Page
1
of
34
Attachment
I:
Draft
Approach
to
Exempting
Certain
PVCP­
PIPs
from
Regulation
under
FIFRA
I.
What
Action
Does
this
Paper
Discuss?

EPA
is
considering
whether
to
establish
an
exemption
under
FIFRA
for
certain
plantincorporated
protectants
that
are
based
on
one
or
more
genes,
or
segments
of
genes,
that
encode
a
coat
protein
of
a
virus
that
naturally
infects
plants.
Such
plant
virus
coat
protein
plantincorporated
protectants
are
hereafter
referred
to
as
"
PVCP­
PIPs."

In
accordance
with
EPA's
regulations
at
40
C.
F.
R.
§
174.21,
three
general
qualifications
must
be
met
for
any
PIP
to
be
exempt
from
the
requirements
of
FIFRA:
(
1)
the
PIP
must
meet
a
set
of
criteria
specific
to
the
particular
PIP
under
consideration,
(
2)
the
residues
of
a
PVCP­
PIP
that
is
intended
to
be
produced
and
used
in
a
crop
used
as
food
must
either
be
exempted
from
the
requirement
of
a
tolerance
under
FFDCA
or
no
tolerance
would
otherwise
be
required
for
the
PVCP­
PIP,
and
(
3)
any
inert
ingredient
contained
in
the
PIP
must
be
codified
at
subpart
X
of
40
CFR
Part
174
 
List
of
Approved
Inert
Ingredients.
The
Agency
is
considering
three
criteria
that
would
allow
a
PVCP­
PIP
to
satisfy
the
first
of
these
general
requirements
(
i.
e.,
the
174.21(
a)
requirement).
Thus,
the
three
criteria
relevant
to
the
174.21(
a)
requirement
are
intended
to
address
three
issues
that
are
associated
with
potential
risks
of
PVCP­
PIPs:
gene
flow
(
criterion
(
a)),
recombination
(
criterion
(
b)),
and
protein
production
(
criterion
(
c)).
The
criteria
under
consideration
referred
to
throughout
this
document
appear
together
in
the
Appendix
of
this
attachment.

The
PVCP­
PIP
would
have
to
meet
all
three
of
the
criteria
in
order
to
meet
the
174.21(
a)
exemption
requirement.
A
PVCP­
PIP
may
be
determined
to
meet
each
criterion
in
one
of
two
ways:
a
product
developer
may
self­
determine
that
paragraph
(
1)
of
the
criterion
is
met
or
the
Agency
may
determine
that
paragraph
(
2)
of
the
criterion
is
met.
Paragraph
(
1)
of
each
criterion
describes
an
objective,
well­
defined
characteristic.
Therefore,
the
developer
may
determine
whether
the
PVCP­
PIP
meets
the
requirement.
Paragraph
(
2)
of
each
criterion
would
be
conditioned
on
an
Agency
determination
because
it
may
involve
analysis
of
several
types
of
information.
Therefore,
an
Agency
review
would
be
necessary
to
evaluate
and
conclude
that
the
PVCP­
PIP
meets
the
requirement
and
is
therefore
of
a
nature
warranting
exemption.

Each
criterion
may
be
satisfied
under
either
paragraph
(
1)
or
paragraph
(
2)
irrespective
of
how
the
other
two
criteria
are
satisfied.
Thus,
there
would
be
no
requirement
that
all
three
criteria
must
be
satisfied
under
either
paragraph
(
1)
or
paragraph
(
2)
in
order
to
qualify
for
the
exemption.

Products
that
fail
to
meet
one
or
more
of
the
exemption
criteria
would
need
to
obtain
a
registration.
EPA
would
evaluate
such
PVCP­
PIPs
under
the
existing
registration
process
and
could
implement
control
measures
on
use
as
appropriate.

If
a
PVCP­
PIP
does
not
satisfy
a
particular
criterion
under
paragraph
(
1),
EPA
envisions
that
in
order
for
a
product
to
qualify
for
an
exemption,
the
product
developer
would
make
a
Page
2
of
34
submission
to
the
Agency
containing
supporting
data
or
other
information
to
demonstrate
that
a
particular
PVCP­
PIP
meets
paragraph
(
2)
of
that
criterion
to
enable
Agency
review
for
that
criterion.

II.
Key
Scientific
Issues
Several
scientific
questions
concerning
risk
issues
associated
with
PVCP­
PIPs
have
been
identified:
(
1)
What
is
the
potential
for
a
PVCP­
PIP
to
endow
plants
with
characteristics
that
could
disrupt
the
existing
network
of
ecological
relationships
in
managed,
semi­
managed,
or
natural
ecosystems,
e.
g.,
through
gene
transfer
to
wild
or
weedy1
relatives?;
(
2)
what
is
the
potential
for
viral
interactions
to
affect
the
epidemiology
or
pathogenicity
of
plant
viruses?;
and
(
3)
what
is
the
potential
for
exposure
of
humans
or
nontarget
organisms
to
PVC­
proteins
with
novel
toxic
or
allergenic
properties?

A.
Potential
for
a
PVCP­
PIP
to
disrupt
ecological
relationships
In
evaluating
whether
a
PVCP­
PIP
could
alter
ecological
relationships
among
plants,
EPA
considered
two
primary
issues:
(
1)
whether
the
PVCP­
PIP
could
endow
a
transgenic
plant
itself
with
the
ability
to
spread
into
natural
or
semi­
managed
habitats
and
(
2)
whether
the
transfer
of
a
PVCP­
PIP
from
a
transgenic
plant
into
wild
or
weedy
relatives
could
disrupt
ecological
relationships.
Whether
gene
transfer
could
disrupt
ecological
relationships
depends
on
several
additional
considerations.
First,
does
the
crop
plant
containing
the
PVCP­
PIP
have
wild
relatives
growing
in
its
vicinity
with
which
it
is
able
to
hybridize?
Second,
is
virus
infection
limiting
the
growth
and/
or
reproduction
of
individual
plants
within
populations
of
wild
or
weedy
relatives
such
that
a
gene
conferring
virus
resistance
is
likely
to
become
a
stable
part
of
the
genome?
Third,
would
stable
introduction
of
a
PVCP­
PIP
into
the
plant
population
(
i.
e.,
introgression)
cause
it
to
become
weedier/
more
invasive
or
lose
its
competitive
ability,
thereby
changing
the
population
dynamics
of
the
plant
community?

1.
Likelihood
that
a
crop
plant
containing
a
PVCP­
PIP
could
itself
disrupt
ecological
relationships
In
considering
whether
a
PVCP­
PIP
could
affect
the
ability
of
a
plant
to
spread
into
natural
or
semi­
managed
habitat
at
the
margins
of
cultivated
fields,
i.
e.,
to
form
feral
or
naturalized
populations,
the
key
consideration
is
whether
viral
infection
is
currently
limiting
the
ability
of
agricultural
crops
to
do
so.
EPA
is
aware
of
no
evidence
suggesting
that
such
is
the
case.
For
1
Throughout
this
document,
EPA
considers
weedy
plants
to
be
those
with
the
characteristics
of
weeds,
i.
e.,
those
that
are
considered
undesirable,
unattractive,
or
troublesome,
especially
when
growing
where
they
are
not
wanted.
Wild
plants
are
those
that
occur,
grow,
and
live
in
a
natural
state
and
are
not
domesticated,
cultivated,
or
tamed.
EPA
considers
a
naturalized
population
to
be
a
population
of
domesticated
plants
that
grows
in
wild
(
non­
cultivated
areas).
EPA
considers
a
native
plant
population
to
be
one
that
originates
in
a
particular
region
or
ecosystem.
Page
3
of
34
example,
field
experiments
with
transgenic
virus
resistant
sugar
beets
revealed
no
competitive
advantage
(
measured
as
seedling
emergence
and
biomass
production)
between
the
transgenic
and
susceptible
control
lines
(
Ref.
1).
There
are
also
no
reports
that
conventional
control
of
virus
pathogens,
e.
g.,
by
virus
resistance
traits
introduced
by
conventional
breeding,
has
resulted
in
increased
weediness
or
invasiveness
of
a
crop
plant
(
Ref.
2).

Although
virus
infection
has
been
shown
to
decrease
growth
and/
or
reproduction
of
some
natural
plant
communities
suggesting
that
acquired
virus
resistance
has
the
potential
to
influence
plant
population
dynamics
(
discussed
below
in
Unit
II.
A.
2),
there
are
many
reasons
to
believe
the
situation
would
be
different
for
crop
plants.
Most
naturalized,
domesticated
crops
generally
are
unable
to
effectively
compete
with
wild
species
in
natural
ecosystems
and
are
unlikely
to
acquire
this
ability
with
genetic
modification
(
Ref.
2).
Plant
breeders
have
capitalized
for
decades
on
the
fact
that
relatively
minor
genetic
changes
can
produce
plants
with
altered
ecological
properties,
but
the
addition
of
pest
resistant
traits
has
not
been
known
to
result
in
increased
weediness
of
widely
used
crops,
the
possible
exception
being
gene
transfer
from
cultivated
traditionally­
bred
sorghum
to
Johnson
grass
(
Ref.
2).
For
domesticated
crops,
the
traits
that
make
them
useful
to
humans
also
reduce
their
competitive
ability
in
nonagricultural
habitats.
EPA
believes
crops
that
have
been
subjected
to
long­
term
breeding
(
e.
g.,
corn,
beans,
maize,
and
wheat)
are
unlikely
to
possess
characteristics
that
would
allow
the
plant
to
compete
effectively
outside
of
managed
ecosystems.
Domesticity
arises
because
many
characteristics
that
would
enhance
weediness
(
e.
g.,
seed
shattering,
thorns,
seed
dormancy,
and
bitterness)
have
been
deliberately
eliminated
from
the
crop
plant
through
intensive
breeding
efforts.
For
example,
lack
of
seed
shattering
and
seed
dormancy
greatly
reduces
the
ability
of
an
annual
crop
to
persist
without
human
intervention.
Without
major
changes
in
its
phenotype,
corn
is
unlikely
to
survive
for
multiple
generations
outside
agricultural
fields
no
matter
what
transgene
it
contains
(
Ref.
3).

The
reassuring
history
for
cultivated
plants
does
not
completely
preclude
a
crop
containing
a
PVCP­
PIP
from
becoming
a
weed,
but
it
suggests
that
the
likelihood
of
that
event
is
small
(
Ref.
2).
Given
the
selective
disadvantage
of
crop
plants
in
natural
ecosystems,
it
is
unlikely
that
acquisition
of
virus
resistance
would
confer
sufficient
competitive
advantage
on
naturalized
populations
of
crops
to
upset
the
network
of
existing
ecological
relationships
given
that
many
other
factors
appear
to
constrain
their
competitiveness
in
non­
cultivated
areas
(
Ref.
1).
A
survey
of
the
weedy
characteristics
of
crop
versus
weed
species
showed
that
weeds
possess
significantly
more
weedy
characteristics
on
average
than
do
crop
plants,
suggesting
that
acquisition
of
any
single
trait
would
be
insufficient
to
make
a
crop
plant
a
weed
(
Ref.
4).

Thus,
EPA
believes
that
the
available
evidence
supports
a
finding
that
there
is
a
low
probability
of
risk
that
a
PVCP­
PIP
would
cause
the
engineered
crop
plant
to
become
wild
or
weedy.
As
a
result,
EPA
believes
that
the
only
condition
on
an
exemption
that
is
necessary
to
ensure
that
crop
plants
containing
PVCP­
PIPs
that
qualify
for
exemption
present
only
a
low
risk
of
disrupting
ecological
relationships
is
a
requirement
that
the
crop
plant
is
not
itself
already
a
weedy
or
invasive
species.

2.
Likelihood
that
a
crop
plant
containing
a
PVCP­
PIP
could
disrupt
ecological
relationships
through
gene
transfer
Page
4
of
34
As
discussed
above
in
Unit
II.
A.
1,
the
available
evidence
suggests
that
there
is
only
a
low
likelihood
that
a
PVCP­
PIP
is
likely
to
cause
the
transgenic
crop
plant
containing
it
to
become
weedy
or
invasive.
However,
the
question
of
whether
gene
transfer
to
naturally
occurring
plants
in
the
agroecosystem
could
lead
to
such
adverse
outcomes
is
a
more
complicated
question
because
it
involves
a
much
broader
range
of
potential
plants.
The
answer
to
this
question
depends
first
on
the
question
of
whether
the
transgenic
crop
plants
could
transfer
a
PVCP­
PIP
to
other
plant
populations.
This
potential
for
transfer
depends
on
the
frequency
of
hybridization
between
domesticated
species
and
their
wild
relatives.
Hybridization
is
affected
by
the
ability
of
plants
to
cross­
pollinate
which
in
turn
is
affected
by
their
timing
of
reproductive
viability
and
the
proximity
of
the
plants.
Hybridization
is
also
affected
by
the
ability
of
pollen
to
fertilize
recipient
plants,
develop
into
viable
seeds,
germinate,
and
grow
into
a
viable
adult
(
Ref.
5).
In
spite
of
these
potential
constraints,
a
survey
of
the
world's
most
important
crops
suggests
that
spontaneous
hybridization
of
domesticated
plants
with
wild
relatives
appears
to
be
a
general
feature
across
at
least
a
portion
of
the
geographic
area
over
which
each
is
cultivated
(
Refs.
6,
7).

The
answer
to
whether
virus
infection
limits
the
growth
and/
or
reproductive
ability
of
wild
or
weedy
plant
populations
appears
to
be
much
more
variable.
In
general,
viruses
appear
to
be
pervasive
in
natural
plant
populations
(
Refs.
8,
9,
10,
11),
although
a
comprehensive
body
of
literature
on
the
presence
and
effect
of
viruses
in
weed
species
is
lacking.
Some
studies
report
that
virus
infection
has
little
or
no
effect
on
the
plants
(
e.
g.,
see
Refs.
10,
12).
However,
other
studies
have
reported
that
infection
reduces
plant
growth
and/
or
fecundity.
For
example,
tobacco
leaf
curl
gemininivirus
infection
increases
mortality
and
has
significant
negative
effects
on
growth
and
seed
output
in
plants
from
wild
populations
of
Eupatorium
chinense
(
Ref.
13),
and
geminivirus
infection
likewise
decreases
growth
and
reproductive
output
of
Eupatorium
makinoi
(
Ref.
14).
Field
experiments
showed
that
wild
cabbage
plants
(
Brassica
oleracea)
infected
with
turnip
mosaic
potyvirus
or
turnip
yellow
mosaic
tymovirus
have
reduced
survival,
growth,
and
reproduction
(
Ref.
15).
Similarly,
cucumber
mosaic
virus
infection
was
found
to
reduce
vegetative
growth
and
flower
production
of
purslane
(
Portulaca
oleracea)
(
Ref.
16).

It
is
difficult
to
predict
the
actual
impact
on
overall
plant
population
dynamics
that
would
result
from
acquisition
of
virus
resistance
by
plants
that
are
in
some
way
negatively
affected
by
virus
infection.
EPA
is
not
aware
of
any
study
that
has
directly
examined
this
question
by
purposefully
freeing
a
weed
species
from
virus
infection
and
investigating
the
resulting
population
dynamics
of
infected
versus
uninfected
plants.
However,
some
studies
show
that
virus
infection
can
in
some
cases
affect
plant
population
dynamics,
suggesting
such
impacts
may
also
result
if
the
population
were
subsequently
freed
from
virus
infection.
For
example,
infection
with
alfalfa
mosaic
virus
substantially
diminished
the
ability
of
certain
medic
cultivars
(
Medicago
polymorpha)
to
compete
with
other
species
such
as
capeweed
(
Arctotheca
calendula),
both
directly
by
decreasing
the
competitive
ability
of
infected
plants,
and
indirectly
by
altering
the
proportions
in
which
the
species
germinated
(
Ref.
17).
In
another
example
of
virus
infection
affecting
plant
population
dynamics,
growth
analysis
of
E.
makinoi
revealed
that
plants
naturally
infected
with
a
geminivirus
had
significantly
reduced
stem
growth
and
plant
height,
along
with
decreased
flowering
and
survivorship.
This
study
suggests
that
such
negative
fitness
attributes
have
a
significant
impact
on
plant
population
dynamics
in
this
species
(
Ref.
18).

Although
relatively
little
research
has
been
conducted
regarding
how
plant
population
dynamics
are
influenced
by
virus
infection,
such
results
as
described
in
the
previous
paragraph
support
the
Page
5
of
34
premise
that
at
least
some
plant
populations
acquiring
virus
resistance
might
be
able
to
outcompete
other
species
(
Ref.
19)
and/
or
spread
to
previously
unsuitable
habitat
(
Ref.
20).
However,
it
has
also
been
discussed
that
acquisition
of
virus
resistance
might
decrease
plant
fitness.
For
example,
barley
yellow
dwarf
virus
was
found
in
at
least
one
year
to
increase
the
fitness
of
its
host
Setaria
viridis
by
approximately
25%
(
Ref.
21).
Such
results
might
be
expected
if
the
plants
become
more
attractive
to
herbivores
when
not
infected
by
viruses,
as
was
found
to
occur
for
seedlings
of
Kennedya
rubicunda
(
Ref.
12).
In
a
coastal
bushland
experimental
site,
virus­
free
seedlings
were
grazed
at
twice
the
rate
as
those
manually
inoculated
with
Kennedya
yellow
dwarf
virus
due
to
increased
palatability
to
herbivores.
Such
considerations
may
be
important
in
evaluating
effects
on
endangered/
threatened
species.

EPA
believes
that
many
PVCP­
PIP/
plant
combinations
are
likely
to
pose
a
low
risk
of
disrupting
the
existing
network
of
ecological
relationships
in
semi­
managed
or
natural
ecosystems
given
that
multiple
factors
must
be
present,
i.
e.,
hybridization
with
a
wild
relative
must
occur,
introgression
of
the
gene
must
occur,
and
acquired
virus
resistance
must
confer
an
advantage
(
or
disadvantage)
to
the
recipient
plant
sufficient
to
alter
plant
population
dynamics.
Nevertheless,
the
research
discussed
above
showing
that
viruses
can
in
some
cases
affect
plant
population
dynamics
highlights
the
difficulty
in
drawing
a
general
conclusion
as
to
whether
all
PVCPPIP
plant
combinations
are
likely
to
pose
a
low
risk
of
disrupting
existing
ecological
networks.
Virtually
any
crop
could
be
modified
to
contain
a
PVCP­
PIP,
including
less
domesticated
forage
crops
and
trees,
and
such
a
wide
range
of
crop
plants
will
be
associated
with
a
concomitantly
wide
range
of
characteristics
and
behaviors.
Ecosystems
are
highly
complex
and
variable,
and
the
factors
that
limit
establishment
of
a
given
plant
species
can
be
subtle
and
are
not
well
understood
(
Ref.
3).
Consequently,
EPA
does
not
believe
that
the
available
body
of
evidence
would
currently
support
a
definitive
conclusion
for
all
PVCP­
PIPs
that
the
potential
transfer
to
wild
or
weedy
relatives
presents
a
low
risk
of
altering
the
network
of
ecological
relationships
in
semi­
managed
or
natural
ecosystems.
Thus,
one
of
the
key
challenges
that
EPA
has
faced
is
how
to
clearly
describe
for
regulatory
purposes
those
situations
in
which
gene
transfer
of
a
PVCP­
PIP
would
not
likely
alter
existing
ecosystem
relationships.

Information
currently
available
does
not
appear
sufficient
to
describe
any
set
of
circumstances
that
would
predict
for
the
wide
variety
of
possible
PVCP­
PIP/
plant
combinations
whether
introgression
of
the
PVCP­
PIP
into
a
wild
or
weedy
relative
could
change
the
population
dynamics
of
the
recipient
plant.
For
example,
it
is
not
possible
to
predict
a
priori
whether
a
possible
fitness
advantage
that
individual
plants
might
acquire
with
a
PVCP­
PIP
would
lead
to
the
plant
population
outcompeting
other
species.
Whether
population
dynamics
would
be
affected
in
a
given
circumstance
is
dependent
on
multiple,
interacting
factors.
In
some
instances,
a
weight­
of­
evidence,
case­
by­
case
review
of
information
such
as
experimental
data
might
allow
such
a
determination;
however,
general
knowledge
of
factors
likely
to
influence
population
dynamics
cannot
be
readily
used
for
regulatory
purposes
to
develop
a
clearly
understandable
criterion
suitable
for
a
categorical
exemption.
Thus,
although
EPA
would
like
to
describe
an
objectively
defined
criterion
that
distinguishes
those
PVCP­
PIP/
plant
combinations
that
are
likely
to
change
plant
population
dynamics
from
those
that
are
not,
EPA
has
concluded
that
this
is
not
currently
feasible.
Instead,
EPA
is
considering
an
exemption
criterion
related
to
concerns
associated
with
gene
flow
based
on
other
relevant
factors.
Paragraph
(
1)
of
criterion
(
a),
a
categorical
exemption
criterion
for
a
subset
of
PVCP­
PIPs,
was
developed
based
on
the
potential
for
the
genetic
material
of
a
PVCP­
PIP
to
flow
to
wild
or
weedy
relatives.
Paragraph
(
2)
of
Page
6
of
34
criterion
(
a),
an
exemption
criterion
conditional
on
Agency
review,
was
developed
based
on
characteristics
of
the
plant
containing
the
PVCP­
PIP
and
characteristics
of
that
plant's
wild
or
weedy
relatives.
Each
part
of
criterion
(
a)
is
discussed
in
more
detail
below.

a.
Categorical
exemption
criterion
For
the
reasons
articulated
above,
EPA
does
not
believe
it
can
develop
a
categorical
exemption
that
is
based
on
whether
a
PVCP­
PIP/
plant
combination
is
likely
to
result
in
changes
in
the
population
dynamics
of
wild
or
weedy
relatives.
However,
EPA
believes
that
a
criterion
for
a
categorical
exemption
could
be
developed
based
on
exposure
potential,
i.
e.,
whether
genetic
material,
including
the
PVCP­
PIP,
could
flow
from
the
engineered
crop
plant
to
related
wild
or
weedy
relatives.
Essentially,
this
criterion
would
focus
on
the
first
of
the
three
events
that
must
occur
for
a
PVCP­
PIP
potentially
to
alter
the
existing
network
of
ecological
relationships
through
gene
flow:
whether
the
plant
containing
the
PVCP­
PIP
has
wild
relatives
growing
within
its
vicinity
with
which
it
can
produce
viable
hybrids.
Basing
a
regulatory
criterion
on
this
consideration
does
not
mean
the
Agency
fails
to
recognize
that
several
events
are
necessary
before
existing
networks
of
ecological
relationships
could
be
disrupted;
rather
it
is
an
attempt
to
create
a
clearly
understandable
regulatory
criterion
suitable
for
an
exemption.

In
developing
any
categorical
exemption
for
a
subset
of
PVCP­
PIPs
in
which
a
developer
could
self­
determine
whether
the
criteria
were
met,
EPA
seeks
to
identify
those
situations
that
clearly
pose
low
risk
with
respect
to
gene
transfer.
Although
the
Agency
recognizes
that
many
events
must
occur
before
transfer
of
a
PVCP­
PIP
would
cause
an
adverse
environmental
impact,
the
inability
of
the
crop
plant
to
hybridize
with
wild
or
weedy
relatives
provides
the
most
straightforward
assurance
in
an
objective
criterion
that
an
adverse
environmental
impact
would
not
occur
for
a
particular
PVCP­
PIP/
plant
combination.

A
PVCP­
PIP
would
meet
criterion
(
a)
under
paragraph
(
1)
if
the
plant
containing
the
PVCP­
PIP
is
one
of
the
following:
almond
(
Prunus
communis),
apricot
(
Prunus
armeniaca),
asparagus
(
Asparagus
officinale),
avocado
(
Persea
americana),
banana
(
Musa
acuminata),
barley
(
Hordeum
vulgare),
bean
(
Phaseolus
vulgaris),
black­
eyed
pea
(
Vigna
unguiculata),
cacao
(
Theobroma
cacao),
celery
(
Apium
graveolens),
chickpea
(
Cicer
arietinum),
citrus
(
Citrus
spp.),
coffee
(
Coffea
arabicua),
corn
(
Zea
maize),
cucumber
(
Cucumis
sativus),
eggplant
(
Solanum
melongena),
guava
(
Psidium
guajava),
kiwi
(
Actinidia
spp.),
mango
(
Mangifera
indica),
nectarine
(
Prunus
persica),
okra
(
Abelmoschus
esculentus),
olive
(
Olea
europaea),
papaya
(
Carica
papaya),
parsley
(
Petroselinum
crispum),
pea
(
Pisum
sativum),
peach
(
Prunus
persica),
peanut
(
Arachis
hypogaea),
pineapple
(
Ananas
comosus),
pistachio
(
Pistacia
vera),
plum
(
Prunus
domestica),
potato
(
Solanum
tuberosum),
soybean
(
Glycine
max),
spinach
(
Spinacia
oleracea),
starfruit
(
Averrhoa
carambola),
taro
(
Colocasia
esculenta),
tomato
(
Lycopersicon
lycopersicum),
or
watermelon
(
Citrullus
lanatus).

Plant
species
on
this
list
were
identified
by
the
October
2004
FIFRA
SAP
as
having
no
wild
or
weedy
relatives
in
the
United
States,
its
possessions,
or
territories
with
which
they
can
produce
viable
hybrids
in
nature.
Thus,
given
the
extremely
low
probability
that
virus
resistance
could
be
transferred
to
another
species,
any
PVCP­
PIP
contained
in
such
a
plant
would
pose
a
low
Page
7
of
34
probability
of
altering
existing
plant
population
dynamics
or
existing
ecological
relationships.
In
addition,
a
list
is
very
straightforward,
providing
an
easy­
to­
understand
criterion.
EPA
thus
believes
that
a
developer
could
self­
determine
eligibility
as
no
further
data
or
information
would
be
needed
to
evaluate
whether
ecological
relationships
could
be
disrupted
through
gene
flow
when
the
plant
containing
the
PVCP­
PIP
is
on
the
list.

The
SAP
noted
(
Ref.
22)
that
some
of
the
plants
on
this
list
were
able
to
escape
cultivation
and
form
occasional
volunteer
populations
(
i.
e.,
asparagus
and
celery).
Upon
further
investigation,
EPA
determined
that
other
plant
species
on
the
list
are
also
able
to
naturalize
in
some
region
of
the
United
States,
its
possessions,
or
territories
(
e.
g.,
soybean
and
corn).
The
ability
to
naturalize
is
an
apparently
common
feature
of
crop
plants
(
see
www.
plants.
usda.
gov).
As
EPA
has
previously
discussed,
naturalized
populations
of
most
crop
plants,
particularly
those
recently
establishing
such
populations,
would
be
expected
to
carry
with
them
a
suite
of
traits
suitable
to
cultivation
in
a
managed
habitat
but
that
confer
a
selective
disadvantage
on
plants
in
the
wild.
For
these
crops,
the
traits
that
make
them
useful
to
humans
also
reduce
their
competitive
ability
in
nonagricultural
habitats
(
see
Unit
II.
A.
1).

b.
Exemption
criterion
conditional
on
Agency
determination
The
Agency
recognizes
that
many
PVCP­
PIPs
would
reasonably
be
expected
to
pose
low
risk
with
respect
to
potential
for
disturbing
existing
ecological
relationships
among
plants
even
though
the
crop
plant
containing
the
PVCP­
PIP
is
not
on
the
above
list.
In
such
cases,
although
there
exists
the
potential
of
the
plant
containing
the
PVCP­
PIP
to
hybridize
with
wild
or
weedy
relatives
in
some
region
of
the
United
States,
its
possessions,
or
territories,
additional
events
are
likely
necessary
for
any
adverse
environmental
outcomes
to
occur,
e.
g.,
acquired
virus
resistance
must
confer
an
advantage
(
or
disadvantage)
on
the
recipient
plant
sufficient
to
alter
plant
population
dynamics.
As
discussed
above
in
Unit
II.
A.
2,
given
the
diversity
of
plants
that
could
contain
a
PVCP­
PIP
and
the
complex
and
variable
nature
of
ecosystems,
EPA's
challenge
is
to
develop
an
objectively
defined
criterion
that
would
describe
for
regulatory
purposes
only
those
PVCP­
PIP/
plant
combinations
that
would
likely
not
significantly
disrupt
natural
plant
population
dynamics.
However,
developing
such
a
criterion
may
currently
not
be
feasible
because
of
insufficient
information
to
make
generic
determinations
regarding
characteristics
of
a
plant
and/
or
PVCP­
PIP
that
would
indicate
such
events
are
unlikely
to
occur.

EPA
does
not
believe
it
can
develop
at
this
time
a
broader
categorical
exemption
criterion
than
that
discussed
above
(
Unit
II.
A.
2.
a)
which
allows
developers
to
self­
determine
whether
their
PVCP­
PIP/
plant
combination
meets
the
criterion.
However,
by
relying
on
a
case­
by­
case
Agency
determination
of
whether
the
PVCP­
PIP/
plant
combination
meets
a
criterion,
EPA
might
be
able
to
expand
any
exemption
to
include
a
larger
set
of
PVCP­
PIP/
plant
combinations
expected
to
present
low
risk.
Nevertheless,
even
with
an
Agency
determination,
EPA
must
still
define
a
criterion
with
sufficient
precision
such
that
the
public
would
be
able
to
understand
what
products
would
qualify
and
evaluate
whether
they
meet
the
standard
for
a
FIFRA
exemption.
Such
a
criterion
is
difficult
to
develop
because
many
characteristics
would
influence
this
determination
in
ways
that
could
only
be
poorly
defined
for
the
entire
class
of
PVCP­
PIPs.
In
addition,
many
relevant
considerations
such
as
the
impact
of
virus
infection
on
wild
plant
populations
and
the
Page
8
of
34
likely
selective
advantage
afforded
by
acquisition
of
virus
resistance
are
currently
poorly
understood
for
the
vast
majority
of
species.
The
scarcity
of
research
in
this
area
makes
it
particularly
difficult
to
construct
criteria
describing
low
risk
groups
with
sufficient
precision
such
that
a
product
developer
or
the
public
would
be
able
to
effectively
determine
whether
a
product
would
qualify
for
the
exemption.

Nonetheless,
in
addition
to
the
categorical
exemption
for
a
subset
of
PVCP­
PIPs
discussed
above
in
Unit
II.
A.
2.
a,
EPA
also
believes
that
a
criterion
conditional
on
Agency
determination
could
be
developed
based
on
whether
the
transgenic
plant
itself
is
a
weedy
or
invasive
species
or
whether
gene
flow
could
occur
from
the
transgenic
plant
to
a
weedy
or
invasive
species
or
endangered/
threatened
species.
Given
that
such
plants
already
are
associated
with
serious
environmental
concerns,
any
disruption
of
their
population
dynamics
could
have
significant
consequences.
EPA
has
therefore
determined
that
PVCP­
PIPs
that
are
in
or
could
potentially
end
up
in
such
plants
through
gene
flow
would
have
to
go
through
the
registration
review
process.
Although
the
Agency
recognizes
that
transfer
of
a
PVCP­
PIP
into
a
nearby
relative
that
is
not
a
weedy
or
invasive
species
also
has
the
potential
for
altering
plant
population
dynamics,
the
changes
in
such
circumstances
are
unlikely
to
rise
to
the
level
requiring
the
regulation
provided
by
the
registration
process
because
the
plant
population
acquiring
the
virus
resistance
trait
is
not
already
weedy
or
invasive.
Species
composition
of
natural
communities
is
likely
a
dynamic
variable,
constantly
changing
in
response
to
diverse
environmental
factors.
The
changes
in
plant
population
dynamics
potentially
introduced
when
a
plant
that
is
not
weedy
or
invasive
acquires
virus
resistance
are
likely
to
be
within
the
range
of
changes
that
happen
naturally
within
plant
communities
without
adverse
effects.

Accordingly,
EPA
is
considering
an
approach
under
which
PVCP­
PIP/
plant
combinations
that
fail
to
meet
paragraph
(
a)(
1)
could
still
meet
criterion
(
a),
subject
to
an
Agency
review
to
determine
whether
they
meet
a
different
set
of
conditions
related
to
this
issue.
Under
such
an
approach,
a
PVCP­
PIP
would
meet
criterion
(
a)
under
paragraph
(
2)
if
the
Agency
determines
that
the
plant
containing
the
PVCP­
PIP
(
i)
is
itself
not
a
weedy
or
invasive
species
outside
of
agricultural
fields
in
the
United
States,
its
possessions,
or
territories,
and
(
ii)
does
not
have
relatives
outside
of
agricultural
fields
in
the
United
States,
its
possessions,
or
territories
that
are
weedy
or
invasive
species
or
endangered/
threatened
species
with
which
it
can
produce
viable
hybrids
in
nature.

Under
such
an
approach,
PVCP­
PIPs
could
qualify
for
exemption
if
they
are
in
a
plant
species
that
is
not
a
weedy
or
invasive
species
and
does
not
have
relatives
that
are
weedy
or
invasive
species
based
on
the
low
probability
that
acquisition
of
a
virus­
resistance
trait
would
confer
sufficient
additional
competitive
advantage
on
plants
that
are
not
already
weedy
or
invasive
to
lead
to
adverse
environmental
outcomes.
Even
in
cases
when
a
plant
population
is
under
intense
disease
selection
pressure,
this
selective
pressure
is
unlikely
to
be
the
only
condition
restraining
the
population.
It
is
unlikely
that
the
use
of
PVCP­
PIPs
would
affect
wild
or
weedy
relatives
differently
than
what
has
occurred
in
the
past
with
virus
resistant
varieties
developed
through
traditional
breeding
and
grown
throughout
the
United
States
over
many
years.
The
source
of
resistance
traits
in
such
conventionally
bred
crops
is
often
a
wild
relative
of
the
crop.
There
is
no
indication
that
growing
such
crop
plants
near
wild
or
weedy
relatives
has
resulted
in
these
relatives
becoming
any
more
of
a
weed
problem
than
they
were
previously
due
to
acquisition
of
the
virus­
resistant
trait
(
Ref.
2).
Page
9
of
34
In
addition,
outbreeding
depression
between
crop
plants
and
their
wild
relatives
appears
to
be
more
common
than
hybrid
vigor
(
Ref.
23).
In
outbreeding
depression,
mating
between
individuals
from
two
different
environments
can
disrupt
gene
combinations
that
are
favored
by
natural
selection
in
each
environment.
Resulting
offspring
may
have
phenotypes
that
are
poorly
adapted
to
the
habitat
of
either
parent.
Thus,
hybrid
offspring
acquiring
a
PVCP­
PIP
are
often
likely
to
be
less
competitive
than
their
wild
parent
in
nature.
While
this
observation
supports
the
contention
that
crop­
wild
hybrids
containing
a
PVCP­
PIP
are
unlikely
to
outcompete
other
plants,
it
also
highlights
the
hazard
crop­
wild
hybridization
may
pose
to
endangered/
threatened
species.
EPA
thus
addresses
endangered/
threatened
species
in
criterion
(
a)(
2)
to
ensure
that
a
PVCP­
PIP
would
not
exacerbate
population
loss
in
such
species.

When
EPA
asked
the
FIFRA
SAP
in
2004
about
the
likelihood
that
plant
populations
freed
from
viral
pressure
could
have
increased
competitive
ability
leading
to
changes
in
plant
population
dynamics,
the
FIFRA
SAP
offered
the
following
opinion:
"[
b]
ased
on
knowledge
obtained
from
observation
of
cultivated
crops
in
the
agroecosystem,
the
majority
of
the
[
2004]
Panel
concluded
that
it
would
be
unlikely
that
a
plant
population
freed
from
viral
pressure
would
give
a
plant
species
a
competitive
advantage"
(
Ref.
22).

EPA
means
by
the
term
"
weedy
or
invasive
species"
plants
that
are:
(
1)
either
non­
native
(
or
alien)
to
the
ecosystem
under
consideration
or
aggressive
competitors
in
their
natural
ecosystems,
and
(
2)
whose
introduction
causes
or
is
likely
to
cause
economic
or
environmental
harm
or
harm
to
human
health.
EPA
considers
a
non­
native
(
or
alien)
species
to
be
synonymous
with
an
introduced
species,
or
one
that
occurs
in
a
region
in
which
it
is
not
native.
EPA
uses
the
terms
endangered
species
and
threatened
species
consistent
with
their
meaning
under
the
Endangered
Species
Act.

During
its
review,
EPA
would
consider
the
most
recent
scientific
information
about
the
plant
species
containing
the
PVCP­
PIP
and
its
wild
or
weedy
relatives
to
evaluate
the
potential
for
weedy
or
invasive
behavior,
including
whether
any
of
these
species
are
extending
their
range.
The
Agency
would
evaluate
a
number
of
sources
including
existing
lists
of
invasive
weeds,
e.
g.,
the
Federal
Noxious
Weed
List.
Examples
of
plants
related
to
crop
species
that
generally
are
considered
to
be
weedy
or
invasive
species
by
a
number
of
organizations
are
animated
oats
(
Avena
sterilis),
johnsongrass
(
Sorghum
halepense),
red
rice
(
Oryza
punctata),
wild
safflower
(
Carthamus
oxyacantha),
and
wild
sugarcane
(
Saccharum
spontaneum).
Inclusion
on
any
given
list
would
be
informative,
but
not
determinative
for
the
Agency's
evaluation.
Examination
of
existing
lists
has
shown
that
different
organizations
use
different
criteria
for
listing
species
depending
on
the
goals
and
missions
of
those
organizations.
Thus
the
Agency
would
use
existing
lists
as
a
resource
much
as
it
would
use
published
literature,
rather
than
as
determinative
sources.
For
example,
plants
that
may
form
volunteer
populations
in
agricultural
fields
are
considered
weeds
by
some
organizations
and
may
appear
on
those
organizations'
weed
lists,
but
for
reasons
described
above
in
this
Unit,
EPA
would
not
consider
propensity
to
volunteer,
i.
e.,
to
grow
in
a
field
from
seeds
dropped
from
the
previous
crop
rotation,
to
be
indicative
of
general
weediness
potential
for
a
plant.
EPA
would
include
consideration
of
whether
the
plant
is
invasive
or
weedy
outside
of
agricultural
fields
to
emphasize
that
the
key
consideration
is
the
plant's
behavior
in
natural
settings,
including
semi­
managed
habitat
surrounding
agricultural
fields
as
opposed
to
its
behavior
within
the
fields
themselves.
Page
10
of
34
EPA
does
not
intend
to
develop
or
maintain
its
own
list
of
weedy
and
invasive
species.
Plants
are
regularly
being
newly
classified
as
weedy
or
invasive
by
various
weed
societies
and
other
organizations
as
more
information
is
acquired
and
as
plants
extend
their
ranges.
Given
the
difficulties
associated
with
developing
and
maintaining
a
comprehensive
list
and
the
many
considerations
comprising
weediness
or
invasiveness,
the
Agency
believes
that
individual
caseby
case
determination
for
each
plant
would
be
preferable.

For
the
purposes
of
paragraph
(
a)(
2),
EPA
would
focus
on
whether
the
recipient
plant
"
can
produce
viable
hybrids
in
nature"
because
the
Agency
believes
that
this
characteristic
is
a
critical
attribute
that
would
determine
the
potential
for
introgression
of
the
PVCP­
PIP
into
a
native
or
naturalized
plant
population.
Although
hybrids
must
be
able
to
reproduce
themselves
in
order
for
introgression
to
occur,
the
Agency
has
chosen
to
focus
on
the
production
of
"
viable"
hybrids
(
i.
e.,
those
that
are
able
to
grow)
because
this
characteristic
may
be
described
more
clearly
in
a
regulatory
standard
than
examining
the
reproductive
potential
of
any
hybrids.
In
many
cases,
reproductive
potential
of
hybrids
has
not
been
fully
investigated.
Given
that
reduced
fertility
in
F1
crop­
wild
hybrids
is
frequently
restored
to
normal
in
subsequent
generations
(
Ref.
20),
measurement
of
hybrid
fertility
involves
consideration
of
several
generations.
In
addition,
viability
is
the
appropriate
standard
because
even
very
low
rates
of
gene
transfer
can
lead
to
introgression
(
Ref.
24),
suggesting
that
any
degree
of
hybrid
fertility
could
indicate
the
potential
for
introgression
to
occur.
The
Agency
recognizes
that
introgression
of
a
trait
such
as
virus
resistance
into
natural
plant
populations
does
not
automatically
confer
a
competitive
advantage
to
the
recipient
population.
However,
at
this
time,
there
is
little
information
available
to
predict
categorically
whether
acquisition
of
such
a
trait
might
affect
the
competitiveness
of
a
specific
plant
population,
and
the
available
information
does
not
allow
the
Agency
to
make
this
determination
a
priori.
EPA
therefore
relied
on
the
ability
to
produce
viable
hybrids
when
developing
this
criterion.
The
language
also
clarifies
that
the
relevant
question
is
whether
the
hybrid
can
be
produced
"
in
nature."
The
fact
that
plants
could
be
crossed
in
the
laboratory
is
not
necessarily
indicative
of
a
plant's
true
reproductive
potential.
The
Agency's
focus
would
be
on
whether
a
viable
hybrid
could
be
produced
under
normal
growing
conditions
in
the
field
or
in
nature,
rather
than
under
controlled
experimental
conditions
that
might
have
little
relevance
to
how
the
product
would
behave
in
the
environment.

For
the
purposes
of
paragraph
(
a)(
2),
EPA
is
considering
whether
to
limit
the
exemption
to
those
PVCP­
PIPs
present
in
plant
that
is
itself
not
a
weedy
or
invasive
species
in
the
United
States,
its
possessions,
or
territories
and
does
not
have
relatives
in
the
United
States,
its
possessions,
or
territories
that
are
weedy
or
invasive
species
or
endangered/
threatened
species.
The
Agency
believes
the
entire
United
States,
including
its
possessions
and
territories
is
the
relevant
scope
of
inquiry
because
an
exemption
would
carry
no
limitations
on
where
the
exempted
PVCP­
PIP
plant
combination
could
be
planted
and
thus
could
be
planted
in
all
areas
subject
to
U.
S.
law.
FIFRA
section
2(
aa)
defines
"
State"
as
"
a
State,
the
District
of
Columbia,
the
Commonwealth
of
Puerto
Rico,
the
Virgin
Islands,
Guam,
the
Trust
Territory
of
the
Pacific
Islands,
and
American
Samoa.

The
Agency's
rationale
for
excluding
from
exemption
plants
that
fail
to
meet
criterion
(
a)(
2)
would
be
the
recognition
that
weedy
and
invasive
species
are
already
associated
with
significant
economic
costs
and
ecological
damage.
For
plants
that
fail
to
meet
criterion
(
a)
and
thus
do
not
qualify
for
exemption,
an
applicant
may
apply
for
a
registration
under
section
3
of
FIFRA.
Page
11
of
34
During
review
of
the
registration
application,
in
addition
to
other
considerations,
the
Agency
may
evaluate
whether
a
plant's
weedy
or
invasive
characteristics
could
be
augmented
by
acquisition
of
a
virus
resistance
trait.
A
case­
by­
case
review
for
registration
would
allow
the
Agency
to
evaluate
in
depth
the
potential
impacts
of
acquisition
of
a
PVCP­
PIP
by
considering,
for
example,
the
effect
of
virus
infection
on
such
species,
the
existence
and
impact
of
any
natural
virus
resistance
in
the
population,
the
overlap
of
the
plant's
distribution
with
crop
cultivation
areas,
and
other
relevant
traits.
As
part
of
registration,
the
Agency
could
also
impose
control
conditions
if
possible
and
appropriate.

c.
Other
approaches
In
1994
EPA
proposed
two
different
alternatives
for
exempting
PVCP­
PIPs
from
FIFRA
requirements.
Although
the
Agency
is
still
considering
these
options,
they
are
no
longer
the
Agency's
preferred
approach
for
a
number
of
reasons.
One
of
these
options
contained
criteria
directed
towards
addressing
concerns
associated
with
gene
transfer.
Under
this
alternative,
the
Agency
defined
a
set
of
criteria
to
identify
those
PVCP­
PIP/
plant
combinations
with
the
lowest
potential
to
confer
selective
advantage
on
wild
or
weedy
plant
relatives.
Only
those
PVCP­
PIPs
so
identified
would
have
been
exempt
from
regulation
under
the
1994
proposal.
In
1994
EPA
described
this
alternative
exemption
as
follows:

"
Coat
proteins
from
plant
viruses
[
would
be
exempt]
if
the
genetic
material
necessary
to
produce
a
coat
protein
is
introduced
into
a
plant's
genome
and
the
plant
has
at
least
one
of
the
following
characteristics:

"(
1)
The
plant
has
no
wild
relatives
in
the
United
States
with
which
it
can
successfully
exchange
genetic
material,
i.
e.,
corn,
tomato,
potato,
soybean,
or
any
other
plant
species
that
EPA
has
determined
has
no
sexually
compatible
wild
relatives
in
the
United
States.

"(
2)
It
has
been
demonstrated
to
EPA
that
the
plant
is
incapable
of
successful
genetic
exchange
with
any
existing
wild
relatives
(
e.
g.,
through
male
sterility,
self­
pollination).

"(
3)
If
the
plant
can
successfully
exchange
genetic
material
with
wild
relatives,
it
has
been
empirically
demonstrated
to
EPA
that
existing
wild
relatives
are
resistant
or
tolerant
to
the
virus
from
which
the
coat
protein
is
derived
or
that
no
selective
pressure
is
exerted
by
the
virus
in
natural
populations"
(
59
FR
60504).

EPA
carefully
reconsidered
this
1994
proposal
in
its
deliberations
and
presented
several
modified
criteria
to
the
FIFRA
SAP
at
the
October
2004
meeting
for
consideration.
In
light
of
comments
received
from
the
FIFRA
SAP
and
additional
scientific
information
available
since
1994,
EPA
no
longer
believes
this
alternative
would
adequately
address
questions
associated
with
weediness
in
a
manner
that
could
be
reasonably
implemented.
However,
EPA
still
considers
that
it
would
be
appropriate
to
limit
an
exemption
based
on
the
concerns
outlined
in
the
earlier
proposal
associated
with
acquisition
of
virus
resistance
through
hybridization
with
a
transgenic
plant
containing
a
PVCP­
PIP.

Although
similar
in
intent
to
the
first
characteristic
of
this
option
proposed
in
1994,
criterion
(
a)
in
this
document
focuses
on
the
potential
to
"
form
viable
hybrids
in
nature"
rather
than
simply
"
exchange
genetic
material"
because
the
former
is
more
critical
for
determining
whether
a
Page
12
of
34
PVCP­
PIP
might
negatively
affect
a
recipient
plant
population.
The
ability
to
exchange
genetic
material,
which
is
often
demonstrated
by
performing
hand
crosses
in
the
laboratory,
may
not
indicate
any
relevant
information
about
how
the
plants
would
behave
in
nature.
The
approach
the
Agency
is
currently
considering
also
expands
the
list
of
plants
meeting
this
condition
beyond
the
four
in
the
1994
proposal.
When
EPA
presented
a
similar
criterion
to
the
2004
SAP,
they
responded
that
"
the
Panel
was
of
the
opinion
that
the
absence
of
a
competent
wild/
weedy
relative
positioned
in
relation
to
the
plant
containing
the
PVCP­
PIP
was
an
appropriate
condition"
(
Ref.
22).

EPA
now
also
believes
that
the
second
characteristic
of
the
option
proposed
in
1994
may
be
insufficient
based
on
the
conclusions
of
the
2004
SAP
that
current
methods
of
bioconfinement
are
imperfect
and
are
unlikely
to
adequately
restrict
gene
flow
(
Ref.
22).
The
Agency
asked
whether
the
condition
that
"
genetic
exchange
between
the
plant
into
which
the
PVCP­
PIP
has
been
inserted
and
any
existing
wild
or
weedy
relatives
is
substantially
reduced
by
modifying
the
plant
with
a
scientifically
documented
method,
(
e.
g.,
through
male
sterility)"
would
be
necessary
and/
or
sufficient
to
minimize
the
potential
for
a
PVCP­
PIP
to
harm
the
environment
through
gene
transfer
from
the
crop
plant
containing
the
PVCP­
PIP
to
wild
or
weedy
relatives.
The
Panel
"
accepted
that
tactics
aiming
at
diminished
gene
exchange
are
highly
desirable
and
even
necessary
but
are
not
sufficient."
EPA
believes
that
criterion
(
a)
in
the
approach
the
Agency
is
currently
considering
more
precisely
defines
those
situations
where
successful
genetic
exchange
is
either
not
possible
(
under
paragraph
(
1))
or
not
of
concern
(
under
paragraph
(
2)).
However,
EPA
is
still
considering
whether
it
would
be
possible
to
construct
a
criterion
involving
significantly
reduced
potential
for
gene
exchange
such
as
that
presented
to
the
2004
SAP
that
would
enable
the
Agency
to
determine
with
review
that
a
product
presents
low
risk
with
respect
to
concerns
associated
with
gene
flow.

EPA
believes
that
the
third
characteristic
of
the
option
proposed
in
1994
is
sound
conceptually,
but
impractical
to
implement
in
an
exemption.
Appropriate
and
generically
applicable
protocols
that
could
be
followed
to
demonstrate
convincingly
that
either
condition
listed
in
characteristic
(
3)
was
met
are
unavailable.
In
particular,
generic
sampling
protocols
are
especially
difficult
to
develop
given
that
plant
species
are
extremely
diverse,
e.
g.,
in
geographic
distribution
and
life
history.
Based
on
subsequent
consideration,
the
Agency
presented
the
following
revision
to
the
2004
SAP
for
their
consideration:
"
all
existing
wild
or
weedy
relatives
in
the
United
States
with
which
the
plant
can
produce
a
viable
hybrid
are
tolerant
or
resistant
to
the
virus
from
which
the
coat
protein
is
derived."
Among
the
Panel
members,
"[
i]
t
was
generally
accepted
that
such
wording
was
not
helpful
for
a
number
of
reasons."
Specifically,
"[
t]
he
Panel
had
particular
difficulty
when
attempting
to
add
precision
to
approaches
that
should
be
followed
when
sampling
wild
and
weedy
relatives
for
the
occurrence
of
specific
virus
tolerance
or
resistance
as
specified
by
the
Agency."
However,
the
Agency
still
recognizes
the
scientific
utility
of
evaluating
the
impact
of
virus
infection
on
wild
or
weedy
plant
populations
that
could
acquire
a
PVCP­
PIP
through
gene
flow
when
attempting
to
determine
whether
a
PVCP­
PIP
presents
low
risk.
If
such
a
criterion
could
be
clearly
articulated
such
that
the
public
and
product
developers
would
have
a
reasonable
understanding
of
which
PVCP­
PIPs
would
qualify
for
exemption
and
which
would
not,
the
Agency
would
consider
incorporating
this
concept
into
an
exemption.

The
other
option
proposed
in
1994
did
not
contain
a
criterion
addressing
concerns
associated
with
gene
flow.
This
option
proposed
a
full
categorical
exemption
for
all
PVCP­
PIPs
(
59
FR
Page
13
of
34
60503).
Although
the
Agency
is
still
considering
this
option,
it
is
no
longer
the
Agency's
preferred
approach
for
a
number
of
reasons.
Specifically,
EPA
has
received
scientific
advice
since
the
1994
proposal
calling
into
question
the
Agency's
1994
rationale
that
all
PVCP­
PIPs
meet
the
FIFRA
25(
b)(
2)
exemption
standard,
including
gene
flow
considerations.
Although
EPA
believes
that
many
PVCP­
PIPs
present
low
risk
and
thus
meet
the
FIFRA
25(
b)(
2)
exemption
standard,
in
order
to
categorically
exempt
all
PVCP­
PIPs,
the
Agency
must
be
able
to
draw
this
conclusion
for
all
PVCP­
PIPs.
Advances
in
scientific
understanding
since
1994
suggest
it
may
not
be
possible
to
support
this
rationale
for
all
PVCP­
PIPs
and
that
certain
PVCP­
PIPs
may
pose
a
greater
level
of
risk
than
is
characteristic
of
the
group
as
a
whole.
For
example,
virus
resistance
is
common
in
natural
plant
populations
as
evidenced
by
conventionally­
bred
virus
resistant
plants
that
are
only
possible
due
to
existing
resistance
in
available
breeding
stock
(
Ref.
25).
This
fact
suggests
that
acquisition
of
virus
resistance
is
often
unlikely
to
introduce
a
novel
trait
into
many
plant
populations.
However,
some
notable
exceptions
to
the
ubiquity
of
virus
resistance
in
natural
plant
populations
exist
including
the
lack
of
successful
conventionally
bred
resistance
to
barley
yellow
dwarf
virus
in
major
crops
and
the
lack
of
natural
resistance
in
some
wild
relatives
of
these
crops
(
Ref.
19).
Such
information
suggests
that
acquisition
of
a
PVCP­
PIP
by
such
wild
relatives
of
these
plants
has
the
potential
to
free
these
wild
relatives
from
what
may
be
an
important
ecological
constraint.
The
conclusions
of
the
2004
FIFRA
SAP
are
consistent
with
the
idea
that
it
may
not
be
possible
to
apply
a
general
exemption
rationale
to
all
PVCP­
PIPs.
The
report
concluded
that
" 
PVCP­
PIPs
[
have]
no
inherent
capacity
to
harm
the
environment."
However,
"[
i]
t
was
recognized
that
knowledge
of
hybridization
potential
was
sparse
and
of
very
unequal
quality
but
the
likelihood
of
serious
economic
harm
was
such
that
some
plants
engineered
to
contain
stress
tolerant
traits
should
not
be
released"
(
Ref.
22).
Similarly,
the
2000
National
Research
Council
(
NRC)
report
recommended
that
because
of
concerns
associated
with
outcrossing
to
weedy
relatives,
"
EPA
should
not
categorically
exempt
viral
coat
proteins
from
regulation
under
FIFRA.
Rather,
EPA
should
adopt
an
approach,
such
as
the
agency's
alternative
proposal ,
that
allows
the
agency
to
consider
the
gene
transfer
risks
associated
with
the
introduction
of
viral
coat
proteins
to
plants"
(
Ref.
26).

B.
Viral
Interactions
A
key
concern
associated
with
PVCP­
PIPs
is
the
question
of
whether
they
could
affect
the
epidemiology
and
pathogenicity
of
plant
viruses.
Given
the
potential
impact
of
virus
infection,
such
changes
might
affect
competitiveness
of
plant
populations
thereby
altering
ecosystem
dynamics,
e.
g.,
through
changes
in
species
composition
of
populations,
resource
utilization,
or
herbivory.

Mixed
viral
infections
are
extremely
common
in
crops
and
other
plants
(
Ref.
27).
In
natural,
mixed
infections,
viral
genomes
from
different
strains
and/
or
different
species
simultaneously
infect
the
same
plant
and
thus
have
opportunities
to
interact
(
e.
g.,
through
recombination,
heterologous
encapsidation,
or
synergy).
In
spite
of
many
opportunities
for
interaction
in
nature,
such
events
rarely
lead
to
any
detectable
adverse
outcome
(
Ref.
28).
However,
such
in
planta
interactions
do
have
the
potential
to
result
in
a
virus
that
causes
increased
agricultural
or
other
environmental
damage.
For
example,
the
epidemic
of
severe
cassava
mosaic
disease
in
Uganda
is
thought
to
be
due
to
the
combination
and/
or
sequential
occurrence
of
several
phenomena
Page
14
of
34
including
recombination,
pseudorecombination,
and/
or
synergy
among
cassava
geminiviruses
(
Ref.
29).

In
transgenic
plants
containing
PVCP­
PIPs,
every
virus
infection
is
essentially
a
mixed
infection
with
respect
to
the
coat
protein
gene
(
Ref.
30).
The
key
question
facing
EPA
is
whether
interactions
between
such
introduced
plant
virus
sequences
and
other
invading
viruses
in
transgenic
plants
may
increase
in
frequency
or
be
unlike
those
expected
to
occur
in
nature
(
Ref.
31).
The
issues
associated
with
recombination,
heterologous
encapsidation,
and
synergy
are
briefly
described
below.
EPA
provides
a
general
overview
of
each
of
the
processes
separately,
followed
by
a
brief
review
of
relevant
field
studies
that
investigated
these
processes.

1.
Recombination
Recombination
is
a
natural
process
that
can
occur
during
replication
of
DNA
or
RNA
whereby
new
combinations
of
genes
are
produced.
Plant
virus
recombination
can
occur
between
members
of
the
same
virus
pathotype
in
natural
infections,
contributing
to
the
number
of
variants
that
exist
within
that
pathotype.
Recombination
can
also
occur
when
different
viruses
coinfect
the
same
plant
and
interact
during
replication
to
generate
virus
progeny
that
have
genetic
material
from
each
of
the
different
parental
genomes.
Although
recombination
likely
occurs
regularly
in
mixed
viral
infections,
recombination
only
rarely
leads
to
viable
viruses
with
truly
novel
behavior
and/
or
characteristics
or
any
detectable
adverse
outcome.
In
order
to
persist
in
the
field,
a
recombinant
virus
must
compete
with
variants
of
the
parental
viruses
that
are
already
highly
adapted
to
existing
conditions,
in
all
stages
of
the
infective
cycle,
for
example
in
transmission,
gene
expression,
replication,
and
assembly
of
new
virions
(
Ref.
28).
An
analysis
of
cucumber
mosaic
virus
(
CMV)
isolates
in
natural
populations
showed
that
viable
recombinants
were
very
rarely
recovered
in
mixed
infections
(
Ref.
32).

However,
laboratory
experiments
suggest
that
viruses
with
increased
pathogenicity
or
altered
epidemiology
can
be
created
through
recombination.
A
pseudorecombinant
strain
created
by
experimentally
combining
regions
of
the
CMV
and
tomato
aspermy
cucumovirus
(
TAV)
genomes
was
found
to
have
more
severe
symptoms
than
either
of
the
parental
genomes,
although
the
recombinant
wasn't
able
to
move
beyond
infection
of
the
initially
infected
cells
(
Ref.
33).
Experiments
have
also
shown
interspecific
recombination
between
CMV
and
TAV
under
conditions
in
which
recombinants
would
not
be
expected
to
have
any
particular
fitness
advantage
(
Ref.
34).
In
another
example,
alteration
of
the
host
range
of
tobacco
mosaic
virus
(
TMV)
occurred
when
a
chimeric
virus
expressed
the
coat
protein
from
alfalfa
mosaic
virus
(
AMV)
instead
of
its
own
(
Ref.
35).

Moreover,
even
though
selection
in
the
field
appears
to
act
against
persistence
of
a
new
recombinant
virus,
recombination
is
thought
to
play
a
significant
role
in
virus
evolution.
Evidence
of
past
recombination
having
led
to
the
creation
of
new
DNA
and
RNA
viruses
has
been
documented
in
a
number
of
different
groups
including
bromoviruses
(
Ref.
36),
caulimoviruses
(
Ref.
37),
luteoviruses
(
Ref.
38),
nepoviruses
(
Ref.
39),
cucumoviruses
(
Ref.
40),
and
geminiviruses
(
Refs.
29,
41).
Sequence
analysis
of
viruses
from
the
family
Luteoviridae
indicated
that
this
family
has
evolved
via
both
intra­
and
interfamilial
recombination
(
Ref.
42).
Page
15
of
34
Several
instances
can
be
cited
in
which
relatively
recent
recombination
events
appear
to
have
resulted
in
the
creation
of
new
viruses.
For
example,
numerous
recombination
events
among
tomato­
infecting
begomoviruses
around
the
Nile
and
Mediterranean
Basins
are
likely
at
least
partially
responsible
for
numerous
whitefly­
transmitted
tomato
diseases
that
have
emerged
in
the
last
20
years
(
Ref.
43).
In
addition,
a
natural
recombinant
between
Tomato
yellow
leaf
curl
Sardinia
virus
and
Tomato
yellow
leaf
curl
virus
was
detected
in
southern
Spain
with
a
novel
pathogenic
phenotype
that
might
provide
it
with
selective
advantage
over
the
parental
genotypes
(
Ref.
44).
Finally,
analysis
of
a
newly
described
Curtovirus
species
associated
with
disease
of
spinach
in
southwest
Texas
suggests
that
it
may
be
the
result
of
recombination
among
previously
described
Curtovirus
species
(
Ref.
45).

In
addition
to
virus­
virus
recombination,
recombination
has
also
been
found
to
occur
between
virus
and
plant
host
RNA.
Sequence
analysis
of
the
5'
terminal
sequence
of
potato
leafroll
virus
(
PLRV)
suggests
that
it
arose
via
recombination
with
host
mRNA
(
Ref.
46).
Evidence
suggests
that
such
recombination
events
can
affect
virus
virulence
(
for
review
see
Ref.
47).
Several
experiments
have
therefore
investigated
whether
a
PVCP­
PIP
integrated
into
a
plant
genome
is
able
to
recombine
with
the
genetic
material
of
an
infecting
virus.
Like
a
plant
host
genome,
viral
transgenes
would
be
available
for
recombination
with
infecting
viruses,
and
portions
of
the
transgene
could
thus
be
incorporated
into
the
replicating
virus.
Laboratory
experiments
with
pseudorecombinant
transcripts
of
papaya
ringspot
virus
(
PRSV)
have
shown
that
recombinant
viruses
that
theoretically
could
be
produced
in
field­
grown
transgenic
papaya
would
be
able
to
affect
the
virulence
of
the
infecting
strains
(
Ref.
48).

Several
laboratory
experiments
have
also
investigated
the
potential
for
recombination
between
viral
transgenes
and
infecting
viruses
of
the
same
species.
These
experiments
show
that
recombination
can
occur
between
viral
transgenes
and
both
RNA
viruses
(
Refs.
49,
50,
51,
52,
53)
and
DNA
viruses
(
Refs.
54,
55,
56,
57).
However,
the
transgenic
plants
used
in
these
DNA
virus
experiments
actually
show
no
viral
resistance;
attempts
to
develop
transgenic
DNA
virusresistant
plants
in
general
have
had
little
success
(
Ref.
27).
In
addition,
to
facilitate
the
detection
of
recombinants,
most
of
these
experiments
were
conducted
under
conditions
of
high
selective
pressure
favoring
the
recombinant,
i.
e.,
only
recombinant
viruses
were
viable.
The
selective
pressure
under
normal
field
conditions
would
likely
favor
the
parental
viruses
rather
than
a
recombinant
as
parental
viruses
will
outnumber
the
new
recombinant
and
will
be
competent
in
all
of
the
functions
needed
for
propagation.

The
above
information
suggests
that
recombination
among
viruses
likely
leads
to
rare
instances
of
adverse
changes
in
virus
epidemiology
and/
or
pathogenicity.
Based
on
the
available
information,
EPA
is
not
able
to
rule
out
the
concern
that
viable
recombinant
viruses
could
arise
in
plants
containing
a
PVCP­
PIP.
The
body
of
existing
scientific
knowledge
supports
the
contention
that
recombination
between
a
PVCP­
PIP
and
an
infecting
virus
could
lead
to
environmental
impacts
in
some
instances.
However,
the
vast
majority
of
such
interactions
are
expected
to
be
no
different
from
those
that
would
occur
in
a
natural
mixed
infection
of
the
respective
viruses
and
would
not
cause
any
adverse
environmental
effects
beyond
what
could
occur
in
the
absence
of
the
PVCP­
PIP.
EPA
believes
that
the
Agency
has
identified
in
this
discussion
those
few
circumstances
in
which
the
potential
recombinants
involving
the
PVCP­
PIP
could
involve
viruses
that
would
otherwise
not
be
expected
to
interact
in
a
mixed
infection
found
in
nature
(
i.
e.,
leading
to
"
novel
viral
interactions")
and
for
which
additional
data
or
information
Page
16
of
34
would
therefore
be
needed
to
make
a
determination
that
the
PVCP­
PIP
could
qualify
for
exemption
from
regulation
under
FIFRA.

2.
Heterologous
encapsidation
Heterologous
encapsidation
occurs
when
the
coat
protein
subunits
of
one
virus
surround
and
encapsidate
the
viral
genome
of
a
different
virus.
The
coat
protein,
possibly
in
conjunction
with
other
viral
factors,
is
essential
for
transmission
and
responsible
for
conferring
the
high
degree
of
vector
specificity.
Therefore,
a
heterologously
encapsidated
viral
genome
may
be
transmitted
by
the
vectors
of
the
virus
contributing
the
coat
protein
rather
than
the
vectors
of
the
virus
contributing
the
viral
genome.
For
many
viruses,
transmission
from
plant
to
plant
occurs
by
insect
vectors,
and
each
virus
tends
to
be
transmitted
by
only
one
type
of
insect
(
Ref.
58).
To
the
extent
that
vectors
visit
different
groups
of
plants,
vectors
carrying
a
heterologously
encapsidated
viral
genome
may
carry
it
to
a
plant
it
does
not
normally
encounter
(
Ref.
30).

Most
evidence
of
heterologous
encapsidation
is
derived
from
laboratory
or
greenhouse
studies.
The
high
frequency
of
mixed
infections
suggests
the
potential
for
heterologous
encapsidation
to
occur
in
nature
is
great,
but
most
mixed
infections
do
not
lead
to
heterologous
encapsidation,
and
those
virus
interactions
that
do
occur
are
very
specific
(
Ref.
59).
Heterologous
encapsidation
is
however
known
to
be
a
regular
occurrence
among
some
plant
viruses.
Its
frequency
depends
on
the
viruses
involved
and
is
more
likely
to
occur
among
close
relatives
(
Ref.
60).
An
expansion
of
aphid
vector
specificity
due
to
heterologous
encapsidation
was
first
observed
in
plants
infected
with
two
different
isolates
of
barley
yellow
dwarf
luteovirus
(
BYDV;
Ref.
61)
and
was
later
shown
to
be
a
general
phenomenon
among
these
viruses
in
natural
populations
of
several
plant
species
(
Ref.
62).
Heterologous
encapsidation
was
also
shown
to
occur
in
potyviruses.
An
isolate
of
zucchini
yellow
mosaic
virus
(
ZYMV)
that
is
normally
non­
aphid
transmissible
due
to
a
transmission­
deficient
coat
protein
was
found
to
be
transmitted
by
the
aphid
vector
due
to
heterologous
encapsidation
when
in
a
mixed
infection
with
another
potyvirus,
papaya
ringspot
virus
(
Ref.
63).
Heterologous
encapsidation
may
sometimes
be
an
important
route
of
disease
transmission.
For
example,
umbraviruses
do
not
encode
a
coat
protein,
and
therefore
transmission
between
plants
occurs
through
encapsidation
by
an
aphid­
transmissible
luteovirus
coat
protein
(
Ref.
64).

Heterologous
encapsidation
is
considered
a
possible
environmental
concern
because
of
the
potential
that
a
virus
may
be
spread
to
plants
it
ordinarily
had
no
means
of
reaching
and
thus
could
not
have
infected.
Such
concerns
are
largely
mitigated
by
several
factors.
First,
the
heterologously
encapsidated
viral
genome
may
not
be
able
to
replicate
in
the
new
host
plant
and
could
therefore
not
actually
infect
it.
Second,
if
replication
is
possible
in
the
new
plant,
the
replicating
viral
genome
would
produce
its
own
coat
protein
rather
than
that
which
heterologously
encapsidated
it.
This
virus
would
not
be
transmitted
by
the
new
vector
which
brought
the
heterologously
encapsidated
nucleic
acid
to
the
plant.
The
epidemiological
consequences
of
such
heterologous
encapsidation
would
thus
be
limited.
Another
consideration
is
that
for
some
viruses,
effective
vector
transmission
may
depend
on
more
than
the
coat
protein
(
Ref.
65),
requiring
other
regions
of
the
viral
genome,
e.
g.,
coat
protein
read­
through
domains
or
helper
factors,
and
a
PVCP­
PIP
producing
other
viral
proteins
would
not
qualify
for
the
Page
17
of
34
exemption
discussed
here.
Thus,
in
such
cases
heterologous
encapsidation
would
not
lead
to
a
change
in
vector
specificity.
Finally,
in
large
monocultures
of
crop
plants,
a
vector
is
most
likely
to
transmit
even
a
heterologously
encapsidated
virus
to
the
same
plant
that
the
virus
is
already
able
to
infect
(
Ref.
65).

EPA
has
evaluated
a
number
of
circumstances
to
determine
whether
heterologous
encapsidation
might
nevertheless
be
of
environmental
concern.
For
example,
EPA
considered
whether
a
virus
that
is
heterologously
encapsidated
and
carried
to
a
new
host
plant
might
be
exposed
to
a
vector
that
feeds
on
the
new
host
plant
and
perhaps
other
plants
the
virus
ordinarily
could
not
access.
EPA
considered
whether
this
new
vector
might
in
some
cases
be
able
to
transmit
the
virus
even
though
it
would
be
encapsidated
in
its
own
coat
protein,
thereby
expanding
the
virus'
vector
range.
A
new
vector
could
possibly
transfer
the
virus
to
new
host
plants,
thus
expanding
the
plant
host
range
as
well
(
Ref.
27).
EPA
considers
expansion
of
host
range
through
heterologous
encapsidation
to
be
an
extremely
unlikely
outcome
because
such
an
outcome
depends
on
each
event
in
a
series
of
rare
events
occurring.
Should
the
probability
of
occurrence
of
any
one
event
in
this
series
be
zero,
the
adverse
event
of
an
expanded
host
range
would
not
occur.
First,
a
virus
must
be
heterologously
encapsidated,
an
event
that
is
not
possible
for
every
viral
genome­
coat
protein
combination.
Second,
the
encapsidated
viral
genome
must
be
transmitted
by
a
new
vector.
Third,
the
transmission
must
be
to
a
new
host
plant.
Fourth,
the
heterologously
encapsidated
viral
genome
must
be
able
to
replicate
in
the
new
host
plant.
Fifth,
the
resulting
virus,
now
encapsidated
in
its
own
coat
protein,
must
be
exposed
to
a
new
vector
the
virus
never
encountered
before
that
is
nevertheless
able
to
transmit
it.
Finally,
the
virus
must
be
transmitted
by
this
vector
to
a
new
plant
that
the
virus'
prior
vectors
never
visited.
For
such
a
series
of
events
to
be
novel,
the
viruses,
vectors,
and
plants
involved
must
have
had
no
previous
opportunity
to
interact,
but
this
requirement
is
rarely
met.
For
example,
it
is
known
that
many
viruses
are
transmitted
by
polyphagous
insects,
which
would
facilitate
introduction
of
the
virus
to
many
potential
hosts
(
Ref.
27),
and
viruses
may
be
transmitted
at
low
frequency
by
a
range
of
species
other
than
their
primary
vector
or
mechanically,
e.
g.,
through
the
practices
of
modern
agriculture
(
Ref.
66).

Another
scenario
EPA
considered
is
that
with
a
high
enough
frequency
of
vector
transmission
to
a
new
host
plant
due
to
heterologous
encapsidation,
secondary
spread
among
new
plant
hosts
might
not
be
required
for
the
phenomenon
to
affect
the
population,
assuming
that
the
virus
is
able
to
decrease
the
new
host
plant's
growth
and/
or
reproduction.
Although
this
scenario
may
be
more
likely
to
occur
than
an
expansion
of
host
range
given
that
fewer
rare
events
would
have
to
occur,
any
impact
on
the
affected
plant
population
would
be
highly
localized
being
confined
to
plants
in
or
near
transgenic
crop
fields.
Such
negative
impacts
are
unlikely
to
be
sufficiently
detrimental
to
require
FIFRA
regulation
given
their
localized
nature
and
the
probability
that
common
agricultural
practices
(
e.
g.,
vector
control)
could
be
used
to
manage
the
problem.
Moreover,
although
isolated
instances
of
transmission
may
occur,
a
significant
proportion
of
a
plant
population
is
unlikely
to
be
infected
in
such
a
scenario.
For
example,
a
field
experiment
(
discussed
in
Unit
II.
B.
4)
showed
that
heterologous
encapsidation
led
to
infection
of
only
2%
of
plants
compared
to
99%
of
plants
infected
under
similar
conditions
by
a
virus
that
is
not
heterologously
encapsidated
(
Ref.
67).
Most
importantly,
the
heterologously
encapsidated
virus
will
still
have
no
way
to
spread
among
or
beyond
the
plants
of
the
affected
population.
In
the
case
where
a
plant
population
contains
relatively
few
individuals
such
that
the
impact
of
single
plant
infections
would
be
magnified,
plant
infections
are
even
less
likely
to
occur
because
in
Page
18
of
34
addition
to
the
inefficient
nature
of
heterologous
encapsidation,
the
vector
would
be
less
likely
to
feed
on
a
rare
plant
and
more
likely
to
feed
on
the
more
abundant
transgenic
crop
plants.
In
some
cases
a
vector
may
have
a
strong
preference
for
a
specific
plant
over
even
closely
related
plants
(
Ref.
68).

Finally,
EPA
considered
that
after
expansion
to
a
new
host,
rapid
selection
of
variants
best
adapted
to
the
new
environment
might
lead
to
the
evolution
of
a
new
virus
(
Ref.
27).
However,
all
viruses
that
are
occasionally
heterologously
encapsidated
and
transmitted
to
a
new
plant
host
have
had
the
opportunity
to
adapt
to
new
plant
environments.
The
opportunities
for
rapid
viral
evolution
presented
by
transgenic
plants
containing
PVCP­
PIPs
would
not,
under
any
reasonably
likely
circumstances,
be
fundamentally
different
from
what
occurs
in
nature
because
it
is
not
dependent
on
the
unique
combination
of
viruses
that
interact
but
rather
the
occurrence
of
a
virus
in
a
new
plant
host,
an
event
that
likely
occurs
in
nature
at
some
frequency
for
most
viruses
either
through
heterologous
encapsidation
or
through
occasional
transmission
that
occurs
mechanically
or
from
secondary
vectors
(
Ref.
66).

Experimental
studies
have
shown
that
the
PVC­
protein
in
transgenic
plants
has
the
ability
to
encapsidate
even
unrelated
infecting
viruses
(
Refs.
69,
70,
71,
72).
However,
heterologous
encapsidation
involving
a
viral
transgene
can
only
occur
if
it
expresses
coat
protein
that
possesses
the
appropriate
physical
parameters
to
encapsidate
the
viral
genome
of
infecting
viruses.
In
transgenic
VCP
plants
that
express
very
little
coat
protein
(
i.
e.,
those
relying
on
posttranscriptional
gene
silencing
to
confer
resistance),
the
probability
of
heterologous
encapsidation
would
be
very
small
except
in
cases
of
suppression
of
gene
silencing.
(
For
a
more
detailed
discussion
of
post­
transcriptional
gene
silencing
and
suppression
of
gene
silencing,
see
Unit
II.
B
of
Appendix
II:
Draft
Approach
to
Exempting
PVC­
Proteins
from
the
Requirement
of
a
Tolerance
under
FFDCA.)
In
addition,
as
with
recombination,
as
long
as
the
VCP
inserted
in
the
transgenic
plant
is
from
a
virus
that
normally
infects
the
plant
in
the
area
where
it
is
planted,
the
outcome
of
any
heterologous
encapsidation
that
may
occur
is
expected
to
be
the
same
in
transgenic
plants
as
in
natural,
mixed
infections.

3.
Synergy
In
synergy,
another
type
of
viral
interaction,
the
disease
severity
of
two
viruses
infecting
together
is
greater
than
expected
based
on
the
additive
severity
of
each
virus
alone.
For
example,
when
a
plant
containing
potato
virus
X
(
PVX)
is
coinfected
with
a
number
of
potyviruses
including
tobacco
vein
mottling
virus,
tobacco
etch
virus,
and
pepper
mottle
virus,
the
disease
symptoms
are
considerably
worsened
and
PVX
accumulates
to
a
greater
concentration
(
Ref.
73).
A
listing
of
reported
viral
synergisms
has
been
compiled
(
Ref.
74).

The
question
EPA
must
address
in
developing
an
exemption
is
whether
an
infecting
virus
might
have
increased
disease
severity
when
infecting
a
plant
containing
a
PVCP­
PIP.
For
this
to
occur,
the
PVC­
protein
must
be
at
least
one
of
the
factors
causing
synergy.
However,
the
coat
protein
is
considered
much
less
likely
to
be
responsible
for
synergism
than
other
parts
of
the
virus
(
Refs.
75,
76),
and
a
PVCP­
PIP
producing
other
viral
proteins
would
not
qualify
for
the
exemption
under
consideration
here.
In
addition,
any
negative
effects
are
expected
primarily
in
the
Page
19
of
34
transgenic
crop
itself,
as
expression
of
the
coat
protein
would
be
necessary
to
produce
the
synergistic
disease.
Furthermore,
any
negative
effects
are
expected
to
be
self­
limiting
because
any
plants
containing
a
PVCP­
PIP
that
is
prone
to
display
synergy
with
viruses
common
in
the
areas
of
planting
would
be
quickly
abandoned
once
such
effects
were
detected,
perhaps
as
early
as
the
field­
testing
stage
of
product
development.
Synergistic
interactions
can
be
evaluated
in
transgenic
plants
before
deployment
by
experimental
inoculation
with
all
of
the
viruses
likely
to
be
encountered
in
the
field
(
Ref.
65).
Developers
have
a
strong
incentive
to
undertake
such
efforts
to
ensure
the
efficacy
of
their
product
after
deployment.

4.
Field
experiments
The
experiments
referenced
in
Units
II.
B.
1­
3
above
investigated
potential
viral
interactions
in
VCP­
transgenic
plants
under
laboratory
conditions.
However,
equally
important
is
consideration
of
the
likelihood
and
potential
impact
of
viral
interactions
under
natural
field
conditions
(
Ref.
77).
Relatively
few
field
studies
have
been
conducted
to
address
the
questions
EPA
is
evaluating,
but
the
Agency
has
carefully
considered
the
available
literature.

A
six­
year
experiment
searched
for
and
failed
to
find
evidence
of
interactions
involving
viral
transgenes
in
25,000
transgenic
potato
plants
transformed
with
various
PLRV
coat
protein
constructs.
Plants
were
exposed
to
infection
by
PLRV
by
direct
inoculation,
plant­
to­
plant
spread,
or
natural
exposure.
In
field
experiments,
plants
were
also
naturally
exposed
to
the
complex
of
viruses
that
occur
in
the
region.
Both
the
greenhouse
and
field
tests
failed
to
show
any
change
in
the
type
or
severity
of
disease
symptoms,
and
all
viruses
isolated
were
previously
known
to
infect
the
plants
and
had
the
expected
transmission
characteristics
(
Ref.
78).
These
results
suggest
that
viral
interactions
leading
to
evolution
of
new
viruses
and/
or
more
severe
viral
disease
are
rare
events.

A
two­
year
experiment
with
transgenic
melon
and
squash
expressing
coat
protein
genes
of
an
aphid­
transmissible
strain
of
CMV
failed
to
find
evidence
that
either
recombination
or
heterologous
encapsidation
enabled
spread
of
an
aphid
non­
transmissible
strain
of
CMV
in
the
field
(
Ref.
79).
A
similar
experiment
used
transgenic
squash
expressing
coat
protein
genes
of
an
aphid­
transmissible
strain
of
watermelon
mosaic
virus
(
WMV).
Plants
were
mechanically
inoculated
with
an
aphid
non­
transmissible
strain
of
ZYMV,
and
subsequent
transmissions
of
the
virus
(
assumed
to
be
vectored
by
aphids)
were
assessed.
Infections
of
ZYMV
were
not
detected
in
nontransgenic
fields,
but
the
virus
infected
up
to
2%
of
plants
in
transgenic
fields.
Several
lines
of
evidence
suggested
ZYMV
infection
was
mediated
by
the
WMV
PVC­
protein
heterologously
encapsidating
the
ZYMV
viral
genome.
However,
the
virus
spread
over
short
distances,
and
transmission
at
a
low
rate
failed
to
lead
to
an
epidemic
of
ZYMV
in
fields
of
WMV­
resistant
transgenic
squash
despite
the
presence
of
optimal
conditions
for
transmission
(
Ref.
67).
These
results
support
the
contention
that
even
if
heterologous
encapsidation
involving
a
PVC­
protein
were
to
occur,
the
impact
is
likely
to
be
limited
because
each
plant
infection
requires
a
rare
event
to
occur.
Natural
processes
of
viral
infection
can
be
at
least
an
order
of
magnitude
more
efficient
and
lead
to
relatively
greater
impacts
(
Ref.
67).
Page
20
of
34
An
experiment
to
assess
the
biological
and
genetic
diversity
of
California
CMV
isolates
sampled
before
and
after
deployment
of
transgenic
melon
containing
the
CMV
coat
protein
gene
documented
only
one
CMV
isolate
that
had
significant
sequence
changes
after
infecting
the
transgenic
squash.
However,
this
isolate
did
not
result
from
recombination;
most
likely
it
reflected
a
random
colonization
on
new
host
plants
from
a
mixed
population
inoculum
(
Ref.
80).
The
only
field
experiment
to
directly
assess
the
effect
of
recombination
in
a
transgenic
plant
containing
a
PVCP­
PIP
found
no
detectable
grapevine
fanleaf
virus
(
GFLV)
recombinants
containing
the
inserted
coat
protein
sequence
over
the
course
of
a
four­
year
study
(
Ref.
81).
Test
plants
consisted
of
nontransgenic
scions
grafted
onto
transgenic
and
nontransgenic
rootstocks
that
were
exposed
over
3
years
to
GFLV
infection
at
two
sites.
Analysis
of
challenging
GFLV
isolates
revealed
no
difference
in
the
molecular
variability
among
isolates
from
190
transgenic
and
157
nontransgenic
plants,
or
from
plants
within
(
253
individuals)
or
outside
(
94
individuals)
of
the
two
test
sites.

5.
Conclusions
regarding
viral
interactions
The
information
in
Units
II.
B.
2­
4
suggests
that
heterologous
encapsidation
very
rarely
leads
to
changes
in
virus
epidemiology
that
could
have
any
large­
scale
impact
and
that
synergy
in
plants
containing
PVCP­
PIPs
also
is
unlikely
to
cause
any
widespread
environmental
harm.
Consistent
with
these
observations,
the
2004
SAP
that
"
except
perhaps
for
a
very
few
cases,
neither
heterologous
encapsidation
nor
synergy
should
be
considered
to
be
of
serious
concern"
(
Ref.
31).
The
Agency
believes
that
even
in
the
very
few
cases
mentioned
by
the
SAP,
concerns
associated
with
these
types
of
viral
interactions
are
likely
to
be
limited
in
scope
(
for
reasons
discussed
in
Units
II.
B.
2­
3)
such
that
the
determination
can
be
made
that
they
pose
low
risk
to
human
health
and
the
environment.
EPA
therefore
concludes
that
PVCP­
PIPs
present
low
risk
with
respect
to
heterologous
encapsidation
and
synergy
and
that
PVCP­
PIPs
could
be
exempted
without
further
qualification/
requirements
to
address
these
endpoints.

However,
EPA
is
not
able
to
draw
the
same
conclusions
regarding
recombination.
Based
on
the
available
evidence
(
discussed
in
Unit
II.
B.
1),
EPA
agrees
with
the
conclusions
of
the
2004
SAP
that
"[
i]
n
contrast
to
heterologous
encapsidation
and
synergy,
at
least
in
theory,
the
impact
of
recombination
could
be
much
greater,
since
there
is
now
abundant
bioinformatic
evidence
that
recombination
has
indeed,
as
long
suspected,
played
a
key
role
in
the
emergence
of
new
viruses
over
evolutionary
time"
(
Ref.
22).

The
few
field
evaluations
conducted
(
discussed
in
Unit
II.
B.
4)
suggest
that
adverse
environmental
effects
due
to
recombination
in
transgenic
plants
containing
PVCP­
PIPs
are
unlikely
to
occur
at
least
on
a
small
scale
over
a
short
time
period.
However,
large
acreages
of
VCP­
transgenic
plants
grown
over
many
years
may
provide
increased
opportunity
for
rare
events
to
occur
that
are
unlikely
to
be
detected
in
experimental
studies
(
Ref.
75).
In
addition,
none
of
the
experimental
systems
described
above
would
be
predicted
to
involve
viruses
that
would
otherwise
not
be
expected
to
interact
in
a
mixed
infection
found
in
nature.
Given
the
limited
amount
of
field
data
available,
particularly
data
relevant
to
the
circumstances
EPA
has
identified
as
being
of
highest
concern
(
i.
e.,
those
that
could
lead
to
novel
interactions),
EPA
would
limit
an
exemption
to
those
PVCP­
PIPs
for
which
novel
viral
interactions
are
unlikely
to
occur.
When
Page
21
of
34
EPA
consulted
the
2004
SAP
about
situations
in
which
novel
viral
interactions
might
be
a
concern,
the
Panel
agreed
"
that
recombination
is
a
concern
when
the
two
contributing
viruses
have
not
previously
had
a
chance
to
recombine"
(
Ref.
22).

In
addition
to
considering
the
potential
for
novel
viral
interactions
to
occur,
EPA
also
considered
whether
transgenic
plants
containing
PVCP­
PIPs
might
present
opportunity
for
a
generally
increased
frequency
of
viral
interactions
given
that
a
transgene
expressed
with
a
constitutive
promoter
could
be
present
in
all
cells
of
the
plant
at
all
times.
In
natural
mixed
infections
viruses
must
simultaneously
replicate
in
the
same
cellular
compartment
for
their
RNA
to
be
able
to
interact.
However,
when
a
virus
invades
a
cell,
it
often
replicates
and
then
moves
to
other
cells
within
the
plant.
The
RNA
remaining
in
the
initially
infected
cell
becomes
encapsidated
and
may
no
longer
be
available
for
interactions
with
another
invading
virus
(
Ref.
82).
When
EPA
presented
this
issue
to
the
2004
SAP,
the
panel
responded
that
"
no
increase
in
heterologous
encapsidation
should
be
anticipated
in
PVCP­
PIP
plants"
and
"
the
important
questions
are
not
the
relative
likelihood
for
recombination
to
occur,
but
rather
whether
recombinants
in
transgenic
plants
are
different
from
those
in
non­
transgenic
plants
and
whether
they
are
viable"
(
Ref.
22).
Thus,
EPA's
current
approach
focuses
on
situations
in
which
novel
recombination
events
could
occur
due
to
the
presence
of
a
PVCP­
PIP.

6.
Categorical
exemption
criterion
In
developing
the
categorical
exemption
for
a
subset
of
PVCP­
PIPs
in
which
a
developer
could
self­
determine
whether
the
criteria
were
met,
EPA
seeks
to
identify
those
situations
that
clearly
pose
low
risk
with
respect
to
viral
interactions,
i.
e.,
those
situations
in
which
recombination
in
a
transgenic
plant
would
involve
segments
of
viruses
that
already
have
the
opportunity
to
recombine
in
a
natural,
mixed
infection.

A
PVCP­
PIP
would
meet
criterion
(
b)
under
paragraph
(
1)
if
the
viral
pathotype
used
to
create
the
PVCP­
PIP
has
naturally
infected
plants
in
the
United
States,
its
possessions,
or
territories
and
naturally
infects
plants
of
the
same
species
as
that
containing
the
PVCP­
PIP.
The
developer
may
make
this
determination.
If
the
viral
pathotype
was
isolated
from
a
plant
in
the
United
States,
its
possessions,
or
territories
that
is
of
the
same
species
as
the
plant
containing
the
PVCP­
PIP
and
was
not
subsequently
modified,
the
PVCP­
PIP
meets
this
criterion.
No
further
data
or
information
would
be
needed
to
evaluate
risks
associated
with
recombination
when
paragraph
(
1)
of
criterion
(
b)
is
satisfied,
and
therefore
no
Agency
review
would
be
necessary.

The
Agency
asked
the
FIFRA
SAP
during
the
October
2004
meeting
to
what
extent
PVCP­
PIPs
in
plants
might
present
a
potential
concern
should
interactions
with
infecting
viruses
occur.
The
Panel
expressed
concern
only
"
about
certain
limited
situations"
and
stated
that
"
in
most
cases
there
is
little
a
priori
reason
to
believe
that
recombinants
between
viruses
and
transgenes
will
be
more
of
a
problem
than
recombinants
between
two
viruses
infecting
the
same
plant,
unless
transgenes
are
derived
from
severe
or
exotic
isolates.
The
general
recommendation
to
use
mild,
endemic
isolates
as
the
source
of
the
transgene
(
e.
g.
Hammond
et
al.
1999)
should
minimize
any
potential
for
creation
of
novel
isolates
that
would
not
equally
easily
arise
in
natural
mixed
infections"
(
Ref.
27).
Page
22
of
34
The
Agency's
approach
in
paragraph
(
b)(
1)
is
consistent
with
the
SAP's
recommendations
because
it
excludes
use
of
exotic
virus
isolates
as
the
source
of
the
PVCP­
PIP
transgene.
When
severe
isolates
are
used,
the
PVCP­
PIP
may
only
meet
paragraph
(
b)(
1)
if
they
were
present
in
the
natural
system
and
therefore
should
pose
no
novel
interactions.
Paragraph
(
b)(
1)
is
also
intended
to
exclude
from
exemption
PVCP­
PIPs
that
are
inserted
into
a
plant
species
that
is
not
naturally
infected
by
the
virus
used
to
create
the
PVCP­
PIP.
Most
PVCP­
PIPs
are
created
from
viruses
that
do
naturally
infect
the
plant
species
into
which
they
are
inserted
because
greater
efficacy
is
achieved
when
a
virus
most
similar
to
the
target
virus
is
used
to
isolate
the
sequence
used
in
the
PVCP­
PIP.
However,
virus­
resistant
transgenic
plants
have
been
created
where
this
is
not
the
case
(
Ref.
83).
In
these
situations,
a
virus
is
introduced
into
a
system
where
it
does
not
naturally
occur,
and
viruses
with
which
it
does
not
otherwise
interact
may
be
present
in
that
system.
The
Agency
cannot
a
priori
determine
that
such
interactions
are
safe
because
there
is
no
experience
upon
which
to
base
such
a
finding.

EPA
means
by
the
term
"
naturally
infect"
to
infect
by
transmission
to
a
plant
through
direct
plant­
to­
plant
contact
(
e.
g.,
pollen
or
seed),
an
inanimate
object
(
e.
g.,
farm
machinery),
or
vector
(
e.
g.,
arthropod,
nematode,
or
fungus).
It
does
not
include
infection
by
transmission
that
occurs
only
through
intentional
human
intervention.
The
Agency
wants
specifically
to
exclude
transmission
that
occurs
only
through
intentional
human
intervention,
e.
g.,
manual
infection
in
a
laboratory
or
greenhouse
setting,
because
such
transmission
would
have
little
relevance
to
normal
human
dietary
exposure.
EPA
intends
to
include
viruses
that
are
likely
to
have
been
part
of
the
human
diet
due
to
their
ability
to
spread
without
intentional
human
intervention.
EPA
recognizes
that
humans
may
play
an
inadvertent
role
in
infection
(
e.
g.,
by
transmitting
the
virus
on
farm
machinery).
Such
unintentional
(
and
often
unavoidable)
transmission
can
be
an
important
means
of
virus
transmission,
and
this
mode
of
transmission
would
be
included
under
"
naturally
infects."

EPA
uses
the
term
"
viral
pathotype"
rather
than
the
more
generic
term
"
virus"
in
response
to
the
FIFRA
SAP
comment
in
October
2004
that
"[
n]
ot
all
isolates
of
a
virus
infect
and
cause
disease
in
all
plant
genotypes
and,
as
a
consequence,
the
unqualified
use
of
the
term
"
virus"
when
setting
a
condition
for
applicants
to
the
Agency
[
is]
not
adequate
in
this
context.
It
is
therefore
appropriate
in
the
context
of
biosafety
as
well
as
virus
epidemiology
to
recognize
the
value
of
defining
specific
viral
pathotypes
or
host
range
variants."

7.
Exemption
criterion
conditional
on
Agency
determination
The
Agency
recognizes
that
many
PVCP­
PIPs
may
pose
low
risk
with
respect
to
recombination
even
though
they
fail
to
satisfy
paragraph
(
1)
of
criterion
(
b).
Therefore,
EPA
is
considering
an
approach
under
which
PVCP­
PIPs
that
fail
to
meet
paragraph
(
b)(
1)
could
still
meet
criterion
(
b),
subject
to
an
Agency
review
to
determine
whether
they
meet
a
different
set
of
conditions
related
to
this
issue.
Under
this
approach,
a
PVCP­
PIP
would
meet
criterion
(
b)
under
paragraph
(
2)
if
the
Agency
determines
either
that
(
i)
the
properties
of
the
viral
pathotype
that
are
determined
by
the
coat
protein
gene
used
to
create
the
PVCP­
PIP
are
substantially
similar
to
the
properties
of
a
viral
pathotype
that
naturally
infects
plants
in
the
United
States,
its
possessions,
or
territories,
and
the
viral
pathotype
used
to
create
the
PVCP­
PIP
naturally
infects
plants
of
the
same
species
Page
23
of
34
as
that
containing
the
PVCP­
PIP,
or
(
ii)
viruses
that
naturally
infect
the
plant
containing
the
PVCP­
PIP
are
unlikely
to
acquire
the
coat
protein
sequence
through
recombination
and
produce
a
viable
virus
with
significantly
different
properties
than
either
parent
virus.

With
an
Agency
determination
under
paragraph
(
2)
of
criterion
(
b),
EPA
would
create
a
criterion
that
would
encompass
a
larger
set
of
those
PVCP­
PIPs
that
pose
low
risk
with
respect
to
viral
interactions
than
are
covered
under
paragraph
(
1).
However,
even
with
an
Agency
determination,
EPA
must
still
define
a
criterion
with
sufficient
precision
such
that
the
public
would
be
able
to
understand
what
products
would
qualify
and
evaluate
whether
they
would
meet
the
standard
for
a
FIFRA
exemption.
EPA
seeks
to
define
a
characteristic
of
a
PVCP­
PIP
that
would
indicate
the
viral
interactions
that
could
occur
would
be
no
different
than
would
occur
in
a
natural,
mixed
infection
found
in
nature.
PVCP­
PIPs
meeting
the
condition
in
paragraph
(
b)(
2)(
i)
would
pose
low
probability
of
risk
with
respect
to
viral
interactions
because
the
viral
interactions
that
EPA
determines
could
occur
in
that
plant
would
not
be
substantively
different
than
what
could
occur
in
a
natural
mixed
infection
in
the
United
States
involving
that
virus.

EPA
believes
that
an
Agency
review
would
be
needed
to
make
this
determination
because
it
is
more
complicated
than
the
relatively
straightforward
determination
of
whether
paragraph
(
b)(
1)
is
met
based
on
knowledge
of
the
plant
from
which
the
viral
pathotype
used
to
create
the
PVCPPIP
was
isolated.
If
the
viral
pathotype
was
isolated
from
outside
the
United
States,
its
possessions,
or
territories
or
the
coat
protein
sequence
was
modified
after
isolation,
a
determination
of
substantial
similarity
would
be
based
on
consideration
of
similarity
in
the
coat
protein
gene
sequence
of
the
pathotype
used
and
of
pathotypes
found
in
the
United
States,
its
possessions,
or
territories.
If
information
is
available,
EPA
would
evaluate
the
extent
to
which
any
significant
sequence
differences
are
likely
to
influence
phenotypic
properties
of
the
virus.
EPA's
review
would
consider
data
from
a
number
of
different
sources
including
virus
coat
protein
sequence
data
from
public
repositories
and
developer­
generated
data
on
the
natural
range
of
variation
of
coat
protein
genes
for
particular
viral
pathotypes.

EPA
believes
that
the
risks
associated
with
recombination
arise
when
the
potential
recombinants
would
be
unlike
those
expected
in
a
natural
mixed
infection
found
in
nature.
The
conditions
in
paragraphs
(
2)(
a)
and
(
2)(
b)(
i)
address
these
concerns
by
ensuring
that
no
novel
viral
interactions
occur.
Under
paragraph
(
2)(
b)(
ii),
a
PVCP­
PIP
could
qualify
for
exemption
even
when
novel
viral
interactions
could
occur,
providing
steps
were
taken
to
significantly
reduce
the
likelihood
that
an
infecting
virus
would
not
acquire
a
portion
of
the
PVCP­
PIP
coat
protein
sequence
through
recombination
and
produce
a
viable
virus
with
significantly
different
properties
than
either
parent
virus.
For
example,
the
following
methods
for
reducing
the
frequency
of
recombination
might
be
relevant
in
evaluating
a
PVCP­
PIP
under
paragraph
2(
b)(
ii):
if
the
PVCP­
PIP
confers
virus
resistance
through
post­
transcriptional
gene
silencing
thereby
greatly
reducing
the
amount
of
RNA
available
for
recombination;
if
the
PVCP­
PIP
construct
is
designed
to
reduce
the
frequency
of
recombination
(
e.
g.,
Refs.
84,
85,
86,
87,
88);
or
if
the
inserted
coat
protein
sequence
is
only
a
relatively
small
portion
of
the
naturally
occurring
sequence
suggesting
that
viruses
acquiring
the
region
are
unlikely
to
acquire
a
novel
phenotype.
EPA
recognizes
the
comments
of
the
2004
SAP
that
"
methods
for
minimizing
recombination
are
only
partially
effective.
For
this
reason,
the
question
remains
whether
novel
recombinants
would
be
created
in
transgenic
plants,
and
simply
reducing
the
frequency
of
these
events
is
not
an
answer
to
the
question"
(
Ref.
31).
However,
a
combination
of
two
or
more
methods,
or
even
perhaps
a
single
Page
24
of
34
method
in
some
cases,
could
be
employed
to
reduce
the
expected
frequency
of
recombination
to
a
level
that
would
support
a
determination
that
a
PVCP­
PIP
would
pose
low
risk
with
respect
to
viral
interactions.
Given
that
there
is
no
universally
applicable
method
for
reducing
recombination
frequency,
EPA
believes
an
Agency
review
is
needed
to
make
this
determination.

8.
Other
approaches
EPA's
proposed
exemption
in
1994
did
not
contain
any
criteria
related
to
viral
interactions.
However,
since
that
time,
many
additional
scientific
papers
and
reviews
have
been
published
on
this
topic.
Most
affirm
the
general
safety
of
PVCP­
PIPs
with
respect
to
viral
interactions,
but
some
call
into
question
assumptions
of
how
generically
this
conclusion
holds
across
all
PVCPPIPs
For
example,
although
the
2000
NRC
report
stated
that,
"[
m]
ost
virus­
derived
resistance
genes
are
unlikely
to
present
unusual
or
unmanageable
problems
that
differ
from
those
associated
with
traditional
breeding
for
virus
resistance,"
the
NRC's
report
also
suggested
that
their
conclusions
were
based
on
the
assumption
that
certain
risk
management
strategies
should
or
would
be
implemented,
e.
g.,
elimination
of
specific
sequences
to
limit
the
potential
for
recombination
(
Ref.
26).
EPA
believes
the
Agency's
1994
conclusion
of
low
probability
of
risk
still
holds
for
most
PVCP­
PIPs,
but
in
order
to
grant
an
exemption
under
FIFRA,
EPA
must
be
able
to
make
such
a
finding
for
all
PVCP­
PIPs
covered
by
the
exemption
and
must
make
its
safety
determination
in
the
absence
of
any
regulatory
oversight
under
FIFRA
that
could
ensure
mitigation
measures,
such
as
those
discussed
in
the
NRC
report,
were
employed.
Therefore,
it
appears
prudent
at
this
time
to
limit
any
exemption
with
a
criterion
that
restricts
the
potential
for
novel
recombination
events,
as
these
have
been
identified
as
the
rare
situation
in
which
viral
interactions
in
plants
containing
a
PVCP­
PIP
may
lead
to
adverse
environmental
effects.

EPA
presented
a
set
of
conditions
to
the
2004
SAP
and
asked
whether
they
would
significantly
reduce
either
the
novelty
or
frequency
of
viral
interactions
in
plants
containing
PVCP­
PIPs
such
that
the
Agency
would
not
need
to
regulate
the
PVCP­
PIP
(
Ref.
22).
The
first
condition
was
that
"
the
genetic
material
of
the
PVCP­
PIP
is
translated
and/
or
transcribed
in
the
same
cells,
tissues,
and
developmental
stages
naturally
infected
by
every
virus
from
which
any
segment
of
a
coat
protein
gene
used
in
the
PVCP­
PIP
was
derived."
EPA
considered
such
a
condition
because
with
a
PVCP­
PIP,
plants
may
express
viral
genes
in
cells
and/
or
tissues
that
the
virus
does
not
normally
infect.
Genetic
promoters
currently
used
in
most
transgenic
plants
cause
constitutive
expression
of
transgenes
at
developmental
stages
that
might
otherwise
be
unaffected
by
viral
infection
and
often
in
tissues
that
the
virus
does
not
normally
infect
(
Ref.
82).
For
example,
luteoviruses
are
normally
expressed
only
in
phloem
tissue,
but
the
cauliflower
mosaic
virus
(
CaMV)
promoter
drives
expression
of
luteoviral
coat
protein
in
all
plant
cells.
Some
evidence
suggests
that
in
natural
infections
different
viruses
have
different
temporal
or
spatial
expression
patterns
that
would
limit
their
interactions
(
Refs.
34,
89,
90).
However,
the
2004
SAP
concluded
that
such
a
condition
would
be
of
limited
utility
because
"[
m]
ost
plant
viruses
are
present
in
a
wide
range
of
cell
and
tissue
types"
(
Ref.
22).

The
second
condition
presented
to
the
2004
SAP
was
that
"
the
genetic
material
of
the
PVCP­
PIP
contains
coat
protein
genes
or
segments
of
coat
protein
genes
from
viruses
established
throughout
the
regions
where
the
crop
is
planted
in
the
United
States
and
that
naturally
infect
the
Page
25
of
34
crop
into
which
the
genes
have
been
inserted."
EPA
considered
the
first
part
of
this
criterion
because
plants
may
be
engineered
with
PVCP
genes
from
an
exotic
strain
of
a
virus
that
may
be
more
virulent
or
have
other
properties
different
from
endemic
isolates.
Interactions
with
such
virus
sequences
could
potentially
change
the
epidemiology
or
pathogenicity
of
viruses
infecting
plants
containing
these
sequences.
The
2004
SAP
concurred
that
"
using
such
an
exotic
PVCP
gene
would
open
possibilities
for
novel
interactions."
EPA's
current
criterion
(
b)
thus
would
exclude
exotic
coat
protein
genes
from
exemption
unless
steps
have
been
taken
to
reduce
the
frequency
of
recombination.
EPA
considered
the
second
part
of
this
criterion
because
in
heterologous
resistance,
a
plant
may
be
resistant
to
infection
by
a
particular
virus
in
spite
of
having
the
coat
protein
gene
of
another
virus
incorporated
into
its
genome.
For
example,
PVCP
genes
from
LMV
were
used
to
provide
resistance
to
PVY
in
tobacco
which
is
not
infected
by
LMV
(
Ref.
91).
In
such
plants,
LMV
might
have
a
new
opportunity
to
interact
with
viruses
that
infect
tobacco.
The
2004
Panel
concluded
that
"[
w]
hat
is
described
here
is
most
often
implemented:
in
designing
a
PVCP
transgene,
better
efficacy
is
often
observed
if
it
is
similar
as
possible
to
the
target
virus."
Nevertheless,
EPA
believes
that
such
a
condition
is
appropriate
given
that
PVCP­
PIPs
may
be
developed
using
heterologous
resistance.
EPA's
current
approach
thus
excludes
from
exemption
PVCP­
PIPs
used
in
plants
that
the
virus
used
to
create
the
PVCPPIP
does
not
naturally
infect
unless
steps
have
been
taken
to
reduce
the
frequency
of
recombination.

The
third
condition
presented
to
the
2004
SAP
was
that
"
the
PVCP­
PIP
has
been
modified
by
a
method
scientifically
documented
to
minimize
recombination
(
e.
g.,
deletion
of
the
3'
untranslated
region
of
the
coat
protein
gene).
As
discussed
above,
EPA
recognizes
the
comments
of
the
2004
SAP
that
"
methods
for
minimizing
recombination
are
only
partially
effective.
For
this
reason,
the
question
remains
whether
novel
recombinants
would
be
created
in
transgenic
plants,
and
simply
reducing
the
frequency
of
these
events
is
not
an
answer
to
the
question"
(
Ref.
31).
However,
EPA
believes
that
a
combination
of
two
or
more
methods,
or
even
perhaps
a
single
method
in
some
cases,
could
be
employed
such
that
the
expected
frequency
of
recombination
would
be
reduced
to
a
level
that
would
support
determination
that
a
PVCP­
PIP
would
pose
low
risk
with
respect
to
viral
interactions.
EPA
intends
that
paragraph
(
b)(
2)(
ii)
would
allow
the
Agency
to
make
this
determination
after
review.

The
fourth
condition
presented
to
the
2004
SAP
was
that
"
the
PVCP­
PIP
has
been
modified
by
a
method
scientifically
documented
to
minimize
heterologous
encapsidation
or
vector
transmission,
or
there
is
minimal
potential
for
heterologous
encapsidation
because
no
protein
from
the
introduced
PVCP­
PIP
is
produced
in
the
transgenic
plant
or
the
virus
does
not
participate
in
heterologous
encapsidation
in
nature."
The
2004
SAP
concluded
that
"[
t]
his
method
can
 
be
considered
seriously
if
deemed
necessary."
However,
the
Agency
concluded
(
as
discussed
above
in
Unit
II.
B.
2)
that
such
methods
are
not
necessary
because
heterologous
encapsidation
is
so
rarely
likely
to
be
of
any
significant
ecological
concern.

C.
Production
of
proteins.
Page
26
of
34
PVCP­
PIPs
contain
plant
virus
coat
protein
sequences
that
may
lead
to
protein
production
in
the
plant
in
which
the
sequences
are
inserted.
EPA
thus
must
consider
the
safety
of
any
potentially
expressed
proteins
when
proposing
criteria
to
evaluate
PVCP­
PIPs
for
possible
exemption.

EPA
has
to
consider
human
dietary,
nontarget,
and
occupational
exposure
risks
in
evaluating
the
safety
of
PVC­
proteins.
Readers
are
referred
to
EPA's
assessment
of
human
dietary
exposure
risks
as
well
as
other
non­
occupational
exposure
in
Attachment
II:
Draft
Approach
to
Exempting
Certain
PVC­
Proteins
from
the
Requirement
of
a
Tolerance
under
FFDCA.
Many,
if
not
all,
of
the
considerations
used
to
evaluate
the
potential
for
novel
exposures
in
nontargets
can
be
directly
extrapolated
from
the
discussion
on
the
history
of
safe
exposure
to
naturally
occurring
plant
virus
coat
proteins
found
in
Attachment
II.

EPA
consulted
the
2004
SAP
about
possible
nontarget
effects
of
PVC­
proteins.
The
panel
confirmed
that
PVC­
proteins
within
the
range
of
natural
variation
of
the
virus
would
not
be
anticipated
to
present
risks
to
nontarget
organisms,
concluding
that,
"[
l]
ethal
effects
in
animal
life
after
feeding
on
PVCP­
PIP
plants
are
highly
unlikely
because
plant
viruses
are
not
known
to
have
deleterious
effects
on
animal
life.
Additionally,
animals
routinely
feed
on
non­
engineered
virus­
infected
plants
and
do
not
die .
[
S]
ublethal
effects
are
not
expected
to
be
manifested
in
animal
life,
again
because
wildlife
and
insects
regularly
feed
on
non­
engineered
virus­
infected
plants
with
no
apparent
sublethal
damage"
(
Ref.
31).

1.
Categorical
exemption
criterion
In
developing
the
categorical
exemption
for
a
subset
of
PVCP­
PIPs
in
which
a
developer
could
self­
determine
whether
the
criteria
were
met,
EPA
seeks
to
identify
those
situations
that
clearly
pose
low
risk
with
respect
to
protein
production
because
the
proteins
produced
would
be
within
the
range
of
natural
variation.
EPA
wants
to
ensure
that
a
long
history
of
safe
human
and
nontarget
exposure
has
occurred
for
any
PVC­
protein
produced
from
a
PVCP­
PIP
that
would
qualify
for
an
exemption.
A
PVCP­
PIP
would
meet
criterion
(
c)
under
paragraph
(
1)
if
a
product
developer
self­
determines
that
the
genetic
material
encodes
only
a
single
contiguous
portion
of
each
unmodified
viral
coat
protein.
This
would
include
multiple
proteins
expressed
from
a
single
PVCP­
PIP
construct,
but
not
chimeric
proteins.

The
requirement
that
the
genetic
material
encode
"
only
a
single
contiguous
portion
of
each
unmodified
viral
coat
protein,"
would
exclude
residues
of
modified
PVC­
proteins.
For
example,
PVC­
proteins
containing
insertions,
internal
deletions,
or
amino
acid
substitutions
would
be
excluded,
as
would
be
chimeric
proteins
that
are
encoded
by
a
sequence
constructed
from
portions
of
two
or
more
different
plant
virus
coat
protein
genes.
EPA
is
considering
whether
to
exclude
such
PVC­
proteins
from
the
self­
determining
part
of
the
exemption
in
response
to
the
advice
of
the
FIFRA
SAP
in
October
2004
that,
"[
t]
here
was
general
agreement
that
an
allergenicity
assessment
would
be
appropriate
for
insertions
or
deletions,
except
perhaps
for
terminal
deletions
that
do
not
affect
overall
protein
structure."
However,
insufficient
information
exists
at
this
time
to
allow
EPA
to
describe
a
priori
a
criterion
that
would
ensure
all
PVCproteins
with
such
modifications
fall
within
the
base
of
experience
supporting
an
exemption.
At
this
point
in
time,
it
is
not
possible
to
make
a
categorical
risk
assessment
finding
that
insertions
Page
27
of
34
or
internal
deletions
are
unlikely
to
change
the
characteristics
of
any
protein
produced.
Thus,
EPA
would
follow
the
prudent
course
for
paragraph
(
1)
of
criterion
(
c)
and
allow
neither
modification.

EPA
believes
the
phrase
"
a
single
contiguous
portion"
conveys
the
concept
that
segments
of
PVC­
proteins
that
are
identical
to
an
unmodified
coat
protein
would
also
be
exempt.
EPA
believes
the
exemption
of
segments
is
supported
by
the
experience
base
EPA
is
relying
on
to
develop
an
exemption
because
it
is
probable
that
segments
of
coat
proteins
exist
in
nature
due
to
processes
such
as
incomplete
translation
of
transcripts
and
partial
degradation
of
proteins.
Incomplete
translation
may
occur
due
to
routine
replication
errors
causing
a
ribosome
to
dissociate
from
an
RNA
transcript
or
if
mutation
introduces
a
premature
stop
codon,
i.
e.,
a
nonsense
mutation.
Truncated
plant
virus
coat
proteins
are
indeed
known
to
occur
in
nature
(
Ref.
92).
Thus,
PVC­
proteins
that
are
truncated
forms
of
naturally
occurring
plant
virus
coat
proteins
would
not
significantly
increase
the
likelihood
of
exposure
to
a
toxic
or
allergenic
protein
since
humans
are
currently
exposed
to
them
in
the
diet
along
with
complete
plant
virus
coat
proteins.

The
Agency
is
considering
whether
also
to
include
in
the
categorical
exemption,
i.
e.,
without
Agency
review,
amino
acid
sequences
containing
terminal
deletion(
s)
and/
or
an
additional
Nterminal
methionine
residue.
The
AUG
codon
for
methionine
initiates
translation
in
eukaryotes
(
Ref.
93).
Among
certain
viruses
such
as
the
Potyviridae,
the
coat
protein
is
produced
as
part
of
a
polyprotein,
so
the
coding
region
for
the
coat
protein
is
excised
from
the
genetic
material
encoding
the
polyprotein
to
create
a
PVCP­
PIP
and
thus
normally
lacks
a
start
codon.
Insertion
of
an
AUG
codon
allows
for
PVC­
protein
expression,
which
may
be
needed
to
confer
virus
resistance.
EPA
believes
the
addition
of
a
single,
N­
terminal
methionine
residue
would
be
unlikely
to
affect
a
PVC­
protein's
toxicity
or
allergenicity
relative
to
a
naturally
occurring
plant
virus
coat
protein.

If
the
genetic
material
encodes
only
a
single
contiguous
portion
of
an
unmodified
viral
coat
protein,
no
novel
exposures
to
humans
or
nontarget
organisms
are
likely
to
occur
because
these
PVC­
proteins
are
identical
to
plant
viral
coat
proteins
that
are
widespread
in
the
plant
kingdom,
as
most
plants
are
infected
by
at
least
one
virus.
EPA
is
relying
on
this
history
of
safe
exposure
when
considering
whether
to
exempt
certain
PVCP­
PIPs
from
regulation
under
FIFRA.
The
Agency
believes
that
when
such
a
PVCP­
PIP
is
used,
the
PVCP­
PIP
would
pose
low
probability
of
risk
with
respect
to
protein
production.
EPA
believes
that
no
further
data
or
information
would
be
needed
to
evaluate
this
issue
when
paragraph
(
1)
of
criterion
(
c)
is
satisfied,
and
therefore
no
Agency
review
would
be
necessary.

2.
Exemption
criterion
conditional
on
Agency
determination
The
Agency
acknowledges
that
many
PVCP­
PIPs
may
pose
low
risk
with
respect
to
concerns
associated
with
protein
production
even
though
they
fail
to
satisfy
paragraph
(
1)
of
criterion
(
c).
EPA
would
review
such
PVCP­
PIPs
that
fail
to
meet
paragraph
(
c)(
1)
under
slightly
different
factors
that
the
Agency
believes
also
ensure
that
qualifying
PVCP­
PIPs
pose
low
risk
with
respect
to
concerns
associated
with
protein
production.
Therefore,
a
PVCP­
PIP
would
also
meet
criterion
(
c)
under
paragraph
(
2)
if
the
Agency
determines
that
the
genetic
material
(
i)
encodes
a
Page
28
of
34
protein
that
is
minimally
modified
from
a
coat
protein
from
a
virus
that
naturally
infects
plants,
or
(
ii)
produces
no
protein.

The
Agency's
rationale
for
concluding
no
novel
exposures
to
proteins
are
associated
with
PVCPPIPs
would
cover
only
those
PVC­
proteins
that
are
not
significantly
different
from
naturally
occurring
plant
viral
coat
proteins.
For
PVCP­
PIPs
that
contain
modified
genetic
material
encoding
a
PVC­
protein
that
is
not
identical
or
not
minimally
modified
from
a
naturally
occurring
plant
virus
coat
protein,
the
base
of
experience
upon
which
EPA
relies
to
support
exempting
such
proteins
would
not
apply.
Therefore,
were
such
a
PVC­
protein
to
be
produced
from
the
PVCP­
PIP,
EPA
would
not
be
able
to
make
the
determination
that
the
PVCP­
PIP
poses
a
low
probability
of
risk
to
humans
and
the
environment
and
will
not
generally
cause
unreasonable
adverse
effects
on
the
environment
even
in
the
absence
of
regulatory
oversight
under
FIFRA.
For
discussion
of
the
concept
of
"
minimally
modified"
see
Unit
II.
D.
2
of
Attachment
II:
Draft
Approach
to
Exempting
Certain
PVC­
Proteins
from
the
Requirement
of
a
Tolerance
under
FFDCA.

EPA
developed
paragraph
(
2)
of
criterion
(
c)
because
the
Agency
recognizes
that
PVCP­
PIP
developers
may
wish
to
modify
PVCP­
PIP
constructs
to
achieve
certain
product
development
goals
such
as
greater
efficacy,
and
such
modifications
might
result
in
changes
to
the
protein(
s)
produced.
Many
modifications
to
the
genetic
material
may
be
so
minor
that
they
are
unlikely
to
cause
changes
to
the
protein
that
would
be
significant
from
a
human
or
nontarget
organism
perspective.
Under
paragraph
(
c)(
2)
EPA
may
consider
such
insertions
or
internal
deletions
on
a
case­
by­
case
basis.
Many
of
the
modifications
are
likely
to
produce
proteins
that
fall
within
the
range
of
natural
variation
of
the
virus.
However,
it
is
not
currently
possible
clearly
to
define
the
range
of
variation
of
viruses
in
general
or
even
of
any
particular
virus
as
discussed
in
Unit
II.
C
of
Attachment
II.
Therefore,
paragraph
(
2)(
i)
of
criterion
(
c)
requires
an
Agency
review
to
determine
qualification.

PVCP­
PIPs
are
known
to
have
at
least
two
mechanisms
to
confer
virus
resistance.
Resistance
may
be
either
protein­
mediated,
in
which
the
level
of
resistance
is
correlated
with
the
level
of
protein
expression,
or
it
may
be
RNA­
mediated,
in
which
the
level
of
resistance
is
not
correlated
with
the
level
of
protein
expression.
In
the
case
of
RNA­
mediated
resistance,
little
to
no
PVCprotein
may
be
produced
from
the
PVCP­
PIP.
In
such
cases,
little
to
no
risk
due
to
protein
production
would
be
associated
with
the
PVCP­
PIP.
However,
the
Agency
believes
that
it
would
not
be
possible
at
this
time
to
describe
a
priori
conditions
that
must
be
satisfied
to
ensure
that
no
protein
is
produced
by
the
PVCP­
PIP.
Therefore,
paragraph
(
2)(
ii)
of
criterion
(
c)
requires
an
Agency
review
to
determine
qualification.

3.
Other
approaches
The
approach
EPA
is
currently
considering
is
consistent
with
what
EPA
has
always
intended.
EPA
has
never
intended
that
any
proposed
exemption
for
PVCP­
PIPs
would
cover
those
that
produce
proteins
significantly
different
from
those
that
occur
naturally
(
November
23,
1994,
59
FR
at
60539;
see
in
particular
July
19,
2001,
66
FR
37865).
EPA's
approach
discussed
here
relies
on
the
known
history
of
safe
exposure
to
coat
proteins
of
naturally
occurring
plant
viruses.
Page
29
of
34
However,
this
rationale
would
only
cover
those
PVC­
proteins
that
are
not
significantly
different
from
naturally
occurring
plant
viral
coat
proteins.
For
modified
PVCP­
PIPs,
the
base
of
experience
upon
which
EPA
relies
for
support
of
an
exemption
might
not
be
relevant.
In
some
cases,
EPA
might
not
be
able
to
make
the
determination
that
the
PVCP­
PIP
poses
a
low
probability
of
risk
to
humans
and
the
environment
and
will
not
generally
cause
unreasonable
adverse
effects
on
the
environment
even
in
the
absence
of
regulatory
oversight
under
FIFRA.

D.
Other
Definitions
Under
this
exemption
approach,
a
plant­
incorporated
protectant
based
on
a
plant
virus
coat
protein
gene
(
PVCP­
PIP)
would
be
defined
to
mean
a
plant­
incorporated
protectant
based
on
one
or
more
genes
that
encode
a
coat
protein
of
a
virus
that
naturally
infects
plants.
This
includes
PVCP­
PIPs
that
produce
no
protein.
A
PVCP­
PIP
may
contain
multiple
plant
virus
coat
protein
genes,
or
segments
thereof,
translated
as
individual
proteins.
A
PVCP­
PIP
may
also
contain
multiple
plant
virus
coat
protein
genes,
or
segments
thereof,
translated
as
a
single,
chimeric
protein.
In
this
context,
the
word
"
segment"
has
the
commonly
accepted
meaning
(
Ref.
94),
i.
e.,
a
"
part
cut
off
from
the
other
parts"
of
the
whole
coat
protein.

The
definition
of
a
PVCP­
PIP
would
contain
the
phrase
"
naturally
infects
plants."
Including
this
phrase
in
the
definition
would
specifically
limit
an
exemption
by
requiring
that
the
virus
coat
protein
gene
upon
which
the
PVCP­
PIP
is
based
come
exclusively
from
a
plant
virus.
This
limitation
is
intended
to
exclude
from
the
definition
any
coat
proteins
of
plant
viruses
that
have
been
modified
with
sequences
from
animal
or
human
viruses.
EPA
includes
this
concept
in
response
to
comment
received
from
the
public
on
earlier
documents
pertaining
to
PVCP­
PIPs.

III.
References
1.
Bartsch
D,
Schmidt
M,
Pohl­
Orf
M,
Haag
C,
Schuphan
I.
Competitiveness
of
transgenic
sugar
beet
resistant
to
beet
necrotic
yellow
vein
virus
and
potential
impact
on
wild
beet
populations.
Molecular
Ecology
1996;
5:
199.
2.
National
Research
Council.
Field
Testing
Genetically
Modified
Organisms.
Washington,
DC:
National
Academy
Press,
1989.
3.
National
Research
Council.
Environmental
Effects
of
Transgenic
Plants:
The
Scope
and
Adequacy
of
Regulation.
Washington,
D.
C.:
National
Academy
Press,
2002.
4.
Keeler
KH.
Can
genetically
engineered
crops
become
weeds?
Bio/
Technology
1989;
7:
1134­
9.
5.
Levin
DA.
The
origin
of
isolating
mechanisms
in
flowering
plants.
Evolutionary
Biology
1978;
11:
185­
317.
6.
Ellstrand
NC,
Prentice
HC,
Hancock
JF.
Gene
flow
and
introgression
from
domesticated
plants
into
their
wild
relatives.
Annual
Review
of
Ecology
and
Systematics
1999;
30:
539­
63.
7.
Ellstrand
N.
Dangerous
Liaisons.
Baltimore:
Johns
Hopkins
University
Press,
2003.
8.
MacClement
WD,
Richards
MG.
Virus
in
wild
plants.
Canadian
Journal
of
Botany
1956;
34:
793­
9.
9.
Barnett
OW,
Gibson
PB.
Identification
and
prevalence
of
white
clover
viruses
and
the
resistance
of
Trifolium
species
to
these
viruses.
Crop
Science
1975;
15:
32­
7.
10.
Hammond
J.
Viruses
occurring
in
Plantago
species
in
England.
Plant
Pathology
1981;
30:
237­
44.
11.
Raybould
AF,
Maskell
LC,
Edwards
ML,
Cooper
JI,
Gray
AJ.
The
prevalence
and
spatial
distribution
of
viruses
in
natural
populations
of
Brassica
oleracea.
New
Phytologist
1999;
141:
265.
Page
30
of
34
12.
Gibbs
A.
A
plant
virus
that
partially
protects
its
wild
legume
host
against
herbivores.
Intervirology
1980;
13:
42­
7.
13.
Yahara
T,
Oyama
K.
Effects
of
virus
infection
on
demographic
traits
of
an
agamospermous
population
of
Eupatorium
chinense
(
Asteraceae).
Oecologia
1993;
96:
310­
5.
14.
Funayama
S,
Hikosaka
K,
Yahara
T.
Effects
of
virus
infection
and
growth
irradiance
on
fitness
components
and
photosynthetic
properties
of
Eupatorium
makinoi
(
Compositae).
American
Journal
of
Botany
1997;
84:
823­
30.
15.
Maskell
LC,
Raybould
AF,
Cooper
JI,
Edwards
ML,
Gray
AJ.
Effects
of
turnip
mosaic
virus
and
turnip
yellow
mosaic
virus
on
the
survival,
growth
and
reproduction
of
wild
cabbage
(
Brassica
oleracea).
Annals
of
Applied
Biology
1999;
135:
401­
7.
16.
Friess
N,
Maillet
J.
Influence
of
cucumber
mosaic
virus
infection
on
the
intraspecific
competitive
ability
and
fitness
of
purslane
(
Portulaca
oleracea).
New
Phytologist
1996;
132:
103­
11.
17.
Jones
RAC,
Nicholas
DA.
Impact
of
an
insidious
virus
disease
in
the
legume
component
on
the
species
balance
within
self­
regenerating
annual
pasture.
The
Journal
of
Agricultural
Science
1998;
131:
155­
70.
18.
Funayama
S,
Terashima
I,
Yahara
T.
Effects
of
virus
infection
and
light
environment
on
population
dynamics
of
Eupatorium
makinoi
(
Asteraceae).
American
Journal
of
Botany
2001;
88:
616­
22.
19.
Power
AG.
Ecological
risks
of
transgenic
virus­
resistant
crops.
In:
Letourneau
DK,
Burrows
BE.
Genetically
Engineered
Organisms:
Assessing
Environmental
and
Human
Health
Effects.
Boca
Raton:
CRC
Press,
2002:
125­
42.
20.
Pilson
D,
Prendeville
HR.
Ecological
effects
of
transgenic
crops
and
the
escape
of
transgenes
into
wild
populations.
Annual
Review
of
Ecology
and
Systematics
2004;
35:
149­
74.
21.
Remold
SK.
Unapparent
virus
infection
and
host
fitness
in
three
weedy
grass
species.
Journal
of
Ecology
2002;
90:
967.
22.
U.
S.
Environmental
Protection
Agency.
Minutes
of
the
October
13­
15,
2004
FIFRA
Scientific
Advisory
Panel
Meeting
on
Issues
Associated
with
Deployment
of
a
Type
of
Plant­
Incorporated
Protectant
(
PIP),
Specifically
those
Based
on
Plant
Viral
Coat
Proteins
(
PVCP­
PIPs).
2004.
23.
Hails
RS,
Morley
K.
Genes
invading
new
populations:
a
risk
assessment
perspective.
Trends
in
Ecology
&
Evolution
2005;
20:
245­
52.
24.
Haygood
R,
Ives
AR,
Andow
DA.
Population
genetics
of
transgene
containment.
Ecology
Letters
2004;
7:
213­
20.
25.
Khetarpal
RK,
Maisonneuve
B,
Maury
Y,
Chalhoub
B,
Dinant
S,
Lecoq
H
et
al.
Breeding
for
resistance
to
plant
viruses.
In:
Hadidi
A,
Khetarpal
RK,
Koganezawa
H.
Plant
Virus
Disease
Control.
St.
Paul:
APS
Press,
1998:
14­
32.
26.
National
Research
Council.
Genetically
Modified
Pest­
Protected
Plants:
Science
and
Regulation.
Washington,
DC:
National
Academy
Press,
2000.
27.
Hammond
J,
Lecoq
H,
Raccah
B.
Epidemiological
risks
from
mixed
virus
infections
and
transgenic
plants
expressing
viral
genes.
Advances
in
Virus
Research
1999;
54:
189­
314.
28.
Falk
BW,
Bruening
G.
Will
transgenic
crops
generate
new
viruses
and
new
diseases?
Science
1994;
263:
1395­
6.
29.
Pita
JS,
Fondong
VN,
Sangare
A,
Otim­
Nape
GW,
Ogwal
S,
Fauquet
CM.
Recombination,
pseudorecombination
and
synergism
of
geminiviruses
are
determinant
keys
to
the
epidemic
of
severe
cassava
mosaic
disease
in
Uganda.
Journal
of
General
Virology
2001;
82:
655­
65.
30.
de
Zoeton
GA.
Risk
assessment:
Do
we
let
history
repeat
itself?
Phytopathology
1991;
81:
585­
6.
31.
U.
S.
Environmental
Protection
Agency.
Viral
Interactions
in
Viral
Coat
Protein
Transgenic
Plants:
A
Literature
Review.
Washington,
DC:
Office
of
Science
Coordination
and
Policy,
Office
for
Prevention,
Pesticides
and
Toxic
Substances,
U.
S.
Environmental
Protection
Agency,
2004.
32.
Fraile
A,
Alonso­
Prados
JL,
Aranda
MA,
Bernal
JJ,
Malpica
JM,
García­
Arenal
F.
Genetic
exchange
by
recombination
or
reassortment
is
infrequent
in
natural
populations
of
a
tripartite
RNA
plant
virus.
Journal
of
Virology
1997;
71:
934­
40.
33.
Salánki
K,
Carrère
I,
Jacquemond
M,
Balázs
E,
Tepfer
M.
Biological
properties
of
pseudorecombinant
and
recombinant
strains
created
with
cucumber
mosaic
virus
and
tomato
aspermy
virus.
J.
Virol.
1997;
71:
3597­
602.
34.
Aaziz
R,
Tepfer
M.
Recombination
between
genomic
RNAs
of
two
cucumoviruses
under
conditions
of
minimal
selection
pressure.
Virology
1999;
263:
282­
9.
35.
Spitsin
S,
Steplewski
K,
Fleysh
N,
Belanger
H,
Mikheeva
T,
Shivprasad
S
et
al.
Expression
of
alfalfa
mosaic
virus
coat
protein
in
tobacco
mosaic
virus
(
TMV)
deficient
in
the
production
of
its
native
coat
Page
31
of
34
protein
supports
long­
distance
movement
of
a
chimeric
TMV.
Proceedings
of
the
National
Academy
of
Sciences
1999;
96:
2549­
53.
36.
Allison
RF,
Janda
M,
Ahlquist
P.
Sequence
of
cowpea
chlorotic
mottle
virus
RNAs
2
and
3
and
evidence
of
a
recombination
event
during
bromovirus
evolution.
Virology
1989;
172:
321­
30.
37.
Chenault
KD,
Melcher
U.
Phylogenetic
relationships
reveal
recombination
among
isolates
of
cauliflower
mosaic
virus.
J.
Mol.
Evol.
1994;
39:
496­
505.
38.
Gibbs
MJ,
Cooper
JI.
A
recombinational
event
in
the
history
of
luteoviruses
probably
induced
by
basepairing
between
the
genomes
of
two
distinct
viruses.
Virology
1995;
206:
1129­
32.
39.
Le
Gall
OL,
Lanneau
M,
Candresse
T,
Dunez
J.
The
nucleotide
sequence
of
the
RNA­
2
of
an
isolate
of
the
English
serotype
of
tomato
black
ring
virus:
RNA
recombination
in
the
history
of
nepoviruses.
J.
Gen.
Virol.
1995;
76:
1279­
83.
40.
Masuta
C,
Ueda
S,
Suzuki
M,
Uyeda
I.
Evolution
of
a
quadripartite
hybrid
virus
by
interspecific
exchange
and
recombination
between
replicase
components
of
two
related
tripartite
RNA
viruses.
Proc.
Natl.
Acad.
Sci.
1998;
95:
10487­
92.
41.
Zhou
X,
Liu
Y,
Calvert
L,
Munoz
C,
Otim­
Nape
GW,
Robinson
DJ
et
al.
Evidence
that
DNA­
A
of
a
geminivirus
associated
with
severe
cassava
mosaic
disease
in
Uganda
has
arisen
by
interspecific
recombination.
Journal
of
General
Virology
1997;
78:
2101­
11.
42.
Moonan
F,
Molina
J,
Mirkov
TE.
Sugarcane
yellow
leaf
virus:
an
emerging
virus
that
has
evolved
by
recombination
between
luteoviral
and
poleroviral
ancestors.
Virology
2000;
269:
156­
71.
43.
Fauquet
CM,
Sawyer
S,
Idris
AM,
Brown
JK.
Sequence
analysis
and
classification
of
apparent
recombinant
Begomoviruses
infecting
tomato
in
the
Nile
and
Mediterranean
Basins.
Phytopathology
2005;
95:
549­
55.
44.
Monci
F,
Sánchez­
Campos
S,
Navas­
Castillo
J,
Moriones
E.
A
natural
recombinant
between
the
Geminiviruses
Tomato
yellow
leaf
curl
Sardinia
virus
and
Tomato
yellow
leaf
curl
virus
exhibits
a
novel
pathogenic
phenotype
and
is
becoming
prevalent
in
Spanish
populations
.
Virology
2002;
303:
317­
26.
45.
Baliji
S,
Black
MC,
French
R,
Stenger
DC,
Sunter
G.
Spinach
curly
top
virus:
A
newly
described
Curtovirus
species
from
southwest
Text
with
incongruent
gene
phylogenies.
Phytopathology
2004;
94:
772­
8.
46.
Mayo
MA,
Jolly
CA.
The
5'­
terminal
sequence
of
potato
leafroll
virus
RNA:
evidence
of
recombination
between
virus
and
host
RNA.
J.
Gen.
Virol.
1991;
72:
2591­
5.
47.
Rubio
T,
Borja
M,
Scholthof
HB,
Jackson
AO.
Recombination
with
host
transgenes
and
effects
on
virus
evolution:
an
overview
and
opinion.
Molecular
Plant­
Microbe
Interactions
1999;
12:
87­
92.
48.
Chiang
CH,
Wang
JJ,
Jan
FJ,
Yeh
SD,
Gonsalves
D.
Comparative
reactions
of
recombinant
papaya
ringspot
viruses
with
chimeric
coat
protein
(
CP)
genes
and
wild­
type
viruses
on
CP­
transgenic
papaya.
Journal
of
General
Virology
2001;
82:
2827­
36.
49.
Lommel
SA,
Xiong
Z.
Reconstitution
of
a
functional
red
clover
necrotic
mosaic
virus
by
recombinational
rescue
of
the
cell­
to­
cell
movement
gene
expressed
in
a
transgenic
plant.
Journal
of
Cell
Biochemistry
1991;
15A:
151.
50.
Greene
AE,
Allison
RF.
Recombination
between
viral
RNA
and
transgenic
plant
transcripts.
Science
1994;
263:
1423­
5.
51.
Borja
M,
Rubio
T,
Scholthof
HB,
Jackson
AO.
Restoration
of
wild­
type
virus
by
double
recombination
of
tombusvirus
mutants
with
a
host
transgene.
Molecular
Plant­
Microbe
Interactions
1999;
12:
153­
62.
52.
Adair
TL,
Kearney
CM.
Recombination
between
a
3­
kilobase
tobacco
mosaic
virus
transgene
and
a
homologous
viral
construct
in
the
restoration
of
viral
and
nonviral
genes.
Archives
of
Virology
2000;
145:
1867­
83.
53.
Varrelmann
M,
Palkovics
L,
Maiss
E.
Transgenic
or
plant
expression
vector­
mediated
recombination
of
Plum
pox
virus.
Journal
of
Virology
2000;
74:
7462­
9.
54.
Gal
S,
Pisan
B,
Hohn
T,
Grimsley
N,
Hohn
B.
Agroinfection
of
transgenic
plants
leads
to
viable
cauliflower
mosaic
virus
by
intermolecular
recombination.
Virology
1992;
187:
525­
33.
55.
Schoelz
JE,
Wintermantel
WM.
Expansion
of
viral
host
range
through
complementation
and
recombination
in
transgenic
plants.
Plant
Cell
1993;
5:
1669­
79.
56.
Wintermantel
WM,
Schoelz
JE.
Isolation
of
recombinant
viruses
between
cauliflower
mosaic
virus
and
a
viral
gene
in
transgenic
plants
under
conditions
of
moderate
selection
pressure.
Virology
1996;
223:
156­
64.
57.
Frischmuth
T,
Stanley
J.
Recombination
between
viral
DNA
and
the
transgenic
coat
protein
gene
of
African
cassava
mosaic
geminivirus.
Journal
of
General
Virology
1998;
79:
1265­
71.
58.
Callaway
A,
Giesman­
Cookmeyer
D,
Gillock
ET,
Sit
TL,
Lommel
SA.
The
multifunctional
capsid
proteins
of
plant
RNA
viruses.
Annual
Review
of
Phytopathology
2001;
39:
419­
60.
Page
32
of
34
59.
Falk
BW,
Passmore
BK,
Watson
MT,
Chin
LS.
The
specificity
and
significance
of
heterologous
encapsidation
of
virus
and
virus­
like
RNAs.
In:
Bills
DD,
Kung
SD.
Biotechnology
and
Plant
Protection:
Viral
pathogenesis
and
disease
resistance.
Singapore:
World
Scientific,
1995:
391­
415.
60.
Tepfer
M.
Viral
genes
and
transgenic
plants:
What
are
the
potential
environmental
risks?
Biotechnology
(
N
Y)
1993;
11:
1125­
32.
61.
Rochow
WF.
Barley
yellow
dwarf
virus:
phenotypic
mixing
and
vector
specificity.
Science
1970;
167:
875­
8.
62.
Creamer
R,
Falk
BW.
Direct
detection
of
transcapsidated
barley
yellow
dwarf
luteoviruses
in
doubly
infected
plants.
J.
Gen.
Virol.
1990;
71:
211­
7.
63.
Bourdin
D,
Lecoq
H.
Evidence
that
heteroencapsidation
between
two
potyviruses
is
involved
in
aphid
transmission
of
a
non­
aphid­
transmissible
isolate
from
mixed
infections.
Phytopathology
1991;
81:
1459­
64.
64.
Taliansky
ME,
Robinson
DJ.
Molecular
biology
of
umbraviruses:
phantom
warriors.
Journal
of
General
Virology
2003;
84:
1951­
60.
65.
Robinson
DJ.
Environmental
risk
assessment
of
releases
of
transgenic
plants
containing
virus­
derived
inserts.
Transgenic
Research
1996;
5:
359­
62.
66.
Robinson
DJ.
Environmental
risk
assessment
of
releases
of
transgenic
plants
containing
virus­
derived
inserts.
Transgenic
Research
1996;
5:
359­
62.
67.
Fuchs
M,
Gal­
On
A,
Raccah
B,
Gonsalves
D.
Epidemiology
of
an
aphid
nontransmissible
potyvirus
in
fields
of
nontransgenic
and
coat
protein
transgenic
squash.
Transgenic
Research
1999;
8:
429­
39.
68.
Losey
JE,
Eubanks
MD.
Implications
of
pea
aphid
host­
plant
specialization
for
the
potential
colonization
of
vegetables
following
post­
harvest
emigration
from
forage
crops.
Environmental
Entomology
2000;
29:
1283­
8.
69.
Osbourn
JK,
Sarkar
S,
Wilson
TMA.
Complementation
of
coat
protein­
defective
TMV
mutants
in
transgenic
tobacco
plants
expressing
TMV
coat
protein.
Virology
1990;
179:
921­
5.
70.
Candelier­
Harvey
P,
Hull
R.
Cucumber
mosaic
virus
genome
is
encapsidated
in
alfalfa
mosaic
virus
coat
protein
expressed
in
transgenic
tobacco
plants.
Transgenic
Research
1993;
2:
277­
85.
71.
Lecoq
H,
Ravelonandro
M,
Wipf­
Scheibel
C,
Monsion
M,
Raccah
B,
Dunez
J.
Aphid
transmission
of
a
non­
aphid­
transmissible
strain
of
zucchini
yellow
mosaic
potyvirus
from
transgenic
plants
expressing
the
capsid
protein
of
plum
pox
potyvirus.
Mol.
Plant
Microbe
Interact.
1993;
6:
403­
6.
72.
Maiss
E,
Koenig
R,
Lesemann
DE.
Heterologous
encapsidation
of
viruses
in
transgenic
plants
and
in
mixed
infections.
In:
Jones
DD.
Biosafety
Results
of
Field
Tests
of
Genetically
Modified
Plants
and
Microorganisms.
Oakland:
University
of
California
,
Division
of
Agriculture
and
Natural
Resources,
1994:
129­
39.
73.
Vance
VB,
Berger
PH,
Carrington
JC,
Hunt
AG,
Shi
XM.
5'
proximal
potyviral
sequences
mediate
potato
virus
X/
potyviral
synergistic
disease
in
transgenic
tobacco.
Virology
1995;
206:
583­
90.
74.
OECD
Environment
Directorate.
Consensus
document
on
general
information
concerning
the
biosafety
of
crop
plants
made
virus
resistant
through
coat
protein
gene­
mediated
protection.
http://
www.
olis.
oecd.
org/
olis/
1996doc.
nsf/
62f30f71be4ed8a24125669e003b5f73/
ce3a104b8ada9e8ac1256
3e2003183bb/$
FILE/
11E63213.
ENG
.
1996.
75.
Miller
WA,
Koev
G,
Mohan
BR.
Are
there
risks
associated
with
transgenic
resistance
to
luteoviruses?
Plant
Disease
1997;
81:
700­
10.
76.
Pruss
GJ,
Ge
X,
Shi
XM,
Carrington
JC,
Vance
VB.
Plant
viral
synergism:
the
potyviral
genome
encodes
a
broad­
range
pathogenicity
enhancer
that
transactivates
replication
of
heterologous
viruses.
The
Plant
Cell
1997;
9:
859­
68.
77.
García­
Arenal
F,
Malpica
JM,
Fraile
A.
Evolution
of
plant
virus
populations:
The
role
of
genetic
exchange.
In:
Fairbairn
C,
Scoles
G,
McHughen
A.
Proceedings
of
the
6th
International
Symposium
on
the
Biosafety
of
Genetically
Modified
Organisms.
Saskatoon,
Canada:
University
Extension
Press,
University
of
Saskatchewan,
2000:
91­
6.
78.
Thomas
PE,
Hassan
S,
Kaniewski
WK,
Lawson
EC,
Zalewski
JC.
A
search
for
evidence
of
virus/
transgene
interactions
in
potatoes
transformed
with
the
potato
leafroll
virus
replicase
and
coat
protein
genes.
Molecular
Breeding
1998;
4:
407­
17.
79.
Fuchs
M,
Klas
FE,
McFerson
JR,
Gonsalves
D.
Transgenic
melon
and
squash
expressing
coat
protein
genes
of
aphid­
borne
viruses
do
not
assist
the
spread
of
an
aphid
non­
transmissible
strain
of
cucumber
mosaic
virus
in
the
field.
Transgenic
Research
1998;
7:
449­
62.
Page
33
of
34
80.
Lin
HX,
Rubio
L,
Smythe
A,
Jiminez
M,
Falk
BW.
Genetic
diversity
and
biological
variation
among
California
isolates
of
Cucumber
mosaic
virus.
Journal
of
General
Virology
2003;
84:
249­
58.
81.
Vigne
E,
Komar
V,
Fuchs
M.
Field
safety
assessment
of
recombination
in
transgenic
grapevines
expressing
the
coat
protein
gene
of
Grapevine
fanleaf
virus.
Transgenic
Research
2004;
13:
165­
79.
82.
Allison
RF,
Schneider
WL,
Deng
M.
Risk
assessment
of
virus
resistant
transgenic
plants.
In:
Schiemann
J.
The
Biosafety
Results
of
Field
Tests
of
Genetically
Modified
Plants
and
Microorganisms.
Berlin:
Biologische
Bundesanstalt
für
Land­
und
Forstwirtschaft,
2000:
186­
8.
83.
Dinant
S,
Blaise
F,
Kusiak
C,
Astier­
Manifacier
S,
Albouy
J.
Heterologous
resistance
to
potato
virus
Y
in
transgenic
tobacco
plants
expressing
the
coat
protein
gene
of
lettuce
mosaic
potyvirus.
Phytopathology
1993;
83:
819­
24.
84.
Miller
J.
Biotech
boosts
natural
bounty.
Today's
Chemist
at
Work
2000;
9:
38­
44.
85.
Nagy
PD,
Bujarski
JJ.
Homologous
RNA
recombination
in
brome
mosaic
virus:
AU­
rich
sequences
decrease
the
accuracy
of
crossovers.
Journal
of
Virology
1996;
70:
415­
26.
86.
Nagy
PD,
Bujarski
JJ.
Silencing
homologous
RNA
recombination
hot
spots
with
GC­
rich
sequences
in
brome
mosaic
virus.
Journal
of
Virology
1998;
72:
1122­
30.
87.
Nagy
PD,
Ogiela
C,
Bujarski
JJ.
Mapping
sequences
active
in
homologous
RNA
recombination
in
brome
mosaic
virus:
prediction
of
recombination
hot
spots.
Virology
1999;
254:
92­
104.
88.
Teycheney
PY,
Aaziz
R,
Dinant
S,
Salánki
K,
Tourneur
C,
Balázs
E
et
al.
Synthesis
of
(­)­
strand
RNA
from
the
3'
untranslated
region
of
plant
viral
genomes
expressed
in
transgenic
plants
upon
infection
with
related
viruses.
Journal
of
General
Virology
2000;
81:
1121­
6.
89.
Gibbs
M.
Risks
in
using
transgenic
plants?
Science
1994;
264:
1650­
1.
90.
Hull
R.
Risks
in
using
transgenic
plants?
Science
1994;
264:
1649­
50.
91.
Dinant
S,
Blaise
F,
Kusiak
C,
Astier­
Manifacier
S,
Albouy
J.
Heterologous
resistance
to
potato
virus
Y
in
transgenic
tobacco
plants
expressing
the
coat
protein
gene
of
lettuce
mosaic
potyvirus.
Phytopathology
1993;
83:
818­
24.
92.
Sacher
R,
Ahlquist
P.
Effects
of
deletions
in
the
N­
terminal
basic
arm
of
brome
mosaic
virus
coat
protein
on
RNA
packaging
and
systemic
infection.
Journal
of
Virology
1989;
63:
4545­
52.
93.
Berg
J,
Tymoczko
J,
Stryer
L,
Clarke
N.
Biochemistry,
5th
ed.
New
York:
W.
H.
Freeman
and
Company,
2002.
94.
New
Collegiate
Dictionary.
Springfield,
Massachusetts:
G.
&
C.
Merriam
Company,
1977.
Page
34
of
34
Appendix:
Index
of
Exemption
Criteria
(
a)
Criterion
a
is
satisfied
if
either
paragraph
1
or
paragraph
2
applies:

(
1)
The
plant
containing
the
PVCP­
PIP
is
one
of
the
following:
almond
(
Prunus
communis),
apricot
(
Prunus
armeniaca),
asparagus
(
Asparagus
officinale)
avocado
(
Persea
americana),
banana
(
Musa
acuminata),
barley
(
Hordeum
vulgare),
bean
(
Phaseolus
vulgaris),
black­
eyed
pea
(
Vigna
unguiculata),
cacao
(
Theobroma
cacao),
celery
(
Apium
graveolens),
chickpea
(
Cicer
arietinum),
citrus
(
Citrus
spp.),
coffee
(
Coffea
arabicua),
corn
(
Zea
maize),
cucumber
(
Cucumis
sativus),
eggplant
(
Solanum
melongena),
guava
(
Psidium
guajava),
kiwi
(
Actinidia
spp.),
mango
(
Mangifera
indica),
nectarine
(
Prunus
persica),
okra
(
Abelmoschus
esculentus),
olive
(
Olea
europaea),
papaya
(
Carica
papaya),
parsley
(
Petroselinum
crispum),
pea
(
Pisum
sativum),
peach
(
Prunus
persica),
peanut
(
Arachis
hypogaea),
pineapple
(
Ananas
comosus),
pistachio
(
Pistacia
vera),
plum
(
Prunus
domestica),
potato
(
Solanum
tuberosum),
soybean
(
Glycine
max),
spinach
(
Spinacia
oleracea),
starfruit
(
Averrhoa
carambola),
taro
(
Colocasia
esculenta),
tomato
(
Lycopersicon
lycopersicum),
or
watermelon
(
Citrullus
lanatus).

(
2)
The
Agency
determines
after
review
that
the
plant
containing
the
PVCP­
PIP
(
i)
is
itself
not
a
weedy
or
invasive
species
outside
of
agricultural
fields
in
the
United
States,
its
possessions,
or
territories,
and
(
ii)
does
not
have
relatives
outside
of
agricultural
fields
in
the
United
States,
its
possessions,
or
territories
that
are
weedy
or
invasive
species
or
endangered/
threatened
species
with
which
it
can
produce
viable
hybrids
in
nature.

(
b)
Criterion
b
is
satisfied
if
either
paragraph
1
or
paragraph
2
applies:

(
1)
The
viral
pathotype
used
to
create
the
PVCP­
PIP
has
naturally
infected
plants
in
the
United
States,
its
possessions,
or
territories
and
naturally
infects
plants
of
the
same
species
as
that
containing
the
PVCPPIP

(
2)
The
Agency
determines
after
review
that
(
i)
the
properties
of
the
viral
pathotype
that
are
determined
by
the
coat
protein
gene
used
to
create
the
PVCP­
PIP
are
substantially
similar
to
the
properties
of
a
viral
pathotype
that
naturally
infects
plants
in
the
United
States,
its
possessions,
or
territories,
and
the
viral
pathotype
used
to
create
the
PVCP­
PIP
naturally
infects
plants
of
the
same
species
as
that
containing
the
PVCP­
PIP,
or
(
ii)
viruses
that
naturally
infect
the
plant
containing
the
PVCP­
PIP
are
unlikely
to
acquire
the
coat
protein
sequence
through
recombination
and
produce
a
viable
virus
with
significantly
different
properties
than
either
parent
virus.

(
c)
Criterion
c
is
satisfied
if
either
paragraph
1
or
paragraph
2
applies:

(
1)
The
genetic
material
encodes
only
a
single
contiguous
portion
of
each
unmodified
viral
coat
protein.
This
would
allow
multiple
PVC­
proteins
that
could
each
separately
qualify
for
the
exemption.
Chimeric
PVC­
proteins
would
not
qualify.

(
2)
The
Agency
determines
after
review
that
the
genetic
material
(
i)
encodes
a
protein
that
is
minimally
modified
from
a
coat
protein
from
a
virus
that
naturally
infects
plants,
or
(
ii)
produces
no
protein.
