15
Intrabundle dynamics were also addressed in the mpk4 mutant (Figs. 7A to L; Video S3) and in
366
this case growth and shrinkage rates as well as catastrophe and rescue frequencies were found
367
similarly reduced. Thus, based on kymographs (Figs. 7E to G) intrabundle microtubule plus ends
368
were growing at 4.45
±3.27
μm/min (mean±SD; n=22; Fig. 7L; Tables S3 and S4; by comparison
369
to plus end dynamics of GFP-MBD-labeled microtubules from wild type cells, p=0.0053) and
370
shrinking at 5.81
±5.20
μm/min (mean±SD; n=32; Fig. 7L; Tables S3 and S4; by comparison to
371
plus end dynamics of GFP-MBD-labeled microtubules from wild type cells, p<0.0001). The
372
respective rates for discernible minus ends were 0.47
±0.63
μm/min (mean±SD; n=37; Fig. 7L;
373
Tables S3 and S4; by comparison to plus end dynamics of GFP-MBD-labeled microtubules from
374
wild type cells, p<0.023) and 0.69
±0.88
μm/min (mean±SD; n=32; Fig. 7L; Tables S3 and S4;
375
by comparison to plus end dynamics of GFP-MBD-labeled microtubules from wild type cells,
376
p=0.11). The catastrophe and rescue frequencies were 0.015 events/s and 0.020 events/s (n=89
377
representing 285 minutes of observation).
378
Side observations on microtubule organization in the mpk4 mutant showed the formation of
379
short, rigid and non-growing microtubule bundles (e.g., Fig. 8A, B, E) consisting of 3-4
380
microtubules as judged by their maximum fluorescence intensity (Fig. 8F). These were often
381
positioned free in the cytoplasm but frequently formed tip-wise attachments with the walls of
382
microtubules or microtubule bundles (Fig. 8E). When attached these bundles were either gliding
383
over short distances (e.g., Fig. 8B to D) or swinging around the attachment point (Fig. 8B).
384
385
386
www.plantphysiol.org
on July 21, 2018 - Published by
Downloaded from
Copyright © 2014 American Society of Plant Biologists. All rights reserved.
16
Discussion
387
General remarks
388
The temporal resolution of intracellular dynamics always progressed together with advances in
389
microscopy (e.g., Waterman-Storer, 1998). In this respect superresolution techniques that were
390
developed and vigorously upgraded during the past two decades (Hensel et al., 2013) are slowly
391
implemented to biological studies of dynamic subcellular events (e.g., Rego et al., 2012). To date
392
platforms of patterned illumination (e.g., STED and SIM; Hell, 2007, Allen et al., 2014) or
393
precision of localization (e.g., PALM and STORM; Small and Parthasarathy, 2014)
394
superresolution microscopies are commercially available, making superresolution imaging
395
widely accessible.
396
Knowledge on microtubule dynamics and organization, and their regulation by various MAPs,
397
was largely advanced by in vitro imaging of purified components (e.g., Dogterom and Surrey,
398
2013). In such assays fluorescently labelled or unlabeled microtubules grow in observation
399
chambers, nearly attached to the coverslip and hence they can be very clearly recorded at video
400
rates. However, the in vitro acquisition of microtubule dynamics is very frequently conflicting
401
with more physiological in vivo studies (Li et al., 2012; Zanic et al., 2013).
402
In plants a fine example of such discrepancy can be found in the elucidation of the dynamics of
403
in vitro assembled and MAP-free carrot tubulin by AVEC-DIC (Moore et al., 1997). Thereby a
404
stunning shortening velocity of 195
μm/min was reported being nearly 10 times higher than the
405
respective rates recorded in vivo on microtubules labeled with various markers such as
406
microinjected fluorophore-conjugated brain tubulin, GFP-MBD, GFP-TUA6, GFP-MAP65-1 or
407
GFP-EB1a (e.g., Zhang et al., 1990; Dhonukshe and Gadella, 2003; Shaw et al., 2003; van
408
Damme et al., 2004; Vos et al., 2004).
409
Subdiffraction microtubule organization
410
Considering the above, SIM was applied to study the organization of cortical microtubule arrays
411
in hypocotyl epidermal cells of Arabidopsis thaliana stably transformed with GFP-MBD and
412
GFP-TUA6 markers. Due to the propensity of plant cortical microtubule arrays to form extensive
413
bundles (e.g., Wasteneys and Ambrose, 2009), SIM observations were extended to hypocotyl
414
www.plantphysiol.org
on July 21, 2018 - Published by
Downloaded from
Copyright © 2014 American Society of Plant Biologists. All rights reserved.
17
epidermal cells of the mpk4 mutant where microtubule bundle formation is more pronounced
415
(Beck et al., 2010).
416
The point spread function of individual microtubules was determined experimentally for both
417
GFP-MBD and GFP-TUA6 labeled microtubules by means of normalized fluorescence intensity
418
profiling and determination of the FWHM for transverse profiles of such microtubules. As
419
expected from the specifications of the SIM platform used (see Materials and Methods) and the
420
theory behind superresolution imaging with SIM (Gustaffsson, 2000), resolution after high
421
performance imaging (see Materials and Methods for details) was nearly twice as good as the
422
ideal diffraction-limited resolution which cannot exceed 200 nm (Abbe’s limit; e.g.,
423
Verdaasdonk et al., 2014), and it reached 106 nm for GFP-MBD labeled microtubules and 117
424
nm for GFP-TUA6 labeled microtubules. These values are very close to the theoretical 100 nm
425
resolution predicted for linear SIM (Gustaffsson, 2000, Verdaasdonk et al., 2014). In the same
426
manner, high quality acquisitions with WF and CLSM were considerably above Abbe’s limit
427
(roughly between 225 and 325 nm) but nevertheless within the limits reported for such
428
microscopies (e.g., Zucker and Price, 1999; Salmon and Waterman, 2011).
429
SIM imaging revealed discontinuous incorporation of GFP-TUA6 fusion proteins to the
430
microtubule lattice, leading to the speckled appearance of such labeled microtubules. As
431
previously reported GFP-TUA6 is indistinguishably assembly-competent with unlabeled tubulin
432
dimers incorporating the native TUA6 (Shaw et al., 2003) or other
α-tubulin isoforms expressed
433
in hypocotyl cells. Based on the normal phenotypes of GFP-TUA6 transformed seedlings and the
434
similar growth/shrinkage rates of GFP-TUA6-labeled microtubules or microtubules labeled by
435
microinjection of Tradescantia stamen hair cells with fluorophore-conjugated brain tubulin
436
(Zhang et al., 1990), it was concluded that the GFP-TUA6 marker is not affecting microtubule
437
dynamics and function (see Shaw et al., 2003).
438
The speckled decoration of intracellular filamentous polymeric structures such as actin filaments
439
or microtubules was first reported when such structures were allowed to assemble in the presence
440
of very low levels (e.g. 0.01%-0.25% of labeled molecules with respect to the total pool) of
441
fluorescently labeled monomers diluted in a pool of unmodified monomers (e.g. Salmon and
442
Waterman, 2011). The stochastic incorporation of either labeled or unlabeled tubulin dimers in
443
www.plantphysiol.org
on July 21, 2018 - Published by
Downloaded from
Copyright © 2014 American Society of Plant Biologists. All rights reserved.
Dostları ilə paylaş: |