RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO.
5, 2010 1515
FOOD COMPOSITION AND ADDITIVES
Methodology for Determining Degree of Hydrolysis of Proteins in
Hydrolysates: A Review
SHANE M. RUTHERFURD
Riddet Institute, Massey University, Palmerston North, New Zealand
Degree of hydrolysis (DH) is defined as the
proportion of cleaved peptide bonds in a protein
hydrolysate. Several methods exist for determining
DH; the most commonly used of these include the
pH-stat, trinitrobenzenesulfonic acid (TNBS),
o-phthaldialdehyde (OPA), trichloroacetic acid
soluble nitrogen (SN-TCA), and formol titration
methods. The pH-stat method is based on the
number of protons released during hydrolysis; the
TNBS, OPA, and formol titration methods are based
on the measurement of amino groups generated
from hydrolysis. The SN-TCA method measures the
amount of TCA-soluble nitrogen, rather than DH.
The pH-stat is the simplest and most commonly
used method, but does not determine peptide
bonds directly. In addition, the accuracy of the
method depends on the type of hydrolytic enzymes
used, the size of the hydrolyzed peptides, and the
reaction temperature. Generally, the TNBS and OPA
methods compare well and do directly determine
DH. However, the assumption that the response
factor for all derivatized N-terminal amino acids is
similar may lead to inaccuracies. In conclusion,
there is no consensus as to the best method for
determining the DH of protein hydrolysates;
consequently, there is a need for a standardized
approach if interstudy comparisons are to be made.
echnological developments in food processing and
T advancements in separation technology have resulted in
the widespread use of hydrolysates of dietary proteins
and their fractions in various food and feed products due to
their desirable nutritional and functional properties. Protein
hydrolysates have been frequently used as the main source of
protein in supplements, sole source foods (infant formulas and
meal replacers), sports nutrition products, and medical foods,
such as products for parenteral administration.
Hydrolysates have been used to improve the digestion
and absorption of poorly digestible protein sources, but
more recently have also been cited as a potent source of
bioactive peptides, exhibiting antihypertensive, antioxidative,
Received June 10, 2010. Accepted by SG June 14, 2010.
Corresponding author’s e-mail: S.M.Rutherfurd@massey.ac.nz
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
hypocholesterolemic, antithrombotic, immunomodulatory,
and antimicrobial effects (1). Hydrolysates are mostly
produced from enzymatic and/or acid hydrolyses and consist
of complex mixtures of free amino acids and peptides of
different chain length. The degree to which a protein source
has been hydrolyzed is a reflection of the number of peptide
bonds broken and, therefore, the average size of the peptides
present. The number of peptide bonds broken as a proportion
of the total number of peptide bonds present is defined as the
degree of hydrolysis (DH). Several techniques are currently
used to determine the DH of food hydrolysates. This review
aims to describe and discuss the suitability of those
techniques and finally draw conclusions in relation to the
need for a standardized analytical approach.
Determining Degree of Hydrolysis
The DH is defined as the proportion of the total number of
peptide bonds that are cleaved during hydrolysis (2) and is
calculated as follows:
DH, % = h/htot ´ 100
where h is the number of hydrolyzed peptide bonds and htot is
the total number of peptide bonds present (3). There is no
standard method for determining DH. Instead, a number of
methods have been developed and are commonly used to
determine the DH of protein hydrolysates. These include the
pH-stat, osmometric, soluble nitrogen after trichloroacetic
acid precipitation (SN-TCA), 2,4,6-trinitrobenzenesulfonic
acid (TNBS), o-phthaldialdehyde (OPA), amino acid
nitrogen, and formol titration methods. Although some of
these methods have been reviewed (4–6), this review aims to
extend that discussion to cover all the commonly used
methods for the determination of the DH of protein
hydrolysates.
pH-Stat
The pH-stat method was first developed by Jacobsen et
al. (7) and is arguably the most commonly used method. In
recent times it has been used to determine the DH in protein
hydrolysates from shrimp (8), soy protein (9, 10), whey (11),
wheat gluten (12), trevally (13), grass carp (14), rapeseed (15),
yellowfin tuna waste (16, 17), bovine serum albumin (18),
hemp (19), atlantic salmon muscle (20), porcine plasma (21),
and egg yolk (22). Márquez and Vázquez (23) also used the
1516 RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO. 5, 2010
pH-stat method to examine the kinetics of hydrolysis of a
hemoglobin-Alcalase hydrolysis model.
During peptide bond hydrolysis a carboxyl and an amino
group are liberated. At neutral or alkaline pH, carboxyl groups
are completely deionized and proton exchange occurs
between the carboxyl group and the amino group (23). In
contrast, at alkali pH, the amino groups will also be either
partially or totally deionized depending on the pH of the
reaction medium and the amino acid in question, as the pK of
the N-terminal amino group of free amino acids ranges from 9
to 10.8. The resulting free protons cause a decrease in the pH
of the reaction mixture, and the addition of base is required to
maintain the pH. The amount of base required has a direct
relationship to the number of peptide bonds hydrolyzed, and
can be used to estimate the DH. The latter relationship has
been described by Adler-Nissen (2) in the equation given
below:
DH = B ´ NB/(Mp ´ a ´ htot) ´ 100
where B is the base consumption, NB is the normality of the
base, Mp is the amount of the protein (which is often
determined as the nitrogen content multiplied by 6.25), htot is
the number of peptide bonds in the protein, and a is the
average degree of dissociation of the a-amino groups which
was in turn described as:
a = 10(pH–pK)/(1 + 10(pH–pK))
where pK is the average pK of the a-amino groups liberated
during hydrolysis.
The main advantage of the method is that it is very
straightforward. As part of the digestion process, the pH of the
reaction mixture must be adjusted to maintain it at the
optimum for the protease enzymes, and the amount of base
used to maintain the pH can simply be recorded, particularly if
an automated pH-stat is used. The pH-stat method eliminates
the need for additional steps to determine the DH, such as the
derivatization steps required for the TNBS, OPA, or
ninhydrin-based methods, or the precipitation and
centrifugation steps required for the SN-TCA method. This
allows for the real-time monitoring of the DH, which can be a
valuable tool if a predefined DH for a digestion process is
required.
Unfortunately, the relationship between the DH and base
consumption is not simple and depends on a number of
variables, including pK of the liberated a-amino group,
reaction mixture temperature, and peptide chain
length (24, 25). The a value is estimated based on the average
pK value for the liberated a-amino groups, yet the actual
N-terminal residues in the hydrolysate peptides are not known
and depend on the amino acid sequence and the specificity of
the protease enzymes used. Dzwolak and Ziajka (26)
suggested that the effect of temperature on pK results from
changes in the enthalpy of ionization of the amino groups.
Camacho et al. (27) conducted a thorough investigation of the
correlation between base consumption and the DH. They
reported that the correct pK value of any a-amino acid group
depends on the pH of the reaction mixture and the protein
being hydrolyzed. This is further complicated because the pK
of any given a-amino acid will be influenced by the nature of
the adjacent amino acids which will differ across hydrolyzed
peptides.
Another potential inaccuracy for determining DH using the
pH-stat, or any of the other methods, is the estimation of the
total number of peptide bonds. Several methods have been
used to estimate this, including the use of nitrogen multiplied
by 6.25 (28) for hemoglobin (23) and salmon muscle (20),
or 5.7 for wheat gluten (12). Others (8, 9, 15, 19) have used
prederived general values based on the work of
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
Adler-Nissen (2). Karamaƒ et al. (29) used the TNBS method
to determine the free amino groups after 6 M HCl hydrolysis
of pea proteins, and Silva et al. (30) used amino acid analysis
and the summation of the individual amino acids to estimate
the total number of peptide bonds. The latter approach, which
is based on amino acid analysis, is more accurate, but is slower
and more laborious; it needs to be conducted before
hydrolysis if methods like pH-stat are to be used for the
real-time monitoring of DH during hydrolysis.
The main drawback of extrapolating the number of peptide
bonds from the total protein content (N ´ 6.25) is that the
average amino acid MW must be used to calculate the number
of amino acid residues, and the latter value will depend on the
amino acid composition which differs between proteins. This
is reflected in the peptide bond content of protein reported by
Adler-Nissen (2), who suggested a range of 7.8–8.6 meqv/g of
protein.
Another drawback of the pH-stat method is that it tends to
underestimate the DH for digests that contain exopeptidases.
Exopeptidases digest protein and peptides from the end of the
polypeptide chain, resulting in a greater amount of amino
acids and di- and tri-peptides. The latter peptides have higher
pK values than larger peptides, which in turn leads to an
overestimation of a and a concomitant underestimation of
DH (2). The pH-stat method may also be unsuitable for
hydrolysis systems with multiple enzyme steps that require
pH changes between steps, or if it is undesirable to have an
alkali endproduct (3).
Osmometric Method
Osmometry is based on the relationship between the
number of hydrolyzed peptide bonds and the osmolality of the
reaction mixture, and is calculated using the following
equation (26):
DH = [DC/(S% ´ fosm) ´ (1/w) ´ (1/htot) ´ 100
where DC = osmolality (mOsm/kg), S% = the substrate
concentration, fosm = the conversion factor calculated based on
the dry matter content (%) in the substrate (DM): fosm =
1000/(100 – DM), htot = total number of peptide bonds (mM/g
protein), and w = the osmotic coefficient for peptides (26).
Dzwolak and Ziajka (26) described a strategy whereby the
pH-stat method was used to determine the DH for alkali
solutions and the osmometric method was used for acidic
 RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO. 5, 2010 1517
Figure 1. Reaction of o-phthaldialdehyde (OPA) with
amino acids.
solutions. These workers also compared the TNBS method
with the pH-stat/osmometric method when applied to casein
and whey protein concentrate (WPC) after digestion with
either trypsin/chymotrypsin (pH 7.65 and 8.0) using the
pH-stat method or pepsin using the osmometric method. They
reported a linear relationship between the TNBS method and
both the pH-stat and osmometric methods for determining the
amount of free amino groups present in the hydrolysate.
However, while the slope of the line for the osmometric
method (osmometric coefficient) approximated unity (0.95),
this was not the case for the pH-stat method where the slope
ranged from 1.6 to 2.5, depending on the substrate protein and
reaction mixture pH. Similarly, Chirife et al. (31) reported that
the osmometric coefficient of a-amino acids was slightly
lower than unity. The relatively wide range in slopes for the
pH-stat method may reflect (1) the degree of dissociation of the
amino groups (above pH 5 all the carboxyl groups will be
dissociated), which is, in turn, related to the reaction mixture
pH, and (2) the inaccuracy of the a value which, as mentioned
above, may reflect the fact that the estimated average pK of the
liberated amino groups may not accurately reflect the actual
average pK.
OPA
OPA has been used for the determination of amino acids for
many years (32) and is a well-known derivatizing agent for
amino groups (Figure 1). The reaction, which requires the
presence of a thiol group, can be conducted in a matter of
minutes at room temperature. The resulting OPA-amino acid
derivative is strongly fluorescent (l excitation, 350 nm,
l emission, 450 nm), while the unreacted OPA does not
fluoresce. OPA does not react with proline and reacts poorly
with cysteine. This method has been used to determine the DH
in protein hydrolysates of sheep milk whey (33), casein (34),
fish protein (35), and cheese (36). The OPA method has also
been used to calibrate the base consumption during hydrolysis
to allow the direct determination of DH using the pH-stat
method (37).
A range of different thiol reagents have been used for the
OPA-amino acid reaction, including N-acetyl-L-cysteine
(24, 33), ethandiol (38), dithiothreitol (3), and
N,N-dimethyl-2-mercaptoethyl ammonium chloride (34).
Frister et al. (38) proposed that ethanediol derivatives were
more stable than the traditional mercaptoethanol derivatives.
In turn, Hernández et al. (39) suggested that
N-acetyl-L-cysteine leads to derivatives that are more stable
than those generated by the use of either mercaptoethanol or
ethanediol.
One advantage of the OPA method is that the
derivatization is rapid and permits the real-time monitoring of
protein hydrolysis. However, given the poor reactivity with
proline and cysteine, OPA may not be suitable for
determining the DH of proteins containing high levels of
these amino acids. Furthermore, the OPA method may not
derivatize insoluble peptides or proteins, and this may lead to
an underestimation of the DH of such material. In contrast,
OPA may overestimate DH since it can react with the amino
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
side chain of lysine. Another disadvantage of the OPA
method is that only one amino acid, often leucine (24, 34) or
serine (3), is used to determine the response factor for
OPA-peptide or OPA-amino acid derivatives. Although this
assumes that the response factor for OPA-leucine or
OPA-serine is representative of all the other OPA-amino
acid/peptide derivatives present, the validity of this
assumption can be challenged.
SN-TCA Method
This method has recently been used to estimate the DH of
pumpkin protein isolate (40), fish offal (41, 42), beans (43),
casein (44), and WPC (45). The method does not determine
the DH directly, as it does not determine the number of
peptide bonds broken. Instead, it measures the TCA-soluble
nitrogen, which is assumed to consist only of amino acids and
small peptides. This method is also known as the
TCA-solubility index or nonprotein nitrogen method, yet
some workers (40, 41) describe the TCA-solubility index as
DH, and this may be somewhat misleading. The SN-TCA
method uses between 3.6 and 13.6% TCA to precipitate the
unhydrolyzed protein that may be present. The precipitated
proteins are centrifuged, and the nitrogen content of the
supernatant and original material are then determined. The
TCA-solubility index, or NPN, is then calculated using the
following equation:
TCA-solubility index, % =
TCA-soluble N/total N in the sample ´ 100
Various methods have been used to determine the total
nitrogen content of protein sources, including the
Kjeldahl (42, 43, 45) and Lowry methods (40, 46). The
accuracy of the latter methods directly affects the accuracy of
the DH determinations.
The main advantage of the SN-TCA method is that it is
simple to perform. However, there are several disadvantages,
including the assumption that TCA-soluble nitrogen is
derived only from small peptides and amino acids, which
may not be correct since not all intact proteins are
precipitated by TCA. Moughan et al. (47) showed that as
much as 70% of proteins greater than 10 kDa were not
precipitated from ileal digesta using TCA. Based on their
work and the work of others it would appear that the
presence of glycoproteins is problematic and may cause an
1518 RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO. 5, 2010
Figure 2. Reaction of 2,4,6-trinitrobenzenesulfonic
acid (TNBS) with amino acids.
overestimation of the DH when the SN-TCA method is used.
Some workers (45, 48) have termed TCA-soluble nitrogen
as the TCA solubility index rather than DH, and this may be a
more appropriate term since in essence this method does not
measure peptide bond breakage.
TNBS Method
In recent years, this method has been used by a number of
workers (10, 49–56) and is based on the reaction of TNBS
with N-terminal amino groups. It was first used to determine
peptide content by Satake et al. (57). The reaction, which
takes only a few minutes at room temperature, converts the
free N-terminal amino acids to trinitrophenyl-amino acid
derivatives, which are then determined using absorbance at
340 nm (Figure 2). The total number of N-terminal amino
groups is then estimated based on either the N content of the
original protein source (50, 53) or the amino acid composition
after hydrolysis in 6 M HCl at 110°C for 24 h (49, 52, 56),
with synthetic leucine used as a representative amino acid to
determine the amount of trinitrophenyl-amino acid
derivatives present. The reaction is conducted in slightly
alkaline conditions and stopped by lowering the pH (58).
Adler-Nissen (58) pointed out several problems with the
TNBS method that contributed to poor repeatability,
including the fact that TNBS reacts slowly with hydroxyl ions
in solution which can cause baseline drift (59), the presence of
insoluble proteinaceous material, which can be problematic
for spectrophotometric analysis, and the relationship between
the color density and the concentration of amino groups,
which is incorrectly assumed to be linear (58).
Adler-Nissen (58) proposed the use of sodium dodecyl sulfate
to solubilize proteins and peptides, and the use of borate rather
than bicarbonate buffer to remove the possibility of carbon
dioxide outgasing. These measures, along with using a 1 h
reaction time, resulted in greater accuracy and repeatability.
Despite these measures, Adler-Nissen (58) noted that some
samples were still difficult to solubilize, particularly those
with interprotein disulfide bonds. Unfortunately, reducing
agents that contain sulfhydryl groups, such as
mercaptoethanol, which can assist solubilization by breaking
disulfide bonds, cannot be used with TNBS as they
themselves react with the TNBS reagent (60).
The main advantages of the TNBS method are that it
directly determines the free N-terminal amino groups in a
hydrolysate and does not rely on a complex relationship to
determine these groups, as does the pH-stat method. In
Figure 3. Titration reaction between amino acids and
formaldehyde in the presence of potassium hydroxide.
contrast, the disadvantages of the TNBS methods are that a
derivatization step must be carried out in addition to the
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
enzymatic hydrolysis reaction, and, as with the OPA method,
TNBS does not react with proline. Furthermore, it is
assumed that the response factor for single amino acids, such
as leucine (58) or glycine (10), which are used as a standard
to quantify all trinitrophenyl (TNP)-amino acids, is
representative of all the other TNP-peptide or TNP-amino
acid derivatives present. An additional disadvantage is that the
e-amino groups of lysine will also react with TNBS, leading to
an overestimation of the DH.
Formol Titration Method
The formol titration method based on the Sörensen
method (5) is described in detail by Taylor (61). This method
has recently been used to determine the DH of fish proteins or
byproducts (62–65), oilseed proteins (66), whey protein (67),
and hemoglobin (48). The mechanics of the reaction are as
follows: At neutral or alkali pH, formaldehyde reacts with the
amino group of an amino acid liberating a proton and
lowering the pKa of the amino acid-formaldehyde product. If
the pH of the reaction mixture is lower than the pKa of the
amino acid, the protons can be titrated with a base giving a
direct measurement of the amount of free amino groups
present (Figure 3).
The two approaches for carrying out the formol titration
method have been termed the direct and indirect methods (61),
but both involve the titration of amino acids with
formaldehyde in the presence of an alkali. Essentially, for the
direct method, formaldehyde is added directly to the test
solution which is then titrated with an alkali to the appropriate
end point depending on the indicator being used. For the
indirect method, the test solution is adjusted to a preselected
pH, and formaldehyde (adjusted to the final pH) is added. The
solution is then titrated to the final pH. Taylor (61) compared
both methods, and reported that the direct method was
superior because the indirect method was not reproducible
and had the propensity to underestimate amino acid N. Many
workers reporting the use of the formol titration method are
not clear as to which method (direct or indirect) they use.
Other Methods
The reaction between ninhydrin and amino groups has also
been used to determine DH and the a-amino N content of
hydrolysates (68–71), although this method is not commonly
used. Other less common methods include using the area
under the peak of hydrolyzed and unhydrolyzed proteins run
 RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO. 5, 2010 1519
Table 1. Advantages and disadvantages of the most common methods for determining DH
Method Advantages
pH-stat Real-time monitoring
Nondenaturing method
Rapid method Relationship between base consumption and DH is complex and may not be accurate for all proteins
No derivatization
TNBS Generally accurate
Derivatization is rapid
TNP-leucine response factor is assumed to represent all amino acids
OPA Derivatization is rapid
Real-time monitoring
SN-TCA Rapid method
Formol titration Rapid method
Real-time monitoring
on a reversed-phase HPLC (72, 73) to calculate the degree of
hydrolysis. However, this method does not determine the DH
as defined by Adler-Nissen (2), but rather the disappearance of
the intact proteins. Infrared spectroscopy has also been used to
determine the DH of a hemoglobin hydrolysate (74).
As can be seen from the discussion above, no single method
is currently being used to determine the DH of protein
hydrolysates. Furthermore, each method has its own
advantages and disadvantages (Table 1). If the commonly used
DH methods generate the same DH values for a wide range of
hydrolysates, there is no need to standardize DH methodology.
In contrast, if data generated from the commonly used DH
methods do not correlate well among methods, there is a need
for a thorough evaluation of the current methods and
standardization of the protocol before a robust comparison of
the extent of hydrolysis for a hydrolysate prepared in different
ways can be made.
Comparison of DH Methodologies
Although most workers tend to use only one of the available
methods for determining DH, a number of reports compare
different methods. Some studies have been conducted with the
specific intent to compare DH methods while for others the
latter comparison was a secondary aim.
Spellman et al. (24) compared the pH-stat, OPA, and
TNBS methods and determined the DH of a WPC digested
with either Debitrase (HYW20) or Alcalase. They found that
after Debitrase digestion, the DH differed across methods by as
much as 60%, depending on which method was being
compared. The pH-stat method gave the lowest value and the
Disadvantages
Accuracy depends on the type of enzyme used
Exogenous proteases lead to an underestimate of DH
Can only be used for alkali reaction mixtures
No real-time monitoring
Only works for soluble material
Toxic compound
Interference from lysine side chains
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
Requires derivatization step
Inaccurate for proline-rich proteins
Inaccurate for cysteine/proline-rich proteins
Only works for soluble material
Interference from lysine side chains
Requires derivatization step
Determines TCA-soluble nitrogen rather than DH
Variable results
Indirect and direct methods do not give different results
TNBS method gave the highest. Spellman et al. (24)
concluded that the low pH-stat value was due to the high
exopeptidase content of the Debitrase preparation, resulting
in a high amount of amino acids and small peptides present.
The difference between the DH values obtained using either
the OPA or TNBS methods was about 15%. To ascertain
which of the DH methods (OPA or TNBS) was inaccurate,
Spellman et al. (24) redigested the WPC with Alcalase, which
does not contain exopeptidases and for which the DH should
be accurate when determined by the pH-stat method. In this
latter study, the TNBS and pH-stat methods agreed well, and
the OPA underestimated the DH by 13%. Spellman et al. (24)
showed that the extinction coefficient for OPA-cysteine was
only 20% of the mean extinction coefficients of the other
OPA-amino acid derivatives. In contrast, the extinction
coefficient for TNP-cysteine was similar to that observed for
most of the other amino acids. They concluded that the
underestimation in DH observed with OPA was due to the
poor reactivity with cysteine in the cysteine-rich WPC.
However, WPC contains about 3% cysteine on a molar
basis (75); therefore, all the cysteine residues would have to
be present as either free amino acids or N-terminal amino
acids to account for the 13% difference between the OPA and
TNBS methods—a situation that would appear unlikely.
Panasiuk et al. (70) used the TNBS, OPA, and ninhydrin
methods to determine the a-amino N in a pea protein
hydrolysate. They reported a good correlation between the
TNBS and OPA methods, but reported that the ninhydrin
method gave much lower (approximately half) values when
compared to the other two methods. In contrast, in a
1520 RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO. 5, 2010
Figure 4. DH determined using the pH-stat (9),
SN-TCA (), and formol titration () methods applied to
hemoglobin digested with 1% Esperase followed by
either 0, 0.5, 1, or 2% Flavourzyme [data taken from In et
al. (48)].
comparison of the TNBS and ninhydrin methods applied to
defatted cheese samples, Clegg et al. (68) found that the
ninhydrin method gave higher a-amino N values than the
TNBS method, but recommended the TNBS method because
it was a simpler procedure.
Cheison et al. (50) used both the pH-stat and TNBS
methods to monitor the hydrolysis of whey proteins in a
tangential-flow filter membrane reactor. They found that
when the DH of the reaction mixture was low (approximately
20%), the DH determined by the pH-stat method was only
slightly lower than that determined by the TNBS method. This
finding was consistent with the studies of Spellman et al. (24)
when Alcalase was used, but not when Debitrase was used.
However, the DH determined in the final hydrolysate was
much higher when determined by the pH-stat method
(60–75%) than by the TNBS method (27–45%). Cheison et
al. (50) concluded that the TNBS values were likely to be
more accurate than those of the pH-stat method when applied
to the final product. They based this conclusion on data
obtained from the MW profile and from the free amino acid
composition of the final hydrolysate, but no explanation for
this discrepancy was given. It has been suggested that whey
protein peptides have a buffering capacity that may affect the
accuracy of the pH-stat method (76), and this may account for
the differences between the pH-stat and TNBS methods
observed by Cheison et al. (50).
Achouri et al. (10) used the TNBS method to determine the
free amino groups and the pH-stat method to determine the
DH for a soy protein isolate hydrolysate. Based on the data
these workers presented, and assuming that values represent
moles of free amino acids/100 g protein (this is not clear in the
article), the DH based on the pH-stat method ranged from 4 to
10%, while the corresponding values determined using the
TNBS method ranged from approximately 27 to 32%. The
reason for this difference was not addressed.
Nielsen et al. (3) conducted a study comparing the TNBS
and OPA methods using soy protein isolate and sodium
caseinate hydrolyzed with Alcalase. They reported a
relationship between the DH determined by the TNBS and
OPA methods as follows:
y = 1.01x + 0.09 (R2 = 0.999) for soy protein isolate
y = 1.08x + 2.88 (R2 = 1.000) for sodium caseinate
The relationship demonstrated a striking correlation
between the two methods. These workers suggested that,
compared to the TNBS method, OPA had the advantage of
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
being more rapid and using more stable and less toxic
reagents. In addition, a good correlation was also reported for
the SN-TCA method when compared to the OPA method (36)
and the TNBS method (77).
Margot et al. (45) compared the pH-stat method with the
SN-TCA method using WPC and porcine pancreatic trypsin.
They found that, unlike the comparison of the TNBS and OPA
methods (3), the pH-stat and SN-TCA methods did not
produce the same results, nor was there a linear relationship
between the two methods. Equations describing the nonlinear
relationship between the DH determined by the two methods
were presented. Margot et al. (45) suggested that a linear
relationship would not be expected since the relationship
between peptide size and solubility in TCA is also likely to be
nonlinear.
In et al. (48) described the hydrolysis of hemoglobin using
a serial digestion with an endopeptidase (Esperase) followed
by exopeptidase (Flavourzyme) at three different
concentrations. They then used the pH-stat, formol titration,
and SN-TCA methods to determine the DH (Figure 4).
Overall, there was very little correlation among the three
methods. The values obtained using the pH-stat method
doubled as Flavourzyme concentration increased, as did the
DH determined by the formol titration method. However, the
latter values were half those reported for the pH-stat method
across all Flavourzyme concentrations. The SN-TCA method
generated similar DH values for all treatments, and these
values were two and four times higher than those obtained
using the pH-stat and formol titration methods, respectively,
for the hydrolysate digested with Esperase only. In contrast,
when 2% Flavourzyme was used in addition to the Esperase,
the SN-TCA and pH-stat method compared well, but both
methods generated DH values twice that of the formol
titration method.
The Need for Standardization of DH Methodologies
DH is defined as the number of free amino groups present
in a hydrolysate as a percentage of the total number of peptide
bonds in a protein substrate. This review has focused on the
most commonly used methods for determining DH. A
number of studies have compared various methods for
determining DH in protein hydrolysates; however, most
studies compare only two or three different methods.
Moreover, the methods being compared differ among studies
and often use unrelated food proteins for the comparison.
 RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO. 5, 2010 1521
This makes a comprehensive assessment of the most suitable
methods for determining DH difficult, and only by
extrapolating across different studies can an evaluation of any
sort be made.
Some methods, such as SN-TCA, do not determine DH
per se, but rather solubility in TCA. Consequently, describing
SN-TCA values as DH values can be misleading. The pH-stat
method, which is most commonly used due to its rapid nature
and simplicity, also does not directly determine the number of
peptide bonds broken. Instead, the base consumption as a
result of peptide bond breakage is measured, and this is used to
extrapolate the number of peptide bonds broken. Osmometric
methods are similar to the pH-stat method except that they are
suitable for acidic reaction mixtures, while the pH-stat method
is suitable for alkali reaction mixtures. The reliance of the
pH-stat method on the estimated pK of the liberated amino
groups to calculate DH reduces the accuracy of this method. In
addition, the presence of exopeptidases appears to significantly
reduce the accuracy of the pH-stat method.
The TNBS and OPA methods determine the number of
peptide bonds broken by measuring the free amino groups
present in a hydrolysate and consequently, from a theoretical
standpoint, accurately determine DH. However, these
methods require a derivatization step, which detracts from
their usability, and they do not determine N-terminal proline
or cysteine accurately. The formol titration method is
straightforward in theory, but in practice yields different
results depending on whether the direct or indirect method is
used.
Overall, there is no consensus as to the best method for
determining the DH of protein hydrolysates. Based on the
literature presented in this review, the pH-stat method
generally did not correlate well with the SN-TCA and formol
titration methods. However, it compared well with the TNBS
in some studies, but not in others. Generally, the TNBS and
OPA methods compared well across most studies. In addition,
for some, but not all, studies, the two latter methods also
correlated well with the SN-TCA method. It would appear that
there is a considerable effect of enzyme, pH, extent of
hydrolysis, and protein source on the suitability of each DH
method. Consequently, careful consideration must be taken
when choosing a DH method for a particular hydrolysis
system. No study has compared all of the commonly used DH
methods across a wide range of food protein hydrolysates, and
it would appear, based on the findings of this review, that such
a study is long overdue. There is generally a poor correlation
among the various DH methods commonly used, and some of
the methods (e.g., SN-TCA) do not even determine DH.
Consequently, to permit the comparison of DH data and to
allow a superior assessment of the extent of hydrolysis in
hydrolysates generated in different laboratories, the use of a
standardized protocol to determine DH is imperative.
References
(1) Rutherfurd, K.J., & Moughan, P.J. (2005) J. AOAC Int. 88,
955–967
(2) Adler-Nissen, J. (1986) Enzymic Hydrolysis of Food
Proteins, Elsevier, London, UK
(3) Nielsen, P.M., Petersen, D., & Dambmann, C. (2001) Food
Chem. Toxicol. 66, 642–646
(4) Nnanna, I.A., & Wu, C. (2007) in Handbook of Food
Products Manufacturing, Y.H. Hui (Ed.), John Wiley &
Sons, Inc., Hoboken, NJ, pp 537–556
(5) Silvestre, M. (1997) Food Chem. 60, 263–271
(6) Wróblewska, B., & Karamaƒ, M. (2003) Acta Alimentaria
32, 193–204
(7) Jacobsen, C.F., Léonis, J., Linderström-Lang, K., & Ottesen,
M. (1957) Methods Biochem. Anal. 4, 171–210
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
(8) Lemos, D., Lawrence, A.L., & Siccardi III, A.J. (2009)
Aquaculture 295, 89–98
(9) Lamsal, B.P., Jung, S., & Johnson, L.A. (2007) Food Sci.
Technol. 40, 1215–1223
(10) Achouri, A., Zhang, W., & Shiying, X. (1998) Food Res.
Int. 31, 617–623
(11) Prieto, C.A., Guadix, A., González-Tello, P., & Guadix,
E.M. (2007) J. Food Eng. 78, 257–265
(12) Kong, X., Zhou, H., & Qian, H. (2007) Food Chem. 102,
759–763
(13) Klompong, V., Benjakul, S., Kantachote, D., & Shahidi, F.
(2007) Food Chem. 102, 1317–1327
(14) Wasswa, J., Tang, J., Gu, X.-H., & Yuan, X.-Q. (2007)
Food Chem. 104, 1698–1704
(15) Chabanon, I., Chevalot, G., Framboisier, X., Chenu, S., &
Marc, I. (2007) Process Biochem. 42, 1419–1428
(16) Guérard, F., Dufossé, L., De La Broise, D., & Binet, A.
(2001) J. Mol. Catal. B: Enzymol. 11, 1051–1059
(17) Li, Z.-Y., Youravong, W., & Kittikun, A.-H. (2010) Food
Sci. Technol. 43, 166–172
(18) Shi, D., He, Z., & Qi, W. (2005) Process Biochem. 40,
1943–1949
(19) Yin, S.-W., Tang, C.-H., Cao, J.-S., Hu, E.-K., Wen, Q.-B.,
& Yang, X.-Q. (2008) Food Chem. 106, 1004–1013
(20) Kristinsson, H.G., & Rasco, B.A. (2000) Process Biochem.
36, 131–139
(21) Liu, Q., Kong, B., Xiong, Y.L., & Xia, X. (2010) Food
Chem. 118, 403–410
(22) Wang, G., & Wang, T. (2009) Int. J. Food Sci. Technol. 44,
763–769
(23) Márquez, M.C., & Vázquez, M.A. (1999) Process Biochem.
35, 111–117
(24) Spellman, D., McEvoy, E., O’Cuinn, G., & Fitzgerald, R.J.
(2003) Int. Dairy J. 13, 447–453
(25) Steinhardt, J., & Beychok, S. (1964) in The Proteins, Vol.
II, H. Neurath (Ed.), Academic Press, New York, NY,
pp 139–304
(26) Dzwolak, W., & Ziajka, S. (1999) J. Food Sci. 64, 393–395
(27) Camacho, F., González-Tello, P., Páez-Dueñas, M.-F.,
Guadix, E.-M., & Guadix, A. (2001) J. Dairy Res. 68,
251–265
(28) García-Carreño, F.L. (1991) Biotechnol. Educat. 2, 150–153
(29) Karamaƒ, M., Amarowicz, R., & Kostyra, H. (2002) Czech.
J. Food Sci. 20, 1–6
(30) Silva, V.M., Park, K.J., & Hubinger, H.D. (2010) J. Food
Sci. 75, C36–C42
(31) Chirife, J., Ferro Fontan, C., & Benmengui, E.A. (1980) J.
Food Technol. 15, 59–70
1522 RUTHERFURD: JOURNAL OF AOAC INTERNATIONAL VOL. 93, NO. 5, 2010
(32) Roth, M., & Hampai, A. (1973) J. Chromatogr. 83, 353–356
(33) Mišún, D., „urda, L., & Jelen P. (2008) Small Ruminant Res.
79, 51–56
(34) Benkhelifa, H., Bengoa, C., Larre, C., Guibal, E., Popineau,
Y., & Legrand, J. (2005) Process Biochem. 40, 461–467
(35) Batista, I., Ramos, C., Coutinho, J., Bandarra, N.M., &
Nunes, M.L. (2010) Process Biochem. 45, 18–24
(36) Rohm, H., Tschager, E., & Jaros, D. (1996) Lebensm. Wiss.
Technol. 29, 191–194
(37) Britten, M., Giroux, H.J., & Gaudin, V. (1994) J. Dairy Sci.
77, 676–684
(38) Frister, H., Meisel, H., & Schlimme, E. (1988) Fresenius’ J.
Anal. Chem. 330, 631–633
(39) Hernández, M.J., Domingo, E.B., Villanueva Camañas,
R.M., & García Alvarez-Cogue, M.C. (1991) J. Dairy Sci.
74, 1779–1785
(40) Živanoviƒ, I., Vaštag, Ž., Popoviƒ, S., Popoviƒ, L., & Periƒin,
D. (2010) Int. J. Biol. Life Sci. 6. 30–34
(41) Ovissipour, M., Abedian, A., Motamedzadegan, A., Rasco,
B., Safari, R., & Shahiri, H. (2009) Food Chem. 115,
238–242
(42) Vega, R.E., & Brennan, J.G. (1988) J. Food Eng. 8, 201–215
(43) Montoya, C.A., Gomez, A.S., Lallès, J.-P., Souffrant, W.B.,
Beebe, S., & Leterme, P. (2008) Food Chem. 106,
1225–1233
(44) Ge, S.-J., Bai, H., Yuan, H.-S., & Zhang, L.-X. (1996) J.
Biotechnol. 50, 161–170
(45) Margot, A., Flaschel, E., & Renken, A. (1994) Process
Biochem. 29, 257–262
(46) Periƒin, D., Raduloviƒ-Popoviƒ, L., Vaštag, Ž.,
Maparev-Popoviƒ, S., & Triviƒ S. (2009) Food Chem. 115,
753–757
(47) Moughan, P.J., Darragh, A.J., Smith, W.C., & Butts, C.A.
(1990) J. Sci. Food Agric. 52, 13–21
(48) In, M.-J., Chae, H.J., & Oh, N.-S. (2002) Bioresource
Technol. 84, 63–68
(49) Bandyopadhyay, K., Misra, G., & Ghosh, S. (2008) J. Oleo
Sci. 57, 47–52
(50) Cheison, S.C., Zhang, S.-B., Wang, Z., & Xu, S.-Y. (2009)
Food Res. Int. 42, 91–97
(51) Govindaraju, K., & Srinivas, H. (2007) Food Sci. Technol.
40, 1056–1065
(52) Jamdar, S.N., Rajalakshmi, V., Pednekar, M.D., Juan, F.,
Yardi, V., & Sharma, A. (2010) Food Chem. 121, 178–184
(53) Quist, E.E., Phillips, R.D., & Saalia, F.K. (2009) Food Sci.
Technol. 42, 1717–1721
(54) Tunçürk, Y., & Zorba, O. (2006) Food Hydrocolloids 20,
475–482
(55) Vorob’ev, M.M. (2009) J. Mol. Catal. B: Enzymol. 58,
146–152
(56) Yust, M.M., Pedroche, J., Megías, M., Girón-Calle, J., Alaiz,
M., Millán, F., & Vioque, J. (2003) Grasas Aceities 54,
419–423
(57) Satake, K., Okuyama, T., Ohashi, M., & Shinoda, T. (1960)
J. Biochem. 47, 654–660
(58) Adler-Nissen, J. (1979) J. Agric. Food Chem. 24, 1090–1093
(59) Fields, R. (1971) Biochem. J. 124, 581–590
(60) Kotaki, A., Harada, M., & Yagi, K. (1964) J. Biochem. 55,
Downloaded from https://academic.oup.com/jaoac/article-abstract/93/5/1515/5655787 by guest on 12 February 2020
553–561
(61) Taylor, W.H. (1957) Analyst 82, 488–498
(62) Daukšas, E., Šližyte, R., Rustad, T., & Storrø, I. (2004) J.
Aquat. Food Prod. Technol. 3, 101–114
(63) Nilsang, S., Lertsiri, S., Suphantharika, M., & Assavanig, A.
(2005) J. Food Eng. 70, 571–578
(64) Šližyte, R., Daukšas, E., Falch, E., Storrø, I., & Rustad, T.
(2005) Process Biochem. 40, 1415–1424
(65) You, L., Zhao, M., Cui, C., Zhao, H., & Yang, B. (2009)
Innovative Food Sci. Emerg. Technol. 10, 235–240
(66) Taha, F.S., & Ibrahim, M.A. (2002) Grasas Aceites 53,
273–281
(67) Pintado, M.E., Pintado, A.E., & Malcata, F.X. (1999) J.
Food Eng. 42, 1–13
(68) Clegg, K.M., Lee, Y.K., & McGilligan, J.F. (1982) J. Food
Technol. 17, 517–520
(69) Frost, D.V., & Heinsen, J. (1945) J. Biol. Chem. 161,
517–521
(70) Panasiuk, R., Amarowicz, R., Kostyra, H., & Sijtsma, L.
(1998) Food Chem. 62, 363–367
(71) Schwartz, T.B., & Engel, F.L. (1950) J. Biol. Chem. 184,
197–202
(72) Kaur, L., Rutherfurd, S.M., Moughan, P.J., Drummond, L.,
& Boland, M.J. (2010) J. Agric. Food Chem. 58, 5074–5080
(73) Roufik, S., Gauthier, S.F., & Turgeon, S.L. (2007) Int. Dairy
J. 17, 471–480
(74) Ruckebusch, C., Duponchel, L., Huvenne, J.-P., Legrand, P.,
Nedjar-Arroume, N., Lignot, B., Dhulster, P., & Guillochon,
D. (1999) Anal. Chim. Acta 396, 241–251
(75) Rutherfurd, S.M., & Moughan, P.J. (1998) J. Dairy Sci. 81,
909–917
(76) Salaun, F., Mietton, B., & Gaucheron, F. (2005) Int. Dairy J.
15, 95–109
(77) Bouton, Y., & Grappin, R. (1994) J. Dairy Res. 61, 437–440