There are many ways of defining a cell type. A cell can be unique because of its function, morphology, or molecular characteristics. By any of these measures, cell type diversity in the nervous system is immense. Technological advances in our ability to examine neuroanatomy and molecular characteristics of single cells has only emphasised this point. So, this is an era of a new scale of challenges in developmental neurobiology, but also of unprecedented technological innovations to understand how the enormous diversity and complexity of nervous systems is encoded by genomes.

Studies in Drosophila have led the way in this field. Principles uncovered in flies are bearing out in other organisms – those with more complex as well as ’simpler’ nervous systems. Here, I will cover one such principle: how neural stem cells (NSC) experience two axes of information – spatial and temporal – to generate diversity.

The developmental output of spatial patterning is a progenitor that can only generate a specific lineage of neurons or glia. It is typically initiated on the neuroepithelium, which, through the regionalised expression of various genes, gives rise to heterogeneous pools of progenitors. These, in turn, can generate distinct neuronal fates. Thus, spatial patterning generates inter-lineage diversity. On the other hand, the developmental output of temporal pattering is a progenitor that can generate different neurons or glia as it ages. This occurs through the expression of a series of genes expressed in a sequence within the NSC, allowing it to generate different identities over time. Temporal patterning, therefore, results in intra-lineage diversity. The two axes acting together in an NSC allows it to generate lineage-appropriate, time-specific neurons, and such inter- and intra-lineage diversity eventually contributes to the generation of cell diversity in nervous systems.

Essential to this is the specification of the NSCs. It creates a cell that not only has the potential to generate numerous progeny, but one that can also integrate molecular information and therefore influence the identity of these progeny. In flies, for example, two post-mitotic neurons are generated at each NSC division (via an intermediate precursor called a GMC) with each successive division generating different neural identities. The status of Notch signalling – whether on or off – in these siblings determines alternate fates within them. This can include anatomical as well as neurotransmitter fates [1], [2]. Immature neurons generated by the NSC finally utilise ’terminal selector transcription factors’ to determine their terminal identities [3], [4]. For example, differential Notch signalling specifies projection neuron (PN) versus local interneuron (LN) fate in a Drosophila antennal lobe NSC. This NSC also generates different types of LNs and PNs over time [5], [6]. Similarly in the mouse spinal cord, progenitor domains with molecularly distinct spatial identities generate heterogeneous pools of differentiated progeny. Notch signalling operates within these pools to generate different fates [7]. For example, in the motoneuron progenitor pool (pMN), Notch signalling specifies motoneurons of the median motor column over hypaxial motoneuron identities [8]. The final molecular and functional identities of the neurons that derive from a single progenitor domain depends on the combinatorial expression of various transcription factors of the bHLH and homeodomain family [7], [9], [10]. Importantly, because the spatial identity of the NSC determines the temporal, hemilineage, and terminal outputs of an NSC these identities are nested fates within the overarching spatial identity of a progenitor.

In this chapter, therefore, I will highlight evidence for spatial and temporal patterning in the Drosophila brain and optic lobe, as well as in the vertebrate cortex and spinal cord. I will quote examples of our current understanding of the molecular underpinnings of both these axes and how they might intersect with each other. And finally, I will highlight what remains to be understood and how emerging technologies might help address these lacunae.