Embodied mechanisms of motor control in the octopus.

Achieving complex behavior in soft-bodied animals is a hard task, because their body morphology is not constrained by a fixed number of jointed elements, as in skeletal animals, and thus the control system has to deal with practically an infinite number of control variables (degrees of freedom). Almost 30 years of research on Octopus vulgaris motor control has revealed that octopuses efficiently control their body with strategies that emerged during the adaptive coevolution of their nervous system and body morphology. In this minireview, we highlight principles of embodied organization that were revealed by studying octopus motor control, and that are used as inspiration for soft robotics. We describe the evolved solutions to the problem, implemented from the lowest level, the muscular system, to the network organization in higher motor control centers of the octopus brain. We show how the higher motor control centers, where the sensory-motor interface lies, can control and coordinate limbs with large degrees of freedom without using body-part maps to represent sensory and motor information, as they do in vertebrates. We demonstrate how this unique control mechanism, which allows efficient control of the body in a large variety of behaviors, is embodied within the animal's body morphology.

mEMbrain: an interactive deep learning MATLAB tool for connectomic segmentation on commodity desktops.

Connectomics is fundamental in propelling our understanding of the nervous system's organization, unearthing cells and wiring diagrams reconstructed from volume electron microscopy (EM) datasets. Such reconstructions, on the one hand, have benefited from ever more precise automatic segmentation methods, which leverage sophisticated deep learning architectures and advanced machine learning algorithms. On the other hand, the field of neuroscience at large, and of image processing in particular, has manifested a need for user-friendly and open source tools which enable the community to carry out advanced analyses. In line with this second vein, here we propose mEMbrain, an interactive MATLAB-based software which wraps algorithms and functions that enable labeling and segmentation of electron microscopy datasets in a user-friendly user interface compatible with Linux and Windows. Through its integration as an API to the volume annotation and segmentation tool VAST, mEMbrain encompasses functions for ground truth generation, image preprocessing, training of deep neural networks, and on-the-fly predictions for proofreading and evaluation. The final goals of our tool are to expedite manual labeling efforts and to harness MATLAB users with an array of semi-automatic approaches for instance segmentation. We tested our tool on a variety of datasets that span different species at various scales, regions of the nervous system and developmental stages. To further expedite research in connectomics, we provide an EM resource of ground truth annotation from four different animals and five datasets, amounting to around 180 h of expert annotations, yielding more than 1.2 GB of annotated EM images. In addition, we provide a set of four pre-trained networks for said datasets. All tools are available from https://lichtman.rc.fas.harvard.edu/mEMbrain/. With our software, our hope is to provide a solution for lab-based neural reconstructions which does not require coding by the user, thus paving the way to affordable connectomics.

Connectomics of the Octopus vulgaris vertical lobe provides insight into conserved and novel principles of a memory acquisition network.

Here, we present the first analysis of the connectome of a small volume of the Octopus vulgaris vertical lobe (VL), a brain structure mediating the acquisition of long-term memory in this behaviorally advanced mollusk. Serial section electron microscopy revealed new types of interneurons, cellular components of extensive modulatory systems, and multiple synaptic motifs. The sensory input to the VL is conveyed via~1.8 × 106 axons that sparsely innervate two parallel and interconnected feedforward networks formed by the two types of amacrine interneurons (AM), simple AMs (SAMs) and complex AMs (CAMs). SAMs make up 89.3% of the~25 × 106VL cells, each receiving a synaptic input from only a single input neuron on its non-bifurcating primary neurite, suggesting that each input neuron is represented in only~12 ± 3.4SAMs. This synaptic site is likely a 'memory site' as it is endowed with LTP. The CAMs, a newly described AM type, comprise 1.6% of the VL cells. Their bifurcating neurites integrate multiple inputs from the input axons and SAMs. While the SAM network appears to feedforward sparse 'memorizable' sensory representations to the VL output layer, the CAMs appear to monitor global activity and feedforward a balancing inhibition for 'sharpening' the stimulus-specific VL output. While sharing morphological and wiring features with circuits supporting associative learning in other animals, the VL has evolved a unique circuit that enables associative learning based on feedforward information flow.

mEMbrain: an interactive deep learning MATLAB tool for connectomic segmentation on commodity desktops.

Connectomics is fundamental in propelling our understanding of the nervous system’s organization, unearthing cells and wiring diagrams reconstructed from volume electron microscopy (EM) datasets. Such reconstructions, on the one hand, have benefited from ever more precise automatic segmentation methods, which leverage sophisticated deep learning architectures and advanced machine learning algorithms. On the other hand, the field of neuroscience at large, and of image processing in particular, has manifested a need for user-friendly and open source tools which enable the community to carry out advanced analyses. In line with this second vein, here we propose mEMbrain, an interactive MATLAB-based software which wraps algorithms and functions that enable labeling and segmentation of electron microscopy datasets in a user-friendly user interface compatible with Linux and Windows. Through its integration as an API to the volume annotation and segmentation tool VAST, mEMbrain encompasses functions for ground truth generation, image preprocessing, training of deep neural networks, and on-the-fly predictions for proofreading and evaluation. The final goals of our tool are to expedite manual labeling efforts and to harness MATLAB users with an array of semi-automatic approaches for instance segmentation. We tested our tool on a variety of datasets that span different species at various scales, regions of the nervous system and developmental stages. To further expedite research in connectomics, we provide an EM resource of ground truth annotation from 4 different animals and 5 datasets, amounting to around 180 hours of expert annotations, yielding more than 1.2 GB of annotated EM images. In addition, we provide a set of 4 pre-trained networks for said datasets. All tools are available from https://lichtman.rc.fas.harvard.edu/mEMbrain/ . With our software, our hope is to provide a solution for lab-based neural reconstructions which does not require coding by the user, thus paving the way to affordable connectomics.

Protocol for controlled behavioral testing of octopuses using a single-arm tactile discrimination two-choice task.

Due to their unique body, standard behavioral testing protocols are often hard to apply to octopuses. Our protocol enables controlled behavioral testing of the sensory systems in single arms while allowing observation of the arm motion. The protocol allows the researcher to exclude the sense of vision without surgical manipulation and selectively test peripheral sensory input-derived learning and motor behavior. Applying the protocol requires systematic and multistage training of octopuses to associate correct maze interaction with food reward. For complete details on the use and execution of this profile, please refer to Gutnick et al. (2020).

Neurotransmission and neuromodulation systems in the learning and memory network of Octopus vulgaris.

The vertical lobe (VL) in the octopus brain plays an essential role in its sophisticated learning and memory. Early anatomical studies suggested that the VL is organized in a "fan-out fan-in" connectivity matrix comprising only three morphologically identified neuron types; input axons from the median superior frontal lobe (MSFL) innervating en passant millions of small amacrine interneurons (AMs), which converge sharply onto large VL output neurons (LNs). Recent physiological studies confirmed the feedforward excitatory connectivity; a glutamatergic synapse at the first MSFL-to-AM synaptic layer and a cholinergic AM-to-LNs synapse. MSFL-to-AMs synapses show a robust hippocampal-like activity-dependent long-term potentiation (LTP) of transmitter release. 5-HT, octopamine, dopamine and nitric oxide modulate short- and long-term VL synaptic plasticity. Here, we present a comprehensive histolabeling study to better characterize the neural elements in the VL. We generally confirmed glutamatergic MSFLs and cholinergic AMs. Intense labeling for NOS activity in the AMs neurites were in-line with the NO-dependent presynaptic LTP mechanism at the MSFL-to-AM synapse. New discoveries here reveal more heterogeneity of the VL neurons than previously thought. GABAergic AMs suggest a subpopulation of inhibitory interneurons in the first input layer. Clear γ-amino butyric acid labeling in the cell bodies of LNs supported an inhibitory VL output, yet the LNs co-expressed FMRFamide-like neuropeptides, suggesting an additional neuromodulatory role of the VL output. Furthermore, a group of LNs was glutamatergic. A new cluster of cells organized as a "deep nucleus" showed rich catecholaminergic labeling and may play a role in intrinsic neuromodulation. In-situ hybridization and immunolabeling allowed characterization and localization of a rich array of neuropeptides and neuromodulators, likely involved in reward/punishment signals. This analysis of the fast transmission system, together with the newly found cellular elements, help integrate behavioral, physiological, pharmacological and connectome findings into a more comprehensive understanding of an efficient learning and memory network.

Use of Peripheral Sensory Information for Central Nervous Control of Arm Movement by Octopus vulgaris.

Octopuses are active predators with highly flexible bodies and rich behavioral repertoires [1-3]. They display advanced cognitive abilities, and the size of their large nervous system rivals that of many mammals. However, only one third of the neurons constitute the CNS, while the rest are located in an elaborate PNS, including eight arms, each containing myriad sensory receptors of various modalities [2-4]. This led early workers to question the extent to which the CNS is privy to non-visual sensory input from the periphery and to suggest that it has limited capacity to finely control arm movement [3-5]. This conclusion seemed reasonable considering the size of the PNS and the results of early behavioral tests [3, 6-8]. We recently demonstrated that octopuses use visual information to control goal-directed complex single arm movements [9]. However, that study did not establish whether animals use information from the arm itself [9-12]. We here report on development of two-choice, single-arm mazes that test the ability of octopuses to perform operant learning tasks that mimic normal tactile exploration behavior and require the non-peripheral neural circuitry to use focal sensory information originating in single arms [1, 10]. We show that the CNS of the octopus uses peripheral information about arm motion as well as tactile input to accomplish learning tasks that entail directed control of movement. We conclude that although octopus arms have a great capacity to act independently, they are also subject to central control, allowing well-organized, purposeful behavior of the organism as a whole.

Motor control pathways in the nervous system of Octopus vulgaris arm.

The octopus's arms have virtually infinite degrees of freedom, providing a unique opportunity for studying movement control in a redundant motor system. Here, we investigated the organization of the connections between the brain and arms through the cerebrobrachial tracts (CBT). To do this, we analyzed the neuronal activity associated with the contraction of a small muscle strand left connected at the middle of a long isolated CBT. Both electrical activity in the CBT and muscle contraction could be induced at low threshold values irrespective of stimulus direction and distance from the muscle strand. This suggests that axons associated with transmitting motor commands run along the CBT and innervate a large pool of motor neurons en passant. This type of innervation implies that central and peripheral motor commands involve the simultaneous recruitment of large groups of motor neurons along the arm as required, for example, in arm stiffening, and that the site of movement initiation along the arm may be determined through a unique interplay between global central commands and local sensory signals.

Evolution of highly diverse forms of behavior in molluscs.

Members of the phylum Mollusca demonstrate the animal kingdom's tremendous diversity of body morphology, size and complexity of the nervous system, as well as diversity of behavioral repertoires, ranging from very simple to highly flexible. Molluscs include Solenogastres, with their worm-like bodies and behavior (see phylogenetic tree; Figure 1); Bivalvia (mussels and clams), protected by shells and practically immobile; and the cephalopods, such as the octopus, cuttlefish and squid. The latter are strange-looking animals with nervous systems comprising up to half a billion neurons, which mediate the complex behaviors that characterize these freely moving, highly visual predators. Molluscs are undoubtedly special - their extraordinary evolutionary advance somehow managed to sidestep the acquisition of the rigid skeleton that appears essential to the evolution of other 'successful' phyla: the exoskeleton in ecdysozoan invertebrates and the internal skeleton in Deuterostomia, including vertebrates.

Pull or Push? Octopuses Solve a Puzzle Problem.

Octopuses have large brains and exhibit complex behaviors, but relatively little is known about their cognitive abilities. Here we present data from a five-level learning and problem-solving experiment. Seven octopuses (Octopus vulgaris) were first trained to open an L shaped container to retrieve food (level 0). After learning the initial task all animals followed the same experimental protocol, first they had to retrieve this L shaped container, presented at the same orientation, through a tight fitting hole in a clear Perspex partition (level 1). This required the octopuses to perform both pull and release or push actions. After reaching criterion the animals advanced to the next stage of the test, which would be a different consistent orientation of the object (level 2) at the start of the trial, an opaque barrier (level 3) or a random orientation of the object (level 4). All octopuses were successful in reaching criterion in all levels of the task. At the onset of each new level the performance of the animals dropped, shown as an increase in working times. However, they adapted quickly so that overall working times were not significantly different between levels. Our findings indicate that octopuses show behavioral flexibility by quickly adapting to a change in a task. This can be compared to tests in other species where subjects had to conduct actions comprised of a set of motor actions that cannot be understood by a simple learning rule alone.

Arm coordination in octopus crawling involves unique motor control strategies.

To cope with the exceptional computational complexity that is involved in the control of its hyper-redundant arms [1], the octopus has adopted unique motor control strategies in which the central brain activates rather autonomous motor programs in the elaborated peripheral nervous system of the arms [2, 3]. How octopuses coordinate their eight long and flexible arms in locomotion is still unknown. Here, we present the first detailed kinematic analysis of octopus arm coordination in crawling. The results are surprising in several respects: (1) despite its bilaterally symmetrical body, the octopus can crawl in any direction relative to its body orientation; (2) body and crawling orientation are monotonically and independently controlled; and (3) contrasting known animal locomotion, octopus crawling lacks any apparent rhythmical patterns in limb coordination, suggesting a unique non-rhythmical output of the octopus central controller. We show that this uncommon maneuverability is derived from the radial symmetry of the arms around the body and the simple pushing-by-elongation mechanism by which the arms create the crawling thrust. These two together enable a mechanism whereby the central controller chooses in a moment-to-moment fashion which arms to recruit for pushing the body in an instantaneous direction. Our findings suggest that the soft molluscan body has affected in an embodied way [4, 5] the emergence of the adaptive motor behavior of the octopus.

Octopus arm movements under constrained conditions: adaptation, modification and plasticity of motor primitives.

The motor control of the eight highly flexible arms of the common octopus (Octopus vulgaris) has been the focus of several recent studies. Our study is the first to manage to introduce a physical constraint to an octopus arm and investigate the adaptability of stereotypical bend propagation in reaching movements and the pseudo-limb articulation during fetching. Subjects (N=6) were placed inside a transparent Perspex box with a hole at the center that allowed the insertion of a single arm. Animals had to reach out through the hole toward a target, to retrieve a food reward and fetch it. All subjects successfully adjusted their movements to the constraint without an adaptation phase. During reaching tasks, the animals showed two movement strategies: stereotypical bend propagation reachings, which were established at the hole of the Perspex box and variant waving-like movements that showed no bend propagations. During fetching movements, no complete pseudo-joint fetching was observed outside the box and subjects pulled their arms through the hole in a pull-in like movement. Our findings show that there is some flexibility in the octopus motor system to adapt to a novel situation. However, at present, it seems that these changes are more an effect of random choices between different alternative motor programs, without showing clear learning effects in the choice between the alternatives. Interestingly, animals were able to adapt the fetching movements to the physical constraint, or as an alternative explanation, they could switch the motor primitive fetching to a different motor primitive 'arm pulling'.

Self-recognition mechanism between skin and suckers prevents octopus arms from interfering with each other.

Controlling movements of flexible arms is a challenging task for the octopus because of the virtually infinite number of degrees of freedom (DOFs) [1, 2]. Octopuses simplify this control by using stereotypical motion patterns that reduce the DOFs, in the control space, to a workable few [2]. These movements are triggered by the brain and are generated by motor programs embedded in the peripheral neuromuscular system of the arm [3-5]. The hundreds of suckers along each arm have a tendency to stick to almost any object they contact [6-9]. The existence of this reflex could pose significant problems with unplanned interactions between the arms if not appropriately managed. This problem is likely to be accentuated because it is accepted that octopuses are "not aware of their arms" [10-14]. Here we report of a self-recognition mechanism that has a novel role in motor control, restraining the arms from interfering with each other. We show that the suckers of amputated arms never attach to octopus skin because a chemical in the skin inhibits the attachment reflex of the suckers. The peripheral mechanism appears to be overridden by central control because, in contrast to amputated arms, behaving octopuses sometime grab amputated arms. Surprisingly, octopuses seem to identify their own amputated arms, as they treat arms of other octopuses like food more often than their own. This self-recognition mechanism is a novel peripheral component in the embodied organization of the adaptive interactions between the octopus's brain, body, and environment [15, 16].

The sea anemone toxin AdE-1 modifies both sodium and potassium currents of rat cardiomyocytes.

AdE-1, a cardiotonic peptide recently isolated from the sea anemone Aiptasia diaphana, contains 44 amino acids and has a molecular mass of 4907 Da. It was previously found to resemble other sea anemone type 1 and 2 Na+ channel toxins, enhancing contractions of rat cardiomyocytes and slowing their twitch relaxation; however, it did not induce spontaneous twitches. AdE-1 increased the duration of the cardiomyocyte action potential and decreased its amplitude and its time-to-peak in a concentration-dependent manner, without affecting its threshold and cell resting potential. Nor did it generate the early and delayed after-depolarizations characteristic of sea anemone Na+ channel toxins. To further understand its mechanism of action we investigated the effect of AdE-1 on the major ion currents of rat cardiomyocytes. In the present study we show that AdE-1 markedly slowed inactivation of the Na+ current, enhancing and prolonging the current influx with no effect on current activation, possibly through direct interaction with the site 3 receptor of the Na+ channel. No significant effect of AdE-1 on the Ca2+ current was observed, but, unexpectedly, AdE-1 significantly increased the amplitude of the transient component of the K+ current, shifting the current threshold to more negative membrane potentials. This effect on the K+ current has not been found in any other sea anemone toxin and may explain the exclusive reduction in action potential amplitude and the absence of the action potential disorders found with other toxins, such as early and delayed after-depolarizations.

How nervous systems evolve in relation to their embodiment: what we can learn from octopuses and other molluscs.

Cephalopods such as the octopus show the most advanced behavior among invertebrates, which they accomplish with an exceptionally flexible body plan. In this review I propose that the embodied organization approach, developed by roboticists to design efficient autonomous robots, is useful for understanding the evolution and development of the efficient adaptive interaction of animals with their environment, using the octopus as the leading example. The embodied organization approach explains adaptive behavior as emerging from the continuous dynamical and reciprocal physical and informational interactions between four elements: the controller, the mechanical and the sensory systems and the environment. In contrast to hierarchical organization, in embodied organization, self-organization processes can take part in the emergence of the adaptive properties. I first discuss how the embodiment concept explains covariation of body form, nervous system organization, and level of behavioral complexity using the Mollusca as an example. This is an ideal phylum to test such a qualitative correlation between body/brain/behavior, because they show the greatest variations of body plan within a single phylum. In some cases the covariation of nervous system and body structure seems to arise independently of close phylogenetic relationships. Next, I dwell on the octopus as an ideal model to test the embodiment concept within a single biological system. Here, the unusual body morphology of the octopus exposes the uniqueness of the four components comprising the octopus' embodiment. Considering together the results from behavioral, physiological, anatomical, and motor control research suggests that these four elements mutually influence each other. It is this mutual interactions and self-organization which have led to their unique evolution and development to create the unique and highly efficient octopus embodiment.

Kinematic decomposition and classification of octopus arm movements.

The octopus arm is a muscular hydrostat and due to its deformable and highly flexible structure it is capable of a rich repertoire of motor behaviors. Its motor control system uses planning principles and control strategies unique to muscular hydrostats. We previously reconstructed a data set of octopus arm movements from records of natural movements using a sequence of 3D curves describing the virtual backbone of arm configurations. Here we describe a novel representation of octopus arm movements in which a movement is characterized by a pair of surfaces that represent the curvature and torsion values of points along the arm as a function of time. This representation allowed us to explore whether the movements are built up of elementary kinematic units by decomposing each surface into a weighted combination of 2D Gaussian functions. The resulting Gaussian functions can be considered as motion primitives at the kinematic level of octopus arm movements. These can be used to examine underlying principles of movement generation. Here we used combination of such kinematic primitives to decompose different octopus arm movements and characterize several movement prototypes according to their composition. The representation and methodology can be applied to the movement of any organ which can be modeled by means of a continuous 3D curve.

AdE-1, a new inotropic Na(+) channel toxin from Aiptasia diaphana, is similar to, yet distinct from, known anemone Na(+) channel toxins.

Heart failure is one of the most prevalent causes of death in the western world. Sea anemone contains a myriad of short peptide neurotoxins affecting many pharmacological targets, several of which possess cardiotonic activity. In the present study we describe the isolation and characterization of AdE-1 (ion channel modifier), a novel cardiotonic peptide from the sea anemone Aiptasia diaphana, which differs from other cnidarian toxins. Although AdE-1 has the same cysteine residue arrangement as sea anemone type 1 and 2 Na(+) channel toxins, its sequence contains many substitutions in conserved and essential sites and its overall homology to other toxins identified to date is low (<36%). Physiologically, AdE-1 increases the amplitude of cardiomyocyte contraction and slows the late phase of the twitch relaxation velocity with no induction of spontaneous twitching. It increases action potential duration of cardiomyocytes with no effect on its threshold and on the cell's resting potential. Similar to other sea anemone Na(+) channel toxins such as Av2 (Anemonia viridis toxin II), AdE-1 markedly inhibits Na(+) current inactivation with no significant effect on current activation, suggesting a similar mechanism of action. However, its effects on twitch relaxation velocity, action potential amplitude and on the time to peak suggest that this novel toxin affects cardiomyocyte function via a more complex mechanism. Additionally, Av2's characteristic delayed and early after-depolarizations were not observed. Despite its structural differences, AdE-1 physiologic effectiveness is comparable with Av2 with a similar ED(50) value to blowfly larvae. This finding raises questions regarding the extent of the universality of structure-function in sea anemone Na(+) channel toxins.

An embodied view of octopus neurobiology.

Octopuses have a unique flexible body and unusual morphology, but nevertheless they are undoubtedly a great evolutionary success. They compete successfully with vertebrates in their ecological niche using a rich behavioral repertoire more typical of an intelligent predator which includes extremely effective defensive behavior--fast escape swimming and an astonishing ability to adapt their shape and color to their environment. The most obvious characteristic feature of an octopus is its eight long and flexible arms, but these pose a great challenge for achieving the level of motor and sensory information processing necessary for their behaviors. First, coordinating motion is a formidable task because of the infinite degrees of freedom that have to be controlled; and second, it is hard to use body coordinates in this flexible animal to represent sensory information in a central control system. Here I will review experimental results suggesting that these difficulties, arising from the animal's morphology, have imposed the evolution of unique brain/body/behavior relationships best explained as intelligent behavior which emerges from the octopus's embodied organization. The term 'intelligent embodiment' comes from robotics and refers to an approach to designing autonomous robots in which the behavior emerges from the dynamic physical and sensory interactions of the agent's materials, morphology and environment. Consideration of the unusual neurobiology of the octopus in the light of its unique morphology suggests that similar embodied principles are instrumental for understanding the emergence of intelligent behavior in all biological systems.

Alternative sites of synaptic plasticity in two homologous "fan-out fan-in" learning and memory networks.

BACKGROUND: To what extent are the properties of neuronal networks constrained by computational considerations? Comparative analysis of the vertical lobe (VL) system, a brain structure involved in learning and memory, in two phylogenetically close cephalopod mollusks, Octopus vulgaris and the cuttlefish Sepia officinalis, provides a surprising answer to this question. RESULTS: We show that in both the octopus and the cuttlefish the VL is characterized by the same simple fan-out fan-in connectivity architecture, composed of the same three neuron types. Yet, the sites of short- and long-term synaptic plasticity and neuromodulation are different. In the octopus, synaptic plasticity occurs at the fan-out glutamatergic synaptic layer, whereas in the cuttlefish plasticity is found at the fan-in cholinergic synaptic layer. CONCLUSIONS: Does this dramatic difference in physiology imply a difference in function? Not necessarily. We show that the physiological properties of the VL neurons, particularly the linear input-output relations of the intermediate layer neurons, allow the two different networks to perform the same computation. The convergence of different networks to the same computational capacity indicates that it is the computation, not the specific properties of the network, that is self-organized or selected for by evolutionary pressure.