Download Runions et al - Oxford Academic

Journal of Experimental Botany, Vol. 57, No. 1, pp. 43–50, 2006
doi:10.1093/jxb/eri289 Advance Access publication 5 October, 2005
FOCUS PAPER
Photoactivation of GFP reveals protein dynamics within the
endoplasmic reticulum membrane
John Runions*, Thorsten Brach†, Sebastian Kühner† and Chris Hawes
Biological and Molecular Sciences, Oxford Brookes University, Oxford OX3 0BP, UK
Received 31 May 2005; Accepted 15 August 2005
Abstract
Components of the plant cell secretory pathway, including the endoplasmic reticulum and Golgi apparatus, are in constant motion. The photoactivation of GFP
has been used to determine that proteins within the
membrane of the ER flow as the ER is remodelled.
Measurement of the rate at which activated GFP moves
away from the activation spot shows that this motion is
much faster than would be expected if membrane
components moved simply by diffusion. Treatment
with latrunculin to depolymerize the actin cytoskeleton
stops ER remodelling and reduces the rate of GFP
movement to that expected from diffusion alone. This
suggests that myosin binds directly or indirectly to ER
membrane proteins and actively moves them around
over the actin scaffold. Tracking of Golgi body movement was used to demonstrate that they move at the
same rate and in the same direction as do photoactivated ER surface proteins. Golgi bodies, therefore,
move with, and not over, the surface of the ER. These
observations support the current theory of continuity
between Golgi bodies and discrete ER exit sites in the
ER membrane.
Key words: Actin cytoskeleton, endoplasmic reticulum, GFP,
Golgi, photoactivatable GFP.
Introduction
The endoplasmic reticulum (ER) and Golgi apparatus are in
constant, actin-dependent motion within the cortical cytoplasm of leaf epidermal cells of Nicotiana tabacum (Quader
et al., 1987; Boevink et al., 1998). Newly folded proteins
move along the anterograde pathway from the ER to Golgi
bodies for modification and secretion to their cellular
destination. Theoretical mechanisms of ER/Golgi body association have been revised and updated frequently (Staehelin
and Moore, 1995; Batoko et al., 2000; Nebenführ and
Staehelin, 2001; Robinson, 2003; Hawes, 2005). Obvious
physical continuity is seldom observed between the ER and
Golgi stacks at the light or electron microscope level, but
there are reports of tubular connections in some cell types
(Brandizzi et al., 2002). A major question to be resolved is
whether the entire ER surface is competent to export proteins
which are then picked up by motile Golgi stacks (Boevink
et al., 1998) or whether the Golgi stacks stop at discrete ER
exit sites (ERES) as proposed by Nebenführ et al. (1999)?
Recent evidence suggests that, in fact, ERES are in connection with the individual Golgi stacks either permanently
(daSilva et al., 2004), or at least transiently (Yang et al.,
2005) and that the ERES/Golgi stack complex moves over
or within the surface of the ER.
Fluorescent protein fusions to structural proteins of the
secretory pathway are an ideal way to study the dynamic
nature of the system (Brandizzi et al., 2004). When green
fluorescent protein (GFP) is targeted to the ER and retained
due to an HDEL signal (Boevink et al., 1996; Haseloff
et al., 1997; Batoko et al., 2000), a polygonal, reticulate
network becomes visible in the cortical cytoplasm of living
cells. This network remodels relatively slowly compared
with the fast moving strands of ER that radiate through it in
a co-planar fashion or that traverse the cell as trans-vacuolar
strands. GFP modified thus, remains within the lumen of
the ER and is, therefore, not a useful tool for observation of
the ER membrane as a separate entity. What is required is
an ER membrane-bound fluorescent protein that will enable
monitoring of ER surface movement contemporaneously
with tracking of Golgi stacks. Such a system would allow
investigation as to whether Golgi bodies move with, or
over the ER surface and, ultimately, would help define the
mechanism at the ER/Golgi interface more precisely.
* To whom correspondence should be addressed. E-mail: [email protected]
y
Present address: Department of Cell Biology, Heidelberg Institute for Plant Sciences, University of Heidelberg, D-69120 Heidelberg, Germany.
ª The Author [2005]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved.
For Permissions, please e-mail: [email protected]
44 Runions et al.
To this end, photoactivatable GFP (PAGFP; Patterson and
Lippincott-Schwartz, 2002) fused to the trans-membrane
domains of Arabidopsis thaliana calnexin (Huang et al.,
1993; Irons et al., 2003) have been utilized. Irons et al.
(2003) observed that calnexin overexpression in the ER
membrane results in flat sheets of ER that are not normally
observed in GFP-HDEL expressing plants. These sheets
remain connected in a network by normal-appearing ER
tubules and plants appear unaffected by the change in ER
morphology. When activated, the calnexin-PAGFP (CXPAGFP) construct fluorescently marks the ER membrane
and, if a small enough region of membrane has been
activated, its movement can subsequently be tracked by
time-lapse imaging. The Golgi apparatus was marked at
the same time using a construct consisting of the transmembrane domain of a rat sialyl transferase (Boevink
et al., 1998) fused to red fluorescent protein (mRFP). Two
different software approaches were required to correlate
movement of the ER membrane (a diffuse structure), and
Golgi bodies (discrete structures). Ultimately, it was possible to quantify the speed and direction of movement of each.
Materials and methods
Marking the ER membrane with photoactivatable GFP
PAGFP was fused to the 39 end of the calnexin trans-membrane
domain to produce CX-PAGFP. In this orientation, PAGFP is
anchored to the ER membrane but is located within the cell cytoplasm
which makes it more readily photoactivatable. The calnexin fragment
was PCR amplified from the vector pVKH18En6 spGFP5CX (Irons
et al., 2003) using the primers JR023 (forward GCTTCTAGAGGTATGATCACGGAACTGATTGAGAAAGCCGAG) to add an
Xba1 restriction site and an ATG start codon, and JR008 (reverse
TGAGGATCCGATATTATCACGTCTCGGTTGCCTTTTGCG) to
remove the TAG stop codon and add a BamH1 restriction site (264
bp). PAGFP (provided by George Patterson and Jennifer LippincottSchwartz) was PCR amplified with the primers JR016 (forward
GCTGGATCCGGTGTGAGCAAGGGCGAGGAGCTGTTC)
to
add a BamH1 restriction site and remove the ATG start codon, and
JR013 (reverse AGCGAGCTCTCTTTACTTGTACAGCTCGTCCATGCC) to add a Sac1 restriction site (788 bp). These two PCR
fragments were digested and subcloned into pGEM-3Z (Promega) to
make pGEM-3Z CX-PAGFP. The CX-PAGFP fragment was PCR
amplified for Gateway cloning (Invitrogen) using the primers JR024
(forward GGGGACAAGTTTGTACAAAAAAGCAGGCTTCAAAAAAATGATCACGGAACTGATTGAGAAAGCCGAG)
and
JR025 (reverse GGGGACCACTTTGTACAAGAAAGCTGGGTCTCACTTGTACAGCTCGTCCATGCCGAGAGT) to add forward
and reverse aatB recombination sites, respectively, as well as a Kozak
consensus sequence (AAA AAA) immediately upstream of the start
codon (1119 bp). This PCR fragment was recombined with the
Gateway pDONR207 vector in a BP reaction to produce pENTR207
CX-PAGFP. Recombination of this vector with the Gateway plant
binary vector pMDC32 (Curtis and Grossniklaus, 2003) in an LR
reaction produced pMDC32 CX-PAGFP. Expression of CX-PAGFP
is under control of a single CaMV 35S promoter in this vector.
Marking Golgi bodies with mRFP
In the plant binary vector pVKH18En6 ST-mRFP (provided by
Federica Brandizzi), the GFP of pVKH18En6 ST-GFP (Saint-Jore
et al., 2002) is replaced with monomeric RFP (provided by Roger
Tsien). ST-mRFP expression is under the control of 63 tandemlyrepeated CaMV 35S promoters. ST-mRFP marks not only the Golgi
bodies but the excess is secreted and fills the apoplastic space
between leaf epidermal cells, therefore outlining them, as well.
Plant transformation
4–6-week-old plants of Nicotiana tabacum SR1 cv. Petit Havana
were transiently transformed with fluorescent constructs following
the protocol of Batoko et al. (2000). Briefly, Agrobacterium
tumefaciens GV3101-pMP90 were transformed by heat-shock with
the binary vectors described above. Colonies were grown to mid-log
phase in 10 ml YEB medium at 28 8C with 100 mg l1 kanamycin
selection. The culture was pelleted and washed twice in infiltration
medium [50 mM MES, pH 5.6, 2 mM Na3PO4, 0.5% D-glucose
(w/v), and 100 lM acetosyringone (Aldrich)]. Resuspended bacteria
were diluted to an optical density (k=600 nm) of 0.1 and 0.05 for CXPAGFP and ST-mRFP, respectively. Small needle marks were made
in the leaves and diluted bacteria were injected into the leaf lamina
through the lower epidermis using a 1 ml syringe (without needle)
and gentle pressure. When co-expression of the PAGFP and mRFP
containing constructs was desired, the bacteria were mixed in
appropriate volumes of infiltration buffer prior to injection into the
leaf. Fluorescent protein expression was studied in lower epidermal
cells between 2–6 d after transformation.
Confocal microscopy
Confocal imaging and activation of photoactivatable GFP was done
with a Zeiss LSM 510 META system. PAGFP emits very little
fluorescence at 510 nm by excitation with the 488 nm line of a argon
laser prior to activation. A 25 mW blue diode 405 nm laser was used
at fairly high output (25–50% transmission) to target small regions (2
lm diameter) of the endoplasmic reticulum using the photobleaching
function of the Zeiss software in time-lapse mode. Generally, 3–5
pulses of the 405 nm laser were sufficient to activate PAGFP so that it
produced very bright fluorescence emission which was detected by
excitation at 488 nm using a 500–530 nm band pass filter. The 488
nm line of the argon laser was used at a very low transmission
percentage (3.2%) for time-lapse imaging of the photoactivated GFP
and in tests on fully activated cells this resulted in no appreciable
photobleaching during the time-course (1–2 min) of most experiments. Golgi body mRFP fluorescence was excited with a 543 nm
HeNe laser and emission was detected using a 565–615 band pass
filter. Most of the images used for subsequent analyses were made
using a 633 1.4 NA oil immersion objective. Pieces (5–10 mm2)
were cut out of the transformed leaf lamina and mounted under a 50
mm long coverslip in water so that the lower epidermis was facing
upwards. The majority of time-lapse imaging of the cortical ER in this
study was done in two dimensions, i.e. time resolved x–y. Movement
of the photoactivated ER membrane was analysed for 15 cells in
which the actin cytoskeleton was functioning and for 5 cells in which
it had been depolymerized by treatment with latrunculin B (see
below).
Latrunculin B treatment
Leaf pieces cut out as described above were placed into 25 lM
latrunculin B (Calbiochem) in an Eppendorf tube for 1 h prior to
imaging to depolymerize the actin cytoskeleton. The working solution of latrunculin B was made immediately prior to use from a frozen
stock solution (10 mM in DMSO).
Image analysis—photoactivated ER membrane
Time-lapse series of the photoactivated GFP were analysed using the
Radial Profile Plot plug-in of ImageJ (freeware available from the
ER membrane dynamics 45
NIH at http://rsb.info.nih.gov/ij/). This plug-in produces a profile plot
of normalized integrated intensities around concentric circles as
a function of distance from a point in the image. Movement of
activated ER membrane away from the activation spot was recorded
as an increase in intensity at a distance from the centre spot or as
a decrease in intensity at the centre spot in time. Intensity at a given
diameter was used to plot intensity decay curves for the time period
from peak intensity after activation. For this report, the curve that
contained the highest post-activation normalized intensity (see
below) was used in the analysis of each activation. Fitting these
curves using SPSS software (SPSS Inc.) enabled the dynamic nature
of ER membrane movement to be estimated. As a first approximation,
a decay curve that is well fit by an exponential equation:
Results
Supplementary movies showing the photoactivation of
PAGFP in the ER membrane can be found at JXB online.
Movie 1: rapidly remodelling ER membrane (Fast ER
membrane (B019-6).avi). Actual elapsed time 40.9 s.
(4.6 Mb). Movie 2: slowly remodelling ER membrane
(Slow ER membrane (B022-8).avi). Actual elapsed time
97.1 s. (3.3 Mb). Movie 3: non-remodelling ER membrane
after treatment with latrunculin B (Latrunculin B ER
membrane (B024-9).avi). Actual elapsed time 120.1 s.
(7.2 Mb).
ðb1ÞðtÞ
f ðtÞ = b0 exp
where b0 and b1 are parameters of the curve, can represent diffusive
movement. When the decay is better fit by other curves, for example,
the inverse curve:
f ðtÞ = b0 + ðb1 =tÞ
it is likely that other factors, for example, energetic processes, are
involved in the movement. Raw intensity data was normalized across
all data sets using the following equation:
In = ½ðIt Imin Þ=ðImax Imin Þ3100
where In is the normalized intensity, It is the intensity at time t, Imin is
the minimum intensity in the data set, and Imax is the maximum
intensity in the data set.
Various curve-fitting models were used to fit the normalized
intensity data and statistics were generated from the equation with the
highest r2 value in each case. The speed of CX-PAGFP movement
was estimated by calculating the half-time of intensity decay (t1/2)
using different equations based on the best curve fit in each case:
(i)
Exponential curve fit: t1=2 = ½lnðI 1=2 Þ lnðb0 Þ=b1
where I1=2 = ðImax Imin Þ=2
(ii) Inverse curve fit: t1=2 = b1 =ðI1=2 b0 Þ
Finally, the mobile fraction (A) of fluorescent protein in the
membrane was estimated by calculating the intensity difference
between photoactivation and the end of the time series:
A = Ia In
where Ia is activation intensity (normalized to 100%), In is intensity at
the final time point.
Image analysis-tracking Golgi body movement
Movement of individual Golgi stacks was analysed using the
classification and particle tracking routine of Volocity version 3.0
(Improvision). This software can track the movement of individual
fluorescent particles in time-resolved 2-, or 3-D. In 5/15 cases where
photoactivation of the PAGFP in the ER membrane was analysed,
Golgi body movement was also analysed. All Golgi bodies within
a large region of interest, that extended at least in a 12 lm radius
beyond the activation spot, were analysed. Time series images were
filtered to remove background and thresholding was used so that the
software could reliably track moving Golgi bodies. Various relevant
statistics about Golgi movement including velocity, track length, and
an index of movement pattern (the ‘meandering’ index) were
generated automatically. In addition, the Volocity software was used
to generate vectorial maps of Golgi movement within the region of
CX-PAGFP activation so that comparisons were possible between
the direction of GFP movement and Golgi body movement.
Photoactivation of the ER membrane: control
GFP-HDEL highlights the reticulate nature of the cortical
ER in cells of the lower leaf epidermis of N. tabacum (Fig.
1a). Reticulate-patterned areas of ER remodel relatively
slowly while trans-vacuolar strands that are seen traversing
the cell flow quickly. Golgi bodies associate with the ER
reticulae and strands (Fig. 1a). Overexpression of CXPAGFP in the ER membrane resulted in a change in ER
morphology such that flat sheet-like regions of ER were
observed (Fig. 1b). Sheets of ER were connected by tubules
that appeared similar to those of the normally reported ER
network. Because the ER is not visible prior to photoactivation of the PAGFP in its membrane, only about 75%
of the cells that were targeted yielded analysable results.
The other cells either had not been transformed with the
CX-PAGFP construct or the ER was not in the focal plane
before the scanning began. Photoactivated GFP in the ER
membrane was found to move away from the site of laser
activation at various speeds (Fig. 1c–j). t1/2 values of
intensity decay ranged from 5.12 to 18.57 s (mean
t1/2=9.2865.39 s, n=15) (Table 1). Visual inspection of
the ER morphology in the activated region indicated
a strong correlation between ER morphology and t1/2.
Fast intensity decay occurred when the targeted area of
membrane was within a fast-moving strand of cytoplasm
and slow intensity decay occurred in areas of tubular and
planar ER that appeared static. The salient fact arising
from this observation is that the ER membrane is a fluid
structure in which CX-PAGFP is mobile. More importantly, the type of curve equation used to derive t1/2 values
gave clues as to the mechanism of membrane movement.
When PAGFP in the ER membrane was fast moving, the
intensity decay curve was always fit better by the inverse
equation (e.g. r2 range 0.82–0.97) than by the exponential
decay equation (Fig. 2a) (n=13). Only in two cases of
the slowest moving PAGFP, did the exponential decay
curve provide better fit (r2=0.96 and 0.97) (Fig. 2b). The
mobile fraction (A) of PAGFP in the activated region (i.e.
the amount of activated GFP that moved out of the
activated region during the course of observation)
ranged from 88.97% to 100.00% (mean A=94.3264.52,
n=15).
46 Runions et al.
Fig. 1. GFP marks the endoplasmic reticulum (green) and RFP marks Golgi bodies (ST-mRFP; red) in tobacco leaf abaxial epidermal cells. (a) The
cortical ER network marked by GFP-HDEL has two different spatial domains; the reticulate ER which is slow to remodel and fast moving strands
(arrowheads) that are co-planar with the reticulate net or that traverse the central vacuole. Golgi bodies are closely associated with the ER tubules and
transvacuolar strands. Scale bar=20 lm. (b) Sheet-like ER results from over-expression of GFP fused to the trans-membrane domains of calnexin (GFPCX). Golgi bodies remain within lacunae of the ER for the most part, but occasionally are seen traversing the ER sheets. (c–n) Time-series images
showing dispersal of photoactivatable GFP (CX-PAGFP) in the ER membrane after activation using a 405 nm laser. Scale bar=5 lm. (c–f ) Fast moving
ER. When the photoactivated region of CX-PAGFP in the ER membrane is within a transvacuolar strand, complete dispersal of the GFP is very quick.
(g–j) Slowly moving ER. GFP in sheet-like regions of ER membrane resulting from overexpression of CX-PAGFP tended to disperse much more slowly
than GFP in transvacuolar strands when photoactivated. (k–n) latrunculin B-treated cells. When the actin cytoskeleton was depolymerized by latrunculin
B treatment, remodelling of the ER ceased and Golgi bodies stopped moving, but the membrane continued to flow. Dispersal of photoactivated GFP was
slower than in the non-treated example above (g–j). Yellow circle, activation spot.
Photoactivation of the ER membrane: latrunculin B
When cells were treated with latrunculin B to depolymerize
the actin cytoskeleton prior to photoactivation of CXPAGFP, the exponential decay equation always gave better
fit to intensity data than the inverse equation (Figs 1k–n,
2c). t1/2 of intensity decay values ranged between 17.94 and
33.33 s (Table 1) (n=5). The mean latrunculin B t1/2 value
was significantly longer than in the control case (23.816
8.32 versus 9.2865.39 s, respectively, P <0.05). The
mobile fraction (A) of PAGFP in the activated region after
latrunculin B treatment ranged from 79.36% to 95.46%
(mean A=88.6766.74%). There was no difference in the
mobile fraction between the control and latrunculin B
treatment.
Golgi body movement
When all of the ER membrane in a cell was photoactivated
at once, individual Golgi bodies were observed to move in
close association with the ER tubules or along the edges of
ER sheets. Golgi bodies were often enclosed within lacunae
of the ER sheets (Fig. 1b), but were mobile as the lacunae
remodelled quickly. Golgi bodies within a region centred
ER membrane dynamics 47
Table 1. Summary statistics for movement of CX-PAGFP within the ER membrane as a function of photoactivatable GFP intensity
change and for Golgi body tracking
t1/2 is the half-time of fluorescence decay after PAGFP activation. Mobile fraction is a measure of the amount of activated PAGFP in the ER membrane
that moved during the course of data collection. Golgi body Meandering Index is derived as displacement/track length and describes the ‘straightness’ of
a track. Control, n=15 cells. Latrunculin B, n=5 cells.
Photoactivatable GFP
t1/2 (s)
Golgi body tracks
Mobile fraction (%)
Control
Latrunculin B
Control
Latrunculin B
9.2865.39
23.8168.32a
94.3264.52
88.6766.74
a
Velocity
(lm s1)
Track length
(lm)
Displacement
(lm)
Meandering
Index
0.2560.39
4.0161.24
2.3460.80
0.5560.10
P <0.05.
on the activation spot were tracked with Volocity version
3.0 (Fig. 3a) (n=5 cells). Several important statistics were
generated including: total number of tracks (178666), track
length (4.0161.24 lm), velocity (0.2560.39 lm s1),
displacement, i.e. linear distance between start and end
point (2.3460.80 lm), and Meandering Index, i.e. displacement/track length (0.5560.10) (Table 1). Large variation in the speed of different Golgi bodies within the same
cell was responsible for the high standard deviation of
velocity. The Meandering Index is derived as a function of
particle displacement (i.e. linear distance between first and
last measured position) and track length. High values
indicate straight tracks and low values indicate a high
degree of direction changing along a track. The software
also rendered movement of individual Golgi bodies pictorially such that relative direction and speed were represented
by vectors of varying length (Fig. 3b, c). In each plot, Golgi
tracks are shifted in x–y so that they appear to share
a common origin. It was possible to inspect the actual track
shape (Fig. 3b) or a directional vector (Fig. 3c) of Golgi
movement visually to correlate the direction and speed of
Golgi movement with movement of the bulk of photoactivated GFP in the ER membrane (Fig. 3a) in each case.
Golgi bodies were always seen to behave in a similar
manner to the PAGFP. In particular, when the PAGFP
moved very quickly (low t1/2) away from the activation
spot, Golgi bodies moved quickly in the same direction and
with high Meandering Index. Low Meandering Index Golgi
tracks were associated with slowly moving areas of PAGFP
in those cases where flow within the ER membrane was
best fit by an exponential decay equation. When cells had
been treated with latrunculin B, Golgi bodies did not move
(Fig. 1k–n); they became stationary at triple junctions of ER
tubules or within lacunae of the planar sheets.
Discussion
ER membrane dynamics
Photoactivatable GFP fused to the trans-membrane domain
of calnexin (CX-PAGFP) marks the ER membrane brightly
when activated with 405 nm laser light. Activated GFP
moves quickly away from the site of activation by a combination of energetic process and diffusion. The actin cytoskeleton is necessary for remodelling of the ER in plants
and in animals (Nebenführ et al., 1999; Saint-Jore et al.,
2002; Voeltz et al., 2002; Poteryaev, 2005) and large
bundles of actin filaments underlie the rapidly moving
strands of ER that traverse areas of tubular ER (Fig. 1)
(Quader et al., 1987; Boevink et al., 1998). When actin
depolymerizing chemicals are used, remodelling of the ER
ceases but the ER retains its form. The implication is that
there is a connection between the ER and the actin/myosin
system but it is, as yet, unknown how they connect. In the
absence of actin/myosin remodelling when cells were
treated with latrunculin B, CX-PAGFP in the membrane
of the ER continues to flow at a reduced rate and disperses
throughout the ER in the cell. Intensity decay curves of
activated GFP in the absence of actin were well fit by the
exponential equation which, as a first approximation, indicates that the mechanism of PAGFP movement is diffusive.
Diffusion speeds of proteins within the ER membrane can
be affected by a number of different factors including
association with other membrane proteins and protein complexes, and aggregation (Nehls et al., 2000). When the
actin/myosin system was functioning, the speed of activated CX-PAGFP dispersal varied from far in excess of that
resulting from this diffusion alone down to, in 2/15 cases,
not significantly different from the rate of diffusion in
latrunculin B-treated cells. This suggests that diffusion of
CX-PAGFP occurs even in those reticulate regions where
fast, actin-driven remodelling is absent. Fast ER membrane
movement curves were better fit by the inverse curve
function than by the exponential decay function and this
allowed an estimation of the speed of fluorescence movement. Components of the movement mechanism, including
effective diffusion coefficients for CX-PAGFP in different
ER regions may be better estimated using more advanced
models for curve-fitting in future experiments. However,
Siggia et al. (2000) describe the large potential for variation
in effective diffusion coefficients based on structure, for
example, tubular versus sheet-like ER, in the bleached
ER region.
48 Runions et al.
Normalised Intensity (%)
a)
Normalised Intensity (%)
b)
Normalised Intensity (%)
c)
120
110
100
90
80
70
60
50
40
30
20
10
0
120
110
100
90
80
70
60
50
40
30
20
10
0
Fast moving ER
(t1/2 = 9.05 sec.)
Exponential (solid line)
Inverse (dashed line)
0
20
30
40
50
60
Slowly moving ER
(t1/2 = 17.62 sec.)
Exponential (solid line)
Inverse (dashed line)
0
10
20
30
40
50
r2 = 0.96
r2 = 0.85
60
Non-moving ER (latrunculin B)
(t1/2 = 17.94 sec.)
120
110
100
90
80
70
60
50
40
30
20
10
0
10
r2 = 0.65
r2 = 0.93
Exponential (solid line)
Inverse (dashed line)
0
10
20
30
40
50
r2 = 0.85
r2 = 0.52
60
Time (sec.)
Fig. 2. Model curve fitting to intensity decay data after photoactivation
of GFP in the ER membrane. Two different curve fit models were tried in
each case; exponential decay (solid line) which would indicate that
intensity decline was by diffusion of GFP out of the activation region and
the inverse curve (dashed line) which allowed estimation of the speed of
intensity decay in cases where it was too fast to fit by the exponential
curve. (a) Fast-moving ER. Intensity decay of GFP in fast-moving
regions of the ER was not amenable to fitting by the exponential decay
equation which, in particular, could not fit the rapid decay that occurred
during the first few seconds after photoactivation. Rapid intensity decay
indicates that an energetic process, in addition to diffusion, moves CXPAGFP within the ER membrane. n=13/15 cells. (b) Slowly moving ER.
In slowly remodelling regions of ER, the intensity decay was much more
gradual and better fit by the exponential decay equation which indicated
that CX-PAGFP was not actively being moved by actin/myosin but was
diffusing. n=2/15 cells. (c) Latrunculin B-treated cells. ER movement
ceased after actin depolymerization, but CX-PAGFP within the ER
membrane continued to diffuse as indicated by good curve fitting with the
exponential decay model. n=5/5 cells.
Fig. 3. Tracking Golgi body movement. (a) Golgi bodies marked by STmRFP (red) can be tracked in time-series data to generate movement
statistics like velocity, distance moved, and Meandering Index. In this
example, photoactivatable GFP (green) has been activated and its
dispersal recorded in a time series. Coloured lines and their associated
track number indicate the path followed by individual Golgi bodies
during the entire time series even though only a single time point is
shown. (b, c). Plots of Golgi body movement. Movement of individual
Golgi bodies can be plotted relative to a common origin (b) or as vectors
relative to a common origin (c) to give an idea of Golgi directionality and
speed. Ultimately, vector data can be correlated with CX-PAGFP
movement data. Here, the majority of Golgi bodies were moving quickly
towards the upper left quadrant as was the CX-PAGFP seen in (a).
As an analogy with the ‘mobile fraction’ calculated in
photobleaching experiments, which describes the proportion of bleached membrane protein that is replaceable, it has
been calculated that most (;90%) of the photoactivated
GFP can diffuse away from the sight of activation. This
suggests that the calnexin sequence, which targets PAGFP
to the membrane, is homogenously distributed and does not
associate with the immobile fraction. Photoactivation of
a mobile fraction protein will tend to increase total cell
fluorescence as it redistributes on a diffusion gradient and
this might account for the small, apparent immobile fraction
observed.
ER membrane dynamics 49
Golgi body tracking
Software for motion tracking greatly simplifies the task of
describing the movement of motile organelles. In a short
time, enough data can be analysed to yield statistically
meaningful results. In this experiment, Golgi bodies were
tracked as discrete units moving during dispersal of the
photoactivated GFP in the ER membrane. On average,
Golgi bodies within a cell move at approximately the same
speed as that previously reported (this study 0.2560.39
lm s1) (Boevink et al., 1998). Large variation in Golgi
body velocity can be accounted for by the types of motion
that occur within a cell: some Golgi bodies remain
relatively stationary or stop and start while others travel
quickly with transvacuolar strands of ER. The fastest Golgi
bodies measured moved at 0.4760.12 lm s1 which is
considerably slower than that reported by Boevink et al.
(1998) who observed Golgi bodies moving at 2.2 lm s1.
Examination of individual Golgi bodies moving within
regions of photoactivated GFP, however, revealed that their
speed and direction was the same as that of the GFP. In
cases of fast dispersing GFP, Golgi velocity was high and
directionality, as judged from vector plots and the Meandering Index, was strongly oriented along the GFP movement
axis (Fig. 3c). Low Golgi body velocities and Meandering
Indices occurred in those regions in which PAGFP was
found to be dispersing by diffusion alone, i.e. in the slowly
remodelling tubular or planar areas.
The ER/Golgi connection
Fluidity and diffusion within the ER membrane are important aspects when considering the existing models of plant
Golgi body movement in relation to the ER. Both the ER
and the Golgi apparatus are known to move over the actin
cytoskeleton (Boevink et al., 1998). Individual Golgi
bodies move in close association with the tubules of ER
without any apparent gap between the two as judged by
confocal microscopy. Nebenführ et al. (1999) proposed
the stop-and-go model to explain the movement of Golgi
bodies over the ER. This type of motion is indeed what
happens. On average, Golgi bodies have a Meandering
Index of ;0.5 which means that they travel twice their
linear displacement. This type of movement seems to be
a function of Golgi location within the ER lattice. Golgi
bodies are relatively stationary with oscillating type motion
within the slowly remodelling regions of ER and move
quickly in straight lines with the CX-PAGFP in the ER
membrane along fast transvacuolar strands of ER. Functionally, protein export from the ER to the Golgi is thought
to occur at the ER exit sites (ERES: Nebenführ et al., 1999).
Discrete ERESs form within the ER membrane in animal
and yeast cells when COPII-coated membrane and/or
vesicles concentrate after initiation involving the GTPase
Sar1p (Bonifacino and Glick, 2004). In plant cells, most of
the COPII vesicle-associated proteins have been identified
(Phillipson et al., 2001; Hawes, 2005) and discrete ER
membrane regions are marked when Sar1p-GFP is coexpressed with a fluorescent marker of the Golgi body
membrane, for example, ST-YFP (Brandizzi et al., 2002;
daSilva et al., 2004; Yang et al., 2005). It has been shown
recently (daSilva et al., 2004) that plant ERESs co-localize
and move with Golgi bodies. Based on this observation, if
it is predicted that ERESs are protein complexes within the
ER membrane, then these data showing that components
can move within the membrane (both by passive and active
processes) support the idea of mobile exit sites attached to
Golgi bodies. Further, the observation that individual Golgi
bodies move along with photoactivated regions of ER
membrane make the existence of an ER/ERES/Golgi body
physical continuity plausible.
When GFP is fused to the calnexin transmembrane domains and expressed in plant cells (Irons et al., 2003), large
sheet-like regions of ER form (Fig. 1b). This surface area
increase probably results simply from overexpression of
a membrane-targeted construct and such sheets are ideal
for observation of the ER/Golgi body interaction. Boevink
et al. (1998) reported that the cortical ER network in
Nicotiana leaf epidermal cells precisely overlaid the actin
network. The ER sheets reported here are scaffolded at their
margins and throughout on actin filaments (not shown) and
the majority of Golgi bodies move around the edges and
within the lacunae of the ER. Occasionally, as well, individual Golgi bodies are seen that traverse the sheets but in
almost no cases do Golgi bodies appear to be disconnected
from the ER. This spatial continuity is under investigation
in more detail following the observations of Brandizzi et al.
(2002) that Golgi bodies can, on occasion, break free of the
ER and move along actin filaments ahead of newly forming
ER tubules. It would seem that either the connection of
Golgi bodies to ERES within the ER membrane has the
ability to be transient or that the ERES/Golgi body complex
can disconnect from the ER membrane and transit along the
actin network.
Supplementary material
Movies that illustrate photoactivation of PAGFP in the ER
membrane can be found at JXB online. They illustrate three
different cases: (i) fast remodelling ER, (ii) slowly remodelling ER, and (iii) non-remodelling ER after treatment of
cells with latrunculin B.
Acknowledgements
Funding was provided by BBSRC to CH. The University of
Heidelberg provided visiting studentship support for SK and TB.
We would like to thank Ian Moore for the use of his Zeiss LSM 510
Meta confocal system and Federica Brandizzi for making ST-mRFP.
George Patterson and Jennifer Lippincott-Schwartz provided the
50 Runions et al.
photoactivatable GFP vector. Mark Curtis and Ueli Grossniklaus
provided the Gateway pMDC32 binary vector. We would also like to
thank Imogen Sparkes for helpful cloning advice.
References
Batoko H, Zheng H-Q, Hawes C, Moore I. 2000. A Rab1 GTPase
is required for transport between the endoplasmic reticulum and
Golgi apparatus and for normal Golgi movement in plants. The
Plant Cell 12, 2201–2217.
Boevink P, Oparka K, Santa Cruz S, Martin B, Betteridge A,
Hawes C. 1998. Stacks on tracks: the plant Golgi apparatus traffics
on an actin/ER network. The Plant Journal 15, 441–447.
Boevink P, Santa Cruz S, Hawes C, Harris N, Oparka K. 1996.
Virus-mediated delivery of the green fluorescent protein to the
endoplasmic reticulum of plant cells. The Plant Journal 10,
935–941.
Bonifacino JS, Glick BS. 2004. The mechanisms of vesicle budding
and fusion. Cell 16, 153–166.
Brandizzi F, Irons S, Johansen J, Kotzer A, Neumann U. 2004.
GFP is the way to glow: bioimaging of the plant endomembrane
system. Journal of Microscopy 214, 138–158.
Brandizzi F, Snapp EL, Roberts AG, Lippincott-Schwartz J,
Hawes C. 2002. Membrane protein transport between the endoplasmic reticulum and the Golgi in tobacco leaves is energydependent but cytoskeleton-independent: evidence from selective
photobleaching. The Plant Cell 14, 1293–1309.
Curtis MD, Grossniklaus U. 2003. A gateway cloning vector set for
high-throughput functional analysis of genes in planta. Plant
Physiology 133, 462–469.
daSilva LLP, Snapp EL, Denecke J, Lippincott-Schwartz J,
Hawes C, Brandizzi F. 2004. Endoplasmic reticulum export sites
and Golgi bodies behave as single mobile secretory units in plant
cells. The Plant Cell 16, 1753–1771.
Haseloff J, Siemering K, Prasher D, Hodge S. 1997. Removal of
a cryptic intron and subcellular localization of green fluorescent protein are required to mark transgenic Arabidopsis plants
brightly. Proceedings of the National Academy of Sciences, USA
94, 2122–2127.
Hawes C. 2005. Cell biology of the plant Golgi apparatus. New
Phytologist 165, 29–44.
Huang L, Franklin AE, Hoffman NE. 1993. Primary structure and
characterization of an Arabidopsis thaliana calnexin-like protein.
The Journal of Biological Chemistry 268, 6560–6566.
Irons S, Evans D, Brandizzi F. 2003. The first 238 amino acids of
the human lamin B receptor are targeted to the nuclear envelop in
plants. Journal of Experimental Botany 54, 943–950.
Nebenführ A, Gallagher LA, Dunahay TG, Frohlick JA,
Mazurkiewicz AM, Meehl JB, Staehelin LA. 1999. Stop-andgo movements of plant Golgi stacks are mediated by the actomyosin system. Plant Physiology 121, 1127–1141.
Nebenführ A, Staehelin LA. 2001. Mobile factories: Golgi dynamics in plant cells. Trends in Plant Science 6, 160–167.
Nehls S, Snapp EL, Cole NB, Zaal KJM, Kenworthy AK,
Roberts TH, Ellenberg J, Presley JF, Siggia E, LippincottSchwartz J. 2000. Dynamics and retention of misfolded proteins
in native ER membranes. Nature Cell Biology 2, 288–295.
Patterson GH, Lippincott-Schwartz J. 2002. A photoactivatable
GFP for selective photolabeling of proteins and cells. Science 297,
1873–1877.
Phillipson BA, Pimpl P, daSilva L, Crofts AJ, Tayler JP,
Movafeghi A, Robinson DG, Denecke J. 2001. Secretory bulk
flow of soluble proteins is efficient and COPII dependent. The
Plant Cell 13, 2005–2020.
Poteryaev D, Squirrell JM, Campbell JM, White JG, Spang A.
2005. Involvement of the actin cytoskeleton and homotypic
membrane fusion in ER dynamics in Caenorhabditis elegans.
Molecular Biology of the Cell 16, 2139–2153.
Quader H, Hofmann A, Schnepf E. 1987. Shape and movement of
the endoplasmic reticulum in onion bulb epidermis cells: possible
involvement of actin. European Journal of Cell Biology 44, 17–26.
Robinson DG. 2003. The Golgi apparatus and the plant secretory
pathway. Oxford, UK: Blackwell Publishing.
Saint-Jore CM, Evins J, Batoko H, Brandizzi F, Moore I, Hawes C.
2002. Redistribution of membrane proteins between the Golgi apparatus and endoplasmic reticulum in plants is reversible and not
dependent on cytoskeletal networks. The Plant Journal 29, 661–678.
Siggia ED, Lippincott-Schwartz J, Bekiranov S. 2000. Diffusion
in inhomogeneous media: theory and simulations applied to whole
cell photobleach recovery. Biophysical Journal 79, 1761–1770.
Staehelin LA, Moore I. 1995. The plant Golgi apparatus: structure,
functional organization and trafficking mechanisms. Annual Review
of Plant Physiology and Plant Molecular Biology 46, 261–288.
Voeltz GK, Rolls MM, Rapoport TA. 2002. Structural organization
of the endoplasmic reticulum. EMBO Report 3, 944–950.
Yang Y-d, Elamawi R, Bubeck J, Pepperkok R, Ritzenthaler C,
Robinson DG. 2005. Dynamics of COPII vesicles and the Golgi
apparatus in cultured Nicotiana tabacum BY-2 cells provides
evidence for transient association of Golgi stacks with endoplasmic
reticulum exit sites. The Plant Cell 17, 1513–1531.