Journal of ImmunologicalMethods, 153 (1992) 167-172 © 1992 ElsevierSciencePublishers B.V. All rights reserved 0022-1759/92/$05.00
167
JIM 06400
A comparison of specific IgG antibody levels to the cell wall mannan of Candida albicans in normal individuals and in patients with primary antibody deficiency J.A. Faux, A.E. Agbarakwe, S.A. Misbah and H.M. Chapel Department of Immunology. ChurchillHospital, Oxford OX3 7LJ. UK (Received9 December 1991,revised received25 February 1992,accepted 3 April 1992)
An enzyme-linked immunosorbent assay (ELISA) has been developed to measure specific IgG antibody to the polysaccharide, cell wall mannan of Candida albicans (mannan). The results were expressed as arbitrary units/ml, with an inter- and intra-assay coefficient of variation of 7-11%. In establishing normal ranges we found that specific IgG to the mannan increased with age, with 18% of healthy children aged 3-10, 48% of healthy children aged 11-19 and 76% of an adult donor population having specific IgG antibody to mannan ( > 30 U / m l ) . We have compared these normal ranges, with a group of patients with primary antibody deficiency (PAD). None of the 23 patients with PAD, which included common variable immunodeficiency, lgG subclass deficiency, and selective IgA deficiency, had titres > 30 U / m l . The patients with PAD had significantly lower levels of specific IgG anti-mannan antibody (median 9 U / m i ) compared to healthy children aged 11-19 (median 26 U / m i ) or adults (median 58 U / m l ) ( p = < 0.001) but not children aged 3-10, (median 1 U / m l ) ( p = 0.08). Key words: ELISA; SpecificIgG; Cell wall mannan; Candida albicans; Primary antibody deficiency
Introduction In patients with suspected panhypogammaglobulinaemia quantitative defects of immunoglobulin synthesis can be diagnosed by measuring total serum immunoglobulin levels (WHO Report, 1989). Schur et al. (1970) reported an association between selective IgG2 subclass deficiency and a susceptibility to recurrent pyogenic infections. Since then patients with other selective subclass deficiencies have been reported (Oxelius, 1979; Stanley et al., 1984; Jefferis and
Correspoade,weto: J.A. Faux, Department of Immunology, Churchill Hospital, Oxford OX3 7LT, UK. Tel.: 0865-22993.
Kumararatne, 1990; Preud'homme and Hanson, 1990). More recently a further important defteiency has been recognized. Some patients with recurrent sinopulmonary infections have normal lgG subclass levels but are unable to mount an antibody response to bacterial polysaceharide antigens (Lane and Maelennan, 1986; Ambrosino et al., 1988; Geha, 1988). The diagnosis of this selective antibody deficiency is currently dependent on the demonstration of a poor specific lgG antibody response to Streptococcus pneumoniae and Haemophilus influenzae polysaccharide antigens (Shackleford et ai., 1987; Austrian, 1989). Another antigen that may give additional information on the ability to mount a specific immune response to polysaccharide antigens is the
168 cell wall mannan of Candida albicans. C. albicans is found as a commensal organism in the normal gastrointestinal tract, vagina and mouth with the result that nearly 50% of normal adults have been shown to have precipitating antibody to this antigen (Faux, 1968). We have developed a highly sensitive and specific ELISA for the quantitative determination of specific IgG antibody to cell wall mannan of C. albicans and compared the levels found in a normal population with a group of patients with primary antibody deficiency.
Materials and methods
Subjects In order to establish normal values for the assay, sera were collected from (i) 28 healthy children age range 3-10 (ii) 31 healthy children age range l l - 1 9 , both collected as part of a genetic linkage study for atopy, (iii) 81 adult blood donors age range 20-59, (iv) 38 elderly subjects age range 60-80 (J. Lea, in preparation) (v) 23 patients with primary antibody deficiency (PAD), which included, 14 with common variable immunodeficiency (CVI), four with IgA deftciency and five with selected IgG subclass deficiency. Antigen extracts (1) A purified extract of the cell wall mannan of Candida albicans (mannan) was provided by the Mycological Reference Laboratory, PHLS, London, and used in all the assays. (2) A cell wall mannan of C. albicans was provided by the National Institute for Biological Standards and Controls which was ased in the specificity assay (mannan NIBSC). Each preparztion was made up at 1 m g / m l in phosphate-buffered saline pH 7.4 plus 0.1% sodium azide and stored at 4°C until required. Using the Ouchterlony double diffusion test a line of identity was shown between these two extracts. Methods Microtitre plates (Dynatech) were sensitized overnight at 4°C with 100 /xl per well of a 5 p.g/ml solution of the mannan in carbonate buffer
(pH 9.6). All washings were performed four times using a washer (Flow Laboratories) and phosphate-buffered saline (pH 7.4) with 0.05% Tween 20 (PBST). The test sera were diluted 1 in 100 using PBST (100 /~1 per well in duplicate) and incubated for 1 h. All incubations were done at 21°C. After washing 100 tzl of a 1 in 10000 dilution of an alkaline phosphatase conjugated goat anti-human IgG (Fc affinity purified, ICN Immunobiologicals) were added to each well. The plates were incubated for 1 h. After washing, 100 /zl per well of p-nitrophenyiphosphate (1 mg/ml, Sigma) in diethanolamine buffer (pH 9.8) were added to each well and allowed to develop for 30 min at 210C in the dark. Stopping solution (50/~i of 3 M NaOH) were added to each well and the plates were read at 405 nm using an ELISA plate reader (Flow Laboratories)
Analysis of ELISA data Seven standards were prepared by diluting an IVlg (GammaGard, Baxter Healthcare) preparation with a negative normal serum. This was designated to contain 500 U / m l of specific lgG to the mannan. The U / m l of the unknown sera were calculated from the standard curve, plotted as l o g . U / m l versus linear OD giving an assay range of 1-200 U / m l . Two sera were routinely included in each assay for quality control, a low and a medium level. Both sera were from normal blood donors. Statistical anal-sis Groups were compared using the Mann Whitney U Test for non-paired data.
Results
Optimum antigen concentration Four sera, with differing binding capacities, were tested against dilutions of the mannan ranging from 20 to 1 tzg/ml to determine the optimum antigen concentration. Using the mean of the optical densities at different dilutions, maximum binding was found at an antigen dilution of 5/.~g/ml and this concentration was adopted for the study. Washed sensitized plates could be covered and stored a t - 20°C for 12 months before
169
use. Fresh and frozen plates were compared using the two control sera; the results for the low control were 13-18 U / m l of anti-mannan specific lgG and for the medium control, 45-65 U / m l of anti-mannan specific IgG. Same day and different day plates (nine assays) gave a coefficient of variation from 7-11%.
~ i
1.5
l.0,
~o.s
Assessmazt of antigenic specificity To test the specificity of the assay, increasing ten fold concentrations of the m a n n a n antigen diluted in PBS were preincubated with aliquots of a positive serum sample for 1 h at room temperature and then centrifuged at 1800 × g for 10 min. The absorbed sera were diluted and assayed as described. It was possible to inhibit binding completely with 10 p , g / m i of either of the mannan extracts (Fig. 1). No inhibition occurred with a m a n n a n extract purified from Saccharomyces cerecisiae at 1 m g / m l (unpublished data) an unrelated m a n n a n antigen.
Relationship between control and test sera Three adult sera were assayed using serial dilutions from 1 / 5 0 to 1/800 and compared with the standards. The resultant graphs demonstrated linearity and gave parallel slopes (Fig. 2).
~.
1.0.
o.0
, 1/50
•
,
.
11100
Standard crave
~
,
.
1i200
" , 11400
d~tioo Bdl ~
.
,
-
11800 BdZ
---4---
Bd3
Fig. 2. Comparisonof the ELISA standard curve with the sera of three adult blood donors; Bdl, Bd2 and Bd3, diluted 1/50 to I/8(X)givinggood linear and parallel slopes.
Non-specific binding to sensitized plate~ The diluent used throughout the assay for the standard curve, unknown sera and anti lgG conjugate was PBS + 0.05% Tween. We investigated whether the addition of either 1% goat serum, 1% horse serum or 1% bovine serum albumin to this assay diluent altered the standard curve or the results of five sera which contained a range of specific IgG to the mannan. All samples were assayed on the same plate. The standard curves and the assay results with each diluent were similar; with specific lgG to the m a n n a n for these five sera ranging from 1 to 5 U / m l , 4-6 U / m l , 1 0 - ! 4 U / m l , 48-54 U / m l and 150-165 U / m l respectively, with no consistent differences in the results using the different diluents.
Antibody lerels to mannan in selected groups o"
o.s
0.0
. . . . . . . . . , .1 1 Antigen
....... 10 concentration
Mannan Extract
•
, 100
ug/ml Mannan NIRSC
Fig. I. Specificity of the ELISA procedure for assaying IgG a n t i - m a n n a n a n t i b o d i e s w a s d e m o n s t r a t e d by i n h i b i t i o n o f b i n d i n g w i t h 1 0 / . t g / m l o f t h e m a n n a n p r o v i d e d by t h e N a tional Institute of Standards and Biological Control and the m a n n a n e x t r a c t u s e d in t h e assays.
In establishing a normal range we found that ther~ was a significant increase in specific IgG to mannan with age (Fig. 3). Only 18% of healthy children, aged 3-10, but 48% of healthy children aged 11-19 and 76% of healthy adults aged 20-80 had a titre greater than 30 U / m l . The median value for adults was 58 U / m l with interquartile range 30-114 U / m l . Within this adult group we found no significant differences in specific lgG to the mannan between male and female adults (Fig. 3). There was a significant increase in specific antibody between young adults (aged 20-39) and middle aged adults (40-59 years) with me-
170 dian values of 44 U / m i and 82 U / m l respectively ( p = 0.05), but the decrease in the median U / m l shown with a further increase in age (60-80 years) was not significant ( p = 0.23) (Table I). None of the patients with PAD had specific lgG of > 30 U / m l to the cell wall mannan (Fig. 4), the median value was 9 U / m l and the interquartile range 2-11 U / m l (Table I). This was significantly different from the normal adults and children aged 11-19 ( p = < 0.001) but not the younger children aged 3-10, with a median value 1 U / m l ( p = 0.08); only one of the patients studied here was in this age range.
AgeGroups (yrs) • 3-1o • 20-39 [ ] 40-59 I:] so.80 • subjects 11-~9 Adult
"~ 75 1" so
i I ~ ~ t~ 25
[] []
male
female
0 Noin 2831453638 5564 group Fig. 3. Percentage of the study group with specific lgG to
mannan > 30 U/ml showing an increase with age, but no significant differencebetween male and female adults. There were comparable numbersin each age group.
•
¢,
g,
"
; ,
-L .
°
i
1 "
i[
Discussion We report here the development of a reproducible and sensitive ELISA for the detection of specific lgG antibody to the cell wall mannan of Candida albicans. We found that 76% of the normal adult donor population but none of the patients with primary antibody deficiency (PAD) had a titre greater than 30 U / m l . This suggests that the measurement of specific lgG antibody to cell wall mannan of C. albicans may be useful in the laboratory assessment of PAD. Recent studies from the United States and Britain have highlighted the clinical significance of specific antibody deficiency to po.lysaccharide antigens (Herrod et al., 1989; Kumararatne et al., 1991). These patients are particularly prone to recurrent sinopulmonary infections with capsulated bacteria and may benefit from immunoglobulin replacement therapy (Knutsen, 1989). At present the antibody
!
180• S
150. 120 ' !
~,
90' 60'
30,
:
± 3-10
T
|
11-19
•
;
[. 20-39
40-59
Agegroups (yrs)
60-80
,
t PAD
Fig. 4. Distribution of specific lgG to mannan according to age and expressed as units/ml with medium values shown. None of the patients with primary antibody deficiency(PAD) had a tilre greater than 3() U/ml.
TABLE i LEVELS OF IgG ANTIBODIES TO MANNAN IN HEALTHY CHILDREN AND ADULTS Group
Heahhy children Healthy adults Antibodydeficiency
Age (years)
no.
3-10 11-19 20-39 40-59 60-80 5-71
28 31 45 36 38 23
SpecificlgG to mannan(U/ml) Median lnterquartile range I 26 44 82 50 9
I- 10 12- 48 23- 92 46-122 23- 100 2- 11
Significance between groups p=
167
JIM 06400
A comparison of specific IgG antibody levels to the cell wall mannan of Candida albicans in normal individuals and in patients with primary antibody deficiency J.A. Faux, A.E. Agbarakwe, S.A. Misbah and H.M. Chapel Department of Immunology. ChurchillHospital, Oxford OX3 7LJ. UK (Received9 December 1991,revised received25 February 1992,accepted 3 April 1992)
An enzyme-linked immunosorbent assay (ELISA) has been developed to measure specific IgG antibody to the polysaccharide, cell wall mannan of Candida albicans (mannan). The results were expressed as arbitrary units/ml, with an inter- and intra-assay coefficient of variation of 7-11%. In establishing normal ranges we found that specific IgG to the mannan increased with age, with 18% of healthy children aged 3-10, 48% of healthy children aged 11-19 and 76% of an adult donor population having specific IgG antibody to mannan ( > 30 U / m l ) . We have compared these normal ranges, with a group of patients with primary antibody deficiency (PAD). None of the 23 patients with PAD, which included common variable immunodeficiency, lgG subclass deficiency, and selective IgA deficiency, had titres > 30 U / m l . The patients with PAD had significantly lower levels of specific IgG anti-mannan antibody (median 9 U / m i ) compared to healthy children aged 11-19 (median 26 U / m i ) or adults (median 58 U / m l ) ( p = < 0.001) but not children aged 3-10, (median 1 U / m l ) ( p = 0.08). Key words: ELISA; SpecificIgG; Cell wall mannan; Candida albicans; Primary antibody deficiency
Introduction In patients with suspected panhypogammaglobulinaemia quantitative defects of immunoglobulin synthesis can be diagnosed by measuring total serum immunoglobulin levels (WHO Report, 1989). Schur et al. (1970) reported an association between selective IgG2 subclass deficiency and a susceptibility to recurrent pyogenic infections. Since then patients with other selective subclass deficiencies have been reported (Oxelius, 1979; Stanley et al., 1984; Jefferis and
Correspoade,weto: J.A. Faux, Department of Immunology, Churchill Hospital, Oxford OX3 7LT, UK. Tel.: 0865-22993.
Kumararatne, 1990; Preud'homme and Hanson, 1990). More recently a further important defteiency has been recognized. Some patients with recurrent sinopulmonary infections have normal lgG subclass levels but are unable to mount an antibody response to bacterial polysaceharide antigens (Lane and Maelennan, 1986; Ambrosino et al., 1988; Geha, 1988). The diagnosis of this selective antibody deficiency is currently dependent on the demonstration of a poor specific lgG antibody response to Streptococcus pneumoniae and Haemophilus influenzae polysaccharide antigens (Shackleford et ai., 1987; Austrian, 1989). Another antigen that may give additional information on the ability to mount a specific immune response to polysaccharide antigens is the
168 cell wall mannan of Candida albicans. C. albicans is found as a commensal organism in the normal gastrointestinal tract, vagina and mouth with the result that nearly 50% of normal adults have been shown to have precipitating antibody to this antigen (Faux, 1968). We have developed a highly sensitive and specific ELISA for the quantitative determination of specific IgG antibody to cell wall mannan of C. albicans and compared the levels found in a normal population with a group of patients with primary antibody deficiency.
Materials and methods
Subjects In order to establish normal values for the assay, sera were collected from (i) 28 healthy children age range 3-10 (ii) 31 healthy children age range l l - 1 9 , both collected as part of a genetic linkage study for atopy, (iii) 81 adult blood donors age range 20-59, (iv) 38 elderly subjects age range 60-80 (J. Lea, in preparation) (v) 23 patients with primary antibody deficiency (PAD), which included, 14 with common variable immunodeficiency (CVI), four with IgA deftciency and five with selected IgG subclass deficiency. Antigen extracts (1) A purified extract of the cell wall mannan of Candida albicans (mannan) was provided by the Mycological Reference Laboratory, PHLS, London, and used in all the assays. (2) A cell wall mannan of C. albicans was provided by the National Institute for Biological Standards and Controls which was ased in the specificity assay (mannan NIBSC). Each preparztion was made up at 1 m g / m l in phosphate-buffered saline pH 7.4 plus 0.1% sodium azide and stored at 4°C until required. Using the Ouchterlony double diffusion test a line of identity was shown between these two extracts. Methods Microtitre plates (Dynatech) were sensitized overnight at 4°C with 100 /xl per well of a 5 p.g/ml solution of the mannan in carbonate buffer
(pH 9.6). All washings were performed four times using a washer (Flow Laboratories) and phosphate-buffered saline (pH 7.4) with 0.05% Tween 20 (PBST). The test sera were diluted 1 in 100 using PBST (100 /~1 per well in duplicate) and incubated for 1 h. All incubations were done at 21°C. After washing 100 tzl of a 1 in 10000 dilution of an alkaline phosphatase conjugated goat anti-human IgG (Fc affinity purified, ICN Immunobiologicals) were added to each well. The plates were incubated for 1 h. After washing, 100 /zl per well of p-nitrophenyiphosphate (1 mg/ml, Sigma) in diethanolamine buffer (pH 9.8) were added to each well and allowed to develop for 30 min at 210C in the dark. Stopping solution (50/~i of 3 M NaOH) were added to each well and the plates were read at 405 nm using an ELISA plate reader (Flow Laboratories)
Analysis of ELISA data Seven standards were prepared by diluting an IVlg (GammaGard, Baxter Healthcare) preparation with a negative normal serum. This was designated to contain 500 U / m l of specific lgG to the mannan. The U / m l of the unknown sera were calculated from the standard curve, plotted as l o g . U / m l versus linear OD giving an assay range of 1-200 U / m l . Two sera were routinely included in each assay for quality control, a low and a medium level. Both sera were from normal blood donors. Statistical anal-sis Groups were compared using the Mann Whitney U Test for non-paired data.
Results
Optimum antigen concentration Four sera, with differing binding capacities, were tested against dilutions of the mannan ranging from 20 to 1 tzg/ml to determine the optimum antigen concentration. Using the mean of the optical densities at different dilutions, maximum binding was found at an antigen dilution of 5/.~g/ml and this concentration was adopted for the study. Washed sensitized plates could be covered and stored a t - 20°C for 12 months before
169
use. Fresh and frozen plates were compared using the two control sera; the results for the low control were 13-18 U / m l of anti-mannan specific lgG and for the medium control, 45-65 U / m l of anti-mannan specific IgG. Same day and different day plates (nine assays) gave a coefficient of variation from 7-11%.
~ i
1.5
l.0,
~o.s
Assessmazt of antigenic specificity To test the specificity of the assay, increasing ten fold concentrations of the m a n n a n antigen diluted in PBS were preincubated with aliquots of a positive serum sample for 1 h at room temperature and then centrifuged at 1800 × g for 10 min. The absorbed sera were diluted and assayed as described. It was possible to inhibit binding completely with 10 p , g / m i of either of the mannan extracts (Fig. 1). No inhibition occurred with a m a n n a n extract purified from Saccharomyces cerecisiae at 1 m g / m l (unpublished data) an unrelated m a n n a n antigen.
Relationship between control and test sera Three adult sera were assayed using serial dilutions from 1 / 5 0 to 1/800 and compared with the standards. The resultant graphs demonstrated linearity and gave parallel slopes (Fig. 2).
~.
1.0.
o.0
, 1/50
•
,
.
11100
Standard crave
~
,
.
1i200
" , 11400
d~tioo Bdl ~
.
,
-
11800 BdZ
---4---
Bd3
Fig. 2. Comparisonof the ELISA standard curve with the sera of three adult blood donors; Bdl, Bd2 and Bd3, diluted 1/50 to I/8(X)givinggood linear and parallel slopes.
Non-specific binding to sensitized plate~ The diluent used throughout the assay for the standard curve, unknown sera and anti lgG conjugate was PBS + 0.05% Tween. We investigated whether the addition of either 1% goat serum, 1% horse serum or 1% bovine serum albumin to this assay diluent altered the standard curve or the results of five sera which contained a range of specific IgG to the mannan. All samples were assayed on the same plate. The standard curves and the assay results with each diluent were similar; with specific lgG to the m a n n a n for these five sera ranging from 1 to 5 U / m l , 4-6 U / m l , 1 0 - ! 4 U / m l , 48-54 U / m l and 150-165 U / m l respectively, with no consistent differences in the results using the different diluents.
Antibody lerels to mannan in selected groups o"
o.s
0.0
. . . . . . . . . , .1 1 Antigen
....... 10 concentration
Mannan Extract
•
, 100
ug/ml Mannan NIRSC
Fig. I. Specificity of the ELISA procedure for assaying IgG a n t i - m a n n a n a n t i b o d i e s w a s d e m o n s t r a t e d by i n h i b i t i o n o f b i n d i n g w i t h 1 0 / . t g / m l o f t h e m a n n a n p r o v i d e d by t h e N a tional Institute of Standards and Biological Control and the m a n n a n e x t r a c t u s e d in t h e assays.
In establishing a normal range we found that ther~ was a significant increase in specific IgG to mannan with age (Fig. 3). Only 18% of healthy children, aged 3-10, but 48% of healthy children aged 11-19 and 76% of healthy adults aged 20-80 had a titre greater than 30 U / m l . The median value for adults was 58 U / m l with interquartile range 30-114 U / m l . Within this adult group we found no significant differences in specific lgG to the mannan between male and female adults (Fig. 3). There was a significant increase in specific antibody between young adults (aged 20-39) and middle aged adults (40-59 years) with me-
170 dian values of 44 U / m i and 82 U / m l respectively ( p = 0.05), but the decrease in the median U / m l shown with a further increase in age (60-80 years) was not significant ( p = 0.23) (Table I). None of the patients with PAD had specific lgG of > 30 U / m l to the cell wall mannan (Fig. 4), the median value was 9 U / m l and the interquartile range 2-11 U / m l (Table I). This was significantly different from the normal adults and children aged 11-19 ( p = < 0.001) but not the younger children aged 3-10, with a median value 1 U / m l ( p = 0.08); only one of the patients studied here was in this age range.
AgeGroups (yrs) • 3-1o • 20-39 [ ] 40-59 I:] so.80 • subjects 11-~9 Adult
"~ 75 1" so
i I ~ ~ t~ 25
[] []
male
female
0 Noin 2831453638 5564 group Fig. 3. Percentage of the study group with specific lgG to
mannan > 30 U/ml showing an increase with age, but no significant differencebetween male and female adults. There were comparable numbersin each age group.
•
¢,
g,
"
; ,
-L .
°
i
1 "
i[
Discussion We report here the development of a reproducible and sensitive ELISA for the detection of specific lgG antibody to the cell wall mannan of Candida albicans. We found that 76% of the normal adult donor population but none of the patients with primary antibody deficiency (PAD) had a titre greater than 30 U / m l . This suggests that the measurement of specific lgG antibody to cell wall mannan of C. albicans may be useful in the laboratory assessment of PAD. Recent studies from the United States and Britain have highlighted the clinical significance of specific antibody deficiency to po.lysaccharide antigens (Herrod et al., 1989; Kumararatne et al., 1991). These patients are particularly prone to recurrent sinopulmonary infections with capsulated bacteria and may benefit from immunoglobulin replacement therapy (Knutsen, 1989). At present the antibody
!
180• S
150. 120 ' !
~,
90' 60'
30,
:
± 3-10
T
|
11-19
•
;
[. 20-39
40-59
Agegroups (yrs)
60-80
,
t PAD
Fig. 4. Distribution of specific lgG to mannan according to age and expressed as units/ml with medium values shown. None of the patients with primary antibody deficiency(PAD) had a tilre greater than 3() U/ml.
TABLE i LEVELS OF IgG ANTIBODIES TO MANNAN IN HEALTHY CHILDREN AND ADULTS Group
Heahhy children Healthy adults Antibodydeficiency
Age (years)
no.
3-10 11-19 20-39 40-59 60-80 5-71
28 31 45 36 38 23
SpecificlgG to mannan(U/ml) Median lnterquartile range I 26 44 82 50 9
I- 10 12- 48 23- 92 46-122 23- 100 2- 11
Significance between groups p=
