Kinematic coordinations capture learning during human-exoskeleton interaction

Human-exoskeleton interactions have the potential to bring about changes in human behavior for physical rehabilitation or skill augmentation. Despite significant advances in the design and control of these robots, their application to human training remains limited. The key obstacles to the design of such training paradigms are the prediction of human-exoskeleton interaction effects and the selection of interaction control to affect human behavior. In this article, we present a method to elucidate behavioral changes in the human-exoskeleton system and identify expert behaviors correlated with a task goal. Specifically, we observe the joint coordinations of the robot, also referred to as kinematic coordination behaviors, that emerge from human-exoskeleton interaction during learning. We demonstrate the use of kinematic coordination behaviors with two task domains through a set of three human-subject studies. We find that participants (1) learn novel tasks within the exoskeleton environment, (2) demonstrate similarity of coordination during successful movements within participants, (3) learn to leverage these coordination behaviors to maximize success within participants, and (4) tend to converge to similar coordinations for a given task strategy across participants. At a high level, we identify task-specific joint coordinations that are used by different experts for a given task goal. These coordinations can be quantified by observing experts and the similarity to these coordinations can act as a measure of learning over the course of training for novices. The observed expert coordinations may further be used in the design of adaptive robot interactions aimed at teaching a participant the expert behaviors.

Comparing Plan Recognition Algorithms Through Standard Plan Libraries

Plan recognition deals with reasoning about the goals and execution process of an actor, given observations of its actions. It is one of the fundamental problems of AI, applicable to many domains, from user interfaces to cyber-security. Despite the prevalence of these approaches, they lack a standard representation, and have not been compared using a common testbed. This paper provides a first step towards bridging this gap by providing a standard plan library representation that can be used by hierarchical, discrete-space plan recognition and evaluation criteria to consider when comparing plan recognition algorithms. This representation is comprehensive enough to describe a variety of known plan recognition problems and can be easily used by existing algorithms in this class. We use this common representation to thoroughly compare two known approaches, represented by two algorithms, SBR and Probabilistic Hostile Agent Task Tracker (PHATT). We provide meaningful insights about the differences and abilities of these algorithms, and evaluate these insights both theoretically and empirically. We show a tradeoff between expressiveness and efficiency: SBR is usually superior to PHATT in terms of computation time and space, but at the expense of functionality and representational compactness. We also show how different properties of the plan library affect the complexity of the recognition process, regardless of the concrete algorithm used. Lastly, we show how these insights can be used to form a new algorithm that outperforms existing approaches both in terms of expressiveness and efficiency.

The Association for the Advancement of Artificial Intelligence 2020 Workshop Program

The Association for the Advancement of Artificial Intelligence 2020 Workshop Program included twenty-three workshops covering a wide range of topics in artificial intelligence. This report contains the required reports, which were submitted by most, but not all, of the workshop chairs.

"PhysIt" - A Diagnosis and Troubleshooting Tool for Physiotherapists in Training

Many physiotherapy treatments begin with a diagnosis process. The patient describes symptoms, upon which the physiotherapist decides which tests to perform until a final diagnosis is reached. The relationships between the anatomical components are too complex to keep in mind and the possible actions are abundant. A trainee physiotherapist with little experience naively applies multiple tests to reach the root cause of the symptoms, which is a highly inefficient process. This work proposes to assist students in this challenge by presenting three main contributions: (1) A compilation of the neuromuscular system as components of a system in a Model-Based Diagnosis problem; (2) The PhysIt is an AI-based tool that enables an interactive visualization and diagnosis to assist trainee physiotherapists; and (3) An empirical evaluation that comprehends performance analysis and a user study. The performance analysis is based on evaluation of simulated cases and common scenarios taken from anatomy exams. The user study evaluates the efficacy of the system to assist students in the beginning of the clinical studies. The results show that our system significantly decreases the number of candidate diagnoses, without discarding the correct diagnosis, and that students in their clinical studies find PhysIt helpful in the diagnosis process.

Reports of the Workshops Held at the 2019 AAAI Conference on Artificial Intelligence

The workshop program of the Association for the Advancement of Artificial Intelligence’s 33rd Conference on Artificial Intelligence (AAAI-19) was held in Honolulu, Hawaii, on Sunday and Monday, January 27–28, 2019. There were fifteen workshops in the program: Affective Content Analysis: Modeling Affect-in-Action, Agile Robotics for Industrial Automation Competition, Artificial Intelligence for Cyber Security, Artificial Intelligence Safety, Dialog System Technology Challenge, Engineering Dependable and Secure Machine Learning Systems, Games and Simulations for Artificial Intelligence, Health Intelligence, Knowledge Extraction from Games, Network Interpretability for Deep Learning, Plan, Activity, and Intent Recognition, Reasoning and Learning for Human-Machine Dialogues, Reasoning for Complex Question Answering, Recommender Systems Meet Natural Language Processing, Reinforcement Learning in Games, and Reproducible AI. This report contains brief summaries of the all the workshops that were held.

Goal and Plan Recognition Design for Plan Libraries

This article provides new techniques for optimizing domain design for goal and plan recognition using plan libraries. We define two new problems: Goal Recognition Design for Plan Libraries (GRD-PL) and Plan Recognition Design (PRD). Solving the GRD-PL helps to infer which goal the agent is trying to achieve, while solving PRD can help to infer how the agent is going to achieve its goal. For each problem, we define a worst-case distinctiveness measure that is an upper bound on the number of observations that are necessary to unambiguously recognize the agent’s goal or plan. This article studies the relationship between these measures, showing that the worst-case distinctiveness of GRD-PL is a lower bound of the worst-case plan distinctiveness of PRD and that they are equal under certain conditions. We provide two complete algorithms for minimizing the worst-case distinctiveness of plan libraries without reducing the agent’s ability to complete its goals: One is a brute-force search over all possible plans and one is a constraint-based search that identifies plans that are most difficult to distinguish in the domain. These algorithms are evaluated in three hierarchical plan recognition settings from the literature. We were able to reduce the worst-case distinctiveness of the domains using our approach, in some cases reaching 100% improvement within a predesignated time window. Our iterative algorithm outperforms the brute-force approach by an order of magnitude in terms of runtime.

Reports of the Workshops of the 32nd AAAI Conference on Artificial Intelligence

The AAAI-18 workshop program included 15 workshops covering a wide range of topics in AI. Workshops were held Sunday and Monday, February 2–7, 2018, at the Hilton New Orleans Riverside in New Orleans, Louisiana, USA. This report contains summaries of the Affective Content Analysis workshop; the Artificial Intelligence Applied to Assistive Technologies and Smart Environments; the AI and Marketing Science workshop; the Artificial Intelligence for Cyber Security workshop; the AI for Imperfect-Information Games; the Declarative Learning Based Programming workshop; the Engineering Dependable and Secure Machine Learning Systems workshop; the Health Intelligence workshop; the Knowledge Extraction from Games workshop; the Plan, Activity, and Intent Recognition workshop; the Planning and Inference workshop; the Preference Handling workshop; the Reasoning and Learning for Human-Machine Dialogues workshop; and the the AI Enhanced Internet of Things Data Processing for Intelligent Applications workshop.

Reports of the Workshops of the Thirty-First AAAI Conference on Artificial Intelligence

The AAAI-17 workshop program included 17 workshops covering a wide range of topics in AI. Workshops were held Sunday and Monday, February 4-5, 2017 at the Hilton San Francisco Union Square in San Francisco, California, USA. This report contains summaries of 12 of the workshops, and brief abstracts of the remaining 5

CRADLE

In exploratory domains, agents’ behaviors include switching between activities, extraneous actions, and mistakes. Such settings are prevalent in real world applications such as interaction with open-ended software, collaborative office assistants, and integrated development environments. Despite the prevalence of such settings in the real world, there is scarce work in formalizing the connection between high-level goals and low-level behavior and inferring the former from the latter in these settings. We present a formal grammar for describing users’ activities in such domains. We describe a new top-down plan recognition algorithm called CRADLE (Cumulative Recognition of Activities and Decreasing Load of Explanations) that uses this grammar to recognize agents’ interactions in exploratory domains. We compare the performance of CRADLE with state-of-the-art plan recognition algorithms in several experimental settings consisting of real and simulated data. Our results show that CRADLE was able to output plans exponentially more quickly than the state-of-the-art without compromising its correctness, as determined by domain experts. Our approach can form the basis of future systems that use plan recognition to provide real-time support to users in a growing class of interesting and challenging domains.

The repair Schwann cell and its function in regenerating nerves

Nerve injury triggers the conversion of myelin and non‐myelin (Remak) Schwann cells to a cell phenotype specialized to promote repair. Distal to damage, these repair Schwann cells provide the necessary signals and spatial cues for the survival of injured neurons, axonal regeneration and target reinnervation. The conversion to repair Schwann cells involves de‐differentiation together with alternative differentiation, or activation, a combination that is typical of cell type conversions often referred to as (direct or lineage) reprogramming. Thus, injury‐induced Schwann cell reprogramming involves down‐regulation of myelin genes combined with activation of a set of repair‐supportive features, including up‐regulation of trophic factors, elevation of cytokines as part of the innate immune response, myelin clearance by activation of myelin autophagy in Schwann cells and macrophage recruitment, and the formation of regeneration tracks, Bungner's bands, for directing axons to their targets. This repair programme is controlled transcriptionally by mechanisms involving the transcription factor c‐Jun, which is rapidly up‐regulated in Schwann cells after injury. In the absence of c‐Jun, damage results in the formation of a dysfunctional repair cell, neuronal death and failure of functional recovery. c‐Jun, although not required for Schwann cell development, is therefore central to the reprogramming of myelin and non‐myelin (Remak) Schwann cells to repair cells after injury. In future, the signalling that specifies this cell requires further analysis so that pharmacological tools that boost and maintain the repair Schwann cell phenotype can be developed.

Schwann cell precursors transplanted into the injured spinal cord multiply, integrate and are permissive for axon growth

There is a strong current interest in the use of cell transplantation for the treatment of spinal cord injuries. We report here the novel and potentially useful properties of an early cell in the Schwann cell lineage, the Schwann cell precursor (SCP). The experiments reveal a striking difference between these cells and Schwann cells when transplanted into the CNS. Unlike Schwann cells, SCPs thrive in the CNS where they initially proliferate rapidly but then fall out of division, thus effectively filling up the large cystic cavities formed following crush injury, while avoiding tumor formation. By 8 weeks, SCPs had started to express S100beta protein, a marker that differentiates Schwann cells from SCPs and had formed an apparently stable, vascularized cell mass, which created a continuous cellular bridge across the cystic cavities. The formation of the surrounding glial scar was reduced by local spread of the transplanted cells into the surrounding CNS tissue, where the cells integrated intimately with astrocytes and attenuated the physical barrier they normally form. SCP transplantation also altered and reduced the expression of chondroitin sulfate proteoglycans around the injury site. Caudal to the SCP transplants there was a large increase in the number of axons, compared with that seen in nontransplanted control tissue, showing that the implants effectively support axonal growth or sprouting. SCPs have advantageous attributes for CNS repair, despite the fact that sticky tape removal and ladder crossing tests at 8 weeks did not reveal significant functional improvements when compared with control animals.

Schwann cell precursors: a favourable cell for myelin repair in the Central Nervous System

Cell transplant therapies are currently under active consideration for a number of degenerative diseases. In the immune-mediated demyelinating-neurodegenerative disease multiple sclerosis (MS), only the myelin sheaths of the CNS are lost, while Schwann cell myelin of the PNS remains normal. This, and the finding that Schwann cells can myelinate CNS axons, has focussed interest on Schwann cell transplants to repair myelin in MS. However, the experimental use of these cells for myelin repair in animal models has revealed a number of problems relating to the incompatibility between peripheral glial cells and the CNS glial environment. Here, we have tested whether these difficulties can be avoided by using an earlier stage of the Schwann cell lineage, the Schwann cell precursor (SCP). For direct comparison of these two cell types, we implanted Schwann cells from post-natal rat nerves and SCPs from embryo day 14 (E14) rat nerves into the CNS under various experimental conditions. Examination 1 and 2 months later showed that in the presence of naked CNS axons, both types of cell form myelin that antigenically and ultrastructurally resembles that formed by Schwann cells in peripheral nerves. In terms of every other parameter we studied, however, the cells in these two implants behaved remarkably differently. As expected from previous work, Schwann cell implants survive poorly unless the cells find axons to myelinate, the cells do not migrate significantly from the implantation site, fail to integrate with host oligodendrocytes and astrocytes, and form little myelin when challenged with astrocyte-rich environment in the retina. Following SCP implantation, on the other hand, the cells survive well, migrate through normal CNS tissue, interface smoothly and intimately with host glial cells and myelinate extensively among the astrocytes of the retina. Furthermore, when implanted at a distance from a demyelinated lesion, SCPs but not Schwann cells migrate through normal CNS tissue to reach the lesion and generate new myelin. These features of SCP implants are all likely to be helpful attributes for a myelin repair cell. Since these cells also form Schwann cell myelin that is arguably likely to be resistant to MS pathology, they share some of the main advantages of Schwann cells without suffering from the disadvantages that render Schwann cells less than ideal candidates for transplantation into MS lesions.

The neurobiology of Schwann cells

This selective review of Schwann cell biology focuses on questions relating to the origins, development and differentiation of Schwann cells and the signals that control these processes. The importance of neuregulins and their receptors in controlling Schwann cell precursor survival and generation of Schwann cells, and the role of these molecules in Schwann cell biology is addressed. The reciprocal signalling between peripheral glial cells and neurons in development and adult life revealed in recent years is highlighted, and the profound change in survival regulation from neuron-dependent Schwann cell precursors to adult Schwann cells that depend on autocrine survival signals is discussed. Besides providing neuronal and autocrine signals, Schwann cells signal to mesenchymal cells and influence the development of the connective tissue sheaths of peripheral nerves. The importance of Desert Hedgehog in this process is described. The control of gene expression during Schwann cell development and differentiation by transcription factors is reviewed. Knockout of Oct-6 and Krox-20 leads to delay or absence of myelination, and these results are related to morphological or physiological observations on knockout or mutation of myelin-related genes. Finally, the relationship between selected extracellular matrix components, integrins and the cytoskeleton is explored and related to disease.

Signals that determine Schwann cell identity

While the signals that direct neural crest cells to choose the glial lineage and generate Schwann cell precursors are still obscure, studies both in vivo and in vitro indicate that the survival and differentiation of these cells to form Schwann cells is regulated by at least two signals, neuregulin-1 and endothelin. We know little about the signals that cause some immature Schwann cells to choose myelin differentiation, while other cells form non-myelinating cells. Three transcription factors, Sox-10, Oct-6 and Krox-20, have been shown to play key roles in the Schwann cell lineage. The transcription factor Krox-20 has been identified as a major target of the signals that induce myelin differentiation. Gene transfer experiments in vitro show that this protein has a remarkable ability to promote a large number of phenotypic changes in immature Schwann cells that characterize the transition of these cells to myelinating cells. Furthermore, Krox-20 shows important functional interactions with neuregulin and transforming growth factor beta (TGFbeta), two factors that have been implicated in the regulation of myelination in postnatal nerves. Another signal of importance in developing peripheral nerves, Desert Hedgehog, secreted by Schwann cells directs formation of the peripheral nerve connective tissue sheaths. Ongoing gene screening experiments are likely to reveal new genes of interest in this system.

Transforming growth factor beta (TGFbeta) mediates Schwann cell death in vitro and in vivo: examination of c-Jun activation, interactions with survival signals, and the relationship of TGFbeta-mediated death to Schwann cell differentiation

In some situations, cell death in the nervous system is controlled by an interplay between survival factors and negative survival signals that actively induce apoptosis. The present work indicates that the survival of Schwann cells is regulated by such a dual mechanism involving the negative survival signal transforming growth factor beta (TGFbeta), a family of growth factors that is present in the Schwann cells themselves. We analyze the interactions between this putative autocrine death signal and previously defined paracrine and autocrine survival signals and show that expression of a dominant negative c-Jun inhibits TGFbeta-induced apoptosis. This and other findings pinpoint activation of c-Jun as a key downstream event in TGFbeta-induced Schwann cell death. The ability of TGFbeta to kill Schwann cells, like normal Schwann cell death in vivo, is under a strong developmental regulation, and we show that the decreasing ability of TGFbeta to kill older cells is attributable to a decreasing ability of TGFbeta to phosphorylate c-Jun in more differentiated cells.

In early development of the rat mRNA for the major myelin protein P(0) is expressed in nonsensory areas of the embryonic inner ear, notochord, enteric nervous system, and olfactory ensheathing cells

The myelin protein P(0) has a major structural role in Schwann cell myelin, and the expression of P(0) protein and mRNA in the Schwann cell lineage has been extensively documented. We show here, using in situ hybridization, that the P(0) gene is also activated in a number of other tissues during embryonic development. P(0) mRNA is first detectable in 10-day-old embryos (E10) and is at this time seen only in cells in the cephalic neural crest and in the otic placode/pit. P(0) expression continues in the otic vesicle and at E12 P(0) expression in this structure largely overlaps with expression of another myelin gene, proteolipid protein. In the developing ear at E14, P(0) expression is complementary to expression of serrate and c-ret mRNAs, which later are expressed in sensory areas of the inner ear, while expression of bone morphogenetic protein (BMP)-4 and P(0), though largely complementary, shows small areas of overlap. P(0) mRNA and protein are detectable in the notochord from E10 to at least E13. In addition to P(0) expression in a subpopulation of trunk crest cells at E11/E12 and in Schwann cell precursors thereafter, P(0) mRNA is also present transiently in a subpopulation of cells migrating in the enteric neural crest pathway, but is down-regulated in these cells at E14 and thereafter. P(0) is also detected in the placode-derived olfactory ensheathing cells from E13 and is maintained in the adult. No signal is seen in cells in the melanocyte migration pathway or in TUJ1 positive neuronal cells in tissue sections. The activation of the P(0) gene in specific tissues outside the nervous system was unexpected. It remains to be determined whether this is functionally significant, or whether it is an evolutionary relic, perhaps reflecting ancestral use of P(0) as an adhesion molecule.

Developmental regulation and overexpression of the transcription factor AP-2, a potential regulator of the timing of Schwann cell generation

There is now evidence from in vivo and in vitro studies that the rate of Schwann cell generation is regulated by the balance of two opposing signals, beta neuregulins and endothelins. The beta neuregulins promote the development of precursors to Schwann cells whereas endothelins retard it through an action on endothelin-B receptors. The present work has shown additional controls of this transition, and implicates AP-2 transcription factors, in particular AP-2 alpha, as negative regulators of Schwann cell generation. We found that both AP-2 alpha and AP-2 gamma are present in early embryonic nerves, whereas AP-2 beta was not. Isoform-specific analysis of AP-2 alpha showed that isoform 3 was most abundant with isoforms 1 and 2 present in lesser amounts; isoform 4 was absent. Maximal AP-2 alpha and AP-2 gamma mRNA expression occurred at embryonic day (E) 12/13 in the mouse and at E14/15 in the rat, which correlates with the presence of Schwann cell precursors in the nerve. In both rats and in mice, in vivo and in vitro, downregulation of AP-2 alpha mRNA and protein coincided with one of the main steps in Schwann cell development, the precursor-Schwann cell transition. Moreover, Schwann cell generation was delayed if this downregulation was prevented by enforced expression of AP-2 alpha in precursors. These studies suggest that AP-2 is involved in the control of the timing of Schwann cell development.

The neuron-glia signal beta-neuregulin promotes Schwann cell motility via the MAPK pathway

Neuregulins constitute a family of related growth factors that play important roles in Schwann cell development and maturation. We investigated the involvement of beta-neuregulin in Schwann cell migration, using a simple in vitro bioassay. Pure Schwann cells were prepared from the sciatic nerves of 5-day-old rats and were grown in defined medium, with or without serum, until a monolayer of confluent cells was formed. A cell-free area was then generated by inflicting a scratch resulting in a 1-mm-wide gap. Schwann cell migration within the gap was monitored microscopically at given time intervals and was quantified using an image analysis system. The extent of cell proliferation was estimated by BrdU incorporation, and cell migration was quantified both in the absence and presence of cytosine arabinoside. We found that, in the absence of serum, beta-neuregulin at a dose submaximal for proliferation increased the rate of Schwann cell migration by 84%. A more moderate effect was observed when beta-neuregulin was applied in the presence of serum which, however, is by itself responsible for increased Schwann cell motility. To assess the signal transduction pathways involved in this procedure we used one inhibitor of MAPK, PD098059, two inhibitors of PI-3-kinase, wortmannin, and LY0294002, and three different PKC inhibitors. Of these PD098059 inhibited the neuregulin-induced enhancement in Schwann cell migration by 40%, the two PI-3-kinase inhibitors yielded an approximately 20% inhibition while the PKC inhibitors were ineffective. Our data indicate that the action of beta-neuregulin on Schwann cell motility is primarily mediated via the MAPK pathway.

Endothelins control the timing of Schwann cell generation in vitro and in vivo

Schwann cell precursors, derivatives of the neural crest, generate Schwann cells in a process that is tightly timed, well characterized, and directly controlled by axonal signals, in particular beta-neuregulins. Here we provide evidence that endothelins (ETs) are also important for survival and lineage progression in this system. We show that ETs promote rat Schwann cell precursor survival in vitro without stimulation of DNA synthesis. Using ET receptor agonists and antagonists, we demonstrate that this action of ET is mediated by the ET(B) receptor. RT-PCR reveals the presence of ET and ET receptor mRNA in the developing rat PNS. We showed previously that in vitro beta-neuregulins promote the generation of Schwann cells from precursors on schedule and that this process can be accelerated by fibroblast growth factor 2. Here we show that although ETs promote long-term precursor survival the transition of precursors to Schwann cells is delayed. Moreover, ETs block the maturation effects of beta-neuregulins. In spotting lethal rats, in which functional ET(B) receptors are absent, we find accelerated expression of the Schwann cell marker S100 in developing nerves. These observations indicate that complex growth factor interactions control the timing of Schwann cell development in embryonic nerves and that ETs act as negative regulators of Schwann cell generation.

Auditory cortex on the human posterior superior temporal gyrus

The human superior temporal cortex plays a critical role in hearing, speech, and language, yet its functional organization is poorly understood. Evoked potentials (EPs) to auditory click-train stimulation presented binaurally were recorded chronically from penetrating electrodes implanted in Heschl's gyrus (HG), from pial-surface electrodes placed on the lateral superior temporal gyrus (STG), or from both simultaneously, in awake humans undergoing surgery for medically intractable epilepsy. The distribution of averaged EPs was restricted to a relatively small area on the lateral surface of the posterior STG. In several cases, there were multiple foci of high amplitude EPs lying along this acoustically active portion of STG. EPs recorded simultaneously from HG and STG differed in their sensitivities to general anesthesia and to changes in rate of stimulus presentation. Results indicate that the acoustically active region on the STG is a separate auditory area, functionally distinct from the HG auditory field(s). We refer to this acoustically sensitive area of the STG as the posterior lateral superior temporal area (PLST). Electrical stimulation of HG resulted in short-latency EPs in an area that overlaps PLST, indicating that PLST receives a corticocortical input, either directly or indirectly, from HG. These physiological findings are in accord with anatomic evidence in humans and in nonhuman primates that the superior temporal cortex contains multiple interconnected auditory areas.

Auditory cortex on the human posterior superior temporal gyrus

The human superior temporal cortex plays a critical role in hearing, speech, and language, yet its functional organization is poorly understood. Evoked potentials (EPs) to auditory click-train stimulation presented binaurally were recorded chronically from penetrating electrodes implanted in Heschl's gyrus (HG), from pial-surface electrodes placed on the lateral superior temporal gyrus (STG), or from both simultaneously, in awake humans undergoing surgery for medically intractable epilepsy. The distribution of averaged EPs was restricted to a relatively small area on the lateral surface of the posterior STG. In several cases, there were multiple foci of high amplitude EPs lying along this acoustically active portion of STG. EPs recorded simultaneously from HG and STG differed in their sensitivities to general anesthesia and to changes in rate of stimulus presentation. Results indicate that the acoustically active region on the STG is a separate auditory area, functionally distinct from the HG auditory field(s). We refer to this acoustically sensitive area of the STG as the posterior lateral superior temporal area (PLST). Electrical stimulation of HG resulted in short-latency EPs in an area that overlaps PLST, indicating that PLST receives a corticocortical input, either directly or indirectly, from HG. These physiological findings are in accord with anatomic evidence in humans and in nonhuman primates that the superior temporal cortex contains multiple interconnected auditory areas.

Schwann cell-derived desert hedgehog signals nerve sheath formation

Reciprocal signaling between axons and Schwann cells during development is well established. The contribution of Schwann cells to the formation and maintenance of the protective nerve sheaths (endo-, peri-, and epineurium) has been less studied. Although mesenchymal cells contribute to all these structures, only perineurial cells contribute to the diffusion barrier between nerves and surrounding tissues. During development, prospective perineurial cells shift from a mesenchymal to epithelial phenotype, forming concentric layers of cells around the nerve fascicles that collectively form a barrier against unwanted molecules and cellular infiltration. We have studied the role of Schwann cells in the formation and maintenance of this barrier. The signaling molecule Desert hedgehog is expressed in Schwann cell precursors, and in Schwann cells until at least postnatal day 10, while its receptor patched is seen in mesenchymal cells surrounding the developing nerve at embryo day 15. In Desert hedgehog knockout mice, the connective tissue sheaths in adult nerves appear highly abnormal by electron microscopy. There is almost no epineurium, and the perineurium is thin and highly abnormal. In addition, perineurial-like cells invade the endoneurial space, forming mini-fascicles around small bundles of nerve fibers similar to those seen in regenerating nerves. Functional tests reveal that the diffusion and cellular infiltration barrier is compromised, demonstrating that Desert hedgehog signaling from Schwann cells to the mesenchyme is involved in the formation of a morphologically and functionally normal perineurium.

Why do Schwann cells survive in the absence of axons?

Schwann cell precursors in embryonic nerves rely for survival on signals from the axons they associate with. A major component of this signal is beta neuregulin. While it can be argued that such paracrine axonal regulation makes biological sense in embryonic nerves, such an arrangement would be problematic postnatally, since nerve damage would then lead to Schwann cell death with adverse consequences for regeneration; in fact, transection of older nerves is not accompanied by a detectable increase in Schwann cell death. Our evidence indicates that this is, at least in part, due to the ability of Schwann cells to support their own survival by autocrine circuits. These circuits are not present in Schwann cell precursors. We have identified insulin-like growth factor, neurotrophin-3 and platelet-derived growth factor-BB as components of the autocrine Schwann cell survival signal.

Schwann cells and their precursors emerge as major regulators of nerve development

It is becoming ever clearer that Schwann cells and Schwann-cell precursors are an important source of developmental signals in embryonic and neonatal nerves. This article reviews experiments showing that these signals regulate the survival and differentiation of other cells in early nerves. The evidence indicates that glial-derived signals are necessary for neuronal survival at crucial periods of development, that they regulate the molecular and functional specialization of axons and that they control the maturation of the perineurial sheath that protects nerves from inflammation and unwanted macro-molecules produced in the surrounding tissues. Furthermore, an autocrine survival circuit enables Schwann cells in postnatal nerves to survive in the absence of axons, a vital requirement for successful nerve regeneration following injury. The molecular identity of these signals and their receptors is currently being determined.

Schwann cell-derived Desert hedgehog controls the development of peripheral nerve sheaths

We show that Schwann cell-derived Desert hedgehog (Dhh) signals the formation of the connective tissue sheath around peripheral nerves. mRNAs for dhh and its receptor patched (ptc) are expressed in Schwann cells and perineural mesenchyme, respectively. In dhh-/- mice, epineurial collagen is reduced, while the perineurium is thin and disorganized, has patchy basal lamina, and fails to express connexin 43. Perineurial tight junctions are abnormal and allow the passage of proteins and neutrophils. In nerve fibroblasts, Dhh upregulates ptc and hedgehog-interacting protein (hip). These experiments reveal a novel developmental signaling pathway between glia and mesenchymal connective tissue and demonstrate its molecular identity in peripheral nerve. They also show that Schwann cell-derived signals can act as important regulators of nerve development.

Schwann cell development in embryonic mouse nerves

Previously we proposed that Schwann cell development from the neural crest is a two-step process that involves the generation of one main intermediate cell type, the Schwann cell precursor. Until now Schwann cell precursors have only been identified in the rat, and much remains to be learned about these cells and how they generate Schwann cells. Here we identify this cell in the mouse and analyze its transition to form Schwann cells in terms of timing, molecular expression, and extracellular signals and intracellular pathways involved in survival, proliferation, and differentiation. In the mouse, the transition from precursors to Schwann cells takes place 2 days earlier than in the rat, i.e., between embryo days 12/13 and 15/16, and is accompanied by the appearance of the 04 antigen and the establishment of an autocrine survival circuit. Beta neuregulins block precursor apoptosis and support Schwann cell generation in vitro, a process that is accelerated by basic fibroblast growth factor 2. The development of Schwann cells from precursors also involves a change in the intracellular survival signals utilized by neuregulins: To block precursor death neuregulins need to signal through both the mitogen-activated protein kinase and the phosphoinositide-3-kinase pathways although neuregulins support Schwann cell survival by signaling through the phosphoinositide-3-kinase pathway alone. Last, we describe the generation of precursor cultures from single 12-day-old embryos, a prerequisite for culture studies of genetically altered precursors when embryos are non-identical with respect to the transgene in question.

Developing Schwann cells acquire the ability to survive without axons by establishing an autocrine circuit involving insulin-like growth factor, neurotrophin-3, and platelet-derived growth factor-BB

Although Schwann cell precursors from early embryonic nerves die in the absence of axonal signals, Schwann cells in older nerves can survive in the absence of axons in the distal stump of transected nerves. This is crucially important, because successful axonal regrowth in a damaged nerve depends on interactions with living Schwann cells in the denervated distal stump. Here we show that Schwann cells acquire the ability to survive without axons by establishing an autocrine survival loop. This mechanism is absent in precursors. We show that insulin-like growth factor, neurotrophin-3, and platelet-derived growth factor-BB are important components of this autocrine survival signal. The secretion of these factors by Schwann cells has significant implications for cellular communication in developing nerves, in view of their known ability to regulate survival and differentiation of other cells including neurons.

Chronic microelectrode investigations of normal human brain physiology using a hybrid depth electrode

Neurosurgeons have unique access to in vivo human brain tissue, and in the course of clinical treatment important scientific advances have been made that further our understanding of normal brain physiology. In the modern era, microelectrode recordings have been used to systematically investigate the cellular properties of lateral temporal cerebral cortex. The current report describes a hybrid depth electrode (HDE) recording technique that was developed to enable neurosurgeons to simultaneously investigate normal cellular physiology during chronic intracranial EEG recordings. The HDE combines microelectrode and EEG recordings sites on a single shaft. Multiple microelectrode recordings are obtained from MRI defined brain sites and single-unit activity is discriminated from these data. To date, over 60 HDEs have been placed in 20 epilepsy surgery patients. Unique physiologic data have been gathered from neurons in numerous brain regions, including amygdala, hippocampus, frontal lobe, insula and Heschl's gyrus. Functional activation studies were carried out without risking patient safety or comfort.

Posterior ventricular catheter burr-hole localizer. Technical note

Proper ventricular catheter placements are associated with improved shunt performance. When placing ventricular catheters via the posterior approach, the surgeon must determine an optimum trajectory and then pass a catheter along that trajectory. The incidence of optimal posterior catheter placements is increased by using a posterior catheter guide (PCG); however, errors may still occur because of poor selection of a posterior burr-hole site. In this report an easy-to-use posterior burr-hole localizer (Localizer) is described that defines the optimum burr-hole location based on geometric relationships involving the ear and supraorbital rims. The basic design principle of the Localizer was formulated and tested by using neuronavigational imaging tools to examine normal adult ventricular anatomy in relation to surface landmarks and by reviewing imaging studies obtained in 50 adult patients with hydrocephalus. Subsequently, the Localizer was used in 28 consecutive patients scheduled to undergo shunt surgery performed by using the PCG. In all cases the catheter entered the ventricle on the first pass and postoperative imaging studies demonstrated successful placement in the ipsilateral anterior horn. There were no catheter-related complications. These early results indicate that the Localizer and PCG devices may be safe and effective when used in combination for placement of posterior ventricular catheters.

Origin and early development of Schwann cells

Cellular events leading to the generation of Schwann cells from the neural crest have recently been clarified and it is now possible to outline a relatively simple model of the Schwann cell lineage in the rat and mouse. Neural crest cells have to undergo three main developmental transitions to become mature Schwann cells. These are the formation of Schwann cell precursors from crest cells, the formation of immature Schwann cells from precursors and, lastly, the postnatal and reversible generation of non-myelin- and myelin-forming Schwann cells. Axonal signals involving neuregulins are important regulators of these events, in particular of the survival, proliferation, and differentiation of Schwann cell precursors. Transcription factors likely to be involved in the developmental transitions are beginning to be identified. These include Oct-6, Krox-20, and Pax-3 but also members of the basic helix-loop-helix family, Sox 10, and the cAMP response element binding protein CREB.

The neuron-glia signal beta neuregulin induces sustained CREB phosphorylation on Ser-133 in cultured rat Schwann cells

beta neuregulins (also called NDF, GGF, ARIA, and heregulins) are neuron-derived molecules that are likely to be responsible for Schwann cell precursor survival, proliferation, and maturation in vivo and in vitro. Although the receptors to which beta neuregulins bind have been defined, little is known about the transcription factors these important ligands activate. Using antibodies, quantitative imaging methods and Western blotting, we show that beta neuregulin induces a high level of phosphorylation of the transcription factor cyclic AMP response element binding protein (CREB) on Ser-133 in cultured rat Schwann cells and that the phosphorylation is prolonged over several hours. In contrast, neurotrophins, CNTF, FGF-2, EGF, and TGF beta induce little or no phosphorylation of CREB despite the fact that receptors for these factors are present on Schwann cells. As expected CREB phosphorylation was detected following cAMP elevation, and it was also induced by elevation of cytoplasmic Ca2+, endothelin 1, and PDGF-BB. The signal was lower than that seen in response to beta neuregulin, and transient, unlike the sustained CREB activation induced by beta neuregulin. Our results suggest that the sustained phosphorylation of CREB on Ser-133 may contribute to the broad spectrum of effects that beta neuregulins have on cells of the Schwann cell lineage and that the CREB pathway may be important for transduction of neuregulin signals in Schwann cells.

Helix-loop-helix proteins in Schwann cells: a study of regulation and subcellular localization of Ids, REB, and E12/47 during embryonic and postnatal development

Although basic helix-loop-helix (bHLH) proteins play an important role in transcriptional control in many cell types, the role of HLH proteins in Schwann cells has yet to be assessed. In this study, we have analyzed the expression of the dominant negative HLH genes, Id1 to Id4 and the class A gene REB, during Schwann cell development. We found that mRNA derived from these genes was present in the Schwann cell lineage throughout development including embryonic precursors and mature cells. The mRNA levels were not significantly regulated during development. Nevertheless, by using antibodies against the four different Id proteins, we found clear regulation of some of these genes at the protein level, in particular Id 2, 4, and REB, both in amount and nuclear/cytoplasmic localization. All these proteins are found in the nuclei of Schwann cell precursors but are not seen in nuclei of Schwann cells of newborn nerves. We observed extensive overlap in Id expression, especially in Schwann cell precursors that co-expressed all four Id proteins and REB. We also showed that Id 1 and 2 were up-regulated as Schwann cells progressed through the cell cycle. These data indicate that HLH transcription factors act as regulators of Schwann cell development and point to the existence of as yet unidentified cell type-specific bHLH proteins in these cells.

An illuminating retractor for intracranial microneurosurgery

During certain microneurosurgical procedures, the illumination provided by current coaxial microscope light sources is suboptimal. One potential solution is the surgical light blade (SLB), a malleable retractor incorporating a flat fiberoptic bundle along its surface. The SLB provides intense illumination within the surgical field without decreasing the amount of limited operating space. A prototype of the SLB was tested in six patients. Results suggest that with further development the SLB could become useful for intracranial microneurosurgery.

An integrated multipurpose lesion-making electrode

PURPOSE

Although excellent results are reported from centers using microelectrode mapping during stereotactic pallidotomy, the recording methods used are time-consuming and technically cumbersome. We sought to develop a new electrode concept that may, when eventually applied in clinical practice, improve the efficiency and safety of microelectrode-guided functional neurosurgical procedures.

CONCEPT

The scout electrode uses a recently developed research technique for simultaneously obtaining multiple microelectrode recordings along the shaft of a macroelectrode.

RATIONALE

By positioning recording sites on either side of a lesion-making contact, it is possible to functionally "scout" the brain tissue surrounding the proposed lesion site, thus eliminating the need for serial movements and electrode interchanges.

DISCUSSION

The feasibility of the scout electrode concept was tested using prototypes placed into cat medial geniculate nucleus. High-quality unit recordings were simultaneously obtained from different regions within the medial geniculate nucleus. This ability to physiologically map sites above and below the lesion-making contact facilitated precise placement of radiofrequency lesions in the center of the medial geniculate nucleus. These results suggest that it may be possible to develop clinically useful devices based on this novel concept.

Helix-loop-helix proteins in Schwann cells: a study of regulation and subcellular localization of Ids, REB, and E12/47 during embryonic and postnatal development

Although basic helix-loop-helix (bHLH) proteins play an important role in transcriptional control in many cell types, the role of HLH proteins in Schwann cells has yet to be assessed. In this study, we have analyzed the expression of the dominant negative HLH genes, Id1 to Id4 and the class A gene REB, during Schwann cell development. We found that mRNA derived from these genes was present in the Schwann cell lineage throughout development including embryonic precursors and mature cells. The mRNA levels were not significantly regulated during development. Nevertheless, by using antibodies against the four different Id proteins, we found clear regulation of some of these genes at the protein level, in particular Id 2, 4, and REB, both in amount and nuclear/cytoplasmic localization. All these proteins are found in the nuclei of Schwann cell precursors but are not seen in nuclei of Schwann cells of newborn nerves. We observed extensive overlap in Id expression, especially in Schwann cell precursors that co-expressed all four Id proteins and REB. We also showed that Id 1 and 2 were up-regulated as Schwann cells progressed through the cell cycle. These data indicate that HLH transcription factors act as regulators of Schwann cell development and point to the existence of as yet unidentified cell type-specific bHLH proteins in these cells.

Expression and regulation of alpha1beta1 integrin in Schwann cells

The interaction of cells with the extracellular matrix plays a critical role in morphogenesis and cell differentiation. To define how Schwann cells might interact with the extracellular matrix, we chose to study the expression of the laminin/collagen receptor alpha1beta1 integrin during nerve development in the rat from embryonic day 14 to maturity. We found that this integrin is expressed predominantly on mature non-myelin-forming cells and only at very low levels on myelin-forming cells. Significant levels of this integrin were not detected on Schwann cell precursors or embryonic Schwann cells in vivo. Experiments using transected and crushed sciatic nerve showed that alpha1beta1 integrin expression is regulated at least in part by axonal contact. Furthermore, Schwann cell culture experiments showed that alpha1beta1 integrin levels are strongly upregulated by transforming growth factor-beta(s) and phorbol esters.

Embryonic Schwann cell development: the biology of Schwann cell precursors and early Schwann cells

The cellular events leading to the generation of Schwann cells from the neural crest have recently been clarified and it is now possible to outline a relatively simple model of the Schwann cell lineage in the rat and mouse. Neural crest cells have to undergo 3 main developmental transitions to become mature Schwann cells. These are the formation of Schwann cell precursors from crest cells, the formation of immature Schwann cells from precursors and, lastly, the postnatal and reversible generation of non-myelin- and myelin-forming Schwann cells. Axonal signals involving neuregulins are important regulators of these events, in particular of the survival, proliferation and differentiation of Schwann cell precursors.

Identification of transcriptionally regulated mRNAs from mouse Schwann cell precursors using modified RNA fingerprinting methods

We have adopted RNA fingerprinting methods to screen for genes that are rapidly up- or down-regulated during normal mammalian development, comparing mRNA from early (embryo day 12) to late (embryo day 13) mouse Schwann cell precursors. The use of total RNA, a reduction of cDNA template for amplification, the detection of RT-PCR products with a sensitive DNA stain and polyacrylamide gel electrophoresis and rigid selection criteria involving three screening steps are significant improvements on previous methods. Of 19 differentially displayed bands, 15 represented novel genes. The four known cDNA fragments (interleukin enhancer binding factor 1, beta3 subunit of phospholipase C, brain beta-spectrin, and P21 polypeptide) consisted of coding sequences, indicating a high chance of obtaining coding regions. A semiquantitative RT-PCR analysis of three of the four known genes and a cDNA fragment randomly selected from the pool of 15 novel sequences, confirmed that they were regulated between embryo days 12 and 13, as predicted by the display gels. Our results suggest that the combination of methods described here will have wide applicability in studies of other developmental systems where precisely timed changes occur and where only small amounts of RNA can be obtained for analysis.

Response of Schwann cells to mitogens in vitro is determined by pre-exposure to serum, time in vitro, and developmental age

We compared the mitogenic response of Schwann cells freshly isolated from adult, neonatal, and embryonic nerves, and compared these cells with cells that had been cultured in serum for 5 days. DNA synthesis in response to growth factors was measured using bromodeoxyuridine and immunocytochemistry. Freshly isolated adult Schwann cells were unresponsive to growth factors with or without forskolin to elevate intracellular cAMP levels. After 5 days of culture in serum, or alternatively in defined medium containing fibroblast growth factor 2 plus forskolin, or neu-differentiation factor beta2, adult cells were responsive to mitogens, whereas cells cultured in defined medium alone remained unresponsive. Serum also increased expression of type 1 fibroblast growth factor receptor. Freshly isolated embryonic and neonatal Schwann cells in contrast responded to growth factors even in the absence of forskolin. This responsiveness changed with time in culture. Neonatal cells cultured for 5 days in defined medium in the presence or absence of serum no longer responded to FGF alone, but required forskolin for a mitogenic response. Thus, the response of freshly isolated cells to mitogens is developmentally regulated; extrinsic signals are required to render adult cells responsive to mitogens; and with time in culture, neonatal cells develop a requirement for cAMP elevation for mitogenic response.

P0 is constitutively expressed in the rat neural crest and embryonic nerves and is negatively and positively regulated by axons to generate non-myelin-forming and myelin-forming Schwann cells, respectively

We show that in the rat, the major gene of PNS myelin, P0, is expressed long before myelination in the neural crest, Schwann cell precursors, and embryonic Schwann cells irrespective of whether they will myelinate or not. This myelin-independent P0 expression is constitutive and likely to serve as a specific marker for the Schwann cell lineage. The much higher P0 expression accompanying myelination is therefore not new gene expression but strong up-regulation of preexisting basal levels. We provide new evidence that the up-regulation to myelination-related levels depends on positive extrinsic signals and therefore does not represent a constitutive phenotype. P0 mRNA is not detectable in mature non-myelin-forming Schwann cells of the sympathetic trunk, but is detectable after transection, indicating that there is a P0-inhibitory signal associated with mature unmyelinated axons. Thus, the regulation of the P0 gene is complex, encompassing extrinsically signaled amplification superimposed on a highly lineage-specific and constitutive basal expression.

Oct-6 (SCIP/Tst-1) is expressed in Schwann cell precursors, embryonic Schwann cells, and postnatal myelinating Schwann cells: comparison with Oct-1, Krox-20, and Pax-3

The POU domain transcription factor Oct-6 (SCIP/Tst-1) is likely to control important stages of Schwann cell development, including the initiation of myelination around birth. Here, we use immunocytochemical and reverse transcriptase-polymerase chain reaction techniques to examine Oct-6 earlier in nerve development, to test the idea that Oct-6 has an additional role in Schwann cell precursors or early embryonic Schwann cells, a possibility raised by previous studies on transgenic mice. Consistent with this, we find low but unambiguous levels of Oct-6 mRNA and protein in Schwann cell precursors of mouse and rat (nerves from 12- and 14-day-old embryos, respectively), with expression levels gradually increasing during early Schwann cell development and towards birth. Unexpectedly, Oct-6 immunoreactivity is clearly present in nuclei of most myelinating cells at least as late as postnatal day 12. Furthermore, many nonmyelinating Schwann cells express Oct-6 in adult life. A comparison of Oct-6 mRNA with other Schwann cell transcription factors-namely, Oct-1, Krox-20, and Pax-3-reveals that each factor exhibits strong developmental regulation and a unique expression pattern in embryonic nerves. Therefore, they are likely to play distinct regulatory roles in early development of the Schwann cell lineage.