Background paper 5: amino acid composition in

relation to protein nutritional quality of meat and

poultry products12
Vernon

R Young,

PhD

and

Peter

L Pellett,

PhD

General

The physiological
role of dietary
proteins
is to provide
substrates
required
for the synthesis of body proteins
and other metabolically
important
nitrogen-containing
compounds.
The most demanding
aspect of the
amino acid utilization
offood
proteins,
however, relates
to protein
synthesis.
For this
reason,
the content
of the nutritionally
indispensable
amino
acids in a protein
or mixture of food proteins
is usually
a primary
determinant
of protein
nutritional
quality.
Hence,
knowledge
of the amino
acid cornposition
of foods serves as a basis for establishing their potential
nutritive
value. It may
also allow evaluation
of changes
in nutritive
value
that
may arise
in the preparation,
processing,
and storage
of foods.
Finally,
it
provides
a means
of predicting
the contribution
of a food protein
toward
meeting
human
amino
acid requirements.
In considering
protein
quality
evaluation
for regulatory
purposes,
the objective
seems
to be assurance
that a particular
food protein
meets
a given,
acceptable
standard.
This
standard
could be based on values obtained
from chemical
or in vitro assay procedures,
or it might be established
from the conduct
of a standardized
bioassay
methodology.
In
the discussion
that follows,
a case will be
developed
for the use of amino
acid data,
where regulation
of meat, poultry,
and their
products
includes
a concern
for protein
nutritional
quality.
This paper
has been developed
as an extension
of the analysis
given in Background
Paper 4, in view of the overlap
of the questions raised by the conveners
of the Expert
Work
Group
for consideration
by the authors of Background
Papers
4 and 5.

For purposes
of regulation,
chemical,
in
vitro enzymatic,
microbiological,
and shortterm rat bioassay
procedures
represent
alternative choices
for monitoring
protein
nutritional
value.
Table
1 lists traditional
and
newer methods.
It is not our intent to discuss these various
approaches
and specific
procedures
or their
applications
in detail,
since we (1 , 2) and
others (3-6) have done this on a number
of
occasions.
Rather,
it is more worthwhile
to
draw some general
points
in evaluating
the
use of amino
acid composition
in protein
quality
assessment
of meat
and
poultry
products.
First, because
a standardized
protein
efficiency
ratio
(PER)
procedure
has been
adopted
in the US (7) and Canada
(8), many
of the proposed
and alternative
methods
have been compared
with PER. As Bender
(9) has recently
pointed
out, this is unsatisfactory.
Indeed,
he, we (1), and others
(4,
10) have recommended
abandoning
PER as
a bioassay
method
and adopting
a procedure
that provides
values for protein
quality
that
are linear
or directly
proportional
to each
other over the range expected
in foodstuffs
(1 1). The relative
net protein
ratio as proposed by McLaughlan
et al (12) would
be a
suitable
choice.
Second,
the purpose
of regulating
the protein quality
of meat and poultry
products
is

The American
Journal of Clinical
Nutrition
40:
1984 American
Society for Clinical
Nutrition

SEPTEMBER

From

points:

assessment

the Department

of procedures

of Nutrition

and

Food

Sci-

ence (VRY),
Massachusetts
Institute
of Technology,
Cambridge
MA 02139
and the Nutrition
Program
(PLP),
Department
of Food
Science and Nutrition,
University
of Massachusetts,
Amherst,
MA 01002.
2Add
reprint
requests
to: Information
and Legislative Affairs, Food Safety and Inspection
Service, Rm
1160 South Building,
Washington,
DC 20250.
1984, pp 737-742.

Printed

in USA

737

Downloaded from www.ajcn.org by guest on March 29, 2011

Introduction

738
TABLE
1
Some methods
NBI

NGI

NU

RNU
NPR

RNPR
RPV

for measuring

protein

AND

quality*

(nitrogen
balance
index): tangent
of the
curve relating
N balance
to absorbed
N
(numerically
similar to BV)
(nitrogen
growth index): slope ofthe line
relating
weight gain to N intake (corresponds to NBI and NRP)
(nitrogen
utilization
value):
wt change
in 14 d plus 10% ofinitial
and final wt
to allow for maintenance
N
(relative
N utilization
value):
NU cxpressed as percentage
of lactalbumin
(net protein
ratio): wt gain of test group
plus wt loss ofnonprotein
group divided
by protein
consumed
(relative
net protein
ratio): NPR compared with reference
protein
(relative
protein
value): Regression
line
relating dose to response
for test protein
compared
with reference
protein
omitting zero protein
level
(protein
efficiency
ratio): wt gain per wt
ofprotein
consumed
in 28 days

Rapid biological
and in vitro methods
Red flour beetle (Tribolium
castaneum)
Confused
flour beetle ( Tribolium
confusum)
Aspergillusfiavus
Leuconostoc
mesenteroides
Insects
Escherichia
co/i
Tetrahymena,
modified
Mealworm
( Tenebrio
mo/itor)
PDR
(pepsin digest residue)
PPD
(pepsin pancreatin
index)
PPDD
(pepsin pancreatin
digest dialyzate)
C-PER
(computed
protein
efficiency
ratio)
* Adapted
from Bender (9).

assumed
to be maintenance
of their nutritional value (in the context
of human
nutrition). Thus,
the relationship
between
measures of protein
quality
and human
protein
nutrition
becomes
an important
consideration in establishing
standards
for regulatory
purposes.
This point is relevant
in considering the appropriateness
of a standard
for
mechanically
separated
(species)
that
requires a PER of no less than 2.5 (13).
There
is now sufficient
evidence
to indicate that soy protein
isolates
and other proteins, whose
standardized
PER values
are
less than 2.5 and as low as 1.8 (14, 15), have
a relatively
high protein
nutritional
value for
children
and adults.
Thus,
although
many
meat and poultry
products,
such as mechanically separated
(species),
achieve
minimum
PER values
of 2.5, as discussed
in Back-

PELLETT

ground
Paper 4, it would
not appear
to be
necessary
or even rational
to set 2.5 as a
minimum
value if PER were retained
as the
official
method
for protein
quality
assurance.
Third,
if a bioassay
procedure
such as
relative
net protein
ratio were accepted
as a
preferable
alternative
to PER, which indeed
it most certainly
is in the context
of bioassay
principles,
another
question
then
arises:
What should the standard
value be to assure
maintenance
ofthe protein
nutritional
value
of meat and poultry
products?
The answer
to this question
has to be judgmental;
it also
has to consider
the extent to which a lowering in the protein
nutritional
quality
of these
products
would
adversely
affect the overall
nutritional
quality
of diets chosen
by population
groups
considered
to be at risk for
dietary
protein
inadequacy.
Therefore,
we
will defer to the Expert
Work Groups
consideration
and judgment
on this complex
problem.
Fourth,
from the listing in Table
1 , it can
be seen that procedures
have been used as
alternatives
to rat assays and to those based
on amino acid composition
data-including
microbiological
assays and growth
methods
using
species
such
as beetles
and
mealworms.
There
are disadvantages
to all of
these alternative
methods
(1); more importantly,
none has been standardized
or sufficiently
investigated
to be suitable
for use in
verifying
protein
nutritional
quality
for regulatory
purposes.
Fifth, from the preceding
series of points
it emerges
that, at this time, a satisfactory
means ofassuring
protein
nutritional
quality
ofmeat
and poultry
products
might be based
on the amino
acid composition
of these
protein
foods.
In the following
sections
we
will examine
this possibility.
Procedures
composition

based
data

on amino

acid

The concept
of assaying
the nutritional
quality
of a protein
from a knowledge
of its
constituent
amino
acids was introduced
in
1946 by Block
and
Mitchell
(16). Their
method
was based on the relationship
they
observed
between
the biological
value and
amino acid composition
of various
food pro-

Downloaded from www.ajcn.org by guest on March 29, 2011

PER

YOUNG

REGULATING

PROTEIN

no evidence
in the published
literature
(19)
or supporting
documents
provided
to the
Expert
Work Group
that either the digestibility of meat and poultry
products
or damage to lysine and/or
sulfur amino
acids are
significant
factors
in the interpretation
of
amino
acid composition
data of these products.
A second general
problem
associated
with
the application
of amino
acid composition
data concerns
the reliability
ofthe analytical
procedures
used
to generate
these
data.
Again, this issue is discussed
in Background
Paper
4, and especially
in reference
to the
reliability
of values
for tryptophan,
sulfur
amino
acids, and hydroxyproline
content
of
meat, poultry,
and their products.
It cannot
be too strongly
emphasized
that there is an
urgent
need to establish
reproducible,
standardized
methods
for determining
and reporting
amino
acid data for these and other
food protein
sources.
For the following
discussion
we assume
that reliable
data are available
on the amino
acid composition
of meat and poultry
products, and that these data indicate
closely the
level of available
amino
acids in these foods.
Use of amino
purposes

acid

data

for regulatory

From
the foregoing
discussion
and the
published
literature,
it can be accepted
that
there is clear evidence
of a relationship
between
the amino
acid composition
of proteins and a measure
of their protein
nutritional
value (20). Indeed,
the analysis
undertaken
in Background
Paper 4 illustrates
the general
ability
of amino
acid composition data to serve as a predictor
of protein
nutritional
value.
Hence,
the use of amino
acid composition
data for regulatory
purposes and in the specific
context
of meat,
poultry,
and their products
will now be considered.
From
Background
Paper
4 a number
of
conclusions
can be drawn:
1 ) mechanically
separated
products
contain
more
collagen
(and elastin)
than muscle
or hand-deboned
meat. On the basis of the amino
acid composition
of collagen,
which contains
low levels of tryptophan,
sulfur
amino
acids,
and
lysine,
it would
be predicted
that collagen

Downloaded from www.ajcn.org by guest on March 29, 2011

teins. This approach

requires
the establishment of appropriate
reference
or standard
amino
acid patterns
as the basis for developing scoring
procedures.
(This topic is discussed in Background
Paper 4.) Thus, determination
of the amino
acid composition
of
a protein
and a comparison
with a reference
pattern
of amino
acids provides,
in theory,
a basis for ranking
proteins
according
to
their potential
nutritive
value; it also permits
a prediction
of a proteins
contribution
toward meeting
human
amino
acid requirements.
Thus,
knowledge
of the amino
acid
composition
of food proteins
is potentially
useful
for quality
assurance
and for assessment ofthe role ofa food protein
or product
in human
protein
nutrition.
However,
the
way in which amino
acid composition
data
are used will differ according
to whether
the
problem
is a matter
of regulation
or one of
evaluating
the efficacy
ofa protein
source in
human
diets.
Background
Paper 4 summarizes
the limitations
of the procedures
involving
application
of amino
acid composition
data as a
means
of determining
protein
nutritional
value.
These
points
need not be repeated
here in detail. However,
a number
of general
considerations
should
first be raised,
particularly in reference
to the analysis
ofthe data
presented
in that working
paper,
in order to
assess the specific
use to which amino
acid
data might be put in relation
to regulation
of meat, poultry,
and their products.
A general
concern
about
use of amino
acid composition
data is that the chemically
determined
content
ofamino
acids does not
necessarily
provide
an index of the content
ofavailable
amino
acids. For proteins
of low
digestibility,
the
chemically
determined
amino
acid content
would
overestimate
the
level of amino
acid available
for meeting
physiological
needs
and,
thus,
its protein
nutritional
quality.
Similarly,
processing
procedures
might
reduce
the amount
of
available
lysine (17) or sulfur
amino
acids
(18); and the reductions
would not necessarily be detected
by chemical
analysis
of the
total amino
acids present
in a food protein.
However,
this concern
does not appear
to
pose any serious
limitations
for the evaluation of protein
nutritional
quality
of meat,
poultry,
and their products.
Thus,
there
is

739

QUALITY

740

YOUNG

AND

cluded
that amino
acid composition
data
offer, at the very least, as sound
a basis as
rat bioassays
for developing
a regulation.
In the final section
we will, therefore,
propose a framework
for considering
use of
amino
acid composition
data if a regulation
is deemed
necessary
or desirable.
From
the observations
on the relationships between
amino
acid content
and PER
determinations
given in Background
Paper
4 and the comments
made above, a possible
approach
to be followed
in developing
an
amino
acid measure
of protein
nutritional
quality
for regulatory
purposes
might
consider the following.
1) Because
ofthe
quantitative
importance
ofthe extracellular
protein,
collagen,
in meat
and poultry
products,
and because
ofits low
protein
nutritional
quality,
it might be reasonable
to establish
an acceptable
upper
limit for collagen
in meat, poultry,
and their
products.
A more
persuasive
reason
for a
collagen
limit, however,
may be consumers
interest
in purchasing
muscle
as opposed
to
tendon
or other
tissues
high in collagen.
Thus,
hydroxyproline
(or proline)
determinations
might be a desirable
component
of
an amino
acid measure
of protein
nutritional value.
2) Although
the concentration
of total
essential
amino
acids
(EAA7
or EAA10)
yielded
equations
that were useful
in predicting
PER values
(Table
16, Background
Paper 4) and these equations
supported
the
protein
quality
requirement
set forth in the
regulation
on mechanically
separated
(species) (14), this does not necessarily
justify
use ofmeasurement
ofthese
amino
acids for
regulatory
purposes.
In fact, data on the total
essential
amino
acid content
of meat and
poultry
products
did not lead to an approach
superior
to that obtained
from a measure
of
the collagen
content.
Thus,
for regulatory
purposes,
total essential
amino
acid data per
se would not appear
to offer any advantage
or be necessary,
unless
analytical
advantages-with
respect
to quality
control
concerns-can
be gained
from determining
a
number
of amino
acids,
followed
by their
summation,
rather
than determining
only
one amino
acid.
3) For meat and poultry
that are mixed
with other
sources
of protein,
a chemical

Downloaded from www.ajcn.org by guest on March 29, 2011

would
be of low protein
nutritional
quality-a
point that is beyond
dispute.
Furthermore, a characteristic
of collagen,
compared
to the proteins
of muscle
fiber, is its high
concentration
of glycine,
proline,
and hydroxyproline.
It follows, then, that meat and
poultry
products
with a high hydroxyproline
concentration
would be of lower nutritional
value than comparable
products
with less
collagen.
This
prediction
is confirmed
in Background
Paper 4 (see Table
1 5), although
it
is worth noting that data shown
in the mdividual tables indicate
that wide variations
in
collagen
content
(varying
over a 2- or 3-fold
range) may have little effect on the PER of
the products
tested.
A partial
explanation
for this undoubtedly
relates to the errors and
variations
associated
with the determination
of collagen
and the estimation
of protein
quality
by means
of the PER technique.
It
would
be far more satisfactory
if data were
available
comparing
the collagen
content
of
meat and poultry
products
with a more reliable bioassay,
such as the relative
net protein ratio.
2) Statistically
significant
correlations
exist between
the content
of combinations
of
various
amino
acids (ie, leucine
and proline;
all seven indispensable
amino
acids) and the
nutritional
quality
as judged
by the PER
method.
It should be noted that the precision
of the predictions
of protein
nutritional
value
of individual
products
from amino
acid data is poor and, again,
probably
to a
significant
extent due to problems
arising in
the chemical
and biological
assay procedures
used for making
these comparisons.
In view
of the significant
drawbacks
of the PER
method,
against
which
all of these amino
acid data are compared,
it seems quite reasonable
to conclude
that use of amino
acid
data can be valid for regulatory
purposes.
3)
The relationships
between
available
data on the amino acid composition
of meat,
poultry,
and their products
and measures
of
protein
quality,
although
extensive,
are not
sufficient
to permit
a definitive
comparison
of amino
acid data versus
a simple
rat
assay procedure
as an approach
for regulatory purposes
at this time.
From this account
and those presented
in
the other background
papers,
it can be con-

PELLETT

REGULATING

Summary

and conclusions

The protein
nutritional
value
of meat,
poultry,
and their products
must be considered in relation
to humans.
Because
direct
assessment
of protein
nutritional
value
in
human
subjects
is impractical
for regulatory
purposes,
methods
based on in vitro (chemical) and animal
bioassays
for assessment
of
protein
quality
have
been
developed.
Herein,
a case is made for the use of amino
acid data as a basis for regulation
of meat
and poultry
products,
where considerations

QUALITY

741

of protein
nutritional
value are included
in
the regulation.
Current
data on the protein
nutritional
quality
of these products
are based
almost
entirely
on the use of the rat PER assay
procedure,
supplemented
by data
on the
amino
acid content
of these products.
Because of the inadequacy
of the PER assay,
the available
data base provides
an madequate
body
of knowledge
from
which
to
draw
final
recommendations
concerning
policies
for regulation
of the protein
nutritional quality
of meat and poultry
products.
However,
ifit is necessary
to include
a measure of protein
nutritional
quality
in a regulation,
these
data
provide
a reasonably
sound
basis for tentatively
recommending
use of amino
acid composition
data rather
than the standard
rat bioassay
procedure.
In such an approach
based on amino
acid
composition
data,
consideration
might
be
given to the contents
of hydroxyproline
(or
possibly
proline),
3-methylhistidine,
lysine,
and nitrogen,
with a view to establishing
a
composite
or minimum
value for the noncollagen
muscle
protein
content
of these
products.
El
References
1. Pellett PL, Young
VR, eds. Nutritional
evaluation
of protein
foods. Tokyo:
United
Nations
University,1980.
2. Pellett
PL. Protein
quality
evaluation
revisited.
Food Technol
1978;30:60-79.
3. McLaughlan
JM, Campbell
JA. Methodology
of
protein evaluation.
In: Munro HN, ed. Mammalian
protein
metabolism.
Vol 3. New York, NY: Academic Press, 1969:391-422.
4. Hegsted
DM. Assessment
of protein
quality.
In:
Harper
AE, Hegsted
DM, eds. Improvement
of
protein nutriture.
Washington,
DC: National
Academy of Sciences/National
Research
Council/Food
and Nutrition
Board,
1974:64-88.
5. Bodwell
CE, Adkins JS, Hopkins
DT, eds. Protein
quality
in humans:
Assessment
and in vitro estimation.
Westport,
CT: AVI Publishing,
1981.
6. Porter MWG,
Rolls BA, eds. Proteins
in human
nutrition.
New York, NY: Academic
Press, 1973.
7. Official methods
of analysis.
13th ed. Washington,
DC: Association
of Official
Analytical
Chemists,
I980:43.I95-.208.
8. Chapman
DO, Castillo
R, Campbell
iA. Evaluation of proteins
in food.
1. A method
for the
determination
of protein
efficiency
ratios. Can i
Biochem
Physiol
I 959;37:679-860.
9. Bender
AE. Evaluation
of protein
quality:
Meth-

Downloaded from www.ajcn.org by guest on March 29, 2011

measure
of the amount
of meat
(muscle)
might be desired.
In this context
it might be
worthwhile
to include
a measurement
of the
amino
acid NT-methylhistidine
(3-methylhistidine)
as an index of the content
of myosin and actin, the major
myofibrillar
proteins of muscle
fibers.
4) Because
the lysine content
of muscle
does not vary greatly
(see Table
3, Background
Paper
4), but is present
in much
lower amounts
in collagen,
it would appear
useful also to monitor
lysine content
in meat
and poultry
products.
An additional
advantage is that this amino
acid could be determined
rapidly
using the short column
for
basic amino
acids,
simultaneously
with 3methylhistidine.
5) A determination
of total nitrogen
and
the relationship
of these amino
acids would
provide
an index of the proportion
of total
protein
contributed
by collagen
(determined
as hydroxyproline
or proline),
muscle
fiber
proteins
(3-Mehis),
and total essential
amino
acids (from lysine).
This suggestion
appears
to warrant
further
attention.
6) The above suggestions
have been raised
in reference
to the specific
problem
of regulation and are not made in the broader
context ofthe
role of meat and poultry
products
in meeting
human
amino
acid
requirements-which
we have judged
was not our
mandate.
If this problem
were to be addressed,
it would
be necessary
to consider,
and in more detail,
tryptophan,
the sulfur
amino
acids, and possibly
threonine.
Furthermore,
human
amino
acid requirements
at various
ages would
need to be discussed
in detail.

PROTEIN

742

10.
11.

12.

13.

14.

AND

odological
considerations.
Proc Nutr
Soc 1982;
41:267-327.
Protein
Advisory
Group.
PAG guideline
no 6. Vol
4. New York, NY: PAG, United
Nations,
1974.
Harper
AE. Importance
of protein
quality
in the
United
States diet. In: Bodwell
CE, Adkins
iS,
Hopkins
DT, eds. Protein
quality
in humans:
Assessment
and in vitro estimation.
Westport,
CT:
AVI Publishing, 1981:19-28.
McLaughlan
JM, Anderson
OH, Hackler
LR, et al.
Assessment
of rat growth
methods
for estimating
protein
quality:
interlaboratory
study.
i AOAC
l980;63:462-7.
US Department
of Agriculture,
Food Safety and
Inspection
Service. Standards
and labeling
requirements
for mechanically
separated
(species)
and
products
in which it is used. Federal
Register.
June
29, l982;47:282l4-58
[9CFR
parts
317,
318,
319].
Young
VR, Scrimshaw
NS, Torun
B, Viteri
F.
Soybean
protein in human
nutrition:
An overview.
J Am Oil Chem Soc 1979:56:110-19.
Torun
B, Pineda 0, Viteri FE, Arroyave
0. Use of

PELLETT

16.

17.

18.

19.

20.

amino
acid composition
data to predict
protein
nutritive
value for children,
with specific reference
to new estimates
oftheir
amino acid requirements.
In: Bodwell
CE, Adkins
iS, Hopkins
DT, eds.
Protein quality in humans:
assessment
and in vitro
estimation.
Westport,
CT:
AVI
Publishing,
1981:374-93.
Block Ri, Mitchell
NH. The correlation
of the
amino-acid
composition
of proteins
with their nutritive value. Nutr Abstr Rev 1946;16:249-78.
Booth VH, Carpenter
Ki. Damage
to lysine in food
processing:
its measurement
and its significance.
Nutr Abstr Rev l973;43:423-5
1.
Cheftel JC. Chemical
and nutritional
modifications
of food proteins
due to processing
and storage.
In:
WhitakeriR,
Tannenbaum
SR, eds. Food proteins.
Westport,
CT: AVI Publishing,
1977:401-45.
Babji AS, Froning
OW, Satterlee
LD. Protein
nutritional
quality
of mechanically
deboned
poultry
meat as predicted
by the C-PER
assay. J Food Sci
l980;45:441-3.
Energy
and protein
requirements.
Geneva:
Food
and Agriculture
Organization/World
Health Organization
(technical
report series no 522), 1973.

Downloaded from www.ajcn.org by guest on March 29, 2011

15.

YOUNG