The tooth root is an integral, functionally important part of our dentition. The formation of a functional root depends on epithelialmesenchymal interactions and integration of the root with the jaw bone, blood supply and nerve innervations. The root development process therefore offers an attractive model for investigating organogenesis. Understanding how roots develop and how they can be bioengineered is also of great interest in the field of regenerative medicine. Here, we discuss recent advances in understanding the cellular and molecular mechanisms underlying tooth root formation. We review the function of cellular structure and components such as Hertwig's epithelial root sheath, cranial neural crest cells and stem cells residing in developing and adult teeth. We also highlight how complex signaling networks together with multiple transcription factors mediate tissue-tissue interactions that guide root development. Finally, we discuss the possible role of stem cells in establishing the crown-to-root transition, and provide an overview of root malformations and diseases in humans.

SUMMARYHair cells of the inner ear sensory organs originate from progenitor cells located at specific domains of the otic vesicle: the prosensory patches. Notch signalling is necessary for sensory development and loss of function of the Notch ligand jagged 1 (Jag1, also known as serrate 1) results in impaired sensory organs. However, the underlying mechanism of Notch function is unknown. Our results show that in the chicken otic vesicle, the Sox2 expression domain initially contains the nascent patches of Jag1 expression but, later on, Sox2 is only maintained in the Jag1-positive domains. Ectopic human JAG1 (hJag1) is able to induce Sox2 expression and enlarged sensory organs. The competence to respond to hJag1, however, is confined to the regions that expressed Sox2 early in development, suggesting that hJag1 maintains Sox2 expression rather than inducing it de novo. The effect is non-cell-autonomous and requires Notch signalling. hJag1 activates Notch, induces Hes/Hey genes and endogenous Jag1 in a non-cell-autonomous manner, which is consistent with lateral induction. The effects of hJag1 are mimicked by Jag2 but not by Dl1. Sox2 is sufficient to activate the Atoh1 enhancer and to ectopically induce sensory cell fate outside neurosensorycompetent domains. We suggest that the prosensory function of Jag1 resides in its ability to generate discrete domains of Notch activity that maintain Sox2 expression within restricted areas of an extended neurosensory-competent domain. This provides a mechanism to couple patterning and cell fate specification during the development of sensory organs.

The tongue and mandible have common origins. They arise simultaneously from the mandibular arch and are coordinated in their development and growth, which is evident from several clinical conditions such as Pierre Robin sequence. Here, we review in detail the molecular networks controlling both mandible and tongue development. We also discuss their mechanical relationship and evolution as well as the potential for stem cell-based therapies for disorders affecting these organs.

Large language models can encode a wealth of semantic knowledge about the world. Such knowledge could be extremely useful to robots aiming to act upon high-level, temporally extended instructions expressed in natural language. However, a significant weakness of language models is that they lack real-world experience, which makes it difficult to leverage them for decision making within a given embodiment. For example, asking a language model to describe how to clean a spill might result in a reasonable narrative, but it may not be applicable to a particular agent, such as a robot, that needs to perform this task in a particular environment. We propose to provide real-world grounding by means of pretrained skills, which are used to constrain the model to propose natural language actions that are both feasible and contextually appropriate. The robot can act as the language model's "hands and eyes," while the language model supplies high-level semantic knowledge about the task. We show how low-level skills can be combined with large language models so that the language model provides high-level knowledge about the procedures for performing complex and temporally extended instructions, while value functions associated with these skills provide the grounding necessary to connect this knowledge to a particular physical environment. We evaluate our method on a number of real-world robotic tasks, where we show the need for real-world grounding and that this approach is capable of completing long-horizon, abstract, natural language instructions on a mobile manipulator. The project's website and the video can be found at say-can.github.io. (a) Large Language Models (LLMs) (b) SayCan