Kriegman S et al. (2021), Kinematic self-replication in reconfigurable or...

XB-ART-58837

Proc Natl Acad Sci U S A 2021 Dec 07;11849:. doi: 10.1073/pnas.2112672118.

Show Gene links Show Anatomy links

Kinematic self-replication in reconfigurable organisms.

Kriegman S, Blackiston D, Levin M, Bongard J.

???displayArticle.abstract???
All living systems perpetuate themselves via growth in or on the body, followed by splitting, budding, or birth. We find that synthetic multicellular assemblies can also replicate kinematically by moving and compressing dissociated cells in their environment into functional self-copies. This form of perpetuation, previously unseen in any organism, arises spontaneously over days rather than evolving over millennia. We also show how artificial intelligence methods can design assemblies that postpone loss of replicative ability and perform useful work as a side effect of replication. This suggests other unique and useful phenotypes can be rapidly reached from wild-type organisms without selection or genetic engineering, thereby broadening our understanding of the conditions under which replication arises, phenotypic plasticity, and how useful replicative machines may be realized.

???displayArticle.pubmedLink??? 34845026
???displayArticle.pmcLink??? PMC8670470
???displayArticle.link??? Proc Natl Acad Sci U S A

Species referenced: Xenopus laevis
Genes referenced: got2 grap2 mrc1 psmd6
GO keywords: reproduction [+]

???attribute.lit??? ???displayArticles.show???

	Fig. 1. Spontaneous kinematic self-replication. (A) Stem cells are removed from early-stage frog blastula, dissociated, and placed in a saline solution, where they cohere into spheres containing ∼3,000 cells. The spheres develop cilia on their outer surfaces after 3 d. When the resulting mature swarm is placed amid ∼60,000 dissociated stem cells in a 60-mm-diameter circular dish (B), their collective motion pushes some cells together into piles (C and D), which, if sufficiently large (at least 50 cells), develop into ciliated offspring (E) themselves capable of swimming, and, if provided additional dissociated stem cells (F), build additional offspring. In short, progenitors (p) build offspring (o), which then become progenitors. This process can be disrupted by withholding additional dissociated cells. Under these, the currently best known environmental conditions, the system naturally self-replicates for a maximum of two rounds before halting. The probability of halting (α) or replicating( 1 − α) depends on a temperature range suitable for frog embryos, the concentration of dissociated cells, the number and stochastic behavior of the mature organisms, the viscosity of the solution, the geometry of the dish’s surface, and the possibility of contamination. (Scale bars, 500 μm.)
	Fig. 2. Amplifying kinematic self-replication. Due to surface tension, reconfigurable organisms naturally develop into ciliated spheroids, but they can be sculpted into nonspheroidal morphologies manually during development to realize more complex body shapes. Progenitor shapes were evolved in silico to maximize the number of self-replication rounds before halting. (A) Shapes often converge to an asymmetrical semitoroid (C-shape; pink) with a single narrow mouth in which dissociated cells (green) can be captured, transported, and aggregated. This evolved shape was fabricated and released in vivo (B), recapitulating the behavior observed in silico (A). Offspring built by wild-type spheroids (C) were smaller than those built by the semitoroids (D), regardless of the size and aspect ratios of the spheroids, and across different concentrations of dissociated cells (E). The maximum of two rounds of self-replication achieved by the spheroids (F) was extended by the semitoroids to a maximum of four rounds (G). (Scale bars, 500 μm.)
	Fig. 3. Evolving self-replication. (A) An evolutionary algorithm, starting with random swarms, evolves swarms with increasing self-replicative ability. (FG = number of filial generations achieved by a given swarm. The fractional part denotes how close the swarm got to achieving another replication round.) The most successful lineage in this evolutionary trial originated from a spheroid that built piles no larger than 74% of the size threshold required to self-replicate (B). A descendent swarm composed of nine flexible tori (C) contained two members that built one pile large enough to self-replicate (two arrows), which, alone, built piles no larger than 51% of the threshold. A descendent of the toroid swarm, a swarm of semitori (D), contained six members (E) that collectively built three piles large enough to mature into offspring (F). One of those offspring built a pile large enough to mature into a second generation offspring (G). An additional 48 independent evolutionary trials (H) evolved self-replicative swarms with diverse progenitor shapes.
	Fig. 4. Forecasting utility. (A) A swarm of self-replicating semitoroidal organisms (gray) was placed inside a partially completed circuit (black) containing two power sources (red dots), four light emitters (circled X; black when OFF, red when ON), and disconnected flexible adhesive wires (black lines). Dissociated stem cells (not pictured), if pushed into piles, develop into offspring (irregularly shaped gray masses). Dissociated cells are replaced every 3.5 s. After 17.5 s of self-replication and circuit building within a single dish, the progenitors are discarded, and all first through fourth filial generation offspring are divided into two equal-sized groups and placed into two new dishes, each containing a partially completed circuit (B and C). If only one offspring is built, one dish is seeded with it. If no offspring are built, bifurcation halts. This process results in an unbalanced binary tree (D). The red edges denote circuits in which at least one light emitter was switched on by closing a circuit from power source to light emitter (OFF/ON inset). The gray edges denote circuits in which no light emitters were switched on. The number of lights switched on increased quadratically with time (E). This differs from k nonreplicative robots that can switch lights on in k Petri dishes per unit of time, resulting in a line with slope k (e.g., a single robot arm could switch on all four lights in its dish at every unit of time [dotted line in E]). With sufficient time, the self-replicative swarm can achieve higher utility than the nonreplicative swarm for any arbitrarily large value of k.
	Fig. S1. Construction of reconfigurable organisms from embryonic Xenopus material. Two methods are used to construct the initial swarm (generation 0) of reconfigurable organisms. The first requires excision of animal cap tissue of Nieuwkoop and Faber stage 10 embryos (24h post fertilization at 14°C) with microsurgery forceps (A). Individual explants are then transferred to a 0.75x saline solution (Marc’s Modified Ringer’s) which allows the tissue to heal into a spheroid of tissue (B) and develops into a mucociliary epithelium, becoming motile after 3-4 days of culture at 14°C (B′). The second method dissociates the animal cap material in calcium free, magnesium free media, and the pigmented superficial ectoderm is discarded (C). The dissociated cells are then transferred to 0.75x MMR and mechanically pushed into a pile, which naturally adheres (D). The aggregates forms into a spheroid of tissue (D′) which becomes motile after 3-4 days of culture at 14°C. Various morphologies can be given to parent organisms via surgical forceps and a microcautery electrode (E), allowing for the production of semi-toroidal shapes [shown in F, next to a spheroid (white arrow head in F) and shaped from reaggregated cells in F′ ], moderately compressed spheroids (G, lateral view G′), and toroids (H). Scale bars indicate 500 microns.
	Fig. S2. Reconfigurable organisms are required for the generation of offspring. Dissociated stem cell layers are produced from animal cap tissue of Nieuwkoop and Faber stage 9 embryos (A), which naturally dissociates when placed in calcium-free, magnesium-free media (B). Pooled and washed cells can then be deposited into dishes at various concentrations (C, C′), providing the necessary material for self replication. This process required reconfigurable organisms to be present: offspring were never produced across three trials with dissociated stem cells only (D-D′′). Any small aggregates fall apart over proceeding days of development, and no motile offspring were observed after 5 days (E). Scale bars indicate 500 microns.
	Fig. S3. Relative size of the self-replicating organisms. (A) It can be difficult to conceptualize 500 microns, so C- and O shaped designs were placed on top of a US dollar bill for comparison. Wild type reconfigurable organisms healed from an animal cap (B), and reconfigurable organisms formed by manually dissociating and reassociating the stem cells contained within a cap (B′), are shown at the same magnification beside a tadpole (C,C′), also at the same magnification: Scale bars indicate 500 microns.
	Fig. S4. Modeling kinematic self replication. [Clockwise from top left:] A swarm of nine virtual wild type spheroids (parents; pink) are placed in a virtual petri dish that is lined with virtual dissociated stem cells (green). As the swarm moves through the dissociated stem cells, piles of stem cells are formed (t=2). The parents are then removed (t=3), and any piles larger than a preselected threshold, develop from piles to motile offspring (green to pink) (t=4). More dissociated cells are injected into empty space in the dish, and pile building restarts. Here, a single filial generation was produced, then replication stopped (t=7). On average, simulated wild type spheroids did not produce piles larger than the selected threshold of two thirds the size of a wild type spheroid. This threshold was set higher than the biological data suggested: piles approximately one fifth the diameter of the initial parents could develop into motile offspring. However, small children are likely to produce even smaller grandchildren, or none at all. Because each filial generation is computationally expensive, we increased the threshold to create a more conservative filter: only the settings that result in the largest offspring and the most replication will pass through the filter and be allotted computational resources.
	Fig. S5. GPU-accelerated simulations. Deepgreen is a high performance computing cluster at the University of Vermont which contains ten Nvidia GPU nodes. Each Nvidia node has eight Tesla V100s that are capable of running the CUDA programming platform, which was a requirement of the employed simulator, voxcraft-sim. We parallelized evolutionary trials across different nodes: On each node, an independent trial maintained a population of 16 designs, which were evaluated in batches of eight designs at a time, in parallel, across the node’s eight GPUs. Each simulation contains a single design, which consists of N voxels. At each time step of simulation (numerical integration), the dynamics (position, velocity and acceleration) of each voxel within a simulation (on the order of 10 4 to 10 5 voxels) were evaluated concurrently on separate threads. Note that the number of voxels that can be updated in parallel will be constrained by the main memory bandwidth well before the number of voxels approaches the total number of potentially independent threads (80×2 11=163,840).
	Fig. S6. Amplifying self replication via morphology optimization in silico. Forty nine optimization trials were conducted (A), each of which starts with the evaluation of a swarm of wild type spheroids, in silico, under random swimming trajectories as derived from a unique set of random cilia forces. These 49 independent random cilia forces were held constant while body shape was optimized. Starting from 49 different randomly generated populations of 16 body shapes, the optimizer randomly removes voxels from the sphere, selecting shapes that result in more self replication. At the end of optimization, the highest amount of self replication produced by each of the 49 trials (B) was compared against the amount of self replication produced by the wild type spheres. The solid blue and red lines indicate mean fitness (whose integer part is the number of filial generations produced; Eqn. 1) across optimization time in silico for the optimized and wild type body shapes, respectively. Ninety-five percent bootstrapped confidence intervals (95%- and 5%-tiles) are drawn as shaded blue regions around the mean fitness of the optimized design; the dotted red line denotes the 95%-tile of fitness for the wild type spheres.
	Fig. S7. Controlling self replication via terrain optimization in silico. There are many tunable parameters that affect the efficacy of kinematic self replication in reconfigurable organisms. In addition to optimizing organism shape to increase self replication, we optimized the structure of the terrain. Black, immovable and unpassable voxels were added along the surface of each simulated petri dish. These barriers act as guide rails, channeling the random swimming of the unsculpted wild type spheres (pink) along certain trajectories. Instead of determining where to carve away tissue from a spherical body, the optimization algorithm now determines where to place black voxels on the surface plane. Random terrains trap the organisms, inhibiting self replication (top row). Optimized terrains reliably increased self replication compared to both random terrains and flat terrains without black voxels.
	Fig. S8. An optimized terrain that amplifies self replication in wild type reconfigurable organisms. One of the optimized terrains (black voxels) that amplified self replication in silico, yielded two filial generations of pile building after the initial swarm. On flat terrain (without black voxels), no replication occurs on average: the average number of filial generations is below one.
	Fig. S9. Recovering self-replication in a cluttered environment in silico. A static grid of unpassable black voxels were placed on the bottom of the simulated dish. In this cluttered environment, the wild type spherical organisms could no longer move enough to build offspring (A). Their ability to spontaneously self-replicate was lost. However, by optimizing organism shape, self replication can be recovered. The results of nine independent evolutionary trials are shown here at two different perspectives: from above (B) and from the side (C). The evolutionary algorithm discovered how to raise the organisms on stilts so they can glide over the top of the clutter and rescue function: aggregating loose stem cells into piles large enough to develop into offspring (D).
	Fig. S10. Increasingly larger offspring in silico. Reconfigurable organisms can create offspring that are larger than parents, and this enlarging process can persist for multiple rounds of replication in silico.
	Fig. S11. The simulated circuit completion task. A swarm of simulated kinematically self-replicating reconfigurable organisms was placed inside a petri dish alongside simulated modular electronic components (A) that can freely move and rotate along a surface plane, and connect on contact. For clarity, the dish is shown in grayscale, without the loose stem cells (B). There are three simulated electronic modules: light emitter (C), wire (D), and power supplies (E).

References [+] :

Blackiston, A cellular platform for the development of synthetic living machines. 2021, Pubmed, Xenbase