Amplification of neural stem cell proliferation by intermediate progenitor cells in Drosophila brain development
© Bello et al.. 2008
Received: 27 November 2007
Accepted: 19 February 2008
Published: 19 February 2008
In the mammalian brain, neural stem cells divide asymmetrically and often amplify the number of progeny they generate via symmetrically dividing intermediate progenitors. Here we investigate whether specific neural stem cell-like neuroblasts in the brain of Drosophila might also amplify neuronal proliferation by generating symmetrically dividing intermediate progenitors.
Cell lineage-tracing and genetic marker analysis show that remarkably large neuroblast lineages exist in the dorsomedial larval brain of Drosophila. These lineages are generated by brain neuroblasts that divide asymmetrically to self renew but, unlike other brain neuroblasts, do not segregate the differentiating cell fate determinant Prospero to their smaller daughter cells. These daughter cells continue to express neuroblast-specific molecular markers and divide repeatedly to produce neural progeny, demonstrating that they are proliferating intermediate progenitors. The proliferative divisions of these intermediate progenitors have novel cellular and molecular features; they are morphologically symmetrical, but molecularly asymmetrical in that key differentiating cell fate determinants are segregated into only one of the two daughter cells.
Our findings provide cellular and molecular evidence for a new mode of neurogenesis in the larval brain of Drosophila that involves the amplification of neuroblast proliferation through intermediate progenitors. This type of neurogenesis bears remarkable similarities to neurogenesis in the mammalian brain, where neural stem cells as primary progenitors amplify the number of progeny they generate through generation of secondary progenitors. This suggests that key aspects of neural stem cell biology might be conserved in brain development of insects and mammals.
Neural stem cells are primary precursors that have the ability to renew themselves at each division such that one of the two daughter cells retains stem cell identity, while the other enters a program of differentiation and contributes to a continuous supply of neural cell types. Understanding how neural stem cells maintain their pluripotent state and how their progeny differentiate into distinct neural fates is of central importance for understanding nervous system development (for recent reviews, see [1–3]). Neural stem cells must exert a tight control over proliferative divisions so as to generate the appropriate number of neural progeny necessary to populate the nervous system but not to produce so many self-renewing daughters that neoplastic overgrowth occurs . Therefore, a better comprehension of the mechanisms that control the behavior of neuronal stem cells and their progeny may also be important for understanding brain tumors [5, 6].
The Drosophila central nervous system is an excellent simple model system for analyzing the molecular mechanisms that control neural stem cell divisions (for recent reviews, see [7, 8]). Drosophila neural stem cells, called neuroblasts (NBs), delaminate as single cells from the neuroectoderm and undergo repeated asymmetric cell divisions, each of which self-renew the NB while producing a smaller neural progenitor cell called a ganglion mother cell (GMC). Compared to the NB, the GMC adopts a radically opposite fate and undergoes a single neurogenic division to produce two cells that exit the cell cycle and differentiate (reviewed in [9–12]). During embryogenesis, each NB produces a lineage of 10–20 primary neural cells that contribute to the functional circuitry of the larva. Following a period of quiescence, most NBs resume their asymmetric mode of proliferative divisions during post-embryonic development and generate the lineage-related clusters of secondary adult-specific neurons that make up the bulk of the adult central brain and thoracic ganglia [13–16].
Mechanisms involved in NB division and neural proliferation during embryogenesis have been studied in great detail (reviewed in [7, 17–19]). NB divisions are known to be molecularly as well as morphologically asymmetric, and a number of key intrinsic and extrinsic factors that control the asymmetrical and self-renewing divisions of these NBs have been identified. Among these, a central role is played by molecular polarity cues that establish the apico-basal polarity of the NB and enable the asymmetric segregation of localized cell-fate determinants from the NB to the GMCs at each asymmetric cell division. Although considerable insight has been attained into the mechanisms by which NB polarity is established and maintained, little is known about the function of the proteins that are asymmetrically localized to the GMC. The best characterized of these fate determinants is the homeodomain protein Prospero, which is synthesized in the NB and localized at the cell cortex in a polarized manner. Upon segregation to the GMC, Prospero acts in the nucleus to repress NB-specific gene expression (including genes required for self-renewal) and activate genes for GMC fate specification and terminal differentiation of post-mitotic neurons [20–23]. Asymmetric segregation of Prospero protein is mediated by the adaptor coiled-coil protein Miranda. Once segregated from the NB to the GMC, Miranda is degraded, thereby releasing Prospero from the cell cortex and allowing it to enter the nucleus [24–26]. Indeed, the nuclear localization of Prospero is one of the first molecular differences between the self-renewing NB and a differentiating cell [27, 28].
During the postembryonic period of neurogenesis, the NBs of the central brain and thoracic ganglia are thought to undergo a similar proliferation program and express many of the asymmetric cell fate determinants that characterize embryonic neurogenesis [29, 30]. Nuclear localization of Prospero is manifest in GMCs and postmitotic neurons of the larval brain, and loss of prospero in somatic clones results in massive overproliferation of cells that express molecular markers of NBs [31–33]. Additionally, numerous other molecular control elements are likely to be required for the continuous mitotic activity of NBs during postembryonic life (reviewed in ).
Controlled neuronal proliferation is especially important for the generation of the adult brain. The mature brain of Drosophila is an exceedingly complex structure with numerous highly organized neuropil assemblies, such as the mushroom bodies, central complex and antennal lobes, as well as other specialized neuropils and major fiber tracts required for complex behavioral functions . Remarkably, approximately 95% of the neurons that make up the adult brain are post-embryonic in origin, and in the central brain all of these neurons are produced by a set of only about 100 bilaterally symmetrical NBs [36, 37]. Given the fact that 100 NB pairs generate the tens of thousands of differentiated, spatially heterogeneous neurons in the adult central brain, sophisticated mechanisms for lineage- and region-specific amplification control of NB proliferation are likely to be required during post-embryonic brain development. However, with the exception of rough estimates, which suggest that each brain NB might undergo between 40 and 60 rounds of post-embryonic mitosis to produce lineages of 100–150 neurons, very little is known about this process and the underlying molecular mechanisms.
Here we report that a striking amplification of neuronal proliferation is achieved by specific brain NBs during postembryonic development through the generation of intermediate progenitor cells (IPs). Using cell lineage-tracing and marker analysis, we show that remarkably large NB lineages develop in the dorsomedial (DM) area of the larval brain. Like any other lineages in the brain, they derive from unique NB precursors that remain associated with their post-mitotic neuronal progeny. In addition, they contain a large pool of cells that do not express neuronal differentiation markers, are engaged in the cell cycle, and show mitotic activity. While some of these mitotically active cells are GMCs, the others express NB-specific molecular markers and divide repeatedly to produce neural progeny, implying that they are IPs. The proliferative divisions of these IPs are morphologically symmetrical, but molecularly asymmetrical in that cell fate determinants such as Prospero and Miranda are segregated into only one of the daughter cells. The IPs are generated by a specific set of NBs that do not segregate Prospero to their smaller daughter cell, thereby allowing this cell to retain proliferative capacity instead of undergoing its final neurogenic division. The amplification of NB proliferation through IPs reported here for Drosophila bears remarkable similarities to mammalian neurogenesis, where neural stem cells as primary progenitors often amplify the number of progeny they generate via symmetrically dividing secondary progenitors (reviewed in ). This suggests that key aspects of neural stem cell biology might be conserved in brain development of flies and mammals.
Large neuroblast lineages are located in the dorsomedial brain hemispheres
Prominent among these were unusually large clones recoverable at the DM margins of the brain hemispheres (Figure 1b). Six NBs located in the most medial position of each hemisphere were found to generate this type of clone, hereafter referred to as 'DM lineages' or 'DM clones'. As detailed below, the parental DM NBs were easily identifiable owing to the signature pattern of Miranda-positive cells that followed the lateral to medial orientation of their progeny in these labeled clones. Morphologically, DM NBs were indistinguishable from other NBs in the central brain or in the ventral ganglia. Thus, cell volume measurements of DM and non-DM NBs in third larval instar brains gave comparable values of 344 ± 94 μm3 (n = 12) and 424 ± 110 μm3 (n = 13), respectively. Preliminary analysis of the axonal tracts suggests that the large NB clones in the dorsal brain correspond to the pl and pm subgroups of the Dorsoposterior medial (DPM) lineages previously described (data not shown) .
To compare the proliferative capacity of the DM NBs with that of other NBs in the larval central nervous system, we quantified the number of cells in DM NB lineages, in mushroom body NB lineages, and in other NB lineages scored randomly in different brain and ventral ganglion regions of the late third instar larvae shortly before pupation (96 h ALH). The number of cells in the DM lineages had an average value of 450 (range 370–580). Remarkably, this was more than twice the average number of cells observed for the larval lineages of the mushroom body NBs (184 ± 17, n = 17) or for other larval NB lineages scored in other areas of the central nervous system (Figure 1c).
To determine the rate of clone size increase during larval central nervous system development, we counted the number of cells in MARCM-labeled DM NB clones, mushroom body NB clones and other dorsal brain NB clones at various larval stages (Figure 1d). Following a quiescent phase in the early developing larva, most NBs had entered mitosis by the late second larval instar stage . Our observations show that at this stage (48 h ALH), NBs in the dorsal brain had generated only a small number of postembryonic cells and that no pronounced lineage-specific differences in progeny number was apparent (Figure 1d, 48 h ALH). However, at 72 h and 96 h ALH, the DM lineages had increased markedly in size when compared to other dorsal brain NB lineages, indicating an approximate four-fold increase in their rate of proliferation (Figure 1d).
To investigate this further, we cultured MARCM-labeled brain explants in 5-bromodeoxyuridine (BrdU) and then used anti-BrdU immunocytochemistry to determine the number of cells engaged in S-phase in DM clones compared to other NB clones of the central brain. Following a 90 minute pulse of BrdU incorporation in L3 brain explants, we found a markedly higher number of BrdU-positive cells in DM clones (38 ± 8 BrdU positive cells, n = 8 clones) than in the other NB clones scored at random in dorsal brain regions of the same specimens (4 ± 1.5, n = 27). (This higher rate of BrdU incorporation in DM clones was also observed at earlier stages and in various conditions of incubation; data not shown.)
These data indicate that a significant amplification of proliferation occurs in the DM lineages when compared to other NB lineages of the central brain (hereafter collectively referred to as 'non-DM' lineages).
DM lineages contain a large population of mitotically active progenitor cells
The large number of cells found in the DM NB clones could, in principle, be due to an unusually high rate of mitotic activity of the DM NBs. However, immunodetection of mitotic DNA in MARCM clones (via the phospho-histone H3 (PH3) epitope) revealed a comparable mitotic frequency in these NBs (22.5%, n = 40) compared to NBs found in dorsal (16.7 %, n = 48) or ventral (21.6 %, n = 97) brain lineages. This prompted us to search for other types of progenitor cells in these lineages. To this aim, we first characterized molecular markers enabling in situ detection of mitotically active versus post-mitotic cells in labeled NB lineages of the larval brain.
Could these smaller Elav-negative cells associated with the NBs be GMCs? To investigate this, we first studied the expression of the coiled-coil protein Miranda. The miranda gene has been reported to be expressed in larval NBs but not in their GMCs ; Miranda expression might, therefore, be a useful marker for differentiating NB-like cells from GMCs. In non-DM lineages, Miranda was strongly expressed in the NBs but only very weakly expressed in the set of smaller, Elav-negative cells associated with the NBs, suggesting that these Elav-negative cells were GMCs (Figure 2a,a'). (Their weak expression of Miranda could be due to perdurance of the protein during cell divisions; see also [29, 30]). In DM lineages, Miranda was strongly expressed in the NB; however, in contrast to non-DM lineages, distinct Miranda expression was also observed in many of the smaller, Elav-negative cells associated with the NBs (Figure 2b,b'). This suggests that the smaller Elav-negative/Miranda-positive cells in the DM lineages might not be GMC-like, but might have properties that are more NB-like. To investigate this further, we next attempted to find other markers for progenitor cells and, thus, examined the expression of Cyclin E (CycE) and PH3 as markers of mitotically active cells.
In green fluorescent protein (GFP)-labeled non-DM NB clones, used as control, a small number of GMCs were observed as small NB-associated cells expressing either CycE or PH3 (Figure 2c,c'). At 96 h ALH we found an average of two CycE-positive cells (range one to five) and a maximum of one cell engaged in mitosis as visualized by anti-PH3 (Figure 2e) . This pattern was consistent with live imaging data obtained in experiments on cultured nervous systems to monitor asymmetric NB divisions . Thus, as in the embryo, these larval NBs divide by a budding process that generates a set of smaller GMCs, each GMC is born adjacent to the previous one, and the division of the 'oldest' GMC is delayed compared to that of the NB.
Contrasting with this simple pattern, DM lineages contained an average of 38 CycE-positive cells located around the NB, and many scattered mitoses, up to 14 per clone, were observed by PH3 immunoreactivity (Figure 2d,d',e). This strikingly high level of ongoing mitotic activity and engagement in the cell cycle in DM lineages compared to other central brain lineages (including mushroom body lineages) was seen at all stages of larval development examined (Figure 2f). These findings indicate that significantly elevated mitotic activity occurs among the numerous small NB-associated cells in larval DM lineages. Moreover, they are in accordance with the idea that these cells do not adopt a GMC fate, but rather remain mitotically active and continue to proliferate. In this case, these cells would have the characteristics of IPs that amplify the proliferation of their parent NBs (primary progenitors) in the DM lineages.
Molecular markers reveal two types of non-neuroblast progenitor cells in DM lineages
If some of the mitotically active cells in DM NB clones are amplifying IPs, they might be expected to have cellular and molecular features in common with proliferating NBs. To investigate this, we first examined the expression patterns of Prospero, Miranda, and CycE in NBs of non-DM lineages, used as control, as well as in the small NB-associated progenitors of the DM lineages. For this, MARCM clones induced at larval hatching were scored at 96 h ALH. Importantly, we further restricted our analysis to cells engaged in mitosis (PH3-positive) in order to identify progenitor cells unambiguously and to obtain valid comparisons, since all markers showed cell-cycle dependent expression (see below). (Clones analyzed at 48 h or 72 h ALH gave comparable results; data not shown.)
As expected, the adaptor protein Miranda formed prominent cortical crescents in dividing NBs of non-DM clones (Figure 3c, asterisks). In the associated GMCs, Miranda was detected at weaker levels with uniform cortical distribution both at interphase and during mitosis (Figure 3c, inset, and Figure 3d, arrowheads). Strikingly, in DM lineages, 36% of the NB-associated cells showed strong and polarized expression of Miranda during mitosis, as described for dividing NBs (Figure 3h,h", arrows). The remaining dividing cells showed weak and uniform cortical localization of Miranda; their Miranda expression pattern was, thus, GMC-like (Figure 3h,h' arrowheads).
To confirm the presence of both NB-like and GMC-like progenitors in the DM NB lineages, we searched for markers of cellular identity that did not rely on the conventional criteria of cell size and/or cortical polarity. Significantly, we found that in non-DM lineages (taken as reference lineages), CycE was detected in virtually all the self-renewing NBs during mitosis (Figure 3e, asterisks; n = 74), but never during the terminal division of the GMCs (Figure 3f, arrowheads; n = 48). This distinctive criterion for cell identity was only applicable during mitosis because all progenitor cells expressed CycE at interphase, irrespective of their size (Figure 3e,f; PH3- nuclei; see also Figure 2c,d). In DM lineages, some of the small PH3-positive cells were negative for CycE but other small PH3-positive cells were positive for CycE (Figure 3i,i', arrow and arrowhead). Thus, in agreement with the data obtained using markers of cell polarity, both NB-like and GMC-like progenitors could be identified simultaneously in the progeny of a single DM NB (Figure 3g–i). Furthermore these two types of progenitors were observed specifically in these lineages and at all larval stages examined. Thus, the small CycE-positive/PH3-positive progenitors represented 55% (n = 64), 45% (n = 93) and 40% (n = 105) of the mitotic cells found in DM NB clones at 48 h ALH, 72 h ALH and 96 h ALH, respectively. The small CycE-positive/PH3-positive progenitors were never found associated with NBs of the ventral brain or the ventral ganglia at the corresponding stages (114 PH3-positive cells in 297 clones examined).
Taken together, these data indicate that the larval DM lineages contain two types of molecularly distinct progenitor cells other than NBs. Although not readily identifiable by their size, approximately two-thirds of these cells have molecular expression patterns of Prospero, Miranda and CycE that are characteristic of GMCs. In contrast, the remaining third have expression patterns of Prospero, Miranda and CycE that are remarkably similar to the patterns found in proliferative NBs. These novel NB-like progenitors are hereafter referred to as IPs. Our data further show that IPs are generated by DM NBs throughout larval neurogenesis in a quantitatively stable and balanced ratio with GMC-like progenitors and post-mitotic neurons.
Intermediate progenitor cells divide repeatedly and produce multicellular neuronal clones
Mitotic recombination was randomly induced in progenitor cells at 24 h and 48 h ALH and progenies were examined in isolated GFP-labeled clones 48 hours later (Figure 5b). As expected, single cell-, two cell-, and NB clones were recovered throughout the central nervous system. Prominent among the latter were the exceptionally large DM NB clones identifiable in the dorsal brain by their medial position and the spatial orientation of the labeled progeny that extend from the typical large cluster of late born Miranda-positive cells (Figure 5d,d'). Consistent with their linear growth rate (Figure 1d), we measured comparable clone sizes for DM NB clones generated during each of the two overlapping 48 hour windows (157 cells ± 33, n = 14 clones, and 220 cells ± 43, n = 16 clones, respectively). Likewise, non-DM NBs selected at random in the dorsal brain also generated comparable, albeit smaller, NB clones in the same time periods (63 cells ± 20, n = 40 clones, and 66 cells ± 23, n = 48 clones, respectively). Importantly, however, numerous clones lacking a NB and consisting of more than two cells were recovered in these experiments. These multicellular non-NB clones were found only in close spatial association with DM NBs and their progeny (Figure 5e,e'). Cell counts revealed a wide range of clone sizes in these lineages. Most clones, however, comprised 6–25 cells and this class was observed at comparable frequency in the two time windows examined (73% and 67%, respectively; Figure 5c). In over 90% of the cases examined, the cells in these multicellular clones expressed Elav, indicating that they were composed exclusively of post-mitotic neurons (Figure 5e,e').
The observed variability in clone size could be due to intrinsic variations in the mitotic capacity of different IPs and/or may result from mitotic recombination occurring in an IP that had already completed a variable number of divisions after its birth. Interestingly, the distribution of clonal cell number appeared remarkably similar when FLP/FRT recombination was induced at 24 h or at 48 h ALH (Figure 5c). This suggests that the mitotic potential of IPs is independent of their birth date from their parental DM NBs during larval development.
These findings imply that IPs in DM lineages can divide several times and produce differentiated progeny in less than 48 hours. Thus, they allow considerable amplification of the number of neurons produced in comparison to the standard mode of division adopted by other lineages in the central brain.
DM neuroblasts do not segregate Prospero protein to their daughter cells
The experiments described above show that DM NBs generate multiply dividing daughter cells that produce neural progeny. Surprisingly, these amplifying IP cells appear to be restricted to the DM lineages. What might explain this restriction? DM and non-DM NBs are not morphologically distinguishable and both divide asymmetrically to generate smaller progeny cells (Figure 3 and below).
Remarkably, and in contrast to all other Drosophila NBs described to date, Prospero was undetectable in the DM NBs during mitosis (Figure 6b,d). In all DM NB clones examined (n = 25), Miranda, but not Prospero, formed a cortical crescent in the dividing NB at metaphase (Figure 6b, asterisk) and segregated to the smaller daughter cell (Figure 6d). As a result, the IPs that derived directly from the DM NB lacked nuclear Prospero. GFP-labeled DM lineages typically contained 28 ± 9 Prospero-negative cells close to the NB (Figure 6b, white dots, n = 14 clones). These are likely to be accumulating IPs in interphase because they showed weak uniform expression of Miranda at the cortex and did not express PH3 (Figure 6b and data not shown). At IP mitosis, however, Prospero was unambiguously detected in these progenitors and showed co-localization with Miranda in a polarized manner (Figure 6b, arrows).
These data identify the DM NBs as a unique subset of neural stem cell-like progenitors that do not express and segregate Prospero during mitosis, thereby generating daughter cells that are molecularly distinct from GMCs.
Intermediate progenitor cell divisions are morphologically symmetrical but molecularly asymmetrical
The asymmetric division of DM NBs is also associated with the unequal segregation of Miranda to the smaller daughter cell (Figure 6b,d). Moreover, the resulting IP divides symmetrically to generate sibling cells of similar size as examined at telophase (n = 14; Figure 7b,d). Thus, in terms of the morphology of their cell divisions, IP cells are more like GMCs than like NBs. However, in sharp contrast to GMCs, mitotic IPs show cortical crescents of Miranda and Prospero (Figure 6b) and unequal partitioning of these two proteins at telophase (Figure 7d; n = 7). Thus, in terms of the segregation of cell fate determinants, dividing IP cells are remarkably more NB-like and differ substantially from GMCs.
Taken together, these findings demonstrate that the proliferative divisions of amplifying IPs in DM lineages have novel cellular and molecular features. These divisions are morphologically symmetrical and lead to two daughter cells of similar size, but molecularly asymmetrical in that the differentiating cell fate determinants Prospero and Miranda are segregated into only one cell. The ensuing absence of these differentiating cell fate determinants in the remaining daughter cell is likely to be a significant factor in the mitotic activity of amplifying IP cells.
The data presented here are consistent with a novel model for neurogenesis exemplified by the DM NBs, which divide asymmetrically in a stem cell mode to self-renew and generate IP daughter cells (Figure 8b). In this process, they do not segregate the cell fate determinant Prospero into the IP cells, which subsequently repeatedly divide symmetrically (in morphological aspects) yet asymmetrically segregate the cell fate determinants Prospero and Miranda during mitosis. The daughter cell that receives the Prospero and Miranda determinants is fated to become a differentiating GMC-like cell, whereas the other daughter cell retains its ability to divide several more times.
This novel model postulates that DM NBs produce exclusively IPs and not GMCs. The alternative notion, that the NB sometimes produces an IP and sometimes a GMC, is unlikely given that Prospero is never detected in the NB and, thus, cannot be segregated to one of its daughter cells as would be required for GMC generation. The model also posits that GMCs are produced by IPs through (functionally) asymmetrical divisions that result in one daughter cell becoming a GMC while the other daughter cell self-renews as an IP. Alternative scenarios, such as one in which IPs first divide symmetrically to expand in numbers and then adopt a GMC fate to generate differentiating neurons, are unlikely given the spatiotemporal pattern of Prospero/Miranda expression and the stable ratio of IPs versus GMCs observed in DM NB clones throughout larval development.
The experimental findings that support this novel model have implications for our understanding of neural stem cells and proliferation control. These are discussed in the following.
The NBs of the developing central brain and ventral ganglia divide asymmetrically in a stem cell mode in which the larger NB self renews and the smaller daughter cell differentiates into a different cell type, usually a GMC (reviewed by [7, 8, 10–12, 18]). This asymmetric division of the parent NB has been thought to be tightly coupled with the asymmetric segregation of cell fate determinants, and central among these molecular determinants is the transcription factor Prospero, which is required in GMCs to inhibit self-renewal and to promote differentiation [20–23, 27, 28]. Our findings indicate that the asymmetric segregation of Prospero does not occur in all dividing brain NBs. Indeed, in the DM NBs the lack of asymmetric segregation of Prospero to the IPs may be a key element in imparting (transient) NB-like features to these proliferating cells.
The GMCs of the developing nervous system divide symmetrically and generate two postmitotic progeny of equal size. Our findings indicate that IP cells also divide symmetrically in morphological terms, although Prospero and Miranda are partitioned to only one of their daughter cells. Thus, the morphologically symmetric cell division of a NB-derived daughter cell does not necessarily engender equal portioning of differentiation factors into both resulting cells. It has been assumed that only cells of a certain critical size show NB-like proliferative properties. The small size of the GMC would be a key factor promoting cell cycle exit and differentiation of its progeny (see ). This simple link between cell size and self renewing/terminal division is also called into question by our findings, since IPs are comparable in size to GMCs and yet they possess a very distinct mitotic potential.
The only repeatedly dividing progenitor cell type identified to date in the central nervous system of Drosophila is the NB. Our studies identify the IP cell as a second progenitor type with the capacity to undergo multiple rounds of divisions. This characteristic is coupled with several cellular and molecular features that are shared with NBs. Among these are the specific expression patterns of Prospero, Miranda and CycE during mitosis as well as the ability to asymmetrically segregate Prospero and Miranda during cell division. The number of divisions that IPs typically carry out is currently not known with precision. Our observations based on quantification of cell number in multicellular clones suggest an average of three-to five divisions as a conservative estimate. If, as assumed by our model, each IP cell division results in the generation of one GMC-like daughter cell, this estimate would predict a three- to five-fold amplification of the number of neuronal progeny in DM lineages compared with other lineages of the central brain and ventral ganglia. This prediction is in reasonable accordance with the amplified cell numbers observed in NB clones of DM versus non-DM lineages. The ultimate fate of the IPs is currently not known. The fact that almost all intermediate precursor-derived multicellular clones are composed exclusively of postmitotic neurons suggests that, after multiple divisions, these cells are either eliminated by programmed cell death or that they terminally divide and differentiate.
Although the DM NBs do not express and segregate Prospero to their daughter intermediate precursors, these daughter cells do express Prospero in a cortical and polarized manner during mitosis. The off/on state of Prospero must be kept under tight control for a controlled amplification of proliferation achieved in DM lineages since complete mutational loss of Prospero in brain clones leads to uncontrolled proliferative activity and brain tumor formation [31–33, 44]. Indeed, our observations on the DM lineages imply that deregulated IPs that fail to express Prospero might be an important source of tumor cells in the brain. Interestingly, region-specific action of tumor suppressor genes in the larval brain has been previously reported using somatic cell clones .
Here we identify a novel intermediate neural progenitor generated by asymmetric division of a subset of the Drosophila brain neuroblasts during the postembryonic phase of neurogenesis. Unlike conventional GMC this intermediate progenitor express molecular markers of self-renewing neuroblasts and undergoes multiple divisions in absence of Prospero. In the dorsomedial brain these progenitors amplify the number of neurons that can be generated by the parental neuroblasts at each round of their divisions. The novel IP described here bears remarkable similarities to the IPs that have been identified recently in mammalian brain development (reviewed by [1–3]). In the developing mammalian brain, primary neural stem cells persist in the ventricular zone through asymmetric self-renewing divisions, and IPs, which are thought to derive from these primary neural stem cells, divide symmetrically in the adjacent subventricular zone [45–47]. The division of IPs in the subventricular zone amplifies the number of cells produced by a given neural stem cell division and may be an important determinant of brain size, since species with larger brains have a larger pool of IPs . The surprising similarities in the patterns of neural stem and IP cell division in Drosophila and mammals suggest that amplification of brain neurogenesis in both groups of animals may rely on evolutionarily conserved cellular and molecular mechanisms.
In mammalian brain development, a two-step model has been proposed where, as a first step, asymmetric divisions of a primary neural stem cell generate a diverse set of molecularly different IPs and, as a second step, multiple symmetric divisions of each of these IPs generate large numbers of neurons of the same subtype . Given this model in mammalian brain development, it will be interesting to investigate if different IPs in a given DM lineage of the Drosophila brain are molecularly diverse, and if the neuronal subpopulations they generate during development acquire distinguishable anatomical and functional characters. If this is the case, the expansion of neurogenesis and the generation of multiple neuronal subclasses may be intimately related in the brain development of animals as diverse as insects and mammals.
Materials and methods
Fly stocks and MARCM analysis
All Drosophila stocks were reared and maintained on standard yeast-cornmeal-agar medium and all experiments were performed at 25°C. Unless otherwise stated, fly stocks carrying transgenes and recombinant chromosomes were obtained from the Bloomington stock center and assembled using standard genetics. To generate positively marked MARCM clones y, w, hsFLP; FRT40A, tubP-GAL80LL10/CyO, ActGFPJMR1; tubP-GAL4LL7, UAS-mCD8::GFPLL6/TM6, Tb, Hu were mated to either w; FRT40A, UAS-mCD8::GFPLL5 (standard cell lineage labeling with membrane-tethered GFP), or UASp-cnnGFP26.1; FRT40A, UAS-cLacZBg4-1-2 (for additional labeling of centrosomes), or w; FRT40A, UAS-cLacZBg4-1-2; UAS-tauGFP12/2/3 (gift of A Brand), for live imaging. Generation of MARCM clones and larval staging was performed as previously described  for this sub-section.
Immunohistochemistry and live imaging
Nervous systems were dissected from larvae, fixed and immunostained as previously described . Primary antibodies were as follows: rabbit anti-PH3 (1:400; Upstate, Charlottesville, Virginia, USA), mouse anti-MIRA (1:50; gift of P Overton), rabbit anti-MIRA (1:200; gift of YN Jan), mouse anti-PROS (1:10; Developmental Study Hybridoma Bank (DSHB), Iowa City, Iowa, USA); mouse anti-ELAV (1:30; DSHB) rat anti-ELAV (1:30; DSHB), mouse anti-CYCE (1:50; gift of H Richardson). Alexa Fluor-conjugated secondary antibodies (Molecular Probes, Invitrogen, Paisley, Renfrewshire, UK) were used at 1:200.
For live imaging, larval brains were dissected in Schneider's Drosophila Medium with 10% fetal bovine serum and mounted in 400-5 mineral oil (Sigma Diagnostic, Inc. St Louis, MO, USA) between a glass coverslip and a gas-permeable plastic foil (bioFOLIE 25, In Vitro System and Services, GmbH, Gottingen, Germany).
Microscopy and image processing
Fluorescently stained nervous systems were imaged using a Leica TCS SP scanning confocal microscope. Z stacks were collected with optical sections at 1–1.5 μm intervals. Pictures in this paper are presented as 'thick-section' merges projected as a flat image using ImageJ . Figures were assembled using Adobe Illustrator and Photoshop. Clone/lineage sizes were determined from confocal Z stacks of sections, spaced by 1 μm. Using ImageJ, cells were marked section-by-section and counted. Typically, 20–50 nervous systems per staining/genotypes/larval stages were examined using 63× oil-immersion objective. Only well isolated clones were recorded from the surface-located NB to the earliest born neurons close to the neuropil. Sample sizes, means and standard deviations for all histograms are indicated in the text and figure legends.
For time lapse, Z stacks made of 1 μm thick slices were collected at 4 minute intervals. Movies were processed and assembled using house made ImageJ plug-ins. Briefly, the sample motions were corrected in X and Y dimensions by manual reference point tracking. A single slice was arbitrarily selected per time point, allowing both some Z dimension drifting correction and the follow up of the most interesting cells within the sample.
We thank A Brand, P Overton, F Matzuzaki, YN Jan, H Richardson, the Bloomington stock center and the DSHB for strains and/or antibodies. We also thank all members of the Reichert lab for discussions and assistance. This work was supported by a grant from the Swiss NSF to HR.
- Gotz M, Huttner WB: The cell biology of neurogenesis. Nat Rev Mol Cell Biol 2005,6(10):777–788.View ArticlePubMed
- Merkle FT, Alvarez-Buylla A: Neural stem cells in mammalian development. Curr Opin Cell Biol 2006,18(6):704–709.View ArticlePubMed
- Kriegstein A, Noctor S, Martinez-Cerdeno V: Patterns of neural stem and progenitor cell division may underlie evolutionary cortical expansion. Nat Rev Neurosci 2006,7(11):883–890.View ArticlePubMed
- Morrison SJ, Kimble J: Asymmetric and symmetric stem-cell divisions in development and cancer. Nature 2006,441(7097):1068–1074.View ArticlePubMed
- Al-Hajj M, Clarke MF: Self-renewal and solid tumor stem cells. Oncogene 2004,23(43):7274–7282.View ArticlePubMed
- Oliver TG, Wechsler-Reya RJ: Getting at the root and stem of brain tumors. Neuron 2004,42(6):885–888.View ArticlePubMed
- Yu F, Kuo CT, Jan YN: Drosophila neuroblast asymmetric cell division: recent advances and implications for stem cell biology. Neuron 2006,51(1):13–20.View ArticlePubMed
- Egger B, Chell JM, Brand AH: Insights into neural stem cell biology from flies. Philos Trans R Soc Lond B Biol Sci 2008,363(1489):39–59.View ArticlePubMed
- Karcavich RE: Generating neuronal diversity in the Drosophila central nervous system: a view from the ganglion mother cells. Dev Dyn 2005,232(3):609–616.View ArticlePubMed
- Skeath JB, Thor S: Genetic control of Drosophila nerve cord development. Curr Opin Neurobiol 2003,13(1):8–15.View ArticlePubMed
- Pearson BJ, Doe CQ: Specification of temporal identity in the developing nervous system. Annu Rev Cell Dev Biol 2004, 20:619–647.View ArticlePubMed
- Technau GM, Berger C, Urbach R: Generation of cell diversity and segmental pattern in the embryonic central nervous system of Drosophila. Dev Dyn 2006,235(4):861–869.View ArticlePubMed
- Truman JW, Bate M: Spatial and temporal patterns of neurogenesis in the central nervous system of Drosophila melanogaster. Dev Biol 1988,125(1):145–157.View ArticlePubMed
- Prokop A, Technau GM: The origin of postembryonic neuroblasts in the ventral nerve cord of Drosophila melanogaster. Development 1991,111(1):79–88.PubMed
- Truman JW, Schuppe H, Shepherd D, Williams DW: Developmental architecture of adult-specific lineages in the ventral CNS of Drosophila. Development 2004,131(20):5167–5184.View ArticlePubMed
- Pereanu W, Hartenstein V: Neural lineages of the Drosophila brain: a three-dimensional digital atlas of the pattern of lineage location and projection at the late larval stage. J Neurosci 2006,26(20):5534–5553.View ArticlePubMed
- Bardin AJ, Le Borgne R, Schweisguth F: Asymmetric localization and function of cell-fate determinants: a fly's view. Curr Opin Neurobiol 2004,14(1):6–14.View ArticlePubMed
- Betschinger J, Knoblich JA: Dare to be different: asymmetric cell division in Drosophila, C. elegans and vertebrates. Curr Biol 2004,14(16):R674–85.View ArticlePubMed
- Wang H, Chia W: Drosophila neural progenitor polarity and asymmetric division. Biol Cell 2005,97(1):63–74.View ArticlePubMed
- Chu-Lagraff Q, Wright DM, McNeil LK, Doe CQ: The prospero gene encodes a divergent homeodomain protein that controls neuronal identity in Drosophila. Development 1991, Suppl 2:79–85.PubMed
- Vaessin H, Grell E, Wolff E, Bier E, Jan LY, Jan YN: prospero is expressed in neuronal precursors and encodes a nuclear protein that is involved in the control of axonal outgrowth in Drosophila. Cell 1991,67(5):941–953.View ArticlePubMed
- Li L, Vaessin H: Pan-neural Prospero terminates cell proliferation during Drosophila neurogenesis. Genes Dev 2000,14(2):147–151.PubMed
- Choksi SP, Southall TD, Bossing T, Edoff K, de Wit E, Fischer BE, van Steensel B, Micklem G, Brand AH: Prospero acts as a binary switch between self-renewal and differentiation in Drosophila neural stem cells. Dev Cell 2006,11(6):775–789.View ArticlePubMed
- Ikeshima-Kataoka H, Skeath JB, Nabeshima Y, Doe CQ, Matsuzaki F: Miranda directs Prospero to a daughter cell during Drosophila asymmetric divisions. Nature 1997,390(6660):625–629.View ArticlePubMed
- Shen CP, Jan LY, Jan YN: Miranda is required for the asymmetric localization of Prospero during mitosis in Drosophila. Cell 1997,90(3):449–458.View ArticlePubMed
- Schuldt AJ, Adams JH, Davidson CM, Micklem DR, Haseloff J, St Johnston D, Brand AH: Miranda mediates asymmetric protein and RNA localization in the developing nervous system. Genes Dev 1998,12(12):1847–1857.View ArticlePubMed
- Knoblich JA, Jan LY, Jan YN: Asymmetric segregation of Numb and Prospero during cell division. Nature 1995,377(6550):624–627.View ArticlePubMed
- Spana EP, Doe CQ: The prospero transcription factor is asymmetrically localized to the cell cortex during neuroblast mitosis in Drosophila. Development 1995,121(10):3187–3195.PubMed
- Akong K, McCartney BM, Peifer M: Drosophila APC2 and APC1 have overlapping roles in the larval brain despite their distinct intracellular localizations. Dev Biol 2002,250(1):71–90.View ArticlePubMed
- Ceron J, Gonzalez C, Tejedor FJ: Patterns of cell division and expression of asymmetric cell fate determinants in postembryonic neuroblast lineages of Drosophila. Dev Biol 2001,230(2):125–138.View ArticlePubMed
- Bello B, Reichert H, Hirth F: The brain tumor gene negatively regulates neural progenitor cell proliferation in the larval central brain of Drosophila. Development 2006,133(14):2639–2648.View ArticlePubMed
- Betschinger J, Mechtler K, Knoblich JA: Asymmetric segregation of the tumor suppressor brat regulates self-renewal in Drosophila neural stem cells. Cell 2006,124(6):1241–1253.View ArticlePubMed
- Lee CY, Wilkinson BD, Siegrist SE, Wharton RP, Doe CQ: Brat is a Miranda cargo protein that promotes neuronal differentiation and inhibits neuroblast self-renewal. Dev Cell 2006,10(4):441–449.View ArticlePubMed
- Maurange C, Gould AP: Brainy but not too brainy: starting and stopping neuroblast divisions in Drosophila. Trends Neurosci 2005,28(1):30–36.View ArticlePubMed
- Strausfeld: Atlas of an insect brain. 1976.
- Younossi-Hartenstein A, Nassif C, Green P, Hartenstein V: Early neurogenesis of the Drosophila brain. J Comp Neurol 1996,370(3):313–329.View ArticlePubMed
- Urbach R, Schnabel R, Technau GM: The pattern of neuroblast formation, mitotic domains and proneural gene expression during early brain development in Drosophila. Development 2003,130(16):3589–3606.View ArticlePubMed
- Ito K, Hotta Y: Proliferation pattern of postembryonic neuroblasts in the brain of Drosophila melanogaster. Dev Biol 1992,149(1):134–148.View ArticlePubMed
- Lee T, Luo L: Mosaic analysis with a repressible cell marker (MARCM) for Drosophila neural development. Trends Neurosci 2001,24(5):251–254.View ArticlePubMed
- Bello B, Holbro N, Reichert H: Polycomb group genes are required for neural stem cell survival in postembryonic neurogenesis of Drosophila. Development 2007,134(6):1091–1099.View ArticlePubMed
- Wu JS, Luo L: A protocol for mosaic analysis with a repressible cell marker (MARCM) in Drosophila. Nat Protoc 2006,1(6):2583–2589.View ArticlePubMed
- Lee CY, Robinson KJ, Doe CQ: Lgl, Pins and aPKC regulate neuroblast self-renewal versus differentiation. Nature 2006,439(7076):594–598.View ArticlePubMed
- Rusan NM, Peifer M: A role for a novel centrosome cycle in asymmetric cell division. J Cell Biol 2007,177(1):13–20.View ArticlePubMed
- Caussinus E, Gonzalez C: Induction of tumor growth by altered stem-cell asymmetric division in Drosophila melanogaster. Nat Genet 2005,37(10):1125–1129.View ArticlePubMed
- Haubensak W, Attardo A, Denk W, Huttner WB: Neurons arise in the basal neuroepithelium of the early mammalian telencephalon: a major site of neurogenesis. Proc Natl Acad Sci U S A 2004,101(9):3196–3201.View ArticlePubMed
- Noctor SC, Martinez-Cerdeno V, Ivic L, Kriegstein AR: Cortical neurons arise in symmetric and asymmetric division zones and migrate through specific phases. Nat Neurosci 2004,7(2):136–144.View ArticlePubMed
- Miyata T, Kawaguchi A, Saito K, Kawano M, Muto T, Ogawa M: Asymmetric production of surface-dividing and non-surface-dividing cortical progenitor cells. Development 2004,131(13):3133–3145.View ArticlePubMed
- Martinez-Cerdeno V, Noctor SC, Kriegstein AR: The role of intermediate progenitor cells in the evolutionary expansion of the cerebral cortex. Cereb Cortex 2006, 16 Suppl 1:i152–61.View ArticlePubMed
- ImageJ: A public domain Java image processing program. [http://rsb.info.nih.gov/ij]
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.