Biomacromolecules 2006, 7, 1253-1260
1253
Synthesis and NMR Characterization of New Hyaluronan-Based
NO Donors
Chiara Di Meo,† Donatella Capitani,‡ Luisa Mannina,‡,§ Enzo Brancaleoni,‡ Devis Galesso,⊥
Gilda De Luca,⊥ and Vittorio Crescenzi*,†
Department of Chemistry, University of Rome “La Sapienza”, p.le Aldo Moro 5, 00185 Rome, Italy, Institute
of Chemical Methodologies, CNR, Research Area of Rome, Via Salaria Km 29,300, 00016 Monterotondo
Stazione, Rome, Italy, S.T.A.A.M. Department, University of Molise, 86100 Campobasso, Italy, and Fidia
Farmaceutici SpA, Via Ponte della Fabbrica 3/a, I-35031 Abano Terme, Padua, Italy
Received November 28, 2005; Revised Manuscript Received January 15, 2006
Nitric oxide (NO) and hyaluronic acid (HA), two species widely different in terms of molecular complexity and
biological competence, are both known to play an important role in the wound healing process. To combine the
properties of HA and NO, we synthesized new NO-donors based on hyaluronic acid derivatives exhibiting a
controlled NO-release under physiological conditions (in vitro tests). Since two molecules of NO can form a
covalent bond with secondary amines to yield structures, named NONO-ates, able to release NO in solution, we
used spermidine bound to HA as the NO-linker. The HA-spermidine derivative was obtained by controlled HA
amidation in aqueous media, activating the biopolymer carboxylate groups with a water soluble carbodiimide.
The resulting derivative, soluble in water, was fully characterized by high field 1H and 13C NMR spectroscopy.
The amount of grafting of spermidine on HA was determined by integration of suitable 1H NMR signals. In
addition, cross-linked derivatives of HA were synthesized by the Ugi’s four-component reaction using formaldehyde,
cyclohexylisocyanide, and spermidine. The HA-spermidine networks were characterized by 13C CP-MAS NMR
spectroscopy. The degree of cross-linking of the networks was also determined. Finally, the release of NO from
the swollen hydrogels freshly saturated with NO, in contact with aqueous media, was monitored by means of UV
spectrophotometric measurements.
Introduction
NO is synthesized in vivo from L-arginine by the enzymes
NOS (nitric oxide synthases) that can be either in the constitutive
(cNOS) or in the inducible (iNOS) form.1 NO is a gas showing
complex biological activities which, just to mention a few,
include a marked influence on the cardiovascular and nervous
systems, on the immunological response, on vasodilation, and
on apoptosis.1 As a consequence, NO-releasing prodrugs, such
as organic nitrates and nitroprusside, are often used in the
treatment of cardio-circulatory failures. Indeed, a large class of
NO-releasing prodrugs is currently employed.2,3
NO also plays a crucial role in wound healing processes,4
provided that the NO production can be increased by the
activation of cNOS and by the synthesis of iNOS. In fact, NO
modulates cytokine activity in the inflammatory process, in
angiogenesis, and in cell proliferation.5 Hyaluronic acid (HA)
is also known to play an important role in the wound healing
process. In the context, we wish to report here the synthesis
and the characterization of new NO-donors based on HA, using
linear and cross-linked derivatives aimed at combining the
properties of HA and NO. In such derivatives, spermidine
molecules are linked via one primary amino group only to HA
(soluble amide derivatives) or via both primary amino groups
to different HA chains (cross-linked amide derivatives). The
* To whom correspondence should
vittorio.crescenzi@uniroma1.it.
† University of Rome “La Sapienza”.
‡ Institute of Chemical Methodologies.
§ University of Molise.
⊥ Fidia Farmaceutici SpA.
be
addressed.
E-mail:
products can readily absorb NO molecules which form specific
covalent bonds with the secondary amine groups of sperimidine
to yield NONO-ate structures.6 Subsequently, the latter are able
to release NO in solution, at certain values of pH and
temperature, with a characteristic kinetics dependent mainly on
the degree of substitution/cross-link density of the HA derivatives.
Materials and Methods
Materials. HA sodium salt, from Fidia Advanced Biopolimers (FAB
srl, Abano Terme, Padua, Italy) Mη ) 200 kDa, was used throughout.
Spermidine trihydrochloride salt was supplied by Fluka, Milan, Italy.
All other chemicals were reagent grade and were used without further
purification.
Synthesis of HA-Spermidine Hydrosoluble Derivative. HA (≈
200 mg) was dissolved in 8 mL of MES [2-(N-morpholino)-ethanesulfonic acid buffer (50 mM, pH ) 4); 144 mg of EDC‚HCl [N-(3dimethylaminopropyl)-N′-ethylcarbodimide hydrochloride)], and 88 mg
of NHS (N-hydroxysuccinimide) and 1 g of spermidine trihydrochloride
salt were added. The reaction was performed at room temperature under
stirring for 24 h; the solution was then dialyzed (cutoff ) 12 kDa)
against a NaCl saturated solution for 1 day and then against distilled
water for 5 days. Finally, the solution was freeze-dried.
Synthesis of HA-Spermidine Hydrogels by Ugi’s Reaction.
Spermidine (≈ 240 mg) trihydrochloride salt was dissolved in 1.2 mL
of distilled water to obtain a 784 mM solution. Samples were obtained
dissolving 120 mg of HA in 2 mL of distilled water (polymer
concentration: 6% w/V). The solutions were acidified with a few drops
of HCl 0.5 M to obtain a pH of 4. Different amounts of spermidine
solution were added to the samples to obtain networks with a theoretical
cross-linking degree (TCD) ranging from 5% up to 25%. TCD is defined
as the stoichiometric ratio between the moles of diamine and the moles
10.1021/bm050904i CCC: $33.50
© 2006 American Chemical Society
Published on Web 03/10/2006
1254
Biomacromolecules, Vol. 7, No. 4, 2006
of the carboxylic group of HA. Then, 50 µL of formaldehyde and 100
µL of cyclohexylisocyanide were added to the solutions which were
stirred for one minute, and then left at rest overnight. The obtained
gels were dialyzed against distilled water for 7 days. Five networks
with a TCD equal to 5%, 8%, 12%, 15%, and 25% respectively were
obtained.
Swelling Measurements. The swelling capacity of the hydrogels,
Sw, is defined as the ratio between the weight of swollen gels (Ws)
after extensive dialysis against distilled water and the weight of the
dry networks (Wd): Sw ) Ws/Wd. Freeze-dried networks at different
cross-linking degree were swollen in distilled water at 25 °C until
constant weight.
NMR in Solution. Samples, ≈ 5 mg, were dissolved in 700 µL of
a phosphate buffered (pD ) 7) D2O solution 0.1 M NaCl.
1H and 13C NMR experiments were performed at 27 °C on a Bruker
AVANCE AQS 600 spectrometer operating at 600.13 and 150.95 MHz,
respectively, and equipped with a Bruker multinuclear, z gradient
probehead. In all of the 1H spectra, a soft presaturation of the HOD
residual signal was applied.7 1H and 13C assignments were obtained
using 1H-1H COSY, 1H-1H TOCSY, and 1H-13C HSQC experiments
with a z gradient coherence selection. All 2D experiments were carried
out using 1024 data points in the f2 dimension and 512 data points in
the f1 dimension; the recycle delay was 1 s. The TOCSY experiment
was performed with a spin-lock duration of 80 ms. The HSQC
experiment was performed using a coupling constant of 150 Hz. The
TOCSY and the HSQC experiments were processed in the phase
sensitive mode (TPPI) with 512 × 512 data points.
1
H and 13C chemical shifts are reported in ppm with respect to 2,2dimethyl-2-silapentane-5-sulfonate sodium salt (DSS) used as an internal
standard.
Pulsed gradient spin-echo (PGSE) measurements were performed
both on a free spermidine and on the HA-spermidine sample; in both
cases 1 mg of sample was dissolved in 700 µL of a 0.1 M NaCl
phosphate buffered (pD)7) D2O solution.
PGSE experiments were performed with a pulsed field gradient unit
producing a magnetic field gradient in the z direction with a strength
of 55.4 G cm-1. The stimulated echo pulse sequence using bipolar
gradients with a longitudinal eddy current delay was used. The strength
of the sine-shaped gradient pulse with a duration of 1.8 ms was
logarithmically incremented in 32 steps, from 2% up to 95% of the
maximum gradient strength, with a diffusion time of 700 ms and a
longitudinal eddy current delay of 25 ms. After Fourier transformation
and a baseline correction, the diffusion dimension was processed using
the DOSY subroutine of the Bruker Xwinnmr software package.
Solid State NMR. Samples were freeze-dried, finely cut, packed
into 4 mm zirconia rotors, and sealed with Kel-F caps. Solid state 13C
CP-MAS spectra were performed at 50.13 MHz on a Bruker ASX-200
spectrometer. The spin-rate was 8 kHz. The π/2 pulse width was 3.5
µs, and the relaxation delay was 3 s; 8000 scans were acquired for
each spectrum. The cross-polarization was performed applying the
variable spin-lock sequence RAMP-CP-MAS;8,9 the RAMP was
applied on the 1H channel, and during the contact time τ, the amplitude
of the RAMP increased from 50 to 100% of its maximum value. This
method allows the motional modulation of the carbon and proton
coupling caused by spinning the sample at high rate to be overcome.
Experiments performed in the cross-polarization mode with a
simultaneous phase inversion (CP-SPI) allowed the different type of
carbons to be selectively observed;10,11 the contact time τ for the crosspolarization was 1 ms, whereas the length of the pulse used for the
phase inversion was 24 µs.
Spectra were acquired using 1024 data points in the time domain,
zero filled and Fourier transformed. The chemical shift was externally
referred to tetramethylsilane (TMS).
Cross-Polarization (CP) Dynamic. The 13C CP-MAS spectra are
not quantitative because the intensity of the carbon resonances depends
on the CP rates, which can be different for different carbon atoms.12
The CP rate depends on the number of abundant spins I (protons) near
Di Meo et al.
Scheme 1. Structures of R-Linked and η-Linked HA-Spermidine
Water Soluble Derivatives
the dilute spin S (carbons) and on their distance from S. The question
rises when we wish to obtain a quantitative analysis of the intensity of
resonances due to carbon of different types. In simple cases, the problem
can be solved by investigating the CP dynamic;12 the kinetic of the CP
dynamic can be analyzed using the two-phase polarization transfer
model13,14 which can be described by the equation
( )[
(
(
)
S
τ
τ
) exp - 1H 1 - s exp S0
TCHR
T1F
(1 - s) exp -
)
(
)]
3τ
τ2
exp 2TCHR
2TCHD2
(1)
where S0 is the area of the resonance at the time τ ) 0, T1H
1F is the
proton spin-lattice relaxation time in the rotating frame, TCHD is the
time constant for the first phase, i.e., CP from covalently bonded
protons, TCHR is the time constant for the second phase, i.e., CP from
the more remote protons, finally, (1 - s)/s is the ratio between the
magnitude of the fast and slow polarization transfer phases (this ratio
is also fitted to the experimental data).15 In homogeneous systems, the
spin-diffusion process averages the T1H
1F values of all carbon resonances; under these circumstances the S0 values obtained fitting the
experimental data to the eq 1 are the “true” areas of the resonances.
Therefore, the CP dynamics were carefully investigated. For all
samples, a series of 13C-CP MAS spectra were carried out, with the
contact time τ ranging from 0.05 up to 12 ms. The intensity of two
selected resonances was reported as a function of the contact time; the
T1H
1F and S0 values were evaluated.
NONO-ate Bond Formation. Freeze-dried samples were placed into
a homemade steel reactor. To remove air and to dry both the reactor
and the samples, N2 was fluxed for 20 min and then pressurized at 5
atm for 4 times. Then, NO was pressurized at 5 atm for 3 days. After
purging, N2 was fluxed again to remove NO traces in the reactor.
Finally, samples were collected and stored at -20 °C before the NO
release measurements.
NO Release. NO release from the samples was measured by
colorimetric Griess assay with an HP 8452A diode array UV-vis
spectrophotometer. The Griess solution was prepared by dissolving 500
mg of SULF (sulfanilamide) and 10 mg of NEDD (N-(1-naphthyl)
ethylenediamine dihydrochloride) in 100 mL of PBS 100 mM, pH )
7.4; 4 mg of the sample was placed in 2.5 mL of Griess solution in a
1 cm quartz cell. The NO release was detected recording the absorbance
increase at λ ) 496 nm as a function of time. The concentration of
NO released was calculated using the molar absorptivity according to
a method reported in the literature (12 500 M-1 cm-1).16
Results and Discussion
Synthesis of Water Soluble HA-Spermidine Derivative.
The carboxylic group activation with a water soluble carbodi-
New Hyaluronan-Based NO Donors
Biomacromolecules, Vol. 7, No. 4, 2006 1255
Figure 1. log Sw vs log TCD in water at 25 °C for the HA-spermidine
Ugi gels with a TCD equal to 5%, 8%, 12%, 15%, and 25%.
Figure 3. Expansion of 1H-1H TOCSY map of HA-spermidine in
phosphate buffered (pD ) 7) D2O solution 0.1 M NaCl at 27 °C. Four
spin-systems, A, B, C, and D are evidenced.
Figure 2. 1H NMR spectrum (600.13 MHz) of HA-spermidine in
phosphate buffered (pD)7) D2O solution 0.1 M NaCl at 27 °C, along
with the assignment of the resonances of spermidine bound to HA.
In the insert, the resonances used for the quantitative evaluation of
the spermidine moieties bound to HA are evidenced.
imide17 and the subsequent reaction with a primary amine (or
diamine) are widely used to obtain HA derivatives.18 We
employed the EDC/NHS system and a large excess of spermidine in order to reduce the probability of cross-links formation.19-22 The HA-spermidine hydrosoluble product was purified by dialysis and recovered by freeze-drying. Note that a
spermidine molecule can be bound to HA in the two possible
ways as shown in Scheme 1.
Synthesis of Networks. We used Ugi’s reaction23 to obtain
hydrogels from different polysaccharides.24,25 Because of the
rapidity of the Ugi condensation, we used spermidine in its more
soluble trihydrochloride form in a 6% w/V HA solution. Gels
with different degrees of cross-linking were easily obtained
adding different amounts of spermidine solution in H2O. The
cross-linking process lasted a few seconds, that is a time much
shorter than the time necessary for the cross-linking of the HAlysine system.26 The resulting gels were dialyzed against distilled
water for several days and then freeze-dried.
Swelling Measurements. In Figure 1, the swelling data for
the HA-spermidine gels with a TCD equal to 5%, 8%, 12%,
15%, and 25% are reported. The log(Sw) vs log TCD plot shows
a linear trend. As expected, when the TCD value is increased,
the swelling diminishes. Fitting the experimental data to the
equation log Sw ) A + B log TCD, the parameters A and B are
found to be 3.08 and 1.09, respectively. The estimated A and B
values for HA-spermidine networks are very similar to those
previously found for HA-lysine networks.26
NMR in Solution.
NMR Characterization of HA-Spermidine along with the
Schemes of a HA-Spermidine Derivatives, namely r and
Figure 4. 1H-13C HSQC map of HA-spermidine in phosphate
buffered (pD ) 7) D2O solution 0.1 M NaCl at 27 °C. The full
assignment is also reported. The 1H and the 13C NMR spectra are
reported as projections in the f2 and f1 dimensions, respectively.
η. Because spermidine is not a symmetric molecule, it is
expected to be bound to HA in two possible ways as shown in
Scheme 1.
The 1H NMR spectrum of HA-spermidine is shown in Figure
2. Besides the resonances of the HA moiety, other resonances
are observed. The 2D-TOCSY and the comparison with the 1H
spectrum of a standard of spermidine allows the presence of R
and η HA-spermidine derivatives to be revealed, see numbers
1 and 2, respectively, in Scheme 1. Four different spin systems,
reported in Figure 3 as A-D, have been identified. The spin
system A shows the correlation among the methylene protons
R1, R2, β, and γ. In fact, after the binding with HA, the
methylene protons R both resonating at 3.099 ppm (R′) in free
spermidine, become fully nonequivalent and downfield shifted,
their chemical shift being at 3.262 and 3.475 ppm, respectively.
After the binding, the methylene protons β and γ are both
upfield shifted, their chemical shifts being at 1.937 and 3.043
1256
Biomacromolecules, Vol. 7, No. 4, 2006
Di Meo et al.
Table 1. 1H and 13C Assignment of HA-Spermidine and Free
Spermidine Trihydrochloride Salt in Phosphate Buffered (pD ) 7)
D2O Solution 0.1 M NaCl at 27 °C
1H
HA-spermidine
1a
2a
3a
4a
5a
6a
6a′
7a
8a
1b
2b
3b
4b
5b
6b
R1
r2
β
γ
δ
E
ζ
η
r
β
γ
δ
E
ζ
η1
η2
free spermidine
r′
β′
γ′
δ′
E′
ζ′
η′
(ppm)
4.543
3.816
3.700
3.490
3.462
3.905
3.739
13C
(ppm)
3.262
3.475
1.937
3.043
3.080
1.775
1.756
3.041
3.100
2.103
3.130
3.089
1.725
1.625
3.166
3.410
102.14
55.89
84.12
70.05
77.12
62.06
62.06
176.04
24.05
104.73
74.10
75.17
81.55
77.94
175.20
38.04
38.04
27.01
46.65
48.49
24.33
25.50
40.38
38.22
25.36
46.02
48.81
24.45
27.05
40.39
40.39
3.099
2.111
3.144
3.097
1.785
1.767
3.035
39.29
26.42
47.14
49.69
25.48
26.64
41.43
2.006
4.446
3.331
3.560
3.716
3.691
ppm, respectively, being at 2.111 (β′) and 3.144 ppm (γ′) in
free spermidine. The spin system B shows the correlation among
protons δ, ǫ, ζ, and η: their chemical shifts are rather unaffected
after the binding, these protons being far away from the binding
site. Therefore, the spin systems A and B fully match with
number 1 in Scheme 1. Two other spin systems are observable
in the TOCSY slice of Figure 3. The spin system C shows the
correlation among the methylene protons η1, η2, ζ, ǫ, and δ.
Note that, again, after binding with HA the methylene protons
η both resonating at 3.035 ppm (η′) in free spermidine become
fully nonequivalent and downfield shifted, being at 3.410 and
3.166 ppm, respectively. The methylene protons ζ are also
upfield shifted at 1.625 ppm, being at 1.767 (ζ′) in free
spermidine. Methylene protons ǫ are slightly upfield shifted from
1.785 (ǫ′) to 1.725 ppm, whereas protons δ are rather unaffected.
Finally, the spin system D shows the correlation among the
methylene protons R, β, and γ, and all of the chemical shifts of
these protons are almost unaffected, these protons being far away
from the binding site. The spin systems C and D fully match
with number 2 in Scheme 1.
Figure 5. 1H-detected DOSY map of HA-spermidine derivative in
phosphate buffered (pD ) 7) D2O solution 0.1 M NaCl at 27 °C. The
1H spectrum is shown as horizontal projection.
To confirm the assignment, an HSQC experiment was also
performed (Figure 4). All of the cross-peaks of spermidine are
well observable: the unequivalent methylenes R1 and R2 as well
as the unequivalent methylenes η1 and η2 are evidenced. The
assignment of all of the other resonances of spermidine bound
to HA is also reported in the map.
The assignment of carbon resonances of free spermidine as
well as the assignment of carbon resonances of spermidine after
the binding with HA are reported in Table 1. For the sake of
clarity, the assignment of the resonances of the HA moiety is
also reported in the same table; the assignment has been made
acting upon literature information27 and our own data collected
using the same solvent as in the actual experiments.
DOSY Measurements. As is well-known from the literature,28 molecular self-diffusion can be encoded into NMR
datasets by means of the pulsed-gradient of the magnetic field
(PFG-NMR). Diffusion-ordered NMR spectroscopy (DOSY)29
is a particularly convenient way of displaying the molecular
self-diffusion information in a bi-dimensional array, with the
NMR spectrum in one dimension and the self-diffusion coefficient in the other one. DOSY has been successfully used for
the analysis of mixtures,30 for the study of intermolecular
interactions,31,32 for the characterization of aggregates,33 for the
molecular weight determination of uncharged polysaccharides,34
and for the optimization of the dialysis process of hyaluronic
acid derivatives.35
As previously shown, TOCSY and HSQC experiments allow
a full assignment of the HA-spermidine. The linkage between
spermidine and the polymer can be further established by means
of a DOSY experiment. Because of the marked difference in
the diffusion coefficients, the DOSY map easily establish the
presence of spermidine free to move and of spermidine bound
to the HA moiety; in fact, after the binding, the diffusion
coefficient of the spermidine bound to HA is the same as the
one of HA, which is, obviously, much slower than the diffusion
coefficient of spermidine free in solution. In fact the diffusion
coefficient measured for the HA-spermidine sample was found
to be D ) 5.51 × 10-12 m2/s, whereas the diffusion coefficient
measured for free spermidine was found to be D ) 1.87 × 10-9
m2/s. Therefore, with the DOSY map, it is also possible to
establish the presence, if any, of low molecular weight
compounds such as free spermidine, which may impair the
Biomacromolecules, Vol. 7, No. 4, 2006 1257
New Hyaluronan-Based NO Donors
determination of the true percentage of the grafting of spermidine on HA.
Here, we report the DOSY map of the HA-spermidine
derivative; the 1H spectrum is shown in the horizontal projection
(Figure 5). The map unequivocally demonstrates that all of the
resonances of HA as well as of spermidine do show the same
diffusion coefficient. The map also indicates that no low
molecular weight impurities are present in the sample. The
absence of impurities allows the true degree of grafting to be
calculated. The degree of grafting, in agreement with Scheme
1, is calculated according to the relationship
I(ζ)
grafting (η) )
2
) 0.11
I(1A) + I(1B)
2
I(ζ) being the area of the signal of the methylene protons ζ
resonating at 1.767 ppm, and I(1A) and I(1B) the area of the
anomeric protons of HA, see the insert in Figure 2.
The degree of grafting, in agreement with Scheme 1, is
calculated according to the relationship
I(ǫ) + I(ǫ) + I(ζ) - I(ζ)
grafting(R) )
4
I(1A) + I(1B)
2
) 0.20
where I(ǫ) + I(ǫ) + I(ζ) is obtained by integrating the signals
of the 1H spectrum resonating between 1.7 and 1.85 ppm (see
insert in Figure 2).
The total percentage of grafting is therefore
(grafting(η) + grafting(R)) × 100 ) 11% + 20% ) 31%
13C
Solid State NMR. The 13C CP-MAS spectrum of an HA
sample is shown in Figure 6a; for the sake of clarity, the
assignment of the resonances is also reported26 in agreement
with the labeling shown in Scheme 2.
In the same figure, the spectra of the spermidine-based
networks with a different TCD are also shown. Besides the
resonances of HA, other resonances are observed in the spectra
of the networks; these resonances are due to carbon atoms
belonging to the chemical bridges between the HA polymeric
chains. Note, however, that the possible presence of spermidine
bound as a pendant group cannot be totally disregarded.
The resonance at 168 ppm is due to the amidic carbon 7; the
resonance centered at 47 ppm is due to the methine carbon 1 of
the cyclohexyl rings and to the methylene carbons γ and δ of
spermidine; the resonance at 33 ppm is ascribed to the methylene
Scheme 2. Structure of HA-Spermidine Ugi Networks
Figure 6. 13C CP-MAS NMR spectra of HA along with the resonance
assignment (a); HA-spermidine Ugi networks with TCD ) 5% (b),
TCD ) 8% (c), TCD ) 12% (d), TCD ) 15% (e), TCD ) 25% (f); the
assignment of the resonances of the networks is also reported.
carbons 2 and 6 of the cyclohexyl rings and to the methylene
carbons R and η of spermidine; finally, methylene carbons 3,
4, and 5 of the cyclohexyl ring and methylene carbons ǫ, β,
and ζ of spermidine resonate at about 24 ppm almost fully
overlapped to the intense resonance of the acetyl carbon of HA.
To further confirm the assignment of the resonances, the CPSPI pulse sequence was applied. Using this sequence, a full
spectral editing of the CP-MAS spectrum was performed. As a
result, a spectrum was obtained where methine carbon resonances are zeroed and methylene carbon resonances are inverted,
whereas resonances due to quaternary and methyl carbon atoms
are intense and positive, see Figure 7. In this figure, the 13C
CP-MAS spectrum (top) of the spermidine-based network with
TCD ) 25% is compared with the CP-SPI spectrum (bottom)
of the same sample. In agreement with the assignment given
above, all of the methylene carbons resonances were assigned.
1258
Biomacromolecules, Vol. 7, No. 4, 2006
Di Meo et al.
Figure 7. Top: 13C CP-MAS NMR spectrum of HA-spermidine gel
with TCD ) 25%. Bottom: 13C CP-SPI spectrum of the same
sample.
By comparing the spectra of the HA networks reported in
Figure 6, a-f, an increment of the intensity of the resonances
centered at 24, 33, 47, and 168 ppm is observable; the intensity
of these resonances definitely increases as the TCD increases.
This trend qualitatively matches with the reported TCD.
The area of the resonance centered at 33 ppm, due to the
methylene carbons 2 and 6 of the cyclohexyl rings and to the
methylene carbons R and η of spermidine, and the area of the
resonance centered at 105 ppm, due to the anomeric carbons of
HA, may be used for evaluating the real cross-linking degree
(RCD). However, since 13C CP-MAS spectra are not quantitative, caution must be used in the quantitative evaluation of the
intensity and/or area of the resonances.36 Nevertheless, in
homogeneous systems, the “true” area of the resonances, S0,
can be obtained fitting the experimental data to eq 1. We recall
that in an homogeneous system, the spin diffusion process
1H
averages all of the T1F
values. Therefore, the CP dynamic was
carefully investigated for all the samples.
13C CP-MAS spectra were run at different contact times, τ,
with τ ranging from 0.05 up to 12 ms. The areas of the
resonances centered at 105, S0(105), and 33 ppm, S0(33),
respectively, were reported as a function of the contact time.
As an example, the correlation between the area of these
resonances and the contact time is shown in Figure 8 for two
spermidine-based networks with TDC ) 12 (a) and 25 (b),
1H
and S0
respectively. Fitting the experimental data to eq 1, T1F
values were obtained.
The values found for the network with TDC ) 25 are
S0(105) ) 1.5 ( 0.2 S0(33) ) 0.91 ( 0.02
1H
1H
(33) ) (5.4 ( 0.3) ms
(105) ) (5.4 ( 0.8) ms T1F
T1F
whereas the values found for the network with TDC)12 are
S0(105) ) 1.7 ( 0.1 S0(33) ) 0.60 ( 0.01
1H
1H
(33) ) (5.4 ( 0.2) ms
(105) ) (5.0 ( 0.3) ms T1F
T1F
Because in both samples the selected resonances show the same
1H
values within experimental errors, the samples can be
T1F
considered as homogeneous systems. As a consequence, the
Figure 8. Area of the resonances centered at 105 ppm S(τ) (b) and
33 ppm S(τ) (O) reported as a function of the contact time for two
spermidine-based networks with TDC ) 12% (a) and 25% (b).
Table 2. TCD Values of the Spermidine-based Networks
Compared with the RCD Values as Obtained from Solid State
NMR
TCD (%)
RCD (%)
25
15
12
8
5
20.0 ( 0.5
13.0 ( 0.4
11.8 ( 0.4
8.0 ( 0.5
4.9 ( 0.4
obtained S0 values can be used to quantitatively evaluate the
RCD (real degree of cross-link)
S0(33)
n
RCD )
S0(105)
m
where n ) 6 is the number of carbon atoms contributing to the
area of the resonance at 33 ppm and m ) 2 is the number of
the anomeric carbon atoms contributing to the area of the
resonance at 105 ppm.
In Table 2, the TCD values of the spermidine-based networks
are compared with the RCD values obtained from NMR data;
the error on the RCD percentage was calculated applying the
error propagation theory.
NO Release. NONO-ate bond formation on secondary amine
after incubation under NO pressure, see Scheme 3, is largely
described in the literature.37,38
New Hyaluronan-Based NO Donors
Biomacromolecules, Vol. 7, No. 4, 2006 1259
Scheme 3. Schematic Representation of NONO-ate Bond
Formation on Secondary Amine after Incubation under NO
Pressure and Subsequent NO Release in Physiological Conditions
Figure 10. Kinetics of NO-release from HA-spermidine gels with
different TCD in a Griess phosphate buffered solution 0.1 M (pH )
7.4) at T ) 37 °C: TCD ) 25% (]), TCD ) 15% (b), TCD ) 12%
(*), TCD ) 8% (2), TCD ) 5% (O).
To obtain covalent linking of NO on HA-spermidine water
soluble and cross-linked derivatives, samples were treated with
NO at 5 atm in a steel reactor for 72 h. Due to the high reactivity
of NO with water to give very corrosive nitrous acid, it is
important to purge the reactor with dry N2 or Ar to eliminate
moisture from the reactor and the samples.
Furthermore, due to the sensitivity of the NONO-ate bond
to temperature, samples have to be stored at -20 °C before
NO-release monitoring.
A variety of analytical procedures to monitor nitric oxides
in air or in solution have been widely used, as chemometric,39
electrochemical,40 spectrometric (EPR),41 or spectrophotometric
methods.42,43 A simple and useful procedure is the Griess
colorimetric assay in aqueous solution16 that uses the nitrosation
of sulfanilamide in the presence of N-(1-naphthyl) ethylenediamine dihydrochloride to give an orange-colored azo-compound.
The quantity of NO released from the samples and reacting with
Griess reagents is easily monitored with an UV-vis spectrophotometer at λ ) 496 nm.
The experiments on HA-spermidine water soluble or crosslinked derivatives were performed at the same concentration,
1.6 mg/mL, in the Griess solution at physiological values of
pH and temperature. The increase of the absorbance at λ ) 496
nm vs the time for a water soluble sample and for all the crosslinked gels is reported in Figures 9 and 10, respectively.
All samples show NO release. Absorbance values are
transformed into NO concentration (µM) using the molar
absorptivity value ǫ ) 12 500 M-1 cm-1 reported in the
literature for the same experimental conditions.16 The assay is
both sensitive and linear over the range of values observed.
The water soluble derivative shows a complete gas release
Figure 9. Kinetic of NO-release from HA-spermidine in a Griess
phosphate buffered solution 0.1 M (pH ) 7.4) at T ) 37 °C.
in about 4 min, with an initial linear trend pointing to a zero
order kinetic (Figure 9).
NO release from the Ugi networks, re-swollen as gels in
aqueous media, shows more complex kinetics (Figure 10). A
rather fast increase of the absorbance is observed in the first
few seconds, and then for several minutes, a slow increase of
the absorbance is observed. Finally, a pseudolinear trend prevails
until the complete gas loss is reached. This behavior can be
explained considering the nonsoluble gel structure: the NO
molecules closer to the surface are, in fact, quickly released in
solution, whereas the NO molecules inside the bulk of the gel
are slowly released. Therefore, the whole process depends on
the swelling of the gels, as well as on the diffusion of the Griess
reagents inside and outside the gel bulk. A qualitative explanation of the increase in the amount of NO released with increasing
degree of gels cross-linking may be made considering that the
higher the TCD, the higher the number of NO-binding sites in
the gels.
Conclusions
New NO donors based on hyaluronic acid derivatives with
modulable NO-release properties in physiological conditions
have been synthesized and structurally characterized. To form
NONO-ate structures able to release NO in solution, we used
spermidine as the amine linking agent. The water soluble
derivative of HA and spermidine was obtained by amidation
reaction in aqueous media using EDC/NHS activation of the
HA carboxylate. The HA-spermidine derivative was fully
characterized by 1H and 13C NMR spectroscopy in solution.
The degree of grafting of spermidine on HA was also obtained.
The cross-linked derivatives of HA were synthesized by Ugi’s
reaction using formaldehyde, cyclohexylisocyanide and spermidine as diamine. The resulting networks having different
cross-linking degrees were studied by 13C CP-MAS NMR
spectroscopy. Furthermore, the study of the cross-polarization
dynamic process allowed the degree of cross-linking to be
quantitatively ascertained.
The HA-spermidine hydrogels were incubated in NO gas
under pressure to obtain the NONO-ate bond formation involving the secondary amine groups of the spermidine residues. The
kinetic of the NO-release from the diazeniumdiolate structures
was monitored in aqueous solution under physiological conditions using the colorimetric Griess assay. Soon, the potential
of the novel HA derivatives for therapeutic formulations will
be ascertained by means of in vivo experiments, especially in
1260 Biomacromolecules, Vol. 7, No. 4, 2006
wound healing processes in which both NO and hyaluronan play
important roles.
Acknowledgment. This work has been carried out with the
financial support of Fidia Farmaceutici, Abano Terme (PD),
Italy.
References and Notes
(1) Wink, D. A.; Mitchell, J. B. Free Radical Biol. Med. 1998, 25 (45), 434-456.
(2) Wang, P. G.; Xian, M.; Tang, X.; Wu, X.; Wen, Z.; Cai, T.; Janczuk,
A. J. Chem. ReV. 2002, 102, 1091-1134.
(3) Napoli, I.; Ignarro, L. J. Annu. ReV. Pharmacol. Toxicol. 2003, 43,
97-123.
(4) Frank, S.; Kämpfer, H.; Wetzler, C.; Pfeilschifter, J. Kidney Int. 2002,
61, 882-888.
(5) Schwentker, A.; Vodovotz, Y.; Weller, R.; Billiar, T. R. Nitric Oxide
2002, 7, 1-10.
(6) Hrabie, J. A.; Keefer, L. K. Chem. ReV. 2002, 102, 1135-1154.
(7) Guerènon, M.; Plateu, P.; Decorps, M. Prog. NMR Spectrosc. 1991,
23, 135-109.
(8) Metz, G.; Wu, X.; Smith, S. O. J. Magn. Reson. A 1994, 110, 219.
(9) Cook, R. L.; Langford, H. H.; Yamadagni, R.; Preston, C. M. Anal.
Chem. 1996, 68, 3979.
(10) Wu, X.; Zilm, K. W. J. Magn Reson. 1993, A102, 205.
(11) Wu, X.; Zilm, K. W. J. Magn Reson. 1993, A104, 119.
(12) Voelkel, R. Angew. Chem., Int. Ed. Engl. 1988, 27, 1468-1483.
(13) Ha, M. A.; Evans, B. W.; Jarvis, M. C.; Apperley, D. C.; Kenwright,
A. M. Carbohydr. Res. 1996, 288, 15-23.
(14) Wu, X.; Zhang, S. Phys. ReV B 1988, 37, 9827-9829.
(15) Jarvis, M. C.; Fenwick, K. M.; Apperley, D. C. Carbohydr. Res.
1996, 288, 1-14.
(16) Nims, R. W.; Darbyshire, J. F.; Saavedra, J. E.; Christodoulou, D.;
Hanbauer, I.; Cox, G. W.; Grisham, M. B.; Laval, F.; Cook, J. A.;
Krishna, M. C.; Wink, D. A. Methods: Compan. Methods Enzymol.
1995, 7, 48-54.
(17) Tomihata, K.; Ykada, Y. J. Biomed. Mater. Res. 1997, 37 (2), 243251.
(18) Crescenzi, V.; Francescangeli, A.; Renier, D.; Bellini, D. Biopolymers
2002, 64 (2), 86-94.
(19) Bulpitt, P.; Aeschlimann, D. J. Biomed. Mater. Res. 1999, 47 (2),
152-169.
(20) Pouyani T.; Kuo, J.; Harbison G.; Prestwich, G. J. Am. Chem. Soc.
1992, 114 (15), 5972-5976.
(21) Kuijpers, A. J.; Engbers, G.; Feijen, J.; De Smedt, S.; Meyvis, T.;
Demeester, J.; Krijgsveld, J.; Zoat, S. A. J.; Dankert, J. Macromolecules 1999, 32 (10), 3325-3333.
Di Meo et al.
(22) Park, S.; Park, J.; Kim, H. O.; Song, M. J.; Suh, H. Biomaterials
2002, 23 (4), 1205-1212.
(23) Ugi, I.; Lohberger, S.; Karl, R. The Passerini and Ugi reactions. In
ComprehensiVe organic synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon Press: London, 1991; Vol. II, pp 1086-1109.
(24) de Nooy, A. E. J.; Masci, G.; Crescenzi, V. Macromolecules 1999,
32, 1318-1320.
(25) de Nooy, A. E. J.; Capitani, D.; Masci, G.; Crescenzi, V. Biomacromolecules 2000, 1 (2), 259-267.
(26) Crescenzi, V.; Francescangeli, A.; Capitani, D.; Mannina. L.; Renier,
D.; Bellini, D. Carbohydr. Polym. 2003, 53, 311-316.
(27) Cowman, M. K.; Hittner, D. N.; Feder-Davis, J. Macromolecules
1996, 29, 2894-2902.
(28) Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1965, 42, 288-292.
(29) Morris, K. F.; Johnson, C. S. J. Am. Chem. Soc. 1992, 114 (8), 31393141.
(30) Morris, K. F.; Stilbs, P.; Johnson, C. S. Anal. Chem. 1994, 66, 211215.
(31) Kapur, G. S.; Cabrata, E. J.; Berger, S.; Tetrahedron Lett. 2000, 41,
7181-7185.
(32) Viel, S.; Mannina, L.; Segre, A. L. Tetrahedron Lett. 2002, 43 (14),
2515-2519.
(33) Morris, K. F.; Johnson, C. S. J. Am. Chem. Soc. 1993, 115, 42914299.
(34) Viel, S.; Capitani, D.; Mannina, L.; Segre, A. L. Biomacromolecules
2003, 4, 1843-1847.
(35) Crescenzi, V.; Francescangeli, A.; Taglienti, A.; Capitani, D.;
Mannina, L. Biomacromolecules 2003, 4, 1045-1054.
(36) Harris, R. K. In Multinuclear Magnetic Resonance in Liquids and
Solids- Chemical Application; Granger, P., Harris, H. K., Eds.; NATO
ASI Series 322; Kluwer: Dordrecht, The Netherlands, 1988; p 291.
(37) Drago, S.; Paulik, F. E. J. Am. Chem. Soc. 1960, 82 (1), 96-98.
(38) Drago, R. S.; Ragsdale, R. O.; Eyman D. P. J. Am. Chem. Soc. 1961,
83, 3 (21), 4337-4339.
(39) Beckman, J. S.; Conger, K. A. Methods: Compan. Methods Enzymol.
1995, 7, 35-39.
(40) Wink, D. A.; Christodoulou, D.; Ho, M.; Krishna, M. C.; Cook, J.
A.; Haut, H.; Randolph, J. K.; Sullivan, M.; Coia, G.; Murray, R.;
Meyer, T. Methods: Compan. Methods Enzymol. 1995, 7, 71-77.
(41) Venkataraman, S.; Martin, S. M.; Schafer, S. Q.; Buettner, G. R.
Free Radical Biol. Med. 2000, 29 (6), 580-585.
(42) Kudo, S.; Bourrassa, J. L.; Boggs, S. E.; Sato, Y.; Ford, P. C. Anal.
Biochem. 1997, 247, 1193-202.
(43) Miles, A. M.; Chen, Y.; Owens, M. W.; Grisham, M. B. Methods:
Compan. Methods Enzymol. 1995, 7, 40-47.
BM050904I