A human kringle domain-based fluorescence

Analytical Biochemistry 451 (2014) 63–68
Contents lists available at ScienceDirect
Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio
A human kringle domain-based ﬂuorescence-linked immunosorbent
assay system
Gu Min Jeong a, Yong Sung Kim b, Ki Jun Jeong a,c,⇑
a
Department of Chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology (KAIST), Yuseong-gu, Daejeon 305-701, Republic of Korea
Department of Molecular Science and Technology, Ajou University, Yeongtong-gu, Suwon 443-749, Republic of Korea
c
Institute for the BioCentury, Korea Advanced Institute of Science and Technology (KAIST), Yuseong-gu, Daejeon 305-701, Republic of Korea
b
a r t i c l e
i n f o
Article history:
Received 6 November 2013
Received in revised form 4 January 2014
Accepted 31 January 2014
Available online 10 February 2014
Keywords:
Kringle domain
FLISA
GFP
Immunoassay
a b s t r a c t
As a non-immunoglobulin protein scaffold, human kringle domain (KD) has attractive properties such as
high speciﬁcity, stability, and production in bacterial hosts. Here, we developed a rapid and sensitive
ﬂuorescence-linked immunosorbent assay (FLISA) system using a ﬂuorescent kringle domain (ﬂuoKD),
a fusion protein of a green ﬂuorescent protein (GFP), and a kringle domain variant (KD548). Two
kinds of ﬂuoKDs in which KD was fused to the N terminus of GFP (N-ﬂuoKD) or the C terminus of GFP
(C-ﬂuoKD) were constructed and characterized. In Escherichia coli host, both ﬂuoKDs were produced in
high yield and solubility and were successfully puriﬁed by a simple procedure. The puriﬁed ﬂuoKDs
exhibited strong ﬂuorescent activities and high afﬁnities to the target antigen. Furthermore, it was
successfully demonstrated that the FLISA with puriﬁed ﬂuoKDs allowed for more rapid detection of target
antigens with higher sensitivity compared with conventional enzyme-linked immunosorbent assay
(ELISA), indicating that a simple, rapid, and sensitive immunoassay system could be developed by using
KD instead of antibody or antibody fragments.
Ó 2014 Elsevier Inc. All rights reserved.
Currently, enzyme-linked immunosorbent assays (ELISAs)1 are
widely used for the detection of speciﬁc molecules, including proteins, peptides, and sugars, as well as various other small molecules
and chemicals. In addition, the immunosorbent assay is an important tool in most protein-based diagnostic platforms [1]. In most
immunosorbent assay systems, antibodies are used as capturing reagents because of their high speciﬁcity and afﬁnity to target molecules. However, the intrinsic properties of antibodies also create
several limitations in the immunosorbent assay systems. In many
cases, antibodies are produced in a mammalian host and the production cost is high due to their low productivity and use of expensive
⇑ Corresponding author at: Department of Chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology (KAIST), Yuseong-gu,
Daejeon 305-701, Republic of Korea. Fax: +82 42 350 3910.
E-mail address: [email protected] (K.J. Jeong).
1
Abbreviations used: ELISA, enzyme-linked immunosorbent assay; KD, kringle
domain; FLISA, ﬂuorescence-linked immunosorbent assay; DR5, death receptor 5;
ﬂuoKD, ﬂuorescent kringle domain; GFP, green ﬂuorescent protein; N-ﬂuoKD, ﬂuoKD
in which anti-DR5 KD548 was fused to the N terminus of GFP; C-ﬂuoKD, ﬂuoKD in
which anti-DR5 KD548 was fused to the C terminus of GFP; PCR, polymerase chain
reaction; IPTG, isopropyl-b-D-thiogalactopyranoside; PBS, phosphate-buffered saline;
PBST, PBS containing 0.05% Tween 20; PBSM, PBS containing 10 wt% skim milk; HRP,
horseradish peroxidase; TMB, tetramethylbenzidine; SDS–PAGE, sodium dodecyl
sulfate–polyacrylamide gel electrophoresis; LOD, limit of detection.
http://dx.doi.org/10.1016/j.ab.2014.01.019
0003-2697/Ó 2014 Elsevier Inc. All rights reserved.
media. Even though bacterial hosts can be used for production of
antibody proteins, the folding efﬁciency and puriﬁcation yield are often poor. In addition, the instability of antibodies in harsh conditions
makes them difﬁcult to use in diagnostic systems [2].
To overcome the limitations of antibodies, exploration of alternative protein reagents with the ability of speciﬁc binding to target
molecules has been stimulated and has led to the development of
non-immunoglobulin protein scaffolds. Compared with typical
antibodies, non-immunoglobulin protein scaffolds, used as an
alternative to antibody proteins, have several advantages: (i) smaller size (<100 amino acids), (ii) simple structure with high stability,
and (iii) economic production in bacterial or yeast hosts [3,4]. During the past decade, several types of protein scaffolds, including
DARPins, Afﬁbody, Avimers, and Fibronectin, have been developed
and successfully engineered for many applications in biotechnology and biomedical research [5]. Recently, we also reported the
development of a new non-immunoglobulin scaffold from human
plasminogen kringle domain 2 (PgnKD2) [6]. The kringle domain
(KD) is composed of 80 amino acids and contains a rigid core
structure with three conserved disulﬁde bonds. Compared with
other non-antibody protein scaffolds, KD has several distinctive
properties, including high thermal stability (active at 65 °C), higher
rigidity via formation of three disulﬁde bonds, and relatively long
64
Human kringle domain-based FLISA system / G.M. Jeong et al. / Anal. Biochem. 451 (2014) 63–68
variable regions in the seven ﬂexible loops ( 45 amino acid residues; see Fig. 1A). In particular, owing to the long variable regions,
KD can be engineered to multivalent variants that might simultaneously interact with two or more targets [7]. The typical structure
of KD is strongly sequence tolerant to mutation in the loop region;
therefore, it has the potential to be engineered for distinct
binding to various target molecules [6,7]. In addition, a functional
KD can be produced in Escherichia coli host [8] and yeast, including
Pichia pastoris and Saccharomyces cerevisiae [6,7], both of which
have high production yields (>5 g/L in E. coli host); therefore, the
low production cost may be more beneﬁcial when compared with
most antibody molecules. In most diagnostic systems based on
immunosorbent reaction, the sensitivity and stability are highly
dependent on the capturing molecules immobilized on the surface
of the diagnostic system. As described earlier, KD has ideal
properties for diagnostic systems, and its application toward the
development of diagnostic systems is highly desired.
In this study, we developed a ﬂuorescence-linked immunosorbent assay (FLISA) system using a KD variant (Fig. 1B). As a model
KD variant, KD548 was used and engineered to have high afﬁnity
to human death receptor 5 (DR5), a receptor protein present on
the surface of cancer cells [6]. For the ﬂuorescent signal, two
ﬂuorescent kringle domains (ﬂuoKDs) were constructed where
anti-DR5 KD548 was fused to the N terminus or C terminus of
green ﬂuorescent protein (GFP) (N-ﬂuoKD or C-ﬂuoKD, respectively) with a ﬂexible (Gly4Ser)2 linker between GFP and KD548.
After production and puriﬁcation in E. coli cultivation, both ﬂuoKDs
were applied to the FLISA system and their ﬂuorescent intensity
and binding activity were analyzed. We also successfully demonstrated that the ﬂuoKD-based FLISA system has high sensitivity
compared with a conventional ELISA system.
Materials and methods
Bacterial strains and plasmid
All bacterial strains used in this work are listed in Table 1. E. coli
XL1-Blue was used as a host cell for gene manipulation and plasmid maintenance. E. coli SHufﬂe Express, which is engineered to
promote disulﬁde bond formation in the cytoplasm by constitutive
expression of DsbC in cytoplasm, was used for production of both
ﬂuoKDs and GFP. Even though KD has three disulﬁde bonds, the
correctly folded KD can be produced in cytoplasm of this strain.
Polymerase chain reaction (PCR) was performed with the C1000
Thermal Cycler (Bio-Rad, Hercules, CA, USA) using PrimeSTAR HS
Polymerase (Takara Bio, Shiga, Japan). The nucleotide sequences
of all primers used in this study are listed in Table S1 of the online
Supplementary material. A GFP gene was ampliﬁed from pGFPmut2 [9] by PCR with primers CF-F4, CF-F5, CF-F6, and GFP-R1.
The PCR product was digested with restriction enzymes XbaI and
HindIII and then cloned into the same enzyme sites of pMoPac1
[10], yielding pGM-GFP. For the expression of N-ﬂuoKD in which
B
A
N-fluoKD
GFP
N
Flexible
linker
N C
C
Target
C-fluoKD
N-fluoKD
Fig.1. (A) Schematic structure of KD (PDB entry code: 1I5K). It has a rigid core via three disuﬁde bonds and seven ﬂexible loops. (B) Concept of KD-based FLISA system. The
ﬂuoKDs in which KD is linked to the N or C terminus of GFP can bind to target antigens with speciﬁc afﬁnity, and the fused GFP gives ﬂuorescent signal for detection of target
molecules.
Table 1
Bacterial strains and plasmids used in this study.
E. coli strains or plasmids
E. coli strains
XL1-Blue
BL21(DE3)
SHufﬂe Express
Plasmids
pMoPacI
pET21b-DR5
pPICZa-KD548
pGFPmut2
pGM-GFP
pGM-N-fKD
pGM-C-fKD
a
b
c
Description
References
(recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F0 proAB lacIqZDM15 Tn10 (Tetr)]
F– ompT gal dcm lon hsdSB(r-B m-B) k(DE3 [lacI lacUV5-T7 gene 1 ind1 sam7 nin5])
fhuA2 [lon] ompT ahpC gal katt::pNEB3-r1-cDsbC (SpecR, lacIq) DtrxB sulA11 R(mcr-73::miniTn10–TetS)2 [dcm]
R(zgb-210::Tn10-TetS) endA1 Dgor D(mcrC-mrr)114::IS10
Stratagenea
Novageneb
New England Biolabsc
ColE1 origin, Cmr, lac promoter, lacIq
6His tag-DR5
human KD548 gene
GFPmut2 gene
pMoPac1 derivative, 6His tag-GFP
pMoPac1 derivative, 6His tag-N-ﬂuoKD
pMoPac1 derivative, 6His tag-C-ﬂuoKD
[10]
[4]
[4]
[9]
This study
This study
This study
Stratagene, Santa Clara, CA, USA.
Novagene, Madison, WI, USA.
New England Biolabs, Ipswich, MA, USA.
Human kringle domain-based FLISA system / G.M. Jeong et al. / Anal. Biochem. 451 (2014) 63–68
KD548 was linked to the N terminus of GFP, KD548 coding gene
was ampliﬁed from pPICZa-KD548 [4] by PCR with primers NFF1, NF-F2, and NF-R2, and GFP coding gene was ampliﬁed from
pGFPmut2 by PCR with primers NF-F3 and NF-R3. Then, each
PCR product was mixed and N-ﬂuoKD gene was synthesized by
overlap PCR with primers NF-F2 and NF-R3. The resulting PCR
product in which KD548 and GFP genes were linked via (Gly4Ser)2
linker was digested with restriction enzymes XbaI and HindIII and
then cloned into pMoPac1, yielding pGM-N-fKD. For the expression
of C-ﬂuoKD in which KD548 was linked to the C terminus of GFP,
KD548 coding gene was ampliﬁed by PCR with primers CF-F7
and CF-R5, and GFP coding gene was ampliﬁed by PCR with primers CF-F4, CF-F5, CF-F6, and CF-R4. Then, each PCR product was
mixed and C-ﬂuoKD gene was synthesized by overlap PCR with
primers CF-F6 and CF-R5. The resulting PCR was digested with XbaI
and HindIII and then cloned into pMoPac1, yielding pGM-C-fKD.
The schematic diagram of plasmids used for this work is shown
in Fig. S1 of the Supplementary material. All DNA manipulations,
including restriction digestion, ligation, and agarose gel electrophoresis, were carried out using standard procedures [11].
65
FACS analysis of GFP-producing cell
The cultured cells were harvested by centrifugation at 4 °C and
6000 rpm for 10 min. The cells were washed with 1 PBS and applied to FACS analysis. MoFlo XDP (Beckman Coulter, Brea, CA,
USA) was used for FACS analysis. The ﬂuorescence reading was obtained using an excitation of 488 nm with an argon laser. The emission signal was measured in FL1 channel centered at 530/40 nm on
excitation. The sample mean ﬂuorescence intensity values and
images were analyzed using SUMMIT software version 5.2. Coulter
Isoton II Diluent (Beckman Coulter) was used in all experiments as
the ﬂow cytometry sheath ﬂuid.
Measurement of ﬂuorescence intensity of ﬂuoKD
After puriﬁcation of ﬂuoKDs and GFP, their ﬂuorescence intensities were measured. The same concentration (4 lM, 50 ll/well)
of puriﬁed GFP and ﬂuoKDs was loaded on the 96-well black plate
(Corning, Tewksbury, MA, USA). The ﬂuorescence intensities of
samples were measured by a microplate reader (Inﬁnite M200
PRO, TECAN, Grodig, Austria). The excitation and emission wavelengths used for detection were 488 and 535 nm, respectively.
Flask cultivation
Conventional ELISA
For the production of both ﬂuoKDs and GFP, E. coli SHufﬂe Express harboring each expression plasmid was inoculated and cultivated at 37 °C in liquid LB medium (BD, Franklin Lakes, NJ, USA)
with 2% glucose. After an overnight cultivation, 5 ml of the inoculums was transferred into 500 ml of fresh medium in a 2-L ﬂask
and incubated at 37 °C while shaking at 200 rpm. When the cell
density (OD600nm) reached approximately 0.6, cells were induced
with isopropyl-b-D-thiogalactopyranoside (IPTG; Sigma–Aldrich,
St. Louis, MO, USA) to a ﬁnal concentration of 1 mM. After induction, cells were further cultivated at 37 °C for 4 h.
For the production of DR5, E. coli BL21(DE3) harboring pET21bDR5 [4] was used. Cells were inoculated into the LB medium with
2% glucose and overnight cultured at 37 °C and 200 rpm. The next
morning, the cells were transferred to fresh LB medium and cultivated at 37 °C. When the cell density (OD600nm) reached approximately 0.6, cells were induced with IPTG to a ﬁnal concentration
of 0.5 mM. The cells were further incubated at 25 °C and 200 rpm
for 16 h, and then cells were harvested by centrifugation
(6000 rpm, 4 °C, 10 min).
Protein puriﬁcation
Following ﬂask cultivation, cells were harvested by centrifugation at 6000 rpm and 4 °C for 10 min. Then, cell pellets were resuspended in 40 ml of the binding buffer (50 mM phosphate–KOH and
300 mM NaCl, pH 8.0) and disrupted by three cycles of sonication
(each for 10 min of on-time at 20% of maximum output;
High-Intensity Ultrasonic Liquid Processors, Sonics & Material,
Newtown, CT, USA). After cell disruption, only the soluble lysate
was collected by centrifugation (10,000 rpm, 4 °C, 10 min) and
was incubated with 1 ml of Ni-NTA (nickel–nitrilotriacetic acid)
resin (GE Healthcare, Buckinghamshire, UK) for 1 h with shaking
at 4 °C. After washing the resin with washing buffer (50 mM phosphate–KOH, 300 mM NaCl, and 5 mM imidazole, pH 8.0), proteins
were eluted with elution buffer (50 mM phosphate–KOH,
300 mM NaCl, and 250 mM imidazole, pH 8.0). After dialysis
against phosphate-buffered saline (PBS: 135 mM NaCl, 2.7 mM
KCl, 4.3 mM Na2HPO4, and 1.4 mM KH2PO4, pH 7.2), the puriﬁed
proteins were stored at 4 °C for future use. The puriﬁed proteins
were quantiﬁed using Bradford assay solution (Bio-Rad). All protein puriﬁcations were carried out at 4 °C.
First, 2 lM, 50 ll/well KD or ﬂuoKDs were coated on the 96well plate at 37 °C for 1 h and washed with 200 ll/well PBST
(PBS containing 0.05% Tween 20) four times. In addition, those
wells were blocked by 200 ll/well PBSM (PBS containing 10 wt%
skim milk) at 37 °C for 1 h and washed with PBST four times. Next,
serially diluted DR5 was applied with 50 ll/well volume to the
wells and then incubated at 37 °C for 1 h. The wells were washed
with PBST four times, and then 100 ll/well of 1:2000 diluted
horseradish peroxidase (HRP)-conjugated anti-FLAG tag–antibody
was added. After 1 h of incubation, the wells were washed with
PBST four times and 50 ll/well of the TMB (tetramethylbenzidine)
peroxidase substrate (BD Biosciences, San Jose, CA, USA) was added
for peroxidase reaction. Finally, 50 ll of 2 M H2SO4 was added into
each well to stop the peroxidase reaction and, the absorbance at
450 nm was measured by a microplate reader (Inﬁnite M200 PRO).
Conventional FLISA
First, 3 lg of DR5 was coated on each well at 37 °C for 1 h, and
then wells were washed with 200 ll/well PBST four times. The
wells were blocked by 200 ll/well PBSM at 37 °C for 1 h and then
washed with PBST four times. Next, serially diluted ﬂuoKDs were
applied with 50 ll/well volume to the wells and then incubated
at 37 °C for 1 h. The wells were washed with PBST four times,
and ﬁnally 50 ll/well PBS was applied. The ﬂuorescent signal
was measured by a microplate reader (Inﬁnite M200 PRO). The
excitation and emission wavelengths were 488 and 535 nm,
respectively.
Competitive FLISA
First, wells were coated with 3 lg/well DR5 at 37 °C for 1 h and
then washed with 200 ll/well PBST four times. The wells were
blocked by 200 ll/well PBSM at 37 °C for 1 h and then washed with
PBST four times. During these steps, 25 ll of serial diluted DR5
(from 0.64 lg/ml) in PBS was incubated with 25 ll of each ﬂuoKD
(each 400 lg/ml) solution for 1 h at 37 °C. The mixed samples were
loaded on the DR5-coated wells. After incubation at 37 °C for 1 h,
each well was washed with 200 ll/well PBST four times to remove
the unbound ﬂuoKDs. Finally, 50 ll/well PBS was added into the
wells, and then the ﬂuorescence signal was measured by a
Human kringle domain-based FLISA system / G.M. Jeong et al. / Anal. Biochem. 451 (2014) 63–68
microplate reader (Inﬁnite M200 PRO). The excitation and emission wavelengths were 488 and 535 nm, respectively.
Competitive ELISA
Competitive ELISA was performed in microtiter well plates
(Nunc, Invitrogen, Carlsbad, CA, USA). First, the wells were coated
with 3 lg/well DR5 at 37 °C for 1 h and washed with 200 ll/well
PBST four times. The wells were blocked by 200 ll/well PBSM at
37 °C for 1 h and washed with PBST four times. During the steps
above, 25 ll of serial diluted DR5 (from 400 lg/ml) in PBS was
incubated with 25 ll of each ﬂuoKD (each 400 lg/ml) solution
for 1 h at 37 °C. The mixed samples were reacted with the wells
at 37 °C for 1 h, and the wells were washed with 200 ll/well PBST
four times. The washed wells were blocked again with PBSM at
37 °C for 1 h, and then the plates were reprobed by 100 ll/well
anti-GFP–HRP-conjugated antibody (ABCam, Cambridge, UK) in
PBST (1:2000) for 1 h at 37 °C. The plates were washed four times
by PBST, and 50 ll of the TMB peroxidase substrate was added for
peroxidase reaction. Finally, 50 ll of 2 M H2SO4 was used to stop
the peroxidase reaction. The absorbance at 450 nm was estimated
using an Inﬁnite M200 PRO microplate reader.
Results
Production and puriﬁcation of ﬂuoKDs
1
2
3
After puriﬁcation, protein concentrations were normalized and
then the ﬂuorescence intensity of each puriﬁed ﬂuoKD was analyzed and compared with that of wild-type GFP. Compared with
wild-type GFP, N-ﬂuoKD and C-ﬂuoKD exhibited relatively lower
intensities (82 and 72% of GFP intensity, respectively; see
Fig. 3A), but both were high enough to be used for ﬂuorometric
detection. Between both ﬂuoKDs, N-ﬂuoKD showed slightly higher
intensity than C-ﬂuoKD (1.1-fold). This result indicates that fusion
of KD to GFP at the N- and C-terminal ends did not have any deleterious effect on the ﬂuorescent activity of GFP. The binding activities of both ﬂuoKDs were also analyzed by ELISA and compared
with that of original KD (no fusion with GFP). We clearly observed
that all examined proteins except BSA (negative control) showed
very similar binding activities against DR5 (Fig. 3B). As described
in the introductory paragraphs, KD has a rigid core structure and
so the binding activity of KD to target molecules is not easily modiﬁed by the fusion with other proteins.
4
5 6
200
150
100
75
50
37
25
To conﬁrm the activities of N- and C-ﬂuoKDs, the conventional
FLISA was carried out. A 96-well black microtiter plate was coated
with DR5 (antigen of KD548), and serially diluted ﬂuoKDs (from
0.0128 to 40 ng) were loaded into each well and their ﬂuorescent
intensities were analyzed. In the case of N-ﬂuoKD, the ﬂuorescent
A
Fluorescence intensity
(Arbitrary unit)
M
Fluorescent intensity of ﬂuoKD
Conventional FLISA with ﬂuoKD
For the production of both ﬂuoKDs (anti-DR5 KD548 fused
GFPs), the constructed plasmids, pGM-N-fKD and pGM-C-fKD,
were transformed into E. coli SHufﬂe Express for production. After
ﬂask cultivation, the production yields and solubility of each ﬂuoKD were analyzed by sodium dodecyl sulfate–polyacrylamide gel
electrophoresis (SDS–PAGE). In the cytoplasm of E. coli, both ﬂuoKDs were highly expressed (N-ﬂuoKD, 17% of total proteins; C-ﬂuoKD, 8% of total proteins) with high solubility (>90%) (Fig. 2). The
production and ﬂuorescent signal intensities of both ﬂuoKDs in
E. coli cells were also analyzed by a ﬂow cytometer. Similar to
the above expression data (Fig. 2), N-ﬂuoKD also exhibited much
higher ﬂuorescent signal intensity than C-ﬂuoKD (see Fig. S2 in
Supplementary material). With cultured cells, puriﬁcation of both
ﬂuoKDs was carried out as described in Materials and Methods,
and we observed that both ﬂuoKDs with a molecular mass of
38 kDa were successfully puriﬁed (Fig. 2). From a 1-L culture,
(kDa)
approximately 20.5 mg of N-ﬂuoKD and 7.0 mg of C-ﬂuoKD could
be puriﬁed with high purity (>90%) and high recovery yields (40–
50%). Between both ﬂuoKDs, N-ﬂuoKD showed a relatively higher
expression level and ﬁnal production yield compared with C-ﬂuoKD (Fig. 2).
12000
10000
8000
6000
4000
2000
0
GFP
N-fluoKD
C-fluoKD
PBS
B
Absorbance at 450 nm
66
20
Concentration of DR5 (µ
µg/ml)
Fig.2. SDS–PAGE analysis of production and puriﬁcation of N-ﬂuoKD (lanes 1–3)
and C-ﬂuoKD (lanes 4–6). Lane M: molecular weight markers (kDa); lanes 1 and 4:
total fraction; lanes 2 and 5: soluble fraction; lanes 3 and 6: puriﬁed protein. Open
and closed arrowheads indicate N-ﬂuoKD and C-ﬂuoKD, respectively.
Fig.3. Characterization of ﬂuoKDs. (A) Measurement of ﬂuorescent intensities of
ﬂuoKDs. The same amount of GFP and ﬂuoKDs was applied to measurement, and
PBS without protein was used as negative control. (B) Measurement of binding
activity of ﬂuoKDs by ELISA. Symbols: j, KD only; d, N-ﬂuoKD; N, C-ﬂuoKD; ., BSA.
Competitive FLISA and competitive ELISA
To validate the application of KD-based FLISA toward an immunoassay tool, the competitive FLISA experiments were carried out
and compared with a conventional ELISA system. In both competitive ELISA and FLISA, N-ﬂuoKD or C-ﬂuoKD was incubated with
serially diluted target molecule (DR5) on a DR5-coated well plate.
After incubation and washing, the remaining ﬂuoKD bound to the
DR5 coated was analyzed by direct detection of ﬂuorescent intensity (FLISA) or by HRP-conjugated anti-GFP antibody (ELISA). In the
case of competitive FLISA using N-ﬂuoKD, the detectable range of
DR5 concentrations was from 100 pg/ml to 64 ng/ml (Fig. 5A).
The competitive FLISA using C-ﬂuoKD showed a slightly higher
detection range (from 500 pg/ml to 320 ng/ml) than that using
N-ﬂuoKD, but it also appeared to be quite sensitive. Competitive
ELISA was also performed to compare its sensitivity with that of
competitive FLISA. As shown in Fig. 5B, both ﬂuoKDs in competitive
ELISA showed much higher detection ranges (from 1.6 to 200 lg/ml)
A
Concentration of DR5 (µ
µg/ml)
B
Relative Signal Intensity (A/A 0)
intensity was positively correlated with the concentration of the NﬂuoKD in the range of 1.28–800 lg/ml (Fig. 4A). In the case of CﬂuoKD, the good correlation between concentration of C-ﬂuoKD
and ﬂuorescent intensity was observed in a similar range of concentration, although it exhibited less ﬂuorescent intensity than
N-ﬂuoKD (Fig. 4B). The relatively lower intensities of C-ﬂuoKD
might be related not to the binding activity of KD but rather to
the relatively lower ﬂuorescent intensity of GFP in the C-ﬂuoKD
format (Fig. 3A). In both cases, the use of ﬂuoKD exhibited a fully
distinguishable signal against speciﬁc antigen DR5 compared with
a negative control using BSA. Considering these results, we concluded that the fusion of KD and GFP did not cause any deleterious
effect on the ﬂuorescent activity of GFP or on the speciﬁc binding
activity of KD independent of fusion positions.
67
Relative Signal Intensity (F/F0)
Human kringle domain-based FLISA system / G.M. Jeong et al. / Anal. Biochem. 451 (2014) 63–68
Concentration of DR5 (µg/ml)
Fig.5. (A) Competitive FLISA data of N-ﬂuoKD and C-ﬂuoKD. (B) Competitive ELISA
data of N-ﬂuoKD and C-ﬂuoKD. Symbols: j, N-ﬂuoKD; ., C-ﬂuoKD.
compared with those in competitive FLISA. Based on competitive
FLISA and ELISA, the limits of detection (LODs) for DR5 determination in FLISA using N-ﬂuoKD and C-ﬂuoKD were estimated as
10,000- and 1000-fold lower than those in competitive ELISAs,
respectively. These results indicate that the FLISA using ﬂuoKD
allows rapid and highly sensitive detection compared with a
traditional ELISA system.
Fluorescence intensity
(arbitrary unit)
A
Discussion
Concentration of N-fluoKD (µg/ml)
Fluorescence intensity
(arbitrary unit)
B
Concentration of C-fluoKD (µg/ml)
Fig.4. (A) Conventional FLISA of N-ﬂuoKD (j). (B) Conventional FLISA of C-ﬂuoKD
(j). In both experiments, BSA (.) was used as negative control.
Since the last decade, there has been considerable progress in
the development of non-immunoglobulin protein scaffolds
[12,13]. In addition to having similar properties as antibodies, protein scaffolds have several useful features such as small size (<100
amino acids), high thermostability, and economic production in
bacterial hosts. Owing to these features, they have been considered
as potential alternatives to antibodies for therapeutic and diagnostic purposes [5]. In many diagnostic platforms, the fusion of protein
with other molecules such as small peptides, proteins, and chemicals has been a useful strategy for improving the performance of
diagnostic systems. However, the fusion of protein with other molecules can cause the conformational change of proteins and, in
many cases, can result in loss of original activities of the fused proteins. Sakamoto and coworkers [14,15] reported the development
of antibody (scFv) fused GFP for FLISA, but the fusion of antibody
to the N terminus of GFP (N-ﬂuobody) caused a signiﬁcant
decrease of ﬂuorescence (600-fold lower activity compared with
C-terminal fusion) and it is supposed that the difference of ﬂuorescent intensity is mainly derived from different ﬂexibility of linker
at the N terminus or C terminus of GFP. This means that the fusion
point can have a signiﬁcant effect on antibody and GFP functions
and that the choice of fusion point should be done very carefully
68
Human kringle domain-based FLISA system / G.M. Jeong et al. / Anal. Biochem. 451 (2014) 63–68
with structure information and experiments. In this aspect, KD has
beneﬁcial structures for protein fusion. KD has seven ﬂexible loops,
and each loop can serve as a binding motif to the target molecules
independent of interaction with other loops. As demonstrated
here, the fusion of KD with GFP at the N or C terminal did not have
any effect on the ﬂuorescent intensity of GFP or on binding afﬁnity
of KD to the target molecules, DR5 (Fig. 3). This result strongly indicates the superiority of KD in the development of the diagnostic
system compared with other binding molecules, including antibody. Moreover, the high solubility of KD in the cytoplasm of the
E. coli host is advantageous over antibody molecules. For biological
activities, most antibodies require disulﬁde bond formation (at
least 2 disulﬁde bonds for scFv or more disulﬁde bonds for other
antibody fragments and full-length antibodies), and those disulﬁde
bonds can be formed only in the periplasm of E. coli [16]. The production of antibodies in the cytoplasm of E. coli results in aggregation of antibodies, and the labor-intensive refolding process is
required for obtaining active antibody molecules [14]. In this format (GFP-fused antibody), GFP is also produced as an inactive form
and can be activated during the refolding process. During this
refolding process, the presence of antibody that has a relatively
large size (30 kDa) may prevent the correct formation of chromophore in GFP, and the efﬁciency of chromophore formation may be
dependent on position (N or C terminal) of antibody fusion. In contrast, KD has a much smaller size (10 kDa) and is highly soluble in
cytoplasm of E. coli. It has been demonstrated that more than 5 g of
KD can be produced in a 1-L scale fed-batch cultivation [8]. The
small size of KD and easy folding in cytoplasm may provide more
preferable conditions for the correct formation of GFP chromophore. Both N-ﬂuoKD and C-ﬂuoKD were produced as soluble
forms in cytoplasm with high ﬂuorescence and could be simply
puriﬁed through single-afﬁnity column chromatography with high
purity and recovery yields (Fig. 2). The elimination of the refolding
process and high solubility of fused proteins can be advantageous
for protein preparation and development of the diagnostic system
[17].
Compared with conventional ELISA, the better performance of
KD-based FLISA has been successfully demonstrated. In the competitive FLISA with N-ﬂuoKD, the LOD was approximately 60 pM,
nearly 10,000-fold lower than that in the competitive ELISA, which
was sufﬁcient to be used for diagnostic purposes (Fig. 5). Currently,
most ELISA systems employ the HRP reaction for detection, and the
lower sensitivity of ELISA can be ascribed to less sensitivity of HRPbased secondary reaction than ﬂuorometric detection with GFP
[18,19]. Additional beneﬁts include the avoidance of toxic and
acidic substrates necessary for HRP reaction as well as the reduced
assay time by the elimination of a secondary reaction. Even though
the current dynamic range (1.5 log scale) is acceptable for a diagnostic system, the narrow dynamic range can be a limitation in a
diagnostic system and needs to be improved. We believe that the
narrow range of ﬂuoKD-based FLISA does not result from GFP fusion but rather results from the intrinsic afﬁnity of KD548 against
target DR5 that is not high enough (KD = 172 nM) [6]. The dynamic
range can be further improved by using the high-afﬁnity KD variant. Similar to antibody engineering, the afﬁnity of KD can be easily
improved by the afﬁnity maturation process, and we believe that,
with the afﬁnity maturated KD variants, a more reliable and sensitive diagnostic system with extended dynamic range can be developed. In addition, it is applicable to develop a sandwich FLISA
system in which target antigen is captured by the immobilized
KD and detected by the other ﬂuoKD. Two different KD variants
against same target molecule can be isolated by screening a
combinatorial KD library, and a sandwich FLISA system with two
KD variants can provide a sensitivity-improved diagnostic system.
In conclusion, we have successfully developed a FLISA system
with KD–GFP fusion protein (ﬂuoKD) for an immunosorbent assay.
Due to several distinctive features, including small size, ﬂexibility,
stability, and easy production, the KD fusion system can provide a
simpler and more sensitive platform for immunosorbent assays
than conventional antibody-based ELISA and FLISA systems. We
showed a single example of KD (KD548) speciﬁc to DR5. KD can,
however, be easily modiﬁed to new derivatives toward other target
molecules using a general combinatorial library screening tool
(e.g., phage display); therefore, the applicability of KD-based FLISA
can be extended to the development of a diagnostic system for various target molecules with engineered KD.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ab.2014.01.019.
References
[1] R.M. Lequin, Enzyme immunoassay (EIA)/enzyme-linked immunosorbent
assay (ELISA), Clin. Chem. 51 (2005) 2415–2418.
[2] W. Wang, S. Singh, D.L. Zeng, K. King, S. Nema, Antibody structure, instability,
and formulation, J. Pharm. Sci. 96 (2007) 1–26.
[3] T. Wurch, A. Pierré, S. Depil, Novel protein scaffolds as emerging therapeutic
proteins: From discovery to clinical proof-of-concept, Trends Biotechnol. 30
(2012) 575–582.
[4] M. Gebauer, A. Skerra, Engineered protein scaffolds as next-generation
antibody therapeutics, Curr. Opin. Chem. Biol. 13 (2009) 245–255.
[5] S. Banta, K. Dooley, O. Shur, Replacing antibodies: engineering new binding
proteins, Annu. Rev. Biomed. Eng. 15 (2013) 93–113.
[6] C.H. Lee, K.J. Park, E.S. Sung, A. Kim, J.D. Choi, J.S. Kim, S.H. Kim, M.H. Kwon, Y.S.
Kim, Engineering of a human kringle domain into agonistic and antagonistic
binding proteins functioning in vitro and in vivo, Proc. Natl. Acad. Sci. U.S.A.
107 (2010) 9567–9571.
[7] C.H. Lee, K.J. Park, S.J. Kim, O. Kwon, K.J. Jeong, A. Kim, Y.S. Kim, Generation of
bivalent and bispeciﬁc kringle single domains by loop grafting as potent
agonists against death receptors 4 and 5, J. Mol. Biol. 411 (2011) 211–219.
[8] S.H. Jang, C.H. Lee, Y.S. Kim, K.J. Jeong, High-level production of a kringle
domain variant by high-cell-density cultivation of Escherichia coli, Appl.
Microbiol. Biotechnol. 92 (2011) 327–336.
[9] S.T. Jung, K.J. Jeong, B.L. Iverson, G. Georgiou, Binding and enrichment of
Escherichia coli spheroplasts expressing inner membrane tethered scFv
antibodies on surface immobilized antigens, Biotechnol. Bioeng. 98 (2007)
39–47.
[10] K.J. Jeong, M.J. Seo, B.L. Iverson, G. Georgiou, APEx 2-hybrid, a quantitative
protein–protein interaction assay for antibody discovery and engineering,
Proc. Natl. Acad. Sci. U.S.A. 104 (2007) 8247–8252.
[11] J. Sambrook, D. Russell, Molecular Cloning: A Laboratory Manual, 3rd ed., Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2001.
[12] J. Löfblom, F.Y. Frejd, S. Ståhl, Non-immunoglobulin based protein scaffolds,
Curr. Opin. Biotechnol. 22 (2011) 843–848.
[13] Y.L. Boersma, A. Plückthun, DARPins and other repeat protein scaffolds:
advances in engineering and applications, Curr. Opin. Biotechnol. 22 (2011)
849–857.
[14] S. Sakamoto, Y. Tanizaki, B. Pongkitwitoon, H. Tanaka, S. Morimoto, A chimera
of green ﬂuorescent protein with single chain variable fragment antibody
against ginsenosides for ﬂuorescence-linked immunosorbent assay, Protein
Expr. Purif. 77 (2011) 124–130.
[15] S. Sakamoto, B. Pongkitwitoon, K. Sasaki-Tabata, W. Putalun, K. Maenaka, H.
Tanaka, S. Morimoto, A ﬂuorescent single domain antibody against plumbagin
expressed in silkworm larvae for ﬂuorescence-linked immunosorbent assay
(FLISA), Analyst 136 (2011) 2056–2063.
[16] A. de Marco, Strategies for successful recombinant expression of disulﬁde
bond-dependent proteins in Escherichia coli, Microb. Cell Fact. 8 (2009) 26.
[17] E.D.B. Clark, Refolding of recombinant proteins, Curr. Opin. Biotechnol. 9
(1998) 157–163.
[18] P. Oelschlaeger, S. Srikant-Iyer, S. Lange, J. Schmitt, R.D. Schmid, Fluorophorlinked immunosorbent assay: a time- and cost-saving method for the
characterization of antibody fragments using a fusion protein of a singlechain antibody fragment and enhanced green ﬂuorescent protein, Anal.
Biochem. 309 (2002) 27–34.
[19] N. Velappan, J. Clements, C. Kiss, R. Valero-Aracama, P. Pavlik, A.R. Bradbury,
Fluorescence linked immunosorbant assays using microtiter plates, J.
Immunol. Methods 336 (2008) 135–141.
Supplementary Information
A human kringle domain-based fluorescence-linked immunosorbent assay
(FLISA) system
Gu Min Jeonga, Yong Sung Kimb, Ki Jun Jeonga,c*
a
Department of Chemical and Biomolecular Engineering, KAIST, 335 Gwahagno, Yuseong-gu, Daejeon
305-701, Republic of Korea
b
Department of Molecular Science and Technology, Ajou University, San 5, Woncheon-dong, yeongtong-
gu, Suwon 443-749, Republic of Korea
c
Institute for the BioCentury, KAIST, 335 Gwahagno, Yuseong-gu, Daejeon 305-701, Republic of Korea
1
Table S1 Oligonucleotides used in this study
Primer
Sequences (5′–3′)
name
GFP-R1
tgtaagcttctattagtcgaccttggatagttcatcca
NF-F1
aagaaggagatatacatatgcatcaccatcaccatcacgaggaatgtcaccagaccca
NF-F2
atattctagaataattttgtttaactttaagaaggagatatacatatgcatcacca
NF-F3
agtaaaggagaagaacttttcactggag
NF-R2
ctccagtgaaaagttcttctcctttactagaaccaccaccaccagaaccaccaccacctggaggtgttgtgcagc
NF-R3
atataagcttctattagtcgaccttggatagttcatcca
CF-F4
acagaattcatgcatcaccatcaccatcacagtaaaggagaagaacttttcactg
CF-F5
tctagaataattttgtttaactttaagaaggagatatacatatgcatcaccatcacca
CF-F6
acatctagaataattttgtttaactttaagaaggagata
CF-F7
tggtggtggtggttctgaggaatgtcaccagacc
CF-R4
agaaccaccaccaccagaaccaccaccaccgtcgaccttggatagttcat
CF-R5
acaaagcttctattatggaggtg
2
Fig. S1. The Schematic diagram of plasmids used for production of GFP (pGM-GFP), NfluoKD (pGM-N-fKD) and C-fluoKD (pGM-C-fKD). In all expression systems, Plac
promoter was used for gene expression and 6xHis tag was fused to N-terminal of each protein
for detection and simple purification. Two restriction enzymes (XbaI and HindIII) sites for
cloning are indicated.
3
Fig. S2. FACS analysis of E. coli cells producing N- or C-fluoKDs. The SHuffle® Express E.
coli strain (no plasmid) was used as a negative control. M means the mean fluorescence
intensity.
4

Download Report