UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
D.
C.
20460
OFFICE
OF
PREVENTION,
PESTICIDES,
AND
TOXIC
SUBSTANCES
June
1,
2006
MEMORANDUM
SUBJECT:
Transmittal
of
Meeting
Minutes
of
the
FIFRA
Scientific
Advisory
Panel
Meeting
Held
March
14
­
15,
2006
on
Event
MIR604
Modified
Cry3A
Protein
Bt
Corn
­
Plant­
Incorporated
Protectant
TO:
James
J.
Jones,
Director
Office
of
Pesticide
Programs
FROM:
Joseph
E.
Bailey,
Designated
Federal
Official
FIFRA
Scientific
Advisory
Panel
Office
of
Science
Coordination
and
Policy
THRU:
Steven
Knott,
Executive
Secretary
FIFRA
Scientific
Advisory
Panel
Office
of
Science
Coordination
and
Policy
Clifford
J.
Gabriel,
Ph.
D.,
Director
Office
of
Science
Coordination
and
Policy
Attached,
please
find
the
meeting
minutes
of
the
FIFRA
Scientific
Advisory
Panel
open
meeting
held
in
Arlington,
Virginia
on
March
14
­
15,
2006.
This
report
addresses
a
set
of
scientific
issues
being
considered
by
the
Environmental
Protection
Agency
pertaining
to
Event
MIR604
Modified
Cry3A
Protein
Bt
Corn
­
Plant­
Incorporated
Protectant.

Attachment
2
cc:

Susan
Hazen
Frank
Sanders
Amy
Farrell
Richard
Keigwin
Anne
Lindsay
William
Jordan
Margie
Fehrenbach
Douglas
Parsons
Janet
Andersen
Enesta
Jones
Debbie
Edwards
Vanessa
Vu
(
SAB)
Steven
Bradbury
Mike
Mendelsohn
William
Diamond
Tessa
Milofsky
Arnold
Layne
Annabel
Fellman
Tina
Levine
OPP
Docket
Lois
Rossi
FIFRA
Scientific
Advisory
Panel
Members
Steven
G.
Heeringa,
Ph.
D.
(
Chair
of
the
FIFRA
SAP)
Kenneth
M.
Portier,
Ph.
D.
(
Session
Chair)
John
Bucher,
Ph.
D.
Janice
E.
Chambers,
Ph.
D.
H.
Christopher
Frey,
Ph.
D.
Stuart
Handwerger,
M.
D.
Gary
Isom,
Ph.
D.

FQPA
Science
Review
Board
Members
Amal
Halim
Assa'ad,
M.
D.
John
C.
Schneider,
Ph.
D.
George
P.
Cobb,
Ph.
D.
Geoffrey
I.
Scott,
Ph.
D.
Brian
A.
Federici,
Ph.
D.
Janice
E.
Thies,
Ph.
D.
Steven
Gendel,
Ph.
D.
Mark
A.
Weaver,
Ph.
D.
Louis
B.
Hersh,
Ph.
D.
Mark
E.
Whalon,
Ph.
D.
George
Barrie
Kitto,
Ph.
D.
Sean
M.
Richards,
Ph.
D.
Shridhar
K.
Sathe,
Ph.
D.
SAP
Minutes
No.
2006­
02
A
Set
of
Scientific
Issues
Being
Considered
by
the
Environmental
Protection
Agency
Regarding:

Event
MIR604
Modified
Cry3A
Protein
Bt
Corn
­
Plant­
Incorporated
Protectant
March
14
­
15,
2006
FIFRA
Scientific
Advisory
Panel
Meeting,
held
at
the
National
Airport
Holiday
Inn
Arlington,
Virginia
2
NOTICE
These
meeting
minutes
have
been
written
as
part
of
the
activities
of
the
Federal
Insecticide,
Fungicide,
and
Rodenticide
Act
(
FIFRA),
Scientific
Advisory
Panel
(
SAP).
The
meeting
minutes
represent
the
views
and
recommendations
of
the
FIFRA
SAP,
not
the
United
States
Environmental
Protection
Agency
(
Agency).
The
content
of
the
meeting
minutes
does
not
represent
information
approved
or
disseminated
by
the
Agency.
The
meeting
minutes
have
not
been
reviewed
for
approval
by
the
Agency
and,
hence,
the
contents
of
these
meeting
minutes
do
not
necessarily
represent
the
views
and
policies
of
the
Agency,
nor
of
other
agencies
in
the
Executive
Branch
of
the
Federal
government,
nor
does
mention
of
trade
names
or
commercial
products
constitute
a
recommendation
for
use.

The
FIFRA
SAP
is
a
Federal
advisory
committee
operating
in
accordance
with
the
Federal
Advisory
Committee
Act
and
established
under
the
provisions
of
FIFRA
as
amended
by
the
Food
Quality
Protection
Act
(
FQPA)
of
1996.
The
FIFRA
SAP
provides
advice,
information,
and
recommendations
to
the
Agency
Administrator
on
pesticides
and
pesticide­
related
issues
regarding
the
impact
of
regulatory
actions
on
health
and
the
environment.
The
Panel
serves
as
the
primary
scientific
peer
review
mechanism
of
the
EPA,
Office
of
Pesticide
Programs
(
OPP),
and
is
structured
to
provide
balanced
expert
assessment
of
pesticide
and
pesticide­
related
matters
facing
the
Agency.
Food
Quality
Protection
Act
Science
Review
Board
members
serve
the
FIFRA
SAP
on
an
ad
hoc
basis
to
assist
in
reviews
conducted
by
the
FIFRA
SAP.
Further
information
about
FIFRA
SAP
reports
and
activities
can
be
obtained
from
its
website
at
http://
www.
epa.
gov/
scipoly/
sap/
or
the
OPP
Docket
at
(
703)
305­
5805.
Interested
persons
are
invited
to
contact
Joseph
E.
Bailey,
SAP
Designated
Federal
Official,
via
e­
mail
at
bailey.
joseph@
epa.
gov
In
preparing
the
meeting
minutes,
the
Panel
carefully
considered
all
information
provided
and
presented
by
the
Agency
presenters,
as
well
as
information
presented
by
public
commenters.
This
document
addresses
the
information
provided
and
presented
by
the
Agency
within
the
structure
of
the
charge.
3
TABLE
OF
CONTENTS
PARTICIPANTS...........................................................................................................
5
INTRODUCTION.........................................................................................................
7
PUBLIC
COMMENTERS
...........................................................................................
8
SUMMARY
OF
PANEL
DISCUSSION
AND
RECOMMENDATIONS
..................
9
PANEL
DELIBERATIONS
AND
RESPONSE
TO
CHARGE
................................
12
REFERENCES.................................................................................................................
38
4
SAP
Minutes
No.
2006­
02
A
Set
of
Scientific
Issues
Being
Considered
by
the
Environmental
Protection
Agency
Regarding:

Event
MIR604
Modified
Cry3A
Protein
Bt
Corn
­
Plant­
Incorporated
Protectant
March
14
­
15,
2006
FIFRA
Scientific
Advisory
Panel
Meeting,
held
at
the
National
Airport
Holiday
Inn
Arlington,
Virginia
Kenneth
M.
Portier,
Ph.
D.
Joseph
E.
Bailey
FIFRA
SAP
Session
Chair
Designated
Federal
Official
FIFRA
Scientific
Advisory
Panel
FIFRA
Scientific
Advisory
Panel
Date:
June
1,
2006
Date:
June
1,
2006
5
Federal
Insecticide,
Fungicide,
and
Rodenticide
Act
Scientific
Advisory
Panel
Meeting
March
14
­
15,
2006
Event
MIR604
Modified
Cry3A
Protein
Bt
Corn
­
Plant­
Incorporated
Protectant
PARTICIPANTS
FIFRA
SAP
Chair
Kenneth
M.
Portier,
Ph.
D.,
Program
Director,
Statistics,
American
Cancer
Society,
Statistics
and
Evaluation
Center,
Atlanta,
GA
Designated
Federal
Official
Joseph
E.
Bailey,
FIFRA
Scientific
Advisory
Panel,
Office
of
Science
Coordination
and
Policy,
Environmental
Protection
Agency
FIFRA
Scientific
Advisory
Panel
Members
Stuart
Handwerger,
M.
D.,
Director,
Division
of
Endocrinology,
University
of
Cincinnati,
Children's
Hospital
Medical
Center,
University
of
Cincinnati,
Cincinnati,
OH
Steven
G.
Heeringa,
Ph.
D.,
Research
Scientist
&
Director
for
Statistical
Design,
University
of
Michigan,
Institute
for
Social
Research,
Ann
Arbor,
MI
FQPA
Science
Review
Board
Members
Amal
Halim
Assa'ad,
M.
D.,
Professor
of
Clinical
Pediatrics,
University
of
Cincinnati,
Children's
Hospital
Medical
Center,
Cincinnati,
OH
George
P.
Cobb,
Ph.
D.,
Associate
Professor,
Environmental
Toxicology,
Texas
Tech
University,
Lubbock,
TX
Brian
A.
Federici,
Ph.
D.,
Professor,
Entomology
and
Interdepartmental
Graduate
Programs
in
Genetics
and
Microbiology,
University
of
California,
Riverside,
CA
Steven
Gendel,
Ph.
D.,
Branch
Chief,
Food
and
Drug
Administration,
National
Center
for
Food
Safety
and
Technology,
Summit­
Argo,
IL
Louis
B.
Hersh,
Ph.
D.,
George
Schwert
Professor
and
Chair,
Department
of
Molecular
and
Cellular
Biochemistry,
College
of
Medicine,
University
of
Kentucky,
Lexington,
KY
6
George
Barrie
Kitto,
Ph.
D.,
Professor
of
Chemistry
and
Biochemistry,
Department
of
Chemistry
and
Biochemistry,
The
University
of
Texas
at
Austin,
Austin,
TX
Sean
M.
Richards,
Ph.
D.,
UC
Foundation
Assistant
Professor,
Department
of
Biological
and
Environmental
Sciences,
The
University
of
Tennessee
at
Chattanooga,
Chattanooga,
TN
Shridhar
K.
Sathe,
Ph.
D.,
D.
K.
Salunkhe
Professor
of
Food
Science,
Distinguished
Teaching
Professor,
Department
of
Nutrition,
Food
and
Exercise
Sciences,
Florida
State
University,
Tallahassee,
FL
John
C.
Schneider,
Ph.
D.,
Professor/
Research
Entomologist,
Department
of
Entomology
&
Plant
Pathology,
Mississippi
State
University,
Mississippi
State,
MS
Geoffrey
I.
Scott,
Ph.
D.,
Director,
Center
for
Coastal
Environmental
Health
&
Biomolecular
Research
at
Charleston,
U.
S.
Department
of
Commerce,
NOAA,
National
Ocean
Service,
Charleston,
SC
Janice
E.
Thies,
Ph.
D.,
Associate
Professor,
Soil
Biology
and
Ecology,
Department
of
Crop
and
Soil
Science,
Cornell
University,
Ithaca,
NY
Mark
A.
Weaver,
Ph.
D.,
Plant
Pathologist,
South
Weed
Science
Research
Unit,
Agricultural
Research
Center,
USDA,
Stoneville,
MS
Mark
E.
Whalon,
Ph.
D.,
Professor,
Center
for
Integrated
Plant
Systems,
Michigan
State
University,
East
Lansing,
MI
7
INTRODUCTION
The
Federal
Insecticide,
Fungicide,
and
Rodenticide
Act
(
FIFRA),
Scientific
Advisory
Panel
(
SAP)
has
completed
its
review
of
Event
MIR604
Modified
Cry3A
Protein
Bt
Corn
­
Plant­
Incorporated
Protectant.
Advance
notice
of
the
meeting
was
published
in
the
Federal
Register
on
January
25,
2006.
The
review
was
conducted
in
an
open
Panel
meeting
held
in
Arlington,
Virginia,
March
14
­
15,
2006.
Dr.
Kenneth
M.
Portier
chaired
the
meeting.
Joseph
E.
Bailey
served
as
the
Designated
Federal
Official.

The
FIFRA
SAP
met
to
consider
and
review
human
health
and
environmental
issues
associated
with
Event
MIR604
Modified
Cry3A
Protein
Bt
Corn
­
Plant­
Incorporated
Protectant.
Syngenta
Seeds,
Inc.
(
Registrant)
submitted
applications
for
FIFRA
section
3
registrations
of
the
plant­
incorporated
protectant
Modified
Cry3A
protein
intended
to
provide
corn
protection
from
western,
northern
and
Mexican
corn
rootworm
larvae.

The
agenda
for
this
SAP
meeting
included
an
introduction
and
background
of
the
issues
under
consideration
provided
by
Mr.
Mike
Mendelsohn
(
Biopesticides
and
Pollution
Prevention
Division,
Office
of
Pesticide
Programs).
Tessa
Milofsky,
MS
(
Biopesticides
and
Pollution
Prevention
Division,
OPP)
provided
an
overview
of
EPA's
ecological
risk
assessment
and
Ms.
Annabel
Fellman
(
Biopesticides
and
Pollution
Prevention
Division,
OPP)
provided
an
overview
of
EPA's
human
health
risk
assessment.
Janet
Andersen,
Ph.
D.
(
Director,
Biopesticides
and
Pollution
Prevention
Division,
OPP)
offered
opening
remarks
at
the
meeting.
8
PUBLIC
COMMENTERS
Oral
statements
were
presented
by:

Dave
Nelson
on
behalf
of
the
National
Corn
Growers
Association
James
D.
Thrift
on
behalf
of
the
Agricultural
Retailers
Association
Demetra
Vlachos,
Alan
Raybould,
and
Fred
Waters,
on
behalf
of
Syngenta
Seeds,
Inc.

Written
statements
were
provided
by:

Todd
Barlow,
Kentucky
Corn
Growers
Association
Wes
Beck,
Jr.,
Farm
Operator
James
E.
Betts,
Ohio
Professional
Applicators
for
Responsible
Regulation
William
Bond,
M.
S.,
C.
A.
E.,
Minnesota
Crop
Production
Retailers
Association
Dan
DeRycke,
DeRycke
Farms
Steve
Ebke,
Nebraska
Corn
Growers
Association
Cresswell
A.
Hizer,
Indiana
Plant
Food
and
Agricultural
Chemicals
Association
John
Kuhfuss,
Illinois
Corn
Growers
Association
Roger
Kvols,
Kvols
Farms
Lloyd
R.
Lee,
Lee
Family
Farms
Dale
R.
Ludwig,
Missouri
Soybean
Association
Gary
D.
Marshall,
Missouri
Corn
Growers
Association
Robert
W.
Prince,
M.
S.,
C.
C.
A.,
Garst
Seed
Company
Mindy
Larson
Poldberg,
Iowa
Corn
Growers
Association
B.
Sachau,
Public
Citizen
Ted
Vinson,
Vinson
Farms
Bernie
Walsh,
Bernie
Walsh
Farms
Jere
White,
Kansas
Corn
Growers
Association
9
SUMMARY
OF
PANEL
DISCUSSION
AND
RECOMMENDATIONS
Panel
conclusions
regarding
the
safety
of
mCry3A
were
based
on
several
factors.
Environmental
Protection
Agency
(
Agency)
summaries
of
the
process
and
the
data
packet
provided
the
overview,
with
specific
details
of
studies
presented
by
Syngenta
Seeds
Inc.
Data
were
reviewed
from
studies
on
ecological
effects
(
non­
target),
human
health,
mCry3A
characterization,
mammalian
toxicity,
and
allergenicity,
as
well
as
a
review
of
requirements
for
field
testing
of
ecological
effects
in
fields
where
MIR604
corn
is
cultivated.
Public
testimony
was
also
considered.

Overall,
the
Panel
found
poor
data
quality
and
inadequate
documentation
in
many
aspects
of
the
registration
packet
for
mCry3A
safety,
from
amino
acid
sequence
reporting
to
documentation
of
natural
versus
plant
incorporated
protein
homology
to
toxicity
testing
protocols.
The
errors
or
omissions
in
some
studies
were
of
a
nature
and
degree
that
many
on
the
Panel
believed
they
should
be
excluded
from
consideration
in
ecological
hazard/
risk
assessment
(
in
particular
the
earthworm,
soil
dissipation,
and
ground
beetle
studies)
since
their
study
designs
and
data
reporting
did
not
meet
basic
Agency
standards.
Omissions
in
several
other
studies
reduced
the
certainty
of
estimates
to
the
point
where
they
may
not
be
useful
even
in
Tier
1
hazard
assessments.
In
several
instances
the
hazard
quotients
or
safety
factors
computed
were
deemed
by
the
Panel
to
be
too
low
and/
or
unreliable.
In
others,
specifically
in
the
rainbow
trout,
mouse,
and
quail
studies,
the
safety
factors
were
sufficiently
high;
but
improper
or
inadequate
replication
weakened
the
validity
of
all
three
tests
and
their
conclusion
of
no
acute
mortality
hazard.

Panel
members
expressed
a
diversity
of
opinions
concerning
the
adequacy
of
the
Agency's
analysis
of
the
ecological
studies
submitted
by
the
Registrant.
These
opinions
ranged
from
nearly
unqualified
acceptance
to
qualified
rejection.
A
number
of
suggestions
for
improving
Tier
1
Maximum
Hazard
Dose
testing
for
effects
of
Plant­
Incorporated
Protectants
(
PIPs)
on
non­
target
organisms
(
NTOs)
were
made
during
the
course
of
the
Panel
discussion
and
are
listed
below.
One
omission
identified
was
that
the
endangered
Hungerford's
crawling
water
beetle
(
Brychius
hungerfordi;
Coleoptera:
Haliplidae),
that
exists
in
some
areas
of
high
corn
production,
was
not
studied
or
otherwise
considered
in
this
assessment.
Panel
members
strongly
supported
the
Agency's
request
for
supplemental
ecological
studies
in
areas
where
MIR604
corn
is
being
cultivated,
and
strongly
recommended
that
these
studies
be
designed
around
sound
scientific
study
principles
in
order
that
data
generated
can
be
analyzed
statistically
in
order
to
better
meet
information
requirements
of
the
Agency.
The
Panel
noted
from
the
public
comments
how
use
of
Bt
corn
may
not
necessarily
lead
to
a
corresponding
reduction
in
costs
associated
with
application
of
conventional
pesticides.
The
ecological
impact
of
these
combined
technologies,
which
was
not
discussed
in
the
assessment,
is
only
now
being
discussed
in
the
scientific
literature,
and
represents
an
aspect
of
PIP
use
that
should
be
studied
further.
10
The
Panel
was
asked
to
comment
on
the
extent
to
which
mCry3A
protein
extracted
from
MIR604
corn
(
LPMIR604­
0103)
is
similar
for
risk
assessment
purposes
to
that
produced
from
recombinant
E.
coli
(
MCRY3A­
0102).
Based
on
bioassays
against
first
instar
western
corn
rootworm
larvae,
the
Registrant
noted
that
the
mCry3A
extract
from
MIR604
corn
was
more
active
than
the
mCry3A
produced
in
E.
coli.
The
Panel
found
that
this
could
be
explained
by
the
different
mCry3A
gene
constructs
used
to
synthesize
mCry3A
in
corn
versus
E.
coli,
and
by
the
subsequent
proteolytic
processing
of
mCry3A
in
corn
or
non­
target
invertebrates,
especially
beetles.
Synthesis
of
mCry3A
protein
in
E.
coli
produced
two
forms
of
the
protein,
one
with
a
mass
of
67.7
kDa,
(
roughly
equivalent
to
LPMIR604­
0103
extracted
from
MIR604
corn)
and
the
other
of
69.5
kDa,
(
a
long
form
containing
sixteen
additional
amino
acids
at
the
N­
terminus).
In
corn,
however,
the
two
forms
of
mCry3A
detected
were
a
67.7­
kDa
protein
and
a
55­
kDa
protein,
the
latter
apparently
being
an
activated
form
of
mCry3A
that
constituted
50%
of
the
mCry3A
in
corn.
It
was
concluded
that
the
mCry3A
proteins
are
substantially
equivalent
with
respect
to
their
amino
acid
sequences,
lack
of
glycosylation,
other
types
of
potential
posttranslational
modifications
and
general
stability.
However,
because
the
surrogate
forms
of
mCry3A
produced
by
E.
coli
were
not
activated,
as
was
apparently
50%
of
the
mCry3A
produced
in
the
mCry3A
corn
noted
above,
it
was
not
certain
that
the
different
forms
of
the
mCry3A
produced
in
corn
versus
E.
coli
could
be
considered
substantially
equivalent
for
the
purpose
of
non­
target
invertebrate
studies.
The
activated
forms
of
mCry3A,
i.
e.,
the
55­
kDa
protein,
were
considered
to
be
substantially
equivalent
regardless
of
whether
the
source
was
corn
or
E.
coli.
Although
no
direct
information
was
provided
about
this
form
from
subsequent
cleavage
of
either
of
the
longer
forms
from
bacterial
production,
one
can
reasonably
make
the
conservative
assumption,
but
it
is
only
an
assumption,
that
the
55­
kDa
form
is
produced
from
both
the
long
and
short
forms
of
E.
coli­
produced
mCry3A.
However,
at
least
one
Panel
member
believed
that
the
mammal
studies
were
inadequate
to
support
the
statement
that
the
activated
forms
were
substantially
equivalent.

Low
mammalian
toxicity
was
deemed
to
be
supported
by
the
test
results
presented.
Most
of
the
Panel
agreed
that
the
exposure
assessment
showed
that
the
amount
of
Cry
protein
that
would
be
consumed
in
a
rather
large
serving
of
corn
kernels
would
be
several
orders
of
magnitude
lower
than
the
highest
dose
tested
in
the
mammalian
toxicity
studies
 
which
had
no
observed
effect.
The
exact
calculation
of
exposure
was
considered
uncertain
due
to
the
mixture
of
short
and
long
forms
of
the
protein
in
the
E.
coli­
derived
material
used
in
the
assay.
It
was
pointed
out
that
non­
oral
routes
of
exposure
were
not
explored
and
that
true
allergenicity
studies
were
not
performed.
The
sequence
analysis
(
homology
search
via
NCBI­
BLASTP)
was
poorly
documented
and
difficult
for
the
Panel
to
interpret
and
as
a
result
is
of
limited
value
in
the
risk
assessment.

A
couple
of
Panel
members
believed
that
no
conclusion
on
safety
could
be
derived
from
the
data
and
that
much
more
information
is
needed
about
acute
effects
and
potential
effects
on
children's
growth
and
development.
At
least
three
Panel
members
expressed
concern
regarding
the
adequacy
of
the
scientific
basis
supporting
a
permanent
exemption
from
the
requirement
for
a
tolerance.
Such
concerns
could
be
alleviated
by
more
rigorous
allergenic
testing/
monitoring
of
mCry3A
among
workers
or
consumers
of
products
generated
from
this
type
of
corn.
11
The
Panel
discussed
the
following
recommendations
to
focus
Agency
evaluations
of
Tier
1
studies
related
to
PIPs.

 
The
Agency
should
utilize
full
technical
reports
rather
than
summary
reports
in
the
assessment
including
full
documentation
of
study
protocols,
implementation
findings
and
inclusion
of
raw
data
in
technical
reports.
 
Rely
less
on
descriptive
statistics
in
technical
reports
and
require
more
formal
statistical
testing.
 
Ensure
that
Tier
1
hazard
assessments
meet
the
minimum
regulatory
requirements
and
have
robust
experimental
designs
and,
in
particular,
use
true
replication
in
study
protocols.
Study
designs
should
represent
good
science
and
be
of
journal
publication
quality.
The
resulting
data
should
support
estimation
of
the
margin
of
exposure
(
MOE)
with
acceptable
95%
confidence
intervals
(
CIs).
 
For
PIPs
which
may
exist
in
multiple
forms,
require
clear
characterization
of
test
materials,
particularly
in
terms
of
structure
and
toxicity
potential.
 
Study
protocols
should
be
at
least
as
precise
and
extensive
as
those
required
in
chemical
pesticide
assessments.
A
finding
in
support
of
a
permanent
exemption
from
the
requirement
for
a
tolerance
for
a
PIP
in
a
specific
crop
should
require,
in
addition
to
true
replication
of
treatments
and
in
some
cases
of
whole
studies,
a
predominance
of
no
effect
findings.
In
this
situation
and
with
major
food
crops
(
such
as
corn),
the
potential
for
risk
must
be
demonstrated
to
be
extremely
low.
 
With
respect
to
hazard
assessment
in
Tier
1
ecological
studies,
the
Agency
should
reevaluate
its
current
approach
of
giving
study
effort
and
weight
to
sub­
lethal
effects
only
when
it
has
been
demonstrated
that
the
LC50<
10X
EEC.
 
Fate
and
persistence
studies
should
be
performed
on
multiple
soil
types,
representative
of
the
soils
on
which
the
majority
of
the
crop
will
be
grown
and
soils
measured
for
sufficient
time
to
demonstrate
how
long
the
protein
persists
under
field
conditions.
 
Develop
protocols
that
clearly
allow
non­
allergenicity
in
PIPs
to
be
demonstrated.
12
PANEL
DELIBERATIONS
AND
RESPONSE
TO
CHARGE
The
specific
issues
addressed
by
the
Panel
are
keyed
to
the
Agency's
background
documents,
references,
and
the
Agency's
charge
questions.

Charge
to
the
Panel
­
MIR604
Environmental
Assessment
The
weight
of
evidence
from
the
reviewed
data
indicates
that
there
will
not
be
a
hazard
to
wildlife
from
the
commercialization
of
Event
MIR604
corn.
Although
the
mCry3A
protein
expressed
by
Event
MIR604
corn
is
known
to
affect
only
coleopteran
insect
species,
EPA
assessed
the
potential
risks,
to
a
wide
variety
of
non­
target
organisms
(
i.
e.
mammals,
birds,
fish,
invertebrates
and
plants),
that
could
potentially
result
from
exposure
to
this
Bt
protein.
The
emphasis
of
this
non­
target
risk
assessment,
however,
was
on
invertebrate
species
that
dominate
corn
agro­
ecosystems.
The
Agency
also
evaluated
a
soil
fate
study
that
was
intended
to
provide
information
on
the
persistence
and
rate
of
degradation
of
mCry3A
protein
in
the
soil
environment.
After
reviewing
data
submitted
in
support
of
the
Event
MIR604
Bt
corn
registration,
EPA
concluded
that
aquatic
and
terrestrial
wildlife,
including
soil
organisms,
were
not
likely
to
be
adversely
affected
and
that
this
Bt
corn
product
is
not
likely
to
threaten
the
long­
term
survival
of
any
non­
target
wildlife
populations.

The
Panel
is
requested
to
comment
on
the
Agency's
analysis
of
the
currently
available
data
on
the
potential
impacts
of
Event
MIR604
corn
on
non­
target
species.

Panel
Response
The
Panel
reviewed
the
supplemental
data
provided
in
support
of
the
Agency's
ecological
risk
assessment.
The
sections
below
describe
the
Panel's
conclusions
with
regard
to
the
data
provided
and
address
specific
components
of
the
assessment
under
each
subtitle.

The
Standard
for
Maximum
Hazard
Dose
Testing
(
MOE>
10).

Several
Panel
members
expressed
the
opinion
that
uncertainty
in
Margin
of
Exposure
(
MOE)
estimates
for
maximum
hazard
dose
testing
should
be
quantified
and
used
in
the
Agency's
ecological
risk
assessment.
Uncertainty
in
estimates
of
MOE
is
by
definition
a
function
of
the
uncertainty
in
Equivalent
Test
Concentration
(
ETC)
estimates
and
Expected
Environmental
Concentration
(
EEC)
estimates
because
MOE=
ETC/
EEC.
When
using
surrogate
test
material,
uncertainty
in
ETC
includes
consideration
of
the
uncertainty
of
the
correction
factor
used
to
adjust
for
the
difference
in
the
effectiveness
between
the
test
material
and
the
protein
as
expressed
in
the
Event
MIR604
corn.
Uncertainty
in
EEC
can
be
addressed
by
making
conservative
assumptions
concerning
the
diets
of
non­
target
organisms
(
NTO)
in
the
field
and
incorporating
the
uncertainty
in
the
estimate
of
mCry3A
protein
concentrations
in
relevant
crop
tissues.
Finally,
computing
the
confidence
interval
(
CI)
for
the
MOE
requires
estimation
of
the
standard
error
of
the
MOE
estimate
which
statistically
is
constructed
as
the
ratio
of
the
two
random
variables.
A
Taylor
series
expansion
approximation
of
the
standard
error
of
this
ratio
can
be
computed
as
demonstrated
in
Cochran
(
1977,
eq
6.13).
13
Several
Panel
members
noted
methodological
and/
or
analytical
inaccuracies
in
estimating
ETCs
and
EECs.
In
some
cases,
it
was
possible
to
estimate
correction
factors
to
account
for
these
inaccuracies.
The
Panel's
comments
on
these
issues
are
organized
under
the
following
headings:
accuracy
and
standard
errors
(
SE)
of
ETCs,
accuracy
and
SEs
of
EECs,
and
95%
CIs
of
adjusted
MOEs.

Accuracy
and
SEs
of
ETCs.
A
primary
issue
discussed
by
the
Panel
was
whether
the
E.
coliderived
MCRY3A­
0102
test
substance
was
equivalent
in
toxicity
to
mCry3A
produced
by
MIR604
corn.
Discussions
of
biochemical
function
suggested
that
there
is
reasonable
cause
to
conclude
that
the
two
proteins
are
not
substantially
equivalent,
with
respect
to
the
NTO
toxicity
tests,
for
the
following
reasons.
About
50%
of
the
mCry3A
in
corn
is
55
kDa,
a
mass
consistent
with
the
activated
form
of
mCry3A;
but
because
the
55­
kDa
form
of
mCry3A
was
not
sequenced,
it
is
not
certain
that
this
is
actually
the
activated
form.
The
bacterially
derived
MCRY3A­
0102,
however,
contains
two
forms,
a
short
form
(
SF)
and
a
long
form
(
LF),
neither
of
which
is
pre­
activated.
Independent
bioassays
of,
respectively,
the
SF
and
LF
against
the
western
corn
rootworm
(
WCRW)
provided
evidence
that
LF
was
much
more
difficult
to
activate,
possibly
due
to
the
additional
amino
acids
at
the
N­
terminus.
It
was
pointed
out
that
to
make
MCRY3A­
0102
equivalent
to
the
MIR604
protein,
about
50%
of
MCRY3A­
0102
proteins
should
have
been
pre­
activated
with
a
cathepsin
(
or
activated
with
some
other
proteases)
in
such
a
way
as
to
yield
about
50%
of
the
protein
in
the
55­
kDa
activated
form
before
being
used
in
the
bioassays.
It
was
also
unclear
whether
the
various
non­
target
organisms,
especially
the
insects
and
other
invertebrates,
have
cathepsins
capable
of
activating
the
long
and
short
forms
of
MCRY3A­
0102.
One
Panelist
noted
that
the
Agency
itself
did
not
seem
to
agree
on
this
issue.
For
example,
the
Agency's
human
health
assessment
considered
the
corn
and
E.
coli­
derived
mCry3A
proteins
to
be
equally
toxic,
whereas
the
Agency's
ecological
risk
assessment
did
not.

The
western
corn
rootworm
LC50s
of
MCRY3A­
0102
and
LPMIR604­
0103,
a
protein
extract
of
mCry3A
from
leaves
of
MIR604
corn,
were
compared
based
on
evaluation
of
the
overlay
of
the
associated
95%
CIs.
The
Panel
concluded
that
this
test
was
inadequate
to
support
the
Registrant's
conclusion
that
the
tested
materials
" 
demonstrated
comparable
activities
 
".
Overlap
of
95%
CIs
is
not
sufficient
to
conclude
statistically
significant
differences
since
the
statistical
significance
of
two
estimates
also
depends
on
effective
sample
sizes
and
the
form
of
the
comparison
test
used.
For
very
low
sample
sizes,
and/
or
when
a
pooled
variance
term
of
the
LC50s
is
used
in
the
hypothesis
test,
substantial
overlap
of
95%
CIs,
even
to
the
degree
that
one
estimate
is
included
in
the
95%
CI
for
the
other
estimate,
can
occur
simultaneously
with
a
finding
of
significant
difference
between
the
population
mean
LC50s.
Also,
there
is
a
curious
asymmetry
in
the
95%
CI
reported
for
the
LC50
of
MCRY3A­
0102.
On
a
log(
dose)
scale,
the
95%
CI
for
the
LC50s
should
be
symmetric.
This
is
approximately
the
case
for
log(
LC50)
of
LPMIR604­
0103
( 
0.7
±
0.3)
but
not
for
log(
LC50)
of
MCRY3A­
0102
( 
0.4 
0.5,
 
0.4+
0.3).
If
one
adjusts
the
lower
limit
of
the
95%
CI
for
LC50
MCRY3A­
0102
to
be
consistent
with
the
upper
limit,
the
95%
CI
changes
from
0.14­
0.94
to
0.20­
0.94
µ
g/
ml,
and
the
clear
inclusion
of
the
LC50
for
LPMIR604­
0103
(
0.20
µ
g/
ml)
within
the
95%
CI
for
MCRY3A­
0102
is
lost.
The
number
of
concentrations
tested
and
the
SEs
of
the
LC50s
are
not
reported,
so
it
is
not
possible
to
perform
the
appropriate
significance
test
on
the
mean
LC50s
or
confirm
the
validity
of
the
correction
of
the
95%
CI.
Conservatively,
one
must
assume
that
the
test
material
MCRY3A­
0102
has
a
toxicity
14
0.46
[=
(
0.20
µ
g
LPMIR604­
0103/
ml
diet)/(
0.43
µ
g
MCRY3A­
0102/
ml
diet)]
times
that
of
mCry3A
from
MIR604
corn.
If
one
makes
the
generous
assumption
that
five
concentrations
were
used
to
estimate
the
LC50s,
then
a
conservative
estimate
of
the
SE
of
this
correction
factor
[
with
the
appropriate
adjustment
for
conversion
from
log(
dose)]
is
0.069
[=
0.46

 
2

(
AVE[((
95%
CIlog)]/
2)/
t0.025(
3)
=
0.46

 
2

0.333/
3.182].
Thus,
except
for
Oncorhyncus,
which
was
dosed
with
feed
formulated
from
MIR604
grain
(
see
below),
the
ETC
values
estimated
by
the
Agency
should
be
multiplied
by
0.46
(
±
0.07)
[
mean
(
±
SE)].

In
the
Eisenia
toxicity
tests,
there
are
several
problems
with
the
ETC
estimates
provided.
First,
solutions
for
the
Leptinotarsa
decemlineata,
Colorado
potato
beetle
(
CPB),
toxicity
tests
to
verify
the
presence
of
mCry3A
in
the
test
soil
were
prepared
by
adding
an
agar
mixture
at
50
°
C
to
the
MCRY3A­
0102­
treated
soil.
This
could
have
reduced
the
effectiveness
of
mCry3A
by
11%
based
on
Registrant
data
for
temperature
stability
of
MCRY3A­
0102.
Second,
toxicity
data
in
the
CPB
tests
show
much
temporal
variation
in
LC50s
that
appears
to
be
contrary
to
standard
acute
dose
responses.
Third,
recovery
of
the
nominal
applied
concentration
of
mCry3A
ranged
from
7­
18%
by
ELISA.
Analyses
that
conform
to
Agency
criteria
generally
require
80­
120%
recovery
(
USEPA,
1996;
see
Tables
13­
15).
Approximate
limits
of
quantification
were
reported
as
0.06
to
0.28
µ
g/
g
for
corn
leaf
and
pollen
(
MRID
4615560­
04).
It
must
be
noted
that
these
limits
are
not
method
of
quantification
limits,
but
rather
instrument
quantification
limits
multiplied
by
a
dilution
factor.
This
does
not
account
for
the
presence
of
interfering
compounds
or
for
matrix
effects
in
the
samples
analyzed,
which
normally
increase
quantification
limits.
One
Panel
member
suggested
that
a
better
method
to
test
for
concentrations
in
samples
is
needed
and
that
quantitative
polymerase
chain
reaction
might
be
a
relevant
approach
to
consider.
Fourth,
the
use
of
crude
extracts
in
analyses
that
depend
on
polyclonal
antibodies
for
toxin
quantification
raised
questions
regarding
interferences.
Finally,
CPB
mortality
did
not
decline
over
time
whereas
the
concentration
of
mCry3A
did.

Un­
aged,
MCRY3A­
0102­
treated
soil
used
in
the
Eisenia
toxicity
test
was
fed
to
CPB
at
rates
of
incorporation
into
the
diet
of
5%
and
10%,
and
the
resulting
mortalities
were
assessed.
In
the
same
experiment,
the
mortality
of
CPB
fed
diet
with
MCRY3A­
0102
added
directly
to
the
test
soil
("
neat")
was
assessed
at
two
concentrations.
(
Note:
the
Agency's
nominal
calculation
of
250
µ
g
MCRY3A­
0102/
g
in
the
test
soil
is
slightly
in
error:
the
correct
value
is
closer
to
the
nominal
value
of
266
µ
g
MCRY3A­
0102/
g
diet.).
The
probit
vs.
log[
dose]
relationship
based
on
the
"
neat"
test
data
suggests
that,
in
the
presence
of
the
test
soil,
MCRY3A­
0102
has
30
(
±
4)%
the
toxicity
it
has
in
the
absence
of
the
soil
(
Appendix
1).
Thus,
in
the
absence
of
dose
verification
with
reasonable
recoveries,
a
case
can
be
made
that
the
ETC
value
estimated
by
the
Agency
for
Eisenia
should
be
multiplied
by
a
factor
of
0.30
(
±
0.04).

In
the
Coccinella
toxicity
test,
Agral
90,
a
wetting
agent
included
in
the
negative
control
treatment,
showed
no
evidence
of
toxicity.
However,
no
data
are
presented
to
support
the
presumption
that
Agral
90
did
not
denature
or
otherwise
reduce
the
toxicity
of
MCRY3A­
0102
in
the
MCRY3A­
0102
treatment.
Agral
90
and
MCRY3A­
0102
should
have
been
tested
against
CPB
as
an
indication
that
mCry3A
toxicity
was
not
reduced
by
Agral
90.
15
In
the
Apis
toxicity
test,
the
test
material
was
administered
as
a
suspension
of
66.4
µ
g
MCRY3A­
0102/
ml
in
50%
aqueous
sucrose
solution.
Viscous
solutions
or
suspensions
are
known
to
increase
the
variability
of
dose
delivery
(
Deicke
and
Suverkrup,
2000;
Elkheshen
et
al.,
1996).
There
are
many
examples
of
concentration
of
the
test
material
in
a
dosing
suspension
being
far
lower
than
the
nominal
concentration
(
ca.
0.7X)
(
Thomas
et
al.,
1987;
Thomas
et
al.,
1990;
RTI
1992;
Brewer
et
al.,
1995;
Ramsey
et
al.,
2003).
Such
an
effect
can
be
caused
by
adsorption
to
solution
containers
and
to
transfer
tools
such
as
pipettes.
Such
errors
lower
the
concentration
of
the
test
material
in
the
test
substrate
and
bias
test
results
toward
a
finding
of
reduced
toxicity.

Accuracy
and
SEs
of
EECs.
Given
the
uncertainties
of
ecological
processes,
the
Agency
made
appropriately
conservative
assumptions
concerning
the
diets
of
NTOs
in
the
field,
and
the
resulting
EECs
are
close
to
maximum
estimates
in
most
cases.
However,
there
was
a
systematic
error
in
correcting
for
extraction
efficiency
of
mCry3A
from
plant
tissues
as
submitted
by
the
Registrant.
The
efficiency
of
extraction
of
mCry3A
from
plant
tissues
using
the
ELISA
assay
was
estimated
by
assuming
that
all
of
the
mCry3A
remaining
in
an
initial
extraction
is
recovered
by
a
second
extraction.
If
one
makes
the
more
realistic
assumption
that
a
second
extraction
recovers
mCry3A
with
the
same
efficiency
as
the
first,
then
estimated
extraction
efficiency
is
reduced,
the
estimated
concentration
of
mCry3A
in
MIR604
corn
tissue
is
increased,
the
estimated
value
of
EEC
is
increased,
and
the
estimated
value
of
MOE
is
decreased
(
Appendix
2).
The
Registrant
reported
extraction
efficiencies
for
leaves,
kernels,
and
silage
of
77.1%,
69.7%,
and
84.5%,
respectively.
Variance
information
for
extraction
efficiency
is
provided
by
the
Registrant
for
leaf
tissue
only.
Assuming
that
the
coefficient
of
variation
(
CV)
for
extraction
efficiency
of
the
other
tissues
is
approximately
that
for
leaves,
then
standard
errors
can
be
estimated
for
all
three.
Applying
the
correction
for
extraction
efficiencies
(
Appendix
2)
and
making
the
appropriate
adjustments
in
standards
errors
for
the
reciprocal
of
random
variables
(
Barford
1967),
the
efficiencies
[
mean
(
±
SE)]
of
first
extraction
of
leaves,
kernels,
and
silage
are
estimated
to
be
70.3
(
±
1.1)%,
56.5
(
±
1.5)%,
and
81.7
(
±
0.8)%,
respectively.
Consequently,
the
Agency
EEC
values
based
on
the
concentration
of
mCry3A
in
MIR604
leaf
tissues
(
Orius,
Coccinella,
Poecilus,
Apis,
and
Eisenia)
should
be
increased
by
10%
[
i.
e.,
multiplied
by
1.10
(
±
0.02)],
those
based
on
kernels
(
Colinus,
Mus,
and
Onchorhyncus,
as
well
as
the
ETC
value
for
the
Onchorhyncus)
by
23%
[
i.
e.,
multiplied
by
1.23
(
±
0.05)],
and
those
based
on
silage
(
Aleochara)
by
3%
[
i.
e.,
multiplied
by
1.03
(
±
0.01)].

The
Agency
made
an
error
in
estimating
EEC
for
Aleochara.
The
Agency's
conservative
assumption
is
that
the
diet
of
Aleochara
in
the
field
is
lepidopteran
larvae
containing
70%
the
concentration
of
mCry3A
observed
in
MIR604
tissue.
Assuming
that
the
relevant
tissue
is
leaf
tissue,
the
resulting
EEC
is
4.7
µ
g/
g.
Thus,
the
Agency's
EEC
estimate
for
Aleochara
of
1.07
µ
g
mCry3A/
g
tissue
should
be
multiplied
by
4.4.

95%
Confidence
Intervals
of
Adjusted
MOEs.
Combining
the
SEs
for
ETCs
and
EECs
for
adjusted
estimates
of
MOEs
(
Barford
1967)
results
in
MOE
<
10
for
five
of
the
nine
NTOs
tested;
Lower
Bound
of
MOE
95%
CI
<
5
for
five
of
the
NTOs;
and
Lower
Bound
of
MOE
95%
16
CI
<
1
for
two
NTOs
(
Table
1).
It
should
be
reiterated
that
the
data
in
Table
1
assume
that
the
nominal
test
concentrations
were
in
fact
the
concentrations
administered.

Table
1.
Adjustment
of
EPA
Estimates
for
Effective
Test
Concentration
(
ETC),
Expected
Environmental
Exposure
(
EEC),
and
Margin
of
Exposure
(
MOE,
MOE
 
ETC/
EEC)
for
Toxicity
Tests
of
Non­
target
Organisms
Submitted
by
the
Registrant
in
Support
of
an
Application
for
Registration
of
MIR604
Corn
Non­
target
Organism
EPA­
Estimated
MOE
Adjusted,
ETC
[(
SE),
µ
g
mCry3A
g­
1
diet]
Adjusted,
EEC
[(
SE),
µ
g
mCry3A
g­
1
diet]
Adjusted
MOE
(
SE)
95%
CI
of
Adjusted
MOE
Orius
10.6
23
(
3)
5.0
(
0.2)
4.6
(
0.7)
2.3­
6.0
Poecilus
11.2
5.5
(
0.8)
5.0
(
0.2)
1.1
(
0.2)
0.55­
1.7
Aleochara
15.6
23
(
3)
3.3
(
0.3)
6.9
(
1.0)
3.6­
10.1
Coccinella
12.3
4.1
(
0.6)
0.78
(
0.04)
5.3
(
0.8)
2.7­
8.0
Apis
36
23
(
3)
1.5
(
0.1)
15
(
2.4)
7.7­
23
Eisenia
46
37
(
7.3)
6.0
(
1.2)
6.2
(
1.7)
0.68­
12
Oncorhyncus
37
0.17
(
0.01)
ETC/
37a
37
­­­­
Colinus
1400
300
(
45)
b
0.55
(
0.05)
c
550
(
100)
230­
860
Mus
5500
1090
(
164)
b
0.52
(
0.05)
c
2100
(
380)
900­
3300
aInsufficient
data
supplied
by
the
Agency
to
calculate.
bTest
dose,
µ
g
mCry3A
g­
1
body
weight.
cDaily
dietary
dose,
µ
g
mCry3A
g­
1
body
weight.

According
to
OPPTS
885.4340
(
Microbial
Pesticide
Test
Guidelines
for
Tier
1
Nontarget
Insect
Testing),
the
minimum
required
concentration
for
maximum
hazard
dose
testing
is
10X
EEC.
The
Panel
was
split
on
the
importance
of
the
Registrant
not
adhering
to
this
standard.
Several
Panel
members
considered
this
a
serious
deficiency
but
a
deficiency
that
easily
could
have
been
avoided.
Other
Panel
members
considered
the
test
concentrations/
doses
adequate
to
support
the
Agency's
finding
of
no
likely
adverse
ecological
effects.

Estimation
of
95%
CIs
for
MOEs,
although
based
on
incomplete
information,
represents
an
enhancement
of
hazard
assessment
that
may
be
useful
in
the
Agency's
overall
risk
assessment
process.
Adding
confidence
intervals
provides
the
basis
to
identify
areas
of
concern
on
which
the
Agency
may
wish
to
have
the
Registrant
focus
additional
research.
In
addition,
it
would
enable
future
Panels
that
may
be
asked
to
consider
related
issues
to
make
more
definitive
decisions
regarding
the
quality
of
the
data
supporting
the
analyses
17
Experimental
Design
and
Statistics.

Most
Panel
members
thought
that
the
experimental
design
and
statistical
analysis/
reporting
in
the
Registrant's
NTO
toxicity
testing
were
frequently
inadequate.
Several
Panel
members
were
of
the
opinion
that
standards
of
experimental
design,
data
analysis,
and
statistical
reporting
in
studies
submitted
in
support
of
requests
for
registration
of
crops
with
PIPs
should
be
no
less
than
those
required
for
publication
in
professional
scientific
journals.
The
Panel's
concerns
are
organized
into
consideration
of
the
following
two
topics:
improper
or
insufficient
replication;
and
failure
to
report
sample
sizes,
SEs,
and
test
statistics/
P­
values
or
even
failure
to
perform
statistical
tests
of
interest.

Replication.
In
toxicity
testing
of
individuals,
one
can
define
a
"
replicate"
as
an
experimental
unit
(
whether
a
single
subject
or
a
group
of
subjects)
from
which
a
value
for
the
random
variable
of
interest
is
generated
independently
of
other
units.
The
experimental
unit
can
be
a
single
subject
when
the
random
variable
of
interest
is
an
individual
trait
(
e.
g.
weight),
or
it
can
be
a
group
of
subjects
when
the
random
variable
of
interest
is
a
group
trait
(
e.
g.
%
mortality).

Mortality
data
can
be
tested
for
differences
among
treatment
groups
using
Fisher's
Exact
Test
or
Chi­
Square 
as
was
done
in
the
vertebrate
NTO
toxicity
tests:
Oncorhyncus,
Colinus,
and
Mus.
Unfortunately,
in
these
tests,
all
individuals
in
a
given
treatment
group
were
housed
in
a
single
tank,
pen,
or
cage,
respectively.
Consequently,
treatment
effects
are
confounded
with
any
differences
between
the
two
groups
due
to
uncontrolled
factors
so
that
the
datum
collected
from
a
given
subject
("
alive"
or
"
dead",
in
this
case)
is
not
independent
of
the
data
collected
from
the
other
subjects.
Although
the
requirement
of
independence
of
data
must
be
qualified
to
require
independence
with
respect
to
sources
of
experimental
error
that
are
considered
likely,
it
was
the
consensus
of
the
Panel
that
the
vertebrate
NTO
toxicity
tests
were
deficient
in
this
respect
(
Table
2).
In
addition,
one
Panel
member
preferred
to
have
replication
of
groups
of
experimental
units
for
tests
involving
percent
mortality
as
an
endpoint,
and
another
expressed
the
opinion
that
none
of
the
NTO
toxicity
tests
satisfied
the
requirement
of
reasonable
independence
of
experimental
units
because
they
had
not
been
repeated.
18
Table
2.
Statistical
Design/
Analysis/
Reporting
in
Non­
target
Organism
and
Auxiliary
Studies
Submitted
by
the
Registrant
in
Support
of
Application
for
Registration
of
MIR604
Corn.

Non­
target
Species/
Study
Valid
Replication
Sufficient
Replication
SEs
Reported
Tests/
Analyses
Performed
and
Valid
Characterization
of
MCRY3A­
0102


X
X
MCRY3A­
0102:
Further
Analysis


X
X
Quantification
of
mCry3a
in
MIR604

X

X
Soil
Degradation
X
X
X
X
Orius
toxicity




Orius
Diet
Validation

X
X
X
Poecilus
Toxicity



X
Poecilus
Diet
Validation

X
X
X
Aleochara
Toxicity




Aleochara
Diet
Validation

X
X
X
Coccinella
Toxicity




Coccinella
Diet
Validation

X
X

Apis

X


Eisenia
Toxicity




Eisenia
Diet
Validation

X
X
X
Oncorhyncus
Toxicity
X


X
Oncorhyncus
Diet
Validation




Colinus
X
X

X
Mus
X
X

X

=
Adequate
X=
Inadequate
Insufficient
replication
reduces
the
value
of
data
presented
as
the
basis
for
making
ecological
risk
assessment
decisions
because
the
power
of
statistical
comparisons
is
insufficient
to
detect
treatment
effects
of
potentially
ecologically­
significant
magnitude.
The
consensus
of
the
Panel
was
that
the
number
of
replicates
or
sample
sizes
in
the
toxicity
tests
of
Colinus
and
Mus
was
too
low:
5
individuals
per
sex
X
treatment
group
combination.
Other
concerns
regarding
insufficient
sample
size
included
the
following:
all
of
the
ELISA
assays
of
NTO
toxicity
test
diets
(
except
Oncorhyncus)
were
unreplicated,
although
in
two
cases
(
Poecilus
and
Coccinella),
units
of
the
test
diet
(
fly
pupae
and
aphids,
respectively)
were
composited
for
the
single
analysis
performed;
the
feed
consumption
rate
observations
for
Colinus
and
Mus
were
unreplicated;
and
the
sample
size
in
the
Apis
toxicity
test
(
N
=
4)
provided
insufficient
statistical
power
given
the
19
high
level
of
variance
observed
in
the
negative
controls
(
Table
2).

Statistics:
Performing
and
Reporting.
Raw
data
were
supplied
for
all
of
the
NTO
toxicity
tests,
so
it
was
possible
for
the
Panel
to
calculate
SEs
and
perform
statistical
tests
even
when
such
statistical
tests
were
not
reported
by
the
Registrant.
However,
in
the
soil
degradation
study
(
except
for
the
microbial
activity
observations)
and
in
all
of
the
CPB
assays
used
in
the
NTO
toxicity
tests
to
verify
the
presence
of
mCry3A
in
the
diets,
the
raw
data
were
not
supplied
and
SEs
were
not
reported.
In
addition,
the
Registrant
apparently
did
not
perform
tests
for
the
sex
X
treatment
effect
for
Poecilus
adult
weight,
the
sex
X
treatment
effect
for
Colinus
weight
gain,
and
differences
in
variance
of
Oncorhycus
weight
gain
over
time
(
Table
2).
As
discussed
above,
a
valid
statistical
test
to
compare
LC50s
of
MCRY3A­
0102
and
LPMIR604­
0103
was
not
performed.
Also,
a
similar
failure
occurred
in
comparing
LC50s
for
MCRY3A­
0102­
SF
and
MCRY3A­
0102­
LF
(
Table
2).
In
the
opinion
of
several
Panel
members,
these
failures
represent
a
serious
deficiency
in
the
value
of
the
evidence
presented
by
the
Registrant.

Non­
lethal
PIP
Effects.

If
one
takes
the
NTO
toxicity
assays
of
this
Tier
I
hazard
assessment
at
face
value,
none
of
them
reported
high
acute
toxicity.
Rather,
in
most
instances
mortality
was
low
or
did
not
occur
at
the
ETCs
achieved.
Although
the
ETCs
frequently
did
not
meet
the
standard
for
maximum
hazard
dose
testing,
these
results
do
support
the
conclusion
of
the
absence
of
a
strong,
toxic
response
in
the
NTOs.
However,
sub­
lethal
endpoints
(
weight
gain)
measured
in
the
toxicity
tests
of
Oncorhyncus
and
Colinus
may
imply
signs
of
stress
that
were
not
translated
into
mortality
within
the
time
frame
and
conditions
of
the
tests
conducted.

The
statistical
analysis
provided
by
the
Registrant
for
the
Oncorhyncus
toxicity
test
is
incomplete.
Although
it
is
stated
that
homogeneity
of
variances
between
treatment
and
control
groups
was
tested,
it
was
not
reported
that
there
was
a
significant
increase
in
the
variances
of
size
and
weight
for
treated
fish
relative
to
untreated
over
the
course
of
the
experiment.
The
variance
of
weight
in
the
treatment
group
was
1.88
times
that
for
the
control
group
by
day
14
(
F
=
1.88;
df
=
39,
39;
P
<
0.05)
and
2.67
times
greater
by
the
end
of
the
test
on
day
28
(
F
=
1.88;
df
=
38,
39;
P
<
0.01)
(
Fig.
1a).
The
variance
of
length
in
the
treatment
group
was
1.60
times
that
for
the
control
group
by
day
14
(
F
=
1.60;
df
=
39,
39;
0.05<
P<
0.10)
and
2.52
times
greater
by
day
28
(
F
=
2.67;
df
=
38,
39;
P
<
0.01)
(
Fig.
1b).
The
increase
in
variance
in
the
treatment
group
could
be
the
result
of
stress
due
to
the
PIP
acting
on
genetic
variability
among
the
test
subjects.
A
test
of
this
hypothesis
can
be
performed
by
examining
the
effect
of
the
PIP
on
a
condition
index
[
CdI
=
Length/(
Wt)
 
]
for
the
data
collected
on
Day
28,
the
day
of
the
final
observation.
For
the
control
group,
CdIC
=
39.92
(
0.133,
40)
[
mean
(
SE,
N)];
and
if
an
outlier
for
CI
is
excluded
from
the
treatment
group
(
Z
=
3.432;
P
=
0.0003),
CdIT
=
39.41
(
0.153,
38).
The
variances
of
CdIT
and
CdIC
are
not
significantly
different
(
F
=
1.27;
df
=
37,
39;
P
=
0.465),
and
H0:
mean[
CdIT]
=
mean[
CdIC]
is
rejected
(
Z
=
2.511;
P
=
0.0142).
Thus,
for
a
given
body
weight,
Oncorhyncus
fed
a
diet
containing
MIR604
corn
grain
were
shorter
than
those
fed
the
control
diet.
20
Fig.
1a.
Weight
and
SD[
Weight]
of
Oncorhyncus
in
Toxicity
Test
Test
Day
AVE[
Wt]
Control
AVE[
Wt}
Trtmt
SD[
Wt]
Control
SD[
Wt]
Trtmt
AVE[
Weight],
g
SD[
Weight],
g
0
10
20
30
0
5
10
0
0.5
1
Fig.
1b.
Length
and
SD[
Length]
of
Oncorhyncus
in
Toxicity
Test
Test
Day
AVE[
Lngth]
Control
AVE[
Lngth}
Trtmt
SD[
Lngth]
Control
SD[
Lngth]
Trtmt
AVE[
Length],
g
SD[
Length],
g
0
10
20
30
40
70
100
0
3
6
The
statistical
analysis
provided
by
the
Registrant
for
the
Colinus
toxicity
test
is
incomplete.
It
was
not
reported
that
there
was
a
significant
interaction
between
Treatment
Group
(
TG)
and
Sex
(
S)
when
weight
change
( 
Wt)
is
the
dependent
variable
in
an
analysis
of
variance
with
TG,
S,
and
TG
x
S
as
sources
of
variation
(
F
=
5.69;
df
=
1,
16;
P
=
0.0298).
For
males
in
21
the
group
fed
feed
containing
MIR604
corn
grain,
 
Wt
was
lower
than
that
in
the
control
group
while
the
reverse
was
true
for
females
(
Fig.
2).

Fig.
2.
Change
in
Weight
of
Colinus
(
by
Sex)
in
NTO
Toxicity
Test
Treatment
Mean
+­
SEM,
g
Control
mCry3A
­
5
0
5
10
Female
Male
In
both
of
the
toxicity
tests
in
which
the
subjects
were
fed
MIR604
corn
itself,
treatment
effects
were
observed.
Whether
these
effects
are
harmful,
beneficial,
or
neutral
for
Oncorhyncus
and
Colinus
is
unclear.
Also,
as
noted
above,
the
validity
of
these
tests
are
in
question
due
to
improper
replication.
However,
several
Panel
members
were
of
the
opinion
that
such
responses
should
trigger
subsequent
hazard
assessments
of
PIPs
at
and
below
1X
EEC.

At
present,
sub­
lethal
endpoints
observed
during
Tier
I
maximum
hazard
dose
testing
of
NTOs
are
given
no
weight
and
follow­
up
testing
is
triggered
only
for
LC50
<
10X
EEC.
However,
sub­
lethal
effects
may
represent
very
important
ecological
information.
For
example,
observations
of
discoloration,
sounding
and
surfacing
in
the
Oncorhyncus
test,
while
not
causing
increased
lab
mortality,
may
significantly
increase
predation
rates
in
the
field.
Redundant
response
across
multiple
taxa,
as
was
observed
here,
should
be
of
particular
concern.
Future
study
designs
should
use
models
that
assess
the
effectiveness
of
copulation,
survival
and
development
of
offspring,
and
subsequent
emergence
success
as
adults.
The
Panel
recommends
that
the
Agency
reevaluate
its
current
approach
with
respect
to
the
weight
given
to
sub­
lethal
effects
observed
in
Tier
1
toxicity
testing.

A
number
of
Panel
members
suggested
that
the
Agency
better
define
its
basic
approach
to
data
analysis
and
interpretation
for
these
studies
(
e.
g.,
is
ANOVA
or
multiple
T­
tests
recommended
for
comparison
of
treatment
groups?).
This
assessment
also
points
to
a
need
to
develop
strategies
for
evaluating
and
using
evidence
in
situations
where
different
statistical
analyses
produce
conflicting
results.

Endangered
Species.
22
Hungerford's
crawling
water
beetle
(
Brychius
hungerfordi;
Coleoptera:
Haliplidae),
an
endangered
beetle
species,
occurs
within
habitats
in
corn
production
areas
of
Michigan's
northern
Lower
Peninsula
where
they
could
encounter
corn
crop
residues.
Both
adults
and
larvae
are
herbivorous
and
live
in
moderate
to
fast­
flowing
streams
of
the
Cheboygan
River
watershed,
but
the
larvae
feed
in
riverbank
soils
affording
potential
exposure
to
plant
residues.
It
is
possible
that
this
endangered
species
could
be
exposed
to
mCry3A
from
MIR604
corn
planted
in
this
region.
Several
panelists
strongly
suggested
that
the
Agency
consider
the
potential
impact
of
MIR604
corn
on
B.
hungerfordi.

Aquatic
Species
Testing.

Oncorhynchus
Toxicity
Test.
Findings
are
generally
in
agreement
with
what
is
known
about
acute
toxicity
of
Bacillus
thuringiensis
(
Bt),
Bacillus
thuringiensis
israeliensis
(
Bti),
and
Bacillus
sphaericus
(
Bs)
in
fish,
fiddler
crabs
and
grass
shrimp.
Typically,
mortality
occurs
at
concentrations
that
are
much
greater
than
the
expected
environmental
concentrations
(
Lee
and
Scott,
1989;
Dee,
1988;
Banks,
1988;
Lee,
1987).
These
studies
estimated
a
MOE
of
approximately
900
for
fish
based
upon
the
NOEC
and
an
EEC
based
on
Bt
application
rates
for
mosquito
control
in
12
inches
of
standing
water.

The
Agency
did
not
consider
the
importance
of
components
of
feed
formulations
on
carrier
molecules
in
the
NTO.
Research
has
shown
that
antioxidants
such
as
ethoxyquin
(
very
common
in
most
commercial
foods)
will
turn
on
carrier
proteins
in
fish,
such
as
p­
glycoproteins
and
multi­
drug
resistance
proteins.
Up­
regulation
of
these
accessory
proteins
could
be
very
important
in
affecting
uptake
of
Bt
and
Cry3A
proteins.
Up­
regulation
may
result
in
either
greater
uptake
or
lowered
uptake
depending
on
the
specific
protein
pathway
affected.
Greater
uptake
may
result
in
an
overestimation
of
toxic
effects
while
down
regulation
may
mean
that
bioeffects
have
been
underestimated.

Waiver
of
Estuarine
and
Marine
Animal
Toxicity
Testing.
The
Agency
maintains
that
estuarine
and
marine
animal
studies
are
not
required
for
this
product,
because
mCry3A
is
not
intended
for
direct
application
to
estuarine
or
marine
environments
and
there
is
very
low
potential
that
these
ecosystems
will
be
exposed
to
mCry3A
protein
in
MIR604
corn.
Although
the
risks
of
exposure
in
marine/
estuarine
systems
may
be
low,
there
could
be
concern
in
areas
where
large
acreages
of
corn
are
grown
in
coastal
plain
habitats
that
discharge
directly
into
receiving
streams.
Based
upon
the
discharge
of
conventional
pesticides
and
plot
studies,
applied
pesticide
losses
are
generally
in
the
<
5%
range,
but
may
approach
20%
during
catastrophic
events
(>
2
inches
of
rain
immediately
after
application).
The
Agency's
overall
risk
assessment
for
aquatic
environments
emphasizes
the
importance
of
pollen
as
the
primary
aquatic
exposure
route,
but
non­
point
source
(
NPS)
runoff
from
corn
fields
has
not
been
excluded
by
direct
study.
The
Agency
should
commission
studies
to
verify
that
indeed
there
is
no
NPS
runoff
risk
from
PIP
corn
into
aquatic
ecosystems
including
marine/
estuarine
ecosystems.
This
could
be
accomplished
most
efficiently
by
sampling
sediments
from
retention
ponds
draining
corn
fields
or
streams
that
border
corn
fields.
An
absence
of
mCry3A
proteins
in
these
sites
would
preclude
their
occurrence
in
more
23
distant
receiving
bodies
of
water.

Some
aquatic
testing
of
marine
organisms
should
be
considered
if
PIP
corn
is
expected
to
be
grown
on
sufficient
acreage
where
events
might
conspire
to
move
a
significant
amount
of
mCry3A
proteins
into
marine
systems
where
it
would
otherwise
not
normally
be
encountered.
Freshwater
fish
and
marine
fish
differ
greatly
in
their
physiology:
marine
fish
may
drink
several
times
their
body
weight
per
day
while
freshwater
fish
do
not.
Agency
Water
Quality
Criteria
documents
have
recognized
this
difference
as
being
extremely
important.
In
addition,
support
for
the
waiver
did
not
appear
to
consider
that
aquatic
plants,
such
as
phytoplankton
and
blue
green
algae,
and
bacteria
have
the
greatest
potential
for
gene
flow
concerns
because
of
their
very
short
growth
cycles.
Finally,
if
the
Agency
does
require
additional
testing
in
marine
systems,
it
would
be
important
to
use
a
species
such
as
the
marine
copepod
Amphiascus
tenuiremis,
which
allows
multi­
generational
assessments
on
growth,
survival,
and
reproduction
as
well
as
population
estimations
using
a
Leslie
matrix
population
model
linked
with
Monte
Carlo
simulation.
However,
in
using
copepods
for
NTO
toxicity
testing
of
PIPs,
it
is
important
to
take
into
account,
by
carefully
defining
control
groups,
that
egg
production
may
be
increased
as
they
apparently
use
the
protein
as
a
source
of
food.

While
much
of
the
presented
data
support
the
claim
for
waiver
of
estuarine
and
marine
animal
toxicity
testing,
the
uncertainties
and
concerns
expressed
above
suggests
that
the
Agency
should
consider
requesting
the
Registrant
provide
additional
information
to
decrease
these
uncertainties
before
consideration
of
a
waiver
request.

Laboratory
Soil
Degradation
Study.

General
Comments.
As
already
identified
by
the
Agency,
this
study
is
not
sufficiently
robust
to
draw
conclusions
with
confidence
regarding
how
long
the
mCry3A
protein
is
likely
to
persist
and
retain
insecticidal
activity
in
different
soils.
Reports
suggest
that:
(
i)
the
Cry1Ab
protein
binds
to
clays
in
soil,
thus
protecting
it
from
bacterial
degradation
and
increasing
its
residence
time
(
Tapp
and
Stotzky,
1998);
(
ii)
the
Cry1Ab
protein
is
released
from
the
roots
of
Bt
corn
(
Saxena
and
Stotzky,
2000);
and
(
iii)
the
rate
of
decomposition
of
Bt
crop
residues
is
lower
compared
to
non­
Bt
crop
residues
(
Saxena
et
al.,
2004;
Flores
et
al.,
2005).
These
reports
have
raised
public
concern
and
increased
scrutiny
in
this
area.
Thus,
the
Registrant
should
supply
a
dataset,
properly
replicated
and
statistically
evaluated,
that
provides
the
likely
upper
and
lower
limits
for
the
longevity
of
the
protein
derived
from
crop
residues
in
soil.
There
is
a
rapidly
expanding
body
of
published
research
that
shows
Bt
residues
containing
other
Cry
proteins
(
e.
g.,
Cry1Ac)
degrade
readily
and
quickly
in
soil
(
Head
et
al.,
2002;
Zwalen
et
al.,
2003).
However,
there
are
also
a
few
studies
reporting
results
to
the
contrary
(
e.
g.,
Palm
et
al.,
1996).
Since
public
concern
is
heightened
on
this
specific
topic,
more
information
should
be
supplied
to
increase
confidence
in
the
low
risk
of
the
protein
persisting
in
soil
post­
cropping.

Reports
supplied
by
the
Registrant
provide
results
of
a
single
study
on
a
single
soil
using
the
purified
bacterially­
derived
MCRY3A­
0102
protein.
As
residues,
and
potentially
root
exudates,
are
a
major
environmental
source
of
this
protein,
soil
degradation
studies
should
be
24
conducted
with
the
residues
themselves,
as
already
proposed
by
the
Agency.
The
characteristic
binding
to
clays,
humic
materials
and
soil
organic
matter
will
differ
between
the
purified
protein
and
its
form
when
released
from,
or
bound
to,
decomposing
residues.
Therefore,
the
likelihood
of
the
protein's
persistence
in
soil
should
be
evaluated
using
the
plant
materials
in
this
case.

Although
this
test
was
meant
to
"
screen"
for
potential
soil
accumulation
and
longer­
term
toxicity
to
susceptible
NTOs,
there
was
insufficient
evidence
supplied
to
project
the
decomposition
kinetics
of
residue­
derived
mCry3A
protein
with
confidence
across
the
range
of
soil
types
on
which
this
crop
may
be
grown.

While
the
Registrant
appears
to
have
followed
many
stipulations
of
the
protocols
accepted
by
the
EPA
for
Tier
1
"
screening",
these
protocols
may
require
some
adjustment
for
future
studies
as
the
Cry
proteins
do
differ
substantially
from
chemical
pesticides
in
their
likely
fate
and
persistence
in
soil.
Although
the
Agency
did
not
appear
to
require
statistically
robust
experimental
designs
in
this
Tier
1
"
screening"
process,
attention
to
detail
and
full
reporting,
at
a
minimum,
should
be
expected.

Specific
comments
about
the
conduct
of
the
laboratory
soil
degradation
study
are
provided
as
follows:

1.
A
single
soil
sample
to
a
depth
of
6"
was
collected
from
a
single
farm
site
and
transported
to
the
laboratory.
a.
Is
this
just
a
`
random'
soil?
Why
was
it
selected?
Information
on
how
the
soil
was
selected
and
collected
is
expected.
From
what
total
area
was
the
soil
collected?
Was
the
sample
a
composite
of
multiple
samples
across
the
field
site?
b.
Information
on
the
soil
history
is
warranted.
What
was
grown
on
this
land
prior
to
sampling?
Was
the
soil
subjected
to
any
treatment
that
might
interfere
with
the
degradation
study
(
enhance
or
reduce
binding
or
metabolism
of
the
protein)?
2.
The
soil
selected
had
several
properties
that
made
it
a
good
choice
as
a
test
soil.
Published
reports
of
the
protein
binding
to
clays
and
soil
organic
matter
(
SOM)
indicate
that
the
protein
is
likely
to
have
increased
persistence
in
soils
with
higher
clay
content,
cation
exchange
capacity
(
CEC)
and
SOM,
thus
higher
clay
soils
should
be
targeted.
The
soil
tested
had
a
clay
content
of
30%,
CEC
of
19.7
meq/
100
g
and
3.8%
SOM,
which
are
within
ranges
where
binding
of
the
Bt
protein
to
the
soil
complex
is
likely.
a.
Five
other
soil
types
which
may
contain
30%
or
greater
clay
contents
(
clay,
silty
clay,
sandy
clay,
clay
loam
and
sandy
clay
loam)
could
be
targeted
for
testing
to
give
better
range
of
persistence
across
varying
environments.
b.
Target
specifically
those
soil
types
that
dominate
the
regions
where
this
crop
will
be
grown.
If
the
dominant
soil
types
have
lower
clay
contents
and
SOM,
then
these
should
be
selected.
c.
Aim
for
a
range
of
soil
types
to
yield
understanding
of
how
the
protein
derived
from
plant
tissue
behaves
in
different
environments.
25
3.
The
soil
was
acclimatized
for
10
days,
which
decreases
sampling
effects
on
soil
respiration
and
stabilizes
the
soil
moisture
content.
All
units
tested
were
sub­
samples
from
this
single
soil
sample.
This
is
typical
for
many
controlled
studies.
4.
The
Registrant
prepared
a
single
aqueous
solution
of
the
protein
in
de­
ionized
water.
Sub­
samples
of
this
solution
were
then
used
to
treat
all
experimental
units.
a.
Is
the
solubility
of
the
MCRY3A­
0102
protein
in
de­
ionized
water
sufficient
to
ensure
that
each
5
ml
sub­
sample
taken
contains
the
concentration
targeted
(
230
µ
g
g­
1
soil)
for
each
experimental
unit
(
50
g
soil)?
5.
Twenty­
four
experimental
units
were
treated
with
the
protein,
and
destructive
sampling
of
2
units
at
0,
1,
3,
7,
12,
30,
45,
60
and
90
days
was
intended.
a.
Data
were
presented
only
for
units
sampled
up
until
30
days.
What
happened
to
the
45,
60
and
90
day
samples?
b.
Samples
from
the
0,
1,
3,
7,
12,
30
day
time­
points
were
frozen
until
used
to
prepare
a
diet
for
follow­
on
toxicity
tests
on
CPB.
6.
Only
1
experimental
unit
was
left
untreated
(
de­
ionized
water
only
added)
as
a
"
control",
and
this
unit
was
repeatedly
sub­
sampled
at
each
time
point.
Thus,
the
control
treatment
was
essentially
not
replicated.
a.
It
is
unclear
how
many
sub­
samples
were
taken
from
this
single
experimental
unit
at
each
time
point.
b.
It
is
unclear
how
many
of
these
control
samples
were
eventually
tested.
c.
How
was
the
mean
value
of
18%
mortality
in
the
untreated
control
derived?
Since
there
is
only
a
single
experimental
unit,
all
sub­
samples
are
repeated
measures,
thus
a
true
variance
for
this
treatment
cannot
be
derived.
7.
An
additional
6
untreated
experimental
units
were
set
up
to
demonstrate
that
there
was
an
active
microbial
community
present
in
the
test
soil.
Substrate
induced
soil
respiration
(
SIR),
using
standard
protocols,
was
measured
on
3
untreated
units
at
0
time.
The
remaining
3
untreated
units
were
incubated
along
with
the
main
body
of
the
experiment
and
then
assayed
for
SIR
at
42
days.
a.
Why
was
the
second
SIR
measurement
made
12
days
after
the
last
reported
data
from
the
main
experiment?
b.
Why
were
only
untreated
soils
tested?
c.
To
demonstrate
that
an
active
microbial
community
potentially
capable
of
degrading
the
protein
is
present
in
the
test
soils,
those
samples
in
which
the
protein
was
incorporated
should
also
be
tested.
It
is
easy
to
do
this
test
and
it
answers
the
question
more
directly.
d.
If
only
3,
1­
g
sub­
samples
from
each
experimental
unit
are
used
for
toxicity
testing
on
CPB,
the
remaining
soil
can
be
used
to
measure
SIR.
e.
Use
of
two
or
more
true
experimental
units,
each
sampled
over
time
would
allow
formal
statistical
analysis
to
estimate
uncertainty
and
test
for
effects/
trends.
8.
At
each
time
point,
2
experimental
units
were
sampled.
Thus,
if
use
of
a
single
aqueous
solution
as
the
starting
treatment
is
considered
sufficiently
independent,
there
are
only
two
replicates
for
each
sampling
time
for
the
treated
soils.
a.
Three,
1­
g
soil
samples
were
taken
from
each
experimental
unit.
These
were
used
to
prepare
CPB
media
that
contained
10%
soil
(
1
g
in
9
ml
of
medium).
26
b.
Each
of
these
preparations
was
used
to
pour
two
Petri
dishes.
Thus,
each
dish
contained
approximately
0.5
g
soil
and
4.5
ml
of
CPB
medium.
c.
For
each
sampling
time,
12
Petri
dishes
were
prepared.
However,
these
represent
duplicate
plates
of
a
single
sample
and
3
sub­
samples
from
each
of
2
experimental
units.
No
matter
how
it
is
viewed,
there
are
never
more
than
2
replicates
per
time
point,
not
the
12
suggested
in
the
report.
9.
Purified
MCRY3A­
0102
protein
was
incorporated
on
its
own
into
the
CPB
medium
as
a
positive
control.
This
treatment
was
replicated
4
times.
No
tests
were
done
on
the
medium
itself
(
without
soil
or
purified
protein
added).
The
mortality
of
CPB
on
the
soil
control
treatment
was
high
(
18%),
but
we
do
not
know
if
this
was
related
to
adding
soil
to
the
medium
or
some
problem
with
the
medium
itself.
It
is
important
to
know
the
level
of
survivorship
under
the
general
experimental
conditions
in
order
to
properly
evaluate
the
mortality
rate
on
the
soil
control
(
and
in
the
treated
soils
and
in
media
to
which
the
protein
was
added
directly).
10.
Both
a
bar
chart
and
a
table
are
presented
with
mortality
values
derived
as
a
straight
mean
of
12
plates
per
time
point
for
the
treated
soils,
a
total
of
4
plates
for
the
purified
protein
across
the
whole
experiment,
and
an
unknown
number
of
control
soil
samples.
The
Agency
and
the
Panel
should
have
access
to
the
full
dataset.
a.
What
was
the
between
plate
and
within
plate
variability
for
the
treated
soils?
b.
What
was
the
variance
observed
between
the
two
actual
replicates?
c.
How
repeatable
are
the
numbers
we
see
in
both
Table
1
and
Figure
1?
d.
The
Registrant
should
supply
all
data,
and
in
particular,
any
data
used
in
a
statistical
analysis.
11.
The
data
used
to
model
the
degradation
kinetics
and
derive
a
DT50
value
are
questionable.
The
"
observed"
values
on
the
graph
were
derived
by
subtracting
the
"
mean"
value
of
the
Day
30
mortality
from
the
"
mean"
mortality
of
CPB
on
treated
soils
at
each
time
point.
This
leads
to
a
calculation
of
zero
mortality
at
Day
30,
but
it
is
not
at
all
certain
that
there
is
no
toxicity
to
CPB
after
30
days
based
on
this
test
system.
a.
A
number
of
different
decline
models
could
have
been
fit
to
these
data
with
the
same
or
similar
precision.
b.
Given
that
the
DT50
value
was
improperly
derived,
the
reported
value
is
not
a
good
indicator
of
what
the
kinetics
of
degradation
are
likely
to
be.
12.
More
details
on
the
CPB
medium
formulation
are
needed.
Four
different
antibiotics
were
added
to
the
medium
at
0.1
ml
g­
1
diet.
No
concentrations
of
these
antibiotics
are
reported.
Thus,
we
have
no
capacity
to
replicate
this
experiment.

Interaction
between
mCry3A
and
Conventional
Insecticides.

Two
additional
recommendations
to
the
Agency
were
made
by
Panel
members
concerning
potential
functional
and
economic
interactions
between
mCry3A
and
conventional
insecticides.

Functional
Interaction.
The
use
of
other
chemicals
in
Bt
corn
needs
to
be
considered.
Studies
of
Bti
and
fenoxycarb
have
shown
greater
than
additive
toxicity
in
grass
shrimp
and
mosquitoes
(
Dee,
1988).
Multiple
factor
interactions,
including
additive
toxicity
with
other
27
chemicals,
must
be
rigorously
evaluated
by
looking
at
conventional
pesticides
already
registered
for
use
on
the
crop
and
conducting
basic
mixture
toxicity
tests.
This
may
happen
with
other
chemicals
used
in
conjunction
with
MIR604
corn
in
the
coastal
zone.

Economic
Interaction.
Information
provided
in
public
testimony
before
the
Panel
indicated
that
the
costs
of
conventional
pesticide
use
has
not
dramatically
changed
with
the
introduction
of
Bt
corn.
While
this
testimony
did
not
provide
quantitative
and
statistical
analyses
of
pesticide
cost/
usage
data,
this
information
is
important
for
assessing
the
relative
risks
and
benefits
of
PIPs
in
general.
The
testimony
described
increased
Bt
corn
yields
as
a
benefit,
but
costs
associated
with
conventional
chemical
pesticide
usage
had
not
dropped
appreciably
in
Bt
corn
production.
While
this
situation
could
be
a
result
of
sequential
corn
planting
without
rotation
to
soybeans,
it
will
affect
the
cost/
benefit
analysis
and
should
be
evaluated.
In
general,
the
Agency
should
consider
quantitatively
assessing
the
effect
PIPs
may
have
on
the
economics
of
conventional
pesticide
use
when
considering
future
PIP
registrations,
if
a
justification
for
discounting
safety
concerns
with
a
PIP
is
asserted
based
upon
decreased
use
of
conventional
pesticides.

Additional
Data
Requirements
The
Agency's
ecological
risk
assessment
states
that
"
The
Agency
has
sufficient
information
to
believe
that
there
is
no
risk
from
the
proposed
uses
of
mCry3A
corn
to
non­
target
wildlife,
aquatic,
and
soil
organisms.
However,
in
response
to
the
August
2002
SAP
recommendations,
the
Agency
is
requesting
supplementary
studies
that
will
evaluate
the
persistence
of
mCry3A
in
the
soil
and
the
long
range
effects
of
cultivation
of
mCry3A
on
the
invertebrate
community
structure
in
corn
fields.
This
will
facilitate
identification
of
potential
adverse
effects
which
may
result
from
long­
term
use
of
this
product."
In
particular,
the
following
two
types
of
supplemental
studies
were
identified:

Ecosystem
effects:
Long­
term
field
studies
should
be
conducted
based
on
recommendations
of
the
August,
2002
SAP.

Soil
fate
studies:
Long­
term
soil
degradation
field
studies
should
be
conducted.
Studies
should
follow
guidelines
outlined
by
the
August
2002
SAP.

Ecosystem
Effects
Studies
The
August
2002
FIFRA
SAP
recommended
field­
scale
census
studies
of
3­
4
years
duration
to
provide
information
on
the
general
ecological
attributes
of
PIP
crops.
In
addition,
they
provided
at
least
13
suggestions
for
improving
the
effectiveness
of
such
studies.
The
consensus
among
current
Panel
members
strongly
supported
the
Agency's
request
for
such
supplemental
studies
because
they
could
provide
an
evaluation
of
the
Agency's
primarily
laboratory­
based
toxicity/
bioassays
and
predictive
modeling
approach
to
environmental
hazard
assessment.
28
Non­
target
impacts
of
PIPs
in
corn
agro­
ecosystems
are
likely
to
affect
ecosystem
function
if
soil
food
web
structures,
ecosystem
working
order,
or
numerical
species
relationships
in
leaf­
and
root­
zone
guilds
are
shifted.
The
temporal
and
spatial
scales
of
these
changes
are
just
as
likely
to
accrue
advantages
as
disadvantages
to
farmers
and
society
but
will
go
undetected
without
sustained
record­
keeping.
Even
though
ecosystem
function
changes
are
likely
to
be
subtle,
they
do
lend
themselves
to
detection
in
routine
Integrated
Pest
Management
(
IPM)
and
Insect
Resistance
Management
(
IRM)
monitoring
systems
tuned
to
observe
and
record
natural
enemies
as
well
as
typical
pest
species
monitored
by
industry
representatives,
IPM
consultants,
extension
agents
and/
or
scouts.
These
effects
compared
and
contrasted
with
similar
measures
in
nearby
conventional
IRM
fields
provide
an
ideal
replicated
contrast.
Such
a
long­
term
ecological
transition
study
may
yield
valuable
insight
especially
where
shifting
national
priorities
lead
to
more
"
corn
on
corn"
acreage
associated
with
biofuels
policies
as
public
testimony
conveyed.

Soil
Fate
Studies
The
August
2002
SAP
recommended
that
the
PIP
persistence
be
monitored
in
the
field
in
at
least
three
different
soil
types
until
they
become
undetectable,
but
for
a
minimum
of
one
growing
season
after
harvest.
The
consensus
among
current
Panel
members
strongly
supported
the
Agency's
request
for
such
supplemental
studies.
However,
the
current
Panel
sees
a
need
for
the
Agency
to
design
the
requisite,
statistically
defensible
tests
that
will
be
used
to
establish
a
condition
of
"
undetectable"
in
these
field
studies.
Public
awareness
and
concern
regarding
the
issue
of
PIP
persistence
in
soil
is
high
and
needs
to
be
addressed.
However,
the
time
and
resources
of
the
Registrant,
the
Agency,
and
independent
reviewers
will
be
completely
wasted
unless
said
field
trials
are
properly
designed,
replicated,
and
sampled
and
unless
the
samples
are
analyzed
in
a
robust
experimental
system
with
proper
design
and
replication.
Attention
to
statistical
and
analytical
designs
in
the
field
and
laboratory
is
urgently
needed.

First,
soil
types
tested
in
proposed
field
trials
should
not
be
restricted
to
those
with
high
clay
contents,
especially
if
MIR604
corn
is
unlikely
to
be
planted
on
these
soil
types.
Soils
chosen
should
represent
at
least
the
three
major
soil
types
on
which
this
crop
will
be
grown.

Second,
to
assess
the
persistence
of
the
mCry3A
protein
(
derived
from
plant
tissue
or
possibly
root
exudates)
through
successive
cropping
years
and
the
duration
of
its
bioactivity
in
soil
requires
that
a
method
be
developed
to
track
proteins
released
from
residues
of
varying
age.
The
Agency
has
requested
that
the
Registrant
sample
fields
where
MIR604
corn
has
been
grown
for
at
least
3
years.
If
the
protein
is
not
bio­
available
(
no
detection)
after
a
specified
time
postcropping
then
age
of
inputs
is
irrelevant.
However,
if
active
protein
is
detected
in
tested
soils,
how
will
the
age
of
the
proteins
detected
be
determined
to
establish
`
longevity'?
How
will
this
potentially
confounding
factor
be
handled?
The
Agency
should
be
sure
the
design
is
sufficiently
robust
to
generate
the
data
requested,
such
that
it
will
satisfy
the
data
requirements
of
the
Agency
and
enable
more
thorough
reviews
by
any
future
panels
that
may
consider
related
issues.
29
Third,
the
temporal
bioavailability
of
Cry
protein
in
soil
is
a
factor
that
will
strongly
influence
toxicity
to
NTOs.
Thus,
results
of
the
CPB
assay
are
more
relevant
in
field
studies
than
are
ELISA
assays.
The
ELISA
results
are
clearly
confounded
by
low
extraction
efficiencies
and
high
interference
from
other
soil
components
and
thus
not
sufficiently
robust
for
use
in
field
tests.
Other
methods
to
track
protein
persistence
in
soil
should
be
developed.

Charge
to
the
Panel
­
MIR604
Human
Health
and
Characterization
­
Protein
Equivalence
Previously
submitted
studies
for
Event
MIR604
demonstrated
the
equivalence
of
the
plant­
and
bacterial­
produced
test
substances
by
showing
similar
molecular
weights,
purity
of
90.3%,
a
lack
of
post­
translational
glycosylation
of
mCry3A
from
either
source,
and
comparable
toxicities
toward
western
corn
rootworm
(
WCRW).
However,
two
forms
of
mCry3A
(
designated
as
mCry3A­
sf
and
mCry3A­
lf)
were
found
in
the
bacterial­
produced
test
material
(
MCRY3A­
0102).
The
molecular
weights
of
the
short
and
long
form
of
mCry3A
were
67.5
kDa
and
69.1
kDa,
respectively,
determined
via
SDS­
PAGE
(
sodium
dodecyl
sulfate
­
polyacrylamide
gel
electrophoresis)
and
MALDI­
TOF
(
Matrix
Assisted
Laser
Desorption/
Ionization­
Time
of
Flight)
mass
spectrometry.
The
lesser
of
the
two
components,
with
the
lower
molecular
weight,
corresponded
to
the
intended
mCry3A
protein
with
598
amino
acids.
The
other
component
contained
the
same
598
amino
acids
as
the
first
component
but
also
contained
an
additional
16
amino
acids
at
the
N 
terminal
end
of
the
protein.
Both
mCry3A
forms
were
insecticidally
active
against
WCRW.
On
this
basis,
and
taking
into
account
the
high
degree
of
structural
homology
(
97.4%
amino
acid
identity),
the
two
forms
of
mCry3A
in
test
material
MCRY3A­
0102
were
considered
to
be
equivalent.

Please
comment
on
the
Agency's
conclusion
that
the
mCry3A
proteins
from
corn
event
MIR604
and
from
recombinant
E.
coli
are
substantially
the
same
for
the
purpose
of
the
Agency's
risk
assessment;
and
that
the
equivalence
is
confirmed
for
the
two
forms
of
the
bacterial­
produced
mCry3A
test
material
(
MCRY3A­
0102).

Panel
Response
Cry3A
is
a
protein
endotoxin
produced
naturally
by
the
tenebrionis
strain
of
Bacillus
thuringiensis
subsp.
morrisoni
(
H
8a8b)
that
upon
ingestion
and
proteolytic
activation
exhibits
a
high
level
of
insecticidal
activity
to
certain
species
of
coleopteran
insects,
but
otherwise
is
apparently
non­
toxic
to
other
types
of
insects,
as
well
as
vertebrates.
The
basic
biology
of
this
protein
is
reasonably
well
known.
The
wild
type
protein
is
produced
as
a
67
kDa
protoxin
protein,
and
has
been
engineered
into
other
subspecies
of
B.
thuringiensis,
E.
coli,
and
a
few
crop
plant
species,
mainly
potato
and
corn.
The
wild
type
protein
has
served
as
a
model
Cry
toxin
for
basic
studies
of
Bt,
and
was
the
first
Bt
endotoxin
for
which
the
3­
dimensional
structure
was
solved.
The
availability
of
this
structure
along
with
sitedirected
mutational
analysis
of
Cry3A
and
other
Cry
proteins
based
on
this
structure
provided
important
insights
into
the
selectivity
and
mode
of
action
of
these
insecticidal
proteins.
Specifically,
these
studies
showed
that
the
selectivity
of
binding
to
receptors
on
the
midgut
microvilli
is
controlled
by
domains
II
and
III
of
this
three­
domain
molecule.
After
ingestion
by
a
sensitive
insect,
the
molecule
must
be
30
activated
by
proteolytic
cleavage
carried
out
by
midgut
proteases.
This
activation
by
proteolytic
cleavage
occurs
at
both
the
N­
and
C­
termini.
Once
the
protoxin
molecule
is
activated,
the
protein
binds
to
specific
receptors
on
midgut
microvilli,
typically
glycosylated
enzymes
or
glycolipids,
after
which
it
inserts
into
the
membrane,
oligomerizes,
and
forms
cation­
selective
pores
that
lead
to
midgut
cell
lysis
and
insect
death.
The
Cry3A
midgut
receptor
in
coleopteran
insects
has
not
been
identified,
but
for
lepidopteran
insects,
depending
on
the
species,
receptors
include
amino
peptidases,
alkaline
phosphatases,
and
cadherins.

The
Event
MIR604
corn
being
considered
for
registration
and
the
subject
of
this
SAP,
contains
a
modified
Cry3A
in
which
a
cathepsin
G
cleavage
site
was
created
beginning
at
amino
acid
155
by
replacing
the
valine­
serine­
serine
sequence
with
alanine­
alanine­
proline
followed
by
an
additional
amino
acid,
phenylalanine.
The
new
cleavage
site
occurs
between
alpha­
helices
3
and
4
of
domain
I,
just
upstream
from
the
trypsin
and
chymotrypsin
sites
present
in
wild
type
Cry3A.
The
Registrant
named
this
genetically
modified
endotoxin
"
mCry3A"
to
differentiate
it
from
the
wild
type
Cry3A.
The
cathepsin
cleavage
site
was
introduced
into
the
molecule
because
it
enhances
the
toxicity
of
Cry3A
against
the
western
corn
rootworm,
Diabrotica
virgifera
virgifera,
which
apparently
uses
a
cathepsin
protease
as
one
of
its
key
midgut
proteases.
In
vitro
digestion
experiments
comparing
cleavage
of
mCry3A
versus
Cry3A
with
chymotrypsin
show
that
both
are
cleaved
to
a
55­
kDa
protein.
However,
this
cleavage
occurs
much
more
quickly,
within
30
minutes,
for
mCry3A,
compared
to
four
hours
for
wild
type
Cry3A.
The
receptor
binding
regions
in
domains
II
and
III
are
the
same
in
Cry3A
and
mCry3A,
and
thus
the
binding
properties
of
these
two
molecules
­
the
primary
regions
of
the
molecule
responsible
for
insect
target
spectrum
­
should
be
substantially
similar,
if
not
the
same.
Thus,
the
addition
of
a
cathepsin
cleavage
site
should
not
alter
the
spectrum
of
activity
of
mCry3A
compared
to
that
of
Cry3A.
It
is,
however,
possible
that
this
cleavage
site
could
decrease
the
environmental
stability
of
mCry3A
compared
to
Cry3A.
In
fact,
the
finding
by
the
Registrant
that
the
dominant
forms
of
mCry3A
in
event
MIR604
corn
are
peptides
of
67
and
55
kDa
in
approximately
equal
amounts
shows
that
the
protein
is
cleaved/
activated
in
corn
tissue,
and
therefore
could
be
considered
"
less
stable"
in
the
plant
compared
to
Cry3A.
This
processing
in
plant
tissue
to
the
activated
form
of
mCry3A
probably
accounts
for
a
substantial
portion
of
the
higher
activity
of
mCry3A
corn
extract
(
LPMIR604­
0103)
compared
to
mCry3A
(
MCRY3A­
0102)
produced
in
E.
coli,
as
determined
by
the
Registrant
in
bioassays
against
first
instar
western
corn
rootworms.

In
addition
to
adding
a
cathepsin
cleavage
site,
to
optimize
synthesis
in
corn,
the
Registrant
constructed
a
synthetic
mCry3A
gene
based
on
plant
codon
usage.
The
wild
type
Cry3A
gene
encodes
an
endotoxin
protein
of
73­
kDa.
Rather
than
synthesize
the
entire
gene,
as
the
full­
length
protein
encoded
by
the
wild
type
gene
is
not
required
for
toxicity,
the
Registrant
synthesized
a
truncated
gene
that
encodes
a
protein
beginning
at
methionine­
48
of
wild
type
Cry3A.
For
expression
in
E.
coli,
the
gene
was
cloned
into
a
pUC
expression
vector
that
used
a
cry1A
promoter
system
that
included
16
amino
acids
upstream
from
the
67.7
mCry3A
open
reading
frame.
Expression
of
this
construct
in
E.
coli
yielded
two
forms
of
the
mCry3A,
one
of
the
predicted
mass,
67.7
kDa,
referred
to
as
the
short
form
(
mCry3A­
SF),
and
a
second
product
of
69.5
kDa,
referred
to
as
the
long
form
(
mCry3A­
LF).
These
occurred
in
a
ratio
of
2
parts
SF
to
3
parts
LF
in
the
mCry3A
inclusions
isolated
from
E.
coli.
Importantly,
the
long
form
of
mCry3A
appeared
significantly
less
toxic
(
LC50
=
361
µ
g/
ml
diet)
than
the
short
form
(
LC50
=
82
µ
g/
ml
diet)
to
first
instar
western
corn
rootworms.
However,
the
Registrant
noted
this
difference
was
within
the
95%
confidence
level
of
the
assay
and
did
not
consider
these
forms
different
in
activity.
The
31
substantially
lower
toxicity
of
the
long
form
in
the
E.
coli­
derived
mCry3A,
especially
compared
to
the
corn­
derived
mCry3A,
of
which
approximately
50%
was
already
in
an
activated
form,
is
another
factor
that
likely
contributed
to
the
approximately
2­
fold
difference
between
the
toxicity
values
of
the
cornderived
mCry3A
(
LPMIR604­
0103,
LC50
=
0.20
µ
g/
ml
diet)
versus
the
E.
coli­
derived
protein
(
MCRY3A­
0102,
LC50
=
0.43
µ
g/
ml
diet)
against
first
instars
of
the
western
corn
rootworm.

Owing
to
the
low
levels
of
mCry3A
synthesized
in
corn
tissue,
it
is
impractical
to
purify
sufficient
quantities
of
this
protein
for
Tier
I
safety
studies
carried
out
against
non­
target
invertebrates
and
vertebrates.
Thus,
the
E.
coli­
derived
mCry3A
is
used
as
a
surrogate
protein
for
the
corn
mCry3A
in
these
studies.
The
quantities
of
E.
coli­
derived
mCry3A
to
be
used
in
these
studies
is
set
by
using
multiples
of
10X
or
higher
than
that
which
the
non­
target
test
species
are
likely
to
encounter
in
nature.
Therefore,
the
equivalence
of
the
E.
coli­
derived
and
corn­
derived
mCry3A
along
with
the
amount
of
mCry3A
that
occurs
in
corn
are
important
determinants
of
the
quantities
of
E.
coli­
derived
mCry3A
to
be
used
in
non­
target
tests.

The
presence
of
two
distinct
forms
of
the
microbially­
produced
test
material
which
differ
in
molecular
weight
is
due
to
the
presence
of
an
additional
16
amino
acids
at
the
N­
terminus
of
the
short
form
of
the
protein.
The
long
form
is
due
to
the
presence
of
an
ATG
start
codon
in
the
promoter
sequence
which
is
in
addition
to
the
desired
start
codon.

It
is
regrettable
that
the
presence
of
this
unwanted
start
codon
was
not
detected
prior
to
the
fullscale
production
of
the
test
substance
and
initiation
of
biological
testing.
It
would
have
been
a
trivial
matter
to
have
scanned
the
sequence
proposed
to
be
incorporated
into
E.
coli
to
detect
any
part
of
the
sequence
that
could
cause
potential
perturbations.
This
would
have
averted
the
need
to
consider
the
potential
negative
effects
of
the
non­
native
(
i.
e.
long
version)
protein
on
health.
It
is
strongly
recommended
that
such
screening
for
untoward
effects
be
carried
out
in
the
future
prior
to
the
full
production
of
the
test
substance.
In
the
same
vein,
it
is
also
strongly
recommended
that
highly
discriminatory
characterization
tools,
such
as
MALDI­
TOF,
be
applied
to
test
substances
such
as
the
mCry3A
protein
prior
to
the
use
of
the
substance
in
risk
assessment
trials.

As
noted
above,
the
test
material
included
two
forms
of
the
protein
 
one
that
was
equivalent
to
the
intended
plant
expression
product
and
one
with
additional
N­
terminal
amino
acids.
Because
the
longer
form
is
not
produced
in
the
plant,
and
will
not
be
consumed,
the
relevant
issue
is
not
the
human
safety
of
this
form,
but
whether
the
presence
of
the
longer
form
affects
our
ability
to
consider
the
bacterially­
produced
mixed
material
as
equivalent
to
the
shorter
protein
in
the
plant.
Since
the
longer
form
was
the
more
predominant
form
derived
from
E.
coli
(
the
long/
short
ratio
was
3:
2),
there
is
concern
that
testing
may
have
been
done
with
a
potentially
less
active
form
of
the
mCry3A
protein.

As
a
general
consideration,
it
would
be
unlikely
that
addition
of
the
extra
N­
terminal
sequence
would
affect
the
biological
activity
of
a
non­
glycosylated
protein
such
as
mCry3A
in
a
major
way,
since
many
previous
studies
on
a
variety
of
proteins
have
indicated
that
short
N­
or
C­
terminal
extensions,
including
materials
such
as
immunological
flags,
markers
such
as
green
fluorescent
protein
and
purification
aids
such
as
His­
tags,
added
at
either
the
N­
or
C­
terminus
typically
have
negligible
effects
on
activity.
However,
one
is
reminded
that
even
single
amino
acid
substitutions
in
a
protein
can
have
major
32
deleterious
effects
on
occasion
(
for
example
the
change
in
sickle
cell
hemoglobin
and
other
single
site
mutations
in
a
variety
of
genetic
diseases).

In
the
present
instance
one
might
anticipate
that
the
presence
of
the
additional
16
amino
acids
at
the
N­
terminus
could
affect
the
folding
of
the
protein.
This
in
turn
might
affect
the
biological
activity
if
the
protein
folded
in
an
incorrect
manner.
In
the
present
instance
this
might
have
involved
differences
in
the
propensity
to
crystallize.
This
issue
was
addressed,
to
a
degree,
by
the
submissions
from
the
Registrant.
However,
it
was
found
that
the
long
form
of
the
protein
was
more
than
four­
fold
less
active
than
the
short
form
in
at
least
one
study
with
WCRW.
This
difference
in
activity
causes
difficulties
in
analysis
of
other
test
results
for
both
the
environmental
and
health
effects
studies.
However,
it
was
stated
that
the
relative
activities
in
the
WCRW
test
were
within
the
95%
confidence
levels.
Given
the
experimental
difficulties
of
working
with
WCRW
it
was
suggested
to
the
Panel
that
from
these
data
the
long
form
and
short
form
of
the
protein
are
substantially
equivalent.

However,
from
the
data
presented
it
is
not
possible
to
ascertain
whether
the
two
forms
are
essentially
identical
in
bioactivity
(
refer
to
Panel
response
to
ecological
risk
assessment).
No
adverse
affects
were
seen
in
a
mouse
study.
Therefore,
when
evaluating
mammalian
data
this
issue
seems
to
be
of
greatest
importance
in
calculating
the
level
of
exposure
in
the
mouse
studies.
Should
human
exposure
calculations
be
based
on
the
total
protein
level,
or
only
the
level
of
the
short
form?
The
most
conservative
approach
would
be
to
calculate
exposures
based
only
on
the
amount
of
the
short
form
present.

Moreover,
with
respect
to
the
equivalence
of
the
two
proteins
based
on
the
toxicity
of
the
cornderived
versus
E.
coli­
derived
mCry3A,
the
approximately
2­
fold
higher
activity
of
the
corn­
derived
mCry3A
toxicity
is
likely
due
to
the
factors
noted
above.
Specifically,
these
factors
are
that
(
1)
approximately
50%
of
the
mCry3A
in
corn
is
activated,
i.
e.,
cleaved
to
the
activated
55­
kDa
form
after
synthesis,
and
(
2)
the
E.
coli­
derived
mCry3A
contains
two
types
of
unactivated
mCry3A,
a
short
form
(
67.7
kDa)
and
a
long
form
(
69.5
kDa),
the
latter
of
which
is
of
considerably
lower
toxicity
than
the
short
form.
Whereas
about
50%
of
the
corn­
derived
mCry3A
is
activated,
all
of
the
E.
coli­
derived
mCry3A
must
be
activated.
The
bioassay
data
noted
above
suggest
that
this
occurs
much
more
slowly
in
the
long
form
of
mCry3A,
and
possibly
to
a
significantly
reduced
extent.
In
the
end,
all
of
the
mCry3A,
whether
corn­
derived
or
E.
coli­
derived,
should
be
processed
to
the
activated
55­
kDa
form.
However,
it
is
reasonable
to
assume,
based
on
the
bioassay
data,
and
aside
from
the
problems
with
the
statistical
analyses
discussed
elsewhere
in
this
report,
that
the
differences
in
the
LC50s
of
the
corn­
derived
and
E.
coli­
derived
mCry3A
are
due
largely
to
the
differences
in
the
activated
state
of
these
proteins
upon
consumption,
and
the
rate
at
which
the
two
forms
of
the
E.
coli­
derived
protein
are
processed
after
ingestion
by
first
instar
western
corn
rootworm.
From
the
standpoint
of
toxicity,
therefore,
it
can
be
concluded
that
the
E.
coli­
derived
and
corn­
derived
mCry3A
are
substantially
equivalent
with
respect
to
their
likely
effects
on
the
western
corn
rootworm.
However,
this
equivalence
may
not
transfer
to
nontarget
species
that
do
not
have
cathepsin
to
readily
activate
Cry
proteins.
33
The
question
of
the
relative
bioactivity
of
the
long
and
short
forms
of
mCry3A
could
have
been
further
addressed
by
the
use
of
insects
such
as
the
Colorado
potato
beetle,
which
are
sensitive
to
mCry3A
but
which
show
markedly
less
variability
in
bioassay
studies.
It
is
also
recommended
that
a
receptor
binding
assay
be
developed
where
the
long
and
short
forms
of
E.
coli
expressed
mCry3A
could
be
more
quantitatively
compared
to
each
other
as
well
as
to
corn
derived
mCry3A.
This
would
not
totally
rule
out
the
possibility
that
once
the
receptor
is
bound
the
long
form
might
still
be
less
active,
but
it
could
provide
additional
criteria
that
the
long
and
short
forms
have
equivalent
biological
activity.

Additionally,
the
Panel
thought
that
the
relationship
between
the
bacterial
test
material
and
the
protein
produced
in
planta
needs
to
be
considered.
The
presence
of
a
significant
amount
of
a
putative
break­
down
product,
apparently
the
activated
form
of
mCry3A,
in
the
plant
material
represents
a
difference
between
the
test
material
and
the
in
planta
material.
No
data
were
presented
suggesting
that
this
break
down
product
was
characterized
in
any
way.
As
far
as
the
Panel
can
tell,
the
only
information
available
is
that
this
form
reacts
with
the
same
antibodies
as
the
intact
form
in
a
Western
blot.
The
Panel
believes
that
this
form
should
be
characterized
to
insure
that
it
is,
in
fact,
a
simple
breakdown
product
and
that
it
has
not
been
modified
in
any
other
way.
Without
more
information
on
this
protein,
it
is
difficult
to
conclude
that
the
test
material
was
fully
equivalent
to
the
plant
material.

In
addition,
whether
the
surrogate
mCry3A
produced
in
E.
coli
and
used
in
all
target
and
nontarget
studies
is
substantially
equivalent
to
the
mCry3A
produced
in
MIR604
corn
is
another
matter.
In
fact,
there
is
reasonable
cause
not
to
consider
these
two
forms
of
mCry3A
substantially
equivalent.
Specifically,
about
50%
of
the
mCry3A
in
corn
is
already
activated,
whereas
the
E.
coli­
derived
mCry3A
contains
two
forms
(
SF,
short
form;
LF,
long
form),
with
neither
form
being
pre­
activated.
Moreover,
the
independent
bioassays
of,
respectively,
the
SF
and
LF
against
the
WCRW
conducted
by
the
Registrant
provide
evidence
that
the
latter
form
is
much
more
difficult
to
activate
perhaps
due
to
the
additional
amino
acids
at
the
N­
terminus.
Therefore,
to
make
the
mCry3A
produced
in
E.
coli
equivalent,
about
50%
of
the
E.
coli­
derived
mCry3A
should
have
been
pre­
activated
with
a
cathepsin
before
being
used
in
bioassays
(
or
activated
with
other
proteases
to
yield
about
50%
of
the
protein
in
the
55­
kDa
activated
form).
Another
reason
for
doing
this
would
be
that
it
is
not
known
whether
the
various
non­
target
organisms,
especially
the
insects
and
other
invertebrates,
have
cathepsins
capable
of
activating
the
long
and
short
forms
of
mCry3A.

The
Panel
had
some
questions
regarding
information
in
MRID
461556­
05.
In
this
document,
it
is
noted
in
the
original
analysis
that
a
single
N­
terminal
sequence
was
obtained
for
the
E.
coli
expressed
mCry3A
protein
which
corresponds
to
the
expected
N­
terminal
sequence.
However
in
an
addendum
it
notes
two
proteins
referred
to
as
the
long
and
short
forms
of
E.
coli
expressed
mCry3A
that
have
different
N­
terminal
sequences.

Although
the
fact
that
the
sequences
of
the
two
forms
of
the
protein
are
97%
identical
is
a
strong
argument
for
equivalence,
it
is
important
to
realize
that
this
similarity
has
a
different
biological
meaning
if
the
differences
are
distributed
throughout
the
protein
or
concentrated
in
one
region.
Interestingly,
the
data
submitted
by
the
Registrant
describing
the
mass
spectrometry
studies
show
that
it
was
necessary
to
34
digest
the
two
forms
with
different
enzymes
to
release
N­
terminal
fragments,
suggesting
that
there
may
be
structural
differences
between
these
forms.

The
fact
that
the
proteins
(
i.
e.
the
long
and
short
forms)
have
97.4%
amino
acid
identity
was
considered
of
less
consequence
than
the
fact
that
both
forms
do
have
substantial
biological
activity,
albeit
with
an
approximately
four­
fold
difference
between
them.

Based
on
the
data
available
it
is
likely,
yet
not
fully
proven,
that
the
two
forms
of
the
mCry3A
are
of
relatively
comparable
biological
activity
for
the
purposes
of
the
human
health
assessments.
That
is,
the
mCry3A
proteins
are
substantially
equivalent
with
respect
to
their
amino
acid
sequences,
lack
of
glycosylation,
other
types
of
potential
post­
translational
modifications
and
general
stability.
There
is,
however,
a
question
as
to
the
equivalence
of
how
readily
they
are
activated.
Thus,
for
the
purposes
of
NTO
vertebrate
(
including
human
health
assessment)
and
non­
insect
invertebrate
toxicity
assessment,
data
from
Tier
I
non­
target
effects
studies
based
on
mCry3A
produced
in
E.
coli,
which
in
no
case
showed
existence
of
a
hazard,
may
be
assumed
to
be
equivalent
to
those
that
would
have
been
obtained
had
mCry3A
from
transformed
corn
been
used
to
conduct
these
studies.
The
situation
would
be
much
improved
if
more
complete
data
were
available.
For
example,
in
order
to
more
firmly
draw
this
conclusion,
the
data
supplied
for
the
plant
and
E.
coli
expression
systems
used
to
produce
mCry3A
and
the
associated
bioassays
using
the
WCRW
would
require
a
more
detailed
analysis
than
that
provided
in
the
various
materials
supplied
to
EPA
by
the
Registrant.

As
regards
the
mammalian
and
human
effects
of
the
mCry3A
protein,
several
Panel
members
noted
that
with
respect
to
the
equivalence
of
both
forms
of
the
protein,
there
are
no
data
in
the
mammalian
testing
that
utilized
each
of
the
forms
separately
(
the
long
form
and
the
short
form).
The
testing
was
done
with
a
mixture
that
included
a
ratio
of
3:
2
of
long
form
versus
short
form.
It
has
been
stated
that
the
form
expressed
in
the
plant
is
the
short
form
of
the
protein.

One
Panel
member
noted
that
no
data
were
provided
in
the
mouse
study
on
equivalence
and
since
there
is
a
16
amino
acid
difference
between
the
two
proteins,
the
following
three
concerns
were
raised:
1)
It
is
well
known
that
two
different
proteins
are
not
guaranteed
to
have
the
same
activity
in
human
and
other
mammals.
2)
While
no
allergenicity
testing
was
done
in
animals,
it
was
noted
that
a
difference
of
one
amino
acid
may
make
the
difference
between
an
allergenic
protein
and
a
non­
allergenic
protein.
This
is
particularly
relevant
if
the
different
proteins
fold
differently,
which
may
bring
different
sequences
of
amino
acids
together
in
proximity
and
form
new
allergenic
epitopes.
3)
A
difference
of
one
amino
acid
may
make
a
difference
in
the
degree
of
toxicity.
Even
proteins
with
completely
similar
structure
(
no
amino
acid
differences)
but
of
different
spatial
configurations
can
have
vast
differences
in
their
toxic
effects
on
humans
and
animals.
So,
without
testing,
one
cannot
assume
equivalence.

This
Panel
member
believed
that,
given
the
research
presented
(
feeding
of
10
mice
with
the
test
substance
that
contained
a
mixture
of
the
short
and
long
forms
of
the
protein
and
that
there
were
no
other
studies
presented
that
utilized
the
short
form
alone
or
compared
the
activity
of
both
forms
in
a
live
mammal),
no
data
were
presented
that
would
allow
them
to
state
that
the
Agency's
conclusion
is
justified,
namely,
that
the
mCry3A
proteins
from
corn
event
MIR604
and
from
recombinant
E.
coli
are
35
substantially
the
same
for
the
purpose
of
the
Agency's
risk
assessment;
and
that
the
equivalence
is
confirmed
for
the
two
forms
of
the
bacterial­
produced
mCry3A
test
material
(
MCRY3A­
0102)
as
regards
toxicity
and
allergenicity
to
mammals
and
humans
within
the
scope
of
the
Agency's
risk
assessment.
That
Panelist
suggested
that
the
Agency
proceed
with
testing
the
short
form
in
a
rigorous
and
well
controlled
study
design.

Charge
to
Panel
­
MIR604
Human
Health
and
Characterization
­
Mammalian
Toxicity
Previously
submitted
studies
demonstrated
the
lack
of
toxicity
of
the
mCry3A
protein
following
acute
oral
high­
dose
exposure
to
mice,
rapid
degradation
of
mCry3A
upon
exposure
to
simulated
mammalian
gastric
fluid,
and
the
lack
of
significant
amino
acid
sequence
homology
of
the
mCry3A
protein
to
proteins
known
to
be
mammalian
toxins
or
human
allergens.
Moreover,
little
to
no
human
dietary
exposure
to
mCry3A
protein
is
expected
to
occur
via
transformed
corn.
Therefore,
dietary
exposure
to
mCry3A
is
not
anticipated
to
pose
any
dietary
risk
to
the
U.
S.
population.

Please
comment
on
the
Agency's
conclusions
regarding
the
lack
of
mammalian
toxicity
and
allergenicity
of
mCry3A.

Panel
Response
The
Panel
agreed
that
the
assessment
of
potential
human
toxicity
rested
on
five
analyses:
acute
studies
in
mice,
evidence
of
rapid
degradation
in­
vitro,
sequence
analysis,
analysis
of
glycosylation,
and
an
exposure
assessment.

Several
Panel
members
believed
that
the
results
of
the
acute
mouse
studies
reasonably
support
the
conclusion
of
the
risk
assessment.
Given
that
the
number
of
mice
tested
is
considered
adequate
by
EPA
as
a
point
estimate
of
LD50,
the
data
show
a
lack
of
adverse
effects
at
the
concentrations
tested.
The
only
issue
that
these
Panel
members
identified
was
the
exact
calculation
of
exposure,
based
on
the
question
of
the
long­
form
of
the
protein
discussed
in
the
previous
question.
These
Panel
members
also
noted
that
it
is
important
to
obtain
data
showing
that
the
shorter
form
of
the
protein
found
in
planta
is,
in
fact,
a
simple
break
down
product
with
the
expected
sequence
and
that
it
has
not
been
modified
in
any
other
way.

One
Panel
member
believed
that
this
study
did
not
support
the
drawing
of
any
conclusions
about
human
safety
and
elaborated
that
only
ten
mice
were
tested,
that
there
were
no
longer­
term
studies
or
chronic
exposure
studies,
and
that
there
was
no
consideration
of
the
possibility
that
the
presence
of
the
test
protein
may
have
contributed
in
some
way
to
the
death
of
the
one
mouse
that
died
as
the
result
of
mechanical
damage
during
gavage.
This
Panel
member
also
believed
that
other,
non­
oral,
routes
of
exposure
such
as
inhalation
should
have
been
considered
and
suggested
that
the
fact
that
several
pollen­
borne
proteins
are
known
allergens
suggests
that,
even
given
the
low
level
expression
in
pollen,
this
issue
should
have
been
considered.
This
Panel
member
also
expressed
doubts
about
the
ability
to
extrapolate
from
animal
studies
to
human
toxicity,
particularly
allergenicity.
36
Most
of
the
Panel
members
believed
that
the
digestion
data
are
indicative,
but
could
have
been
presented
in
a
stronger
manner.
Sensitivity
to
degradation
is
a
relative
measure
 
one
protein
is
more
or
less
sensitive
than
another
under
certain
conditions.
The
data
provided
do
not
compare
sensitivity
of
the
modified
protein
to
the
native
protein,
or
to
any
other
protein.
Further,
these
studies
were
done
at
a
relatively
high
enzyme
to
protein
ratio.
It
would
have
been
more
informative
if
the
test
could
have
been
carried
out
with
lower
enzyme
concentrations,
so
that
potential
breakdown
products
could
have
been
detected.
Given
the
poor
quality
of
the
images
in
the
documents,
it
is
difficult
to
unequivocally
determine
that
there
were
no
stable
breakdown
products
produced.
For
example,
one
can
compare
undigested
protein
extracted
from
plants
shown
in
Fig.
1,
Lane
3
in
MRID
461556­
03
(
characterizations
study),
with
a
loading
of
3.9
µ
g
protein,
to
Lanes
3
and
4
of
Figure
2
of
MRID
461556­
07
(
digestion
study)
which
would
have
1
µ
g
of
the
protein.
The
second
gel
does
not
show
any
of
the
many
lower
molecular
weight
bands
that
are
seen
in
the
former
(
most
of
which
should
be
visible
even
taking
the
different
loadings
into
account).
This
makes
it
difficult
to
feel
great
assurance
that
any
breakdown
products
would
have
been
observed
in
the
digestion
study.
This
could
be
an
issue
related
to
image
reproduction
rather
than
with
the
actual
data,
but
it
is
of
interest.
However,
given
that
digestion
stability
does
not
correlate
with
actual
allergenicity,
these
shortcomings
are
not
significant.

One
Panel
member
suggested
that
the
Agency
consider
the
fact
that
some
humans
may
have
lower
than
normal
acidity
in
the
stomach.
Those
people
may
be
exposed
to
intact,
or
partially
degraded,
protein.

The
Panel
believed
that
the
sequence
analysis
as
described
was
difficult
to
fully
evaluate
because
the
document
was
more
a
summary
abstract
than
a
technical
description.
The
construction
and
contents
of
the
allergen
sequence
database
were
not
fully
described.
For
example,
it
is
not
clear
what
is
meant
by
"
redundant
sequences
were
removed
from
the
database."
Does
redundant
mean
100%
sequence
identity,
or
some
lower
percentage
homology?
How
were
iso­
allergens
treated?
How
many
food
and
non­
food
allergens
are
included
in
the
database?
The
short
sequence
(
8
mer)
testing
was
carried
out
using
a
propriety
program,
but
no
information
is
given
on
how
it
worked
or
how
it
was
validated.
The
long
sequence
testing
(
80
mer)
was
carried
out
using
only
a
single
scoring
system
with
FASTA.
It
is
known
that
changes
in
the
scoring
matrix
used
can
make
a
difference
in
the
ability
of
the
program
to
find
regions
of
homology.
Also,
homology
is
not
an
absolute
situation.
Although
the
Registrant
used
the
cut­
off
value
suggested
by
the
FAO/
WHO
expert
consultation,
it
should
be
noted
that
this
value
has
not
in
fact
been
adopted
by
FAO/
WHO.
It
would
have
been
much
nicer
to
see
the
actual
results
of
the
testing.

The
Panel
also
believed
it
is
worth
mentioning
that
the
sequence
analysis
related
to
similarity
to
known
toxins
provided
even
less
information.
For
example,
it
was
not
specified
which
version
of
the
National
Center
for
Biotechnology
Information
database
was
tested,
nor
were
descriptions
provided
of
the
parameters
used
in
the
Basic
Local
Alignment
Search
Tool
­
Protein
(
a
protein
sequence
analysis
tool).
The
information
as
actually
presented
makes
it
37
difficult
to
appropriately
analyze
this
study
and
therefore
to
scientifically
support
a
permanent
exemption
from
the
requirement
for
a
tolerance.

One
Panel
member
indicated
that,
in
general,
sequence
comparison
is
of
limited
value.
This
Panel
member
suggested
that
difference
of
one
amino
acid
may
make
the
difference
between
an
allergenic
protein
and
a
non­
allergenic
protein.
This
would
be
particularly
relevant
if
the
different
proteins
fold
differently,
which
may
bring
different
sequences
of
amino
acids
together
in
proximity
and
form
new
allergenic
epitopes.
This
Panel
member
also
said
that
a
one
amino
acid
difference
may
make
a
difference
in
the
degree
of
toxicity.

The
Panel
had
no
comments
on
the
glycosylation
study.

Most
of
the
Panel
believed
that
the
exposure
assessment
appeared
to
be
straightforward.
It
showed
that
the
amount
of
Cry
protein
that
would
be
consumed
in
a
rather
large
serving
of
corn
kernels
would
be
several
orders
of
magnitude
lower
than
the
highest
dose
tested
in
the
mammalian
toxicity
studies
 
which
had
no
observed
effect.
Even
taking
into
account
the
questions
about
quantification
that
were
raised
previously,
this
assessment
indicates
that
the
level
of
human
exposure
presents
a
reasonable
certainty
of
no
harm.

Having
said
all
this,
most
Panel
members
believed
that
the
data
did
not
raise
"
red
flags"
regarding
human
toxicity.
On
the
other
hand,
a
couple
of
Panel
members
agreed
that
no
conclusion
on
safety
could
be
derived
from
these
data
and
that
much
more
information
is
needed
about
acute
effects
and
potential
effects
on
children's
growth
and
development.
At
least
three
Panel
members
indicated
that
far
too
little
data
was
made
available
to
support
scientifically
a
permanent
exemption
from
the
requirement
of
a
tolerance,
especially
in
a
widely
used
food
crop
such
as
corn.

There
was
a
broad
consensus
among
the
Panel
members
that
there
were
significant
data
quality
issues.
In
many
cases,
it
was
not
possible
to
fully
evaluate
the
results
because
the
methods
were
inadequately
described
in
the
documents
given
to
the
Panel.
In
other
cases,
the
methods
were
used
in
an
inconsistent
manner.
For
example,
two
documents
(
MRID
462656­
05
and
MRID
462656­
06)
reported
Western
Blot
studies
that
appear
to
have
been
performed
within
a
short
period
of
time
that
used
different
detection
systems,
with
clearly
different
sensitivities.
This
raises
questions
about
whether
all
relevant
bands
were
detected.
The
Panel
also
suggested
that
better
quality
reproductions
of
data
images
be
used
in
the
future
and
that
all
of
the
data
obtained
(
such
as
for
sequence
analyses)
be
included
in
the
material
presented
to
the
Panel.
38
REFERENCES
Banks,
J.
1988.
The
acute
toxicity
of
diflubenzuron,
methoprene,
temephos,
fenoxycarb,
Bacillus
thuringiensis
var.
israelensis
(
BTI),
and
a
fenoxycarb/
Bti
mixture
on
the
nontarget,
saltmarsh
fiddler
crab,
Uca
pugilator,
with
an
emphasis
on
the
sublethal
effects
of
fenoxycarb
and
the
fenoxycarb/
Bti
mixture
on
energy
metabolism.
Master
Thesis,
University
of
South
Carolina,
School
of
Public
Health:
72
pp.

Barford,
N.
1967.
Experimental
Measurements:
Precision,
Error
and
Truth.
Addison­
Wesley,
London,
England.

Brewer
RA,
Carlock
LL,
Brewer
LW,
Hooper
MJ,
Cobb
GP,
R.
J.
Kendall.
1995.
Toxicity
Survivability
and
Activity
Patterns
of
Northern
Bobwhite
Quail
Dosed
with
the
Insecticide
COUNTER
®
.
Environ.
Toxicol.
Chem.,
15:
570­
573.

Cochran,
W.
G.
1977.
Sampling
Techniques,
3rd
Edition,
John
Wiley
and
Sons.
Equation
6.13.

Dee,
J.
1988.
The
acute
toxicity
of
four
commonly
used
mosquito
control
larvicides,
a
microbial
larvicide
(
BTI)
and
pesticide
mixture
on
the
grass
shrimp,
Palaemonetes
pugio.
Master
Thesis,
University
of
South
Carolina,
School
of
Public
Health:
63
pp.

Deicke,
A.;
Suverkrup,
R.
2000.
Dose
uniformity
and
redispersibility
of
pharmaceutical
suspensions.
2.
Assessment
of
three
commercial
erythromycin
ethyl
succinate
oral
liquids.
European
Journal
of
Pharmaceutics
and
Biopharmaceutics
49(
1),
73­
78.

Elkheshen,
Seham
A.;
Badawi,
Sabry
S.;
Badawi,
Alia
A.
1996.
Optimization
of
a
reconstitutable
suspension
of
rifampicin
using
24
factorial
design.
Drug
Development
and
Industrial
Pharmacy.
22(
7),
623­
630.

Flores,
S.,
Saxena,
D.,
Stotzky,
G.,
2005.
Transgenic
Bt
plants
decompose
less
in
soil
than
non­
Bt
plants.
Soil
Biology
and
Biochemistry
37,
1073­
1082.

Head
G.,
J.
B.
Surber,
J.
A.
Watson,
J.
W.
Martin,
and
J.
J.
Duan.
2002.
No
detection
of
Cry1Ac
protein
in
soil
after
multiple
years
of
transgenic
Bt
cotton
(
Bollgard)
use.
Environmental
Entomology
31:
30­
36.

Lee,
B.
1987.
The
acute
toxicity
of
four
chemical
mosquito
larvicides
(
temephos,
fenoxycarb,
diflubenzuron,
and
methoprene)
and
a
microbial
mosquito
larvicide
(
Bacillus
thuringiensis
var.
israelensis)
on
a
nontarget,
estuarine
fish
species
­
the
mummichog
(
Fundulus
heteroclitus).
Master
Thesis,
University
of
South
Carolina,
School
of
Public
Health:
53
pp.
39
Lee,
B.
&
G.
Scott.
1989.
The
acute
toxicity
of
4
chemical
mosquito
control
larvicides
and
a
microbial
larvicide
(
BTI)
on
the
estuarine
fish,
Fundulus
heteroclitus.
Bulletin
of
Env.
Contamination
and
Toxicology
43:
827­
832.

Palm,
C.
J.,
Schaller,
D.
J.,
Donegan,
K.
K.,
Seidler,
R.
J.,
1996.
Persistence
in
soil
of
transgenic
plant
produced
Bacillus
thuringiensis
var.
kurstaki
d­
endotoxin.
Canadian
Journal
of
Microbiology
42,
1258­
1262.

Ramsey
Janine
M;
Cruz­
Celis
Adriana;
Salgado
Liliana;
Espinosa
Luis;
Ordonez
Rosalinda;
Lopez
Rene;
Schofield
C
J.
2003.
Efficacy
of
pyrethroid
insecticides
against
domestic
and
peridomestic
populations
of
Triatoma
pallidipennis
and
Triatoma
barberi
vectors
of
Chagas'
disease
in
Mexico.
Journal
of
Medical
Entomology
40(
6),
912­
20.

RTI
(
1992).
Developmental
toxicity
evaluation
of
N,
N'­
methylenebisacrylamide
administered
by
gavage
to
Swiss
(
CD­
1)
mice
on
gestational
days
6
through
17.,
Research
Triangle
Institute.
RTP,
NC
198
pp.

Saxena,
D.,
Stotzky,
G.,
2000.
Insecticidal
toxin
from
Bacillus
thuringiensis
is
released
from
roots
of
transgenic
Bt
corn
in
vitro
and
in
situ.
FEMS
Microbiology
Ecology
33,
3539.

Saxena,
D.,
C.
N.
Stewart,
I.
Altosaar,
Q.
Shu,
and
G.
Stotzky.
2004.
Larvicidal
Cry
proteins
from
Bacillus
thuringiensis
are
released
in
root
exudates
of
transgenic
corn,
potatoes
and
rice
but
not
of
Bacillus
thuringiensis
canola,
cotton
and
tobacco.
Plant
Physiology
and
Biochemistry
42:
383­
387.

Tapp,
H.,
Stotzky,
G.,
1998.
Persistence
of
the
insecticidal
toxin
from
Bacillus
thuringiensis
subsp.
kurstaki
in
soil.
Soil
Biology
and
Biochemistry
30,
471476.

Thomas,
Peter
T.;
Ratajczak,
Helen
V.;
Eisenberg,
Walter
C.;
Furedi­
Machacek,
Marianna;
Ketels,
Kathleen
V.;
Barbera,
Peter
W.
1987.
Evaluation
of
host
resistance
and
immunity
in
mice
exposed
to
the
carbamate
pesticide
aldicarb.
Fundamental
and
Applied
Toxicology
9(
1),
82­
9.

Thomas
P;
Ratajczak
H;
Demetral
D;
Hagen
K;
Baron
R.
1990.
Aldicarb
immunotoxicity:
functional
analysis
of
cell­
mediated
immunity
and
quantitation
of
lymphocyte
subpopulations.
Fundamental
and
applied
toxicology.
15(
2),
221­
30.

Thompson,
S.
K.
1992.
Sampling.
Chapter
7.
John
Wiley
and
Sons,
Inc.
New
York.

USEPA.
1996.
Microbial
pesticide
test
guidelines
for
tier
I
nontarget
insect
testing.
Office
of
Prevention,
Pesticides
and
Toxic
Substances.
Harmonized
Testing
Guideline
885.4340.
EPA
712­
C­
96­
280.
40
Zwahlen,
C.,
A.
R.
Hilbeck,
P.
Gugerli,
and
W.
Nentwig.
2003.
Degradation
of
Cry1Ab
protein
within
transgenic
Bacillus
thuringiensis
corn
tissue
in
the
field.
Molecular
Ecology
12:
765­
775
41
Appendix
1
Effect
of
the
Eisenia
Test
Diet
on
Toxicity
of
MCRY3A­
0102
j
0
1
..
:=
DoseNeat
j
12.5
50
:=
MortNeat
j
0.87
0.97
:=

Concentration
of
MCRY3A­
0102
("
DoseNeat",
µ
g/
g)
and
resulting
mortality
("
MortNeat",
fraction)
following
"
neat"
(
without
soil)
addition
to
CPB
test
diet.

LogDoseNeat
j
log
DoseNeat
j
(
)
:=
PrbtMortNeat
j
qnorm
MortNeat
j
0
,
1
,
(
)
:=

LogDoseNeat
1.097
1.699
   
   
=
PrbtMortNeat
1.126
1.881
   
   
=

Log
base­
10
of
DoseNeat
("
LogDoseNeat")
and
probit
mortality
of
MortNeat
("
PrbtMortNeat"),
where
probit
mortality
is
the
inverse
of
the
cumulative
distribution
for
the
underlying
standard
normal
tolerance
distribution
(+
5).

1
 
   
   
   
   

j
PrbtMortNeat
1
PrbtMortNeat
0
 
(
)

LogDoseNeat
1
LogDoseNeat
0
 
(
)
:=
1
 
   
   
1.253
1.253
   
   
=

Slope
of
the
probit
mortality­
log(
dose)
plot,
where
the
variable
name
used
acknowledges
that
the
slope
is
equal
to
the
reciprocal
of
the
standard
deviation
of
the
underlying
standard
normal
tolerance
distribution.

µ
 
   
   
   
   

j
PrbtMortNeat
j
1
 
   
 

  
 

  
 

  

j
LogDoseNeat
j
 
 
:=
µ
 
   
   
   
   
0.248
 

0.248
 
   
   
=

Y­
intercept
of
the
probit
mortality­
log(
dose)
plot,
where
the
variable
name
used
acknowledges
that
the
y­
intercept
is
equal
to
the
ratio
of
the
mean
to
the
standard
deviation
of
the
underlying
standard
normal
tolerance
distribution.

i
0
100
..
:=
LogDose
i
i
50
:=
PrbtMort
i
1
 
   
   
   
   

1
LogDose
i
 
µ
 
   
   
   
   

1
   
   
+
:=

Definition
of
the
probit
mortality­
log(
dose)
line
between
the
two
"
neat"
observations.

MortSoil
j
0.67
0.83
:=
PrbtMortSoil
j
qnorm
MortSoil
j
0
,
1
,
(
)
:=

Mortality
observed
following
addition
of
MCRY3­
0102
in
test
soil
to
the
CPB
test
diet
("
MortSoil",
fraction)
and
probit
mortality
of
MortSoil
("
PrbtMortSoil").
42
LogDoseSoil
j
PrbtMortSoil
j
µ
 
   
 

  
 

  
 

  

1
 

   
 

  

1
 
   
   
   
   

1
:=

Calculation
of
the
Log(
DoseSoil)
("
LogDoseSoil")
values
predicted
by
the
probit
mortalitylog
dose)
relationship
based
on
the
"
neat"
data
and
the
observed
PrbtMortSoil
values.

0
0.5
1
1.5
2
1
0
1
2
3
PrbtMort
i
PrbtMortNeat
j
PrbtMortSoil
j
LogDosei
LogDoseNeat
j
,
LogDoseSoil
j
,

Plot
of
the
probit
mortality­
log(
dose)
relationship
based
on
the
"
neat"
data
and
the
observed
PrbtMortSoil
values.

DoseSoil
j
10
LogDoseSoil
j
:=
DoseSoil
3.54
9.109
   
   
=

Antilog(
LogDoseSoil)
to
give
the
effective
concentrations
of
MCRY3­
0102
added
with
soil
to
CPB
test
diet
("
DoseSoil",
µ
g/
g).

NomDoseSoil
j
13.3
26.6
:=

FracNomDoseSoil
j
DoseSoil
j
NomDoseSoil
j
:=
FracNomDoseSoil
0.266
0.342
   
   
=

Nominal
concentrations
of
MCRY3­
0102
added
with
soil
to
CPB
test
diet
("
NomDoseSoil",
µ
g/
g)
and
DoseSoil
as
a
fraction
of
NomDoseSoil
("
FracNomDoseSoil").

CF
mean
FracNomDoseSoil
(
)
:=
CF
0.304
=
Average
concentration
of
MCRY3­
0102
relative
to
nominal
concentration
("
CF").

SECF
Stdev
FracNomDoseSoil
(
)

2
(
)
.5
:=
SECF
0.038
=

Standard
error
of
CF.
43
Appendix
2
Estimation
of
Extraction
Efficiency
from
a
Series
of
Two
Extractions
The
Registrant
estimated
extraction
efficiency
("
EE")
of
mCry3A
from
tissues
of
MIR604
corn
using
the
following
formulation:
EE
=
X/(
X+
Y)
(
X
 
concentration
mCry3A
recovered
in
first
extraction)
(
Y
 
concentration
mCry3A
recovered
in
second
extraction)

If
alternatively
one
assumes
that
the
fraction
of
residual
mCry3A
recovered
in
successive
extractions
is
constant,
then
EE
overestimates
the
true
extraction
efficiency
("
AEE").

For
total
extractable
mCry3A
present
in
the
corn
tissue
("
T"),
one
has
the
following:
AEE
=
X/
T
(
definition
of
AEE)
Y
=
(
X/
T)

(
T­
X)
(
applying
constant
extraction
efficiency
assumption)
Y
=
X

[
1­(
X/
T)]
(
rearranging)
X/
T
=
(
X­
Y)/
X
=
2 
EE­
1
(
solving
for
X/
T
and
substituting
Registrant's
def.
for
EE)
AEE
=
2 
EE­
1
i
0
100
..
:=
EE
i
0.5
i
100
+
:=
AEE
i
2
1
EE
i
 
:=

Calculation
of
AEE
for
various
tissues
of
MIR604
corn:
Leaf
tissue:

EE
0.771
:=

AEE
2
1
EE
 
:=
AEE
0.703
=

Kernel
tissue:

EE
0.697
:=

AEE
2
1
EE
 
:=
AEE
0.565
=

Silage
tissue:

EE
0.845
:=

AEE
2
1
EE
 
:=
AEE
0.817
=

Factor
by
which
the
Registrant's
estimates
for
EEC
should
be
multiplied
to
give
EEC
based
on
AEE
("
AEEC"):

AEEC/
EEC
:=
(
X/
AEE)/(
X/
EE)
:=
EE/
AEE
44
0.5
0.6
0.7
0.8
0.9
1
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
EEi
AEEi
EEi
Graph
of
factor
to
convert
from
EEC
to
AEEC.

Leaf
tissue:
EE/
AEE
=
0.771/
0.703
=
1.10
Kernel
tissue:
EE/
AEE
=
0.697/
0.565
=
1.23
Silage
tissue:
EE/
AEE
=
0.845/
0.817
=
1.03
