Serological testing in malaria*

Memoranda Serological testing in malaria* The main purpose of this paper is to evaluate, in a critical manner, various serological tests with general emphasis on their value in the epidemiological assessment of malaria. Several tests have been employed in the past. However, the present memorandum will deal only with the methods that have been widely used recently-i.e., indirect immuno- fluorescence (IFA), passive haemagglutination (IHA), and gel-diffusion. The three immunoglobulins most commonly involved in these tests are IgG, IgM, and-to a lesser extent-IgA. NECESSITY FOR SEROLOGICAL TESTING IN MALARIA In many present-day circumstances accurate appre- ciation of the prevalence of malaria is difficult without detailed knowledge of serological indices. There is thus need for a more appropriate classifica- tion of endemicity based on indices that would faci- litate the monitoring of change with time in malaria endemicity and would be of value in indicating the correct priority to be accorded to malaria in health planning. In areas where malaria is or has been endemic, serology may be especially useful for the following purposes: (1) For the establishment of malarial endemicity rates including species prevalence (in particular, age-specific indices). (2) To assess changes in the degree of malaria transmission, usually during or after malaria eradi- cation or control operations. (3) To permit specific epidemiological assessment of malaria-e.g., to delineate malarious areas; to show the altitude delineation of malaria; and to identify and follow up foci of malaria. (4) To identify areas or individuals requiring action with regard to malaria, especially during the later stages of malaria control programmes. To gain the maximum benefit, the serological and parasitological information should be collected and evaluated together. * This Memorandum was drafted by the signatories listed on page 532. Reprints can be obtained from: Research and Technical Documentation, Division of Malaria and Other Parasitic Diseases, World Health Organization, 1211 Geneva 27, Switzerland. A French translation of the Memorandum will be published in a future issue of the Bulletin.S - q > 0 In areas where malaria is not endemic, serology can be of use for the following purposes: (a) For case-detection and identification, in some instances, of the species of malaria parasite respon- sible for the infection. (b) For the screening of blood donors. (c) To exclude the diagnosis of malaria in patients with symptoms such as pyrexia of unknown origin, hepatosplenomegaly, anaemia, and nephrotic syn- drome. COLLECTION OF SERUM AND PLASMA SAMPLES Serum and plasma can both be used in the tests discussed here. In the many instances where vene- puncture is inadvisable, it is possible to collect blood by finger puncture into heparinized capillary tubes. It is generally agreed that, wherever possible, the capillary-tube method should be used for the collection and transport of samples, although under certain conditions the only practical way may be to use the filter paper technique. Capillary tubes The finger is cleaned with an alcohol-moistened swab and is allowed to dry in order to prevent haemolysis of the sample. A disposable lancet is then used to puncture the finger and the first drop of blood is wiped off. One or more labelled heparin- ized capillary tubes (with a minimum internal dia- meter of 1.1 mm and a length of 75 mm), preferably containing dried thiomersal, are then filled. The capillary tubes are sealed at one end by plugging with plasticine or by heat. They should be centrifuged within a few hours with an electrical centrifuge or 3220 -527- 528 MMEMORANDA effective alternative. The capillaries are then cut at the plasma-cellular interface. The plasma section is again sealed at both ends with plasticine and stored. Filter paper Under particular conditions in the field, the blood samples collected in capillary tubes may be absorbed on filter paper from which the serum can be extracted prior to serological testing. In this case, it will' be necessary to absorb precisely measured quantities of blood on the filter paper-e.g., by means of capillary tubes. The impregnated filter papers must not come into contact with fixing agents and should be dried as soon as possible, avoiding temperatures exceeding 56°C. Although these dried blood samples on filter paper are known to withstand normal environmental temperatures in subtropical condi- tions for at least a fortnight without loss of sero- reactivity, it is advisable to store them as soon as possible at -20°C (experience has shown that they can be preserved at that temperature for at least a year). For immunofluorescence tests, the eluate is used. With 70-pl blood samples, such as those taken with standard haematocrit capillary tubes, each filter paper is placed in 0.7 ml of buffer solution of pH 7.2, and is left overnight at a temperature of 4°C. This corresponds to a dilution of 1/20 on an assessed haematocrit value of 50%. Even when the haematocrit values are very low, over-estimation of the titre value is not great. Recent investigations suggest that absorption of blood on filter paper may destroy the reactivity of antibodies of the IgM fraction. Seroepidemiological investigations in which these antibodies may play a role should therefore be conducted with fresh plasma samples. The presence of haemoglobin in plasma eluate obtained from filter paper specimens reduces fluorescence-hence the need to use at least two dilutions. STORAGE OF SERUM There appear to be no reports of any formal studies on the stability of malaria antibodies in human serum during storage, but general experience suggests the following: (1) Ideally tests should be carried out on fresh serum. (2) Bacterial growth should be avoided. This can be achieved by handling the serum so as to avoid gross contamination and by adding sodium azide or thiomersal to the blood or serum sample. (3) Serum should be stored in the cold. For long- term storage, it should be frozen (-700C is better than -20°C). (4) Thawing and refreezing may be harmful. Therefore all tests should be performed simul- taneously on any batch of thawed serum, or aliquots should be frozen separately to be thawed as required for tests performed at different times. (5) IgM antibodies or " early " antibodies may be liable to loss of activity on storage. (6) Freeze-drying under the carefully controlled conditions used for the preparation of biological standards appears to be a good way to preserve the properties of the serum. However, this is not a method for general application. (7) The effectiveness of a storage procedure should be assessed by comparing the properties of the serum before and after storage, or after different conditions of storage, by means of standardized tests. INDIRECT IMMUNOFLUORESCENCE The antigen The antigen consists of thick or thin blood films on glass slides and is made from the blood of malaria- infected animals or people. From a practical point of view, thick smears are preferable. Antigen preparations should contain mature schizonts when- ever possible, since these give maximum sensitivity to the test. For the detection of antibodies, the homologous antigens should be used, except in the case of antibodies to Plasmodium malariae where either the homologous species or P. brasilianum can be employed. A mixed-species antigen will give maximum sensitivity, but is very difficult to prepare. Suitable P. falciparum antigen can be obtained by the in vitro cultivation of heavily infected human blood.a Suitable P. vivax and P. malariae antigens are less readily obtained direct from human sources because the parasite densities of these infections are often too low. Antigen preparations can be made from the already adapted strains of P. falciparum, P. vivax, or P. mala- riae in splenectomized Aotus trivirgatus (dourocoulis or owl monkeys), especially in the subspecies A. t. griseimembra from Colombia. A. t. trivirgatus from a Bull. Wld Hlth Org., 47: 357-373 (1972). SEROLOGICAL TESTING IN MALARIA 529 Colombia and Peru will support the development of a more limited number of strains. If homologous antigens are not available, then P.fieldi from infected splenectomized rhesus monkeys is probably the best alternative general-purpose antigen. P. cynomolgi (also in rhesus monkeys) can be used as a second-best antigen for P. vivax. Other species of Plasmodium are considered to be un- suitable. Monkeys are infected with the relevant species of malaria parasite and, when the parasitaemia reaches a density of 1-5% schizont-infected erythro- cytes, blood is taken from the animal with an anti- coagulant. The blood is centrifuged at 1 500 g for 10 min, the supernatant fluid is discarded, and the erythrocytes are resuspended in phosphate-buffered saline. The centrifuging and washing procedure is repeated three times. The erythrocytes are then made up to a 2-4% cell suspension in phosphate-buffered saline. It is best to complete the preparation of antigen slides on the day the blood is collected. Storage of the whole blood prior to use as antigen leads to loss of potency. Before large batches of antigen are prepared it is advisable to use a few slides in an IFA test with positive and negative control serum. If this prelimi- nary test is satisfactory, thick smears with 1-16 spots on a slide may be prepared in bulk. The antigen slides are air-dried and placed in a desiccator (calcium chloride, silica gel, or other desiccant) for at least 3 h before being individually wrapped in absorbent paper. Small packets of slides are sealed in plastic bags or plastic foil and are stored at a low tem- perature. It has been shown that antigen slides can be satisfactorily stored for several years at -70°C and the present evidence suggests that storage under dry conditions at -20°C is satisfactory for at least 6 months. Slides may be stored at ambient tem- peratures for several weeks if great care is taken to ensure that they are kept completely dry. This facilitates transport. The conjugate The sensitivity and specificity of the IFA test is largely dependent on the conjugate and the dilution at which it is used. Conjugated antisera prepared against the total immunoglobulins or against spe- cific immunoglobulins can be used. No chemical or biological parameters have yet been found to be satisfactory indicators of the potency of such anti- sera. Conjugates should be assessed in malaria IFA tests against one of the standardized reagents.a Rheumatoid factor-an autoantibody of IgM class reactive with IgG-may combine with IgG antibody attached to parasite antigen and so give a falsely positive result in the rFA test for JgM antibodies. There is no simple method of separating this activity from antimalarial IgM antibody. Hence, serum that is positive for rheumatoid factor should be regarded as suspect. Test procedure Annex 1 describes one way in which the IFA test may be carried out and read. Standardization Malaria IFA tests can be made comparable in different places and at different times only if refer- ence antisera and antigens are exchanged and if the different conjugates used are specified in terms of established reference conjugates. Merits of the test (1) The preparation of comparable batches of antigen is relatively simple. (2) The whole infected cell, morphologically iden- tifiable, is used as antigen. (3) Many studies have shown that the test can reflect experience of malaria by an individual or community. (4) The results of the test can be used to show differences in malarial endemicity between localities, and to detect transmission. (5) The test is adequately sensitive. (6) At higher titres, the test is virtually always specific for malaria and sometimes can be used to indicate species-prevalence. (7) Any laboratory with facilities for carrying out IFA tests for other diseases can perform the test for malaria if the antigen is provided. Limitations (a) The reading of results is subjective at present. (b) Malaria-parasite carriers can occasionally give negative reactions. This has been observed especially with children. a Information concerning fluorescein-conjugated anti- human IgG may be obtained on application to the Secretariat of the Standardization Committee, International Union of Immunological Societies, c/o Wellcome Research Labora- tories, Langley Court, Beckenham, Kent, England. MEMORANDA (c) The necessity for specialized equipment and personnel limits the test to major laboratories. (d) Antigens are available from only a few centres and their storage requires considerable refrigeration space. (e) The transport of antigen can present problems. INDIRECT OR PASSIVE HAEMAGGLUTINATION TEST The antigen The antigen is a soluble extract of erythrocytes that are heavily infected with malaria parasites in the schizont stage of development. in the case of P. falciparum, satisfactory antigen has been obtained from the blood of experimentally infected monkeys (dourocoulis) and from heavily infected human placentae. Antigen preparations from rhesus mon- keys infected with various simian malaria parasites have in general been less satisfactory. Test procedure Annex 2 describes a method of carrying out the IHA test. Merits of the test (1) The test is simple to carry out and no special- ized equipment or highly qualified personnel are needed. (2) The test is adequately sensitive and reprodu- cible, and it rarely gives false-positive reactions. (3) Large numbers of sera can easily be handled in this test. (4) The test reagents can be prepared in a central laboratory and can be used in many areas, thus achieving a considerable degree of comparability. (5) The test is eminently suitable for field use and action can be taken on the results the day the sera are collected. (6) All the reagents can be stored in a small space and can be transported easily. Limitations (a) Malaria-parasite carriers, especially children or individuals tested during the initial attack, can show negative reactions to the test. (b) Small variations in the test procedure can drastically affect the results. The extract used con- tains a large number of different antigens ofunknown physicochemical nature, and variations or modifi- cations of technique may lead to a preferential absorption of certain antigens to the red cell. (c) There are at present important differences in the preparation of reagents and in the performance and reading of the tests in different laboratories. (d) Antigens are available from only a few centres. GEL-DIPFUSION TEST This involves the diffusion in agar gel of soluble malarial antigens against antibody-containing sera. Where an antigen encounters its specific precipitating antibody, immunoprecipitation occurs, with the formation of a precipitin line. Studies on the use of this technique in malaria have so far been directed towards P. falciparum infections, and only antigens prepared from asexual erythrocyte forms of this parasite have been used. A method of carrying out the test, the mode of interpretation, and details on the storage of antigen are contained in Annex 3. Antibody patterns detected by gel-diffusion in popu- lation studies made in highly endemic areas The pattern of prevalence of precipitins in sera detected by unheated antigen extracts closely re- sembles that of antibodies detected by the IFA technique. Prevalence is high in the newborn, falls in the ensuing weeks of life, and then slowly increases to reach high levels, approximating to 100% by the 5th year of life. These levels are maintained through- out later life. The number of lines also tends to increase as age advances. Malarial antigens present in aqueous extracts have been broadly classified according to their suscepti- bility to heat. L (labile) antigens are destroyed by heating to 56°C for 30 min. These antigens have been further subdivided into subgroups-e.g., La 1, La 2, and Lb. R (resistant) antigens survive heating to 56°C, but are destroyed within 5 min at 100°C. They also have been subdivided into subgroups- e.g., R1 and R . S (stable) antigens survive boil- ing for 5 min; they, too, have been subdivided into subgroups-e.g., S 1, S 2, etc. So far, some 30 distinct antigens have been noted in association with falciparum malaria. The use of heated antigen extracts has shown that antibodies to different groups of antigens follow distinctive distribution patterns. For example, La antibody prevalence tends to closely resemble the precipitin pattern detected by whole, unheated antigen extracts. On the other hand, antibodies to 530 SEROLOGICAL TESTING IN MALARIA R and S antigens have been shown to have much more restricted distribution patterns. Standardization of the test This presents problems, since individual sera and also extracts made from different batches of infected blood may vary considerably in the number and identity of the antibodies and antigens, respectively, that they contain. Standardization would therefore appear to require first the isolation and purification of individual antigens and subsequently their ade- quate storage. Until this has been achieved, efforts should be made to prepare infected-blood samples containing approximately equivalent numbers of parasites of similar maturity. Merits of the test (1) It requires only simple equipment. (2) Dried stained gels constitute a durable record of results. (3) Large numbers of antisera can easily be studied daily for the presence or absence of precipitins. (4) The identity of antibodies in sera can be estab- lished, as can the identity of different antigens in extracts of infected blood. (5) The test can be used for antigen and antibody analysis in multiple systems. For example, since antigens present in an aqueous extract of infected blood may be destroyed or may survive heating to different temperatures, the use of heated and un- heated extracts can be used to classify antibodies present in serum. (6) The test has also been used to detect the pre- sence of soluble malarial antigens in malarious serum. Limitations (a) Placentae suitable for preparation of the anti- gen are scarce. (b) Determination of the antibody titre is time- consuming and greatly increases the consumption of antigen. (c) Immunoprecipitation is a relatively insensitive method of detecting antibody. Consequently the test may give poor results in areas of low malarial endem- icity, where specific antibody concentrations in sera are low. (d) The test has been noted to yield negative results in the presence of patent parasitaemia in young children. NEW TESTS AN:D IMPROVEMENTS IN ESTABLISHED TECHNIQUES Quantification and automation of immunofluorescence techniques Preliminary studies suggest that the use of a fibre- optic system for the quantitative assessment of malarial immunofluorescence tests eliminates some of the disadvantages inherent in the subjective read- ings of the tests. With this equipment, it is possible to compensate for many of the variables due to technical factors. At present this system is not suitable for routine screening. Other fluorophoto- metric devices are being developed, but have not yet been used in studies on malaria. Pulse lasers can be used for fluorescence excitation, and this reduces the rate of decay of fluorescent preparations. Methods are also being developed for continuous immunofluorescence measurements for use on mov- ing systems. Enzyme-labelled antibody techniques Antibodies conjugated with enzymes (e.g., per- oxidase and phosphatase) are commercially available and can be used in place of fluorochrome conjugates. In this way, the sites of reaction of conjugates can be detected visually by means of a conventional microscope. But at the present time such techniques, when applied to malaria, have proved useful only in immunochemical studies on tissues. However, there is reason to hope that a practical serological test may be devised for use with the same type of conjugates with soluble antigens in an enzyme- linked immunosorbent assay. The results of such a test would be read by means of a colorimeter. Radioimmune assay techniques These techniques are being investigated. Gel-precipitation Preliminary studies with counter-current electro- phoresis suggest that this may be a useful serological technique for malaria. The indications are that it lacks high sensitivity, but its rapidity and possible species-specificity seem promising. Crossed field immunoelectrophoresis techniques have been used successfully in resolving multiple antigen/antibody systems. Differential light-scatter methods and fibre optics may be useful in the qualitative and quantitative assessment of gel-precipitation methods. 531 532 MEMORANDA CONCLUSIONS (1) The establishment of malaria reference sera is of great importance, and such sera should be distributed. The British National Institute for Bio- logical Standards and Controls holds these materials at present and is willing to assist in the implementa- tion of this proposal. (2) For the malaria WFA test, it is recommended that (a) an initial serum dilution of 1: 20 be adopted and (b) the possibility of using only two serum dilutions, i.e., 1: 20 and 1: 80, be considered for large-scale screening. (3) The preparation of a reference batch of IHA reagents should be instituted and assessed in different laboratories. (4) There is an urgent need for comparative studies of the various serological tests in areas of differing malarial endemicity. (5) Studies are required to determine the effects on malarial serology of methods of collection, transport, storage, and treatment of serum specimens, with particular reference to IgM activity. (6) It would be of interest to continue a limited number of critical studies on the relationship between serum levels of malarial antibodies detected by the different techniques and the concentration of other serum components. (7) The possibility of establishing suitable refer- ence antigen preparations should be explored. (8) Studies should be made on the isolation, puri- fication, and characterization of antigens. (9) Techniques are required for the production of antigen from low-density parasitaemias. (10) The possibility of devising methods for the storage and transport of malarial antigens at tem- peratures other than -70°C should be investigated. (11) Because the supply of Aotus trivirgatws and particularly of A. t. griseimembra is limited, it is recommended that laboratories cooperate in their use. The search for, and study of, other nonhuman primate hosts for the human malarias should be further encouraged. The feasibility of breeding these animals in captivity should be investigated. (12) Antigens from primate malaria parasites in addition to P. falciparum need to be studied, par- ticularly by gel-precipitation techniques, in order to establish the patterns related to species-specificity. (13) Alternative techniques, including serological methods for the diagnosis of subpatent parasit- aemias, should be investigated. * * * P. Ambroise-Thomas, Laboratory of Parasitology and Exotic Pathology, Faculty of Medicine and Pharmacy, University Hospital Centre, 38700-La Tronche, France P. Bertagna, Research and Technical Documentation, Division of Malaria and Other Parasitic Diseases, World Health Organization, Geneva, Switzerland W. E. Collins, Unit on Primate Malaria, Laboratory of Parasitic Diseases, National Institutes of Health, Cham- blee, Ga., USA T. Godal, Immunology, Division of Non-Communicable Diseases, World Health Organization, Geneva, Switzer- land G. Gramiccia, Assessment and Training, Division of Malaria and Other Parasitic Diseases, World Health Organization, Geneva, Switzerland J. Haworth, Research and Technical Documentation, Division of Malaria and Other Parasitic Diseases, World Health Organization, Geneva, Switzerland N. Kent, Epidemiological Methodology and Clinical Pathology, Division of Malaria and Other Parasitic Diseases, World Health Organization, Geneva, Switzer- land I. A. McGregor, Head, Laboratory for Tropical Com- munity Studies, National Institute for Medical Research, Mill Hill, London, England J. H. E. Th. Meuwissen, WHO Regional Reference Centre for Malaria Serology, Laboratory of Medical Parasito- logy, University of Nijmegen, Netherlands D. S. Rowe, WHO Immunology Research and Training Centre, Lausanne, Switzerland A. Voller, WHO Collaborating Laboratory for the Development of Malaria Serological Techniques, Nuf- field Institute of Comparative Medicine, Zoological Society of London and London School of Hygiene and Tropical Medicine, London, England W. Wernsdorfer, Research and Technical Documentation, Division of Malaria and Other Parasitic Diseases, World Health Organization, Geneva, Switzerland A. I. 0. Williams, Department of Chemical Pathology, University of Ibadan, Nigeria ACKNOWLEDGEMENTS Thanks are due to Dr A. Voller for collating the material and revising the completed manuscript and to Dr T. Illeni, Department of Bacteriological Biotechnology, Karolinska Institute, Stockholm, Sweden, for providing the information on some of the new tests. SEROLOGICAL TESTING IN MALARIA Annex I THE INDIRECT IMMUNOFLUORESCENCE TEST Test procedure Packets of antigen slides are removed from the freezer and are allowed to warm to room temperature in a desiccator or in their plastic covering before being unwrapped. This avoids damage to the antigen by hae- molysis. Serial dilutions of the test plasma or serum samples are made in phosphate-buffered saline (pH 7.2). A dilution of 1: 20 is a commonly used starting point. Measured quantities of the serum dilutions are then transferred to each antigen spot. The subsequent pro- cessing of the slides is carried out as follows in the range 20°C to 37°C: (1) Incubate with serum dilutions in a humid chamber for 30 min. (2) Wash with phosphate-buffered saline with agitation, 3 changes of 5 min each. (3) Pour off excess saline and dry slides, except for the antigen spots. (4) Apply the diluted fluorescein-labelled antiglobulin conjugate. Incubate in a humid chamber for 30 min. (5) Wash as in (2). (6) Mount slides with coverslips in 10% glycerol in phosphate-buffered saline (or with a commercial moun- tant, particularly if permanent preparations are required). (7) Examine with a fluorescence microscope. Optical systems Various optical systems have been found to be satis- factory and are in use. At the present time it is difficult to assess the relative merits of each. The introduction of reference reagents should permit clarification of this point. Reading of the test The end-point of the IFA titration is the last serum dilution that results in more intense fluorescence of the schizonts than that observed with schizonts reacting with negative control serum. The day-to-day sensitivity and specificity of the test must be monitored by the inclusion of positive and nega- tive control sera. nex 2 THE INDIRECT HAEMAGGLUTINATION TEST Antigen Defibrinated blood or blood taken with an anti- coagulant is obtained from infected monkeys or from human placentae. The cells are centrifuged and the supernatant is discarded. The cells are then washed in phosphate-buffered saline (pH 7.2) and are resuspended to the original blood volume. To eliminate the white blood cells, an equal volume of 3 % high molecular weight of dextran in phosphate-buffered saline is added to the cell suspension and the erythrocytes are allowed to sedi- ment for 1 h. The supernatant containing white cells is discarded. The aggregated erythrocytes are resuspended to the original volume in phosphate-buffered saline. The dextran treatment is then repeated. The sedimented erythrocytes are resuspended in phosphate-buffered saline and centrifuged at 1 100 g for 5 min at 4°C. Resuspension and centrifugation are repeated twice more. The packed erythrocytes are then stored in 0.1-mi aliquots in rubber-stoppered vials at - 70°C. Control antigens are prepared in the same way from uninfected erythrocytes preferably obtained, where a monkey is being used, from the same animal before infection. On the day of the actual test, an ampoule is removed from the refrigerator and 0.9 ml of phosphate-buffered saline (pH 5.5) is added to the cells in the ampoule. The cells in the suspension are then disrupted either by sonication or, less satisfactorily, by freezing and thawing. The suspension is then centrifuged at 8 500 g for 15 min at 4°C and the supernatant is used, undiluted, as the antigen. Preparation of sensitized cells Fixation of erythrocytes. Blood is collected from a sheep in Alsever's solution and is stored at 4°C for up to 8 days. (In the following procedures, all solu- tions are kept sterile and are maintained at a tem- perature of 4°C.) The erythrocytes are packed by centrifugation and washed 3 times in sterile phosphate- buffered saline. Commercially available 25% stabil- 533 MEMORANDA ized glutaraldehyde solution, kept in a refrigerator at 4°C, is diluted to 1 % with a buffer containing 1 volume of 0.15 M phosphate-buffered saline (pH 8.2) + 9 vol- umes of 0.15 M sodium chloride + 5 volumes of distilled water. The sheep erythrocytes are mixed to a 1-2% suspension with this diluted glutaraldehyde solution. The suspension is incubated at 4°C for 30 min with intermittent stirring. The erythrocytes are sedimented by centrifugation at 400 g for 5 min at room temperature. They are then resuspended to give a 20% suspension and are washed 5 times with phosphate-buffered saline (pH 7.2) and 5 times with distilled water. The fixed cells are then made up to a 30% suspension in distilled water and are stored in aliquots of 1 ml at 4°C. Tannic acid treatment of fixed cells. When the fixed sheep erythrocytes, prepared as above, are to be used for sensitization, the supernatant is decanted and discarded. One millilitre of phosphate-buffered saline (pH 7.2) is added to the erythrocytes in each ampoule and the vial is shaken mechanically; a further 8 ml of the saline solu- tion are then added and the suspension is centrifuged at 400 g for 5 min. This procedure is repeated. Then 0.3 ml of packed cells is resuspended in 9.7 ml of tannic acid solution (1: 40 000 in phosphate-buffered saline, pH 7.2) and incubated at 4°C for 15 min. The suspension is centrifuged at 400 g for 5 min and the supernatant is discarded. The cells are resuspended in phosphate-buffered saline (pH 7.2) and centrifuged. Finally the cells are again resuspended in 9.7 ml of the saline solution. Antigen titration procedure. The cell suspension is divided into 5 aliquots and these are centrifuged. The cells are then washed in phosphate-buffered saline (pH 5.5) and centrifuged again, the supernatant being discarded. The contents of the 5 tubes are resuspended in phosphate-buffered saline (pH 5.5) to a total vol- ume of 0.9 ml containing 0, 60, 100, 200, and 300 pi of antigen, respectively. The stoppered tubes are incubated in a water bath at 37°C for 30 min. The tubes are then centrifuged and the cells resuspended in phosphate-buffered saline (pH 7.2), containing 1 % inactivated rabbit serum (i.e., diluent). The cells are then packed by centrifuging, resuspended, washed in the same diluent, and recentrifuged. Four millilitres of diluent phosphate-buffered saline (pH 7.2) are added to each tube. An aliquot of 0.5 ml from one tube is used to determine the haematocrit value of the suspension. The cells in each tube are then made up to exactly 1.3% suspension in the same diluent. Five series of 2-fold dilutions are made, in the micro- titre system, of one negative and three positive reference sera in the range of 1: 40-1: 5 120 with the same diluent. It is imperative to use U-type permanent lucite (Cooke) plates or U-type disposable 6 (Linbro) plates. Volumes of 0.025 ml from each of the batches of sensitized cells are added to one of the series of diluted sera and to a control well containing only diluent. The microtitre plates are mechanically shaken immedi- ately for 2 min. The haemagglutination patterns are read after incubation at room temperature for 1 h and the plates are stored overnight in a refrigerator. The optimum quantity of sensitizing antigen is the smallest amount that shows the highest haemaggluti- nation titre with the set of positive reference sera, while the sensitized cells tested with the negative reference sera and the unsensitized cells tested with either sera or plain diluent show a negative pattern. A haemagglutination pattern is considered to be nega- tive when the well shows a button of sedimented cells. The end-point of the test is read as the point where a ring pattern replaces the button pattern. Sensitization. Cells to be used for the actual test are sensitized with the optimum quantity of antigen as determined in the antigen titration test. This is done for both the test cells-i.e., cells sensitized with malaria antigen-and the cells meant for the preparation of con- trol cells-i.e., cells sensitized with antigen prepared from noninfected erythrocytes (control antigen). Lyophilization of test and control cells Sensitized fixed cells suspended in diluent as 5-20% suspension can be lyophilized when they are not needed for immediate use. For each 3 ml of packed red cells present in the suspension, 0.5 ml of a 12.5% solution of Tween 80 in phosphate-buffered saline (pH 7.2) is added. By means of 5-ml Pyrex ampoules, 1-ml aliquots of the suspension are prepared and freeze-dried for about 18 h. The ampoules can be vacuum-sealed or sealed after filling with dry nitrogen gas. The lyophilized preparation remains stable for over a year at ambient temperatures, even under tropical conditions. When the lyophilized test and control cells are used, diluent must be added to the lyophilized contents of the ampoules. Optimum reconstitution of the cell suspensions from the lyophilized preparations requires that the contents of the ampoules should be mixed carefully so as to avoid any foam for- mation, which would have a deleterious effect on the quality of the cells. Initially, only half the quantity of diluent should be added to the lyophilized control cells, as these cells are first to be used for the absorption procedure. Examination of samples by the IHA test In the actual test, the first step is an absorption pro- cedure for elimination of nonspecific agglutination ac- tivity from the plasma samples that are to be investigated. In the first well of each row of the microtitre plate, 2 drops (i.e., 0.05 ml) of the plasma or serum samples diluted 1: 10 are mixed with an equal volume of control cell suspension in double strength, i.e., 2.6%. With only this first row of the microtitre plate filled, the cell sus- pensions are shaken mechanically for 2 min and incubated at room temperature for 1 h. After that the control cells will have sedimented to the bottom of the well. With serum diluters, 0.025 ml of the supernatant (i.e., 1: 20- 534 SEROLOGICAL TESTING IN MALARIA 535 diluted absorbed sample) is taken from the first well for further 2-fold serial dilution in diluent. These dilutions of the samples are used for mixing with the 1.3% sus- pension of test cells. Another volume of 0.025 ml absorbed sample is serially diluted to 1: 40 and 1: 80 for later addition of 0.025 ml of 1.3% suspension of control cells as a control on the persistence of nonspecific agglutina- tion factors in the sample after absorption. Each plate also includes, as extra control wells, 0.025 ml test and control cell suspensions added to different wells with only 0.025 ml of diluent. The plates are shaken mechanically for 2 min and incubated exactly as described for the antigen titration procedure. For the estimation of the specific antibody titre of the malarious sera in the wells in which the sample dilutions were mixed with the test cells, the sedimentation patterns of all control wells must be negative. Annex 3 THE GEL-DWFFUSION TEST Test procedure Double diffusion by the method of Ouchterlony may be carried out in 1.5% Noble agar, in either 0.3 M phosphate buffer (pH 8.0) or 0.05 M veronal buffer (pH 8.6) applied to microscope slides. Gels prepared with veronal buffer tend to give better resolution of complex precipitin reactions. Wells 2 mm in diameter placed 5 mm (centre to centre) apart have been found satisfactory. Antigen is extracted from infected blood by disinte- gration procedures such as freezing and thawing, sonica- tion, or pressure disintegration with a Hughes Press or similar apparatus. Blood infections should be dense (about 50% erythrocytes infected) and most of the para- sites should be mature or nearly mature schizonts. Extracts are centrifuged (at 28 000 g or more for 20 min and the clear supematant is taken as antigen. As mature asexual erythrocytic forms of P. falciparum are rarely encountered in the peripheral blood of malaria patients, supplies of requisitely infected blood may be obtained (a) from infected human placentae and (b) by in vitro cul- tivation in proprietary media to parasite maturity of tro- phozoite-infected blood obtained from malaria patients.a In each instance the schizont density of infected blood may be increased by slow centrifugation (at 400 g) for 15 min and by taking the schizont-rich chocolate- coloured layer that forms beneath the buffy coat. Ery- throcytic lysis prior to antigen extraction is not recom- mended. Antigen and antisera are diffused in gels at controlled temperatures-usually 4°C-for 48 h in a humid chamber, and slides are then washed in at least two changes of 1 % sodium chloride containing about 0.01 M sodium azide for up to 48 h, care being taken to avoid lifting or breaking the gel. When washed, the wet gels may be a Bull. Wid Hlth Org., 47: 357 (1972). examined by transillumination for the presence of pre- cipitins. The slides are then rinsed in distilled water for 1 h, a strip of wet lint-free filter paper is applied to cover the gel surface, and the slide is allowed to dry slowly at room temperature (usually 16-24 h). Slides may then be stained with an appropriate dye, e.g., amido black. Stained slides should be examined for precipitin reactions, both macroscopically and with the aid of a hand lens (x 8- x 10) in good illumination against a white back- ground. As in other antibody tests, appropriate positive and negative control sera should be included in selected test systems. Mode of interpretation (1) The simplest is to record the negative and positive results. (2) The number of lines present between individual sera and the antigen well may be counted. (3) The titre of antibody can be established by appro- priate dilution of the test sera. Antigen storage Although aqueous extracts of antigen can be stored at - 70°C for many weeks with little apparent deteriora- tion in antigen content, it is probably best to store infected blood at -70°C and to disintegrate aliquots of it when the need for antigen extracts arises. Lyophilized infected blood has been stored in this way for 5 years without apparent deleterious effects. At room temperature, lyophilized infected blood appears to keep satisfactorily for at least I month. Extracts made from infected blood have been noted to lose potency slowly on storage at -20°C and rapidly at 4°C. Loss of activity of the labile group of antigens usually occurs the earliest. 3