Resource
Global Analysis of mRNA Localization
Reveals a Prominent Role in Organizing
Cellular Architecture and Function
Eric Le´ cuyer,1,3 Hideki Yoshida,1,3 Neela Parthasarathy,1,2,3 Christina Alm,1,2,3 Tomas Babak,1,2,3
Tanja Cerovina,1,3 Timothy R. Hughes,1,2,3 Pavel Tomancak,4 and Henry M. Krause1,2,3,*
1Banting and Best Department of Medical Research
2Department of Medical Genetics and Microbiology
3Terrence Donnelly Centre for Cellular and Biomolecular Research
University of Toronto,Toronto, Canada
4Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany
*Correspondence: h.krause@utoronto.ca
DOI 10.1016/j.cell.2007.08.003
SUMMARY
Although subcellular mRNA trafficking has
been demonstrated as a mechanism to control
protein distribution, it is generally believed
that most protein localization occurs subsequent
to translation. To address this point, we
developed and employed a high-resolution
fluorescent in situ hybridization procedure to
comprehensively evaluate mRNA localization
dynamics during early Drosophila embryogenesis.
Surprisingly, of the 3370 genes analyzed,
71% of those expressed encode subcellularly
localized mRNAs. Dozens of new and striking
localization patterns were observed, implying
an equivalent variety of localization mechanisms.
Tight correlations between mRNA distribution
and subsequent protein localization and
function, indicate major roles for mRNA localization
in nucleating localized cellular machineries.
A searchable web resource documenting
mRNA expression and localization dynamics
has been established and will serve as an invaluable
tool for dissecting localization mechanisms
and for predicting gene functions and
interactions.
INTRODUCTION
Virtually all cells are polarized, partitioning their contents
to a variety of organelles, compartments and membrane
interfaces that execute specialized biological and regulatory
functions. Since the discovery of the signal peptide by
Blobel and colleagues (Blobel and Dobberstein, 1975), the
targeting of most proteins to these various subcellular
destinations has been thought to occur after translation.
More recently, it has been shown that protein localization
can also be controlled by localizing the mRNA transcript
prior to translation (Bashirullah et al., 1998; Czaplinski
and Singer, 2006; Kloc et al., 2002; St Johnston, 2005).
A potential advantage of this mechanism is its cost effectiveness.
Each localized mRNA can facilitate many rounds
of protein synthesis, thereby avoiding the significant energy
costs of moving each protein molecule individually
(Jansen, 2001). This process also helps to ensure that
proteins do not appear where their effects would be
detrimental.
Localized mRNAs can serve many biological functions,
including the establishment of morphogen gradients
(Driever and Nusslein-Volhard, 1988; Ephrussi et al.,
1991; Gavis and Lehmann, 1992), the segregation of
cell-fate determinants (Broadus et al., 1998; Gore et al.,
2005; Hughes et al., 2004; Li et al., 1997; Long et al.,
1997; Melton, 1987; Neuman-Silberberg and Schupbach,
1993; Simmonds et al., 2001; Takizawa et al., 1997; Zhang
et al., 1998), and the targeting of protein synthesis to specialized
organelles or cellular domains (Adereth et al.,
2005; Lambert and Nagy, 2002; Lawrence and Singer,
1986; Mingle et al., 2005; Zhang et al., 2001).
While the list of known localized mRNAs has grown
steadily over the past two decades (Bashirullah et al.,
1998; Czaplinski and Singer, 2006; Kloc et al., 2002;
St Johnston, 2005), the prevalence, variety and overall
importance of mRNA localization events is unknown. Previous
in situ screening efforts in Drosophila have established
speculative estimates of the proportion of localized
mRNAs, ranging from one to ten percent (Dubowy
and Macdonald, 1998; Tomancak et al., 2002). However,
the detection methods used in past studies were of
insufficient resolution to observe intricate subcellular
patterns.
To assess mRNA subcellular localization dynamics on
a global scale, a high-resolution fluorescent in situ hybridization
(FISH) procedure was developed and applied to
early developmental stages of Drosophila embryogenesis.
174 Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc.
Of the genes expressed during this developmental
window, a surprising 71% were found to encode mRNAs
exhibiting clear subcellular distribution patterns. The frequency
and variety of localization events suggests that virtually
all aspects of cellular function are impacted by RNA
trafficking pathways. We conclude that mRNA localization
is a major mechanism for controlling cellular architecture
and function.
RESULTS
Method Development, Screening,
and Localization Database
To circumvent the deficiencies of existing in situ hybridization
protocols, considerable effort was made to develop
a procedure with optimal subcellular resolution, sensitivity,
consistency, throughput and economy. Typical results
obtained with the resulting method (Le´ cuyer et al., 2007),
versus traditional alkaline phosphatase-based probe detection,
are illustrated in Figures 1A and S1 in the Supplemental
Data available with this article online. Reassuringly,
control analyses using probes with increasing sequence
divergence indicate that the occurrence of false positive
signals due to cross-hybridization to mRNAs with similar
sequences is highly unlikely (Figure S2).
Following high-throughput FISH, samples were
mounted and analyzed using epifluorescence microscopy.
For each expressed gene, representative low and
high magnification images were captured at key developmental
stages and incorporated within a relational database.
The first 4.5 hr of Drosophila development, spanning
embryonic stages 1–9, was chosen for analysis, as this interval
is manageable in terms of data annotation and encompasses
major developmental landmarks such as the
midblastula transition (MBT), gastrulation and the specification
of many cell types. The MBT is the period during
which developmental regulation switches from control
by maternally synthesized gene products to control by zygotically
transcribed genes (Tadros et al., 2007b). Importantly,
our FISH method enables the unambiguous distinction
between maternal and zygotic mRNA populations.
Maternally provided transcripts are generally cytoplasmic,
exhibit FISH signal intensities above background in stage
1 embryos and decrease in intensity during later stages.
Zygotic mRNAs, on the other hand, are always first
detected in nascent transcript foci within subsets of
Figure 1. Embryonic Gene Expression Dynamics Revealed by High-Resolution FISH
(A and B) The optimized FISH procedure reveals localization patterns not readily discernible with traditional detection methods and enables the
unambiguous distinction of maternal and zygotic mRNA populations. Examples of patterns observed are shown for maternal Bsg25D transcripts (A),
and for zygotically expressed CG4500 and Trn-SR transcripts (B), detected using optimized FISH (mRNAs in green/nuclei in red), or standard
alkaline phosphatase-based detection ([A] left panel, image obtained from the BDGP in situ database, Tomancak et al., 2002).
(C) General summary of observed and projected gene expression and mRNA localization events.
(D) Comparison of maternal and zygotic transcripts and their respective gene ontology (GO) term enrichments.
(E) Expression and localization dynamics of maternal and zygotic transcripts during stages 1–9 of embryogenesis.
Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc. 175
embryonic nuclei and are not generally observed until
stage 4. Examples of maternal and zygotic mRNAs are
illustrated in Figures 1A and 1B.
An annotation term hierarchy has been created to document
stage-specific expression, localization and degradation
dynamics of each transcript. In addition, an RNA localization
database resource, containing annotation terms
and representative images of all transcripts detected, has
been established using previously described web-based
tools (Tomancak et al., 2002) and can be accessed
through a searchable web-browser at: http://fly-fish.ccbr.
utoronto.ca.
Embryonic Gene Expression Dynamics
Of the 3370 genes ($25% of genome) successfully
screened, 2314 were expressed during the developmental
window analyzed (Figure 1C). Of these, 2198 were maternally
provided, 504 were zygotically expressed, and 388
were expressed both maternally and zygotically (Figures
1C and 1D). The percentage of genes that we find to be
maternally expressed (65%) is more than twice that of earlier
predictions ($30%; Arbeitman et al., 2002), although it
more closely resembles estimates from a recent microarray-based
study ($55%; Tadros et al., 2007a). Our higher
values likely reflect the thorough gene-by-gene approach
that was used and the improved sensitivity of the detection
procedure. Consistent with the importance of
transcript degradation in the transition from maternal to
zygotic control of embryogenesis (Tadros et al., 2007a),
the majority ($65%) of these maternal mRNAs are no
longer detectable by stages 8–9 (Figure 1E).
Gene ontology (GO) term enrichment analysis reveals
that maternal transcripts are enriched for genes involved
in RNA metabolism (Figures 1D, S3, S4A, and Table S1),
including, for example, many components of the spliceosome
(crn, prp8, SmD3, snRNP69D, snRNP70K, U2af50).
This finding is consistent with the large and diverse maternal
mRNA contingent, and the need to organize aspects of
their processing, translation, stability and localization. Interestingly,
these terms are more strongly enriched among
the more stable maternal mRNA subsets that continue to be
detected through stages 6–9 (Figure S4A and Table S1), in
agreement with recent observations (Tadros et al., 2007a).
The first major burst of zygotic transcription occurs at
stages 4–5 as the Drosophila embryo begins to transition
from syncytial to cellular growth (Figure 1E). Approximately
half (230) of the zygotically expressed genes detected
are initiated during stages 4–5 and continue to be
expressed throughout the developmental period analyzed
(Figure S5A). Interestingly, several of the earliest transcribed
mRNAs, which were unexpectedly detected as
early as stages 1–2, are encoded by transposable/repetitive
elements (copia, Doc, Ste12DOR). The significance of
these early expression events will be addressed further
below. As a group, the zygotically expressed genes are
strongly enriched for functions relating to transcriptional
regulation, cell fate determination, tissue/organ development
and morphogenesis (Figures 1D, S4B, and Table
S1), consistent with the rapid patterning and morphogenetic
processes that follow the MBT. These functional enrichments
are in agreement with a recent study by De
Renzis et al. (2007), who identified 1158 putative zygotically
expressed genes using microarray analyses of chromosome
deletion mutants (De Renzis et al., 2007). However,
their dataset does not include 337 of the 504
genes that we unambiguously found to be zygotically
expressed. Thus, by extrapolation, we predict the number
of zygotically expressed genes during this period to be
significantly higher ($2043 versus 1158; Figure 1C). Altogether,
our projections suggest that >9000 genes are
expressed either maternally or zygotically during the early
stages of Drosophila embryogenesis (Figure 1C), representing
a vast and complex set of regulatory interactions.
Transcript Localization
Of the 2314 mRNAs expressed, a remarkable 71% are
subcellularly localized (Figure 1C), with a peak in localization
frequency observed between embryonic stages 4–7
(Figure 1E). This peak in localized transcript numbers
may reflect a relatively high demand for localization events
during the conversion from syncytial to cellular environments.
Alternatively, it may be skewed by the relative
ease of detection of localization in large embryos versus
small cells. Further analyses of transcript localization in
other tissues and developmental stages, and using higher
resolution microscopy techniques, will undoubtedly reveal
many additional localization events. Hence, our numbers
represent a conservative estimate of the total number of
localized mRNAs encoded in the fly genome.
The 1644 localized transcripts observed were grouped
into $35 localization categories, most of which are listed
in Table 1. The following sections will focus on some of
the more diverse and striking of these localization patterns
and their functional implications.
Subembryonic Localization Patterns
The most prevalent localization patterns observed are the
sub-embryonic ‘exclusionary’ categories, where mRNAs
are excluded from parts of the embryo, such as the peripheral
apical cytoplasm or the germline pole cells that
form at the posterior tip (Table 1). Although these patterns
might easily be missed in smaller somatic cells, their importance
is clear. For example, the pole cell-excluded
subgroup is specifically enriched for transcripts encoding
ribosomal constituents and factors involved in mRNA metabolism
and processing (Table S2). This is consistent with
the general need to prevent transcript synthesis and translation
in germline cells during early embryogenesis (Seydoux
and Dunn, 1997; Van Doren et al., 1998), and provides
new insights into the mechanisms responsible.
Similar mechanisms are likely to be used in somatic cells
to prevent translation or related processes in portions of
the cytoplasm.
Perhaps the most easily detected of the subembryonic
patterns are the previously documented anterior and
posterior categories. Notably, the number of anterior
176 Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc.
Table 1. Summary of mRNA Localization Patterns
Localization Patterns
Number of Genesa
(Projectedb
)
Percent of
Expressed
Number of Genes/Stage Range
St. 1–3 St. 4–5 St. 6–7 St. 8–9
Expressed 2314 (9379) 100.0 2213 2183 1592 1154
Localized 1644 (6663) 71.0 233 1578 1135 187
Subembryonic patterns 1468 (5950) 63.4 207 1431 1041 147
Exclusionary patterns 1397 (5662) 60.3 123 1371 976 8
Pole plasm/Pole cell exclusion 1272 (5156) 54.9 83 1235 920
Apical exclusion 1145 (4641) 49.5 NA 1142 747 6
Basal exclusion 256 (1038) 11.1 NA 225 112
Yolk cortex exclusion 207 (839) 8.9 58 175 6
Yolk cortex localization 277 (1123) 12.0 9 171 172 3
Posterior localization 198 (803) 8.6 74 195 106
Pole/Germ cell localization 195 (790) 8.4 NA 192 105 62
Pole plasm localization 70 (284) 3.0 70 NA NA NA
Pole buds 54 (219) 2.3 54 NA NA NA
RNA islands 44 (178) 1.9 44 NA NA NA
Anterior localization 5 (20) 0.2 5 3
Subcellular localization
patterns
366 (1483) 15.8 42 135 270 107
Basal localization 209 (847) 9.0 NA 25 190 50
Nuclei-associated localization 82 (332) 3.5 27 43 36 24
Perinuclear localization 81 (328) 3.5 25 37 35 24
Perinuclear yolk nuclei 59 (239) 2.5 6 25 35 22
Perinuclear cortical nuclei 27 (109) 1.2 8 21 2
Intranuclear accumulation 14 (57) 0.6 2 14 4 4
Apical localization 78 (316) 3.4 NA 59 54 15
Diffuse apical localization 36 (146) 1.6 NA 29 18 12
Discrete apical foci 27 (109) 1.2 NA 20 18 2
Apical clusters 24 (97) 1.0 NA 15 20 0
Apical in neuroblasts 6 (24) 0.3 NA NA NA 6
Cytoplasmic foci 45 (182) 1.9 14 38 14 15
Cell division apparatus 33 (134) 1.4 16 29 1 1
Microtubule-associated 10 (41) 0.4 9 6 1
Spindle midzone localization 10 (41) 0.4 1 10
Centrosomal localization 6 (24) 0.3 6 4 1 1
Chromatin-associated 14 (57) 0.6 2 12
Cell junction-associated 12 (49) 0.5 4 10 9 6
Membrane-associated 8 (32) 0.3 5 5 5 2
Polar body-associated 4 (16) 0.2 4 NA NA NA
NA, not applicable to these embryonic stages; St., stages.
a
All screened genes encoding localized mRNAs in either of the analyzed stages; patterns are not necessarily mutually exclusive.
b
Projected number of localized transcripts encoded in the Drosophila genome, out of a total of 13,659 coding genes.
Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc. 177
transcripts is far outnumbered by those that are posteriorly
localized (5 versus 198; Table 1). For both of these
categories, the sensitivity and resolution of detection
enabled the identification of distinct subgroups of patterns
(Figure 2). For example, among the anteriorly localized
mRNAs, which include bcd, CycB, lok, milt and asp, only
bcd exhibits tight anterior localization (Figure 2A), consistent
with its extensively characterized function as the
primary determinant of anterior cell fate specification
(Ephrussi and St Johnston, 2004). The other four mRNAs
exhibit a more diffuse gradient of anterior enrichment (Figure
2B). Interestingly, three of these encode proteins
involved in cytoskeleton organization and microtubulebased
processes (Figure 2F and Table S3).
The posterior group of mRNAs could be clearly subdivided
into three categories: (1) transcripts that localize to
early pole plasm (Figure 2C), a specialized region of cytoplasm
that directs the formation of germline pole cells; (2)
transcripts that reside in the pole plasm and then localize
further into distinctive rings around pole cell nuclei
(Figure 2D); and (3) transcripts that only begin to localize
posteriorly in early stage 4 embryos (Figure 2E). The biological
significance of these subcategories is underscored
by their specific GO term enrichments (Figure 2F
and Table S3). For example, the category 1 and 2 pole
plasm localized transcripts, which include mRNAs such
as aret, eIF5, gcl, Imp, nos, osk, orb, pAbp, pum, spir,
and Tm1, are strongly enriched for cell development,
Figure 2. Anterior/Posterior Patterns and Functional Enrichments
(A–E) Sagittal views of entire embryos (A and B) or of the posterior region (C–E) between stages 2–5, following FISH with probes to bcd (A), asp (B), osk
(C), orb (D), or grp (E) transcripts (mRNA green/nuclei red). (A and B) Varieties of anterior patterns, with bcd mRNA (A) showing tight anterior localization
and asp transcripts (B), a more diffuse anterior enrichment. (C–E) Early and late posterior localization patterns. While both osk and orb transcripts
localize to the posterior pole plasm in stage 1–2 embryos ([C and D] arrowheads), orb mRNA forms distinctive rings around pole cell nuclei at
stage 3 ([D] arrow). In contrast, grp transcripts localize in the posterior yolk plasm in early stage 4 embryos ([E] arrow). All of these transcripts localize to
the pole cells at stage 5.
(F) GO term enrichments exhibited by transcripts found within annotation categories pertaining to anterior and posterior localization in stage 1–5
embryos (column categories 1 and 2 refer to stages 1–3 and 4–5, respectively). The ‘‘hot metal’’ color scale reflects statistical significance (Àlog 10
of the p value) of the GO term enrichments.
178 Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc.
translation regulation, pole plasm assembly and RNA
localization functions (Figure 2F). In contrast, the late
posterior group, which includes genes such as aur,
CG14030, CycA, grp, gwl, Rbp-1, Su(var)3-9, and ttk,
is enriched for protein kinases and negative regulators
of gene expression, again consistent with previous findings
that germ cells are transcriptionally silent (Seydoux
and Braun, 2006; Seydoux and Dunn, 1997; Van Doren
et al., 1998). Taken together, these observations suggest
the existence of distinct early and late pathways for
posterior transcript localization.
Notably, no maternal transcripts were identified as being
either dorsally or ventrally localized. Instead, differential
distribution of transcripts along the dorso-ventral axis
was always a consequence of localized zygotic transcription.
The preponderance of transcripts localized to
the posterior pole of the embryo, in comparison to the
other embryonic poles, seemingly reflects special requirements
for germ cell specification and the sufficiency
of existing zygotic mechanisms to define the other coor-
dinates.
Subcellular Categories: Apicobasal Patterns
Besides the subembryonic localization patterns, a large
collection of mRNAs, either of maternal or zygotic origin,
were found to exhibit intricate subcellular localization
patterns. Classic examples include the subset of mRNAs
that localize to the apical cytoplasm within the embryonic
epithelium (Figure 3). Although apical mRNAs have been
characterized previously and considered as a homogeneous
group (Davis and Ish-Horowicz, 1991; Simmonds
et al., 2001; Tepass et al., 1990), many distinctive subgroups
of apical transcripts could be distinguished, ranging
from broad gradients of apical enrichment to tightly
localized clusters or foci (Figures 3A–3E). Likewise,
a large number of basally localized mRNAs were identified,
which also fall into a number of subgroups (Figures
3G–3I). Other patterns that vary along the apico-basal
axis include transcripts that are excluded from the apical
cytoplasm or from the entire blastoderm layer (Figures 3J
and 3K).
The functional relevance of these subgroup classifications
is underlined by the GO term enrichments observed.
Figure 3. Varieties of Apicobasal Localization Patterns and Their Functional Enrichments
(A–L) Sagittal views through the embryonic epithelium of embryos hybridized with the indicated probes. Several distinctive subcategories of apical
(A–E), basal (G–I), or exclusionary (J and K) patterns are shown. (F and L) CG14896 transcripts are apical in posterior epithelial cells (F) and in later
arising neuroblasts ([L] arrowheads). For all images, mRNAs are green and nuclei red.
(M) GO term enrichments observed for different subcategories of apical mRNAs. Enrichment scores are depicted using a hot metal color scale conveying
statistical significance (Àlog 10 of the p value). Column categories 2–4 refer to embryonic stages 4–5, 6–7, and 8–9, respectively.
Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc. 179
For example, the apical clusters category, which includes
mRNAs such as Ama, bib, Btk29A, crb, fra, Gp150, htl, Ptr,
scb, sog, and smo, is enriched for GO categories for
plasma membrane and signaling pathway components
(Figure 3M and Table S4). In contrast, the diffuse apical
group, which contains several pair-rule gene transcripts
(hairy, odd, prd, run), is enriched for functions associated
with transcriptional regulation and pattern/axis specifica-
tion.
In addition to the apicobasal patterns detected in the
embryonic epithelium, several mRNAs were observed
with asymmetric patterns in neuroblasts. This category includes
transcripts such as asp, Gp150, mira, odd, pros,
and wg, some of which have been observed previously
(Broadus et al., 1998; Schuldt et al., 1998), and not surprisingly,
show GO term enrichments for asymmetric cell
division functions (Figure 3M). We also identified mRNAs
from uncharacterized genes, such CG14896, which
exhibit apical localization both in the posterior embryonic
epithelium underlying the pole cells and in neuroblasts
that arise later in embryogenesis (Figures 3F and 3L).
This example suggests the likelihood that many
localization mechanisms will operate in a variety of cell
types.
Membrane-Associated Patterns
Also remarkable are transcripts that localize to membrane-associated
structures prior to and following cellularization
(Figure 4). For example, cno and anillin mRNAs
(Figures 4A and 4B) associate with the embryonic cortex
or perinuclear clouds as early as stage 3, and then evolve
into polygonal mosaic networks shortly thereafter (Figures
4A and 4C). These patterns resemble subsequent actin
filament distributions and dynamics and precede cell
junction formation. Several other mRNAs encoding cell
junction components, such as Patj (Figure 4D) and dlg1
(Figure 4E), localize at specific sites along the basolateral
membrane. In contrast, mira transcripts localize throughout
the lateral membrane of embryonic epithelial cells
(Figure 4F). Accordingly, this category is enriched for
GO terms related to cytoskeleton organization and biogenesis
(Table S5). These observations imply a significant
role for mRNA localization in the nucleation and positioning
of cytoskeletal networks and membrane-associated
structures.
Cell Division and Nuclei-Associated Patterns
Many of the most striking subcellular patterns observed
occur during nuclear or cellular division (Figure 5). These
include transcripts that localize to spindle poles, centrosomes/centrioles,
astral microtubules, or along the mitotic
spindles themselves during anaphase and telophase (Figures
5A–5H). Furthermore, several mRNAs that are zygotically
transcribed in early stage 4 embryos concentrate
around metaphase chromosomes during mitosis and often
become associated with spindle midbodies (Figure
5D). Intriguingly, many of these mRNAs encode transcriptional
regulators (dpld, nullo, odd, rib, run, stwl, Taf4).
The genes in these categories show GO term enrichments
for cell division related processes (Figure S3 and
Figure 4. Membrane-AssociatedPatterns
(A–F) Surface plane (upper panels) and sagittal
(lower panels) views of stage 3 (A and B), 4 (C),
5 (E and F), and 6 (D) embryos hybridized with
probes for the transcripts indicated in lower
panels (mRNA green/nuclei red). cno transcripts
localize within cortical polygonal networks
([A] arrowhead), while anillin mRNA is
first perinuclear ([B] arrowhead) and then
evolves into a cell junction type pattern (C).
(D–F) Patj, dlg1, and mira transcripts localize
at different positions along the lateral membrane
(arrowheads).
180 Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc.
Table S5), implying important roles for localized mRNAs in
the establishment, function and regulation of cell division
machineries.
Interestingly, several of the earliest zygotically expressed
mRNAs, such as those encoded by the Doc
and copia transposons, exhibit intricate chromatin-associated
patterns. Indeed, Doc-element RNA localizes in
the vicinity of centromeres, either along the ‘rosettes’
formed by the polar body chromosomes or in dividing diploid
nuclei (Figures 5I and 5J). In contrast, mRNA encoded
by the Ste12DOR gene is found within large chromatinassociated
foci that localize to telomeric regions during
anaphase (Figures 5K and 5L). Notably, this gene is highly
homologous and in close proximity to the tandemly repeated
Stellate gene cluster on the X chromosome, which
has been implicated in the maintenance of male fertility
through an RNA interference process involving the Su(Ste)
gene cluster located on the Y chromosome (Aravin et al.,
2001). Finally, roX1, a noncoding RNA involved in X chromosome
dosage compensation (Park et al., 2002), localizes
to the basal side of blastoderm nuclei, where the
X chromosome presumably resides (Figure 5M). As has
been demonstrated for roX1, these assorted DNA-associated
RNAs may be functioning to help organize, establish
and/or maintain chromatin domains. For Doc and Ste12DOR,
the chromosome-associated RNA may be
Figure 5. Cell Division and Nuclei-Associated Transcripts
(A–P) Surface plane (A–L and O) or sagittal (M, N, and P) views of stage 1–5 embryos hybridized with the indicated probes (mRNA green/nuclei red).
(A–H) Examples of mRNAs that localize to different sections of the cell division apparatus, including spindle poles (A), microtubule networks and centrosomes
(B, C, E, and F), the spindle midzone (G and H), or in proximity to metaphase chromosomes (D). (I and J) Doc-element transcripts localize to
centromeric chromatin regions on polar body chromosomes (I) and during mitosis in diploid nuclei (J). (K and L) Ste12DOR transcripts localize in chromatin-associated
foci during metaphase (K), which then become telomeric during anaphase (L). (M) roX1 RNA shows polarized enrichment in the
basal portion of blastoderm nuclei (arrowhead). (N) Bsg25D transcripts exhibit perinuclear localization around peripheral blastoderm and yolk nuclei.
(O and P) Several mRNAs exhibit nuclear retention; (O) cas transcripts are retained in groups of ventral nuclei following zygotic expression in stage 4
embryos, and (P) CG15634 mRNA is retained in nuclei situated just below the peripheral layer (arrowhead).
Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc. 181
functioning in the ‘repeat-associated small interfering
RNA’ (rasiRNA) pathway, which acts in part to suppress
transposable element activity (Slotkin and Martienssen,
2007). If so, this autoregulation would add a new dimension
to our understanding of this process.
Finally, other nuclei-associated mRNAs were observed
that range from mRNAs with tight perinuclear localization
(Figure 5N) to those that appear to be uniformly localized
throughout the nucleus (Figures 5O and 5P). These include
cas, CG15634, CG8552, Eip71CD, hb, Jra, kuk,
mfas, and scw. Interestingly, some of these are only retained
within nuclei that appear to be dropping out of
the blastoderm layer (Figures 5P and 6E). As these nuclei
are generally observed in pairs, and ‘nuclear fallout’ may
be a consequence of unsuccessful nuclear divisions
(Rothwell et al., 1998), this suggests a potential function
of these mRNAs in nuclear migration, sorting and/or
apoptosis.
Colocalization of RNAs and Proteins
In many of the cases cited above, where details are known
about protein localization or gene function, there is a striking
correlation between transcript localization and the
patterns or functions of the encoded proteins (Table S6).
To further illustrate some of these relationships, doublestaining
for selected transcripts, their protein products
or relevant markers was carried out (Figure 6). Examples
are shown for mRNAs that are apically localized (Figure
6A), cell division apparatus-associated (Figures 6B and
6C), or that reside at cell junctions (Figure 6D). In each
of these cases, mRNA localization is noted prior to the
appearance of protein, consistent with the view that
transcript localization generally predetermines protein
distribution at most subcellular destinations. Another intriguing
example is CG14438 mRNA, which appears to localize
at the level of centrioles and is found nestled within
structures labeled with the pericentriolar marker CNN
(Figure 6C). These types of examples support the notion
that localized transcripts, and the proteins they encode,
play central nucleation functions within the cell. These observations
further suggest that the necessary translation
and secretory machineries are generally available at each
of these sites.
Interestingly, a reverse correlation exists for mRNAs
that show nuclear retention, as shown for kuk and cas
transcripts (Figures 6E and 6F). Indeed, while Kuk protein
exhibits robust nuclear envelop localization in cortical
nuclei where the mRNA is cytoplasmic, no protein is
observed in or around the yolk nuclei in which kuk mRNA
is retained (Figure 6E). Similarly, in stage 9 neuroblasts,
Figure 6. Correlations in mRNA and
Protein Distribution Patterns
(A–F) Stage 4 (B–E), 5 (A) and 9 (F) embryos hybridized
with probes for the indicated apical
(A), cell division-associated (B and C), membrane-associated
(D), and nuclear retained (E
and F) mRNAs (red signal, left panels) and colabeled
with antibodies against the indicated
protein products (green signal, middle panels).
Overlaid mRNA and protein signals are shown
in the right panels. Nuclei are shown in blue in
the left and middle panels in (A)–(E). Arrowheads
in (E) and (F) indicate nuclei showing
an accumulation of kuk and cas mRNAs,
respectively.
182 Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc.
Cas protein is expressed robustly in cells with diffuse
cytoplasmic cas mRNA, but shows little or no protein expression
in cells where cas mRNA is nuclear (Figure 6F).
These examples suggest that nuclear mRNA retention
serves as a means of precisely timing or coordinating
protein expression to related cellular processes (Brandt
et al., 2006; Grosskortenhaus et al., 2006; Kambadur
et al., 1998; Pilot et al., 2006). Although this mechanism
of translational control has only once been documented
before (Prasanth et al., 2005), it may prove to be a
relatively common form of post-transcriptional gene
regulation.
DISCUSSION
The striking diversity and frequency of localized mRNAs
observed in this study, and the numerous correlations between
mRNA and protein distribution and function uncovered,
show that mRNA localization plays a far greater role
in coordinating cell physiology and anatomy than ever
previously suspected. Over the years, mRNA localization
has primarily been thought to coordinate specialized biological
processes such as morphogen gradient formation
and asymmetric cell division (Kloc et al., 2002; St Johnston,
2005). Our findings, however, necessitate a change
in perspective, implicating mRNA localization as a means
of regulating a vast number, if not the majority, of cell func-
tions.
mRNAs as Nucleators of Localized Complexes
The major inference of this study is that, since mRNA localization
generally occurs prior to encoded protein production,
and is so pervasive in scope, it must play a major
role in the nucleation and assembly of protein complexes
and organelles. One such example is anillin mRNA, which
encodes an actin-interacting protein and is localized in
dynamic patterns that closely resemble actin filament
distributions that form subsequently (Field and Alberts,
1995; Karr and Alberts, 1986). Another example is the
set of chromatin-associated RNAs, which may be acting
like some siRNAs to recruit chromatin remodeling complexes.
Other organelles likely to be controlled by mRNA
localization include various subcomponents of the cell
division apparatus. An example is CG14438 mRNA,
which localizes to centrosome-associated foci that
move during cell division to nestle within CNN-containing
clusters (Figure 6C). Notably, CG14438 encodes a
protein containing 21 zinc finger domains of the C2H2
type, which could readily serve as a nucleating scaffold
for RNA-containing centrosomal/centriolar complexes.
Our findings are consistent with recent studies
suggesting a role for RNAs in the regulation of centrosome
dynamics (Alliegro et al., 2006; Blower et al.,
2005; Lambert and Nagy, 2002). Moreover, our study extends
these previous findings, by revealing a large collection
of cell-division apparatus associated mRNAs that
await further characterization.
Localized RNAs as Functional Components
of RNP Complexes
In some cases, localized mRNAs are likely to have roles in
addition to targeting protein synthesis. For example, they
could well serve as structural or catalytic components
of RNP complexes, as seen for the many noncoding
RNA molecules found in ribosomes and spliceosomes
(Eddy, 2001; Erdmann et al., 2001; Prasanth and Spector,
2007). Indeed, a structural function for RNAs that localize
to the vegetal pole of Xenopus oocytes has recently been
uncovered (Kloc et al., 2005), and genetic studies of the
oskar gene in Drosophila have shown that protein null alleles
yield only a subset of the phenotypic traits exhibited
by RNA nulls (Jenny et al., 2006). This may also be the
case for many other genes for which protein null alleles
are considered to be genetic knockouts. These additional
roles would be consistent with the recent explosion of
newly recognized RNA-dependent processes, and with
the hypothesis that RNAs preceded proteins evolutionarily
in many cellular functions (Gilbert, 1986). Taken further,
the high proportion of the genome that is transcribed but
noncoding (Mattick, 2004; Willingham and Gingeras,
2006) suggests that we may only now be looking at the
tip of an iceberg.
RNA Localization and Disease
As our data suggests that mRNA localization has a key
role in targeting various cellular machineries, it can be
imagined that the integrity of transcript localization pathways
will be essential for ensuring appropriate cell growth
and differentiation, while also preventing cellular transformation
and tumorigenesis. Indeed, inappropriate targeting
of mRNAs would lead to aberrant protein distributions
within the cell, interference with normal regulatory pathways
and altered complex stoichiometries. In support of
this general view, several of the localized mRNAs characterized
in this study (ex: mira, dlg1, raps) encode wellknown
regulators of cell polarity and asymmetric cell division
with demonstrated tumor suppressor functions
(Caussinus and Gonzalez, 2005; Pagliarini and Xu, 2003;
Siegrist and Doe, 2005). Other examples include the
CG6522 mRNA, which exhibits an early cell membraneassociated
pattern. Homology searches reveal that this
mRNA likely encodes the Drosophila ortholog of Testin,
a cytoskeleton-associated LIM-domain protein that has
been identified as a tumor suppressor in mice and humans
(Drusco et al., 2005; Mueller et al., 2007). This example
typifies the potential usefulness of our database in identifying
genes with human disease relevance. Altogether, our
findings force a reassessment of many cancer pathways
for which modeling has been entirely concentrated on
events and interactions occurring at the post-translational
level.
Database Usage
The accompanying mRNA localization database will impact
many different biological disciplines. For example,
it will augment existing resources currently being used to
Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc. 183
study functional and regulatory genetic relationships.
Since our strategy provides additional levels of detail
with regards to spatio-temporal expression dynamics, this
dataset will also serve as a powerful tool for deciphering
and validating gene regulatory networks.
As there is a tight correlation between mRNA localization
and protein function, a more specific use for this resource
will be its predictive value in assigning functions
for uncharacterized genes. Indeed, 919 of the subcellularly
localized mRNAs in this study are encoded by
poorly characterized genes designated only by ‘‘CG’’ numbers.
For many of these, we can now postulate a function
with significantly higher confidence due to our knowledge
of where the mRNA is localized in the cell and
which other mRNAs, proteins and organelles it temporally
and spatially colocalizes with. These predictive
values may also suggest new functions for well-characterized
genes. Thus, this approach may help identify
or confirm many new components of regulatory complexes
and pathways. We expect that previously uncharacterized
macromolecular complexes will also be
discovered, many of which will be spatially and developmentally
regulated.
Another area of research that will benefit greatly from
this dataset will be the study of RNA cis-regulatory elements
and trans-acting factors that dictate the localization
of different mRNA subpopulations. In the past,
these studies have been tedious and complicated, due
in part to the complexity of RNA regulatory elements,
trans-acting complexes and the limitations of RNA structure
prediction algorithms. Unlike DNA cis-elements,
RNA regulatory elements tend to possess complex
features of both sequence and structure. The richness
of the data set provided here will reveal common sequence
and/or structural elements. Together with the
growing list of Drosophila genome sequences available
to assess sequence and structure conservation, rapid
advances in the identification of conserved cis-elements
and trans-acting machineries should be possible. In
point of fact, a consensus localization motif for mRNAs
in the diffuse apical category has been identified using
exactly this approach (dos Santos et al., 2007). Transcripts
that colocalize are also likely to share other
types of regulatory elements. For example, similarly localized
mRNAs may share common regulatory elements
used to couple translation and stability control.
Future Prospects and Considerations
An important point that remains to be confirmed is
whether the extent and variety of mRNA localization
events observed here occur in other tissues and organisms.
The amazing level of conservation between Drosophila
genes and developmental processes, and those
of higher organisms, suggests that these mRNA localization
frequencies and mechanisms will also be highly conserved.
In fact, it may have been the prior existence of
these elements and machineries that made early Drosophila
syncytial development an evolutionary possibility.
This would also be consistent with suggestions that RNAbased
processes have played a major role in promoting
phenotypic variation and complexity in higher eukaryotes
(Mattick, 2004).
EXPERIMENTAL PROCEDURES
Probe Production
The Drosophila gene collection 1 and 2 (DGC1 and 2) bacterial cDNA
libraries (Rubin et al., 2000; Stapleton et al., 2002) were obtained as
bacterial glycerol stocks in 96-well plates. Overnight cultures served
as templates for PCR using universal primers to amplify library cDNA
sequences containing flanking bacteriophage promoter elements
(T7, T3, or Sp6). Following PCR product purification on filter plates
(Whatman Inc., Clifton, NJ, USA), fragments were concentrated by
ethanol precipitation and resuspended in nuclease-free water. Antisense
RNA probes labeled with digoxigenin (DIG) were synthesized
as described by Le´ cuyer et al. (2007). The efficiency and accuracy of
all PCR amplifications and transcription reactions was systematically
verified by agarose gel electrophoresis after each step. Samples for
which no PCR product was obtained, or that contained multiple fragments,
were excluded from further analysis.
Embryo Collection, Fixation, and Fluorescent In Situ
Hybridization (FISH)
Wild-type Oregon R flies were maintained at 25
C and 50% humidity
in large plexiglass cages (60 3 60 3 60 cm). Following a 1 hr preclearing
step, embryos were collected on fresh food plates for 4.5 hr and
processed for fixation and storage as described by Le´ cuyer et al.
(2007). Conditions and methods for FISH and double FISH are also
described in detail by Le´ cuyer et al. (2007). The following antibodies
were used for immunostaining: mouse anti-BicD 4C2 and mouse
anti-Crumbs Cq4 (Developmental Studies Hybridoma Bank, Iowa
City, IA, USA), rabbit anti-Anillin (Provided by Dr. Julie Brill, The Hospital
for Sick Children, Toronto, ON, Canada), rabbit anti-Cas (Kambadur
et al., 1998), rabbit anti-CNN (Provided by Dr. Thomas Kaufman,
Indiana University, Bloomington, IN, USA), and rabbit anti-Kuk (Brandt
et al., 2006). Mouse and rabbit antisera were used at 1:10 and 1:100
dilutions, respectively.
Sample Imaging Procedure
Samples were analyzed on a Leica DMRA2 epifluorescence microscope
equipped with a rotating stage and a Q-imaging Retiga EX
digital camera (Quorum Technologies Inc., Guelph, ON, Canada)
and Openlab imaging software (Improvision Ltd., Coventry, England).
For each positive sample, a combination of low (103) and high magnification
(203 and 403) images were captured to document transcript
dynamics. As often as possible, care was taken to capture
images with embryos in the traditional orientation, with anterior to
the left and dorsal to the top. Controls probes were included in
each experiment to control for experimental variations in staining
levels. An autoexposure function was used as a semiquantitative
measure of maternal transcript abundance in stage 1–2 embryos,
with exposure times used to categorize expression levels as strong
(1–20 ms), intermediate (21–40 ms), weak (41–80 ms), or nonexpressed
(>81 ms). A maximum exposure time of 80 ms was used
at 203 magnification, as this provided a comparative standard, in
particular for transcripts that are degraded over time. The autoexposure
function was also used for DAPI images. Tyramide-Cy3 and
DAPI images were false-colored in green and red, respectively, using
Openlab, as this color scheme was found to provide the best contrast.
All overlaid images were saved as high-resolution TIF files. Figures
were constructed using Photoshop (Adobe Systems Inc., San
Jose, CA, USA).
184 Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc.
Data Annotation and Database Setup
All of the image and annotation data were organized within a MySQL
database using Perl-based annotation tools adapted from a previous
study (Tomancak et al., 2002). Images were sorted into appropriate
stage ranges (stages 1–3, 4–5, 6–7, 8–9, >10) and saved with a numerical
identifier. An annotation term hierarchy was created to document
mRNA localization, degradation, and zygotic expression characteristics
within each stage range. For each gene, relevant annotation terms
were selected and submitted to the database along with specific annotator
comments regarding staining quality and/or experimental obser-
vations.
Computational Analysis
The annotation data was converted into a binary matrix, containing
genes on one axis and localization terms on the other, where the presence
of a localization feature for a given gene was indicated numerically
as ‘‘1,’’ while lack of a feature was annotated as ‘‘0.’’ This matrix
was then used for GO term enrichment analysis. Functional GO annotations
for all genes were downloaded from Flybase (http://flybase.bio.
indiana.edu/genes/lk/function/). Annotations were up-propagated using
the GO hierarchy (Ashburner et al., 2000), and calculations were restricted
to genes that were both GO annotated and analyzed in this
study (1651 genes). The hypergeometric distribution was used to calculate
probabilities of overlap between each localization category
against all GO categories containing three or more genes. The Benjamini-Hochberg
procedure (Benjamini and Hochberg, 1995) was used
to control for multiple testing by computing a P-value threshold corresponding
to a false discovery rate (FDR) of 0.25. Transcript subgroups
were also analyzed independently for GO term enrichments using
EASE (Hosack et al., 2003). EASE scores of less that 0.05 were considered
significant, as reported previously (Tadros et al., 2007a).
Supplemental Data
Supplemental Data include five figures and six tables and can be found
with this article online at http://www.cell.com/cgi/content/full/131/1/
174/DC1/.
ACKNOWLEDGMENTS
We thank J. T. Westwood and colleagues at the Canadian Drosophila
Microarray Centre for the DGC stocks, as well as J. Brill and P. Goldbach
for providing Anillin antisera. We thank B.J. Blencowe, H. D. Lipshitz,
and C. A. Smibert for reviewing the manuscript prior to submission.
We acknowledge funding from the National Cancer Institute
of Canada and the Canadian Institutes of Health Research (CIHR).
E.L. is supported by a CIHR fellowship and H.M.K. by a CIHR Senior
Scientist award.
Received: May 8, 2007
Revised: July 30, 2007
Accepted: August 2, 2007
Published: October 4, 2007
REFERENCES
Adereth, Y., Dammai, V., Kose, N., Li, R., and Hsu, T. (2005). RNAdependent
integrin alpha3 protein localization regulated by the
Muscleblind-like protein MLP1. Nat. Cell Biol. 7, 1240–1247.
Alliegro, M.C., Alliegro, M.A., and Palazzo, R.E. (2006). Centrosomeassociated
RNA in surf clam oocytes. Proc. Natl. Acad. Sci. USA
103, 9034–9038.
Aravin, A.A., Naumova, N.M., Tulin, A.V., Vagin, V.V., Rozovsky, Y.M.,
and Gvozdev, V.A. (2001). Double-stranded RNA-mediated silencing
of genomic tandem repeats and transposable elements in the D. melanogaster
germline. Curr. Biol. 11, 1017–1027.
Arbeitman, M.N., Furlong, E.E., Imam, F., Johnson, E., Null, B.H.,
Baker, B.S., Krasnow, M.A., Scott, M.P., Davis, R.W., and White,
K.P. (2002). Gene expression during the life cycle of Drosophila melanogaster.
Science 297, 2270–2275.
Ashburner, M., Ball, C.A., Blake, J.A., Botstein, D., Butler, H., Cherry,
J.M., Davis, A.P., Dolinski, K., Dwight, S.S., Eppig, J.T., et al. (2000).
Gene ontology: tool for the unification of biology. The Gene Ontology
Consortium. Nat. Genet. 25, 25–29.
Bashirullah, A., Cooperstock, R.L., and Lipshitz, H.D. (1998). RNA
localization in development. Annu. Rev. Biochem. 67, 335–394.
Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery
rate: A practical and powerful approach to multiple testing. J. Roy.
Statist. Soc. Ser. B. Met. 57, 289–300.
Blobel, G., and Dobberstein, B. (1975). Transfer of proteins across
membranes. I. Presence of proteolytically processed and unprocessed
nascent immunoglobulin light chains on membrane-bound
ribosomes of murine myeloma. J. Cell Biol. 67, 835–851.
Blower, M.D., Nachury, M., Heald, R., and Weis, K. (2005). A Rae1containing
ribonucleoprotein complex is required for mitotic spindle
assembly. Cell 121, 223–234.
Brandt, A., Papagiannouli, F., Wagner, N., Wilsch-Brauninger, M.,
Braun, M., Furlong, E.E., Loserth, S., Wenzl, C., Pilot, F., Vogt, N.,
et al. (2006). Developmental control of nuclear size and shape by
Kugelkern and Kurzkern. Curr. Biol. 16, 543–552.
Broadus, J., Fuerstenberg, S., and Doe, C.Q. (1998). Staufen-dependent
localization of prospero mRNA contributes to neuroblast daughter-cell
fate. Nature 391, 792–795.
Caussinus, E., and Gonzalez, C. (2005). Induction of tumor growth by
altered stem-cell asymmetric division in Drosophila melanogaster.
Nat. Genet. 37, 1125–1129.
Czaplinski, K., and Singer, R.H. (2006). Pathways for mRNA localization
in the cytoplasm. Trends Biochem. Sci. 31, 687–693.
Davis, I., and Ish-Horowicz, D. (1991). Apical localization of pair-rule
transcripts requires 30
sequences and limits protein diffusion in the
Drosophila blastoderm embryo. Cell 67, 927–940.
De Renzis, S., Elemento, O., Tavazoie, S., and Wieschaus, E.F. (2007).
Unmasking Activation of the Zygotic Genome Using Chromosomal Deletions
in the Drosophila Embryo. PLoS Biol. 5, e117. 10.1371/journal.
pbio.0050117.
dos Santos, G., Simmonds, A.J., and Krause, H.M. (2007). A stem-loop
structure in the wingless transcript defines a consensus motif for apical
RNA transport. Development, in press.
Driever, W., and Nusslein-Volhard, C. (1988). The bicoid protein determines
position in the Drosophila embryo in a concentration-dependent
manner. Cell 54, 95–104.
Drusco, A., Zanesi, N., Roldo, C., Trapasso, F., Farber, J.L., Fong, L.Y.,
and Croce, C.M. (2005). Knockout mice reveal a tumor suppressor
function for Testin. Proc. Natl. Acad. Sci. USA 102, 10947–10951.
Dubowy, J., and Macdonald, P.M. (1998). Localization of mRNAs to
the oocyte is common in Drosophila ovaries. Mech. Dev. 70, 193–195.
Eddy, S.R. (2001). Non-coding RNA genes and the modern RNA world.
Nat. Rev. Genet. 2, 919–929.
Ephrussi, A., Dickinson, L.K., and Lehmann, R. (1991). Oskar organizes
the germ plasm and directs localization of the posterior determinant
nanos. Cell 66, 37–50.
Ephrussi, A., and St Johnston, D. (2004). Seeing is believing: the bicoid
morphogen gradient matures. Cell 116, 143–152.
Erdmann, V.A., Barciszewska, M.Z., Hochberg, A., de Groot, N., and
Barciszewski, J. (2001). Regulatory RNAs. Cell. Mol. Life Sci. 58,
960–977.
Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc. 185
Field, C.M., and Alberts, B.M. (1995). Anillin, a contractile ring protein
that cycles from the nucleus to the cell cortex. J. Cell Biol. 131, 165–
178.
Gavis, E.R., and Lehmann, R. (1992). Localization of nanos RNA
controls embryonic polarity. Cell 71, 301–313.
Gilbert, W. (1986). The RNA world. Nature 319, 618.
Gore, A.V., Maegawa, S., Cheong, A., Gilligan, P.C., Weinberg, E.S.,
and Sampath, K. (2005). The zebrafish dorsal axis is apparent at the
four-cell stage. Nature 438, 1030–1035.
Grosskortenhaus, R., Robinson, K.J., and Doe, C.Q. (2006). Pdm and
Castor specify late-born motor neuron identity in the NB7–1 lineage.
Genes Dev. 20, 2618–2627.
Hosack, D.A., Dennis, G., Jr., Sherman, B.T., Lane, H.C., and Lempicki,
R.A. (2003). Identifying biological themes within lists of genes
with EASE. Genome Biol. 4, R70.
Hughes, J.R., Bullock, S.L., and Ish-Horowicz, D. (2004). Inscuteable
mRNA localization is dynein-dependent and regulates apicobasal
polarity and spindle length in Drosophila neuroblasts. Curr. Biol. 14,
1950–1956.
Jansen, R.P. (2001). mRNA localization: message on the move. Nat.
Rev. Mol. Cell Biol. 2, 247–256.
Jenny, A., Hachet, O., Zavorszky, P., Cyrklaff, A., Weston, M.D., Johnston,
D.S., Erdelyi, M., and Ephrussi, A. (2006). A translation-independent
role of oskar RNA in early Drosophila oogenesis. Development
133, 2827–2833.
Kambadur, R., Koizumi, K., Stivers, C., Nagle, J., Poole, S.J., and
Odenwald, W.F. (1998). Regulation of POU genes by castor and
hunchback establishes layered compartments in the Drosophila
CNS. Genes Dev. 12, 246–260.
Karr, T.L., and Alberts, B.M. (1986). Organization of the cytoskeleton in
early Drosophila embryos. J. Cell Biol. 102, 1494–1509.
Kloc, M., Wilk, K., Vargas, D., Shirato, Y., Bilinski, S., and Etkin, L.D.
(2005). Potential structural role of non-coding and coding RNAs in
the organization of the cytoskeleton at the vegetal cortex of Xenopus
oocytes. Development 132, 3445–3457.
Kloc, M., Zearfoss, N.R., and Etkin, L.D. (2002). Mechanisms of subcellular
mRNA localization. Cell 108, 533–544.
Lambert, J.D., and Nagy, L.M. (2002). Asymmetric inheritance of centrosomally
localized mRNAs during embryonic cleavages. Nature 420,
682–686.
Lawrence, J.B., and Singer, R.H. (1986). Intracellular localization of
messenger RNAs for cytoskeletal proteins. Cell 45, 407–415.
Le´ cuyer, E., Parthasarathy, N., and Krause, H.M. (2007). Fluorescent In
Situ Hybridization Protocols in Drosophila Embryos and Tissues. In
Methods Mol. Biol., C. Dahmann, ed. (Totowa, NJ: Humana Press,
Inc.), pp. 289–302.
Li, P., Yang, X., Wasser, M., Cai, Y., and Chia, W. (1997). Inscuteable
and Staufen mediate asymmetric localization and segregation of prospero
RNA during Drosophila neuroblast cell divisions. Cell 90, 437–447.
Long, R.M., Singer, R.H., Meng, X., Gonzalez, I., Nasmyth, K., and Jansen,
R.P. (1997). Mating type switching in yeast controlled by asymmetric
localization of ASH1 mRNA. Science 277, 383–387.
Mattick, J.S. (2004). RNA regulation: a new genetics? Nat. Rev. Genet.
5, 316–323.
Melton, D.A. (1987). Translocation of a localized maternal mRNA to the
vegetal pole of Xenopus oocytes. Nature 328, 80–82.
Mingle, L.A., Okuhama, N.N., Shi, J., Singer, R.H., Condeelis, J., and
Liu, G. (2005). Localization of all seven messenger RNAs for the actin-polymerization
nucleator Arp2/3 complex in the protrusions of
fibroblasts. J. Cell Sci. 118, 2425–2433.
Mueller, W., Nutt, C.L., Ehrich, M., Riemenschneider, M.J., von Deimling,
A., van den Boom, D., and Louis, D.N. (2007). Downregulation of
RUNX3 and TES by hypermethylation in glioblastoma. Oncogene 26,
583–593.
Neuman-Silberberg, F.S., and Schupbach, T. (1993). The Drosophila
dorsoventral patterning gene gurken produces a dorsally localized
RNA and encodes a TGF alpha-like protein. Cell 75, 165–174.
Pagliarini, R.A., and Xu, T. (2003). A genetic screen in Drosophila for
metastatic behavior. Science 302, 1227–1231.
Park, Y., Kelley, R.L., Oh, H., Kuroda, M.I., and Meller, V.H. (2002).
Extent of chromatin spreading determined by roX RNA recruitment
of MSL proteins. Science 298, 1620–1623.
Pilot, F., Philippe, J.M., Lemmers, C., Chauvin, J.P., and Lecuit, T.
(2006). Developmental control of nuclear morphogenesis and anchoring
by charleston, identified in a functional genomic screen of Drosophila
cellularisation. Development 133, 711–723.
Prasanth, K.V., Prasanth, S.G., Xuan, Z., Hearn, S., Freier, S.M., Bennett,
C.F., Zhang, M.Q., and Spector, D.L. (2005). Regulating gene
expression through RNA nuclear retention. Cell 123, 249–263.
Prasanth, K.V., and Spector, D.L. (2007). Eukaryotic regulatory RNAs:
an answer to the ‘genome complexity’ conundrum. Genes Dev. 21, 11–
42.
Rothwell, W.F., Fogarty, P., Field, C.M., and Sullivan, W. (1998). Nuclear-fallout,
a Drosophila protein that cycles from the cytoplasm to
the centrosomes, regulates cortical microfilament organization. Development
125, 1295–1303.
Rubin, G.M., Hong, L., Brokstein, P., Evans-Holm, M., Frise, E., Stapleton,
M., and Harvey, D.A. (2000). A Drosophila complementary DNA
resource. Science 287, 2222–2224.
Schuldt, A.J., Adams, J.H., Davidson, C.M., Micklem, D.R., Haseloff,
J., St Johnston, D., and Brand, A.H. (1998). Miranda mediates asymmetric
protein and RNA localization in the developing nervous system.
Genes Dev. 12, 1847–1857.
Seydoux, G., and Braun, R.E. (2006). Pathway to totipotency: lessons
from germ cells. Cell 127, 891–904.
Seydoux, G., and Dunn, M.A. (1997). Transcriptionally repressed germ
cells lack a subpopulation of phosphorylated RNA polymerase II in
early embryos of Caenorhabditis elegans and Drosophila melanogaster.
Development 124, 2191–2201.
Siegrist, S.E., and Doe, C.Q. (2005). Microtubule-induced Pins/Galphai
cortical polarity in Drosophila neuroblasts. Cell 123, 1323–1335.
Simmonds, A.J., dosSantos, G., Livne-Bar, I., and Krause, H.M. (2001).
Apical localization of wingless transcripts is required for wingless
signaling. Cell 105, 197–207.
Slotkin, R.K., and Martienssen, R. (2007). Transposable elements and
the epigenetic regulation of the genome. Nat. Rev. Genet. 8, 272–285.
St Johnston, D. (2005). Moving messages: the intracellular localization
of mRNAs. Nat. Rev. Mol. Cell Biol. 6, 363–375.
Stapleton, M., Liao, G., Brokstein, P., Hong, L., Carninci, P., Shiraki, T.,
Hayashizaki, Y., Champe, M., Pacleb, J., Wan, K., et al. (2002). The
Drosophila gene collection: identification of putative full-length cDNAs
for 70% of D. melanogaster genes. Genome Res. 12, 1294–1300.
Tadros, W., Goldman, A.L., Babak, T., Menzies, F., Vardy, L., OrrWeaver,
T., Hughes, T.R., Westwood, J.T., Smibert, C.A., and Lipshitz,
H.D. (2007a). SMAUG is a major regulator of maternal mRNA destabilization
in Drosophila and its translation is activated by the PAN GU
kinase. Dev. Cell 12, 143–155.
Tadros, W., Westwood, J.T., and Lipshitz, H.D. (2007b). The Motherto-Child
Transition. Dev. Cell 12, 847–849.
Takizawa, P.A., Sil, A., Swedlow, J.R., Herskowitz, I., and Vale, R.D.
(1997). Actin-dependent localization of an RNA encoding a cell-fate
determinant in yeast. Nature 389, 90–93.
186 Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc.
Tepass, U., Theres, C., and Knust, E. (1990). crumbs encodes an EGFlike
protein expressed on apical membranes of Drosophila epithelial
cells and required for organization of epithelia. Cell 61, 787–799.
Tomancak, P., Beaton, A., Weiszmann, R., Kwan, E., Shu, S., Lewis,
S.E., Richards, S., Ashburner, M., Hartenstein, V., Celniker, S.E., and
Rubin, G.M. (2002). Systematic determination of patterns of gene
expression during Drosophila embryogenesis. Genome Biol. 3,
RESEARCH0088.
Van Doren, M., Williamson, A.L., and Lehmann, R. (1998). Regulation of
zygotic gene expression in Drosophila primordial germ cells. Curr. Biol.
8, 243–246.
Willingham, A.T., and Gingeras, T.R. (2006). TUF love for ‘‘junk’’ DNA.
Cell 125, 1215–1220.
Zhang, H.L., Eom, T., Oleynikov, Y., Shenoy, S.M., Liebelt, D.A.,
Dictenberg, J.B., Singer, R.H., and Bassell, G.J. (2001). Neurotrophin-induced
transport of a beta-actin mRNP complex increases
beta-actin levels and stimulates growth cone motility. Neuron 31,
261–275.
Zhang, J., Houston, D.W., King, M.L., Payne, C., Wylie, C., and Heasman,
J. (1998). The role of maternal VegT in establishing the primary
germ layers in Xenopus embryos. Cell 94, 515–524.
Cell 131, 174–187, October 5, 2007 ª2007 Elsevier Inc. 187