Journal of Experimental Botany, Vol. 68, No. 15 pp.

4075–4087, 2017
doi:10.1093/jxb/erx243  Advance Access publication 23 August 2017
This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details)

RESEARCH PAPER

Plasmolysis-deplasmolysis causes changes in endoplasmic

reticulum form, movement, flow, and cytoskeletal
association
Xiaohang Cheng1,*, Ingeborg Lang2, Opeyemi Samson Adeniji1,† and Lawrence Griffing1,‡
1 
Biology Department, Texas A&M University, 3258 TAMU, College Station, TX 77843, USA

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

2 
Cell Imaging and Ultrastructure Research, University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria

* Current address: Department of Biology, University of Massachusetts, Amherst, MA.

†
Part of undergraduate thesis; Current address: Integrated Biomedical Sciences (IBMS), University of Texas Health San Antonio, 7703
Floyd Curl Drive, San Antonio, TX 78229.
‡
Correspondence: [email protected]

Received 23 January 2017; Editorial decision 20 June 2017; Accepted 10 August 2017

Editor: Richard Napier, University of Warwick, UK

Abstract
Plasmolysis of hypocotyl cells of transgenic Arabidopsis thaliana and Nicotiana benthamiana diminishes the dynam-
ics of the remodeling of the endoplasmic reticulum (ER) in the central protoplast, namely that withdrawn from the cell
wall, and more persistent cisternae are formed, yet little change in the actin network in the protoplast occurs. Also,
protein flow within the ER network in the protoplast, as detected with fluorescence recovery after photobleaching
(FRAP), is not affected by plasmolysis. After plasmolysis, another network of strictly tubular ER remains attached to
the plasma membrane-wall interface and is contained within the Hechtian strands and reticulum. FRAP studies indi-
cate that protein flow within these ER tubules diminishes. Actin is largely absent from the Hechtian reticulum and the
ER becomes primarily associated with altered, branched microtubules. The smaller volume of the central protoplast
is accompanied by decreased movement rates of tubules, cisternae, and spheroid organelles, but this reduced move-
ment is not readily reversed by the increase in volume that accompanies deplasmolysis.

Key words:  Cytoplasmic streaming, endoplasmic reticulum, Hechtian reticulum, Hechtian strand, membrane contact sites,
persistency mapping, plasmolysis.

Introduction
Plasmolysis is generally a reversible decrease in the volume than that necessary to abolish the hydrostatic pressure of
of a walled plant cell protoplast caused by water flow down a turgor, namely the point of incipient plasmolysis, resulting
gradient along the chemical potential of water when the cell in volume change as water flows out of the protoplast and
is exposed to hyperosmotic external solute concentrations. it shrinks away from the surrounding wall (Nobel, 2009).
In highly vacuolated cells, such as the elongated hypocotyl Shrinkage away from the wall is, however, limited in many
cells examined here, this decrease in cellular volume comes cell types by a physical linkage between the plasma mem-
primarily from water transport out of the vacuole and is brane (PM) and the wall, a linkage that is strong enough to
achieved when the water potential outside the cell is lower remain intact even as the bulk of the protoplast shrinks. If

Abbreviations: ER, endoplasmic reticulum; FRAP, fluorescence recovery after photobleaching; HR, Hechtian reticulum; HS, Hechtian strands; MCS, membrane
contact sites; MES, 2-(N-morpholino) ethanesulfonic acid; PM, plasma membrane.
© The Author 2017. Published by Oxford University Press on behalf of the Society for Experimental Biology.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which
permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
4076 | Cheng et al.

the wall-PM anchor sites are small, then plasma membrane the ER in the protoplast but does change the flow in the
tubules as small as 50–100  nm in diameter (Oparka et  al., Hechtian strands and reticula.
1994) connect the withdrawn protoplast to the wall and Likewise, the association of cortical ER with the cytoskel-
are called Hechtian strands (Hecht, 1912). If the anchored eton changes in the Hechtian reticulum. Most cortical ER
area of the plasma membrane is large, then the unanchored dynamics are regulated by the actin-myosin system and not by
plasma membrane of the shrinking protoplast retracts microtubules, since latrunculin B, but not oryzalin, increases
around the anchored area, producing a thin plasma mem- the persistency of tubules and cisternae (Sparkes et al., 2009).
brane ‘sandwich’, around 100–200  nm in thickness, which There are some instances where cortical ER can track on
under light microscopy appears tubular or cisternal and is microtubules in the presence of latrunculin B (Hamada et al.,
called the Hechtian reticulum. 2014; Griffing et  al., 2016) and these events are much less
Hechtian strands and the Hechtian reticulum have both dynamic than the actin-myosin mediated movement (Sparkes
been labeled with DiOC6 (Oparka et  al., 1994; Lang- et al., 2009; Ueda et al., 2010). However, as described below,
Pauluzzi and Gunning, 2000), a carbocyanine fluorescent we find that the ER remaining in the Hechtian reticulum
dye that accumulates in response to membrane potential colocalizes with altered, branched microtubules found there
(Terasaki, 2000). Although its selectivity for different orga- and the ER tubules appear to track exclusively along these

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

nelles varies, it is often used to stain the mitochondria and altered microtubules and not the filamentous actin network.
endoplasmic reticulum (ER). Furthermore, some, but not
all, Hechtian strands form at plasmodesmata, where the
plasma membrane is not only attached to the wall but trav- Materials and methods
erses it, and is associated with the ER desmotubule within Transgenic plant growth and plasmolysis
the plasmodesmata. The ER attached to the desmotubule Nicotiana benthamiana and Arabidopsis thaliana seedlings were
is presumably within these Hechtian strands. However, grown in half-strength modified basal salts of Murishige and Skoog
because DiOC6 can be a non-specific label (Terasaki et al., media (Caisson Laboratories Inc., North Logan, UT USA) with
2001) and not all Hechtian strands form at plasmodesmata, 1% (w/v) agar at room temperature in 17.5-hour-long days for 7 to
14 d.  For tobacco experiments, N.  benthamiana line 16c (Haseloff
the question of the ubiquity of ER in the Hechtian strands
et al., 1997; Ruiz et al., 1998) was used. It constitutively expresses
and Hechtian reticulum has been unresolved. Here, we GFP-HDEL in the ER lumen. For Arabidopsis experiments, sev-
show that ER is in both the Hechtian strands and Hechtian eral different transgenic lines were used. Arabidopsis constitutively
reticulum using plants expressing ER-localized fluores- expressing the GFP-tagged actin-binding domain of fimbrin1
cent proteins with carboxy-terminal ER retention signal (GFP-FABD) was used (Voigt et  al., 2005); seeds were generously
gifted by Prof. Jozef Samaj. In addition, transgenic Arabidopsis
(HDEL) fusions.
plants constitutively expressing a double YFP-fusion to the carboxy
Analysis of the ER morphology in the protoplast after and amino terminal regions of the actin binding domain of fim-
plasmolysis was examined using persistency mapping brin (35S: YFP-ABD2-YFP, courtesy of Elison Blancaflor, Noble
(Sparkes et  al., 2009; Griffing et  al., 2014), whereby those Foundation, Ardmore, OK, USA) and a GFP-TUA6 labeling alpha-
tubular or cisternal features that persist for longer than tubulin (Ueda et  al., 1999) were used. The latter two Arabidopsis
lines were transformed by floral dip with Agrobacterium with a
1.5–5  s are analyzed. Increases in persistent ER cisternae
binary vector containing a signal-transit-sequence-mCherry-HDEL
are characteristic of treatments that interfere with myosin fusion re-engineered from a yeast plasmid courtesy of Eric Snapp,
XI-K, 1, and C, (Myo11E, Myo11F and Myo11C1; Park Janelia Research Campus, Ashburn, VA, USA. Individuals express-
and Nebenfuhr, 2013; Griffing et  al. 2014). In young and ing strong fluorescence in the T2 generation were identified prior to
therefore small hypocotyl and root cells the ER is also more experimentation.
Plasmolysis experiments were carried out on whole seedlings
cisternal (Ridge et al., 1999). In addition, young and small
mounted in 10 mM 2-(N-morpholino) ethanesulfonic acid (MES)
cells have slower rates of streaming than older, large cells buffer (Sigma-Aldrich, St. Louis MO USA, pH 5.8) without (con-
(Stefano et al., 2014). Can changing the size of a cell’s proto- trol) and with plasmolyticum (0.75 M sorbitol). Following analysis
plast change the form of its ER? In this study, we show that without plasmolyticum, the MES buffer was completely removed by
in addition to decreasing the rate of streaming within the absorption with Kimwipes (Kimtech Science, USA) at the edge of
the coverslip, followed by the addition of plasmolyticum to com-
cell, persistent ER cisternae also increase upon plasmolysis.
pletely immerse the seedling. Seedlings lose turgor during the first 15
However upon deplasmolysis, rates of streaming remain low min of hypertonic solution treatment, causing a loss of focal plane,
and a high level of persistent cisternae occurs, uncoupling which was then adjusted. To minimize frame shifting and loss of
these events from cell size. focus, some of the experiments were performed with a coverslipped
The flow within the ER, as distinct from remodeling or seedling in 35 mm uncoated No. 1.5 glass-bottom dishes (made
in-house).
translational movement of tubules or cisternae, has been
shown to be relatively insensitive to drug treatments that
depolymerize microtubules and actin microfilaments and to Fluorescent dye labeling
treatments that interfere with myosin XI-K activity (Sparkes N. benthamiana seedlings constitutively expressing GFP-HDEL were
et al., 2009). Consequently, although there may be an under- stained with the following dyes and imaged with the 488 nm argon
lying organizational role for the cytoskeleton, the directional ion laser line, with emission at 500–530 nm, and the laser excitation
described for each probe. For periplasmic region labeling, 2.5  mg/
flows that occur within the plant ER (Runions et  al., 2006) mL lucifer yellow CH (dipotassium salt, Sigma-Aldrich, St. Louis,
may be more influenced by other factors. We also find that a MO, USA) was included in the plasmolyticum, following treatment
plasmolysis-deplasmolysis cycle does not change flows within for 45 min in 0.75 M sorbitol solution in 10 mM MES buffer without
 Plasmolysis changes ER dynamics and cytoskeletal association  |  4077

lucifer yellow to first achieve stable plasmolysis. When using the final plugin in Image J (Hardin J. 2016. FRAP profiler plugin in
488 nm argon ion laser line for excitation, lucifer yellow fluorescence ImageJ. http://worms.zoology.wisc.edu/research/4d/4d.html#frap,
was weaker than the relatively bright ER network, thereby pro- last accessed 4 July 2017). Recovery halftime and percent mobility
viding a convenient difference in contrast that allowed distinction was calculated from the fitted curves from each sequence. For ER
between the ER and the periplasm. To label the plasma membrane, luminal protein photobleaching analysis, each ER component at
20 μM FM4-64 (ThermoFisher Scientific, Waltham MA USA) was different stages was compared by pairs using Tukey-Kramer HSD
included in the plasmolyticum during plasmolysis experiments and analysis, with a confidence level of 95% (P<0.05). For microtu-
excited with a 543nm HeNe solid state laser, with emission at 650– bule photobleaching analysis, both recovery halftime and percent
750nm. For cell wall labeling, seedlings were treated with 10 μg/mL mobility from each microtubule component, namely polymerized
propidium iodide (PI, ThermoFisher Scientific, Waltham MA USA) microtubules in normal cytoplasm, depolymerized microtubules in
for 15 min before transfer into plasmolyticum and imaged with the protoplast, and microtubules in Hechtian reticulum, were compared
543nm NeNe laser, with emission at 570–670nm. by pairs using Tukey-Kramer HSD analysis, with a confidence level
of 95% (P<0.05).

Confocal microscopy and image processing

Fluorescent live images were acquired using an Olympus FluoView 3D reconstruction
1000 with a numerical aperture (NA) 1.2 UPLSAPO water immer- Z-stack images for 3D reconstruction were taken on the Olympus
sion 60x objective and an Andor laser spinning-disc confocal micro- FV1000 laser scanning confocal microscope with a 1.2 NA

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

scope on an inverted Olympus IX81 stand with either a 40x (NA 1.3) UPLSAPO water immersion 60x objective, with a step size of half
or 100x (NA 1.53) oil immersion objective. For persistency mapping of the Z-resolution. Pixel lateral dimensions were adjusted to fit the
the Olympus FluoView was used. A 70-frame time-lapse movie with Nyquist criterion and pixel dwell time was adjusted to maximize
frame interval 0.32 s was acquired for each region-of-interest and fluorescence while maintaining fast scanning speed. 3D images were
the first 50 frames of used for persistency mapping. Seedlings were generated using surface rendering with the ImageJ 3D viewer plugin
kept stable during the period of time-lapse image recording and an in ImageJ (Schmid B, Longair M, Schindelin J. 2013. ImageJ 3D
auto-correlation image stabilization program was used to compen- Viewer, https://imagej.nih.gov/ij/plugins/3d-viewer/. http://3dviewer.
sate for drift (Lucas-Kanade algorithm implemented in ImageJ, Li neurofly.de, last accessed 4 July 2017) and the open source software
K. 2008. The image stabilizer plugin for ImageJ. http://www.cs.cmu. package, Visualization Tool Kit (VTK, Kitware, Inc.) as described
edu/~kangli/code/Image_Stabilizer.html, last accessed 4 July 2017). earlier (Enloe and Griffing, 2000). Images of 3D reconstructed struc-
Persistency mapping was carried out using an in-house macro in tures were later exported as tiff files and edited in Adobe Illustrator
ImageJ using the procedure described previously (Sparkes et al., CS6 ® (Adobe Systems Inc.) for labeling and annotation.
2009).
The relative movement of the ER is calculated by dividing the
integrated density of the sum of the displaced frame difference
(DFD) by the total membrane area. The sum of the DFD is calcu- Results
lated by summing the difference images between every fifth frame.
Total membrane area is the area of the pixels above background in ER undergoes increasing persistent cisternalization
the sum of all frames in the image sequence. during osmotic shock
Organelle streaming analysis was carried out on the DIC images
of the image sequence following fluorescence recovery after pho- Osmotic shock carried out by treating seedlings with 0.75 M
tobleaching (FRAP). After generating a substack of the DIC image sorbitol solution in 10mM MES buffer at pH 5.8 was used
sequence, the two fastest moving, highly refractile lipidbodies that to induce plasmolysis. Intact seedlings of N.  benthamiana
did not move out of the frame during the analysis sequence were plants, stably expressing GFP-HDEL to label the ER, were
chosen and tracked manually by using the Manual Tracking plugin
examined before and during osmotic shock. Cortical ER of
in ImageJ (Cordelières F. 2005. Manual Tracking Plugin for ImageJ,
https://imagej.nih.gov/ij/plugins/track/track.html, last accessed 4 a single hypocotyl cell, specifically the fifth cell from root-
July 2017). The velocity based on the displacement between two shoot junction, was analyzed with persistency mapping of a
sequential frames was calculated by the plugin, then averaged in time-lapse video of 70 frames. Shrinking of the entire seed-
Microsoft Excel over the entire sequence and statistics calculated. ling ensued within minutes of plasmolyticum addition but the
Statistical analysis was done using the Student’s t-test in Microsoft
shrinking rate diminished and stabilized after plasmolysis.
Excel. Measurements for each treatment were pooled, averages were
calculated and the Student’s t-test (P<0.05) used to compare each Prior to plasmolysis, the cells have a fine polygonal, cortical
treatment by pair. ER network made up of thin tubules with a few cisternae seen
at some of the tubule three-way junctions (Fig. 1A). As the
protoplast withdraws from the wall during plasmolysis, usu-
FRAP
ally seen by 30 min following treatment with plasmolyticum,
Photobleaching was done using the 405 nm diode SIM laser on the
Olympus FluoView 100, set to 100% transmission and 10 micro- its ER cisternalizes (Fig. 1A, C) as more persistent and non-
second dwell time. FRAP of ER-GFP was carried out during a persistent cisternae form. As early as 31 min after addition of
40-frame (41.4x41.4 μm, 200x200 pixel) video with a frame interval plasmolyticum, the protoplast containing the ER pulls away
of 0.32 s. The bleach started after recording the first 10 frames using from the wall (blue arrows in Fig. 1A–C). By 45 min, distinct
a bleach ROI of 4.14 x 4.14 μm (20 x 20 pixels). FRAP of micro- Hechtian strands and reticula were usually seen.
tubules during plasmolysis was carried out during a 20-frame (53 x
53 μm, 200 x 200 pixels) video with a frame interval of 0.8 s. The At different times during osmotic shock, persistency maps of
bleach started after recording the first 5 frames in a circular ROI the two structural components of ER, the cisternae and tubules,
(diameter 20 pixels). The videos were acquired as 16-bit gray-scale were used to calculate the ratio of persistent membrane to total
images. Each analyzed video was first processed with the Image sta- membrane area (Sparkes et  al., 2009). Both persistent tubules
bilizer plugin of ImageJ (Li K. 2008. The image stabilizer plugin (Fig. 1B, D) and persistent cisternae (Fig. 1C, E) with a mem-
for ImageJ. http://www.cs.cmu.edu/~kangli/code/Image_Stabilizer.
html). FRAP analysis was then carried out using FRAP profiler brane area >0.3  μm2 were measured. Persistent cisterna area
4078 | Cheng et al.

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

Fig. 1.  Change in amount of persistent ER cisternae and tubules after osmotic shock in N. benthamiana seedlings with ER lumen labeled with GFP-
HDEL. A–C) Z-projections and persistency maps of time-lapse movies taken at different time points indicated by the numbers on each column showing
the minutes after starting 0.75 M sorbitol treatment. Recovery is within 20 min of sorbitol wash out. Blue arrows indicate where the protoplast withdraws
from the cell wall. Darker blue colors indicate the highest persistency, while yellow indicates the lowest persistency. Scale bar,10 μm. A) Projection of
the sum of the movie frames showing the general structure of ER in black. B) Persistency map of ER tubules. C) Persistency map of ER cisternae. D)
Tubule persistency in three different representative cells, indicated by different colors and symbols, over the times indicated after treatment with sorbitol.
E) Cisternal persistency in the same three cells. F) Change in the persistent cisternal area over time as percentage of total membrane imaged. Three
time points were chosen for comparison: control (before treatment), plasmolysis (after 48 min of treatment), and during recovery. Horizontal lines show
significant differences with P<0.05. Bars show average values. Error bars are standard deviation, n=10 cells.

increased over the time course of the osmotic treatment (Fig. 1E), 3D reconstructions. Figure 2B and C show reconstruc-
while the persistent tubule area remained relatively constant tion views from outside-in (0°) and from inside-out (180°),
(Fig. 1D). Significant differences (Fig. 1F) in the increasing area respectively, of non-plasmolyzed cells, while Fig. 2D and E
of the persistent cisternae exist among cells before, during, and show similar views of deplasmolyzed cells with increased
after the first 20 min of recovery in MES buffer. The increase in cisternalization. Figure 2F shows how the imaged hypoco-
persistent cisternae is probably a combination of both a decrease tyl cells look in the context of the entire hypocotyl. Figure
in the translational movement of cisternae (see below), as well as 2G–I show plasmolyzed cells at lower magnification (Fig.
an increase in the amount of membrane in cisternae. 2G, H) and higher magnification (Fig. 2I) in several dif-
The total volume of the protoplast is reduced during plas- ferent rotational angles, which display the organization
molysis, as revealed by the filling of the periplasmic space with of the reconstructed Hechtian reticulum and strands. In
the diffuse fluorescence of the wall-permeant probe, Lucifer the face-on view (0°, Fig. 2I), the tubules of the Hechtian
Yellow (Fig.  2A). In the periplasmic space, the remaining reticulum look kinked with short branches. As the view
membranes form an altered network, the Hechtian reticulum is rotated through 45° and then 90° degrees (Fig. 2I), the
(HR), adjacent to the wall that contains GFP-HDEL-labeled strands of the Hechtian reticulum foreshorten and fall in
ER tubules. The wall-adjacent HR network is shown in the the plane just beneath the wall, while the Hechtian strands
combined top four optical sections (first panel, sections 1–4, that connect the wall with the protoplast, show a continu-
Fig.  2A), but is absent from the combined next four optical ous connection with the ER in the protoplast (Fig. 2I, 90°
sections that contain the underlying periplasmic space marked rotation). Rotating the view through an additional 180°,
with Lucifer Yellow (second panel, sections 5–8, Fig.  2A). the Hechtian reticulum branches away from the observer,
As the cell is more deeply optically sectioned, tubules marked separate from the Hechtian strand that is connected to the
by the ER label occur in the periplasmic space and correspond protoplast ER (Fig. 2I, 270° rotation). The 3D reconstruc-
to the Hechtian strands (HS, Fig. 2A, sections 9–12 and 13–16). tion of the cell wall and ER confirms that the Hechtian
The organization of the ER-containing Hechtian reticu- reticulum forms at the vicinity of the inner side of cortical
lum and strands can be more clearly distinguished using cell wall (Fig. 3A, B).
 Plasmolysis changes ER dynamics and cytoskeletal association  |  4079

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

Fig. 2.  ER shape change during plasmolysis in N. benthamiana hypocotyl cells with ER lumen labeled with GFP-HDEL. A) Seedlings were treated for 45
min with 0.75 M sorbitol then incubated in sorbitol + 2.5 mg/mL Lucifer Yellow. Each image is the sum of four slices with a step size between slices of
0.53 μm. Slice range is indicated at bottom right of each frame. HS, Hechtian strands; HR, Hechtian reticulum; N, nucleus. Images are psuedocolored
with the Orange Hot look-up table in ImageJ. Scale bar, 10 μm. The periplasmic space is indicated by the yellow regions outside the protoplasts that
have a relatively lower intensity than that of the cell wall. B–C) 3D reconstruction of ER in control cells. Orange lines indicate cell wall. Scale bar, 6
μm*6 μm*15 μm. D–E) 3D reconstruction of ER during the recovery after plasmolysis. Scale bar, 6 μm*6 μm*6 μm. F) Illustration of the orientation of
the hypocotyl epidermal cell used to generate 3D reconstruction. The face-on view toward the apical side wall of the epidermal cell is described as 0°.
G–H) 3D reconstruction of ER during plasmolysis. Yellow lines indicate the border of withdrawing protoplasts. Representations of Hechtian strands and
reticulum are indicated by magenta and blue lines, respectively. Scale bar, 6 μm*6 μm*5.7 μm. I) 3D reconstructions of ER in the specific periplasmic
region in G–H (purple square) where Hechtian reticulum forms. Small inset of each frame indicates the angle of view toward the apical surface of the
hypocotyl cell side wall.
4080 | Cheng et al.

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

Fig. 3.  3D ER reconstructions in relation to the position of cell wall labeled with propidium iodide or plasma membrane labeled with FM4-64 in
N. benthamiana hypocotyl cells expressing GFP-HDEL. A) Confocal images of ER (green) and cell wall labeled with propidium iodide (magenta) at
two different focal planes in control (left panel) and plasmolyzed cells (right panel). B) 3D reconstructions of cells shown in (A). ER (green) and cell wall
(magenta) in control (left 2 images) and plasmolyzed cells (right 2 images). Scale bar, 4μm*4μm*10.6μm for control, 4μm*4μm*14.31μm for plasmolysis.
C) 3D reconstructions of a cell undergoing concave plasmolysis, showing the ER (green) and plasma membrane (red). Scale bar, 6μm*6μm*15.3μm.
Black arrows indicate Hechtian strands surrounded by plasma membrane. D) 3D reconstructions of two cells undergoing convex plasmolysis and an
adjacent cell undergoing concave plasmolysis, showing the ER (green) and plasma membrane (red). Black arrows indicate Hechtian reticulum with
plasma membrane label on both the inside and outside. Adjacent cells often pull away in the same region but some show concave plasmolysis (blue
arrow in 180° Merge) and others show convex plasmolysis (blue arrow in 0° Merge). Scale bar, 6μm*6μm*12μm. Yellow lines indicate withdrawing
protoplast border, orange lines indicate cell wall border. CW, cell wall; PM, plasma membrane.

The formation of the Hechtian strands and Hechtian retic- plasmolysis, where the protoplast forms a convex structure,
ulum occurs in regions of both concave plasmolysis (Fig. 2G, often forms at the cell’s apex, while concave plasmolysis
H) and convex plasmolysis (Fig. 3A, B). The Hechtian retic- often occurs along the side of the cell. In an effort to see the
ulum occurs most commonly at the outer and inner pericli- plasma membrane covering the Hechtian strands, cells were
nal walls (see Supplementary Fig. S1 at JXB online). Convex labeled with the plasma membrane and endosomal dye, FM
 Plasmolysis changes ER dynamics and cytoskeletal association  |  4081

4-64, and analyzed by 3D reconstruction. Those regions with slowed with further plasmolysis (Fig.  4C). There is a pos-
Hechtian strands (green, continuous with ER labeling) had sibility that this is simply the result of artificially diminish-
FM 4-64 labeled plasma membrane (red) surrounding the ing the cell size and that streaming rates correlate with cell
strands in the 180° reconstruction (black arrows in Fig. 3C, size (Tominaga et al., 2013; Stefano et al., 2014). However,
D), indicating that they are coated with plasma membrane upon 20  min of recovery during deplasmolysis, when the
which can be more clearly seen with electron microscopy cell resumes its previous size, the streaming rates remain
(Oparka, 1994), whereas the Hechtian reticulum appeared to low (Fig. 4C). This corresponds to the decreased movement
be associated with the wall-plasma membrane interface that rates of the ER (Fig. 4B).
still had FM 4–64 labeling. In regions, the Hechtian reticu-
lum may be covered by a sheet of plasma membrane partially Changes in movement within the ER lumen during
folding back on itself (Fig. 3C). The 3D reconstructions also plasmolysis and deplasmolysis
show the large amount of cisternalization of the protoplast
ER (Fig. 3C, D). Although measurement of the movement within the ER is
sometimes combined with measurement of the translational
movement of the ER when using some types of optic flow
Decreased movement of the ER accompanies

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

analysis (Ueda et al., 2010; Stefano et al., 2014), here we inde-
increased cisternal persistency
pendently assess the movement within the ER lumen using
The relative movement of the entire protoplast ER mem- FRAP. In previous studies where both FRAP and photoacti-
brane is calculated from the difference in pixel intensities vation were used to analyze the flow within the ER (Sparkes
that occurs between every five frames, that is every 1.6  s. et al., 2009), changes in membrane surface flow did not cor-
Those that move most show the highest integrated intensity relate with increased persistency of the ER. Here we exam-
values of the displaced frame difference, providing a pixel- ine the luminal flow of GFP-HDEL. Although ER-Golgi
based optic flow measurement of the translational move- recycling of this luminal protein may have an effect on this
ment of the ER cisternae and tubules (Griffing et al., 2014). flow analysis, all of the half-times of recovery are one to two
During osmotic shock, the relative movement of the total orders of magnitude, namely 0.5–5 s, faster than the half-time
protoplast ER diminishes (Fig.  4A). There is a significant of recovery of the ERD2-GFP marker (105 +/- 21.5 s, daSilva
difference between the amount of relative movement of the et al., 2004). Consequently, the influence of recycling on flow
ER before osmotic shock and 48  min after osmotic shock rates is minimal.
when it reaches stable plasmolysis (Fig. 4B). This decreased While plasmolysis increases cisternal persistency (Fig.  1)
movement corresponds to increased persistent cisternaliza- and decreases ER translational movement in the protoplast
tion (Fig. 1C, E). The even further reduction of ER move- (Fig.  4), it does not change the half-time of fluorescence
ment in the first 20  min of recovery when plants were recovery or the percentage mobility of ER luminal protein
transferred back into pH 5.8 MES buffer without plasmo- in the protoplast (Fig. 5). Figure 5A and B show representa-
lyticum (Fig. 4B) corresponds with the lack of recovery from tive regions. Boluses of condensed GFP-HDEL form upon
persistent cisternalization (Fig. 2D, E). deplasmolysis (Fig. 5A). Experimental data from many cells
As the ER movement diminishes during plasmolysis, so indicate that the half-time for recovery does not change in
the streaming of other organelles slows. Using highly refrac- the protoplast during plasmolysis but increases several fold
tile 1–2  µm diameter lipid bodies identified by differential in boluses of GFP-HDEL that form upon deplasmolysis
interference contrast (DIC) microscopy, the rates of stream- (Fig. 5B). Likewise, the percentage mobility of GFP-HDEL
ing were analyzed before, during, and after plasmolysis, does not change in the protoplast ER but there are significant
Fig.  4C. A  typical streaming rate of about 1  µm/s before decreases in the Hechtian strands, Hechtian reticulum, and
and during the first 15 min of plasmolysis was considerably boluses (Fig. 5B).

Fig. 4.  Movement of ER and other organelles during plasmolysis in N. benthamiana seedlings expressing GFP-HDEL. A) Movement of ER analyzed with
displaced frame difference (DFD) in three separate representative cells indicated by different colors and symbols. B) Movement (DFD) of ER before and
after plasmolysis and during recovery (n=10). Bars indicate averages and error bars represent standard deviation. Horizontal lines are between pairs that
show significant differences (P<0.05). C) Streaming of organelles identified with DIC microscopy in the same cells analyzed for ER luminal flow (Fig. 5).
Horizontal lines are between pairs that show significant differences (P<0.05). Bars represent averages and error bars represent standard deviation. n=40
for each group. (This figure is available in colour at JXB online)
4082 | Cheng et al.

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

Fig. 5.  FRAP of different regions of the ER before and during plasmolysis and following recovery from plasmolysis in tobacco seedlings. A) Images
showing representative regions being assessed by FRAP (red rectangles) Scale bar=10 μm. B) Recovery half-time of GFP-HDEL in tubules, cisternae,
Hechtian strands and reticulum, and boluses that form upon deplasmolysis. Horizontal line between bracketed groups indicate a significant difference
(P<0.05). n=18, 12, 13, 23, 9, 11, 24, 18, and 12 for each bar from left to right. C) Percentage mobility of the GFP-HDEL in tubules, cisternae, Hechtian
strands and reticulum, and boluses that form upon deplasmolysis. Horizontal lines show significant differences between the indicated selected or
bracketed groups (P<0.05, except for Hechtian strands and plasmolysis cisternae where P=0.0507). n=18, 12, 13, 23, 9, 11, 24, 18, and 13 for each bar
from left to right. Bars are average values. Error bars represent standard deviation.

Hechtian strands and Hechtian reticulum contain ER Hechtian reticulum differs from the other two conditions in
and appear to colocalize with the cytoskeleton in the that the tubulin has a very low percentage mobility but the
periplasmic region part that does recover takes a longer time to do so, with a
half-time of 3.1 s (Fig. 6D). This indicates that the recovery
The presence of microtubules in the Hechtian reticulum and that does occur is a consequence of the movement of tubulin
strands has been previously described (Lang-Pauluzzi and with a lower rate of diffusion than that found in the ‘back-
Gunning, 2000; Lang et  al., 2014) but here we show that ground’ of the control.
the ER tracks along microtubules in the Hechtian reticu- As with the microtubules, actin filaments (GFP-FABD2)
lum, as seen in plasmolyzed Arabidopsis hypocotyl cells were present in the Hechtian strands (Fig.  7A, D) and the
that are dual-labeled with mCherry-HDEL and GFP-TUA6 polymerized actin forms a cap on the withdrawing protoplast.
(Fig. 6A–C). In the region of the Hechtian reticulum, which However, the appearance of actin in the Hechtian reticulum
is formed predominately on the outer and inner pericli- is more problematic. In Fig. 7B, the protoplast (outlined in
nal walls (Supplementary Fig. S1), not only do the ER and cyan) ER tubules track extensively on actin. However, in the
the microtubules colocalize but the microtubules take on a region of the Hechtian reticulum (Fig.  7C, D; outlined in
branching morphology. This branching morphology differs cyan in Fig. 7C), the ER that is in the Hechtian reticulum is
from the normal morphology of microtubules (Fig. 6E), so in a region devoid of actin. This is in marked contrast with
an analysis of the behavior of their movements was under- the presence of tubulin in the Hechtian reticulum (Fig. 6).
taken (Fig.  6D). Although there is low percentage mobility
in the control (Fig.  6D), that which is mobile shows rapid
recovery, namely a half-time of 1  s, typical of the mobile Discussion
unpolymerized tubulin in the background. The low mobility
of the polymerized tubulin (Fig. 6D) is expected and is typi- Although the presence of the ER in components of the cell
cal of microtubules undergoing treadmilling (Ehrhardt and ‘left behind’ at the cell wall by the process of plasmolysis
Shaw, 2006). The predominance of depolymerized tubulin has been indicated by electron microscopy (Oparka, 1994)
in the plasmolyzed protoplast shows fairly rapid and com- and vital staining (Lang-Pauluzzi and Gunning, 2000), the
plete recovery, with a half-time of 2.3 s and a completion of use of an ER-resident protein, GFP-HDEL (Sparkes et al.,
77.3% (Fig. 6D). However, FRAP dynamics of tubulin in the 2009) definitively demonstrates this (Fig. 2). The ER remains
 Plasmolysis changes ER dynamics and cytoskeletal association  |  4083

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

Fig. 6.  Microtubule behavior during plasmolysis in Arabidopsis hypocotyl cells with ER lumen labeled by mCherry and microtubule labeled by tubulin-
GFP. A) The Hechtian reticulum of a plasmolyzed cell, showing the separate ER (red) channel, the microtubule (green) channel, and the merged image.
B) 3D reconstruction of the merged volumes of the ER (red) and microtubule (green) in the Hechtian reticulum in (A) to show their colocalization. Scale
bar, 7μm*7μm*10.6μm. C) 185° flip of (B). D) FRAP analysis of microtubules before plasmolysis (MT-control) and after plasmolysis in the protoplast
(MT-protoplast) and in the Hechtian reticulum (MT-HR). The left-hand vertical axis shows recovery half-time; n=38, 49, and 55, respectively. The right-
hand vertical axis shows the percentage mobility; n=38, 49, and 57, respectively. Horizontal bars indicate which pairwise differences are significant
(P<0.05). Bars are averages. Error bars are standard deviation. E) Representative confocal optical sections of microtubules for FRAP analysis shown in
(D). Orange circles indicate photobleached area. Scale bar, 10 μm.

attached to the PM-cell wall interface during plasmolysis media at 960 mOsmolar does not reduce the mobile fraction
and is an integral component of the Hechtian reticulum and of non-glycosylated YFP-KDEL, but when two glycosyla-
Hechtian strands, thereby providing evidence for its strong tion sites are engineered into YFP-KDEL, the mobile frac-
attachment. The strength of the attachment at membrane tion is greatly diminished from 98% mobility to 38% mobility
contact sites (MCS) between the ER and the PM in plants is (Nagaya et al., 2008). Examination of the mobility of glyco-
supported by experiments showing that ER/PM attachments sylated proteins that have an ER retention signal in plants
persist over time (Sparkes et al., 2009) and following centrifu- would be of interest.
gation (Quader et al., 1987). The relative luminal flow as determined by the half-time
The movement of luminal protein (GFP-HDEL) within of FRAP recovery does not change in the withdrawing pro-
ER is one to two orders of magnitude faster than the recycling toplast before, during, or after plasmolysis (Fig. 5). This is
of the putative HDEL-receptor, ERD2-GFP, from the ER in contrast with the flow within the tubules in the Hechtian
through the Golgi (daSilva et al., 2004). This is similar to the strands and reticula which change both in half-time of recov-
diffusion rate difference between YFP-KDEL and KDEL- ery and percentage mobility. During plasmolysis, the per-
receptor-YFP in animal cells (Nagaya et  al., 2008), so the centage mobility drops to about 70% in both the Hechtian
flows we see are probably confined to luminal flows through strands and reticula, significantly below the near 100% in
the network that are relatively uninfluenced by recycling protoplast tubules and cisternae. Not only is there a smaller
dynamics through the Golgi. In animal cells, hyperosmotic fraction of motile GFP-HDEL but the diffusion within the
4084 | Cheng et al.

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

Fig. 7.  ER and actin filaments interaction during plasmolysis in Arabidopsis. A) Actin formation at different focal planes (lower left) of a plasmolyzed
hypocotyl cell with a 3D reconstruction showing several Hechtian strands, one of which is labeled HS. Regions containing the Hechtian reticulum as
identified with DIC are shown as HR. B–D) Confocal images of hypocotyl cells showing the separate green channel with actin filaments labeled with YFP-
ABD2, the red channel showing the ER labeled with mCherry-HDEL, and a merged image. B) ER tracks on actin within the protoplast during plasmolysis.
The outline of the protoplast is in cyan. C) Actin doesn’t show up in the Hechtian reticulum outlined in cyan. D) Actin forms a big Hechtian strand within
the periplasmic region, labeled HS, but is excluded in the fine network of the Hechtian reticulum, labelled HR. Adj. cell, adjacent cell. Scale bar, 10 μm.

tubules slows, as shown by the doubling of the half-time for limited connectivity with the rest of the network, namely low
recovery in strands and tripling of the half-time for recov- percentage mobility.
ery in reticula. This decrease in diffusion is not surprising The situation for luminal flow in the Hechtian reticulum
in strands, since they may be ‘pulled out’ and thinned by the is quite different. Since it is not stretched like the ER in the
process of plasmolysis. It has been shown that overexpression Hechtian strand, there is no a priori reason to think that it
of reticulons decreases the mobility of co-expressed GFP- is thinned. However, its half-time of recovery is even longer
HDEL (Tolley et al., 2008), while also producing condensed than that in the Hechtian strands; hence, apparent diffusion
GFP-HDEL at three-way junctions and in other regions of is considerably reduced. This may be a consequence of being
the network. This has been interpreted to be the result of encased in the folds of the plasma membrane, which may
thinning the ER tubule as well. Although the strands have wrap around it. In that situation, the ER might have at least
decreased mobility of GFP-HDEL within their lumen, the two sides of its membrane bound to the plasma membrane
only condensed GFP-HDEL, in boluses, that we see is fol- (Fig. 8C) and that may alter the flow within the lumen if sur-
lowing deplasmolysis (Fig. 5). We interpret these boluses to face flow is linked to luminal flow.
be protein in isolated Hechtian strands, pulled away from the Several proteins that have been shown to be present at or
rest of the network, and upon deplasmolysis the strand either near the ER/PM MCS, namely Vamp-associated protein 27
forms a vesicle or condensed GFP-HDEL and remains with (VAP27), NET3C (Wang et  al., 2014) and synaptotagmin 1
limited diffusion, namely long half-times of recovery, and (Syt 1; Pérez-Sancho et al., 2015; Levy et al., 2015) may act as
 Plasmolysis changes ER dynamics and cytoskeletal association  |  4085

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

Fig. 8.  Summary and models for the change in ER and cytoskeleton during plasmolysis. A) ER, vacuole, and PM prior to plasmolysis. B) ER, vacuole,
and PM after 15 min of treatment with 0.75 M sorbitol - a temporary state of incipient plasmolysis. During this time, it is proposed that an increase in
cytosolic calcium is accompanied by a change in the number and distribution of MCS between the ER and the PM. It is also proposed that there is
a release of the PM from the cell wall in regions enriched for MCS that are not adherent to the wall, while those ER/PM contact sites with high affinity
to the wall start the process of Hechtian strand and reticulum formation. C) ER, PM, and vacuole membranes after 30–45 min in 0.75 M sorbitol. The
Hechtian strands and Hechtian reticulum have formed at sites where the ER and PM and cell wall remain tethered. ER-PM contact sites on the PM of the
withdrawing protoplast cause higher levels of internal cisternalization of the ER. D) Organization of microtubules and actin prior to treatment with sorbitol.
E) It is proposed that an increase in cytosolic calcium ions that accompanies osmotic shock produces fewer polymerized microtubules but those that do
polymerize become bundled and assume a different conformation. F) During the withdrawal of the protoplast from the cell wall after 30–45 min in 0.75
M sorbitol, actin becomes primarily associated with Hechtian strands and not the Hechtian reticulum. An actin cap forms on the withdrawing protoplast.
Meanwhile, the microtubules reform as macrotubulin-containing tree-like structures in the Hechtian reticulum. ER, endoplasmic reticulum; CW, cell wall;
MCS, membrane contact sites; PD, plasmodesma; PM, plasma membrane.

tethers between the ER and the PM. Syt 1 is an ER-resident plants (Fig. 8B, C) after it reaches a certain critical level, as
protein that is similar to extended synaptotagmins (E-Syts). it does in animal cells. The generation of more extensive PM/
E-Syts have extended calcium-sensitive domains that bind ER MCS associated with regions of the PM that have low
the ER to the PM (Giordano et al., 2013). When calcium re- affinity for the cell wall is consistent with the presence of per-
uptake by the ER is inhibited in animal cells (Poteser et al., sistent, that is probably associated with the PM, cisternal ER
2016), with the concomitant increase in cytosolic calcium structures seen in the withdrawing protoplast (Figs 1 and 8C).
(Giordano et al., 2013), ER/PM MCS increase (Poteser et al., A likely function of MCS may be to generate subdomains
2016) and ER-localized Syt1 is recruited to ER/PM MCS that compartmentalize the biochemical activity of the ER
(Giordano et al., 2013). As cytosolic calcium increases upon and PM (Dittman and Menon, 2017). Furthermore, there
hyperosmotic treatment (Choi et al., 2014), it can be hypoth- appear to be MCS ‘master switches’ that can change the
esized that this changes the nature of the PM/ER MCS in abundance of some MCS subdomains relative to other MCS
4086 | Cheng et al.

subdomains (Elbaz-Alon et al., 2015). Two separate tethers, Lang et  al., 2014), perhaps serving a stabilizing role. These
Syt1 and VAP27, have recently been shown to be in separate regions, however, are also not in direct contact with the cell
MCS subdomains in Arabidopsis (Siao et  al., 2016). The wall. The absence of detectable actin in the Hechtian retic-
presence of ER in Hechtian strands and reticula would be ulum when actin-associated proteins are present in some
consistent with a separate subdomain of the ER/PM that has MCS (NET3C, Wang et al., 2014) would be consistent with
high affinity for the cell wall (Fig. 8C). With increasing cal- the model that different MCS subdomains separate upon
cium, a separate set of ER/PM tether(s) would be recruited plasmolysis, as described above. The actin-containing MCS
to a subdomain of the PM with high affinity to the wall and would be those giving rise to persistent cisternae in the proto-
the tubules would thereby be stabilized at the cell cortex. The plast, while the tubulin-associated MCS would be those in the
nature of the connection between the cell wall, the PM, and Hechtian reticulum. Unlike the cisternalization that occurs
the ER is not known, but we suggest that it becomes more in the presence of latrunculin or myosin dominant negative
adherent in these regions of tubular ER, which then form the experiments (Sparkes et al., 2009), it is not accompanied by
Hechtian reticula and Hechtian strands. As described below increased persistent tubulation (Fig. 1A), thereby supporting
and in Fig. 8E and F, the cortical microtubules are also stabi- the conclusion that the persistent cisternalization caused by
lized in this region and undergo some transformation (Fig. 6). plasmolysis is different from that caused by inactivation of

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

During plasmolysis, cortical microtubules depolymerize the actin-based cytoskeleton.
and bundle in the withdrawing protoplast (Fig. 6E). However In summary, plasmolysis reveals different kinds of asso-
tubulin, in a different state as determined by FRAP, is left ciation between the ER with the PM. The PM of the with-
behind in the Hechtian reticulum (Fig. 6D, E). These are not drawing protoplast is associated with persistent ER cisternae
the treadmilling polymers in the cortex of non-plasmolyzed and actin. The PM of the Hechtian reticulum is associated
cells. No individual microtubules are seen treadmilling; with the cell wall and with persistent ER tubules that track
instead, relatively non-mobile bundles or ‘trees’ of tubulin along macrotubular tubulin. The mobility of luminal GFP-
form, some of which is slowly diffusively exchangeable, with HDEL is different in these two ER/PM associations. Our
a large half-time of recovery from FRAP (Fig.  6D). These model is that these MCS subdomains separate when hyper-
tubulin ‘trees’ have been seen before using immunocytochem- osmotic treatment causes a calcium influx. The separation
istry of the fixed Hechtian reticulum (Lang-Pauluzzi, 2000; of the subdomains may be achieved through the differential
Lang-Pauluzzi and Gunning, 2000) or GFP-tubulin cell lines calcium binding of ER tethers to the PM. This model can
(Lang et  al., 2014). Interestingly, as shown here, the ER in be tested in the future by examining cytoplasmic calcium
the Hechtian reticulum tracks the tubulin ‘trees’. The general during plasmolysis as the protoplast withdraws from the cell
depolymerization of microtubules in the retracting protoplast, wall and by analyzing the subcellular distribution of ER/PM
which we speculate to be driven by increased cytoplasmic cal- MCS characterized by separate tether populations following
cium ion concentration, is accompanied by the appearance of plasmolysis. 
these wavy, treelike, bundled polymers (Fig. 8E, F).
The nature of this tubulin is unclear but we can rule out
the possibility that it is all completely depolymerized and Supplementary data
trapped in pockets of folded over plasma membrane encas- Supplementary data are available at JXB online.
ing it as there is little mobility of the tubulin, as determined Fig. S1. Hechtian reticulum forms extensively against
by FRAP (Fig.  6). It could be depolymerized but have low outer and inner periclinal walls.
mobility because it is bound tightly to the plasma membrane.
Alternatively, it could be polymerized in a different state than
normal microtubules. In this case, the tubulin would be bound Acknowledgements
to other tubulin subunits but in a way that allows local diffu-
The research was supported by funds from Texas A&M University and with
sion and substitution of monomers, unlike normal microtu- technical support and instrumentation provided by the Microscopy and
bules. Such microtubules have previously been detected after Imaging Center at Texas A&M University. We also wish to thank Baochou
plasmolysis and are called ‘macrotubules’, which apparently Ton and Kathryn Ryan for assistance and Rene Garcia for access to facilities.
assemble differently from normal microtubules because they
have a larger diameter in electron micrographs (Komis et al.,
2002a). References
The extensive and dynamic tubular reticulum in the Choi WG, Toyota M, Kim SH, Hilleary R, Gilroy S. 2014. Salt stress-
retracting protoplast is correlated with an active actin net- induced Ca2+ waves associated with rapid, long-distance root-to-shoot
signaling in plants. Proceedings of the National Academy of Sciences,
work (Fig.  7). Also indicative of an active actin network is USA 111, 6497–6502.
the maintenance of cytoplasmic streaming in the plasmolyz- daSilva LL, Snapp EL, Denecke J, Lippincott-Schwartz J, Hawes C,
ing protoplast (Fig.  4). The association of actin with the Brandizzi F. 2004. Endoplasmic reticulum export sites and Golgi bodies
PM on the retracting protoplast is strong in areas that are behave as single mobile secretory units in plant cells. The Plant Cell 16,
1753–1771.
separated from the cell wall, forming a cap of actin on api-
Dittman JS, Menon AK. 2017. Speed limits for nonvesicular intracellular
cally plasmolyzing cells (Fig. 7). Actin filament bundles are sterol transport. Trends in Biochemical Sciences 42, 90–97.
common in large Hechtian strands and in cytoplasmic con- Ehrhardt DW, Shaw SL. 2006. Microtubule dynamics and organization in
nections between subprotoplasts (Komis et al., 2002b, 2003; the plant cortical array. Annual Review of Plant Biology 57, 859–875.
 Plasmolysis changes ER dynamics and cytoskeletal association  |  4087
Elbaz-Alon Y, Eisenberg-Bord M, Shinder V, Stiller SB, Shimoni Perez-Sancho J, Vanneste S, Lee E, McFarlane HE, Esteban Del
E, Wiedemann N, Geiger T, Schuldiner M. 2015. Lam6 regulates the Valle A, Valpuesta V, Friml J, Botella MA, Rosado A. 2015. The
extent of contacts between organelles. Cell Reports 12, 7–14. Arabidopsis synaptotagmin1 is enriched in endoplasmic reticulum-plasma
Enloe L, Griffing L. 2000. Improved volume rendering for the visualization membrane contact sites and confers cellular resistance to mechanical
of living cells examined with confocal microscopy. Procedings of the SPIE stresses. Plant Physiology 168, 132–143.
3960, 385–392. Poteser M, Leitinger G, Pritz E, Platzer D, Frischauf I, Romanin C,
Giordano F, Saheki Y, Idevall-Hagren O, et al. 2013. PI(4,5)P(2)- Groschner K. 2016. Live-cell imaging of ER-PM contact architecture by a
dependent and Ca(2+)-regulated ER-PM interactions mediated by the novel TIRFM approach reveals extension of junctions in response to store-
extended synaptotagmins. Cell 153, 1494–1509. operated Ca2+-entry. Science Reports 6, 35656.
Griffing LR, Gao HT, Sparkes I. 2014. ER network dynamics are Quader H, Hofmann A, Schnepf E. 1987. Shape and movement ofthe
differentially controlled by myosins XI-K, XI-C, XI-E, XI-I, XI-1, and XI-2. endoplasmic reticulum in onion bulb epidermis cells: possible involvement
Frontiers in Plant Science 5, 218. of actin. European Journal of Cell Biology 44, 17–26.
Griffing LR, Lin C, Perico C, White RR, Sparkes I. 2016. Plant ER Ridge RW, Uozumi Y, Plazinski J, Hurley UA, Williamson RE. 1999.
geometry and dynamics: biophysical and cytoskeletal control during Developmental transitions and dynamics of the cortical ER of Arabidopsis
growth and biotic response. Protoplasma 254, 43–56. cells seen with green fluorescent protein. Plant and Cell Physiology 40,
Hamada T, Ueda H, Kawase T, Hara-Nishimura I. 2014. Microtubules 1253–1261.
contribute to tubule elongation and anchoring of endoplasmic reticulum, Ruiz MT, Voinnet O, Baulcombe DC. 1998. Initiation and maintenance
resulting in high network complexity in Arabidopsis. Plant Physiology 166, of virus-induced gene silencing. The Plant Cell 10, 937–946.

Downloaded from https://academic.oup.com/jxb/article/68/15/4075/4093040 by guest on 18 March 2022

1869–1876. Runions J, Brach T, Kuhner S, Hawes C. 2006. Photoactivation of GFP
Haseloff J, Simerling KR, Prasher DC, Hodge S. 1997. Removal of a reveals protein dynamics within the endoplasmic reticulum membrane.
cryptic intron and subcellular localization of green fluorescent protein are Journal of Experimental Botany 57, 43–50.
required to mark transgenic Arabidopsis plants brightly. Proceedings of the
Siao W, Wang P, Voigt B, Hussey PJ, Baluska F. 2016. Arabidopsis
National Academy of Sciences, USA 94, 2122–2127.
SYT1 maintains stability of cortical endoplasmic reticulum networks and
Hecht K. 1912. Studien über den Vorgang der Plasmolyse. Beiträge zur VAP27-1-enriched endoplasmic reticulum-plasma membrane contact
Biologie der Pflanzen 11, 133–189. sites. Journal of Experimental Botany 67, 6161–6171.
Hu J, Prinz WA, Rapoport TA. 2012. Weaving the web of ER tubules. Sparkes I, Runions J, Hawes C, Griffing L. 2009. Movement and
Cell 148, 832–832. remodeling of the endoplasmic reticulum in nondividing cells of tobacco
Komis G, Apostolakos P, Galatis B. 2002a. Hyperosmotic stress leaves. The Plant Cell 21, 3937–3949.
induces formation of tubulin macrotubules in root-tip cells of Triticum Stefano G, Renna L, Brandizzi F. 2014. The endoplasmic reticulum
turgidum: Their probable involvement in protoplast volume control. Plant exerts control over organelle streaming during cell expansion. Journal of
and Cell Physiology 43, 911–922. Cell Science 127, 947–953.
Komis G, Apostolakos P, Galatis B. 2002b. Hyperosmotic stress-
Terasaki M. 2000. Dynamics of the endoplasmic reticulum and Golgi
induced actin filament reorganization in leaf cells of Chlorophyton
apparatus during early sea urchin development. Molecular Biology of the
comosum. Journal of Experimental Botany 53, 1699–1710.
Cell 11, 897–914.
Komis G, Apostolakos P, Galatis B. 2003. Actomyosin is involved in
the plasmolytic cycle: gliding movement of the deplasmolyzing protoplast. Terasaki M, Loew L, Lippincott-Schwartz J, Zaal K. 2001.
Protoplasma 221, 245–256. Fluorescent staining of subcellular organelles: ER, Golgi complex, and
mitochondria. Current Protocols in Cell Biology Chapter 4, Unit 4–4.
Lang-Pauluzzi I. 2000. The behaviour of the plasma membrane during
plasmolysis: a study by UV microscopy. Journal of Microscopy 198, 188–198. Tolley N, Sparkes IA, Hunter PR, Craddock CP, Nuttall J, Roberts
LM, Hawes C, Pedrazzini E, Frigerio L. 2008. Overexpression of a plant
Lang-Pauluzzi I, Gunning BES. 2000. A plasmolytic cycle: the fate of reticulon remodels the lumen of the cortical endoplasmic reticulum but
cytoskeletal elements. Protoplasma 212, 174–185. does not perturb protein transport. Traffic 9, 94–102.
Lang I, Sassmann S, Schmidt B, Komis G. 2014. Plasmolysis: Loss of Tominaga M, Kimura A, Yokota E, Haraguchi T, Shimmen T,
turgor and beyond. Plants 3, 583–593. Yamamoto K, Nakano A, Ito K. 2013. Cytoplasmic streaming velocity as
Levy A, Zheng JY, Lazarowitz SG. 2015. Synaptotagmin SYTA forms a plant size determinant. Developmental Cell 27, 345–352.
ER-plasma membrane junctions that are recruited to plasmodesmata for
Ueda H, Yokota E, Kutsuna N, Shimada T, Tamura K, Shimmen T,
plant virus movement. Current Biology 25, 2018–2025.
Hasezawa S, Dolja VV, Hara-Nishimura I. 2010. Myosin-dependent
Nagaya H, Tamura T, Higa-Nishiyama A, et al. 2008. Regulated motion endoplasmic reticulum motility and F-actin organization in plant cells.
of glycoproteins revealed by direct visualization of a single cargo in the Proceedings of the National Academy of Science, USA 107, 6894–6899.
endoplasmic reticulum. The Journal of Cell Biology 180, 129–143.
Ueda K, Matsuyama T, Hashimoto T. 1999. Visualization of
Nobel PS. 2009. Physicochemical and environmental plant physiology, Ed microtubules in living cells of transgenic Arabidopsis thaliana. Protoplasma
4. UK: Elsevier Inc, Linacre House, Jordan Hill, Oxford OX2 8DP 206, 201–206.
Oparka K. 1994. Plasmolysis: New insights into an old process. New Voigt B, Timmers ACJ, Samaj J, Muller J, Baluska F, Menzel D.
Phytologist 126, 571–591 2005. GFP-FABD2 fusion construct allows in vivo visualization of the
Oparka KJ, Prior DAM, Crawford JW. 1994. Behaviour of plasma dynamic actin cytoskeleton in all cells of Arabidopsis seedlings. European
membrane, cortical ER and plasmodesmata during plasmolysis of onion Journal of Cell Biology 84, 595–608.
epidermal cells. Plant, Cell and Environment 17, 163–171. Wang P, Hawkins TJ, Richardson C, Cummins I, Deeks MJ, Sparkes
Park E, Nebenführ A. 2013. Myosin XIK of Arabidopsis thaliana I, Hawes C, Hussey PJ. 2014. The plant cytoskeleton, NET3C, and
accumulates at the root hair tip and is required for fast root hair growth. VAP27 mediate the link between the plasma membrane and endoplasmic
PLoS One 8, e76745. reticulum. Current Biology 24, 1397–1405.