Experimental and computational studies of the recognition of amino acids by galactosyl-imine and -amine derivatives: An attempt to understand the lectin-carbohydrate interactions

Article abstract of DOI:10.1021/jo0700979

(Chemical Equation Presented) A galactosyl-naphthyl-imine-based derivative, 1-(β-D-galactopyranosyl-1′-deoxy-1′-iminomethyl)-2- hydroxynaphthalene (GNI), and a galactosyl-naphthyl-amine-based derivative, 1-(galactopyranosyl-1′-deoxy-1′-aminomethyl)-2-hydroxynaphthalene (GNA), possessing an ONO binding core were studied for their recognition of naturally occurring amino acids using fluorescence and absorption spectroscopy, and the corresponding association constants were derived for the complexes formed. The complexes formed between GNI/GNA and amino acids were supported by electrospray ionization mass spectrometry (ESI/MS). The structures of the complexes were optimized by computational studies using density functional theory, and stabilization energies were computed for the complexes to substantiate the interactions present between GNI/GNA and amino acid. The interactions were found to be primarily hydrogen bonding in nature. These interactions are reminiscent of those present in the lectin-carbohydrate and glycosidase substrate. Thus, the carbohydrate moiety present in GNI shows high specificity toward the -COOH group of the amino acid, which may be relevant to such interactions present between the carbohydrates and the polypeptides.

Full text of DOI:10.1021/jo0700979

Experimental and Computational Studies of the Recognition of
Amino Acids by Galactosyl-imine and -amine Derivatives: An
Attempt to Understand the Lectin-Carbohydrate Interactions
†
†
†
‡
Rohit Ahuja, Nitin K. Singhal, Balaji Ramanujam, Maddula Ravikumar, and
,
†
Chebrolu P. Rao*
Bioinorganic Laboratory, Department of Chemistry, Indian Institute of Technology Bombay,
Mumbai 400 076, India, and Analytical Chemistry DiVision, Indian Institute of Chemical Technology,
Hyderabad 500 007, India
cprao@iitb.ac.in
ReceiVed January 17, 2007
A galactosyl-naphthyl-imine-based derivative, 1-(â-D-galactopyranosyl-1′-deoxy-1′-iminomethyl)-2-
hydroxynaphthalene (GNI), and a galactosyl-naphthyl-amine-based derivative, 1-(galactopyranosyl-1′-
deoxy-1′-aminomethyl)-2-hydroxynaphthalene (GNA), possessing an ONO binding core were studied
for their recognition of naturally occurring amino acids using fluorescence and absorption spectroscopy,
and the corresponding association constants were derived for the complexes formed. The complexes
formed between GNI/GNA and amino acids were supported by electrospray ionization mass spectrometry
(ESI/MS). The structures of the complexes were optimized by computational studies using density
functional theory, and stabilization energies were computed for the complexes to substantiate the
interactions present between GNI/GNA and amino acid. The interactions were found to be primarily
hydrogen bonding in nature. These interactions are reminiscent of those present in the lectin-carbohydrate
and glycosidase substrate. Thus, the carbohydrate moiety present in GNI shows high specificity toward
the -COOH group of the amino acid, which may be relevant to such interactions present between the
carbohydrates and the polypeptides.
Introduction
important roles in the immune system found exclusively on
pathogens/cell surfaces. As a result of this, lectins are
2
Lectins are carbohydrate-binding proteins of nonimmune
origin that agglutinate cells and/or precipitate glycoconjugates
and are isolated mainly from plant and animal sources. These
potentially involved in a wide range of applications in the fields
3
of pharmacology, immunology, and cancer therapy. In all these,
1
the main property of the lectin is to recognize carbohydrates
serve many different biological functions and are known to play
(1) (a) Ahuja, R.; Singhal, N. K.; Rao, C. P. Khim./Chem.: Bulgarian
*
To whom correspondence should be addressed. Phone: 91 22 2576 7162.
J. Chem. Educ. 2006, 15, 275 and references cited therein. (b) Sharon, N.
Trends Biochem. Sci. 1993, 18, 221. (c) Rini, J. M. Annu. ReV. Biophys.
Biomol. Struct. 1995, 24, 551. (d) Audette, G. F.; Olson, D. J. H.; Ross, A.
R. S.; Quail, J. W.; Delbaere, L. T. J. Can. J. Chem. 2002, 80, 1010.
Fax: 91 22 2572 3480.
†
Department of Chemistry, Indian Institute of Technology Bombay.
Analytical Chemistry Division, Indian Institute of Chemical Technology.
‡
10.1021/jo0700979 CCC: $37.00 © 2007 American Chemical Society
3
430
J. Org. Chem. 2007, 72, 3430-3442
Published on Web 03/30/2007

Recognition of Amino Acids by Galactosyl DeriVatiVes
a
SCHEME 1
a
Schematic representation of the synthesis of G-NH2, GNI, and GNA from G (i.e., galactose): (i) MeOH, NH3 for 10 h at rt, store at +4 °C for 3 days;
(
ii) galactosyl-C1-deoxy-C1-amine (G-NH2, 3.584 g, 20.02 mmol), 2-hydroxy-1-napthaldehyde (3.612 g, 20.88 mmol), EtOH, reflux, 8 h, yield of 4.91 g,
7
4%; (iii) galactose (1.80 g, 10 mmol), 1-aminomethyl-2-naphthol (1.78 g, 10.3 mmol), EtOH, reflux, 8 h, yield of 2.2 g, 66%.
4
selectively. Recent literature witnessed reports on purification
and characterization of a number of lectins isolated from
different sources and comprising varying carbohydrate specifici-
spectroscopy, where the naphthyl moiety acts as an intrinsic
chromophore as well as a fluorophore. The species formed
between these derivatives and the amino acids were confirmed
on the basis of electrospray ionization mass spectrometry (ESI/
MS). The complexes formed were further studied by compu-
tational methods using Gaussian 98 and/or Gaussian 03 at
varying levels of calculations, such as semiempirical (PM3),
ab initio (HF), and density functional theory (B3LYP).
5
ties. Of these, several are galactose-based ones. On the basis
of few crystal structures reported in the literature, it is reasonable
to conclude that the recognition of carbohydrates by lectins is
primarily through hydrogen bonding, though such interactions
6
were yet to be proven in a large number of lectins. On the
other hand, glycosidases selectively hydrolyze glycosidic bonds
where the substrate glycoside is expected to interact with the
polypeptide in the active site. To model the interactions
displayed by lectins and glycosidases toward carbohydrates, it
would become critical to study the interactions present between
carbohydrates and amino acids using both experimental and
computational methods.
In our journey to understand the basis of the interactions of
carbohydrates with lectins and glycosidases, we have taken up
a model study involving the galactosyl-naphthyl-imine-based
derivative, viz., 1-(â-D-galactopyranosyl-1′-deoxy-1′-imino-
methyl)-2-hydroxynaphthalene (GNI), and the galactosyl-naph-
thyl-amine-based derivative, viz., 1-(galactopyranosyl-1′-deoxy-
Results and Discussion
The galactosyl derivatives, viz., GNI and GNA, were
synthesized as per that shown in Scheme 1. The details of the
synthesis and the characterization data are presented in the
Experimental Section.
A. Fluorescence Spectral Studies. Although free GNA
exhibits one intense fluorescence emission peak centered at 355
nm, free GNI exhibits two weak emissions centered at 355 and
4
45 nm. Comparison of the fluorescence emission spectra of
naphthalene, â-naphthol, 2-hydroxy-1-napthaldehyde, and GNA
with that of GNI leads to the assignment of the 355 nm band
to that arising from naphthalene and of the 445 nm band to
that arising from a naphthalene in conjugation with an imine
moiety (Supporting Information, S1). The presence of a lone
pair on the nitrogen of -CdN- in GNI results in photoinduced
electron transfer (PET) that quenches the fluorescence in GNI
as compared to that in GNA. Both GNI and GNA were titrated
against all 20 naturally occurring L-amino acids by fluorescence
spectroscopy according to the titration details given in the
Experimental Section.
1
′-aminomethyl)-2-hydroxynaphthalene (GNA), and naturally
occurring amino acids using absorption and fluorescence
(
2) (a) Lis, H.; Sharon, N. Chem. ReV. 1998, 98, 637. (b) Zhu-Salzman,
K.; Hammen, P. K.; Salzman, R. A.; Koiwa, H.; Bressan, R. A.; Murdock,
L. L.; Hasegawa, P. M. Comp. Biochem. Physiol., B 2002, 132, 327. (c)
McGreal, E. P.; Martinez-Pomares, L.; Gordon, S. Mol. Immunol. 2004,
4
1, 1109. (d) Rabinovich, G. A.; Rubinstein, N.; Toscano, M. A. Biochim.
Biophys. Acta, General Subjects 2002, 1572, 274.
3) (a) Dougherty, T. J.; Gomer, C. J.; Henderson, B. W.; Jori, G.; Kessel,
D.; Korbelik, M.; Moan, J.; Peng, Q. J. Natl. Cancer Inst. 1998, 90, 889.
b) Bonnett, R. Chem. Soc. ReV. 1995, 24, 19. (c) Klyashchitsky, B. A.;
Nechaeva, I. S.; Ponomaryov, G. V. J. Controlled Release 1994, 29, 1.
4) (a) Weis, W. I.; Drickamer, K. Annu. ReV. Biochem. 1996, 65, 441.
b) Drickamer, K. Structure 1997, 5, 465. (c) Sharma, V.; Surolia, A. J.
(
(
Fluorescence Titration of GNI with Amino Acids. Fluo-
rescence spectra obtained during the titration of GNI with Glu
are shown in Figure 1a. The titration results in quenching the
fluorescence of the 355 nm band as the ratio of the added Glu
increases, however, show no significant change in the fluores-
cence of the 445 nm band, as can be noticed from the relative
fluorescence ratio (I/I0) plot given in Figure 1b. The results are
clearly suggestive of the interaction of Glu with GNI. On the
other hand, the titration of GNI with His, Leu, and Pro resulted
in an increase of fluorescence intensity of both bands to varying
extents (Supporting Information, S2), indicating that the inter-
acting strengths with GNI vary among these amino acids. The
I/I0 plots for these titrations are shown in Figure 2. Thus, the
fluorescence intensity was either enhanced or quenched depend-
ing upon the amino acid used in the titration of GNI.
(
(
Mol. Biol. 1997, 267, 433. (d) Elgavish, S.; Shaanan, B. Trends Biochem.
Sci. 1998, 22, 462.
(
5) (a) Gupta, N.; Narula, A.; Srivatava, P. S. J. Plant Biochem.
Biotechnol. 2004, 13, 141. (b) Prasad, K. N.; Podder, U. S. Indian J. Agric.
Biochem. 2004, 17, 81. (c) Bertrand, O.; Cochet, S.; Ctron, J. P. J.
Chromatogr., A 1998, 822, 19. (d) Nomura, K.; Ashida, H.; Uemura, N.;
Kushibe, S.; Ozaki, T.; Yoshida, M. Phytochemistry 1998, 49, 667. (e) Latha,
V. L.; Rao, R. N.; Nadimpalli, S. K. Protein Expression Purif. 2006, 45,
2
96. (f) Padma, P.; Komath, S. S.; Nadimpalli, S. K.; Swamy, M. J.
Phytochemistry 1999, 50, 363.
6) (a) Loris, R.; Imberty, A.; Beeckmans, S.; Van, D. E.; Read, J. S.;
Bouckaert, J.; De, G. H.; Buts, L.; Wyns, L. J. Biol. Chem. 2003, 278,
6297. (b) Svensson, C.; Teneberg, S.; Nilsson, C. L.; Kjellberg, A.;
(
1
Schwarz, F. P.; Sharon, N.; Krengel, U. J. Mol. Biol. 2002, 321, 69. (c)
Jeyaprakash, A. A.; Katiyar, S.; Swaminathan, C. P.; Sekar, K.; Surolia,
A.; Vijayan, M. J. Mol. Biol. 2003, 332, 217-28. (d) Kulkarni, K. A.; Sinha,
S.; Katiyar, S.; Surolia, A.; Vijayan, M.; Suguna, K. FEBS Lett. 2005, 579,
All the remaining 16 amino acids showed no appreciable
change in the fluorescence intensity of either band of GNI
6
4
775. (e) Liu, Y.; Misulovin, Z.; Bjorkman, P. J. J. Mol. Biol. 2001, 305,
81. (f) Sanders, D. A. R.; Moothoo, D. N.; Raftery, J.; Howard, A. J.;
Helliwell, J. R.; Naismith, J. H. J. Mol. Biol. 2001, 310, 875. (g) Velloso,
L. M.; Svensson, K.; Pettersson, R. F.; Lindqvist, Y. J. Mol. Biol. 2003,
(Figure 3). The results thus clearly indicate selective association
between GNI and the four amino acids, viz., Glu, His, Leu,
3
34, 845. (h) Hamodrakas, S. J.; Kanellopoulos, P. N.; Pavlou, K.; Tucker,
P. A. J. Struct. Biol. 1997, 118, 23.
and Pro, whereas the nature of the interaction seems to be
J. Org. Chem, Vol. 72, No. 9, 2007 3431

Ahuja et al.
FIGURE 1. Fluorescence titration of GNI with Glu: (a) spectral traces at different additions of Glu; (b) plot of relative fluorescence intensity
o
(I/I ) vs [Glu]/[GNI] mole ratio. Open circles are for the 445 nm band and filled triangles are for the 355 nm band.
FIGURE 2. Plots of (I/I
o
) vs mole ratio during the titration of GNI with amino acids: (a) His; (b) Pro; (c) Leu. Symbols have the same meanings
as those in Figure 1.
TABLE 1. Ka and Slopes Obtained from Benesi-Hildebrand Plots
in the Titration of GNI with the Interacting Amino Acids (aa)
3
55 nm band
Ka (M-1)
slope
nda
445 nm band
aa
Ka (M-1
)
slope
His
Glu
Pro
Leu
381 ( 30
28733 ( 3500
10567 ( 2200
0.806 ( 0.01
a
a
0.45 ( 0.05
nd
nd
a
a
nd
nd
1420 ( 160
22914 ( 520
1.06 ( 0.05
1.02 ( 0.01
30911 ( 700
1.01 ( 0.01
a
nd: not determinable, as the changes observed in fluorescence intensity
during the titration were not significant.
enhancing or in quenching the fluorescence, the titrations with
GNA resulted only in quenching the fluorescence in the case
of the interacting amino acids (Supporting Information, S3). The
titration of GNA with Ala, Cys, or Lys results in quenching
the fluorescence intensity as the ratio of the added amino acid
increases. The (I/I0) plots clearly indicate a rapid quenching as
shown in Figure 4, suggesting that these amino acids interact
with GNA. However, all the remaining 17 amino acids showed
no appreciable change (Figure 5). The extent of fluorescence
quenching follows an order, Lys > Cys > Ala. Thus, the results
clearly suggest that there exists a selective association between
the GNA and the three amino acids, viz., Ala, Cys, and Lys,
and this selectivity is different from that exhibited by GNI, as
expected, owing to the presence of amine (-HN-CH2-) in
the former and imine (-CdN-) in the latter (Scheme 1). On
going from the presence of imine (in GNI) to amine (in GNA),
both the structural and the chemical functional differences
emerge, and this leads to the differences in their recognition
toward amino acids.
o
FIGURE 3. Histogram of (I/I ) at saturation for the titration of GNI
with amino acids. Positive values on the y-axis indicate enhancement
fold in the fluorescence, and the negative values indicate quenching
fold. All the titrations were repeated four times, and the variations
obtained were indicated by error bars in the case of the interacting
amino acids, viz., Glu, His, Pro, and Leu.
somewhat different in the case of Glu as compared to the other
amino acids. This difference in the interaction has also been
noted even in the absorption studies as reported in this paper.
The association constants (Ka) were calculated on the basis
of the Benesi-Hildebrand equation, and the results are listed
in Table 1. The Ka values suggest that the strength of the GNI‚
‚
‚aa interaction follows a trend, viz., Leu > His > Pro based
on the 455 nm band and Leu g Glu . His g Pro based on the
55 nm band. The slopes of these plots yielded a 1:1 complex
3
for Leu, His, and Pro and a 2:1 complex for Glu, where one
Glu interacts with two GNIs.
Fluorescence Titration of GNA with Amino Acids. Al-
though the titration of GNI with amino acids resulted either in
It was observed that the fluorescence emission quenching was
too rapid to fit to the Benesi-Hildebrand equation, and hence
Stern-Volmer analysis was employed to analyze the data. The
occurrence of a nonlinear Stern-Volmer plot implied that a
3432 J. Org. Chem., Vol. 72, No. 9, 2007

Recognition of Amino Acids by Galactosyl DeriVatiVes
FIGURE 4. Plots of (I/I
o
) vs amino acid mole ratio: (a) Ala; (b) Lys; (c) Cys.
TABLE 2. Stern-Volmer Parameters (kq and fa) for the Titration
of GNA with Interacting Amino Acids
kq (M- s-1
)
1
aa
fa
Lys
Cys
Ala
0.7 ( 0.03
0.58 ( 0.03
0.27 ( 0.02
1142 ( 150
3831 ( 350
949 ( 100
4
.92 ns based on the time-resolved fluorescence spectroscopy,
and the data are listed in Table 2.
B. Absorption Spectral Studies. Absorption experiments
were performed to confirm the binding of amino acids with GNI
and GNA, and the details of the titration are given in the
Experimental Section. The absorption bands observed in GNI
and GNA were assigned upon comparison with the absorption
spectra of naphthalene, â-naphthol, and 2-hydroxy-1-napthal-
dehyde (Supporting Information, S4). The bands observed at
220 and 300 nm in the case of GNI and those observed at 220,
277, and 334 nm were assignable to the naphthalene moiety
FIGURE 5. Histogram showing the relative fluorescence intensity at
the saturation during the titration of GNA with different naturally
occurring amino acids. For those amino acids that exhibit quenching,
0
reciprocal (I/I ) was shown which represents the number of times
quenching occurs. The titrations were repeated four times, and the
variations obtained were indicated by error bars in the case of the
interacting amino acids, viz., Cys, Ala, and Lys.
â
and the OH attached to the C of the naphthalene moiety. The
bands observed at 400 and 420 nm were distinctly present in
GNI and not in GNA and hence are assignable to the imine
function in GNI.
proportion of the observed fluorescence is not susceptible to
quenching. Hence, a modified Stern-Volmer expression was
used
Absorption Titration of GNI with Amino Acids. Titration
of GNI with Glu showed a gradual decrease followed by
saturation of the absorbance of the 300 nm band as a function
of added Glu, but no significant changes were observed in the
400 nm band. On the other hand, an increase in the absorbance
was observed in both the bands in the case of the titration with
other amino acids, viz., His, Leu, and Pro. Corresponding
spectral traces were given in the Supporting Information (S5),
and the absorbance plots are shown for all four cases in Figure
7. This behavior is analogous to that observed in the case of
the fluorescence studies. The binding constants derived from
the corresponding log plots are listed in Table 3 and were found
to be in the order Glu > Pro > Leu based on the 300 nm band
and His > Pro > Leu based on the 400 nm band.
F0
1
1
fa
)
+
F0 - F
([Q]K f )
sv a
where F0 is the fluorescence intensity of the free GNA; F is
the same in the presence of an amino acid (quencher); [Q] is
the concentration of the quencher; Ksv is the Stern-Volmer
constant; and fa is the fraction of fluorescence that is susceptible
to quenching. Ksv ) kqτ0, where the kq is the quenching constant
and τ0 is the lifetime of the free ligand GNA. The corresponding
modified Stern-Volmer plots are shown in Figure 6. All these
were well fit with straight lines and hence support the assump-
tion that the species were 1:1. From these plots, fa and kq were
derived using the experimentally measured τ0 for GNA being
From the absorption titrations, it was noted that the intensity
of the 220 nm band decreases as the added [His] increases, as
FIGURE 6. Modified Stern-Volmer plots for the interaction of GNA with Lys, Cys, and Ala. F
0
/(F
0
- F) is plotted against 1/[amino acid].
J. Org. Chem, Vol. 72, No. 9, 2007 3433

Ahuja et al.
FIGURE 7. Absorbance of 300 nm (circles) and 400 nm (filled triangles) bands in the titration of GNI with amino acid: (a) Glu; (b) His; (c) Pro;
and (d) Leu.
FIGURE 8. (a) Absorption spectral traces in the range 200-275 nm for the titration of GNI with His. (b) Relative absorbance of the titration
shown in (a) as a function of [His]/[GNI] mole ratio.
TABLE 3. Association Constants (Ka) Derived from 300 and 400
nm Bands in the Absorption Spectra of the Titration of GNI with
Interacting Amino Acids
either GNI or GNA. Thus, the absorption studies reported in
this paper clearly demonstrate the binding selectivity of a
particular amino acid with GNI or GNA.
Ka (M-1)
C. Time-Resolved Fluorescence Spectroscopy. The fluo-
rescence decay of GNI in water can be fitted to biexponential,
yielding two lifetimes (τ1 and τ2) where the 9.06 ns has a higher
contribution to the lifetime. The lifetime data for the titration
of GNI with the interacting amino acids are presented in Table
aa
300 nm band
400 nm band
Glu
His
Pro
Leu
35481 ( 3200
nda
8317 ( 1100
3162 ( 400
1778 ( 150
a
nd
5011 ( 700
1995 ( 200
5
. Significant increases in the lifetimes were observed in the
presence of amino acids, and they exhibited a trend, viz., Leu
His > Pro, synonymous with steady-state flouorescence
a
nd: not determinable, as the changes observed in OD during the titration
were not significant.
>
results. Although the species with a lower lifetime is present in
a high proportion in the case of Leu, it is the higher-lifetime
species that is present in maximum proportion in the cases of
His and Pro. On the other hand, GNI exhibited quenching of
steady-state fluorescence in the presence of Glu and also
exhibited a concomitant decrease in the lifetime of the species
where both the species were present to almost an equal extent.
Thus, the fluorescence lifetime studies clearly indicated a
difference in the decay of the GNI‚‚‚aa complexes from that
of the simple GNI.
shown in Figure 8a, and the plot shown in Figure 8b yields a
-
1
Ka value of 38 018 M . However, no significant change was
observed in this band with the other three amino acids, viz.,
Glu, Pro, and Leu. Therefore, the change observed in the
absorbance of this band could be a result of the interaction of
the amino acid with the naphthyl moiety. Indeed, the compu-
tational calculations provided a long-range interaction between
the imidazole nitrogen and skeletal carbons of the naphthyl
moiety, as discussed later in this paper.
Absorption Titration of GNA with Amino Acids. Titration
of GNA with Ala showed a gradual decrease followed by
saturation in the absorbance of both the 277 and 334 nm bands.
However, Cys and Lys exhibited an increase in the absorbance
of both of the bands. Corresponding spectral traces were given
in the Supporting Information (S6), and the absorbance plots
are shown for all three cases in Figure 9. The changes in the
absorbance follow a trend, Lys . Cys g Ala. The binding
constants derived from the corresponding log plots are listed in
Table 4.
The fluorescence decay of GNA in MeOH can be fitted to a
single exponential yielding a lifetime of 4.92 ns. As the medium
is changed from MeOH to water, the lifetime increases to 5.18
ns (2:1 volume ratio of MeOH to H2O) followed by 5.38 ns
(1:2 volume ratio of MeOH to H2O) and fits better to a
biexponential decay. In a 1:29 volume ratio of MeOH to H2O,
the decay curve could be fitted to yield two species with
lifetimes of 2.28 and 4.42 ns. This change is indicative of the
dissociation of naphthyl-OH in water. In the titration of GNA
with Ala, the lifetime decreases considerably as the ratio of Ala
to GNA increases and approaches 3.98 ns at a 100:1 ratio. On
Titration of a noninteracting amino acid, viz., Met, exhibited
no significant change in the absorbance of any of the bands of
3434 J. Org. Chem., Vol. 72, No. 9, 2007

Recognition of Amino Acids by Galactosyl DeriVatiVes
FIGURE 9. Absorbance of 277 nm (filled triangles) and 334 nm (circles) bands in the titration of GNA with amino acid: (a) Ala; (b) Cys; and
(c) Lys.
TABLE 4. Ka Obtained for 277 and 334 nm Bands Based on the
Absorption Spectra in the Titration of Interacting Amino Acids with
GNA
Ka (M-1)
aa
277 nm band
334 nm band
Ala
Cys
Lys
6309 ( 700
1778 ( 250
3162 ( 300
nda
5011 ( 350
2344 ( 250
a
nd: not determinable, as the changes observed in OD during the titration
were not significant.
TABLE 5. Fluorescence Lifetime Data for GNI‚‚‚aa Systemsa
system
τ1 (ns)
τ2 (ns)
ø2 b
GNI (water)
GNI + Leu
GNI + His
GNI + Pro
GNI + Glu
0.70 (21%)
2.26 (85%)
1.48 (36%)
1.29 (34%)
0.44 (46.5)
9.06 (79%)
15.2 (15%)
9.23 (64%)
7.99 (66%)
5.46 (54%)
1.195
1.087
1.250
1.321
1.304
a
The values given in brackets indicate the percent of species correspond-
b
2
ing to that lifetime. ø represents goodness of fit.
the other hand, Lys and Cys did not exhibit any significant
change in the lifetimes of GNA.
D. Electrospray Ionization Mass Spectral Studies. GNI‚
‚aa Complexes. An ESI mass spectrum of GNI shows
‚
abundant ions at m/z 334, 667, and 1000 corresponding to the
protonated monomer, dimer, and trimer of GNI, respectively,
as marked in Figure 10a. Mass spectra were also recorded for
the reaction mixtures of GNI and amino acids. All these spectra
+
show the [M + H] ion of amino acid and GNI. In addition to
+
the [M + H] ions, the spectra also include the ions at m/z
4
65, 489, and 449 corresponding to the 1:1 complex of GNI
with Leu, His, and Pro, respectively (Figure 10b-d). However,
the peak observed at m/z 813 in the case of Glu titration is
assignable to a 2:1 complex formed between GNI and Glu
+
FIGURE 10. ESI mass spectra of (a) GNI (m/z ) 334 [M + H] ,
+
+
6
67 [2M + H] ), 1000 [3M + H] ); (b) GNI‚‚‚Leu (1:1 at 465); (c)
GNI‚‚‚His (1:1 at 489); (d) GNI‚‚‚Pro (1:1 at 449); (e) GNI‚‚‚Glu‚‚
‚GNI (813).
(Figure 10e). Thus, the mass spectral data supports the results
obtained from the fluorescence studies.
NH3+ moiety + Ala), indicating the formation of a 1:1 species
between GNA and the amino acid. The mass spectra obtained
for the titration mixtures of amino acids with GNA were given
in the Supporting Information (S7).
E. Computational Modeling Studies of the Interaction of
GNI and GNA with Amino Acids. Whether a particular amino
acid among the naturally occurring ones interacts with GNI/
GNA has been identified on the basis of fluorescence and
absorption titration studies. The ratio of the complex formed
between the carbohydrate derivative and the amino acid has
GNA‚‚‚aa Complexes. Unlike in the case of GNI, GNA
+
failed to give the [M + H] ion in the ESI experiments and
rather gives an abundant peak at m/z of 157 corresponding to
the naphthyl moiety that is formed by the cleavage of the -C-
N- amine bond. Another ion with low abundance was observed
+
at m/z 174 corresponding to the naphthyl-NH3 moiety that is
formed by the cleavage of the C1(carb)-N bond. Although the
ESI spectrum of Lys titration with GNA showed a peak at m/z
3
04 (157 + 147, i.e., naphthyl fragment + Lys), that of the
Ala (MW 89) exhibited a peak at 263 (174 + 89, i.e., naphthyl-
J. Org. Chem, Vol. 72, No. 9, 2007 3435

Ahuja et al.
TABLE 6. Energies Obtained at Various Theories and Basis Sets of Calculations for GNI‚‚‚aa Complexes
Ec
(au)
Eaa*
(au)
EGNI*
(au)
p
s
q
s
r
s
theory/basis set
PM3
aa
∆E (kcal/mol)
∆E (kcal/mol)
∆E (kcal/mol)
Pro
Leu
His
-0.4897188
-0.5083708
-0.4317982
-0.9091673
-0.1603063
-0.1821584
-0.1001654
-0.3016028
-0.3249927
-0.3087684
-0.3250396
-0.3027776
-2.8
-10.9
-4.1
#
Glu
-0.1
-
0.3046010
HF/3-21g
Pro
Leu
His
Glu
-1548.572810
-1588.548620
-1694.501016
-2849.499224
-396.5422555
-436.5207011
-542.4658042
-545.452826
-1151.990046
-1151.961252
-1151.985266
-1151.991956
-25.4
-41.8
-31.3
-45.0
-
1151.982736
B3LYP/3-21g
B3LYP/6-311g
Pro
Leu
His
Glu
-1557.908234
-1598.237086
-1704.689125
-2866.494713
-398.9423409
-439.2437758
-545.7117657
-548.5554430
-1158.907427
-1158.908178
-1158.908178
-1158.913280
-36.7
-53.4
-43.4
-79.5
-
1158.899298
Pro
Leu
His
Glu
-1566.439449
-1606.991102
-1714.041436
-2882.209291
-401.1302801
-441.6556082
-548.7189270
-551.5876602
-1165.284342
-1165.285594
-1165.281281
-1165.276336
-15.6
-31.3
-25.9
-42.5
-7.8
-23.1
-17.8
-25.3
-8.0
-19.9
-15.1
-23.5
-1165.277522
*
#
Single-point energy calculation of that moiety as taken from the optimized complex. There are two GNI units for each Glu because the complex formed
p
q
is 2:1. Both the optimizations and single-point calculation in B3LYP/6-311G. Optimizations in B3LYP/6-311G and single-point calculations in B3LYP/
-31+G(2d,2p). Both the optimizations and single-point calculations in B3LYP/6-31+G(2d,2p).
r
6
TABLE 7. Energies Obtained at Various Theories and Basis Sets of Calculations for GNA‚‚‚aa Complexes
Ec
(au)
Eaa*
(au)
EGNA*
(au)
p
s
q
s
r
s
theory/basis set
PM3
aa
∆E (kcal/mol)
∆E (kcal/mol)
∆E (kcal/mol)
Ala
Cys
Lys
Ala
Cys
Lys
Ala
Cys
Lys
Ala
Cys
Lys
-0.5097037
-0.4989198
-0.1549692
-0.1419150
-0.3427969
-0.3519648
-7.5
-3.1
-0.5376265
-0.1759187
-0.3533305
-5.3
HF/3-21g
-1473.281240
-1868.811785
-1644.448230
-1482.107285
-1878.346565
-1654.455628
-1490.256626
-1888.427421
-1663.526843
-320.0618100
-715.6202512
-491.2329172
-321.9413370
-718.1616621
-494.2850140
-323.7162677
-721.8963067
-497.0006997
-1153.137045
-1153.157994
-1153.154548
-1160.072495
-1160.098448
-1160.107469
-1166.470768
-1166.486485
-1166.491479
-51.7
-21.0
-38.1
-58.6
-54.2
-39.6
-43.7
-28.0
-21.8
B3LYP/ 3-21g
B3LYP/ 6-311g
-27.3
-21.0
-15.3
-24.4
-8.7
-18.4
*
p
Single-point energy calculation of that moiety as taken from the optimized complex. Both the optimizations and single-point calculation in B3LYP/
q
r
6
-311G. Optimizations in B3LYP/6-311G and single-point calculations in B3LYP/6-31+G(2d,2p). Both the optimizations and single-point calculations in
B3LYP/6-31+G(2d,2p).
been derived on the basis of the titration data and was further
confirmed on the basis of mass spectral studies. All the
computational calculations were performed using Gaussian 98
optimized using hybrid density functional theory (DFT) with
B3LYP/3-21G. The model obtained at this stage was finally
optimized using B3LYP/6-311G. Thus, the calculations were
carried out in a cascade fashion by going through PM3 f HF/
7a
and/or Gaussian 03.7b Prior to assuming the initial guess model
for computational calculations, the interactions present between
the carbohydrates and the amino acids were examined by
analyzing several crystal structures of lectin‚‚‚carbohydrate
complexes, and it was found that more than 50% of the
interactions were indeed through -COOH and -CONH2
moieties of Asp, Glu, Asn, and Gln. Further, it has been noted
that the carbohydrate moiety mainly interacts through O4 and
O6 and to some extent through O3.7c Therefore, the initial
models for the complexes were generated by keeping all these
factors in mind and were used for the geometry optimization
initially at the semiempirical PM3 level of calculations. The
resulting geometry output from these calculations always
exhibited certain weak interactions between the GNI or GNA
and the corresponding amino acid. The model obtained from
the PM3 calculations was further optimized using ab initio HF/
3
-21G f B3LYP/3-21G f B3LYP/6-311G. At every level of
calculation, the stabilization energy of the complex formed (∆Es)
was computed using the formula ∆Es ) Ec - E(GNI/GNA) - Eaa,
where Ec is the total energy of the complex, E(GNI/GNA) is the
energy of the glyco-conjugate, and Eaa is that of the amino acid,
all taken at the same level of calculations. E(GNI/GNA) and Eaa
were obtained by single-point calculations upon taking the
corresponding moiety from the optimized complex. Energy data
obtained at all these theories are shown in Tables 6 and 7 for
GNI and GNA complexes, respectively (Supporting Informa-
tion, S8). The data suggest that the complexation stabilities
p
s
(
∆E ) follow an order of PM3 < HF/3-21G < B3LYP/3-21G
>
B3LYP/6-311G. As the stabilization energies obtained at this
stage were found to be higher than that expected based on the
observed interactions, further computations were carried out
using B3LYP/6-31+G(2d,2p) which takes care of the diffusion
and polarization effects. Use of such functions yielded better
3-21G. The outcome of this calculation provided a better picture
of the interactions present between the GNI or GNA and the
amino acid. The outcome obtained at this stage was further
3436 J. Org. Chem., Vol. 72, No. 9, 2007

Recognition of Amino Acids by Galactosyl DeriVatiVes
TABLE 8. Hydrogen-Bond Data for the GNI‚‚‚aa Complexes at the B3LYP/6-31+G(2d,2p) Level of Calculations
D‚‚‚H
(Å)
H‚‚‚A
(Å)
D‚‚‚A
(Å)
D‚‚‚H‚‚‚A
(deg)
complex
description of H-bonding
GNI‚‚‚Glu (2:1)
CdO(Glu)‚‚‚H-O-C2(carb) GNI-1
N-H(Glu)‚‚‚O-C2(carb) GNI-1
C-O-H(side chain-Glu)‚‚‚O-C4(carb) GNI-2
CdO(side chain-Glu)‚‚‚O-C6(carb) GNI-2
CdO(His)‚‚‚H-O-C3(carb)
0.97
1.01
0.99
0.98
0.97
0.99
0.98
1.01
0.96
0.98
2.18
2.43
1.81
1.80
1.81
1.69
1.80
1.60
2.52
1.91
2.94
3.24
2.80
2.76
2.76
2.67
2.78
2.60
3.18
2.82
134.7
135.3
171.6
165.4
163.5
174.5
175.57
176.0
125.3
153.1
GNI‚‚‚His (1:1)
GNI‚‚‚Leu (1:1)
GNI‚‚‚Pro (1:1)
C-O-H(His)‚‚‚O-C4(carb)
CdO(Leu)‚‚‚H-O-C6(carb)
C-O-H(Leu)‚‚‚O(napthyl)
C-O(Pro)‚‚‚H-O-C6(carb)
C-O-H(Pro)‚‚‚C4-O(carb)
FIGURE 11. Stereoviews of the optimized structures [at B3YLP/6-31+G(2d,2p)] of the complexes formed between GNI and amino acids: (a)
His and (b) Leu. Identification of atom type: C (gray); O (red); N (blue); H (small circle). The same labeling scheme was used in all the figures.
results in the case of carbohydrate systems as reported in the
literature.7 The corresponding computations were carried out
d-g
in two stages: (a) in the first stage, the stabilization energies
q
s
(
∆E ) were computed using the geometry obtained from
B3LYP/6-311G and the single-point energy calculations carried
out for individual components using B3LYP/6-31+G(2d,2p);
(
(
b) in the second stage, the complex stabilization energies
r
∆E ) were computed where both the optimization and the
s
FIGURE 12. Stereoview of the optimized structure [at B3YLP/6-
1+G(2d,2p)] of the complex formed between GNI and Pro.
(
7) (a) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
3
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.,
Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.;
Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.7; Gaussian,
Inc.: Pittsburgh, PA, 1998. (b) Frisch, M. J.; Trucks, G. W.; Schlegel, H.
B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.,
Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.;
Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda,
R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken,
V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.;
Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski,
V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D.
K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui,
Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith,
T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.;
Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople,
J. A. Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
single-point energy calculations were performed using B3LYP/
6
-31+G(2d,2p).
GNI‚‚‚aa Complexes. The calculations were carried out
between GNI and His, Leu, or Pro as 1:1 complexes and with
Glu as a 2:1 complex at all levels of theories mentioned earlier,
and the corresponding formation (stabilization) energies of these
p
s
q
s
r
s
complexes (∆E , ∆E , and ∆E ) are reported in Table 6.
Different complexes of GNI‚‚‚aa exhibited different hydrogen
bond interactions where these interactions differ in their number,
type, and strength, as can be noticed from the data listed in
Table 8 for the structures obtained at B3LYP/6-31+G(2d,2p).
The complexes of GNI‚‚‚His and GNI‚‚‚Leu were mainly
formed through the interaction of the carboxylic moiety of the
amino acid with GNI but altogether in a different fashion, as
can be seen from Figure 11. Thus, the GNI‚‚‚His complex is
stabilized through two hydrogen bond interactions formed
between the carboxylic function of His and C3-OH and C4-
OH moieties of the carbohydrate portion of the GNI to result
in a nine-membered ring as shown in Figure 11a. In these two
hydrogen bonds, although the -COOH group acts as both a
donor and an acceptor, C3-OH acts as only a donor and C4-
OH acts as only an acceptor. The imidazole moiety of His is
placed right on the top of the naphthyl moiety where the distance
between Nimidazole and Cimine, C1, C8, and C9 (of naphthyl) is
(
c) Prakash C. J.; Rao, C. P., unpublished results. (d) Barrows, S. E.; Dulles,
F. J.; Cramer, C. J.; French, A. D.; Truhlar, D. G. Carbohydr. Res. 1995,
76, 219. (e) Csonka, G. I. J. Mol. Struct. (Theochem) 2002, 584, 1. (f)
Lii, J.-H.; Ma, B.; Allinger, N. L. J. Comput. Chem. 1999, 20, 1593. (g)
Allinger, N. L.; Schleyer, P. V. R. J. Comput. Chem. 1993, 14, 1.
2
J. Org. Chem, Vol. 72, No. 9, 2007 3437

Ahuja et al.
FIGURE 13. Stereoview of the optimized structure [at B3YLP/6-31+G(2d,2p)] of the 2:1 complex formed between GNI and Glu.
FIGURE 14. Stereoviews of the optimized structures [at B3YLP/6-31+G(2d,2p)] of the complexes formed between GNA and amino acids: (a)
Ala and (b) Lys.
TABLE 9. Hydrogen-Bond Data for the GNA‚‚‚Amino Acid Complexes at the B3LYP/6-31+G(2d,2p) Level of Calculations
D‚‚‚H
(Å)
H‚‚‚A
(Å)
D‚‚‚A
(Å)
D‚‚‚H‚‚‚A
(deg)
complex
description of H-bonding
GNA‚‚‚Ala (1:1)
CdO(Ala)‚‚‚H-O-C6(carb)
C-O-H(Ala)‚‚‚O-C6-O(carb)
N(Ala)‚‚‚H-O(napthyl)
0.98
0.98
0.99
1.02
0.97
1.01
1.09
1.01
0.99
0.99
1.77
1.85
1.75
2.11
2.40
2.16
2.54
2.48
1.85
1.74
2.74
2.82
2.75
3.07
3.35
3.12
3.59
3.19
2.82
2.70
169.2
170.0
177.5
154.7
165.9
158.1
161.9
126.7
165.7
160.5
N-H(Ala)‚‚‚N(GNA)
GNA‚‚‚Cys (1:1)
GNA‚‚‚Lys (1:1)
S(Cys)‚‚‚H-O-C6(carb)
CdO(Cys)‚‚‚H-N(GNA)
â
C -H‚‚‚Onaph
CdO(Lys)‚‚‚H-N
N(Lys)‚‚‚H-O-C6
C-O-H(Lys)‚‚‚O(napthyl)
3
.9 ( 0.4 Å. On the other hand, in the GNI‚‚‚Leu complex,
from Figure 13. Thus, Glu moves toward the hydrophilic
carbohydrate moiety to a greater extent. It is evident from the
crystal structures of lectin‚‚‚carbohydrate complexes that the
-COOH side chain of Glu and Asp interacts with the
carbohydrate by using one or both of its oxygen centers similar
the -COOH group of Leu is also involved in two H-bonds but
is extended to the C6-OH moiety of the carbohydrate (CdOLeu‚
‚
‚H-O-C6) and the naphthylic-OH (C-O-HLeu‚‚‚Onaph) re-
sulting in a 16-membered H-bonded network as shown in Figure
1b.
Unlike in the cases of GNI‚‚‚His and GNI‚‚‚Leu, in the GNI‚
‚Pro complex, the -COOH group of Pro interacts only through
6
a,c
1
to that found in the present case.
GNA‚‚‚aa Complexes. The H-bond interactions found be-
tween GNA and the interacting amino acids were listed in Table
9. In the complexes of GNA‚‚‚Ala and GNA‚‚‚‚‚‚Lys, both the
‚
its CdO and not through its -OH. Thus, among the 1:1
complexes of GNI‚‚‚aa, the combination with Pro results in a
weak complex interacting through two H-bonds extending
between the OH group of the carboxylic moiety and C4-OH
and C6-OH groups in a cyclic fashion to result in a six-
R
C -NH2 and the -COOH moieties of Ala and Lys interact
with GNA, but differently, as can be seen from Figure 14. The
NH2 of Ala interacts with the naphthylic-OH and the amine
membered H-bond network as shown in Figure 12. Thus, among
r
s
these three 1:1 complexes, the ∆E follows the order GNI‚‚‚
Leu > GNI‚‚‚His > GNI‚‚‚Pro, and this trend was also reflected
in the Ka obtained from fluorescence data.
In the case of the GNI‚‚‚Glu complex, the side chain -COOH
of Glu interacts with one GNI through two H-bonds extended
from the C4-OH and C6-OH groups. The second GNI interacts
rather weakly through the C2-OH moiety with the CdO and
amine function of Glu. All these interactions can be viewed
FIGURE 15. Stereoview of the optimized structure [at B3YLP/6-
31+G(2d,2p)] of the complex of GNA and Cys.
3438 J. Org. Chem., Vol. 72, No. 9, 2007

Recognition of Amino Acids by Galactosyl DeriVatiVes
FIGURE 16. Schematic representation of the interactions present in the 1:1 complexes of GNI with (a) Pro, (b) His, and (c) Leu.
nitrogen of GNA to result in an eight-membered ring possessing
N-HAla‚‚‚NGNA and O-Hnaph‚‚‚NAla. The -COOH of Ala
interacts with C6-OH to result in a six-membered ring possess-
ing C6-O-H‚‚‚OdCAla and C6-O‚‚‚H-O-CAla interactions.
On the other hand, the GNA‚‚‚Lys complex is stabilized through
three hydrogen bonds (Figure 14b) resulting in 10- and
12-membered rings, viz., N-HGNA‚‚‚OdCLys, H2NLys‚‚‚H-O-
C6, and C-OHLys‚‚‚Onaph. Thus, the GNA acts as a H-donor
for two H-bonds and as an acceptor for one.
In the complex of GNA‚‚‚Cys, the amino acid interacts
â
through its -COOH, -SH, and C -H and forms three weak
H-bonds with GNA, viz., CdOCys‚‚‚H-NGNA, SCys‚‚‚H-O-
â
C6GNA, and C -H‚‚‚Onaph, as can be seen from Figure 15.
Carbohydrate Ring Conformation. In the case of GNI and
its amino acid complexes, the carbohydrate moiety was found
to be in 4C1 chair conformation based on Cremer-Pople
FIGURE 17. Schematic representation of the interactions present in
the 2:1 complexes of GNI with Glu.
8
parameters where the θ value is close to 0° but varies from
3
.7 to 9.9°. In the case of GNA and its amino acid complexes,
imine) in the case of GNI and its aa complexes. On the other
hand, the dihedral angle about -C-N- in GNA and its amino
acid complexes substantially deviates from 180°. The deviation
is 83° in the case of simple GNA, and the deviations are 15°,
this angle varies from 2.1 to 25.4° indicating a higher puckering
of the carbohydrate moiety in GNA complexes as compared to
the GNI complexes. All the corresponding data were given in
the Supporting Information (S9).
Relative Orientation of the Carbohydrate and Naphthyl
Moieties. In free GNI, the average plane of the carbohydrate
makes an obtuse angle with respect to the naphthyl plane where
the distance between C6-O and Ophen is about 5.92 Å, and in
its complexes of amino acids, the distance varies in the range
2
8°, and 15°, respectively, in the cases of the GNA‚‚‚Lys, GNA‚
‚
‚Ala, and GNA‚‚‚Cys complexes, suggesting that in the
presence of amino acids -C-N- comes into the plane of the
naphthyl ring.
Conclusions and Correlations
5.5-6.9 Å indicating direct access for the amino acid to interact
with GNI. On the other hand, in free GNA, the average plane
of the carbohydrate makes an acute angle, where the C6-O to
Ophen distance is only 2.85 Å, and in its amino acid complexes,
this distance varies in the range 5.5-7.3 Å to provide access to
the amino acid for the interaction. Thus, the angle goes from
acute to obtuse upon going from free GNA to its amino acid
complexes.
The inclination of the average plane of the carbohydrate
moiety with respect to the naphthyl plane increases in the order
GNI e GNI‚‚‚Pro < GNI‚‚‚His < GNI‚‚‚Leu < GNI‚‚‚Glu,
wherein it approaches a perpendicular fashion in the case of
the GNI‚‚‚Glu complex. This can be gauged from the dihedral
angle CN-N-C1carb-C2carb or CN-N-C1carb-Ocarb (Support-
ing Information, S9).
In the 1:1 complexes of GNI‚‚‚aa, the -COOH group of the
amino acid primarily interacts with the carbohydrate moiety
through either one or both of its oxygens (Figure 16a,b) or spans
between both the carbohydrate and naphthyl units (Figure 16c).
The imine nitrogen is not involved in any interaction in the
complexes.
A similar trend in the interaction of the -COOH group
toward the carbohydrate moiety continues to be present even
R
δ
in the 2:1 complex of GNI and Glu where both the C and C
carboxylic groups of Glu were involved in such interactions as
shown in Figure 17. Thus, the -COOH moiety of amino acids
exhibits high affinity toward carbohydrate in the imine-linked
system, viz., GNI. Similar interactions were noted in the
literature at the active site of R-galactosidase isolated from rice,
where the substrate galactose interacts through the -COOH
Positioning of N with Respect to the Naphthyl Plane.
Although the nitrogen in GNI and its complexes is in the plane
of the naphthyl moiety as expected, it deviates from this plane
in the case of GNA by 1.008 Å as the -HCdN- present in
GNI is reduced to -H2C-NH-. The deviation of N from the
naphthyl plane is higher in the case of the GNA‚‚‚Ala complex
9
group of three different aspartate moieties. Even the interactions
of the substrate, N-acetylgalactose, with the enzyme, N-
acetylgalactosaminidase, isolated from chicken were found to
1
0
be the same. Also noted from the crystal structures of lectin-
carbohydrate complexes is the involvement of Glu and Asp in
(1.46 Å) and lower in the cases of GNA‚‚‚Cys (1.04 Å) and
GNA‚‚‚Lys (0.82 Å) when compared to GNA. The dihedral
angle about the -CdN- bond is very close to 180° (trans-
(
9) Zui, F.; Satoshi, K.; Mitsuru, M.; Hideyuki, K.; Hiroshi, M. J. Biol.
Chem. 2003, 278, 20313.
10) Garman, S. C.; Hannick, L. Z. A.; Garboczi, D. N. Structure 2002,
10, 425.
(
(8) Cremer, D.; Pople, J. A. J. Am. Chem. Soc. 1975, 97, 1354.
J. Org. Chem, Vol. 72, No. 9, 2007 3439

Ahuja et al.
FIGURE 18. Schematic representation of the interactions present in the 1:1 complexes of GNA with (a) Ala, (b) Lys, and (c) Cys.
FIGURE 19. H-bond data correlations of the amino acid complexes of GNI and GNA: (a) D‚‚‚A and H‚‚‚A; (b) H‚‚‚A vs D-H‚‚‚A; (c) D‚‚‚A
vs D-H‚‚‚A.
the interaction with carbohydrate.7c Of course, even the side
chains of Asn and Ser/Thr were also involved in the lectin-
carbohydrate interactions, besides the involvement of the peptide
unit. In some cases, Phe and Tyr extended hydrophobic
interactions using their aromatic portions.6
On the other hand, in the case of GNA‚‚‚aa complexes, the
amino acid moiety interacts with both the units of GNA as
shown in Figure 18. The interacting amino acids do not exhibit
any specificity in their binding toward either unit of GNA as
can be noticed from this figure. Unlike GNI‚‚‚aa complexes,
those of GNA exhibit interactions through the amine nitrogen
in all its complexes. Thus, it is the carboxylic group of the amino
acid that exhibits high affinity toward the carbohydrate moiety
in GNI complexes but not in GNA, owing to the changes in
their molecular geometries. On the other hand, the nitrogen
center extends interaction with the amino acid in the case of
GNA complexes and not in the GNI complexes. The involve-
ment of the nitrogen center in the interactions could be a possible
reason for the loss of the specificity of the carbohydrate moiety
toward the carboxylic group of the amino acid in the case of
GNA as compared to GNI in their complexes.
s
FIGURE 20. Plot of ∆E estimated using the literature (ref 14) vs
r
∆
E obtained from the computation of all GNI/GNA‚‚‚aa complexes.
s
The characteristics developed on the basis of emission and
absorption spectral studies of the interaction of GNI and GNA
with amino acids reported in this paper would be of great
advantage in the detection and evaluation of amino acids in
proteins, particularly those present at C- and N-terminal regions.
Specifically, GNI would be well suited to detect Pro, His, Leu,
and Glu present at the C-terminal region and perhaps Glu present
in the interior of the oligo- and polypeptides over its counterpart,
GNA. We have recently demonstrated the utility of GNI in
recognizing Cu(II) in HEPES buffer, whereas GNA is ineffec-
The hydrogen bond data obtained from the computational
calculations at the B3LYP/6-31+G(2d,2p) level (Tables 8 and
9
) exhibited excellent linear correlation between D‚‚‚A and H‚
‚
‚A distances and fairly linear correlations among H‚‚‚A vs the
D-H‚‚‚A angle and D‚‚‚A vs the D-H‚‚‚A angle as shown in
Figure 19.
The stabilization energies of all these complexes in their
optimized geometry [B3LYP/6-31+G(2d,2p)] were estimated
using the plot given in the literature for H‚‚‚A vs binding
energy.14 These estimated energies were compared with the
computationally obtained ones (Tables 6 and 7; Supporting
Information, S10), and a fairly linear correlation was found, as
shown in Figure 20. The H-bond energies found in the case of
the GNI/GNA‚‚‚aa complexes suggest that these are primarily
moderate H-bonds.
11
tive under the same experimental conditions. This paper, thus,
certainly evokes interest in understanding carbohydrate‚‚‚aa
interactions of the synthetic carbohydrate analogues in the
context of lectin‚‚‚carbohydrate and glycosidase‚‚‚substrate
interactions.
(11) Singhal, N. K.; Ramanujam, B.; Mariappanadar, V.; Rao, C. P. Org.
Lett. 2006, 8, 3525.
3440 J. Org. Chem., Vol. 72, No. 9, 2007

Recognition of Amino Acids by Galactosyl DeriVatiVes
Experimental Section
quartz cells. Amino acid titrations were carried out by maintaining
the final [GNI] at 6.67 µM and [GNA] at 1.67 µM with the addition
of varying volumes of amino acid bulk solution to result in varying
amino acid to ligand mole ratios and by maintaining a solution
volume of 3 mL at each titration where the requisite dilution was
carried out using MeOH. All the amino acid titrations were repeated
at least three to four times.
D-Galactose (G) and 2-hydroxy-1-naphthaldehyde were procured
from a commercial source. 2-Hydroxy-1-aminomethyl naphthalene
was synthesized in the lab to make GNA. G-NH , GNI and GNA
2
were synthesized and characterized as given below. All the solvents
used were of AR grade and dried using standard procedures
immediately before use.
Absorption Studies. To prepare the bulk solutions, GNI and
GNA were initially dissolved in about 1 mL of DMSO followed
by diluting with MeOH to a 10 mL solution to give concentrations
2
Synthesis and Characterization of G-NH (â-D-Galactopy-
ranosyl-C-1-deoxy-C-1-amine). Through a chilled (using an ice-
salt mixture) suspension of dried galactose (10 g) in methanol was
bubbled dry ammonia gas for about 8-9 h. Then the reaction
mixture was further stirred under an ammonical atmosphere for an
additional 24 h at room temperature. After 3-4 days of refrigeration
at 4 °C, a solid product was formed which was then isolated through
filtration, and the residue was washed with methanol. Filtrate was
washed and kept in a refrigerator to obtain a second crop of product
-
2
-3
of 10 and 10 M, respectively, i.e., the bulk solutions whose
concentrations are 10-fold higher than those prepared for the
fluorescence studies. The bulk solutions of the amino acids were
-2
-3
made in deionized water both at 10 and 10 M. All the absorption
studies were carried out on Jasco V-570, and the spectra were
measured in the 200-500 nm region using the solutions filled in 1
cm quartz cells by maintaining the final [GNI] at 0.25 mM and
[GNA] at 0.025 mM in a total solution volume of 2 mL achieved
by diluting with MeOH. Before the dilution, the amino acid
solutions were added to give varying amino acid to ligand mole
ratios similar to that mentioned in the fluorescence studies. All the
amino acid titrations were repeated at least two to three times.
(3.95 g, 39.5%). IR (KBr): 3373(b) υ(O-H) and υ(N-H), 2920(S)
-
1 1
and 2910(S) υ(C-H), 1611(S) δ(N-H) cm . H NMR (DMSO-d
.36 (S, 2H, NH ), 2.86-2.97 (m, H, C2-H), 3.03-3.31 (m, 4H,
C3-H, C4-H, C6-H), 3.88 (d, H, JC1-H-C2-H ) 8.4 Hz, C1-H),
.95-4.00 (m, H, 5H), 4.72-5.63 (m, 4H, C2-OH, C3-OH, C4-
6
):
2
2
3
3
1
3
OH, C6-OH) ppm. C NMR (DMSO-d ): δ 92.65 (C1), 68.84-
6
Computational Studies. The structures of galactosyl derivatives
and amino acids, viz., GNI, GNA, Pro, Leu, Glu, His, Ala, Cys,
and Lys, were generated using Gaussview V.3.09, and their
optimized structures were obtained using Gaussian 98 and/or
Gaussian 03. Structures were first optimized using the semiempirical
PM3 level of calculation to obtain low-energy starting structures,
which were then geometry optimized using the HF/3-21G, B3LYP/
3-21G, B3LYP/6-311G, and B3LYP/6-31+G(2d,2p) level of
calculations in a cascade fashion where the input for the higher
level of calculation is the output of the previous level. The above
methodology was also used to model the complexes resulting from
the interaction of galactosyl derivatives with the amino acids. In
the initial input, the amino acid was placed at a noninteracting
distance (more than 3 Å) from the galactosyl derivative.
+
7
9
7
0.40 (C3-C6), 60.67(C2) ppm. FABMS: m/z 180 ([M + H] ,
+
0%), 179 ([M ], 30%). Anal. calcd for C17
.32; N, 7.83. Found: C, 39.38; H, 6.82; N, 7.62.
Synthesis and Characterization of GNI. To a suspension of
G-NH (3.584 g, 20.02 mmol) in ethanol (45 mL) was added
19 6
H O N: C, 40.00; H,
2
â-hydroxy napthaldehyde (3.612 g, 20.88 mmol), and the reaction
mixture was allowed to reflux for 6 h. During the course of the
reaction, a yellow solid was formed. The reaction mixture was
allowed to cool to room temperature and was left as such overnight.
Some more solid was formed which was then separated by filtration,
washed with a small portion of methanol followed by petroleum
ether, and dried under a vacuum. Product yield: 4.91 g (74%). IR
(
(
3
KBr): 3394(b) υ(O-H) and υ(N-H), 2938(S), 2935(S), and 2920-
S) υ(C-H), 1631(S) δ(CHdN) cm 1. H NMR (DMSO-d
): 3.12-
-
1
6
Mass Spectrometry. ESI spectra were recorded for reaction
mixtures of GNI and GNA with amino acids using a QSTAR XL
mass spectrometer. The concentrations of the ligands were kept
.57 (m, 6H, C2-H, C3-H, C-4H, C6-H), 3.72-3.75 (m, H, C5-
H), 4.36-5.38 (m, 4H, C2-OH, C3-OH, C-4OH, C6-OH), 4.62 (d,
3
H, JC1-H-C2-H ) 11.02 Hz, C1-H), 6.77-7.81 (m, 6H, Ar-H),
-4
constant at 10 M in the final solution, and two samples
1
3
8
.11 (S, H, CHdN), 14.21 (S, H, Phenol-OH) ppm. C NMR
corresponding to titration after the addition of 5 and 10 equiv of
amino acids were prepared and analyzed.
Time-Resolved Fluorescence Measurements. Time-resolved
data frequently contain more information than is available from
the steady-state data. Time-resolved fluorescence experiments were
performed with a time-domain fluorescence spectrometer model
(
DMSO-d
6
): δ 91.10 (C1), 60.68-77.76 (C2-C6), 106.20-137.52
+
(Ar-10C), 158.10 (CHdN) ppm. FABMS: m/z 334 ([M + H] ,
+
7
0%), 333 ([M ], 40%). Anal. calcd for C17
19 6
H O N: C, 61.32; H,
5.72; N, 4.21. Found: C, 61.22; H, 6.00; N, 3.94.
Synthesis and Characterization of GNA. Dried galactose (G,
1.804 g, 10.02 mmol) and 1-methylamino-2-naphthol (1.782 g, 10.3
199 which uses a gated hydrogen discharge lamp as the excitation
mmol) were suspended in 25 mL of ethanol. The reaction mixture
was stirred for 24 h at room temperature followed by reflux for 8
h. The solid obtained was separated by filtration and dried under a
source and an EG & G ORTEC single-photon-counting (SPC) data
acquisition system, interfaced with an LSI-11/23 computer. The
observed fluorescence decay function F(t) was a convolution of
vacuum. Product yield: 2.20 g (66%). IR (KBr): 3394(S) υ(O-H
)
)
the true fluorescence decay function G(t) [G(t) ) Σ
i
B
i
i
exp(-t/τ ),
and υ(N-H), 3208(S), 3027(S), and 2930(S) υ(C-H), 1628(S) δ(CH-N
where B is the pre-exponential factor and τ is the fluorescence
i
i
-
1
1
cm . H NMR (DMSO-d
6
): 3.36-3.70 (m, 6H, C2-H, C3-H,
lifetime for the ith component] and the instrument response function
I(t) and was analyzed by using an appropriate reconvolution
program employing a nonlinear iterative least-squares fit method.
The mono- or multiexponential behavior of the true decay associated
with the observed fluorescence decay function and the correspond-
C-4H, C6-H), 3.74-3.79 (m, H, C5-H), 4.33-4.92 (m, 4H, C2-
OH, C3-OH, C-4OH, C6-OH), 4.39-4.41 (S, 2H, nap-CH
2
), 4.92
(
S, H, C1-H), 6.16 (S, H, N-H), 7.30-8.13 (m, 6H, Ar-H) ppm.
13
C NMR (DMSO-d
C6), 110.97-154.52 (Ar-10C) ppm. FABMS: m/z ([M - 3] ,
0%), 332. Anal. calcd for C17 N: C, 60.89; H, 6.32; N, 4.21.
Found: C, 60.52; H, 6.56; N, 3.96.
6
): δ 32.81 (nap-CH2), 68.85-79.76 (C2-
+
2
ing computer fit was evaluated by the minimum reduced ø value
3
19 6
H O
as well as by the distribution of the weighted residuals among the
12
data channels and the Durbin-Watson parameter. The time
13
Fluorescence Studies. To prepare the bulk solutions, GNI and
GNA were initially dissolved in about 1 mL of DMSO followed
by diluting with MeOH to a 10 mL solution to give concentrations
resolution of the SPC unit, determined by Zimmermann’s method,
was found to be ca. 100 ps.
-
3
-4
of 10 and 10 M, respectively. The bulk solutions of the amino
-3
(12) O’Connor, D. V.; Philips, D. Time-Correlated Single Photon
Counting; Academic Press: New York, 1984.
acids were made in deionized water to give a 10 M concentration.
All the fluorescence studies were carried out on a Perkin-Elmer
LS55 at 320 nm excitation, and the emissions were measured in
the range 330-500/450 nm using 3 mL solutions filled in 1 cm
(13) Zimmermann, T.; Weigerber, P. Proc. International Workshop on
Mining Software Repositories (MSR 2004), 2004; p 2.
(14) Grabowski, S. J. Annu. Rep. Prog. Chem.: Sect. C 2006, 102, 131.
J. Org. Chem, Vol. 72, No. 9, 2007 3441

Ahuja et al.
Acknowledgment. C.P.R. acknowledges the financial sup-
port by DST, CSIR, and BRNS-DAE. We thank Dr. Haridas
Pal, Radiation & Photochemistry Division, BARC, for allowing
us to use the time-resolved fluorescence instrument. We thank
Dr. M. Vairamani, IICT, Hyderabad, for making the mass
spectrometer available for this work. We thank both the referees
for their valuable suggestions.
Supporting Information Available: The fluorescence data
(S1-S3), absorption data (S4-S6), ESI/MS data (S7), and com-
putational data (S8-S10). This material is available free of charge
via the Internet at http://pubs.acs.org.
JO0700979
3442 J. Org. Chem., Vol. 72, No. 9, 2007