Coordinated control of Notch/Delta signalling and cell cycle progression drives lateral inhibition-mediated tissue patterning

Abstract

Coordinating cell differentiation with cell growth and division is crucial for the successful development, homeostasis and regeneration of multicellular tissues. Here, we use bristle patterning in the fly notum as a model system to explore the regulatory and functional coupling of cell cycle progression and cell fate decision-making. The pattern of bristles and intervening epithelial cells (ECs) becomes established through Notch-mediated lateral inhibition during G2 phase of the cell cycle, as neighbouring cells physically interact with each other via lateral contacts and/or basal protrusions. Since Notch signalling controls cell division timing downstream of Cdc25, ECs in lateral contact with a Delta-expressing cell experience higher levels of Notch signalling and divide first, followed by more distant neighbours, and lastly Delta-expressing cells. Conversely, mitotic entry and cell division makes ECs refractory to lateral inhibition signalling, fixing their fate. Using a combination of experiments and computational modelling, we show that this reciprocal relationship between Notch signalling and cell cycle progression acts like a developmental clock, providing a delimited window of time during which cells decide their fate, ensuring efficient and orderly bristle patterning.

Keywords: Cell cycle; G2 phase; Lateral inhibition; Notch signalling; Patterning.

Fig. 1.

Spatiotemporal patterning of notum cell divisions. (A) Pupal notum expressing Shotgun^GFP (cell boundaries), and nGMCA (SOPs) over time. Posterior to the left, anterior to the right. Scale bar: 25 µm. (B) SOP ‘neighbourhood’: SOP (pink) with primary (1N; blue) and secondary (2N; orange) neighbours. Scale bar: 10 µm. (C) Time of cell division for the genotype shown in A; N=2 pupae. (D) Line graph of the data shown in C. (E) Mean ratio of local SOP neighbourhood division timing, genotype as in A. N=2 nota; n=20 SOPs, 109 1Ns, 127 2Ns. (F) Division timing of SOPs, ‘switch’ cells (neu-GMCA-expressing cells that switch to EC fate) and ECs and their respective 1Ns in shotgun^GFP; neu-GMCA pupae (N=3). ***P<0.001 (unpaired two-tailed t-test for pairs indicated). Mean±s.d. shown. (n)=number of cells.

Fig. 2.

Cell division timing depends on Notch signalling. (A) N^sfGFP expression pattern at 12 h AP. H2B^mRFP labels nuclei. Scale bar: 50 µm. (A′) Higher magnification image of A. Scale bar: 5 µm. (B) False-coloured panel of N^sfGFP -expressing ECs. Asterisk indicates SOP. Primary (1) and secondary (2) neighbours are indicated by dashed boxes. Scale bar: 5 µm. (B′) Time series of nuclear ROIs for cells 1 and 2 until nuclear envelope breakdown (NEBD; indicated by black boxes), leading to transient depletion of signal. (C) N^sfGFP dynamics in ECs (n=29 each, N=3). (D) Rate of N^sfGFP increase for the data shown in C. (E) Maximum normalized N^sfGFP signal for the data shown in C. (F) Mean ratio of local SOP neighbourhood N^sfGFP signal (n=27 SOP, 133 each EC type; N=3). (G-I) neur-GAL4 expression of Delta^DN reduces Notch signalling in wild-type 1N (G) or 2N (H) cells (n=16, N=2) versus control (UAS-lifeActRuby, n=30, N=3) and delays cell division timing in Shotgun^GFP; neu-GAL4, UAS-GMCA>UAS-Delta^DN pupae (I) (N=3). (J) Cell division timing in Shotgun^GFP; pnrGAL4>UAS-Su(H) RNAi pupae relative to control (N=2). ECs, epithelial cells in regions lacking differentiating SOPs. Mean±s.e.m. for C,F,G,H; mean±s.d. for D,E,I,J. n.s., not significant; **P<0.01; ***P<0.0001 by unpaired, two-tailed, t-test as indicated compared with control of the same type (i.e. RNAi-1N to control-1N). (n)=number of cells.

Fig. 3.

Regulation of notum division timing. (A) Cell division timing in Shotgun^GFP; pnrGAL4>UAS-string RNAi pupae (N=3). n.s., not significant by one-way ANOVA. (A′) Percentage of dividing cells in the same genotype as A. (B,C) Cell division timing in Shotgun^GFP; pnrGAL4>UAS-Wee1 RNAi (B) (N=3) or UAS-Myt1 RNAi (C) pupae (N=3). (D,E) N^sfGFP dynamics in primary (D) and secondary (E) neighbour ECs expressing UAS-stg RNAi (red; n=20, N=2), UAS-Wee1 RNAi (blue; n=20, N=2), or control (UAS-lifeActRuby, black; n=30, N=3) under pnr-GAL4. Vertical dashed lines indicate mean cell division timing for cell position and genotype. (F) Time series of nuclear ROIs for string RNAi-expressing cells 1 and 2 (indicated by dashed boxes), showing failure to downregulate signal. NEBD does not occur. Asterisk indicates SOP. **P≤0.01; ***P≤0.001 by unpaired, two-tailed, t-test to control of the same type. Mean±s.d. for A-C; mean±s.e.m. for D,E. (n)=number of cells.

Fig. 4.

Cell division timing is crucial for SOP patterning. (A) Model output for ‘wild type’ simulation (K_R=200, q=5). Average spacing is the mean±s.e.m. distance between each SOP (red) and its ten nearest SOPs. (B) Simulation results for cell division timing in 1N and 2N for the wild-type model described in A. (C) Ratio of mean time of division for 1N and 2N in the model. (D) Model output when Notch signalling and division timing are uncoupled, p_d=0.005 (any non-Delta cell [D<1] divides with probability p_d). (E) Model output when SOPs are forced to divide early (Delta cells [D>1] divide with probability p_d=0.0001). Red, Delta expressing SOPs (D>1); grey, Notch-expressing ECs. (F) Final SOP pattern in tissues with precocious SOP division. Scale bar: 50 µm. (G) Cell division timing in Shotgun^GFP; neu-GAL4, UAS-GMCA>UAS-string pupae (N=3, mean±s.d.). Control: Shotgun^GFP; neu-GAL4, UAS-GMCA (N=3). (H,I) SOP ‘twins’ and secondary neighbour cell switching (asterisks in I), as a consequence of the precocious SOP division shown in F. Yellow, divided cells. Scale bars: 10 µm. **P≤0.01, ***P≤0.001, unpaired, two-tailed, t-test to control of same type. (n)=number of cells.