology. The present invention overcomes many of the disadvantages of the conventional methodology for detecting measles-specific antibodies by allowing one to readily make recombinant measles antigens that serve as sensitive detectors in enzyme immunosorbent assays ("EIAs") and other immunological assays. Due to the teachings of the present invention the advantages of a recombinant measles virus glycoprotein for the detection of measles antibodies including the likely greater stability of the recombinant antigen in comparison to whole virus due to the well-known heat-lability of measles virus, and potentially eliminating the need for refrigeration of the antigen can now be fully appreciated. Furthermore, strain specific modification of the sequence can easily be introduced into the detection system. In general, the method of the invention involves constructing a plasmid DNA, or the like, in which a DNA sequence that encodes a protein of antigenic potential which serves as template for in vitro synthesis of recombinant RNA which can be used both for direct transfection of appropriate host cells as well as production of recombinant virus that only expresses the heterologous portion of a recombinant protein. Within the cells (e.g. after transfection with the recombinant SFV RNA or after infection with the recombinant virus) the recombinant RNA drives its own replication and capping and its own transcription of RNA molecules and leads to massive production of the antigen of interest while competing out the host protein synthesis. See, e.g., Liljestrom and Garoff (1991) BioTechnology 9:1356-1361; Berglund et al. (1993) BioTechnology 11:916-920. Methods for cloning genes, for manipulating the genes to form expression vectors, and for expressing the protein encoded by the gene in a heterologous host are well known, and it will be appreciated that these techniques can be used in accordance with the invention to provide the expression vehicles, host cells, and the like, for WO 99/12038 8 PCT/EP98/05546 expressing cloned genes encoding measles antigens in a host to produce recombinant antigens for use in diagnostic assays, among others. See, e.g., Sambrook et al. MOLECULAR CLONING, A LABORATORY MANUAL,. Second edition, Vol.1-3 (Cold Spring Harbour Laboratory, 1989), Liljestr6m, P., and Garoff, H. 1995. In F.M. Ausubel, et al. (ed.), Current Protocols in Molecular Biology, p16.20.1-16.20.16, Greene Publishing Associates and Wiley Interscience, New York. A variety of expression systems may be used to produce measles antigens in accordance with the invention. For instance, a variety of expression vectors suitable to producing proteins in systems including vaccinia virus (Wild et al. 1992, J. Gen. Virol. 73:359-367; Taylor et al. 1991, J. Virol. 65:4263-4274; Drillien et al. 1988, Proc. Natl. Acad. Sci. U. S. A. 85:1252-1256), canarypox virus (Taylor et al. 1992, Virology 187:321-328), adenovirus (Alkhatib and Briedis, 1988, J. Virol. 62:2718 2727) other systems (Gerald et al. 1986, J. Gen. Virol. 67:2695-2703; Beauverger et al. 1993 J. Virol. Methods 44:199-210) or a togavirus expression system (see below). Antigens produced in accordance with the invention can be used in a variety of immunological assays to detect anti-measles antibodies in an individual which was exposed to or infected by measles virus. In fact, it will be readily appreciated by those of ordinary skill that antigens according to the invention can be used in place of natural virus in practically any immunological assay for detection of measles specific antibodies. The assays include direct and indirect assays, sandwich assays, solid phase assays such as, e.g., those using plates, beads, electronic sensory devices such as electronic sensory chips or any other solid support, and liquid phase assays, inter alia. Assays suitable for use in the invention include those that use primary and secondary antibodies, and those that use antibody binding reagents such as protein A. Moreover, a variety of detection methods can be used in the invention, including colorimetric, fluorescent, phosphorescent, chemiluminescent, luminescent and radioactive, dye concentrating methods, and other detection methods. The present invention is further described by reference to the following, illustrative examples. In these examples, a variety of antibodies were used. The antibodies were WO 99/12038 PCT/EP98/05546 9 produced according to conventional methods and could have been replaced by similar antibodies. In a preferred embodiment of the method of the invention said measles virus specific glycoprotein is the hemagglutinin (H) or the membrane fusion protein. Also comprised by this embodiment is the consecutive or simultaneous use of both glycoproteins for the identification of measles virus specific antibody. The measles virus genome is a negative single-stranded RNA encoding for a small number of viral proteins. During the immune response antibodies are generated against the different proteins including the hemagglutinin protein encoded by the H gene, the fusion protein encoded by the F gene and the nucleoprotein encoded by the N gene (Giraudon and Wild 1985 Virology 144:46-58. Malvoisin and Wild 1990 J. Virol. 64:5160-4. Graves et al. 1984 J. Virol. 49:409. Norrby et al. 1981 Infect Immun. 34:718). Antibodies can prevent measles virus proliferation in vitro as measured by neutralisation assay. In humans, infusion of a solution containing measles-specific antibodies or passive transfer of such antibodies from the mother onto the child can protect against measles disease (Albrecht et al. 1977, J. Pediatrics; Janeway 1949, Bull N.Y. Acad. Med. 21:202-222). Neutralisation of measles virus in a tissue culture and protection against measles by immunoglobulin is due to antibodies directed against the hemagglutinin protein and against the fusion protein (Giraudon and Wild 1985 Virology 144:46-58. Malvoisin and Wild 1990 J. Virol. 64:5160-4). Hemagglutinin protein and fusion protein are associated with the measles virus surface membrane. In contrast to most other measles proteins, both, the hemagglutinin protein and the fusion protein combine several properties relevant to this invention: they are (1) membrane proteins, (2) glycosylated, (3) target proteins of virus eliminating antibodies; (4) potentialy more stable than the whole virus; (5) further more immunoglobulins against these proteins occur during measles infection; (6) antibodies against these proteins persist life long in an individual having been exposed to measles virus either during measles or after vaccination. In accordance with the present invention, it was surprisingly also found that the antibody response against the hemagglutinin reflects the antibody response against the whole measles virus.

WO 99/12038 PCT/EP98/05546 10 Currently the most reliable diagnostic assays for measles are based on the detection of IgM by ELISA using whole MV or MV-infected cells as antigens (Boteler et al. 1983 J. Clin. Microbiol. 17:814-818; Erdman et al. 1991 J. Clin. Microbiol. 29:1466-1471; Weigle et al. 1984 J Clin. Microbiol. 19:376-379). ELISAs based on recombinant proteins detect only a fraction of the total MV antibodies. The time of appearance of protein-specific antibodies is critical for the sensitivity of the assay early after onset of rash. Nucleoprotein specific antibodies may appear somewhat earlier during measles infection and are thought to be the most abundant antibodies early after onset of rash (Graves et al. 1984 J. Virol. 49:409-412; Norrby and Gollmar 1972 Infection and Immunity 6:240-247; Pedersen et al. 1982 Acta Path. Microbiol. Immunol. Scand. Sect. B. 90:153-160), but this protein shows considerable sequence variability (Rima et al. 1995 Vet. Microbiol. 44:127-134). In contrast, lower levels of H-specific antibodies were detected by immunoprecipitation (Graves et al. 1984 J. Virol. 49:409-412), by competition ELISA with H-specific mAbs (Pedersen et al. 1982 Acta Path. Microbiol. Immunol. Scand. Sect. B. 90:153-160) and by comparing complement fixation with immunodiffusion tests (Norrby and Gollmar 1972 Infection and Immunity 6:240-247). In addition these antibodies seemed to develop only later after onset of rash (Graves et al. 1984 J. Virol. 49:409-412). These observations suggested that H-specific antibodies may not be sufficient for diagnosis. By applying the method of the present invention it could surprisingly be demonstrated that considerable amounts of H-specific antibodies are produced within the early days after onset of rash which are readily detectable with an ELISA based on a mammalian expressed recombinant protein. In another preferred embodiment of the present invention said high expression system is based on a togavirus expression system. In a particularly preferred embodiment of the present invention said togavirus is an alphavirus. In an even more preferred embodiment said alphavirus is Semliki Forest virus or Sindbis virus.

WO 99/12038 PCT/EP98/05546 11 The high yield expression systems of the invention based, e.g., on the Semliki Forest Virus system (Liljestrom, P. and H. Garoff. 1991 BioTechnology 9:1356-1361; Berglund et al. 1993, BioTechnology 11:916-920) can be used to express large quantities of foreign proteins and, in accordance with this invention, can provide the necessary processing. For instance, this can be achieved by RNA produced in vitro or belonging to a recombinant SFV particle, encoding the antigen, wherein preferably recombinant RNA comprises the SFV replicase-encoding gene nsP1-P4 followed by a promoter for subgenomic transcription of the antigen RNA sequence. This recombinant RNA allows the expression of the antigen in a wide range of mammalian cells after direct transfection of cells with (a) the recombinant SFV RNA produced in vitro or (b) a DNA vector which directs in vivo the synthesis of the recombinant SFV RNA transcripts or after infection of cells with a recombinant RNA packaged in vivo into SFV particles using cotransfection with packaging-deficient helper RNA molecule. Within the cells (after direct transfection with (a) or (b) or after infection with the recombinant virus) the recombinant RNA drives its own replication and capping and its own transcription of RNA molecules and leads to massive production of the antigen of interest while competing out the host protein synthesis. Liljestrom and Garoff (1991) BioTechnology 9:1356-1361; Berglund et al. (1993) BioTechnology 11:916-920. The Semliki Forest Virus replicon, consisting of a self replicating and self-transcribing molecule using host cell transcriptional machinery system is most preferred because of the high yield of expression. The SFV expression system is preferably used in combination with hamster kidney cell lines such as BHK-21 (see appended examples or e.g. ATCC CCL 10, ATCC CRL 8544). Most preferred in accordance with the method of the invention is that the Semliki Forest virus expression system comprises plasmid pSFV1-MVH, the construction of which is shown in Example 1. In an additionally preferred embodiment of the method of the invention, said sample is derived from a body fluid.

WO 99/12038 PCT/EP98/05546 12 In a particularly preferred embodiment said body fluid is serum, plasma, saliva or cerebrospinal fluid. In a further most preferred embodiment said body fluid is obtained from a patient infected with measles virus, a convalescent having recovered from measles virus infections, a subject immunized with a measles vaccine or a sero-negative individual. Measles virus-specific antibodies are mainly found in the blood (i.e. the plasma or the serum) and to a lesser extent in mucosal fluids such as saliva. The presence of MV-neutralizing and hemagglutination inhibiting antibodies is thought to reflect the immune status of the individual. In particular, these permit to distinguish between individuals with and without immunity against measles. In accordance with the invention it has surprisingly been found that glycoprotein specific antibodies are a better measure of measles immunity than whole measles virus specific antibodies, since specificity, accuracy, positive as well as negative predictive value were better. Both assays did not significantly differ by the numbers of false negative sera. In contrast, the whole virus based ELISA had significantly more false positive sera than the H-ELISA. Measurements of specific measles-lgG is, most importantly, supposed to predict susceptibility to measles infection. In any given cohort, only few individuals will be seronegative for measles. Among these, false positive donors are at risk of disease and can support viral circulation. Identification of such false positive sera would require retesting most sera by using a different assay. On the other hand, individuals tested false negative have no enhanced risk and are epidemiologically irrelevant. Also, rare false negative sera could in principle be retested (with another assay) or such individuals could simply be (re-)vaccinated. For these reasons, false negative results can be better tolerated than false positive results, which were absent in the assay performed in accordance with the invention. Among vaccinated individuals, it is important to detect those with vaccine failure which, in other words, showed no sero-conversion. In accordance with the present invention, it was confirmed that the H-ELISA, in contrast to the MV-ELISA, detects no WO 99/12038 PCT/EP98/05546 13 false positive sera, indicating that the recombinant assay may be more efficient in detecting vaccine failure than conventional assays, which is the main purpose of a IgG detection assay. After vaccination only few individuals will not seroconvert; they will be at risk of disease and support viral transmission. If they are tested false positive, they will not be detected. In contrast, a false negative result represents no risk for the individual concerned nor for his contacts. The few persons with false negative sera could simply be revaccinated together with true negative individuals. An important observation is, that the difference in antibody levels detected between late convalescents and vaccinees by MV-ELISA and NT (Table 4) disappears in the H-ELISA: Median values of late convalescents and vaccinees are essentially the same whether raw or net H-ELISA data are compared. This explains that in vaccinees the sensitivity of the H-ELISA is essentially the same as for the late convalescents (98.9% vs. 99.1%; Table 3), whereas MV-ELISA is less sensitive for vaccinees (99.5% vs. 95.8%). Furthermore, this finding (and the small although significant difference in HI titers), suggests that only in relative but not absolute terms H-specific antibodies differ between late convalescents and vaccinees. Since H-specific antibodies are functionally the most important antibodies, their contribution to protection in the two groups is probably the same. For measles diagnosis, the development of MV-derived antigens is necessary, which can be incorporated into a simplified field test. We show here that some of the properties of the recombinant antigen described here are potentially compatible with testing under field conditions. Because of relatively stable unspecific binding, background subtraction does not seem to be necessary. The short time required for the development represents another considerable advantage. Even when using 30 min raw data no false positive serum is found. Although under these conditions the number of false negative sera increases in comparison to net data (Table 3), it still compares favorably with the MV-ELISA. In an additionally preferred embodiment said antibody is an IgG, IgA or an IgM antibody.

WO 99/12038 PCT/EP98/05546 14 Specific IgM develop only after a first measles virus infection or vaccination. Therefore the detection of specific IgM antibody can be used to diagnose a fresh measles infection, while the detection of IgG indicates that immunity against measles has developed. Specific IgA develops in the blood and on mucosal surfaces such in saliva. Thus, the detection of IgA, IgG and IgM permits distinguishing between fresh measles infection and immunity. In an additionally preferred embodiment said glycoprotein is affixed to a solid support. In a particularly preferred embodiment said solid support is a well of a microtitre plate, a bead, an electronic sensory device or any support compatible with a rapid test format. These embodiments of the present invention find particular application in standard immune assays like EIAs, ELISAs or RIAs. The person skilled in the art is in the position to extend the method of the invention to any other conventional or non conventional assay that is known in the art. In an additionally preferred embodiment said solid support, is comprised in a crude cell extract. In accordance with the present invention, it was demonstrated that expression systems and preferably the SFV expression system which is preferably used in conjunction with BHK 21 cells do not contain components which interfere with the detection of antibodies in the method of the invention in the sense that they would artificially enhance signals: as a result the number of false positive sera remains very low. Further purification of the antigen, although not necessary would only improve the characteristics of the assay. Alternatively, the cell extract may be at least partially purified and an extract comprising the measles glycoprotein(s) be used in the method of the invention. In an additionally preferred embodiment of the method of the invention said detection is obtained without the computation of background values.

WO 99/12038 PCT/EP98/05546 15 Inexpensive and simple tests which could be used under field conditions, as an alternative to whole virus-based ELISA would represent an important step towards measles control. ELISAs based on recombinant proteins would potentially benefit from enhanced stability of the antigen which at the same time provide expressive and reliable data and which can be incorporated into a simplified field test. Such a test requires antigen preparations which give a constant background so that a background subtraction is not necessary, and the result is obtained as a single value read-out. The present invention demonstrates that the combination of the expression system with the antigen and the antigen preparation results in an assay with a stable unspecific binding where background subtraction is not necessary. The short time required for the developement represents another considerable advantage. Even when using 30 min raw data no false positive serum was found. Although under these conditions the number of false negative sera increases in comparison to net data, it still compares favorably with the MV-ELISA. Accordingly, important properties of the recombinant antigen described here are potentially compatible with testing under field conditions. In a further preferred embodiment of the method of the invention, said identification comprises the identification of specific measles strains. Differences in strain characteristics may be due to the local origin of said virus. Since it is well known that measles virus strains such as from different origins of the world have, as a rule, at least slightly different immunological characteristics such as at least slightly different epitopes, the panel of antibodies generated thereto varies with virus strains. Accordingly, antigens, such as H antigens recombinantly produced from different MV strains will yield different read-out intensities when tested in e.g. conventional immunoassays, such as ELISAs with one and the same serum from e.g. a sero-positive individual. The interpretation of the data obtained with such a test system will allow the person skilled in the art to draw conclusions with respect to the nature and/or origin of the virus strain. The invention also relates to a kit comprising WO 99/12038 PCT/EP98/05546 16 (a) recombinantly produced glycoproteins as identified in any of the preceeding methods either in solution or immobilized on a solid support; and (b) reagents suitable for the detection of human antibodies bound to said glycoproteins. Such reagents may be, for example, antibodies specific for human antibodies which are coupled to a detectable marker. The kit of the invention may contain the H protein of one or several vaccine strains and/or of one or several wild-type strains. The combination of H protein molecules derived from different viruses will serve to identify different measles virus strains which have induced immunity in a given individual. In addition, the kit of the invention may be used for any of the purposes described herein above. The figures show: Figure 1 is a schematic diagram showing the strategy for the construction of recombinant plasmids used in the reconstitution of the full-length cDNA of the recombinant H protein and its transfer to the expression vector pSFV-1. Restriction endonuclease digests were performed as indicated by hatched and strippelt areas representing coding regions of the measles H protein and the vectors or plasmids used. Figure 2 shows the detection of recombinant H protein (panel A) on transfected BHK-21 cells (BHK-H) with an anti-MV mAb (BH47; Fournier et al. 1997. J Gen. Virol. In press) by flow cytometry. An MV fusion protein-specific mAb (A352; anti MV-F) served as an irrelevant control antibody; beta-galactosidase transfected BHK-21 cells (BHK-gal, panel B) served as negative control cells. Data are expressed in arbitrary fluorescence units (AFU). Figure 3 shows the binding of MV-H antigen preparation to ELISA plates after coating with increasing concentrations of total membrane protein (ng/per well).

WO 99/12038 PCT/EP98/05546 17 Recombinant H protein was measured using MV-H specific mAb (BH216, Ziegler et al. 1996 J. Gen. Virol. 77:2479-2489) an irrelevant mAb, a positive (HS6111) and a negative (HS3071) human measles immune serum (HS). Data are shown as net mO.D. measured after 1 hr. Figure 4 shows the correlation between the H-ELISA (A, B) or the MV-ELISA (C, D) with HI (A, C) and NT (B, D) titers (expressed as log2). Each point corresponds to one test serum (n=228 vaccinees). The coefficient of determination was calculated on the basis of the best polynomial trendline (order 2). Figure 5 shows the correlation between H-ELISA and MV-ELISA of all sera (A; n=228), of sera which are HI/NT double negative (B; n=11) or HI/NT double positive (C; n=212). Values below and above the gray zone (H-ELISA: 80-120 mO.D.; MV ELISA: 100-200 mO.D.) are considered negative and positive respectively. Values inside the gray zone are undefined. HI/NT titers (expressed as log2 dilutions) of false negative sera are shown in () (panel C). The X- and Y-axis of panel C are truncated at 500 and 1400 mO.D. Figure 6 shows HI and NT titers (expressed as log2 dilutions) of sera negative by H ELISA (panel A) or MV-ELISA (panel C) and of sera positive by H-ELISA (B) or MV sera (D). Numbers (n) of sera are shown. Figure 7 shows the standardization of the H-ELISA with dilutions (in dilution buffer) of the 2nd International Standard for anti-measles serum. The undiluted Standard contains 5000 mlU. The highest concentration tested corresponds to the dilution used for measuring the test sera (1:300). Figure 8 shows the correlation between raw O.D. values (0.D. BHK-H) and net O.D. values (O.D.BHK-H-O.D.BHK-ga) of the same H-ELISA of 388 vaccinated first year high school children (R 2 =0.986).

WO 99/12038 18 PCT/EP98/05546 Figure 9 shows the correlation of IgM measured with a certified commercial IgM assay based on whole MV with the H-ELISA. mO.D. of the MV-ELISA below 200 are considered negative. Sera were obtained from acute and convalescent phase measles patients. Both panels show the same individuals using either an enzyme linked monoclonal antibody specific for human IgM (panel A) or a polyclonal m chain specific conjugate. Figure 10 shows a case definition by IgM in MV-ELISA, CDC criteria, or increase in HI, NT and specific IgG levels (found in all paired sera) of 70 measles patients from which single, paired (or multiple) sera were available. For most patients CDC criteria were not evaluated. (A) and (B) correspond to patients A and B of Figure 11. Figure 11 shows a comparison of IgM values by H-ELISA and MV-ELISA of 112 serum samples drawn from 70 acute phase measles patients. Thresholds for positivity are 147 mOD in the H-ELISA (dotted line, as defined in text) and 200 mOD for the MV-ELISA (IgM-Enzygnost). The grey zone (100 to 200 mOD) is undefined according to the manufacturer. Single and paired sera of patient A, A' and B, B', respectively are indicated. The different symbols correspond to sera drawn before onset of rash (n=4, A), day 0 to 19 (0) and day 20 to 59 (0) after onset of rash (n=108). Figure 12A shows IgM values by H-ELISA (0) and MV-ELISA (0) before (n=4) and after onset of rash (n=108). Panel A displays the moving average based on a period of 4, for H-ELISA (solid line) and MV-ELISA (broken line). In Figure 12B grey bars represent total number sera tested by H-ELISA and MV-ELISA for each time interval; open and closed bars describe the number of sera tested positive by the MV-ELISA and the H-ELISA, respectively. Day 0 corresponds to the day of onset of rash. Figure 13 shows IgG values by H-ELISA (0, solid line) and MV-ELISA (O, broken line) before (n=4) and after onset of rash (n=108). The trendline represents the moving average was based on 4 sera.

WO 99/12038 PCT/EP98/05546 19 Figure 14A shows IgA values by H-ELISA before (n=4) and after onset of rash (n=108). 16 negative sera are shown for comparison (categorie before day 0). Figure 14B shows IgA versus IgG values by H-ELISA of the sera of panel A (without the 16 negative sera). The examples illustrate the invention. Example 1: Production of recombinant measles virus hemagglutinin protein (Figure 1) Cloning of full-length cDNA encoding H protein. Three overlapping cDNA fragments (A, B, C) of the MV-H protein were obtained by RT-PCR from total RNA of virus-infected Vero cells (Edmonston strain). They were cloned separately into the pAMP10 vector (pAMP10AHMV, pAMP10BHMV, pAMP10CHMV) by standard cloning techniques. These three cDNA fragments were transferred into the unique BamH I site of pUC-18, to facilitate further manipulations (PUCAMV, PUCBMV, PUCCMV). After appropriate digestion of these recombinant pUC-18 vectors, fragments were ligated together to generate a pUC-18 containing the full-length cDNA in its BamH I site (pUC-HMVrev), which was then transferred into the BamH I site of the pSFV1 plasmid (PSFV1-HMV). The plasmids were sequenced by asymmetric PCR on a ABI Prism 377 sequencer (Perkin Elmer).

WO 99/12038 PCT/EP98/05546 20 The full length hemagglutinin cDNA insert was 1857 kb in length, originating from 3 basepairs upstream from the first in-frame AUG, and extending through the stop codon. The first basepair (nucleotide A) directly upstream from ATG was changed into a C nucleotide to have an better ribosome binding site (Kozak 1989, J. Cell. Biol. 108:229-241) In comparison with the nucleotide sequence of the wild type Edmonston strain (EMBL-ID: PANP) sequencing revealed 3 silent changes: A to G at nucleotide 631, T to C at nucleotide 1083; G to A at nucleotide 1306; A to T at nucleotide 1441. A mutation at position 1649, where a C was replaced by a T, resulted in a phenylalanin instead of a serine. Nucleotide A of the first in-frame ATG corresponds to position +1. In vitro transcription and electroporation. As described by Liljestrom and Garoff (1995, In F.M. Ausubel et al. (ed.), Current Protocols in Molecular Biology, p16.20.1 16.20.16, Greene Publishing Associates and Wiley Interscience, New York.) pSFV1 MVH and pSFV3-lacZ (Liljestroem, P. and H. Garoff. 1991 BioTechnology 9:1356 1361) were linearized by Spe I digestion and purified by standard phenol/chloroform extraction and ethanol precipitation. Transcription reactions were carried out at 37*C for 1 hr in a 50 pl volume containing 1-5 pg of linearized DNA, 40 mM Hepes-KOH (pH7.4), 6 mM MgCI 2 , 2 mM spermidine-HCI, 5 mM dithiothreitol (DTT), 1 mM of ATP, CTP and UTP, 0.5 mM of GTP, 1 mM of m 7 rG(5')ppp(5')G, 50 units of RNAsin and 30 units of SP6 RNA polymerase. After incubation, synthesized RNA was purified by isopropanol precipitation and 1 ml of RNA was analyzed on a 0.5% agarose gel. Parallel transfections were performed with RNA generated from the control plasmid pSFV-lacZ. The latter transfectants served as negative antigen control cells. Electroporations were performed as described by Liljestr6m and Garoff 1995. After trypsinisation, late log-phase cells were washed in PBS (without Mg 2 + or Ca 2 +) resuspended at 107 cells/mI. 800 ml of cells were mixed with the 5-10 pg RNA purified from the transcription reaction and were transferred to a 0.4 cm chilled electroporation cuvette. Electroporation was carried out at room temperature with two consecutive pulses of 850 kV/25 mF (Bio-Rad Gene Pulser). Cells were then diluted 20-fold in complete medium and transferred back into tissue culture flasks and kept under standard tissue culture conditions. BHK-21 cells transfected with the H- WO 99/12038 PCT/EP98/05546 21 recombinant SFV1 RNA are referred to as BHK-H, control cells mock-transfected with b-galactosidase SFV RNA are called BHK-gal. Example 2: Expression and production of hemagglutinin antigen from in vivo packaged recombinant SFV particles. In vitro transcription of RNA from recombinant pSFV-MVH and the pSFV helper 2 plasmids were performed as described in example 1. The two RNA transcripts (mixed in a 1:1 ratio) were used to transfect BHK-21 cells as described in example 1. As a result of trans-complementation, recombinant RNAs were packaged into SFV particles which are released into the medium. Berglund et al. (1993) BioTechnology 11: 916-920. 24 hours later, the supernatant was collected and cleared by centrifugation (15 min., 2000xg). The recombinant virus initiates only a single round of intracellular replication. Therefore, the titer of the packaged stock cannot be determined by conventional plaque assay, but was tested by indirect immunofluorescence. Cells were infected with different dilutions of virus stock and hemagglutinin expression was detected by flow cytometry as described in example 3. The supernatant containing the recombinant virus was aliquoted, frozen in liquid nitrogen and stored at -80*C. Alternatively, the hemagglutinin recombinant SFV was concentrated and purified from the medium by sucrose gradient centrifugation. Medium collected from transfected BHK-21 cells was layered on a sucrose gradient (consisting of 1ml of 55% sucrose and 3 ml of 20% sucrose) and centrifuged for 90 min. at 160000xg. The virus was collected from the 20/55% sucrose interphase. The medium was aspirated from above with 0.8 ml of the 55% sucrose. From the bottom of the tube recombinant virus was harvested in a total volume of 1 ml and then aliquoted and frozen as described above. For infecting tissue culture cells, recombinant virus was activated with 0,5 volume of a-chymotrypsine solution (10 mg/ml PBS and 10 mM MgCI 2 and 20 mM CaCI 2 ; 20 min, room temperature). a-chymotrypsine was inactivated with 0,5 volume of 2 mg/ml of aprotinin. Activated virus was added to washed BHK-21 cells grown to 80% confluency.

WO 99/12038 PCT/EP98/05546 22 The infected cells expressed high levels of measles hemagglutinin and their crude membrane extract was an antigen preparation suitable for the detection of measles immunity. Example 3: Detection of measles virus hemagglutinin antibodies by flow cytometry (Figure 2) Transfected cells were harvested after 18 hrs, washed and resuspended in FACS buffer (PBS containing 0.5 % of bovine serum albumin and 0.05% of sodium azide). 50 ml of 10 6 -2x10 6 cells/ml were incubated with dilutions of MV-H specific mAbs. These mAbs included some which recognized conformational dependent epitopes of the native protein, others recognized only denatured protein in its reduced or non reduced form. Binding of the mAbs was monitored with FITC-conjugated goat anti mouse IgG antibodies. Cells incubated with an irrelevant mAb or no mAb served as negative controls. Figure 2 demonstrates that by flow cytometry both a high transfection efficiency (96%) was obtained and that transfected cells expressed high levels of recombinant H protein. BHK-21 transfected with RNA transcripts from the negative control plasmid pSFV3-IacZ were negative. 19 MV-H specific mAbs were used to study the structural and conformational integrity of the protein. They were divided into four groups: I) Six mAbs recognizing conformational-dependent epitopes (BH 81, BH 67, 1.44, mAb 15, 55, 85, (Giraudon and Wild 1985 Virology 46-57. Giraudon and Wild 1981 J. Gen. Virol. 54:325-332)); II) three mAbs of the hemagglutinin noose epitope (BH6, BH21, BH216 (Ziegler et al. 1996 J. General Virol 77:2479-2489)); Ill) three others recognizing the two linear epitopes 236-250 (BH47, BH59, BH129; Fournier et al. 1997 J Gen Virol 78, 1295-1302). IV) five mAbs which recognize only denatured MV-H (BH1, BH164, BH189, BH171, BH195). All mAbs of the groups I-111 recognized the MV-H protein expressed on the cell surface of the H-transfected BHK-21 cells. In contrast, the mAbs of the group IV did WO 99/12038 PCT/EP98/05546 23 not bind H-transfected BHK-21 cells. None of the mAbs bound to the cell surface of BHK-21 transfected with RNA transcripts from the negative control plasmid pSFV3 lacZ. In conclusion, this example demonstrates that high levels of the recombinant hemagglutinin protein were expressed with high transfection efficiency on the cell surface of transfected BHK-21 cells. The protein was indistinguishable from that of the measles virus with regard to its conformational integrity as defined by conformational dependent mAbs and was solely detectable in its native conformation. Hemagglutinin transfected BHK-21 cells were also suitable to detect measles virus antibodies in human sera. Example 4: Rosetting assay. Transfected BHK-21 cells were incubated for 1 hr at 37 0 C with monkey erythrocytes (100 erythrocytes/BHK cells). Rosette formation was clearly seen in the hemagglutinin transfected BHK-21 cells but not in BHK-21 transfected with RNA from the pSVFI-lacZ which served as a negative control (data not shown). Thus the recombinant H protein exhibit an important functional activity of the measles viral hemagglutinin. Monkey erythrocytes were also agglutinated by a suspension of crude membranes derived from hemagglutinin transfected BHK-21 cells and this reaction was inhibited by the addition of sera samples of individuals with measles immunity (data not shown) Example 5: Example of a simple preparation of hemagglutinin antigen for the detection of measles hemagglutinin specific antibodies. 24 hr post-transfection, 4x10 7 cells were resuspended in an ice-cold hypotonic solution (2 ml of 10 mM Tris-HCI, pH 7.6, 0.5 mM MgCI 2 ,) containing 10 pg/ml of leupeptin, 10 pg/ml aprotinin, 1 mM PMSF and 1.8 mg/ml of iodoacetamide (solution A) and disrupted with 30-40 strokes in a potter, keeping the nuclei intact. To restore the tonicity, 0.7 ml of solution A supplemented with 0.6 M NaCI was added. Nuclei WO 99/12038 PCT/EP98/05546 24 and large debris were sedimented at 500g (5 min., 40C). Supernatants were collected and EDTA was added to a final concentration of 5 mM and centrifuged at 110 000xg (45 min, 4*C). The pellet containing insoluble particles and total membrane was resuspended in PBS containing 0.5% NP-40. The suspension was again centrifuged at 14000xg (15 min, 4*C). The supernatant containing the total membrane fraction was kept at 4 0 C after addition of 0.05% azide. Protein concentration was determined by the method of Bradford. Example 6: Detection of measles hemagglutinin specific IgG by enzyme linked immunosorbant assay using immobilized recombinant hemagglutinin (late convalescents) (Figure 3-7). Immobilization of recombinant hemagglutinin protein. Specific IgG levels were measured using an ELISA based on recombinant MV-H. Maxisorp microtiter plates (NUNC, DK) were coated overnight at 40C with 50 ml/well of a 2 mg/ml (unless stated otherwise) total protein of the above crude membrane preparation in 0.1 M bicarbonate buffer (pH 9.6). The plates were washed with 0.23 M NaCI containing 0.6 % Tween 20 ("washing buffer"), and wells were blocked 90 minutes at room temperature with 1% BSA in Tris-buffered (15 mM) saline (136 mM NaCI, 2 mM KCI), pH 7.4. Conditions for measuring hemagglutinin-specific IgG. After a washing step, 50 pl of a 1:300 dilution of sera or of diluted mAbs (as controls) in Tris-buffered saline containing 0.1% Tween 20, 1% BSA (dilution buffer) were incubated for 90 min at room temperature. After the plates were washed, 50 pl of a dilution of alkaline phosphatase-conjugated goat anti-human IgG (1:700) or goat anti-mouse IgG (1:750; human adsorbed; Southern Biotechnology Associates, USA) in dilution buffer were added to the corresponding wells and the mixture were incubated for 90 minutes at room temperature. Unbound conjugate was removed by washing, and bound enzyme was detected by the addition of 100 pl of a solution of 0.05% p-nitrophenyl phosphate in 1 mM 2-amino-2-methyl-1-propanol and 0.1 mM MgCI 2 .6H 2 0 (pH 10.2) per well. After 30, 60, 90, 120 min, A405 values were determined. The BHK-gal background were 129±21, 205+41, 302+68, 353±82 respectively. Unless stated otherwise, data are expressed as net 90-min mO.D. values by computing for each WO 99/12038 PCT/EP98/05546 25 individual serum or mAb mO.D. (BHK-H) - mO.D.(BHK-gal). In these cells b-galactosidase is produced as a soluble, cytosolic irrelevant protein and mO.D. (BHK-H) /mO.D.(BHK-gal). was always >0.8. Statistics. Data were analyzed by 2-tailed Student's t-test. The coefficient of determination (R 2 ) was obtained by regression analysis (MS Excel). The following parameters were used to assess the assays (7): * Sensitivity (%) = [true positives / (true positives + false negatives)] x 100 * Specificity (%) = [true negatives / (true negatives + false positives)] x 100 * Accuracy (%) = [(true positives + true negatives) / all] x 100 * Prevalence (%) = [positive sera for HI and/or NT / all] x 100 * Positive predictive value (%) = (prevalence x sensitivity)/ [(prevalence x sensitivity) + (1 - prevalence) x (1 - specificity)] x 100 * Negative predictive value (%) = (1 - prevalence) x specificity / [(1 - prevalence) x specificity + (1 - sensitivity) x prevalence] x 100 Except for the prevalence, undefined sera were excluded from these calculations, unless otherwise specified. Optimizing antigen coating. The optimal signal was reached when plates were coated with 63-250 ng/50 ml of protein and per well (Figure 3). MAbs specific for native protein (produced according to standard technology) or measles immune sera gave a 20-80-fold signal above the one of a negative serum or BH1 which recognizes only denatured H protein. A higher signal to noise ratio (90-160) was found with 32 ng/50 ml, although the signal was lower under these conditions. After 2 hours, signal to noise ratios up to 130 were found when plates were coated with 32-250 ng/50 ml. This confirmed the integrity of the protein conformation after coating. Measuring hemagglutinin-specific IgG in late convalescent donors. To test whether the above assay could be used for detection of MV-H specific IgG in human sera, ELISA plates where coated with 100 ng/well total membrane protein from BHK H cells. Serologically characterized sera obtained from measles late convalescent adults served as a test panel. Figure 4 compares the correlation of the recombinant WO 99/12038 PCT/EP98/05546 26 H-ELISA (read-out after 90 min) and of a certified commercial MV-ELISA with HI and NT titers. The best correlations with HI (R 2 =0.64 vs. 0.48) and NT (R 2 =0.66 vs. 0.52) were found for the recombinant H assay probably because H protein is the most important target for HI and NT antibodies. When the 60 min-values of the recombinant H assay was used in these comparisons, the correlation was the same (HI: R 2 =0.64; NT: R 2 =0.66). These correspond to the correlation found between NT and HI for these sera (R 2 =0.67). These experiments demonstrate that the recombinant H antigen preparation used here is suitable for the detection of the functionally important MV-antibodies and measles immunity. Analysis of false positive and false negative sera. Figure 5A shows the correlation between the recombinant H-ELISA and the commercial MV-ELISA for all 228 sera tested. The two ELISAs showed a reasonably good correlation (R 2 = 0.45), considering that the MV-ELISA detects also antibodies which recognize other MV proteins. The data of Figure 5A was analyzed to determine the number of false positive and false negative sera for both assays. Positivity and negativity was defined on the basis of HI/NT titers. Figure 5B focuses only on sera which are double negative for HI/NT (titer <1:24). On the basis of these values (mean ± S.D. = 54 ± 26) (and the calibration curve obtained with the ISMS, cf. Figure 7) positivity was defined >120 (= mean + 2.5 S.D.), negativity <80 mO.D (= mean + 1 S.D.), with 80-120 mO.D. being a gray zone were sera are undefined. The corresponding values for the commercial MV-ELISA are defined by the supplier as <100 and >200 mO.D. By these criteria, significantly more HI/NT double negative sera were false positive in the MV-ELISA (6/11) than in the H ELISA (0/11; p<0.05; Figure 5B). Figure 5C which represents HI/NT double positive sera shows that the H-ELISA seems to detect as many false negative sera as the MV-ELISA (2/212 versus 1/212; p-value not significant). The HI/NT values of the false negative sera are low ( 1:265, cf. Figure SC). Thus, a positive serum has 209/212 (98.6 %; one serum being undefined) or a 211/212 (99.5%) chance of being tested positive and a negative serum has an 10/11 (90.9%; one negative serum being undefined) or a 3/11 (27.3 %; two sera being undefined) chance of being tested negative by H-ELISA or MV-ELISA respectively.

WO 99/12038 PCT/EP98/05546 27 Since HI is thought to be less sensitive than the NT assay (2, 34, 50), the above analysis based on HI/NT titers potentially excludes weakly positive sera. Figure 6 shows the HI/NT values of sera positive or negative for each one of the two ELISAs. Of the sera which were negative by H-ELISA, 10/13 were HI/NT double negative (Figure 6A). All sera which were positive by H-ELISA were also positive for NT, 4/213 being HI-/NT+ and 209/213 HI+/NT+ (Figure 6B). In the MV-ELISA, four sera were negative; three of these were HI/NT double negative (Figure 6C). Among the sera positive by MV-ELISA 6/222 were HI/NT double negative, an additional 5 were HI-/NT+, all others i.e. 211/222 were double positive (Figure 6D). 209/213 (98.1%) and 211/222 (95.0%) of the sera which were seropositive by H- or MV-ELISA were HI/NT double positive, and 10/13 (76.9%) and 3/4 (75.0%) of sera which were negative by H- or MV-ELISA were double negative by HI/NT. Characteristics of the assay. In parallel, dilutions (in dilution buffer) of the 2nd International Standard for anti-Measles serum (66/202) gave an estimate of the range covered by the assay in International Units (Figure 7). Given the above threshold for positivity (>120 mO.D.), when serum is diluted 1:300, the assay detects 215 mlU/ml in the undiluted serum; negativity (<80 mO.D.) corresponds to 113 mlU. In absolute values the limit of detection is 0.38-0.72 mlU. These results demonstrate that the recombinant hemagglutinin antigen preparation used is suitable for quantifying measles virus-antibodies and therefore for monitoring measles immunity. Example 7: Detection of measles hemagglutinin specific IgG by enzyme linked immunosorbant assay using immobilized recombinant hemagglutinin (vaccinees) (Figure 8). The above assay (example 6) was also used to determine the immunity of measles vaccinees using a panel of sera collected from first year high-school children. Detection of MV-specific IgG The H-ELISA was used to determine the immunity of measles vaccinees. Table 1 compares the H-ELISA and the MV-ELISA with NT and HI titers respectively. The coefficients of determination between H-ELISA and HI or NT titers were considerably WO 99/12038 PCT/EP98/05546 28 higher than those between MV-ELISA and HI or NT. The coefficients of determination did not vary with time of substrate development . They were similar to coefficient obtained when HI and NT titers of the same serum panel were compared with each other (R 2 = 0.51). Background correction: raw versus net O.D. values. The above data were based on net mO.D. values which are obtained by computing for each individual serum mO.D. (BHK-H) - mO.D.(BHK-gal). For a simplified assay single value measurements would represent a considerable advantage. Therefore, raw values obtained with BHK-H antigen were compared with the corresponding net values. An excellent correlation between net values and raw values was observed, with a slight decrease in R 2 values (0.991 to 0.983) between 30 and 120 min of substrate development (90 min: R 2 =0.986; Figure 1). Moreover, coefficients of determination did not substantially decrease when raw mO.D. values (Table 2) were compared with HI or NT instead of net values (Table 1). These data indicate that, in a simplified version of the H-ELISA, background correction may not be necessary. Characteristics of the H-ELISA based on raw O.D. values. In our earlier study, background corrected net O.D. values (mO.D.(BHK-H) - mO.D.(BHK gal)) (of sera from late convalescent donors) were measured after 90 min. In the following, we have analyzed the characteristics of the assay using raw O.D. values (i.e. uncorrected mO.D.(BHK-H)). As before, HI and NT served as standard and also the same definitions were applied for negativity and positivity. As in the previous study, no false positive results were recorded, independent of whether raw or net data were considered and independent of the time points (30, 60, 90 and 120 min; Table 3). However, in comparison with net data, raw data seemed to be associated with more false negative results (e.g. 90 min; Table 3). In contrast, the MV-ELISA registered among 10 HI/NT double negative sera 6 as false positive (0/10 vs. 6/10 P<0.02). Among 378 double positive sera more false negative tended to be found with the MV-ELISA than with the H-ELISA (15/378 vs. 9 WO 99/12038 PCT/EP98/05546 29 or 10/378; P>0.05) and the number of undefined sera was significantly higher in the MV- than the H-ELISA (24/388 vs. 7-9/388; P=0.003 or <0.02). Moreover, by most parameters shown in Table 3 the H-ELISA prevails over the MV-ELISA. Thus. non responders and seroconverted individuals were most accurately defined with background corrected H-ELISA data. But even when uncorrected O.D. values were used, the assay competed well with the MV-ELISA. H-specific antibodies in sera of late convalescent donors and vaccinees. By EIA based on whole virus the level of detected MV-antibodies is considerably higher in late convalescents than in vaccinees (P<0.001). In contrast, by H-ELISA similar levels of antibodies were measured in these two groups whether raw or net values were considered (Table 4). Example 8: Detection of measles hemaggluinin specific IgM by enzyme linked immunosorbant assay using directly immobilized recombinant hemagglutinin (Figure 9). For the diagnosis of measles, it is important to detect specific IgM antibodies. A panel of sera from acute phase and convalescent phase measles patients were used to test whether the H-antigen preparation described in this invention is also suitable for measuring specific IgM. Similarly to the detection of measles specific IgG antibodies, IgM antibodies were detected using an alkaline phosphatase conjugated goat anti human IgM antibody (Figure 9). In Figure 9, the detection of IgM with a commercial IgM detection assay based on whole MV and with the H-ELISA are compared. The coefficient of determinaton of panel A and B are 0.52 and 0.50 respectively. Sera with mO.D.>200 for the MV-ELISA are considered positive for IgM. Figure 9 shows that on the basis of this definition all IgM-positive sera are detected with the H-ELISA (no false negative result) and one (panel A) or two (panel B) sera are false positive. Example 9: Detection of measles hemagglutinin specific IgM by ELISA using indirectly coated recombinant hemagglutinin.

WO 99/12038 PCT/EP98/05546 30 The measles virus hemagglutinin preparation described here was also tested for its suitability to serve as an antigen in an IgM capture ELISA. Serum panel. From 1996 to 1997, 112 serum samples were collected from 70 patients (age range 1.1-35.4 years; median 8.2 years) during a major outbreak and from several isolated cases of measles in Luxembourg. Paired sera were obtained from 31 patients; four patients were bled three times and one patient four times within 59 days after onset of rash. All patients were confirmed measles cases by specific IgM antibodies in a certified commercial ELISA ("MV-ELISA"; Behring Diagnostics, Marburg, DE). In some patients measles was in addition confirmed by an increase in specific antibodies in paired sera (n=36) and/or the clinical case definition of the Centers for Disease Control and Prevention (CDC, Atlanta, GA) (n=24) (Fig. 10). The CDC-criteria include (a) generalized maculopapular rash of 3 days or more; (b) fever of 38.3°C, if measured; (c) at least one of the following symptoms, cough, coryza, or conjunctivitis (Centers for Disease Control 1983 MMWR 31:707-711.). All patients presented with a typical rash. Four of the 112 samples were obtained between 2 and 14 days before onset of rash; these cases were also confirmed by increased hemagglutination inhibition (HI) test and neutralization (NT) test titers, by MV-specific IgG and IgM antibodies in paired samples drawn after onset of rash. 35 sera that were IgG-positive and IgM-negative by MV-ELISA served to determine the threshold for positivity of the IgM H-ELISA. These sera were positive by HI and NT assays. They were collected from 13 individuals who were vaccinated at least 12 months before (median, 5.9 years; median age of vaccination, 15.8 months) and from 22 late convalescent donors. Microtiter plates (Maxisorp, NUNC, Roskilde, Denmark) were coated with 50 pl of a mixture of three conformational-dependent H-specific monoclonal antibodies (BH81, BH97, BH125; 5mg/ml) in 0.1 M sodium bicarbonate buffer (pH 9.6). The monoclonal antibodies were derived from mice immunized with Edmonston strain MV. The plates were washed three times with 1% Tween 20 in Tris-buffered saline (15 mM, pH 8.0) and incubated for 75 min at room temperature with 50 pi/well of the above H-antigen (10 mg protein/ml) or negative control antigen. The plates were blocked with 1% bovine serum albumin in Tris-buffered saline (15 mM, pH 7.4). Test sera were diluted WO 99/12038 PCT/EP98/05546 31 1:10 in Gullsorb (Gull Laboratories, Louvain-La Neuve, Belgium) to eliminate interference of IgG. Sera were further diluted to a final dilution of 1:25 in a modified commercial dilution buffer (Enzymum-test, Boehringer, Mannheim, Germany) and added for 75 min at room temperature to the antigen coated microtiter plates. Plates were washed three times with the above wash buffer. Alkaline phosphatase conjugated goat anti-human IgM (1:1000; Southern Biotechnology Associates, Birmingham, Ala) and p-nitrophenylphosphate (0.5 mg/ml; 100ml/well) (Sigma, St. Louis, USA) were used to develop the assay. Optical density was measured at 405 nm following a 2 hours incubation at 37 0 C. Data are expressed as milli-optical density (mOD). The threshold for positivity was defined as the mean mOD + 2 standard deviations (SD) measured after 2 hours of the IgM-negative sera. Definition of threshold for positivity. Specific IgM was measured by H-ELISA with sera obtained from measles late-convalescent adults and vaccinees; all of these individuals were negative for IgM by MV-ELISA. IgG levels of these donors were between 270 and 2750 mOD by MV-ELISA. By the H-ELISA, the mean ±SD of IgM values for the convalescents and the vaccinees were 39±161 (range -77 to 131) and 23±49 (range -45 to 105), respectively. T-test detected no significant difference in the mean of vaccinees and convalescents. When IgM values of sera with IgG<1000 mOD and >1000 mOD were compared no difference was detectable (26±58.2 versus 35 ± 57.2, P value not significant). Based on the above IgM-negative sera from vaccinees and late convalescents donors, the threshold for IgM-positivity was defined as the mean + 2 SD (= 33 + 2*57 = 147). Experiments with unadsorbed sera gave sometimes false-positive IgM titers in the H-ELISA (data not shown). Thus, in the present format of the H-ELISA preadsorption with anti-human IgG (GuIISORB) gave a more reliable estimation of IgM levels. Detection of MV-specific IgM in measles patients. In 108 serum samples (from 70 different patients) drawn at different time intervals after onset of rash, levels of specific IgM were analysed both by H-ELISA and by MV-ELISA. Figure 11 shows that the MV-ELISA detects 107 IgM-positive sera (99.1%), whereas the H-ELISA detects IgM in only 81 sera (75%) on the basis of the above threshold. All (additional) sera WO 99/12038 PCT/EP98/05546 32 collected before onset of rash (on day: -14, -9, -8, -2; n=4) were IgM-negative in both ELISAs (Fig. 12). Most of the sera that were IgM-positive in both the MV-ELISA and the H-ELISA were found before day 20 after onset of rash. Among 68 sera collected from 50 individuals between day 0 and 19 after onset of rash, 67 had detectable measles-specific IgM by MV-ELISA and by H-ELISA (98.5%) (Fig. 12B). The one serum that was negative in the MV-ELISA was drawn on the day of onset of rash (day 0) and was positive by H-ELISA (372 mOD, patient A in Fig. 11). This patient was IgG negative by both ELISAs and developed measles according to CDC criteria. A second sample (serum A' in Fig. 11) drawn on day 12 from the same patient was IgM positive by both ELISAs (791 and 1142 mOD by MV-ELISA and H ELISA, respectively). This patient experienced a significant increase in HI and NT titers and in specific IgG (6 to 748 mOD by IgG MV-ELISA and 228 to 726 mOD by IgG H-ELISA). This suggests that on day 0 the IgM was accurately measured positive by the H-ELISA, in contrast to the MV-ELISA. The serum obtained on day 1 from a patient B (serum B in Fig. 11) who was IgM negative by H-ELISA and weakly positive by MV-ELISA (266 mOD) was the first of paired sera. The second sample of this patient (drawn on day 13, serum B' in Fig. 11) was positive by both IgM ELISAs and this patient experienced a significant increase in HI and NT titers and in specific IgG (14 to 1077 mOD by IgG MV-ELISA and 142 to 1533 by IgG H-ELISA). This may suggest that serum B was measured false negative on day one by H-ELISA. An additional 40 sera were drawn from 37 patients after day 19 of onset of rash. Twenty of these patients were bled twice or more and independently confirmed at least by increase in specific IgG. Nineteen of the latter patients were found IgM positive by H-ELISA either before day 20 (17 sera) or after day 20 (4 sera) of onset of rash. One patient (bled on day 20 and 32) was twice IgM negative by H-ELISA. Another patient which was bled only oni day 23 but confirmed by CDC criteria was also IgM negative by H-ELISA. Thus among these 24 sera which were independently confirmed by MV-ELISA and at least one other criteria, only 9 sera were IgM positive for H-ELISA (37.5%).

WO 99/12038 PCT/EP98/05546 33 Among the 16 (single) sera drawn after day 19 that were IgM-positive by MV-ELISA and not independently confirmed clinically or by IgG increase, 5 (29.4%) were positive by H-ELISA. Thus, among the sera obtained after day 19, no significant difference was observed in the proportion of sera positive in the H-ELISA whether IgM-positivity in the MV-ELISA was independently confirmed or not (37.5% versus 29.4%, P-value not significant). Detection of IgG in measles patients by H-ELISA. As for the MV-ELISA, IgG can be detected in all samples after day 3 (no false negative sample). When the first serum sample (n=16) was drawn between day 0 and 15, the second serum (drawn 7 to 54 days later) exhibited a mean increase of 1013 mOD (range 146 to 2076 mOD) in the H-ELISA and 959 mOD (range 215 to 1963 mOD) in the MV-ELISA. Thus, no significant difference between the two assays became apparent (Fig. 13). Our study demonstrates considerable amounts of H-specific antibodies within the early days after onset of rash which are readily detectable with an ELISA based on a mammalian expressed recombinant protein. The initial time course of IgM seemed to be similar in the H- and the commercial MV-ELISA. Maximal average levels of specific IgM were reached between day 5 and 10 after onset of rash. This is in agreement with earlier studies with total IgM isolated by sucrose gradient tested by HI (day 10) (Centers for Disease Control 1982 Morbid. Mortal. Weekly Rep. 31:396 402) or by IgM-Enzygnost (day 5-10) (Pederson et al. 1986 Vaccine 4:173-178; Rossier et al. 1991 J. Clin. Microbiol. 29:1069-1071). The early appearance of H-specific IgM translates into a high sensitivity (98.5%) of the H-ELISA within the first 19 days, which matches that of the Enzygnost MV-ELISA (98.5%) for the same sera. When an interval from 0 to 30 days is considered the sensitivity of the H-ELISA was 92.7% and of the MV-ELISA 98.8%. In another cohort the sensitivity between day 0 and 30 was reported to be 91.8% for the Behring test (specificity 98.2%), 93.3% for the Gull test (specificity 90.5%) and 85.5% for the Incstar test (specificity 95.2%) (Arista et al. 1995 Res. Virol. 146:225-232). When an even longer interval is considered the sensitivity of the H-ELISA rapidly deteriorated WO 99/12038 PCT/EP98/05546 34 (day 0-59: 75%) while the sensitivity of the MV-ELISA did not (98.1%) because it served as a gold standard. The high sensitivity seems to agree with Ozanne and d'Halewyn (Ozanne and d'Halewyn 1992 J. Clin. Microbiol. 30:564-569) who reported for the Enzygnost a sensitivity of 91.9% between day 1 and 56. However, the overall sensitivity strongly depends on the number of sera collected during the later time points. In some studies only 0%, 8%, 5% or 4.5% of samples were collected after day 15 (Rossier et al. 1991 J. Clin. Microbiol. 29:1069-1071), day 19 (Mayo et al. 1991 J. Clin. Microbiol. 29:2865-2867), day 25 (Ozanne and d'Halewyn 1992 J. Clin. Microbiol. 30:564-569) or day 30 (Arista et al. 1995 Res. Virol. 146:225-232) respectively. In contrast, our study included 37% of samples drawn after day 20 or 25% collected after day 30. It is therefore important to compare sensitivities later after onset of rash. In the H-ELISA the sensitivity was 64,3% (n=14) and 19.2% (n=26) between day 20 to 29 and day 30 to 59 respectively. In the study of Arista et al. (Arista et al. 1995 Res. Virol. 146:225 232), the sensitivity of the Enzygnost was 40% between day 31 to 35. After day 19, less than half of the sera (16 of 40) were solely confirmed by MV ELISA. Therefore, the assessment of the sensitivity of the H-ELISA after day 19 may depend on the performance of IgM-Enzygnost. However, essentially no difference in sensitivity of the H-ELISA was found whether the sera were confirmed by the commercial ELISA only (29.4%) or by at least one additional parameter (37.5%). Thus, in our cohort, the difference in sensitivity between the H-ELISA and the MV ELISA cannot be explained by excessive false positive results (low specificity) of the MV-ELISA, but rather by accelerated waning of H-specific IgM in comparison to MV IgM. The threshold that gave a sensitivity of 98.8% in the H-ELISA (day 0-19) was associated with a specificity of 100% in the IgM-negative panel. However, this needs to be confirmed in a panel of IgM-negative sera, which is independent of the definition of the threshold. Lowering the threshold of the H-ELISA would decrease the specificity but increase the sensitivity above 95% within the first 30 days. A high specificity, i.e. the low percentage of false positive results was also reported by several authors for the Behring test (Arista et al. 1995 Res. Virol. 146:225-232; WO 99/12038 PCT/EP98/05546 35 Ozanne and d'Halewyn 1992 J. Clin. Microbiol. 30:564-569; Rossier et al. 1991 J. Clin. Microbiol. 29:1069-1071). Measles is most contagious within one week before and after onset of rash (Katz 1985 Measles and subacute sclerosing panencephalitis. p. 1062, 18th ed. Appleton Century-Crofts, New-york, U.S.A.). Serological assays before onset of rash are not available, but during the contagious period after onset of rash the H- and MV-ELISA perform equally well. Most studies which rely on IgM for measles diagnosis recommend serum samples drawn within about 3 weeks after onset of rash (Arista et al. 1995 Res. Virol. 146:225-232; Erdman et al. 1991 J. Clin. Microbiol. 29:1466 1471; Lievens and Brunell 1986 J. Clin. Microbiol. 24:391-394; Ozanne and d'Halewyn 1992 J. Clin. Microbiol. 30:564-569; Rossier et al. 1991 J. Clin. Microbiol. 29:1069-1071). Thus an optimal performance within 19 days seems to be sufficient for measles diagnosis (Helfand et al. 1997 J. Infect. Dis. 175:195-199; Ozanne and d'Halewyn 1992 J. Clin. Microbiol. 30:564-569). Nonetheless, we currently investigate whether further optimizing the assay could increase the sensitivity beyond day 19. We have also studied the development of H-specific IgG in measles patients. No false negative serum was detected and increases between paired sera were as significant as for the MV-ELISA. In light of these results and the overall high sensitivity and specificity of the H-ELISA for IgG (Bouche et al. 1998 J. Clin. Microbiol. 36:721-726; Bouche et al. 1998 J. Virol. Methods. In press), we conclude that this assay is as reliable to diagnose measles in paired sera as an ELISA based on whole MV. Despite the longer persistence of maximal levels of IgM in MV-ELISA and early waning of H-specific IgM, the initial time course of H- and MV-specific IgG are similar. Since the H protein can be efficiently produced in a mammalian system (Bouche et al. 1998 J. Clin. Microbiol. 36:721-726; Bouche et al. 1998 J. Virol. Methods. In press) and since recombinant proteins may benefit from enhanced stability, an ELISA based on this antigen seems like an interesting alternative in the search for a low cost, rapid diagnostic system for measles.

WO 99/12038 PCT/EP98/05546 36 Example 10: Detection of measles hemagglutinin specific IgA by ELISA using indirectly coated recombinant hemagglutinin. The assay described in Example 9 was also used to detect IgA using alkaline phosphatase-conjugated goat anti-human IgA (1:1000, Southern Biotechnology Associates).Essentially the same panel of measles patients was used as in Example 9. In addition, 16 measles virus seronegative donors were used. Figure 14A shows that 100% of the IgG seronegative sera were also negative for IgA. Also prior to onset of rash and on day 0-2after onset of rash IgA titers were negative or very low. Later after onset of rash most patients became IgA positive. Figure 14B compares the MV-H specific IgG with MV-H specific IgA. These results demonstrate that the H ELISA also efficiently and reliable detects H-specific IgA. A similar result was obtained when the H protein was directly coated to the microtiter plate.

WO 99/12038 PCT/EP98/05546 37 Table 1. Correlation between H-ELISA (net O.D. values) and MV-ELISA with HI or NT respectively. H-ELISA (net values) MV-ELISA 30' min 60 min 90 min 120 min HI 0.612) 0.60 0.60 0.59 0.36 NT 0.65 0.66 0.64 0.64 0.36 1 time of read-out 2) coefficient of determination R 2 WO 99/12038 PCT/EP98/05546 38 Table 2. Correlation between H-ELISA (raw O.D. values) and HI or NT respectively. H-ELISA (raw values) 30 min' 60 min 90 min 120 min HI 0.6121 0.59 0.58 0.58 NT 0.64 0.65 0.62 0.62 ' time of read-out 2) coefficient of determination R 2 WO 99/12038 PCT/EP98/05546 39 0)0 6- CD c Gi a;. aif C) cD a) 0t v% C' z 00 a Oco w 0D 0 0 - 0 0) 0) 0Y 0 0) 0 a.) () Go w ) 0) 0) co co > pD (D r-: -: e -C0 ) C0)0)) o Cl) cc wQ CD C z m-* 0 CY )0) 0 P 0 C) 00- C)N046 6. C 0. < u~ 00).- 00)- 00) 0O 00) DN 0j 06 CO c- N 0 C '0 a D ) 0)F CD . 0(DC0 0 CI.N C0a) CD 0 A)C 0~' N )'-- C') C' CLC 0.C 0 CL c C. Cu a u " V co- D DU) A V 0 O C-j <~ /- 0. l -Jc I Cc j < CO 0) u ccu C'> D C C-U) 0 E j 0 Cu5 0)00 C~~C CL) CD0 N (

Claims (15)

1. A method for the identification of measles virus specific antibody in sample comprising (a) contacting a sample suspected of containing measles virus specific antibody with a measles virus specific glycoprotein recombinantly produces in mammalian cells using a high expression system; and (b) detecting the presence or absence of said measles virus specific antibody in said sample.

2. The method of claim 1 wherein said measles virus specific glycoprotein is the hemagglutinin or the membrane fusion protein.

3. The method of claim 1 or 2 wherein said high expression system is based on a togavirus expression system.

4. The method of claim 3 wherein said togavirus is an alphavirus.

5. The method of claim 4 wherein said alphavirus is Semliki Forest virus.

6. The method of any one of claims 1 to 5 wherein said sample is derived from a body fluid.

7. The method of claim 6 wherein said body fluid is serum, plasma, saliva or cerebrospinal fluid.

8. The method of claim 6 or 7 wherein said body fluid is obtained from a patient infected with measles virus, a convalescent having recovered from measles virus infections, a subject immunized with a measles vaccine or a sero negative individual. WO 99/12038 41 PCT/EP98/05546

9. The method of any one of claims 1 to 8 wherein said antibody is an IgG, IgA or an IgM antibody.

10. The method of any one of claims 1 to 9 wherein, prior to step (a), said glycoprotein is affixed to a solid support.

11. The method of claim 10 wherein said solid support is a well of a microtitre plate, a bead, an electronic sensory device or any support compatible with a rapid test format.

12. The method of claim 10 or 11 wherein said glycoprotein, when affixed to said solid support, is comprised in a crude cell extract.

13. The method of any one of claims 1 to 12 wherein said detection is obtained without the computation of background values.

14. The method of any one of claims 1 to 13 wherein said identification comprises the identification of the virus strain.

15. Kit comprising (a) recombinantly produced glycoproteins as identified in any of the preceeding claims either in solution or immobilized on a solid support; and (b) reagents suitable for the detection of human antibodies bound to said glycoproteins.