Further investigations on Plasmodium gallinaceum as an antigen in the diagnosis of human malaria.

P. GALLINACEUM AS ANTIGEN IN DIAGNOSIS OF MALARIA: FURTHER STUDIES 617 Further Investigations on Plasmodium gallinaceum as an Antigen in the Diagnosis of Human Malaria * by A. KIELMANN,1 G. SARASIN,2 A. BERNHARD & N. WEISS Several authors have investigated cross-reactions between Plasmodium gallinaceum and human and simian malaria species. Whereas Lippincott (1945) demonstrated some cross-reactions using the comple- ment-fixation test, Ingram et al. (1961) and Voller (1962) failed to show any reactions between P. galli- naceum and the primate malaria species. According to El-Nahal (1967) the exoerythrocytic forms seem to be even more specific than the blood forms. As P. gallinaceum, in contrast to the simian plasmodia, is easily maintained and therefore readily available in most laboratories, we have re-examined its use as antigen in human malaria, and in contrast to the above-mentioned authors, were more successful (Kielmann & Weiss, 1968a, 1968b; Kielmann et al. 1970). Thus of 30 cases of parasitologically proven malaria (13 P. falciparum, 12 P. vivax, 3 P. malariae and 2 P. ovale) all yielded positive titres ranging from 1/10 to 1/640. Similarly, we detected antibodies to malaria in a fair proportion of persons returning from malaria endemic zones by use of this antigen. In the present study, our aim was to extend our investigations to another strain of P. gallinaceum and to compare P. falciparum, P. cynomolgi bastia- nellii and P. gallinaceum as antigen for the diagnosis of human malaria. Furthermore we attempted to elaborate possible reasons for the differences existing between our own results and those of other authors by examining the influences which the following variables could exert on the suitability of P. gallinaceum as antigen for the detection of anti- bodies to human malaria: (1) differences in the maintenance of the indi- vidual strain, i.e., cyclic versus acyclic transmission; (2) time chosen during parasitaemia for the preparation of blood film antigens; * From the Swiss Tropical Institute, Basle, Switzerland. lPhysician, Swiss Tropical Institute. Present address: Department of International Health, Johns Hopkins Uni- versity, School of Hygiene and Public Health, Baltimore, Md 21205, USA. ' Present address: Agrochemical Department, CIBA Ltd., Basle, Switzerland. (3) age of the host animals; and (4) production of antibodies by P. gallinaceum infection in the host animals. Materials and methods Experimental animals. Hybro chickens, a strain maintained locally but of Dutch origin, were used throughout. Antigens. The P. gallinaceum strain used con- sisted of 2 variants that had been kept at CIBA Ltd., Basle, since 1954, One of them was maintained there exclusively by blood passages, whereas the other had been kept cyclically at the Swiss Tropical Institute (STI) since 1959 as described previously (Kielmann et al 1970). For reasons of simplicity we shall refer to the former as the CIBA strain and to the latter as the STI strain. Another P. gallinaceum strain originated from the Weilcome Laboratories, but had been kept at the Liverpool School of Tropical Medicine since 1952, where it had undergone more than 200 mosquito passages. It was sent to us in 1969 by Professor B. G. Maegraith. Blood films infected with P. cynomolgi bastianellii were obtained from Dr P. Ambroise-Thomas, Department of Parasitology, University of Lyon, France. Thin smears of P. falciparum-infected monkey blood were provided by Dr A. Voller, Nuffield Institute of Comparative Medicine, London, Eng- land. Parasitaemia of the infected blood smears was estimated on Giemsa-stained slides prepared at the same time as those used as antigens. Antisera. Sera from parasitologically confirmed cases of malaria due to P. fakciparum, P. vivax, P. malariae and P. ovale were used. Sera from in- fected chickens were collected at different intervals following infection by puncture of the major vein of the wing. All sera were stored at - 20°C until use. The fluorescein-isothiocyanate labelled anti- human serum, obtained from the Institut Pasteur, Paris, was used at a dilution of 1/40, whereas the labelled anti-chicken serum, obtained from 2586G NOTES Nutritional Biochemicals Corporation, USA, could only be used at a dilution of 1/5. The preparation of the antigen smears and the execution of the fluorescent antibody (FA) technique as well as the microscope and equipment used have been described in our previous paper (Kielmann et al. 1970). Results To begin with, we compared the value of different plasmodial species as antigens for the diagnosis of human malaria. As not all the antigenic materials mentioned above were available at the same time, we performed two series of tests. In the first one, we compared the STI-strain of P. gallinaceum with P. c. bastianellii (Table 1), whereas in the second one the STI and the Wellcome strains of P. gailinaceum were tested simultaneously with P. falciparum (Table 2). Of the tests with P. c. bastianellii as antigen, 21 were performed in Lyon, and the remaining tests were done by us with material supplied from there. For two of the sera, presented in Table 1 (87 and 190), the antibody titres had also been determined in a homologous system, before the sera were sent to us. These homologous titres were found to be 1/80 in both cases. The results of this series indicate that no over-all superiority could be demonstrated for either the primate or the avian antigen. In the second series of tests (Table 2), the two strains of P. gallinaceum yielded, on an average, practically identical results. No differences in titre of more than 1 dilution were observed in any case. The human titres arrived at, using P. falciparum as antigen, were, even in the homologous system, on average only about 1 dilution higher than those obtained with the avian antigen. In the second part of our work, the influence of various factors on the antigenic reactivity of P. gallinaceum were examined. First we wanted to find out if the quality of the inoculum, as influenced by the maintenance of the strain, and the moment during parasitaemia chosen for the preparation of the antigen were of any importance. For this purpose, 8-10-day-old chickens, in groups of 5, were infected with the following inocula: Group I-sporozoites from the STI strain Group II-blood, parasitized by the STI strain Group III-blood from the 862nd blood passage of the CIBA strain. TABLE 1 COMPARISON BETWEEN THE STI STRAIN OF PLASMODIUM GALLINACEUM AND P. CYNOMOLGI BASTIANELLII AS ANTIGENS Reciprocal titres Serum Antiserum to With With P. cynomolgi no. P. gallina- bastianellii ceum I STI Basle Lyon 11 18 196 201 381 389 398 412 6 7 8 156 161 221 190 7996 8364 87 2 31 146 220 283 2440 10 Sera P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. falciparum P. falciparum P. fakiparum P. falciparum P. falciparum P. falciparum P. malariae P. malariae P. malariae P. ovale Undetermined Undetermined Undetermined Undetermined Undetermined Undetermined Undetermined 80 320 40 20 40 160 160 40 20 40 20 10 160 20 80 80 40 40 10 10 20 40 80 40 neg. 160 20 negative 160 20 10 negative negative negative negative 80 ±10 40 ±10 80 40 40 40 320 40 40 negativea 20 20 40 negative 80 negative a Initial dilution in Lyon, France = 1/20. As soon as parasitaemia exceeded 1 %, thin smears to be used as antigens were prepared almost daily until the end of parasitaemia or death of the animal host. In Group I, parasites could only be detected between day 8 and day 12 after infection, para- sitaemia never exceeding 50%. In Group II, para- sitaemia started on day 4, reached a maximum . ~ ~~ ~ 618 P. GALUNACEUM AS ANTIGEN IN DIAGNOSIS OF MALARIA: FURTHER STUDIES TABLE 2 COMPARISON OF THE STI AND THE WELLCOME STRAINS OF PLASMODIUM GALWNACEUM AND P. FALCIPARUM AS ANTIGENS Reciprocal titres with Serum Antiserum no. to P. gallinaceum P. falci- STI Wellcome parum a 6 7 8 52 54 447 448 449 18 196 302 381 384 398 410 412 413 418 434 460 492 44 60 620 7 996 8 364 P. fakiparum P. falciparum P. falciparum P. fakiparum P. falciparum P. falciparum P. falciparum P. falciparum P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. vivax P. ovale P. ovale P. ovale P. malariae P. malariae 20 20 40 80 160 160 160 160 40 20 +20 20 160 80 40 40 20 80 80 20 40 40 20 20 160 20 40 20 20 80 320 320 160 160 80 20 20 20 160 80 80 80 40 80 40 20 20 40 20 ±20 160 ±20 80 160 160 80 320 320 320 320 160 80 80 40 80 160 ND 80 ND 640 160 20 20 20 ND 20 160 40 a ND = Not done. (70%-100%) around day 10 and declined slowly thereafter. Only in 2 animals was a recrudescence of peripheral parasitaemia observed between day 17 and day 19. Parasitaemia in Group III followed a similar pattern, but never exceeded 80% and, in this group, only 1 animal had a recrudescence of parasitaemia. Typical courses of the infection from one animal of each group are given in Table 3. All the other birds within the same groups exhibited very similar, at times identical, patterns. As we did not have sufficient quantities of any single serum to test the antigen smears of all chickens prepared on all the different days, we were forced to use the following three sera: No. 7 from a P. falciparum infection with a reciprocal titre of 40, No. 357 and No. 395 from two P. vivax cases both ofwhich had reciprocal titres of 160,when tested routinely against P. gallinaceum as antigen as des- cribed in our previous paper (Kielmann et al. 1970). Very soon we realized that titres obtained with antigens prepared on different days varied very little. Therefore we did not carry out reactions with all the slide antigens prepared from all the birds of the same group, but limited the examination to 46 separate preparations, made between day 6 and day 19 after infection. Some of these we tested against two different antisera. The titres of our sera, reacting with the various antigens, never differed by more than 1 dilution. As there was no definite pattern in these small TABLE 3 TYPICAL COURSES OF P. GALLINACEUM INFECTION FOLLOWING INOCULATION BY SPOROZOITES FROM STI STRAIN (GROUP I), BY BLOOD PARASITIZED BY STI STRAIN (GROUP II) AND BLOOD FROM THE 862nd BLOOD PASSAGE OF THE CIBA STRAIN (GROUP 1II) No. of days Percentage of red cells parasitized a after infection Group I Group II Group III 4 NE <1 NE 5 NE 5 NE 6 NE 10 <1 7 0 30 <1 8 <1 80 <1 9 30 90 20 10 50 100 80 11 20 70 80 12 <1 50 50 13 0 20 20 14 0 20 10 17 NE <1 0 19 NE 0 0 21 NE 0 0 a NE = not examined. 619 NOTES fluctuations, they had to be ascribed to the technique. It thus seems that all our antigen preparations, whether made early or late during either blood- or sporozoite-induced parasitaemia, were equally sensitive for the detection of antibodies to P. falci- parum and P. vivax. In a second experiment, we examined the influence of the age of our animal hosts on the course of infection, on the antibody production and on the antigenic quality of blood films. Six-month-old chickens of the same breed were thus selected: we assumed that they would be more resistant to infection by nature of their age. Todorovic et al. (1968) had infected 5-month-old birds intravenously with 2 ml of heavily parasitized blood and we decided to give identical inocula to our chickens. Chickens no. 1-3 received the STI strain and chickens no. 4-6 received the CIBA strain. When we wanted to prepare the first blood films, 4 days after inoculation, chickens no. 2 and 6 had succumbed to their infections. Blood isolated at this time by cardiac puncture revealed parasitaemias between 80% and 100%. Three of the surviving animals were in rather poor general condition and, on examination, exhibited parasitaemia ranging from 50% to 100%. To prevent further deaths, they were therefore treated with 1 mg/kg bodyweight of pyrimethamine the first day, and, as little improve- ment had occurred 24 hours later, received another 0.5 mg/kg of the same drug. By contrast, chicken no. 5 showed no parasitized red blood cells until day 6 after infection and survived without treat- ment. In Table 4 the courses of parasitaemia are given for the surviving birds. All serological reactions in this experiment were carried out with the P. vivax, antiserum no. 395. Similar and well-comparable titres were achieved with all antigen preparations from our 4 chickens. Thus, in most instances, the previously recorded reciprocal titre of 1/160 was obtained. Neither the different times of their preparation during parasitaemia nor the age of the animal hosts seemed to have had any apparent influence. Only a few reactions gave randomly distributed titres of 1/320 or 1/80. In the last experiment, we determined antibody formation in the animals used as antigen source. From the chickens which were infected when they were 8-10 days old, serum samples were taken between the 15th and the 60th day after infection. To our surprise, none of these contained any anti- bodies, measurable by the FA technique. Three TABLE 4 COURSE OF PARASITAEMIA IN 6-MONTH-OLD CHICKENS RECEIVING 2 ml OF HEAVILY PARASITIZED BLOOD Percentage of red cells parasitized No. of days after Chicken Chicken Chicken Chicken infection no. 1 no. 3 no. 4 no. 5 (STI) (STI) (CIBA) (CIBA) 4 50a, b 100a, b 60a, b 0 5 50a, b 60a, b 50a, b 0 6 20 20 20 loa 7 <1 a <1 a <1 20 a 8 <1 <1 <1 30 a 9 0 0 0 50 a 10 0 0 0 80 11 0 < <1 20 a 12 0 <1 a <1 10 13 <1 <1 a <1 a 0 14 <1 1 <1 0 15 <1 1 0 0 18 0 0 0 0 19 0 0 0 0 20 0 0 0 0 a Times at which blood films to be used as antigens were prepared. b Treated with pyrimethiamine 1 mg/kg on day 4 and 0.5 mg/kg on day 5. birds, one from each group of experiment 1, were inoculated a second time with 1 ml of heavily parasitized blood 1 month after the first infection. This second infection produced in all cases only a slight and transient parasitaemia. But again no antibodies could be detected in the serum samples taken 12 and 24 days later. By contrast, all older chickens, except no. 5, readily exhibited antibodies in their serum, when examined 11, 18 and 27 days after infection. Chicken no. 5 was found to react positively only on days 18 and 27. Discussion The purpose of this investigation was to corro- borate the previously observed cross-reactions between P. gallinaceum and human malarial anti- bodies and thus to prove that P. gallinaceum was a valid antigen for the diagnosis of human malaria. 620 P. GALLINACEUM AS ANTIGEN IN DIAGNOSIS OF MALARIA: FURTHER STUDIES From the results we obtained with the Wellcome strain of P. gallinaceum it seems that these cross- reactions are not limited to our particular strain, but may probably be obtained with any strain of this parasite. The antibody titres we arrived at, using P. c. bastianeliii or P. falciparum as antigens, were unexpectedly low. The simian antigen proved to be superior in only 2 out of 18 parasitologically confirmed cases of human malaria. Six of these sera from confirmed malaria patients yielded simi- lar titres with the simian and the avian antigen and for the remaining 10 proven cases the titres with P. gallinaceum were even higher. The latter results were so surprising, that we repeated these tests several times using both antigens simultaneously. The outcome was always the same. The rather pronounced differences, seen in a few instances, will have to be further investigated. With P. fakciparum the titres were generally some- what higher than those obtained with the two strains of P. gallinaceum. The differences were practically the same whether homologous or hetero- logous sera were used and amounted, on an average, to about only 1 dilution. But as 1 dilution is accepted by many authors as being within the range of tech- nical errors, a superiority of P. falciparum could, at best, be proved with a larger number of sera. We were surprised to see that the titres observed in the homologous P. falciparum system were not decidedly higher than those found in heterologous systems. But we had already recorded a similar relationship between homologous and heterologous titres for the P. malariae serum no. 190, which showed identical titres against P. malariae, P. c. bastianellii and P. gailinaceum. Also the P. ovale serum no. 87 yielded similar results with the homologous and the avian antigen. It might thus be speculated that, in all instances, we were working with a genus specific antigen-antibody system and that the species specific antigens were not involved at all. We cannot, as yet, give any explanation for the fact that, except for the homologous systems, we got, in some cases, even higher titres with P. gailinaceum than with the primate antigens. The quality of the antigens may have had some influence. The primate antigens had been prepared abroad and may have deteriorated during shipment. In contrast the P. gallinaceum antigen-slides were made on the spot and used under optimum conditions. The most important point for us, however, was that all the sera from parasitologically confirmed cases of malaria again gave positive reactions with P. gallinaceum. In the few instances where the titres remained below the specific level of 1/20, we were dealing with sera which were taken during the first days of a primary attack of malaria, and it was shown by Tobic et al. (1966) that measurable amounts of antibody appear only some days after the onset of parasitaemia and clinical illness. The additional finding, that there was very little difference between the titres obtained with either primate or avian antigens, constitutes a further indication for the use of P. gallinaceum for the serological diag- nosis of human malaria. Since during our earlier work we had always used antigen smears prepared between 9 and 11 days after a blood-induced infection with the STI strain, we thought that by varying these experimental conditions, it might be possible to explain the con- flicting opinions regarding these cross-reactions. In this respect we failed, as the cross-reactions persisted despite all the variations we applied. It is well known that the acyclic transmission by blood passages may alter a parasite. However, as the differ- ent variants of our strain yielded identical results, it became clear that the antigen we were dealing with was not involved in any possible variations due to such causes. On the contrary, it seemed to be remarkably stable. The other questions that we were investigating were all in connexion with the formation of anti- bodies by the infected chicken. Thus they will be discussed together. As Kreier & Ristic (1964) had observed that in P. berghei-infected mice an antigen-antibody complex formed in vivo beginning with the 10th day after infection, we wanted to know if a similar phenome- non occurred in P. gallinaceum-infected chickens. We thought that the fixation of such antibodies on to the plasmodia could have the same effect as an absorption. This would then either reduce or abolish the capacity of the parasites to react with human antibodies. As the decline of parasitaemia is often thought to be due to the appearance of antibodies, we expected to get antigen smears of good quality up to the peak of the parasitaemia and smears with reduced or no antigenic activity thereafter. But as we were able to demonstrate, the antigenic activity of our parasites was never reduced, even when the smears were prepared as late as the 19th day after infection. This could possibly be explained by the fact that, when we examined the sera of our young birds, we could not 9 621 622 NOTES detect any fluorescent antibodies, even after a second inoculation with parasitized blood, to which they were, nevertheless, highly resistant. By contrast, Kreier & Ristic (1964) were able to demonstrate significant antibody levels in P. berghei- infected mice, although in these hosts the disease usually ends fatally and parasitaemia can be observed to increase until death. Kuvin & Voller (1963) had previously warned, however, that in malaria as in other parasitic diseases, the level of serum anti- bodies did not necessarily reflect the degree of resistance to the parasite. In the sera of our older chickens, antibodies to the P. gallinaceum infection could be detected by the 11th day. Slides prepared from them on days 12 and 13 showed very few parasitized red blood cells. These few cells, however, gave identical titres with our standard sera. This could be the result of two opposite factors: the abnormally small amount of antigen present would presumably be responsible for an increase in titre, whereas the antibodies produced by the animal host would bring the titres back to normal. Garin et al. (1966) observed that in a P. c. bastia- nelli infection the antigenic value of the parasites decreased as the infection progressed. But they did not attempt to relate this observation to the presence of antibodies in their monkeys. It is, thus, not yet possible to state conclusively, that the presence of antibodies does alter the quality of an antigen. Voller (1962, 1964) in his publications on the cross-reactions between the different plasmodia did not give any technical details about the mainte- nance of the strain, the dilutions of the sera, etc. From our work, however, it seems that experimental condi- t:ons were not responsible for our contradictory results. ACKNOWLEDGEMENTS We are greatly indebted to Professor B. G. Maegraith, Dean of the Liverpool School of Tropical Medicine, England, for his advice and encouragement as well as for supplying us with his strain of P. gallinaceum. Our sincere thanks also go to Dr P. Ambroise-Thomas and Dr T. Kien Truong, University of Lyon, France, for carrying out part of the investigations and for giving us some P. c. bastianelli-infected blood films. We further thank Dr A. Voller, Nuffield Institute of Comparative Medicine, London, England, and Dr P. Suter, CIBA Ltd, Basle, Switzerland, for providing us with their strains of P. falciparum and P. gallinaceum, respectively. REFERENCES El-Nahal, H. M. S. (1967) Bull. Wld Hlth Org., 37, 154 Garin, J. P., Rey, M. & Ambroise-Thomas, P. (1966) Bull. Soc. Path. exot., 59, 316 Ingram, R. L., Otken, L. B. & Jumper, J. R. (1961) Proc. Soc. exp. Biol. (N.Y.), 106, 52 Kielmann, A. & Weiss, N. (1968b) Trans. roy. Soc. trop. Med., Hyg., 62, 458 Kielmann, A. & Weiss, N. (1968a) Acta tropica, 25, 185 Kielmann, A., Weiss, N. & Sarasin, G. (1970) Bull. Wld Hlth Org., 43, 612 Kreier, J. P. & Ristic, M. (1964) Amer. J. trop. Med. Hyg., 13, 6 Kuvin, S. F. & Voller, A. (1963) Brit. med. J., 2, 477 Lippincott, S. W. et al. (1945) J. clin. Invest., 24, 362 Tobie, J. D. et al. (1966) Amer. J. trop. Med. Hyg., 15, 676 Todorovic, R., Ferris, D. & Ristic, M. (1967) Ann. trop. Med. Parasit., 61, 117 Voller, A. (1962) Bull. Wld Hlth Org., 27, 283 Voller, A. (1964) Amer. J. trop. Med. Hyg., 13, 204