Molecules of messenger RNA, or mRNA for short, contain the instructions needed to make proteins. Many mRNAs are only found in certain parts of the cell to ensure that the corresponding proteins are only produced where they are actually needed. The mRNAs are delivered to their final location in the cell by motor proteins that move along tracks made of filaments called microtubules.

In female fruit flies, a mRNA called bicoid is transported to front end of a developing egg cell, while another mRNA called oskar is moved to the rear end. When the egg is fertilized, the region that contains bicoid mRNA develops into the head of the embryo, while the other end gives rise to the abdomen. A motor protein called kinesin-1 transports the oskar mRNA to the rear end, but how bicoid mRNA moves to the front end is not clear.

Trovisco et al. used microscopy to study how bicoid mRNA moves. The experiments show that another motor protein called Dynein moves bicoid mRNA along microtubules. However, unlike oskar mRNA, bicoid mRNA moves along microtubules in all directions and is not biased towards the front end of the cell. Trovisco et al. hypothesized that when bicoid mRNA reaches the front end of the egg it is trapped there by other factors.

Further experiments found that bicoid mRNA is indeed anchored at the front end of the cell. The mRNA does not seem to be trapped at the ends of the microtubules along which it is transported, nor does it form large clumps. Instead, it forms small, well-defined particles that remain the same size as the egg develops. The findings of Trovisco et al. raise the possibility that bicoid mRNA is packaged into these particles in order to be transported and anchored at the front end of the egg cell. Future work is needed to understand how particles containing bicoid mRNA are tethered at the front end of the egg cell and whether other mRNAs are also packaged in a similar manner.

mRNA localisation is a widely-used mechanism for targeting proteins to the regions of the cell where they are required and is often coupled to translational repression to prevent expression of the encoded protein until after its transcript is localised (Lécuyer et al., 2007; Jambor et al., 2015). This is particularly important during axis formation in organisms such as Drosophila and Xenopus where mRNAs localise during oogenesis to provide the primary patterning signals for the embryo. In Drosophila, the anterior-posterior axis is determined by the microtubule-dependent localisation of bicoid (bcd) and oskar (osk) mRNAs to the anterior and posterior poles of the oocyte, respectively (Pokrywka and Stephenson, 1991; Clark et al., 1994; Roth et al., 1995). bcd mRNA is translationally repressed during oogenesis and is only translated when the egg is laid, providing a local source of Bcd protein, which diffuses to form a morphogen gradient that patterns the anterior half of the embryo (Ephrussi and St Johnston, 2004). By contrast, osk mRNA is translated when it reaches the posterior of the oocyte to produce long and short isoforms of Oskar protein (Markussen et al., 1995; Rongo et al., 1995). Long Oskar anchors its own RNA, whereas short Oskar nucleates the polar granules, leading to the posterior recruitment of the germ line determinants and the abdominal determinant, nanos mRNA (Wang and Lehmann, 1991; Ephrussi and Lehmann, 1992; Vanzo and Ephrussi, 2002).

Both bcd and osk mRNAs are transcribed in the nurse cells within the germline cyst and are then transported along microtubules through the ring canals into the oocyte by Dynein (Clark et al., 2007; Mische et al., 2007). The localisation of osk mRNA to the posterior of the oocyte requires the plus end-directed microtubules motor protein, Kinesin-I (Brendza et al., 2000). Live imaging of fluorescently-labelled osk mRNA reveals that it forms particles that undergo rapid movements in all directions with a slight posterior bias, indicating that the RNA takes a biased random walk to the posterior cortex, where it is then anchored (Zimyanin et al., 2008). Since almost all osk mRNA movements depend on Kinesin-I, the microtubule cytoskeleton appears to be largely disordered, with a small excess of microtubule plus ends pointing posteriorly. This is consistent with measurements of the direction of growing microtubule plus ends, which reveal that most grow from the anterior/lateral cortex and extend in random directions with a weak orientation bias that is stronger close to the posterior pole (Parton et al., 2011).

How bcd mRNA is targeted to the anterior of the oocyte at mid-oogenesis is less well understood. Disrupting the Dynein/Dynactin complex by over-expressing Dynamitin causes either a posterior spreading or complete delocalisation of bcd mRNA, suggesting that the RNA is localised by Dynein-dependent, minus end-directed transport along microtubules (Clark et al., 1997; Duncan and Warrior, 2002; Januschke et al., 2002). However, as Dynein is also required for bcd mRNA transport into the oocyte, it is hard to distinguish direct from indirect effects. Furthermore, injected naïve bcd mRNA accumulates at the nearest region of the anterior/lateral cortex to its site of origin and not specifically at the anterior, consistent with the observation that microtubules ends are anchored or nucleated from all of the cortex except the very posterior pole (Cha et al., 2001). Only pre-treatment with nurse cell cytoplasm renders in vitro transcribed RNA competent to localise specifically to the oocyte anterior, and this conditioning requires the the pseudonuclease Exuperantia (Exu) (Cha et al., 2001). The role of Exu in the localisation of bcd mRNA requires its homo-dimerisation and RNA binding (Lazzaretti et al., 2016). Finally, computer simulations of the microtubule network in the anterior region of the oocyte suggest that it has little orientation bias, making it unlikely that RNA movement towards microtubules minus ends can account for the rapid anterior accumulation of the RNA (Trong et al., 2015).

The number of genes required for bcd mRNA localisation increases as oogenesis proceeds, suggesting distinct mechanisms localise the RNA at different stages. exu is required at all stages of localisation, whereas swallow, the γ-tubulin ring complex (γ-TURC), staufen and the ESCRT-II complex are only needed from stage 10b onwards (Berleth et al., 1988; St Johnston et al., 1989; Ferrandon et al., 1994; Irion and St Johnston, 2007; Schnorrer et al., 2000, 2002; Weil et al., 2006). The γ-TURC forms a new microtubule organising centre (MTOC) in the middle of the anterior cortex at this stage and this coincides with the re-localisation of bcd mRNA from an anterior ring into a central disc adjacent to the MTOC. Furthermore, live imaging of stage 10b-12 oocytes reveals that bcd mRNA particles move towards the anterior in a Dynein-dependent manner (Weil et al., 2006, 2008). Since the RNA localisation is labile, it has been proposed that it is not anchored at the anterior at this stage and continually diffuses away, to be re-localised by Dynein-mediated transport (Weil et al., 2008).

Here we use fast live imaging, fluorescence recovery after photo-bleaching, photo-conversion and super-resolution microscopy to investigate the mechanism of the initial localisation of bcd mRNA to the anterior at stage 9. Our results indicate that at this stage the mRNA is not localised by continual directed transport, but by random active transport and anterior anchoring.

In order to visualise bcd mRNA in living oocytes, we took advantage of the MS2-system for fluorescently labelling RNA in vivo, in which the RNA contains multiple MS2 stem-loops that are bound by the MS2 coat protein (MCP) coupled to a fluorescent reporter (Bertrand et al., 1998; Forrest and Gavis, 2003; Weil et al., 2008). The original genomic bicoid-MS2 transgenes expressed full-length bcd mRNA from its endogenous promoter fused to 6 copies of the MS2 stem loop (Weil et al., 2006), but the relatively low numbers of MCP-GFP bound per RNA and the low expression levels made the RNA hard to image, particularly in fast moving particles. We therefore generated a construct in which the maternal α4 tubulin promoter drives the expression of the bcd 3’UTR fused to 11 copies of the MS2 stem-loop (bcdMS2), as the 3’UTR is sufficient for all steps in bcd RNA localisation (Macdonald and Struhl, 1988). The removal of the coding region allowed us to express this RNA at higher levels without disrupting embryonic development by expanding the Bcd morphogen gradient (Namba et al., 1997). Co-expression of bcdMS2 with MCP-GFP (bcd*GFP) gave strong labelling of bcd mRNA, which showed an identical localisation to the endogenous transcript at all stages of oogenesis (Figure 1A–B, data not shown).

Figure 1 with 1 supplement see all

Download asset Open asset

Kinesin-I antagonises the Dynein-dependent transport of bcd mRNA

Fast, high magnification wide-field imaging of bcd*GFP in stage 9 oocytes revealed many small RNA particles that moved at speeds of up to 2.2 μm/sec (Figure 1C–D, Video 1). All movements were abolished by treatment with the microtubule-depolymerising drug, Colcemid, whereas the actin depolymeriser, Cytocholasin D, caused premature cytoplasmic streaming but had no effect on particle motility (Figure 1—figure supplement 1A, Video 2, data not shown). These results are consistent with the observation that bcd mRNA localisation at all stages of oogenesis is disrupted by microtubule-depolymerising drugs, and supports the view that particle movements play a role in delivering the mRNA to the oocyte anterior (Pokrywka and Stephenson, 1991; Weil et al., 2006).

Video 1

Download asset

Video 2

Download asset

The bcd mRNA particles in oocytes with one copy of bcdMS2 moved with an average velocity of 0.64 μm/sec, which is significantly faster than osk mRNA particles (0.47 μm/sec) imaged under equivalent conditions (Zimyanin et al., 2008) (Table 1, Table 1—source data 1, Figure 1D). This difference was even more marked when we imaged egg chambers expressing two copies of bcdMS2, with the mean velocity increasing to 0.78 μm/s (Table 1, Table 1—source data 1, Figure 1D). This increase is presumably because the signal from the fastest particles is spread across more camera pixels per frame (6 pixels for particles moving at 2 μm/s, imaged for 0.25 s), making them harder to detect. Doubling their brightness therefore increases the efficiency of detection of the fastest particles.

Table 1

Genotype (2x bcd*GFP)	Tracks (n)	Movies (n)	Oocytes (n)	Anterior Mov (%/n)	Binomial P-value	a)	Speed ±S.E.M. (µm/s)	Wilcoxon P-value	b)	Mixed-effects P-value	c)	Track distance ±S.E.M. (µm)	Anterior displacement ±S.E.M. (µm/s)	Wilcoxon P-value	d)
Wild-type	1181	32	9	52.6 / 621	0.040	*	0.78 / 0.01					1.41 / 0.03	0.02 / 0.02	0.200
Dhc6-10/8-1	669	7	5	50.8 / 340	0.350		0.50 / 0.01	<2.2E-16	***	6.90E-10	***	1.61 / 0.05	0.01 / 0.01	0.500
exu1	215	17	4	53.0 / 114	0.207		0.47 / 0.02	<2.2E-16	***	3.10E-09	***	1.39 / 0.07	0.03 / 0.02	0.230

Genotype (1x bcd*GFP)	Tracks (n)	Movies (n)	Oocytes (n)	Anterior Mov (%/n)	Binomial P-value	a)	Speed ±S.E.M. (µm/s)	Wilcoxon P-value	b)	Mixed-effects P-value	c)	Track distance ±S.E.M. (µm)	Anterior displacement ±S.E.M. (µm/s)	Wilcoxon P-value	d)
Wild-type	450	22	13	54.4 / 245	0.033	*	0.64 / 0.01					0.98 / 0.05	0.02 / 0.02	0.135
Khc27 GLC	921	14	6	53.7 / 495	0.013	*	0.94 / 0.01	<2.2E-16	***	1.60E-06	***	1.55 / 0.04	0.01 / 0.02	0.272
Khc17 GLC	639	21	12	53.2 / 340	0.057		0.89 / 0.02	<2.2E-16	***	2.00E-04	***	1.60 / 0.05	0.02 / 0.02	0.401
Khc23 GLC	612	15	5	53.9 / 330	0.029	*	1.01 / 0.02	<2.2E-16	***	2.20E-06	***	1.67 / 0.06	0.06 / 0.03	0.044	*
Dhc8-1/+	141	7	4	- - -	- - -		0.38 / 0.02	3.61E-16	***	3.00E-04	***	1.15 / 0.08	0.06 / 0.02	- - -
Dhc8-1 GLC	43	16	7	- - -	- - -		0.25 / 0.02	<2.2E-16	***	3.10E-07	***	0.72 / 0.06	-0.05 / 0.03	- - -

1. a) Binomial test for the frequency of anterior-directed movements being >50% (one-tailed)

2. b) Wilcoxon rank sum test for speed comparisons - comparison to wild-type (2x bcd*GFP or 1x bcd*GFP)

3. c) Mixed-effects linear model (LMER) test for speed comparisons – comparison to wild-type (2x bcd*GFP or 1x bcd*GFP). Fixed Effect: Genotype; Random Effects: Variability between oocytes and movies

4. d) Wilcoxon 1-sample test for the net anterior displacement. Null hypothesis: mean=0 (two-tailed)

5. *p<0.05; **p<0.01; ***p<0.001

6. - - - Not applicable / Not done

Table 1—source data 1

Tracking of bcd*GFP particles in wild-type and mutant stage 9 oocytes.

Includes the data in: Table 1; Figure 1 panels D–F; Figure 2, panels A–E; Figure 2—figure supplement 1, panel C.

https://doi.org/10.7554/eLife.17537.008

Download elife-17537-table1-data1-v4.xlsx

The observation that bcd mRNA particles move at nearly twice the average speed of osk mRNA particles suggests that they are transported by different motor proteins. The most likely candidate for a motor that transports bcd mRNA is cytoplasmic Dynein, since putative microtubule minus end markers are enriched anteriorly (Clark et al., 2007, 1997). It is not possible to test null mutations in Dynein components, as these block oocyte determination. We therefore used a combination of hypomorphic alleles of the Dynein heavy chain, Dhc^6-10/Dhc^8-1, that has previously been shown to reduce the speed of mRNA movement towards the microtubule minus ends in the embryo (Bullock et al., 2006). bcd mRNA particles moved with a mean velocity of 0.50 μm/sec in Dhc^6-10/Dhc^8-1 compared with 0.78 μm/sec in wild-type oocytes (Table 1, Table 1—source data 1, Figure 1E, Figure 1—figure supplement 1C). We observed a similar reduction in the velocities of bcd mRNA particles in Dhc^8-1/+ heterozygotes (59% of wild-type; Table 1, Table 1—source data 1, Figure 1E, Figure 1—figure supplement 1C), suggesting that Dhc^6-10 has little effect on motor speed and that Dhc^8-1 has a dominant negative effect. Consistent with this, homozygous mutant germline clones of Dhc^8-1 showed an even greater reduction in particle velocity to 39% of the wild-type speed (Table 1, Table 1—source data 1, Figure 1E, Figure 1—figure supplement 1C). Thus, this allele produces a functional motor protein that still moves, but significantly more slowly than the wild-type protein. The strong reduction in the speed of bcd mRNA particles in Dhc^8-1 homozygotes indicates that the majority are transported by Dynein. The Dhc^8-1 mutant also significantly reduces the amount of bcd mRNA localised to the anterior (Figure 1—figure supplement 1D). This does not seem to be due to reduced frequency of movement because the mobile fraction of bcd mRNA particles is unaffected in Dhc^6-10/Dhc^8-1 mutants (26% in wild-type versus 22% in mutant; Table 2, Table 2—source data 1, Figure 1G). Thus, slower Dynein-dependent transport, presumably both from the nurse cells into the oocyte and within the oocyte, impairs the delivery of bcd mRNA to the oocyte anterior.

Table 2

Genotype	Oocytes	Mobile fraction/5 s	T-test P-value	a)
bcd*GFP/+	5	0.22 ± 0.03
bcd*GFP/+;Khc27 GLC	5	0.47 ± 0.02	0.0004	***
bcd*GFP/+;Khc23 GLC	4	0.25 ± 0.02	0.624
bcd*GFP/bcd*GFP;Dhc6-10/8-1	4	0.26 ± 0.05	0.556
bcd*GFP/bcd*GFP;exu1	4	0.07 ± 0.01	0.006	**

1. a) T-test for comparison of mobile fractions (two-tailed) - comparisons to bcd*GFP/+ (wild-type)

2. **p<0.01; ***p<0.001

Table 2—source data 1

Mobile fraction of bcd*GFP particles in wild-type and mutant stage 9 oocytes.

Includes the data in: Table 2; Figure 1, panel G.

https://doi.org/10.7554/eLife.17537.010

Download elife-17537-table2-data1-v4.xlsx

Null mutations in the Kinesin heavy chain (Khc) also disrupt the localisation of bcd mRNA, with the majority of oocytes showing spreading of the RNA along the anterior and lateral cortex (Januschke et al., 2002) (Figure 1—figure supplement 1E). It is unclear whether this phenotype arises because Kinesin-I plays a direct role in the transport and/or anchoring of bcd mRNA. Kinesin-I transports Dynein to the oocyte posterior, indicating that the two motors can associate in the same complex, and Kinesin-I could therefore affect bcd mRNA indirectly, for example by recycling Dynein to the oocyte posterior for further rounds of minus end-directed transport, or by modulating the activity of Dynein (Januschke et al., 2002; Palacios and St Johnston, 2002). To test the role of Kinesin-I directly, we analysed the movement of bcd mRNA particles in germline clones of the null allele, Khc²⁷(Brendza et al., 2000). Surprisingly, the velocity of bcd mRNA particle movements was significantly increased in the absence of Kinesin-I, with an average speed of 0.98 μm/sec, compared to 0.64 μm/sec in wild-type (Table 1, Table 1—source data 1, Figure 1E–F, Figure 1—figure supplement 1C). This increase could be explained if a fraction of the bcd mRNA particles are transported at low speeds by Kinesin-I, so that its loss raises the average velocity by removing the slow population of moving particles. If so, one would expect the fraction of mobile particles to be reduced in Khc²⁷. In wild-type, approximately 20% of the bcd mRNA particles move over a five second period (Table 2, Table 2—source data 1, Figure 1G), which is nearly twice the proportion observed for osk mRNA (Zimyanin et al., 2008). In Khc²⁷ mutants, the mobile fraction more than doubled to 47% (Table 2, Table 2—source data 1, Figure 1G). Thus, both the speed and frequency of bcd mRNA particle motility are increased in the absence of Kinesin-I, making it highly unlikely that this motor is responsible for a significant proportion of the movements. Furthermore, bcd mRNA particles showed virtually no reversals of movement in either wild-type or Khc²⁷ mutants (1.9% versus 1.5%, respectively), suggesting the RNA does not alternate between transport by motors of opposing polarity.

Another way to test the role of Kinesin-I is to analyse two mutants, Khc¹⁷ and Khc²³, with single amino acid changes in the motor domain that reduce the speed of Kinesin-I movement without affecting its other properties (Brendza et al., 1999; Serbus et al., 2005; Zimyanin et al., 2008). Like the null allele, germline clones of Khc¹⁷ and Khc²³ also caused a significant increase in the mean velocity of bcd mRNA particle movements (0.89 μm/sec and 1.0 μm/sec, respectively) (Table 1, Table 1—source data 1, Figure 1E). This rules out the possibility that Kinesin-I is responsible for the slow movements of bcd mRNA particles, as this would result in a decrease in the average velocity in the mutants. Kinesin-I must therefore act by some other mechanism to reduce the speed of movement by another motor, presumably Dynein, for example by engaging in a tug of war. The slow allele, Khc²³, has little effect on the fraction of mobile particles (25%), compared to 47% in the null allele (Table 2, Table 2—source data 1, Figure 1G), indicating that the effects on speed and frequency are separable. One possible explanation for this difference is that the slow allele still transports the Dynein/Dynactin complex to the posterior pole of the oocyte, albeit more slowly, whereas the null allele does not (Januschke et al., 2002) (Figure 1H). The absence of Dynein transport to the posterior in the null mutant will therefore increase the concentration of Dynein at the anterior of the oocyte, which could account for the more than doubling of the frequency of particle movement in this region.

The localisation of bcd mRNA at all stages depends on Exu protein (Berleth et al., 1988) and we therefore also examined the behaviour of bcd mRNA particles in the mutant, exu¹, which reduces the affinity of Exu to RNA (Lazzaretti et al., 2016). Very few particles were visible in exu¹ homozygous oocytes, and these moved significantly less frequently than normal (7%) and at a reduced speed (Table 1, Table 1—source data 1, Table 2, Table 2—source data 1, Figure 1E,G, Figure 1—figure supplement 1C). Thus, Exu is required for both the formation of bcd mRNA particles and their efficient transport on microtubules, consistent with the dimerisation of Exu (possibly leading to the dimerisation of bcd mRNA) and the results obtained from bcd mRNA injections (Cha et al., 2001; Lazzaretti et al., 2016). The residual bcd mRNA motility in the mutant may explain why a small amount of RNA is still diffusely localised at the anterior of exu mutant oocytes (Figure 1—figure supplement 1F).

Lack of a strong directional bias in bcd mRNA particle movement

We next assessed the directionality of bcd mRNA particle movements to determine if it could account for the anterior accumulation of the mRNA. We found only a slight excess of movements towards the anterior compared to the posterior over a region up to 40 μm from the anterior of the oocyte (52.6% versus 47.4%, p=0.04) (Figure 2A). The particles were tracked on near-surface optical sections, where they were better detected, but a similarly weak directional bias was also observed in deeper optical sections (Figure 2—figure supplement 1A–B). The velocity of the movements in each direction was not significantly different, and the net displacement was also not significantly different from zero (Table 1, Table 1—source data 1, Figure 2A–C). To test whether this bidirectional movement reflected motors moving in opposite directions along a strongly polarised microtubule cytoskeleton, or mainly unidirectional transport along a weakly polarised cytoskeleton, we measured the velocity in each direction in slow Dynein mutants. All Dhc mutant combinations reduced the velocity of posterior movements to the same extent as of anterior movements (Figure 2B,D). Thus, Dynein is responsible for the majority of particle movements both towards and away from the anterior cortex, and the absence of a strong bias in the direction of bcd mRNA transport is due to the very weak polarisation of the microtubule cytoskeleton in this anterior region. This is in good agreement with tracking of plus-ends of microtubules (Parton et al., 2011) and computer simulations of the oocyte microtubule network, which predict almost no bias in the orientation of microtubules near the anterior and a stronger bias in the posterior (Trong et al., 2015). The exu mutant caused a similar reduction in speed in both directions (Figure 2B), whereas Khc mutants increased the speed in both directions, consistent with the unpolarised nature of the microtubule network at the anterior (Figure 2D–E).

Figure 2 with 1 supplement see all

Download asset Open asset

Figure 2—source data 1

Tracking of bcd*Tom particles in wild- stage 9 oocytes.

Includes the data in: Figure 2—figure supplement 1, panel B.

https://doi.org/10.7554/eLife.17537.012

Download elife-17537-fig2-data1-v4.xlsx

We classified the particle movements according to their distance from the anterior cortex to determine if the bias varies across this region (Figure 2A). This revealed that there is a strong anterior bias of 12% in the movements of particles that are more than 10 μm from the anterior cortex (39.6% versus 27.6%), but this decreases to 6% in the region 5–10 μm from the anterior and reverses in the 0-5 μm region (20.8% versus 23.9%) to give a 3% excess of movements away from the anterior (Figure 2A). Similar results were obtained when tracking bcd mRNA particles deeper in the oocyte (Figure 2—figure supplement 1A–B, Figure 2—source data 1). Our analysis therefore indicates that although Dynein-dependent transport is required for bcd mRNA localisation, it is not sufficient to explain its robust anterior accumulation, because Dynein moves the RNA in and out of the anteriormost region at similar rates. This is incompatible with a model in which bcd mRNA is maintained by continuous anteriorly-directed transport, as has been proposed to occur at later stages of oogenesis (Weil et al., 2006). Instead, the data suggest that bidirectional transport facilitates the delivery of bcd mRNA particles to the anterior where they are specifically sequestered by some other mechanism.

To further test whether bcd mRNA localisation is independent of a polarised microtubule cytoskeleton, we examined the phenotype of shot^2A2 (Chang et al., 2011). shot^2A2 disrupts the anchoring of microtubules to the anterior/lateral cortex of the oocyte, resulting in a largely unpolarised microtubule network and the failure to localise osk mRNA to the posterior (Figure 2F) (Nashchekin et al., 2016). bcd mRNA localises normally in shot^2A2 mutant oocytes, despite the lack of directional bias in microtubule orientation (Figure 2G). For comparison, we analysed the behaviour of hu-li tai shao (hts) mRNA, which also localises anteriorly, but to a somewhat broader region than bcd mRNA and with different genetic requirements (Ding et al., 1993). Unlike bcd mRNA, the anterior enrichment of hts mRNA is largely lost in shot^2A2 (Figure 2H). Thus, hts mRNA localises by a different mechanism to bcd mRNA that depends on the weakly polarised microtubule cytoskeleton in the vicinity of the anterior.

Anterior anchoring of bcd mRNA

The proposal that there is a mechanism that retains or anchors bcd mRNA once it reaches the anterior predicts that the mRNA should be relatively stable at the anterior, whereas the continual transport model predicts a rapid turn-over of the localised RNA. To distinguish between these possibilities, we performed Fluorescent Recovery After Photobleaching (FRAP) experiments on localised bcd mRNA in stage 9 oocytes. The rate of recovery was best fit by two exponential curves, suggesting the existence of fast and slow recovering populations (Figure 3A,D, Video 3). Furthermore, only the slow population was affected by microtubule-depolymerisation with Colcemid (Figure 3B,D, Video 3), which abolishes the active transport of bcd mRNA particles (Figure 1—figure supplement 1A, Video 2). The fast population is therefore likely to correspond to highly-diffusive, nonspecific signal, most likely from autofluorescent background and/or free MCP-GFP. Consistent with this, FRAP on the cytoplasm of nurse cells, which have very low levels of bcd mRNA, or at the anterior of oocytes expressing MCP-GFP alone, yielded very fast recoveries that fit single exponential curves (Table 3, Table 3—source data 1, Figure 3C–D). We therefore used the recovery in the nurse cells to fit the FRAP data to a bi-exponential and then removed the nonspecific, fast component (see Material and Methods). The remaining specific signal recovered to 33% over an hour in untreated oocytes, a value that is reduced to 10% by Colcemid treatment (Table 3, Table 3—source data 1, Figure 3E). Thus, most bcd mRNA is stably anchored at the anterior cortex at stage 9, and the limited recovery is predominantly due to microtubule-dependent delivery of mRNA.

Figure 3 with 1 supplement see all

Download asset Open asset

Figure 3—source data 1

Photo-conversion data for bcd*Dendra2 and grk*Dendra2 (only timepoints 10 min and 60 min after photo-conversion).

Includes the data in: Figure 3—figure supplement 1, panel D.

https://doi.org/10.7554/eLife.17537.015

Download elife-17537-fig3-data1-v4.xlsx

Table 3

Sample	Mobile fraction @ 20 min	Fluorescence Half-time (min)	Oocyte (n)
MCP-GFP - St9	0.76	3.7	6
bcd*GFP / Nurse cell - St9	0.71	2.0	8

Sample	Mobile fraction @ 55 min	Fluorescence Half-time (min)	Oocytes (n)	F-test P-value a)
bcd*GFP - St9 - Mock	0.33	52.2	9
bcd*GFP - St9 - Colcemid	0.10	75.6	9	<0.0001

1. a) F-test for pairwise comparison of fluorescence recovery curves - comparisons to bcd*GFP/+ (wild-type)

Table 3—source data 1

FRAP data for MCP-GFP and bcd*GFP in stage 9 oocytes. Includes the data in: Table 3; Figure 3, panels D–E.

https://doi.org/10.7554/eLife.17537.018

Download elife-17537-table3-data1-v4.xlsx

Video 3

Download asset

The FRAP experiments cannot distinguish whether the recovery is due to the de novo delivery of newly-synthesised bcd mRNA or to the recycling of previously localised RNA from outside the bleached region, as predicted by the continual transport model. To distinguish between these possibilities, we generated a transgene expressing MCP fused to the photo-convertible protein Dendra2 (MCP-Dendra2) so that we could label only the RNA that is already localised (Gurskaya et al., 2006; Chudakov et al., 2007). 63% of photo-converted bcd mRNA (bcd*Dendra2) remained localised in the same small region over a 55 min period, in good agreement with the FRAP data (Table 4, Table 4—source data 1, Figure 3F,J, Video 4). Moreover, there was very little spreading of the photo-converted RNA along the anterior margin, arguing against continual re-localisation of the RNA (Figure 3F).

Table 4

Sample	Mobile fraction @ 20 min	Fluorescence Half-time (min)	Oocytes (n)
bcd*Dendra2 / Nurse cell - St9	0.93	3.2	10

mRNA	Mobile fraction @ 55 min	Fluorescence Half-time (min)	Oocytes (n)	F-test P value
bcd*Dendra2 - St9	0.37	26.0	10
osk*Dendra2 - St9	0.20	173.6	5	<0.0001	a)
hts*Dendra2 - St9	0.59	31.7	10	<0.0001	a)
bcd*Dendra2 - St10	0.64	16.1	8	<0.0001	a)
bcd*Dendra2 - St13	0.39	24.5	7	0.81	a)
bcd*Dendra2 - St14	0.24	138.5	10	<0.0001	a)
bcd*Dendra2 / grk2E12/2B6 - St9	0.40	33.2	11	0.91	a)
bcd*Dendra2 / 2x bcdMS2 - St9	0.44	51.4	5	0.4	a)

mRNA	Mobile fraction @ 55 min	Fluorescence Half-time (min)	Oocytes (n)	F-test P value
bcd*Dendra2 - St9 - Mock	0.35	27.9	5	0.86	a)
bcd*Dendra2 - St9 - Colcemid	0.22	152.1	5	<0.0001	a),b)

1. a),b) F-test for pairwise comparison of fluorescence recovery curves. Comparisons to (a) bcd*Dendra2 - St9 or (b) bcd*Dendra2 - St9 - Mock

Table 4—source data 1

Photo-conversion data for all samples (MCP-Dendra2; bcd*Dendra2, osk*Dendra2 and hts*Dendra2 in wild-type stage 9 oocytes; bcd*Dendra2 in stages 10b, 13 and 14 wild-type oocytes; bcd*Dendra2 in wild-type stage9 oocytes expressing 2 copies of the bcdMS2 transgene; bcd*Dendra2 in grk mutant stage9 oocytes; bcd*Dendra2 in wild-type stage 9 oocytes treated with Colcemid or mock Control).

Includes the data in: Table 4; Figure 3, panels J-K; Figure 3—figure supplement 1, panels B, F; Figure 4, panel C.

https://doi.org/10.7554/eLife.17537.021

Download elife-17537-table4-data1-v4.xlsx

Video 4

Download asset

We also analysed the behaviour of MS2-tagged hts mRNA labelled with MCP-Dendra2 (hts*Dendra2). Unlike bcd mRNA, photo-converted hts*Dendra2 showed marked spreading along the anterior cortex and was significantly more labile, with less than half (41%) of the signal remaining by the end of the time course (Figure 3H,I, Video 5). Both the spreading and the lower retention of localised hts mRNA are consistent with the idea that its anterior enrichment depends on continual transport, unlike bcd RNA at stage 9 of oogenesis.

Video 5

Download asset

As bcd RNA has been proposed to be localised by continual active transport beginning at stage 10b, we examined if the retention of localised bcd mRNA varies during oogenesis, by performing similar photo-conversion experiments on stage 10b, stage 13 and stage 14 oocytes. The RNA is significantly more mobile at stage 10b, with only 36% remaining in the ROI after 55 min (Figure 3K), consistent with the results of Weil et al. (2006). Stage 13 showed similar levels of retention to stage 9, but the RNA was significantly more stable at stage 14, coincident with the formation of larger aggregates (Figure 3J–K) (Weil et al., 2008). bcd mRNA therefore appears to be localised by a distinct mechanism at stage 10b, when fast cytoplasmic streaming starts and the RNA relocalises from an anterior ring into a central disc (Theurkauf et al., 1992; Schnorrer et al., 2000, 2002).

Since our results suggest that bcd mRNA is anchored in some way at the anterior of the oocyte at stage 9, we compared its behaviour to that of osk and gurken (grk) mRNAs, which are both anchored to the cytoskeleton after their localisation (Delanoue et al., 2007; Vanzo and Ephrussi, 2002). Although osk RNA is more stably anchored than bcd mRNA, with 80% of the signal remaining at the end of the experiment (Figure 3G,I), grk and bcd mRNAs showed almost identical retention rates (Figure 3—figure supplement 1C–D, Figure 3—source data 1). The greater retention of osk mRNA compared to bcd and grk mRNAs could reflect distinct anchoring mechanisms, but might also be due to different conditions at the posterior of the oocyte relative to anterior, where bcd and grk RNAs localise. To directly test of the effects of position within the oocyte, we examined bcd mRNA stability in strong grk mutants, in which bcd mRNA localises to both anterior and posterior poles (González-Reyes et al., 1995; Roth et al., 1995). Photo-converted bcd mRNA at the posterior of grk mutants yielded decay kinetics indistinguishable from those of RNA at the anterior of wild-type oocytes (Figure 3—figure supplement 1E–F, Table 4, Table 4—source data 1). Thus, the stability of localised bcd mRNA is intrinsic and not a consequence of the local geometry of the oocyte.

We next investigated mechanisms that might retain bcd mRNA at the anterior cortex. grk mRNA is anchored at the dorsal anterior corner of the oocyte by the binding of static Dynein to minus ends of microtubules, a mechanism that also anchors pair rule transcripts apically in the blastoderm embryo (Delanoue and Davis, 2005; Delanoue et al., 2007). We therefore tested whether the anterior retention of bcd mRNA is microtubule-dependent by culturing the egg chambers in the presence of Colcemid. Although the microtubules were completely depolymerised after ten minutes, as monitored with the microtubule binding protein, Tau-GFP, bcd mRNA labelled with MCP-Tomato (bcd*Tom) remained tightly localised after 1 hr (Figure 4A, Video 6). Furthermore, performing the same experiment with photo-converted RNA revealed that the immobile fraction increased from 65% to 78% in the absence of microtubules (Table 4, Table 4—source data 1, Figure 4B–C). Thus, the anchoring of bcd mRNA is microtubule-independent, and microtubule-dependent processes enhance its depletion from the anterior cortex. It is notable that bcd mRNA is as stably localised as osk mRNA in the absence of microtubules, with only about 20% loss over the period of 55 min. hts mRNA was more sensitive than bcd mRNA to the depletion of microtubules at stage 9, with only a small amount of RNA remaining localised (Figure 4D).

Figure 4 with 1 supplement see all

Download asset Open asset

Figure 4—source data 1

Van Steensel co-localisation analyses of bcd*GFP and mCherry-Patr.

Includes the data in: Figure 4, panel G.

https://doi.org/10.7554/eLife.17537.025

Download elife-17537-fig4-data1-v4.xlsx

Figure 4—source data 2

Van Steensel co-localisation analyses of bcd*Tom and Tau-GFP.

Includes the data in: Figure 4—figure supplement 1, panel B.

https://doi.org/10.7554/eLife.17537.026

Download elife-17537-fig4-data2-v4.xlsx

Video 6

Download asset

By contrast to stage 9, treating stage 10b egg chambers with Colcemid led to the loss of most bcd mRNA from the anterior, except for a small amount around the oocyte nucleus (Figure 4E). This is consistent with the lower anterior retention of photo-converted bcd mRNA at stage 10b. This reduction in stability coincides with the assembly of new MTOC in the middle of the anterior cortex and the relocalisation of the RNA from an anterior ring to a central disc (Schnorrer et al., 2000, 2002). The microtubule-dependence of bcd mRNA retention is transient, however, and microtubule depolymerisation had little effect on the stability of the localisation at stage 11 and stage 13 (Figure 4E).

bcd RNA is not anchored to microtubules

Although Colcemid depolymerises all detectable microtubules in the oocyte, short 'stumps' of microtubules may persist where the minus ends are attached to the cortex, which could provide anchors for bcd mRNA. We therefore examined the distribution of bcd mRNA particles relative to the microtubule minus ends, using the microtubule minus end-binding protein Patronin as a marker (Goodwin and Vale, 2010; Hendershott and Vale, 2014). Although both Patronin-labelled microtubule minus ends and bcd mRNA particles are most concentrated at the anterior corners of the oocyte, they very rarely overlap (Figure 4F). Analysing their distributions using the van Steensel method gives a Pearson’s correlation coefficient of 0.1 at zero displacement, indicating that only a very small proportion of bcd mRNA particles co-localise with microtubule minus ends (Figure 4G, Figure 4—source data 1) (van Steensel et al., 1996). We obtained similar results when using the microtubule-associated protein, Tau, as marker for stable microtubule ends in the presence of Colcemid (Figure 4—figure supplement 1A–B, Figure 4—source data 2) (Parton et al., 2011). The small degree of co-localisation makes it unlikely that the RNA is anchored to minus ends by static Dynein.

Since microtubules do not appear to anchor bcd mRNA, we turned to cortical actin, which has been implicated in keeping the RNA localised at later stages (Weil et al., 2006, 2010). To explore an earlier role of actin on bcd mRNA localisation, we cultured stage 9 oocytes in the presence of the actin depolymerising drug, Cytochalasin D. The treatment interfered with the cytoplasmic actin mesh, as it induced premature cytoplasmic streaming (data not shown) (Dahlgaard et al., 2007), but did not significantly affect the cortical actin or the distribution of bcd mRNA (Figure 5A). Since the drug treatment experiment was not conclusive, we applied two-colour Stimulated Emission Depletion (STED) super-resolution microscopy to investigate whether bcd mRNA is anchored on actin, but detected virtually no co-localisation between them (Figure 5B–C). These data suggest that cortical actin is unlikely to act as a direct anchor for bcd mRNA during stage 9.

Figure 5

Download asset Open asset

In many cases, RNAs are retained in a specific location by their incorporation into large particles, such as the polar granules at the posterior of the Drosophila oocyte or the P granules in C. elegans (Little et al., 2015; Trcek et al., 2015; Elbaum-Garfinkle et al., 2015). bcd mRNA has been shown to be associated with P-bodies, which may act to prevent its translation during oogenesis (Weil et al., 2012). This raises the possibility that sequestration in P-bodies also plays a role in anchoring bcd mRNA at the anterior. We observed that bcd mRNA particles partially co-localise with P-bodies, particularly at the very anterior of the oocyte. First, van Steensel co-localisation analysis of bcd mRNA (bcd*GFP) and the P-body component, Trailerhitch (Tral-mRFP), gave Pearson’s correlation coefficients of 0.3 and 0.2 at the very anterior and adjacent cytoplasm of the oocyte, respectively (Figure 6A–B, Figure 6—source data 1). Second, two-colour STED super-resolution microscopy revealed that bcd mRNA particles are enriched within P-bodies, particularly in the larger aggregates, but are also found free in the cytoplasm (Figure 6C). P-bodies are ubiquitous in the cytoplasm, however, and are most abundant in the posterior pole plasm (Figure 6D). Thus, incorporation of bcd RNA into P-bodies could play a role in retaining the RNA at the anterior, but there must be some additional mechanism to ensure that the RNA is only sequestered once it has reached its destination.

Figure 6

Download asset Open asset

Figure 6—source data 1

Van Steensel co-localisation analyses of bcd*GFP and Tral-mRFP.

Includes the data in: Figure 6, panel B.

https://doi.org/10.7554/eLife.17537.031

Download elife-17537-fig6-data1-v4.xlsx

To test the role of P-bodies more directly, we examined bcd mRNA localisation in germline clones of a null mutation in the core P-body component, Ge-1, which has been reported to disrupt P-body structure (Fan et al., 2011). The localisation of bcd mRNA appeared normal in Ge-1^∆5 clones, however, suggesting that P-body integrity is not required for bcd mRNA anchoring (Figure 6E).

Exu is required for the assembly of bcd mRNPs

The highly structured 3’UTR of bcd mRNA contains dimerisation/oligomerisation domains (stem-loop III) that are required for its efficient transport and apical localisation in the syncytial blastoderm embryo (Ferrandon et al., 1997; Wagner et al., 2001; Snee et al., 2005). This raises the possibility that the RNA is retained at the anterior by aggregating into larger and less diffusive particles. We therefore used STED microscopy to visualise MS2/MCP-labelled RNA particles with high resolution. Imaging of stage 9 oocytes revealed that bcd, grk and hts mRNAs form regular particles (Figures 5B-C,6C,7A–B, data not shown). We then used a fluorescence intensity curve fitting method to estimate the sizes of the particles (see Material and Methods, Figure 7—figure supplement 1A). bcd and grk mRNAs particles averaged 112 nm and 106 nm, respectively, whereas hts mRNA particles were significantly smaller, averaging 77 nm (Table 5, Table 5—source data 1, Figure 7A-B,G).

Figure 7 with 1 supplement see all

Download asset Open asset

Table 5

	Genotype	Particles (n)	Particle size ±S.E.M. (nm)	Mixed-effects test P-value	a)	Mixed-effects test P-value	b)	Summed fluorescence ±S.E.M. (a.u.)	Mixed-effects test P-value	c)	Mixed-effects test P-value	d)
MS2-labelling	bcd*GFP - St9	732	111.9 ± 1.1	- - -		0.14		- - -	- - -		- - -
	grk*GFP - St8-9	376	106.5 ± 1.6	0.41		- - -		- - -	- - -		- - -
	hts*GFP - St9	284	76.6 ± 1.0	0.002	**	0.56		- - -	- - -		- - -
	bcd*GFP / exu1 - St9	224	97.9 ± 1.8	0.046	*	- - -		- - -	- - -		- - -
	bcd*GFP - St14 (isolated)	165	109.6 ± 2.5	0.56		- - -		- - -	- - -		- - -
	bcd*GFP – St9 / 1x bcdMS2	293	100.2 ± 1.5	0.07		0.78		155.1 ± 5.3	- - -		0.11
	bcd*GFP – St9 / 2x bcdMS2	292	105.3 ± 1.6	0.19		- - -		149.3 ± 5.1	0.50		- - -
smFISH	bcd - St9 endogenous (yw)	901	125.9 ± 1.0	- - -		<0.0001	***	204.4 ± 5.4	- - -		0.13
	bcd – St14 (isolated) endogenous (yw)	749	116.3 ± 1.2	0.044	*	- - -		687.7 ± 30.0	<0.0001	***	- - -
	bcd – St14 (clustered) endogenous (yw)	125	124.6 ± 2.6	0.84		- - -		- - -	- - -		- - -
	bcd – St9 endogenous + 1x bcdMS2	935	128.6 ± 1.0	0.28		- - -		264.8 ± 7.4	<0.0001	***	- - -
	bcd – St9 endogenous + 2x bcdMS2	1509	124.6 ± 0.8	0.74		- - -		328.5 ± 7.4	<0.0001	***	- - -

1. a) Mixed effects linear model (LMER) test for comparison of RNA particle sizes (FWHM). Fixed effect: mRNA / Genotype; Random effect: variability between oocytes. Compared to bcd*GFP - St9 or bcd - St9 endogenous (yw)

2. b) Mixed effects linear model (LMER) test to analyse the effect of the distance from the anterior on the RNA particle sizes (FWHM). Fixed effect: Distance from anterior; Random effect: variability between oocytes.

3. c) Mixed effects linear model (LMER) test for comparison of the summed fluorescence of RNA particles. Fixed effect: mRNA / Genotype; Random effect: variability between oocytes. Compared to bcd*GFP - St9 / 1x bcdMS2, or bcd - St9 endogenous (yw)

4. d) Mixed effects linear model (LMER) test for comparison of the summed fluorescence of RNA particles. Fixed effect: Distance from anterior; Random effect: variability between oocytes

5. - - - Not applicable / Not determined

6. a.u. arbitrary units

7. *p<0.05; **p<0.01; ***p<0.001

Table 5—source data 1

Properties of RNA particles from STED super-resolution imaging.

Includes the data in: Table 5; Figure 7, panels G–K; Figure 8, panles A–D.

https://doi.org/10.7554/eLife.17537.035

Download elife-17537-table5-data1-v4.xlsx

bcd mRNA forms larger aggregates at stage 14 of oogenesis, but super-resolution imaging revealed that these are still composed of small, discrete RNA particles (Figure 7C). As the particles in clusters are about 200 nm apart in the XY dimension (224 nm and 202 nm mean distance to nearest neighbour in stage 9 and 14 oocytes, respectively), each ~700 nm optical Z-section is likely to include more than one particle. This causes a high and irregular background, leading to overestimation of fluorescence intensities and unreliable curve fittings. Nevertheless, the estimated size of isolated bcd mRNA particles at stage 14 was comparable to those in stage 9 oocytes (Table 5, Table 5—source data 1, Figure 7G). This suggests that bcd mRNA particles remain relatively homogeneous in size throughout oogenesis, despite of their clustering into large, semi-ordered aggregates at stage 14.

To confirm these findings, we also performed STED imaging on endogenous bcd RNA labelled by single molecule FISH (smFISH) with probes spanning the 3’UTR. This technique also revealed that bcd RNA forms particles that remain approximately the same size throughout oogenesis, although they appear slightly larger than those seen with MS2-GFP labelling, presumably because smFISH labels the entire bcd 3’UTR, not just the MS2 sites in the RNA (Table 5, Table 5—source data 1, Figure 7E–F). To explore if particle remodelling plays a role in anchoring bcd mRNA at the anterior, we compared the properties of particles at different distances from the anterior margin of stage 9 oocytes. The average size of bcd mRNA particles measured by both MS2-labelling and smFISH did not change substantially with the distance from the anterior (Table 5, Table 5—source data 1, Figure 7J–K), arguing against their coalescence upon localisation.

The uniform size of bcd mRNA particles, regardless of their location or the stage of oogenesis, was unexpected and suggests that the RNA is incorporated into a well-defined structure rather than assembling into aggregates of variable size depending on the RNA concentration. We tested this hypothesis by comparing the sizes of the particles formed by the endogenous RNA in wild-type oocytes with those formed in oocytes expressing either one or two additional copies of bcdMS2, which raises the RNA levels to 1.75x and 3.25x the endogenous level, respectively (Figure 7I). The size of the bcd RNA particles remained constant with increasing RNA concentration (Table 5, Table 5—source data 1, Figure 7G–H), but we observed significantly more particles (172%) in oocytes expressing 2 copies of bcdMS2 compared to just the endogenous RNA alone. Consistent with this, extra bcdMS2 RNA did not affect the decay kinetics in photo-conversion experiments, indicating that the diffusion characteristics of the particles were also constant under differing RNA concentrations (Figure 7—figure supplement 1B). These experiments demonstrate that the size of bcd mRNA particles is insensitive to the concentration of mRNA, supporting the view that the RNA is assembled into a distinct structure of uniform size. Indeed, the only condition that altered the size of the bcd mRNA particles was the exu¹ mutant, which strongly reduces the affinity of Exu for RNA (Lazzaretti et al., 2016). In this case, the few particles that were detected were slightly, but significantly, smaller (98 nm versus 112 nm) (Table 5, Table 5—source data 1, Figure 7D–E). Thus, Exu may provide part of the scaffolding for the assembly of bcd mRNA particles.

To determine whether the particles also contain the same amount of RNA throughout oogenesis and at different bcd gene dosages, we used the summed fluorescent intensities of the particles as a measure of their RNA content. The average fluorescent intensity of the particles detected by both MS2-labelling and smFISH did not change with distance from the anterior, reinforcing the conclusion that the particles do not fuse upon localisation and anchoring at the anterior (Figure 8A–B). By contrast, the mean fluorescence intensity of the particles increased when bcd mRNA levels were raised by expressing one or two copies of the bcdMS2 transgene, with 30% and 60% more fluorescence, and thus more RNA, per particle respectively (Table 5, Table 5—source data 1, Figure 8C). Furthermore, stage 14 particles contained more than 3 times as much RNA as those at stage 9 (Table 5, Table 5—source data 1, Figure 8D). It has previously been shown that most bcd RNA enters the oocyte during nurse cell dumping at stage 10b-12 (Weil et al., 2006), suggesting that much of this extra RNA is incorporated into pre-existing RNA particles. Thus, the bcd RNA particles have a variable RNA content, despite their constant size.

Figure 8

Download asset Open asset

It has been generally assumed that bcd mRNA is localised by directed transport along a polarised microtubule cytoskeleton (Wolpert et al., 2015; St Johnston, 2005). Indeed, studies on the later stages of oogenesis indicate that the RNA is continually transported towards microtubule minus ends by Dynein and then is increasingly more anchored as oogenesis progresses (Pokrywka and Stephenson 1991; Weil et al., 2006, 2008). Here we provide evidence that the RNA is localised by a different mechanism at other stages.

Live imaging of fluorescently-labelled bcd mRNA in stage 9 oocytes revealed that the RNA forms particles that undergo frequent active movements along microtubules. The speed was strongly reduced in Dhc mutants, consistent with bcd mRNA being transported predominantly by Dynein. Furthermore, bcd mRNA moved significantly more frequently and faster in a null mutant in Khc. Since Kinesin-I transports Dynein to the oocyte posterior, the more frequent movements in the Khc null mutant may be explained by the higher levels of Dynein at the anterior of the oocyte, which should increase the probability of its binding to bcd RNPs. This cannot account for the increased velocity of bcd mRNA particle movements, however, as this also occurs in the “slow” Khc alleles, which still localise Dynein posteriorly. Thus, Kinesin-I somehow slows down Dynein through a mechanism that depends on its full motor activity. One possibility is that Kinesin-I engages in a tug of war with Dynein and therefore exerts a drag that slows down Dynein movement. Interestingly, inhibition of Dynein increases both the velocity of Kinesin-I-driven ooplasmic streaming and osk mRNA particle transport, indicating that this antagonistic relationship between Dynein and Kinesin-I is reciprocal (Serbus et al., 2005; Zimyanin et al., 2008).

osk mRNA is transported to the posterior of the oocyte by Kinesin-I-dependent movements along a weakly polarised microtubule network, in which approximately 15% more of the microtubules have their plus-ends pointing posteriorly than anteriorly (Zimyanin et al., 2008; Parton et al., 2011). We found that bcd mRNA particles show a reciprocal anterior bias near the middle of the oocyte, with 12% more movements towards the anterior cortex than away from it. This supports the idea that bcd mRNA is mainly transported towards microtubule minus ends by Dynein. The bias becomes increasingly weak close to the anterior, however, and even reverses in the region 0-5 μm from the anterior cortex. Since the small bias in the anterior region is unaffected in the Khc null mutant and there are virtually no directional reversals, it is not due to bidirectional transport along a polarised microtubule network. Instead, there seem to be approximately equal numbers of microtubules pointing anteriorly and posteriorly near the anterior, with most microtubules running parallel to the anterior cortex. This fits well with experimental measurements of microtubule polarity in the oocyte, in which the orientation bias decreases from posterior to anterior, and with 3-dimensional computer simulations of the microtubule organisation, which give negligible orientation bias close to the anterior (Parton et al., 2011; Trong et al., 2015). Thus, directed transport cannot account for the localisation of bcd mRNA to the very anterior of the oocyte, although it can deliver the mRNA to a broader anterior region.

In light of these observations, we propose that bcd mRNA is localised by rapid, bidirectional Dynein-dependent transport in and out of the anterior region, coupled to some mechanism that specifically retains or anchors the RNA at the anterior cortex (Figure 9A). This random transport and anterior anchoring model predicts that the RNA will only turn over slowly at the anterior cortex. This is confirmed by FRAP and photo-conversion experiments, which show that more than 60% of the RNA remains stably localised at the anterior over a period of 55 min. This turn-over rate is the same as that measured for grk mRNA, which has previously been shown to be specifically anchored at its localisation site above the oocyte nucleus (Delanoue et al., 2007; Jaramillo et al., 2008). Further support for non-directional transport and anterior anchoring comes from the observation that bcd mRNA is more efficiently retained at the anterior when the microtubules are depolymerised, indicating that microtubule-based transport plays a role in removing the RNA from the anterior, as well as delivering it.

Figure 9

Download asset Open asset

Unlike bcd mRNA, the behaviour of hts mRNA at stage 9 fits well with the predictions of the continual active transport model. It is localised to a broader anterior region than bcd mRNA, turns-over significantly more rapidly than bcd RNA in photo-conversion experiments and spreads along the anterior margin after photoconversion. Furthermore, hts mRNA localisation is strongly reduced after 90 min of Colcemid treatment and in the shot^2A2 mutant, in which the oocyte microtubule cytoskeleton is not polarised, whereas bcd mRNA localisation is largely unaffected in both conditions. Thus, the contrast between the two RNAs reinforces the view that bcd mRNA cannot be explained solely by directed transport and must involve an anterior anchoring step.

This model may help to explain the observations of Cha et al. (2001), who showed that bcd mRNA injected into the oocyte localises to the nearest region of anterior/lateral cortex, whereas RNA that is exposed to nurse cell cytoplasm before injection localises specifically to the anterior. Although the authors proposed that the “nurse cell-conditioned” RNA gains the capacity to discriminate between microtubules emanating from the anterior and lateral cortex, a simpler explanation is that both untreated and conditioned RNA are transported by Dynein along microtubules, but only the latter becomes competent to be retained at the anterior. The untreated RNA therefore concentrates near microtubule minus ends, much like hts RNA (which is more biased towards the anterior than injected RNA because it enters from the nurse cells), whereas the conditioned RNA localises specifically to the anterior. Thus, factors such as Exu loaded on the RNA in the nurse cells may licence the RNA for anterior anchoring (Figure 9A).

The retention of bcd mRNA at the anterior varies over the course of oogenesis, with the RNA being much less stably localised at stage 10b. This fits well with the observations of Weil et al., 2006, who measured very similar fluorescence recovery rates at stage 10b to those reported here. This decrease in anterior retention coincides with a redistribution of the RNA from an anterior ring to a disc at the centre of the anterior cortex, and with the formation of a new MTOC in this region (St Johnston et al., 1989; Schnorrer et al., 2002; Vogt et al., 2006). Thus, the anterior anchoring mechanism seems to be specifically inactivated at this stage to allow the remodelling of the RNA distribution (Figure 9B). During this period, bcd mRNA localisation is consistent with continual active transport along the polarised microtubule network formed by the new anterior MTOC. This is only transient, however, as the RNA becomes more stable at the anterior at stage 13, and is very efficiently anchored at stage 14, which is important to keep bcd mRNA localised until fertilisation, so that it can act as the source of the Bcd morphogen gradient in the embryo (Figure 9C).

The mechanism that retains bcd RNA at the anterior is unclear. We can rule out anchoring by Dynein to microtubule minus ends, as has been reported for grk mRNA in the oocyte and pair-rule transcripts in the embryo (Delanoue and Davis, 2005; Delanoue et al., 2007), since the anterior retention of bcd mRNA is not microtubule-dependent and the RNA does not co-localise with microtubules. The mRNA could be tethered to cortical actin, which would be consistent with the anchoring defect in late oocytes seen in swallow mutants, which disrupt the actin cortex (Weil et al., 2010). However, bcd RNA does not show a significant co-localisation with F-actin, although this tethering could be indirect. Another possibility is that the RNA is maintained at the anterior by sequestering it in P-bodies (Weil et al., 2012) (Figure 9). P-bodies are ubiquitous throughout the oocyte, and there would therefore have to be some mechanism that induces bcd RNA incorporation into these structures specifically at the anterior.

Super-resolution imaging revealed that bcd mRNA forms 110–120 nm particles throughout oogenesis, regardless of whether the RNA is localised or not. Even the large aggregates of RNA at stage 14 are still formed of individual particles of similar size, although their protein composition is different from stage 9, as Staufen and ESCRT-II are recruited to bcd mRNA only at stage 10b (Martin et al., 2003; Weil et al., 2006) (Figure 9). Importantly, over-expression of the mRNA does not alter particle size, but instead leads to more particles, which have higher average RNA content. The same occurs at stage 14 of oogenesis, when the bcd mRNA content of the oocyte is much higher following nurse cell dumping. Thus, bcd mRNA seems to assemble into a structure of defined size, almost like a virus particle, which can incorporate more or less mRNA molecules depending on availability. An exu mutant that affects RNA binding affinity causes a large reduction in the number of detectable bcd mRNA particles and a small, but significant reduction in the size of the few particles that form. This suggests that Exu, which forms homodimers that probably bind two bcd mRNA molecules (Lazzaretti et al., 2016), plays a role in scaffolding the assembly of the particles (Figure 9). Loss of Exu also strongly reduces both the speed and frequency of bcd mRNP movement, as well as its anterior anchoring at all stages of oogenesis. Particle formation may therefore be a prerequisite for all of these processes, explaining the pleiotropic effects of exu mutants.

The invariant size of bcd RNA particles make them fundamentally different from other well-characterised RNA granules, such as the P-granules in C. elegans, which form by the aggregation of RNA and proteins into droplets that phase-separate from the surrounding cytoplasm (Brangwynne et al., 2009; Saha et al., 2016). P-granules have variable size that depends on the RNA concentration and readily fuse with each other when juxtaposed. By contrast, bcd RNA particles stay the same size as the RNA concentration increases, even though they incorporate more RNA, and they do not appear to fuse when tightly clustered in aggregates. The exact nature of the particles will require the identification of more of their components, but their behaviour is compatible with a model in which they consist of a rigid protein framework that contains multiple RNA-binding sites. In future, it will be interesting to determine whether other localised RNAs are packaged into similar structures.

1. 1. Baumbach J
   2. Novak ZA
   3. Raff JW
   4. Wainman A

   (2015) Dissecting the function and assembly of acentriolar microtubule organizing centers in Drosophila cells in vivo

   PLoS Genetics 11:e1005261.

   https://doi.org/10.1371/journal.pgen.1005261

   • PubMed
   • Google Scholar
2. 1. Berleth T
   2. Burri M
   3. Thoma G
   4. Bopp D
   5. Richstein S
   6. Frigerio G
   7. Noll M
   8. Nüsslein-Volhard C

   (1988)

   The role of localization of bicoid RNA in organizing the anterior pattern of the Drosophila embryo

   The EMBO Journal 7:1749–1756.

   • PubMed
   • Google Scholar
3. 1. Bertrand E
   2. Chartrand P
   3. Schaefer M
   4. Shenoy SM
   5. Singer RH
   6. Long RM

   (1998) Localization of ASH1 mRNA particles in living yeast

   Molecular Cell 2:437–445.

   https://doi.org/10.1016/S1097-2765(00)80143-4

   • PubMed
   • Google Scholar
4. 1. Bianco A
   2. Poukkula M
   3. Cliffe A
   4. Mathieu J
   5. Luque CM
   6. Fulga TA
   7. Rørth P

   (2007) Two distinct modes of guidance signalling during collective migration of border cells

   Nature 448:362–365.

   https://doi.org/10.1038/nature05965

   • PubMed
   • Google Scholar
5. 1. Bottanelli F
   2. Kromann EB
   3. Allgeyer ES
   4. Erdmann RS
   5. Wood Baguley S
   6. Sirinakis G
   7. Schepartz A
   8. Baddeley D
   9. Toomre DK
   10. Rothman JE
   11. Bewersdorf J

   (2016) Two-colour live-cell nanoscale imaging of intracellular targets

   Nature Communications 7:10778.

   https://doi.org/10.1038/ncomms10778

   • PubMed
   • Google Scholar
6. 1. Brangwynne CP
   2. Eckmann CR
   3. Courson DS
   4. Rybarska A
   5. Hoege C
   6. Gharakhani J
   7. Jülicher F
   8. Hyman AA

   (2009) Germline P granules are liquid droplets that localize by controlled dissolution/condensation

   Science 324:1729–1732.

   https://doi.org/10.1126/science.1172046

   • PubMed
   • Google Scholar
7. 1. Brendza KM
   2. Rose DJ
   3. Gilbert SP
   4. Saxton WM

   (1999) Lethal kinesin mutations reveal amino acids important for ATPase activation and structural coupling

   Journal of Biological Chemistry 274:31506–31514.

   https://doi.org/10.1074/jbc.274.44.31506

   • PubMed
   • Google Scholar
8. 1. Brendza RP
   2. Serbus LR
   3. Duffy JB
   4. Saxton WM

   (2000) A function for kinesin I in the posterior transport of oskar mRNA and Staufen protein

   Science 289:2120–2122.

   https://doi.org/10.1126/science.289.5487.2120

   • PubMed
   • Google Scholar
9. 1. Bulgakova NA
   2. Grigoriev I
   3. Yap AS
   4. Akhmanova A
   5. Brown NH

   (2013) Dynamic microtubules produce an asymmetric E-cadherin-Bazooka complex to maintain segment boundaries

   The Journal of Cell Biology 201:887–901.

   https://doi.org/10.1083/jcb.201211159

   • PubMed
   • Google Scholar
10. 1. Bulinski JC
    2. Odde DJ
    3. Howell BJ
    4. Salmon TD
    5. Waterman-Storer CM

    (2001)

    Rapid dynamics of the microtubule binding of ensconsin in vivo

    Journal of Cell Science 114:3885–3897.

    • PubMed
    • Google Scholar
11. 1. Bullock SL
    2. Nicol A
    3. Gross SP
    4. Zicha D

    (2006) Guidance of bidirectional motor complexes by mRNA cargoes through control of dynein number and activity

    Current Biology 16:1447–1452.

    https://doi.org/10.1016/j.cub.2006.05.055

    • PubMed
    • Google Scholar
12. 1. Buszczak M
    2. Paterno S
    3. Lighthouse D
    4. Bachman J
    5. Planck J
    6. Owen S
    7. Skora AD
    8. Nystul TG
    9. Ohlstein B
    10. Allen A
    11. Wilhelm JE
    12. Murphy TD
    13. Levis RW
    14. Matunis E
    15. Srivali N
    16. Hoskins RA
    17. Spradling AC

    (2007) The carnegie protein trap library: a versatile tool for Drosophila developmental studies

    Genetics 175:1505–1531.

    https://doi.org/10.1534/genetics.106.065961

    • PubMed
    • Google Scholar
13. 1. Cha BJ
    2. Koppetsch BS
    3. Theurkauf WE

    (2001) In vivo analysis of Drosophila bicoid mRNA localization reveals a novel microtubule-dependent axis specification pathway

    Cell 106:35–46.

    https://doi.org/10.1016/S0092-8674(01)00419-6

    • PubMed
    • Google Scholar
14. 1. Chang CW
    2. Nashchekin D
    3. Wheatley L
    4. Irion U
    5. Dahlgaard K
    6. Montague TG
    7. Hall J
    8. St Johnston D

    (2011) Anterior-posterior axis specification in Drosophila oocytes: identification of novel bicoid and oskar mRNA localization factors

    Genetics 188:883–896.

    https://doi.org/10.1534/genetics.111.129312

    • PubMed
    • Google Scholar
15. 1. Chou TB
    2. Perrimon N

    (1996)

    The autosomal FLP-DFS technique for generating germline mosaics in Drosophila melanogaster

    Genetics 144:1673–1679.

    • PubMed
    • Google Scholar
16. 1. Chudakov DM
    2. Lukyanov S
    3. Lukyanov KA

    (2007) Using photoactivatable fluorescent protein Dendra2 to track protein movement

    BioTechniques 42:553-563.

    https://doi.org/10.2144/000112470

    • PubMed
    • Google Scholar
17. 1. Clark A
    2. Meignin C
    3. Davis I

    (2007) A Dynein-dependent shortcut rapidly delivers axis determination transcripts into the Drosophila oocyte

    Development 134:1955–1965.

    https://doi.org/10.1242/dev.02832

    • PubMed
    • Google Scholar
18. 1. Clark I
    2. Giniger E
    3. Ruohola-Baker H
    4. Jan LY
    5. Jan YN

    (1994) Transient posterior localization of a kinesin fusion protein reflects anteroposterior polarity of the Drosophila oocyte

    Current Biology 4:289–300.

    https://doi.org/10.1016/S0960-9822(00)00068-3

    • PubMed
    • Google Scholar
19. 1. Clark IE
    2. Jan LY
    3. Jan YN

    (1997)

    Reciprocal localization of Nod and kinesin fusion proteins indicates microtubule polarity in the Drosophila oocyte, epithelium, neuron and muscle

    Development 124:461–470.

    • PubMed
    • Google Scholar
20. 1. Costes SV
    2. Daelemans D
    3. Cho EH
    4. Dobbin Z
    5. Pavlakis G
    6. Lockett S

    (2004) Automatic and quantitative measurement of protein-protein colocalization in live cells

    Biophysical Journal 86:3993–4003.

    https://doi.org/10.1529/biophysj.103.038422

    • PubMed
    • Google Scholar
21. 1. Dahlgaard K
    2. Raposo AA
    3. Niccoli T
    4. St Johnston D

    (2007) Capu and Spire assemble a cytoplasmic actin mesh that maintains microtubule organization in the Drosophila oocyte

    Developmental Cell 13:539–553.

    https://doi.org/10.1016/j.devcel.2007.09.003

    • PubMed
    • Google Scholar
22. 1. Delanoue R
    2. Davis I

    (2005) Dynein anchors its mRNA cargo after apical transport in the Drosophila blastoderm embryo

    Cell 122:97–106.

    https://doi.org/10.1016/j.cell.2005.04.033

    • PubMed
    • Google Scholar
23. 1. Delanoue R
    2. Herpers B
    3. Soetaert J
    4. Davis I
    5. Rabouille C

    (2007) Drosophila Squid/hnRNP helps Dynein switch from a gurken mRNA transport motor to an ultrastructural static anchor in sponge bodies

    Developmental Cell 13:523–538.

    https://doi.org/10.1016/j.devcel.2007.08.022

    • PubMed
    • Google Scholar
24. 1. Ding D
    2. Parkhurst SM
    3. Lipshitz HD

    (1993) Different genetic requirements for anterior RNA localization revealed by the distribution of Adducin-like transcripts during Drosophila oogenesis

    PNAS 90:2512–2516.

    https://doi.org/10.1073/pnas.90.6.2512

    • PubMed
    • Google Scholar
25. 1. Driever W
    2. Siegel V
    3. Nüsslein-Volhard C

    (1990)

    Autonomous determination of anterior structures in the early Drosophila embryo by the bicoid morphogen

    Development 109:811–820.

    • PubMed
    • Google Scholar
26. 1. Duncan JE
    2. Warrior R

    (2002) The cytoplasmic dynein and kinesin motors have interdependent roles in patterning the Drosophila oocyte

    Current Biology 12:1982–1991.

    https://doi.org/10.1016/S0960-9822(02)01303-9

    • PubMed
    • Google Scholar
27. 1. Elbaum-Garfinkle S
    2. Kim Y
    3. Szczepaniak K
    4. Chen CC-H
    5. Eckmann CR
    6. Myong S
    7. Brangwynne CP

    (2015) The disordered P granule protein LAF-1 drives phase separation into droplets with tunable viscosity and dynamics

    PNAS 112:7189–7194.

    https://doi.org/10.1073/pnas.1504822112

    • PubMed
    • Google Scholar
28. 1. Emmons S
    2. Phan H
    3. Calley J
    4. Chen W
    5. James B
    6. Manseau L

    (1995) Cappuccino, a Drosophila maternal effect gene required for polarity of the egg and embryo, is related to the vertebrate limb deformity locus

    Genes & Development 9:2482–2494.

    https://doi.org/10.1101/gad.9.20.2482

    • PubMed
    • Google Scholar
29. 1. Ephrussi A
    2. Lehmann R

    (1992) Induction of germ cell formation by oskar

    Nature 358:387–392.

    https://doi.org/10.1038/358387a0

    • PubMed
    • Google Scholar
30. 1. Ephrussi A
    2. St Johnston D

    (2004) Seeing is believing: the bicoid morphogen gradient matures

    Cell 116:143–152.

    https://doi.org/10.1016/S0092-8674(04)00037-6

    • PubMed
    • Google Scholar
31. 1. Fan S-J
    2. Marchand V
    3. Ephrussi A

    (2011) Drosophila Ge-1 promotes P body formation and oskar mRNA localization

    PLoS One 6:e20612.

    https://doi.org/10.1371/journal.pone.0020612

    • Google Scholar
32. 1. Ferrandon D
    2. Elphick L
    3. Nüsslein-Volhard C
    4. St Johnston D

    (1994) Staufen protein associates with the 3'UTR of bicoid mRNA to form particles that move in a microtubule-dependent manner

    Cell 79:1221–1232.

    https://doi.org/10.1016/0092-8674(94)90013-2

    • PubMed
    • Google Scholar
33. 1. Ferrandon D
    2. Koch I
    3. Westhof E
    4. Nüsslein-Volhard C

    (1997) RNA-RNA interaction is required for the formation of specific bicoid mRNA 3' UTR-STAUFEN ribonucleoprotein particles

    The EMBO Journal 16:1751–1758.

    https://doi.org/10.1093/emboj/16.7.1751

    • PubMed
    • Google Scholar
34. 1. Forrest KM
    2. Gavis ER

    (2003) Live imaging of endogenous RNA reveals a diffusion and entrapment mechanism for nanos mRNA localization in Drosophila

    Current Biology 13:1159–1168.

    https://doi.org/10.1016/S0960-9822(03)00451-2

    • PubMed
    • Google Scholar
35. 1. Gepner J
    2. Li M
    3. Ludmann S
    4. Kortas C
    5. Boylan K
    6. Iyadurai SJ
    7. McGrail M
    8. Hays TS

    (1996)

    Cytoplasmic dynein function is essential in Drosophila melanogaster

    Genetics 142:865–878.

    • PubMed
    • Google Scholar
36. 1. González-Reyes A
    2. Elliott H
    3. St Johnston D

    (1995) Polarization of both major body axes in Drosophila by gurken-torpedo signalling

    Nature 375:654–658.

    https://doi.org/10.1038/375654a0

    • PubMed
    • Google Scholar
37. 1. Goodwin SS
    2. Vale RD

    (2010) Patronin regulates the microtubule network by protecting microtubule minus ends

    Cell 143:263–274.

    https://doi.org/10.1016/j.cell.2010.09.022

    • PubMed
    • Google Scholar
38. 1. Gurskaya NG
    2. Verkhusha VV
    3. Shcheglov AS
    4. Staroverov DB
    5. Chepurnykh TV
    6. Fradkov AF
    7. Lukyanov S
    8. Lukyanov KA

    (2006) Engineering of a monomeric green-to-red photoactivatable fluorescent protein induced by blue light

    Nature Biotechnology 24:461–465.

    https://doi.org/10.1038/nbt1191

    • PubMed
    • Google Scholar
39. 1. Hazelrigg T
    2. Watkins WS
    3. Marcey D
    4. Tu C
    5. Karow M
    6. Lin XR

    (1990)

    The exuperantia gene is required for Drosophila spermatogenesis as well as anteroposterior polarity of the developing oocyte, and encodes overlapping sex-specific transcripts

    Genetics 126:607–617.

    • PubMed
    • Google Scholar
40. 1. Hendershott MC
    2. Vale RD

    (2014) Regulation of microtubule minus-end dynamics by CAMSAPs and Patronin

    PNAS 111:5860–5865.

    https://doi.org/10.1073/pnas.1404133111

    • PubMed
    • Google Scholar
41. 1. Irion U
    2. St Johnston D

    (2007) bicoid RNA localization requires specific binding of an endosomal sorting complex

    Nature 445:554–558.

    https://doi.org/10.1038/nature05503

    • PubMed
    • Google Scholar
42. 1. Jambor H
    2. Surendranath V
    3. Kalinka AT
    4. Mejstrik P
    5. Saalfeld S
    6. Tomancak P

    (2015) Systematic imaging reveals features and changing localization of mRNAs in Drosophila development

    eLife 4:05003.

    https://doi.org/10.7554/eLife.05003

    • PubMed
    • Google Scholar
43. 1. Januschke J
    2. Gervais L
    3. Dass S
    4. Kaltschmidt JA
    5. Lopez-Schier H
    6. St Johnston D
    7. Brand AH
    8. Roth S
    9. Guichet A

    (2002) Polar transport in the Drosophila oocyte requires Dynein and Kinesin I cooperation

    Current Biology 12:1971–1981.

    https://doi.org/10.1016/S0960-9822(02)01302-7

    • PubMed
    • Google Scholar
44. 1. Jaramillo AM
    2. Weil TT
    3. Goodhouse J
    4. Gavis ER
    5. Schupbach T

    (2008) The dynamics of fluorescently labeled endogenous gurken mRNA in Drosophila

    Journal of Cell Science 121:887–894.

    https://doi.org/10.1242/jcs.019091

    • PubMed
    • Google Scholar
45. 1. Khuc Trong P
    2. Doerflinger H
    3. Dunkel J
    4. St Johnston D
    5. Goldstein RE

    (2015) Cortical microtubule nucleation can organise the cytoskeleton of Drosophila oocytes to define the anteroposterior axis

    eLife 4:e06088.

    https://doi.org/10.7554/eLife.06088

    • PubMed
    • Google Scholar
46. 1. Lazzaretti D
    2. Veith K
    3. Kramer K
    4. Basquin C
    5. Urlaub H
    6. Irion U
    7. Bono F

    (2016) The bicoid mRNA localization factor Exuperantia is an RNA-binding pseudonuclease

    Nature Structural & Molecular Biology 23:705–713.

    https://doi.org/10.1038/nsmb.3254

    • PubMed
    • Google Scholar
47. 1. Little SC
    2. Sinsimer KS
    3. Lee JJ
    4. Wieschaus EF
    5. Gavis ER

    (2015) Independent and coordinate trafficking of single Drosophila germ plasm mRNAs

    Nature Cell Biology 17:558–568.

    https://doi.org/10.1038/ncb3143

    • PubMed
    • Google Scholar
48. 1. Lowary PT
    2. Uhlenbeck OC

    (1987) An RNA mutation that increases the affinity of an RNA-protein interaction

    Nucleic Acids Research 15:10483–10493.

    https://doi.org/10.1093/nar/15.24.10483

    • PubMed
    • Google Scholar
49. 1. Lowe N
    2. Rees JS
    3. Roote J
    4. Ryder E
    5. Armean IM
    6. Johnson G
    7. Drummond E
    8. Spriggs H
    9. Drummond J
    10. Magbanua JP
    11. Naylor H
    12. Sanson B
    13. Bastock R
    14. Huelsmann S
    15. Trovisco V
    16. Landgraf M
    17. Knowles-Barley S
    18. Armstrong JD
    19. White-Cooper H
    20. Hansen C
    21. Phillips RG
    22. Lilley KS
    23. Russell S
    24. St Johnston D
    25. UK Drosophila Protein Trap Screening Consortium

    (2014) Analysis of the expression patterns, subcellular localisations and interaction partners of Drosophila proteins using a pigP protein trap library

    Development 141:3994–4005.

    https://doi.org/10.1242/dev.111054

    • PubMed
    • Google Scholar
50. 1. Lécuyer E
    2. Yoshida H
    3. Parthasarathy N
    4. Alm C
    5. Babak T
    6. Cerovina T
    7. Hughes TR
    8. Tomancak P
    9. Krause HM

    (2007) Global analysis of mRNA localization reveals a prominent role in organizing cellular architecture and function

    Cell 131:174–187.

    https://doi.org/10.1016/j.cell.2007.08.003

    • PubMed
    • Google Scholar
51. 1. Macdonald PM
    2. Struhl G

    (1988) cis-acting sequences responsible for anterior localization of bicoid mRNA in Drosophila embryos

    Nature 336:595–598.

    https://doi.org/10.1038/336595a0

    • PubMed
    • Google Scholar
52. 1. Markussen FH
    2. Michon AM
    3. Breitwieser W
    4. Ephrussi A

    (1995)

    Translational control of oskar generates short OSK, the isoform that induces pole plasma assembly

    Development 121:3723–3732.

    • PubMed
    • Google Scholar
53. 1. Martin SG
    2. Leclerc V
    3. Smith-Litière K
    4. St Johnston D

    (2003) The identification of novel genes required for Drosophila anteroposterior axis formation in a germline clone screen using GFP-Staufen

    Development 130:4201–4215.

    https://doi.org/10.1242/dev.00630

    • PubMed
    • Google Scholar
54. 1. McGrail M
    2. Hays TS

    (1997)

    The microtubule motor cytoplasmic dynein is required for spindle orientation during germline cell divisions and oocyte differentiation in Drosophila

    Development 124:2409–2419.

    • PubMed
    • Google Scholar
55. 1. McKim KS
    2. Joyce EF
    3. Jang JK

    (2009) Cytological analysis of meiosis in fixed Drosophila ovaries

    Methods in Molecular Biology 558:197–216.

    https://doi.org/10.1007/978-1-60761-103-5_12

    • PubMed
    • Google Scholar
56. 1. Meijering E
    2. Dzyubachyk O
    3. Smal I

    (2012) Methods for cell and particle tracking

    Methods in Enzymology 504:183–200.

    https://doi.org/10.1016/B978-0-12-391857-4.00009-4

    • PubMed
    • Google Scholar
57. 1. Micklem DR
    2. Dasgupta R
    3. Elliott H
    4. Gergely F
    5. Davidson C
    6. Brand A
    7. González-Reyes A
    8. St Johnston D

    (1997) The mago nashi gene is required for the polarisation of the oocyte and the formation of perpendicular axes in Drosophila

    Current Biology 7:468–478.

    https://doi.org/10.1016/S0960-9822(06)00218-1

    • PubMed
    • Google Scholar
58. 1. Mische S
    2. Li M
    3. Serr M
    4. Hays TS

    (2007) Direct observation of regulated ribonucleoprotein transport across the nurse cell/oocyte boundary

    Molecular Biology of the Cell 18:2254–2263.

    https://doi.org/10.1091/mbc.E06-10-0959

    • PubMed
    • Google Scholar
59. 1. Nakamura A
    2. Amikura R
    3. Hanyu K
    4. Kobayashi S

    (2001)

    Me31B silences translation of oocyte-localizing RNAs through the formation of cytoplasmic RNP complex during Drosophila oogenesis

    Development 128:3233–3242.

    • PubMed
    • Google Scholar
60. 1. Namba R
    2. Pazdera TM
    3. Cerrone RL
    4. Minden JS

    (1997)

    Drosophila embryonic pattern repair: how embryos respond to bicoid dosage alteration

    Development 124:1393–1403.

    • PubMed
    • Google Scholar
61. 1. Nashchekin D
    2. Fernandes AR
    3. St Johnston D

    (2016) Patronin/Shot cortical foci assemble the noncentrosomal microtubule array that specifies the Drosophila anterior-posterior axis

    Developmental Cell 38:61–72.

    https://doi.org/10.1016/j.devcel.2016.06.010

    • PubMed
    • Google Scholar
62. 1. Neuman-Silberberg FS
    2. Schüpbach T

    (1993) The Drosophila dorsoventral patterning gene gurken produces a dorsally localized RNA and encodes a TGF alpha-like protein

    Cell 75:165–174.

    https://doi.org/10.1016/S0092-8674(05)80093-5

    • PubMed
    • Google Scholar
63. 1. Palacios IM
    2. St Johnston D

    (2002) Kinesin light chain-independent function of the Kinesin heavy chain in cytoplasmic streaming and posterior localisation in the Drosophila oocyte

    Development 129:5473–5485.

    https://doi.org/10.1242/dev.00119

    • PubMed
    • Google Scholar
64. 1. Parton RM
    2. Hamilton RS
    3. Ball G
    4. Yang L
    5. Cullen CF
    6. Lu W
    7. Ohkura H
    8. Davis I

    (2011) A PAR-1-dependent orientation gradient of dynamic microtubules directs posterior cargo transport in the Drosophila oocyte

    The Journal of Cell Biology 194:121–135.

    https://doi.org/10.1083/jcb.201103160

    • PubMed
    • Google Scholar
65. 1. Pokrywka NJ
    2. Stephenson EC

    (1991)

    Microtubules mediate the localization of bicoid RNA during Drosophila oogenesis

    Development 113:55–66.

    • PubMed
    • Google Scholar
66. 1. Pokrywka NJ
    2. Stephenson EC

    (1995) Microtubules are a general component of mRNA localization systems in Drosophila oocytes

    Developmental Biology 167:363–370.

    https://doi.org/10.1006/dbio.1995.1030

    • PubMed
    • Google Scholar
67. Book

    1. R Core Team

    (2013)

    R: A Language and Environment for Statistical Computing

    Vienna, Austria: R Foundation for Statistical Computing.

    • Google Scholar
68. 1. Rongo C
    2. Gavis ER
    3. Lehmann R

    (1995)

    Localization of oskar RNA regulates oskar translation and requires Oskar protein

    Development 121:2737–2746.

    • PubMed
    • Google Scholar
69. 1. Roth S
    2. Neuman-Silberberg FS
    3. Barcelo G
    4. Schüpbach T

    (1995) cornichon and the EGF receptor signaling process are necessary for both anterior-posterior and dorsal-ventral pattern formation in Drosophila

    Cell 81:967–978.

    https://doi.org/10.1016/0092-8674(95)90016-0

    • PubMed
    • Google Scholar
70. 1. Saha S
    2. Weber CA
    3. Nousch M
    4. Adame-Arana O
    5. Hoege C
    6. Hein MY
    7. Osborne-Nishimura E
    8. Mahamid J
    9. Jahnel M
    10. Jawerth L
    11. Pozniakovski A
    12. Eckmann CR
    13. Jülicher F
    14. Hyman AA

    (2016) Polar positioning of phase-separated liquid compartments in cells regulated by an mRNA competition mechanism

    Cell 166:1572–1584.

    https://doi.org/10.1016/j.cell.2016.08.006

    • PubMed
    • Google Scholar
71. 1. Schindelin J
    2. Arganda-Carreras I
    3. Frise E
    4. Kaynig V
    5. Longair M
    6. Pietzsch T
    7. Preibisch S
    8. Rueden C
    9. Saalfeld S
    10. Schmid B
    11. Tinevez JY
    12. White DJ
    13. Hartenstein V
    14. Eliceiri K
    15. Tomancak P
    16. Cardona A

    (2012) Fiji: an open-source platform for biological-image analysis

    Nature Methods 9:676–682.

    https://doi.org/10.1038/nmeth.2019

    • PubMed
    • Google Scholar
72. 1. Schmittgen TD
    2. Livak KJ

    (2008) Analyzing real-time PCR data by the comparative C(T) method

    Nature Protocols 3:1101–1108.

    https://doi.org/10.1038/nprot.2008.73

    • PubMed
    • Google Scholar
73. 1. Schnorrer F
    2. Bohmann K
    3. Nüsslein-Volhard C

    (2000) The molecular motor dynein is involved in targeting swallow and bicoid RNA to the anterior pole of Drosophila oocytes

    Nature Cell Biology 2:185–190.

    https://doi.org/10.1038/35008601

    • PubMed
    • Google Scholar
74. 1. Schnorrer F
    2. Luschnig S
    3. Koch I
    4. Nüsslein-Volhard C

    (2002) Gamma-tubulin37C and Gamma-tubulin ring complex protein 75 are essential for bicoid RNA localization during Drosophila oogenesis

    Developmental Cell 3:685–696.

    https://doi.org/10.1016/S1534-5807(02)00301-5

    • PubMed
    • Google Scholar
75. 1. Schupbach T
    2. Wieschaus E

    (1986) Germline autonomy of maternal-effect mutations altering the embryonic body pattern of Drosophila

    Developmental Biology 113:443–448.

    https://doi.org/10.1016/0012-1606(86)90179-X

    • PubMed
    • Google Scholar
76. 1. Serbus LR
    2. Cha BJ
    3. Theurkauf WE
    4. Saxton WM

    (2005) Dynein and the actin cytoskeleton control kinesin-driven cytoplasmic streaming in Drosophila oocytes

    Development 132:3743–3752.

    https://doi.org/10.1242/dev.01956

    • PubMed
    • Google Scholar
77. 1. Shaner NC
    2. Campbell RE
    3. Steinbach PA
    4. Giepmans BN
    5. Palmer AE
    6. Tsien RY

    (2004) Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein

    Nature Biotechnology 22:1567–1572.

    https://doi.org/10.1038/nbt1037

    • PubMed
    • Google Scholar
78. 1. Snee MJ
    2. Arn EA
    3. Bullock SL
    4. Macdonald PM

    (2005) Recognition of the bcd mRNA localization signal in Drosophila embryos and ovaries

    Molecular and Cellular Biology 25:1501–1510.

    https://doi.org/10.1128/MCB.25.4.1501-1510.2005

    • PubMed
    • Google Scholar
79. 1. Sprague BL
    2. Pego RL
    3. Stavreva DA
    4. McNally JG

    (2004) Analysis of binding reactions by fluorescence recovery after photobleaching

    Biophysical Journal 86:3473–3495.

    https://doi.org/10.1529/biophysj.103.026765

    • PubMed
    • Google Scholar
80. 1. St Johnston D
    2. Driever W
    3. Berleth T
    4. Richstein S
    5. Nüsslein-Volhard C

    (1989)

    Multiple steps in the localization of bicoid RNA to the anterior pole of the Drosophila oocyte

    Development 107:13–19.

    • PubMed
    • Google Scholar
81. 1. St Johnston D

    (2005) Moving messages: the intracellular localization of mRNAs

    Nature Reviews. Molecular Cell Biology 6:363–375.

    https://doi.org/10.1038/nrm1643

    • PubMed
    • Google Scholar
82. 1. Theurkauf WE
    2. Smiley S
    3. Wong ML
    4. Alberts BM

    (1992)

    Reorganization of the cytoskeleton during Drosophila oogenesis: implications for axis specification and intercellular transport

    Development 115:923–936.

    • PubMed
    • Google Scholar
83. 1. Trcek T
    2. Grosch M
    3. York A
    4. Shroff H
    5. Lionnet T
    6. Lehmann R

    (2015) Drosophila germ granules are structured and contain homotypic mRNA clusters

    Nature Communications 6:7962.

    https://doi.org/10.1038/ncomms8962

    • PubMed
    • Google Scholar
84. 1. Van Doren M
    2. Williamson AL
    3. Lehmann R

    (1998) Regulation of zygotic gene expression in Drosophila primordial germ cells

    Current Biology 8:243–246.

    https://doi.org/10.1016/S0960-9822(98)70091-0

    • PubMed
    • Google Scholar
85. 1. van Steensel B
    2. van Binnendijk EP
    3. Hornsby CD
    4. van der Voort HT
    5. Krozowski ZS
    6. de Kloet ER
    7. van Driel R

    (1996)

    Partial colocalization of glucocorticoid and mineralocorticoid receptors in discrete compartments in nuclei of rat hippocampus neurons

    Journal of Cell Science 109:787–792.

    • PubMed
    • Google Scholar
86. 1. Vanzo NF
    2. Ephrussi A

    (2002)

    Oskar anchoring restricts pole plasm formation to the posterior of the Drosophila oocyte

    Development 129:3705–3714.

    • PubMed
    • Google Scholar
87. 1. Vicente NB
    2. Zamboni JED
    3. Adur JF
    4. Paravani EV
    5. Casco VH

    (2007) Photobleaching correction in fluorescence microscopy images

    Journal of Physics: Conference Series 90:012068.

    https://doi.org/10.1088/1742-6596/90/1/012068

    • Google Scholar
88. 1. Vogt N
    2. Koch I
    3. Schwarz H
    4. Schnorrer F
    5. Nüsslein-Volhard C

    (2006) The gammaTuRC components Grip75 and Grip128 have an essential microtubule-anchoring function in the Drosophila germline

    Development 133:3963–3972.

    https://doi.org/10.1242/dev.02570

    • PubMed
    • Google Scholar
89. 1. Wagner C
    2. Palacios I
    3. Jaeger L
    4. St Johnston D
    5. Ehresmann B
    6. Ehresmann C
    7. Brunel C

    (2001) Dimerization of the 3'UTR of bicoid mRNA involves a two-step mechanism

    Journal of Molecular Biology 313:511–524.

    https://doi.org/10.1006/jmbi.2001.5057

    • PubMed
    • Google Scholar
90. 1. Wang C
    2. Lehmann R

    (1991) Nanos is the localized posterior determinant in Drosophila

    Cell 66:637–647.

    https://doi.org/10.1016/0092-8674(91)90110-K

    • PubMed
    • Google Scholar
91. 1. Weil TT
    2. Forrest KM
    3. Gavis ER

    (2006) Localization of bicoid mRNA in late oocytes is maintained by continual active transport

    Developmental Cell 11:251–262.

    https://doi.org/10.1016/j.devcel.2006.06.006

    • PubMed
    • Google Scholar
92. 1. Weil TT
    2. Parton R
    3. Davis I
    4. Gavis ER

    (2008) Changes in bicoid mRNA anchoring highlight conserved mechanisms during the oocyte-to-embryo transition

    Current Biology 18:1055–1061.

    https://doi.org/10.1016/j.cub.2008.06.046

    • PubMed
    • Google Scholar
93. 1. Weil TT
    2. Parton RM
    3. Herpers B
    4. Soetaert J
    5. Veenendaal T
    6. Xanthakis D
    7. Dobbie IM
    8. Halstead JM
    9. Hayashi R
    10. Rabouille C
    11. Davis I

    (2012) Drosophila patterning is established by differential association of mRNAs with P bodies

    Nature Cell Biology 14:1305–1313.

    https://doi.org/10.1038/ncb2627

    • PubMed
    • Google Scholar
94. 1. Weil TT
    2. Xanthakis D
    3. Parton R
    4. Dobbie I
    5. Rabouille C
    6. Gavis ER
    7. Davis I

    (2010) Distinguishing direct from indirect roles for bicoid mRNA localization factors

    Development 137:169–176.

    https://doi.org/10.1242/dev.044867

    • PubMed
    • Google Scholar
95. 1. Whittaker KL
    2. Ding D
    3. Fisher WW
    4. Lipshitz HD

    (1999)

    Different 3' untranslated regions target alternatively processed hu-li tai shao (hts) transcripts to distinct cytoplasmic locations during Drosophila oogenesis

    Journal of Cell Science 112:3385–3398.

    • PubMed
    • Google Scholar
96. Book

    1. Wolpert L
    2. Tickle C
    3. Arias AM

    (2015)

    Principles of Development

    Oxford University Press, USA.

    • Google Scholar
97. 1. Zimyanin VL
    2. Belaya K
    3. Pecreaux J
    4. Gilchrist MJ
    5. Clark A
    6. Davis I
    7. St Johnston D

    (2008) In vivo imaging of oskar mRNA transport reveals the mechanism of posterior localization

    Cell 134:843–853.

    https://doi.org/10.1016/j.cell.2008.06.053

    • PubMed
    • Google Scholar

Author details

1. Vítor Trovisco

   1. The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom
   2. Department of Genetics, University of Cambridge, Cambridge, United Kingdom

   Contribution

   VT, Conception and design, Performed most of the experiments and analysed the data, drafted and revised the manuscript with DStJ

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0002-8728-0281

2. Katsiaryna Belaya

   1. The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom
   2. Department of Genetics, University of Cambridge, Cambridge, United Kingdom

   Present address

   The Weatherall Institute of Molecular Medicine, University of Oxford, Oxford, United Kingdom

   Contribution

   KB, Started the project, performed and analysed the initial bcd mRNA particle tracking experiments

   Competing interests

   The authors declare that no competing interests exist.

3. Dmitry Nashchekin

   1. The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom
   2. Department of Genetics, University of Cambridge, Cambridge, United Kingdom

   Contribution

   DN, Performed the analysis of RNA localization and microtubule cytoskeleton organisation in the shot^2A2 mutant, provided the Patronin-YFP and mCherry-Patronin transgenics prior to publication, drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

4. Uwe Irion

   1. The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom
   2. Department of Genetics, University of Cambridge, Cambridge, United Kingdom

   Present address

   Max Planck Institute for Developmental Biology, Tübingen, Germany

   Contribution

   UI, Created the bcdMS2 transgene, Commented on the manuscript

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0003-2823-5840

5. George Sirinakis

   1. The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom
   2. Department of Genetics, University of Cambridge, Cambridge, United Kingdom

   Contribution

   GS, Built the STED microscope and assisted with the collection and analysis of the super-resolution data

   Competing interests

   The authors declare that no competing interests exist.

6. Richard Butler

   The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom

   Contribution

   RB, Assisted with the analysis of the bcd mRNA particle tracking data, assisted with the analysis of the co-localization data using his unpublished Fiji plugin EsCOP, wrote the Fiji plugin Profiler for the analysis of RNA particle properties from the super-resolution images

   Competing interests

   The authors declare that no competing interests exist.

7. Jack J Lee

   Department of Molecular Biology, Princeton University, Princeton, United States

   Contribution

   JJL, Created the MCP-Dendra2 transgenic line

   Competing interests

   The authors declare that no competing interests exist.

8. Elizabeth R Gavis

   Department of Molecular Biology, Princeton University, Princeton, United States

   Contribution

   ERG, Designed the MCP-Dendra2 construct, made extensive comments on the manuscript and revisions

   Competing interests

   The authors declare that no competing interests exist.

9. Daniel St Johnston

   1. The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom
   2. Department of Genetics, University of Cambridge, Cambridge, United Kingdom

   Contribution

   DSJ, Conceived and designed the project with VT and KB, analysed and interpreted the data with VT, drafted and revised the manuscript with VT

   For correspondence

   d.stjohnston@gurdon.cam.ac.uk

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0001-5582-3301

• 4,096

  views

• 769

  downloads

• 39

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.