Movement and Remodeling of the Endoplasmic Reticulum in ......2009/12/29 · ER membrane and...

Movement and Remodeling of the Endoplasmic Reticulum inNondividing Cells of Tobacco Leaves W

I. Sparkes,a J. Runions,a C. Hawes,a and L. Griffingb,1

a School of Life Sciences, Oxford Brookes University, Oxford OX3 0BP, United KingdombBiology Department, Texas A&M University, College Station, Texas 77843-3258

Using a novel analytical tool, this study investigates the relative roles of actin, microtubules, myosin, and Golgi bodies on

form and movement of the endoplasmic reticulum (ER) in tobacco (Nicotiana tabacum) leaf epidermal cells. Expression of a

subset of truncated class XI myosins, which interfere with the activity of native class XI myosins, and drug-induced actin

depolymerization produce a more persistent network of ER tubules and larger persistent cisternae. The treatments

differentially affect two persistent size classes of cortical ER cisternae, those >0.3 mm2 and those smaller, called punctae.

The punctae are not Golgi, and ER remodeling occurs in the absence of Golgi bodies. The treatments diminish the mobile

fraction of ER membrane proteins but not the diffusive flow of mobile membrane proteins. The results support a model

whereby ER network remodeling is coupled to the directionality but not the magnitude of membrane surface flow, and the

punctae are network nodes that act as foci of actin polymerization, regulating network remodeling through exploratory

tubule growth and myosin-mediated shrinkage.

INTRODUCTION

The endoplasmic reticulum (ER) is the port of entry into the

secretory pathway, yet very little is understood about the dy-

namic nature of this pleomorphic structure (Sparkes et al.,

2009a). Actomyosin-based directional movement and cytoplas-

mic streaming drives the movement and dispersion of endo-

membrane compartments, such as the Golgi apparatus in plants

(Avisar et al., 2008, 2009; Peremyslov et al., 2008; Prokhnevsky

et al., 2008; Sparkes et al., 2008). Plant myosins, especially class

XI myosins, including XIK, motorize Golgi bodies (and mitochon-

dria and peroxisomes) just as myosin-V motorizes endomem-

brane organelles in yeast (Estrada et al., 2003) and nerve cells

(Brown et al., 2004).While expression of the tail domain of certain

Arabidopsis thaliana class XI myosins perturb organelle move-

ment, these fluorescent fusions do not collocate to the organelles

whose movement they effect, thus indicating titration of acces-

sory factors or low levels of organelle binding. Here, analytical

software has been developed to address questions of whether

the increased movement and dispersion that characterize plant

organelles, relative to organelles of animal cells, also pertain to

the ER and how this movement and dispersion is driven by the

actomyosin system.

As in cultured animal cells and yeast, the ER of higher plants is

made of two primary domains, a polygonal tubule network, and

cisternae (flattened membrane saccules) that are coextensive

with the network. The movement of the network is not an entire

shifting of a stationary set of polygons relative to, for instance, the

plasmamembrane. Indeed, the network appearsmore anchored

to the plasmamembrane than are other organelles (Quader et al.,

1987). In the absence of global shifting of the network,movement

within the network is either the movement of persistent network

structures or the remodeling of the network through the forma-

tion or movement of network nodes and branches (Wolfram,

2002). Therefore, movement of the ER can be assessed by

determining the persistency of network structures, such as

nodes and branches. Here, this is done with persistency maps,

images that show the level of persistency of different network

structures in the ER, including tubules (branches and polygons),

nodes (punctae or cisternae with an area between 0.1 and 0.3

mm2), and cisternae (membrane sheets with an area more than

0.3 mm2).

ER movement includes movement of proteins within the

ER membrane and network remodeling. Directional, actin-

dependent movement of membrane protein in the ER has re-

cently been demonstrated in plants (Runions et al., 2006). The

remodeling of the cisternal domain into tubules through the

recruitment of reticulons, proteins that drive tubulation, may lead

to the dynamic transition between tubular and cisternal domains

(Tolley et al., 2008). In yeast (De Craene et al., 2006) and

mammalian cells (Voeltz et al., 2006), the nuclear envelope is

the most reliably visible cisternal domain of the ER. In plants,

tubular-cisternal transitions are developmental (Ridge et al.,

1999) and may generally represent regions of differing ER func-

tion (Staehelin, 1997).

The plant ER network overlies the actin cytoskeleton (Boevink

et al., 1998), but does the actomyosin system regulate the forms

of movement described above? Here, changes in the persis-

tency of the cortical ER of tobacco (Nicotiana tabacum) leaf

epidermal cells are determined after treatments that depoly-

merize actin filaments (latrunculin b) or microtubules (oryzalin),

1 Address correspondence to [email protected] author responsible for distribution of materials integral to thefindings presented in this article in accordance with the policy describedin the Instructions for Authors (www.plantcell.org) is: L. Griffing([email protected]).WOnline version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.109.072249

This article is a Plant Cell Advance Online Publication. The date of its first appearance online is the official date of publication. The article has been

edited and the authors have corrected proofs, but minor changes could be made before the final version is published. Posting this version online

reduces the time to publication by several weeks.

The Plant Cell Preview, www.aspb.org ã 2009 American Society of Plant Biologists 1 of 13

perturb organelle motility (coexpression of a truncated class XI

myosin) or remove Golgi bodies (Brefeldin A). Although the

results indicate that some form of linkage between the actin

cytoskeleton and the ER membrane might drive ER movement

and change the mobile fraction of ER surface proteins, they do

not support the notion that the Golgi bodies, which are intimately

associatedwith the ER (Boevink et al., 1998; Runions et al., 2006;

Sparkes et al., 2009b) and which are also driven by the acto-

myosin system (Boevink et al., 1998; Nebenfuhr et al., 1999),

mediate ER movement.

RESULTS

ER Network Persistence

Image maps showing the persistence of tubules and cisternae

are called persistency maps. Two different control tobacco

epidermal cells, representative of 30 such cells, expressing the

fluorescent marker green fluorescent protein (GFP)-HDEL (see

Supplemental Movie 1 online) are shown in Figure 1, which

includes the first and last frames of a 50-framemovie (Figures 1A,

1A’, 1B, and 1B’), the total imaged membrane over the time of

imaging (image sum) used for normalization (Figures 1C and 1C’),

and tubule (Figures 1D, 1D’, 1F, and 1F’) and cisternal persis-

tency maps (Figures 1G, 1G’, 1I, and 1I’). Persistence is dis-

played as a gray level gradient (Figures 1D, 1D’, 1G, and 1G’) or

discrete gray levels (Figures 1E, 1E’, 1H, and 1H’). The darkest,

most persistent tubules above a certain size (Figures 1F and 1F’)

are quantified (see below). Likewise, themost persistent cisternal

structures, punctae that are smaller than 0.3 mm2 and cisternae

that are larger than 0.3 mm2, are quantified (see below). Tubule

and cisternal persistence are lower in some cells (Figures 1A’ to

1I’) than in others (Figures 1A to 1I). Both contain fast lanes of ER

movement with low persistence gray regions adjacent to the

arrow (Figures 1E and 1E’), where tubules change rapidly and

active streaming occurs in the direction indicated. Neither ex-

ample has persistent polygons within the network, nor do less

persistent polygonal structures move (no gray tubules remain as

polygons in Figures 1E and 1E’). There are few persistent

cisternae, but less persistent ones (dark gray, particularly punc-

tae, Figures 1H and 1H’) move. Therefore, as expected, and thus

validating the modeling system, the movement of the network is

not movement of polygonal subdomains of the network but of

small nodal punctae and tubule branches that form, grow, and

shrink or generate cisternae.

Actomyosin Requirement for ER Network Remodeling

To ascertain the contribution of the cytoskeleton to ER remod-

eling, tobacco epidermal cells were treated with drugs to depo-

lymerize either actin filaments or microtubules or subjected to

overexpression of class XI myosin tail domains, XIK or XIJ, fused

to enhanced yellow fluorescent protein (eYFP). To confirm that

latrunculin b and oryzalin depolymerize actin filaments and

microtubules, respectively, pieces of tobacco leaf epidermal

cells expressing fluorescent markers for both were treated and

are shown in Supplemental Figure 1 online. Previous studies

have shown that XIK tail fusions perturb Golgi, peroxisome, and

mitochondria movement (Sparkes et al., 2008). Proposed expla-

nations for these effects include competition for myosin effectors

and cargo binding sites; however, since little is known about the

latter, the formal proof for the mode of action of plant myosin tail

domains remains elusive (Sparkes et al., 2008; Avisar et al.,

2009). This fusion, along with other class XI tail domain fusions,

does not collocate to these organelles, is present in motile

puncta, and is diffuse throughout the cytosol. This diffuse cyto-

plasmic distribution could reflect a low level of binding to their

target organelles. Similarly, the tail domain of XIJ fused to eYFP is

diffuse throughout the cytosol but has a negligible affect on Golgi

movement (Avisar et al., 2009). Therefore, as a control for XIK

specificity on ER modeling, the effects of the tail domain of XIJ

fused to eYFP were tested in parallel.

ER network movement is quite different when cells coexpress

the myosin XIK tail domain (Figure 2; see Supplemental Movie 2

online, representative of 27 cells) or are treated with latrunculin b

(Figure 3; see Supplemental Movie 3 online, representative of 14

cells). The pattern of tubules and cisternae changes less (Figures

2, 3A, and 3B; see Supplemental Movies 2 and 3 online) than in

controls (Figures 1A, 1B, 1A’, and 1B’; see Supplemental Movie

1 online). The tubule persistencymaps of the treatments (Figures

2D to 2Fand 3D to 3F) show a connected network of persistent

tubules that include polygonal rings not found in control cells

(Figures 1D to 1F and 1D’ to 1F’). The most persistent tubules

(Figures 2E, 2F, 3E, and 3F) are longer and more abundant

throughout the network than in control cells (Figures 1E, 1F, 1E’,

and 1F’). The two treatments producemore and longer persistent

tubules than control (Figure 4A). However, there is little difference

in persistent tubules between the latrunculin b treatment and

myosin XIK tail coexpression (Figure 4A). Both treatments pro-

duce more persistent cisternae (Figures 2G to 2I and 3G to 3I)

than found in the controls (Figures 1G to 1I and 1G’ to 1I’).

Although there is little difference between the number of persis-

tent punctae per 100 mm2 of imaged membrane in the control

and cells coexpressing truncated myosin XIK, there is a signif-

icant increase in punctae in the latrunculin b treatment compared

with the control (Figure 4B).

After myosin XIK tail coexpression, high persistency punctae

are near the periphery of the larger persistent and less persistent

cisternae (arrows, Figure 2I), whereas the lower persistency

punctae subtend tubular networks (arrows, Figure 2H). Many

punctae seen in the latrunculin b treatment are not somuch nodal

at the branching or bending points of tubules, as they are on the

periphery of the large cisternae (arrows, Figures 3A, 3E, and 3I).

Less persistent punctae sometimes appear at the blind end of

tubules (arrowheads, Figures 3E and 3H). Persistent tubules in

control cells terminate into a network of lower persistency

tubules (Figures 1E and 1E’), whereas short persistent tubules

in myosin XIK tail coexpression and latrunculin b treatments do

not (Figures 2E and 3E).

Cisternae increase in both area and number after truncated

myosin XIK coexpression and latrunculin b treatment (Figure 4C;

compare Figures 1G and 1G’ with 2G and 3G). Although both

treatments show a similar increase in the average area of the

cisternae over control, there is a larger increase in the number of

2 of 13 The Plant Cell

persistent cisternae after latrunculin b treatment than after trun-

cated myosin XIK coexpression (Figure 4C).

To ascertain whether microtubules and another class XI my-

osin (XIJ) affect ER remodeling, movies were captured of cells

coexpressing eYFP-XIJ tail domain and GFP-HDEL and after

oryzalin treatment of cells expressing GFP-HDEL (see Supple-

mental Figures 2 and 3 and Supplemental Movies 4 and 5 online).

Oryzalin has no significant effect on tubule length and persis-

tency (Figure 4A) but does slightly increase persistent cisternae

size compared with the control (Figure 4C; see Supplemental

Figure 2 online for persistency maps). Prior to quantification, it

was apparent that coexpression of the XIJ tail domain did not

perturb ER remodeling. A small sample size was analyzed and

indicated that the XIJ tail domain did not affect ER cisternal size

or number, but there were no persistent tubules above the

threshold level of 0.3mm2 (see Supplemental Figure 3 online for

persistency maps).

ActomyosinRequirement forERMembraneProteinMobility

Fluorescence recovery after photobleaching (FRAP) and photo-

activation studies reveal the dynamics of ERmembrane proteins

after disruption of the actomyosin system. Overexpression of

GFP-CXN generates large ER cisternae over time (Figure 5A).

Cisternalization provides large two-dimensional sheets desirable

for FRAP photobleaching. FRAP analysis of cisternal GFP-CXN

yielded rates (t1/2 values) of protein movement and a measure of

the mobile fraction of membrane protein. Control cells (Figures

5A to 5C), latrunculin b–treated cells (Figures 5D to 5F), and cells

coexpressing GFP-CXN and the eYFP-XIK tail domain (Figures

5G to 5I) had similar recovery rates (t1/2 =;2.4 s; Table 1, Figure

5L). However, coexpression of GFP-CXN and the eYFP-XIK tail

domain resulted in a significantly lower mobile fraction of fluo-

rescent protein, for example, exchange of bleached molecules

with fluorescent molecules outside of the spot bleach region

(79.09% 6 13.20%, Tukey’s HSD P < 0.001) compared with

control and latrunculin b treatments, which were the same

(87.72% 6 10.18% and 89.14% 6 7.39%, respectively; Table

1, Figure 5L).

Photoactivation studies using CXN-paGFP (Runions et al.,

2006) measured protein mobility and mobile fraction indepen-

dently of FRAP. Protein mobility is measured as fluorescence

intensity decline as photoactivated CXN-paGFP disperses from

the activation spot (Figure 6). The t1/2 values of ;2 s for

movement away from the activation spot were the same in all

cases (Figures 6A to 6K; see also Table 2). Mobile fractions,

Figure 1. Persistency of Tubules, Punctae, and Cisternae of Cortical ER

in Two Different Control Tobacco Epidermal Cells Expressing GFP-

HDEL.

The first ([A] and [A’]) and last frame ([B] and [B’]) of a 50-frame 80-s

movie. Summed frames of the entire movie ([C] and [C’]), showing total

imaged membrane (gray plus black) and the brighter punctae in the

summed image (black). The tubule persistency map ([D] and [D’]) shows

more persistent tubules as darker values. The posterized gray-level

tubule persistency map ([E] and [E’]) shows high persistency (black),

mid-range persistency (dark gray), and low persistency (light-gray gra-

dient). The black arrow is positioned adjacent to an area of active

streaming, showing the direction of streaming. High persistency tubules

([F] and [F’]) from (E) and (E’) that were >0.3 mm2 in area are the tubules

counted in Figure 4A. In (F), the arrow indicates a persistent tubule that is

subtended by persistent punctae ([I], below). The persistency map ([G]

and [G’]) shows more persistent cisternae and punctae as darker values.

The posterized gray-level cisternal persistency map ([H] and [H’]) shows

high persistency (black), mid-range persistency (dark gray), and low

persistency (light gray gradient) cisternae and punctae (arrowheads [H’]

show midrange persistency punctae). The black arrow is positioned

adjacent to an area of active streaming, showing the direction of

streaming. The high persistency map ([I] and [I’]) from (G) and (G’)

shows both the punctae (0.1 to 0.3 mm2, circularity > 0.5) and cisternae

(0.3 mm2). Arrowheads show persistent punctae connected by a persis-

tent tubule. Bar = 10 mm.

Cytoskeletal Control of Plant Endoplasmic Reticulum 3 of 13

however, of cells coexpressing the mRFP-XIK tail domain

(91.27% 6 6.12%) and those treated with latrunculin b

(89.42% 6 7.04%) were significantly lower than in control cells

(97.63% 6 4.26%, Tukey’s HSD, P < 0.001; Figure 6L).

Are Golgi Bodies Involved in ER Remodeling and ER

Membrane Mobility?

In various plant tissues, Golgi bodies appear to be associated

with the ER and move either over, or with the ER membrane, but

as seen here do not collocate with persistent cortical ER puncta

(Figures 7A to 7C).

To test the putative role of Golgi bodies in ER remodeling and

membrane flow, cells were treated with Brefeldin A (BFA), which

deconstructs Golgi and results in varying degrees of cisternal-

ization of the ER (Boevink et al., 1999). This cisternalization

precludes comparison of tubule and cisternal persistency with

control cells because drug effects could be confused with the

direct effect of Golgi body absence.

Cells coexpressing GFP-HDEL and the Golgi marker ST-

mRFP were treated with BFA for 3 h, resulting in disappearance

of Golgi bodies and redistribution of ST-mRFP to the ER (Figures

7D to 7F). Stills from a representativemovie (Figures 7G to 7I; see

Supplemental Movie 6 online) show that the ER remodels in the

absence of Golgi bodies. Photoactivation analysis of cells

coexpressing CXN-paGFP and ST-mRFP after treatment with

BFA for 3 h reveals t1/2 values for movement away from the

activation spot (1.79 6 0.98 s) similar to control (2.01 6 1.15 s).

The mobile fraction (91.88% 6 8.08%) was significantly lower

after BFA than in control samples (97.63% 6 4.26%, Tukey’s

HSD P = 0.001; Table 2, Figures 6I to 6K).

DISCUSSION

Visualization of Cortical ER Network Movement and

Cytoplasmic Streaming

The nature of cortical ER network movement in plant cells is not

widely understood. In higher plants, the ER cisternal/tubular

network does not move as a whole or as cisternal/tubular

network fragments, as is often illustrated for the involvement of

ER in charophyte cytoplasmic streaming (Kachar and Reese,

1988; Lodish et al., 2008). Instead, about 5% of the total imaged

ER is very persistent (Figure 4A) and forms the framework upon

which movement occurs. This framework is revealed with per-

sistencymaps, visual maps of the ER that show its recent degree

of persistency. Themaps show ER features that remain relatively

static for periods of longer than 8 s (five time-lapse frames) in an

80-s data set (Figures 1D, 1D’, 1I, and 1I’; see Supplemental

Movie 1 online). These persistencymaps do not have persistently

connected networks of polygonal rings (Figures 1D, 1D’, 1E, and

1E’); polygonal subunits do not persist. Polygon ring formation


in a Tobacco Epidermal Cell Coexpressing GFP-HDEL and the eYFP-XIK

Tail Domain.

(A) to (I) Same as in Figure 1. Because there are fewer or no fast lanes in

cells coexpressing myosin-XIK tails and treated with latrunculin b, the

gray regions in (E) indicate regions of remodeling cisternae that are

differentially skeletonized. Arrows in (H) indicate less persistent punctae

associated with tubules. Arrows in (I) indicate more persistent punctae

associated with cisternae. Bar = 10 mm.


in a Tobacco Epidermal Cell Expressing GFP-HDEL and Treated with

Latrunculin b.

(A) to (I) Same as in Figure 1. Arrows in (A), (E), and (I) indicate similar

location near the periphery of cisternae where a persistent puncta

occurs. Bar = 10 mm.


and closure are two of the more transient events in the network.

Until now, most analysis of the tubular network has focused on

these transient events that generate and remove the tubule

polygons that make up the network (Lee and Chen, 1988; Knebel

et al., 1990; Lichtscheidl and Url, 1990), but some have noted

that punctate structural features persist (Knebel et al., 1990;

Lichtscheidl and Url, 1990). Observations using persistency

maps confirm that there is a population of distributed, persistent

punctae (cisternae between 0.1 and 0.3 mm2; Figures 1G to 1I

and 1G’ to 1I’) that could function as the nodal framework upon

which network movement takes place. Although these punctae

may serve to anchor persistent features in the otherwise dynamic

network, they are less persistent and may move (gray punctae

near arrowheads, Figure 1H’) in regions of rapid streaming.

Therefore, instead of referring to them as anchor sites, here we

prefer to describe them as anchor/growth sites, and we predict

that they are the same anchor sites around which the ER tubules

can be remodeled manually using laser-trapped Golgi bodies

(Sparkes et al., 2009b).

Figure 4. Quantification of Persistent ER Tubules and Cisternae under

Various Treatment Conditions.

(A) Persistent tubules in control GFP-HDEL–expressing cells (black,

Control), GFP-HDEL, and eYFP-XIK tail coexpressing cells (white, XIK),

latrunculin b–treated GFP-HDEL–expressing cells (light gray, LATB), and

oryzalin-treated GFP-HDEL–expressing cells (gray). Columns show av-

erage value, and error bars show SD in total number of movies evaluated,

n (Control, n = 30; XIK, n = 27; LATB, n = 14; oryzalin, n = 16). Percentage

of tubule area is the percentage of the total imaged membrane (from

summed images, refer to Figures 1 to 3C, black and gray regions) that is

persistent tubules (high persistency tubules, refer to Figures 1 to 3F).

Number/100 mm2 is the number of persistent tubules per 100 mm2 of total

imaged membrane. Length average is the average length in micrometers

of the persistent tubules. The differences between all except the control-

oryzalin pair (all categories) and the XIK-LATB pair (tubule length cate-

gory) are significant (P < 0.05, Tukey’s HSD multiple comparisons of

means, 95% family-wise confidence level).

(B) Persistent punctae in control GFP-HDEL–expressing cells (black,

Control), GFP-HDEL, and eYFP-XIK tail coexpressing cells (white),

latrunculin b (light gray), and oryzalin (medium gray) treated GFP-

HDEL–expressing cells, and GFP-HDEL and eYFP-XIJ tail coexpressing

cells (dark gray). Columns show average value, and error bars show SD in

total number of movies evaluated, n (Control, n = 30; XIK, n = 27; LATB,

n = 14; oryzalin, n = 16; XIJ, n = 4). Percentage of cisternal area is the

percentage of the total imaged membrane (from summed images, refer

to Figures 1 to 3C, black and gray regions) that is persistent punctae

(high persistency cisternae with areas between 0.1 and 0.3 mm2 and

circularity > 0.5; refer to Figures 1 to 3I; see Supplemental Figures 2 and

3F online). Number/100 mm2 is the number of persistent punctae per 100

mm2 of total imaged membrane. The LATB treatment is significantly

different (P < 0.05) from all of the other treatments, if the XIJ data are not

included in the set because of low sample size. The other treatments do

not significantly differ from each other (Tukey’s HSDmultiple comparison

of means, 95% family-wise confidence level).

(C) Persistent cisternae in control GFP-HDEL–expressing cells (black),

GFP-HDEL and eYFP-XIK tail coexpressing cells (white), latrunculin

b–treated (light gray) and oryzalin-treated (medium gray) GFP-HDEL–

expressing cells, and GFP-HDEL and eYFP-XIJ tail coexpressing cells

(dark gray). Columns show average value, and error bars show SD in total

number of movies evaluated, n (same as [B]). Percentage of cisternal

area is the percentage of the total imaged membrane (from summed

images, refer to Figures 1 to 3C, black and gray regions) that is persistent

cisternae (high persistency cisternae >0.3 mm2, refer to Figures 1 to 3I;

see Supplemental Figures 2 and 3F online). The Control-LATB and -XIK

pairs are significantly different, as are the Oryzalin-LATB and -XIK pairs

and the XIJ-LATB pair (P < 0.05). Number/100 mm2 is the number of

persistent cisternae per 100 mm2 of total imaged membrane. All pair

comparisons except the Control-XIJ, Oryzalin-XIJ, XIK-XIJ, and Oryzalin-

XIK pairs are significant (P < 0.05). Area average is the average area in

mm2 of the persistent cisternae. The only pair comparisons that are

significant (P < 0.05) are Control-LATB, Control-XIK, and Oryzalin-XIK.

Tukey’s HSD multiple comparisons of means were done at 95% family-

wise confidence level.


The Role of the ER-Golgi Interface in ER Remodeling

There is an intimate association of Golgi bodies with moving ER

exit sites (daSilva et al., 2004), ER tubule growth appears to

follow Golgi bodymovement (Brandizzi et al., 2002), and the flow

of proteins within the ERmembrane corresponds in direction and

magnitude to that of streaming nearby Golgi (Runions et al.,

2006). These observations prompted the notion that Golgi-

associated motors might mediate the actomyosin dependence

of ER movement (Brandizzi et al., 2002; Hawes et al., 2008).

However, the results presented here indicate that Golgi bodies

do not affect ER remodeling (Figure 7) because when the Golgi

have been deconstructed with BFA, new ER tubules still actively

form and retract (see Supplemental Movie 6 online). The rate of

movement of ER membrane protein (t1/2 values) after BFA

treatment, as estimated by photoactivation (Figures 6I to 6L),

did not differ from other treatments, but the mobile fraction was

lower, matching myosin XIK tail coexpression values most

closely (Figure 6L). Although Golgi bodies do not appear to drive

ER, an as yet unanswered question is, does ER drive Golgi

bodies? However, the ER does appear to exert an effect on Golgi

stack location in terms of constraining them to the tubular ER or

to the rims of cisternal regions (Figure 7) (Boevink et al., 1998;

Runions et al., 2006; Hawes et al., 2008). While a comprehensive

study of all the Arabidopsis myosins, through transient expres-

sion studies in tobacco epidermal cells, has been implemented,

it has been shown that the XIK tail domain, but not the XIJ tail,

significantly affects movement of Golgi and mitochondria (Avisar

et al., 2009). Similarly, we show a comparable differential effect

on ER movement with the same myosin tail domains, perhaps

indicating an interdependent motion of the two organelles driven

by the same myosin.

The Effect of the Actomyosin System and Microtubules on

ER Remodeling and Movement

Both eYFP-XIK myosin tail coexpression and latrunculin b treat-

ment affect the appearance and dynamics of the ER (Figures 2

and 3; see Supplemental Movies 2 and 3 online). The eYFP-XIK

myosin tail domain perturbs streaming of Golgi, mitochondria,

and peroxisomes (Peremyslov et al., 2008; Sparkes et al., 2008),

presumably by interferingwith the activity of endogenousmyosin

XIK through titration of accessory factors required for binding

cargo or through titration of active myosin dimers required to

bind cargo. Latrunculin b treatment reduces the pool of poly-

merized F-actin by complexing G-actin monomers so that actin

filaments disappear (Coue et al., 1987; seeSupplemental Figures

1C and 1D online). Not only do both these treatments result in an

increase in the presence of persistent ER tubules (Figure 4A;

number/100 mm2 and length average), but they also produce a

pattern of more persistent polygonal ring structures (cf. Figures

2, 3D, and 3E with Figures 1D, 1D’, 1E, and 1E’), resulting in a

persistency map of polygonal ring structures connected by

persistent tubules. Furthermore, although both treatments give

rise to increased persistent cisternal area (Figure 4C, area

average), latrunculin b treatment appears to generate more of

Figure 5. The Effect of the eYFP-XIK Tail Domain and Latrunculin b on

FRAP of Tobacco Leaf Epidermal ER Marker GFP-CXN.

FRAP of a bleached spot area in control cells ([A] to [C]), cells treated

with latrunculin b ([D] to [F]), and those coexpressing eYFP-XIK tail

domain ([G] to [I]) were quantified where prebleach ([A], [D], and [G]),

bleach ([B], [E], and [H]) and 1.21 s postbleach ([C], [F], and [I]) are

shown. Coexpression of GFP-CXN and the eYFP-XIK tail domain ([J];

merged GFP-CXN image in [K]) prior to the bleach experiment shown in

(G) to (I) is indicated. The bleached area of GFP-CXN (circle) was

monitored over time, and the fluorescence recovery was plotted ([L];

where the green line is control, black line is latrunculin b treatment, and

red line is coexpression of GFP-CXN and the eYFP-XIK tail domain).

Bars = 2 mm.

Table 1. ER Membrane Dynamics Determined with FRAP of GFP-CXN

under Various Treatment Conditions

Treatment t1/2 (s)

Mobile Fraction

(Plateau) (%)

Control (n = 57) 2.37 6 0.49 87.72 6 10.18

Latrunculin B (n = 16) 2.45 6 0.44 89.14 6 7.39

eYFP-XIK tail (n = 48) 2.33 6 0.58 79.09 6 13.20


an increase in the number of persistent punctae (Figure 4B,

number/100 mm2) and persistent cisternae (Figure 4C; number/

100 mm2) than truncated myosin XIK coexpression.

The increase in the size of persistent ER cisternae (Figure 4C,

area average) may arise through a common mechanism, (i.e.,

inhibition of dynamic or exploratory tubular outgrowth of mem-

branes from anchor sites on cisternae that would otherwise

diminish cisternal area). The observation that persistent punctae

are redistributed to cisternae and their periphery after XIK tail

coexpression (arrows, Figure 2I) and latrunculin b treatment

(arrowhead, Figure 3A, 3E, and 3I) indicates that the nodal

punctaemay not only be anchor sites but may also act as growth

sites, stopped at some point in the tubule growth cycle.

We observed no significant effect of oryzalin on ER dynamics

other than an increase in cisternae. Oryzalin at the concentration

and time used is effective at depolymerizing microtubules (see

Supplemental Figures 1A and 1B online). Recent studies have

highlighted that oryzalin can have a secondary effect leading to

ER nodulation (Langhans et al., 2009), which could explain the

apparent increase in small persistent cisternae observed here.

Unlike Knebel et al. (1990), who also noted some ER nodulation,

Langhans et al. (2009) also reported an effect on ER dynamics

but attributed this to the drug and not to the absence of

microtubules. Our evidence supports that of Knebel et al.

(1990); microtubule depolymerization has no effect on ER dy-

namics. In addition, the small persistent cisternae may actually

be swollen anchor/growth sites, as suggested by Knebel et al.

(1990). We should caution, however, that this by no means

excludes a role for microtubules exerting a regulating influence

on the dynamics of the ER at other stages of the cell cycle.

Indeed, it is quite likely that during the transition to metaphase,

while the nuclear envelope resorbs into the ER, which then forms

aggregates at the spindle poles, there is a switch over from an

actin-based movement to microtubule-based movement on

spindle microtubules (Gupton et al., 2006). The association of

ER membrane with spindle microtubules in plants is well estab-

lished (Hepler and Wolniak, 1984).

AModel for the Involvement of Actin andMyosin in ER Form

and Remodeling

Observations of ER form reveal a dynamic structure that un-

dergoes rapid remodeling in an actin-dependent manner. Here,

we present the development of analytical software in order to

quantify these observations, with some surprising revelations

relating to persistent growth/anchor sites and the movement of

the ER surface in relation to the active remodeling structure.

Figure 6. The Effect of mRFP-XIK Tail Domain, Latrunculin b, and BFA

on the Mobility of CXN-Photoactivated GFP in Tobacco Leaf Epider-

mal ER.

A spot activated area (circle) of CXN-photoactivatable GFP was gener-

ated, and its mobility out of the activation spot was monitored over time

([A] to [K]) and quantified ([L]; where the green line is control, black line is

latrunculin b treatment, red line is coexpression of myosin XIK tail

domain, and the blue line is BFA treatment). Mobility of CXN-paGFP in

control cells ([A] to [C]), cells treated with latrunculin b ([D] and [E]),

those coexpressing CXN-paGFP and the mRFP-XIK tail domain ([F] to

[H]) or treated with BFA ([I] to [K]) were quantified where activation ([B],

[D], [G], and [J]) and 1.93 s postactivation ([C], [E], [H], and [K]) are

shown. An example of preactivation (A) highlights the lack of fluores-

cence prior to photoactivation in control cells. Expression of the mRFP-

XIK tail domain (F) and the effect of relocation of ST-mRFP back to

the ER (I) in the activation experiments shown ([F] to [K]) are indicated.

Bars = 2 mm.

Table 2. ER Membrane Dynamics Determined with Photoactivation of

CXN-paGFP under Various Treatment Conditions

Treatment t1/2 (s) Plateau (%)

Mobile

Fraction (%)

Control (n = 45) 2.01 6 1.15 2.37 6 4.26 97.63 6 4.26

Latrunculin B (n = 20) 2.31 6 1.22 10.58 6 7.04 89.42 6 7.04

mRFP-XIK tail (n = 31) 1.85 6 1.11 8.73 6 6.12 91.27 6 6.12

BFA (n = 28) 1.79 6 0.98 8.12 6 8.08 91.88 6 8.08


Based on these results, we propose a model for how the

actomyosin system regulates ER remodeling by interacting

with key structural elements, including growing, shrinking, and

stabilized tubules, three-way junctions, anchor/growth sites, and

tubule/cisternal transitions (Figure 8). In plants, extremely little is

understood regarding any of these key elements, and only

recently has it emerged that reticulons are involved in constrict-

ing the ER, and we have proposed a role for them in modulating

cisternalization versus tubulation (reviewed in Tolley et al., 2008;

Sparkes et al., 2009a).

ER Tubule Growth/Shrinkage

Since it is not clear how ER tubules and microfilaments mutually

orient themselves during growing/shrinking cycles, the relative

roles of myosin and actin polymerization in ER tubule remodeling

remain speculative. However, generating a model that helps

explain the results in the context of what is currently known about

cytoskeletal dynamics is desirable. The central features of the

model (Figure 8) are that the ER is centered on anchor/growth

sites and moves along microfilaments, mainly through its at-

tachment to a myosin such as XIK, but also via actin polymer-

ization. This generates tubule nonpersistency through cycles of

tubule elongation and shrinkage (Figure 8, processes 1 to 3) as

well as mutual sliding of tubules generating polygonal ring

closure (Figure 8, process 5). Both myosin XIK tail coexpression

with GFP-HDEL and latrunculin b treatments diminish dynamic,

exploratory tubules by blocking either tubule shrinkage or growth

(Figure 8, processes 1 to 3). This is supported by Figure 4A,

where both latrunculin b and myosin XI-K expression increase

the length and number of persistent tubules. Inhibition of grow-

ing/shrinking cycles of tubules after F-actin has depolymerized is

supported by the large number of blind ended tubules in Figures

2 and 3, whichmay have started a cycle but not completed it. We

further postulate that there is a polarity for ER tubule remodeling

and the actomyosin system (Figure 8, processes 1 to 3 and 5);

growth is mainly driven by actin polymerization, as supported by

the observation that the rate of in vivo actin polymerization (1.7

mm/s; Staiger et al., 2009) is very close to the rate of tubule

growth previously reported (1.34 6 0.54 mm/s; Sparkes et al.,

2009b), while shrinkage is mediated by (+) end-directed myosin

processivity. Myosins are classically defined as (+) end-directed

motors. Although this is true for a tobacco myosin (Tominaga

et al., 2003), the processivity and directionality of Arabidopsis

myosins is not known at this time.

Based on the BFA studies presented here, ER tubule remod-

eling still occurs in the absence of Golgi bodies. Therefore, the

potential role of ER tubule growth inGolgi bodymovement is also

depicted whereby a physical link, be it direct membrane conti-

nuity or interacting protein tethers, between the two organelles

results in interdependent movement (Figure 8, process 4).

Three-Way Junction and Anchor Sites Are Stabilization

Factors to Maintain Tubules

Three-way junctions and anchor/growth sites are evident under

all treatment conditions. Through micromanipulating Golgi bod-

ies, we showed that the underlying ER did indeed remodel and

resulted in the identification of ER anchor sites (Sparkes et al.,

2009b). However, it is unclear at this timewhether all anchor sites

are indeed growth sites and vice versa. The data presented here

indicate two populations of anchor sites/growth sites; a stable

persistent group and a transitory nonpersistent group. Upon

latrunculin b treatment, the latter become persistent, resulting in

a significant increase in number over the other treatment condi-

tions (Figure 4B). Therefore, perhaps an actin-dependent pro-

cess is required to determine the extent of movement of anchor

sites (Figure 8, process 3). Myosins may also be involved in the

movement of tubules toward moving anchor points but do not

affect the movement of anchor sites as much as latrunculin b

because the number of persistent punctae do not increase after

truncated myosin XIK expression (Figure 4B).

Proteinaceous factors regulating tubule stability can also be

tested using this system. For example, in mammals, CLIMP63

(cytoskeletal linking membrane protein p63) is required to inter-

act with microtubules to stabilize the cortical ER network (see

review in Vedrenne and Hauri [2006] and references therein;

Klopfenstein et al., 1998). Growth of tubules in animal systems is

also dependent on integral ER membrane proteins (e.g., stromal

interaction molecule 1) that bind the plus end of growing

Figure 7. Golgi Bodies Do Not Collocate with Persistent ER Puncta in

Tobacco Epidermal Cells and Do Not Affect ER Membrane Remodeling.

Cells coexpressing GFP-HDEL ([A]; green) and ST-mRFP ([B]; magenta)

treated with latrunculin b highlight the persistent puncta on the cisternal

ER and show that the Golgi bodies do not collocate to these regions (C).

Cells coexpressing GFP-HDEL and ST-mRFPwere treated with BFA (3 h,

100 mg/mL) ([D] to [I]), and remodeling of the ER network was monitored

over time. The Golgi marker ST-mRFP has relocated to the ER after BFA

treatment, indicating Golgi body absence ([E] and merged image in [F]).

Consecutive images ([G] to [I]) show that the ER network is able to

remodel and tubule outgrowth still occurs (arrowhead) after BFA treat-

ment. Bars = 2 mm.


microtubules (Grigoriev et al., 2008). By isolating analogous

actin-dependent factors in plants, such as the hypothetical

minus-end tip associated complex in Figure 8, persistency

maps can determine which proteins are required for tubule,

anchor point, and three-way junction stability.

The model presented for cortical ER dynamics in higher plants

relies heavily on myosin and actin dynamics. Therefore, one

assumption would be that upon actin depolymerization, the ER

network would disintegrate; however, it remained in place over

the time frame of the experiment albeit with altered tubule length/

dynamics/cisternalization (see Boevink et al., 1998). We hypoth-

esize that the anchor points, which might be connected to the

plasma membrane, are required to hold the ER network in place

and that the tubules are stabilized by accessory scaffolding or

coat proteins in the absence of actin. This is shown in the model

in Figure 8, where the tubules forming the central polygon are

tethered between anchor/growth sites in the absence of directly

underlying microfilaments.

Tubule/Cisternal Transitions

The actomyosin system also appears to be involved to some

extent in tubulation versus cisternalization as latunculin b treat-

ment and overexpression of the myosin XIK tail domain resulted

in increased persistent cisternae (Figure 4C) compared with

control and other treatment conditions. Based on the large

number of observed persistent polygonal structures (Figures 2

and 3) compared with the control (Figure 1), we propose that the

actomyosin system is also required for polygon remodeling. This

is achieved by the relative lateral sliding of newly formed

interconnected tubules (Figure 8, process 5). Actomyosin in-

volvement in polygon ring closure (Figure 8, process 5) may have

a flip side. In the absence of functional actomyosin, branch sites

making up the polygon ring may become unstable and the inner

side of the tubule corner, where there are high degrees of

curvature, would fill with membrane, causing more cisternaliza-

tion (Figure 4C) through polygon filling (Figure 8, process 6).

Figure 8. Model of the Different Activities of F-Actin and Myosin at Anchor/Growth Sites in Tubule Dynamics at Nonanchored Sites and in Determining

the Mobile Fraction of ER Membrane Proteins.

Process 1: Tubules grow outwards, driven by actin polymerization at an anchor/growth site. The anchor site is proposed to be a region where the

plasmamembrane and ER are closely associated. These sites may contain actin nucleation and growth proteins. The mechanism of actin attachment to

the growing tubule is represented diagrammatically as a hypothetical protein associated with the growing tubule tip at the pointed (�) end of the actin

filament. Process 2: Tubules shrink or retract in the (+) direction in a myosin-dependent fashion. This is supported by Figure 4A, where both latrunculin b

and myosin XI-K expression increase the length and number of persistent tubules. We interpret this as arising from an absence of shrinkage and

stabilization of untethered tubules. Process 3: Less persistent anchor/growth sites can move, perhaps recruited at branch points in the actin network,

indicated here by the presence of an ARP2/3 complex. There are fewer driven junctions after latrunculin b and myosin XI-K expression (Figures 2 and 3),

as indicated by persistent kinks. Blind-end tubules in Figures 2 and 3 may also be anchor sites that cannot move far from the point of origin but have

tubules attached. Note that three-way junctions do not require tethering to an anchor site. Process 4: Golgi do not regulate all of the growth or shrinkage

of ER tubules. After BFA treatment, which deconstructs the Golgi, new ER tubules still grow and shrink (Figure 7). However, since the moving ER and

moving Golgi remain closely associated, an interdependence of ER membrane movement and Golgi movement seems likely. Process 5: Myosin

mediates ring closure of polygons. The association of myosin with an unanchored three-way junction is postulated, and the special nature of the

junction is inferred. The observation that intact rings of the polygonal network are persistent after myosin XIK tail expression (Figures 2D to 2F) and

latrunculin b treatment (Figures 3D to 3F) is evidence to support this part of the model. Process 6: Actomyosin dependence of polygon filling. Removal of

myosin may result in the collapse of the corner with high angle of curvature and polygon filling, leading to cisternae and giving rise to the larger cisternal

area and larger number of cisternae after myosin XIK tail expression (Figure 4C).


The Effect of the Actomyosin System on ERMembrane

Protein Mobility

The data presented here, based on the mobility of the trans-

membrane domain of calnexin (CXN) fused to GFP, indicate that

ER remodeling and the direction of movement of the surface

membrane could be linked but are not linked to the magnitude of

the diffusive movement. Initial membrane diffusion rates (t1/2) are

similar in control to a latrunculin b–treated immotile ER network

system. Latrunculin treatment and overexpression of myosin XIK

tail domain remove the directional restriction on diffusion, while

restricting the available amount of molecules available for diffu-

sion (mobile fraction, Tables 1 and 2).

We hypothesize that native myosin XIK might naturally main-

tain a high mobile fraction of the CXN-GFP construct in the ER

membrane by destabilizing actin-membrane linkages during

shrinkage (Figure 8, process 2) or outgrowth of the membrane

(Figure 8, process 3). The eYFP-XIKmyosin tail domain is present

throughout the cytosol and could therefore directly interact with

the cortical ER, producing a lower mobile fraction of ER mem-

brane proteins by interfering with the activity of native myosin

XIK. Alternatively, instead of interfering directly with native my-

osin, the tail domain could remove an active component required

for membrane flow (through titration of accessory factors) or

introduce a rigid locked cytoskeletal scaffold on the ER surface

causing a barrier to free diffusive movement of proteins.

Latrunculin b treatment decreased the mobile fraction of ER

membrane CXN fused to paGFP/GFP in the photoactivation

experiments but not in the FRAP experiments. Photoactivation

may target either tubules or cisternae (both are invisible prior to

photoactivation), while FRAP targets preselected cisternae.

Latrunculin b treatment produces more persistent tubules (Fig-

ures 2 to 4), stabilized perhaps in a way that would diminish the

mobile fraction of protein in the membrane. Latrunculin b also

produces cisternae with many peripheral punctae (Figures 3H

and 3I, arrows) that may be stabilized by myosins, while the

center of the cisternae remains relatively unaffected, with no

apparent decrease in mobile fraction, as detected by FRAP.

Comparison of the results of this studywith an earlier one using

a similar technique to evaluate ERmembrane dynamics (Runions

et al., 2006) shows an apparent disparity in the diffusion rate of

CXN-paGFP. Half-times of recovery in the previous study were

reported to be 5 to 10 times longer than in this study, whilemobile

fraction percentages are comparable. Reexamination of the

older data shows that in the previous study, entire small cisternal

regions of ER were photoactivated so that diffusion of activated

paGFP was only possible through connecting tubules of small

radius. This resulted in a bottleneck effect and slower movement

of GFP away from the activation region than measured in this

study, where FRAP and photoactivation experiments were

performed on the tubular network of ER or larger cisternal sheets

that allowed a steeper diffusion gradient of GFP.

This study builds on the previous work by establishing the

diffusive, nonmotorized nature of movement of ER membrane

protein even during rapid ER remodeling. Previous work

(Runions et al., 2006) demonstrated a directionality to this

movement that can be seen in Figures 6B and 6C (control) and

Figures 6J and 6K (BFA treatment). The directionality of this

diffusion is changed when the actomyosin system is perturbed

(Runions et al., 2006; Figures 6D, 6E, 6G, and 6H). Here, we show

that the average rate of movement itself is not affected by the

actomyosin network, indicating nonmotorized and diffusive

movement. However, occasional fast, apparently nondiffusive

movements were monitored but were infrequent in comparison

to diffusional movements. A better quantification is required

to determine vectorial, directional diffusive versus apparent

active movements. The fact that directional diffusion occurs

after BFA treatment helps establish that directional diffusion is

not a property of tubules alone. This process of directional

diffusion within tubules (near one-dimensional) and cisternae

(near two-dimensional) may provide a mechanism whereby

proteins and metabolites can be more efficiently transferred

through the cell with minimal expenditure of energy at rates that

far exceed those of nondirectional diffusion in three dimensions

(Adam and Delbruck, 1968; Loverdo et al., 2008; Mirny, 2008;

Pickard, 2003).

In conclusion, we have found that the cortical ER network in

tobacco epidermal cells undergoes rapid remodeling that is

coupled to the directionality but not the magnitude of ER surface

flow. ER remodeling is an actomyosin-dependent process and

is independent of microtubules. Quantification of ER tubule

growth, cisternalization, and recognition of anchor/growth sites

has led to a model defining the basic principles behind remod-

eling of the plant cortical ER in leaf epidermal cells. Additionally,

ER network remodeling occurs in the absence of Golgi bodies,

thereby addressing the long-held postulate that Golgi body

movement might drive ER tubule growth. Future work using

this model and persistency mapping will help identify factors

required for macro- and microscopic changes in ER network

remodeling.

METHODS

Plant Material and Constructs

Nicotiana tabacum plants were grown according to Sparkes et al. (2005).

Fluorescent protein fusion constructs including the ERmarkerGFP-HDEL

(Batoko et al., 2000), GFP-FABD2 (Sheahan, et al., 2004), GFP-TUA6

(Ueda et al., 1999), eYFP- and mRFP-myosin XIK tail domain (eYFP-XIK

and mRFP-XIK; Sparkes et al., 2008), eYFP-XIJ tail domain (eYFP-XIJ;

Avisar et al., 2009), transmembrane domain of calnexin (CXN-paGFP)

(Runions et al., 2006), GFP transmembrane domain of CXN (GFP-CXN;

Irons et al., 2003), and sialyl transferase signal anchor sequence-mRFP

(ST-mRFP; Saint-Jore et al., 2002) were all infiltrated according to

Sparkes et al. (2006) with an optical density of 0.1 except for GFP-

FABD2 at 0.02 and GFP-HDEL, CXN fusions, and ST-mRFP, which

required 0.04 optical density. The mRFP-XIK tail domain-pB7WGR2

construct was generated by Gateway cloning using the entry vector XIK

tail-pDonor 207 (Sparkes et al., 2008).

Sample Preparation and Image Acquisition

The ER in the outermost cortical region of adaxial leaf epidermal pave-

ment cells was imaged. Where stated, leaf segments of ;25 mm2 were

incubated either in BFA (100 mg/mL) for 3 h or in latrunculin b (25 mm) or

oryzalin (14 mm) for 30 min prior to imaging.

Dual imaging of YFP and GFP was done using multitracking in line

switchingmode on a Zeiss LSM510Meta confocal microscope. GFPwas


excited with the 458-nm line, and eYFP with the 514-nm line of an argon

ion laser using a 458/514 dichroic mirror, and the subsequent emission

was detected using 470- to 500-nm and 560- to 615-nm band-pass

filters. Similarly, GFP and mRFP emission was captured using multi-

tracking in line switching mode. GFP was excited with a 488-nm argon

laser andmRFPwith a 543-nm laser and their emissions detected using a

488/543 dichroic mirror and 505- to 530-nm and 560- to 615-nm band-

pass filters. All imaging was performed using a 363 1.4–numerical

aperture oil immersion objective.

For persistency mapping, time-lapse images of GFP-HDEL were

captured using a 2- to 3-mm pinhole, 512 3 512 pixel resolution, and

32.3 digital zoom. To reduce noise, 43 line averaging was used. The

scan rate was increased by imaging a 955- to 960-mm2 region of interest

so that 50 frames per 80 s were captured (0.63 frames/s). For samples

where cells were coexpressing a fluorescent myosin tail domain and

GFP-HDEL, coexpression was verified before time-lapse imaging of the

GFP-HDEL alone was performed.

FRAP and Photoactivation

For FRAP experiments, five prebleach scans were captured using set-

tings as for GFP-HDEL (above) with the 488-nm laser transmission set to

1% transmission prior to bleaching of a 10- to 11-mm2 spot using 50

iterations of a 405-nmdiode laser set to 100% transmission. Fast intensity

recovery required a high scan rate (4.57 frames/s); therefore, time-lapse

imaging was performed using 2563 256 pixel resolution and 2.83 digital

zoom.

For photoactivation of CXN-paGFP, 30 prephotoactivation scans were

captured using settings as for GFP-HDEL (above) with the 488-nm laser

transmission set to 1% transmission prior to activation of the 8- to 10-mm2

spot using five iterations of the 405-nm diode laser set at 100% trans-

mission. Inherent high mobility of the ER required a high scan rate.

Postactivation time-lapse imaging was therefore performed using 128 3

128 resolution (6.8 frames/s) and 2.83 digital zoomwith the 488-nm laser

set at 1% transmission.

Mean fluorescence intensity was calculated (Zeiss LSM software)

within the bleach (FRAP) or activation (photoactivation) region during a

30- to 40-s postbleach or postactivation period. For FRAP, recovery of

fluorescence occurred on a scale normalized from 0% for the first

postbleach scan to 100% for the prebleach mean fluorescence level

within the region of interest. For photoactivation, normalized intensity

occurred on a scale between 0%, which represented mean preactivation

intensity, and 100%, the immediate postactivation intensity (Graumann

et al., 2007).

Normalized data from the time of bleaching/photoactivation were

plotted against time using Prism 5 software (GraphPad Software). Non-

linear regression was used to fit curves.

The exponentially increasing curve used to fit the FRAP data took the

form:

IðtÞ ¼ AeðB=ðtþ1ÞÞ ð1Þ

The exponentially decreasing curve used to fit photoactivation data

took the form:

IðtÞ ¼ Aeð2BtÞ þ ðC=ðtþ 1ÞÞ; ð2Þ

where A, B, and C are parameters of the curve and e is the base of the

natural logarithms.

Individual data sets within each experiment were fit independently to

assess variation within groups, and data points from all sets within a

group were fitted by sharing curve parameters in order to derive a master

fit for graphical presentation. Comparison across experimental groups

was doneby deriving themobile fraction of bleached/activatedmolecules

(plateau) and the half-time of fluorescence recovery (t1/2). The plateau

value was the highest-intensity value postbleach for FRAP and the lowest

intensity value for photoactivation in each data set. To determine the

plateau, the calculated parameters A, B, and C, which gave the best fit in

Equations 1 and 2 were used to fit a theoretical curve to the duration of

each data set. Half time of fluorescence recovery/decay is the time at

which half of the fluorescence has recovered/decayed after bleaching/

photoactivation. Half-fluorescence intensity (I1/2) was half the difference

between minimum and plateau fluorescence values. Half-time of fluo-

rescence recovery was interpolated at I1/2 using Prism 5 software

(GraphPad Software).

Multivariate analysis of variance (MANOVA) was used to compare the

families of curves derived in each of the experiments based on their

plateau and t1/2 values usingSPSS software.Wilks’ lambdawas used as a

test statistic for differences between experiments, and post-hoc tests to

compare differences between individual pairs of experiments were done

by computing Tukey’s HSD statistic.

Modeling the Cortical ER in Tobacco Epidermal Cells

Persistency maps were generated in Image J (version 1.37v; Wayne

Rasband, National Institutes of Health, Bethesda, MD). Beginning at

frame number 5 of 50, every fifth frame was compared with frames 8 s

(five frames) earlier with a Boolean “AND” operation and by subtraction.

The subtraction result was subsequently subtracted from the AND result

(with minor improvement), resulting in a set of 45 frames with objects that

persisted, unmoving for at least 8 s. Binary conversion of objects in the 13

to 255 (the brightest pixels) gray level range was followed either by a

closing operation (default in ImageJ, one iteration, one pixel count) that

connected tubules that may have become discontinuous during the

second subtraction operation or by an opening operation that removed

tubules between cisternae. The closed binary images were skeletonized

and summed, producing the tubule persistency map. The opened binary

imageswere summed, producing the cisternal persistencymap. To count

and measure persistent tubules and cisternae, eight-bit persistency

maps were segmented between 0 and 180 gray levels of 256 (the darkest

tubules and cisternae). For tubule maps, the number and length (half the

perimeter) of resultant persistent tubules were measured. For cisternae

maps, the number and area of the resultant cisternae were measured.

To generate percentage area values and number per 100mm2, the total

imaged membrane for each movie was determined by summing the

frames and segmenting between 15 and 255 gray levels (gray and black

areas in Figures 1 to 3C). The brightest regions, between 50 and 255 gray

levels (black areas in Figures 1 to 3C) look somewhat like persistent

cisternae and punctae (Figures 1 to 3H) but do not show good spatial

correlation. These regions may either be persistent or have high local

fluorophore concentration.

Quantitation of themost persistent tubules, punctae, and cisternaewas

done, normalizing each movie to the total membrane area imaged (panel

C in Figures 1 to 3) in the percentage area and number/100 mm2 values.

Statistical comparison of the data was done using R version 2.7.1 (http://

www.r-project.org/) using analysis of variance and Tukey’s HSD two-way

comparison at 95% confidence interval.

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure 1. Oryzalin and Latrunculin b Depolymerize

Microtubules and Actin, Respectively.

Supplemental Figure 2. Persistency of Tubules, Punctae, and Cis-

ternae of Cortical ER in a Tobacco Epidermal Cell Expressing GFP-

HDEL Treated with Oryzalin.


Supplemental Figure 3. Persistency of Tubules, Punctae, and Cis-

ternae of Cortical ER in a Tobacco Epidermal Cell Coexpressing GFP-

HDEL and eYFP-XIJ Myosin Tail Construct.

Supplemental Movie 1. Cortical ER Network Remodeling in Two

Control Tobacco Epidermal Cells.

Supplemental Movie 2. Cortical ER Network Remodeling in Tobacco

Epidermal Cells Coexpressing GFP-HDEL and the eYFP-XIK Tail

Domain.


Epidermal Cells Treated with Latrunculin b.


Epidermal Cells Expressing GFP-HDEL Treated with Oryzalin.


Epidermal Cells Coexpressing GFP-HDEL and the eYFP-XIJ Tail

Domain.


Epidermal Cells Treated with BFA.

ACKNOWLEDGMENTS

We thank Janet Evins for technical support during the work, D.W.

McCurdy for the GFP-FABD2 constructs, and T. Hashimoto for GFP-

TUA6 constructs. I.S. was supported by a Leverhulme Trust grant

(F/00382/G) to C.H. J.R. was funded by a Biotechnology and Biological

Science Research Council grant (BB/F014074/1), and L.G. was funded

by the International Research Travel/Award Grant and Faculty Devel-

opment Leave programs at Texas A&M University.

Received October 16, 2009; revised November 19, 2009; accepted

December 7, 2009; published December 29, 2009.

REFERENCES

Adam, G., and Delbruck, M. (1968). Reduction of dimensionality in

biological diffusion processes. In Structural Chemistry in Molecular

Biology, A. Rich and N. Davidson, eds (San Francisco, CA: Freeman),

pp. 198–215.

Avisar, D., Abu-Abied, M., Belausov, E., Sadot, E., Hawes, C., and

Sparkes, I.A. (2009). A comparative study of the involvement of 17

Arabidopsis myosin family members on the motility of Golgi and other

organelles. Plant Physiol. 150: 700–709.

Avisar, D., Prokhnevsky, A.I., Makarova, K.S., Koonin, E.V., and

Dolja, V.V. (2008). Myosin XI-K is required for rapid trafficking of Golgi

stacks, peroxisomes and mitochondria in leaf cells of Nicotiana

benthamiana. Plant Physiol. 146: 1098–1108.

Batoko, H., Zheng, H.Q., Hawes, C., and Moore, I. (2000). A Rab1

GTPase is required for transport between the endoplasmic reticulum

and Golgi apparatus and for normal Golgi movement in plants. Plant

Cell 12: 2201–2218.

Boevink, P., Martin, B., Oparka, K., Santa Cruz, S., and Hawes, C.

(1999). Transport of virally expressed green fluorescent protein

through the secretory pathway in tobacco leaves is inhibited by

cold shock and brefeldin A. Planta 208: 392–400.

Boevink, P., Oparka, K., Santa Cruz, S., Martin, B., Betteridge, A.,

and Hawes, C. (1998). Stacks on tracks: The plant Golgi apparatus

traffics on an actin/ER network. Plant J. 15: 441–447.

Brandizzi, F., Snapp, E.L., Roberts, A.G., Lippincott-Schwartz, J.,

and Hawes, C. (2002). Membrane protein transport between the

endoplasmic reticulum and the Golgi in tobacco leaves is energy

dependent but cytoskeleton independent: Evidence from selective

photobleaching. Plant Cell 14: 1293–1309.

Brown, J.R., Stafford, P., and Langford, G.M. (2004). Short-range

axonal/dendritic transport by myosin-V: A model for vesicle delivery to

the synapse. J. Neurobiol. 58: 175–188.

Coue, M., Brenner, S.L., Spector, I., and Korn, E.D. (1987). Inhibition

of actin polymerization by latrunculin A. FEBS Lett. 213: 316–318.

daSilva, L.L.P., Snapp, E.L., Denecke, J., Lippincott-Schwartz, J.,

Hawes, C., and Brandizzi, F. (2004). Endoplasmic reticulum export

sites and Golgi bodies behave as single mobile secretory units in plant

cells. Plant Cell 16: 1753–1771.

De Craene, J.O., Coleman, J., Estrada de Martin, P., Pypaert, M.,

Anderson, S., Yates III, J.R., Ferro-Novick, S., and Novick, P.

(2006). Rtn1p is involved in structuring the cortical endoplasmic

reticulum. Mol. Biol. Cell 17: 3009–3020.

Estrada, P., Kim, J.W., Coleman, J., Walker, L., Dunn, B., Takizawa,

P., Novick, P., and Ferro-Novick, S. (2003). Myo4p and She3p are

required for cortical ER inheritance in Saccharomyces cerevisiae. J.

Cell Biol. 163: 1255–1266.

Graumann, K., Irons, S.L., Runions, J., and Hawes, C. (2007).

Retention and mobility of the mammalian lamin B receptor in the

plant nuclear envelope. Biol. Cell 99: 553–562.

Grigoriev, I., Gouveia, S.M., van der Vaart, B., Demmers, J., Smyth,

J.T., Honnappa, S., Splinter, D., Steinmetz, M.O., Putney, J.W., Jr.,

Hoogenraad, C.C., and Akhmanova, A. (2008). STIM1 is a MT-plus-

end-tracking protein involved in remodeling of the ER. Curr. Biol. 18:

177–182.

Gupton, S.L., Collings, D.A., and Allen, N.S. (2006). Endoplasmic

reticulum targeted GFP reveals ER organization in tobacco NT-1 cells

during cell division. Plant Physiol. Biochem. 44: 95–105.

Hawes, C., Osterrieder, A., Hummel, E., and Sparkes, I. (2008). The

Plant ER-Golgi interface. Traffic 9: 1571–1580.

Hepler, P.K., and Wolniak, S.M. (1984). Membranes in the mitotic

apparatus: Their structure and function. Int. Rev. Cytol. 90: 169–238.

Irons, S.L., Evans, D.E., and Brandizzi, F. (2003). The first 238 amino

acids of the human lamin B receptor are targeted to the nuclear

envelope in plants. J. Exp. Bot. 384: 943–950.

Kachar, B., and Reese, T.S. (1988). The mechanism of cytoplasmic

streaming in characean algal cells: Sliding of endoplasmic reticulum

along actin filaments. J. Cell Biol. 106: 1545–1552.

Klopfenstein, D.R., Kappeler, F., and Hauri, H.P. (1998). A novel direct

interaction of endoplasmic reticulum with microtubules. EMBO J. 17:

6168–6177.

Knebel, W.H., Quader, H., and Schnepf, E. (1990). Mobile and immo-

bile endoplasmic reticulum in onion bulb epidermis cells: Short- and

long-term observations with a confocal laser scanning microscope.

Eur. J. Cell Biol. 52: 328–340.

Langhans, M., Niemes, S., Pimpl, P., and Robinson, D. (2009).

Oryzalin bodies: In addition to its anti-microtubules properties, the

dinitroaniline herbicide oryzalin causes nodulation of the endoplasmic

reticulum. Protoplasma 236: 73–84.

Lee, C., and Chen, L.B. (1988). Dynamic behavior of endoplasmic

reticulum in living cells. Cell 54: 37–46.

Lichtscheidl, I.K., and Url, W.G. (1990). Organisation and dynamics of

cortical endoplasmic reticulum in inner epidermal cells of onion bulb

scales. Protoplasma 157: 203–215.

Lodish, H., Berk, A., Kaiser, C.A., Krieger, M., Scott, M.P., Bretscher,

A., Ploegh, H., and Matsudaira, P. (2008). Molecular Cell Biology.

(New York: W.H. Freeman and Co.).

Loverdo, C., Benichou, O., Moreau, M., and Voituriez, R. (2008).

Enhanced reaction kinetics in biological cells. Nat. Phys. 4: 134–137.


Mirny, L. (2008). Cell commuters avoid delays. Nat. Phys. 4: 93–95.

Nebenfuhr, A., Gallagher, L.A., Dunahay, T.G., Frohlick, J.A.,

Mazurkiewicz, A.M., Meehl, J.B., and Staehelin, L.A. (1999).

Stop-and-go movements of plant Golgi stacks are mediated by the

acto-myosin system. Plant Physiol. 121: 1127–1141.

Peremyslov, V.V., Prokhnevsky, A.I., Avisar, D., and Dolja, V.V.

(2008). Two class XI myosins function in organelle trafficking and

root hair development in Arabidopsis thaliana. Plant Physiol. 146:

1109–1116.

Pickard, W.F. (2003). The role of cytoplasmic streaming in symplastic

transport. Plant Cell Environ. 26: 1–15.

Prokhnevsky, A.I., Peremyslov, V.V., and Dolja, V.V. (2008). Over-

lapping functions of the four class XI myosins in Arabidopsis growth,

root hair elongation and organelle motility. Proc. Natl. Acad. Sci. USA

105: 19744–19749.

Quader, H., Hofmann, A., and Schepf, E. (1987). Shape and movement

of the endoplamic reticulum in onion bulb epidermis cells:possible

involvement of actin. Eur. J. Cell Biol. 44: 17–26.

Ridge, R.W., Uozumi, Y., Plazinski, J., Hurley, U.A., and Williamson,

R.E. (1999). Developmental transitions and dynamics of the cortical

ER of Arabidopsis cells seen with green fluorescent protein. Plant Cell

Physiol. 40: 1253–1261.

Runions, J., Brach, T., Kuhner, S., and Hawes, C. (2006). Photo-

activation of GFP reveals protein dynamics within the endoplasmic

reticulum membrane. J. Exp. Bot. 57: 43–50.

Saint-Jore, C.M., Evins, J., Batoko, H., Brandizzi, F., Moore, I., and

Hawes, C. (2002). Redistribution of membrane proteins between the

Golgi apparatus and endoplasmic reticulum in plants is reversible and

not dependent on cytoskeletal networks. Plant J. 29: 661–678.

Sheahan, M.B., Staiger, C.J., Rose, R.J., and McCurdy, D.W. (2004).

A green fluorescent protein fusion to actin-binding domain 2 of

Arabidopsis fimbrin highlights new features of a dynamic actin cyto-

skeleton in live plant cells. Plant Physiol. 136: 3968–3978.

Sparkes, I.A., Frigerio, L., Tolley, N., and Hawes, C. (2009a). The plant

endoplasmic reticulum: A cell-wide web. Biochem. J. 423: 145–155.

Sparkes, I.A., Hawes, C., and Baker, A. (2005). AtPEX2 and AtPEX10

are targeted to peroxisomes independently of known endoplasmic

reticulum trafficking routes. Plant Physiol. 139: 690–700.

Sparkes, I.A., Ketelaar, T., Ruijter, N.C.A., and Hawes, C. (2009b).

Grab a Golgi: Laser trapping of Golgi bodies reveals in vivo interac-

tions with the endoplasmic reticulum. Traffic 10: 567–571.

Sparkes, I.A., Runions, J., Kearns, A., and Hawes, C. (2006). Rapid,

transient expression of fluorescent fusion proteins in tobacco plants

and generation of stably transformed plants. Nat. Protoc. 1: 2019–

2025.

Sparkes, I.A., Teanby, N.A., and Hawes, C. (2008). Truncated myosin

XI tail fusions inhibit peroxisome, Golgi and mitochondrial movement

in tobacco leaf epidermal cells: A genetic tool for the next generation.

J. Exp. Bot. 59: 2499–2512.

Staehelin, A. (1997). The plant ER: A dynamic organelle composed

of a large number of discrete functional domains. Plant J. 11:

1151–1185.

Staiger, C.J., Sheahan, M.B., Khurana, P., Wang, X., McCurdy, D.W.,

and Blanchoin, L. (2009). Actin filament dynamics are dominated by

rapid growth and severing activity in the Arabidopsis cortical array. J.

Cell Biol. 184: 269–280.

Tolley, N., Sparkes, I.A., Hunter, P.R., Craddock, C.P., Nuttall, J.,

Roberts, L.M., Hawes, C., Pedrazzini, E., and Frigerio, L. (2008).

Overexpression of a plant reticulon remodels the lumen of the cortical

endoplasmic reticulum but does not perturb protein transport. Traffic

9: 94–102.

Tominaga, M., Kojima, H., Yokota, E., Orii, H., Nakamori, R.,

Katayama, E., Anson, M., Shimmen, T., and Oiwa, K. (2003). Higher

plant myosin XI moves processively on actin with 35nm steps at high

velocity. EMBO J. 6: 1263–1272.

Ueda, K., Matsuyama, T., and Hashimoto, T. (1999). Visualisation of

microtubules in living cells of transgenic Arabidopsis thaliana. Proto-

plasma 206: 201–206.

Voeltz, G.K., Prinz, W.A., Shibata, Y., Rist, J.M., and Rapoport, T.A.

(2006). A class of membrane proteins shaping the tubular endoplas-

mic reticulum. Cell 124: 573–586.

Vedrenne, C., and Hauri, H.P. (2006). Morphogenesis of the endo-

plasmic reticulum: Beyond active membrane expansion. Traffic 7:

639–646.

Wolfram, S. (2002). A New Kind of Science. (Champaign, IL: Wolfram

Media).


DOI 10.1105/tpc.109.072249; originally published online December 29, 2009;Plant Cell

I. Sparkes, J. Runions, C. Hawes and L. GriffingMovement and Remodeling of the Endoplasmic Reticulum in Nondividing Cells of Tobacco Leaves

This information is current as of March 8, 2021

Supplemental Data /content/suppl/2009/12/17/tpc.109.072249.DC1.html

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

http://www.plantcell.org/cgi/alerts/ctmain

http://www.plantcell.org/cgi/alerts/ctmain

http://www.aspb.org/publications/subscriptions.cfm

Movement and Remodeling of the Endoplasmic Reticulum in ......2009/12/29 · ER membrane and...

Documents

Transcript of Movement and Remodeling of the Endoplasmic Reticulum in ......2009/12/29 · ER membrane and...