Introduction
All cells require efficient mRNA translation, but protein synthesis
capacity varies by cell type [1]. Cell type-specific translation
kinetics determine proliferation potential and maintain unique cellular
properties. Cancer cells may alter translation initiation to globally
increase protein synthesis and sustain proliferation [2]. A delicate
balance of translation activity is also critical for hematopoietic stem
cell (HSC) homeostasis, with increased or decreased global translation
impairing HSC function [3]. Variation in ribosome concentration is
one mechanism cells use to control translation output [4]. In
addition to influencing growth and proliferative capacity, ribosome
concentration can affect cell fate through biased translation of certain
mRNAs [4, 5]. Calibration of ribosome abundance and altered protein
synthesis has been described in several developmental contexts.
Ribosomal protein levels decrease when mouse embryonic stem cells
differentiate to embryoid bodies while global translation efficiency
increases [6]. Similarly, in the Drosophila female germline
production of ribosomal RNA (rRNA) and ribosome assembly factors
decreases along the differentiation pathway [7, 8] yet germline stem
cell differentiation is associated with increased global protein
synthesis [8]. These studies suggest that the relationship between
ribosome abundance and cell fate is more complex than simply meeting the
metabolic needs of progenitors versus differentiated progeny.
Regulation of ribosomal RNA synthesis is one way of tuning global
translation capacity. Decreased rRNA synthesis can affect developmental
transitions in the Drosophila female germline [7] and
mammalian tissue culture cells [9]. In eukaryotes, rRNA is
transcribed by RNA polymerase I from tandem repeats of a gene encoding
precursor rRNA (pre-rRNA). Pre-rRNA is processed into individual 28S,
18S, and 5.8S rRNAs while a fourth rRNA, 5S rRNA, is transcribed by RNA
polymerase III. Ribosomal RNA transcription occurs at a specialized site
in the nucleus called the nucleolus. Differences in rRNA transcription
can be recognized via changes in nucleolus size (a larger nucleolus
usually indicates more rRNA transcription) and via direct detection of
nascent rRNA. For example, Drosophila female germline stem cells
have large nucleoli and high rates of rRNA transcription but their
differentiated progeny have smaller nucleoli and reduced rRNA synthesis
[7]. Downregulation of rRNA synthesis also occurs during
differentiation in the mammalian forebrain [10]. Decreased ribosome
abundance appears to be a common feature of neural differentiation:
additional studies have shown that ribosomal protein production [11]
and rRNA synthesis [12] decreases in post-mitotic neurons.
Compartment-specific changes, such as loss of ribosomes from mature
axons, also occurs during neural development [13, 14]. These
findings raise interesting questions regarding how ribosome biogenesis
is regulated to meet the mRNA translation needs of neurons. One
possibility is that components of the translation machinery, including
rRNAs, are primarily synthesized in progenitors then passed to neurons
during differentiating divisions. The absence of cytokinesis in neurons
and the long half-life of ribosomes (days to weeks) could establish a
pool of ribosomes sufficient to meet the protein synthesis needs of
neurons in the absence of any autonomous ribosome production [15].
Drosophila neural stem cells, called neuroblasts, undergo
multiple rounds of asymmetric self-renewing divisions to ultimately
produce neurons and glia. There are two main types of neuroblast in the
larval brain: type I neuroblasts produce a transient progenitor, the
ganglion mother cell (GMC), at each division while type II neuroblasts
produce intermediate neural progenitors (INPs) that undergo multiple
rounds of asymmetric divisions, self-renewing and producing a GMC
[16]. In both lineages the GMC divides once to produce post-mitotic
progeny. Previous work has shown that the nucleolus is smaller in
differentiated cells (INPs, GMCs, neurons) compared to neuroblasts
[17, 18], suggesting that rRNA synthesis is restricted upon
differentiation. The transcription factor Myc is likely a crucial
regulator of this restriction. Myc is expressed at high levels in
neuroblasts but is absent from INPs, GMCs and neurons [18]. Myc
promotes cell growth and proliferation via several pathways, including
transcriptional activation of genes encoding RNA polymerase I subunits
[19].
While decreased nucleolus size suggests rRNA synthesis is restricted
upon neural differentiation in Drosophila , multiple questions
remain. First, does direct measurement of rRNA synthesis confirm this
prediction? Second, to what degree is rRNA synthesis restricted along
the differentiation pathway? Third, since the absence of Myc is
predicted to limit rRNA synthesis, how do differentiated progenitors and
neurons obtain the necessary amount of rRNA to support their translation
needs? Here we show that high levels of nascent rRNAs are present in
neuroblasts, INPs and GMCs but rRNA synthesis is severely restricted in
neurons. Our data reveal that neural progenitors pass rRNA to their
progeny during cytokinesis and suggest that the rRNA in INPs and GMCs is
at least partly derived from their neuroblast parent. Ultimately, the
rRNA in GMCs is inherited by neurons at cytokinesis. Finally, we show
that progenitor-derived rRNAs are sufficient to support brain
development and normal protein synthesis in neurons. This work supports
a model in which rRNA inheritance establishes cell type-specific
translation programs along the neural differentiation pathway.