Collagen cross linking agents: design and development of a multifunctional cross linker.

Abstract

A new cross linking reagent based on the first-generation polyamidoamine dendrimer (G.1 PAMAM) has been synthesized by reaction of the PAMAM with eight equivalents of p-nitrophenyl diazopyruvate. The resulting water-soluble octadiazopyruvoyl PAMAM (8G.1 DAP, 1.3) was shown to undergo Wolff rearrangements upon photolysis in methanol at lambda > 300 nm to yield the methyl esters of the ketenes formed from the loss of nitrogen. 8G.1 DAP also forms strong bonds with dehydrated collagen with glass as high as 36 N cm(-2). Collagen to collagen bonds with tensile strengths as high as 92 N cm(-2) were observed with fully dehydrated tissues. The bonding decreased rapidly with increasingly hydrated tissue possibly due to the increased distance between the collagen fibrils and the competition of H2O for the free ketene functions.

Full text

Page 1

Collagen Cross Linking Agents: Design and Development of a
Multifunctional Cross Linker
Richard S. Givens*1,2, Abraham L. Yousef1, Shaorong Yang1and George T. Timberlake1,2
1Department of Chemistry, University of Kansas, Lawrence, KS
2Department of Ophthalmology, Anatomy & Cell Biology, University of Kansas Medical Center, Kansas City, KS
Received 27 July 2007, accepted 26 August 2007, DOI: 10.1111⁄j.1751-1097.2007.00218.x
ABSTRACT
A new cross linking reagent based on the first-generation
polyamidoamine dendrimer (G.1 PAMAM) has been synthe-
sized by reaction of the PAMAM with eight equivalents of
p-nitrophenyl diazopyruvate. The resulting water-soluble octa-
diazopyruvoyl PAMAM (8G.1 DAP, 1.3) was shown to
undergo Wolff rearrangements upon photolysis in methanol at
k > 300 nm to yield the methyl esters of the ketenes formed
from the loss of nitrogen. 8G.1 DAP also forms strong bonds
with dehydrated collagen with glass as high as 36 N cm)2.
Collagen to collagen bonds with tensile strengths as high as
92 N cm)2were observed with fully dehydrated tissues. The
bonding decreased rapidly with increasingly hydrated tissue
possibly due to the increased distance between the collagen
fibrils and the competition of H2O for the free ketene
functions.
INTRODUCTION
An effective cross linking agent for Type I collagen to replace
sutures or attach tissue to other materials is a research
objective of this and other research groups (1–8). However,
application of these techniques to tissue cross linking has been
particularly challenging because of difficulties arising from
biological incompatibility as well as the restrictions on physical
and chemical properties which hamper effective covalent
bonding to collagenous protein. Commonly bi-functional
reagents such as glutaraldehyde (9) or bis-isocyanates (10)
have been explored but these reagents have deleterious side
effects such as reacting slowly with glycosaminoglycans and
other endogenous nucleophiles. Furthermore, diffusion of
these cross linking agents and the interstitial penetration have
been problematic (11).
Recently, new photochemically activated approaches for
cross linking have been reported (12,13). For example,
bifunctional naphthalimide reagents (14,15) which can be
activated at 457.9 nm by an argon laser cross link human
meniscus, human articular cartilage (16) and rabbit skin (17)
at tensile strengths as high as 20 N cm)2
corneal stroma (18) with tensile strengths of 30 N cm)2.
Photosensitization with rose bengal (19–22) at 514 nm sealed
full-thickness corneal incisions as well as penetrating kera-
and human
toplasties in rabbits (20). Riboflavin has also served as a
photosensitizer for collagen cross linking in the treatment
ofkeratoconusandcorneal
Pedone and Brocchini showed that polyethylene glycol
(PEG) can be conjugated with lysozyme and ribonuclease
by a photochemical approach (25).
Approaches using light-activated cross linking agents are
intrinsically appealing because they provide a greater measure
of spatial and temporal control over the bonding event which
should minimize the amount of physical trauma to the
surrounding tissue. However, the mechanisms of the cross
linking processes for most of the photochemical agents are
unknown and, more seriously, often produce unknown
byproducts.
Our approach has been the design and synthesis of reagents
that rely on known, well-characterized photochemical reac-
tions of a-diazopyruvates. This reagent undergoes a clean
Wolff rearrangement to form a transparent peptide bond via a
reactive ketene intermediate (Scheme 1). This approach to
cross linking collagen must meet the following criteria. (1) The
photochemical activation of the reagent must occur at wave-
lengths greater than 330 nm to avoid damage to the tissue. (2)
The photoactivated chromophore must be completely removed
or destroyed by the photolysis reactions. (3) Any reactive
intermediates (in this case, and a-ketoketene) must be suffi-
ciently reactive to covalently bond to a functional group or
groups on collagen. (4) Any new functional groups that are
generated must be transparent in the near-UV visible range. In
considering the functional groups available for cross linking
Type I collagen, we have focused our attention on the
nucleophilicity of c-amino group on the lysyl side chain of
collagen. In previous studies we established (26) that typical 1?
and 2? amine fluorescent labeling reagents such as o-phthal-
aldehyde or dansyl chloride can be covalently attached to Type
I collagen.
To selectively capture the target c-amino lysine function,
we designed a reagent that photochemically generated ketene
through a Wolff rearrangement. Ketenes react rapidly with
1? and 2? amines, thereby providing the desired connectivity
to collagen. Lawton and coworkers (27–29) have demon-
strated the feasibility of this strategy in earlier studies on
photoaffinity labeling of proteins and in at least one instance
that cross linking of a protein is possible (27,30). We
captured this chemistry by designing a bis-substituted PEG
diazopyruvamide, i.e. N,N¢-bis-(3-diazopyruvoyl)-2,2¢-(ethy-
ulcers(23,24).Finally,
*Corresponding author email: givensr@ku.edu (Richard S. Givens)
?2007TheAuthors.JournalCompilation.TheAmericanSocietyofPhotobiology 0031-8655/07
Photochemistry and Photobiology, 2007, 83: 1–8
1

Page 2

lenedioxy)bis(ethylamine, DPD 1.1) as our first cross linking
agent. The photochemical activation and rearrangement are
shown in Scheme 1 (27,31).
Photolysis of DPD 1.1 at 320–550 nm sequentially gener-
ates two ketene amides (see 1.2) through the photo-Wolff
rearrangement. Nucleophilic attack on the ketene by a c-amino
group on lysine or hydroxylysine residue forms a malonamide,
i.e. a peptide bond. We have demonstrated that collagen cross
linked by photoreaction with 1.2 strongly binds two strips of
dehydrated bovine corneal tissue with tensile strengths up to
46.5 N cm)2(26). We extended the study to include dehy-
drated corneal tissue strips bonded to gel strips or to glass with
comparable strengths. Unfortunately, these initial studies only
produced measurably strong bonds with fully dehydrated and
partially hydrated tissue, suggesting that further design mod-
ifications were required. Critical features such as low aqueous
solubility, the short length of the PEG tether, the rather
modest selectivity of the ketene for amines over H2O and
hydroxyl groups and the low number of reactive groups per
cross linking agent⁄tether required modification of our
original design.
Toward this end, we are now building functionalized
poly(amidoamine) PAMAM dendrimers (32) for the tether.
These were selected because of their ready availability in a
variety of sizes (or generations), the well-known chemistry of
dendrimers and the multifunctional nature of the reagents.
Dendrimers exhibit good water solubility properties and
possess multiple sites for attachment of the diazopyruvoyl
chromophore that increase exponentially with each generation.
Furthermore, the backbone functional structure of the den-
drimer is composed of peptide linkages which are near-UV
transparent. In fact, recent application of another dendrimer
has been reported by Wathier et al. (33) employing lysine-
based peptide dendrons, to generate hydrogels for repair of
corneal incisions. Amino-terminated cysteine dendrons-PEG
dialdehyde polymers have also been employed in a hydrogel to
seal corneal incisions in an enucleated eye (34). As noted for
earlier studies, mechanisms of these wound-sealing processes
are not known. In addition, the hydrogel approaches offer
little or no temporal and spatial control that are possible with
photoinitiated reactions (34).
Based on our previous DPD results (26) and the structure
and properties of the PAMAM dendrimers, we report a new,
multifunctional PAMAM diazopyruvamide to serve as a more
versatile, biocompatible light-activated cross linking agent for
Type I collagen. Our initial studies with the multifunctional
diazopyruvoyl dendrimers generated from readily available
first-generation (G.1) PAMAM dendrimer (Scheme 2) are
reported here.
N
H
O
O
O
N2
O
O
N2
O
N
H
collagen
Lys
N
H
C
O
O
H2NH2CH2CH2CH2C
h
320-500 nm
-N2
H
N
N2
O
O
O
O
h
-N2
repeat
CH2CH2CH2CH2NH
N
H
O
O
O
Lys
O
H
N
OO
collagen
Lys
HNH2CH2CH2CH2C
collagen
DPD, 1.1
1.2
Scheme 1.
N
HN
N
N
NH
N
HN
O
O
O
N
H
N
N
H
HN
NH
NH
HN
NH
HN
O
O
O
O
O
NH
HN
NH
N
O
HN
N
H
H
N
O
O
NH
N2
O
O
O
O
O
O
N2
N2
N2
O
O
N2
O
O
N2
O
O
N2
O
O
NH
HN
O
O
O
N2
N
HN
N
N
NH
N
HN
O
O
O
N
H
N
NH2
HN
NH2
NH
H2N
NH
HN
O
O
O
O
O
NH2
H2N
NH
N
O
HN
N
H
H2N
O
O
NH2
NHO
NH2
O2NO
O
O
N2
(p-nitrophenyl-3-diazopyruvate)
1.4
chloroform, rt, 42 h
8G.1 PAMAM
1.3 8G.1 DAP
Scheme 2. Synthesis of diazopyruvoyl PAMAM (8G.1 DAP, 1.3).
2Richard S. Givens et al.

Page 3

MATERIALS AND METHODS
General. All starting materials were purchased commercially and
used without further purification unless otherwise noted.1H and13C
NMR spectra were obtained on a Bruker 400 MHz instrument
unless otherwise noted. Samples were dissolved in chloroform-d
(CDCl3), dimethyl sulfoxide-d6 (DMSO-d6), N,N¢-dimethylforma-
mide-d7(DMF-d7) or deuterium oxide (D2O) and chemical shifts are
reported in parts per million (d). Precoated silica plates from Merck
were used for thin layer chromatography and spots were visualized
under a UV lamp. Chromatographic purification was effected with
flash chromatography using standard grade (32–63 lm) silica gel
(Sorbent Technologies, Atlanta, GA). IR spectra were obtained on a
Shimadzu FTIR-8400S spectrophotometer; results are reported in
cm)1. HPLC analyses were conducted with a C18 Econosphere
250 · 4.6 mm analytical column (Altech Associates, Inc., Deerfield,
IL) connected to a Rainin dual pump system. UV⁄Vis spectra were
measured on a Cary 100 spectrophotometer (Varian, Inc., Palo Alto,
CA). Irradiations were performed with a Novacure 2000 near-UV
(320–550 nm) light source (Exfo, Missasagua, ON, Canada) unless
otherwise noted.
Synthesis of octa-substituted diazopyruvoyl PAMAM (8G.1 DAP,
1.3). A 10 mL round bottom flask was flame dried and charged with
0.3 mL of a solution of (G.1) PAMAM dendrimer (Aldrich, 20
wt.% in methanol). The methanol was removed under reduced
pressure(6–7 h).Theresulting
0.042 mmol) and dry chloroform (3 mL) was added, along with p-
nitrophenyl 3-diazopyruvate (30) (1.4, 119 mg, 0.506 mmol) and the
solution stirred at RT in the dark. About 36 h later, the solution
together with the entire residue on the flask wall were transferred to
125 mL of a separatory funnel. The organic layer was discarded and
the water layer was further extracted with 10 mL of chloroform
three to six times until the water layer appeared colorless. The
aqueous solution was cooled by dry ice for 30 min and lyophilized
(24 h) to afford octa-diazopyruvoyl PAMAM (1.3, 8G.1 DAP) as a
brown solid: 51 mg, 55%;1H NMR (400 MHz, D2O) d = 6.41 (s,
4.27H), 6.15 (s, 2.65H), 5.53 (s, 0.60H), 5.39 (s, 0.48H), 3.49–3.40
(bm, 9.07H), 3.37–3.24 (bm, 33.4H), 3.17 (bm, 14.4H), 3.06–2.99
(bm, 19.9H), 2.55 (bm, 23.3H);
d = 182.14, 171.63, 171.29, 160.76, 54.91, 54.41, 52.24, 49.81, 38.46;
IR (KBr, cm)1) 3304, 3080, 2943, 2122 (C=N=N), 1636 (C=O),
1535, 1364, 1119, 779; UV–Vis (H2O) kmax(log e (M)1cm)1)) 253
(4.92), 300 (4.76); Mass Spec. Calc’d for C86H128N42O28: 2196.99;
Found: ESMS(m⁄z) (M+H) 2197.53 ± 0.78; MALDI-TOF m⁄z
2203.
Photochemistry of octa-diazopyruvoyl PAMAM (8G.1 DAP, 1.3)
in MeOH. A 10 mL Pyrex tube was charged with 8G.1 DAP (1.3,
1.4 mg, 0.64 lmol) in 5 mL of methanol. After dissolution was
complete, the tube was transferred to a Rayonet photoreactor fitted
with four RPR-3000 A˚lamps and photolyzed for 5 min. Sam-
ple aliquots (0.4 mL) were taken after 0, 2 and 5 min of photolysis.
Thealiquotsweretransferred to
diluted with methanol to the mark, and analyzed by UV (200–
400 nm). The disappearance of the kmax300 nm (log e 4.67) band of
1.3 indicated complete conversion of 1.3 to the methyl ester (see
Fig. 3).
Glass slides. Gold Seal?Microscope Slides (Clay Adams, Parsip-
pany, NJ) were cut into ca 0.5 · 2.5 cm pieces and cleaned in a
Piranha solution (70:30 vol⁄vol concentrated H2SO4:30% H2O2) for
12 h at RT.†After cleaning in the Piranha bath, the glass slides
were washed in deionized water for at least 12 h and then dried and
stored in a petri dish until use.
Bovine cornea. Corneas were dissected from ca 2 h postmortem
bovine eyes obtained from a local abattoir. Epithelial and endothelial
cell layers were removed by scraping, and the corneas were cut into two
ca 5 · 15 mm strips. Each strip was cut in half, parallel to the corneal
surface using a fine scalpel. The four resulting strips were weighed and
placed stromal side up on glass slides, covered with another slide and
dehydrated in a vacuum lyophilizer for varying amounts of time prior
to use. The partially dehydrated collagen tissue was weighed just
before use.
residue wasweighed (60 mg,
13C NMR (500 MHz, DMF-d7)
a10 mLvolumetric flask,
Cross linking with bovine corneal tissue. A plastic snap-cap vial was
charged with 8G.1 DAP (1.3, 23 mg, 10 lmol) and water (52 lL) and
the material was dissolved with gentle warming and shaking in the
dark. The solution was kept in an amber jar between experiments to
protect the compound from incident light. Three microliters of 8G.1
DAP solution was used for each experiment. For the co-monomer
experiments,2,2¢-(ethylenedioxy)
6.8 mmol) was dissolved in water (10 mL) in a volumetric flask. In a
separate snap-cap vial 12 lL of 0.2 M
of the freshly prepared EDEA solution to provide 2.2 equivalents of
EDEA.
Corneal samples were prepared as described earlier. For irradiation
and bonding studies, one corneal sample was placed stromal side up on
a quartz slide with the other sample placed on top, forming an overlap
area of ?5 · 5 mm. Three microliters of a solution of 0.2 M
in water was instilled between the two corneal samples after which a
second quartz slide placed on top. Light fingertip pressure was applied
evenly to the quartz slides which were then stabilized for further
manipulation with small clamps.
The glass slides were then exposed to UV–Vis irradiations. The
exposure time was adjusted to provide an exposure of 333 J for the
bonding process in accordance with the calibrated power of the light
source. Tensile strength testing was carried out as described previously
(31). The overlap areas were defined by a light tan residue from the
photoreaction of 8G.1 DAP.
Bonding cornea to glass. The glass slides were prepared as described
above. A clean glass (ca 0.5 · 2.5 cm) slide was put on a quartz slide
and on one end of it was added 3 lL of an aqueous solution of 0.2 M
8G.1 DAP, then the cornea strip (ca 0.5 · 2.5 cm) was placed on top
with the stromal side in contact with the glass, thus creating an
overlapping area of ?5 · 5 mm2. A second quartz slide was applied on
the top of the glass slides with two clamps helping to assure full contact
of the glass slides. The overlapping area was then exposed to the UV
light source for the requisite time to provide an exposure of 333 J for
the cornea-to-glass bonding. Tensile strength testing was determined as
reported previously (31). The overlap areas were defined by a light tan
residue from the photoreaction of 8G.1 DAP.
Bonding glass to glass. The glass slides were prepared as described
above. A clean glass slide (ca 0.5 · 2.5 cm) was placed on a quartz
slide and 3 lL of an aqueous solution of 0.2 M
at one end. A second glass slide (ca 0.5 · 2.5 cm) was placed on top,
thus creating an overlapping area of ?5 · 5 mm2and a second quartz
slide applied on the top of the glass slides with two clamps providing a
minimum amount of pressure, to assure surface contact. The overlap-
ping area was then irradiated (333 J). Tensile strengths were measured
as reported previously (31) except that O-rings were attached to the
nonbonded end of each glass slide with cyanoacrylate glue to provide a
clamping surface for the mechanical force measurements. The overlap
area measurements are described later.
diethylamine(EDEA,1 mL,
M 8G.1 DAP was mixed with 8 lL
M 8G.1 DAP
M
M 8G.1 DAP was applied
RESULTS
Synthesis and exploratory photochemistry
Synthesis of 8G.1 DAP shown in Scheme 2 was accomplished
using 10–12 equivalents of p-nitrophenyl 3-diazopyruvate in
chloroform for each equivalent of 8G.1 PAMAM. The
reaction was complete within 42 h (1H NMR) giving 8G.1
DAP in 55% yield.
The incorporation of eight diazopyruvoyl groups per
PAMAM dendrimer was established by1H NMR and mass
spectral analyses. A calculated mass of 2196.99 was exactly
matched by the experimental value of 2197 (ESMS). The1H
NMR, however, was more intriguing. The spectrum of 8G.1
DAP (1.2, Fig. 1) showed four characteristic downfield dia-
zomethine hydrogen singlets at d 6.41, 6.15, 5.53 and
5.39 p.p.m. in a ratio of 1.0:0.62:0.13:0.11. The chemical shifts
are in agreement with the expected
similar structures reported by Goodfellow et al. (27) and our
earlier studies. The unusual number of diazomethine hydrogen
1H NMR results from
†Caution: Piranha solution reacts violently with organic compounds and should
be handled carefully behind a blast shield (35).
Photochemistry and Photobiology, 2007, 833

Page 4

signals is attributable to four of eight possible rotational
isomers arising from the restricted rotation of the conjugated
amide, the ketone carbonyl and diazomethane groups (i.e. the
dihedral angles a, b and c in RNHcCObCOaCHN2). We have
not attempted to assign these signals to the individual
conformers although consideration of the steric and dipole–
dipole interactions would suggest that an anti-anti arrange-
ment of the three groups would be preferred (Fig. 2).
The up field region of the spectrum is characterized by a
number of broad signals from 2.4 to 3.5 p.p.m., corresponding
to the methylene protons of the dendritic backbone. The ratio
of the total integration of these signals to those of the four
downfield methine protons was approximately 11.7:1, which
closely matches the predicted ratio of 12.5:1 (complete
derivatization) indicating a yield of 93.6% derivatization of
all eight amino groups (see 1.3 in Scheme 2).
The
resonances at d 182, 172 and 161 p.p.m. corresponding to the
diazoketone and amide functions. Other significant spectral
features were an intense IR band for the diazo group at
2122 cm)1and the UV absorption (H2O) maxima at 253 (log e
4.92) and 300 nm (log e 4.86), similar to those reported by
Goodfellow et al. (27) and by us (26) for other diazopyruva-
mides.
The photochemical rearrangement of 8G.1 DAP 1.3 was
followed by UV. A sample of 0.13 mM
irradiated at 300 nm and monitored at 0, 2 and 5 min (Fig. 3)
showed a rapid, efficient disappearance of the diazoketo
chromophore. No new absorptions appeared above 220 nm
indicating quantitative conversion of 8G1 DAP to product
(Scheme 3).
13C NMR spectrum of 8G.1 DAP showed carbon
M 1.3 in methanol
Dendrimer cross linking studies
Unlike the bifunctional DPD, G.1 DAP was moderately
soluble in methanol and very soluble in water. The latter
solvent was chosen for use in cross linking studies with cornea
as it is the more biologically benign of the solvents and would
be more compatible with the interfibrillar matrix that sur-
rounds collagen.
Glass-to-cornea bonding with 8G.1 DAP. Corneas were pre-
pared as described earlier for bonding cornea to glass with
DPD. Aqueous solutions of 8G.1 DAP were administered
between the tissue and glass and then irradiated for an
exposure of 333 J. The results of several runs are shown in
Fig. 4 and Table 1.
8G.1 DAP glass-to-corneal stroma bonding exhibits the
same general trends that were observed for DPD glass-to-
corneal stroma bonding, i.e. the strongest bonding occurs at
lower hydration levels. In addition, all samples tested with
8G.1 DAP produced a measurable bond at low hydration
values (H < 2).
Glass-to-glass bonding with 8G.1 DAP. The strength of bond-
ing glass-to-glass was also tested. The generation of N2was
readily observed in these experiments and the effervescence
apparently resulted in the formation of small bubbles between
the plates as illustrated in Fig. 5. This disruption of the surface
contact may be the source of the weak bonding, especially for
the glass-to-cornea bonding. Methods for minimizing this
effect are under investigation.
Cornea-to-cornea bonding with G.1 DAP (1.3). The results
from bonding of two cornea tissue samples are shown in Fig. 6
and Table 2. Hydration levels were varied for the tissue
samples as shown in the Table and Figure. The results of the
N
R
O
O
N2
H
H
N
R
O
O
N2
H
H
anti-anti-anti conformer
syn-anti-anti conformer
Figure 2. Examples of restricted rotations for the diazopyruvoyl
moiety.
Figure 1. 1H NMR spectrum of 8G.1 DAP 1.3 in D2O at 400 MHz
(the absorption at 4.9 p.p.m. is due to HOD).
Wavelength (nm)
200220240260 280 300320340
Abs
0.0
0.1
0.2
0.3
0.4
0.5
t = 0 min
t = 2 min
t = 5 min
Figure 3. UV spectral analysis of photolysis of 8G.1 DAP 1.3 at
300 nm in methanol.
HN
H
N2
O
O
8
h , 300nm
methanol
dendrimer
OMe
N
H
OO
8
dendrimer
1.3
Scheme 3.
4 Richard S. Givens et al.

Page 5

tensile strength measurements paralleled those of the DPD
bonding studies in that the strongest bonds were obtained at
lower hydration levels. The maximum tensile strengths with
8G.1 DAP (92 N cm)2) were nearly twice the value of
47 N cm)2obtained for DPD.
Although the measured tensile strengths at intermediate
hydrationlevels(H = 1–3)
(6.4 ± 3.9 N cm)2), they were within the experimental error
identical to those for DPD at similar hydration levels
(5.9 ± 6.2 N cm)2). Even the addition of 2.2 equivalents of
EDEA as a co-monomer with 8G.1 DAP resulted in no
significant improvement in the tensile strength (Fig. 6).
were slightlyhigher
Concentration dependence of cornea-to-cornea bonding. In
order to optimize the bonding, the concentration of 8G.1
DAP was varied. For these studies, two cornea strips (about
0.5 cm wide) were cut into two pieces lengthwise, thus giving
two strips of half the original width with the same hydration
level. As shown in Fig. 7, the tensile strengths of the bonded
strips of cornea decreased with increasing hydration. For
1 < H < 2, the data clearly showed 0.2 M
better bonding than 0.03 MM concentrations. For H > 2, the
tensile strengths remain low. With H-values between 2 and 3.1,
M 8G.1 DAP gave
12
Tensile strength (N/cm2)
0
3
5
10
15
20
40 6080100
050100
0
50
100
0123
% Hydration
H
Figure 4. Maximum tensile strengths of glass-to-corneal stroma
bonds with 0.2 MM 8G.1 DAP 1.3 in H2O as a function of tissue
hydration value. Inset shows that the ‘‘dry’’ tissue (H < 1) resulted
in bond strengths of 11.5–57.5 N cm)2. No bonds formed with H2O
only.
Table 1. Tensile bond strengths for glass-to-corneal stroma G.1 DAP
1.3.
Sample no.H
Tensile strength
(N cm)2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
0
0
0
0.09
0.09
0.18
0.46
0.72
1.23
1.59
1.69
2.14
2.61
2.62
3.19
3.20
3.24
3.53
3.58
21.49
29.11
36.77
11.57
22.29
21.81
11.54
12.26
2.43
6.59
5.30
0
0
0
0
0
5.13
0
0
H = hydration value of corneal sample.
2 mm
Figure 5. Glass-to-glass bonds with 0.2 M
large dark areas are nonbonded regions generated by N2 from
photolysis of the diazo group. The lines outlining the cavities are the
bonding regions. The tensile strength is 60 N cm)2.
M 8G.1 DAP 1.3 in H2O. The
H
0123
Maximum stress (N/cm2)
0
4
20
40
60
80
100
% Hydration
02040 6080 100
DAP/H2O
DAP + PEG
H2O only
Figure 6. Maxium stress of cornea-to-cornea bonding with aqueous
0.2 MM 8G.1DAP 1.3 versus tissue hydration value (H). Circles (s)
indicate level of hydration. Triangles (n) indicate sample pairs in
which the bonding solution contained 2.2 equivalents of the co-
monomer EDEA (2,2¢-(ethylenedioxy)diethylamine). Squares (h)
indicate H2O control runs.
Photochemistry and Photobiology, 2007, 835