Construction of a 19F-lectin biosensor for glycoprotein imaging by using affinity-guided DMAP chemistry

Article abstract of DOI:10.1016/j.bmcl.2011.06.038

In this study, assisted by affinity-guided DMAP strategy, we developed a novel ¹⁹F-modified lectin as a biosensor for specific detection and imaging of glycoproteins. Exploited the large chemical shift anisotropy property of ¹⁹F nucl

Full text of DOI:10.1016/j.bmcl.2011.06.038

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4393–4396
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Construction of a 19F-lectin biosensor for glycoprotein imaging by using
affinity-guided DMAP chemistry
Yedi Sun a, Yousuke Takaoka a, Shinya Tsukiji b, Michiko Narazaki c, Tetsuya Matsuda c, Itaru Hamachi a,d,
⇑
a Department of Synthetic Chemistry and Biological Chemistry, Graduate School of Engineering, Kyoto University, Katsura, Kyoto 615-8510, Japan
b Top Runner Incubation Center for Academia-Industry Fusion, Nagaoka University of Technology, 1603-1 Kamitomioka, Nagaoka, Niigata 940-2188, Japan
c Department of Systems Science, Graduate School of Informatics, Kyoto University, 36-1, Yoshida-Honmachi, Sakyo-ku, Kyoto 606-8501, Japan
d Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, 5 Sanbancho, Chiyoda-ku, Tokyo 102-0075, Japan
a r t i c l e i n f o
a b s t r a c t
In this study, assisted by affinity-guided DMAP strategy, we developed a novel 19F-modified lectin as a
biosensor for specific detection and imaging of glycoproteins. Exploited the large chemical shift anisot-
ropy property of 19F nuclei, glycoproteins detected by our 19F-biosensor are signatured by broadened
peaks in 19F NMR, hence enabled the distinction between glycoproteins and small molecule saccharides.
Such signal on/off switching was also applied to glycoprotein imaging by 19F MRI.
Ó 2011 Elsevier Ltd. All rights reserved.
Article history:
Received 9 May 2011
Revised 7 June 2011
Accepted 10 June 2011
Available online 17 June 2011
Keywords:
Protein–Protein interaction
Glycosylation
Biosensor
19F NMR/MRI
Protein labeling
19F NMR/MRI is anticipated to serve as an intriguing detection
alternative to the 1H NMR/MRI owing to its high sensitivity (83%
relative to 1H) and the absent background signals in animal
bodies.1–3 In addition, high environmental sensitivity is another
attractive property of 19F nuclei. Taking advantages of these
strengths, research on protein-based 19F biosensors started sprout-
ing in recent years, where genetic or non-genetic methods were
elaborately utilized to site-specifically incorporate 19F-probes into
proteins.4,5 In most cases, such biosensors sensitively monitored
protein conformational change and/or ligand binding events
through the chemical shift change in 19F NMR spectroscopy.
We herein propose a novel 19F-based lectin biosensor for selec-
tive detection of glycoproteins. Protein glycosylation, one of the
most important and complicated post-translational modifications,
is known to be involved in controlling a wide range of protein sta-
bilities and functions.6 Therefore, biosensors those are able to spe-
cifically detect glycoprotein are undoubtedly important from both
fundamental science and clinical diagnosis perspective. In our
strategy, lectins, a family of sugar-binding proteins, were em-
ployed as the desired sensor scaffold owing to their remarkable
specificity for saccharides. We intended to exploit the large chem-
ical shift anisotropy property of 19F nuclei which intrinsically
yields in the intimate relationship between the transverse relaxa-
tion time of 19F NMR signals and the apparent molecular mass
(Mr) to 19F-probe.7 It is thus reasonably expected that glycopro-
teins detection could be carried out through the transverse relaxa-
tion time change of 19F-labeled lectins, and we also expect to
distinguish glycoproteins from other small molecule saccharides
on the basis of their Mr difference (Fig. 1a).
Assisted by affinity-guided DMAP (AGD) strategy previously re-
ported by us (Fig. 1b),8,9 construction of this unprecedented 19F
biosensor was achieved by incorporating a 19F-probe with strict
site-specificity in order to generate the sharp NMR signal of 19F-
modified lectin. Congerin II (CongII), a galactin isolated from skin
mucus of conger eel,10 was selected as the model lectin for the pur-
pose of principle validation. Based on the knowledge that CongII
selectively binds to Lactose/LacNAc with moderate affinity, the
AGD reagent consisting of lactose for specific CongII recognition
and DMAP moiety for acyl-transfer catalysis was designed (1,
Fig. 1b).8 To fulfill the intention of introducing 19F probes to lectin,
two thioester-type acyl donors containing 3,5-bis(trifluoro-
methyl)benzene moiety that carries six magnetically equivalent
19F nuclei were synthesized and tested (2 and 3, Fig. 1b). Labeling
reaction was carried out in buffer solution at 37 °C. After 4 h incu-
bation, the labeling yield monitored by MALDI-TOF MS achieved
60% using acyl donors 2 whereas the labeling yield associated with
shorter linker type acyl donors 3 was only about 30% (Fig. S1). Sub-
sequent NMR measurements revealed 2-labeled CongII as a sharp
single peak at À62.7 ppm (Fig. 2a) whereas 3-labeled CongII failed
to produce any NMR peaks corresponding to the 19F probe (Fig. 2b).
In the contrary to 2-labeled CongII, randomly modified CongII pre-
⇑
Corresponding author. Tel.: +81 75 383 2754; fax: +81 75 383 2759.
E-mail address: ihamachi@sbchem.kyoto-u.ac.jp (I. Hamachi).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.06.038

4394
Y. Sun et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4393–4396
Figure 1. AGD catalyst-mediated selective and site-specific chemical protein labeling. (a) Schematic illustration of 19F based lectin biosensor construction strategy. Signal of
19F probe is expected to change upon the specific glycoprotein binding due to the apparent Mr change whereas no such NMR signal change is assumed upon the small sugar
molecule binding. (b) Chemical structures of AGD catalyst 1, and acyl donors 2 and 3 used in this study.
Figure 2. (a–c) 19F NMR spectra of purified 19F-labeled CongII (50
prepared with an excess amount of 2 alone for long time incubation. (d,e) 19F NMR spectra of 2-CongII upon glycoprotein binding experiments. (d) 19F NMR spectra of 2-
lM); (a) 2-CongII prepared with 1 and 2, (b) 3-CongII prepared with 1 and 3, (c) randomly-labeled CongII
CongII (50
CongII (50
lM) alone, in the presence of ASF (50 l
M) and in the presence of both ASF and Lactose (2 mM) (top, middle and bottom, respectively). (e) 19F NMR spectra of 2-
l
M) alone and in the presence of RiboB (50 M) (top and bottom, respectively). All experiments were performed in 50 mM HEPES buffer (pH 8.0, 0.2 mM TFA, 10%
l
D2O (v/v)) at 25 °C.

Y. Sun et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4393–4396
4395
Figure 3. Magnetic resonance images of the samples containing 2-CongII. (a) 1H MR images (top), 19F MR images corresponding to TFA (around at À75.6 ppm, middle) and 19
F
MR images corresponding to 2-CongII (around at À62.7 ppm, bottom) of the samples containing 2-CongII alone (100
lM) (i), in the presence of ASF (100
l
M) (ii) and in the
presence of both ASF and lactose (4 mM) (iii) (b) 19F MR images corresponding to 2-CongII (around at À62.7 ppm) of the samples containing 2-CongII alone (100
lM) (iv), in
the presence of RiboB (100 lM) (v) and in the presence of both RiboB and lactose (4 mM) (vi). All the experiments were performed in 50 mM HEPES buffer (pH 8.0, 0.2 mM
TFA) at 25 °C.
pared by incubating CongII with excessive amount of acyl donor 2
gave multiple weakly-intense NMR peaks (Fig. 2c), strongly indi-
cating the importance of site-specific probe incorporation. Assisted
by tyrpsin-digestion in combination with MS/MS analysis, labeling
site of 2-CongII was determined to be dominantly Tyr51 which lo-
cates at the sugar-binding pocket (Fig. S2).8 These results alto-
gether implied that DMAP-catalyzed protein engineering method
works successfully for 19F-labeled lectin construction.
Taking labeling efficiency as well as NMR peak quality into con-
sideration, 2-labeled CongII was employed for following glycopro-
tein detection studies. Glycoproteins that possess galactoside-rich
saccharides at glycan terminals are anticipated to interact with
CongII.10a As expected, gentle mixing of 2-CongII with asialofetuin
(ASF), a N-type glycoprotein with binding-affinity of 1.3 Â 107 MÀ1
towards CongII,11 resulted in remarkable peak broadening depicted
in 19F NMR spectrum (Fig. 2d). Similar result was obtained when 2-
CongII biosensor was used for another known CongII-binding gly-
coprotein ovalbumin (OVA)12 detection (Fig. S3a). In contrast, 19F
NMR peak intensity of 2-CongII remained unaffected in the pres-
ence of ribonuclease B (RiboB) (Fig. 2e), a mannose rich glycopro-
tein,13 or b-galactosidase (b-Gal) (Fig. S3b), a non-glycoprotein
with large molecular weight. This unmistakable 19F NMR differ-
ence hence strongly supported that the distinctive peak broaden-
ing resulted from ASF or OVA binding is ascribed to specific
interaction between 2-CongII and its binding glycoprotein part-
ners, demonstrating that 2-CongII works successfully for specific
glycoprotein detection. These signal changes were observed in a
saturation manner, suggesting that the present method would be
theoretically applicable for quantitative determination of the gly-
coprotein concentration. Interestingly, the broadened 19F NMR
peak of 2-CongII with ASF was almost completely recovered to
its natural shrarpness again by the subsequent addition of an ex-
cess amount of lactose (Fig. 2d, bottom). These results therefore
not only provided the evidence of specific interaction between 2-
CongII and galactoside sugars on ASF, but also brought up the
importance of apparent molecular mass (Mr) on the peak broaden-
ing phenomenon. Indeed, Mr of 2-CongII-ASF complex is estimated
to be 100 kDa, which is more than threefold larger than 2-CongII
alone or 2-CongII-lactose complex (about 30 kDa, CongII exist in
a homodimer10a in the solution). Such significant Mr increment
therefore substantially reduced the transverse relaxation time
(T2) of 19F nucleus according to chemical shift anisotropy princi-
ple,7 which in turn was reflected in NMR as broadened peak. Since
small CongII-binding molecules such as lactose cause negligible Mr
increment, NMR peak sharpness was preserved. In fact, T2 mea-
surements further supported the proposed glycoprotein detection

4396
Y. Sun et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4393–4396
mechanism, that is, the T2 value of 2-CongII reduced from 0.34 sec
to 0.14 sec upon the binding of ASF and recovered to 0.62 sec right
after the subsequent addition of lactose.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.06.038.
Moving on from 19F NMR experiments, more visually apprehen-
sible and clinically applicable MRI studies were carried out on gly-
coprotein detections with the employment of our 19F-based lectin
biosensor. For each set of experiment, MR images of proton (1H,
Fig. 3a, top), 19F attributed to lectin biosensor (Fig. 3b, bottom)
and 19F contributed by internal standard TFA (Fig. 3a, middle) were
collected from the samples containing 2-CongII. Prior to addition of
glycoproteins, all 1H and 19F images of samples represented them-
selves as well-recognizable images (samples (i) and (iv)). After
adding ASF, the image attributed to 19F biosensor diminished, leav-
ing only background signal level to be revealed (Fig. 3a, bottom,
sample (ii)). Such phenomenon is due to significantly shortened
T2 value as proposed previously and it well agreed with the corre-
sponding NMR results. The intensity was resumed later in the pres-
ence of excess amount of lactose (Fig. 3a, bottom, sample (iii)). On
the other hand, mixing of 19F biosensor and RiboB caused no much
impact on the image intensity (Fig. 3b, samples (v) and (vi)), a pre-
dictable outcome judging from NMR results. Therefore, these re-
sults clearly indicated that our turn-off type 19F-biosensor is a
powerful tool for visualizing glycoprotein specifically.
References and notes
1. (a) Rothbacher, U.; Moats, R.; Jacobs, R. E.; Fraser, S. E.; Meade, T. J. Nat.
Biotechnol. 2000, 18, 321–325; (b) Sosnonik, D. E.; Weissleder, R. Curr. Opin.
Biotechnol. 2007, 18, 4–10; (c) Kiessling, F.; Morgentstern, B.; Zhang, C. Curr.
Med. Chem. 2007, 14, 77–91.
2. (a) Danielson, M. A.; Falke, J. J. Ann. Rev. Biophys.Biomol.Struct. 1996, 25, 163–
195; (b) Yu, J.; Kodibagkar, V. D.; Cui, W.; Mason, R. P. Curr. Med. Chem. 2005,
12, 819–848.
3. (a) Yu, J.; Otten, P.; Ma, Z.; Cui, W.; Liu, L.; Mason, R. P. Bioconjugate Chem. 2004,
15, 1334–1341; (b) Higuchi, M.; Iwata, N.; Matsuba, Y.; Sato, K.; Sasamoto, K.;
Saido, T. C. Nat. Neurosci. 2005, 8, 527–533; (c) Mizukami, S.; Takikawa, R.;
Sugihara, F.; Hori, Y.; Tochio, H.; Wälchli, M.; Shirakawa, M.; Kikuchi, K. J. Am.
Chem. Soc. 2008, 130, 794–795; (d) Takaoka, Y.; Sakamoto, T.; Tsukiji, S.;
Narazaki, M.; Matsuda, T.; Tochio, H.; Shirakawa, M.; Hamachi, I. Nat. Chem.
2009, 1, 557–561.
4. (a) Jackson, J. C.; Hammill, J. T.; Mehl, R. A. J. Am. Chem. Soc. 2007, 129, 1160–
1166; (b) Hammill, J. T.; Miyake-Stoner, S.; Hazen, J. L.; Jackson, J. C.; Mehl, R. A.
Nat. Protoc. 2007, 2, 2601–2607.
5. (a) Tsukiji, S.; Miyagawa, M.; Takaoka, Y.; Tamura, T.; Hamachi, I. Nat. Chem.
Biol. 2009, 5, 341–343; (b) Takaoka, Y.; Sun, Y.; Tsukiji, S.; Hamachi, I. Chem. Sci.
2011, 2, 511–520.
6. (a) Rudd, P. M.; Elliott, T.; Cresswell, P.; Wilson, I. A.; Dwek, R. A. Science 2001,
291, 2370–2376; (b) Bertozzi, C. R.; Kiessling, L. L. Science 2001, 291, 2357–
2364; (c) Ohtsubo, K.; Marth, J. D. Cell 2006, 126, 855–867.
7. Grage, S. L.; Dürr, U. H. N.; Afonin, S.; Mikhailiuk, P. K.; Komarov, I. V. Ulrich A. S.
J. Magn. Reson. 2008, 191, 16–23.
8. Koshi, Y.; Nakata, E.; Miyagawa, M.; Tsukiji, S.; Ogawa, T.; Hamachi, I. J. Am.
Chem. Soc. 2008, 130, 245–251.
9. (a) Hamachi, I.; Nagase, T.; Shinkai, S. J. Am. Chem. Soc. 2000, 122, 12065–
12066; (b) Nakata, E.; Koshi, Y.; Koga, E.; Katayama, Y.; Hamachi, I. J. Am. Chem.
Soc. 2005, 127, 13253–13261; (c) Ishii, M.; Matsumura, S.; Toshima, K. Angew.
Chem. Int. Ed. 2007, 46, 8396–8399.
10. (a) Kamiya, H.; Muramoto, K.; Goto, R. Dev. Comp. Immunol. 1988, 12, 309–318;
(b) Mitsuyama, C. S.; Ito, Y.; Konno, A.; Miwa, Y.; Ogawa, T.; Muramoto, K.;
Shirai, T. J. Mol. Biol. 2005, 347, 385–397.
11. Muramoto, K.; Kagawa, D.; Sato, T.; Ogawa, T.; Nishida, Y.; Kamiya, H. Comp.
Biochem. Phys. Part B 1999, 123, 33–45.
12. Yamashita, K.; Tachibana, Y.; Kobata, A. J. Biol. Chem. 1978, 253, 3862–3869.
13. Berman, E.; Walters, D. E.; Allerhand, A. J. Biol. Chem. 1981, 256, 3853–3857.
In conclusion, we succeeded to develop the 19F-based lectin bio-
sensor for specific detection of glycoproteins by using the saccha-
ride-tethered DMAP catalyst and 19F-acyl donors. Constructed
biosensor could distinguish glycoproteins from small molecule
saccharides for the peak broadening in 19F NMR directly relates
to the Mr of each glycoconjugate. Furthermore, this distinct signal
on/off switching was applied to visualize the lectin-glycoprotein
interaction in a 19F MRI phantom. We envision further applications
of such chemically modified 19F-biosensor for detection of various
protein–protein interaction pairs.
Acknowledgments
Y.S. and Y.T. acknowledge JSPS Research Fellowshios for Young
Scientists. This work was partly supported by the CK Integrated
Medical Bioimaging Project (MEXT) and CREST (Japan Science
and Technology Agency).