In vivo programming of endogenous antibodies via oral administration of adaptor ligands

Article abstract of DOI:10.1016/j.bmc.2017.09.010

Vaccination is a reliable method of prophylaxis and a crucial measure for public health. However, the majority of vaccines cannot be administered orally due to their degradation in the harsh gut environment or inability to cross the GI tract. In this stud

Full text of DOI:10.1016/j.bmc.2017.09.010

Accepted Manuscript
In vivo Programming of Endogenous Antibodies via Oral Administration of
Adaptor Ligands
Masanobu Nagano, Nancy Carrillo, Nobumasa Otsubo, Wataru Hakamata,
Hitoshi Ban, Roberta F. Fuller, Nasir K. Bashiruddin, Carlos F. Barbas III
PII:
S0968-0896(17)30969-0
https://doi.org/10.1016/j.bmc.2017.09.010
BMC 13970
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
10 May 2017
24 August 2017
8 September 2017
Please cite this article as: Nagano, M., Carrillo, N., Otsubo, N., Hakamata, W., Ban, H., Fuller, R.F., Bashiruddin,
N.K., Barbas, C.F. III, In vivo Programming of Endogenous Antibodies via Oral Administration of Adaptor Ligands,
Bioorganic & Medicinal Chemistry (2017), doi: https://doi.org/10.1016/j.bmc.2017.09.010
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
In vivo Programming of Endogenous
Antibodies via Oral Administration of
Adaptor Ligands
Leave this area blank for abstract info.
Masanobu Nagano, Nancy Carrillo, Nobumasa Otsubo, Wataru Hakamata, Hitoshi Ban, Roberta F. Fuller, Nasir
K. Bashiruddin, Carlos F. Barbas, III
The Skaggs Institute for Chemical Biology and the Departments of Chemistry and Molecular Biology, The
Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037
Department of Chemistry, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, 113-
0033 Tokyo, Japan
NO2
O2N
DNP
Endogenous
anti-hapten Abs
Programmed
Abs
O
O
DK
hapten
hapten
Biotin
O
O
S
oral
Ph
Adaptor
ligand
Pre-immunized
mice by hapten
VS

Bioorganic & Medicinal Chemistry
In vivo Programming of Endogenous Antibodies via Oral Administration of Adaptor
Ligands
,
Masanobu Nagano a, † , Nancy Carrillo a , Nobumasa Otsubo a , Wataru Hakamata a , Hitoshi Ban a ,
Roberta F. Fuller a , Nasir K. Bashiruddin b and Carlos F. Barbas, III a,
‡
a The Skaggs Institute for Chemical Biology and the Departments of Chemistry and Molecular Biology, The Scripps Research Institute, 10550 North Torrey
Pines Road, La Jolla, California 92037, USA
b Department of Chemistry, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, 113-0033 Tokyo, Japan
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
Vaccination is a reliable method of prophylaxis and a crucial measure for public health.
However, the majority of vaccines cannot be administered orally due to their degradation in the
harsh gut environment or inability to cross the GI tract. In this study, we report the first proof-of-
concept study of orally producible chemically programmed antibodies via specific conjugation
of adaptor ligands to endogenous antibodies, in vivo. Pre-immuniztion with 2,4-dinitrophenyl
(DNP), or the reactive hapten, 1,3-diketone (DK), or a novel reactive hapten, vinyl sulfone (VS)
in mice, followed by oral administration of adaptor ligands composed of the hapten and biotin to
the pre-immunized mice resulted in successful in vivo formation of the biotin-hapten-antibody
complexes within 2 hours. Pharmacokinetic evaluations revealed that apparent serum
concentrations of programmed antibodies were up to 144 nM and that the serum half-lives
reached up to 34.4 h. These findings show promise for the future development of orally
bioavailable drug-hapten-antibody complexes as a strategy to quickly and easily modulate
immune targets for aggressive pathogens as well as cancer.
Available online
Keywords:
Hapten
Instant immunization
Vaccine
Oral administration
Immunoconjugate
2009 Elsevier Ltd. All rights reserved.
1. Introduction
composed of the hapten conjugated to a carrier protein.4 Once
endogenous anti-hapten antibodies are produced, injection of a
Currently, active immunization is the most powerful method to
provide immunity against disease-related viruses and pathogens
and is said to prevent 6 million deaths worldwide every year.1
Unlike antibody therapies using passive immunization (e.g.
hyperimmune globulin, therapeutic monoclonal antibodies and
antibody-drug conjugates), active immunization elicits
endogenous antibodies with a variety of immunoglobulin
isotypes as well as long-lasting adaptive immunity via memory
B-cells.2 However, usage of common vaccine antigens (i.e.
weakened or inactive forms of pathogens or their surface
proteins) often take days to weeks to develop immunity to their
corresponding pathogens, thus, limiting their efficacy against
aggressive pathogens and rapidly acting toxins.
pathogenic target protein-binding compound conjugated to the
hapten (adaptor ligand) results in instant formation of adaptor
ligand-antibody complexes (immunoconjugates) in vivo which
can be used to target pathogens or even cancer cells (Figure 1a).
There have been successful examples for formation of anti-
cancer immunoconjugates in vivo. Mice pre-immunized with
fluorescein or 2, 4-dinitrophenyl (DNP) as haptens followed by
the intravenous (i.v.) administration of the adaptor ligands
composed of cancer-targeting small molecules linked to
corresponding haptens resulted in successful in vivo formation of
immunoconjugates.5,6,7,8 Further, a treatment for renal cell
carcinomas where immunization with the synthetic vaccine,
fluorescein-conjugated KLH (keyhole limpet hemocyanin)
followed by the injection of folate-conjugated fluorescein adaptor
ligand had reached Phase II clinical trials (although terminated
due to low subject accrual).9 Our group has previously shown
that reactive immunization with the hapten, 1,3-diketone (DK),
conjugated to KLH enables the production of catalytic aldolase
antibodies10,11,12 These antibodies were able to covalently bind in
To address the relatively slow process of vaccine-mediated
acquisition of immunity, a combination of elements from active
and passive immunization have been explored to expedite the
process of pathogen clearance.3 This approach first induces an
immune state for a designated immune-response eliciting small
molecule, or hapten, via the injection of a synthetic vaccine
———
 Corresponding author. Tel.: +81-3-5841-8372, fax: +81-3-5841-8372; mnagano@chem.s.u-tokyo.ac.jp (M. Nagano).
†Current address: Department of Chemistry, Graduate School of Science, The University of Tokyo, Tokyo 113-0033, Japan.ꢀ
‡C.F. Barbas III passed away on June 24th, 2014, during the preparation of the manuscript.

drug (bio n)
a
c
O
hapten
b
hapten
drug
HN
H
NH
H
N
NO2
N
N
H
H
H
O
N
N
N
O
S
O
3
3
O
O
O2N
O
DNP-Biotin
(1)
Adaptor ligand
O
an -hapten an body
O
HN
NH
H
H
O
O
H
N
O
N
O
S
N
H
N
3
H
N
N
O
3
DK-Biotin
(2)
O
HN
NH
HO
N
N
O
H
H
O
O
H
H
N
H
N
N
S
N
O
S
N
H
3
O
O
O
Programmed an body
(Immunoconjugate)
VS-Biotin
(3)
❶ Immuniza on to produce an -hapten an bodies
❷ Subsequent oral administra on of adaptor instantly
make producible programmed an bodies in vivo
❸ Analysis of the sera by ex vivo ELISAs
HRP
hapten
In vivo
matura on
an -hapten
Programmed
Abs
an -hapten
Abs
Germ Abs
Abs
SA
Carrier
protein
hapten
Bio
n
oral
Pre-immunized
mouse
Naïve mouse
Adaptor ligand
Figure 1. (a) A composition of programmed antibody. (b) Structures of adaptor ligands, DNP-biotin (1), DK-biotin (2) and VS-biotin (3). (c) The concept of
this study. Active immunization against haptens followed by oral administration of the hapten-biotin adaptor ligand instantly produces programmed
antibodies as anti-streptavidin antibodies in vivo.
vitro various adaptor ligands composed of DK-fused drugs
against flu, HIV or cancer, in order to chemically program
antibody specificity and thus biological activity.13 Furthermore,
the injection of DK fused integrin-binding compounds into mice
pre-immunized against DK successfully led to the in vivo
formation of programmed polyclonal antibodies that significantly
reduced melanoma size by targeting cancer-related integrins.14
Thus the development of instant chemically induced immunity
by way of hijacking the immune system through programming of
endogenous antibodies with adaptor ligands is a significant leap
from current vaccinnology. However, administration of adaptor
ligands requires multiple injections over a given time course
depending on the ailment to be treated. Such treatments would
require multiple or prolonged visits to the clinic. Furthermore,
orally bioavailable adaptor ligands would offer a number of
significant advantages over current biological vaccination
methodologies, such as instant immunity, targeting non-
mucosomal pathogens, suitability for long term storage, a
needless for trained personnel, low cost and suitable for mass
production.
Here, we report a proof-of-concept study demonstrating the in
vivo production of immunoconjugates via oral administration of
hapten-biotin adaptor ligands in mice actively immunized by
these adaptor ligands (Figure 1b). Further, to investigate the
potency of immunoconjugates formed in vivo, the previously
reported haptens DK and DNP as well as a novel reactive hapten,
vinyl sulfone (VS), were chosen and chemically conjugated to
biotin (Figure 1c). These adaptor ligands were subsequently
orally administered to mice pre-immunized to the corresponding
hapten. After 2 h, in vivo formation of the corresponding biotin-
hapten immunoconjugates were detected at sub-micromolar
concentrations in serum. Further pharmacokinetic analysis of
these novel immunoconjugates revealed that, of the 3 haptens
tested, DNP produced the most stable immunoconjugates with
apparent serum half-lifes of up to 34.4 h. To the best of our
knowledge, this is the first report of orally programmed
antibodies to rapidly modify antigen specificity and studies of
hapten potency forming stable immunoconjugates in vivo. Future
studies using these haptens linked to potent pathogen or cancer-
cell targeting small molecules may open new avenues for the
modulation of active immunity via oral administration.
2. Results and discussion
2.1 Synthesis of the haptens and adaptor ligands
As hapten moieties, we have chosen DNP, DK and VS.
Humans have naturally occurring anti-DNP antibodies and these
are expected to bind DNP in a non-covalent manner.15 As
described above, reactive immunization with DK is able to
generate aldolase antibodies and are therefore expected to
produce covalently bound antibody-hapten conjugates through
enaminone formation. The ,-unsaturated sulfone structure of
VS is known to irreversibly inhibit cysteine proteases via
Michael addition16,17 and it was expected that reactive
immunization with VS would behave similarly to the DK and
produce covalent antibody-hapten conjugates. In this study, we
chose to synthesize the aforementioned haptens conjugated to
biotin as adaptor ligands due to the wide variety of detection and
quantification methods via streptavidin.
The adaptors, DNP-biotin (1), DK-biotin (2) and VS-biotin (3)
and N-hydroxy succinimide esters of these haptens were prepared
for the programming of endogenous antibodies and conjugations

Scheme 2. Preparation of hapten-carrier protein conjugates as antigens.
(a) Preparation of DNP antigen (b) Preparation of DK antigen (c)
Preparation of VS antigen. X represents predicted number of lysines on
a surface of protein from Thermo Scientific. ND, not determined.
Scheme 1. Synthesis of hapten derivatives. (a) Synthesis of DNP (b)
Synthesis of DK (c) Synthesis of VS (d) Synthesis of biotinylated
haptens.
desired BSA and KLH conjugates (Scheme 2a and 2b). The
hapten density of BSA conjugates were estimated by MALDI-
TOF mass spectrometry and gave roughly 12.5 and 11.7
modifications per protein for the DNP and DK derivatives,
respectively. Since the hapten density of oligomeric KLH is
difficult to measure by MALDI-TOF due to its high molecular
weight (4-8 MDa), enzyme-linked immunosorbent assays
(ELISAs) were performed for both DNP-KLH and DK-KLH to
determine whether the haptens were successfully conjugated to
KLH. Specific binding of anti-DNP and anti-DK monoclonal
antibodies to DNP-KLH or DK-KLH, respectively, were detected
and both conjugates showed comparable levels of hapten-
conjugation (Supplementary Figure 2).
to carrier proteins. Aromatic substitution of 2,4-dinitiro-1-
chlorobenzene by 5-aminopetanoic acid, followed by
condensation with azide-peg-amine or N-hydroxysuccinimide
gave DNP-azide (5) or DNP-NHS ester (6), respectively
(Scheme 1a). DK-azide (8) or DK-NHS ester (9) were prepared
in the same manner starting from an acid of 1,3-diketone (7)10
(Scheme 1b). Synthesis of VS was successfully prepared by
slight modifications to the procedure reported previously for the
preparation of the cysteine protease inhibitor, K-777
(Supplementary Figure 1).16,18,19 Based on reported SAR
information, the chirality of homophenylalanine at the P1 site
was inverted and the phenylalanine at the P2 site, which is
essential for binding to proteases, was replaced with a less
inhibitory serine to avoid the undesired reaction of K-777 to
cysteine protease homologs. Additionally, the P3 site was
replaced with a PEG linker to increase solubility in water
(Scheme 1c). Oxidation of 6-azidehexanol, using iodobenzene
diacetate, followed by amidation with a TBS protected serine
methyl ester (10) gave TBS protected VS-azide (11). Hydrolysis
of 11 rendered an acid, and was subjected to coupling with a
phenyl vinyl sulfonated compound (13), obtained from (R)-
homophenylalanine to afford TBS-protected VS-azide (14). Acid
catalyzed deprotection of the silyl group yielded VS-azide (15).
Finally, the azide conjugated haptens were linked with
biotinylated alkynes using copper catalyzed cycloaddition in the
presence of THTPA20, to give the desired adaptor ligands, DNP-
biotin (1), DK-biotin (2) and VS-biotin (3) (Scheme 1d).
For VS, BSA was selected as the carrier protein for
immunization studies while ovalbumin (OVA) was chosen as the
carrier protein for downstream antibody-binding studies. To
prepare the VS-carrier protein conjugates, BSA and OVA were
coupled with the alkyne moiety 16 by EDC to give, 13.2 and 9.9
alkyne modificaitions, respectively (Scheme 2c). Subsequent
click reactions of conjugates with VS-azide (15) gave the desired
VS-BSA or VS-OVA with 4.4 and 0.5 hapten densities,
respectively, as determined via MALDI-TOF MS analysis.
2.3 Active immunization with hapten-carrier protein
conjugates
To produce anti-hapten antibodies against DNP, DK and VS,
balb/c mice were interperitonealy (i.p.) injected with DNP-KLH,
DK-KLH and VS-BSA, respectively, on days 1, 22 and 43
(Figure 3a). On days 29 and 50, serum samples were obtained by
tail vein bleeds and analyzed by ELISA using DNP-BSA, DK-
BSA and VS-OVA (Figure 3b) to determine antibody titers. Anti-
DNP titers peaked at bleed-1 at over 20,000, while anti-DK titers
reached a peak of 20,000 on bleed-2. Anti-VS titers showed a
value of only 5,000 at bleed-2, thus, a fourth booster shot was
administered and after 77 days (bleed-3) a maximum titer of near
25,000 was reached. In addition, anti-sera from DNP-KLH and
DK-KLH immunized mice showed negligible levels of anti-KLH
titers. In addition, anti-BSA titers were undetectable from VS-
BSA immunized mice. Next, to investigate the cross-reactivity of
the anti-hapten antibodies, we performed ELISAs using the
2.2 Preparation of hapten-carrier protein conjugates
To induce an immune responses against haptens, conjugation
with appropriate carrier proteins is required. Therefore, for both
DK and DNP, KLH was selected as the carrier protein for
immunization purposes while bovine serum albumin (BSA) was
selected as the carrier protein for downstream antibody-binding
studies. Conjugates of DNP and DK were produced by coupling
the amines of lysines on the surface of KLH and BSA via the
hapten-NHS esters 6 and 9, followed by gel filtration to give the

after washing with acid due to covalent bond formation between
38C2 and DK on BSA. Next, anti-DK and anti-VS sera against
DK-BSA and VS-OVA were tested. The signal intensity of both
anti-DK and -VS sera drastically decreased when exposed to
acidic conditions, in comparison to the neutral wash at pH 7.4.
These results suggest that the main binding of anti-DK and VS
antibodies produced via reactive immunization is non-covalent
rather than covalent. It has been reported in a previous study that
reactive immunization with DK resulted in only 2 aldolase
antibodies out of the 20 highest affinity anti-DK antibodies
produced in the study.10,11 This may explain the low abundance of
covalently-binding antibodies in this study.
a
b
2nd
bleed
3rd
bleed
1st
bleed
An gen An gen
An gen
71
An gen
Balb/c-mice
1
22
29
43
50
77
Days
3
2
1st bleed
2nd bleed
3rd bleed
1
0
2.5 In vitro programming of endogenous anti-hapten
antibodies from immunized mice with adaptor ligands
Since it was shown that high titers against the haptens were
achieved, we proceeded to investigate in vitro programming of
anti-hapten antibodies in the antisera of the corresponding
immunized mice. Anti-sera were incubated with different
quantities of the corresponding adaptor ligands (1-3) followed by
incubation in streptavidin coated wells for ELISA analysis
(Supplementary Figure 4). Treatment of anti-DNP sera with
DNP-biotin (1) resulted in strong absorption signals confirming
successful antibody programming while treatment of sera from
non-immunized naïve mice showed no significant adsorption in
ELISA. Similarly, treatment with DK-biotin (2) and VS-biotin
(3) detected a specific signal only when the respective anti-sera
were used. Thus, findings from these studies indicate that all anti-
hapten antibodies produced by immunized mice were capable of
being programmed as desired biotinylated immunoconjugates by
the adaptors in an in vitro manner.
A
A
A
S
A
S
B
A
V
H
A
H
L
K
A
V
S
S
L
S
B
B
B
-
Immobilized
protein
K
B
O
O
-
-
K
P
S
D
N
V
D
Immunogen
DNP-KLH
DK-KLH
VS-BSA
c
3
2
1
0
anti-DNP serum
anti-DK serum
anti-VS serum
A
S
A
A
Immobilized
protein
V
O
S
B
B
-
-
S
-
K
P
V
D
N
D
Figure 2. Active immunization with synthetic vaccines. (a)
Immunization schedule. Only VS-BSA required an additional booster.
(b) Anti-serum titers derived from vaccinated mice (n = 4 - 6) were
individually determined by ELISA. Dilution range is up to 25,600.
Error bars, s.d. (n = 4 - 6). (c) Cross reactivity checked by ELISA. The
individual anti-serum derived from the last bleed were pooled. Error
bars, s.d. (n = 3). The anti-serum was diluted to 1/2,500 in PBS.
2.6 Pharmacokinetic analysis of an intraperitoneally
injected biotinylated hapten-monoclonal antibody
To assess the pharmacokinetic properties of
immunoconjugates formed via oral administration (p.o.) of
adaptor ligands, it is essential to determine the level of
immunoconjugates in vivo without the bias of antibody titer and
oral availability of the adaptors. To achieve this,
immunoconjugates formed in vitro were intraperitoneally (i.p.)
injected into non-immunized mice to analyze their
haptens conjugated to carrier proteins that were not used for
immunization (Figure 3c). When pooled DNP anti-sera were
incubated with DNP-BSA, DK-BSA or VS-OVA coated wells,
only DNP-BSA showed high absorbance while low signal
intensities were detected for DK-BSA and VS-OVA.
Additionally, DK and VS antisera showed specific binding to
DK-BSA and VS-OVA, respectively. It should be noted that
although anti-DNP antibodies are known to exist in mammals
independent of immunization, DK and VS anti-sera did not show
significant binding to DNP-BSA.7 Overall, immunization with
hapten-carrier protein conjugates successfully elicited the
production of anti-hapten antibodies specific to each hapten.
pharmacokinetics. Since anti-DNP mouse mAb (9H8.1) and
catalytic anti-DK mouse mAb (38C2) are commercially
available, we prepared pre-immunoconjugates with the adaptor
ligands DNP-biotin (1) and DK-biotin (2), respectively. It is
important to note that both monoclonal antibodies are of the IgG
isotype which are primarily induced by active immunization.14
To construct the immunoconjugates, the monoclonal antibodies
were incubated with 2.2 equivalents of the adaptors 1 and 2 in
phosphate-buffered saline solution (pH7.4) at room temperature
for 2 h. Quantitative conversion to the DK-biotin (2)-38C2
2.4 Evaluation of the binding modes of anti-DK and VS
antibodies
a
b
200
150
100
50
150
100
50
Next, we assayed for the presence of covalent antibodies
against DK and VS. Previously, we have demonstrated that
aldolase monoclonal antibodies that bind covalently to DK
through the formation of enaminone are stable at low pH (pH
2.0), whereas non-covalent complexes readily dissociate.14
Therefore, this acid-stable binding provides a means to detect the
presence of covalent antibodies from sera. Thus, we used acid-
wash ELISA to determine whether covalent antibodies were
produced in DK and VS immunized mice (Supplementary Figure
3). When using the catalytic mAb 38C2 against DK-BSA as a
positive control there was no significant loss of signal intensity
0
0
0
20
40
60
80
100
0
20
40
60
80
100
Time (h)
Time (h)
Figure 4. IP injection of in vitro programmed antibodies to non-
immunized mice. (a) Biotin-DNP (1)-anti-DNP mAb (b) Biotin-DK (2)-
38C2. The circles (●) are from averaged values (n = 2). The graphs were
created by GraphPad Prism version 6.

complex was observed via methodol assay by monitoring the
lack of retro-aldol catalytic activity in 38C2, while unreacted
38C2 showed high retro-aldol activity resulting in a strong
fluorescence signal (Supplementary Figure 5). 21 MALDI-TOF
mass spectrometry confirmed the complete conversion to the
DK-biotin (2)-38C2 complex by an increase in mass that was
indicative of the conjugation of two DK-biotins per antibody
(one DK-biotin per active-site lysine) (Supplementary Figure 5b).
ELISA assays against immobilized streptavidin with both DNP
and DK conjugates gave similar signal intensities when using the
same quantities of each complex (Supplementary Figure 5c).
These results suggest that both DNP and DK conjugates near-
quantitatively bound the adaptor ligands. For the quantification
of immunoconjugates formed in vivo, standard curves were
prepared using these in vitro-synthesized immunoconjugates for
both DNP and DK by ELISAs (Supplementary Figure 6). These
conjugates were used without purification for further in vivo
experiments since only negligible amounts of the adaptor ligands
remained. The immunoconjugates were i.p. injected into non-
immunized balb/c mice with a single dose of 1 mg/kg each. Tail
vein blood samples were collected at 2, 4, 8, 24, 48 and 72 h after
injection. For each time point serum concentrations were
determined by ELISA and the data was fitted to a one phase
decay (y = y0 – yplateau*exp( - k*x) + yplateau), The apparent half-
life can serve as an indicator of serum stability of the
be 55.5 h while DK-biotin exhibited an apparent t-half of 14.7 h
(Figure 4).
2.7 Oral programming of endogenous anti-hapten
antibodies via oral administration of adaptor ligands
Based on the success with in vitro programming using anti-
sera, even though the molecular features of the adaptor ligands
do not adhere to Lipinski’s rule of 5,22 we examined if oral
administration of the hapten-biotin adaptor ligands could result in
in vivo formation of immunoconjugates. The adaptor ligands 1-3,
formulated with PEG300, were orally administered at a single
dose of 100 mg/kg to pre-immunized or non-immunized mice,
which had been rested for 6 months after their final injection.
Tail vein blood was collected at 2, 4, 8, 24, 48, 72 and 96 h after
oral administration and the quantity of immunoconjugates at each
time point was evaluated by ELISA (Figure 5). Intriguingly,
ELISA analysis showed that all three adaptor ligands were able
to successfully program endogenous anti-hapten antibodies
forming immunoconjugates in vivo without the addition of any
adjuvants. In contrast, oral administration of the adaptor ligands
to non-immunized mice were undetectable. These promising
results suggest that the combination of active immunization
followed by the oral administration of adaptor ligands can lead to
formation of immunoconjugates.
These results suggest that the tested biotin-hapten adaptor
ligands were able to cross the membrane of GI tract, however, it
still remains unclear how these adaptor ligands permeate through
immunoconjugates. According to this method, we determined the
apparent half life of DNP-biotin monoclonal immunoconjugate to
a
b
c
250
200
150
100
50
0
0
20
40
60
80
100
Time (h)
Figure 5. In vivo Programming of endogenous Abs by oral administration of adaptor lingands (1−3). (a) Biotin-DNP (1) (b) Biotin-DK (2). (c) Biotin-
VS. The circles (●) represent injections to immunized mice (n = 4-6) and the squares (■) represent injections to naïve mice (n = 4-6). All data points
were acquired from each individual mice and individual time points were acquired in triplicate (n = 4-6). Formation of biotin-VS (3)-antibody
conjugates was approximately quantified by applying the standard curve of biotin-DK (2)-38C2 conjugate. See supporting information.
Table 1
Properties of biotinylated immunoconjugates prepared in vivo or in vitro.
DNP-biotin (1)
DK-biotin (2)
DNP-biotin (1)
DK-biotin (2)
VS-biotin (3)
-anti-DNP mAbb
-38C2 mAbb
p.o.
p.o.
p.o.
i.p.
i.p.
Administration
Formulation
PEG300
100
PEG300
100
PEG300
100
PBS (pH 7.4)
PBS (pH 7.4)
1
1
Dose (mg/kg)
DNP-KLH
12,800 ± 4,525
DK-KLH
3,466 ± 1,222
VS-BSA
3,466 ± 1,093
−
−
Immunogen
Titer before injectiona
N.D.
N.D.
Programmed Ab
Cmax (nM)
127 ± 31
34.4
93 ± 24
10.4
144 ± 69
19.2
141 ± 9
55.5
122 ± 20
14.7
Programmed Ab
apparent half-lives (h)
a The titers were measured 30 min before the time of injection. C-max values were acquired at 2 h after injection.
b Monoclonal antibodies conjugated in PBS were administered via I.P. injection (1 mg/kg) to naïve mice (n = 2). N.D. means Not Determined.

the gut. It has been reported that biotinylated compounds utilize
carrier-mediated uptake in the gut through sodium-dependent
multiple vitamin transporter (SMVT). Unlike the biotin
transporter, which has narrow substrate specificity, SMVT has
broad substrate specificity and, therefore, has been used to
increase the oral bioavailablity of biotinylated drugs-of-
interest.23,24,25,26 Encouraged by the success of oral formation of
immunoconjugates with biotin-hapten adaptor ligands, we
investigated instant oral HIV vaccination in which adaptor
ligands composed of the same haptens but linked to the HIV
entry inhibitor, apraviroc,27 instead of biotin, however, no
detectable sign of forming anti-HIV immunoconjugates were
observed (Data not shown). Given this evidence, SMVT may
have contributed to the oral bioavailability of the adaptors used in
this study. In addition, several mechanisms exist for the transport
of mucosal IgAs between the gut and bloodstream.28 However,
results from an ELISA using an alkaline-phosphatase-conjugated
anti-mouse IgA antibody as secondary antibody against anti-DNP
serum showed no detectable signal, therefore, excluding the
possibility of the haptens in this study getting into the
Thus, the VS-immunoconjugates could result in short serum half-
life due to the non-existence of a covalent reservoir and high
reactivity of VS, similar to DK-immunoconjugates. Nonetheless,
VS showed adequate immunogenicity as well as in vivo
formation of immunoconjugates with apparent t-halfs longer than
that of DK validating it as a novel bona fide hapten.
Orally produced DNP-based immunoconjugates showed the
longest apparent t-half (34.4 h) of all tested conjugates and
compared to the DK-immunoconjugates, the DNP-
immunoconjugates showed longer lasting t-halfs for both in vitro
and in vivo synthesized immunoconjugates. DNP is known to
interact nonspecifically with membranes and albumin due to its
hydrophobicity.31,32 However, DNP may lack the strong off-target
interactions which can be considered with covalently binding DK
and VS. It is worth noting that among the tested haptens, DNP
was the most immunogenic and DNP-immunized mice
maintained relatively high titers prior to oral administration of
adaptor. According to Le Chatelier’s principle, high titers lead to
a shift in the binding equilibrium, leading to reduced clearance of
DNP antigen base adaptor ligand and thus longer apparent t-half.
These hydrophobic, less reactive and high immunogenic factors
of DNP may have contributed to the prolonged half-life of the
DNP-immunoconjugates. We expect that further engineering of
hapten structures will lead to more suitable adaptor ligands
beyond DNP, which is capable of conferring better oral
availability and more serum longevity to in vivo-synthesized
immunoconjugates. Thus, we demonstrated the potential of in
vivo programming of endogenous antibodies by orally
bloodstream via binding mucosal IgAs (Fig S7 in SI).
Analysis of the pharmacokinetics of the biotin bound
immunoconjugates (Table 1) revealed that the apparent t-halfs of
the in vivo-synthesized DK-based immunoconjugates (10.4 h)
were shorter than the apparent t-half of DNP-based
immunoconjugates (34.4 h) and this tendency was consistent
with the aforementioned studies comparing the apparent t-halfs
of in vitro-synthesized DK and DNP immunoconjugates. Of note,
in vitro-synthesized pre-conjugate experiments were based on
high affinity monoclonal antibodies, in contrast to in vivo-
synthesized immunoconjugates which were based on polycolonal
antibodies.
administrated reactive and non-reactive haptens, and believe that
further investigation of oral availability of the adaptor ligands
can be generalized to a drug possessing adaptor ligand toward to
an instant oral vaccination.
Intriguingly, injection of the in vitro-synthesized DK-based
immunoconjugates (apparent t-half = 14.7 h), in which the
ligand and the catalytic monoclonal antibody were reversibly
connected via enaminone bond, showed very similar apparent t-
halfs with the DK-based immunoconjugates formed in vivo
which were shown to be mostly non-covalently conjugated
(apparent t-half = 10.4 h). We previously reported that an i.v.
injection of in vitro-synthesized anti-Flu-immunoconjugate
which 38C2 was irreversibly conjugated to a β-lactam hapten-
zanamivir adaptor ligand through an amide bond, was found to
have a t-half of 72 h.29 This result suggests that the reversibility
of enaminone bond formation with catalytic antibody greatly
affects the stability of the conjugate in vivo.30 Thus, we postulate
that the dissociation of the DK-based immunoconjugates in vivo
may have also led to promiscuous binding of reactive DK to
nucleophilic amino acid residues of other proteins. To evaluate
the off-target reactivity of DK to proteins in blood, we conducted
SDS-PAGE analysis of rat serum mixed with DK (Fig S8 in SI).
DK-azide (9) was incubated in rat serum at 37 oC for 2h, and then
reacted with TAMRA-alkyne by click reaction. This result shows
that the DK reacted with various serum proteins, especially
albumin, thus depleting DK-biotin (2) concentration in the blood
stream.
3. Conclusion
Here, we report the successful generation of biotin bound
immunoconjugates by using a combination of active
immunization and oral administration of adaptor ligands as well
as analysis of their pharmacokinetic properties. This is proof-of-
concept study demonstrate the feasibility of programming
endogenous antibodies via oral administration of a synthetic
molecule and brings us closer to the realization of a novel
methodology of oral vaccination. Modification of the adaptor
with other orally bioavailable therapeutic drug moieties for the
oral production of drug-conjugated antibodies is currently
underway. We believe these findings could lead to new
innovative approaches in vaccinology.
4. Experimentals
4.1 Chemical synthesis
Detailed information on synthesis as well as analytical data
can be found in the Supporting Information.
4.2 Preparastion of hapten-carrier protein conjugates
DNP-KLH: 14 mg of KLH (Thermo) was dissolved in 1.4 mL
of H2O to have a concentration of 10 mg/mL in PBS,
Next we focused on the orally programmed VS-based
subsequently 5.6 mL of 100 mM sodium phosphate buffer (pH
7.2) was added. To 6 mL of resulting 2.0 mg/mL KLH was added
600 uL of 1mg/mL DNP-NHS (6) in DMF and the whole was
incubated at rt for 16 h, and then purified through PD-10 column.
Concentration was determined by BCA assay and 4.8 mg/mL 500
µL yellow suspended conjugate was used as an antigen for
immunization.
DNP-BSA: To 1.0 mL of 5 mg/mL BSA (Calbiochem) in H2O
(pH6.0) was added 40 µL of 66 mM DNP-NHS (6) in DMSO
and incubated at rt for 16 h. The resulting solution was purified
immunoconjugates. As described in Fig S3 in SI, the binding
mode of reactive VS hapten to anti-VS antibodies in mice were
mainly non-covalent. The apparent t-half of orally programmed
VS-based immunoconjugates (t-half = 19.2 h) were longer than
those of DK-based immunoconjugates (t-half = 10.4 h), even
though their titers were similar prior to oral administration. As
seen with DK, the off-target reactivity of the VS was also
analyzed and showed similar tendencies of promiscuous binding
to off-target proteins even though the structure of VS was based
on a relatively inert antiparasitic agent K-777 (Fig S1 in SI).

by PD-10 column and concentrated with ultra filtration (Amicon
Ultra MWCO10,000) and centrifuged (3,000 rpm, 10 min) to less
than 500 µL. The concentration of remaining solution was
measured by BSA assay, then 5.7 mg/mL yellow suspended
solution was used for immunizations. The number of modified
residues were determined by MALDI-TOF MS and found to
have 12.5 Lys modified.
Protein concentrations were determined by a modification of
Lowry’s method, using bicinchonic acid (Pierce BCA Protein
Assay Kit, Thermo Fisher), with BSA standards. Samples were
incubated at 37 oC for 30 min before analysis on TECAN plate
reader at 562 nm.
4.4 Animal experiment
DK-KLH and DK-BSA were prepared as described.10 DK-
KLH: 14 mg of KLH (Thermo, #77600) was dissolved in 1.4 mL
of H2O to be 10 mg/mL in PBS, then added to 5.6 mL of 100 mM
Sodium phosphate buffer (pH 7.2). To 6 mL of resulting 2.0
mg/mL KLH was added 600 µL of 1mg/mL DK-NHS (9) in
DMF and incubated at rt for 16 h, and then purified through a
PD-10 column. Concentration was determined by BCA assay and
4.2 mg/mL 500 µL desired conjugate was used as an antigen for
immunization.
DK-BSA: To 1.0 mL of 5 mg/mL BSA (Calbiochem) in H2O
(pH 6.0) was added 40 uL of 66 mM DK-NHS (9) in DMSO and
incubated at rt for 16 h. Resulting solution was purified by PD-10
column and concentrated with ultra filtration (Amicon Ultra
MWCO10,000) and centrifuged (3,000 rpm, 10 min) to less than
500 µL. The concentration of the remaining solution was
measured by BSA assay, then a 5.6 mg/mL solution was used for
immunization. The number of modified residue was determined
by MALDI-TOF MS and showed that 11.7 Lys were modified.
BSA-alkyne: 25 g of BSA was dissolved 5 mL of H2O (pH
6.0), and filtered through Whatman PURADISC (25 mm, 0.45
µm). To the filtrate was added mixture of 4.83 mg of EDC-HCl
2O (pH 6.0) and 2.66 mg of alkyne-acid (16) in
All procedures were approved by The Scripps Research Institute
Animal Care and Use Committee and were performed according
to national and institutional guidelines for the humane treatment
of animals. See supporting info for the details.
4.5 Active immunization and ELISAs
7 week old Balb/c mice (n = 4 for DNP-KLH, DNP-KLH, and
KLH, n = 6 for VS-BSA and BSA) were intraperitoneally
injected 100 µg/mouse in Sigma Adjuvant System, with 200
µL/mouse as a primary injection. Subsequently, 3 weeks later,
the same amount of antigen were injected as a secondary
injection (The day 22). A third injection was performed with 100
µg/mouse with ALUM at 200 µg/mouse (The day 43). For the
immunization of VS, the fourth injection was performed with 100
µg/mouse with ALUM, 100 µL/mouse (The day 71). Individual
DK-antiserum from DK-KLH-immunized mice was collected on
days 28, and 49 (and 77 for VS) and used for in vitro assays. For
ELISA, Costar 96-well ELISA plates (Corning) were coated with
200 ng of DNP-BSA, DK-BSA and VS-OVA in 50 µL PBS and
incubated overnight at 4 °C. After blocking with 150 uL of
PBS/3% BSA for 2 h at 37 °C, 50 µL of different dilutions (from
1 : 500 to 1 : 64,000) of individual sera (4 or 6 mice for hapten)
was added into each well and the plates were incubated for 2 h at
37 °C. Washing and detection were performed as described using
HRP-conjugated goat anti-mouse IgG antibody (diluted 1:3,000
in PBS/1 % BSA). In some experiments, additional incubation
with 50 µL of 0.05 M citric acid, pH 2.5 (acid wash) for 15 min
at rt was performed after the initial washing step. From
reaction mixture was filtered through Whatman PURADISC (25
mm, 0.45 µm), and filtrate was freeze-dried to afford 5.0 mg of
BSA-alkyne. MALDI-TOF: 68,973 (MALDI-TOF of BSA
starting material; 66,435). Molecular weight increase: 2538,
Molecular weight alkyne: 211, 2538/(211-18)=13.2, The BSA-
Alkyne has 13.2 residues of alkyne.
VS-BSA: 5.0 mg of BSA-alkyne obtained above was
individual bleed samples, the titers were measured by making 8
serial 2-fold dilutions starting at 1 : 200 through 1 : 25,600, and
calculated as the dilution where 50 % of the max absorbance
value was achieved.
dissolved in 5.0 mL of 100 mM sodium phosphate (pH 7.0), and
mixed well. The mixture was centrifuged (10,000 rpm, 5 min),
and then supernatant was filtered through Whatman PURADISC
(25 mm, 0.45 µm). The protein concentration of this filtrate was
1.0 mg/mL (Nanodrop, A280, BSA mode). To 1297.5 µL of this
BSA-alkyne solution was added 30 µL of 20 mM compound VS-
azide (15) in DMSO, 22.5 µL of premixed solution of 20 mM
4.6 Cross reactivity ELISA
PBS containing 0.1 % tween and 5% BSA in TBS containing 0.1
% tween were used as washing buffer or binding buffer. 200
ng/well/50 µL of corresponding BSA-DNP, BSA-DK, OVA-VS
conjugate was coated to half well ELISA plates (Corning)
overnight at 4 oC. After wells were rinsed with washing buffer
and subsequently blocked with 5% BSA in TBS containing 0.1 %
tween for 1.5 h at 37 oC, sample conjugates diluted in binding
buffer were added and incubated for 1 h at 37 oC. After washing,
secondary antibody Anti-Mouse IgG whole molecule alkaline-
phosphate conjugate were added and incubated for 1h at 37 oC.
Subsequently, the bound sample was detected according to the
manufactures (Sigma) recommendations.
CuSO4-5H2
75 µL of
100 mM aminoguanidine hydrochloride in H2O and 75 µL of
sodium L-ascorbate in H2O, and then mixed well, kept at rt for
1h. The reaction mixture was applied to ultra filtration (Amicon
ultra MWCO 10,000) and centrifuged (3,000 rpm, 10 min). 3.5
mL of H2O was added to this solution, and centrifuged (3,000
rpm, 20 min) to remove small molecule reagents. This procedure
was repeated 3 times, and the desired BSA-VS solution in H2O
(13.7 mg/mL, 118 µL) was obtained. MALDI-TOF MS gave
molecular weight of VS-BSA = 71,313. Molecular weight
increase from BSA-Alkyne: 4,828, Molecular weight VS-azide:
527, 4,828/527=4.4. The VS-BSA has 4.4 residues of VS.
VS-OVA: VS-OVA was prepared with the similar manner
with VS-BSA via OVA-alkyne. Briefly, MALDI-TOF of OVA:
44,411, OVA-Alkyne: [M + H] = 46,318. Molecular weight
increase: 1907, Molecular weight alkyne: 211, 1,907/(211-
18)=9.9, The OVA-Alkyne has 9.9 residues of alkyne.
4.7 In vitro Programming of anti-hapten antibodies in anti-
serum by adaptor ligands
50 µL of diluted anti-serum in PBS (1 : 500) was mixed with 10
µL of the adaptors (1, 2 or 3
incubated for 3 h at rt. Coastar 96-well ELISA plate (Corning)
were coated with 200 ng of Streptavidin in 50 µL PBS and
incubated overnight at 4 oC. After blocking with 150 µL of 3%
BSA in PBS for 2 h at 37 oC, 60 µL of anti-serum treated with
the adaptors was added into each well and the plate was
incubated for 2 h at 37 oC. Washing and detection were
VS-OVA MALDI-TOF: 46,571, Molecular weight increase of
mono VS-OVA from OVA-Alkyne: 591, Molecular weight VS-
azide: 527, 591/527=1.1. The VS-OVA has 1.1 residues of VS.
4.3 BCA assay for determination of protein concentration

performed essentially as described in our previous paper14 using
alkaline phosphatase-conjugated goat anti-mouse IgG antibody
(diluted 1 : 1,000 in 1 % BSA/PBS).
Acknowledgments
This study was supported in part by The Skaggs Institute for
Chemical Biology and NIH pioneer award. This work was
supported by the National Institutes of Health (DP1CA174426).
MN thanks the Uehara Memorial Foundation. We are grateful to
Diane M. Kubitz for animal work, and Shigehiro Asano and
Tsubasa Inokuma for collecting supporting data. We appreciate
Yuki Goto, Nathan O’Brien and Richard Obexer for valuable
discussions. CFB passed away during a writing of the
manuscript. MN has devotion and truly appreciation to dearest
professor.
4.8 In vitro programming of monoclonal antibodies with
adaptor ligands and their i.p. injections
The in vitro programming of mAb 38C2 with DK-biotin (2)
was performed as previously described.21 anti-DNP monoclonal
antibody (9H8.1) was purchased from Merk-Milllipore
(MAB2223). The 9H8.1 antibody with the adaptors DNP-biotin
(1) were incubated with 2.2 equivalents of the adaptors (1) in
phosphate-buffered saline solution (pH 7.4) at rt for 2 h. ELISA
assays against immobilized streptavidin with both DNP and DK
conjugates. To estimate a concentration of biotin-hapten antibody
conjugate in vivo, the standard curves were prepared for both
DNP and DK by ELISAs. These conjugates were used without
purification for further in vivo experiments. The
immunoconjugates were i.p. injected into non-immunized balb/c
mice with a single dose of 1 mg/kg each, respectively. Tail vein
blood samples were collected at 2, 4, 8, 24, 48 and 72 h after
injection. Formation of serum immunocomplexes were quantified
by ELISA. PBS containing 0.1 % tween and 5% BSA in TBS
containing 0.1 % tween was used as washing buffer or binding
buffer. 200 ng/well of streptavidin was coated to half well
ELISA plates (Corning) overnight 4 oC. After wells were rinsed
with washing buffer and subsequently blocked with 5 % BSA in
TBS containing 0.1 % tween for 1 h at 37 oC, sample diluted in
binding buffer were added and incubated for 1 h at 37 oC. After
washing, secondary antibody Anti-Mouse IgG whole molecule
alkaline-phosphate conjugate were added and incubated for 1h at
37 oC. Subsequently, the bound sample was detected according to
the manufactures (Sigma) recommendations.
Supplementary Material
Supplementary material （details for chemical synthesis,
figures related in vitro animal study, NMR and MALDI-TOF MS
spectra）associated with this article can be found.
References and notes
1.
2.
3.
Ehreth, J. Vaccine 2003, 21, 596–600.
Baxter, D. Occup. Med. (Chic. Ill). 2007, 57, 552–556.
Mcenaney, P. J.; Parker, C. G.; Zhang, A. X.; Spiegel, D. A. ACS
Chem. Biol. 2012, 7, 1139–1151.
4.
5.
Jones, L. H. Nat. Chem. 2015, 7, 952–960.
Lu, Y.; Low, P. S. Cancer Immunol. Immunother. 2002, 51, 153–
162.
6.
7.
Lu, Y.; You, F.; Vlahov, I.; Westrick, E.; Fan, M.; Low, P. S.;
Leamon, C. P. Mol. Pharmaceurics 2007, 4, 2432–2443.
Dubrovska, A.; Kim, C.; Elliott, J.; Shen, W.; Kuo, T. H.; Koo, D.
I.; Li, C.; Tuntland, T.; Chang, J.; Groessl, T.; Wu, X.; Gorney, V.;
Ramirez-Montagut, T.; Spiegel, D. A.; Cho, C. Y.; Schultz, P. G.
ACS Chem. Biol. 2011, 6, 1223–1231.
8.
Rullo, A. F.; Fitzgerald, K. J.; Muthusamy, V.; Liu, M.; Yuan, C.;
Huang, M.; Kim, M.; Cho, A. E.; Spiegel, D. A. Angew. Chemie -
Int. Ed. 2016, 55, 3642–3646.
Amato, R. J.; Shetty, A.; Lu, Y.; Ellis, P. R.; Mohlere, V.;
Carnahan, N.; Low, P. S. J. Immunother. 2014, 37, 237–244.
Wanger, J.; Lerner, R. A.; Barbas, C. F., Ⅲ Science 1995, 270,
1797–1800.
4.9 Oral administration
The adaptors 1-3 were solubilized in PEG 300 at a dose of 100
mg/kg against 30 g weight mouse. Then, these soluble adaptors
were orally administrated to the pre-immunized or non-
9.
10.
11.
12.
immunized mice, which were rested for 6 months after the final
injection. Tail vein blood was collected at 2, 4, 8, 24, 48, 72 and
96 h after oral administration and the amount of biotinylated
immunoconjugates at each time point was evaluated by ELISA.
For ELISAs, PBS containing 0.1 % tween and 5 % BSA in TBS
containing 0.1 % tween were used as washing buffer and binding
buffer, respectively. 200 ng/well of streptavidin was coated to
half well ELISA plates (Corning) overnight 4 oC. After wells
were rinsed with washing buffer and subsequently blocked with 5
% BSA in TBS containing 0.1 % tween for 1 h at 37 oC, sample
diluted in binding buffer were added and incubated for 1 h at 37
oC. After washing, secondary antibody Anti-mouse IgG whole
molecule alkaline-phosphate conjugate were added and incubated
for 1 h at 37 oC. Subsequently, the bound sample was detected
according to the manufactures (Sigma) recommendations.
Zhong, G.; Lerner, R. A.; Barbas, C. F., Ⅲ Angew. Chemie Int. Ed.
1999, 38, 3738–3741.
Zhong, G.; Shabat, D.; List, B.; Anderson, J.; Sinha, S. C.; Lerner,
R. A.; Barbas, C. F., Ⅲ Science 1998, 273, 2481–2484.
Rader, C. Trends Biotechnol. 2014, 32, 186–197.
Popkov, M.; Gonzalez, B.; Sinha, S. C.; Barbas, C. F., Ⅲ Proc.
Natl. Acad. Sci. 2009, 106, 4378–4383.
Ortega, E.; Kostovetzky, M.; Larralde, C. Mol. Immunol. 1984, 21,
883–888.
Palmer, J. T.; Rasnick, D.; Klaus, J. L.; Brömme, D. J. Med. Chem.
1995, 38, 3193–3196.
Renslo, A. R.; McKerrow, J. H. Nat. Chem. Biol. 2006, 2, 701–710.
Choe, Y.; Brinen, L. S.; Price, M. S.; Engel, J. C.; Lange, M.;
Grisostomi, C.; Weston, S. G.; Pallai, P. V.; Cheng, H.; Hardy, L.
W.; Hartsough, D. S.; McMakin, M.; Tilton, R. F.; Baldino, C. M.;
Craik, C. S. Bioorganic Med. Chem. 2005, 13, 2141–2156.
Jaishankar, P.; Hansell, E.; Zhao, D. M.; Doyle, P. S.; McKerrow,
J. H.; Renslo, A. R. Bioorganic Med. Chem. Lett. 2008, 18, 624–
628.
13.
14.
15.
16.
17.
18.
19.
4.10 Isotyping-ELISA
20.
21.
22.
23.
24.
Hong, V.; Presolski, S. I.; Ma, C.; Finn, M. G. Angew. Chemie - Int.
Ed. 2009, 48, 9879–9883.
Sinha, S. C.; Das, S.; Li, L.; Lerner, R. A.; Barbas, C. F., Ⅲ Nat.
Protoc. 2007, 2, 449–56.
Lipinski, C.A.; Lombardo, F.; Dominy, B.W.; Feeney, P. J. Adv.
Drug Deliv. Rev. 2012, 64, 4–17.
Al-Hilal, T. A.; Alam, F.; Byun, Y. Adv. Drug Deliv. Rev. 2013, 65,
845–864.
Ramanathan, S.; Pooyan, S.; Stein, S.; Prasad, P. D.; Wang, J.;
Leibowitz, M. J.; Ganapathy, V.; Sinko, P. J. Pharm. Res. 2001, 18,
950–956.
Chae, S. Y.; Jin, C.; Shin, H. J.; Youn, Y. S.; Lee, S.; Lee, K. C.
Bioconjugate Chem. 2008, 19, 334–341.
Water and 3 % BSA in PBS were used as washing buffer and
binding buffer, respectively. 25 µL of 2.5 mg/ml BSA-DNP
conjugate solution was coated to half well ELISA plates
(Corning) overnight at 4 oC. After wells were rinsed with
washing buffer they were blocked with 3 % BSA in PBS for 1.5
h at 37 oC. The 2nd bleed titer samples diluted in 1 % BSA in PBS
buffer were added and incubated for 1.5 h at 37 oC. After
washing, Anti-mouse IgA-alkaline phosphate antibody produced
in goat in 1% BSA in PBS were added and incubated for 1 h at
37 oC. Subsequently, the bound sample was detected according to
the manufactures (Sigma) recommendations.
25.

26.
27.
Janoria, K. G.; Boddu, S. H. S.; Wang, Z.; Paturi, D. K.; Samanta,
S.; Pal, D.; Mitra, A. K. J. Ocul. Pharmacol. Ther. 2009, 25, 39–49.
Ban, H.; Nagano, M.; Gavrilyuk, J.; Hakamata, W.; Inokuma, T.;
Barbas, C. F., Ⅲ Bioconjug. Chem. 2013, 24, 520–532.
Corthésy, B. Front. Immunol. 2013, 4, 1–11.
30.
Bhat, A.; Laurent, O.; Lappe, R. In Fusion Protein Technologies for
Biopharmaceuticals: Applications and Challenges, Chapter 38;
2013.
Jung, H.; Yang, T.; Lasagna, M. D.; Shi, J.; Reinhart, G. D.;
Cremer, P. S. Biophys. J. 2008, 94, 3094–3103.
Zhang, A. X.; Murelli, R. P.; Barinka, C.; Michel, J.; Cocleaza, A.;
Jorgensen, W. L.; Lubkowski, J.; Spiegel, D. A. J. Am. Chem. Soc.
2010, 132, 12711–12716.
31.
32.
28.
29.
Hayakawa, M.; Toda, N.; Carrillo, N.; Thornburg, N. J.; Crowe, J.
E.; Barbas, C. F., Ⅲ ChemBioChem 2012, 13, 2191–2195.