JOURNAL OF CLINICAL MICROBIOLOGY, June 1991, p. 1236-1242 Vol. 29, No. 6
0095-1137/91/061236-07$02.00/0
Copyright C 1991, American Society for Microbiology
Detection of Antigens and Antibodies in the Urine of Humans with
Plasmodium falciparum Malaria
MARINA RODRIGUEZ-DEL VALLE,1 ISABELLA A. QUAKYI,12 JOHN AMUESI,2 JOHN T. QUAYE,2
FRANCIS K. NKRUMAH,2 AND DIANE W. TAYLOR'*
Department of Biology, Georgetown University 37th and 0 Streets, N. W., Washington, D.C. 20057,1 and
Departments of Haematology and Child Health, University of Ghana Medical School, Accra, Ghana2
Received 19 October 1990/Accepted 22 March 1991
Humans infected with Plasmodium falciparum frequently have elevated levels of proteins in their urine, but
it is unclear if any of these proteins are parasite antigens or antimalarial antibodies. To resolve this question,
urine samples from malaria patients and controls livihg in Thailand and Ghana were evaluated. Urine samples
from 85% of the patients had elevated protein levels and contained proteins with Mrs ranging from <29,000
to >224,000 as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Antisera were
produced against urine from infected and control subjects. Antisera raised against infected, but not control,
urine were positive by indirect immunofluorescence on P. falciparum parasites and immunoprecipitated
approximately 12 unique bands from extracts of parasites metabolically labeled with 35S-methionine. These
data suggest that a variety of P. falciparum antigens are released into urine during acute infection. It is also
likely that anti-P. falciparum antibodies are present in the urine of malaria patients because samples from these
patients, but not controls, were positive in indirect immunofluorescence assays and immunoprecipitated at least
19 P. falciparum antigens from extracts of metabolically labeled parasites. The detection of malarial antigens
and antibodies in urine may lead to a new approach for the diagnosis of malaria.
Individuals with fever commonly have elevated levels of MATERIALS AND METHODS
protein in their urine (8, 10, 16, 25). It is therefore not
surprising that proteinuria has been reported in patients Urine samples. Urine samples were collected in the Chant-
infected with Plasmodium falciparum, a disease character- aburi region of Thailand from adults with P. falciparum
ized by cyclical fevers (3, 7, 17, 19). Recently, Ehrich and infections. Clinical histories were taken by the attending
Horstmann examined urine samples from seven malaria physicians, and parasitemias were confirmed by blood
patients by sodium dodecyl sulfate-polyacrylamide gel elec- smears. Clean-catch urine samples were then obtained from
trophoresis (SDS-PAGE) and found that P. falciparum pa- the patients before the administration of antimalarial chemo-
tients excreted a variety of proteins (5). It is unknown, therapy. Urine samples were transferred to sterile test tubes,
however, if any of these proteins are parasite antigens or immediately frozen at -20°C, and shipped to Georgetown
potentially antimalarial antibodies. University on dry ice. A total of 30 urine samples from 10
Antigens and/or antibodies to a variety of pathogens female and 20 male patients were obtained. The patients had
including Leishmania donovani (14), Trypanosoma cruzi a mean age of 32 years (standard deviation, 15 years) and a
(12), Mycobacterium leprae (11), and human immunodefi- mean parasitemia of 1.1%, with counts ranging from 0.014 to
ciency virus type 1 (HIV-1) (4) have been found in urine. The 5.4%. Many of the subjects reported having had malaria
identification of malarial antigens in urine would have prac- multiple times. For comparison, 19 negative control urine
tical benefits. First, there may be a correlation between the samples were obtained from Thai adults residing in Bangkok
number and molecular weights of malarial antigens released where P. falciparum is not transmitted. Samples and clinical
into urine and the level of renal pathology present in the data were obtained in 1988 by H. Wilde through the DiaTech
host. Such a relationship would be of clinical value. Sec- Program, Program for Appropriate Technology and Health
ondly, the detection of malarial antigens or antibodies in (P.A.T.H.), Seattle, Wash.
urine could provide new approaches for the diagnosis of Urine samples were also collected from 16 children resid-
malaria. ing in Accra, Ghana, who were infected with P. falciparum.
In this study, we sought to determine if P. falciparum Midstream urine samples were obtained whenever possible
antigens and/or antibodies were excreted into human urine. from 10 females and 6 males, with a mean age of 5 years
To identify malarial antigens, mice were immunized with (standard deviation, 2.5 years) and a mean parasitemia of
urine from infected and normal human subjects. The result- 2.4% with counts ranging from 0.3 to 7.3%. After being
ing antisera were used in immunoblotting, immunoprecipita- collected, urine samples were centrifuged, the fluid phase
tion, and immunofluorescence assays for antigen character- was collected, and 0.1% sodium azide was added prior to
ization. At least 12 proteins of parasite origin were detected freezing. All samples were kept frozen during transport and
in the urine of infected patients. Immunofluorescence and maintained at -20°C until used. In addition, midstream urine
immunoprecipitation studies also demonstrated the presence samples were collected from adolescents who had Schisto-
of antimalarial antibodies to a variety of malarial antigens in soma haematobium infections but were negative by blood
the urine of patients with acute P. falciparum malaria. smear for malaria. Urine from patients with S. haematobium
was selected as a negative control because these patients
have proteinuria due to a parasitic infection. Ghanaian
samples were collected in 1983 as part of the Coordinated
* Corresponding author. Research Programme on Nuclear Techniques for the Detec-
1236
Downloaded from http://jcm.asm.org/ on October 4, 2018 by guest
VOL. 29, 1991 P. FALCIPARUM ANTIGENS AND ANTIBODIES IN URINE 1237
tion of Parasite Antigens in Body Fluids sponsored by the standards were purchased from Bethesda Research Labora-
International Atomic Energy Agency. tories (Bethesda, Md.).
Several additional negative control urine samples were Immunoblot analysis. The transfer of proteins was per-
studied. These included 10 clean-catch urine samples from formed by using the method of Towbin et al. (23). In brief,
healthy Americans and a 24-h sample from an American with proteins were transferred onto nitrocellulose sheets
multiple myeloma containing Bence-Jones proteins. These (0.45-,um pore size; Bio-Rad Laboratories) by using a mini
urine samples were frozen and treated in the same manner as Trans-blot Electrophoresis Transfer Cell (Bio-Rad) and
those from Thai adults. following the directions of the manufacturer. After transfer,
Pooled urine samples were used in many of the studies. To nitrocellulose papers were soaked in blocking solution
prepare these, equal volumes of urine from the 30 Thai (0.3% bovine serum albumin and 0.3% Tween-20 in 0.1 M
patients were combined; the pool is referred to as pooled (+) PBS [pH 7.4]) overnight. The blots were then incubated with
Thai urine. A second pool was prepared by combining urine a 1:500 dilution of sera from immunized mice in blocking
samples from 16 Ghanaian malaria patients and is referred to solution diluted 1:100 with 0.1 M PBS (pH 7.4) for 4 h. Blots
as pooled (+) Ghanaian urine. A similar pool was prepared were then washed three times with 0.05% Tween-20 in
from uninfected Thai, Ghanaian, and American controls and 0.1 M PBS (pH 7.4) and incubated for 2 h with a 1:500
is referred to as pooled control urine. dilution of horseradish peroxidase-labeled goat anti-mouse
Urinalysis. Individual urine samples were thawed, and polyvalent immunoglobulin (Sigma) in blocking solution.
urinalysis was performed by using Multistix 10 SG reagent After three additional washes, the blots were developed by
strips (Miles Diagnostics, Elkhart, Ind.). The analysis in- using the 4-Cl-1-naphthol substrate (60 mg of horseradish
cluded measurements of glucose, bilirubin, ketones, specific peroxidase color development reagent [Bio-Rad] in 20 ml of
gravity, blood, pH, protein, urobilinogen, nitrites, and leu- methanol combined with 100 ml of 0.015% H202 in 0.1 M
kocytes. PBS [pH 7.4]). All incubations and washes were performed
Preparation of antisera. Female BALB/c mice, 6 to 10 at 4°C, with samples being agitated at the lowest speed
weeks old (Harlan Sprague-Dawley, Frederick, Md.) were possible on a clinical rotator (Thomas Scientific, Swedes-
immunized three times with (i) pooled (+) Thai urine, (ii) boro, N.J.).
pooled (+) Ghanaian urine, or (iii) pooled control urine. In Metabolic labeling of parasites. Asynchronous cultures of
the first immunization, mice were injected with 200 ,ul of P. falciparum NF54 parasites were maintained in vitro as
urine emulsified in Freund complete adjuvant (Calbiochem, described by Trager and Jensen (22). For metabolical label-
La Jolla, Calif.) in a single subcutaneous site. Approximately ing studies, erythrocytes which had been kept in the refrig-
14 days later, they were injected intraperitoneally with 200 erator for 3 weeks were used to ensure that leukocytes were
RI of urine in Freund incomplete adjuvant. The third boost no longer viable. Since mature erythrocytes do not incorpo-
consisted of 200 RI of urine (without adjuvant) injected rate amino acids, [35S]methionine ([35S]Met)-labeled pro-
intraperitoneally. Fourteen to 30 days later, the mice were teins in this study were synthesized by the parasite.
anesthetized with sodium pentobarbital and blood was col- For metabolic-labeling studies, parasites were cultured in
lected directly from the heart. Sera were stored at -20°C media deficient in methionine by using Selectamine RPMI-
until used. 1640 (Gibco Laboratories, Grand Island, N.Y.). Medium
IFA assay. Sera from immunized mice and unconcentrated was supplemented with 10% normal human A' serum, 30
urine samples from patients with P. falciparum were tested mM HEPES (N-2-hydroxyethylpiperazine-N'-2-ethane-
for the presence of antimalarial antibodies by indirect immu- sulfonic acid), 0.27% sodium bicarbonate, and 10 ,ug of
nofluorescence antibody (IFA) analysis. Routine in vitro hypoxanthine (Sigma) per liter. [35S]Met (185 MBq) (ICN
cultures of P. falciparum parasites of the NF54 (Netherlands Biomedicals Inc., Irvine, Calif.) was added at a concentra-
Airport strain) and Malayan Camp (MC) strains were main- tion of 100 ,uCi/ml. Cultures were gassed with 3% 02-6%
tained by using the method described by Trager and Jensen C02-91% N2 and maintained in 75-cm2 flasks at a 5%
(22). Blood smears of cultured parasites were prepared, air hematocrit. The cells were cultured with the radioactive
dried, and fixed in acetone. The IFA method of Voller was amino acid for 12 h. Final parasitemias were between 7 and
followed (24). In brief, serum samples were diluted in 0.1 M 10%.
phosphate-buffered saline (PBS) (pH 7.4) and applied to the After metabolic labeling, infected erythrocytes were lysed
parasite smears for 20 min at room temperature. Smears with buffer containing 1% Triton X-100 in 0.1 M PBS (pH
were then washed in three changes of 0.1 M PBS (pH 7.4), 7.4-1 mM iodoacetamide (Sigma)-l mM phenylmethylsul-
and a 1:40 dilution of fluorescein isothiocyanate-conjugated fonyl fluoride (Sigma) for 15 min at 4°C and then centrifuged
goat anti-mouse immunoglobulin (Sigma Chemical Co., St. at 1,125 x g for 10 min. The Triton X-100-soluble material
Louis, Mo.) was applied for 20 min at room temperature. was used in the immune precipitations studies described
Smears were washed in three changes of 0.1 M PBS (pH below.
7.4), mounted in barbital-glycerol buffer (0.1 M sodium Immune precipitation. Immune precipitations were per-
barbital, 50% glycerol [pH 8.6]) and examined by use of a formed by use of the method suggested by Kessler (13) with
Laborlux epifluorescence microscope (Leitz, Wetzlar, modifications. To prevent nonspecific binding of malarial
Germany) equipped with a band pass filter (450 to 490 nm). proteins to protein A, 1 ml of the Triton X-100-soluble
SDS-PAGE. Proteins in pooled urine samples were sepa- extract was incubated with 0.5 ml of a 50% slurry of protein
rated under nonreducing conditions by SDS-PAGE by using A-Sepharose (Pharmacia, Piscataway, N.J.) for 30 min at
the method described by Laemmli (15). In brief, urine room temperature. Aliquots (50 RI) of the preadsorbed
samples were diluted 1:1 in nonreducing sample buffer, extract were then incubated with (i) 10 RI of sera from mice
incubated at 95°C for 4 min, and then separated on an immunized with control human urine, (ii) 10 [lI of sera from
SDS-10% polyacrylamide slab gel by using a Mini-protein II mice immunized with pooled (+) Ghanaian urine, (iii) 50 [lI
Electrophoresis Cell (Bio-Rad Laboratories, Richmond, of pooled control human urine, or (iv) 50 RI of pooled (+)
Calif.). Proteins were stained by using the silver stain Ghanaian urine for 16 h at 4°C. The aliquots were then
method of Merril et al. (18). Prestained molecular-weight incubated with 75 RI of a 50% slurry of protein A-Sepharose
Downloaded from http://jcm.asm.org/ on October 4, 2018 by guest
1238 RODRIGUEZ-DEL VALLE ET AL. J. CLIN. MICROBIOL.
TABLE 1. Summary of urinalysis results
No. of subjects with:
Subjects Protein' Bilirubinb Bloodb
- tr 3m0g-/1d0l0 1m0g0/-d3l00 30 -m0g-/2d,l000 + ++ +++ - +_ + ++
P. falciparum infected
Thailand (n = 30) 0 4 12 13 1 18 8 3 1 22 3 2 3
Ghana (n = 16) 1 2 6 5 2 14 1 1 0 14 1 0 1
Uninfected
Thailand (n = 19) 14 5 0 0 0 19 0 0 0 19 0 0 0
United States (n = 10) 6 4 0 0 0 10 0 0 0 10 0 0 0
Proteinuria controls
S. haematobium' (n = 8) 0 2 3 2 1 8 0 0 0 7 1 0 0
Myelomad(n = 1) 0 0 0 1 0 1 0 0 0 1 0 0 0
-, negative. The lowest level detectable is 15 to 30 mg of albumin per dl.
b Qualitative estimates expressed as negative (-), trace (+ -), positive (+), high (++), and very high (+++ +). The lowest levels detectable are as follows:
bilirubin, 0.4 to 0.8 mg/dl; blood, 0.015 to 0.062 mg of hemoglobin per dl.
Patients with S. haematobium from Ghana.
d American patient with multiple myeloma.
for 1 h at room temperature. The samples were washed once urological finding of proteinuria. There were numerous pro-
with 1% bovine serum albumin (Sigma) in NETT buffer tein bands present in urine samples of malaria patients from
(0.5% Triton X-100, 150 mM NaCI, 50 mM Tris base [pH Thailand and Ghana that were not seen in the urine of
7.4], 3 mM NaN3, 1 mM iodoacetamide, 1 mM phenylmeth- control subjects (Fig. 1, lanes 1 and 2 versus lanes 3 to 6).
ylsulfonyl fluoride), twice with NETT buffer, once with 650 The total number (approximately 15) and the Mrs of these
mM NaCl in NETT buffer, and a finally again in NETT bands (ranging from <29,000 to >224,000) were similar in
buffer. Immunoprecipitated samples were combined with 25 pooled (+) urine samples from both Thai adults and Ghana-
[LI of reducing sample buffer, heated at 95°C for 5 min, and ian children. Urine from uninfected, normal Thais (lane 3)
separated on an SDS-10% polyacrylamide gel. Gels were and Americans (lane 6) showed only a few weak bands.
incubated in Resolution (E.M. Corp., Chestnut Hill, Mass.), Urine from proteinuria-control patients showed a clear band
dried, and exposed to X-Omat AR film (Eastman Kodak Co., of Bence-Jones proteins (dimer of light chains) in the indi-
Rochester, N.Y.) for 1 to 5 days at -65°C. vidual with multiple myeloma (lane 4) and a single distinct
band at an Mr of 224,000 in subjects with S. haematobium
RESULTS
Summary of urinalysis. Elevated levels of proteins (.30
mg/dl) were found in the urine of MALARIA26 of 30 (87%) of the adult CONTROLS
patients from Thailand and 13 of 16 (81%) of the infected 1 2 3 4 5 6
children from Ghana (Table 1). Protein levels varied among
the patients, with _ -.. -the majority of specimens containing 30 to
300 mgldl. Proteinuria was not observed in the normal -224
American control subjects but was recorded for individuals
with S. haematobium infections and the adult with multiple
myeloma (100 to 300 mg/dl). Increased levels of hemoglobin - 109
(.0.045 mg/dl) were detected in the urine of only 5 of 30 and
1 of 16 malaria patients from Thailand and Ghana, respec- 1 VW1
tively (Table 1). Bilirubin, which is common in P. falciparum - 71.9
patients with severe renal complications (20), was observed
in 12 of 30 (40%) Thai and 2 of 16 (13%) Ghanaian patients.
All urine samples from control and infected individuals - 45.8
had values within the normal range for specific gravity, pH,
urobilinogen, and glucose. Bacteriuria was measured by the
detection of nitrites, and approximately 12% of the urine
samples from infected patients were found to be positive
(i.e., .0.1 mg/dl). Urine samples from the control group 28.9
were negative for bilirubin, hemoglobin, nitrites, and leuko-
cytes.
SDS-PAGE analysis of urine samples. In initial studies, 5 ,ul FIG. 1. SDS-PAGE analysis of urine samples. Pooled urine
of (i) pooled (+) Thai, (ii) pooled (+) Ghanaian, and (iii) samples were diluted 1:1 in nonreducing sample buffer, separated on
control urine samples were separated under nonreducing an SDS-10% polyacrylamide gel, and visualized by using silver
conditions by SDS-PAGE and visualized by silver staining stain. Each lane contains 5 ,ul of urine. Lanes: 1, pooled (+)
(Fig. 1). Results clearly show that the concentration of Ghanaian urine; 2, pooled (+) Thai urine; 3, pooled control urine
protein in infected urine samples was fromconsiderably greater Thailand; 4, urine from a patient with multiple myeloma; 5,
than in uninfected urine pooled control urine fromsamples, Ghanaianthus patientssupporting withthe above schistosomiasis;and 6, pooled control urine from normal Americans.
Downloaded from http://jcm.asm.org/ on October 4, 2018 by guest
VOL. 29, 1991 P. FALCIPARUM ANTIGENS AND ANTIBODIES IN URINE 1239
AN T I A N T I AN T I -
(+) TH A (+)G H ANA CONTROL
U R I NE U R I N E URINE
1 2 3 4 1 2 3 4 4 3
200 "M q*pm ..i mo.... ...
...
9 7- z. g
**~~~N
"i .. a
4 S - gif
FIG. 3. Immunoblot analysis of pooled urine samples. Urine
samples were separated by SDS-PAGE and transferred to nitrocel-
lulose. Urine samples were from Thai patients (lane 1), uninfected
FIG. 2. Representative patterns of IFA produced by sera from Thais (lane 2), Ghanaian patients (lane 3), and Ghanaians who were
mice immunized with urine from patients with P. falciparum ma- malaria negative but had schistosomiasis (lane 4). Blots were treated
laria. Acetone-fixed P. falciparum blood smears were treated with with sera from mice immunized with pooled (+) Thai urine, pooled
sera from mice immunized with pooled (+) Ghanaian urine (anti- (+) Ghanaian urine, or pooled control urine. Dotted lines denote
body titer, 1:500). Similar patterns were produced by using sera proteins normally found in all urine. Some of the other bands may be
from mice immunized with pooled (+) Thai urine. Sera from mice of parasitic origin.
immunized with pooled control urine were routinely negative at 1:50
dilution.
of both groups of patients. Sera from mice immunized with
control human urine recognized approximately five bands in
infections (lane 5). Thus, SDS-PAGE results demonstrate urine samples from patients with malaria and schistosomia-
both a high concentration of protein and unique protein sis (represented by dotted lines). On the basis of the Mrs,
bands in pooled urine samples from malaria-infected individ- some of these bands most likely are albumin, transferrin, and
uals. immunoglobulin.
Detection of malarial antigens in urine. Antisera from mice To determine if the unique bands seen above were de-
immunized with pooled (+) Thai and pooled (+) Ghanaian tected in the urine of individual malarial patients, urine
urine were tested by IFA analysis (Fig. 2). Results showed samples from 21 infected and 7 control Thai adults were
that sera from these mice, but not from mice immunized with assessed by immunoblotting. The results are summarized in
normal urine, reacted specifically with P. falciparum para- Table 2. In total, 15 bands were observed in the 21 patients
sites (Fig. 2). Thus, mice produced antibodies to malarial and 4 bands were observed in the 7 controls. For simplicity,
antigens present in pooled (+) Thai and pooled (+) Ghanaian
urine samples.
In general, the antisera produced several different patterns TABLE 2. Detection of antigens in individual urine samples
of fluorescence depending on the developmental stage of the
asexual parasite (Fig. 2). Staining was seen as diffuse fluo- % of urine samples froma:
in infected but not normal erythrocytes (Fig. 2A), Antigen band Approx Mrrescence falciparum- Noninfected(103) P.
associated with the parasite itself (Fig. 2B and C), in packets
designation infected patients controls
=
in the cytosol of infected erythrocytes (Fig. 2D), and (n
= 21) (n 7)
sur-
rounding freed merozoites (data not shown). IFA patterns A 250 100 18
were similar when sera from mice immunized with pooled B 224 44 0
(+) Thai or (+) Ghanaian urine were used and when the two C 210 78 0
different strains of P. falciparum parasites were used. Sera D 175 62 0
from mice immunized with control human urine were rou- E 155 80 00
tinely negative for parasite reactivity. F 112 48G 103 100 0
When the above-described antisera were used, unique H 90 48 0
antigens were detected by immunoblot analysis in the urine I 83 100 96
of patients with malaria from Thailand and Ghana. As can be J 70 10 0
seen in Fig. 3, lanes 1 and 3, approximately 9 to 12 bands, K 67 100 100
with Mrs ranging from 20,000 to 200,000, were observed in L 59 81 0
the urine from infected patients, many of which were not M 40 71 50
present in the matched control urine samples (lanes 2 and 4). N 35 38 0
Similar results 32 100 0were obtained when antisera raised against 0
pooled (+) Thai and pooled (+) Ghanaian urine were uti- a Percentages of the urine samples in which the antigen was detected by
lized, indicating the presence of similar proteins in the urine immunoblotting.
Downloaded from http://jcm.asm.org/ on October 4, 2018 by guest
1240 RODRIGUEZ-DEL VALLE ET AL. J. CLIN. MICROBIOL.
0 SERU M Li R IN E
1 2 3 4
2 2 4 .-- 2 2 4
"k s- .
a,S.
. am
1 09 -- -109
4mW -.1111,
E-"a - 7
71.9 1,9-
......
- 45.8
45,8
28,9-i_ _ . .1 28,9
FIG. 4. Immune precipitation of proteins from P. falciparum
parasites metabolically labeled with [35S]Met. (A) Parasite antigens
were immunoprecipitated by using sera from mice immunized with
pooled control urine (lane 1) and pooled (+) Ghanaian urine (lane 2).
(B) Parasite antigens were immunoprecipitated by using pooled (+)
urine from Ghanaian patients (lane 3) and pooled control urine (lane
4). Dotted lines designate antigens that may appear in urine in the
form of immune complexes.
FIG. 5. Indirect immunofluorescence patterns produced by urine
the bands have been assigned letters (Table 2). Bands A, I, from P. falciparum patients. Acetone-fixed P. falciparum blood
K, and M appear to represent normal proteins commonly smears were treated with 5-,u urine samples from malaria patients.
found in urine, whereas the remaining 11 bands were de-
tected only in samples from malaria patients. Table 2 shows
that many of the unique bands were frequently found in urine reacted with normal (uninfected) erythrocytes were detected
samples (i.e., bands B, C, D, E, L, M, and 0). The results in only 3 of 26 patients (Fig. 5C).
suggest that some of the unique bands may be of parasitic To determine the Mr of the parasite antigens recognized by
origin and that they are routinely excreted during acute P. antibodies in urine, immune precipitation studies were per-
falciparum infection. formed by using pooled (+) Ghanaian urine as a source of
Identification of parasite proteins. To determine whether antibodies. The results are shown in Fig. 4B. Approximately
any of the unique bands detected were of parasite origin, 20 [35S]Met-labeled protein bands, with Mrs ranging from
antisera to pooled (+) Ghanaian and control urine were used <25,000 to 160,000, were identified (lane 3). Urine from
to immunoprecipitate antigens from extracts of P. falci- negative controls precipitated only one band, which had an
parum parasites metabolically labeled with [35S]Met. The Mr of 74,000. Thus, it appears that antibodies that reacted
results are shown in Fig. 4A. At least 12 unique proteins with at least 19 different malarial polypeptides could be
were precipitated by the antisera of mice immunized with detected in human urine. A comparison of the immunopre-
urine from malaria patients (lane 2). The antigens had cipitation results shown in Fig. 4A (lane 2) and 4B (lane 3)
approximate Mrs of 215,000, 187,000, 179,000, 139,000, suggests that some antigens (indicated by dotted lines) and
103,000, 62,000, 55,000, 52,000, 40,000 (faint), 34,000, 30,000 their corresponding antibodies may both be excreted into
and 25,000. Only the band of Mr 74,000 was precipitated urine during acute malaria infections.
from the extracts by sera from mice immunized with control
urine (lane 1).
Detection of antimalarial antibodies in urine and identifica- DISCUSSION
tion of their corresponding malarial antigens. To determine if Data from previous studies strongly suggested that malar-
antimalarial antibodies were present in the urine of patients ial antigens and possibly antibodies might be released into
with P. falciparum malaria, urine samples from 20 Thai and urine during acute infection. Immunohistological studies of
6 Ghanaian patients and from 3 controls were tested by IFA. renal tissue from patients with P. falciparum had revealed
Antibodies were detected in all 26 patients; the 3 controls the presence of immune complexes, including C3, antigen,
were negative. Three different patterns of fluorescence were and antibody lining the glomeruli (2, 6, 20) and intact
observed: (i) fluorescence restricted to the parasite itself parasitized erythrocytes sequestered in the nephrons (21).
(Fig. 5A to C), (ii) diffuse fluorescence associated with Occasionally a few intact erythrocytes were seen in the urine
infected but not normal erythrocytes (Fig. SD), and (iii) as of patients (1, 7, 20). These results, along with the docu-
packets of fluorescence (Fig. SE and F). Antibodies that mented reports of proteinuria (3, 7, 17, 19), made it likely
Downloaded from http://jcm.asm.org/ on October 4, 2018 by guest
VOL. 29, 1991 P. FALCIPARUM ANTIGENS AND ANTIBODIES IN URINE 1241
that parasite antigens and antibodies would be excreted into pigment-laden macrophages have been found in the glomer-
urine. ular capillary loop during uncomplicated malaria (20). These
Results from the current study support this conclusion. changes are not associated with severe renal failure and are
Elevated levels of proteins (>30 mg/dl) were present in the thought to be responsible for the mild proteinuria observed
urine of 84% of the patients studied (Table 1). Protein levels during acute infection (3, 7, 17, 19). One therefore wondered
in these patients were approximately equivalent to those if antigens or antibodies would be released when renal
observed in proteinuria controls, namely individuals with pathology was minimal. In the patients included in this
chronic S. haematobium infections and multiple myeloma. study,
When mice were immunized with pooled urine from malaria parasitemias ranged from 0.014 to 5.4% in Thai adults
patients, they produced antibodies that reacted with P. and from 0.3 to 7.3% in Ghanaian children. Serum blood
falciparum parasites in the IFA assay. Mice immunized with urea nitrogen and creatinine data demonstrated that severe
pooled urine from control subjects failed to produce antibod- kidney complications were not present in these patients
ies that reacted with malarial parasites. These data suggest (data not shown). The finding of both high- (>200,000) and
that malarial antigens are present in the urine of P. falci- low-molecular-weight proteins in urine by SDS-PAGE dem-
parum patients. onstrates glomerulotubular involvement (5) but does not
Additional studies showed that antisera from mice immu- necessarily indicate severe renal pathology. Thus, it appears
nized with pooled (+) urine immunoprecipitated a set of that renal dysfunction was not present in the majority of
[355]Met-labeled P. falciparum proteins. Since only malarial patients studied and that it is not a prerequisite for the
proteins are labeled in in vitro cultures, the bands shown in release of P. falciparum antigens or antibodies into urine. In
Fig. 4 (lane 2) represent proteins synthesized by P. falci- another infection where polyclonal activation is also domi-
parum parasites. A single malarial protein was immunopre- nant, antibody to HIV-1 has been demonstrated in the urine
cipitated by using antisera from mice immunized with a pool of patients with HIV-1 infections in the absence of demon-
of control urine (Fig. 2, lane 1). This band may be the result strable proteinuria (4).
of nonspecific binding or may represent a cross-reactive Identification of parasite proteins and antimalarial anti-
epitope between malarial parasites and an altered host bodies in urine suggests that a urine-based assay for diagno-
protein present in urine. On the basis of these results, it sis of malaria may be feasible. An assay that detects antigen
seems logical to conclude that malarial antigens are released or antibody in urine would be noninvasive and thus would
into urine during acute infection and that most of the unique prevent direct exposure of health-care workers to blood. It is
bands depicted in Fig. 4 (lane 2) represent the molecular not known how long malarial antigens or antibodies are
weights of these antigens within the parasite itself. Undoubt- excreted into urine, but the observation that is
edly, the numbers and molecular weights of these antigens proteinuriatransient and rapidly returns to normal following chemother-
could differ in urine. apy (5) suggests that there maybe a direct correlation be-
Clearly, the antisera used in immunoblotting studies react tween
with both malaria- and host-derived antigens (Fig. 2 and
the presence of malarial antigens or antibodies and
Table 2). The antisera to control urine detected a set of acute infection. Recently, Kohanteb et al. (14) reported
bands in the urine of both malaria and control subjects. finding antigens and antibodies in concentrated urine sam-
Conversely, anti-pooled (+) Thai and anti-pooled (+) Gha- ples of patients with L. donovani. In this case, antileishma-
naian antisera reacted with proteins in control and infected nia antibodies persisted longer in urine than parasite antigens
urine. The common reactivities are shown by dotted lines in following chemotherapy. Similar studies are needed in P.
Fig. 3. However when anti-pooled (+) urine antiserum was falciparum patients. An additional important consideration
used, additional reactivities were observed in the urine of in a urine-based diagnostic approach is the quantity of
malaria patients that are absent in the controls. Because antigen or antibody present. Parasite antigens have been
proteins could be diluted out in a "pool," individual urine detected in the urine of patients with T. cruzi (12) and M.
samples were assessed (Table 2). As expected, several of the leprae (11), but the urine was concentrated 25 and 100 times,
antigens were found in most urine samples, but others were respectively. In the current study, straight (unconcentrated)
detected only in the urine of malaria patients (Table 2). Some urine was used throughout. Thus, the concentration of
of the unique proteins were found in the majority of malaria malarial antigens or antibodies present in freshly collected
patients tested. Although it is possible that some of the urine samples should be sufficient for direct use in a diag-
unique bands in Table 2 represent reactivities to elevated nostic assay. These considerations along with the finding of
levels of host proteins (e.g., acute-phase proteins), data from similar antigens in the urine of children in Africa and adults
IFA and immunoprecipitation studies strongly suggest that in Southeast Asia (Fig. 3) suggest that a diagnostic assay
many are parasite proteins. based on the detection of specific antigens or antibodies
It is also likely that antimalarial antibodies of multiple should be considered.
antigenic specificities are excreted during acute P. falci-
parum infection. Urine from Ghanaian children with acute
malaria, but not from adolescents with chronic schistosomi- ACKNOWLEDGMENTS
asis, contained antibodies that reacted with P. falciparum
parasites by IFA and specifically immunoprecipitated We thank I. Gray for his helpful contributions and H. Wilde (Thai
[35S]Met-labeled parasite antigens. The finding of antibody Red Cross Society) for collecting and characterizing the urinethe excellent technical
in urine is not surprising since SDS-PAGE analysis of urine samples from Thailand and acknowledgeassistance of C. B. Evans.
from malaria patients showed that they contained large This project was supported by the DiaTech Program, P.A.T.H.
amounts of high-molecular-weight proteins in the size range Funds for collection of the urine samples from Ghana were provided
of immunoglobulin (Fig. 1). by the International Atomic Energy Agency Research (contract no.
During P. falciparum infection, both mild and severe renal 3233/RB, August 1982-July 1983) and in part by the United Nations
pathological changes have been demonstrated (1, 2, 3, 6, 9, Development Program/World Bank/World Health Organization Spe-
19, 20). Histologically, parasitized erythrocytes and malaria cial Program for Research and Training in Tropical Diseases.
Downloaded from http://jcm.asm.org/ on October 4, 2018 by guest
1242 RODRIGUEZ-DEL VALLE ET AL. J. CLIN. MICROBIOL.
REFERENCES nation test. Exp. Parasitol. 68:208-215.
1. Berger, M., L. M. Birch, and N. F. Conte. 1967. The nephrotic 13. Kessler, S. W. 1981. Use of protein A-bearing staphylococci for
syndrome secondary to acute glomerulonephritis during falcip- the immunoprecipitation and isolation of antigens from cells.
arum malaria. Ann. Intern. Med. 67:1163-1171. Methods Enzymol. 73:442-471.
2. Bhamarapravati, N., S. Boonpucknavig, V. Boonpucknavig, and 14. Kohanteb, J., S. M. Ardehali, and H. R. Rezai. 1987. Detection
C. Yaemboonruang. 1973. Glomerular changes in acute Plasmo- of Leishmania donovani soluble antigen and antibody in the
dium falciparum infection. Arch. Pathol. 96:289-293. urine of visceral leishmaniasis patients. Trans. R. Soc. Trop.
3. Boonpucknavig, V., and V. Sitprija. 1979. Renal disease in acute Med. Hyg. 81:578-580.
Plasmodiumfalciparum infection in man. Kidney Int. 16:44-52. 15. Laemmli, U. K. 1970. Cleavage of the structural proteins during
4. Cao, Y., A. E. Friedman-Kien, J. V. Chuba, M. Mirabile, and B. the assembly of the head of the bacteriophage T4. Nature
Hosein. 1988. IgG antibodies to HIV-1 in urine of HIV-1 (London) 227:680-682.
seropositive individuals. Lancet ii:831-832. 16. Marks, M. I., P. N. McLaine, and K. N. Drummond. 1970.
5. Ehrich, J. H. H., and R. D. Horstmann. 1985. Origin of Proteinuria in children with febrile illnesses. Arch. Dis. Child.
proteinuria in human malaria. Tropenmed. Parasitol. 36:39-42. 45:250-253.
6. Futrakul, P., V. Boonpucknavig, S. Boonpucknavig, C. Mitra- 17. McCabe, M. E. 1966. Malaria: a medical problem yet with us.
kul, and N. Bhamarapravati. 1974. Acute glomerulonephritis Med. Serv. J. Can. 22:313-316.
complicating Plasmodium falciparum infection. Clin. Pediatr. 18. Merril, C. R., M. L. Dunau, and D. Goldman. 1981. A rapid
13:281-283. sensitive silver stain for polypeptides in polyacrylamide gels.
7. Giglioli, G. 1962. Malaria and renal disease with special refer- Anal. Biochem. 110:201-207.
ence to the British Guiana. I. Introduction. Ann. Trop. Med. 19. Sitprija, V. 1970. Renal involvement in malaria. Trans. R. Soc.
Parasitol. 56:101-109. Trop. Med. Hyg. 64:695-699.
8. Hemmingsen, L., and P. Skaarup. 1977. Urinary excretion of ten 20. Sitprija, V. 1988. Nephropathy in falciparum malaria. Kidney
plasma proteins in patients with febrile diseases. Acta Med. Int. 34:867-877.
Scand. 201:359-364. 21. Spitz, S. 1945. The pathology of acute falciparum malaria. Mil.
9. Houba, V. 1975. Immunopathology of nephropathies associated Surg. 99:555-572.
with malaria. Bull. W.H.O. 52:199-207. 22. Trager, W., and J. B. Jensen. 1976. Human malaria parasites in
10. Jensen, H., and K. Henriksen. 1974. Proteinuria in nonrenal continuous culture. Science 193:673-675.
infectious diseases. Acta Med. Scand. 196:75-82. 23. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic
11. Kaldani, R. R. J., K. Maasho, R. Ohman, D. Reitz-Vick, S. transfer of proteins from polyacrylamide gels to nitrocellulose
Britton, and M. J. Lefford. 1987. Methods for the detection of a paper: procedure and some applications. Proc. Natl. Acad. Sci.
specific Mycobacterium leprae antigen in the urine of leprosy USA 76:4350-4354.
patients. Scand. J. Immunol. 25:37-43. 24. Voller, A. 1964. Fluorescent antibody methods and their use in
12. Katzin, A. M., A. Marcipar, H. Freilij, R. Corral, and J. F. malaria research. Bull. W.H.O. 30:343-354.
Yanovsky. 1989. Rapid determination of Trypanosoma cruzi 25. Welty, J. W. 1937. Febrile albuminuria. Am. J. Med. Sci.
urinary antigens in human chronic Chagas' disease by aggluti- 194:70-74.
Downloaded from http://jcm.asm.org/ on October 4, 2018 by guest