Download PilB localization determines the direction of twitching

1
PilB localisation correlates with the direction of twitching motility in the
2
cyanobacterium Synechocystis sp. PCC 6803
3
4
5
Nils Schuergers1, Dennis J. Nürnberg2,3, Thomas Wallner1, Conrad W. Mullineaux2,4,
6
Annegret Wilde1
7
8
1Molecular
9
Germany
Genetics, Institute of Biology III, University of Freiburg, 79104 Freiburg,
10
2School
11
London E1 4NS, UK
12
3Department
13
4Freiburg
14
Freiburg, Germany
of Biological and Chemical Sciences, Queen Mary University of London,
of Life Sciences, Imperial College London, London SW7 2AZ
Institute for Advanced Studies (FRIAS), University of Freiburg, 79104
15
16
*
17
76120397828; Fax +49 (0) 761-203-2745
For Correspondence. E-mail Annegret.Wilde@biologie.uni-freiburg.de; Tel. +49 (0)
18
19
Running title: PilB localisation in Synechocystis
20
21
Contents category: Cell and Molecular Biology of Microbes
22
23
Summary: 225 words
24
Main text: 3345 words
25
Tables: 0
26
Figures: 4
27
Abbreviations: T4P, type IV pili.
28
1
29
SUMMARY
30
31
Twitching motility depends on the adhesion of type IV pili to a substrate, with cell
32
movement driven by extension and retraction of the pili. The mechanism of twitching
33
motility, and the events that lead to a reversal of direction, are best understood in
34
rod-shaped bacteria such as Myxococcus xanthus. In M. xanthus, the direction of
35
movement depends on the unipolar localisation of the pilus extension and retraction
36
motors PilB and PilT to opposite cell poles. Reversal of direction results from
37
relocalisation of PilB and PilT. Some cyanobacteria utilise twitching motility for
38
phototaxis. Here, we examine twitching motility in the cyanobacterium Synechocystis
39
sp. PCC 6803, which has a spherical cell shape without obvious polarity. We use a
40
motile Synechocystis sp. PCC 6803 strain expressing a functional GFP-tagged PilB1
41
protein to show that PilB1 tends to localise in “crescents” adjacent to a specific region
42
of the cytoplasmic membrane. Crescents are more prevalent under the low light
43
conditions that favour phototactic motility, and the direction of motility strongly
44
correlates with the orientation of the crescent. We conclude that the direction of
45
twitching motility in Synechocystis sp. PCC 6803 is controlled by the localisation of
46
the type IV pilus apparatus, as it is in M. xanthus. The PilB1 crescents in the
47
spherical cells of Synechocystis can be regarded as being equivalent to the leading
48
pole in the rod-shaped cells.
49
50
Introduction
51
52
Many prokaryotes are able to move in response to various signals. They use diverse
53
mechanisms and molecular machines for this movement, resulting not only in
54
different speeds but also different responses to environmental cues. Swimming using
55
flagellar motors and the molecular basis of chemotaxis in Proteobacteria is now
56
probably one of the best understood behavioural systems in prokaryotes (Wadhams
57
& Armitage, 2004). A second motility apparatus is represented by type IV pili (T4P),
58
appendages which are present in many bacteria, including some which also possess
59
flagella (Pelicic, 2008). T4P are evolutionarily related to type II secretion systems
60
(Peabody et al., 2003) and they are implicated not only in twitching motility but also in
61
DNA uptake, biofilm formation, pathogenesis and can act as conductive nanowires
62
(Gorby et al., 2006; Mattick, 2002). The cyanobacterium Synechocystis sp. PCC
2
63
6803 (hereafter Synechocystis) shows T4P-dependent phototactic behaviour in
64
response to unidirectional light irradiation. Speed as well as direction of movement
65
depend on intensity and quality of light. Low intensity red and green light induce
66
positive phototaxis on agar plates towards a light source while high intensity blue
67
light and UV-A illumination lead to a negative phototactic response (Choi et al., 1999;
68
Ng et al., 2003). Three distinct phases of light-induced movement of Synechocystis
69
cells can be distinguished (Bhaya et al., 2006). In phase I, single cells move mostly
70
independently of light direction. With higher cell density, phototactic behaviour
71
increases and results in an accumulation of cells on the front edge of a colony (phase
72
II). Phase III is characterized by a concerted movement of groups of cells along the
73
light gradient thereby generating macroscopic finger-like colony extensions on the
74
agar surface. T4P are needed for all phases of cell movement. Synechocystis cells
75
are densely covered by at least two types of appendages. It was demonstrated that
76
thick pili (6-8 nm diameter; 4-5 µm length) resemble T4P, whereas the biochemical
77
nature and genetic origin of the thinner pili (3-4 nm diameter; 1 µm length) are not
78
understood so far (Bhaya et al., 2000; Yoshihara et al., 2001). Genes encoding pilus
79
subunits and proteins involved in pilus biogenesis are conserved between
80
cyanobacteria and other gram-negative bacteria (Yoshihara & Ikeuchi, 2004). In
81
relation to the well-studied T4P apparatus from other gram negative bacteria such as
82
Pseudomonas aeruginosa (Burrows 2012), the following model can be assumed for
83
functioning of the Synechocystis T4P. The major structural component of the pilus,
84
the major pilin PilA1 and potentially several minor pilins (PilA2 to PilA11) are
85
synthesised as prepilins and incorporated into the cytoplasmic membrane after
86
maturation by the prepilin peptidase PilD. Pilins then polymerise outside the inner
87
membrane probably as a result of a conformational change of the integral membrane
88
protein PilC. This energy-requiring step is driven by the secretion ATPase PilB. The
89
growing pilus is transported through the outer membrane via a PilQ pore complex
90
and attaches to a solid surface. Movement of cells is achieved by retraction and
91
depolymerisation of the pilus catalysed by a second secretion ATPase, PilT. In
92
contrast to flagella-based motility, which depends on control of the direction and
93
speed of rotation of the fixed motor, twitching motility is regulated by dynamic
94
localisation of the two motor proteins PilB and PilT. In Myxococcus xanthus, a rod-
95
shaped bacterium, T4P proteins localise to both cell poles while the motor ATPases
96
show a unipolar localisation to opposite poles with PilB at the leading and PilT at the
3
97
lagging cell pole. Oscillation of the motor ATPases leads to a switch in cell polarity
98
and a reversal of movement (Bulyha et al., 2009). It is likely that the PilT paralogue
99
PilU localises only to the piliated cell pole in Pseudomonas aeruginosa (Chiang et al.,
100
2005). Thus, motility in these rod-shaped bacteria is coupled to cell polarity.
101
102
Synechocystis is a coccoid bacterium without any obvious cell polarity. Here, we
103
analyse localisation of PilB in this organism in relation to single cell movement. This
104
organism has two copies of the pilB gene, pilB1 and pilB2. However, pilus assembly
105
is only lost in pilB1 mutants whereas pilB2 mutants retain pili on the cell surface and
106
are still motile, suggesting that PilB1 is the motor ATPase of the T4P (Yoshihara et
107
al., 2001). Nearly all cyanobacterial PilB1 proteins differ from those of other bacteria
108
by a C-terminal extension without sequence similarity to other PilB homologs
109
(Schuergers et al., 2014). This C-terminal domain is especially interesting because of
110
a conserved four cysteine-containing motif resembling a zinc finger. This C-terminal
111
extension of PilB1 was shown to be involved in binding of the putative RNA
112
chaperone Hfq in Synechocystis. PilB1 is important for correct localisation of Hfq and
113
for its function. Furthermore, correct localisation of Hfq at the pilus base is also
114
important for twitching motility and for formation of both types of pili in Synechocystis
115
(Schuergers et al., 2014). Here, we use a Synechocystis strain expressing a
116
functional GFP-tagged PilB1 protein to show that PilB1 tends to concentrate in
117
“crescents” adjacent to the plasma membrane in a single region of the cell periphery.
118
The direction of twitching motility strongly correlates with the orientation of the
119
crescent, which can be regarded as equivalent to the leading pole in the twitching
120
motility of rod-shaped bacteria.
121
122
METHODS
123
124
Growth conditions, bacterial strains, and mutant analysis. Cultures of motile
125
Synechocystis wild-type (originally obtained from S. Shestakov, Moscow State
126
University, Russia) and mutant strains were propagated on 0.75% (w/v) BG11 agar
127
plates (Rippka et al., 1979) at 30°C under white light of 50 µmol photons m -2 s-1. For
128
induction of the petJ promoter, CuSO4 was omitted from the medium. The
129
construction of a ∆pilB1 mutant and an integrative plasmid for the expression of a C-
130
terminal GFP-tagged PilB1 fusion protein under the control of the copper-sensitive
4
131
PpetJ promoter was described earlier (Linhartová et al., 2014; Schuergers et al.,
132
2014). Absence of entire wild-type copies of pilB1 was tested using following primers:
133
pilB-up:
134
5'-CCAGATTCATGTCGATGGAG-3'. The construction of the pixJ1 mutant is
135
described in Fiedler et al. (2005). Immunoblot analysis was done on whole cell
136
protein extracts following standard procedures using anti-GFP antibody (Abcam).
5'-AATCACCGATGGTCACCATTG-3'
and
pilB-rev:
137
138
Phototaxis assay. Phototactic movement was analysed on 0.5% (w/v) BG11 agar
139
plates supplemented with 0.2% glucose and 10 mM TES buffer (pH 8.0). Cell
140
suspensions were spotted on the plates and incubated for 2-3 days under diffuse
141
white light (50 μmol photons m–2 s–1) before being placed into non-transparent boxes
142
with a one-sided opening (about 5 μmol photons m–2 s–1) for 7 days.
143
144
Epifluorescence microscopy. Micrographs were captured at room temperature
145
(20°C) using an upright Nikon Eclipse Ni-U microscope fitted with a 40x (numerical
146
aperture 0.75) and a 100x oil immersion objective (numerical aperture 1.45). For
147
visualisation of epifluorescence a GFP-filter block (excitation 450-490 nm; emission
148
500-550 nm; exposure time 200 ms) or a Cy3-filter block (excitation 530-560 nm;
149
emission 575-645 nm; exposure time 5 ms) were used. Cells were picked from a
150
moving colony of a phototaxis plate grown as described above, resuspended in a
151
small volume of fresh BG11 medium and 3 µl of the suspension was spotted on 0.3%
152
(w/v) BG11 agarose plates supplemented with 0.2% glucose and 10 mM TES buffer
153
(pH 8.0). Single cell movement was captured at 1 frame per 3 s and fluorescence
154
distribution and cell movement were analysed with Nikon BR software.
155
156
Confocal fluorescence microscopy. Micrographs were recorded with a Leica TCS-
157
SP5 laser-scanning confocal microscope with a 63x oil-immersion objective (NA 1.4).
158
Excitation was at 488 nm from an Argon laser. The confocal pinhole was set to give
159
resolution in the z-direction of ~ 1.5 µm and monochromator-defined emission
160
windows were at 503-515 nm for GFP and 670-720 nm for chlorophyll a. Images
161
were recorded at 12-bit resolution (512x512 pixels) at a line frequency of 400 Hz with
162
6x line averaging. Cells were grown on motility plates as for the phototaxis assay,
163
except that the agar concentration was increased to 1% (w/v), Pieces of agar with the
164
front of the moving colony were cut out and mounted under a cover-slip in a custom5
165
built sample holder. As a control, cells were fixed on the motility plates by careful
166
addition to the surface of the agar of a 5 µl drop of 0.25% glutaraldehyde solution in
167
0.125 M Sørensen’s phosphate buffer pH 7.2.
168
169
RESULTS
170
GFP-tagged PilB1 is functional for phototaxis
171
In order to visualise PilB1 in Synechocystis cells, we constructed a C-terminally GFP
172
tagged PilB1 protein as described in Schuergers et al. (2014) using the superfolder
173
gfp gene (gfp) and the psk9 based expression system which supports the integration
174
of the expression cassette into a neutral site of the chromosome (Kuchmina et al.,
175
2012). PilB1-GFP expression is induced under copper-limited conditions and
176
repressed at 2.5 µM Cu2+. For expression we used a non-motile ΔpilB1 strain. ΔpilB1
177
cells lack natural competence for uptake of DNA via transformation, presumably
178
because of loss of T4P function. Therefore wild-type cells were first transformed with
179
the construct bearing the pilB1-gfp fusion gene, followed by transformation with
180
genomic DNA from a previously described ΔpilB1 strain (Linhartová et al., 2014;
181
Schuergers et al., 2014). PCR analysis showed that the resulting ΔpilB1/pilB1-gfp
182
strain was fully segregated (Fig. 1a) and therefore did not retain any functional copies
183
of the wild-type pilB1 gene. To verify the correct GFP tagging of PilB1, we performed
184
immunoblot analysis on whole cell extracts using α-GFP antibody (Fig. 1b), using as
185
a control a strain (WT/gfp) expressing a FLAG-tagged version of the superfolder GFP
186
protein (29.9 kDa) under the same copper-repressible promoter. A protein of around
187
100 kDa, corresponding to the predicted size of the PilB1-GFP fusion protein (102
188
kDa), was detected specifically in ΔpilB1/pilB1-gfp under inducing conditions (Fig.
189
1b). No smaller bands were detected in the immunoblot, indicating that all detectable
190
GFP in the cell was linked to the full-length PilB1 protein. As we do not have an α-
191
PilB1 antiserum, we were not able to directly compare levels of PilB1 in the wild type
192
and the ΔpilB1/pilB1-gfp strain.
193
For assessment of the phototactic motility of the ΔpilB1/pilB1-gfp strain, cells were
194
spotted on 0.5% BG11-agar that was either Cu2+-free or contained 2.5 µM Cu2+ to
195
inhibit pilB1-gfp expression. When cells were grown under weak directional
196
illumination, the wild-type colony formed finger-like projections extending towards the
197
light source that are characteristic of T4P-based phototactic motility of Synechocystis
198
(Fig. 1c). The ΔpilB1/pilB1-gfp strain displayed phototactic motility only under the
6
199
copper-limited conditions that induce the expression of PilB1-GFP (Fig. 1c),
200
indicating that the expression of PilB1-GFP from the PpetJ promoter restores
201
phototactic motility to the non-motile ΔpilB1 cells. This indicates that T4P are
202
assembled and functional in ΔpilB1/pilB1-gfp under inducing conditions. However,
203
the phototactic motility of ΔpilB1/pilB1-gfp is slower than that of the wild type (Fig.
204
1c).
205
206
PilB1-GFP shows a specific dynamic localisation pattern
207
Epifluorescence microscopy was used to observe the distribution of PilB1-GFP in
208
ΔpilB1/pilB1-gfp cells grown under inducing conditions and incubated under weak
209
illumination. Under these conditions we found a strong tendency for GFP
210
fluorescence to be concentrated at the periphery of the cell, in “crescents” located
211
specifically at one side of the cell (Fig. 2). Such a pattern of GFP localisation is not
212
observed in cells expressing free GFP, or in cells expressing other GFP-tagged
213
proteins associated with the cytoplasmic membrane (Bryan et al., 2014). This
214
indicates that the crescent-like distribution reflects a specific aspect of PilB1
215
behaviour.
216
To further examine PilB1-GFP distribution and dynamics we used confocal
217
fluorescence microscopy for better wavelength resolution and sharper resolution in
218
the z-direction (Fig. 3). Cells were grown on motility plates and placed in a sample
219
holder under the microscope stage. Images of chlorophyll a and GFP fluorescence
220
were recorded at 5 min intervals with negligible light exposure between frames. Fig.
221
3a,b show images of a field of cells packed densely enough to prevent lateral cell
222
movement on these timescales. The chlorophyll fluorescence image shows the
223
location of the thylakoid membranes, which are irregular and asymmetrical in
224
Synechocystis (Liberton et al., 2006). The lack of any significant change in the
225
chlorophyll fluorescence image shows that the cells neither move laterally nor rotate
226
on a 5-min timescale (Fig. 3a,b). Nevertheless, we could observe significant
227
redistribution of GFP fluorescence in some cells during the 5 min dark interval. Fig. 3
228
a,b highlight examples of cells where crescents form or disperse during the 5 min
229
interval. Some cells clearly display relocation of the PilB cresecents to a different part
230
of the cell periphery, suggesting that PilB dissociates from the cytoplasmic
231
membrane and then re-associates with a different area of the membrane (Fig. 3a,b).
232
As a control, we recorded similar images for a field of cells that had been fixed by
7
233
careful application of a small drop of glutaraldehyde solution to the surface of the
234
agar. We could not detect any redistribution of PilB-GFP in the glutaraldehyde-fixed
235
cells (Fig. 3c,d). These results indicate that PilB distribution is dynamic on a
236
timescale of 5 min or less: PilB concentrations at the plasma membrane can form,
237
disperse and relocate on these timescales.
238
239
The direction of twitching motility correlates with the localisation of PilB
240
To determine whether the localisation of PilB is connected with the direction of
241
twitching motility, ΔpilB1/pilB1-gfp cells grown under inducing conditions as
242
described for the phototaxis assays were spotted onto the surface of agar motility
243
plates and illuminated with weak red light on the epifluorescence microscope stage.
244
Cell density was kept low enough to ensure that cells could move freely on the agar
245
surface with only infrequent collisions. GFP fluorescence was observed intermittently
246
with excitation at 470 nm from the condenser light: we took care to minimise the
247
intensity and duration of exposure to this light, as exposure to bright blue light can
248
inhibit phototactic motility (Choi et al., 1999). Under these conditions, cells showed
249
twitching motility in apparently random directions, typically moving one cell diameter
250
(~ 3 µm) in about 45 s (Fig. 4a; Supplementary Movie 1). A strong correlation
251
between the orientation of PilB1-GFP crescents and the direction of movement is
252
immediately apparent (Fig. 4a; Supplementary Movie 1). To quantify this correlation
253
we tracked single cells showing a discernible crescents from three independent
254
experiments (n=58). We quantified this correlation by plotting a frequency distribution
255
for the angle between the direction of cell displacement over a 30 s time-window and
256
the orientation of the PilB1 crescent (a line drawn from the centre of the cell to the
257
middle of the crescent) (Fig. 4b). The mean deviation between the direction of
258
movement and the orientation of the crescent was -2.8° ± 27.4°, showing that almost
259
all cells move roughly in the direction of the crescent (Fig. 4b).
260
Mutants lacking the blue/green light receptor PixJ1 show negative phototaxis under
261
conditions when the wild type shows positive phototaxis (Bhaya et al., 2001; Fiedler
262
et al., 2005; Yoshihara & Ikeuchi, 2004). In order to test whether such mutants still
263
show twitching towards the PilB1 patches, we constructed a ΔpixJ1 mutant in a
264
ΔpilB1/pilB1-gfp background. Although the triple mutant shows negative phototaxis
265
on motility plates (not shown), individual cells still display PilB1-GFP with crescent-
8
266
like shape and the direction of their movement is strongly correlated with the
267
orientation of the crescent (Supplementary Movie 2).
268
269
9
270
271
Discussion
272
In the rod-shaped bacterium M. xanthus, the direction of twitching motility is
273
governed by the localisation of PilB and PilT, ATPases that are suggested to power
274
the extension and retraction of T4P by translocation of the PilA pilin subunits (Bulyha
275
et al., 2009; Jakovljevic et al., 2008). A fully functional apparatus for T4P-based
276
motility is formed by association of PilB and PilT with membrane-spanning protein
277
complexes located at the cell poles, which include PilC in the inner membrane and
278
PilQ in the outer membrane. A switch in direction involves the relocation of PilB and
279
PilT from one pole to the other, while the membrane complexes remain static and
280
located at both poles (Bulyha et al., 2009). Here we explore the localisation of the
281
twitching motility apparatus in Synechocystis, a cyanobacterium with spherical cells,
282
by visualisation of GFP-tagged PilB1. Expression of PilB1-GFP restored phototactic
283
motility to a non-motile ΔpilB1 mutant, confirming that PilB1 is essential for twitching
284
motility in Synechocystis and that PilB1-GFP is functional for motility. However, the
285
speed of long-distance phototaxis was lower in the ΔpilB1/pilB1-gfp strain than in the
286
wild type (Fig. 1), raising the possibility that either the GFP tag or the level of PilB1-
287
GFP expression lower the efficiency of pilus extension or the control of motility
288
direction.
289
As expected for a protein that has been shown in other bacteria to be associated with
290
the cytoplasmic membrane, we found that PilB1-GFP is concentrated around the
291
periphery of the cell (Fig. 2; Fig. 3), where we previously showed interaction and
292
partial co-localisation with Hfq, a protein that modulates transcript accumulation of
293
some RNA species (Schuergers et al., 2014). Under conditions that promote motility,
294
we found a strong tendency of PilB1-GFP to concentrate at one side of the cell, in a
295
configuration resembling a crescent-like shape (Fig. 2; Fig. 3). The localisation of
296
PilB1-GFP is quite dynamic, and the “crescents” sometimes form, disperse, or
297
relocate to another region of the cytoplasmic membrane during a 5 min dark
298
incubation (Fig. 3). PilB and PilT in M. xanthus and P. aeruginosa (Bulyha et al.,
299
2009; Chiang et al., 2005) are restricted to two possible localisations at the
300
cytoplasmic membrane at the two cell poles. By contrast, the range of localisation
301
patterns and relocation behaviour of PilB1 in Synechocystis (Fig. 3) suggests that
302
patches of PilB1 could potentially form anywhere around the cytoplasmic membrane.
303
It is plausible that the T4P membrane-spanning complexes in Synechocystis are
304
dispersed all over the cell periphery, allowing PilB1 (and PilT) to form complexes
10
305
functional for T4P extension and retraction anywhere on the cell surface. Our results
306
indicate that PilB1 distribution is heterogeneous and dynamic, suggesting that, as in
307
M. xanthus, the direction of twitching motility is controlled by relocalisation of the T4P
308
extension and retraction motors. The potential for PilB1 and PilT concentrations to
309
form anywhere at the cell periphery could give Synechocystis the ability to move in
310
any direction, in contrast to M. xanthus which is restricted to a choice of two
311
directions depending on which is the leading pole and which is the lagging pole.
312
To test the idea that the localisation of the T4P motors controls the direction of
313
Synechocystis motility, we simultaneously monitored cell movement and PilB1-GFP
314
localisation, finding a very strong correlation between the position of PilB1 crescents
315
and the direction of movement (Fig. 4; Supplementary Movie 1). Thus the localisation
316
of the T4P extension apparatus coincides with the direction of motility. We have not
317
yet been able to determine the localisation of the PilT retraction motors in
318
Synechocystis, but given that they readily relocate in M. xanthus, it is likely that they
319
can also localise to the same regions of the membrane that are defined by to the
320
PilB1-GFP patches. Given the dynamic nature of PilB1 localisation in Synechocystis
321
(Fig. 3), it is very likely that switches of direction result from relocation of the T4P
322
motors to a different region of the cytoplasmic membrane.
323
Previous studies of Synechocystis phototaxis at the single-cell level suggest that
324
individual cells have ability to perceive and respond to the direction of illumination
325
(Choi et al., 1999; Ng et al., 2003). Photoreceptors including the blue/green light
326
absorbing proteins PixJ1 and Cph2, the blue-light receptor PixE and the UV-A
327
receptor UirS are involved in the regulation of phototaxis (Okajima et al., 2005a;
328
Song et al., 2011; Wilde et al., 2002; Yoshihara et al., 2000). Inactivation of pixJ1,
329
pixE as well as uirS (pixA) led to a reversal of cell movement on agar plates (Okajima
330
et al., 2005b; Song et al., 2011; Yoshihara et al., 2000). However, we found that loss
331
of PixJ1 had no obvious effect on the formation of PilB1-GFP crescents, and ΔpixJ1
332
cells still show a strong correlation between the direction of movement and the
333
orientation of the PilB1-GFP crescents (mean deviation 3.93° ± 39.8°, n=29)
334
(Supplementary Movie 2). Therefore ΔpixJ1 cells retain the same basic mechanism
335
of motility, but must be specifically impaired in the perception of light direction. Both
336
PixJ1 and UirS contain transmembrane domains and therefore these photoreceptors
337
could be in close proximity to the T4P motors. However, the details of the signalling
338
mechanism, and the mechanism of light direction sensing remain to be determined.
11
339
340
Acknowledgements
341
We thank Werner Bigott and Anne Leisinger for excellent technical assistance. DJN
342
was supported by a Queen Mary University of London studentship.
343
344
References
345
346
347
Bhaya, D., Bianco, N. R., Bryant, D. & Grossman, A. (2000). Type IV pilus
biogenesis and motility in the cyanobacterium Synechocystis sp. PCC6803. Mol
Microbiol 37, 941–951.
348
349
350
Bhaya, D., Takahashi, A. & Grossman, A. R. (2001). Light regulation of type IV
pilus-dependent motility by chemosensor-like elements in Synechocystis
PCC6803. Proc Natl Acad Sci USA 98, 7540–7545.
351
352
353
Bhaya, D., Nakasugi, K., Fazeli, F. & Burriesci, M. S. (2006). Phototaxis and
impaired motility in adenylyl cyclase and cyclase receptor protein mutants of
Synechocystis sp. strain PCC 6803. J Bacteriol 188, 7306–7310.
354
355
356
357
Bryan, S. J., Burroughs, N. J., Shevela, D., Yu, J., Rupprecht, E., Liu, L.-N.,
Mastroianni, G., Xue, Q., Llorente-Garcia, I. & other authors. (2014).
Localisation and interactions of the Vipp1 protein in cyanobacteria. Mol
Microbiol, in press
358
359
360
361
Bulyha, I., Schmidt, C., Lenz, P., Jakovljevic, V., Höne, A., Maier, B., Hoppert, M.
& Søgaard-Andersen, L. (2009). Regulation of the type IV pili molecular
machine by dynamic localization of two motor proteins. Mol Microbiol 74, 691–
706.
362
363
Burrows, L. L. (2012). Pseudomonas aeruginosa twitching motility: Type IV pili in
action. Annu Rev Microbiol 66, 493–520.
364
365
366
Chiang, P., Habash, M. & Burrows, L. L. (2005). Disparate subcellular localization
patterns of Pseudomonas aeruginosa Type IV pilus ATPases involved in
twitching motility. J Bacteriol 187, 829–839.
367
368
369
370
Choi, J. S., Chung, Y. H., Moon, Y. J., Kim, C., Watanabe, M., Song, P. S., Joe, C.
O., Bogorad, L. & Park, Y. M. (1999). Photomovement of the gliding
cyanobacterium Synechocystis sp. PCC 6803. Photochem Photobiol 70, 95–
102.
371
372
373
Fiedler, B., Börner, T. & Wilde, A. (2005). Phototaxis in the cyanobacterium
Synechocystis sp. PCC 6803: role of different photoreceptors. Photochem
Photobiol 81, 1481–1488.
374
375
376
Gorby, Y. A., Yanina, S., McLean, J. S., Rosso, K. M., Moyles, D., Dohnalkova,
A., Beveridge, T. J., Chang, I. S., Kim, B. H. & other authors. (2006).
Electrically conductive bacterial nanowires produced by Shewanella oneidensis
12
377
378
strain MR-1 and other microorganisms. Proc Natl Acad Sci USA 103, 11358–
11363.
379
380
381
Jakovljevic, V., Leonardy, S., Hoppert, M. & Søgaard-Andersen, L. (2008). PilB
and PilT are ATPases acting antagonistically in type IV pilus function in
Myxococcus xanthus. J Bacteriol 190, 2411–2421.
382
383
384
Liberton, M., Howard Berg, R., Heuser, J., Roth, R. & Pakrasi, H. B. (2006).
Ultrastructure of the membrane systems in the unicellular cyanobacterium
Synechocystis sp. strain PCC 6803. Protoplasma 227, 129–138.
385
386
387
388
Linhartová, M., Bučinská, L., Halada, P., Ječmen, T., Šetlík, J., Komenda, J. &
Sobotka, R. (2014). Accumulation of the Type IV prepilin triggers degradation of
SecY and YidC and inhibits synthesis of Photosystem II proteins in the
cyanobacterium Synechocystis PCC 6803. Mol Microbiol, in press
389
390
Mattick, J. S. (2002). Type IV pili and twitching motility. Annu Rev Microbiol 56, 289–
314.
391
392
Ng, W.-O., Grossman, A. R. & Bhaya, D. (2003). Multiple light inputs control
phototaxis in Synechocystis sp. Strain PCC6803. J Bacteriol 185, 1599–1607.
393
394
395
396
Okajima, K., Yoshihara, S., Fukushima, Y., Geng, X., Katayama, M., Higashi, S.,
Watanabe, M., Sato, S., Tabata, S. & other authors. (2005a). Biochemical and
functional characterization of BLUF-type flavin-binding proteins of two species of
cyanobacteria. J Biochem 137, 741–750.
397
398
399
400
Okajima, K., Yoshihara, S., Fukushima, Y., Geng, X., Katayama, M., Higashi, S.,
Watanabe, M., Sato, S., Tabata, S. & other authors. (2005b). Biochemical and
functional characterization of BLUF-type flavin-binding proteins of two species of
cyanobacteria. J Biochem 137, 741–750.
401
402
403
Peabody, C. R., Chung, Y. J., Yen, M.-R., Vidal-Ingigliardi, D., Pugsley, A. P. &
Saier, M. H. (2003). Type II protein secretion and its relationship to bacterial
type IV pili and archaeal flagella. Microbiology 149, 3051–3072.
404
Pelicic, V. (2008). Type IV pili: e pluribus unum? Mol Microbiol 68, 827-837.
405
406
407
Rippka, R., Deruelles, J., Waterbury, J. B., Herdman, M. & Stanier, R. Y. (1979).
Generic assignments, strain histories and properties of pure cultures of
cyanobacteria. J Gen Microbiol 111, 1–61.
408
409
410
411
Schuergers, N., Ruppert, U., Watanabe, S., Nürnberg, D. J., Lochnit, G., Dienst,
D., Mullineaux, C. W. & Wilde, A. (2014). Binding of the RNA chaperone Hfq to
the type IV pilus base is crucial for its function in Synechocystis sp. PCC 6803.
Mol Microbiol 92, 840–852.
412
413
414
415
Song, J. Y., Cho, H. S., Cho, J. Il, Jeon, J. S., Lagarias, J. C. & Park, Y. Il. (2011).
Near-UV cyanobacteriochrome signaling system elicits negative phototaxis in
the cyanobacterium Synechocystis sp. PCC 6803. Proc Natl Acad Sci USA 108,
10780–10785.
13
416
417
Wadhams, G. H. & Armitage, J. P. (2004). Making sense of it all: bacterial
chemotaxis. Nat Rev Mol Cell Biol 5, 1024–1037.
418
419
Wilde, A., Fiedler, B. & Börner, T. (2002). The cyanobacterial phytochrome Cph2
inhibits phototaxis towards blue light. Mol Microbiol 44, 981–988.
420
421
422
423
Yoshihara, S., Suzuki, F., Fujita, H., Geng, X. X. & Ikeuchi, M. (2000). Novel
putative photoreceptor and regulatory genes required for the positive phototactic
movement of the unicellular motile cyanobacterium Synechocystis sp. PCC
6803. Plant Cell Physiol 41, 1299–1304.
424
425
426
427
Yoshihara, S., Geng, X., Okamoto, S., Yura, K., Murata, T., Go, M., Ohmori, M. &
Ikeuchi, M. (2001). Mutational analysis of genes involved in pilus structure,
motility and transformation competency in the unicellular motile cyanobacterium
Synechocystis sp. PCC 6803. Plant Cell Physiol 42, 63–73.
428
429
430
Yoshihara, S. & Ikeuchi, M. (2004). Phototactic motility in the unicellular
cyanobacterium Synechocystis sp. PCC 6803. Photochem Photobiol Sci 3, 512–
518.
431
432
14
433
Figure legends
434
Fig. 1: The PilB1-GFP fusion protein is properly expressed and restores
435
motility of a ΔpilB1 mutant in a phototaxis assay. (a) Complete segregation of
436
mutant ΔpilB1 copies in the ΔpilB1/pilB1-gfp strain was verified via PCR (b)
437
Immunoblot analysis of total protein from ΔpilB1/pilB1-gfp cells in comparison to a
438
strain expressing the free GFP protein using the PpetJ based expression cassette
439
under inducing (no Cu2+) and repressing (2.5 µM Cu2+) conditions. (c) Synechocystis
440
wild-type and ΔpilB1/pilB1-gfp cells were spotted onto 0.5% agar motility plates with
441
or without copper and incubated for 7 days under unidirectional illumination. Dotted
442
lines show the initial position of spotted cells.
443
444
Fig. 2: PilB1-GFP tends to localise in a non-uniform crescent-like pattern along
445
the cell periphery. ΔpilB1/pilB1-gfp cells grown on copper-free BG11 plates were
446
resuspended in fresh BG11 medium, spotted on glass slides and observed by
447
epifluorescence microscopy under dim illumination (< 0.5 μmol photons m –2 s–1). (a)
448
Phase contrast image (b) chlorophyll fluorescence (c) GFP fluorescence (d) merged
449
signals from GFP and chlorophyll fluorescence.
450
451
Fig. 3: Dynamic redistribution of PilB-GFP fluorescence within cells.
452
ΔpilB1/pilB1-gfp cells were grown under weak directional illumination on copper-free
453
1% agar motility plates. Pieces of agar were cut from the plates, mounted under a
454
coverslip and imaged by confocal fluorescence microscopy, with simultaneous
455
visualization of GFP (green) and chlorophyll (red) fluorescence. The images show a
456
region where cells were too densely packed to move. (a) Image recorded
457
immediately after mounting the sample. (b) Image of the same field of cells recorded
458
5 min later, with the cells kept under negligible light in the interval. The coloured dots
459
highlight cells where various kinds of redistribution of GFP fluorescence were
460
observed: formation or dispersal of bright patches of GFP fluorescence at the cell
461
periphery, or relocation of patches of GFP fluorescence (as indicated by
462
simultaneous loss of GFP fluorescence at one part of the cell periphery, and
463
increased GFP fluorescence at another part of the cell periphery). Note, however,
464
that in the case of fluorescence losses we cannot exclude the possibility that these
465
are due to local GFP bleaching rather than redistribution of PilB-GFP. (c,d) As a
466
control images were recorded with a 5 min interval as in (a,b) except that cells were
15
467
fixed with a drop of 0.25% glutaraldehyde solution before cutting agar blocks from the
468
motility plates.
469
Fig. 4: PilB1-GFP localisation correlates with the direction of single cell
470
motility. ΔpilB1/pilB1-gfp cells grown on copper-free BG11 phototaxis plates were
471
resuspended in fresh medium, spotted on 0.3% agarose motility plates and
472
illuminated with red light (cut-off filter, LEE filters Nr. 106) at 0.5 μmol photons m –2
473
s-1. After 10 min, PilB1-GFP localisation and cell movement were monitored for 1 min
474
by imaging fluorescence at 3 s intervals. (a) Sequential fluorescence images of two
475
representative cells. The initial position of the moving cells is indicated by a dashed
476
circle. See also Supplementary Movie 1. (b) Frequency histogram of the angle (α)
477
between the centre of PilB1 orientation at 0 s and the cell displacement after 30 s
478
(see box). 59 cells showing single cell motility and a clear crescent-like PilB1
479
localisation from three different motility plates were analysed. A fitted normal
480
distribution is shown by the dashed line.
481
482
Supplementary Movie 1: Twitching motility of ΔpilB1/pilB1-gfp cells and subcellular
483
distribution of PilB1-GFP. Frames were recorded at 3 s intervals over a 60 s period.
484
See also Fig. 4.
485
486
Supplementary Movie 2: Twitching motility of ΔpixJ1/ΔpilB1/pilB1-gfp cells and
487
subcellular distribution of PilB1-GFP. Frames were recorded at 3 s intervals over a
488
60 s period.
489
490
16