Journal of Immunological Methods, 131 (1990) 99-104 99

Elsevier

JIM 05627

Covalently linked peptides for enzyme-linked

immunosorbent assay
Jan Sonderg/ird-Andersen a, Edgar Lauritzen 2, Klaus Lind 1 and Arne H o l m 3
Research Center for Medical Biotechnology, 1Mycoplasma Laboratory/Neisseria Department, 2 H I V Laboratory/Rubella Department,
Statens Seruminstitut, Amager Boulevard 80, DK-2300 Copenhagen S, Denmark, and 3 Chemical Laboratory II,
The H. C. Orsted Institute, University of Copenhagen, Denmark
(Received 6 October 1989, revised received 24 January 1990, accepted 29 January 1990)

A general method is described, by which synthetic peptides are covalently linked via their carboxyl
group to microtiter plates (CovaLink) for enzyme-linked immunosorbent assay (ELISA). Plates were
prepared by this method with an angiotensin II peptide and with an HIV-2 peptide and attachment
detected by rabbit anti-angiotensin serum and with a positive serum from an HIV-2-infected patient,
respectively, using the common ELISA procedure in the last steps. The method is simple to perform, it
constitutes an alternative to the common ELISA method, and eliminates the risk of inadvertent loss of
peptide during the procedure. The method is highly reproducible and has a high sensitivity. It may be used
for either antigen or antibody detection.

Key words: Immobilization; ELISA; Synthetic peptide; CovaLink; Angiotensin II; HIV-2

Introduction peptide with a panel of buffers with different pH

Since the introduction of solid-phase peptide
synthesis by Merrifield (1963), synthetic peptides
have greatly facilitated both the study of antigen-
antibody interaction (Hodges et al., 1983; Geysen,
1985) and the mapping of antigenic determinants
(Goudsmit, 1988), and they have a promising fu-
ture in the field of synthetic vaccines (Arnon,
1987) and specific diagnostic tests for micro-
organisms (Gnann, 1987a).
However, when working with synthetic peptides
a major problem is the test system used for mea-
suring their reactivity with antibodies.
It is possible to coat synthetic peptides directly
onto a microtiter plate, but it involves testing each
Correspondence to: J. S~nderg~trd-Andersen, Mycoplasma
Lab., Neisseria Dept., Bygn. 7, Statens Seruminstitut, Amager
Boulevard 80, DK-2300 Copenhagen S, Denmark.
and ionic strength for optimal binding to the solid
phase (Geerlings et al., 1988). Still the method
does not work equally well for all peptides, and
loss of peptides during the ELISA procedure can-
not be excluded. Therefore, when synthetic
peptides are applied directly to a microtiter plate
and reacted with antibodies, it is difficult to verify
if a negative result is due to lack of binding
between peptide and antibody, or to a lack of
primary attachment of the peptide to the surface.
Radioactively labelled peptides have been used to
test for attachment, but besides the drawbacks of
working with radioactive materials, a considerable
workload is involved when dealing with hundreds
or thousands of peptides as described by some
workers (Geysen, 1985).
Another approach is to use peptides coupled to
a carrier like bovine serum albumin (BSA) which
is known to attach to the solid phase (Shirahama
et al., 1985). Besides being time consuming, it is

0022-1759/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

100
difficult to control how the peptide is coupled to
the carrier. Results may vary according to whether
the coupling is to a terminal amino acid, to one or
more of the side chains, or whether the coupling
method itself has modified functional groups of
the peptide (Briand et al., 1985).
Synthesis of peptides directly onto the solid
support has also been used (Geysen et al., 1984).
This has obvious advantages in comparison with
other methods, since peptides are coupled to the
surface in a controlled manner. By this technique
overlapping peptides may be synthesized in a single
microtiter plate, and since peptides are covalently
coupled to the surface, bound antibodies can be
removed, and the peptides retested with a new set
of antibodies. However, by this method the iden-
tity and purity of the peptides cannot be verified
directly.
The need for simple and reliable methods to
covalently couple purified peptides to a surface in
a controlled manner seems obvious. A method
where peptides are covalently coupled to nitrocel-
lulose has recently been described (Lauritzen et
al., 1990). In the present report we wish to de-
scribe a general method by which peptides can be
coupled covalently via the carboxyl group to mi-
crotiter plates (CovaLink, Nunc, Roskilde, Den-
mark).
Materials and methods
Immunoplates
Nunc Immuno Modules, CovaLink NH, cat.
no. 478042, were used. These plates have been
made from polystyrene plates by covalent attach-
ment of the linker shown in Fig. 1 (5). A total
activated surface area corresponding to 100 #1
O O
EDc
RCOOH * N-OH • -OCOR
3 plates, which were then incubated for 30 min at
a \b
1 2 4
CH3 ~ CH3
5 6
Fig. 1. Covalent coupling of peptide to CovaLink. EDC:
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide.
tilling volume, which again corresponds to ap-
proximately 1 cm 2, is available. The number of
linking groups on the surface has been radioac-
tively estimated to approximately 1014 g r o u p s / c m 2
corresponding to 1.67 x 10 -a° m o l / c m 2 (patent
pending).
Peptides
The angiotensin II peptide, with the sequence
H - A s p - A r g - V a l - T y r - I l e - H i s - P r o - P h e - O H , was
purchased from Sigma, and the HIV-2 peptide
with the sequence H-Leu-Asn-Ser-Trp-Gly-Cys-
Ala-Phe-Arg-Gln-Val-Cys-OH ( G n a n n et al.,
1987b) was from Cambridge Research Biochem-
icals, London, U.K.
A n tisera
Anti-angiotensin II antiserum production in
rabbits was performed as described previously by
coupling angiotensin to the thyroglobulin carrier
(Sofreniev et al., 1978). Antiserum and prevaccina-
tion serum from rabbit K-529 was used. Human
serum reactive against the HIV-2 peptide was
from a confirmed HIV-2 seropositive patient
(Poulsen et al., 1989). Normal human serum was
from a healthy, anti-HIV-1 and anti-HIV-2 nega-
tive person.
Coupling of peptides to CovaLink plates
To 5 /tl of the HIV-2 peptide (5.0 m g / m l in
H 2 0 ) were added 5 ~1 of 0.1 M aqueous N-hy-
droxysuccinimide (NHS) and 1-ethyl-3-(3-dimeth-
ylaminopropyl)carbodiimide, HC1 (EDC), both
from Sigma. To 18/tl of the angiotensin II peptide
(1000 # g / m l in H 2 0 ) were added 1 8 / d of N H S +
EDC. After 30 min at room temperature the
activated peptides were diluted from 20 to 0.001
/~g/ml in icecold 0.1 M carbonate buffer p H 8.6
and 100 /~1 applied per well of the CovaLink
4 o C, adapted from an earlier coupling procedure
(Cuatrecasas and Parikh, 1972). To test for the
ability of the peptides to bind noncovalently to
the plates, distilled water was added to the peptides
instead of N H S + EDC in the activation step. As
a control for unspecific binding of antibodies to
the plates, N H S + EDC was added to distilled
water instead of peptide in the activation step.

ELISA procedure
The plates were washed four times in PBS p H
7.2, postcoated with PBS p H 7.2 with 5% ( w / v )
skimmed milk powder (PBS-M) for 40 min at
r o o m temperature on a rocking table, washed four
times and incubated rocking at 3 7 ° C for 1.5 h
with antisera diluted in PBS-M. H u m a n anti-HIV-
2 serum and normal h u m a n serum were diluted
from 1 / 5 0 to 1/400. Rabbit anti-angiotensin II
serum and prevaccination serum were diluted
1 / 1 0 0 to 1/32,000 and 1/100 to 1/2000, respec-
tively. Plates were washed four times and in-
cubated rocking for 45 min at 3 7 ° C with per-
oxidase labeled anti-antibodies (P214 (anti-human
IgG) and P217 (anti-rabbit Ig), both Dakopatts,
Denmark) diluted 1/1000 in PBS-M. The plates
were washed four times, and colour developed for
10 min with o-phenylenediamine (0.5 m g / m l ) in
citrate buffer p H 5.0 with 0.05% H202. Color
development was stopped with 150/~1 1 N H2SO 4.
The plates were read on an I m m u n o R e a d e r (Inter-
Med N J-2000) at 490 nm. In addition to con-
centrations of peptide and serum, the following
parameters were varied during optimalization of
the test: (a) the amount of N H S + E D C in the
final solution added to the plates; (b) the duration
of peptide activation with N H S + EDC; (c) the
3000] , A
2500I
~oo~~
10 5 1 0.1 0.01 0.001
Peptide conc. in j~g/ml
Fig. 2. A: rabbit serum diluted 1/1000 and angiotensin II peptide in serial dilutions. B: angiotensin II peptide 1 #g/ml and serial
dilutions of rabbit serum. © ©, rabbit anti-angiotensin II serum against covalently coupled peptide. []
anti-angiotensin II serum against non-covalently bound peptide. •
without peptide. • O, normal rabbit serum against covalently coupled peptide. OD: optical density. Standard deviation is
indicated on A (n = 6).
101
activation and coupling of peptide by N H S and
E D C alone; (d) the temperature at which peptide
was coupled to the plates; (e) the duration of
peptide coupling to the plates; (f) the type of
coupling buffer; (g) the addition of Tween 20 to
the washing buffer and PBS-M.
Results
Angiotensin H
Checkerboard titrations of angiotensin II
peptide against rabbit anti-angiotensin II serum
showed, that 1 t t g / m l peptide (Fig. 2A) and a
1/1000 dilution of antiserum (Fig. 2B) gave an
optimal distinction between covalently coupled
peptide, and unspecifically bound peptide. The
background from antiserum bound non-covalently
to plates without peptide was minimal, and nor-
mal rabbit serum showed no reaction at all with
the peptide.
Each well contains 1014 functional groups or
1.67 × 10-1° mol on approximately 1 cm 2 corre-
sponding to a filling volume of approximately 100
/~1 (see above). With an average molecular weight
of the peptide of 1000, 0.167/~g/100/~1 of peptide
would be needed to occupy all functional groups
250O , i , , r
8
2ooc
1500
i
1000
5OO
1/1-00 1/5-00 1/1000 112000 1/,~000 1/1~)00 1/1~000 1132000
Antibody dilution
O, rabbit
•, rabbit anti-angiotensin II serum against immunoplates
102
in the well. The highest concentration of peptide
used was 10/xg/ml, and at this concentration the
dose-response curve seemed to flatten (Fig. 2A).
Therefore a surplus of peptide was needed to
occupy all functional sites in the well, but die
non-covalent binding of peptide was high at this
concentration. However, at our working dilution
of 1/~g peptide/ml, not all functional sites would
be occupied.
When the duration of peptide activation by
NHS + EDC exceeded 30 min, reactivity of the
activated peptide dropped. Coupling of peptide to
the CovaLink plates was optimal with 30 min at
4°C. If coupling of peptide was performed at
room temperature a higher background from
peptide binding non-covalently to the plates was
observed.
When the 0.1 M carbonate buffer pH 8.6 used
for coupling of peptide to the plates was sub-
stituted with PBS pH 7.2 or carbonate buffer pH
9.6, a drop in the reactivity of the plates was
observed. The activating solution (peptide and
NHS + EDC) should be diluted preferably 100
times in carbonate buffer. Otherwise the higher
concentration of NHS + EDC would result in in-
creased background from unspecific binding of
serum to the plates.
It was observed, that peptide, that had not been
activated with NHS + EDC, could bind noncova-
lently to the plates. The effect of Tween 20 on the
noncovalently bound peptide was investigated.
When the microtiter plates were treated with PBS
pH 7.2 with 0.05-1% Tween 20 before coupling of
peptide, the plates lost their reactivity. The reac-
tivity observed with covalently bound peptide in-
creased when 0.05% Tween 20 was added to the
washing buffer and PBS-M, but the reactivity of
non-covalently bound peptide increased equally.
Tween 20, therefore, did not give a better distinc-
tion between covalently and non-covalently bound
peptide. If Tween 20 was added only to the wash-
ing buffer applied just after coupling of peptide to
the plates, the same general increase in reactivity
was observed.
After coupling of peptide to the plates, they
may be washed, dried, and stored in a refrigerator
overnight without loss of reactivity.
Earlier results with Nunc immunoplates Maxi-
Sorp showed an almost complete desorption of
2500
2ooo~ T
0
20 10 1 (~1 0.01 0001 0.0001
Peptide ¢on¢. in jJg/ml
Fig. 3. Serum from an HIV-2-infected individual diluted 1/100
and serial dilutions of HIV-2 peptide, o o, anti-HIV-2
serum against covalently coupled peptide. [] n, anti-
HIV-2 serum against non-covalently coupled peptide.
• • , normal human serum against covalently coupled
peptide. • • , anti-HIV-2 serum against immunoplates
without peptide. OD: optical density. Standard deviation is
indicated on the figure (n = 6).
angiotensin II and a C terminal extension of
angiotensin II the decapeptide called angiotensin I
from the plates during the ELISA procedure. These
peptides were desorbed after having been coated
to the plates overnight using the carbonate buffer
pH 9.6.
HIV-2
By checkerboard titrations of HIV-2 peptide
and the serum from an HIV-2 infected individual,
it was shown, that 10-20 /tg peptide/ml and a
1/100 dilution of serum gave the best distinction
between covalently coupled and non-covalently
bound peptide (Fig. 3). As for angiotensin II the
addition of Tween 20 to the buffers increased the
overall signal, but did not give a better distinction
between covalently and non-covalently bound
peptide.
Discussion
In this paper we have described a general
method by which peptides containing a carboxylic

acid group, terminal or side chain, are coupled
covalently to CovaLink microtiter plates.
The secondary amino group, to which the
peptide is coupled, is placed at the end of a spacer
(Fig. 1 (5)). Besides the covalent linkage the spacer
includes 11 carbon atoms and 3 O/NH groups
and is relatively hydrophilic. If the spacer is as-
sumed to be fully extended into the aqueous sur-
roundings then the N H group is estimated to be
approximately 2 nm from the polystyrene surface
wall of the well. This, together with the specific
manner in which peptides are coupled to the func-
tional group, increases the probability that anti-
bodies specific for the peptides will bind to them.
If peptides were absorbed onto the surface in an
unspecific way, as is the case with the commonly
used methodology, the probability exists, that
specific antibodies would be unable to bind be-
cause of steric hindrance or inability of the
peptides to assume a conformation necessary for
binding. With the present method we cannot ex-
clude, of course, that highly lipophilic covalently
bound peptides will still have a tendency to bind
by hydrophobic forces to the polystyrene surface.
We should also point out, that the presence of
side chain carboxylic acid groups, due to the amino
acids aspartic acid and glutamic acid, will compete
with the terminal carboxyl acid group during co-
valent coupling. Therefore, even though all
peptides will be covalently coupled to the Co-
vaLink plates by our method, peptides that have
one or more side chain carboxylic acid groups will
probably be presented to antibodies in more than
one way in the same well, and this should be taken
into consideration when examining the results.
To achieve optimal coupling of peptide to the
CovaLink plates and optimal reaction with anti-
serum, certain steps in the procedure are im-
portant.
Coupling of peptide (Fig. 1 (1)) is obtained by
using w a t e r soluble 1-ethyl-3-(3-dimethyl-
aminopropyl)carbodiimide (EDC (3)) as coupling
agent in the presence of N-hydroxysuccinimide
(NHS (2)). This procedure has been used previ-
ously for activation and coupling of peptide seg-
ments, where absence of N-hydroxysuccinimide
led to a partially racemized product, and excess of
both this reagent and the water soluble carbodii-
mide lowered the yield (Sheppard, 1980). In the
103
present investigation we have observed that
activation with EDC alone gives rise to a much
lower covalent coupling (6) than in the presence
of NHS. Clearly, in order to achieve an optimal
formation of the activated, intermediate peptide
N-hydroxysuccinimide ester (4), the N H S + EDC
concentration must not be too low during the
activation step. On the other hand it must not be
too high in the solution when peptide is coupled
to the plates, since we have observed, that this
results in an increased background from antibod-
ies binding unspecifically to the plates. Thus, high
initial concentrations of peptide, N H S and EDC is
used followed by dilution in carbonate buffer prior
to peptide coupling to the plate. Furthermore,
activation of peptide should not exceed 30 min as
reactivity drops with longer incubation time, most
likely due to hydrolysis of the activated ester (4).
Optimal coupling of peptide was achieved with 30
min incubation. That covalent attachment of the
peptide to the plate has taken place is demon-
strated by the fact, that more peptide is bound to
the plate when covalent conditions are used con-
trary to non-covalent conditions (Fig. 2A).
At high concentrations of peptide there was a
considerable non-covalent binding of peptide to
the plates. To achieve a good distinction between
peptide covalently coupled to the plates, and
peptide that bound non-covalently, checkerboard
titrations of peptide and antiserum should be used
(Figs. 2A and 2B). Also, coupling of peptide to
the plates at 4 o C, instead of at room temperature,
resulted in a better distinction, as the reactivity of
noncovalently bound peptide increased more than
the reactivity of covalently bound peptide at room
temperature. With patients' sera other than those
mentioned above, we observed that with Tween 20
added to all of the buffers certain of these sera
gave a very high background by binding unspecifi-
cally to the plates (unpublished results). This is in
contrast to the usual effect of non-ionic detergents
like Tween 20, which will normally reduce the
nonspecific binding of antibodies to microtiter
plates. Thus, if Tween 20 is added to one or more
of the buffers, it is of paramount importance to
include tests for background of each serum.
The use of the CovaLink plates for testing the
reaction between synthetic peptides and antiserum
has several advantages. First of all the risk of
104
w a s h i n g o u t the c o v a l e n t l y b o u n d p e p t i d e is a b -
sent. S e c o n d l y , if the p e p t i d e is a t t a c h e d to a
m i c r o t i t e r p l a t e via a p r o t e i n c a r r i e r to test for
r e a c t i v i t y w i t h specific a n t i b o d i e s , this carrier, as
well as the l i n k e r u s e d for c o u p l i n g , m u s t b e
d i f f e r e n t f r o m the o n e s u s e d for i m m u n i z a t i o n , as
a n t i b o d i e s will b e p r o d u c e d a g a i n s t b o t h ( B r i a n d
et al., 1985; G e e r l i g s et al., 1988). T h u s the n e e d
for c o u p h n g to a c a r r i e r for the E L I S A is
ehminated with the present methodology. Thirdly,
the m e t h o d is h i g h l y r e p r o d u c i b l e a n d has a sensi-
t i v i t y c o m p a r a b l e to p r e s e n t l y u s e d m e t h o d s
( G n a n n et al., 1987a; G e e r l i g s et al., 1988). F i -
n a l l y , as the p e p t i d e s are c o v a l e n t l y c o u p l e d , it
m a y b e p o s s i b l e to elute b o u n d a n t i b o d i e s a n d
retest the p e p t i d e s w i t h a n e w set o f a n t i b o d i e s .
References
Arnon, R. (Ed.) (1987) Synthetic Vaccines, Vols. I and II. CRC
Press, Boca Raton, FL.
Briand, J.P., Muller, S. and Van Regenmortel, M.H.V. (1985)
Synthetic peptides as antigens: pitfalls of conjugation
methods. J. Immunol. Methods 78, 59.
Cuatrecasas, P. and Parikh, I. (1972) Adsorbents for affinity
chromatography. Use of N-hydroxysuccinimide esters of
agarose. Biochemistry 11, 2291.
Geerhgs, H.J., Weijer, W.J., Bloemhoff, W., Welling, G.W. and
Welling-Wester, S. (1988) The influence of pH and ionic
strength on the coating of peptides of Herpes simplex virus
type 1 in an enzyme-linked immunosorbent assay. J. Im-
munol. Methods 106, 239.
Geysen, M.H. (1985) Antigen-antibody interactions at the
molecular level: adventures in peptide synthesis. Immunol.
Today 6, 364.
Geysen, M.H., Meloen, R.H. and Barteling, S.J. (1984) Use of
peptide synthesis to probe viral antigens for epitopes to a
resolution of a single amino acid. Proc. Natl. Acad. Sci.
U.S.A. 81, 3998.
Gnann, Jr., J.W., Schwimmbeck, P.L., Nelson, J.A., Truax,
A.B. and Oldstone, M.B.A. (1987a) Diagnosis of AIDS by
using a 12-amino acid peptide representing an im-
munodominant epitope of the human immunodeficiency
virus. J. Infect. Dis. 156, 261.
Gnann, Jr., J.W., McCormick, J.B., Mitchell, S., Nelson, J.A.
and Oldstone, M.B.A. (1987b) Synthetic peptide im-
munoassay distinguishes HIV type 1 and HIV type 2 infec-
tions. Science 237, 1346.
Goudsmit, J. (1988) Immunodominant B-cell epitopes of the
HIV-1 envelope recognized by infected and immunized
hosts. AIDS 2 (suppl. 1), $41.
Hodges, R.S., Heaton, R.J. and Parker, J.M.R. (1988)
Antigen-antibody interaction. Synthetic peptides define lin-
ear antigenic determinants recognized by monoclonal anti-
bodies directed to the cytoplasmic carboxyl terminus of
rhodopsin. J. Biol. Chem. 263, 11768.
Lauritzen, E., Masson, M., Rubin, I. and Holm, A. (1990) Dot
immunobindingand immunoblotting of small peptides onto
activated nitrocellulose. Submitted.
Merrifield, R.B. (1963) Solid phase peptide synthesis I. The
synthesis of a tetrapeptide. J. Am. Chem. Soc. 85, 2149.
Poulsen, A.G., Kvinesdal, B., Aaby, P., Molb~ek, K., Frederik-
sen, K., Dias, F. and Lauritzen, E. (1989) Prevalence and
mortality from human immunodeficiency virus type 2 in
Bissau, West Africa. Lancet, ~pril 15, 827.
Sheppard, R.C. (1980) In: E. Gross and J. Meienhofer (Eds.),
The Peptides. Analysis, Synthesis, Biology, Vol. 2. Marcel
Dekker, New York, p. 463.
Shirahama, H. and Suzawa, T. (1985) Adsorption of bovine
serum albumin onto styrene/acrylic acid copolymer latex.
Colloid Polym. Sci. 263, 141.
Sofreniev, M.V., Madler, M., MUller, A.O. and Seriba, P.C.
(1978) A method for the consistent production of high
quality antisera to small peptide hormones. Frezenius 2.
Anal. Chem. 290, 163.