Analytical Biochemistry 293, 53–59 (2001)

doi:10.1006/abio.2001.5068, available online at http://www.idealibrary.com on

An Improved Assay for the N-Acetyl-D-glucosamine

Reducing Ends of Polysaccharides
in the Presence of Proteins

Trias Asteriou,* Brigitte Deschrevel,* Bertrand Delpech,† Philippe Bertrand,†

Florence Bultelle,* Cyrille Merai,* and Jean-Claude Vincent*
IFR “Systèmes Intégrés,” No. 61 CNRS, Université de Rouen, 76821 Mont-Saint-Aignan cedex, France; *Laboratoire
“Polymères, Biopolymères, Membranes,” UMR 6522 Université de Rouen, CNRS, 76821 Mont-Saint-Aignan cedex, France;
and †Laboratoire d’Oncologie Moléculaire, Centre Henri Becquerel, Rue d’Amiens, 76038 Rouen, France

Received November 3, 2000; published online April 23, 2001

Hyaluronan (HA) 1 is a linear negatively charged

Hyaluronan concentration and hyaluronidase activ-
ity can be assayed by using different techniques in-
cluding turbidimetry, viscosimetry, ELISA, chroma-
tography, and colorimetry. The most popular
colorimetric method is that of J. Reissig et al. (1955,
J. Biol. Chem. 217, 959 –966), in which the color results
from a reaction between the Ehrlich’s reagent (DMAB)
and the N-acetyl-D-glucosamine reducing end of each
hyaluronan chain. Nevertheless, there are problems
with this method when proteins are present in the
medium. Here we propose a new interpretation of the
Reissig signal for estimating such reducing ends in
media containing enzymes or other proteins. This in-
terpretation is based on the fact that the absorbance
obtained by using the Reissig method results from two
factors: a turbidity due to the formation of polysaccha-
ride–protein complexes and a color resulting from the
action of DMAB on the reducing end of the polysac-
charide chains. The turbidity at 585 nm, the wave-
length at which the color intensity is maximal, may be
estimated by curve fitting the spectrum between 450
and 650 nm. Subtracting the turbidity from the absor-
bance gives the colorimetric intensity which repre-
sents the concentration of polysaccharide chains.
Moreover, the turbidity may give additional informa-
tion about the existence of polysaccharide–protein
complexes and their nature. © 2001 Academic Press
Key Words: Hyaluronan assay; hyaluronidase kinet-
ics; N-acetyl-glucosamine reducing end assay; Reissig
assay; hyaluronan; hyaluronidase; presence of pro-
teins; colorimetry; turbidity; polysaccharide–protein
complexes.
polysaccharide composed of D-glucuronic acid-␤(1,3)-N-
acetyl-D-glucosamine disaccharide units linked
through ␤(1,4) bonds. HA is present in both pro-
karyotes and eukaryotes and is one of the main com-
ponents of the extracellular matrix (ECM) of higher
animals. Its size ranges from a few hundred thousand
to 10 million Daltons according to its origin. HA is the
main substrate of hydrolytic enzymes, called hyalu-
ronidases (HAases). Both lysosomal and testicular
HAases (EC 3.2.1.35) hydrolyze ␤(1,4) glycosidic bonds,
producing HA fragments with a N-acetyl-D-glu-
cosamine at the reducing extremity (1).
Several techniques for assaying HAase activity have
been described (2, 3), including turbidimetry (4 – 6),
capillary viscosimetry (7, 8), ELISA-like assays (9, 10),
HPLC-SEC chromatography (11–13), and colorimetric
assays (12, 14 –18). One of the most frequently used
colorimetric assays is based on the measure of the
concentration of reducing ends appearing during hy-
drolysis. In the Morgan-Elson assay (17), one of the
first used, the N-acetyl-D-glucosamine reducing end is
successively transformed into chromogens I and II un-
der alkaline conditions at 100°C and then into chromo-
gen III by the action of concentrated hydrochloric and
sulfuric acids, and finally chromogen III reacts with
p-dimethyaminobenzaldehyde (DMAB) (Ehrlich’s re-
agent) to give a red colored product which can be de-
tected at 585 nm (19). Reissig et al. (18) have improved
this assay by adding borate in the first step to increase
1
Abbreviations used: HA, hyaluronan; ECM, extracellular matrix;
HAase, hyaluronidase; DMAB, p-dimethylaminobenzaldehyde; BSA,
bovine serum albumin.

0003-2697/01 $35.00 53
Copyright © 2001 by Academic Press
All rights of reproduction in any form reserved.
54 ASTERIOU ET AL.
the production of chromogens and thus to enhance the
signal at 585 nm.
The method of Reissig et al. has been mainly used to
investigate HAase activity. The classical procedure
consists in incubating a solution of HA and HAase at a
constant temperature and measuring the concentra-
tion of reducing ends, via the absorbance at 585 nm,
after a given incubation time (20). This procedure has
been used, for example, to study HAase activity as a
function of pH or ionic strength (10, 20 –24). However,
a non-zero absorbance at the first time point following
addition of HAase has been reported in some cases by
those performing kinetic measurements (19).
The method of Reissig et al. is particularly suitable
for studying the kinetics and the mechanism of the
HA–HAase reaction because it gives the rate of glyco-
sidic bond cleavage directly. We have used it to obtain
the entire time course of this reaction in order to in-
vestigate the influence of parameters, such as pH, ionic
strength, enzyme, and substrate concentrations, on the
rate of HA hydrolysis catalyzed by HAase. Certain HA
and HAase concentrations resulted in very complex
time courses of absorbance at 585 nm, which could not
be described by classical kinetic equations. Moreover,
the absorbance at time zero depended on the concen-
tration of both HA and HAase. These observations may
be explained by the hypothesis in which absorbance in
the Reissig et al. assay is affected by HAase acting not
only as an enzyme but also as a protein binding non-
specifically to HA (14, 20).
Gacesa et al. (20) have adapted the method of Reissig
et al. to use the assay in the presence of serum or
proteins: ethylacetate was added at the final stage of
the assay and the sample was centrifuged to eliminate
precipitated proteins. However, an additional step is
needed which increases the duration and the complex-
ity of the assay. This step may also introduce errors in
estimating the concentration of the reducing ends if
some HA is complexed to proteins and so discarded
with them. Here we propose a different assay for
the determination of N-acetyl-D-glucosamine reducing
ends in the presence of proteins that is also based on
the method of Reissig et al. but that does not risk losing
any HA. Indeed, our assay permits a new interpreta-
tion of the spectrophotometric signal of the Reissig et
al. method taking the influence of proteins into ac-
count. This interpretation also gives information about
the nature of the polysaccharide–protein complexes.
EXPERIMENTAL PROCEDURE
1. Materials
Bovine testicular HAase with a specific activity of
990 units per mg was obtained from Sigma (H 3884, lot
76H8025). Sodium hyaluronate (HA) from bovine tra-
chea was purchased from Fluka (366047/1) and con-
tained less than 3% protein. Bovine serum albumin
(BSA) was purchased from Sigma (A 3675). HA, BSA,
and HAase were used without further purification.
The Reissig method needs two solutions: (i) the bo-
rate solution prepared by dissolving 4.94 g boric acid
(Sigma B 7660) and 1.98 g potassium hydroxide in 100
mL Milli-Q water (Waters) and (ii) the 0.1 g mL ⫺1
DMAB solution prepared by dissolving 5 g DMAB
(Sigma D 8904) in 6.25 mL hydrochloric acid, 12 N
(Sigma H 7020), made up to a final volume of 50 mL
with glacial acetic acid (Sigma A 6283). The latter
solution was diluted 10-fold with glacial acetic acid just
before use (and at least 15 min before use).
Absorbance was measured with an Uvikon 860
KONTRON spectrophotometer that is equipped with a
temperature-controlled chamber and connected to a
PC. Adjustments of pH were carried out using a
Metrohm 632 pH meter equipped with a Tacussel
XC111 pH electrode.
2. Methods
Measurement of HAase reaction rate was based on
the Reissig method (18), which determines the concen-
tration of reducing ␤-N-acetyl-D-glucosamine ends gen-
erated from HA hydrolysis; N-acetyl-D-glucosamine
(Sigma A 8625) was used as a standard. The HA solu-
tion (1 mg mL ⫺1), containing 5 mM ammonium acetate,
was placed in a reactor, adjusted to the chosen pH with
either acetic acid or ammoniac, stirred, and main-
tained at 37°C. The reaction was started by adding
concentrated HAase (10 mg mL ⫺1 in ammonium ace-
tate, 5 mM, pH 5). At each time point, a 200-␮L aliquot
of the reaction mixture was removed from the reactor
and added to 50 ␮L of borate solution in a glass tube.
The solution in the tube was immediately vortexed,
heated in a boiling water bath for exactly 3 min, and
then placed in a cold water bath at approximately 10°C
until 10 aliquots had been treated. Then, 1.5 mL of the
diluted DMAB solution was added to each of these
tubes, vortexed, and placed at 37°C for exactly 15 min.
This was transferred to a plastic cuvette of 1 cm path-
length, immediately scanned against water between
400 and 700 nm and the spectrum saved on the com-
puter (absorbances at 450, 585, and 650 nm were at
least required). Absorbance was linearized between
450 and 650 nm in order to estimate turbidity at 585
nm. Turbidity was then subtracted from the total ab-
sorbance at 585 nm to give a value proportional to the
N-acetyl-D-glucosamine reducing end concentration.
RESULTS
1. Kinetics of the HA Hydrolysis Catalyzed by HAase
A series of hydrolysis experiments was performed
using different concentrations of HA and/or HAase.
 ASSAY FOR N-ACETYL-GLUCOSAMINE REDUCING ENDS
FIG. 1. Kinetics of the HA (1 mg mL ⫺1) hydrolysis catalyzed by
HAase at pH 6.3 and 37°C in 5 mM ammonium acetate: 200-␮L
aliquots of the reaction medium are treated by using the method of
Reissig at adequate intervals of time. The absorbance at 585 nm is
plotted against time for two HAase concentrations: (■) 0.5 mg mL ⫺1
and (F) 2 mg mL ⫺1.
Aliquots were analyzed throughout the reaction using
the method of Reissig et al. and the first measure was
taken immediately after the addition of HAase (time
zero). The absorbance at 585 nm was then plotted
against time. In many cases, the evolution of the ab-
sorbance at 585 nm is complex (Fig. 1) and does not
correspond to any classical equation of kinetics. At a
low HAase concentration (0.5 mg mL ⫺1), the absor-
bance increases continuously (presumably as reducing
ends are produced). However, at a high HAase concen-
tration (2 mg mL ⫺1), the absorbance decreases during
the first 10 min and then increases continuously. The
relationship between HAase concentration and initial
absorbance was then determined (Fig. 2). Initial absor-
bance depended strongly on the HAase concentration
and was maximal at 2 mg mL ⫺1. Hence, absorbance at
time zero is a function of not only the initial HA con-
centration but also the initial HAase concentration.
2. The Reissig Method in the Presence of Proteins
To investigate whether the differences in initial ab-
sorbance were due to a possible nonspecific action of
HAase, HAase was replaced by BSA, a protein that
does not catalyze HA hydrolysis. A series of solutions
with various concentrations of HA and BSA were ana-
lyzed using the Reissig method. The results (Fig. 3)
show that, as expected, the signal at 585 nm increased
as a function of HA concentration in the absence of
BSA. The signal also increased with HA concentration
55
provided that the BSA concentration was constant.
However, the signal also increased at a constant HA
level when the BSA concentration was increased. This
means that the Reissig signal at 585 nm did not only
depend on the concentration of reducing ends. To in-
vestigate whether the effect of BSA was due to it in-
teracting with the reducing ends of the HA chains, the
absorbance of N-acetyl-D-glucosamine in the presence
of BSA was studied. This revealed that increasing con-
centrations of N-acetyl-D-glucosamine gave a Reissig
signal that was not perturbed by the presence of BSA
(data not shown). Hence the simultaneous presence of
proteins and long chains of HA were required for the
increased Reissig signal. In fact, it has long been
known that BSA added to long HA chains under acidic
conditions results in an increased turbidity that can be
used to assay HA (3, 6, 25, 26). The conclusion of these
authors was that this turbidity reflects the formation of
HA–protein complexes. Moreover, the effect of the
presence of other proteins on absorbance at 585 nm is
known to perturb HA/HAase assays based on the
Reissig method (26 –29). Formation of HA–protein
complexes would explain these perturbations under
the acidic conditions used in the Reissig method and
would explain our results with HAase, particularly at
time zero. In other words, part of the Reissig signal we
obtain can be attributed to the turbidity of HA–protein
complexes.
FIG. 2. Dependence of the Reissig signal as a function of HAase
concentration. During the kinetics of the HA (1 mg mL ⫺1) hydrolysis
catalyzed by HAase at pH 6.3 and 37°C in 5 mM ammonium acetate,
200-␮L aliquots of the reaction medium are treated by using the
method of Reissig at adequate intervals of time. The initial absor-
bance at 585 nm is plotted against the HAase concentration.
56 ASTERIOU ET AL.

FIG. 3. Dependence of the Reissig signal as a function of both HA and BSA concentrations: 200-␮L aliquots of HA–BSA mixtures are
treated by using the method of Reissig and the absorbance at 585 nm is plotted against both HA and BSA concentrations.

3. A New Analysis of the Reissig Signal

Instead of physically eliminating the contribution of
turbidity to the Reissig signal, we propose a new vari-
ant of the method that gives additional information
about the interactions between HAase and HA. In the
Reissig reaction, the absorbance due to turbidity is
easy to distinguish from that due to the color change as
the dye reacts with the reducing ends. In kinetic ex-
periments of HA/HAase interactions, samples taken at
successive times were treated by the Reissig method
and recordings were taken over the 400- to 700-nm
range instead of the usual 585 nm. The superposition
of these spectra (Fig. 4) clearly shows that (i) the ex-
pected double peak of the Reissig reaction with two
maxima at 540 and 585 nm progressively appears, (ii)
at 585 nm, the absorbance does not increase continu-
ously but again decreases during an initial period, and
(iii) the absorbance measured outside this double peak
range does not remain constant. In fact, this behavior FIG. 4. Kinetics of the HA (1 mg mL ⫺1) hydrolysis catalyzed by
can be explained by a continuous increase in the color- HAase (2 mg mL ⫺1): 200-␮L aliquots of the reaction medium are
imetric signal together with a variable turbidity. Each treated by using the method of Reissig at adequate intervals of time.
400- to 700-nm spectra are plotted as a function of time and clearly
scan of the sample between 400 and 700 nm gives a show two different behaviors: turbidity spectra predominant at the
figure for absorbance that corresponds to the sum of beginning of the reaction and color spectra predominant during the
the colorimetric intensity (essentially a double peak) second period.
 ASSAY FOR N-ACETYL-GLUCOSAMINE REDUCING ENDS 57

FIG. 5. Estimation of the turbidity and color contributions of the FIG. 6. Estimation of the turbidity and color contributions of the
Reissig signal at 585 nm by using a linear interpolation of the Reissig signal at 585 nm by using a curvilinear interpolation of the
turbidity between 450 and 650 nm. turbidity between 450 and 650 nm.

and the turbidity (which decreases as the wavelength 4. Complementary Information Available
increases). from the Reissig Assay
Our variant of the Reissig assay is to measure the The turbidity at 585 nm results from the ability of
relative levels of both color and turbidity at 585 nm. the polysaccharide chains to form insoluble polysac-
We propose two ways to achieve this: charide–protein complexes under the acidic conditions
of the Reissig assay. In other words, the turbidity does
—A rapid procedure in which absorbancies at the
three wavelengths 450, 585, and 650 nm are measured.
Since the absorbance measured at 450 and 650 nm is
due to turbidity alone, the turbidity at 585 nm can be
estimated by assuming a linear relationship between
absorbance and wavelength in the 450- to 650-nm in-
terval (Fig. 5). This procedure can be readily performed
on a standard spectrophotometer.
—A more sophisticated approach is to estimate tur-
bidity precisely over the 450- to 650-nm range by fit-
ting a hyperbola to the two ends of the spectrum ob-
tained experimentally between 400 and 450 nm and
between 650 and 700 nm. This procedure requires a
computer-associated spectrophotometer to allow curve
fitting (Fig. 6).
Subtracting turbidity (as estimated above) from the
total absorbance at 585 nm gives the colorimetric in-
tensity and hence the concentration of the reducing
ends, that is, the number of polysaccharide chains per
volume unit. The time course of the colorimetric inten-
sity at 585 nm now corresponds to the kinetics of the
enzyme-catalyzed hydrolysis reaction (Fig. 7) and ex- FIG. 7. Kinetics of the HA (1 mg mL ⫺1) hydrolysis catalyzed by
trapolation to time zero gives a low value which is now HAase (2 mg mL ⫺1) at pH 7.0: 200-␮L aliquots of the reaction me-
dium are treated by using the method of Reissig at adequate inter-
independent of the HAase concentration. The slope of vals of time. The color part of the absorbance at 585 nm is plotted
the curve (colorimetric intensity against time) at time against time: the absorbance at time zero is very close to zero and its
zero is now equal to the true initial reaction rate. time course corresponds to a more classical kinetic shape.
58 ASTERIOU ET AL.
FIG. 8. Kinetics of the HA (5 mg mL ⫺1) hydrolysis catalyzed by
HAase (2 mg mL ⫺1) at pH 4.0 and in presence of NaNO 3, 0.15 M:
40-␮L aliquots of the reaction medium are treated by using the
method of Reissig at adequate intervals of time. The color part of the
absorbance at 585 nm is plotted against time: the absorbance at time
zero is close to zero and its time course corresponds to a classical
kinetic shape.
not correspond to HA–protein complexes formed under
the particular conditions of pH, ionic strength, etc., of
the HA–HAase reaction system studied here but cor-
responds to complexes subsequently formed under the
acidic pH conditions of the Reissig assay. Turbidity is
an indicator of the presence of long HA polysaccharides
while colorimetric intensity is an indicator of the aver-
age polysaccharide chain length. Together, turbidity
and color thus give an idea of the distribution of HA
chain lengths. The complexes revealed by the Reissig
assay probably reflect nonspecific protein–polysaccha-
ride interactions and are therefore different from en-
zyme–substrate complexes (which, in any case, would
be dissociated under the conditions of the Reissig as-
say). During HA hydrolysis, the reaction rate must
depend on the local conditions of the microenviron-
ment. When these conditions result in a rapid reaction,
the HA chains are cut and very quickly transformed
into short chains unable to form insoluble complexes
under acidic conditions: turbidity is thus very low
throughout the reaction (Fig. 8). However, when the
reaction is slow, the chain length remains high for a
few minutes and complexes are easily formed: the ini-
tial turbidity is thus high only to decrease progres-
sively as hydrolysis of the HA chains causes dissocia-
tion of the complexes (Fig. 7). Turbidity should
disappear when chains are too short to form complexes
and the relationship between turbidity and hydrolysis
of HA therefore provides information about the distri-
bution of the lengths of the polysaccharide chains.
DISCUSSION AND CONCLUSION
Many authors who have used the method of Reissig
et al. have mentioned problems related to the presence
of proteins and their ability to form insoluble com-
plexes with long HA chains. Experimental solutions to
this problem have been proposed. Some of those work-
ing with HA in presence of serum or proteins have
introduced a step to solubilize the complex under alka-
line conditions (20) or a step involving centrifugation of
the sample (19) in order to avoid proteins interfering
with the assay. Despite these modifications, extrapola-
tions to time zero were found to be impossible. More-
over, errors in the calculation of the numbers of reduc-
ing ends may result from changes in the color of the
DMAB complex or its stability due to either alkaline
conditions (20) or addition of ethylacetate (19), while
errors may also result from the elimination of material
in the centrifugation step (19) since reducing ends of
HA molecules involved in complexes are eliminated
from the supernatant.
Here, we propose a new interpretation of the Reissig
assay that may be particularly useful when the reduc-
ing ends of polysaccharide are estimated in media con-
taining proteins or enzymes. This interpretation is
based on the fact that the absorbance given by the
Reissig assay results from, first, a turbidity due to the
formation of polysaccharide–protein complexes and,
second, the color resulting from the reaction of DMAB
on the reducing end of the polysaccharide chains. Tur-
bidity at 585 nm, the wavelength at which the color is
most intense, can be estimated by curve fitting. The
difference between the absorbance and turbidity at this
wavelength gives the colorimetric intensity that actu-
ally corresponds to the concentration of the reducing
ends. Linearization, the simplest way to estimate the
turbidity gives a satisfactory result in 98% of the cases.
Our approach has the advantages that it neither re-
moves material from the sample nor requires addi-
tional treatments that may alter the spectrum of the
final colored product.
In our analysis, the time course of the colorimetric
intensity reflects the changing concentration of poly-
saccharide chains. The initial reaction rate can then be
calculated by using the entire time course to extrapo-
late to time zero and hence to study its variations
under given experimental conditions. Influences of pH,
ionic strength, substrate concentration, and enzyme
concentration may thus be studied by using this
method over a large range of parameters. Moreover,
the estimates of turbidity, afforded by our analysis,
may give useful information about the nature of po-
 ASSAY FOR N-ACETYL-GLUCOSAMINE REDUCING ENDS
lysaccharide–protein interactions and the existence of
polysaccharide–protein complexes.
Physico-chemical methods such as viscosimetry or
turbidimetry may be used to measure the hydrolysis of
HA catalyzed by HAase indirectly. However, the
method of Reissig et al. is particularly suitable for
studying this catalysis since it gives a direct measure
of the concentration of reducing ends and hence the
quantity of hydrolyzed glycosidic bonds, on condition
that the signal is interpreted properly, especially when
proteins are present in the medium.
ACKNOWLEDGMENTS
We acknowledge Dr. V. Norris for fruitful discussions and lan-
guage suggestions. We are grateful to the “Conseil Régional de Haute
Normandie” for fellowship accorded to T. Astériou. We also acknowl-
edge F. Gouley and S. Munch for their participation to the experi-
mental work.
REFERENCES
1. Meyer, K. (1971) in The Enzymes (Boyer, P. D., Eds.), 3rd ed.,
Vol. 5, pp. 307–320, Academic Press, New York.
2. Herd, J. K., Tschida, J., and Motycka, L. (1974) The detection of
hyaluronidase on electrophoresis membranes. Anal. Biochem.
61, 133–143.
3. Hynes, W. L., and Ferretti, J. J. (1994) Assays for hyaluronidase
activity. Methods Enzymol. 235, 606 – 616.
4. Dorfman, A., and Ott, M. (1948) A turbidimetric method for the
assay of hyaluronidase. J. Biol. Chem. 172, 367–375.
5. Tolksdorf, S., McCready, M. H., McCullagh, D. R., Schwenk, E.,
and Bloomfield, N. J. (1949) The turbidimetric assay of hyaluron-
idase. J. Lab. Clin. Med. 34, 74 – 89.
6. Ibrahim, A. N., and Streitfeld, M. M. (1973) The microassay of
hyaluronic acid concentration and hyaluronidase activity by cap-
illary turbidity (CT) and capillary turbidity reduction (CTR)
tests. Anal. Biochem. 56, 428 – 434.
7. Dorfman, A. (1948) The kinetics of the enzymatic hydrolysis of
hyaluronic acid. J. Biol. Chem. 172, 377–387.
8. Vercruysse, K. P., Lauwers, A. R., and Demeester, J. M. (1995)
Absolute and empirical determination of the enzymatic activity
and kinetic investigation of the action of hyaluronidase on hya-
luronan using viscosimetry. Biochem. J. 306, 153–160.
9. Maeda, H., Fujita, H., Sakura, Y., Miyazaki, K., and Goto, M.
(1999) A competitive enzyme-linked immunosorbent assay-like
method for the measurement of urinary hyaluronan. Biosci. Bio-
technol. Biochem. 63, 892– 895.
10. Delpech, B., Bertrand, P., and Chauzy, C. (1987) An indirect
enzymoimmunological assay for hyaluronidase. J. Immunol.
Methods 104, 223–229.
11. Cramer, J. A., and Bailey, L. C. (1991) A reversed-phase ion-pair
high-performance liquid chromatography method for bovine tes-
ticular hyaluronidase digests using postcolumn derivatization
with 2-cyanoacetamide and ultraviolet detection. Anal. Biochem.
196, 183–191.
12. Turner, R. E., and Cowman, M. K. (1985) Cationic dye binding by
hyaluronate fragments: Dependence on hyaluronate chain
length. Arch. Biochem. Biophys. 237, 253–260.
59
13. Vercruysse, K. P., Lauwers, A. R., and Demeester, J. M. (1994)
Kinetic investigation of the degradation of hyaluronan by hyal-
uronidase using gel permeation chromatography. J. Chro-
matogr. B 656, 179 –190.
14. Vercruysse, K. P., Lauwers, A. R., and Demeester, J. M. (1995)
Kinetic investigation of the action of hyaluronidase on hyaluro-
nan using the Morgan-Elson and neocuproine assays. Biochem.
J. 310, 55–59.
15. Benchetrit, L. C., Pahuja, S. L., Gray, E. D., and Edstrom, R. D.
(1977) A sensitive method for the assay of hyaluronidase activ-
ity. Anal. Biochem. 79, 431– 437.
16. Homer, K. A., Denbow, L., and Beighton, D. (1993) Spectropho-
tometric method for the assay of glycosaminoglycans and glycos-
aminoglycan-depolymerizing enzymes. Anal. Biochem. 214, 435–
441.
17. Elson, L. A., and Morgan, W. (1933) A colorimetric method for
the determination of glucosamine and chondrosamine. Biochem.
J. 27, 1824 –1828.
18. Reissig, J., Strominger, J., and Leloir, L. (1955) A modified
colorimetric method for the estimation of N-acetylamino sugars.
J. Biol. Chem. 217, 959 –966.
19. Muckenschnabel, I., Bernhardt, G., Spruss, T., Dietl, B., and
Buschauer, A. (1998) Quantitation of hyaluronidases by the Mor-
gan-Elson reaction: Comparison of the enzyme activities in the
plasma of tumor patients and healthy volunteers. Cancer Lett.
131, 13–20.
20. Gacesa, P., Savitsky, M. J., Dodgson, K. S., and Olavesen, A. H.
(1981) A recommended procedure for the estimation of bovine
testicular hyaluronidase in the presence of human serum. Anal.
Biochem. 118, 76 – 84.
21. De Salegui, M., Plonska, H., and Pigman, W. (1967) A compari-
son of serum and testicular hyaluronidase. Arch. Biochem. Bio-
phys. 121, 548 –554.
22. Gacesa, P., Savitsky, M. J., Dodgson, K. S., and Olavesen, A. H.
(1979) Effect of ionic strength and serum on the activity profile of
bovine testicular hyaluronidase [proceedings]. Biochem. Soc.
Trans. 7, 1287–1289.
23. Gorham, S. D., Olavesen, A. H., and Dodgson, K. S. (1975) Effect
of ionic strength and pH on the properties of purified bovine
testicular hyaluronidase. Connect. Tissue Res. 3, 17–25.
24. Maingonnat, C., Victor, R., Bertrand, P., Courel, M. N., Mau-
noury, R., and Delpech, B. (1999) Activation and inhibition of
human cancer cell hyaluronidase by proteins. Anal. Biochem.
268, 30 –34.
25. Pigman, W., Gramling, E., and Holley, H. (1961) Interactions of
hyaluronic acid with serum albumin. Biochim. Biophys. Acta 46,
100 –107.
26. Niedermeier, W., and Gramling, E. A. (1969) The Pi complex. A
review of hyaluronate-protein complexes. Ala. J. Med. Sci. 6,
239 –246.
27. Grinberg, V. Y., and Tolstoguzov, V. B. (1997) Thermodynamic
incompatibility of proteins and polysaccharides in solutions.
Food Hydrocolloids 11, 145–158.
28. Laurent, T. C. (1977) Interaction between proteins and glycos-
aminoglycans. Fed. Proc. 36, 24 –27.
29. Schmitt, C., Sanchez, C., Desobry-Banon, S., and Hardy, J.
(1998) Structure and technofunctional properties of protein–
polysaccharide complexes: A review. Crit. Rev. Food Sci. Nutr.
38, 689 –753.