Pio-Lopez L et al. (2022), Active inference, morphogenesis, and computatio...

XB-ART-59503

Front Comput Neurosci 2022 Nov 24;16:988977. doi: 10.3389/fncom.2022.988977.

Show Gene links Show Anatomy links

Active inference, morphogenesis, and computational psychiatry.

Pio-Lopez L, Kuchling F, Tung A, Pezzulo G, Levin M.

???displayArticle.abstract???
Active inference is a leading theory in neuroscience that provides a simple and neuro-biologically plausible account of how action and perception are coupled in producing (Bayes) optimal behavior; and has been recently used to explain a variety of psychopathological conditions. In parallel, morphogenesis has been described as the behavior of a (non-neural) cellular collective intelligence solving problems in anatomical morphospace. In this article, we establish a link between the domains of cell biology and neuroscience, by analyzing disorders of morphogenesis as disorders of (active) inference. The aim of this article is three-fold. We want to: (i) reveal a connection between disorders of morphogenesis and disorders of active inference as apparent in psychopathological conditions; (ii) show how disorders of morphogenesis can be simulated using active inference; (iii) suggest that active inference can shed light on developmental defects or aberrant morphogenetic processes, seen as disorders of information processing, and perhaps suggesting novel intervention and repair strategies. We present four simulations illustrating application of these ideas to cellular behavior during morphogenesis. Three of the simulations show that the same forms of aberrant active inference (e.g., deficits of sensory attenuation and low sensory precision) that have been used to explain psychopathological conditions (e.g., schizophrenia and autism) also produce familiar disorders of development and morphogenesis when implemented at the level of the collective behavior of a group of cells. The fourth simulation involves two cells with too high precision, in which we show that the reduction of concentration signaling and sensitivity to the signals of other cells treats the development defect. Finally, we present the results of an experimental test of one of the model's predictions in early Xenopus laevis embryos: thioridazine (a dopamine antagonist that may reduce sensory precision in biological systems) induced developmental (anatomical) defects as predicted. The use of conceptual and empirical tools from neuroscience to understand the morphogenetic behavior of pre-neural agents offers the possibility of new approaches in regenerative medicine and evolutionary developmental biology.

???displayArticle.pubmedLink??? 36507307
???displayArticle.pmcLink??? PMC9731232
???displayArticle.link??? Front Comput Neurosci

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

	Figure 1. Representation of active inference. (A) Standard (non-pathological) behavior. We have one target morphology. During morphogenesis, cells start undifferentiated and at the same locations (middle-left panel). Given their generative model they believe they have to reach a specific planar target morphology (upper panel). They will receive different kind of sensory information during morphogenesis (bioelectrical pattern, biochemical and mechanical signals, see lower panel). This gathers sensory evidence that allows the cells to update their internal model of their target identity. Ultimately, they develop the appropriate target morphology (see middle-right panel) by action (migration, differentiation, release of signals, see second lower panel). (B) Dysfunctional behavior. In this case, the prior on the target morphology is much higher for one specific type of cells (intestinal in the example, see upper-left panel). The cells have a strong (rigid) belief that they have to be this type of cell and therefore won't take into account contrary sensory evidence (lower panel) and the prediction errors; in turn, this will lead to a dysfunctional update of the model and an inappropriate development (see middle panels).
	Figure 2. Schematic of variational Bayesian simulation of simple body shape morphogenesis illustrated by the example of regenerative patterning observed in planarian flatworms. (A) When dissecting out the center piece of a planarian flatworm, the constituent cells will remodel into a whole new worm with normal body axis. For the purpose of illustrating our simulation and for simplicity, cells that form different tissue types were grouped together as one cell. (B) Expected Signal concentrations (background color hues) at each target final position (colored stars) in the target morphology encode the cellular model of inference, with the same color encoding cell type from (A). (C) The target morphology for the arrangement of cells is encoded by expectations of external signals ec* for any given position ex* in the defined target morphology that constitutes the final configuration of cells. Each row in ec* corresponds to a different signaling type, while every column represents the signal expression states for a different cell. Adapted from Kuchling et al. (2020).
	Figure 3. (A) Time-lapse movie montage of simulation of normal morphogenesis. The eight initially unspecified cell types perform chemotaxis and update their posterior beliefs in order to infer the correct target morphology. (B) Differentiation profile of the eight cells during morphogenesis. (C) Free-energy minimization during morphogenesis.
	Figure 4. (A) Time-lapse movie montage of simulation of morphogenesis with a deficit of sensory attenuation high precision on the sensory signals received by the cells. The eight initially unspecified cell types perform chemotaxis and update their posterior beliefs in order to infer the correct target morphology. (B) Differentiation profile of the eight cells during morphogenesis. (C) Free-energy minimization during morphogenesis.
	Figure 5. (A) Results of the simulations for one to eight cells having a too high precision. (B) Results of the simulations for one to eight cells having a too high prior on their identity, the cells think they are intestinal cells.
	Figure 6. (A) Time-lapse movie montage of simulation of morphogenesis with a deficit of sensory attenuation via high prior on identity of the cells. The eight initially unspecified cell types perform chemotaxis and update their posterior beliefs in order to infer the correct target morphology. (B) Differentiation profile of the eight cells during morphogenesis. (C) Free-energy minimization during morphogenesis.
	Figure 7. (A) Time-lapse movie montage of simulation of morphogenesis with a deficit of sensory attenuation via low sensory precision of the cells. The eight initially unspecified cell types perform chemotaxis and update their posterior beliefs in order to infer the correct target morphology. (B) Differentiation profile of the eight cells during morphogenesis. (C) Free-energy minimization during morphogenesis.
	Figure 8. Timelapse of the rescue of two cells with too high precision via reduction of concentration signals and sensitivity to cell signals of the other cells.
	Figure 9. Experimental results showing the effects of thioridazine on the development of Xenopus laevis embryos. Compared to normal embryos in (A), treated embryos exhibit significant levels of abnormalities that include hypopigmentation, kinked bodies, edemas, abnormal face shapes, malformed guts, and cleft cement glands (B). The percent of embryos that had each defect is shown in (C). Treated embryos had higher incidents of hypopigmentation (7.17%), edemas (12.67%), abnormal face shapes (13.5%), and cleft cement glands (4.5%) than their control counterparts, summarized in (C). Each point represents one separate trial of 100 embryos each and is graphed as a mean with SEM and *p < 0.01 and p < 0.05.

References [+] :

Abouchar, Fly wing vein patterns have spatial reproducibility of a single cell. 2014, Pubmed