1
FEB
09
2005
Position
Paper:
Scientific
Issues
Associated
with
the
Human
Health
Assessment
of
the
Cry34Ab1
Protein
Introduction
Dow
AgroSciences
(
Dow)
and
Pioneer
Hi­
Bred
International,
Inc.
(
Pioneer)
have
submitted
applications
for
FIFRA
§
3
registration
of
the
plant­
incorporated
protectant
(
PIP)
Bacillus
thuringiensis
(
Bt)
Cry34Ab1
and
Cry35Ab1
proteins
and
the
genetic
material
necessary
for
their
production
in
corn.
These
products
are
intended
to
provide
protection
against
western
and
northern
corn
rootworm
larvae.

Reviews
have
been
completed
on
product
characterization
and
human
health
and
can
be
found
in
memoranda
dated
December
6,
2004
and
February
4,
2005.
A
preliminary
safety
assessment
is
presented
in
the
memo
dated
February
4,
2005.

Since
Cry34Ab1
and
Cry35Ab1
are
proteins,
allergenic
potential
was
considered
in
the
safety
assessment.
EPA
uses
a
weight
of
evidence
approach
suggested
by
scientists
at
the
1994
Allergenicity
Conference,
hosted
by
EPA,
the
Food
and
Drug
Administration,
and
the
U.
S.
Department
of
Agriculture,
where
characteristics
of
a
protein
are
compared
with
characteristics
of
known
allergens.
More
recently,
this
approach
was
outlined
in
the
Annex
to
the
Codex
Alimentarius
"
Guideline
for
the
Conduct
of
Food
Safety
Assessment
of
Foods
Derived
from
Recombinant­
DNA
Plants."
Currently,
no
definitive
tests
exist
for
determining
the
allergenicity
potential
of
proteins.
EPA
considers
the
following
factors
to
provide
assurance
that
a
protein
is
unlikely
to
elicit
an
allergic
reaction:
1)
whether
the
source
of
the
trait
is
associated
with
any
reports
of
allergenicity;
2)
amino
acid
sequence
comparison
with
known
allergens,
both
overall
similarity
and
stepwise
contiguous
amino
acid
searches;
3)
biochemical
properties
of
the
protein,
including
in
vitro
digestibility
in
simulated
gastric
fluid
(
SGF),
heat
stability,
and
glycosylation;
4)
prevalence
in
food;
and
5)
specific
serum
screening
(
e.
g.,
for
proteins
with
sequence
similarity
with
a
known
allergen).
Because
no
single
factor
is
fully
predictive,
EPA
considers
all
of
the
available
information
in
the
assessment.

As
part
of
their
weight
of
evidence
assessment,
Dow
and
Pioneer
submitted
several
in
vitro
digestibility
studies
in
support
of
their
applications
for
registration
and
tolerance
exemption
of
Cry34Ab1
and
Cry35Ab1.
From
analysis
of
these
studies,
as
well
as
published
studies
and
previous
guidance
from
Scientific
Advisory
Panels,
EPA
has
concluded
that
Cry35Ab1
is
rapidly
digested
and
Cry34Ab1
is
moderately
digested
in
simulated
gastric
fluid
(
SGF).
Cry34Ab1
appears
to
digest
slower
than
other
Bt
proteins
that
have
been
registered
for
food
use
and
many
other
proteins
that
are
not
considered
allergens
but
faster
than
most
previously
tested
allergens.
Dow
and
Pioneer
also
submitted
data
indicating
that
both
Cry34Ab1
and
Cry35Ab1
are
inactivated
by
heat,
are
not
glycosylated,
do
not
have
any
sequence
similarity
to
known
allergens,
and
will
only
be
present
at
low
levels
in
food.
To
further
analyze
the
digestion
of
Cry34Ab1,
the
registrants
have
developed
a
kinetic
approach
2
to
assess
protein
degradation
as
part
of
their
weight
of
evidence
evaluation
of
Cry34Ab1.
EPA
is
asking
the
Panel
to
comment
on
1)
the
usefulness
of
the
kinetic
approach
for
moderately
digestible
proteins
and
what
assay
conditions
are
appropriate
for
comparing
the
digestion
of
different
proteins;
2)
how
digestion
assays
should
be
used
in
the
overall
weight
of
evidence
approach
to
allergenicity
assessment;
and
3)
EPA's
allergenicity
assessment
of
Cry34Ab1.

Background
Pepsin
digestibility
is
the
focus
of
this
position
paper
and
Scientific
Advisory
Panel
(
SAP)
meeting
because
it
is
the
only
significant
issue
that
has
arisen
during
the
evaluation
of
Cry34Ab1
and
Cry35Ab1.
A
correlation
between
resistance
to
in
vitro
digestion
by
the
enzyme
pepsin
and
allergenic
potential
has
been
demonstrated
(
Astwood,
et
al.,
1996).
Therefore,
pepsin
digestibility
is
one
component
that
is
considered
as
part
of
the
weight
of
evidence
approach
for
assessing
the
allergenicity
potential
of
proteins.
The
correlation
between
resistance
to
digestion
and
allergenicity,
however,
is
not
absolute:
some
allergens
are
rapidly
digested
and
some
nonallergens
(
i.
e.,
proteins
not
known
to
cause
allergic
reactions)
appear
to
be
resistant
to
digestion
(
Fu,
et
al.,
2002).

One
rationale
for
using
digestibility
assays
in
allergenicity
assessments
is
the
belief
that
proteins
that
are
rapidly
degraded
in
the
digestive
system
are
less
likely
to
induce
an
immune
response.
It
is
generally
accepted
that
a
protein
must
be
stable
in
the
stomach
for
a
sufficient
period
of
time
to
sensitize
an
individual.
Small
peptides
are
believed
to
be
incapable
of
causing
allergic
sensitization.
However,
there
are
reports
of
cases
where
digested
proteins
were
capable
of
eliciting
an
allergic
response
in
individuals
who
have
already
been
sensitized
(
Nilsson,
et
al.,
1999).
Some
researchers
have
suggested
that
stability
to
pepsin
digestion
may
reflect
resistance
to
cleavage
of
the
protein
by
intracellular
proteases
during
processing
for
presentation
to
T
lymphocytes
and
that
inherent
susceptibility
of
proteins
to
enzymatic
digestion
may
influence
the
nature
of
the
immune
response
and
therefore
whether
allergic
sensitization
will
develop
(
Dearman,
et
al.,
2002).

In
Vitro
Digestibility
Assays
Different
protocols
exist
for
assessing
a
protein's
susceptibility
to
digestion
by
pepsin.
The
conditions
used
in
assessing
in
vitro
digestibility
are
important
because
a
protein
can
appear
to
be
resistant
to
digestion
or
rapidly
digested
depending
on
the
pH
of
the
SGF,
the
ratio
of
pepsin
to
substrate
protein,
the
concentrations
of
pepsin
and
substrate
protein,
the
purity
of
the
proteins,
the
specific
activity
of
the
pepsin,
and
the
sensitivity
of
the
detection
method
(
Thomas,
et
al.,
2004).
Typically,
the
SGF
is
prepared
as
specified
in
the
U.
S.
Pharmacopeia
(
USP),
giving
a
pH
of
1.2
and
a
pepsin
concentration
of
3.2
mg/
mL.
However,
the
concentration
of
the
substrate
protein
is
not
specified,
and
the
specific
activity
for
pepsin
has
only
been
specified
in
recent
editions
of
the
USP.
In
addition,
a
2001
Joint
FAO/
WHO
Expert
Consultation
on
Allergenicity
of
Foods
Derived
from
Biotechnology
report
titled
"
Evaluation
of
Allergenicity
of
Genetically
Modified
Foods"
provides
a
protocol
for
conducting
digestibility
assays
where
the
pH
of
the
SGF
is
2.0,
rather
than
1.2.
This
protocol
also
specifies
the
pepsin
concentration
as
0.32%
(
w/
v)
3
(
3.2
mg/
mL)
and
the
substrate
protein
concentration
as
2.5
mg/
mL
(
500

g
in
200

L
SGF).
The
FAO/
WHO
protocol,
however,
has
not
been
tested,
so
it
is
unknown
whether
or
not
a
correlation
between
allergenicity
and
digestibility
would
be
observed
with
this
protocol.
In
addition,
there
is
no
database
of
digestion
times
for
known
allergens
and
non­
allergens
for
comparison
with
new
proteins
tested
under
the
conditions
of
this
protocol.
Recently,
a
number
of
organizations,
companies,
and
government
researchers
published
a
joint
study
where
the
reproducibility
of
a
common
protocol
was
tested
across
multiple
laboratories
(
Thomas,
et
al.,
2004).
This
protocol
used
SGF
solutions
of
both
pH
1.2
and
2.0
with
a
final
pepsin
concentration
of
0.72
mg/
mL
(
3,460
units/
mg)
and
a
ratio
of
10
units
of
pepsin
activity
per

g
test
protein
(
3:
1
pepsin
to
protein,
w/
w).
Registrants
have
used
a
variety
of
conditions
for
testing
in
vitro
digestibility
of
currently
registered
PIPs
(
see
Table
1).

Typically,
the
time
it
takes
for
the
test
protein
or
its
digestion
fragments
to
become
undetectable
is
monitored.
Two
approaches
have
been
used:
1)
separate
reactions
are
set
up
for
each
of
the
time
points
and
quenched
by
the
addition
of
base
at
the
appropriate
time,
or
2)
a
single
reaction
is
set
up
for
each
replicate,
and
samples
are
removed
at
various
time
points
and
quenched
by
the
addition
of
base.
The
samples
are
then
subjected
to
sodium
dodecyl
sulfate
polyacrylamide
gel
electrophoresis
(
SDS­
PAGE),
and
the
protein
bands
on
the
gels
are
visualized
either
by
staining
the
gel
or
using
western
blot
analysis.

Kinetic
Approach
Dow
has
developed
a
new
kinetic
approach
for
assessing
a
protein's
in
vitro
digestibility
based
on
estimating
the
rate
of
pepsin
digestion.
1
Dow
uses
a
protocol
similar
to
those
described
above.
A
single
reaction
is
set
up
for
each
replicate,
and
the
test
proteins
are
incubated
in
SGF
containing
pepsin
at
a
concentration
of
0.3
%
(
w/
v),
pH
1.2,
at
37

C,
with
constant
shaking.
Samples
are
removed
at
various
time
points
and
analyzed
by
SDS­
PAGE.
After
the
gels
are
stained,
the
relative
amount
of
protein
or
protein
fragment
remaining
at
each
time
point
is
monitored
by
determining
the
band
density
using
gel
densitometry.
The
band
density
is
assumed
to
be
directly
proportional
to
the
protein
concentration.
The
degradation
over
time
is
then
assessed
using
a
first­
order
(
exponential)
decay
model.
For
Cry34Ab1,
Dow
used
linear
regression
of
the
natural
logarithm
of
the
percent
protein
remaining
versus
time
to
determine
the
first­
order
rate
constant
(
MRID
455845­
02;
Herman,
et
al.,
2003).
In
subsequent
analyses,
Dow
has
used
non­
linear
regression
of
a
3­
parameter
exponential
model
([
S]
=
[
S0]
e­
KT
+
B,
where
[
S]
is
the
band
density
at
time
T,
[
S0]
+
B
is
the
Y­
intercept,
K
is
the
first­
order
rate
constant,
and
B
is
the
asymptote
or
background
estimate)
to
determine
the
first­
order
rate
constant
(
MRID
463886­
01).
Half­
lives
are
then
calculated
by
dividing
the
natural
logarithm
of
0.5
by
the
firstorder
rate
constant
for
each
protein
or
fragment.

1The
kinetic
approach
to
pepsin
digestion
was
developed
by
scientists
at
Dow
AgroSciences.
Applications
by
both
Dow
and
Pioneer
for
registration
of
their
Cry34Ab1
and
Cry35Ab1
products
rely
on
these
data.
Because
Dow
scientists
developed
this
approach,
the
explanation
of
the
kinetic
approach
only
mentions
Dow.
4
Dow
has
asserted
that
first­
order
decay
is
predicted
based
on
enzyme
theory
as
long
as
the
pepsin
concentration
is
high
and
the
substrate
concentration
is
low
(<<
Km).
Dow
has
also
stated
that
the
first­
order
rate
constant
determined
under
these
conditions
is
equal
to
Vmax/
Km,
which
is
a
measure
of
the
inherent
efficiency
of
an
enzymatic
reaction.
As
long
as
first­
order
conditions
are
met,
first­
order
rate
constants
and
half­
lives
are
unaffected
by
changes
in
protein
substrate
concentration.
Therefore,
first­
order
rate
constants
can
be
used
to
predict
relative
digestion
efficiencies
for
proteins,
even
if
protein
concentrations
are
varied
among
experiments.
In
addition,
Dow
has
stated
that
the
enzyme
concentration
is
saturating
when
the
USP
concentration
of
pepsin
(
0.32%)
is
used,
making
the
rate
constants
relatively
insensitive
to
changes
in
enzyme
concentration
(
Herman,
et
al.,
2005).
Dow
has
stressed
that
the
kinetic
approach
should
improve
the
accuracy
of
determining
pepsin
digestibility,
and
relying
on
the
disappearance
of
a
protein
band
on
a
gel
to
measure
digestibility
of
proteins
as
an
endpoint
may
be
problematic.
The
disappearance
results
depend
on
a
number
of
factors
including
the
affinity
of
different
proteins
for
the
dye
or
antibody
used
to
visualize
the
gel,
the
amount
of
protein
loaded
on
the
gel,
type
of
gel
and
dye
used,
and
development
time.

EPA's
Assessment
of
the
Kinetic
Approach
A
method
for
assessing
pepsin
digestion,
such
as
the
kinetic
approach,
that
is
not
dependent
on
detection
method
may
be
an
improvement
over
relying
on
the
substrate
disappearance
endpoint
in
cases
where
digestion
does
not
appear
to
be
rapid.
However,
EPA
is
confident
that
the
endpoint
method
that
has
been
used
for
assessing
digestion
of
previously
registered
proteins
has
been
adequate.
In
addition,
caution
is
warranted
in
interpreting
results
and
comparing
the
kinetic
rates
of
pepsin
digestion
of
different
proteins
determined
using
the
kinetic
approach.
Many
of
the
same
factors
that
affect
the
endpoint
method
can
also
affect
the
digestion
rates
such
as
enzyme
activity,
purity
of
protein
substrate,
concentrations
of
pepsin
and
substrate,
pH,
temperature,
and
whether
or
not
the
reaction
mixture
is
shaken
or
stirred.
Also,
the
digestion
rates
calculated
using
Dow's
approach
depend
on
the
fit
to
first­
order
kinetics.
Pepsin
hydrolysis
is
a
multi­
step/
multi­
reaction
process,
and
the
kinetics
may
not
be
simple
to
predict.

Pepsin
has
been
shown
to
hydrolyze
proteins
using
different
mechanisms,
depending
on
reaction
conditions
(
Choisnard,
et
al.,
2002)
and
presumably
depending
on
the
protein
substrate.
One
pepsin
molecule
can
degrade
one
protein
substrate
molecule
at
a
time
(
i.
e.,
the
one­
by­
one
or
processive
mechanism);
a
pepsin
molecule
can
also
move
from
one
protein
substrate
molecule
to
another,
cleaving
as
it
goes,
generating
intermediate
peptide
products
(
i.
e.,
the
zipper
mechanism);
or
it
can
use
a
mechanism
that
is
in
between
these
two
extremes.
Choisnard
et
al.
(
2002)
demonstrated
that
at
pH
4.5,
pepsin
hydrolyzed
native
hemoglobin
using
the
one­
by­
one
mechanism,
while
pepsin
hydrolysis
of
denatured
hemoglobin
proceeded
by
the
zipper
mechanism.
It
is
unclear
whether
the
rate
of
decay
of
the
starting
substrate
would
follow
first
order
kinetics
regardless
of
the
mechanism.
It
likely
depends
on
what
the
rate­
limiting
step
of
the
reaction
is,
which
may
depend
on
reaction
conditions
and
the
protein
substrate.
Also,
during
pepsin­
catalyzed
hydrolysis
of
many
proteins,
intermediate
digestion
fragment
peptides
are
formed.
Presumably,
in
some
cases,
these
fragments
could
compete
with
starting
substrate
and
5
inhibit
the
rate
of
decay
of
starting
substrate;
the
decay
rate
might
not
follow
first
order
kinetics
for
these
cases.

Dow
and
Pioneer
have
submitted
two
comparison
studies
using
a
number
of
allergens
and
nonallergens
(
i.
e.,
proteins
that
are
not
known
to
cause
allergic
reactions)
to
test
the
kinetic
approach.
The
first
submitted
study
(
MRID
461239­
20,
reviewed
in
the
memorandum
from
R.
Edelstein
to
M.
Mendelsohn
dated
August
17,
2004)
was
designed
for
a
different
purpose
and
used
conditions
(
pH
1.2
and
2.0;
pepsin
concentration:
0.72
mg/
mL;
test
protein
concentration:
0.25
mg/
mL;
pepsin:
test
protein
ratio:
3:
1,
w/
w)
that
did
not
allow
comparison
with
the
previously
submitted
digestion
study
on
Cry34Ab1.
Most
of
the
proteins
digested
either
too
quickly
or
too
slowly
for
their
digestion
rates
to
be
determined.
Of
the
digestions
that
were
analyzed
kinetically,
some
demonstrated
good
fit
to
first­
order
kinetics,
while
some
had
poor
fit.
The
study
showed
a
correlation
between
resistance
to
pepsin
digestibility
and
allergenicity.
The
second
comparison
study
(
MRID
463886­
01)
used
the
same
conditions
that
were
used
in
the
digestibility
studies
on
Cry34Ab1
(
pH
1.2,
pepsin
concentration:
0.32%,
test
protein
concentration:
0.002
mM,
pepsin:
test
protein
ratio
of
approximately
20:
1,
mol/
mol).
In
this
second
comparison
study,
half­
lives
for
the
proteins
tested
ranged
from
<
30
seconds
to
>
60
minutes.
Allergens
tended
to
be
more
stable
in
SGF
than
non­
allergens,
but
a
strong
correlation
between
digestion
rate
and
allergenicity
was
not
observed
for
the
set
of
proteins
tested
(
seven
allergens
and
eight
non­
allergens).
However,
to
determine
half­
lives
using
Dow's
protocol,
the
test
protein
must
be
stable
enough
in
SGF
for
it
to
be
measured
over
several
time
points.
Therefore,
the
non­
allergens
tested
in
this
study
were
those
known
to
digest
slower
than
many
other
previously
tested
non­
allergens.
The
data
fit
well
to
a
first­
order
decay
model,
except
for
early
time
points,
and
half­
lives
calculated
using
initial
substrate
concentrations
that
differed
by
5­
fold
were
fairly
consistent.

While
Dow's
kinetic
approach
is
less
dependent
on
detection
method
than
the
end­
point
method
typically
used
for
assessing
pepsin
digestibility,
it
is
only
applicable
to
proteins
with
moderate
digestibility.
Because
of
the
high
enzyme
and
low
substrate
concentrations
used,
many
test
proteins
are
digested
so
quickly
that
they
are
undetectable
at
the
first
time
point.
In
addition,
the
method
depends
on
the
fit
to
first­
order
decay.
While
the
digestion
of
a
number
of
proteins
appears
to
fit
the
model,
in
the
comparison
study
described
above,
early
time
points
for
several
of
the
proteins
were
omitted
to
obtain
a
good
fit
to
the
model.

Dow
has
stated
that
the
first­
order
rate
constant
obtained
under
these
conditions
is
equal
to
Vmax/
Km.
It
is
well­
known
that
the
first­
order
rate
constant
from
a
substrate
concentration­
time
profile
is
equal
to
Vmax/
Km
when
the
initial
substrate
concentration
is
much
less
than
Km,
and
catalytic
quantities
of
enzyme
are
used
(
Segel,
1975).
However,
under
conditions
of
high
enzyme
concentration,
although
it
appears
pseudo­
first­
order
kinetics
is
still
predicted
(
Schnell
and
Mendoza,
2004;
Schnell
and
Maini,
2000;
Tzafriri,
2003),
it
is
unclear
whether
the
firstorder
rate
constant
equals
Vmax/
Km.
The
standard
assumptions
(
e.
g.,
steady­
state
assumption:
d[
ES]/
dt
~
0)
used
to
derive
rate
equations
for
enzyme
reactions
do
not
apply
under
conditions
of
high
enzyme
concentration
(
Segel,
1975;
Schnell
and
Mendoza,
2004;
Schnell
and
Maini,
2000;
Tzafriri,
2003).
Different
assumptions
must
be
made
to
derive
rate
equations
under
these
6
conditions.
In
addition,
Dow
did
not
determine
Km
values
for
any
of
the
proteins
tested.
Therefore,
it
is
unknown
whether
the
concentrations
used
are
much
less
than
the
Km
values
for
each
of
the
proteins.

EPA's
Assessment
for
Cry34Ab1
and
Cry35Ab1
Data
have
been
submitted
demonstrating
the
lack
of
mammalian
toxicity
at
high
levels
of
exposure
to
pure
Cry34Ab1
and
Cry35Ab1
proteins.
Three
acute
oral
toxicity
studies
on
Cry34Ab1
and
Cry35Ab1
in
mice
were
submitted:
1)
Oral
toxicity
of
Cry34Ab1
alone
(
MRID
452522­
07);
2)
Oral
toxicity
of
Cry35Ab1
alone
(
MRID
452522­
08);
and
3)
Oral
toxicity
of
Cry34Ab1
and
Cry35Ab1
combined
(
MRID
452522­
09).
All
animals
survived
the
two­
week
studies,
and
no
treatment­
related
effects
were
observed.
The
results
of
these
studies
demonstrate
the
safety
of
the
proteins
at
levels
well
above
maximum
possible
exposure
levels
that
are
reasonably
anticipated
in
the
crops.

Since
Cry34Ab1
and
Cry35Ab1
are
proteins,
allergenic
potential
was
also
considered.
Several
in
vitro
digestibility
studies
were
submitted.
In
the
first
study
(
MRID
452422­
12),
Cry34Ab1
and
Cry35Ab1
were
incubated
in
SGF
(
pepsin
concentration:
0.3
%
(
w/
v);
pH
1.2;
37

C)
with
a
pepsin
to
protein
substrate
ratio
of
approximately
20:
1,
mol/
mol
(
equivalent
to
60:
1,
w/
w
for
Cry34Ab1
and
17:
1,
w/
w
for
Cry
35Ab1).
Samples
taken
at
1,
5,
7,
15,
20,
30,
and
60
minutes
were
analyzed
by
sodium
dodecyl
sulfate
polyacrylamide
gel
electrophoresis
(
SDS­
PAGE)
and
western
blot.
Cry35Ab1
was
no
longer
visible
at
the
five­
minute
time­
point
using
both
SDSPAGE
stained
with
Coomassie
Brilliant
Blue
and
western
blot
detection.
Cry34Ab1
was
visible
on
the
stained
gel
for
the
15­
minute
sample,
but
not
in
later
sample
time
points.
In
the
western
blot
analysis,
Cry34Ab1
was
visible
in
the
20­
minute
sample,
but
not
in
later
sample
time
points.
In
conclusion,
this
first
study
showed
that
Cry34Ab1
was
digested
within
30
minutes
and
Cry
35Ab1
was
digested
within
5
minutes
in
SGF
under
the
conditions
of
the
study.

Because
Cry34Ab1
appeared
to
be
somewhat
resistant
to
SGF
in
the
study
described
above,
Dow
submitted
a
second
study
on
the
in
vitro
digestibility
of
Cry34Ab1
in
SGF
(
MRID
455845­
02).
The
digestion
was
performed
under
the
same
conditions
as
the
previous
study
except
that
reaction
mixtures
were
shaken
during
incubation,
and
samples
were
analyzed
at
1,
2,
3,
5,
7.5,
10,
15,
and
20
minutes.
The
previous
study
on
pepsin
digestibility
of
Cry34Ab1
and
Cry35Ab1,
as
well
as
other
pepsin
digestibility
studies
used
in
allergenicity
assessments,
focused
on
the
time
required
for
the
protein
to
become
undetectable,
and
therefore,
the
results
are
dependent
on
the
detection
limit
of
the
analytical
method
used.
In
this
second
study,
Dow
determined
the
rate
of
pepsin
digestion
of
Cry34Ab1
by
measuring
the
relative
amounts
of
Cry34Ab1
at
each
of
the
time
points
based
on
SDS­
PAGE
densitometry
estimates.
Under
the
conditions
of
the
study,
the
rate
of
decay
fit
a
first­
order
model
(
with
respect
to
Cry34Ab1
concentration),
and
Dow
estimated
the
DT50
(
half­
life)
and
DT90
(
time
until
90%
decay)
to
be
1.9
minutes
and
6.2
minutes,
respectively.
In
this
experiment,
Cry34Ab1
was
visible
on
gels
and
blots
in
15
minute
time
point
samples
but
not
in
20
minute
time
point
samples.

In
the
comparison
study
where
Dow
and
Pioneer
used
the
kinetic
approach
to
assess
the
7
digestibility
of
a
number
of
allergens
and
non­
allergens
under
the
same
conditions
as
those
used
in
the
digestibility
studies
on
Cry34Ab1
(
MRID
463886­
01),
two
allergens
and
two
nonallergens
were
shown
to
digest
similarly
to
Cry34Ab1.
From
the
digestibility
studies
that
Dow
submitted
as
well
as
published
studies,
and
previous
guidance
from
Scientific
Advisory
Panels,
EPA
has
concluded
that
Cry35Ab1
is
rapidly
digested
and
Cry34Ab1
is
moderately
digested
in
SGF.
Cry34Ab1
appears
to
digest
slower
than
other
Bt
proteins
that
have
been
registered
for
food
use
and
many
other
proteins
that
are
not
considered
allergens
but
faster
than
most
previously
tested
allergens.

Submitted
studies
on
heat
stability
of
the
Cry34Ab1
and
Cry35Ab1
proteins
demonstrate
that
these
proteins
are
inactivated
when
heated
for
30
minutes
at
90

C
and
60

C,
respectively
(
MRIDs
453584­
01,
455845­
01,
458086­
01,
and
458602­
01).
A
comparison
of
amino
acid
sequences
of
Cry34Ab1
and
Cry35Ab1
with
known
allergens
showed
no
overall
sequence
similarities
or
homology
at
the
level
of
8
contiguous
amino
acid
residues
(
MRID
452422­
05).
Cry34Ab1
and
Cry35Ab1
expressed
in
corn
were
shown
not
to
be
glycosylated
(
MRID
461239­
06).
Expression
level
analysis
indicated
that
the
proteins
are
present
at
relatively
low
levels
in
corn;
on
a
dry
weight
basis,
Cry34Ab1
is
present
at
a
concentration
of
approximately
70
ng/
mg
in
grain,
and
Cry35Ab1
is
present
at
a
concentration
of
approximately
1
ng/
mg
in
grain
(
MRID
461239­
04).
In
addition,
Bacillus
thuringiensis
is
not
considered
to
be
an
allergenic
source.
EPA
has
concluded
that
the
weight
of
the
evidence
indicates
that
the
Cry34Ab1
and
Cry35Ab1
proteins
are
unlikely
to
be
food
allergens.

Cry34Ab1
appears
to
be
moderately
digested
in
SGF,
rather
than
rapidly
digested.
Considering
all
of
the
available
information
 
Cry34Ab1
originates
from
a
non­
allergenic
source,
has
no
sequence
similarity
with
known
allergens,
is
not
glycoslyated,
is
inactivated
by
heat,
is
moderately
digested
in
SGF,
and
will
only
be
present
at
low
levels
in
food
 
EPA
has
concluded
that
Cry34Ab1
is
unlikely
to
be
a
food
allergen.

Questions
for
the
Panel
Protocols
for
Digestibility
Assays
1)
Dow
has
stated
that
enzyme
kinetic
theory
predicts
first
order
kinetics
for
pepsin
hydrolysis
under
conditions
of
high
enzyme
and
low
substrate
concentrations
and
has
demonstrated
that
the
rate
of
substrate
disappearance
under
these
conditions
follows
first­
order
kinetics
for
a
number
of
proteins.
However,
for
several
proteins,
initial
time
points
were
omitted
to
achieve
a
good
fit
to
the
model.
Dow
states
that
the
data
were
not
included
"
based
on
theoretical
considerations,
which
include:
potential
zero­
order
or
mixed
order
kinetics
due
to
high
substrate
concentration,
possible
presence
of
denatured
and
highly
digestible
protein
contaminating
the
native
protein
preparation,
or
the
possibility
of
an
initial
burst
phase
or
transient
phase
preceding
the
first­
order
phase
of
digestion
(
Schnell
and
Maini,
2000;
Milgrom
et
al.,
1998)."
8
The
Panel
is
requested
to
comment
on
whether
the
explanation
justifies
omitting
early
time
points
or
whether
the
poor
fit
of
early
time
points
indicates
a
problem
with
the
model.

2)
Dow
has
asserted
that
first­
order
decay
is
predicted
based
on
enzyme
theory
as
long
as
the
pepsin
concentration
is
high
and
the
substrate
concentration
is
low
(<<
Km)
and
that
the
firstorder
rate
constant
determined
under
these
conditions
is
equal
to
Vmax/
Km.
Dow
has
also
stated
that
as
long
as
first­
order
conditions
are
met,
first­
order
rate
constants
and
half­
lives
are
unaffected
by
changes
in
substrate
protein
concentration
and
that
first­
order
rate
constants
can
be
used
to
predict
relative
digestion
efficiencies
for
proteins
even
if
the
protein
concentration
is
varied
among
experiments.
In
addition,
Dow
has
stated
that
at
the
USP
concentration
for
pepsin
of
0.32%,
the
enzyme
concentration
is
saturating
and
can
also
be
varied
between
experiments
without
affecting
the
first­
order
rate
constant.

The
Panel
is
asked
to
comment
on
these
statements.
How
much
can
the
pepsin
or
protein
substrate
concentrations
vary
without
affecting
the
kinetics
of
pepsin
digestion
and
first­
order
rate
constants?

3)
Typically,
for
comparing
the
in
vitro
digestibility
of
different
proteins,
researchers
have
used
fixed
concentrations
of
pepsin
and
substrate
protein
on
a
weight
basis
(
mg/
mL)
rather
than
adjusting
for
molecular
weight
of
the
substrate
protein,
presumably
because
larger
proteins
likely
have
more
potential
pepsin
cleavage
sites.
However,
Dow
states
that
"
while
multiple
pepsinlabile
sites
may
occur
within
a
protein,
a
single
site
is
often
responsible
for
limiting
digestion
rates,
and
thus
the
number
of
molecules,
rather
than
total
weight,
is
most
often
more
influential
in
determining
the
kinetics
that
describe
decay."

The
Panel
is
asked
to
comment
on
Dow's
statement.
To
compare
the
rate
of
pepsin
digestion
of
different
proteins,
is
it
more
appropriate
for
the
concentration
of
test
protein
to
be
constant
on
a
weight
basis
(
mg/
mL)
or
a
mole
basis
(
mol/
L)?

4)
Typically,
researchers
have
looked
at
the
effect
of
pepsin
to
substrate
ratio
rather
than
concentrations
on
digestion
(
Karamac,
et
al.,
2002).
How
do
varying
the
ratios
and/
or
concentrations
affect
the
rate
of
hydrolysis?

5)
Different
assays
exist
for
determining
pepsin
activity.
A
pepsin
activity
assay
based
on
measuring
the
trichloracetic
acid­
soluble
products
of
pepsin
hydrolysis
of
hemoglobin
is
provided
in
USP,
2004
under
the
entry
for
pepsin.
However,
the
entry
in
USP,
2004
for
"
gastric
fluid,
simulated"
references
the
Food
Chemicals
Codex
for
pepsin
activity,
which
provides
an
assay
that
measures
pepsin
digestion
of
egg
albumen.

The
Panel
is
asked
to
comment
on
the
appropriateness
of
using
a
fixed
concentration
of
pepsin
versus
using
a
fixed
specific
activity
of
pepsin
in
digestibility
protocols.
How
would
the
use
of
different
pepsin
activity
assays
affect
the
measured
pepsin
activity
units?
9
6)
Typically,
scientists
have
used
SDS­
PAGE
with
staining
or
western
blot
analysis
for
monitoring
digestion
reactions.
HPLC
is
also
sometimes
used.

The
Panel
is
asked
to
comment
on
the
pros
and
cons
of
the
different
methods
that
could
be
used
for
monitoring
digestion
reactions.

7)
Some
researchers
have
used
one
digestion
reaction
and
removed
aliquots
at
various
times
for
monitoring,
while
others
have
set
up
separate
reactions
for
each
of
the
time
points.

What
are
the
pros
and
cons
of
these
approaches?

8)
Under
the
current
protocol,
Dow's
kinetic
approach
is
only
applicable
to
moderately
digestible
proteins
(
i.
e.,
using
Dow's
protocol,
many
proteins
digest
too
quickly
and
some
too
slowly
to
obtain
an
adequate
number
of
data
points
for
quantitative
kinetic
analysis).

Please
comment
on
the
usefulness
of
the
kinetic
approach
for
proteins
that
are
not
rapidly
degraded.

Allergenicity
Assessment
9)
The
2001
FAO/
WHO
report
and
2003
Codex
guidelines
both
recommend
using
in
vitro
digestibility
in
assessing
the
allergenicity
potential
of
a
protein.
The
FAO/
WHO
report
provides
a
"
decision
tree"
approach,
while
the
Codex
guidelines
suggest
a
weight
of
evidence
approach.
Codex
guidelines
state
"
resistance
of
a
protein
to
degradation
in
the
presence
of
pepsin
under
appropriate
conditions
indicates
that
further
analysis
should
be
conducted
to
determine
the
likelihood
of
the
newly
expressed
protein
being
allergenic,"
and
"
it
should
be
taken
into
account
that
a
lack
of
resistance
to
pepsin
does
not
exclude
that
the
newly
expressed
protein
can
be
a
relevant
allergen."
The
Codex
guidelines,
however,
don't
specify
how
a
protein
should
be
further
evaluated
if
it
is
"
resistant"
to
degradation,
and
"
resistant"
is
not
defined.

a)
What
weight
should
in
vitro
digestibility
studies
be
given
in
the
overall
assessment
compared
with
other
criteria
such
as
sequence
homology?

b)
The
Panel
is
asked
to
comment
on
the
appropriateness
of
setting
acceptable/
unacceptable
limits
for
digestibility
in
assessing
the
safety
of
a
protein.

10)
Stable
digestion
fragments
are
often
formed
during
pepsin
digestion
of
proteins,
and
Dow
has
used
the
kinetic
approach
to
estimate
the
half­
lives
of
several
digestion
fragments.

Please
comment
on
the
significance
of
the
rate
of
digestion
of
protein
fragments
for
allergenicity
assessments.
10
Cry34Ab1
Assessment
11)
Cry34Ab1
appears
to
be
moderately
digested
in
SGF,
rather
than
rapidly
digested.
Considering
all
of
the
available
information
 
Cry34Ab1
originates
from
a
non­
allergenic
source,
has
no
sequence
similarity
with
known
allergens,
is
not
glycoslyated,
is
inactivated
by
heat,
is
moderately
digested
in
SGF,
and
will
only
be
present
at
low
levels
in
food
 
EPA
has
concluded
that
Cry34Ab1
is
unlikely
to
be
a
food
allergen.

Please
comment
on
the
Agency's
conclusions
regarding
the
allergenicity
of
Cry34Ab1.

References
Astwood,
J.
D.,
Leach,
J.
N.,
Fuchs,
R.
L.,
"
Stability
of
food
allergens
to
digestion
in
vitro,"
Nat.
Biotech.,
1996,
14,
1269­
1973.

"
Assessment
of
Possible
Allergenicity,"
Annex
to
"
Guideline
for
the
Conduct
of
Food
Safety
Assessment
of
Foods
Derived
from
Recombinant­
DNA
Plants,"
Developed
by
the
Codex
Ad
Hoc
Intergovernmental
Task
Force
on
Foods
Derived
from
Biotechnology,
2003.

Choisnard,
L.,
Froidevaux,
R.,
Nedjar­
Arroume,
N.,
Lignot,
B.,
Vercaigne­
Marco,
D.,
Krier,
F.,
Dhulster,
P.,
Guillochon,
D.,
"
Kinetic
study
of
the
appearance
of
an
anti­
bacterial
peptide
in
the
course
of
bovine
haemoglobin
peptic
hydrolysis,"
Biotechnol.
Appl.
Biochem.
2002,
36,
187­
194.

Dearman,
R.
J.,
Caddick,
H.,
Stone,
S.,
Kenna,
J.
G.,
Basketter,
D.
A.,
Kimber
I.,
"
Immunogenic
properties
of
rapidly
digested
food
proteins
following
gavage
exposure
of
mice:
a
comparison
of
ovalbumin
with
a
potato
acid
phosphatase
preparation,"
Food
Chem.
Toxicol.
2002,
40,
625­
633.

"
Evaluation
of
Allergenicity
of
Genetically
Modified
Foods,"
Report
of
a
Joint
FAO/
WHO
Expert
Consultation
on
Allergenicity
of
Foods
Derived
from
Biotechnology,
FAO,
Rome,
Italy,
22­
25
January
2001.

Food
Chemicals
Codex,
5th
edition,
Committee
on
Food
Chemicals
Codex,
Eds.,
Food
and
Nutrition
Board,
Institute
of
Medicine
of
the
National
Academies,
The
National
Academies
Press,
Washington,
D.
C.,
2004.

Fu,
T.­
J.,
Abbott,
U.
R.,
Hatzos,
C.,
"
Digestibility
of
food
allergens
and
nonallergenic
proteins
in
simulated
gastric
fluid
and
simulated
intestinal
fluid
 
a
comparative
study,"
J.
Agric.
Food
Chem.
2002,
50,
7154­
7160.

Herman,
R.
A.,
Korjagin,
V.
A.,
Schafer,
B.
W.,
"
Quantitative
measurement
of
protein
digestion
in
simulated
gastric
fluid,"
Reg.
Toxicol.
Pharmacol.
2005,
article
in
press.
11
Herman,
R.
A.,
Schafer,
B.
W.,
Korjagin,
V.
A.,
and
Ernest,
A.
D.,
"
Rapid
Digestion
of
Cry34Ab1
and
Cry35Ab1
in
Simulated
Gastric
Fluid,"
J.
Agric.
Food
Chem.
2003,
51,
6823­
6827.

Karamac,
M,
Amarowicz,
R.,
Kostyra,
H.,
"
Effect
of
Temperature
and
Enzyme/
Substrate
Ratio
on
the
Hydrolysis
of
Pea
Protein
Isolates
by
Trypsin,"
Czech
J.
Food
Sci.
2002,
20,
1­
6.

Milgrom,
Y.
M.,
Murataliev,
M.
B.,
Boyer,
P.
D.,
"
Bi­
site
Activation
Occurs
with
Native
and
Nucleotide­
Depleted
Mitochondrial
F1­
ATPase,"
Biochem.
J.
1998,
330,
1037­
1043.

Nilsson,
C.,
Oman,
H.,
Hallden,
G.,
Lilja,
G.,
Lundberg,
M.,
Harfast,
B.,
"
A
case
of
allergy
to
cow's
milk
hydrolysate,"
Allergy
1999,
54,
1322­
1326.

Schnell,
S.,
and
Maini,
P.
K.,
"
Enzyme
Kinetics
at
High
Enzyme
Concentration,"
Bull.
Math.
Biol.
2000,
62,
483­
499.

Schnell,
S.,
and
Mendoza,
C.,
"
The
condition
for
pseudo­
first­
order
kinetics
in
enzymatic
reactions
is
independent
of
the
initial
enzyme
concentrations,"
Biophys.
Chem.
2004,
107,
165­
174
Segel,
I.
H.,
"
Enzyme
Kinetics:
Behavior
and
Analysis
of
Rapid
Equilibrium
and
Steady­
State
Enzyme
Systems,"
John
Wiley
&
Sons,
Inc.,
New
York,
1975,
Wiley
Classics
Library
Edition,
1993,
p.
72­
74.

Thomas,
K.,
Aalbers,
M.,
Bannon,
G.
A.,
Bartels,
M.,
Dearman,
R.
J.,
Esdaile,
D.
J.,
Fu,
T.­
J.,
Glatt,
C.
M.,
Hadfield,
N.,
Hatzos,
C.,
Hefle,
S.
L.,
Heylings,
J.
R.,
Goodman,
R.
E.,
Henry,
B.,
Herouet,
C.,
Holsapple,
M.,
Ladics,
G.
S.,
Landry,
T.
D,
MacIntosh,
S.
C.,
Rice,
E.
A.,
Privalle,
L.
S.,
Steiner,
H.
Y.,
Teshima,
R.,
van
Ree,
R.,
Woolhiser,
M.,
Zawodny,
J.,
A
multi­
laboratory
evaluation
of
a
common
in
vitro
pepsin
digestion
assay
protocol
used
in
assessing
the
safety
of
novel
proteins,
Reg.
Tox.
Pharm.
2004,
39,
87­
98.

Tzafriri,
A.
R.,
"
Michaelis­
Menton
Kinetics
at
High
Enzyme
Concentrations,"
Bull.
Math.
Biol.
2003,
65,
1111­
1129.

United
States
Pharmacopeia
27,
in
The
National
Formulary,
Board
of
Trustess,
Eds.,
The
United
States
Pharmacopeial
Convention,
Inc.,
Rockville,
MD.
2004,
p.
2700
and
2728.
12
Table.
Comparison
of
Pepsin
Digestibility
Studies
Protein
MRID
pH
Pepsin:
Pro
tein
ratioa
(
w/
w)
Pepsin:
Pro
tein
ratioa
(
mol/
mol)
Disappear
ance
Time
(
Western
Blot)
Disappear
ance
Time
(
SDSPAGE
Digestion
Fragment
Disappear
ance
Time
Cry1Ac
431452­
14
1.2
1600:
1
3370:
1
<
30
sec
<
7
min
(
western)

Cry1Ac
43995­
03
1.2
352:
1
740:
1
<
2
min
(<
1%
remaining)

Cry1Ac
43995­
03
1.2
3.5:
1
7.4:
1
<
5
min
(<
1%
remaining)

Cry1Ab
433236­
06
1.0­
1.2
3:
1
10:
1
<
2
min
Cry1Ab
433236­
06
1.0­
1.2
0.007:
1
0.027:
1
<
5
min
Cry1Ab
451144­
01
1.5
15:
1
56:
1
<
15
min
<
15
min
PAT
439995­
03
1.2
93:
1
61:
1
trace
2
min
PAT
439995­
03
1.2
0.93:
1
0.61:
1
<
5
min
PAT
458084­
16
1.2
6.5:
1
4.3:
1
<
30
sec
<
30
sec
Cry1F
455423­
18
1.2
2:
1
<
1
min
<
1
min
VIP3A
458358­
06
1.0­
1.2
(
not
measured)
1.3:
1
1.5:
1
immediate
disappearan
ce
Cry9C
451144­
01
1.5
15:
1
<
1
hour
<
1
hour
(<
10%
remaining
at
30
min)

Cry9C
451144­
02
1.2
32:
1
<
1
hour
<
1
hour
Cry9C
442581­
08
2.0
Not
given
Not
given
>
4
hours
>
4
hours
Cry2Ab2
449666­
03
1.2
20:
1b
<
15
sec
(<
2%
remaining)
<
15
sec
Cry3Bb1
454240­
06
1.2
20:
1b
<
60
sec
(<
1%
remaining)
<
15
sec
<
8
min
Cry35Ab1
452422­
12
1.2
17:
1
20:
1
<
5
min
<
5
min
(<
3%
remaining)

Cry34Ab1
452422­
12
1.2
60:
1
20:
1
<
30
min
(<
0.38%
remaining)
<
20
min
Cry34Ab1
455845­
02
1.2
60:
1
20:
1
<
15
min
(<
0.38%
remaining)
<
10
min
13
