Department of Computer Science, Tufts University, Medford, MA, USA.
Allen Discovery Center at Tufts University and Department of Biology, Medford, MA, USA; Wyss Institute for Biologically Inspired Engineering at Harvard University, Boston, MA, USA.
Biosystems. 2022 Nov;221:104762. doi: 10.1016/j.biosystems.2022.104762. Epub 2022 Sep 3.
The standard view of embryogenesis is one of cooperation driven by the cells' shared genetics and evolutionary interests. However, numerous examples from developmental biology and agriculture reveal a surprising amount of competition among body cells, tissues, and organs for both metabolic and informational resources. To explain the existence of such competition we had hypothesized that evolution uses limiting "reservoirs" of resource molecules as a communication medium - a global scratchpad, to enable tissues across the body to coordinate growth. Here, we test this hypothesis via an evolutionary simulation of embryogeny in silico. Genomes encode state transition rules for cells, such as proliferation, differentiation, and resource use, enabling virtual embryos to develop a specific large-scale morphology. An evolutionary algorithm operates over the genomes, with fitness defined as a function of specific morphological requirements for the final embryo shape. We found that not only does such an algorithm rapidly discover rules for cellular behavior that reliably make embryos with specific anatomical properties, but that it discovers the strategy of using finite resources to coordinate development. Given the option of using finite or infinite reservoirs (which determine cells' ability to carry out specific actions), evolution preferentially uses finite reservoirs, which results in higher fitness and increased consistency (without needing direct selection for morphological invariance). We report aspects of anatomical, physiological/transcriptional, and genomic analysis of evolved virtual embryos that help understand how evolution can use competition among genetically identical subunits within a multicellular body to coordinate reliable, complex morphogenesis. Our results suggest that under some conditions, composite multi-scale systems will promote conflict and artificial scarcity for their components.
胚胎发生的标准观点是由细胞的共同遗传和进化利益驱动的合作。然而，来自发育生物学和农业的众多例子揭示了身体细胞、组织和器官之间为代谢和信息资源而进行的惊人竞争。为了解释这种竞争的存在，我们假设进化使用资源分子的有限“储库”作为一种沟通媒介——一个全局的草稿本，使身体各处的组织能够协调生长。在这里，我们通过对胚胎发生的进化模拟进行了测试。基因组为细胞编码状态转换规则，例如增殖、分化和资源利用，使虚拟胚胎能够发展出特定的大规模形态。进化算法在基因组上运行，适应度定义为最终胚胎形状的特定形态要求的函数。我们发现，这种算法不仅能够快速发现可靠地使胚胎具有特定解剖学特性的细胞行为规则，而且还发现了使用有限资源来协调发育的策略。给定使用有限或无限储库的选择（决定细胞执行特定动作的能力），进化优先使用有限储库，这导致更高的适应度和更高的一致性（无需直接选择形态不变性）。我们报告了进化后的虚拟胚胎的解剖学、生理/转录和基因组分析的各个方面，这些方面有助于理解进化如何利用多细胞体内遗传相同的亚基之间的竞争来协调可靠的、复杂的形态发生。我们的结果表明，在某些条件下，复合多尺度系统将促进其组件之间的冲突和人为稀缺。
