Download Anti-c-myc antibody 9E10 - Protein Engineering, Design and Selection

Protein Engineering vol.14 no.10 pp.803–806, 2001
Anti-c-myc antibody 9E10: epitope key positions and variability
characterized using peptide spot synthesis on cellulose
K.Hilpert1, G.Hansen1, H.Wessner1, G.Küttner1,
K.Welfle2, M.Seifert3 and W.Höhne4
1Institut
für Biochemie, Universitätsklinikum Charité, Humboldt-Universität
zu Berlin, Monbijoustr. 2, 10117 Berlin, 2Max-Delbrück-Centrum für
Molekulare Medizin, Robert-Rössle-Straβe 10, 13122 Berlin, 3Institut für
Medizinische Immunologie, Universitätsklinikum Charité,
Humboldt-Universität zu Berlin, Monbijoustr. 2, 10117 Berlin, Germany
4To
whom correspondence should be addressed.
E-mail: [email protected]
The 9E10 antibody epitope (EQKLISEEDL) derives from
a protein sequence in the human proto-oncogen p62c-myc
and is widely used as a protein fusion tag. This myc-tag is
a powerful tool in protein localization, immunochemistry,
ELISA or protein purification. Here, we characterize the
myc-tag epitope by substitutional analysis and length variation using peptide spot synthesis on cellulose. The key
amino acids of this interaction are the core residues LISE.
The shortest peptide with a strong binding signal is KLISEEDL. Dissociation constants of selected peptide variants
to the antibody 9E10 were determined. scFv constructs with
the shortest possible myc-tags were successfully detected by
Western blot and ELISA, giving a signal comparable to
that of the original myc-tag.
Keywords: epitope characterization/MAB 9E10/myc-tag/
peptide spot synthesis/substitutional analysis
Introduction
The monoclonal antibody 9E10, obtained by immunizing a
BALB/c mouse with a peptide from the C-terminal region
(amino acid 408–439) of the human c-myc protein (Evan et al.,
1985), is an immunoglobulin of the subclass IgG1/κ that binds
specifically to human c-myc. The c-myc gene product, a 62
kDa proto-oncoprotein (p62c-myc), is mainly localized in the
nucleus of the cell and its expression levels are elevated in a
variety of hematopoietic tumors. Therefore, p62 is extensively
used in cancer research and it has been proven to be a diagnostic
parameter in various types of cancer (Fuchs et al., 1997).
The 9E10 epitope has also become a well-known affinity
tag used in recombinant protein expression systems such as
bacteria, yeast or insect cells. Together, this myc-tag and the
9E10 antibody can be used to detect recombinant proteins by
immuno-staining of blots or in ELISAs, or to determine the
localization of a protein in a cell. In addition, a protein labeled
with the myc-tag can be purified in a single-step procedure
using a 9E10 antibody-coupled affinity column (Kramer et al.,
1997). Numerous publications describe the successful application of this protein fusion tag (Hoogenboom et al., 1991;
Dubel et al., 1993, 1995; Kleymann et al., 1995; Kontermann
et al., 1995; Manstein et al., 1995; TerBush and Novick, 1995;
Bach et al., 1996; Tanaka et al., 1996; Schiweck et al.,
1997). The myc-tag amino acid sequences commonly found
in practical use are: EQKLISEEDLN (Hoogenboom et al.,
1991) and EEQKLISEEDL (Fuchs et al., 1997). The Fv
© Oxford University Press
fragment of the monoclonal antibody 9E10 has been sequenced
and also cloned and expressed in Fv, Fab and scFv formats
(Fuchs et al., 1997; Schiweck et al., 1997).
Materials and methods
Peptide synthesis
Cellulose-bound peptides for binding experiments were prepared using a pipetting robot (Abimed, Langenfeld, Germany)
or by pipetting by hand using Whatman 50 cellulose membranes
(Whatman, Maidstone, UK), as described previously (Kramer
et al., 1994; Kramer and Schneider-Mergener, 1998). Deviating
from these protocols, the protection group cleavage procedure
was changed as follows: incubation of the membrane for 3 h in
90% TFA (Fluka, Deisenhofen, Germany), 2% triisopropylsilan
(Lancaster Synthesis GmbH, Mühlheim am Main, Germany),
5% water and 3% phenol (Fluka), change of the solution after
1 h without washing, shaking gently. The peptides are coupled
to the cellulose support by their C-terminus with two β-alanine
as a spacer in between. The peptides used for Kd determination
were prepared by Interactiva (Ulm, Germany), or in the
laboratory of J.Schneider-Mergener according to standard
Fmoc machine protocols using a multiple peptide synthesizer
(Abimed).
Binding experiments with cellulose-bound peptides
The cellulose membranes were washed with 96% ethanol for
5 min and equilibrated with Tris-buffered saline (20 mM Tris–
HCl and 0.5 M NaCl, pH 7.5) containing 0.01% Tween 20
(TTBS). The membranes were blocked by incubation with
blocking agent (Genosys, Cambridge, UK) in 0.1 M Tris
buffer, pH 7.5, for 2 h at room temperature, washed three
times with TTBS for 3 min each. After that the membrane
was incubated in a solution containing 2 µg/ml 9E10, a
sheep anti-mouse IgG horseradish peroxidase (HRP) conjugate
(Amersham, Freiburg, Germany) (dilution 1:500) and 3%
Gelifundol (Biotest Pharma GmbH, Dreieich, Germany) in
TTBS for 2 h at room temperature or overnight at 4°C. The
membranes were finally washed with TTBS three times for
10 min and then incubated with a chemiluminescence substrate
(Pierce, IL, USA) for 3 min or incubated in diaminobenzidine
solution. The light arising from spots with bound 9E10 antibody
(in the case of the chemiluminescence reaction) was measured
with a chemiluminescence detector (Boehringer, Mannheim,
Germany). A diaminobenzidine solution containing 3,3⬘-diaminobenzidine tetrahydrochloride (0.8 mg/ml) (Sigma,
Deisenhofen, Germany) in 0.1 M Tris–HCl, pH 7.5, with 0.2%
NiCl2 and 0.03% H2O2 was used to detect peroxidase activity
(in the case of Figure 1a).
The 9E10 antibody was produced and purified from
hybridoma cell line 9E10 in our group.
scFv fragment construction, recombinant expression and
purification
The single chain antibody fragment 1F9 specifically binding
to turkey egg white lysozyme (Küttner et al., 1998; Ay et al.,
803
K.Hilpert et al.
Fig. 1. Substitutional matrix of (a) the original and (b) a shortened myc-tag
peptide. The first column on the left represents peptides with the wild-type
sequence: EQKLISEEDLN (a) and KLISEEDL (b). The lines represent
peptides substituted at that position by 19 coded amino acids. The bound
antibody 9E10 was detected by HRP-labeled anti-mouse IgG and incubation
in a diaminobenzidin solution (a) or with a chemiluminescence substrate (b).
The light signal from (b) is shown as a negative image, the strongest
binding resulting in the darkest spot.
2000) was used as a model protein. Base sequences coding
for two shortened myc-tags (LISEEDL and LISEFEL) containing an EcoRI restriction site at the C-terminus were
introduced by PCR. These tags were fused to the C-terminus
(framework IV—KLEIK) of the scFv. The gene for the antibody
fragment was recloned then into pHEN 1 (Hoogenboom et al.,
1991) using SfiI and EcoRI for digestion. The mutant constructs
were transformed into Escherichia coli TG 1 cells and screened
for expression of the scFv product. The correct nucleotide
sequence of all constructs was confirmed by sequencing using
the dideoxy method (Sanger et al., 1977). The cells were
grown overnight in 2⫻ TY medium containing ampicillin and
1% glucose for catabolite repression. After centrifugation of
the cells the pellet was suspended in fresh medium containing
ampicillin and 0.1 mM isopropyl β-D-thiogalactopyranoside
(IPTG). The expression was continued for 4 h at 25°C. After
cell lysis the screening of total lysates was carried out by
Western blotting using the anti-myc-tag 9E10 antibody and an
anti-mouse HRP conjugate as the secondary antibody. 3,3⬘,5,5⬘Tetramethylbenzidine blue (TMB) (Seramun GmbH,
Dolgenbrodt, Germany) was used for the color development.
The clones showing the highest expression for both of the
mutants were used for scFv production. The cells were cultured
and induced as described above and the periplasmic fraction
was isolated by incubation of the cell pellet with 1/100 of the
culture volume using 0.1 M Tris–HCl pH 8.0, 0.5 M saccharose,
1 mM ethylenediamintetraacetic acid (EDTA). The periplasmic
fractions were dialyzed against 50 mM Tris–HCl pH 8.0, 0.15
M NaCl and bound to a BrCN Sepharose 6 column containing
coupled turkey egg white lysozyme. After washing with 0.2
M glycine, 0.2 M NaCl pH 5.0, the bound antibody fragment
was eluted by 0.2 M glycine, 0.2 M NaCl pH 2.0.
The pooled fractions were dialyzed against phosphate buffered saline and concentrated using Microcon 10 columns
(Millipore GmbH, Eschborn, Germany). The purified proteins
804
were diluted to the same concentration each (concentration
was measured using absorption at 280 nm) and the band
intensity of the Western blots was visually compared. The
Western blot analysis was carried out using 17% SDS polyacrylamide gels and blotting onto polyvinylidene fluoride
(PVDF) membranes. After blocking the membrane for 1 h
with 5% dry milk in TTBS and three times washing with
TTBS the membrane was incubated with 1 µg/ml 9E10
antibody and anti-mouse Fcgamma antibody-HRP conjugate
1:2000 dilution (Sigma GmbH, Taufkirchen, Germany) simultaneously in TTBS containing 5% gelatine for 1.5 h at room
temperature and washed three times with TTBS. For the color
development TMB blue was used.
Kd measurements with peptides
The microtiter plates were coated with the anti-pre-S2 (HBV)
scFv 1F6 containing the fused myc-tag sequence EQKLISEEDLN (Kuttner et al., 1999) (50 µl per well at 1.5 µg/ml)
overnight at 4°C with a 0.1 M carbonate buffer pH 9.6. The
binding of the 9E10 antibody (0.25 µg/ml) to this scFv was
competed by different concentrations of the peptides. The
bound antibody was detected by a sheep anti-mouse IgGHRP conjugate (Amersham) (dilution 1:1000). The Kd was
determined with the method described by Friguet (Friguet
et al., 1985).
Results and discussion
The myc-tag sequence EQKLISEEDLN included in vectors
such as the pHEN antibody phage display vector (Hoogenboom
et al., 1991) was used for epitope characterization of the 9E10
antibody. The C-terminal asparagine in this tag sequence,
which is a leucine at this position in the c-myc protein, is a
product of vector construction (Munro and Pelham, 1986). To
investigate the myc-tag/9E10 interaction we carried out a
substitutional analysis of the sequence EQKLISEEDLN using
the method of peptide spot synthesis on cellulose sheets (Frank,
1992; Kramer et al., 1994). Each amino acid of the peptide
was substituted by the other 19 coded amino acids to evaluate
the importance of each position of the sequence. The cellulose
sheet was incubated in 9E10 solution and after a washing
procedure the bound antibody was detected by a HRP-labeled
anti-mouse antibody (see Figure 1a).
Peptide positions where amino acids can only be exchanged
by a reduced set of, mainly physico-chemically similar, amino
acids are referred to as key positions; the results show that
these are localized in the core of the sequence. Leucine at
position 4 can only be substituted by itself, whereas isoleucine
(position 5) can also be replaced by physico-chemically similar
amino acids, and serine (position 6) by glycine, proline and
alanine. Furthermore, glutamic acid at position 7 can be
changed to glutamine and also to several hydrophobic amino
acids (phenylalanine, tryptophan, methionine, tyrosine) but not
to aspartic acid. Interestingly, the amino acids at positions 8
and 9 do not show interaction specificity, whereas leucine in
the next position (10) is again more specific. The introduction
of proline in any one of the positions from 3 to 10, with the
exception of position 6, is unfavorable for this interaction. The
substitution of serine at position 6 by proline or glycine may
indicate that the peptide adopts a turn conformation at the
antibody binding site.
A next step involved a truncation of the antibody epitope
by altering the length of the sequence EQKLISEEDLN synthesized on cellulose. The sequence was shortened C- or
Epitope variability of the anti-c-myc antibody 9E10
Table I. Affinity determination for peptides in solution
Peptide sequence
Kd⫻107 (M) (ELISA)
EQKLISEEDLN
KLISEEDL
FLISEEDL
QLISEEDL
KLISDEDL
KLISEFEL
QKLISEFELN
EQKLISEFELN
TMFLISEEDLQ
5.6 ⫾
260 ⫾
200 ⫾
820 ⫾
⬎2000
73 ⫾
14 ⫾
1.9 ⫾
12 ⫾
1.9
190
110
430
Kd values were measured by competition ELISA: for the original myc
peptide EQKLISEEDLN; a peptide with randomly chosen amino acids at
position 1, 2, and 11, and a change at position 3 from lysine to
phenylalanine; and different length variants of the core sequence of the
mutated myc-tag KLISEFEL. Additional peptide variants are included from
substitutional analysis of KLISEEDL.
Fig. 2. Western blot of purified scFv 1F9 variants with myc-tag fusions.
Shortened (snmyc) and shortened and mutated (mutmyc) myc-tags are
compared to sc-Fv containing the original myc-tag sequence
(EQKLISEEDLN). The scFv variants with the changed myc-tags have as
short a tag as possible, the last three amino acids from the light chain
serving as the first amino acids of the tag. The scFv with the original myctag has a spacer between the light chain and the myc sequence comprising
three alanine residues resulting from the vector construction. The scFvs
were purified and used at the following concentrations (determined
by absorption at 280 nm): 1 ⫽ 140, 2 ⫽ 70, 3 ⫽ 35, 4 ⫽ 17.5 and 5 ⫽
8.75 µg/ml. The scFv fusion proteins were detected by a combination of
9E10/anti-mouse Fcgamma HRP-labeled antibodies and visualized by the
color reaction of a HRP substrate.
N-terminally or at both ends (data not shown). Whereas the
binding signal decreases in the case of the peptide without
asparagine at position 11, the N-terminally shortened peptides
both without glutamic acid of position 1 and glutamine of
position 2 lead to a binding signal comparable to that of the
untruncated peptide. A drastic effect amounting to the complete
loss of detectable binding signal is associated with further
shortening by omitting the lysine at position 3. The shortest
peptide giving a strong binding signal, although weaker than
the complete 11-mer myc-tag sequence, is the 8-mer peptide
KLISEEDL. Additionally, we synthesized a substitutional
analysis for the truncated peptide to compare with the wildtype substitutional analysis (Figure 1b). The results show that
the key amino acids remain the same as in the wild-type myctag. Lysine at position 3, glutamic acid at position 7 and
leucine at position 10 (numbering according to the initial
11-mer peptide) appear to become more specific, since fewer
substitutions are tolerated. However, this is probably due to
the overall reduced binding affinity of the 8-mer (Table I),
thus diminishing the saturation effect and revealing intrinsic
differences between the signals. Therefore, some peptide
variants from the substitutional analysis of the shortened
peptide KLISEEDL were selected to measure the Kd values in
solution by competition ELISA: FLISEEDL with a binding
signal similar to the wild-type, QLISEEDL with a somewhat
lower binding signal and KLISEDEL with no binding signal
at all. The Kd values obtained (Table I) reflect the results from
the binding studies of substitutional analysis very well.
The substitutional analysis of the shortened peptide KLISEEDL was inspected for variants with higher affinity. Interestingly, stronger binding signals resulted from substituting
glutamic acid at position 8 by phenylalanine and aspartate at
position 9 by glutamate (Figure 1b). Substitutional analysis of
a new peptide with the glutamate at position 8 exchanged for
phenylalanine, KLISEFDL, again showed a slightly higher
binding signal at position 9 for the substitution of aspartic
acid to glutamic acid (data not shown). For the peptide
sequence exchanged both in positions 8 and 9, KLISEFEL,
the substitutional analysis did no more show any significantly
stronger binding signal (data not shown).
Several peptides with different lengths based on the optimized core sequence KLISEFEL were now synthesized and the
affinities of the soluble peptides measured by competition
ELISA (see Table I). The 8-mer peptide KLISEFEL showed
higher affinity than KLISEEDL, but, as expected, decreased
affinity to 9E10 compared to the original myc-tag EQKLISEEDLN, as did the 10-mer QKLISEFELN. However, the
11-mer EQKLISEFELN bound to the antibody somewhat
better than the original myc-tag. In addition, a peptide with
randomly chosen amino acids at positions 1, 2, and 11 and,
based on the substitution analysis results of the shortened myc
sequences, a change at position 3 from lysine to phenylalanine,
(TMFLISEEDLQ) was synthesized and its affinity to the 9E10
antibody was also determined (Table I). The affinity decreased
only by a factor of approximately 2 compared to the original
myc-tag. The results demonstrate that it is possible to substitute
positions 1, 2, 11 and with respect to the data from Figure 1b
also position 3.
The Kd determination showed that the original myc peptide
sequence EQKLISEEDLN has a 7-fold lower affinity to the
antibody 9E10 compared to the fluorescent Abz-EQKLISEEDLN peptide which was determined by fluorescence
titration to a Kd ⫽ 8 ⫾ 0.5⫻10–8 M (Schiweck et al., 1997).
Because of that deviation we performed an independent affinity
determination by isothermal titration calorimetry for the peptide
EQKLISEEDLN, resulting in a Kd of 5.3 ⫾ 0.3⫻10–7 M. This
is very close to the value obtained from the ELISA experiment.
The deviation between the fluorescence titration result and our
data may be due to an additional interaction between the
2-aminobenzoyl group and the antibody combining site thus
contributing to affinity. To look for specific interactions of
additional amino acids of a C- and N-terminally elongated
myc-tag, a substitutional array of the peptide AEEQKLISEFELLRK was synthesized and analyzed (data not shown), but
no further key amino acids or substitutionally sensitive positions were detected.
Taking into account the information accumulated so far we
designed single chain Fv fragments with a VH/VL orientation
where the last three amino acids of the scFv framework
correspond to the first three amino acids of the myc-tag. The
resulting tags were EIKLISEEDL, and the mutated myc-tag
EIKLISEFEL, where EIK derive from the light chain variable
region of the scFv and the N-terminal asparagine was omitted.
This means that the myc-tag of the recombinant protein is
kept as short as possible, thus minimizing antigenicity effects
and the influence of these residues on the protein itself. This
can be an advantage for example in protein crystallization,
measurements of intracellular protein function or activity, or
5.6
2.1
0.36
1.4
805
K.Hilpert et al.
in the case of recombinant proteins for therapeutical use.
Western blot analyses showed that based on the results of the
substitutional analysis it is possible to shorten and mutate the
original myc-tag without loss of sensitivity as determined by
a dilution series with the scFv variants of the 9E10 antibody
(Figure 2). A possible cross-reactivity between the murine
scFv and anti-mouse-HRP secondary antibody used for 9E10
detection was avoided by use of an anti-mouse Fcgamma-specific
secondary antibody.
The sensitivity of the different myc-tags against proteolyses
was analyzed by a 6-day incubation of the purified scFvs
mixed with a cytoplasmic extract of E.coli TG1 cells at 37°C.
No proteolytic cleavage of the three myc-tags was seen over
this time period as confirmed by Western blotting (data
not shown).
Our results show that is possible to modify and optimize
the commonly used myc-tag for specific applications using
information gained through substitutional analysis. For
example, based on our results Paschke et al. (Paschke et al.,
2001) developed a mutated myc-tag that contains a protease
cleavage site. Other options include substitutions leading to
reduced affinity, which may be an advantage in protein affinity
purification where more gentle elution conditions can then
be applied.
Acknowledgements
We thank D.Reinhardt for technical assistance, G.Zahn for advice and critical
reading of the manuscript and J.Schneider-Mergener for providing synthesized
peptides. This work was partially supported by a grant of the FAZIT Stiftung.
References
Ay,J., Keitel,T., Küttner,G., Wessner,H., Scholz,C., Hahn,M. and Höhne,W.
(2000) J. Mol. Biol., 301, 239–246.
Bach,M., Sander,P., Haase,W. and Reilander,H. (1996) Receptors Channels,
4, 129–139.
Dubel,S., Breitling,F., Fuchs,P., Braunagel,M., Klewinghaus,I. and Little,M.
(1993) Gene, 128, 97–101.
Dubel,S., Breitling,F., Kontermann,R., Schmidt,T., Skerra,A. and Little,M.
(1995) J. Immunol. Methods, 178, 201–209.
Evan,G.I., Lewis,G.K., Ramsay,G. and Bishop,J.M. (1985) Mol. Cell Biol., 5,
3610–3616.
Frank,R. (1992) Tetrahedron, 48, 9217–9232.
Friguet,B., Chaffotte,A.F., Djavadi-Ohaniance,L. and Goldberg,M.E. (1985)
J. Immunol. Methods, 77, 305–319.
Fuchs,P., Breitling,F., Little,M. and Dubel,S. (1997) Hybridoma, 16, 227–233.
Hoogenboom,H.R., Griffiths,A.D., Johnson,K.S., Chiswell,D.J., Hudson,P. and
Winter,G. (1991) Nucleic Acids Res., 19, 4133–4137.
Kleymann,G., Ostermeier,C., Heitmann,K., Haase,W. and Michel,H. (1995)
J. Histochem. Cytochem., 43, 607–614.
Kontermann,R.E., Liu,Z., Schulze,R.A., Sommer,K.A., Queitsch,I., Dubel,S.,
Kipriyanov,S.M., Breitling,F. and Bautz,E.K. (1995) Biol. Chem Hoppe
Seyler, 376, 473–481.
Kramer,A. and Schneider-Mergener,J. (1998) Methods Mol. Biol., 87, 25–39.
Kramer,A., Schuster,A., Reinecke,U., Malin,R., Volkmer-Engert,R.,
Landgraf,C. and Schneider-Mergener,J. (1994) Methods: Comp. Methods
Enzymol., 6, 388–395.
Kramer,A., Keitel,T., Winkler,K., Stocklein,W., Hohne,W. and SchneiderMergener,J. (1997) Cell, 91, 799–809.
Küttner,G., Keitel,T., Gieβman,E., Wessner,H., Scholz,C. and Höhne,W. (1998)
Mol. Immunol., 35, 89–194.
Küttner,G. et al. (1999) Mol. Immunol., 36, 669–683.
Manstein,D.J., Schuster,H.P., Morandini,P. and Hunt,D.M. (1995) Gene, 162,
129–134.
Munro,S. and Pelham,H.R. (1986) Cell, 46, 291–300.
Pascke,M., Zahn,G., Warsinke,A. and Höhne,W. (2001) Biotechniques, 30,
720–724.
Sanger,F., Nicklen,S. and Coulson,A.R. (1977) Proc. Natl Acad. Sci. USA,
74, 5463–5467.
Schiweck,W., Buxbaum,B., Schatzlein,C., Neiss,H.G. and Skerra,A. (1997)
FEBS Lett., 414, 33–38.
806
Tanaka,K., Nagayama,Y., Yamasaki,H., Hayashi,H., Namba,H., Yamashita,S.
and Niwa,M. (1996) Biochem. Biophys. Res. Commun., 228, 21–28.
TerBush,D.R. and Novick,P. (1995) J. Cell. Biol., 130, 299–312.
Received March 6, 2001; revised June 12, 2001; accepted July 18, 2001