METHODS TO DETERMINE AMINO ACID DIGESTIBILITY IN

CORN BY-PRODUCTS

Hans H. Stein, Ameer A. Pahm, and Carsten Pedersen

Department of Animal and Range Sciences
South Dakota State University, Brookings 57007

l. INTRODUCTION

While corn contains only limited quantities of crude protein and amino acids, corn by-products
often contain crude protein and amino acids in appreciable amounts. The reason for this
difference is that corn by-products are often produced after processing that primarily removes the
nitrogen-free extracts (i.e., the starch) from the corn kernel. Because starch is approximately
70% of the total corn kernel, the quantity of crude protein and other nutrients is concentrated in
the product that is left over after processing (Table 1). Corn by-products, therefore, are valuable
sources of crude protein and amino acids in diets for monogastric animals. As is the case with
most other by-products, variations in the concentrations and the digestibility of the amino acids
in corn by-products exist. The main reason for these variations is differences in processing
technology and procedures. Currently, the most used corn by-product in the feeding of
monogastric animals is distillers dried grain with solubles (DDGS). This product is produced by
the ethanol industry and contains mainly the crude protein, fat, fiber and minerals from the corn
kernel whereas the majority of the starch has been removed in the fermentation process. In
general, the concentrations of crude protein, amino acids, fiber and minerals in DDGS are close
to three times that than in corn. Thus, the crude protein concentration in DDGS is usually 26 –
28% and the concentration of amino acids is also approximately three times higher than in corn
(Table 1). Because of the relatively high concentration of crude protein and amino acids, DDGS
contributes significantly to the dietary concentration of amino acids if included in diets for pigs
and poultry. It is, therefore, necessary to estimate the digestibility of these amino acids.

It is the objective of the current contribution to summarize existing knowledge about the methods
that are available to measure amino acid digestibility in corn by-products. Because of the
importance of DDGS in the feeding of pigs and poultry in the US, the summary will be based on
experiences with DDGS. However, the concepts discussed may also be applied to other corn by-
products.

2. AMINO ACID AVAILABILITY

Only amino acids that can be incorporated into tissue proteins are bioavailable. Bioavailability is
defined as the proportion of dietary amino acids that are absorbed in a chemical form that is
suitable for protein synthesis (Batterham, 1992; Lewis, 1992). Bioavailability of AA may be
measured using slope ratio techniques in which the response of an animal to increased intake of
an amino acid is measured (Batterham, 1992). The diets used for this procedure needs to be
deficient in only the amino acid in question and all levels of this amino acid needs to be fed
below the requirement of the animal. However, the slope ratio procedure is tedious and costly
and the determined values for amino acid availability are unique only to the experimental
procedures used and may not be additive in a mixed diet (Gabert et al., 2001). Therefore, for
practical feed formulation, amino acid availability values are not used. In stead, the digestibility
coefficients for amino acids are measured and used as an indication of the quantities of dietary
amino acids that are available to the animal (Stein and Nyachoti, 2003). The term “amino acid
digestibility” does not refer to the digestion of amino acids – it only refers to the digestion of the
peptide bonds connecting the amino acids in a dietary protein (Fuller, 2003).

Table 1. Concentration of energy and nutrients in corn and corn by-productsa.

Corn gluten Corn gluten
Item Product: Corn DDGS feed meal
Energy, ME kcal/kg 3,420 2,820 2,605 3,820
Crude protein, % 8.3 27.7 21.5 60.2
Crude fat, % 3.9 8.4 3.0 2.9
NDF, % 9.6 34.6 33.3 8.7
ADF, % 2.8 16.3 10.7 4.6
Calcium, % 0.03 0.20 0.22 0.05
Phosphorus, % 0.28 0.77 0.83 0.44
Amino acids, %
Arginine 0.37 1.13 1.04 1.93
Histidine 0.23 0.69 0.67 1.28
Isoleucine 0.28 1.03 0.66 2.48
Leucine 0.99 2.57 1.96 10.19
Lysine 0.26 0.62 0.63 1.02
Methionine 0.17 0.50 0.35 1.43
Cysteine 0.19 0.52 0.46 1.09
Phenylalanine 0.39 1.34 0.76 3.84
Tyrosine 0.25 0.83 0.58 3.25
Threonine 0.29 0.94 0.74 2.08
Tryptophan 0.06 0.25 0.07 0.31
Valine 0.39 1.30 1.01 2.79
a
Values from NRC (1998).

3. AMINO ACID DIGESTIBILITY – GENERAL PRINCIPLES

Because the amino acids of undigested dietary proteins entering the large intestine may be
metabolized by hind-gut microbes before they are excreted in the fecal material from the animal,
values for total tract digestibility of amino acids are not accurately predicting amino acid
absorption by the animal. To avoid the manipulation by hind-gut microbes, the digestibility of
amino acids by monogastric animals is most correctly measured at the end of the small intestine
and is referred to as ileal digestibility values (Sauer and de Lange, 1992). This creates a need for
using techniques that enable researchers to capture ileal fluids at the end of the small intestine.
Several techniques have been proposed for this and have been discussed previously (Moughan,
2003b; Stein, 2003). In North America, the installment of a T-cannula in the distal ileum of pigs
(10 to 15 cm prior to the ileo-cecal valve) appears to be the procedure of choice. This procedure
has proven to be accurate and has a minimal trial to trial variation. Because the T-cannula, like
most other procedures that are used for ileal fluid collections, does not allow for the total
collection of the ileal output from the animal, an indigestible marker needs to be included to
calculate changes in amino acid concentrations. Chromic oxide is often used for this purpose, but
other markers exist. Ileal digestibility values are calculated using Eq. 1.

AID = (1 –[(Aad / AAf) x (Mf / Md)] x 100% [1]

where AID is the apparent ileal digestibility of an amino acid, AAd is the amino acid
concentration in the ileal digesta DM (g / kg DM), AAf is the amino acid
concentration in the feed DM (g / kg DM), Mf is the marker concentration in the feed
DM (g / kg DM), and Md is the marker concentration in the ileal digesta DM (g / kg
DM).

Amino acid digestibility values calculated using this procedure are called “apparent ileal
digestibility values” to reflect the fact that this value is calculated simply by subtracting the ileal
amino acid output from the intake of amino acids. The ileal digesta contain not only un-absorbed
dietary amino acids, but also amino acids of endogenous origin; i.e., amino acids that were
absorbed from the intestines and then re-secreted into the intestinal tract in the form of
endogenous proteins such as mucin, sloughed cells, enzymes etc. The value for AID represents,
therefore, only the apparent digestibility of the dietary proteins.

The endogenous amino acids that are secreted into the intestinal tract may be divided into basal
endogenous secretions and diet-specific endogenous secretions (Jansman et al., 2002). The basal
endogenous amino acids consist of amino acids that are secreted into the GI-tract of fasted
animals in addition to amino acids that are secreted in response to the DM intake of animals.
These losses are usually measured as g per kg DM intake. Recent evidence suggests that the
value for endogenous losses (in g per kg DM intake) depends on the DM intake of the animal
and declines as DM intake increases because of the relatively lower influence pr kg DM of the
fasting endogenous loss (Moter and Stein, 2004). As a consequence, only values for endogenous
losses that are measured in animals given free access to feed are applicable to commercial
conditions.

In addition to the basal endogenous loss, most feed ingredients also introduce a diet-specific
endogenous loss which is mainly caused by fibers and anti-nutritional factors in the ingredient.
The diet-specific losses may vary from almost zero in purified ingredients such as casein and
lactose to values that exceed the basal endogenous losses in ingredients that are high in fiber and
anti-nutritional factors such as canola meal and wheat middlings. Because DDGS contains
relatively high quantities of fiber, it is expected that this ingredient will elicit relatively large
diet-dependent endogenous losses. Comprehensive reviews of endogenous losses and factors
influencing endogenous losses have been published (Taminga et al., 1995, Boisen and Moughan,
1996; Nyachoti et al., 1997). The variation in the total quantities of endogenous proteins that
have been reported in the literature also has been published (Jansman, 2002).

Techniques to measure basal endogenous losses include feeding a protein-free diet, the peptide
alimentation procedure, and the regression procedure (Stein, 2003). In North America, the
protein-free diet is most commonly used. Using this approach, the basal endogenous losses are
calculated from ileal digesta collected from animals that are fed the protein-free diet according to
Eq. 2:

IAAend = [AAd x (Mf/Md)] [2]

where IAAend is the basal endogenous loss of an amino acid at the distal ileum (mg
per kg DM intake), AAd is the concentration of that amino acid in the digesta DM,
Mf is the marker concentration in the feed DM, and Md is the marker concentration
in the ileal digesta DM.

There are no techniques available to measure the diet-specific endogenous losses, but total
endogenous losses (basal and diet-specific) may be estimated using the homoarginine technique
or the N15 isotope dilution technique. Both of these procedures have several limitations and
further work to refine these techniques is required before they can be routinely employed in
animal nutrition studies (Stein, 2003).

The introduction of endogenous amino acids into the GI-tract of pigs introduces several
challenges in accurately estimating amino acid absorption because these secretions are included
in the calculations of apparent ileal digestibility values according to Eq. 1. The endogenous
amino acids comprise a relatively larger portion of the total amino acid output from pigs fed low-
protein feed ingredients (i.e., cereal grains) as compared to pigs fed ingredients with a higher
protein concentration such as soybean meal. Therefore, the values for AID that are measured for
low-protein feed ingredients are usually under-estimated (Donkoh et al., 1994; Fan et al., 1994).
Because of this underestimation, values for AID that are measured in individual feed ingredients
usually do not add up in a mixed diet that contains both low-protein feed ingredients and high-
protein ingredients (Stein et al., 2005c). However, if the values for ileal digestibility are
corrected for the basal endogenous losses of amino acids, then this underestimation may be
avoided. By doing that, the standardized ileal digestibility coefficients are calculated according
to Eq. 3:

SID = [AID + (IAAend / AAf)] [3]

where SID is the standardized ileal digestibility of an amino acid (%), AID and
IAAend are calculated according to Eq. [1] and Eq. [2], respectively, and AAF is the
amino acid concentration in the feed DM (g / kg DM). It was recently demonstrated
that values for standardized ileal digestibility are additive in mixed diets even if low
protein feed ingredients are incorporated in the diets (Stein et al., 2005c). Thus, by
using standardized ileal digestibility values, the problems associated with using
apparent ileal digestibility values are eliminated. The feed industry is, therefore,
moving towards using values for standardized ileal digestibility of amino acids rather
than apparent ileal digestibility values. Comprehensive discussions of the principles
behind calculations of standardized ileal digestibility values are available (Mosenthin
et al., 2000; Jansman et al., 2002).
4. IN VIVO MEASUREMENT OF AMINO ACID DIGESTIBILITY IN CORN BY-
PRODUCTS

The amino acid digestibility in corn by-products is usually measured following the procedures
outlined above. For DDGS, values for both apparent and standardized ileal digestibility of amino
acids have been determined in 14 samples originating from 13 different ethanol plants in
Minnesota and South Dakota (Stein et al., 2005a and b). The diets used to measure the apparent
ileal digestibility values in these experiments consisted of 67% DDGS, 27% cornstarch, 1%
soybean oil, 3% sugar, and vitamins and minerals. The basal endogenous losses were determined
using a protein-free diet and the standardized ileal digestibility values were calculated according
to Eq. 3. The results of the experiments showed that some variations exist for the amino acid
digestibility among different samples of DDGS (Table 2). This is true in particular for lysine that
is more variable than all other amino acids in terms of digestibility. The reason for this variation
is believed to be that lysine may have been heat-damaged in some of the samples of DDGS
which in turn has lowered the calculated digestibility of lysine in these samples. Further work is
needed to identify the reasons for this heat damage and to establish procedures for the production
that allow ethanol plants to dry the products without heat damaging it. Nonetheless, the amino
acids in DDGS have a medium digestibility and, except for lysine, the variability among
different samples is within the normal range of variation found in other feed ingredients. Values
for apparent and standardized ileal digestibility in five sources of DDGS originating from other
parts of the US also have been published (Fastinger and Mahan, 2005). These values confirmed
that lysine is the most variable amino acid in DDGS in terms of digestibility.

The digestibility of amino acids in other corn by-products has been less extensively researched
than in DDGS, but values for apparent and standardized ileal digestibility of amino acids in corn
gluten meal and corn gluten feed have been published (NRC, 1998). The amino acid digestibility
in corn gluten feed may be determined as outlined for DDGS and with a similar type of diet.
Because of the higher protein concentration in corn gluten meal, only 30 to 35% of corn gluten
meal would be needed in the diet used to determine the amino acid digestibility in order to
formulate a test diet containing 18 to 20% crude protein. Amino acid digestibility values in other
corn by-products may be measured using a similar approach and with similar types of diets.

5. IN VITRO PROCEDURES TO ESTIMATE AMINO ACID DIGESTIBILITY IN

CORN BY-PRODUCTS

Because in vivo procedures are expensive and time consuming to conduct, current work at South
Dakota State University is focusing on identifying in vitro procedures that are fast and
inexpensive and able to predict the digestibility of amino acids in DDGS. A total of six different
in vitro procedures have been tested at this point. For each procedure, values obtained using the
in vitro procedure have been correlated with the standardized ileal digestibility values that were
obtained in vivo for the 14 samples that have been measured (Table 2).

The six in vitro procedures include two enzymatic procedures (i.e., a one-step pepsin procedure,
and a two-step pepsin-pancreatin procedure), a colorimetric procedure, and three chemical
procedures (i.e., KOH solubility, the furosine procedure, and the reactive lysine procedure).
Table 2. Standardized ileal digestibility (%) of amino acids in 14 samples of DDGS
by growing pigsa.
Standard Lowest Highest
Digestibility: Average deviation value value

Item Crude protein: 70.5 3.96 63.5 77.6

Indispensable amino acids
Arginine 78.6 3.80 74.1 89.9
Histidine 75.2 4.64 70.0 88.9
Isoleucine 72.7 4.36 67.4 84.2
Leucine 81.9 4.16 75.5 91.3
Lysine 59.0 7.19 43.9 78.4
Methionine 80.3 3.98 73.9 89.2
Phenylalanine 79.0 3.78 73.5 89.4
Threonine 69.9 5.96 62.2 87.2
Valine 71.7 4.30 67.3 87.2
Dispensable amino acids
Alanine 75.7 4.27 69.7 87.6
Aspartic acid 66.9 4.66 59.4 80.2
Cysteine 72.4 4.46 66.0 85.8
Glutamic acid 76.7 6.11 67.4 91.4
Glycine 57.2 6.36 46.8 76.7
Proline 66.9 9.40 57.8 88.8
Serine 72.1 5.10 59.6 80.9
Tyrosine 80.9 3.74 74.6 89.9
a
Results reported as the average of 14 samples, the standard deviation, and the range.

5.1. One-Step Pepsin Procedure

The one-step pepsin procedure involves the incubation of a DDGS sample at 37oC in pepsin for 8
or 18 hours at pH 2.5 (procedure 971.09, AOAC, 2000). The sample is filtrated after incubation
and the undigestible portion of the protein is believed to be present in the filtrate. By analyzing
this portion and relating it to the total protein content in the sample, the disappearance of protein
is calculated and assumed to be equal to the digestibility of the protein. Results obtained using
this procedure are presented in Figure 1. The correlation between the in vitro digestibility of
protein and the value for standardized ileal digestibility obtained in vivo is only 0.52. It is
assumed that the digestibility of protein is well correlated to the digestibility of all amino acid
which for a protein that may have been heat damaged is somewhat questionable. Based on the
results that have been obtained this far, it is questionable if this procedure is accurate enough to
predict amino acid digestibility in DDGS.
 80

In vivo SID 75

70

65 y = 1.59x - 61.90
R2 = 0.52

60
80 81 82 83 84 85 86
In vitro pepsin digestion

Figure 1. Correlation between the standardized ileal digestibility values

obtained in vivo and by the one-step pepsin procedure in 14 samples
of DDGS.

5.2. Two-Step Pepsin-Pancreatin Procedure

The two-step pepsin-pancreatin procedure involves sample incubation at 39oC at pH 2 with

pepsin followed by the incubation with pancreatin (a mixture of pancreatic proteolytic enzymes)
at pH 6.8. Following the incubation, all samples are filtrated and washed in ethanol and acetone
and dehydrated. The filtrate is then analyzed for the concentration of crude protein. The
procedure was initially developed for predicting amino acid digestibility in mixed diets (Boisen
and Fernandez, 1995). Alternatives to this procedure including longer incubation times were later
introduced to predict amino acid digestibility in meat and bone meal (Qiao et al., 2004). Based
on a series of experiments in our laboratory, it was concluded that to obtain a reasonably high
correlation between data obtained using this procedure and standardized ileal digestibility values
measured in vivo, it is necessary to incubate the samples with pepsin for 24 hours followed by a
96-hour incubation with pancreatin. An R2 value of 0.79 has been obtained using this procedure
(Figure 2). As was the case with the one-step pepsin procedure, this procedure only estimates the
protein digestibility which is assumed to be well correlated to the amino acid digestibility.
However, current work at South Dakota State University is focusing on analyzing the filtrate for
individual amino acids to investigate the possibility for predicting individual amino acid
digestibility with the two-step pepsin-pancreatin procedure.
 80

75
SID in vivo

70

65 y = 0.63x + 25.79
2
R = 0.79
60

55
55 60 65 70 75 80 85
SID in vitro
Figure 2. Correlation between the standardized ileal digestibility values obtained
in vivo and by the two-step pepsin-pancreatin procedure in 14
samples of DDGS.

5.3. Colorimetric Procedures

Colorimetric assays may be used to predict amino acid digestibility in DDGS because it is
believed that processing that is detrimental to amino acid digestibility may result in darker colors
of the product. This procedure has previously been correlated to amino acid digestibility in
DDGS by chickens (Cromwell et al., 1993). Work at South Dakota State University was
completed using two different instruments to measure color in the samples (i.e., Hunterlab and
Minolta). It was concluded from the initial work that the Minolta instrument gave a better
correlation to in vivo digestibility than did the Hunterlab. Further experiments involved
standardizing the distance from the instrument to the sample using a glass plate or measuring the
color without the glass plate. Samples were either used as-is or were finely ground prior to the
measurements. It was concluded from these experiments that the best correlation between color
scores and in vivo amino acid digestibility is obtained if no glass plate is used. However, there is
no difference between the correlations obtained in DDGS samples as-is and the finely ground
samples (Table 3). The correlations between the color scores and the in vivo amino acid
digestibility for the majority of the amino acids have been between 0.60 and 0.75 indicating that
although there appears to be some correlation between color and amino acid digestibility this
correlation is not particularly strong. Surprisingly, the amino acid with the lowest correlation is
lysine. The biggest advantage of the colorimetric procedure is that it is very fast, easy to perform,
and inexpensive. Future work at South Dakota State University will focus on refining the
procedure to possibly find better correlations.
Table 3. Correlation (%) between standardized ileal digestibility in DDGS by growing pigs
and the color in DDGS measured by Minolta in raw and ground samples.
Item Sample: DDGS, as-is DDGS, ground
Glass plate - + - +
Crude protein 0.67 0.60 0.54 0.53
Indispensable amino acids
Arginine 0.70 0.65 0.76 0.63
Histidine 0.67 0.62 0.76 0.67
Isoleucine 0.60 0.52 0.53 0.49
Leucine 0.76 0.69 0.77 0.71
Lysine 0.49 0.45 0.60 0.53
Methionine 0.71 0.69 0.68 0.55
Phenylalanine 0.70 0.64 0.66 0.59
Threonine 0.65 0.63 0.79 0.61
Valine 0.62 0.55 0.57 0.53
Dispensable amino acids
Alanine 0.73 0.66 0.73 0.65
Aspartic acid 0.66 0.63 0.44 0.45
Cysteine 0.73 0.71 0.78 0.68
Glutamic acid 0.70 0.66 0.86 0.71
Glycine 0.53 0.46 0.42 0.42
Proline 0.55 0.45 0.51 0.40
Serine 0.52 0.47 0.24 0.30
Tyrosine 0.74 0.73 0.71 0.61

5.4. KOH-Solubility

It has been suggested that the KOH-solubility of a feed stuff is well correlated to the digestibility
of protein (Araba and Dale, 1990). The procedure involves the incubation of a sample with a
0.2% KOH solution for 20 min at room temperature. Following this incubation, the sample is
centrifuged and the supernatant is analyzed for the nitrogen concentration. The nitrogen that is
left in the supernatant is believed to be indigestible to the animal; therefore, the digestible
quantity of protein may be calculated as the difference between the protein in the sample and the
protein in the supernatant. Results obtained for DDGS using this procedure have not been
encouraging (Figure 3) and at this point, it is not recommended that the procedure is used to
estimate amino acid digestibility in DDGS.
 81

76
SID in vivo of CP

71

66

y = 1.39x - 55.43
61
R2 = 0.47
56
75 80 85 90 95
Solubility of N in KOH
Figure 3. Correlation between the standardized ileal digestibility values
obtained in vivo and by the KOH solubility procedure.

5.5. Furosine Concentration Procedure

Heat treatment is used extensively in the production of DDGS to dehydrate the products coming
off the fermentation process. During heat treatment of proteins, the epsilon amino group of free
lysine and protein-bound lysine may react with reducing sugars in the Maillard reaction
(Maillard, 1912, 1916). This reaction generates both early and late Maillard products. The early
Maillard products are structurally altered lysine derivatives that are called Amadori compounds,
deoxy-ketosyl derivatives, or blocked lysine, while the late Maillard products are called
melanoidins. The melanoidins will not be identified in the regular analysis for amino acids and
will only result in lower lysine to crude protein ratios in the product. The lysine to crude protein
ratio in corn is approximately 0.031. In non-heat damaged DDGS, the ratio is about the same, but
in heat damaged DDGS, ratios of less than 0.025 have been detected. The calculation of the
lysine to crude protein ratio, therefore, gives an indication of the extent of conversion of lysine to
melanoidins, and therefore, of the degree of heat damage that has occurred. The melanoidins do
not interfere with the normal analysis for lysine and have no influence on the digestibility values
that are calculated – they only result in lower concentrations of lysine in the sample, and
therefore, in lower quantities of lysine being absorbed.

The Amadori compounds (early Maillard products), on the other hand, interfere with the amino
acid analysis and give inaccurate lysine concentrations in the sample being analyzed. These
products are derived from the reaction of lysine with lactose, fructose, aspartate, or glutamate
and the resulting derivatives are called lactulosyllysine, fructosyllysine, aspartyllysine and
glutamyllysine, respectively (Friedman, 1996). The lysine that is bound in these compounds is
called “blocked lysine” and is biologically unavailable because it is resistant to gastrointestinal
enzymatic breakdown (Maga, 1981). However, during acid hydrolysis, 40 to 50% of the blocked
lysine is released as lysine (Hurrell and Carpenter, 1981) while the rest is released as furosine
and pyridosine. The resulting ileal digestibility values will, therefore, overestimate the digestible
lysine content in the sample, but will underestimate the digestibility coefficient of the un-blocked
lysine (Moughan, 2003a). Because lysine, furosine and pyridosine are released from the blocked
lysine in a constant ratio, the total amount of blocked lysine may be calculated by analyzing
either the furosine or the pyridosine concentration in the sample. Work in our laboratory has
demonstrated that furosine is easily analyzed using high performance liquid chromatography
(HPLC). Subsequently, the concentration of blocked lysine has been calculated assuming that the
furosine concentration equals 20% of total blocked lysine. The correlation between the
calculated blocked lysine and the in vivo standardized digestibility of lysine in DDGS was
calculated at 0.71 (Figure 4). Future work will focus on further developing this procedure which
seems to hold some promise as an accurate method to estimate the digestibility of lysine in
DDGS. The digestibility of amino acids other than lysine can, however, not be predicted using
this procedure.

Correlation of Furosine
and SID lysine

90.00

80.00 R2 = 0.7138

70.00

60.00
SID lysine

50.00

40.00

30.00

20.00

10.00

0.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Blocked lysine, % of total lysine value

Figure 4. Correlation between blocked lysine (calculated from furosine) and

standardized ileal digestibility of lysine in 14 samples of DDGS.

5.6. Reactive Lysine Procedure

While the furosine procedure can be used to estimate the quantity of blocked lysine in heat-
damaged feed samples, the reactive lysine method estimates the amount of lysine that is not
bound to any food component in a manner that hinders lysine digestibility or availability. This
lysine is called the reactive lysine and is available for absorption and utilization by the animal.
For feedstuffs that are not heated, the total lysine content in the sample is not contaminated with
Amadori compounds and all the analyzed lysine in the sample is expected to be available for
absorption and utilization. In such a feed ingredient the reactive lysine would be expected to be
equal to the total analyzed concentration of lysine. However, in heat damaged proteins, a portion
of the analyzed lysine may consist of Amadori products and introduce errors in the calculation of
digestible lysine as discussed above (Darragh and Hodgkinson, 2003). In the reactive lysine
procedure, the quantity of lysine that is not bound in the Amadori compounds is quantified.
Because this is the only portion of the analyzed lysine that is available for absorption, a
quantification of this portion will be directly related to the digestible quantities of lysine from the
animal (Moughan and Rutherfurd, 1996). There are at least three methods available to determine
the reactive lysine content of a feed sample. All three methods involve the chemical conversion
of the reactive lysine in the feed ingredient to a different chemical compound (i.e., dinitrophenyl
lysine, trinitrophenyl lysine, or homoarginine). For samples that may contain starch, the
conversion of reactive lysine to homoarginine is believed to be the most accurate. This procedure
involves the reaction of the feed sample with O-methyl-isourea which will convert the reactive
lysine (but not the blocked lysine) to homoarginine. By measuring the conversion of total lysine
to homoarginine, the quantity of reactive lysine can be estimated. Because the quantity of
reactive lysine is supposed to be directly related to the digestibility of lysine, the conversion of
lysine to homoarginine is believed to be correlated to the digestibility of lysine. Preliminary
work in our laboratory has resulted in a medium correlation to the standardized ileal digestibility
of lysine in samples of DDGS (Figure 5). However, there are several methodological aspects of
this procedure that still need to be researched and it is possible that the reactive lysine procedure
may be refined and eventually yield better correlations to in vivo data.

Correlation of SID lysine (%) and Lysine conversion to

Homoarginine (%)
R2 = 0.6557
90.00
80.00
70.00
60.00
SID lysine, %

50.00
40.00
30.00
20.00
10.00
0.00
76.00 78.00 80.00 82.00 84.00 86.00 88.00
% Conversion to Hom oarginine

Figure 5. Correlation between the conversion of lysine to homoarginine and

standardized ileal digestibility of lysine in 14 samples of DDGS.
Because the furosine procedure and the reactive lysine procedure are measuring the blocked
lysine and the reactive lysine, respectively, the values obtained for these procedures theoretically
will add up to the total analyzed lysine concentration in the feed ingredient. Based on the limited
amount of work that has been conducted in our laboratory at this time it is concluded that for
most samples of DDGS, this is also the case. There is, therefore, reason to believe that both of
the two procedures may be used in the future to predict the digestibility of lysine in DDGS and
other corn by-products that have been heat damaged.

6. CONCLUSIONS

The digestibility of amino acids in corn by-products may be measured in vivo in pigs based on
principles for calculating the ileal digestibility. This approach involves a procedure that allows
for the collection of ileal digesta at the distal end of the ileum. A T-cannula is most commonly
used for this purpose, but other procedures exist and have been shown to be equally effective.
Because of the contribution of endogenous amino acids to the ileal output, it is necessary to also
estimate the quantities of endogenous amino acids reaching the distal ileum. This is most
commonly accomplished by feeding the animals a protein-free diet. By subtracting the
endogenous amino acids from the total ileal output, the standardized ileal digestibility of amino
acids may be calculated. Values for standardized ileal digestibility have been shown to be
additive in mixed diets fed to pigs. Therefore, the values for standardized ileal digestibility
should be used in practical feed formulation.

Because of the variation in digestibility of amino acids that has been reported for different
samples of DDGS, a need for a rapid and inexpensive in vitro procedure to predict amino acid
digestibility has emerged. Several approaches have been researched to identify such a procedure.
Based on work with 14 samples of DDGS is has been concluded that a two-step pepsin-
pancreatin procedure may yield results that are reasonable well correlated to the in vivo
digestibility of protein. However, questions related to how well the protein digestibility is related
to the digestibility of individual amino acids still exist. In particular, there seems to be a
relatively low correlation between protein and lysine digestibility in a heat treated feed
ingredient such as DDGS. Therefore, it is not expected that the data obtained using the pepsin-
pancreatin procedure are well correlated with the digestibility of lysine, but this question still
remains to be answered. Future work will focus on using the pepsin-pancreatin procedure to
predict the digestibility of all individual amino acids including lysine. A one-step pepsin
digestibility procedure was shown to be less accurate than the pepsin-pancreatin procedure in
predicting protein digestibility in DDGS. Likewise, a procedure based on measuring protein
solubility following the incubation of DDGS with KOH seems to be somewhat ineffective in
predicting protein digestibility in DDGS.

The furosine procedure and the reactive lysine procedure are similar in that both procedures may
be used to calculate the quantity of lysine that is available for absorption in a feed sample.
However, both procedures only predict the digestible contents of lysine. If the digestible contents
of other amino acids are to be predicted, other procedures are needed. Preliminary work with
these two procedures has given results that hold some promise for future research. If it turns out
that both procedures are equally effective in predicting the quantities of digestible samples in
DDGS, then the furosine procedure may be the method of choice because it is less tedious and
more inexpensive than the reactive lysine procedure.

To get an accurate prediction of the digestibility of all amino acids in a feed ingredient a
combination of the pepsin-pancreatin procedure and the furosine procedure may be needed.

7. REFERENCES

AOAC. 2000. Official methods of analysis (17th Ed.). Association of Official Analytical Chemists, Arlington, VA.
Araba, M. and N. M. Bale. 1990. Evaluation of protein solubility as an indicator of overprocessing of soybean meal.
Poult. Sci. 69:76.
Batterham, E.S. 1992. Availability and utilization of amino acids for growing pigs. Nutr. Res. Rev. 5:1-18.
Boisen, S. and J.A. Fernandez. 1995. Prediction of the apparent ileal digestibility of protein and amino acids in
feedstuffs and feed mixtures for pigs by in vitro analyses. Anim. Feed Sci. Technol. 51: 29-43.
Boisen, S. and P. J. Moughan. 1996. Dietary influences on endogenous ileal protein and amino acid loss in the pig -
A review. Acta Agric. Scand., Sect A, Animal Sci. 46:154-164.
Cromwell, G.L., K.L. Herkelman, and T.S. Stahly. 1993. Physical, chemical, and nutritional characteristics of
distillers dried grain with solubles for chicks and pigs. J. Anim. Sci. 71:679-686.
Darragh, A.J. and S.M. Hodgkinson. 2000. Quantifying the digestibility of Protein. J. Nutr. 130:1850S.
Donkoh, A. and P.J. Moughan. 1994. The effect of dietary crude protein content on apparent and true ileal nitrogen
and amino acid digestibilities. Br. J. Nutr. 72:59-68.
Fan, M.Z., W.C. Sauer, R.T. Hardin, and K.A. Lien. 1994. Determination of apparent ileal amino acid
digestibility in pigs: Effect of dietary amino acid level. J. Anim. Sci. 72:2851-2859.
Fastinger, N.D. and D.C. Mahan. 2005. Apparent and true ileal amino acid and energy digestibility and weanling
pig performance of five sources of distillers dried grain with solubles. Abstract presented at the Midwestern
Section, ASAS 2005 meeting, Des Moines, IA. Abstract #98.
Friedman, M. 1996. Food Browning and its Prevention: an overview. J. Agric. Food Chem. 4:631.
Fuller, M. 2003. Amino acid bioavailability – a brief history. p. 183-198 in Ball, R.O. (Ed.): Digestive Physiology in
pigs, Proc. 9th intl. Symposium. Vol. 1.
Gabert, V.M., H. Jørgensen, and C.M. Nyachoti. 2001. Bioavailability of amino acids in feedstuffs for swine. P 151
– 186 in Lewis, A.J. and L.L. Southern (Ed.). Swine Nutrition, 2nd Ed. CRC Press, Washington, DC.
Hodgkinson, S.M. and P.J. Moughan. 2000. Amino Acids: Digestibility, availability and metabolism. p. 125-132 in
Moughan, P. J., M.W.A. Verstegen, and M.I. Risser-Reyneveld (Ed.): Feed evaluation – Principles and practice.
Wageningen Press, Wageningen, The Netherlands.
Hurrell, R.F. and K.J. Carpenter, 1981. The estimation of available lysine in foodstuffs after Maillard reactions.
Prog. Food Sci. 5: 159-176.
Jansman, A.J.M., W. Smink, P. van Leeuwen, and M. Rademacher. 2002. Evaluation through literature data of the
amount and amino acid composition of basal endogenous crude protein at the terminal ileum of pigs. Anim.
Feed Sci. and Technol. 98:49-60.
Lewis, A.J. 1992. Determination of the amino acid requirements of animals. p. 67-85 in Nissen, S. (Ed.): Modern
methods in protein nutrition and metabolism. Academic Press Inc., San Diego, CA.
Maga, J.A. 1981. Measurement of available lysine using the guanidination reaction. J. Food Sci. 46:132-34.
Maillard, L.C. 1916. Synthéses des matiéres humiques par action des acides aminés sur les sucres réducteurs. Ann.
Chim. 6:258-317.
Maillard, L.C. 1912. Action des acides aminés sur les sucres: formation des mélanoidines per voie méthodologique.
Cr. Acad. Sci. 154:66-68.
Mosenthin, R., W.C. Sauer, R. Blank, J. Huisman, and M.Z. Fan. 2000. The concept of digestible amino acids in diet
formulation for pigs. Livest. Prod. Sci. 64:265-280.
Moter, V. and H.H. Stein. 2004. Effect of feed intake on endogenous losses and amino acid and energy digestibility
by growing pigs. J. Anim. Sci. 82:3518-3525.
Moughan, P.J. 1993. Towards an improved utilization of dietary amino acids by the growing pig. p. 117-136 in
Cole, D.J.A., W. Haresign, and P.C. Garnsworthy (Ed.): Recent developments in pig nutrition 2. Nottingham
University Press, Nottingham, UK.
Moughan, P.J. 2003a. Amino acid availability: aspects of chemical analysis and bioassay methodology. Nutrition
Research Reviews. 16:127-141.
Moughan, P.J. 2003b. Amino acid digestibility and availability in food and feedstuffs. p. 199-221 in Ball, R.O.
(Ed.): Digestive Physiology in pigs, Proc. 9th intl. Symposium. Vol. 1.
Moughan P.J. and S.M. Rutherfurd. 1996. A new method for determining digestible reactive lysine in foods. J. Agri.
Food Chem. 44:2202-2209.
NRC. 1998. P. 110–123 in Nutrient Requirements of Swine. 10th Rev. Ed. Natl. Acad. Press, Washington, DC.
Nyachoti, C.M., C.F.M. de Lange, B.W. McBride, and H. Schulze. 1997. Significance of endogenous gut nitrogen
losses in the nutrition of growing pigs: A review. Can. J. Anim. Sci. 77:149-163.
Qiao, Y., X. Lin, J. Odle, A. Whittaker, and T.A.T.G. van Kempen. 2004. Refining in vitro digestibility assays:
Fractionation of digestible and indigestible peptides. J. Anim. Sci. 82:1669-1677.
Sauer, W.C. and K. de Lange. 1992. Novel methods for determining protein and amino acid digestibilities in
feedstuffs. p. 87-120 in Nissen, S. (Ed.): Modern methods in protein nutrition and metabolism. Academic
Press Inc., San Diego, CA.
Stein, H.H. 2003. Digestible amino acids: Determination and use. p. 231-257 in Proc. 64th Minnesota Nutr. Conf.,
Sep. 16-17, 2003. Minneapolis, MN.
Stein, H.H. and M. Nyachoti. 2003. Animal effects on ileal amino acid digestibility. p. 223-241. In Ball, R.O. (Ed.):
Digestive Physiology in pigs, Proc. 9th Intl. Symposium, Vol. 1.
Stein, H.H., M.L. Gibson, C. Pedersen, and M.G. Boersma. 2005a. Amino acid and energy digestibility in ten
samples of distillers dried grain with solubles by growing pigs. J. Anim. Sci. 83:XXXX (Accepted).
Stein, H.H., C. Pedersen, and M.G. Boersma. 2005b. Energy and nutrient digestibility in dried distillers grain with
solubles by growing pigs. Abstract presented at the Midwestern Section, ASAS 2005 meeting, Des Moines, IA.
Abstract #199.
Stein, H.H., C. Pedersen, A.R. Wirth, and R.A. Bohlke. 2005c. Additivity of values for apparent and standardized
ileal digestibility of amino acids in mixed diets fed to growing pigs. J. Anim. Sci. 83:XXXX (In Press).
Tamminga, S., H. Schulze, J. van Bruchem, and J. Huisman. 1995. The nutritional significance of endogenous N-
losses along the gastro-intestinal tract of farm animals. Arch. Anim. Nutr. 48:9-22.