Encoding and decoding time in neural development

Abstract

The development of a multicellular organism involves time-dependent changes in molecular and cellular states; therefore 'time' is an indispensable mathematical parameter of ontogenesis. Regardless of their inextricable relationship, there is a limited number of events for which the output of developmental phenomena primarily uses temporal cues that are generated through multilevel interactions between molecules, cells, and tissues. In this review, we focus on neural stem cells, which serve as a faithful decoder of temporal cues to transmit biological information and generate specific output in the developing nervous system. We further explore the identity of the temporal information that is encoded in neural development, and how this information is decoded into various cellular fate decisions.

Keywords: cell cycle; neural stem cell; temporal code; timing; transcriptional network.

Figure 1

Temporal parameters used in development. The entire developmental time scale period can be plotted onto a physical time scale. Along this axis, the timer (bold brown arrow) determines the timing (thin brown arrow) of a specific developmental event (i.e. event ‘e’) relative to a reference event (i.e. event ‘a’) along the directional temporal axis. Temporal cues are also produced through periodic events such as oscillations (light blue). In this case, the amplitude, phase, period or frequency (blue) may serve as temporal codes.

Figure 2

Temporal cell types in various neural systems. (A) In Drosophila N7‐1 lineage, neuroblasts expresses distinct transcription factors, Hb, Kr, Pdm and Cas to regulate U1‐U5 neuron specification and temporal identity transitions. (B) In retina, specification of subtypes is encoded in part through expression of transcription factors. Photoreceptor cells (PR) compose of cone and rod cells, in which cone cells are generated at an earlier time window than rod cells. (C) In the cerebral cortex, neural stem (progenitor) cells produce Cajal‐Retzius (CR) cells, deep‐layer (DL) neurons and upper‐layer (UL) neurons sequentially. Foxg1 has been shown as a temporal identity factor that switches neural stem cell competence from CR cells to DL neurons. AC, amacrine cells; BC, bipolar cells; HC, horizontal cells; Mu, muller glia cells; PR, photoreceptor cells; RGC; retinal ganglion cells.

Figure 3

Comparison of the timing of neurogenesis between in vivo and in vitro models. In vivo and in vitro neurogenesis of mouse (top panels) and human (bottom panels) cortical stem cells reveal similar temporal scale in neuronal subtype production. In vitro time scale is calculated based on the days (mouse) or weeks (human) of culture. For human in vitro data, only limited time points have been observed and the start and end period of each subtype neurogenesis has not been determined. CR cell (green), deep‐layer neuron (blue), upper‐layer neuron (red).

Figure 4

Correlation between cell cycle number and neuronal output in the cerebral cortex. (A) Based on BrdU and thymidine double‐label birthdating studies in mice (Takahashi et al.1999), cortical progenitor cells undergo 11 cell cycles after embryonic day (E)11, which sequentially produce CR cells, deep‐layer neurons and upper‐layer neurons. (B) MADM clonal analysis (Gao et al. 2014) reveals that cortical progenitor cells undergo an average of 2.6 symmetric cell divisions followed by 8–9 cell divisions, resulting in a similar number of total cell cycles as in study (A). (C) Manipulation of cell cycle through overexpression of Cdk4 and CyclinD (4D) results in 10% reduction in total cell cycle length. If neural progenitor cells use number of cell cycles as timer, then the prediction is that each temporal transitions are accelerated; however, the number of each cell types does not change (a). If temporal identity transitions progress independently of cell cycle (cell cycle independent timer) then the number of each cell type is expected to increase (b). CR cell, Deep‐layer neuron, Upper‐layer neuron.

Figure 5

Temporal codes utilized in neural development. (A) Encoding and decoding of temporal codes in neural development. Interplay between molecules, cells and tissues generate temporal signals that are decoded by cells or tissues to generate specific outputs. (B) Examples of temporal codes that are generated through negative feedback signaling. Protein‐to‐gene and ion‐to‐transporter feedback generates periodic signals (oscillation), neuron‐to‐progenitor cell feedback triggers switch in cell types. (C) Distinct parameters of periodic signals can be used as temporal codes, such as number of cycles or oscillation frequency. (D) Linear temporal codes. Hierarchical cascade of transcription factors triggers sequential gene expression and transition in cell states. Cumulative signals can act as threshold‐driven timer to trigger temporal events. (E) Various oscillation encoded temporal cues in neural development. Oscillations are observed at ion, gene and cell levels, which are decoded into distinct cellular outputs.