Regression-based identification of behavior-encoding neurons during large-scale optical imaging of neural activity at cellular resolution

The advent of methods for optical imaging of large-scale neural activity at cellular resolution in behaving animals presents the problem of identifying behavior-encoding cells within the resulting image time series. Rapid and precise identification of cells with particular neural encoding would facilitate targeted activity measurements and perturbations useful in characterizing the operating principles of neural circuits. Here we report a regression-based approach to semiautomatically identify neurons that is based on the correlation of fluorescence time series with quantitative measurements of behavior. The approach is illustrated with a novel preparation allowing synchronous eye tracking and two-photon laser scanning fluorescence imaging of calcium changes in populations of hindbrain neurons during spontaneous eye movement in the larval zebrafish. Putative velocity-to-position oculomotor integrator neurons were identified that showed a broad spatial distribution and diversity of encoding. Optical identification of integrator neurons was confirmed with targeted loose-patch electrical recording and laser ablation. The general regression-based approach we demonstrate should be widely applicable to calcium imaging time series in behaving animals.     

Medial temporal lobe dysfunction during encoding and retrieval of episodic memory in non-demented APOE ε4 carriers

Presence of the apolipoprotein E (APOE) ε4 allele is linked to an increased risk to develop Alzheimer's dementia (AD). However, there are controversial data concerning the impact of the APOE genotype on cognitive functioning and brain activity in healthy subjects. We used event-related functional magnetic resonance imaging (fMRI) to investigate the effects of APOE genotype on spatial contextual memory encoding and retrieval success in healthy older adults. Eighteen subjects (eight APOE4 heterozygotes (ε4+) and 10 non-carriers (ε4−), mean age 60.0±5.0 years) were included in the present analysis. Behaviorally, ε4+ subjects performed significantly worse than ε4− subjects in item memory and spatial context retrieval. fMRI data revealed that ε4+ subjects, compared to ε4-subjects, predominantly showed an increase of neural activity specific to encoding of items and their spatial context in prefrontal, temporal and parietal regions. In contrast, ε4+ subjects showed activity decreases in the right amygdala during successful item recognition and in the prefrontal cortex bilaterally during spatial context retrieval when compared to ε4− subjects. While the activity increases during encoding may reflect compensatory activity in the attempt to maintain normal performance, the decreases during retrieval indicate incipient neural decline in ε4+ subjects. These data highlight that preclinical ApoE-related changes in neural activity are not unidirectional but dissociate depending on the memory phase, i.e., encoding or retrieval.     

Spatial attention enhances the selective integration of activity from area MT

Distinguishing which of the many proposed neural mechanisms of spatial attention actually underlies behavioral improvements in visually guided tasks has been difficult. One attractive hypothesis is that attention allows downstream neural circuits to selectively integrate responses from the most informative sensory neurons. This would allow behavioral performance to be based on the highest-quality signals available in visual cortex. We examined this hypothesis by asking how spatial attention affects both the stimulus sensitivity of middle temporal (MT) neurons and their corresponding correlation with behavior. Analyzing a data set pooled from two experiments involving four monkeys, we found that spatial attention did not appreciably affect either the stimulus sensitivity of the neurons or the correlation between their activity and behavior. However, for those sessions in which there was a robust behavioral effect of attention, focusing attention inside the neuron's receptive field significantly increased the correlation between these two metrics, an indication of selective integration. These results suggest that, similar to mechanisms proposed for the neural basis of perceptual learning, the behavioral benefits of focusing spatial attention are attributable to selective integration of neural activity from visual cortical areas by their downstream targets.     

Manoeuvres Affecting Sympathetic Outflow in Human Muscle Nerves

Multi-unit sympathetic activity was recorded in human muscle nerves, together with measurements of intra-arterial blood pressure and forearm or calf blood flow, during manoeuvres causing circulatory adjustments. Manoeuvres causing an increased vascular resistance in the forearm or calf were regularly associated with an increase in sympathetic outflow, proving that the neural activity was dominated by vasoconstrictor impulses. The inverse changes of blood pressure and sympathetic activity observed during Valsalva's manoeuvre and mental stress are explicable in terms of baroreflex control of the neural outflow. However, during muscle work and hyperventilation the relation between blood pressure and sympathetic activity was more complex, indicating that the baroreflex influence was superimposed or modified by other regulatory mechanisms. Some manoeuvres known to affect sympathetic outflow in skin nerves caused no appreciable change in sympathetic muscle nerve activity.     

Ageing-related changes of neural activity associated with spatial contextual memory

Neuropsychological studies provide evidence for an ageing-related decline of memory for contextual information related to remembered items. Using event-related fMRI we investigated the neural correlates of ageing-related changes during encoding and retrieval of spatial contextual memory. Eighteen young and 17 older subjects were included in the analysis (mean age 24 and 60 years, respectively). Although young and older subjects recognised the same amount of items during retrieval, spatial context memory for remembered items was superior in younger subjects. In both groups, left parahippocampal activity during encoding predicted contextual memory performance during retrieval. During encoding, an interaction between age and success of spatial context encoding was found in the left fusiform gyrus. During retrieval, the left hippocampal formation showed higher activity for successful than for unsuccessful spatial context retrieval as well as an interaction between age and spatial context judgement. Both findings are likely to underlie the contextual memory deficit observed in older subjects.     

Sex-dependent changes in anxiety, memory, and monoamines following one week of stress

Chronic restraint stress alters performance of rats on cognitive tasks, and anxiety measurements, and these stress-induced behavioral alterations are sexually dimorphic. Following a long stress period (21 days restraint) males show cognitive impairments while females are either not affected or enhanced on the same tasks. The current study examined whether sexually differentiated responses are also induced following shorter restraint stress durations. Male and female Sprague Dawley rats, aged 2.5 months, served as controls or received restraint stress (6 h/day, 7 days) and were tested for anxiety (plus maze), non-spatial memory (object recognition), and spatial memory (object placement). Plus maze performance was altered by sex and stress exposure. Stress impaired male object recognition but did not affect female performance. Stress did not affect male spatial memory; however, control females could not significantly discriminate between the old and new locations, but stress exposure enhanced female performance. Following behavioral testing, monoamines and metabolites were measured in prefrontal cortex (PFC), hippocampus (CA1, CA3), and amygdala. Notably, PFC and CA3 indices for noradrenergic activity (MHPG levels and MHPG/NE ratios) were increased in stress females, but decreased in males, and similar changes were found in CA1 and BLA dopaminergic indices. Thus, these sexually dimorphic neurochemical changes following stress may underlie the behavioral differences. Current results show that short-term restraint elicits sex-dependent behavioral and neural changes different from those previously reported for longer term stresses and suggest that the temporal relationship between the change from adaptive to maladaptive responses to stress is shorter in male than female rats.     

Neural basis and recovery of spatial attention deficits in spatial neglect

The syndrome of spatial neglect is typically associated with focal injury to the temporoparietal or ventral frontal cortex. This syndrome shows spontaneous partial recovery, but the neural basis of both spatial neglect and its recovery is largely unknown. We show that spatial attention deficits in neglect (rightward bias and reorienting) after right frontal damage correlate with abnormal activation of structurally intact dorsal and ventral parietal regions that mediate related attentional operations in the normal brain. Furthermore, recovery of these attention deficits correlates with the restoration and rebalancing of activity within these regions. These results support a model of recovery based on the re-weighting of activity within a distributed neuronal architecture, and they show that behavioral deficits depend not only on structural changes at the locus of injury, but also on physiological changes in distant but functionally related brain areas.     

The effect of early astigmatism on the visual resolution of gratings

1. Orientational differences in visual resolution were measured at a number of different luminance levels on two subjects with high astigmatism that had remained optically uncorrected until the age of 10. Because of their astigmatism both of these subjects see vertical contours more clearly than horizontal contours with the unaided eye.2. The measurements were made using sinusoidal gratings generated on the face of an oscilloscope with the refractive error carefully corrected with lenses and with the gratings viewed through 3 mm artificial pupils.3. Visual resolution was found to be much better for vertical than for horizontal gratings for both these subjects under these conditions. The difference between the contrast sensitivities for vertical and horizontal gratings was even evident with gratings having spatial frequencies as low as 1 c/deg, but became progressively more pronounced at higher spatial frequencies. In one of the subjects the visual acuity (the cut-off spatial frequency) for horizontal gratings was more than 3/4 of an octave lower than that for vertical gratings.4. This is very different from the results obtained from normal subjects who typically show only a slight reduction in contrast sensitivity for oblique gratings but resolve vertical and horizontal gratings equally well.5. The quantitative differences between the contrast sensitivities for vertical and horizontal gratings of both high and low spatial frequencies cannot be accounted for by either errors of focus in one meridian or by the presence of meridional aniseikonia.6. In order to completely eliminate any optical explanations for these findings measurements of contrast sensitivity were made using sinusoidal interference fringes formed directly on the retina, thereby bypassing the eye's optics. Since the orientational differences in resolution persisted with this method it must be concluded that they are of neural origin.7. By analogy with the effects on cortical physiology that follow early selective visual deprivation in cats and monkeys, it is argued that these orientational differences in resolution are a consequence of changes induced in the neural organization of the astigmat's visual system by the distorted visual input provided by the uncorrected astigmatism early in life. It is furthermore argued that the smaller orientational differences in resolution observed in normal eyes might similarly be induced by certain asymmetries in the early visual input.     

Modalities of Thinking: State and Trait Effects on Cross-Frequency Functional Independent Brain Networks

Functional states of the brain are constituted by the temporally attuned activity of spatially distributed neural networks. Such networks can be identified by independent component analysis (ICA) applied to frequency-dependent source-localized EEG data. This methodology allows the identification of networks at high temporal resolution in frequency bands of established location-specific physiological functions. EEG measurements are sensitive to neural activity changes in cortical areas of modality-specific processing. We tested effects of modality-specific processing on functional brain networks. Phasic modality-specific processing was induced via tasks (state effects) and tonic processing was assessed via modality-specific person parameters (trait effects). Modality-specific person parameters and 64-channel EEG were obtained from 70 male, right-handed students. Person parameters were obtained using cognitive style questionnaires, cognitive tests, and thinking modality self-reports. EEG was recorded during four conditions: spatial visualization, object visualization, verbalization, and resting. Twelve cross-frequency networks were extracted from source-localized EEG across six frequency bands using ICA. RMANOVAs, Pearson correlations, and path modelling examined effects of tasks and person parameters on networks. Results identified distinct state- and trait-dependent functional networks. State-dependent networks were characterized by decreased, trait-dependent networks by increased alpha activity in sub-regions of modality-specific pathways. Pathways of competing modalities showed opposing alpha changes. State- and trait-dependent alpha were associated with inhibitory and automated processing, respectively. Antagonistic alpha modulations in areas of competing modalities likely prevent intruding effects of modality-irrelevant processing. Considerable research suggested alpha modulations related to modality-specific states and traits. This study identified the distinct electrophysiological cortical frequency-dependent networks within which they operate.     

Temporally staggered forelimb stimulation modulates barrel cortex optical intrinsic signal responses to whisker stimulation

Characterization of neurovascular relationships is critical to accurate interpretation of functional neuroimaging data. We have previously observed spatial uncoupling of optical intrinsic signal imaging (OIS) and evoked potential (EP) responses in rodent barrel cortex following simultaneous whisker and forelimb stimulation, leading to changes in OIS response magnitude. To further test the hypothesis that this uncoupling may have resulted from "passive" overspill of perfusion-related responses between functional regions, we conducted the present study using temporally staggered rather than simultaneous whisker and forelimb stimulation. This paradigm minimized overlap of neural responses in barrel cortex and forelimb primary somatosensory cortex (SI), while maintaining overlap of vascular response time courses between regions. When contrasted with responses to 1.5-s lone-whisker stimulation, staggered whisker and forelimb stimulation resulted in broadening of barrel cortex OIS response time course in the temporal direction of forelimb stimulation. OIS response peaks were also temporally shifted toward the forelimb stimulation period; time-to-peak was shorter (relative to whisker stimulus onset) when forelimb stimulation preceded whisker stimulation and longer when forelimb stimulation followed whisker stimulation. In contrast with OIS and EP magnitude decreases previously observed during simultaneous whisker/forelimb stimulation, barrel cortex OIS response magnitude increased during staggered stimulation and no detectable changes in underlying EP activity were observed. Spatial extent of barrel cortex OIS responses also increased during staggered stimulation. These findings provide further evidence for spatial uncoupling of OIS and EP responses, and emphasize the importance of temporal stimulus properties on the effects of this uncoupling. It is hypothesized that spatial uncoupling is a result of passive overspill of perfusion-related responses into regions distinct from those which are functionally active. It will be important to consider potential influences of this uncoupling when designing and interpreting functional imaging studies that use hemodynamic responses to infer underlying neural activity.     

Self-mixing microprobe for monitoring microvascular perfusion in rat brain

Measuring functional activity in brain in connection with neural stimulation faces technological challenges. Our goal is to evaluate, in relative terms, the real-time variations of local cerebral blood flow in rat brain, with a convenient spatial resolution. The use of laser Doppler flowmetry (LDF) probes is a promising approach but commercially available LDF probes are still too large (450 μm) to allow insertion in brain tissue without causing damage in an extension that may negatively impact local measurements. The self-mixing technique coupled to LDF is herein proposed to overcome limitations of the minimal diameter of the probe imposed by non-self-mixing probes (commercial available probes). Our Monte Carlo simulations show that laser photons have a mean penetration depth of 0.15 mm, on the rat brain with the 785 nm laser light microprobe. Moreover, three self-mixing signal processing methods are tested: counting method, autocorrelation method, power spectrum method. The perfusion signal computed shows a good linearity with the scatterers velocity, for the three methods (a determination coefficient close to one is obtained), for the in vitro measurements. Furthermore, we believe that these indicators can be used to monitor local blood flow changes in the rat brain.     

Neural firing in ventrolateral thalamic nucleus during conditioned vocal behavior in cats

 The ventrolateral (VL) thalamus in mammals is a site well-situated to show vocalization-related neural activity if there is general or classical motor system involvement in vocal production. It receives input from both the basal ganglia and cerebellum, and forms reciprocal connections with motor cortical areas. The current study examined the activity in cat VL thalamus neurons during instrumentally conditioned vocalization. Units in our sample showed irregular spontaneous firing which could be modulated by slowly occurring fluctuations in intensity of vocalization task performance. Two main types of behavioral events were associated with changes in neural firing rate. The first of these was the ingestion of food reward. More than half of all recordings showed phasic bursting patterns during licking; a similar number had increases in firing preparatory to this phasic activity. The second behavioral event modulating unit responses was vocalization. Approximately 60% of recordings showed activity changes time-locked to vocalization. These responses were almost always excitatory, and often involved changes in firing that preceded vocalization onset. No spatial organization of differences in firing pattern between neurons could be distinguished. Our results suggest that VL thalamus may well be involved in mediating vocal behavior, although its functional role remains an object of speculation. Results are compared with previous studies of vocalization-related activity and of VL thalamus activity. Received: 15 November 1995 / Accepted: 16 December 1996     

The event-related optical signal: a new tool for studying brain function.

This paper presents an overview of a new method for the non-invasive measurement of brain function, the event-related optical signal (EROS). This technique is based on measures of the optical properties of cortical brain tissue, which change while the tissue is active. These changes are likely to be due to changes in light scattering, and are very rapid and localized, being related to phenomena occurring within or around the neuronal membrane. EROS, therefore yields images of cortical activity that combine spatial specificity (i.e. they can be related to patches of tissue less than a cubic centimeter in size) with temporal resolution (i.e. they depict the time course of the neural activity in the cortical areas under measurement). A limitation of this technique is its reduced penetration into the head (less than 3-5 cm). EROS appears to be a suitable technique for studying the time course of activity in selected cortical areas, and for providing a bridge between hemodynamic and electrophysiological imaging methods.     

Detecting neural-state transitions using hidden Markov models for motor cortical prostheses

Neural prosthetic interfaces use neural activity related to the planning and perimovement epochs of arm reaching to afford brain-directed control of external devices. Previous research has primarily centered on accurately decoding movement intention from either plan or perimovement activity, but has assumed that temporal boundaries between these epochs are known to the decoding system. In this work, we develop a technique to automatically differentiate between baseline, plan, and perimovement epochs of neural activity. Specifically, we use a generative model of neural activity to capture how neural activity varies between these three epochs. Our approach is based on a hidden Markov model (HMM), in which the latent variable (state) corresponds to the epoch of neural activity, coupled with a state-dependent Poisson firing model. Using an HMM, we demonstrate that the time of transition from baseline to plan epochs, a transition in neural activity that is not accompanied by any external behavior changes, can be detected using a threshold on the a posteriori HMM state probabilities. Following detection of the plan epoch, we show that the intended target of a center-out movement can be detected about as accurately as that by a maximum-likelihood estimator using a window of known plan activity. In addition, we demonstrate that our HMM can detect transitions in neural activity corresponding to targets not found in training data. Thus the HMM technique for automatically detecting transitions between epochs of neural activity enables prosthetic interfaces that can operate autonomously.     

The relationship between voltage-sensitive dye imaging signals and spiking activity of neural populations in primate V1

Voltage-sensitive dye imaging (VSDI) is a powerful technique for measuring neural population responses from a large cortical region simultaneously with millisecond temporal resolution and columnar spatial resolution. However, the relationship between the average VSDI signal and the average spiking activity of neural populations is largely unknown. To better understand this relationship, we compared visual responses measured from V1 of behaving monkeys using VSDI and single-unit electrophysiology. We found large and systematic differences between position and orientation tuning properties obtained with these two techniques. We then determined that a simple computational model could explain these tuning differences. This model, together with our experimental results, allowed us to estimate the quantitative relationship between the average VSDI signal and local spiking activity. We found that this relationship is similar to the previously reported nonlinear relationship between average membrane potential and spike rate in single V1 neurons, suggesting that VSDI signals are dominated by subthreshold synaptic activity. This model, together with the VSDI measured maps for spatial position (retinotopy) and orientation, also allowed us to estimate the spatial integration area over which neural responses contribute to the VSDI signal at a given location. We found that the VSDI-integration area is consistent with a Gaussian envelope with a space constant of ～230 μm. Finally, we show how this model and estimated parameters can be used to predict the pattern of population responses at the level of spiking activity from VSDI responses.     

Sequential neural processes of tactile-visual crossmodal working memory

Working memory is essential to learning and performing sensory-motor behaviors that in many situations require the integration of stimuli of one modality with stimuli of another. In the present study, we focused on the neural mechanisms underlying crossmodal working memory. We hypothesized that in performance of the tactile crossmodal working memory task, there would be sequentially discrete task-correlated neural activities representing the processes of crossmodal working memory. Scalp-recorded event-related potentials were collected from 15 electrodes in humans performing each of four tasks: tactile-tactile unimodal delayed matching-to-sample task, tactile-visual crossmodal delayed matching-to-sample task, tactile unimodal control spatial task, and tactile crossmodal control spatial task. Two positive event-related potential peaks were observed during the delay of the task. One peak (late positive component-1) was at about 330 ms after the onset of the tactile stimulus, and the other (late positive component-2) was at about 600 ms. Late positive component-1 was observed in all four tasks. There was no significant difference in late positive component-1 either between the unimodal tasks, or between the crossmodal tasks, but late positive component-1 was significantly larger in the crossmodal tasks than that in the unimodal tasks, and showed a specific pattern of larger activity over parietal areas than activity over frontal areas. Late positive component-2 was not observed in the unimodal matching task but was observed in all other three tasks over parietal areas. During the late delay (1000 ms-1500 ms), significant differences in negative potentials (late negative component) were found between the tasks. The present study shows sequential changes in event-related potentials during the retention period of working memory tasks. It indicates that in performance of a crossmodal working memory task, there are sequentially discrete neural processes that may represent neural activities related to different cognitive functions, such as crossmodal transfer of information, and the working memory of the stimulus.     

N-methyl-D-aspartate and dopamine receptor involvement in the modulation of locomotor activity and memory processes

In this study we report on the effects of N-methyl-d-aspartate (NMDA)- and dopamine (DA)-receptor manipulation on the modulation of one-trial inhibitory avoidance response and the encoding of spatial information, as assessed with a non-associative task. Further, a comparison with the well-known effects of the manipulation of these two receptor systems on locomotor activity is outlined. It is well assessed that NMDA-receptor blockage induces a stimulatory action on locomotor activity similar to that exerted by DA agonists. There is evidence showing that the nucleus accumbens is involved in the response induced by both NMDA antagonists and DA agonists. We show results indicating a functional interaction between these two neural systems in modulating locomotor activity, with D2 DA-receptor antagonists (sulpiride and haloperidol) being more effective than the D1 antagonist (SCH 23390) in blocking MK-801-induced locomotion. A different profile is shown in the effects of NMDA antagonists and DA agonists in the modulation of memory processes. In one-trial inhibitory avoidance response, NMDA antagonists (MK-801 and CPP) impair the response on test day, while DA agonists exert a facilitatory effect; furthermore, sub-effective doses of both D1 (SKF 23390) and D2 (quinpirole) are able to attenuate the impairing effect in a way similar to that induced by NMDA antagonists. The effects of NMDA- and DA-acting drugs on the response to spatial novelty, as assessed with a task designed to study the ability of animals to react to discrete spatial changes, are in good accord with the effects observed on passive avoidance. The results show that NMDA as well as DA antagonists, at low doses, selectively impair the reactivity of mice to spatial changes. In a last series of experiments, the possible role of NMDA receptors located in the nucleus accumbens was investigated regarding reactivity to spatial novelty. The experiments gave apparently contrasting results: while showing an impairing effect of focal administrations of NMDA antagonists in the nucleus accumbens on reactivity to spatial novelty, no effect of ibotenic acid lesions of the same structure was observed.     

Network mechanisms of hippocampal laterality,place coding,and goal-directed navigation

The hippocampus and associated structures are responsible for episodic memory in humans. In rodents, the most prominent behavioral correlate of hippocampal neural activity is place coding, which is thought to underlie spatial navigation. While episodic memory is considered to be unique to humans in a restricted context, it has been proposed that the same neural circuitry and algorithms that enable spatial coding and navigation also support episodic memory. Here we review the recent progress in neural circuit mechanisms of hippocampal activity by introducing several topics: (1) cooperation and specialization of the bilateral hippocampi, (2) the role of synaptic plasticity in gamma phase-locking of spikes and place cell formation, (3) impaired goal-related activity and oscillations in a mouse model of mental disorders, and (4) a prefrontal-thalamo-hippocampal circuit for goal-directed spatial navigation.     

Spatiotemporal scales and links between electrical neuroimaging modalities

Recordings of brain electrophysiological activity provide the most direct reflect of neural function. Information contained in these signals varies as a function of the spatial scale at which recordings are done: from single cell recording to large scale macroscopic fields, e.g., scalp EEG. Microscopic and macroscopic measurements and models in Neuroscience are often in conflict. Solving this conflict might require the developments of a sort of bio-statistical physics, a framework for relating the microscopic properties of individual cells to the macroscopic or bulk properties of neural circuits. Such a framework can only emerge in Neuroscience from the systematic analysis and modeling of the diverse recording scales from simultaneous measurements. In this article we briefly review the different measurement scales and models in modern neuroscience to try to identify the sources of conflict that might ultimately help to create a unified theory of brain electromagnetic fields. We argue that seen the different recording scales, from the single cell to the large scale fields measured by the scalp electroencephalogram, as derived from a unique physical magnitude—the electric potential that is measured in all cases—might help to conciliate microscopic and macroscopic models of neural function as well as the animal and human neuroscience literature.     

Measures of Spatial Similarity and Response Magnitude in MEG and Scalp EEG

Sensor selection is typically used in magnetoencephalography (MEG) and scalp electroencephalography (EEG) studies, but this practice cannot differentiate between changes in the distribution of neural sources versus changes in the magnitude of neural sources. This problem is further complicated by (1) subject averaging despite sizable individual anatomical differences and (2) experimental designs that produce overlapping waveforms due to short latencies between stimuli. Using data from the entire spatial array of sensors, we present simple multivariate measures that (1) normalize against individual differences by comparison with each individual’s standard response; (2) compare the similarity of spatial patterns in different conditions (angle test) to ascertain whether the distribution of neural sources is different; and (3) compare the response magnitude between conditions which are sufficiently similar (projection test). These claims are supported by applying the reported techniques to a short-term word priming paradigm as measured with MEG, revealing more reliable results as compared to traditional sensor selection methodology. Although precise cortical localization remains intractable, these techniques are easy to calculate, relatively assumption free, and yield the important psychological measures of similarity and response magnitude.