A large particle associated with the perimeter of the nuclear pore complex

Abstract

The three-dimensional structure of the nuclear pore complex has been determined to a resolution of approximately 90 A by electron microscopy using nuclear envelopes from Xenopus oocytes. It is shown to be an assembly of several discrete constituents arranged with octagonal symmetry about a central axis. There are apparent twofold axes perpendicular to the octad axis which suggest that the framework of the pore complex is constructed from two equal but oppositely facing halves. The half facing the cytoplasm is in some instances decorated by large particles, similar in appearance and size to ribosomes.

Full text

A
Large
Particle
Associated
with
the
Perimeter
of
the
Nuclear
Pore
Complex
P
.
N
.
T
.
UNWIN
and
R
.
A
.
MILLIGAN
Department
of
Structural
Biology,
Stanford
University
Medical
School,
Stanford,
California
94305
;
and
Medical
Research
Council
Laboratory
of
Molecular
Biology,
Cambridge, England
ABSTRACT

The
three-dimensional
structure
of the
nuclear
pore
complex
has
been
determined
to
a
resolution
of
-90
A
by
electron
microscopy
using
nuclear
envelopes
from
Xenopus
oocytes
.
It
is
shown
to
be an assembly
of
several
discrete
constituents
arranged with
octagonal
symmetry
about
a
central
axis
.
There
are
apparent
twofold
axes
perpendicular
to
the
octad
axis
which
suggest
that
the
framework
of the
pore
complex
is
constructed
from
two
equal
but
oppositely
facing halves
.
The
half
facing
the
cytoplasm
is
in
some
instances
decorated
by
large
particles,
similar
in
appearance
and
size to
ribosomes
.
The
nuclear pore
complex
is
an
organelle,
ubiquitous
to
eucar-
yotic
cells,
which
serves
as
a
pathway
through
the
nuclear
envelope
for
a
variety
of nuclear
and
cytoplasmic molecules
.
Microinjection
experiments
involving substances
such
as
dex-
trans
(24),
colloidal
gold
(10),
and
proteins
(4, 7,
16)
suggest
that
it
forms
an
opening
for
passive
movement
of
molecules
up
to
-90
A
in
diameter
.
Biochemical
studies
suggest
that
it
may
be
composed
of
only
a
few
major
polypeptides
(19)
and
may
contain
RNA
(28),
although
no
direct
identification
has
yet
been
achieved
of
these
components
within the confines of
the
structure
.
Besides
this
knowledge,
very
little
information
is
available
on
the
chemical
or
physical
nature of
the
constituents
or
of
their
mechanisms
of
action
.
Electron
microscopy
shows
that
the
pore
complex
is
a
cylin-
drical
assembly
spanning
the
two
nuclear
membranes
and
having
components
arranged with
octagonal
symmetry
around
its
central
axis
(6, 9, 12,
14)
.
However,
the
details
described
by
different
authors vary
considerably,
probably
because
of
real
differences
associated
with
the
cell
cycle
and
also
because
of
artifactual
differences
related
to
quality
of
preservation
.
Here
we
have
investigated
the structure
of pore
complexes
from
Xenopus
oocytes using
Fourier
averaging
methods
to
reveal
the
details
.
Additional information
obtained
by
selec-
tively
releasing
some
of the
constituents
has
led
to
a
clearer
picture
of
their
organization
and
of
the
overall
three-dimen-
sional
configuration
.
Our
results
emphasize
the subunit
nature
of
the pore
complex
and
we
show
that
particles
similar
in
appearance
and
size
to
ribosomes
associate
with
its
cytoplasmic
perimeter
.
THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
93
APRIL
1982
63-75
©
The
Rockefeller
University
Press
"
0021-9525/82/04/0063/13
$1
.00
MATERIALS
AND
METHODS
General
Nuclear
envelopes
used
were
from
the
oocytes
of
Xenopus
laevis
.
The
oocytes
were
kept
in
small
ovary
pieces in
modified
Barth's
saline
solution (15) at
l7°C
for
periods
up
to
several
days
.
Chemicals
were
obtained
from
the
following
sources
:
triethanolamine
chloride
from
BDH
Chemicals Ltd
.
(Poole,
England)
;
poly-L-Lysine
(40,000
mol
wt)
and
gold
thio-glucose
from
Sigma
Chemical
Co
.
(St
.
Louis,
MO)
;
glutaraldehyde
(ultrapure)
from
Emscope
Laboratories
(London,
England)
;
Triton
X-100
from
Bio-Rod
Laboratories
(Richmond,
CA)
.
Low
salt
medium
was
0
.5
mM
MgCl,
in
1
mM
triethanolamine
chloride
adjusted
to
pH
8
with
KOH
.
High
salt
medium
was400
mM
KCI,
5
mM
MgCl2,
20
mM
triethanolamine
chloride,
adjusted
to
pH
8
with
KOH
.
All
solutions
were
prepared with
double-distilled
water
.
Isolation
of
Nuclear
Envelope
Nuclei
were
isolated
from mature
oocytes
directly into
the
Barth
medium
by
extruding
them
through
small
holes
punched
in
the
center of
the oocyte
black
hemispheres
.
Each
nucleus,
after
isolation,
was
cleaned
by
several
passes
up
and
down
a narrow
capillary
and
immediately
transferred
to
an
electron
microscope
grid
.
Isolation
of
Ribosomes
To
compare
particles
on
the
nuclear
envelope
with
particles
known
to
be
ribosomes,
we
used
as
markers
ribosomes
isolated
from
hypothermic
chick
embryos
(which
have
about
the
same
molecular
weight
as
those
from Xenopus
[20])
.
The
ribosomes
were
isolated
as
tetramers
from a
postmitochondrial
super-
natant
following,
with
modifications,
the
procedure
described
in
(23),
and
they
were
shown
to
be
undegraded
in
terms
of
their
two-dimensional
gel
electropho-
resis
patterns
(Milligan
and
Unwin,
manuscript
in
preparation)
.
63
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

Sample
Preparation
The
pore
complex
constituents
were
examined
in
situ,
attached
to
the
nuclear
membranes,
and
also in
isolation,
detached
from them
.
For
the
first
purpose
we
used
600-mesh
grids
overlaid with a
holey
carbon
support
film,
and
the
capillary
was
withdrawn
to leave the
nucleus
behind
within
asmall
hemispherical
droplet
on
one
surface
.
With
two
fine,
but
blunted,
glass
needles,
the
envelope
was
disrupted
to
release
the
nuclear contents
and
spread
across the
surface,
cytoplas-
mic
face
in
contact
.
Once
firmly
adhered,
the
envelope
was
washed
in
low
salt
medium
(
"
30
s),
fixed
with
2.5%
glutaraldehyde
in the
same
medium
(1
min),
and
postfixed
with
1%
osmium
tetroxide
(1
min)
.
These
and
subsequent
steps
were
carried
out
with the
grid
completely
immersed
so
that
both
sides
received
the
same
treatments
.
In
some
experiments,
we
carried out
an
additional
incuba-
tion
for
l
min
in
high
salt,
or in
low
salt
with
the
MgC1
2
replaced
by
I
mM
ethylenediaminetetraacetic acid
(EDTA),
between
the
washing
and
fixation
steps
.
For
the
second
purpose
we
used
normal
carbon-coated
grids
which
had
been
rendered
hydropilic
on
one
surface
by
bathing
in 0.1% polylysine
(34)
.
Each
grid
was
placed
on
parafilm,
hydrophilic
side
uppermost,
and
immersed
in a
droplet
of
low
salt
medium
(sometimes
containing
0
.1%
Triton
X-100)
before the
nucleus
was
delivered
onto
it
.
After
its
attachment
to
the
grid,
the
envelope
was
disrupted,
partially
spread
as
described
above,
and
centrifuged
at
500
g for
1
min
.
The
envelope
was
then
partially
detached
from
the grid
by
vigorous
swirling
in
high
salt
medium
(or
low
salt
medium
when
Triton
X-100
was
used),
leaving
behind
various
constituents
.
A
small proportion
of
pore
complexes
were
released
intact
when
Triton
X-100
was
used
.
Staining
was
done
in the usual
way
(8)
using
2%
gold
thio-glucose
or
1%
uranyl
acetate
.
The
former
stain
gave
noticeably
better
preservation
ofthe
nuclear
envelope
;
otherwise,
at
the
resolution
of
this
study,
both
stains
gave
equivalent
results
.
Fixation
with
glutaraldehyde
was
avoided
when
uranyl
acetate
was
used
.
Electron
Microscopy
and
Image
Processing
Micrographs
were
recorded
at
x
12,000-20,1x10
using
a Philips
EM301
or
EM400
electron
microscope
operating
at
80 kv
.
A
goniometer
stage
was
used
for
the
tilting
experiments
.
Nuclear
envelopes
were viewed
with
their
cytoplasmic
side
nearest
to
the
electron
source
.
Images of
the
pore
complex
were
analyzed
by
numerical
Fourier
methods
.
Those
images
displaying the
most
perfect
eightfold rotational
symmetry
were
assumed
to
best
represent
the
true
structure
.
By
this
criterion,
and
for
pore
complexes
attached
to the
nuclear
membranes,
good
preservation
was
achieved
only
in regions
overlying
holes in the
carbon
support
film,
and
in
parts
where
the
stain
was
sufficiently
deep
that
the
space
between
the pores
was
of
uniform
density
.
k
-++++----
++
++++

+ +
+++++++++
--
+
___-___+++--
+--
_________
__
+-__
____-______
+
-__
+++
_______
+++++--_
++++---
---
+++
--
++++---
+
----+++
---
++---
++
++-----+++
+
++---+++++++
-----++
-+
++++-++++++++
-----+
- +
+++++++
-+++ -----
--
+
- - - - - -

-
- -
- - -
-
-
-
- -
-
-
-

- -
____+++___
___
__ _- ++
+----++
-++
-++---++
++
+++--++++++
++--
++
----
h
FIGURE
1

Plot
of
centrosymmetric
phases
(plus
and
minus
are
0°
and
180°)
in
the
Fourier
transform
calculated
from an image
of
a
pore
complex
attached
to
the
nuclear
membranes
.
The
plot
was
obtained
by
sampling
the
continuous
transform
on
a
square
lattice
at
intervals
of
4.34
x
10
-°
A
-
'
and
averaging the
complex
values
at
these
points,
(h, k),
with
additional values
collected
at
the
eightfold
related
positions
.
The
projection
computed
using
these phases
and
averaged
amplitudes
is
identical
to
the
one
in Fig
.
9
a,
derived
by
harmonic
analysis
of
the
same
data
.
6
4

THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
93,
1982
TABLE
I
Details
of
Three-dimensional
Data
Refinement
FIGURE
2

Isolated
nuclear
envelope from X
.
laevis
spread
over
a
carbon
support
film
.
The
pore
complexes
are
usually
irregularly
organized
but
sometimes
form
linear
and
square
arrays
.
Unstained
.
x
8,000
.
The
images
were
densitometered
with
a
modified
Nikon
comparator
(3)
or
Perkin-Elmer
101
OA
automatic
microdensitometer (Perkin-Elmer
Corp
.,
Instru-
ment
Div
.,
Norwalk,
CT)
to
convert
the
optical
densities
into
numerical
arrays
.
The
step
size
was
30
pin,
corresponding
to 18
A
at the
specimen,
and
the array
size
was
256
x
256
points,
or,
in
the case of the
three-demensional
study,
512
x
512
points
.
Circular regions in these
arrays,
enclosing
just
the
pore
complex,
were
boxed
off
and
Fourier
transformed,
using
a
program
of the
type
described
by
DeRosier
and
Moore
(8)
.
The
transforms,
rather
than the
images,
were used
for
subsequent manipulations
.
Power
spectra
and
rotationally
averaged
projections
were
calculated following
Crowther
and
Amos
(5)
.
The
power
spectra are
derived
by
calculating
the
weight
of
a
specified
n-fold
rotational
harmonic
forthe best
image
center
consistent
with
n-fold
rotational
symmetry,
and
repeating
this
calculation
over
a
range
of
values
of
n
.
Displayed
are the
relative
strengths
of
the
rotational
harmonics
contributing
to the
image,
allowing
an
objective
assessment
of
the
degree
of
preservation to be
made
.
The
projection
maps
were
derived
by
Fourier-Bessel synthesis
of
just
those
harmonics
consistent
with
the
observed
eightfold
rotational
symmetry
.
A
three-dimensional
map
was
calculated
from
an
image
of
a
pore
complex
attached
to
the
nuclear
membranes
(Fig
.
7)
to
determine
the
distribution
of
matter
in the
direction
perpendicular
to the
membrane
plane
.
Complex
super-
position
of
detail
precluded
the
possibility
of
obtaining
this
information
directly
Number
of
images
10
Range
in
angle
of
tilt
0-52
°
Resolution
cut-off
130
A
Average phase
error,
based
on
comparison
of
individ-
29
°
ual
values
along
lattice
lines
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
3

Constituents
of
the
nuclear
envelope
released
after
its
partial
detachment
from
a
polylysine-treated
carbon
film
in
the
presence
of
0
.1%
Triton
X-100
(see
Methods
and
Materials)
.
Those
constituents
most
obviously
related to
the
pore
complex
are
:
rings
(R),
central
plug
(
C),
spokes
(S),
and
particles
(
P),
occasionally
observed
around
the
rings
.
As
the
lower micrograph
shows,
the
rings
are
sometimes
obtained
in
large
numbers
by
themselves
.
Uranyl acetate
stain
.
X
60,000
.
from
different
views
.
We
manipulated
the
Fourier
transforms
as
if
they
were
derived
from
a
two-dimensional
crystal
by
creating
an
artificial
reciprocal
lattice
consisting
of
lines
arranged
on
a
square
grid
and
oriented
parallel
to
the
octad
axis
.
Images
from
a
tilt
series
were used
to
provide
many
estimates of
amplitude
and
phase
at
different
points
along
these
lines,
allowing
the
continuous
variations
to
be
mapped
out
along
them
.
The
continuous
curves,
sampled
at
regular
intervals,
provided
the
Fourier
terms
from
which
to
calculate
the
structure
.
We
combined
the
transform
measurements
starting
with
the projection
data,
Fig
.
1,
and
refined the phases,
image
by
image,
in
order of
increasi
.
.g
tilt
.
Each
image,
on
account
of
the
octagonal
symmetry,
provided
up
to
four
independent
UNWIN AND
MILLIGAN

Particles
around
Nuclear
Pore
Complex

65
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
4

Detached
pore
complexes
released
onto
the
microscope
grid
from
a
nuclear
envelope
immersed
in
low
salt
medium
containing
0
.1%
Triton
X-100
.
Pore
complexes
within
square
arrays
are
better
preserved
than
others
.
A,
B, C,
and
D
refer
to
images
from
which
the
results
in Fig
.
6were
obtained
;
the
contributions
from
the
eightfold
harmonics
are
respectively
51%,
47%, 46%,
and
43%
of
the
total
power
associated with
azimuthally
varying
components
.
Also
shown
are
oblique,
"O'
;
and
edge-on
views
(inset)
of
the
pore
complex
.
The
drawing
indicates
how
the
view,
"O",
can be
interpreted
in
three-dimensions
in
terms
of
two
coaxial
rings
with
matter
lying
in
between
them
.
These
rings
are
not
obvious
in
the
en
face
views
since
they
are
thin
in
comparison
with
the
rest
of
the
structure
and
hence
do
not
contribute
much
contrast
when
viewed
from
this
direction
.
Uranyl
acetate
stain
.
x
75,000
.
THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
93,
1982
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
5

Detached
nuclear
pore
complex
tilted
about
the
octad
axis
.
The
tilt
angles
are
indicated
.
The
position
of
the
putative
twofold
axis
is
drawn
for
the
untilted
view
.
Uranyl
acetate
stain
.
X
120,000
.
estimates
of
amplitude
and
phase
along
each
lattice
line
.
The
positions
of
the
lattice
(and
eightfold
related)
points in
the
transforms
were
calculated
from
the
known
magnitude of
tilt
and
the
direction
of
the
tilt
axis
on
the
micrograph
.
Only
those
phases
for
which
the
corresponding
amplitudes
were
fairly
strong
(-25%
of
the
total
number)
and
which
along
any
given
lattice
line
lay within
2
.17
x
10
-
"
A
-
'
of
each
other
were
used
in
the
refinement
comparisons
.
The
required
Fourier
terms were
collected
from
the
smooth
amplitude
and
phase
curves
derived
from
these
data
by
sampling
at
intervals
of
4
.34
x
10
-°
A
-
'
along
the
lattice
lines
.
Errors
involved
in
drawing out
these
smooth
curves
caused
only
small departures
from
perfect
octagonal
symmetry
.
Further
details
are
given
in
Table
1
.
RESULTS
Pore
complexes
in
the nuclear
envelopes
of
Xenopus
oocytes
occupy
a
large
fraction
of
the
total
membrane
surface
.
Their
organization within the
membranes
is
usually
rather
irregular,
but
they
sometimes
form
lines
or,
more
rarely,
square
arrays
(Fig
.
2)
.
Such
motifs are
probably
a
result
of
their
interaction
with
the
thin
nuclear
lamina
(1,
19)
and
with each
other
.
A
complete
description
of the
pore
complex,
within
the envelope,
is
derived
below
by
bringing
together
information
obtained
from
several
types
of
experiment
.
Plugs,
Spokes,
Particles,
and
Rings
We
find
that
the
pore
complex
is
constructed from, or
related
to,
several
discrete
constituents
.
These
are
most
easily
recog-
nized following
release
with
the
detergent,
Triton
X-100
(see
Methods
and
Materials),
when
both
the separated
constituents
and
intact
pore
complexes appear
next
to
the
envelope
skeleton
(Fig
.
3)
.
Clearly
visible
are
:
(a)
"rings"
which
have an
inside
diameter
close
to
800
A
and
an
outer
diameter
the
same
as
that
ofthe
pore
complex
itself
(-
1,200
A)
;
(b)
large
particles
forming
"plugs"
at
the
centers
of
the
pore
complexes
;
(c)
smaller
(-x220
A)
particles
occasionally
arranged
around
the
circumference
of
the
rings
;
and
(d)
"spokes,"
matter
in intact
pores
extending
radially
outwards
from
the plugs
towards
the periphery
.
The
plugs
have
a
diameter
of
up
to
350
A,
depending
possibly
on
their state
of
preservation
.
The
rings
are
composed
of
globular
subunits
and
their
power
spectra display
weak
eightfold
com-
ponents
(results
not
shown)
.
Detached
Pore
Complexes
To
learn
about
the
three-dimensional
arrangement
of
some
of
these constituents
we
investigated
further
the
structure
of
detached pore
complexes
.
Viewed
en
face
(i.e
.,
from
a
direction
perpendicular
to
the
plane
in
which
the
membranes
would
lie)
the
detached
pore
complex
is
divisible
into
eight parts
which
are
approximately
equivalent
and
symmetrical
about
lines
drawn
radially
(Fig
.
4
;
see
also
Fig
.
6)
.
It
therefore
seems
to
display
elements
of
both
octagonal
and
mirror
symmetry
.
The
appearance
of
true
mirror
symmetry
would
suggest
that
it is
composed
of
two
equal
but
oppositely facing
halves
(i
.e
.,
halves
related
by
twofold
axes
perpendicular
to
the octad
axis
and
lying
in
the
central
plane)
.
Other
views
are
in
accord
with
this
configuration
.
For
example
the
oblique
view,
"0",
in
Fig
.
4,
shows
two
circular
outer
rims
which
are
coaxial
and
equal
in
diameter
and
thickness
.
Views
perpendicular
to
the
octad
axis
(Fig
.
5,
and
inset
to
Fig
.
4)
show
a
central
zone
of matter
flanked
by
two
lines
(the
rims seen
edge-on)
which
are
equally
prominent
at
their
extremities
and
symmetrically disposed
on
either
side
.
When
this
structure
is
tilted
about
the
octad
axis (Fig
.
5)
the
rims
produce
only small
variations
in
contrast
and
apparent
diameter,
except
at
high
tilts
where
flattening
effects
(11)
become
most
significant
.
This
indicates
that
the rims
are
rings
of
approximately
constant
thickness
rather than,
say,
circular
arrays
of
the
-220
A-diameter
particles
.
Moreover
the
diameter
of the rims
corresponds
with
that
of the
rings
in Fig
.
3
.
We
thus
suppose
the
pore
complex
to
be
framed by
two
equal
rings
facing
toward
the
nucleus
and
the
cytoplasm
of the
cell,
respectively
.
The
separation
between
the
rings
varies
(300-600
A)
probably
because
of
rather
flexible,
or
easily
distorted,
connections
(most
obvious
in
the
-18°
and
-9°
tilts)
to
the
rest
of
the
assembly
.
In
the
parts
of
the
pore
complex
excluding
the
rings
the
changes
in
contrast
with
angle of
tilt
are
more
pronounced
.
At
-27°
(Fig
.
5),
for
example,
the
central
zone
displays
a
peak
of
density
over the
octad
axis,
whereas
at
0°
the
density
is
more
evenly
distributed
.
The
distribution
of
matter
at
the
extremities
of
this
zone,
where
it
links
up
with
the
rings,
changes
to
a
similar
degree
.
These
variations
are
consistent
with
features
varying
octagonally
around
the
axis
of
the
pore complex,
although
poor
preservation
prevents
an
exact
correspondence
of
views
differing
in
tilt
by
45°
.
Power
spectra
calculated
from
images
of
en
face
views
which
display
strong
octagonal
symmetry
indicate
that
preservation
is
best
in
"crystalline" regions,
where
the
pore
complexes
lie
closely
apposed
.
Pore
complexes
by
themselves
are
apparently
more
subject
to
staining
distortions,
involving
differential
shrinkage,
as
described
by
Moody
(22)
.
Best
preserved pore
complexes
show
evidence
of
16-
and
24-fold
harmonics
in
addition
to
the
basic
eightfold
harmonic
(Fig
.
6)
.
In
general,
the
stronger
the
higher order
contributions
relative to
the
background
the
more
perfect
the
mirror
symmetry-a
phenom-
enon
one
would
expect
if
indeed
the
pore
complexes
are
constructed
from
two
equal
and
oppositely
facing
halves
.
Consistent
features
of
the
projection
maps
(Fig
.
6)
are
a
round
central
plug, eight
prominent
spokes
emanating
from
it
radially,
and
a
partitioning
of
each
spoke
into
characteristic
regions
or
domains
.
We
distinguish
an
inner
domain
at
radii
between
220
and
400
A
where
the
spoke
is
narrowest
and
tending
to
connect
up
circumferentially
with
its
equivalent
UNWIN
AND
MILLIGAN

Particles
around
Nuclear Pore
Complex

67
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
6
Projection
maps
obtained
from
the
detached
pore
complexes,
A,
e,
C,
and
D
in
Fig
.
4,
and
their
rotational
power
spectra,
plotted
on
a
logarithmic
scale
.
Improvements
of
the
power
spectra
in
terms
of
enhanced
16-fold
(n
=
16)
and
24-fold
(n
-
24)
harmonics
relative
to
the
background
level
are
correlated
with
a
stronger
tendency
towards
mirror
symmetry
.
The
resolution
is
--90,$
.
The
consistently
observed
broad and
sharp peaks
of
density
(shaded)
are
at radii of
450
and 550
A,
respectively
.
The broken
lines
drawn
in
D
indicate
the
positions
of
the
membrane
border
and
the
particulate
matter
shown
in Fig
.
9
;
the
two
radial
lines
correspond
to
the
putative
twofold
axes
which
give
rise
to
the
appearance
of mirror
symmetry
in
projection
.
neighbors,
and
an
outer
domain
composed
of
two
peaks,
one
rather
broad
and
the
other
rather
sharp
at
radii
of
450 and
550
A,
respectively
.
Correlation
with
the
edge-on view
suggests that
the
inner
domain and
plug
are
located
in
the
central
plane of
the
pore
whereas
the
outer
domain
encompasses
the region
where
the
two
oppositely
facing
halves
of
the
spokes
diverge
from
this
plane
to link
up
with
the
rings
.
The
broad
peak
(outer
domain)
appears
to
be
the
part
of
the structure
where
the
rings
and
the
matter
in
the
central
plane
superimpose
.
The
68

THE
JOURNAL
Of
CELL
BIOLOGY
"
VOLUME
93,
1982
rings
themselves
contribute
little
contrast
because,
viewed
in
this
direction,
they
are
very
thin in
comparison
to
the
rest
of
the
structure
.
The
^-220
A-diameter
particles
observed
in
Fig
.
3
are
not a
part
of
the
isolated
pore
complex
.
Pore
Complexes
Attached
to
Nuclear
Membranes
The
pore
complex
attached
to
the nuclear
membranes
(Fig
.
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

7)
has
additional
contrast
due
to
the
two
membrane
layers
which
are
contiguous
and
come
together
at
the pore
periphery
(33)
.
We
find that
the
borders of
these
membranes
are
made
conspicuous
by
incubating the nuclear
envelope
either
in
high
salt
medium
(400
mM
KCl)
or
in
low
salt
medium
in
which
the
MgCl2
is
replaced
by
1
mM
EDTA
(Fig
.
8)
.
The
former
treatment,
although
introducing
some
disorder,
largely
pre-
serves the
integrity
of
the
pore
complex
.
The
latter
treatment
leads
to
its
dissociation,
causing
also
changes
in
the
size
of
the
membrane
openings
.
Comparison
of Figs
.
7
and
8
shows
that
the
increased
clarity
of
the
membrane
borders
produced by
high
salt
or
EDTA
is
associated
with
the
detachment
of
particles
from around
the
FIGURE
6
Cand
D
perimeters
of
the pore
complex
and
from
intervening
spaces
.
The
effect
of
the
presence
of
these
particles
and
of
the
membrane on
the
appearance
of
the
projection
maps
is
signif-
icant
only
at
high
radius
.
In
projection
maps
calculated
from
pore
complexes
attached
to
the
nuclear
membranes,
Fig
.
9,
the
spokes
_have
more
matter
associated
with
them
in
their
outer
domain
than
previously
(Fig
.
6)
and
appear
against
a
stronger
background
density
(the
membrane)
beginning
abruptly
at
a
radius
of
^-450
A
.
Particles
around
Cytoplasmic
Perimeter
A
low
resolution
three-dimensional
map
of
a
pore
complex
attached
to
the
nuclear
membranes
was
calculated
from
a
series
UNWIN
AND
MILLIGAN

Particles
around
Nuclear
Pore
Complex

69
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
7

Isolated
nuclear
envelope
spanning
holes
in a
carbon
support
film
.
For
the
pore
complexes
encircled,
the
contributions
from
the
eightfold
harmonics
are
(top of
page
to
bottom)
:
23%, 21%, 24%, 27%, 25%, 20%, 15%, 25%, 32%,
and
32%
of
the
total
power
associated
with azimuthally
varying
components
.
The absence
of
a
central
plug
in
some
of
the
pore
complexes
may
be
a
consequence
of
the
isolation
procedure
.
The
inset
is
of
the
image from which
the
results
in Fig
.
10
were
obtained
.
Gold
thio-
glucose
stain
.
x
58,000
.
Inset,
x
73,000
.
of
images
taken with
different
tilts
(see
Materials
and
Methods)
to
observe
how
the
particulate
matter
superimposing
with
the
spokes
is
distributed
in
the
direction
of the octad
axis
.
Some
details
of
this
map
are
given
in
Fig
.
10
.
Central
sections
perpendicular
to
the
plane
of
the
membranes
(Fig
.
10 a)
show
a
marked
departure
from
the
putative
twofold
relationship
described
earlier
(Fig
.
5)
.
The
zone
contributed
by
the
plug
and
the inner
domain
of the
spokes
is
now
flanked,
at
high
radius,
by
matter
more
heavily
weighted
towards
the
cytoplas-
7
0

THE
JOURNAL
OF CELL
BIOLOGY
"
VOLUME
93,
1982
mit
half
.
This
additional
matter
gives
rise
to
strong
eightfold
modulations
in
sections parallel
to
the
membrane
plane
and
a
strong
asymmetry
in
terms ofthe
variation
in
eightfold
contrast
at
this
radius
with
distance
through
the structure
(Fig
.
10
b)
.
Since
both
sides
of
the
pore
complex
were
exposed
equally
to
the
stain
(see
Materials
and
Methods)
it
is
most
unlikely
that
the
asymmetry
is
due
to this
treatment
.
We
interpret the
details
to
indicate that
the
pore
complex
has,
on
its
cytoplasmic
perimeter,
particles-presumably
the
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
8

Isolated
nuclear
envelopes
after
incubation
in
(a)
in
high
salt
medium
(400
mM
KCI)
and
(b)
1
mM
EDTA
for
1
min
(see
Materials
and
Methods)
.
Loss of
matter
associated
with
these
treatments
exposes
the
membrane
borders
of
the
nuclear
pores
.
The
diameter
of
the
opening
delineated
by
these
borders
becomes
variable
after
treatment
with
EDTA
.
Gold
thio-glucose
stain
over
holes
in
the
carbon
support
film
.
x
45,000
.
same
as
those
in
Fig
.
3-which
overlay
the
membrane,
the
ring
and
the
spokes
.
It
is
unlikely
that
other
maps
would
show
other
additional
features
(e.g
.,
particles
also
on
the
nuclear
side)
since,
of the 58
pore
complexes
attached
to
membranes
which
we
analyzed
in
projection,
the
one
from
which
the
map
was
constructed
exhibits
the
best
power
spectrum
in
terms
of
reso-
lution
and
the
strongest eightfold
harmonic
.
On
the other
hand,
inspection
of
the
micrographs
suggests that
pore
complexes
often
have
less
than
eight
particles
on
their
perimeter,
and
sometimes
none
.
We
confirmed
the
presence
of
these
particles
on
the
cyto-
plasmic
side,
independent
of
the structure
analysis,
by
releasing
them
with high
salt
from
envelopes pressed
cytoplasmic
face
downwards
onto
polylysine-coated
carbon
films
(see
Materials
and
Methods)
.
The
"finger-prints"
thus
obtained
show up
to
eight
particles
arranged
in 1000
A
diameter
circles
(Fig
.
11)
.
Now
isolated
from
the
envelope
and
the
rings,
the
particles
are
easily
distinguished
from
the
round
central
plug,
on
the
basis
of
their
smaller
size
and
more
angular
shape,
but
seem
to
be
identical
to
the
particles
in the spaces
between
the
rings
.
They
also
correlate
closely
with
inactive
ribosomes
given the
same
treatments
(Fig
.
12),
in
terms
of
their
shape
and
size
.
DISCUSSION
The
nuclear
pore
complex
is
an
assembly
of
several
discrete
constituents
.
We
have
investigated
their
three-dimensional
or-
ganization
by
visualizing
them,
at
a
resolution
of
-90
A,
both
in
the
presence
of
and
isolated
from
the nuclear
membranes
.
We
used
Fourier
analysis
methods
to
evaluate
and
average
the
images
and
to
derive
three-dimensional
information
from
dif-
ferent
views
.
The
schematic
diagram,
Fig
.
13,
summarizes
our
results
.
We
find
the
pore
complex
to
be
a
symmetrical
structure
framed by
two
widely
separated,
coaxial
rings
.
The
rings
attach
to
the
two
nuclear
membranes
so
that
one
faces
the
nucleus
and
the
other the
cytoplasm
of the
cell
.
Connected
to
these
rings
and
extending
radially
inwards
from
them, along
a
central
plane,
are
elongated
structures
which
we
call
spokes
.
They
approach
and
appear
to
contact
a
central
large,
approximately
spherical
particle,
the
plug
.
Cytoplasmic
particles,
also
ob-
served
decorating
the
perimeter
of
many
ofthe
pore
complexes,
are
probably
not
integral
components
since
they
are
easily
detached
and
are
only
there
when
the nuclear
membranes
are
present
.
The
structural
framework
of the assembly,
i
.e
.,
the
rings,
the
spokes,
and
their
connecting
links,
appears
to
be
arranged with
octagonal
symmetry
about
the
central
axis
perpendicular
to
the
plane
of
the
membranes
and
with
twofold
symmetry
about
axes
lying
in
this
plane
(giving
the dihedral point
group
D
g
[822])
.
A
configuration
like
this,
in
which
the
two
halves
of the
assembly
face
in
opposite
directions,
would
account
simply
for
observations
of
single
pore
complexes
spanning
the
two
equiv-
alent,
but
oppositely
facing,
membranes
of
ER
cisternae
(13)
.
An
example
of
another
two-layered
membrane
system
incor-
porating
this
design
principle
would
be
the
gap
junction
(29)
.
Some
earlier
models
for
the
pore
complex
show
eight
equal
"granular
subunits"
around
its
perimeter
on
both
the
nuclear
and
cytoplasmic
sides
(12,
26)
and
so
may
be
construed
as
suggesting
a
twofold
symmetry
relationship
.
However,
the
quality
of
preservation
achieved
in
the
earlier
studies
was
not
assessed
and
alternative
interpretations
could
not
therefore
be
discounted
.
We
suggest
that
the
granular
subunits
and
likewise
the
proposed
tubular
subunits
(30)
or microcylinders
(35)
should
be
identified
with
the
detail
which
we
observe
in
the
region
where
the
spokes
connect
to
the
rings
.
It is
easy
to
see
how
this
detail
could be
interpreted
in
various
ways
according
to
the
state
of
preservation
.
The
features
we
find
in
the
periph-
eral
region of the
pore
complex
(Figs
.
4and
5)
are
particularly
distinct
because
we
have
exposed
them
by
taking
away
the
membranes
.
UNWIN AND
MILLIGAN

Particles
around
Nuclear
Pore
Complex

7
1
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
9

Typical
projection
maps
from
pore
complexes
attached
to
the
nuclear
membranes,
and
their
rotational
power
spectra
.
The
resolution
is
about
90
,$
.
The
shading
indicates
the
position
of
the
particulate
matter not
present
in
images
of
the
isolated
pore
complexes
.
The
broken
line
in
(A) indicates the
position of
the
membrane
border
.
The
view
is
from
the
cytoplasm towards
the
nucleus
.
Power
spectra
of
pore
complexes
attached
to
the
nuclear
membranes
show
8-
and
24-fold
harmonics
but the
16-fold
harmonic
is
always
very
weak
or
absent
.
Several
independent
lines
of
evidence
have
led
to
our
con-
clusion,
Fig
.
13,
that in
the
cell
there
are
sometimes
large
(-220
A-diameter)
particles
decorating the
pore
complex
around
its
perimeter
on
the
cytoplasmic
side
.
First,
such
particles
were
present
on
occasions
around
the
rings,
following
their
detach-
ment from
the
nuclear
envelope
with Triton
X-100
(Fig
.
3)
.
Second,
these
particles
were
too
large
to be
accommodated
as
part
of
the structure
of
the
pore
complex
itself,
yet did
give
rise
to
additional
density
over
the
rings
and
spokes
of
pore
com-
plexes
which
had
not
been
detached
from
the
nuclear envelope
.
Third,
we
demonstrated
by
a
three-dimensional
analysis
that
the
additional
density
was
concentrated
toward
the
cytoplasmic
side
rather
than
the
central
plane
or the
side
of the
nucleus
.
72

THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
93,
1982
Fourth,
we
were
able to detach
these
particles (identified
by
their
appropriate
circular
configuration)
by
contact
of
the
cytoplasmic
surface
of
the nuclear
envelope
against
the
micro-
scope
grid
.
At
least
two
reports
(6,
17)
have
clearly
demonstrated
deco-
ration
of the
pore
perimeter
by
particles
which might
be
the
same
as
those
we
observe
.
In
other
careful studies
(e
.g
.,
refer-
ence
31),
such
particles
have
not
been
detected,
despite
condi-
tions
being
used
which
should
have
allowed
their
retention
.
Thus
the
particles
appear
to
be
there
in
vivo
on
some
occasions
but not
on
others
.
Accordingly,
their
presence
may
be
related
to
a process
such
as
the
activity
of the pore
in
mediating
transfer
of
molecules
between
nucleus
and
cytoplasm
.
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
10

Details
of
a
low
resolution
three-dimensional
analysis
of
a
pore
complex
attached
to
the
nuclear
membranes
.
(a)
A
projection of
25-X
thickness
through
the
three-dimensional
map
built
up from
sections
perpendicular
to
the
plane
of
the
membranes
and
intersecting
the
maxima
in
the
eightfold
density
modulations
.
The
full
lines
are
the
positive
contours
indicating
the
regions
where
the
biological
matter
is
concentrated
.
The broken
lines
are
negative
contours
.
The
shading
indicates
the
estimated
positions
of
the
twofold
related
features
described
earlier
(Fig
.
5)
.
The two
membrane
layers
are not
resolved
.
Features
on
the
cytoplasmic
side
give
rise
to strong
eightfold
modulations,
but
those
on
the
nuclear
side
do
not
.
(b)
A
plot
of
the
variation
in
contrast
associated
with
the
eightfold
modulations
at a
radius of
500
A
.
The
vertical
scale
is
the
same
as
in a
.
The
contrast
was
estimated
as
the
difference
between
the
maximum
and
minimum
densities
in
sections
perpendicular
to
the
octad
axis
.
The
line at
the
500-A
radius
along
which
the densities
were measured
is
indicated
in
the
section
giving
maximum
contrast,
at
the
bottom
of
the
figure
.
There
is
marked
asymmetry
in
contrast
compared
to
similar
views
from detached
pore
complexes
.
This
analysis
gives
only
a
qualitative
idea of
the
relative
levels
and
strengths
of features
.
The
Fourier
terms
giving
the
variation
in
mean
density
along
the
direction
of
the
octad
axis
are
not
included
.
Dimensions
of features
in this
direction
are
also
affected
by
flattening
distortions
(19)
.
Because
of
the
variable
preservation
of
the
central
plug,
we
attach
no
significance
to
details
in this
region
.
UNWIN AND
MILLIGAN

Particles
around
Nuclear
Pore
Complex

73
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURES
11 and
12

"Finger-print" of material
detached from
the
cytoplasmic
surface
of
the
nuclear
envelope
(top of
picture)
onto
a
polylysine-coated
carbon
support
film
.
The
detachment
was
achieved
using
high
potassium
concentration
(400
mM
;
see
Methods
and
Materials)
.
The
somewhat
angular
particles
arranged
in
rings
(circle)
or
randomly
(square)
can be
identified
with
those
in Fig
.
7
around
the
perimeter
of
the
pore
complexes
and
in
intervening
spaces
.
These
particles
are
easily
distinguished
from
the
larger,
round,
central
plug
(arrows)
.
Gold
thio-glucose
stain
.
x
25,000
.
Inset,
x
55,000
.
Fig
.
12
:
Ribosomes
isolated
from
hypothermic
chick
embryos,
after
incubation
in
high
salt,
fixation,
and
staining
as
in
Fig
.
11
.
Both
the
tetrameric
and
single
ribosomes
appear
rather
angular
(arrows),
possibly
as a result
of
detachment
of
the
small
subunit
.
x
55,000
.
74

THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
93,
1982
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982

FIGURE
13
Diagram
of
the
nuclear
pore
complex
:
(a)
In
central
cross-section
and
(
b) in
projection
down
the
octad
axis
.
The
major
constituents
are
the
central
plug (C),
the
spokes
(S),
and
the
rings
(R)
.
The
spokes
are
connected
to
the
rings
near
to
the
maximum
radius
of
the
assembly
(-600
A)
.
Superposition
of
detail
in
projec-
tion gives
rise
to
a
characteristic
hollow
or
bilobed
region of
density
near
the
outer
extremity
of
the
spokes
(see
Fig
.
6)
.
The
broken
lines
outline
the
positions of
the
additional
features
present
with pore
complexes
embedded
within
the
nuclear
envelope
:
the
two
mem-
brane
layers
which
come
together
at
the
pore
complex
and
octag-
onally
arranged
particles
(P),
resembling
ribosomes
.
The
position of
the
membrane
border
in
a
corresponds
with
the
thin-section
view
(e
.g
.,
reference
6)
.
The
particles
are
easily
detached
and
are
not
always
present
.
Our
observations
on
the
character
of
these
particles
are
consistent
with
the
surmise
that
ribosomes
(21,
31),
or
alter-
natively
ribosomal
precursors,
are
sometimes
associated
with
the
nuclear
pore
complex
.
The
particles
resemble
inactive
ribosomes
stained
under
identical
conditions (Figs
.
11
and
12)
.
They
also
resemble
the
membrane-bound
particles
in
the
spaces
between
the
pores,
which
from
studies
of
nuclear
envelopes
in
other
cells
one
would
presume
to
be
ribosomes
(25,
31,
32)
.
That
they
are
detached
from
theperiphery
of
the
pore
complex
by
the
same
biochemical
treatments
(addition
of
Triton X-100,
EDTA,
or
a
high
concentration
of
potassium
ions)
that
release
inactive
ribosomes
from
membranes
in
secretary
cells
(2,
27)
could
reflect
the
fact
that
they
are
attached
directly to
the
nuclear
membrane
in
the
immediate
vicinity
of
the
pore
com-
plex
and
interact
only
weakly
with
the
pore
complex
itself
.
We
are
grateful
to
John
Gurdon and
Rick
Bram
for
generously
providing
us
with
oocytes,
and
to
Tony
Crowther
and
Linda
Amos
for
the use
of
their
rotational
averaging
computer
program
.
We
thank
Roger
Karnberg,
John
Murray,
and
John
Kilmartin
for
valuable
discussions
and
comments
.
Guido
Zampighi
suggested
using
gold
thio-
glucose
as
a
negative
stain
.
Received
for
publication
14 July
1981,
and
in
revised
form
21
October
1981
.
REFERENCES
I
.
Aaronson,
R
.
P
.,
and
G
.
Blobel
.
1975
.
Isolation
of
nuclear
pore
complexes
in
association
with
a
lamina
.
Proc
.
Nail
.
Acad
.
Scl
.
U,
S
.
A
.
72
:1007-1011
.
2
.
Adelman,
M
.
R
.,
D
.
D
.
Sabatini,
and
G
.
Blobel
.
1973
.
Ribosome
membrane
interaction
.
Nondestructive
disassembly
of
rat
liver
rough
microscmes
into
ribosomal
and
membranous
components
.
J
.
Cell Biol
.
56
:202-229
.
3
.
Arndt,
U
.
W
.,
J
.
Barrington-Leigh,
J
.
F,
W
.
Mallet,
and
K
.
E
.
Twinn
.
1969
.
A
mechanical
microdensitometer
.
J
.
Sci
.
Jnstrurn
.
2
:385-387
.
4
.
Bonner,
W
.
M
.
1975
.
Protein
migration
into nuclei
I .
Frog
oocyte
nuclei
in
vivo
accumulate
microinjected
histones,
allow
entry
to
small
proteins,
and
exclude
large
proteins
.
J
.
Cell
Biot
64
:421-430
.
5
.
Crowther,
R
.
A
.,
and L
.
A
.
Amos
.
1971
.
Harmoni
c
analysis of
electron
microscope
images
with
rotational
symmetry
.
J
.
Mot
.
Biol
.
60
:123-130
.
6
.
Daniels
.
E
.
W
.,
J
.
M
.
McNiff,
and
D
.
R
.
Ekberg
.
1969
.
Nucleopore
s
of
the
giant
amoeba,
Pelomyxa
carolinensis
.
Z
.
Zeltforsch
.
Mikrosk
.
Anal
.
98
:357-368
.
7
.
DeRobenis, E
.
M
.,
R
.
F
.
Longthorne,
and
J
.
B
.
Gurdon
.
1978
.
Th
e
intracellular
migration
of
nuclear
proteins in
Xenopus
oocytes,
Nature
(fond)
.
272
:254-256
.
8
.
DeRosier,
D
.
J
.,
and P
.
B
.
Moore
.
1970
.
Reconstructio
nof
three-dimensional
images
from
electron
micrographs
of
structures
with
helical
symmetry
.
J
.
Mot
Biol
.
52
:355-369
.
9
.
Faberg6,
A
.
C
.
1973
.
Direct
demonstration
of
eightfold
symmetry
in
nuclear
pores,
Z
.
Zet
forsch
.
Mikrosk
.
Anal
.
136
:183-190
.
10
.
Feldherr,
C
.
M
.,
and
1
.
M
.
Marshall
.
1962
.
Th
e
use
of
colloidal
gold
for
studies
of
intracellular
exchanges
in
ameba
Chaos
chaos
.
J
.
Cell Biol
.
12
:641-6645
.
11
.
Finch,
1
.
T
.,
and
A
.
Klug
.
1965
.
The
structure
of
viruses
of
the
Papilloma-Polyoma
type
.
III
.
Structure
of rabbit
Papilloma
virus
.
J
.
Mol
.
Biol
.
13
:12
.
12
.
Franke,
W
.
W
.,
and
U
.
Scheer
.
1970
.
The
ultrastructure
of
the
nuclear
envelope
of
amphibian
oocytes
:
a
reinvestigation
.
J
.
Ultrastrucr
.
Res
.
30
:288-316
.
13
.
Franke,
W
.
W
.,
U
.
Scheer,
and
H
.
Fritsch
.
1972
.
Intranuclear
and
cytoplasmic
annulate
lamellae
in
plant
cells
.
J
.
Cell Biol
.
53
:823-827
.
14
.
Gall,
J
.
G
.
1967
.
Octagonal
nuclear
pores
.
J
.
Cell
Biol
.
32
:391-399
.
15
.
Gunton,
1
.
B
.
1976
.
Injected nuclei
in
frog
oocytes
:
fate,
enlargement,
and
chromatin
dispersal
.
J
.
Embryol
.
Exp
.
Morpho
l
36
:523-540
.
16
.
Gurdon,
J
.
B
.
1970
.
Nuclear
transplantation
and
the
control
of
gene
activity
in
animal
development
.
Proc
.
R
.
Soc
.
Lond
.
B
Biol
.
176
:303-314
.
17
.
Hoeijmakers,
J
.
H
.
J
.,
J
.
H
.
N
.
Schel,
and F
.
Wanka
.
1974
.
Structure
of
the
nuclear
pore
complex
in
mammalian
cells
.
Two
annular
components
.
Exp
.
Cell
Res
.
87
:195-206
.
18
.
Huxley,
H
.
E
.,
and
G
.
Zubay
.
1960
.
Electron
microscope
observations
on
the
structure
of
microsomal
particles
from
Escherichia
coli
.
J
Mot
.
Biol
.
2
:10-18
.
19
.
Krohne,
G
.,
W
.
W
.
Franke,
and
U
.
Scheer
.
1978
.
The
major
polypeptides
of
the
nuclear
pore
complex
.
Exp
.
Cell
Res
.
116
:85-102
.
20
.
Martini,
D
.
H
.
W
.,
and
H
.
J
.
Gould
.
1975
.
Molecula
r
weight
distribution
of
ribosomal
protein
from
several
vertebrate
species
.
Mot
Gen
.
Genet
.
142
:317-331
.
21
.
Mepham,
R
.
H
.,
and
G
.
R
.
Lane
.
1969
.
Nucleopore
s
and
polyribosome
formation
.
Nature
(Land
.)
.
221
:288-289
.
22
.
Moody,
M
.
F
.
1971
.
Application
of
optical
diffraction
to
helical
structures
in
the
bacterio-
phage
tail
.
Philos
.
Trans
.
R
.
Soc
.
Lond
.
B
Biol
.
Sci
.
261
:181-195
.
23
.
Morimoto,
T
.,
G
.
Blobel,
and
D
.
D
.
Sabatini
.
1972
.
Ribosom
e
crystallization
in
chicken
embryos
.
1
.
Isolation,
characterization,
and
is vivo
activity
of
ribosome
tetramers
.
J
.
Cell
Biol
.
52
:338-354
.
24
.
Paine,
P
.
L
.,
L,
C
.
Moore, and
S
.
B
.
Horowitz
.
1975
.
Nuclea
r
envelope
permeability
.
Nature
(Land
.)
.
254
:109-114
.
25
.
Palade,
G
.
E
.
1955
.
A
small
particulate
component
of
the
cytoplasm
.
J
.
Biophys
.
Biochem
.
Cyto
l
1
:59-68
.
26
.
Roberts,
K
.,
and
D
.
H
.
Northcote
.
1970
.
Structur
e
of
the
nuclear
pore
in
higher
plants
.
Nature
(Land
.)
.
228
:385-386
.
27
.
Sabatini,
D
.
D
.,
Y
.
Tashiro,
and
G
.
E
.
Palade
.
1966,
On
the
attachment
of
ribosomes
to
microsomal
membranes
.
J
Mol
.
Biol
.
19
:503-524
.
28
.
Scheer,
U
.
1972
.
The
ultrastructure
of
the
nuclear
envelope
ofamphibian
oocytes
.
IV
.
On
the
chemical
nature
of
the
pore
complex
material
.
Z
.
Zeltforsch
.
Mikrosk
.
Anal
.
127
:127-
148
.
29
.
Unwin,
P
.
N
.
T
.,
and
G
.
Zampighi
.
1980
.
Structure
of
the
junction
betweencommunicating
cells
.
Nature
(Load
.)
.
283
:545-549
.
30
.
Vivier,
E
.
1967,
Observations
ultrastructurates
sur
l'envelope
nucleaire
et
ses
pores
chez
des
sporozoaires
.
J
.
Microscopie
.
6
:371-390
.
31
.
Watson,
M
.
L
.
1959
.
Further
observations
of
the
nuclear
envelope
of
the
animal
cell
.
J
.
Biophys,
Biochem
.
Cytol
.
6
:147-156
.
32
.
Watson,
M
.
L
.
1955
.
The
nuclear
envelope
.
Its
structure
and
relation
to
cytoplasmic
membranes
.
J
.
Biophys
.
Biochem
.
Cytol
. 1
:257-270
.
33
.
Watson,
M
.
L
.
1954
.
Pores
in
the
mammalian
nuclear
membrane
.
Bischim
.
Biophys
.
Acts
.
15
:475-479
.
34
.
Williams,
R
.
C
.
1977
.
Use of
polylysine
for
adsorption
of
nucleic acids
and enzymes
to
electron
microscope
specimen
films
.
Proc
.
Nail
.
Acad
.
Sci
.
U
.
S
.
A
.
74
:2311-2315
.
35
.
Wischnitzer,
S
.
1958
.
An
electron
microscope study
of
the
nuclear
envelope
ofamphibian
oocytes
.
J
.
Uhrastruct
.
Res
. 1
:201-222
.
UNWIN
AND
MILLIGAN

Particles
around
Nuclear
Pore
Complex

75
on July 14, 2011jcb.rupress.orgDownloaded from
Published April 1, 1982