Glycobiology vol. 10 no. 3 pp. 273–281, 2000
Microanalysis of enzyme digests of hyaluronan and chondroitin/dermatan sulfate by
fluorophore-assisted carbohydrate electrophoresis (FACE)
Anthony Calabro1, Maria Benavides, Markku Tammi2,
Vincent C.Hascall and Ronald J.Midura
Department of Biomedical Engineering/ND20, Lerner Research Institute, The
Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195,
USA and 2Department of Anatomy, University of Kuopio, P.O. Box 1627,
SF-70211, Kuopio, Finland
Received on July 14, 1999; revised on September 3, 1999; accepted on
September 9, 1999
Hyaluronan and chondroitin/dermatan sulfate are
glycosaminoglycans that play major roles in the biomechanical properties of a wide variety of tissues, including
cartilage. A chondroitin/dermatan sulfate chain can be
divided into three regions: (1) a single linkage region
oligosaccharide, through which the chain is attached to its
proteoglycan core protein, (2) numerous internal repeat
disaccharides, which comprise the bulk of the chain, and
(3) a single nonreducing terminal saccharide structure.
Each of these regions of a chondroitin/dermatan sulfate
chain has its own level of microheterogeneity of structure,
which varies with proteoglycan class, tissue source, species,
and pathology. We have developed rapid, simple, and sensitive protocols for detection, characterization and quantitation of the saccharide structures from the internal
disaccharide and nonreducing terminal regions of
hyaluronan and chondroitin/dermatan sulfate chains.
These protocols rely on the generation of saccharide structures with free reducing groups by specific enzymatic treatments (hyaluronidase/chondroitinase) which are then
quantitatively tagged though their free reducing groups
with the fluorescent reporter, 2-aminoacridone. These
saccharide structures are further characterized by modification through additional enzymatic (sulfatase) or chemical
(mercuric ion) treatments. After separation by fluorophore-assisted carbohydrate electrophoresis, the relative
fluorescence in each band is quantitated with a cooled,
charge-coupled device camera for analysis. Specifically, the
digestion products identified are (1) unsaturated internal
∆disaccharides including ∆DiHA, ∆Di0S, ∆Di2S, ∆Di4S,
∆Di6S, ∆Di2,4S, ∆Di2,6S, ∆Di4,6S, and ∆Di2,4,6S; (2) saturated nonreducing terminal disaccharides including DiHA,
Di0S, Di4S and Di6S; and (3) nonreducing terminal
hexosamines including glcNAc, galNAc, 4S-galNAc, 6SgalNAc, and 4,6S-galNAc.
Key words: chondroitin sulfate/dermatan sulfate/fine structure/
hyaluronan/microanalysis
1To
whom correspondence should be addressed
© 2000 Oxford University Press
Introduction
Sulfated glycosaminoglycans such as chondroitin sulfate or
dermatan sulfate are composed of three regions, a linkage
oligosaccharide, connecting the chain to the core protein, a variably sulfated disaccharide repeat structure within the chain and a
non-reducing terminus. These regions are of current interest
since they are suggested to confer biologic functions on particular chain populations (see Introduction to Calabro et al., 2000).
Many microanalytical techniques are now available to separate
and quantitate nanogram amounts of the nonreducing termini
(Otsu et al., 1985; Hascall et al., 1995; Midura et al., 1995;
Plaas et al., 1996, 1997), unsaturated disaccharides (Carney and
Osborne, 1991; Karamanos et al., 1994; Midura et al., 1994;
Deutsch et al., 1995; Kitagawa et al., 1995) and linkage
oligosaccharides (Sugahara et al., 1988; Shibata et al., 1992)
produced by chondroitinase digestion of chondroitin/dermatan
sulfate chains. The analyses include capillary zone electrophoresis (Carney and Osborne, 1991; Deutsch et al., 1995;
Kitagawa et al., 1995) and high-performance liquid chromatography (Otsu et al., 1985; Shibata et al., 1992; Karamanos et al.,
1994; Midura et al., 1994, 1995; Hascall et al., 1995; Plaas et
al., 1996, 1997). However, these existing analytical methods
have distinct disadvantages. They normally require long, labor
intensive preparation times which include: (1) isolation of the
chondroitin sulfate chains from tissues or cells, (2) purification
of the chondroitin sulfate chains from other macromolecules, (3)
purification of the chondroitinase digestion products from the
enzyme after digestion, (4) purification of the chondroitinase
digestion products from reducing agents used to stabilize the
digestion products during analysis, and/or (5) purification of
derivatized chondroitinase digestion products from unreacted
fluorescent tag used as a reporter of mass. In addition, they
normally allow for analysis of only one sample at a time with
each analysis taking on the order of hours. 1H-NMR and mass
spectroscopy have the additional disadvantage of requiring large
quantities of starting material (Sugahara et al., 1988). The
present work was therefore undertaken to provide a new methodological approach, which would allow for rapid, simple, and
sensitive detection and quantitation of internal disaccharide and
nonreducing terminal structures of chondroitin/dermatan sulfate
and hyaluronan without the disadvantages of previous methods.
Results
Fluorescent derivatization and polyacrylamide gel
electrophoresis of the internal ∆disaccharide structures derived
from hyaluronan and chondroitin/dermatan sulfate chains
The internal disaccharides of hyaluronan and chondroitin/
dermatan sulfate chains yield unsaturated ∆disaccharides after
273
A.Calabro et al.
Fig. 2. FACE analyses of hyaluronan and chondroitin sulfate ∆disaccharides
derivatized with AMAC as described in Materials and methods. Each lane
contains the following AMAC-derivatized ∆disaccharide standard: ∆DiHA
from hyaluronan (lane 1), and ∆Di0S, ∆Di2S, ∆Di4S, ∆Di6S, ∆Di2,4S,
∆Di2,6S, ∆Di4,6S, and ∆Di2,4,6S from chondroitin sulfate (lanes 2–9,
respectively). The standard lanes (S7) contain a mixture of seven AMACderivatized ∆disaccharides from top to bottom: ∆DiHA, ∆Di0S, ∆Di6S,
∆Di4S, ∆Di2S, ∆Di4,6S, and ∆Di2,4,6S.
Fig. 1. Schematic showing the structure of a 4-sulfated tetrasaccharide
(Di4SDi4S), and its two chondroitinase digestion products (reaction 1). The
left column shows the products expected following subsequent mercuric ion
treatment (reaction 2), and AMAC derivatization (reaction 3) of the saturated
disaccharide product (Di4S), which is representative of other nonreducing
terminal disaccharide structures of chondroitin sulfate chains. The right
column shows the products expected following similar treatments of the
unsaturated ∆disaccharide product (∆Di4S), which is representative of other
internal disaccharide structures of chondroitin sulfate chains. The structure of
the fluorotag, 2-aminoacridone (AMAC), is shown at the bottom left.
chondroitinase digestion (Figure 1). Standard hyaluronan and
chondroitin/dermatan sulfate ∆disaccharides were derivatized
with 2-aminoacridone (AMAC) as described in Materials and
methods (Figure 1). Figure 2A shows the positions of the two
unsulfated and three monosulfated AMAC-derivatized ∆disaccharides in the FACE analysis: ∆DiHA from hyaluronan (lane
1), and ∆Di0S, ∆Di2S, ∆Di4S, and ∆Di6S from chondroitin/
dermatan sulfate (lanes 2–5, respectively). Figure 2B shows the
positions of the three disulfated and one trisulfated AMAC-derivatized ∆disaccharides from chondroitin/dermatan sulfate:
∆Di2,4S (or ∆DiB), ∆Di2,6S (or ∆DiD), ∆Di4,6S (or ∆DiE), and
∆Di2,4,6S (or ∆DiTriS) (lanes 6–9, respectively). With the
exception of the ∆Di2,4S and ∆Di2,4,6S, which run at the same
position in the electrophoresis front, all of these ∆disaccharides
are clearly resolved. Figure 3 shows standard curves for four
∆disaccharides, ∆DiHA, ∆Di0S, ∆Di4S, and ∆Di6S. The four
curves superimpose indicating that the AMAC fluorotag gives
the same molar fluorescence value for each derivative with a
free reducing terminus. The curves show linearity on a log-log
plot over the range from 6.25 to 100 pmol, (∼2–50 ng for these
∆disaccharides) as measured by hexuronic acid assay.
274
Fig. 3. Quantitation of AMAC-derivatized hyaluronan and chondroitin sulfate
∆disaccharides after separation by FACE. Mixtures containing from 6.25 to 100
pmol each of the indicated AMAC derivatives were separated by FACE. The gel
(inset) was then imaged using a Quantix cooled-CCD camera, and the images
analyzed using the Gel-Pro Analyzer program as described in Materials and
methods. The relative fluorescence for each band is plotted versus pmoles of
∆disaccharides as determined by hexuronic acid analysis (R = 0.997, P = 0.00022).
Fluorescent derivatization and polyacrylamide gel
electrophoresis of hyaluronan ladders
Figure 4 shows FACE analyses of AMAC-derivatized products from partial digestion of a constant amount of hyaluronan
for a constant time with serial dilutions of testicular hyaluronidase. This enzyme is a hydrolase with endo-hexosaminidase
specificity, and gives a ladder of oligomer digestion products
that differ by one disaccharide repeat. The partial digests at the
lower enzyme concentrations (lanes 4 and 5) reveal a ladder
extending above 50 disaccharides. As the enzyme concentration increases, the primary end products HA4 and HA6
Microanalysis of hyaluronan and chondroitin/dermatan sulfate
Fig. 5. FACE analyses of hyaluronan and chondroitin sulfate ∆disaccharides
treated with mercuric ion, and then AMAC derivatized. A shift in mobility of a
band from that of the original derivatized ∆disaccharide structure to that of its
derivatized product(s) after mercuric ion treatment is indicated by arrows.
Equal signs indicate that co-migrating bands are the same structure. The
standard lanes labeled S5 contain the following mixture of five AMACderivatized hexosamines listed as they appear from top to bottom: galNAc,
glcNAc, 6S-galNAc, 4S-galNAc, and 4,6S-galNAc.
Fig. 4. FACE analyses of AMAC-derivatized products from partial digestion of
a constant amount of hyaluronan (100 µg) for 4 h at 37°C with 1:3 serial
dilutions of testicular hyaluronidase starting at 1000 U/ml (lanes 2–5). The
relative positions of the saturated hyaluronan oligomers containing 1 (HA2), 2
(HA4), 3 (HA6), 4 (HA8), 5 (HA10), 10 (HA20), 15 (HA30), 20 (HA40), and 25
(HA50) disaccharides are indicated. Lane 1 contains a standard mixture of three
purified, AMAC-derivatized hyaluronan oligomers (HA10, HA14, and HA18)
used to index the ladder.
increase. They no longer migrate on the basis of size and show
an inversion in mobility, with the HA6 overlapping the HA8
and the HA4 migrating at the level of HA14. The digests at the
higher enzyme concentrations (lanes 2 and 3) also show a
small amount of the HA2 (DiHA) disaccharide, a minor
product of this enzyme, which comigrates with ∆DiHA (data
not shown). This shift in mobility away from one based solely
on size is presumably the result of the smaller oligosaccharides
interacting with borate in the electrophoresis buffer. The presence of borate in the Glyko gel running buffer was inferred
from experiments in which a gel running buffer from Novex
(#LC6675) containing 89 mM Tris, 89 mM boric acid, and
2 mM EDTA, pH 8.3 similar to that described in Jackson et al.
(1991, 1994) was substituted, and produced identical results
(data not shown). The interaction with borate makes possible
the separation of AMAC-derivatized oligosaccharides with
similar molecular weights, but different chemistries as seen in
Figures 2, 5, 6, and 7. A standard mixture of three purified,
AMAC-derivatized hyaluronan oligomers, HA10, HA14 and
HA18, is shown in lane 1 to index the ladder.
Fluorescent derivatization and polyacrylamide gel
electrophoresis of the nonreducing terminal monosaccharide
structures derived from chondroitin/dermatan sulfate chains
The nonreducing ends of some chondroitin/dermatan sulfate
chains yield substituted galNAc after chondroitinase digestion.
Mercuric ion treatment quantitatively removes the ∆hexuronic
acid from the ∆disaccharides, releasing the hexosamine
portion (Ludwigs et al., 1987), Figure 1. This chemistry was
used to assign elution positions for all the potential galNAc
non-reducing termini. Figure 5 shows FACE analyses of the
one hyaluronan and eight chondroitin/dermatan sulfate derived
∆disaccharides analyzed in Figure 2 after treatment with, and
subsequent removal of mercuric ion, followed by derivatization with AMAC. ∆DiHA, ∆Di0S, ∆Di4S, ∆Di6S, and
∆Di4,6S yield exclusively AMAC-derivatized glcNAc,
galNAc, 4S-galNAc, 6S-galNAc, and 4,6S-galNAc, respectively, each of which migrates to a unique location relative to
their original derivatized ∆disaccharide.
∆Di2S, ∆Di2,4S, and ∆Di2,4,6S yielded quantitatively derivatized galNAc, 4S-galNAc, and 4,6S-galNAc, respectively,
as expected. However, in each case they also yielded a major
unknown band (X1), which migrates at a position behind derivatized 4S-galNAc, as well as two minor, more rapidly
migrating unknown bands (X2 and X3). The sum of the intensities of these unexpected bands in each case is nearly equal to
the intensity of the respective derivatized galNAc product.
Thus, they appear to be products from the 2S-substituted ∆hexuronic acid moiety that are stable to the mercuric ion treatment,
and that contain a free reducing group that reacts with the
fluorotag. In the absence of a 2S-substituted ∆hexuronic acid
only minor amounts of the X1 product are detected (see lanes
1, 2, and 8). As indicated in the analysis of the products from
∆Di2,6S (lane 7), this major new band (X1) comigrates with
derivatized 6S-galNAc.
Fluorescent derivatization and polyacrylamide gel
electrophoresis of the nonreducing terminal saturated
disaccharide structures derived from chondroitin sulfate
chains
The nonreducing ends of some chondroitin sulfate chains yield
saturated disaccharides after chondroitinase digestion, which
are not altered by mercuric ion treatment, Figure 1. Thus,
FACE analyses were used to identify the locations for derivatized Di0S, Di4S, and Di6S. Tetrasaccharides were prepared
from testicular hyaluronidase digests of chondroitin sulfate
isolated from Swarm rat chondrosarcoma aggrecan (almost
exclusively 4-sulfated) and from human adult cartilage
aggrecan (almost exclusively 6-sulfated) as described in Materials and methods. Since chondroitinase is an eliminase, digestion of these tetrasaccharides with this enzyme gives equal
quantities of both an unsaturated and a saturated disaccharide,
275
A.Calabro et al.
Fig. 6. FACE analyses of a Di4SDi4S tetrasaccharide prepared from testicular
hyaluronidase digests of chondroitin sulfate chains isolated from rat
chondrosarcoma aggrecan as described in Materials and methods. Samples
were derivatized directly with AMAC (lanes 1), chondroitinase ABC digested
then AMAC derivatized (lanes 2), or chondroitinase ABC digested followed by
mercuric ion treatment then AMAC derivatized (lanes 3). Analyses of the
major tetrasaccharide (Di4SDi4S) and minor tetrasaccharides (Di0SDi4S and
Di4SDi0S) are highlighted in (A) and (B), respectively. The positions of
nonreducing terminal structures are indicated (x). The standard lanes labeled
S8 contain a mixture of eight AMAC-derivatized ∆disaccharides from top to
bottom: ∆DiHA, ∆Di0S, ∆Di6S, ∆Di4S, ∆Di2S, ∆Di4,6S, ∆Di2,6S, and
∆Di2,4,6S.
Figure 1. FACE analyses are shown in Figures 6 and 7 for the
4-sulfated tetrasaccharide, Di4SDi4S, and the 6-sulfated
tetrasaccharide, Di6SDi6S, respectively. Analyses were made
on intact derivatized tetrasaccharides (lanes 1), and for the
derivatized products of chondroitinase digests of the tetrasaccharides either directly (lanes 2) or after treatment with
mercuric ion (lanes 3).
In Figure 6, the undigested 4-sulfated tetrasaccharide sample
shows a single dominant band representative of derivatized
Di4SDi4S (panel A, lane 1). After chondroitinase digestion,
the 4-sulfated tetrasaccharide gives a broad band composed of
two major, overlapping bands, one of which is the previously
identified ∆Di4S, and the other of which is Di4S, which
migrates slightly ahead (panel A, lane 2). The identities of
these two bands were confirmed after mercuric ion treatment
(panel A, lane 3). The saturated Di4S was resistant to the
mercuric ion treatment and therefore did not change mobility.
However, the mercuric ion treatment removed the ∆hexuronic
acid from the ∆Di4S shifting its elution position to that of 4SgalNAc.
In Figure 6, a minor tetrasaccharide band migrates slower
than the Di4SDi4S band. This band contains 2 tetrasaccharides, Di4SDi0S and Di0SDi4S (panel B, lane 1). The presence
of the Di4SDi0S is indicated by the production of identifiable
∆Di0S after chondroitinase digestion (panel B, lane 2) that
gives rise to galNAc after mercuric ion treatment (panel B,
lane 3). The presence of the Di0SDi4S is indicated by the
production of Di0S (panel B, lane 2) which is not altered by the
mercuric ion treatment (panel B, lane 3). A small amount of
∆DiHA is also present, indicating some contamination of the
276
Fig. 7. FACE analyses of a Di6SDi6S tetrasaccharide prepared from testicular
hyaluronidase digests of chondroitin sulfate chains isolated from human adult
cartilage aggrecan as described in Materials and methods. Samples were
derivatized directly with AMAC (lanes 1), chondroitinase ABC digested then
AMAC derivatized (lanes 2), or chondroitinase ABC digested followed by
mercuric ion treatment then AMAC derivatized (lanes 3). Analyses of the
major tetrasaccharide (Di6SDi6S) and minor tetrasaccharides (Di6SDi4S and
Di4SDi6S) are highlighted in (A) and (B), respectively. The positions of
nonreducing terminal structures are indicated (x).
original chondroitin sulfate sample with hyaluronan (panel B,
lane 2). In Figure 6, several additional minor peaks are also
apparent, indicating some heterogeneity in the size and in the
sulfation patterns of the undigested oligosaccharides. The presence of 4S-galNAc in the fluorotagged chondroitinase products (panel A, lane 2) indicates the presence of
oligosaccharides with nonreducing terminal 4S-galNAc in the
preparation. This indicates that some of the minor bands in the
preparation (panel A, lane 1) are trisaccharides and/or
pentasaccharides.
In Figure 7, the analysis of the 6-sulfated tetrasaccharide
reveals similar information. The undigested 6-sulfated tetrasaccharide sample shows a single dominant band representative of derivatized Di6SDi6S (panel A, lane 1). The ∆Di6S and
Di6S bands in the chondroitinase digest overlap, with the latter
migrating slightly ahead (panel A, lane 2). The former gives
rise to 6S-galNAc, while the latter is unaffected by the
mercuric ion treatment (panel A, lane 3). A minor tetrasaccharide band migrates just ahead of the Di6SDi6S band. This band
contains 2 tetrasaccharides, Di6SDi4S and Di4SDi6S (panel B,
lane 1). The presence of the Di6SDi4S is indicated by the
production of identifiable ∆Di4S after chondroitinase digestion (panel B, lane 2) that gives rise to 4S-galNAc after
mercuric ion treatment (panel B, lane 3). The presence of the
Di4SDi6S is indicated by the production of Di4S (panel B,
lane 2) which is not altered by the mercuric ion treatment
(panel B, lane 3). As with the Di4SDi4S tetrasaccharide in
Figure 6, several additional minor peaks are also apparent in
the Di6SDi6S preparation, indicating some heterogeneity in
the size and in the sulfation patterns of the undigested oligosaccharides. Further, the presence of 6S-galNAc in the fluoro-
Microanalysis of hyaluronan and chondroitin/dermatan sulfate
Fig. 8. FACE analyses of sulfated chondroitin sulfate ∆disaccharides digested
with or without chondro-4-sulfatase (4Sase) and/or chondro-6-sulfatase
(6Sase) prior to AMAC derivatization (lanes 1–8) or with mercuric ion
treatment (Hg2+) after sulfatase digestion, but prior to AMAC derivatization
(lanes 9–12). The standard lanes (S11) contain a mixture of eleven AMACderivatized saccharides from top to bottom: galNAc, glucose, ∆DiHA, ∆Di0S,
6S-galNAc, 4S-galNAc, ∆Di6S, ∆Di4S, ∆Di2S, ∆Di4,6S, and ∆Di2,4,6S.
tagged chondroitinase products (panel A, lane 2) indicates that
some of the minor bands are trisaccharides and/or pentasaccharides with nonreducing terminal 6S-galNAc. As predicted
from the chemistry (Figure 1), the intensities of the bands for
both derivatized tetrasaccharides, and their respective derivatized saturated and unsaturated disaccharides should be equivalent, and all are within 5% or less of each other.
Specificity of chondro-4-sulfatase and chondro-6-sulfatase for
sulfated ∆disaccharides and sulfated galNAc structures
The suitability of the sulfated ∆disaccharides, and the sulfated
galNAc structures to serve as substrates for chondro-4-sulfatase and chondro-6-sulfatase was tested. In all cases, both the
chondro-4-sulfatase and chondro-6-sulfatase showed only
specificity for sulfate at the 4- and 6-positions of the
hexosamine, respectively, with neither enzyme showing
specificity for sulfate at the 2-position of the hexuronic acid.
Importantly, 4S-galNAc, 6S-galNAc, and 4,6S-galNAc were
not substrates for either of the sulfatases (data not shown). The
results for only those ∆disaccharides that proved to be
substrates for the two sulfatases are shown in Figure 8. Digestion with the chondro-6-sulfatase alone specifically and quantitatively removed sulfate from the 6-position of ∆Di6S,
∆Di2,6S, and ∆Di4,6S yielding ∆Di0S, ∆Di2S, and ∆Di4S,
respectively (lanes 4 – 6). Digestion with the chondro-4-sulfatase alone specifically and quantitatively removed sulfate from
the 4-position of only ∆Di4S, yielding ∆Di0S (lane 1). The
chondro-4-sulfatase only partially removed the sulfates from
the 4-position of ∆Di2,4S and ∆Di4,6S, yielding minor
amounts of ∆Di2S, and ∆Di6S, respectively (lanes 2 and 3).
However, when the chondro-4-sulfatase and chondro-6-sulfatase were used together, the sulfates from both the 4- and 6positions were quantitatively removed from ∆Di4,6S, yielding
∆Di0S (lane 7).
Detection of the expected products after sulfatase digestion
of ∆Di2,4,6S required mercuric ion treatment prior to derivatization with AMAC, since after derivatization the expected
products, ∆Di2,4S and ∆Di2,6S, migrate at or near the electrophoresis front. The migration of derivatized ∆Di2,4,6S at the
electrophoresis front is shown in lane 8. The expected 4,6SgalNAc, X1, X2, and X3 products are seen after mercuric ion
treatment of ∆Di2,4,6S (lane 9). Digestion with the chondro-6sulfatase alone specifically and quantitatively removed sulfate
from the 6-position of ∆Di2,4,6S yielding ∆Di2,4S which,
after mercuric ion treatment, yields 4S-galNAc along with the
X1, X2, and X3 bands (lane 10). However, the chondro-4sulfatase was unable to remove the sulfate from the 4-position
of ∆Di2,4,6S as indicated by the unchanged, derivatized 4,6SgalNAc band (lane 11). Any minor amounts of the 6S-galNAc
resulting from chondro-4-sulfatase digestion of ∆Di2,4,6S to
yield ∆Di2,6S is undetectable since, after mercuric ion treatment, 6S-galNAc comigrates with the X1 band. When both the
chondro-4-sulfatase and chondro-6-sulfatase were used, the
sulfate from the 6-position was quantitatively removed from
∆Di2,4,6S, but the sulfate from the 4-position was only
partially removed yielding a mixture of ∆Di2S and ∆Di2,4S
which, after mercuric ion treatment, yields galNAc and 4SgalNAc, respectively, along with the X1, X2, and X3 bands
(lane 12). These results confirm that sulfation at either the 6position of the hexosamine or the 2-position of the hexuronic
acid inhibits the chondro-4-sulfatase activity. Higher enzyme
concentrations or longer incubation times did not change these
results (data not shown). The incomplete nature of the
mercuric ion treatment in lanes 9–12 is discussed below in
Important considerations.
Important considerations
This section describes important considerations for FACE
analysis for which the data are not shown. A major consideration throughout these analyses was pH. The chondroitinase
enzyme showed incomplete digestion of hyaluronan and unsulfated chondroitin sulfate when the pH was above 7. For this
reason 0.0005% phenol red was added to the sodium acetate or
ammonium acetate buffers at pH 7 to insure that the pH of
resuspended samples was at or below 7. This was done visually
based on the color of the phenol red indicator dye, which
appears yellow at a pH of 6.8 and red at a pH of 8.2. Resuspended samples which appeared pink to red were titrated back
to a pH of 6.8 to 7.0 by adding 1 µl aliquots (total of 10
maximum) of either a 0.1, 1, or 10 M acetic acid solution as
appropriate until the dye first turned yellow. The phenol red
was reduced to a colorless compound in the presence of
cyanoborohydride, and therefore was not detected in the FACE
gels. The amount of 0.0005% phenol red in 100 µl of solution
is ∼1 nmol, which is minor compared to the 50,000 nmol of
cyanoborohydride present in the derivatization reaction. Maintaining an acidic pH was also necessary because the unsaturated disaccharides produced by chondroitinase are unstable at
alkaline pH prior to derivatization. Also for this reason,
samples were always frozen on dry ice and then lyophilized,
rather than concentrated, to avoid a potential rise in pH as a
result of the acetate ion evaporating faster than the ammonium
ion. The major danger of low pH is desulfation, but even in the
presence of acetic acid at a pH of 3.5 (see below) there was
little evidence of desulfation using our protocols.
The pH in the final derivatization reaction was approximately
6.8 due to the buffer formed from the acetic acid and
cyanoborohydride. If while maintaining the acetic acid concentration, the cyanoborohydride concentration was reduced to
125 mM, a 10-fold molar excess of cyanoborohidride above
277
A.Calabro et al.
that of the AMAC, then the pH of the final derivatization solution was ∼3.5. Under these reaction conditions the derivatization proceeded to completion. However, if the acetic acid was
removed and the cyanoborohydride concentration was left at
1.25 M, then the resulting pH was ∼10.5 and no derivatized
products were detected. This was independent of the nature of
the oligosaccharide to be derivatized and not the result of alkaline degradation of unsaturated disaccharide structures, since
glucose showed similar results.
Previous methods have used derivatization reaction volumes
of as little as 10 µl (Jackson, 1991, 1994). This allows for a
greater percentage of the total sample to be loaded on the FACE
gel. However, we found a high degree of variability in samples
as a direct result of the difficulty in complete resuspension of
lyophilized oligosaccharides in such small volumes. Therefore,
we use a derivatization reaction volume of 80 µl, which was
determined to be the minimum volume required to access easily
the entire inner surface of a 1.5 ml microcentrifuge tube by
simple vortexing, and thus confidently resuspend all of a
lyophilized sample. If lower reaction volumes are required, then
the concentration of AMAC should be increased proportionally
to maintain a 10-fold molar excess above that of the oligosaccharides to be derivatized. We routinely load only 5 µl or 1/
20th of the 100 µl of the final derivatization reaction. At the salt
concentrations we use, higher loading volumes caused the
faster moving bands to narrow as they entered the gel. Larger
loading volumes of from 10 to 20 µl were possible, but require
that the salt concentration be proportionally lowered. This can
be accomplished without decreasing the reaction volume by
decreasing both the cyanoborohydride and acetic acid concentrations to the same extent so as to maintain the proper pH. The
cyanoborohydride can be decreased to 125 mM while still
maintaining a 10-fold molar excess above that of the AMAC.
Sodium acetate can be used in place of ammonium acetate
for the chondroitinase or sulfatase digestion steps and,
although it is not volatile, its presence after lyophilization does
not affect the derivatization reaction (see Figure 8, lanes 1–8).
However, it is important to use a volatile buffer such as ammonium acetate for those samples that are to be treated with
mercuric ion after chondroitinase digestion. This is because the
nonvolatile sodium acetate buffer, pH 7, remains behind after
lyophilization and thus increases the pH of the mercuric
acetate solution above the normal pH of 5 used for the
mercuric ion treatment. This results in incomplete removal of
the ∆hexuronic acid from ∆disaccharides and production of
unidentified products (see Figure 8, lanes 9–12). The mercuric
ion treatment is done prior to derivatization, and then the
mercuric ion is removed by treatment with Dowex H+ resin,
because the AMAC-derivatized ∆disaccharides are unstable to
mercuric ion. When 50 nmol aliquots of AMAC-derivatized
∆disaccharides were treated with mercuric ion at concentrations of from 0.15 to 35 mM for 30 min at room temperature
only the 4 mM mercuric ion concentration produced the
expected AMAC-derivatized hexosamine products. At lower
mercuric ion concentrations only the original AMAC-derivatized ∆disaccharides were seen indicating insufficient reagent,
while at higher mercuric ion concentrations no AMAC-derivatized material was detected indicating either loss of the AMAC
and/or destruction of the ∆disaccharide. This is in contrast to
mercuric ion treatment prior to derivatization, which is quantitative over a wide concentration range.
278
Discussion
We have developed fast, simple and sensitive protocols for
determining the fine structure of hyaluronan and chondroitin/
dermatan sulfate chains. These protocols take advantage of
bacterial eliminases, such as chondroitinase ABC or hyaluronidase SD, which are specific for chondroitin/dermatan sulfate
and/or hyaluronan. These enzymes generate specific oligosaccharide products with free reducing aldehydes of known chemistry that were previously blocked in the intact chains. These
newly generated reducing aldehydes are then quantitatively
derivatized with the fluor, 2-aminoacridone (AMAC), by a
Schiff’s base reaction which is stabilized by reduction with
cyanoborohydride. This provides for the same molar fluorescence per reducing aldehyde for each derivatized saccharide
independent of its chemistry, and for detection in the 1 pmol or
less range when imaged with our cooled CCD camera system
after separation by fluorophore-assisted carbohydrate electrophoresis (FACE).
Two additional chemistries are used in our protocols. Since
the bacterial enzymes are eliminases, they release internal
disaccharides with unsaturated hexuronic acid residues, while
any nonreducing terminal disaccharides contain saturated hexuronic acid residues (see Figure 1). This provides a means for
distinguishing these two structures, since only the internal
disaccharides with their unsaturated hexuronic acid are sensitive to mercuric acid treatment. In addition, chondro-4-sulfatase
and chondro-6-sulfatase, enzymes which selectively remove
sulfate esters from chondroitin/dermatan sulfate derived disaccharides (see Figure 8), can be used to confirm the sulfation
pattern of the digestion products analyzed by FACE.
Our standard protocols for FACE analysis of an unknown
sample involves a minimum of four steps. Step 1 involves
direct AMAC derivatization of the starting material. This identifies any structures with reducing groups that are present in the
starting material, whether expected, such as in the tetrasaccharide preparations shown in Figures 6 and 7, or unexpected,
such as glucose, which is often seen in samples obtained
directly from tissue extracts or culture medium (see Calabro et
al., 2000). Step 2 involves AMAC derivatization of the chondroitinase digestion products. This establishes the fine structure of the mass of the chondroitin sulfate chain (i.e., internal
∆disaccharide composition), and some nonreducing termini
that are clearly resolved (i.e., galNAc, 4S-galNAc, and 6SgalNAc). Step 3 involves AMAC derivatization of the chondroitinase digestion products after treatment with mercuric ion.
This unmasks the remaining nonreducing termini including
saturated disaccharides and 4,6S-galNAc, and helps confirm
the identities of the internal ∆disaccharides. Step 4 involves
AMAC derivatization of the chondroitinase digestion products
after treatment with chondro-4-sulfatase and/or chondro-6sulfatase. This confirms the identity of any sulfated disaccharides, and allows for resolution of comigrating bands as
described below. Together these four protocols allow for
complete fine structure analysis of chondroitin/dermatan
sulfate chains and for measurement of hyaluronan content. In
most cases, they provide for several independent measurements of each saccharide structure. Application of these protocols for the analysis of the hyaluronan and chondroitin sulfate
on aggrecan from cartilage is shown in an accompanying
article (Calabro et al., 2000).
Microanalysis of hyaluronan and chondroitin/dermatan sulfate
The specific chondroitinase digestion products which were
identified with these protocols are the unsaturated internal
∆disaccharides including ∆DiHA, ∆Di0S, ∆Di2S, ∆Di4S,
∆Di6S, ∆Di2,4S, ∆Di2,6S, ∆Di4,6S, and ∆Di2,4,6S (see
Figure 2), the saturated nonreducing terminal disaccharides
including DiHA (see Figure 4), Di0S, Di4S, and Di6S (see
Figures 6 and 7), and the nonreducing terminal hexosamines
including glcNAc, galNAc, 4S-galNAc, 6S-galNAc, and 4,6SgalNAc (see Figure 5). In addition, we have identified the relative position of AMAC-derivatized glucose (see Figure 8, lane
S11). Of these 19 AMAC-derivatized saccharides, all clearly
resolve by FACE with the exception of: (1) ∆Di0S from Di0S
and ∆Di4S from Di4S; (2) ∆Di6S from both Di6S and 4,6SgalNAc; (3) ∆Di2,4S, ∆Di2,6S, and ∆Di2,4,6S from each other
(The resolution of ∆Di2,6S from ∆Di2,4S and ∆Di2,4,6S at the
electrophoresis front was variable depending on the gel lot as
seen in the standards (lanes S8) in Figures 6 and 7); (4) ∆DiHA
from both glcNAc and DiHA; and (5) 6S-galNAc and the X1
product. The oligosaccharide composition of specific samples
will determine which of these problems must be addressed as
described below.
The lack of resolution of AMAC-derivatized ∆Di0S and
∆Di4S from AMAC-derivatized Di0S and Di4S, respectively,
is easily resolved by mercuric ion treatment (step 3 above).
Treatment with mercuric ion converts the typically abundant
internal ∆disaccharide structures, ∆Di0S and ∆Di4S, to
galNAc and 4S-galNAc, respectively, which both migrate at
different and unique positions in the FACE analysis. This
permits subsequent quantitation of any nonreducing terminal
Di0S and Di4S structures whose chemistries and therefore
relative mobilities are unaffected by mercuric ion treatment.
Since mercuric ion treatment contributes galNAc and 4SgalNAc from ∆Di0S and ∆Di4S, respectively, the amounts of
galNAc and 4S-galNAc originally present as nonreducing
terminal structures are measured in the original chondroitinase
digest (step 2 above).
The lack of resolution of AMAC-derivatized ∆Di6S from
AMAC-derivatized Di6S is complicated by the comigration of
AMAC-derivatized 4,6S-galNAc. Treatment with mercuric
ion (step 3 above) converts the typically abundant internal
∆disaccharide structure, ∆Di6S, to 6S-galNAc, which migrates
at a different and unique position in the FACE analysis. Since
mercuric ion treatment contributes 6S-galNAc from ∆Di6S,
the amount of 6S-galNAc originally present as a nonreducing
terminal structure is measured using the original chondroitinase digest (step 2 above). However, unlike for Di0S and Di4S
above, direct quantitation of Di6S in samples treated with
mercuric ion is only possible if no detectable 4,6S-galNAc is
present in the sample. Samples containing 4,6S-galNAc as
either a non-reducing terminal structure or as internal ∆Di4,6S,
which subsequently appears as 4,6S-galNAc after mercuric ion
treatment, require an additional calculation. The amount of
4,6S-galNAc resulting from mercuric ion treatment of ∆Di4,6S
is easily estimated from the ∆Di4,6S band in the chondroitinase digest (step 2 above), and this value subtracted. Generally, since the remaining Di6S and 4,6S-galNAc are both
nonreducing terminal structures, their lack of resolution is not
a major problem, because the estimation of number averaged
chain length for chondroitin sulfate chains is unaffected
(Calabro et al., 2000). However, if individual quantitation of
Di6S and 4,6S-galNAc is desired then the chondroitinase
digests (step 2 above) can be digested with both chondro-4sulfatase and chondro-6-sulfatase prior to derivatization to
convert all the saturated and unsaturated disaccharides to Di0S
and ∆Di0S, respectively. This permits the non-reducing
terminal 4,6S-galNAc to be measured directly (step 4 above),
and the amount of Di6S is then calculated by subtraction.
The lack of resolution of AMAC-derivatized ∆Di2,4S,
∆Di2,6S, and ∆Di2,4,6S which run at or near the electrophoresis front, is partially resolved by select sulfatase digestion
of the chondroitinase digest (step 4 above). For example, the
appearance of ∆Di2S following chondro-4-sulfatase digestion
alone or chondro-6-sulfatase digestion alone indicates the presence of ∆Di2,4S, or ∆Di2,6S, respectively, in the electrophoresis front (see Figure 8). However, due to the incomplete
nature of digestion of ∆Di2,4S and ∆Di2,4,6S with chondro-4sulfatase (see Figure 8), quantitation of only the ∆Di2,6S may
be possible. We are currently investigating different running
buffers, gel buffers, gel compositions, and electrophoresis
conditions in an effort to resolve our current problems with
resolution of these ∆disaccharides.
The lack of resolution of AMAC-derivatized ∆DiHA from
glcNAc and DiHA is primarily in the estimation of number
averaged chain length for hyaluronan chains. The internal
∆disaccharide from hyaluronan (∆DiHA), unlike those from
chondroitin sulfate, migrates close to its mercuric ion treatment
product, glcNAc, so that mercuric ion treatment (step 3 above)
does not unmask the nonreducing terminal DiHA structure.
Fortunately, quantitation of total hyaluronan is unaffected
since all three structures contribute to the total mass of
hyaluronan, and their comigration means they do not interfere
with structures related to chondroitin/dermatan sulfate.
The advantages of the FACE procedure over previous methodologies include sensitivity, speed and simplicity. As
mentioned above, AMAC gives the same molar fluorescence
value for every derivatized saccharide independent of chemistry. Figure 2 shows that the standard curves for four ∆disaccharides superimpose. In this case, the curves show linearity on
a log-log plot over the range from 6.25–100 pmol. With the
new, highly sensitive Quantix CCD camera, the standard curves
are linear and readily quantitated at one log lower concentrations, i.e., in the 0.5 pmol range. The FACE analyses in Figures
6 and 7, from digestion of the samples to electrophoresis, image
capture and quantitation, can now be completed in 2 days, and
require only a few micrograms of starting material. With the
exception of the mercuric ion treatment, which requires a
Dowex step to remove the reagent, all steps are carried out in a
single tube, and enzymes and reagents are not removed before
the electrophoresis. In comparison, the previous HPLC based
method required a week or more to complete, as derivatized
samples need to be purified from enzymes and reagents, and
each must be analyzed separately on the Dionex column.
Further, the FACE resolves products that are either lost or
poorly quantitated in the previous protocol, namely ∆Di0S,
Di0S, ∆DiHA, DiHA, and 6S-galNAc.
Materials and methods
2-Aminoacridone, HCl (AMAC) was purchased from Molecular Probes. Mercuric acetate, glacial acetic acid (99.99+%),
dimethylsulfoxide (DMSO, 99.9%), sodium cyanoborohydride
279
A.Calabro et al.
(95%), and glycerol (99.5%) were from Aldrich-Sigma.
Phenol red (0.5% w/v) was from Gibco. MONO® composition
gels (#60100) and MONO® gel running buffer (#70100) were
purchased from Glyko. Dowex AG50W-X8 (200–400 mesh)
was from Bio-Rad. Chondroitinase ABC, chondro-4-sulfatase,
chondro-6-sulfatase, and unsaturated hyaluronan and chondroitin sulfate disaccharide standards were purchased from
Seikagaku, America. Bovine testicular hyaluronidase (Type
VI-S) was from Aldrich-Sigma. High purity hyaluronan was
obtained from Pharmacia (Healon®).
Preparation of ∆disaccharide standards for fluorescent
derivatization
Standard ∆disaccharides (in ultrapure water) from Seikagaku,
including ∆DiHA from hyaluronan, and ∆Di0S, ∆Di2S,
∆Di4S, ∆Di6S, ∆Di2,4S, ∆Di2,6S, ∆Di4,6S, and ∆Di2,4,6S
from chondroitin/dermatan sulfate, were processed for derivatization as follows. Five identical aliquots of each ∆disaccharide were frozen on dry ice, and lyophilized until dry on a
vacuum concentrator. Each aliquot contained no more than
50 nmol of disaccharide reducing equivalents based on hexuronic acid (Blumenkrantz and Asboe-Hansen, 1973). One
aliquot of each ∆disaccharide was derivatized directly. The
second aliquot of each ∆disaccharide was resuspended in
100 µl of 17.5 mM mercuric acetate, 50 mM sodium acetate,
pH 5.0, and incubated 30 min at room temperature (Ludwigs et
al., 1987). The mercuric ion was removed by addition of 30 µl
of a 50% slurry of Dowex H+ resin, and the Dowex H+ resin
was removed by filtration through a glass wool plugged pipette
tip (Plaas et al., 1996). One hundred microliters of ultrapure
water was used to rinse the reaction tube and glass wool
plugged pipette tip. Sample volume trapped in the glass wool
plug was recovered by centrifugation at 2000 × g. The
mercuric ion treated samples were then frozen on dry ice, and
lyophilized until dry on a vacuum concentrator prior to derivatization.
The three remaining aliquots of each ∆disaccharide were
resuspended in 100 µl of 0.0005% phenol red, 100 mM sodium
acetate, pH 7.0. Each aliquot was digested for 1 h at 37°C with
chondro-4-sulfatase alone, chondro-6-sulfatase alone or
chondro-4-sulfatase and chondro-6-sulfatase together (100 mU
of each enzyme/ml). The three sulfatase digests of each ∆disaccharide were frozen on dry ice, and lyophilized until dry on
a vacuum concentrator. The sulfatase digests of all the ∆disaccharides, except ∆Di2,4,6S, were derivatized directly. The
three sulfatase digests of ∆Di2,4,6S were mercuric ion treated
as described above prior to derivatization.
Preparation of purified tetrasaccharides for fluorescent
derivatization
Tetrasaccharides from 68-year-old human aggrecan (predominantly 6-sulfated, Di6SDi6S) and rat chondrosarcoma
aggrecan (predominantly 4-sulfated, Di4SDi4S) were prepared
by testicular hyaluronidase digestion, MicroCon 10 filtration
and Toyopearl HW40S chromatography as previously
described (Plaas et al., 1996). The tetrasaccharide preparations
were generous gifts of Dr. Anna Plaas (Shriners Hospital for
Crippled Children, Tampa, FL). Tetrasaccharide preparations
(in ultrapure water) were processed for derivatization as
follows. Three identical aliquots from each preparation were
frozen on dry ice, and lyophilized until dry on a vacuum
280
concentrator. Each aliquot contained no more than 50 nmol of
potential disaccharide reducing equivalents. Each aliquot was
resuspended in 100 µl of 0.0005% phenol red, 100 mM ammonium acetate, pH 7.0. One aliquot was immediately frozen on
dry ice, and lyophilized until dry on a vacuum concentrator
prior to derivatization. The remaining two aliquots were
digested with chondroitinase ABC (100 mU/ml) for 3 h at
37°C, and then frozen on dry ice, and lyophilized until dry on
a vacuum concentrator. One chondroitinase digest was derivatized directly. The other chondroitinase digest was treated with
mercuric ion as described above prior to derivatization.
Preparation of hyaluronan ladders for fluorescent
derivatization
Hyaluronan ladders were generated from partial testicular
hyaluronidase digests as follows. A 1 mg/ml solution of high
purity hyaluronan in 0.0005% phenol red, 100 mM ammonium
acetate, pH 7.0 was made from Healon® (10 mg/ml), and the
concentration confirmed by hexuronic acid assay (Blumenkrantz and Asboe-Hansen, 1973). Four 100 µl aliquots
containing 100 µg of hyaluronan were each digested for 4 h at
37°C with 37, 111, 333, or 1000 U/ml of bovine testicular
hyaluronidase. The hyaluronidase digests were then immediately frozen on dry ice, and lyophilized until dry on a vacuum
concentrator prior to derivatization.
Fluorescent derivatization with 2-aminoacridone,
fluorophore-assisted carbohydrate electrophoresis, gel
imaging, and data analysis
All samples were derivatized by addition of 40 µl of 12.5 mM
AMAC (500 nmol) in 85% DMSO/15% acetic acid followed
by incubation for 15 min at room temperature. Then 40 µl of
1.25 M sodium cyanoborohydride (50,000 nmol) in ultrapure
water was added followed by incubation for 16 h at 37°C
(Jackson, 1991, 1994). After derivatization, 20 µl of glycerol
(20% final concentration) was added to each sample prior to
electrophoresis. If necessary, samples were diluted prior to
electrophoresis with a solution containing 0.0005% phenol red,
6% acetic acid, 20% glycerol, 34% DMSO, and 480 mM
sodium cyanoborohydride. All derivatized samples were
stored in the dark at -70°C. Some samples formed a precipitate
during storage, which was dissolved by heating the samples to
60°C for 5–10 min.
MONO® composition gels were stored at 4°C. MONO® gel
running buffer was initially dissolved in ultrapure water at
room temperature for 1 h, and then stored at 4°C overnight
with mixing just prior to use. The wells of each gel were rinsed
extensively with running buffer at 4°C, and the glass plates of
each gel were thoroughly cleaned with ultrapure water just
prior to use. The assembled electrophoresis apparatus from
Glyko containing the electrophoresis buffer and one or two
gels was placed in a large tank, and packed in ice to equilibrate
the buffer to 4°C or less at the start of electrophoresis. All eight
lanes of a gel were loaded simultaneously with 5 µl aliquots of
sample using a Hamilton 8-channel glass syringe. The samples
were electrophoresed for 80 min at a constant 500 V with a
starting current of ~25 mA/gel, and a final current of ~10 mA/
gel. The final temperature never exceeded 10°C. Prior to
loading samples, the gels were briefly electrophoresed to
ensure the correct starting current. After electrophoresis, one
gel at a time was removed from the apparatus for imaging with
Microanalysis of hyaluronan and chondroitin/dermatan sulfate
the second gel (if any) left in the cool tank buffer to minimize
diffusion of bands. The glass plates of each gel were thoroughly cleaned with ultrapure water, and the area around the
wells covered with dark tape to mask the intense fluorescence
from the unreacted AMAC, which remains in the wells during
electrophoresis. The gels were imaged while still in their glass
support plates.
The gels were illuminated with UV light (365 nm) from an
Ultra Lum Transilluminator, and imaged with a Quantix cooled
CCD camera from Roper Scientific/Photometrics. Camera
specifications included a research grade Kodak CCD chip, a 5
mega pixel/s transfer rate, a 1317×1035 pixel resolution, and
6.8 µm2 pixels at a 12 bit depth. The camera was fitted with an
11.5–69 mm f1.4 manual zoom lens, a 4× diopter, and an
ethidium bromide orange barrier filter. The images were
analyzed using the Gel-Pro Analyzer® program version 3.0
from Media Cybernetics. The digital images shown in the
Results section depict over saturated pixel intensity for the
major derivatized structures in order to allow visualization of
less abundant derivatized structures. Quantitation was done
with images having all pixels within a linear 12-bit depth range.
Acknowledgments
We thank John Coletta, Dr. Jane Grande, John Mako and Stacy
Stephenson (Department of Biomedical Engineering), Dr.
Preenie Senanayake (Eye Institute), and Carol del la Motte
(Colorectal Surgery), all from the Cleveland Clinic Foundation, Cleveland, OH, for their contributions to this work. This
work was supported in part by a Mizutani Foundation for
Glycoscience grant, NIH Grant HD34831, and funds from the
Lerner Research Institute, Cleveland Clinic Foundation.
Abbreviations
glcA, glucuronic acid; glcNAc, N-acetylglucosamine; galNAc,
N-acetylgalactosamine; 4S-galNAc, N-acetylgalactosamine-4sulfate; 6S-galNAc, N-acetylgalactosamine-6-sulfate; 4,6SgalNAc, N-acetylgalactosamine-4,6-di-sulfate; DiHA or glcAβ1,3-glcNAc, 2-acetamido-2-deoxy-3-O-(β-D-glucopyranosyluronic acid)-D-glucose; Di0S or glcA-β1,3-galNAc, 2acetamido-2-deoxy-3-O-(β-D-glucopyranosyluronic acid)-Dgalactose; Di4S or glcA-β1,3–4S-galNAc, 2-acetamido-2deoxy-3-O-(β-D-glucopyranosyluronic
acid)-4-O-sulfo-Dgalactose; Di6S or glcA-β1,3–6S-galNAc, 2-acetamido-2acid)-6-O-sulfo-Ddeoxy-3-O-(β-D-glucopyranosyluronic
galactose; ∆DiHA or ∆glcA-β1,3-glcNAc, 2-acetamido-2deoxy-3-O-(β-D-gluco-4-enepyranosyluronic acid)-D-glucose;
∆Di0S or ∆glcA-β1,3-galNAc, 2-acetamido-2-deoxy-3-O-(βD-gluco-4-enepyranosyluronic acid)-D-galactose; ∆Di2S or
2-acetamido-2-deoxy-3-O-(2-O2S-∆glcA-β1,3-galNAc,
acid)-D-galactose;
sulfo-β-D-gluco-4-enepyranosyluronic
∆Di4S or ∆glcA-β1,3–4S-galNAc, 2-acetamido-2-deoxy-3-Oacid)-4-O-sulfo-D-galac(β-D-gluco-4-enepyranosyluronic
tose; ∆Di6S or ∆glcA-β1,3–6S-galNAc, 2-acetamido-2deoxy-3-O-(β-D-gluco-4-enepyranosyluronic acid)-6-O-sulfoD-galactose; ∆Di2,4S or 2S-∆glcA-β1,3–4S-galNAc or ∆DiB,
2-acetamido-2-deoxy-3-O-(2-O-sulfo-β-D-gluco-4-enepyranosyluronic acid)-4-O-sulfo-D-galactose; ∆Di2,6S or 2S-
∆glcA-β1,3–6S-galNAc or ∆DiD, 2-acetamido-2-deoxy-3-O(2-O-sulfo-β-D-gluco-4-enepyranosyluronic acid)-6-O-sulfoD-galactose; ∆Di4,6S or ∆glcA-β1,3–4,6S-galNAc or ∆DiE, 2acetamido-2-deoxy-3-O-(β-D-gluco-4-enepyranosyluronic
acid)-4,6-di-O-sulfo-D-galactose; ∆Di2,4,6S or 2S-∆glcAβ1,3–4,6S-galNAc or ∆DiTriS, 2-acetamido-2-deoxy-3-O-(2O-sulfo-β-D-gluco-4-enepyranosyluronic
acid)-4,6-di-Osulfo-D-galactose; AMAC, 2-aminoacridone; DMSO, dimethylsulfoxide; FACE, fluorophore-assisted carbohydrate electrophoresis; CCD, charge-coupled device; HA2, DiHA.
References
Blumenkrantz,N. and Asboe-Hansen,G. (1973) New methods for quantitative
determination of uronic acids. Anal. Biochem., 54, 484–489.
Calabro,A., Hascall,V.C. and Midura,R.J. (2000) Adaptation of FACE methodology for microanalysis of total hyaluronan and chondroitin sulfate composition from cartilage. Glycobiology, 10, 283–293.
Carney,S.L. and Osborne,D.J. (1991) The separation of chondroitin sulfate
disaccharides and hyaluronan oligosaccharides by capillary zone electrophoresis. Anal. Biochem., 195, 132–140.
Deutsch,A.J., Midura,R.J. and Plaas,A.H. (1995) Stucture of chondroitin sulfate on aggrecan isolated from bovine tibial and costochondral growth
plates. J. Orthop. Res., 13, 230–239.
Hascall,V.C., Midura,R.J., Sorrell,J.M. and Plaas,A.H. (1995) Immunology of
chondroitin/dermatan sulfate. Adv. Exp. Med. Biol., 376, 205–216.
Jackson,P. (1991) Polyacrylamide gel electrophoresis of reducing saccharides
labeled with the fluorophore 2-aminoacridone: subpicomolar detection
using an imaging system based on a cooled charge-coupled device. Anal.
Biochem., 196, 238–244.
Jackson,P. (1994) High-resolution polyacrylamide gel electrophoresis of fluorophore-labeled reducing saccharides. Methods Enzymol., 230, 250–265.
Karamanos,N.K., Syrokou,A., Vanky,P., Nurminen,M. and Hjerpe,A. (1994)
Determination of 24 variously sulfated galactosaminoglycan- and hyaluronan-derived disaccharides by high-performance liquid chromatography.
Anal. Biochem., 221, 189–199.
Kitagawa,H., Kinoshita,A. and Sugahara,K. (1995) Microanalysis of glycosaminoglycan-derived disaccharides labeled with the fluorophore 2-aminoacridone by capillary zone electrophoresis and high-performance liquid
chromatography. Anal. Biochem., 232, 114–121.
Ludwigs,U., Elgavish,A., Esko,J.D., Meezan,E. and Rodén,L. (1987) Reaction
of unsaturated uronic acid residues with mercuric salts. Cleavage of the
hyaluronic acid disaccharide 2-acetamido-2-deoxy-3-O- (β-D-gluco-4enepyranosyluronic acid)-D-glucose. Biochem. J., 45, 795–804.
Midura,R.J., Salustri,A., Calabro,A., Yanagishita,M. and Hascall,V.C. (1994)
High-resolution separation of disaccharide and oligosaccharide alditols
from chondroitin sulphate, dermatan sulphate and hyaluronan using CarboPac PA1 chromatography. Glycobiology, 4, 333–342.
Midura,R.J., Calabro,A., Yanagishita,M. and Hascall,V.C. (1995) Nonreducing end structures of chondroitin sulfate chains on aggrecan isolated from
Swarm rat chondrosarcoma cultures. J. Biol. Chem., 270, 8009–8015.
Otsu,K., Inoue,H., Tsuzuki,Y., Yonekura,H., Nakanishi,Y. and Suzuki,S.
(1985) A distinct terminal structure in newly synthesized chondroitin sulphate chains. Biochem. J., 227, 37–48.
Plaas,A.H., Hascall,V.C. and Midura,R.J. (1996) Ion exchange HPLC microanalysis of chondroitin sulfate: quantitative derivatization of chondroitin
lyase digestion products with 2-aminopyridine. Glycobiology, 6, 823–829.
Plaas,A.H., Wong-Palms,S., Roughley,P.J., Midura,R.J. and Hascall,V.C.
(1997) Chemical and immunological assay of the nonreducing terminal
residues of chondroitin sulfate from human aggrecan. J. Biol. Chem., 272,
20603–20610.
Shibata,S., Midura,R.J. and Hascall,V.C. (1992) Structural analysis of the
linkage region oligosaccharides and unsaturated disaccharides from chondroitin sulfate using CarboPac PA1. J. Biol. Chem., 267, 6548–6555.
Sugahara,K., Yamashina,I., De Waard,P., Van Halbeek,H. and Vliegenhart,J.F. (1988) Structural studies on sulfated glycopeptides from the carbohydrate-protein linkage region of chondroitin-4-sulfate proteoglycans of
Swarm rat chondrosarcoma. Demonstration of the structure gal (4-O-sulfate)β1,3galβ1,4xylβ1-O-ser. J. Biol. Chem., 263, 10168–10174.
281
A.Calabro et al.
282