Journal of Immunological Methods 268 (2002) 149 – 157
A hemolytic assay for the estimation of functional
mannose-binding lectin levels in human serum
Saskia Kuipers a,*, Piet C. Aerts a,b, Anders G. Sjoholm c,
Theo Harmsen , Hans van Dijk
Eijkman-Winkler Center for Microbiology, Infectious Diseases, and Inflammation, University Medical Center Utrecht G04.614,
Heidelberglaan 100, 3584 CX Utrecht, The Netherlands
Department of Pediatric Infectious Diseases, Wilhelmina Children’s Hospital, University Medical Center Utrecht, Utrecht, The Netherlands
Institute of Laboratory Medicine, Section for Microbiology, Immunology, and Glycobiology, Lund University, Lund 2100, Sweden
Received 7 December 2001; received in revised form 13 May 2002; accepted 23 May 2002
A simple assay was developed to estimate functional mannose-binding lectin (MBL) levels in serum based on the
principle of yeast-induced bystander lysis of chicken erythrocytes (ChE). The assay is sensitive to inhibition by ethylene glycol
bis-(h-aminoethyl ether)-N,N,NV,NV-tetraacetic acid (EGTA) (which allows alternative pathway activation), ethylene diamine
tetraacetic acid (EDTA), mannose, N-acetylglucosamine and C1 esterase inhibitor (C1-INH), whereas it was not inhibited by
galactose. A high-titer human anti-mannan antibody-containing serum with 0.06 Ag MBL/ml gave a functional signal
corresponding to 0.12 Ag equivalents MBL/ml, indicating that anti-mannan antibodies are poorly hemolytic in the assay. The assay
is well suited for the large-scale testing of patient samples for a functional MBL pathway of complement activation.
D 2002 Elsevier Science B.V. All rights reserved.
Keywords: MBL; Lectin pathway; Functional assay; Bystander hemolysis; Saccharomyces cerevisiae
Abbreviations: AP50, hemolytic alternative pathway activity;
CH50, hemolytic overall complement activity; ChE, chicken Mannose-binding lectin (MBL) is a high-molec-
erythrocytes; C1-INH, C1 esterase inhibitor; EDTA, ethylene
diamine tetraacetic acid; EGTA, ethylene glycol bis-(h-aminoethyl
ular-weight protein present in blood plasma at low
ether)-N,N,NV,NV-tetraacetic acid; ELISA, enzyme-linked immuno- concentrations (1.7 Ag/ml). Together with the human
sorbent assay; HPS, human pooled serum; MAp19, MBL-associated proteins CL-L1 (Ohtani et al., 2001), CL-P1 (Ohtani
protein of 19 kDa; MASP, MBL-associated serine proteinase; MBL, et al., 1999), and lung surfactant proteins A and D
mannose-binding lectin; MoAb, monoclonal antibody; PMN, and the bovine proteins conglutinin, collectin-43
polymorphonuclear granulocyte; VBS, veronal-buffered saline.
Corresponding author. Tel.: +31-302506536; fax: +31-
(Holmskov, 2000) and collectin-46 (Hansen et al.,
302541770. 2001), the protein belongs to the family of C1q-
E-mail address: firstname.lastname@example.org (S. Kuipers). related ‘‘collagenous lectins’’ (collectins). MBL is
0022-1759/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 1 7 5 9 ( 0 2 ) 0 0 1 9 2 - 8
150 S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157
the first component of the lectin pathway (LP) of for the P/Q promoter polymorphism is probably not
complement activation and binds microorganisms useful (Madsen et al., 1995, 1998).
and foreign particles via specific sugars (mannose, Over the past 12 years, several enzyme immuno-
N-acetylglucosamine, and fructose) by its three iden- assays have been developed to estimate antigenic
tical lectin moieties, thereby acting as an opsonin MBL levels in serum. One common disadvantage of
(Turner, 1996). When foreign bodies are bound by these antigenic assays, however, is that they fail to
MBL, three MBL-associated serine proteins (MASP- measure functional MBL, i.e. the joint activity of the
1, MASP-2, and MAp19) (Matsushita and Fujita, MBL-MASPs complex. Recently, functional MBL
1992; Thiel et al., 1997, 2000; Stover et al., 1999) assays based on either the lysis of mannan-coated
become coordinately activated, resulting in the gen- erythrocytes (Suankratay et al., 1998) or the mannan-
eration of the active form of MASP-2, the LP- induced C4b deposition were developed (Petersen et
dependent C4/C2 convertase. The role of a recently al., 2001). In the present paper, we describe a new
discovered fourth MBL-associated protein (MASP-3) hemolytic test for the estimation of functional MBL in
probably involves down-regulation of MASP-2 as serum and compare the data obtained with those from
shown by in vitro experiments (Dahl et al., 2001). a competitive MBL enzyme-linked immunosorbent
Substrate-bound C4bC2a is a C3 convertase, which assay (ELISA) presented earlier (Bax et al., 1999).
allows the conversion of C3 into C3a and C3b, the In short, the yeast Saccharomyces cerevisiae was used
key reaction leading to terminal pathway initiation. as an MBL activator in a dilution series of human
Besides indirect C3 activation, the lectin pathway serum, to which MBL-deficient serum was added as a
may also be involved in direct C3 activation (Kuhl- source of all complement components except MBL.
man et al., 1989): MASP-1 can cleave C3 independ- Chicken erythrocytes (ChE) were used as the target
ently (Matsushita and Fujita, 1995). An ex vivo for hemolysis. The influence of classical pathway
model has shown that MBL decreases proinflamma- activation by anti-mannan IgG antibodies on the out-
tory cytokine production in meningococcal disease come of the assay was studied using an MBL-defi-
(Jack et al., 2001). cient serum with high anti-mannan antibody levels.
MBL deficiencies are quite common in man. In MBL-specific oligosaccharides mannose and N-ace-
fact, low-level haplotypes can occur in up to 30% of tyl-D-glucosamine (GlcNAc), but not galactose, sub-
the human population (Ezekowitz, 2001) and are stantially reduced hemolysis, which is in line with the
associated with all kinds of infectious and infec- idea that the former two sugars compete with yeast
tion-related diseases. MBL treatment of individuals cells for MBL binding. The MBL regulatory function
with recurrent infections secondary to severe MBL of C1 esterase inhibitor (C1-INH), which regulates the
deficiency has proven useful in certain cases (Valdi- lectin pathway by binding to MASP-1 and MASP-2 in
marsson et al., 1998), which strongly suggests that an equimolar manner, could be confirmed using our
there is a future for MBL substitution therapy. The assay (Matsushita et al., 2000).
genetic basis of MBL deficiency is found in variant
alleles and promoter polymorphisms. All mbl gene
codon mutations are found in exon 1 on chromo- 2. Materials and methods
some 10. Single-base substitutions, found at codons
52, 54, and 57, are indicated as the variant alleles D, 2.1. Chemicals
B, and C, respectively. These variant alleles lower
the serum MBL levels (Madsen et al., 1995). D (+)-Mannose (Mr 180.2) was obtained from
Changes in the promoter region of the MBL gene, Sigma (Zwijndrecht, the Netherlands) (catalogue
called H (high), L (low), X, and Y, also have a number M-4625), GlcNAc (Mr 221.2) from Serva
profound impact on MBL serum levels, with LX (Heidelberg, Germany) (catalogue number 10290),
leading to low MBL levels, LY to intermediate levels D(+)-galactose (Mr 180.2) from Sigma (catalogue
and HY to high levels. The P promoter polymor- number G-0750), and C1-INH from Calbiochem-
phism is found in both high-expressing and low- Novabiochem (La Jolla, CA, USA) (1 mg/ml, cata-
expressing haplotypes. Therefore, an isolated search logue number 204883).
S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157 151
2.2. Buffers reagent serum as a source of all complement com-
ponents except MBL, and 109 ChE as the target cells
Veronal-buffered saline (VBS+ + ): 5 mM VBS, pH of bystander hemolysis. Direct hemolysis of the
7.4, containing 0.15 mM Ca2 + and 0.5 mM Mg2 + . erythrocytes induced by yeast cells was excluded
EGTA-VB: VBS containing 8 mM EGTA and 2.5 mM by incubating the erythrocytes and yeast cells
Mg2 + . EDTA-VB: VBS containing 10 mM EDTA. together just with reagent serum. Human pooled
serum (MBL titer 1.67 Ag/ml as determined by
2.3. Human serum ELISA) (Bax et al., 1999) was used as the MBL
reference sample in the assay. A total of 100 Al of
Blood collected from 20 healthy workers of our test mixture was added to each test well containing
laboratory and from 121 donors to a blood bank was serum dilutions, and the microtiter plate was put on a
allowed to clot for 30 min and was subsequently water-based incubator operating at 37 jC (Klerx et
centrifuged at 1500 Â g for 10 min. Individual sam- al., 1983). After incubation, the erythrocytes were
ples were stored at À 70 jC until further use. The sera spun down and 50 Al samples of each supernatant
from the laboratory workers were subsequently were transferred to a flat-bottom plate, with 200 Al
pooled [human pooled serum (HPS)] and stored samples of water in each well. Hemoglobin release
similarly. Serum from a subject known to have a very was measured in an ELISA reader operating at a
low MBL level (Bax et al., 1999) was used as the wavelength of 405 nm. Percentages of hemolysis
MBL reagent serum in the functional assay. Serum were calculated using controls for 100% (water
with a high IgG anti-mannan antibody titer (IgG 8.7 lysed) and 0% (buffer control) hemolysis. The per-
mg/l and IgM 5.6 mg/l) from an MBL-deficient centages were then transformed to the number of
individual (antigenic MBL level of 0.06 Ag/ml) was active sites per cell according to the equation pub-
used to study the influence of anti-mannan antibodies lished by Borsos and Rapp (1963):
on very low MBL levels. Informed consent was
obtained from all donors. Z ðnumber of active sites per cellÞ
¼ Àlnð1 À fraction erythrocytes lysedÞ
2.4. Baker’s yeast
Titers were read at Z = 0.200. Functional MBL
S. cerevisiae cells were cultured on Sabouraud agar titers were defined as titers in microgram equivalent
plates (Merck, Darmstadt, Germany) for 48 h at 37 jC MBL per ml relative to the known reference. Sera
before being used as lectin pathway activators. Yeast were also tested in EGTA-VB to study alternative
particles were enumerated microscopically by the pathway activation and in EDTA that inhibits the
Thoma slide counting chamber method (W. Schreck, three complement activation pathways. For this and
Hofheim, Germany). all other results presented, a minimum of three
replicate assays was performed. The interassay and
2.5. Functional assay for mannose-binding lectin intraassay variations (n z 10) were determined as
The functional MBL assay, based on the principle
of MBL-dependent bystander hemolysis, was exe- 2.6. Inhibition of MBL activation in the functional
cuted in U-well microtiter plates (Greiner, Fricken- assay by competing carbohydrates
hausen, Germany). Firstly, sera to be tested were
serially diluted in the test plate in order to obtain 50 This was done by preincubating test serum sam-
Al samples in a dilution series of 10 À 0.5 (1/10 – 1/ ples with mannose, GlcNAc, or galactose at 37 jC for
3162), starting with a serum dilution of 10% in 30 min, starting with a concentration of 2.5 mg
VBS+ + . Then, a 10 ml test mixture in VBS+ + was carbohydrate per well. A checkerboard titration was
prepared per microtiter plate. Such a mixture con- done in the following manner: HPS was diluted in all
tained a standardized amount of freshly cultured the horizontal rows of the microtiter plate, and di-
baker’s yeast S. cerevisiae (3.0 Â 107 cells), 150 Al lutions (1/3) of the carbohydrates were tested (833,
152 S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157
278, 92.6, 30.9, 10.3, and 3.4 Ag/well) in the vertical 2.10. Human anti-mannan antibodies
A serum sample from an MBL-deficient subject
2.7. The effect of C1-INH in the functional assay (ELISA titer 0.06 Ag/ml), with a high anti-mannan
antibody titer (IgG 8.7 mg/l and IgM 5.6 mg/l,
C1-INH with a concentration of 1 mg/ml was A.G.S.), was tested for functional MBL in order to
added to the functional assay. The highest concen- study the influence of classical pathway activation on
tration tested was 350 Ag/ml, which is the physiolog- the outcome of the assay.
ical concentration of C1-INH. Final concentrations of
175 and 87.5 Ag/ml C1-INH were also tested. In each 2.11. Statistical methods
well, 350 Ag C1-INH or a dilution thereof was mixed
with serial dilutions of human pooled serum. The The Pearson correlation coefficient was calcu-
mixture was allowed to incubate at 37 jC for 30 lated and used to analyze data generated by both
min before the addition of yeast cells, MBL-deficient the functional MBL test and the competitive MBL
serum and chicken erythrocytes. Direct hemolysis by ELISA that was applied to the sera of 121 blood
C1-INH was ruled out by adding C1-INH to eryth-
rocytes and MBL-deficient serum only.
2.8. Anti-MBL monoclonal antibody (MoAb) and
purified human MBL
The commercially available anti-MBL IgG1 anti-
body clone 131-1 (1 mg/ml, lot number 01044PA01)
was obtained from the Statens Serum Institute
(Copenhagen, Denmark). Human recombinant MBL,
which was a kind gift from Dr. R.A.B. Ezekowitz
(Boston, MA, USA), was more than 95% pure.
2.9. MBL-specific competitive ELISA
MBL levels in human sera were estimated using a
protein-specific competitive ELISA. This was exe-
cuted as follows: ELISA plates were first coated with
1.5 Ag/ml recMBL in phosphate-buffered saline (PBS)
and blocked with 4% skim milk in PBS. Then, mixtures
were prepared consisting of 1/5000-diluted anti-MBL
monoclonal antibody in PBS and equal volumes of
serial 10 À 0.5 dilutions of test samples and allowed to
incubate for 1 h at room temperature. After incubation,
the amounts of bound antibody were tested by 1/6000-
diluted peroxidase-labeled anti-mouse IgG antibody in
PBS (Nordic, Tilburg, the Netherlands). Tetramethyl-
benzidine was used as the chromogenic substrate. After Fig. 1. The effects of 1% (a) and 1.7% (b) MBL-deficient serum, as
each incubation step, the wells were thoroughly the reagent serum in the functional MBL assay, on the apparent
washed. Two sera with MBL levels of 0.396 and functional activities of human pooled serum and MBL-deficient
serum, expressed in arbitrary units (AU) per milliliter. Note that the
1.248 Ag/ml (a kind donation of Dr. P. Garred, Copen- alternative pathway activity of 1.7% MBL-deficient serum (meas-
hagen, Denmark) were used as the intermediate and ured in EGTA-VB) exceeds the apparent MBL activity (measured in
high level MBL references, respectively. VBS + + ). Vertical bars indicate the standard error of the mean.
S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157 153
the sample standard deviation and x the sample
The known MBL-deficient donor serum was tested
in the functional assay against the human serum pool,
which had an established antigenic MBL level of 1.67
Ag/ml. The level of antigenic MBL in our reagent
serum was 0.206 Ag/ml. Several reagent serum con-
centrations were tested initially in the functional
assay, but a final concentration of 1% reagent serum
Fig. 2. Functional MBL serum levels are read at a Z value of 0.2. proved to yield the best results with the least contri-
bution (equal to an MBL concentration of 0.02 Ag/ml)
of the alternative pathway measured with EGTA-VB
donors. Interassay and intraassay coefficients of instead of VBS+ + (Fig. 1). Based on the MBL-
variation were estimated from 10 measurements of deficient serum from the regular donor, a Z value of
HPS using the equation: (s/x) Â 100%, where s is 0.200 proved to best estimate functional MBL titers
Fig. 3. Comparison of functional MBL serum levels obtained with the functional assay (shown on the y-axis) with the data obtained with a
competitive ELISA (shown on the x-axis). Sera tested were from donors to a blood bank (n = 121).
154 S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157
(Fig. 2). The intraassay variation was 5.6% and the
interassay variation was 5.2%.
Sera from 121 donors to a blood bank were tested
with both a competitive ELISA and the functional
assay. A highly significant ( P < 0.0001) positive
correlation (r = 0.869) was found between the MBL
ELISA results and the functional MBL levels, as
shown in Fig. 3. The intercept of the line is at about
0.2 Ag/ml equivalent and not at zero.
The MBL-deficient serum with high anti-mannan
antibody titers scored 0.06 Ag MBL/ml when tested
with ELISA and 0.12 Ag equivalent MBL/ml when
tested with our functional assay. We also studied Fig. 5. The influence of C1-INH on MBL serum levels in a
the inhibition of lectin pathway activation by the functional MBL assay. C1-INH is a known inhibitor of MBL
activation. HPS (MBL 1.67 Ag/ml) without C1-INH is used as a
carbohydrates GlcNAc, D(+)-mannose, and D(+)-gal- control (C).
actose using our functional assay. GlcNAc (Fig. 4a)
and mannose (not shown here) fully blocked MBL-
mediated hemolysis, while galactose did not (Fig. 4b). When the MASP-1 and MASP-2 binding C1-
INH was added to the MBL functional assay,
hemolysis was blocked in a concentration-dependent
way (Fig. 5).
This paper describes a new hemolytic assay for
estimating the functional activity of MBL, the leading
component of the lectin complement pathway. Baker’s
yeast cells were chosen as MBL-activating particles
for this assay, because their cell wall is rich in mannan
and because S. cerevisiae was the first reported
microorganism to be associated with MBL (Bull and
Turner, 1984). Before MBL was recognized as a
complement component, defective yeast opsonization
had already been described in 25% of children with
frequent unexplained infections (Soothill and Harvey,
1976). This defective yeast opsonization had also
been noted in the sera from children with an increased
incidence of infection or atopy (Richardson et al.,
1983). Defective yeast opsonization was functionally
measured using opsonized baker’s yeast as the sub-
strate for polymorphonuclear granulocyte (PMN)-
dependent phagocytosis (Soothill and Harvey, 1976;
Turner , 1986) by studying neutrophil iodination
response (Roberton et al., 1981) and by measuring
chemoluminescence emitted by PMNs upon incuba-
Fig. 4. Saccharide-mediated inhibition of MBL activity by the tion with yeast cells opsonized with normal or defi-
saccharide GlcNAc (a) but not by galactose (b). cient sera (Turner, 1986). Tests involving PMN
S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157 155
function, however, suffer from the disadvantage that The two carbohydrates with equatorial 3-OH and
their results may differ from PMN donor to donor due 4-OH groups (mannose and GlcNAc) inhibited MBL-
to the many Fc receptor polymorphisms that occur in dependent bystander lysis in a dose-dependent man-
the population. ner. In contrast, galactose, which does not fulfill these
Although another assay has been described for the spatial demands, showed less inhibition of functional
measurement of MBL-initiated hemolysis based on MBL, thereby confirming earlier findings by Holm-
the lysis of mannan-coated erythrocytes (Suankratay skov et al. (1994). These researchers demonstrated the
et al., 1998), the test we describe here seems to be less saccharide selectivity of collectins using a solid-phase
complex and less laborious, while the accuracy is assay, in which human MBL was best inhibited by
acceptable (about 5% interday and intraday variabil- GlcNAc, less by mannose, and not at all by galactose.
ity). The test we present here appears to be very useful C1-INH, which is a known inhibitor of complement
for measuring functional MBL not only in the normal activation via the classical pathway subcomponents
serum-level range but also in case of deficiencies. C1r and C1s (Sim and Reboul, 1981), has been
Other less sensitive assays, e.g. those based on neph- recognized as an inhibitor of the lectin pathway
elometry, often lack linearity at the low level range. A through binding to MASP-1 and MASP-2 (Wong et
strange but reproducible observation was that the al., 1999; Petersen et al., 2000). In our functional assay,
functional assay does not detect MBL levels below profound inhibition was seen at levels of 350 and 175
0.20 Ag MBL equivalent/ml. This intrinsic property of Ag/ml, which confirms the finding by Matsushita et al.
the functional assay is not easy to explain, unless we (2000) that C1-INH is the only potent inhibitor of
suppose that the chains encoded by variant MBL MBL-MASPs in a hemolytic system. Results obtained
alleles have some residual functional activity. with the putative lectin pathway inhibitors a2-macro-
As far as the MBL activator in this functional assay globulin (Storgaard et al., 1995; Terai et al., 1995;
is concerned, all yeasts are efficient activators of the Armstrong and Quigley, 1999) and aprotinin were
alternative complement pathway, but the advantage of inconclusive and therefore not described in this paper.
S. cerevisiae is that it lacks pathogenicity. With regard Recently, using this functional MBL assay, we
to a possible contribution to the eventual MBL titer of tested more than 750 serum samples from young
the alternative pathway constituents in the MBL children suffering from unexplained repeated episodes
reagent serum, we found that this is not more than of otitis media due to pneumococcal infection. Our
5%, at least when using 1% MBL-deficient serum, results showed a very high incidence of MBL defi-
which we consider to be negligible. ciency in these children as measured by the functional
The apparently low MBL activity found in MBL- assay. The cutoff values used (0.2 Ag MBL equiv-
deficient serum with high anti-mannan immunoglo- alents/ml) was based on an epidemiological study in
bulin levels indicates that classical pathway activation the Dutch population (submitted for publication) (very
by anti-mannan antibodies is not likely to contribute low serum MBL < 0.2 Ag MBL equivalents/ml, low
to major false-positive MBL results. This is in line serum MBL 0.2– 0.42 Ag MBL equivalents/ml). The
with findings by Super et al. (1990), who did not incidence of functional MBL deficiency ranged from
observe any correlation between the levels of IgG1, 24.6% to 35.3%, depending on the age group tested,
IgG3, or IgM anti-mannan antibodies and level of C4 versus 4.9% in the control group (unpublished
or C3bi binding. This allowed us to conclude that, in results). Based on these and the results mentioned
an experimental system using low serum concentra- above, this novel functional MBL assay could be very
tions, MBL cleaves complement component C4 in an useful for large-scale testing in the medical immunol-
antibody-independent manner. In samples with low ogy laboratory.
C4 serum levels, which theoretically is possible in
homozygous C4A or C4B deficiency, MBL activation
still leads to hemolysis in the assay due to MASP-1- Acknowledgements
induced C3 cleavage. The latter ‘‘bypass mechanism’’
will not be found in a C4b binding assay as described The authors want to thank Mr. Henny de
previously (Petersen et al., 2001). Kruijff (Van Miert’s Poultry Slaughterhouse, Breu-
156 S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157
kelen, the Netherlands) for providing the chicken Matsushita, M., Fujita, T., 1992. Activation of the classical comple-
erythrocytes. ment pathway by mannose-binding protein in association with a
novel C1s-like serine protease. J. Exp. Med. 176, 1497.
Matsushita, M., Fujita, T., 1995. Cleavage of the third component of
complement (C3) by mannose-binding protein-associated serine
References protease (MASP) with subsequent complement activation. Im-
munobiology 194, 443.
Armstrong, P.B., Quigley, J.P., 1999. a2-macroglobulin: an evolu- Matsushita, M., Thiel, S., Jensenius, J.C., Terai, I., Fujita, T., 2000.
tionary conserved arm of the innate immune system. Dev. Proteolytic activities of two types of mannose-binding lectin-
Comp. Immunol. 23, 375. associated serine protease. J. Immunol. 165, 2637.
Bax, W.A., Cluysenaer, O.J.J., Bartelink, A.K.M., Aerts, P.C., Eze- Ohtani, K., Suzuki, Y., Eda, S., Kawai, T., Kase, T., Shimada, T.,
kowitz, R.A.B., van Dijk, H., 1999. Association of familial Keshi, H., Sakai, Y., Fukuoh, A., Sakamoto, T., Wakamiya, N.,
deficiency of mannose-binding lectin and meningococcal dis- 1999. Molecular cloning of a novel human collectin from liver
ease. Lancet 354, 1094. (CL-L1). J. Biol. Chem. 274, 13681.
Borsos, T., Rapp, H.J., 1963. Chromatographic separation of the Ohtani, K., Suzuki, Y., Eda, S., Kawai, T., Kase, T., Keshi, H.,
first component of complement and its assay on molecular basis. Sakai, Y., Fukuoh, A., Sakamoto, T., Itabe, H., Suzutani, T.,
J. Immunol. 91, 851. Ogasawara, M., Yoshida, I., Wakamiya, N., 2001. The mem-
Bull, F.G., Turner, N.J., 1984. A serum mannan-binding protein and brane-type collectin CL-P1 is a scavenger receptor on vascular
candidiasis. Sabouraudia 22, 347. endothelial cells. J. Biol. Chem. 276, 44222.
Dahl, M.R., Thiel, S., Matshushita, M., Fujita, T., Willis, A.C., Petersen, S.V., Thiel, S., Jensen, L., Vorup-Jensen, T., Koch,
Christensen, T., Vorup-Jensen, T., Jensenius, J.C., 2001. C., Jensenius, J.C., 2000. Control of the classical and the
MASP-3 and its association with distinct complexes of the man- MBL pathway of complement activation. Mol. Immunol. 37,
nan-binding lectin complement activation pathway. Immunol- 803.
ogy 15, 127. Petersen, S.V., Thiel, S., Jensen, L., Steffensen, R., Jensenius, J.C.,
Ezekowitz, R.A.B., 2001. Mannose-binding lectin in prediction of 2001. An assay for the mannan-binding lectin of complement
susceptibility to infection. Lancet 358, 598. activation. J. Immunol. Methods 257, 107.
Hansen, S., Holm, D., Vitved, L., Bendixen, C., Reid, K.B.M., Richardson, V.F., Larcher, V.F., Price, J.F., 1983. A common con-
Skjoedt, K., Holmskov, U., 2001. CL-46, a novel collectin genital immunodeficiency predisposing to infection and atopy in
highly expressed in bovine thymus. Abstract. Proceedings of infancy. Arch. Dis. Child. 58, 799.
the Fifth International Workshop on C1, the First Component Roberton, D.M., Dhanjal, N.K., Levinsky, R.J., Mowbray, J.F.,
of Complement, and Collectins. Universitat Mainz, Mainz,
¨ Turner, M.W., 1981. Polymorphonuclear neutrophil iodination
Germany, p. 5. response as an estimate of defective opsonization. Clin. Exp.
Holmskov, U., 2000. Collectins and collectin receptors in innate Immunol. 43, 208.
immunity. APMIS, Acta Pathol. Microbiol. Immunol. Scand. Sim, R.B., Reboul, A., 1981. Preparation and properties of human
100, 1. C1 inhibitor. Methods Enzymol. 80, 43.
Holmskov, U., Malhotra, R., Sim, R.B., Jensenius, J.C., 1994. Col- Soothill, J.F., Harvey, B.A., 1976. Defective opsonization. A com-
lectins: collagenous C-type lectins of the innate immune defense mon immunity deficiency. Arch. Dis. Child. 51, 91.
system. Immunol. Today 15, 67. Storgaard, P., Nielsen, E.H., Skriver, E., Andersen, O., Svehag, S.-
Jack, D.L., Read, R.C., Tenner, A.J., Frosch, M., Turner, M.W., E., 1995. Mannan-binding protein forms complexes with a2-
Klein, N.J., 2001. Mannose-binding lectin regulates the inflam- macroglobulin. A proposed model for interaction. Scand. J. Im-
matory response of human professional phagocytes to Neisseria munol. 42, 373.
meningitidis serogroup B. J. Infect. Dis. 184, 1152. Stover, C.M., Thiel, S., Thelen, M., Lynch, N.J., Vorup-Jensen, T.,
Klerx, J.P.A.M., Beukelman, C.J., van Dijk, H., Willers, J.M.N., Jensenius, J.C., Schwaeble, W.J., 1999. Two constituents of the
1983. Microassay for colorimetric estimation of complement initiation complex of the mannan-binding lectin activation path-
activity in guinea pig, human and mouse serum. J. Immunol. way of complement are encoded by a single structural gene. J.
Methods 63, 215. Immunol. 162, 3481.
Kuhlman, M., Joiner, K., Ezekowitz, R.A.B., 1989. The human Suankratay, C., Zhang, X.-H., Zhang, Y., Lint, T.F., Gewurz, H.,
mannose-binding protein functions as an opsonin. J. Exp. 1998. Requirement for the alternative pathway as well as C4 and
Med. 169, 1733. C2 in complement-dependent hemolysis via the lectin pathway.
Madsen, H.O., Garred, P., Thiel, S., Kurtzhals, J.A.L., Lamm, L.U., J. Immunol. 160, 3006.
Ryder, L.P., Svejgaard, A., 1995. Interplay between promoter Super, M., Levinsky, R.J., Turner, M.W., 1990. The level of
and structural variants control basal serum level of mannan- mannan-binding protein regulates the binding of comple-
binding protein. J. Immunol. 155, 3013. ment-derived opsonins to mannan and zymozan at low serum
Madsen, H.O., Satz, M.L., Hogh, B., Svejgaard, A., Garred, P., concentrations. Clin. Exp. Immunol. 79, 144.
1998. Different molecular events result in low protein levels Terai, I., Kobayashi, K., Matsushita, M., Fujita, T., Matsuno, K.,
of mannose-binding lectin in populations from Southeast Africa 1995. a2-macroglobulin binds to and inhibits mannose-binding
and South America. J. Immunol. 161, 3169. protein-associated serine protease. Int. Immunol. 7, 1579.
S. Kuipers et al. / Journal of Immunological Methods 268 (2002) 149–157 157
Thiel, S., Vorup-Jensen, T., Stover, C.M., Schwaeble, W.J., Laursen, moluminescence with selected yeasts and bacteria using sera of
S.B., Poulsen, K., Willis, A.C., Eggleton, P., Hansen, S., Holm- different opsonic potential. Immunology 58, 111.
skov, U., et al., 1997. A second serine protease associated with Turner, M.W., 1996. Mannose-binding lectin: the pluripotent mole-
mannan-binding lectin that activates complement. Nature 386, cule of the innate immune system. Immunol. Today 17, 532.
506. Valdimarsson, H., Stefansson, M., Vikingsdottir, T., Arason, G.J.,
Thiel, S., Petersen, S.V., Vorup-Jensen, T., Matsushita, M., Fujita, Koch, C., Thiel, S., Jensenius, J.C., 1998. Reconstitution of
T., Stover, C.M., Schwaeble, W.J., Jensenius, J.C., 2000. Inter- opsonizing activity by infusion of mannan-binding lectin
action of C1q and mannan-binding lectin (MBL) with C1r, C1s, (MBL) to MBL-deficient humans. Scand. J. Immunol. 48, 116.
MBL-associated serine proteases 1 and 2, and the MBL-associ- ´
Wong, N.K.H., Kojima, M., Dobo, J., Ambrus, G., Sim, R.B., 1999.
ated protein Map19. J. Immunol. 165, 878. Activities of the MBL-associated serine proteases (MASPs) and
Turner, M.W., 1986. Evaluation of C3b/C3bi opsonization and che- their regulation by natural inhibitors. Mol. Immunol. 36, 853.