Plasmodium falciparum: assay of antigens and antibodies by means of a solid phase radioimmunoassay with radioiodinated staphylococcal protein A

421 TRANSACTIONS OPTHEROYALSOCIETY OPTROPICALMEDICINEANDHYGIENE,VOL.~~, No.3, 1981

Plasmodium falciparum: assay of antigens and antibodies by means of a solid phase radioimmunoassay with radioiodinated staphylococcal protein A* HAVA AVRAHAM,~ JACOB GOLENSER,~DAN T. SPIRA 2 AND Dov SULITZEANU~ 1The Lameberg Center for General and Tumor Immunology, The Hebrew University-Hadassah Medical School, 2Dept. of Protozoology,

Jerusalem, Israel The S. F. Kuvin Center for Study of Infectious and Tropical Diseases, The Hebrew University-Hadassah Medical School, Jerusalem, Israel

Summary Human red blood cells (RBC) infected in vitro with Plasmodium falciparum were employed to prepare several types of-antigens (sonicated, infected RBC and uurified. sonicated merozoites and schizonts). Thkse antigens, as well as control preparations derived from non-infected RBC, were used to coat plastic tubes, which were subsequently tested for capacity to bind anti-P. falciparum antibodies. Binding was detected by means of radio-iodinated staphylococcus protein A. Sera from patients with recent disease or patients who had a history of P. falciparum infection gave strong binding, while sera of normal individuals had only a low binding activity. Some of the antibodies in the positive sera were directed against RBC, since they could bind to tubes coated with normal RBC antigens and could be removed by absorption with RBC. The specifici ty of the P. fakiparum antibodies was confirmed by inhibition tests: preparations derived from infected blood but not from normal blood inhibited the binding activity of the positive sera, to antigen coated tubes. Introduction Several reports have been published in recent years, describing attempts to increase the sensitivity of assays for antibodies against plasmodial antigens using isotopic or enzyme immunoassays (VOLLER et al., 1977; HUNTER ef al., 1979). Yet, these useful techniques do not seem to have been applied to the detection of the corresponding antigens. In a nrevious communication, we described a highly sensitive solid-phase antibody bindinginhibition test. for the assav of Plasmodium berphei antigens and antibodies (ABRAHAM et al., 1980).” In this test, antibodies bound to antigen-coated tubes are detected by means of radio-iodinated staphylococcus protein A (SPA), while antigens are detected by their capacity to inhibit the binding of the antibodies. In the present work we have adapted this technique to the study of P. falciparum antigens and antibodies. Cultures of P. falciparum were used as the source of antigen, while serum from patients with malaria served as the source of antibody. Materials and Methods Diluent Borate buffered saline (BBS), pH 8 was used, unless otherwise stated.

Sera Sera of patients known to have been infected with P. falciparum were obtained from the serum collection kept by the Department of Protozoology. P. falciparum cultures Parasites obtained from Dr. W. H. G. Richards of the Wellcome Laboratories, Beckenham, England, were cultured following the technique of TRAGER & JENSEN (1976). The parasites were grown in 35 mm Petri dishes in either blood group A or 0 erythrocytes, in medium containing loo/, blood group A serum. The medium was RPM1 1640 (Gibco) supplemented with 25 mM Hepes, 0.2% NaHCOs buffer and 50/hg/ml gentamicin (Schering Corporation, USA). Fresh erythrocytes were washed three times with the medium and passed through cellulose powder columns to eliminate the leucocytes The medium was changed (HOMEWOOD, 1976). once a day. When about 8% of the erythrocytes became infected, they were washed three times, were diluted in a fresh, leucocyte-free 7.5% erythrocyte suspension, to O*l-0.5% parasitized cells and were returned to culture dishes. P. falciparum antigens 1. SIRC (sonicated, infected red cells) : 1 ml packed infected RBC were suspended in 5.5ml BBS and sonicated at 100 W for one min in an AMC ultrasonic vibrator (England). The sonicate was centrifuged to remove insoluble material and the supernatant was used as antigen. 2. Merozoite antigen (AgB): Cultures were centrifuged at 270g for 5 min to remove the RBC and were then centrifuged at 27,000g for 30 min to sediment the merozoites. The pellet, a highly enriched merozoite preparation, was washed twice, resusuended in lml BBS and sonicated for one min. AThe sonicate was cleared of insoluble debris by centrifugation. 3. Schizont antigen (AgC) : The preparation of this antigen will be described in full in a forthcoming publication (Golenser et al., in preparation). Briefly, the procedure was as follows : Infected RBC were washed once in RPM1 and once in saline. A 50”/6 suspension of RBC was prepared in 72% Percoll (Pharmacia, Uppsala). lml of this suspension was placed in a tube with lOm1 Percoll solution *This work comprises part of the Ph.D. thesis of H. Avraham.

P. falciparum:

422

ASSAY

OF

ANTIGENS

AND

ANTIBODIES

and centrifuged for one hour at 27,000g. The layer of the gradient above the bulk of the packed RBC was enriched in parasitized RBC (approximately 20% parasitaemia, largely schizonts). This layer was suspended in 10 volumes BBS, was sonicated for one min and was centrifuged to remove insoluble material. Control antigens 1. Sonicated,normal red cells (SNRC) : Suspensions

(10%) of A or 0 RBC were sonicated as described above. Sonicates were also prepared from RBC kept in culture for five days, but not infected with parasites. 2. Polymerized, sonicated RBC: This was prepared following the technique of AVRAMEAS & TERNYNCK (1969). 25ml of normal human serum and 25 ml of sonicated normal RBC were dialysed for 48 hours at 4°C against cold saline. To the dialysed sonicate, 5 ml of 1 M acetate buffer pH 5 were added, followed by 15ml of a 2 ‘5% aqueous sohttion of glutaraldehyde. The mixture was left for three hours at room temperature and the resulting gel homogenized and washed 10 times with BBS.

1:25

1 :125

1:625 SERUM

1: 3125 DILUTION

1:15,625

1:78.125

Fig. 1. SPA binding CUIWS for two positive and one control serum are shown. Titration was carried out using 5-fold serum dilutions in 10% FCS. Tubes were coated with antigen B (sonicated merozoites).

Absorption of sera

Serum samples were added to equal volumes of washed packed, normal RBC and incubated for 30 min at 37°C followed by 24 hours at 4%. This absorption was repeated when indicated. Absorptions with polymerized RBC sonicate were carried out in the same manner, but the incubation at 37°C was carried out for three hours. Assay technique

This has been described in detail in a previous publication (AVRAHAM et al., 1980). Briefly, antibodies were assayed by their capacity to bind to antigen-coated tubes. The bound antibodies were detected by treating the tubes with Protein A (SPA, Pharmacia), radio-iodinated as instructed by ZELTZER & SEEGER (1977). Antigens were assayed by their ability to inhibit binding of antibodies to the antigen-coated tubes. Indirect

Fluorescent Antibody

tests (IFA)

Group 0 erythrocytes were collected at the peak of infection (approximately 8% parasitaemia), when the schizont rate was relatively high. The erythrocytes were washed three times by centrifugation and diluted 1: 10 in phosphate buffered saline, after which 2~1 drops were placed on microscope slides and left to dry. Samples were treated with the test serum, followed by fluorescent, rabbit anti-human globulin (Institut Pasteur, Paris, France) diluted 1 : 20, according to the method of JOHNSON & HOLBOROW (1973). Fluorescence was estimated by examination under UV light in a Reichert Zetopan microscope using incident light illumination. Results Antibody content of sera of patients with I?. falciparum

Several sera were tested for reactivity with P. antigens by both the SPA and the IFA

falciparum

G::: 0 1

I 1:10

I

I

1:80 1:160 IFA TITRE

I -1:320

I

I

1:6&O

Fig. 2. Correlation of SPA and fluorescent antibody assays. In the SPA test, sera were used at I:25 dilution. Binding tests were performed with tubes coated with AgC. Initials refer to names of serum donors.

tests. Fig. 1 shows‘dose response curves obtained with strongly reactive sera in binding tests. Significant binding was obtained up to a dilution of 1: 78,125 (highest dilution tested). There was a high degree of correlation between the SPA and IFA titres (Fig. 2), but the radioactive test was about x 100 fold more sensitive (titres of 1: 78,000 and 1: 640, respectively). Specificity

controls

The specificity of the binding of patients’ sera to P. falciparum antigens was confirmed by three types of experiments: (a) absorption with non-infected, human RBC; (b) binding to tubes coated with sonicated, non-infected RBC and (c) inhibition of binding by P. falciparum antigens.

H.

Table I-Effect antigens

of absorption

et al.

AVRAHAM

of sera with NRBC,

on their

423 reactivity

with P. falciparum

and SNRC

Percentage Binding

--

Serum absorbed with Non-absorbed serum

Pooled RBC x 1

Pooled RBC x 2

Pooled RBC x 2 + Polymerized RBC

4.3 5.0

4.0 4.5

3.9 4.5

;:;

Ag B SNRC

35.9 3.0

31.1 2.8

30.0 2.5

28.0 2.7

P.S. 30

Ag B SNRC

38.2 6.0

36.0 5-O

33.9 4.2

30.5 4.0

S.S.

AgB SNRC

22.5 7.5

20.6 5.6

18.3 4.2

15.5 4.0

SNRC & B

4.8 15.7

3.9 9.0

2.5 7.0

EE:

Serum

Antigen

NHS

&B SNRC

E.Z.

P.S. 5

Note: Sera were absorbed once or twice with pooled A and 0 RBC or with pooled RBC followed by polymerized RBC to remove antibodies reacting with normal RBC components. Tubes were coated with AgB (sonicated merozoites) or with sonicated, normal RBC. Sera were used at 1: 25 dilution.

Table II-Binding of anti-P. falciparum antibodies normal RBC or with P. falciparum antigens Percentage Serum NHS K* 30 is: 5 P.S. 15

SIRC

A@

binding A&

to tubes

coated

with

different

preparations

of

to tubes coated with: Sonicated A

Sonicated 0

SNRC

SCRC

4.5

4.2

2.8

3.0

3.1

3.0

3.3

25.8 26.3 15.9 14.6

25.0 13.7 14.0

29.8 33.0 17.0 16.0

1.5 ;:; 7.8

2.0 7.5 7.8 10

7.0 2.3

;:;

;::

;:3”

Note: Binding tests were performed with tubes coated with plasmodial (SIRC, AgB, AgC) or normal RBC antigens (Sonicated A, Sonicated 0). SNRC was a pool of sonicated A and 0 RBC. SCRC was a sonicate of normal RBC kept in culture for 5 days, under the same conditions as the infected RBC.

Absorption experiments Sera were absorbed with pooled RBC (one or two absorptions) or with pooled RBC followed by polymerized, normal RBC. Absorbed and nonabsorbed sera were then tested for ability to bind to tubes coated with AgB or SNRC (Table I). All sera showed at least some reduction in their binding capacity following absorption, and this was taken as evidence for the presence of autoantibodies to RBC. This was further confirmed by the results of binding tests with tubes coated with SNRC. However, absorbed sera which had lost all the reactivity with SNRC still contained antibodies binding strongly to tubes coated with the plasmodial antigens. In view of these results, all further experiments were carried out with absorbed sera.

Binding of absorbed sera to normal RBC and to P. falciparum antigens To rule out as completely as possible any likelihood of erroneous results being obtained because of residual antibodies against RBC antigens, experiments were set up with tubes coated with several preparations derived from infected and non-infected RBC. These included: sonicated 0 and A RBC, mixtures of 0 and A sonicates and also a mixture of 0 and A sonicates prepared from normal RBC which had been kept in culture for five days, under the same conditions as the P. falciparum-infected RBC. Binding to the normal antigens never exceeded 8% (Table II). Binding of the strongly positive sera to the plasmodial antigens ranged between

P.

424

4

f&iparUm: ASSAYOF ANTIGENS AND ANTIBODIES

o----o o-----o

SNRC-0 $NRC-A

o-*-o A-*

SIRC AgEl

t

10-6 1o-7

10-5 10-6

10-L 1o-5

1o-3 lo-L

lo-* lOA

10-l lo-*

100 10"

10' 100

lo-"

10-5

10-L

1o-3

ml-*

10"

100

10'

lo* SNRC(‘lo) 10’ SIRC(% parasitaemia) IO* AgB (~g/ml)

Fig. 3. Inhibition of binding of anti- P. falciparum antibodiesby variousantigens. A posidveserumwaspreincubatedwith plasmodialantigens(SIRC or AgB) or normalantigens(sonicated0 or A RBC) beforeperformingthe binding tests,usingtubescoatedwith AgB.

25 and 33%, clearly indicating specific for the parasite.

that the reaction was

Inhibition of binding by P. falciparum antigens Inhibition tests were set up as described in the methods section, by mixing aliquots of a positive serum with sonicates of normal or infected RBC or with sonicated merozoite antigen (AgB). The normal RBC had been kept in culture for% days before sonication. The results are seen in Fie. 3. The noninfected RBC had no inhibitory caiacity whatsoever. Sonicates of infected RBC inhibited up to a dilution of 1:lOOO. Since the non-diluted antigen had been adjusted to 1% parasitaemia, the test could detect at least O*OOl% parasitaemia. Antigen B inhibited binding up to a concentration of 0. I pglml. These experiments provided fmal proof of the specificity of the test and information regarding its sensitivity for the detection of parasites in the blood. Discussion The SPA radioimmunoassay described in this work would appear to be quite useful for the investigation of P. falciparum antigens and antibodies. The&say is reliable, reproducible, simple to perform. suecific and highlv sensitive. The reagents needkd ire easily obtaynable, the antigen from parasites grown in vitro and the antibodies from patients

with malaria. However, the sera have to be thoroughly absorbed with RBC, to remove autoantibodies-td RBC which may cause erroneous results. Practicallv all sera tested (including one from a nonAfrican datient, who hah been i:fected during a temporary residence in Africa) had auto-antibociies to RBC. These findinns confirm urevious observations by ROSENBERGet al. (19f3) who detected IgM antibodies to normal RBC in patients with P. falciparum, using the fluorescent antibody test. Some of the sera gave very high binding titresup to 1: 78,000. Although titres in the fluorescent antibody tests were much lower, there was a remarkable correlation between the two tests (Fig. 2). It should be noted that the SPA and fluorescent antibody tests were performed by different workers, at different times. Three antigenic preparations were used in these preliminary experiments (sonicated RBC, sonicated merozoites and sonicated schizonts), and all three appeared to be effective, both fo;-preparing the solid Dhase adsorbent and as inhibitors in the inhibiiion tests. Although further work may reveal that use of more highly purified antigens would improve the results, it is worth stressing that simple so&cation of the infected RBC yields apreparation which is highly effective for the detection of the parasite. In fact, a parasitaemia of at least 0.001% was easily detectable.

H.

AVFtAHAJd

It has been repeatedly pointed out in the past that the ideal immunodiagnostic test for malaria should be based on the detection of antigen in the blood. This reauirement is dictated bv the fact that. in endemic areas, practically all inhabitants have antibodies against the parasite and, consequently, antibody titres cannot be used as indicators of infection. The experiments detailed in this communication show that, in principle, assays for the detection of antigen in the blood can be developed. Further work is needed, however, to determine whether this tvne of antibodv-inhibition techniaue can be employed as the basii for the development of a test (radio-isotope or ELISA-based) to be used in the immunodiagnosis of malaria. Acknowledgements

The authors are indebted to Dr. A. I. 0. Williams Dept. of Chemical Pathology, University College Hospital, Ibadan, Nigeria and to Dr. S. Baruch, Soroca Hospital, Beer Sheva, for their gifts of sera of patients with P. falciparum. These investigations were supported by the Lautenberg Endowment Fund and by the Kuvin Center for Study of Infectious and Tropical Diseases. References

Avraham, H., Spira, D. T., Gorsky,Y. & Sulitzeanu, D. (1980). A solid-phase antibody bindinginhibition test, for the assayof Plasmodium berghei antigen and antibodies, using radioiodinated protein A. Journal of Immunological Methods, 32, 151-155. Avrameas, S. & Ternynck, T. (1969). The crosslinking of proteins with glutaraldehyde and its use for the preparation of immuno-absorbents. Immunochemistry,

6, 53-‘6.

et al.

425

Barrett, M. J. (1977). Covalently bound biological substances to plastic materials and use in r&oassav. US natent No. 4.001.583. Homewood, &A. (1976). A comparison of methods used for the removal of white cells from malariainfected blood. Annals of Tropical Medicine and Parasitology, 70, 249-251. Hunter, K. W., Finkelman, F.D., Strickland, G. T., Sayles, F. C. & Scher, I. (1979). Defective resistance to Plasmodit& yoelii in CBA/N mice. 7ournal of Immunologv. 123. 133-142. Jo”hnson, 6. D. & Holl%ow, E. J. (1973). Immunofluorescence. In: Handbook of Experimental Immunology. D. M. Weir (Editor). Oxford: Blackwell Scientific Publications, p. 181. Rosenberg, E. G., Strickland, G. T., Yang, S. L. & Whalen, G. (1973). IgM autoantibodies to red cells and autoimmune anemia in patients with malaria. American Journal of Tropical Medicine and Hygiene, 22, 146-152. Trager, W. & Jensen, J. B. (1976). Human malaria parasites in continuous culture. Science, 193,674675. Voller, A., Bidwell, D. E., Bartlett, A., & Edwards, R. (1977). A comparison of isotopic and enzymeimmune assays for tropical parasitic diseases. Transactions of the Royal Society of Tropical Medicine amd Hygiene, 71, 431-437.

Zeltzer, I?. M. & Seeger, R. C. (1977). Microassay using radioiodinated protein A from Staphylococcus aureus for antibodies bound to cell surface antigens of adherent tumor cells. Journal of Immunological

Methods, 17, 163-175.

Accepted for publication

20th August, 1980.