key: cord-0008659-6jjnhhot authors: Herrmann, John E. title: Enzyme-Linked Immunoassays for the Detection of Microbial Antigens and Their Antibodies date: 2008-04-15 journal: Adv Appl Microbiol DOI: 10.1016/s0065-2164(08)70445-3 sha: 0c04f5329845045f23305c2d68c38debec951ffe doc_id: 8659 cord_uid: 6jjnhhot Antibodies could be labeled with enzymes for use in histochemical staining procedures by enzyme-immunoassay (EIA). The use of EIA is an extension of previously used serological tests, using enzyme-labeled antibody or antigen to determine antibody content. Direct detection of antigen by EIA represents a more dramatic departure from previous methods based on culture. Also, the method has enabled detection of infectious agents that are difficult to cultivate, such as hepatitis A virus and rotavirus, or agents that cannot be cultivated, such as hepatitis B. The use of EIA tests for detection of microbial antigens provides an alternative to culture as a means for direct identification of a specific microbial agent. It also provides a means to detect microbial agents which have not been successfully propagated. The detection of circulating antigen or detection of antigen in other body fluids by EIA is more difficult than detection of antibody because of the sensitivity required, and because of interfering substances in specimens such as feces and respiratory secretions. For this reason, very few antigen detection assays have the sensitivity and specificity required to be used as a primary diagnostic test. The number of tests that have been developed, however, is impressive and because of the possibilities for rapid, specific diagnosis, the interest in antigen detection by EIA remains high. there is little difference in the sensitivity or specificity of the two assays (Sarkkinen et al., 1981~) . The advantages of enzyme labels over radioactive ones are mainly convenience in use, in that the labeled immunoreagents are stable for long periods, and the precautions and disposal procedures required for radioisotopes are unnecessary. In addition, the use of chromogenic substrates for the enzyme labels permits visual interpretation of test results in some cases. The only real disadvantages of EIA tests are the loss of antibody reactivity that may result from conjugation to enzymes, and the limits of substrate detection. For example, use of enzymes that have molecular weights higher than that of IgG molecules such as P-D-galactosidase (MW 540,000 Da) can cause steric hindrance of antibody activity (Herrmann and Morse, 1974) . With regard to limits of substrate detection, improvement of enzyme detection by use of fluorogenic, luminescent, or radioactive substrates (reviewed by Yolken, 1982) has been proposed. The general principles of EIA tests and details of earlier studies have been reviewed a number of times (Yolken, , 1982 Hildebrand, 1979; Voller et al., 1981) and will not be repeated in detail here. Rather, the major emphasis will be on current developments in EIA methodology and the application of EIA to diagnosis of infectious diseases. This will include tests for both antigen and antibody detection in viral, rickettsial, bacterial, and mycotic infections. EIA tests for diagnosis of parasitic agents, hormones, and other antigens have been described but will not be discussed here. The solid-phase or heterogeneous EIA requires immobilization of antigens or antibodies on a solid surface as a means of separating antigenantibody complexes. Solid-phase surfaces used to immobilize antigens or antibodies have for most applications been polystyrene beads, tubes, and wells of microtiter plates, or wells of polyvinyl chloride microtiter plates. Coupling of proteins to these surfaces is usually done by passive adsorption. More recently, adsorption of antigens or antibodies to nitrocellulose membranes has been adapted to detection of viruses by EIA (Bode et al., 1984) . Covalent linkage of antibodies or antigens to a variety of surfaces, including porous glass (Lynn, 1975) , nylon (Hendry and Herrmann, 1980) , cellulose (Ferrua et al., 1979) , agarose (Streefkerk and Deelder, 1975) , and polyacrylamide (Avrameas and Guilbert, 1971 ) has been described. The majority of solid-phase EIA tests that have been found to be clinically DETECTION OF MICROBIAL ANTIGENS AND ANTIBODIES 273 useful utilize plastic microtiter plates or beads, with antigen or antibody passively adsorbed to the solid phase. In situations where a given antigen does not readily attach, antibody to the antigen is applied first. For some antigens, nonspecific adsorbents such as poly-L-lysine have been used to enhance antigen adsorption, or the Clq component of complement to capture antigen-antibody complexes (Yolken, 1982) . Most of the assays described are of the noncompetitive type, although a number of competitive assays have been described. The disadvantage of many competitive assays for antigen detection is that they use labeled antigen, which is usually more difficult to prepare than labeled antibody. With the advent of monoclonal antibodies, competitive assays for specific antibodies are becoming more common. A number of different formats of EIA tests for antigens and antibodies are possible. These are discussed in Section II,C and D. To avoid the need for separation of antigen-antibody complexes, homogeneous EIA tests were developed (Rubinstein et al., 1972) . Homogeneous assays are based on the reaction of antigen with an antibody-enzyme complex. This results in steric hindrance of the enzyme, which causes a decrease in product after reaction with enzyme substrate. The major advantage of homogeneous assays is that they do not require the separation and washing steps required in heterogeneous assays. The major disadvantage of this type of EIA is that it has been difficult to apply it to detection of high-molecularweight antigens with the degree of sensitivity required. Thus, the homogeneous assay has been used mainly for detection of hormones, drugs, and other low-molecular-weight substances. Means to improve the sensitivity of homogeneous assays have been devised (review by Yolken, 1982) but have not been widely used. Thus, the applications discussed below will be limited to EIA tests of the solid-phase type. Enzyme immunoassays for antibodies to microbial agents have been utilized for almost all of the common infectious diseases, because the sensitivity required is well within the range of EIA. A summary of the procedures most often used is given in Table I . The choice of assay depends on the sensitivity required, the availability of reagents, and whether a class-specific test is desired. For detection of IgM, either the noncompetitive EIA or class capture methods can be used. In the noncompetitive EIA, enzyme-labeled antiglobulin (step 3) specific for IgM is used. The advantage of IgM capture If indirect test, enzyme-labeled antiglo1)rrlin is added 6. Enzyme substrate is added 7 . Substrate hydrolysis product is proportional to the amount of IgM present methods is that there is less problem with sera containing rheumatoid factor (RF), although precautions must be taken in designing the test to avoid weakly reactive sera containing RF (Parry, 1984; Briantais et al., 1984) . The use of enzyme-labeled antigen in tests for IgM has been useful for diagnosis of a number viral infections, e.g., cytomegalovirus (Schmitz et al., 1980) , Epstein-Barr virus (Schmitz, 1982) , and flaviviruses (Schmitz and Emmerich, 1984) but requires purified antigen for labeling. Thus, it is limited to those agents where production and purification of antigen is relatively simple and offers improved diagnosis as well. The majority of class-specific EIA tests are for IgM (sometimes referred to as MAC-ELISA for M antibody capture-ELISA), because IgM is the most important serological marker indicating recent infection when only single serum specimens are available. However, substitution of other class-specific markers can be used. Methods for detecting antigen by EIA can be done by competitive and noncompetitive formats, as with assays for antibodies. The lower limit of sensitivity for detecting antigen by EIA in most studies is approximately 100 pg to I ng, although lower limits have been described (Kato et al., 1975) . This level of sensitivity is sufficient to detect virtually all culture-propagated infectious agents, but is not always sensitive enough to detect antigen directly in clinical specimens. The type of formats most often used are shown in Table 11 . A number of factors determine haw efficient an assay is in detecting antigens and antibodies. Some are inherent and cannot be controlled, e.g., the amount of antigen that is usually present in a positive clinical specimen, and others, such as test design, can be controlled. Some of the more important variables that can be controlled are discussed here. Most EIA formats require covalent coupling of enzymes to antibody or antigen. A number of enzymes and coupling techniques have been tried (reviewed by Yolken, 1982) . The most consistent results have been obtained with horseradish peroxidase coupled by use of periodate (Nakane and Kawaoi, 1974) and alkaline phosphatase coupled by use of glutaraldehyde (Avrameas, 1969) . Most of the assays found to be clinically useful in diagnosing infectious diseases use chromogenic substrates, although fluorogenic and radioactive substrates have been described (Yolken, 1982 ; Avrameas and Guesdon, 1982) , as have luminescent ones (review by Seitz, 1984) . A more recent development in EIA which has found application in diagnostic microbiology is the use of avidin and biotin (Guesdon et al., 1979) . The test is based on the high affinity constant for binding biotin to avidin. The most common method is to use specific antibody labeled with biotin; the indicator system is enzyme-labeled avidin. A recent adaptation of this method using biotinillated beta-lactamase in combination with avidin was effective for detecting rotavirus antigen (Yolken and Wee, 1984) . In addition to antibody affinity and the sensitivity of the indicator system used, the sensitivity of many antigen detection systems depends on the amount of antibody that can be immobilized effectively on a solid phase. Methods for immobilization of antibody on plastic surfaces are usually based on simple adsorption, although convalent-linking methods have been utilized. The amount of immunoglobulin that can be immobilized on various plastics is given in Table I11 (Herrmann, 1981) . Increasing the amount of antibody bound to a solid-phase surface should result in increased sensitivity of the EIA. However, it has been noted that increasing the concentration of antibody for coating surfaces beyond 10 kg/ml does not give an increase in immunoassay sensitivity. This is apparently due to desorption of antibody from the plastic surface, and steric hindrance of antibody that is adsorbed. Solid-phase surfaces other than plastic tubes, beads, or particles, such as porous glass (Lynn, 1975) have been used to immobilize antibody. However, the majority of EIA tests for microbial infections use either plastic plates or beads. A more recent development mentioned above that may be applicable to a variety of microbial antigens and antibodies is the use of nitrocellulose membrane disks as a solid phase. This was successfully developed as a visual readout method for detecting adenovirus antigens and antibodies (Bode et al., 1984). The use of antibodies that are highly specific and have high affinity is the most critical aspect of most EIA techniques. An example of how the reagents used determines EIA effectiveness was shown in the two papers discussed below on detection of Clostridium dz$n'le toxin, where the sensitivity was increased from 58.6 to 95% by changing the immunoreagents used (Laughon et al., 1984) . The diluents used for antigen preparation can also alter the sensitivity. Disrupting microbial agents with detergents or other chemicals may increase the sensitivity of some assays but decrease others (Yolken, 1982) . A number of diluents not usually used in EIA tests was tested by Conroy and Esen (1984) for adsorbing a plant protein to polystyrene. These included detergents, acids, alcohols, and urea. Use of alcohols or urea in-creased the EIA substantially. Whether this would be applicable to microbial antigens remains to be tested. The diagnosis of infectious disease has been accomplished in the past by either isolation of the infectious agent, or by measuring serological conversion to a given agent. Serological conversion can be measured by a high level of IgM, or by an increase in total antibody in convalescent sera compared with acute sera. The use of EIA is an extension of previously used serological tests, using enzyine-labeled antibody or antigen to determine antibody content. Direct detection of antigen by EIA represents a more dramatic departure from previous methods based on culture. Also, the method has enabled detection of infectious agents that are difficult to cultivate, such as hepatitis A virus and rotavirus, or agents that cannot be cultivated, such as hepatitis B virus (Section V). A summary of the bacterial and mycotic infections for which EIA serological tests have been devised is given in Table IVA report various degrees of sensitivity and specificity. In a review by Hill and Matsen (1983) a sensitivity greater than 95% was reported for some assays, hut the sensitivity was as low as 50% in many others. Thus, each test must be carefully examined to determine how useful it is for a specific infection. In addition to problems with low sensitivity, the major drawback to many assays is the lack of standardization. Without the availability of standard serum samples for evaluation of assays, new assays require testing by a number of investigators before their validity can be assessed. In many instances, however, serological diagnosis is the only means available to many laboratories for diagnosing some of the more exotic diseases. For example, for diagnosis of Lyme disease, culture ofthe causative spirochete is possible, but is often ineffective. Use of an EIA test for specific IgM and IgG response in patients with proven Lyine disease was diagnostic in 11 of 12, and the EIA gave no false positive results in 40 control subjects (Craft et aE., 1984) . Early recognition of disease is another area where EIA tests for specific IgM and IgG may be the best or only method available to some, such as diagnosis of tuberculous meningitis (Hernandez et al., 1984) , although new- Dittmar et al. (1979) Hopkins et al. (1982) Mathiesen et al. (1978) Feinstone et al. (1979) Gilman and Docherty (1977) Leinikki and Passila (1977 Monath et al. (1984) Hofmann et al. (1979) Forghani et al. (1978) er developments in use of latex particle agglutination for diagnosis of this disease may prove more useful than serology (Krambovitis et al., 1984) . The applications of antibody EIA for diagnosis of viral and rickettsia1 infections are given in Table I V , B . As discussed above for bacterial and mycotic infections, serological diagnosis by EIA for most of the agents listed is still largely experimental. The most frequently used applications are in screening for immune status, such as rubella testing, for cytomegalovirus antibody, and for antibodies to hepatitis B antigens. Serological diagnosis of infectious mononucleosis is also the method of choice. The heterophile antibody test used is not specific for Epstein-Barr virus, and the test is usually done by agglutination, but EIA tests may be more sensitive (Halbert, 1982) . Specific EIA tests for antibody to Epstein-Barr virus components have also been described (Hopkins et al. , 1982) but are not yet widely used. Because many viruses are difficult to isolate, or haven't yet been cultivated, serological tests are often the most useful for diagnosis. The use of IgM capture EIA for determining recent viral infection is becoming more common, and may provide aid in diagnosis where antigen detection methods are not available. The use of EIA tests for detection of microbial antigens provides an alternative to culture as a means for direct identification of a specific microbial agent. It also provides a means to detect microbial agents which have not been successfully propagated. The detection of circulating antigen or detection of antigen in other body fluids by EIA is more difficult than detection of antibody because of the sensitivity required, and because of interfering substances in specimens such as feces and respiratory secretions. For this reason, very few antigen detection assays have the sensitivity and specificity required to be used as a primary diagnostic test. The number of tests that have been developed, however, is impressive and because of the possibilities for rapid, specific diagnosis, the interest in antigen detection by EIA remains high. The tests developed for bacterial infections are primarily for diseases which have causative agents difficult to culture, or where rapid diagnosis will permit prompt treatment. As can be noted by comparing Tables IV,A and V,A, there are far fewer EIA antigen detection tests than antibody tests, for reasons cited above. The efficiencies of the assays reported are variable, but none is as sensitive as the corresponding culture technique. Only one test is commercially available at this writing, an EIA for detecting Neisseria gonorrhoeae antigens. The EIA has been evaluated by anumber of laboratories (Table V,A). In general, almost all reports have found that the EIA was equivalent to culture for detection of gonorrhoeae in males. In females, sensitivities have ranged from 74.4 (Papasian et al., 1984) to 90.9% (Danielson et al., 1983) and specificities from 86.5 (Manis et al., 1984) to 100% (Danielson et al., 1983) . Two other extensive studies reported specificities of 98% (Stamm et al., 1984; Demetriou et al., 1984) . Thus, the reliability of EIA appears to depend on the efficiency of the culture method used, and perhaps variability in performing the test itself. EIA tests have also been developed for another agent of sexually transmitted disease, Chlamydia trachomutis. This agent is more difficult to cultivate than N . gonorrhoeae, in that cell cultures are required. Only one EIA has Williams et al. (1984) Khan (1984) Segal et al. (1979) Torsch (1980, 1981) Pronovost et al. (1982) Mathiesen et al. (1978) Wolters et al. ( ) Miranda et al. (1977 ; Grillner and Landqvist (1983); Land et al. (1984) ; Lawrence et al. (1984) ; Morgan and Smith (1984) ; Nerukar et a1. (1984) ; Warford et al. ( ) Berg et al. (1980 Harmon et al. (1983) ; Sarkkinen et al. (1982) Yolken et al. (1977, 1980) ; Sarkkinen et al. (1980 ) Ziegler (1984 been examined with a significant number of samples, a commercially produced test under development (Chlamydiazyme, Abbott Laboratories). Premarket evaluation of this test on 416 patients showed the EIA had a sensitivity of 83% (63176) and a specificity of 94% (Jones et d., 1984) . In a larger study involving 2384 specimens the EIA had a sensitivity of 83% and a specificity of 94% . Mycoplasma hominis may also be involved in sexually transmitted disease, and diagnosis by culture requires expertise. An antigen EIA has been developed (Miettinen et al., 1984) and was positive for six specimens positive by culture. More extensive evaluation is required to determine the utility of the assay as a screening procedure. Detection by EIA of bacterial antigens in cerebrospinal fluid (CSF) and respiratory tract secretions has also been attempted, with good results for some antigens. Yolken et al. (1984) reported 100% sensitivity for detecting pneumococcal antigen in 25 CSF specimens, but others have found difficulty in differentiating pneumococcal antigens from meningococcal antigens in CSF by EIA . Use of monoclonal antibody for detecting group A meningococcal antigens has been described (Sugasawara et al., 1984) but was 84% as sensitive as polyclonal serum for detecting antigen in the same CSF samples. Detection of Haemophilous influenzae B by EIA has been shown to be effective in limited clinical trials. Drow et al. (1979) developed an EIA which was 100% sensitive on 11 positive CSF samples, , using a similar EIA system, detected 17 of 20 samples that were positive for Haemophilus by counterimmunoelectrophoresis or coagulation. There were 17 positive by culture. Antigen detection by EIA for respiratory infections has been described for Legionella pneumophila, tuberculosis, and streptococcus group A infection. The most extensive study for detecting L. pneumophila antigens in urine was done by Sathapatayavongs et al. (1982) , who obtained a 82.9% (39/47) sensitivity and a specificity of 100% in 178 urines from patients with other diseases. An inhibition EIA for detection of streptococcus group A antigen is throat swabs was also effective, giving a sensitivity of 97.0% and a specificity of 97.9% (Knigge et al., 1984) . Confirmed diagnosis of tuberculosis is difficult because of the long period required to culture the causative organism. Preliminary results of an EIA developed by Sada et al. (1983) , utilizing rabbit antibody to BCG, showed a sensitivity of 81.2% in 16 samples from patients with tuberculosis meningitis. Because antibodies to BGG may cross react with other mycobacteria, as well as with species of Nocardia and Corynebacterium, the test needs further evaluation for specificity. There have also been EIA tests developed for detection of bacterial toxins in clinical samples, most notably assays for Clostridium dijjficile toxins. Lyerly et al. (1983) developed an EIA for C. di,fficiZe A toxin which was 100% specific in 31 samples, but only 58.6% (17/29) sensitive. An improved assay 284 JOHN E. HERRMANN for this toxin and for B toxin was reported by Laughon et al. (1984) . Of 79 tissue-culture-positive specimens, 91% were positive for toxin A and 80% were positive for toxin B. Combined, 95% were positive for either A or B toxin. Thus, this is one EIA test that appears to be a marked improvement over the difficult tissue culture toxin assay. Detection by EIA in stool of toxins from other Clostridium species, C. perfringens A (McClane and Strouse, 1984) and C. botulinum A and B (Dezfulian et al., 1984) has also been reported. Two other unrelated infections have been diagnosed by antigen EIA, Bacteroides fragilis and Yersinia pestis infections. A test for B . fragilis in urine was 100% specific, and detected antigen in 73% (11/15) of individuals shown to be infected with B . fragilis (Rissing et al., 1984) . Use of monoclonal antibody against the F1 antigen of Y . pestis was insensitive in an EIA, detecting antigen in 20% (2/10) sera from patients with acute bubonic plague (Williams et al., 1984) . Because antiviral agents are becoming available for some of the respiratory virus infections, rapid methods of diagnosis are essential for prompt treatment. Rapid diagnosis by EIA has been proposed for a number of respiratory viruses. Several have been described for diagnosis of respiratory syncytial virus. Specificity does not appear to be a problem with any EIA reported, but the sensitivity is less than that found by culture. Compared with culture, sensitivities have been found to be 79.3% (23/29) (Chao et al., 1979) , 60.9% (25/41) (Hornsleth et al., 1981) , 78.7% (37147), (Hornsleth et al., 1982) , and 82.8% (77193) (McIntosh et al., 1982) . Diagnosis of viral influenza by EIA has also been reported, with variable results. Compared with culture, Harmon and Pawlik (1982) reported a sensitivity of 53% (21/40). A later report by Harmon et al. (1983) on an EIA using fluorogenic substrates gave a sensitivity of 87% (27131). By use of a radioactive substrate, Coonrod et al. (1984) were able to detect influenza virus in nasal washes, but the maximum sensitivity at any given day of infection was 48% (12/25). A similar assay described by Yolken (1980) on samples from 12 volunteers gave sensitivities of 78 to loo%, depending on the day tested. Tests for adenovirus in respiratory secretions have also been developed. Harmon and Pawlik (1982) compared an EIA with tissue culture isolation and were able to detect by EIA 62% (13/21) of adenovirus-positive specimens. The interest in sexually transmitted herpesvirus and the availability of treatment have led to development of a number of EIA tests for rapid diagnosis of herpes infection. Most lack sufficient sensitivity to be used as a substitute for culture. Two evaluations of a commercial EIA for herpes genital infection (Ortho Diagnostic Systems, Inc.) have been reported. Morgan and Smith (1984) The two most important gastroenteritis viral agents for which EIA tests have been developed are rotavirus and enteric adenoviruses. An EIA test for rotavirus was first developed by Yolken et al. (1977) , and commercial assays are now available. A recent evaluation of two commercial products (Ro-tazyme, Abbott Laboratories; Enzygnost, Behring) showed the sensitivity of Rotazyme to be 88% and Enzygnost, 98% (Morinet et al., 1984) . The standard for comparison was electron microscopy (EM). Both EIA methods appear suitable for use if EM is not available, although Rotazyme is known to cause false positive results in samples from neonates Chrystie et al., 1983) and is insensitive in samples from adults (Herrmann et al., 1985) . Enteric adenoviruses (types 40 and 41) are difficult to isolate; therefore EIA methods would be preferable if the sensitivity was satisfactory. Preliminary results from Johansson et al. (1980) suggest that development of an EIA specific for enteric adenoviruses is possible. The use of monoclonal antibodies in EIA tests offers two potential advantages: (1) improved specificity due to the nature of monoclonal antibodies, and (2) improved sensitivity by allowing for clearer EIA cut-off values. Sensitivity could also be increased by increasing the amount of detector antibody used in an EIA. However, because monoclonal antibodies react with only one epitope of a given antigen, more than one monoclonal antibody may be needed to achieve the desired sensitivity. In practice, monoclonal antibodies have been used successfully in latex agglutination tests and also in immunofluorescence techniques (Nowinkski et al., 1983) . Their use in EIA has been limited to date, but the number described for microbial antigens to date suggest that applications in clinical diagnosis will be increasing. The EIA tests that have been developed look promising. For EIA detection of adenovirus group antigens in stools, monoclonal antibodies were as sensitive as polyclonal ones, and were more sensitive for detecting noncultivatable adenoviruses (presumably enteric serotypes). All 12 stool samples positive by EM were monoclonal EIA positive (Anderson et al., 1983) . Use of monoclonal antibodies for detection by EIA of microbial antigens in cerebrospinal fluid (CSF) also look promising. In a comparison of polyclonal and monoclonal EIA tests for group A meningococcal antigens in CSF, 21 of 25 CSF specimens positive by polyclonal EIA were positive by monoclonal EIA (Sugasawara et al., 1984) . In a preliminary study, 5/5 CSF specimens positive for group B streptococcal antigen reacted in a monoclonal EIA (Morrow et al., 1984) . For diagnosis of rotavirus infection, we have found that a monoclonal EIA was 100% sensitive and specific for samples from adults and neonates as well as young children (Herrmann et nl., 1985) . This was possible due to the high affinity and broad group specificity of the monoclonal antibody used (Cukor et al., 1984) . Preliminary results of a monoclonal EIA for diagnoses of legionellosis showed positive correlations in 3 / 3 cases (Bibb et al., 1984) . However, in a preliminary study on using a monoclonal EIA for diagnosis of bubonic plague, only 2 of 10 were positive (Laughon et al., 1984) . In this situation, where the sensitivity is low, use of polyclonal sera in a control EIA would be desirable. This would help determine if the problem was the monoclonal antibody or the amount of antigen present in the clinical sample. Although many of the current EIA tests for microbial antigens and antibodies have not realized their potential, there are reasons to believe this situation will improve. For detection of antibodies, the major problems are standardization of reagents and EIA methodology. This should improve when more reagents become commercially available, and when a standard method is selected from the variety of procedures now available. For detection of antigen, which offers a rapid and direct means of diagnosing microbial infections, the major problems has been lack of sensitivity. Increasing the sensitivity of polyclonal EIA tests by using more concentrated immunoreagents or more sensitive enzyme substrates has often resulted in a loss of specificity. From the reports available to date, it appears that use of the appropriate monoclonal antibodies may solve the problem of sensitivity for detecting many infectious agents in clinical samples. With the increasing number of monoclonal antibodies available for varius microbial antigens, we can expect that more of them will be utilized for EIA detection systems. If the affinities of monoclonal antibodies can be increased, the EIA tests may be sufficiently sensitive. Another approach which is being taken for rapid diagnosis is the use of nucleic acid probes in hybridization techniques (review by Richman et al., 1984) . To date, most of the probes have used radioactive labels (32P) and require 1 or 2 days for assay, which makes them impractical for clinical laboratories. The use of biotin labels coupled with enzyme markers may improve both the speed of the assay and the sensitivity of the assay (Richman et al., 1984) . Whether this technique will become useful and practical for direct detection of microorganisms in clinical specimens remains to be determined. Lancet 1, 1364 Diagnostic Procedures for Viral, Rickettsial, and Chlamydia1 Infections Rapid Diagnosis in Infectious Disease Immobilized Enzymes, Antigens, Antibody, and Peptides Immunoassays for the 80's