Left panel: graphical representation of a differentiation landscape with three sequential differentiation steps. Green circles: small diffusible molecule 3O-C6-HSL. Purple diamonds: small diffusible molecule 3O-C14-HSL. Cell colour reflects the active molecular program: undifferentiated cells are gray, sender cells are green, receiver cells are blue, receivers exposed to 3O-C6-HSL become red, receivers exposed to 3O-C14-HSL become yellow (while still maintaining their underlying blue identity). Right panel: representative microscopy image of E. coli colonies engineered with the 3-step sequential differentiation program, coloured to highlight the three steps. Channels: I - brightfield; II - brightfield, green, blue ; III - brightfield, green, red ; IV - brightfield, green, yellow.

Toggle switch-based 1-step differentiation system

a) Genetic circuit of the 1-step differentiation system, including interactions and chemical inducers (LuxI in the GFP-LuxI fusion protein is not necessary for the 1st step, but was included, as it will be important for the 2nd step). Coloured boxes: genes; L-shaped arrows: promoters; T-shaped lines: repressions. b) Microscopy pictures of colonies harbouring the toggle switch. Cells in the initial state mixed were pre-cultured in presence or absence of inducers (indicated on the left side of the picture), then plated and grown overnight in absence of inducers. Channels: brightfield, green, blue. c) Green and blue population percentages in liquid cultures conditioned with different inducer concentrations. Starting from 3 different initial states (green, mixed, blue) cells were pre-cultured in liquid, then the population ratio was quantified via flow cytometry. For the initial state green and blue, cells were taken from colonies that were homogeneously green or blue, while for the initial state mixed, cells were taken from a colony obtained upon transformation of the circuit. Dots represent experimental points (3 biological replicates per condition), lines represent fitted curves (see Methods). d) Representative 3D energy landscape corresponding to an almost symmetrical condition (initial state: mixed, induced with 0.003 mM IPTG). Valleys represent the green and blue stable states, red line represents the separatrix. The depth of the valleys corresponds to the density of events in the flow cytometry data from a single replicate in c. e) Flow cytometry density plots of 5 representative conditions from c (initial state: mixed, inducer concentration indicated below each plot). Black lines represent the projection of the plane crossing the local minima of the green and blue populations. Axes (GFP and mCerulean fluorescence) in the leftmost plot apply to all plots. Inserts represent the 1D energy surface corresponding to each plot, with local minima corresponding to stable states, or valleys. Each 1D plot is the cross section from the corresponding 3D energy landscape (as in d) with the plane crossing the local minima of the green and blue populations, for a single replicate. The colour gradient reflects the inclination of each condition towards the green or the blue state.

Implementation of a 2-step sequential differentiation system.

a) Schematic representation of the 2-step differentiation system. The toggle switch bifurcates the system into two mutually exclusive stable states, receiver (R, blue) and sender (S, green), that are coupled to the two QS components LuxR and LuxI respectively. b) Illustration of the 2-step system functioning: an undifferentiated cell (gray) can become either a green sender or a blue receiver. Upon detection of the 3O-C6-HSL produced by senders, receivers can activate production of the red reporter mCherry. c) Genetic circuitry of the 2-step differentiation system, including interactions and chemical inducers. Coloured boxes: genes; L-shaped arrows: promoters; T-shaped lines: repressions; dashdot arrows: activations; green dots: small diffusible molecule 3O-C6-HSL.

Characterization of the sender and receiver states.

a) Top: schematic representation of the plate assay: cells containing only the C6 device were spread homogenously in the plate in order to have well separated reporter colonies. Either pure 3O-C6-HSL or sender cells (pre-cultured with 100 nM aTc) were spotted in the centre of the plate. After incubation, we took images at increasing distance from the centre of the plate. Bottom: microscopy images of reporter colonies growing on plates in absence of 3O-C6-HSL (I° row), in presence of pure 3O-C6-HSL (2 µl, 50 µM, II° row) or of sender cells (III° row, the senders correspond to the green colony). Channels: brightfield, red, green. b) Quantification of red fluorescence intensity from the colonies in a, as a function of the distance from the plate centre. Hexagons, circles and triangles represent experimental points; black lines represent exponential curves fitted to experimental points.

The 2-step sequential differentiation system forms a simple self-organized spatial pattern.

a) Microscopy images of the pattern generated by the 2-step system in the immediate surroundings of green sender colonies. Channels: brightfield, red, green, blue. b) Quantification of the red fluorescence intensity of receiver colonies as a function of the distance from the closest sender colony. Each dot represents the average fluorescence intensity of a colony, n = 126 colonies across 14 different images, dots coming from the same image are filled with the same colour. Red line represents an exponential curve fitted to experimental points. c) Mathematical simulation of the spatial pattern at 24 h. Colonies are colour-coded, from left to right, according to: TS state (green-senders, blue-receivers), mCherry fluorescence intensity, full pattern. d) Quantification of the red fluorescence intensity of receiver colonies as a function of the distance from the closest sender colony. Dots come from 10 mathematical simulations; black line represents an exponential curve fitted to simulation points.

The 2-step sequential differentiation system can generate a variety of spatial patterns

a) Quantification of the three populations abundances from mathematical simulations of the system, in function of the initial blue:green ratio. Dots represent simulation points (n=3 simulations for each initial condition), lines connect averages. b) Quantification of the three populations abundances (green-senders, blue-receivers, red-receivers) in microscopy images. Cells harbouring the 2-step system were pre-cultured with 11 different inducer concentrations (indicated on the x axis). Boxplots of 3 biological replicates (3 images per replicate) represent 1st, 2nd, 3rd quartile of the distribution, whiskers extend to the rest of the distribution or to 1.5·IQR (interquartile range), dots represent individual data points (including outliers). c) Representative microscopy images of the spatial patterns generated by cells harbouring the 2-step system, pre-cultured with different inducer concentrations (indicated at the bottom of each column). Rows (from top to bottom): GFP channel, CFP channel, mCherry channel, composite image. d) Mathematical simulation of spatial patterns at 24 h, starting from different blue:green ratios (indicated below each simulation). Colonies are colour-coded according to their simulated green, blue and red fluorescence intensity. Initial blue:green ratios were chosen to match experimental blue:green ratios in c.

Implementation of a 3-step sequential differentiation system.

a) Genetic circuitry of the 3-step differentiation system, including interactions and chemical inducers. Coloured boxes: genes; L-shaped arrows: promoters; T-shaped lines: repressions; dashed and dashdot arrows: activations; green dots and purple diamonds: small diffusible molecules 3O-C6-HSL and 3O-C14-HSL, respectively. b) Schematic representation of the 3-step system functioning: a green sender cell produces 3O-C6-HSL via LuxI, which stimulates nearby blue receivers to turn red and to activate production of 3O-C14-HSL via CinI. When the level of 3O-C14-HSL is above threshold, receivers can also turn on production of the yellow reporter mCitrine. c) Demonstration of the 3 sequential steps. The left panel shows the active (black) and inactive (gray) steps of the differentiation cascade. The right panel shows microscopy images of receiver colonies (small, in blue, on the right) grown in presence or absence of sender colonies (big, in green, on the left), with complete or incomplete circuitry, in accordance with the left panel. I° row: receivers only. II° row: senders and receivers, the LuxI synthase is missing from the circuit. III° row: senders and receivers, the CinI synthase is missing from the circuit. IV° row: senders and receivers, complete circuity. Channels: brightfield, red, yellow, green, blue. d) Representative microscopy image of the 3-step sequential differentiation system. Channels: I - brightfield, green, blue ; II - brightfield, red, green ; III - brightfield, yellow, green ; IV - brightfield, red, yellow, green, blue. e) Quantification of red (left panel) and yellow (right panel) fluorescence intensity as a function of the distance from a green sender colony. Each dot represents the average fluorescence intensity of a colony, n = 304 colonies across 15 different images, dots coming from the same image are filled with the same colour. Red and yellow lines represent exponential curves fitted to experimental points.

Fitted parameters describing the toggle-switch response in Figure 2c under different initial induction states (blue, mixed, and green).

The linear IPTG–aTc conversion in Eq. 1 used a common parameter of α = 20 nM/mM.