Meat By-product Protein Composition and Functional

Properties in Model Systems

R.O. NUCKLES, D.M. SMITH, and R.A. MERKEL

ABSTRACT creases, the need to understand, modify and control protein

Composition and functional properties of selected meat by-products functionality becomes more important (Smith, 1988).
were studied. Mechanicallv deboned chicken meat (MDCM) and meat Proteins are denaturedand aggregateinto a cross-linked three
by-products (pork lung lodes, pork liver, beef lung lobes, bdef spleen, dimensional gel matrix during cooking (Foegeding, 1988). Fat
beef heart) varied in their proximate composition, amount of the three and water are physically or chemically trapped inside this ma-
major protein fractions [low ionic strength soluble (LIS), high ionic trix. The texture and cook yield of restructured products are
strength soluble (HIS), insoluble] and collagen content. Cooked batter directly related to this matrix formation. The characteristics of
mode1 system reheat yield was positively correlated with the quantity the gel formed are a function of the type and amount of proteins
of HIS proteins and the percentagemyosin and actin in the HIS protein present and are influenced by environmental (e.g., pH, salt)
fraction. Stress and strain at failure of the cooked batters were in- and processing factors during heating (Whiting, 1988).
versely correlated to the amount of LIS proteins and positively cor-
related with the percentage of myosin and actin in the by-products. Meat by-products exhibit large differences in protein content
and distribution and consequently vary widely in protein func-
tionality and bind values. Bind values are based on the emul-
INTRODUCTION sion capacity of the soluble protein fraction. The type and
amount of meat by-products in a formulation influence the
MEAT BY-PRODUCTS are commonly classified as edible or texture, fat emulsifying capacity and water-holding properties
inedible (Oliveros et al., 1982). As a general definition, by- of the finished product. Each by-product contributes different,
products are “everything of economic value, except for the unique textural attributes to the finished product. Information
carcass,that are incidental to the slaughter of animals” (Booren on the functional and physicochemical properties of by-product
and Weiss, 1988). Meat by-products are underutilized and low- proteins is extremely limited. The protein content and com-
priced because they are regarded as an inferior protein source position of some beef by-products have been reported (Schae-
compared to skeletal muscle meat (Oliveros et al., 1982). The fer and Schierhorn 1974; Oliveros et al., 1982; Gorska et al.,
utilization of meat by-products has been avoided due to un- 1988), but the salt soluble protein and stroma protein content
desirable sensory quality, low biological value of the proteins of by-products from other species are not known. The objective
and high microbial contamination (Kosiba, 1983). Mishan- of this study was to develop a knowledge base of selected meat
dling, poor sanitation and slow chilling decrease quality and by-product composition and function for use in restructured
increase the microbial load of meat by-products (Booren and meat products.
Weiss, 1988). The increasing price of meat and processedmeat
products is causing the food industry to evaluate the utilization MATERIAL & METHODS
of all protein sources, including by-products (Gorska et al., Materials
1988).
Restructuring techniques, such as formed meat chunk tech- Frozen mechanically deboned chicken meat (MDCM) prepared from
nology, allow meat by-products to be utilized in value added the whole carcass was purchased from Nottawa Gardens Corp. (Ath-
ens, MI). By-products (pork lung lobes, pork liver, beef lung lobes,
products. Formed meat chunk technology is used to restructure beef spleen, beef heart) obtained from 15 market weight hogs or 15
meats that are difficult to incorporate into other further-processed market weight steers slaughtered on three different days at the Mich-
products. Meat chunks are formed using nonmeat binders. The igan State University Meat Laboratory were used in this study. Five
process is used commercially in Canada for human foods and animals were slaughtered each day and like by-products combined and
in the United Statesfor premium pet foods. Formed meat chunk designated as one replicate. By-products were removed within 1 hr
technology for human consumption is limited in the United after exsanguination. By-products were ground sequentially through
States due to consumer perceptions of undesirable sensory the 10, 6 and 3 mm plate of a Hobart grinder (Model A-200, Troy,
quality. Understanding the composition and function of by- OH), vacuum-packaged (2.5 mil polyethylene) and frozen until used
product proteins may allow for the replacement of nonmeat within three months. Proximate composition (fat, protein and mois-
ture) were determined by AOAC (1984) procedures. The pH was
binders with combinations of by-products to achieve the de- determined after blending log of MDCM’o; by-product with ‘1OOgof
sired bind, texture and mouthfeel. double distilled water on setting 3 for 1 min in a Sorvall Omni-Mixer
Further-processed meats utilize proteins as the principal (Model 17105, Newtown, CT7 and was measured using a Corning
functional and structural components. The proteins determine 145 pH meter. All experiments were performed in triplicate.
the characteristictexture, water-holding and appearanceof these
products (Hermansson, 1985). Protein gelation, fat binding Timed emulsification
and water-holding are the most important functional properties
Timed emulsification was performed as described by Galluzzo and
in further-processed products. Cooked product stability is cur- Regenstein (1978) with modifications because a cream layer was not
rently not thought to rely upon protein emulsification (Regen- formed using the recommended protein concentration of 5 mg/mL.
stein, 1988). The individual product, processing method and MDCM and by-products were evaluated in triplicate at total protein
stage of processing will determine the relative importance of concentrations of 10 to 150 mg/mL. The lowest concentration of pro-
each functional property. As the variety of meat products in- tein required to produce a cream layer after centrifugation at 30,000
x g for 15 min was recorded.

Authors Nuckles, Smith and Merkel are with the Dept. of Food Quantitation of protein fractions
Science & Human Nutrition, Michigan State Univ., East Lansing, AI1 extraction procedures were performed in a 4°C cold room. A
Ml 48824- 1224. weighed portion of MDCM or by-product was blended (Waring Blen-

640-JOURNAL OF FOOD SCIENCE-Volume 55, No. 3, 1990

 Table 1 -Proximate composition and collagen content of meat by-products
Collagen Collagen
Moisture Fat Protein (% of total (w&t
Sample 1%) I”/.) w protein) sample)
Mechanically deboned chicken meat 65.6~~ 14.2b 17.4b 39 6.8C
Beef heart (cap on) 66.5~~ 17.51 15.4= 5.5b 8.5b
Beef lung (lobes only) 79.7’b 1.9d 17.7b 5.5b 9.7’
Beef spleen 79.9Db 3.5’ 15.35 2.3d 3.5d
Pork liver 71.7b 1 .8d 22.11 1.5” 3.4d
Pork lung (lobes only) 82.5’ 2.0d 15.5c 6.1a 9.40
SEM 1.5 1.0 ., 0.4 0.3 0.4
~AvJ~~ Means within columns followed by the same letter do not differ significantly (PC 0.05)

dor, Model 1120, Winsted, CT) for 1 min with 4 volumes of 0.05M the reheated, cooled frankfurter weight divided by the original, cooked
Na phosphate buffer, pH 7.4, stirred for 3 hr (Talboy T-line Labo- frankfurter weight.
ratory Stirrer, Model 107, Emerson, NJ) and centrifuged at 23,000 x Apparent shear stressand strain at failure (Hamann, 1988) of model
g for 15 min. The supernatantwas saved and the residue re-extracted system frankfurters cores were measured using an Instrdn (Model
for 1 hr. Following centrifugation, the supernatants containing the 4202. Canton. MA) at a crossheadsoeed of 10 mm/min with a 50 N
0.0.5M Na phosphate soluble proteins were combined, designated as compression cell. The cores were compressed between two parallel
low ionic strength soluble (LIS) proteins and quantitated. The precip- plates. A 1.5 cm diameter central core was removed along the lon-
itate was mixed with 4 volumes of 0.6M NaCl, 0.0.5M Na phosphate gitudinal axis of the model system frankfurter using a cork borer and
buffer, pH 7.4 and extracted twice as described previously. The su- cores were cut to 1.5 cm lengths using a razor blade and template.
pernatants, designated as high ionic strength soluble (HIS) proteins, Cores were equilibrated to 1°C prior to testing. Six cores were eval-
were combined and quantitated. The precipitate was designated in- uated per replicate.
soluble proteins and weighed. The protein content of each fraction
was determined by Kjeldahl procedures (AOAC, 1984). Nonprotein
nitrogen was measured’on the low ionic strength soluble proteins ac- Statistics
cording to Helander (1957). Nonprotein nitrogen content was sub- The statistical design was completely randomized. Basic statistics,
tracted from the low ionic strength soluble nitrogen before calculating analysis of variance (ANOVA) and correlations were performed using
protein (N x 6.25). M-STAT (1988). Two way ANOVA was performed to test signifi-
The protein fraction percentageswere calculated by dividing the cance within replications and between treatments. Meat separations
weight of total protein in each fraction by the weight of total protein (significance) were tested using Tukey’s test.
in the meat. Percentageprotein in a weighed portion of each protein
fraction multiplied by the total fraction weight equaled the total protein
in each fraction. RESULTS & DISCUSSION
THE PROXIMATE COMPOSITION of MDCM and by-prod-
Collagen determination ucts is shown in Table 1. MDCM was used in this study be-
Collagen was determined in triplicate on each by-product, MDCM, cause it has been extensively studied, is readily available and
the HIS protein fraction and the insoluble protein fraction according is currently used in a wide variety of processedmeat products.
to Bergman and Loxley (1963). Prior to collagen determination, MDCM, In contrast, only a small amount of mechanically separatedred
by-products and insoluble proteins were frozen with dry ice and ground meats are produced in the U.S. and much of the production is
in an IKH-Universal mill M20 (Jankel and Kunkel Co., West Ger- used in pet foods (Field, 1988). Pork and beef lungs, beef
many).
spleen and pork liver generally contained higher moisture and
lower fat than beef heart and MDCM. Pork liver contained
Electrophoresis significantly more (PC 0.05) total protein than the other by-
Sodium dodecvl sulfate-polvacrvlamide gel electrophoresis (SDS-
products. Beef heart, beef spleen and pork lungs contained the
PAGE) of the HIS and insohrbg protein fracsons was performed‘using lowest (P < 0.05) protein content. The proximate composition
12% oolvacrvlamide aels accordina to Smith and Brekke f19851. The of beef heart (with cap) and pork liver are consistent with
values published by Porteous (1979) and Wiley et al. (1979).
.< < 1 I

insoluble protein fra;tions were solubilized prior to electrophoresis

according to Wu et al. (1982). One hundred micrograms of the HIS The percentage fat and moisture in beef heart published by
protein fraction and 200 ug of the insoluble protein extract were loaded Oliveros et al. (1982) is significantly lower becauseour analy-
on separate gels. Protein bands were stained with Coomassie Blue sis included the heart cap. Nonprotein nitrogen (NPN) content
(Sigma Chem. Co., St. Louis, MO). Myosin, actin and other protein of MDCM and by-products were not significantly different and
bands were identified by comparing relative retention times to molec- contained an average of 2.38% NPN. Collagen content (mg/g
ular weight standards(MW-SDS-200, Sigma Chemical Co., St. Louis, meat) was significantly higher (P < 0.05) in beef and pork lung
MO). Myosin:actin molar ratios of MDCM and by-product protein
fractions were calculated according to Beas et al. (1988). The protein compared to MDCM and other by-products. The beef heart
bands were quantitated by densitometric scans at 580 nm (Schimadzu and liver collagen contents were comparable to the published
CS-930, Kyoto, Japan). values of Porteous (1981). Khalili and Zarkadas (1988) re-
ported 1.96 to 3.08% collagen in chicken breast and 5.63 to
6.87% in leg muscles.
Model system frankfurters The protein distribution varied among the by-products ex-
Batters were prepared in triplicate using a silent cutter as described amined (Table 2). Mechanically deboned chicken meat con-
by Smith and Brekke (1984), except that the batter was stuffed into tained nearly twice as much HIS protein as meat by-products.
50 mL conical centrifuge tubes, capped and heated in a 75°C water The HIS protein percentagesfor beef heart, beef lung and beef
bath to an internal temperature of 72°C. Formulations were prepared spleen are lower than percentagesreported by Oliveros et al.
using a 5050 blend (based on the protein content) of MDCM and (1982). The differences may be explained by different extrac-
mea; by-product. Formulations were standardized to 56% moisture, tion times (7 days versus 4 hr). Beef heart and beef lung con-
30% fat, 12% protein and 2% salt by the addition of pork backfat or tained the highest quantity of insoluble proteins, while pork
ice.
Cook yield of the model system frankfurters was calculated by liver contained the highest quantity of LIS proteins. Beef heart
dividing the drained, cooked frankfurter weight (drained of cook-out contained the highest percentage of collagen in the HIS frac-
fat and moisture) by the original batter weight. Reheat yield was tion. The sum of the collagen in the HIS and insoluble protein
calculated after heating a 20 g sample in 100 mL of water at 95°C for fractions was consistent with the amount of total collagen in
10 min and cooling at room temperature for 5 min. Reheatyield equals the samples. Soluble collagen was measured in the HIS pro-

Volume 55, No. 3, 1990-JOURNAL OF FOOD SCIENCE-641

MEAT BY-PRODUCT PROTEIN COMPOSITION. ..
Table Z-Protein fractions as a percentage of total protein and collagen content of the low ionic strength soluble (L/S), high ionic strength soluble (HIS)
and insoluble protein fractions of by-products
Collagen in
LIS HIS Collagen in insoluble
Proteins Proteins HIS fraction Insoluble fraction
Sample (%I W) W) proteins w
Mechanically deboned chicken meat 28.V 40.4" 1.9' 31.0d 12.0b
Beef heart (cap on) 30.ld 21.2b 5.1' 48.7b lO.lf
Beef lung (lobes only) 36.3c 9.w 3.3* 54.1' lO.lC
Beef spleen 53.6b 21.7b 3.ld 24.70 8.0d
Pork liver 76.1' 14.8c 3.8" 9.1‘ 13.2'
Pork lung (lobes only) 51.6b 10.2d 4.3b 38.2~ 13.3'
SEM 2.7 1.6 0.2 2.4 0.3
a.b~.d.e.fMeans within columns followed by the same letter do not differ significantly (PC 0.05).

Molecular
Protein Weight
Myosin -
heavy
chain

-45,000
Actin -

Myosin
light
chains

uuuuuuu
1 2. 3 4 -5 6 7
Fig. 1 -Representative sodium dodecyl sulfate electrophoretegram of by-product salt soluble proteins (1 = beef heart, 2 = beef lung,
3 = pork lung, 4 = molecular weight standards, 5 = mechanically deboned chicken meat, 6 = pork liver, 7 = beef spleen)

teins because large amounts of collagen within this fraction Table 3-Percentage of myosin and ectin and the myosin:actin ratio of
the by-product high ionic strength soluble protein fraction
may adversely affect functionality.
A representative SDS-PAGE electrophoretegram of the by- Myosin Actin Actin:myosin
Samole 1%) I%) Mole ratio
product HIS proteins is shown in Fig. 1. Myosin or actin was
Mechanically deboned 50.3' 22.3" 5.0
not present on electrophoretegrams of the insoluble proteins. chicken meat
MDCM and beef heart had significantly higher (P < 0.05) per- Beef heart 47.9b 20.6b 5.0
centages of myosin and actin and a lower percentage of other (cap on)
proteins compared to the other by-products (Table 3). MDCM Beef lung 37&C 10.5c 3.2
(lobes only)
and beef heart contained hvice as much myosin and actin as Beef spleen 22.6d 10.15 4.5
beef spleen and pork liver. The significant differences in the Pork liver 24.4d 9.2d 4.3
percentages of myosin, actin and other proteins might be re- Pork lung 38.6" 10.6~ 3.0
lated to differences between skeletal, cardiac and smooth mus- (lobes only)
SEM 2.7 1.3
cle types. Myosin and actin are present in cardiac and smooth
*~b~c~dMean~ within columns followed by the same letter do not differ significantly
muscle and are extractable at high ionic strength (Garrels and
(PC 0.05).
Gibson, 1976; Cross et al., 1988). MDCM and beef heart had
lower myosin:actin molar ratios (1.0:5.0, Table 3) when com-
pared to the other by-products (1.0:3.0 to 1.0:4.5). Potter (1974)
and Murakami and Uchida (1985) reported that the molar ratio Greaser, 1983). Murakami and Uchida (1985) also reported a
of myosin to actin in skeletal muscle was l-0:6.0. The molar ratio of myosin to actin of 1.0:4.1 for cardiac muscle
myosin:actin ratio in rabbit skeletal muscle was 1:6.6 based and 1.0:16.5 for smooth muscle (chicken gizzard). Yasui and
on amino acid analysis rather than densitometry (Yates and Ishioroshi (1980) found that a weight ratio of myosin:actin of

642-JOURNAL OF FOOD SCIENCE-Volume 55, No. 3, 1990

Table 4-yield and textural characteristics of model system franMurters exhibited the lowest reheat yield. The pH values of the frank-
prepared from mechanically deboned chicken meat and meat by-prod-
ucts (1: 1 based on total protein contentjO
furter batters were not significantly different.
Changes in the modified timed emulsification test were pos-
Appar-
ent Appar- itively correlated with the quantity of by-product insoluble pro-
Cooked Reheat stress at ent teins and the percentage collagen contained within the HIS
yield yield failure strain at proteins and negatively correlated with the quantity of HIS
10~1 1%) IkPal failure
proteins (as HIS proteins increased, less protein was needed
Mechanically &boned 91.3d 82.6b 42.80b 0.76!= to form a stable cream layer, Table 5). The quantity of protein
chicken meat
Beef heart 92.5c 60.1C 42.22c 0.76’~
required to form a stable cream layer in the modified timed
(cap on)
emulsification test was not highly correlated with the percent-
Beef lung 93.0b 78.6d 20.35” 0.6gc age of by-product LIS proteins, myosin or actin within the HIS
(lobes only) proteins, collagen in the insoluble proteins or any parameters
Beef spleen 93.7b 74.08 13.63’ 0.536
Pork liver 91.2d 74.48 10.54g 0.566
measured in the model system frankfurter test.
Pork lungs 92.3c 69.8’ 17.37’ 0.69c Reheat yield was positively correlated with the quantity of
(lobes onlv) HIS proteins and the percentage of myosin and actin within
SCM 0.2 0.7 0.59 0.01 the HIS proteins fractions (Table 5). Reheat yield was in-
a Batter formulation: 12% protein, 30% fat, 56% moisture and 2% salt. versely correlated with LIS proteins. Apparent shear stressand
b.C.d.*.‘.g Means within columns followed by the fame letter do not differ significantly strain at failure of the model system frankfurters were nega-
(P-z 0.05)
tively correlated with LIS proteins and positively correlated
with HIS proteins and the percentageof myosin and actin within
the HIS fractions.
Table 5-Correlation coefficients between meat by-product protein com-
position and model system properties
The HIS proteins are generally considered to impart the most
functionality to processedmeats. In this study, larger quantities
Modified Apparent Apparent
Timed Reheat stress at strain at of HIS proteins were related to larger reheat yields and better
Emulsification Yield failure failure textural properties in the model system frankfurters. The quan-
LIS proteins -0.31 - 0.71 - 0.85 -0.82 tity of insoluble proteins in the meat by-productswas not strongly
HIS proteins -0.71 0.78 0.80 0.61 correlated with the parameters measured in the frankfurters.
Mvosin in 0.36 0.64 0.93 0.98 Timed emulsification was not an adequatetool to evaluate by-
HIS fraction
Actin in -0.07 0.79 0.98 0.74
products.
HIS fraction These results indicate that the protein distribution of meat
Collagen in 0.72 -0.37 -0.07 0.08 by-products was related to the yield and texture imparted to
HIS fraction comminuted meat products. Results suggest that possibly a
Insoluble 0.79 0.16 0.14 0.30
proteins
certain percentageof HIS proteins or ratio of protein fractions
Insoluble 0.04 -0.24 -0.04 0.24 are necessary to produce the desired bind in frankfurter for-
proteins, mulations.
% collagen
Apparent -0.09 0.63 0.83
strain at
failure
REFERENCES
Apparent -0.35 0.84
stress at AOAC. 1984. Official Methods of Analyses 14th ed. Association of Official
failure Analytical Chemists, Washington, DC.
Asghar: A. Samejima, K., and Yasui, T. 1985. Functionality of muscle
Reheat vield - 0.41 rotems in gelation mechanisms of restructured meat products. Crit.
F(av. Food Sci. Nutr. 22: 21.
Beas, V.E., Crupkin, M., and Truccq, R.E. 1988. Gelling pro erties of ac-
tomyosin from pre- and post-spawnmg Hake (Merluccius hu ii bsl). J. Food
Sci. 53: 322.
15 formed more rigid gels, whereas Asghar et al. (1985) stated Bergman, I. and Loxloy, R. 1963. Two improved and simplified methods
that the myosin:actin ratio rather than the actomyosin:actin foo;ptrophotometric determmation of hydroxyproline. Anal. Chem. 35:
ratio may be related to gel strength. Booren’, AM. and Weiss, G.M. 1988. Lean skeletal meat trimmings inci-
The timed emulsification procedure of Galluzzo and Regen- dental to slaughter. Ch. 9. In Edible Meat By-Products. Advances in
stein (1978), developed to study the emulsifying properties of Meat Research Vol. 5. A.M. Pearson, and T.R. Dutson, (Ed.), p. 219.
Elsevier Ap 1. Sci. Publ., New York.
skeletal meat proteins, was modified to evaluate meat by-prod- Cross, R.A., 8.iti, S., and Kendrick-Jones, J. 1988. How phosphorylation
ucts. The minimum protein concentration necessaryto produce controls the self-assembly of vertebrate smooth and non-muscle proteins.
Biochem. Sot. Trans. 16: 501.
a cream layer after centrifugation was used as an index of Field, R.A. 1988. Mechanical1 separated meat, poultry and fish. Ch. 4 In
protein functionality among the by-products. MDCM, beef Edible Meat By-Products, A l vances in Meat Research Vol. 5. A. M. Pear-
son, and T. R. Dutson, (Ed.), p. 83. Elsevier Appl. Sci. Publ., NY.
spleen and pork liver formed cream layers (10, 20 and 40 mg Foegeding, E. A. 1988. Thermally induced changes in muscle proteins.
protein/ml, respectively) but beef heart, beef lung and pork Food Technol. 42(6): 58.
lung did not form cream layers at protein concentrations up to Galluzzo, S.J. and Regenstein, J.M. 1978. Emulsion capacity and timed
emulsification of chicken breast muscle myosin. J. Food Sci. 43: 1757.
150 mg/mL. Oliveros et al. (1982) reported that beef spleen Garrels, J. I. and Gibson, W. 1976. Identification and characterization of
emulsified more oil than beef heart or beef lung. multiple forms of actin. Cell 9: 793.
Gorska, I., Szmanko, T., and Krasnowska, G. 1988. Preliminary physico-
Cooked yield, reheat yield, apparent shear stress and ap- chemical and histological characteristics of beef gullet meat tissue. Food
parent strain at failure of model system frankfurters were in- Chem. 27: 131.
Hamann, D.D. 1988. Rheology as a means of evaluating muscle function-
fluenced by meat by-product protein composition (Table 4). ality of processedfoods. Food Technol. 42(6): 66.
Apparent strain is highly correlated to sensory and texture pro- Helander, E. 1957. On quantitative muscle protein determination, Acta
file analysis (TPA) cohesivenesswhile apparent stress at fail- Physiol. &and. 41: 141.
Hermansson, A.M. 1985. Water-and fat holding. Ch. 6. In Functional Prop-
ure is correlated to TPA hardness and sensory firmness er& ofFood Macromolecules, J.R. Mitchell and D.A. Ledward, (Ed.), p.
(Montejano et al., 1985). Model system frankfurter percentage 273. Elsevier Ap 1. Sci. Pub., New York.
Khalili. A.D. and 8 arkadas. C.G. 1988. Determination of mvoiibrillar and
cooked yield ranged from 91.2 to 93.7% while reheat yield connective tissue protein contents of youn and adult a&n (Gallus do-
ranged from 69.8 to 82.6%. The model system frankfurters mesticus) skeletal muscles and the N-met 5 ylhistidine
. content of avian
actins. Poultry Sci. 67: 1593.
produced with by-products had significantly lower (P< 0.05) Kosiba, E. 1983. Uses of animal products from the animal industry. Gas
percentage reheat yield and compressive force at failure com- Miesna. 3:Q. Quoted in Gorska, I., Szmanko, T., and Krasnowska, 6:
pared to MDCM frankfurters. Among the by-product frank- 1988. Preliminary physiocochemical and histological characteristics of
beef gullet meat tissue. Food Chem. 27: 131.
furters, beef heart had the highest reheat yield and pork lung Montejano, J.G. Hamann, D.D., and Lanier, T.C. 1985. Comparison of two
-Continued on page 682
Volume 55, No. 3, 19904OURNAL OF FOOD SCIENCE-643