Two-Color Volumetric Imaging of Neuronal Activity of Cortical Columns

Behavior-relevant top-down cross-modal predictions in mouse neocortex

Animals adapt to a constantly changing world by predicting their environment and the consequences of their actions. The predictive coding hypothesis proposes that the brain generates predictions and continuously compares them with sensory inputs to guide behavior. However, how the brain reconciles conflicting top-down predictions and bottom-up sensory information remains unclear. To address this question, we simultaneously imaged neuronal populations in the mouse somatosensory barrel cortex and posterior parietal cortex during an auditory-cued texture discrimination task. In mice that had learned the task with fixed tone-texture matching, the presentation of mismatched pairing induced conflicts between tone-based texture predictions and actual texture inputs. When decisions were based on the predicted rather than the actual texture, top-down information flow was dominant and texture representations in both areas were modified, whereas dominant bottom-up information flow led to correct representations and behavioral choice. Our findings provide evidence for hierarchical predictive coding in the mouse neocortex.

Recent strategies for neural dynamics observation at a larger scale and wider scope

The tremendous technical progress in neuroscience offers opportunities to observe a more minor or/and broader dynamic picture of the brain. Moreover, the large-scale neural activity of individual neurons enables the dissection of detailed mechanistic links between neural populations and behaviors. To measure neural activity in-vivo, multi-neuron recording, and neuroimaging techniques are employed and developed to acquire more neurons. The tools introduced concurrently recorded dozens to hundreds of neurons in the coordinated brain regions and elucidated the neuronal ensembles from a massive population perspective of diverse neurons at cellular resolution. In particular, the increasing spatiotemporal resolution of neuronal monitoring across the whole brain dramatically facilitates our understanding of additional nervous system functions in health and disease. Here, we will introduce state-of-the-art neuroscience tools involving large-scale neural population recording and the long-range connections spanning multiple brain regions. Their synergic effects provide to clarify the controversial circuitry underlying neuroscience. These challenging neural tools present a promising outlook for the fundamental dynamic interplay across levels of synaptic cellular, circuit organization, and brain-wide. Hence, more observations of neural dynamics will provide more clues to elucidate brain functions and push forward innovative technology at the intersection of neural engineering disciplines. We hope this review will provide insight into the use or development of recent neural techniques considering spatiotemporal scales of brain observation.

Practical considerations in an era of multicolor optogenetics

The ability to control synaptic communication is indispensable to modern neuroscience. Until recently, only single-pathway manipulations were possible due to limited availability of opsins activated by distinct wavelengths. However, extensive protein engineering and screening efforts have drastically expanded the optogenetic toolkit, ushering in an era of multicolor approaches for studying neural circuits. Nonetheless, opsins with truly discrete spectra are scarce. Experimenters must therefore take care to avoid unintended cross-activation of optogenetic tools (crosstalk). Here, we demonstrate the multidimensional nature of crosstalk in a single model synaptic pathway, testing stimulus wavelength, irradiance, duration, and opsin choice. We then propose a "lookup table" method for maximizing the dynamic range of opsin responses on an experiment-by-experiment basis.

Learning cortical hierarchies with temporal Hebbian updates

A key driver of mammalian intelligence is the ability to represent incoming sensory information across multiple abstraction levels. For example, in the visual ventral stream, incoming signals are first represented as low-level edge filters and then transformed into high-level object representations. Similar hierarchical structures routinely emerge in artificial neural networks (ANNs) trained for object recognition tasks, suggesting that similar structures may underlie biological neural networks. However, the classical ANN training algorithm, backpropagation, is considered biologically implausible, and thus alternative biologically plausible training methods have been developed such as Equilibrium Propagation, Deep Feedback Control, Supervised Predictive Coding, and Dendritic Error Backpropagation. Several of those models propose that local errors are calculated for each neuron by comparing apical and somatic activities. Notwithstanding, from a neuroscience perspective, it is not clear how a neuron could compare compartmental signals. Here, we propose a solution to this problem in that we let the apical feedback signal change the postsynaptic firing rate and combine this with a differential Hebbian update, a rate-based version of classical spiking time-dependent plasticity (STDP). We prove that weight updates of this form minimize two alternative loss functions that we prove to be equivalent to the error-based losses used in machine learning: the inference latency and the amount of top-down feedback necessary. Moreover, we show that the use of differential Hebbian updates works similarly well in other feedback-based deep learning frameworks such as Predictive Coding or Equilibrium Propagation. Finally, our work removes a key requirement of biologically plausible models for deep learning and proposes a learning mechanism that would explain how temporal Hebbian learning rules can implement supervised hierarchical learning.

Dual-resonant scanning multiphoton microscope with ultrasound lens and resonant mirror for rapid volumetric imaging

A dual-resonant scanning multiphoton (DRSM) microscope incorporating a tunable acoustic gradient index of refraction lens with a resonant mirror is developed for high-speed volumetric imaging. In the proposed microscope, the pulse train signal of a femtosecond laser is used to trigger an embedded field programmable gate array to sample the multiphoton excited fluorescence signal at the rate of one pixel per laser pulse. It is shown that a frame rate of around 8000 Hz can be obtained in the x-z plane for an image region with a size of 256 × 80 pixels. Moreover, a volumetric imaging rate of over 30 Hz can be obtained for a large image volume of 343 × 343 × 120 μm3 with an image size of 256 × 256 × 80 voxels. Moreover, a volumetric imaging rate of over 30 Hz can be obtained for a large image volume of 256 × 256 × 80 voxels, which represents 343 × 343 × 120 μm3 in field-of-view. The rapid volumetric imaging rate eliminates the aliasing effect for observed temporal frequencies lower than 15 Hz. The practical feasibility of the DRSM microscope is demonstrated by observing the mushroom bodies of a drosophila brain and performing 3D dynamic observations of moving 10-μm fluorescent beads.

Three-dimensional-generator U-net for dual-resonant scanning multiphoton microscopy image inpainting and denoising

A dual-resonant scanning multiphoton (DRSM) microscope incorporating a tunable acoustic gradient index of refraction lens and a resonant mirror is developed for rapid volumetric bioimaging. It is shown that the microscope achieves a volumetric imaging rate up to 31.25 volumes per second (vps) for a scanning volume of up to 200 × 200 × 100 µm³ with 256 × 256 × 128 voxels. However, the volumetric images have a severe negative signal-to-noise ratio (SNR) as a result of a large number of missing voxels for a large scanning volume and the presence of Lissajous patterning residuals. Thus, a modified three-dimensional (3D)-generator U-Net model trained using simulated microbead images is proposed and used to inpaint and denoise the images. The performance of the 3D U-Net model for bioimaging applications is enhanced by training the model with high-SNR in-vitro drosophila brain images captured using a conventional point scanning multiphoton microscope. The trained model shows the ability to produce clear in-vitro drosophila brain images at a rate of 31.25 vps with a SNR improvement of approximately 20 dB over the original images obtained by the DRSM microscope. The training convergence time of the modified U-Net model is just half that of a general 3D U-Net model. The model thus has significant potential for 3D in-vivo bioimaging transfer learning. Through the assistance of transfer learning, the model can be extended to the restoration of in-vivo drosophila brain images with a high image quality and a rapid training time.

Characterization of red fluorescent reporters for dual-color in vivo three-photon microscopy

Significance: Three-photon (3P) microscopy significantly increases the depth and resolution of in vivo imaging due to decreased scattering and nonlinear optical sectioning. Simultaneous excitation of multiple fluorescent proteins is essential to studying multicellular interactions and dynamics in the intact brain. Aim: We characterized the excitation laser pulses at a range of wavelengths for 3P microscopy, and then explored the application of tdTomato or mScarlet and EGFP for dual-color single-excitation structural 3P imaging deep in the living mouse brain. Approach: We used frequency-resolved optical gating to measure the spectral intensity, phase, and retrieved pulse widths at a range of wavelengths. Then, we performed in vivo single wavelength-excitation 3P imaging in the 1225- to 1360-nm range deep in the mouse cerebral cortex to evaluate the performance of tdTomato or mScarlet in combination with EGFP. Results: We find that tdTomato and mScarlet, expressed in oligodendrocytes and neurons respectively, have a high signal-to-background ratio in the 1300- to 1360-nm range, consistent with enhanced 3P cross-sections. Conclusions: These results suggest that a single excitation wavelength source is advantageous for multiple applications of dual-color brain imaging and highlight the importance of empirical characterization of individual fluorophores for 3P microscopy.

Optogenetic Methods to Investigate Brain Alterations in Preclinical Models

Investigating the neuronal dynamics supporting brain functions and understanding how the alterations in these mechanisms result in pathological conditions represents a fundamental challenge. Preclinical research on model organisms allows for a multiscale and multiparametric analysis in vivo of the neuronal mechanisms and holds the potential for better linking the symptoms of a neurological disorder to the underlying cellular and circuit alterations, eventually leading to the identification of therapeutic/rescue strategies. In recent years, brain research in model organisms has taken advantage, along with other techniques, of the development and continuous refinement of methods that use light and optical approaches to reconstruct the activity of brain circuits at the cellular and system levels, and to probe the impact of the different neuronal components in the observed dynamics. These tools, combining low-invasiveness of optical approaches with the power of genetic engineering, are currently revolutionizing the way, the scale and the perspective of investigating brain diseases. The aim of this review is to describe how brain functions can be investigated with optical approaches currently available and to illustrate how these techniques have been adopted to study pathological alterations of brain physiology.

Breaking trade-offs: development of fast, high-resolution, wide-field two-photon microscopes to reveal the computational principles of the brain

Information in the brain is represented by the collective and coordinated activity of single neurons. Activity is determined by a large amount of dynamic synaptic inputs from neurons in the same and/or distant brain regions. Therefore, the simultaneous recording of single neurons across several brain regions is critical for revealing the interactions among neurons that reflect the computational principles of the brain. Recently, several wide-field two-photon (2P) microscopes equipped with sizeable objective lenses have been reported. These microscopes enable large-scale in vivo calcium imaging and have the potential to make a significant contribution to the elucidation of information-processing mechanisms in the cerebral cortex. This review discusses recent reports on wide-field 2P microscopes and describes the trade-offs encountered in developing wide-field 2P microscopes. Large-scale imaging of neural activity allows us to test hypotheses proposed in theoretical neuroscience, and to identify rare but influential neurons that have potentially significant impacts on the whole-brain system.

Non-telecentric two-photon microscopy for 3D random access mesoscale imaging

Diffraction-limited two-photon microscopy permits minimally invasive optical monitoring of neuronal activity. However, most conventional two-photon microscopes impose significant constraints on the size of the imaging field-of-view and the specific shape of the effective excitation volume, thus limiting the scope of biological questions that can be addressed and the information obtainable. Here, employing a non-telecentric optical design, we present a low-cost, easily implemented and flexible solution to address these limitations, offering a several-fold expanded three-dimensional field of view. Moreover, rapid laser-focus control via an electrically tunable lens allows near-simultaneous imaging of remote regions separated in three dimensions and permits the bending of imaging planes to follow natural curvatures in biological structures. Crucially, our core design is readily implemented (and reversed) within a matter of hours, making it highly suitable as a base platform for further development. We demonstrate the application of our system for imaging neuronal activity in a variety of examples in zebrafish, mice and fruit flies.

Electrically tunable lenses - eliminating mechanical axial movements during high-speed 3D live imaging

The successful investigation of photosensitive and dynamic biological events, such as those in a proliferating tissue or a dividing cell, requires non-intervening high-speed imaging techniques. Electrically tunable lenses (ETLs) are liquid lenses possessing shape-changing capabilities that enable rapid axial shifts of the focal plane, in turn achieving acquisition speeds within the millisecond regime. These human-eye-inspired liquid lenses can enable fast focusing and have been applied in a variety of cell biology studies. Here, we review the history, opportunities and challenges underpinning the use of cost-effective high-speed ETLs. Although other, more expensive solutions for three-dimensional imaging in the millisecond regime are available, ETLs continue to be a powerful, yet inexpensive, contender for live-cell microscopy.

Fast, cell-resolution, contiguous-wide two-photon imaging to reveal functional network architectures across multi-modal cortical areas

Fast and wide field-of-view imaging with single-cell resolution, high signal-to-noise ratio, and no optical aberrations have the potential to inspire new avenues of investigations in biology. However, such imaging is challenging because of the inevitable tradeoffs among these parameters. Here, we overcome these tradeoffs by combining a resonant scanning system, a large objective with low magnification and high numerical aperture, and highly sensitive large-aperture photodetectors. The result is a practically aberration-free, fast-scanning high optical invariant two-photon microscopy (FASHIO-2PM) that enables calcium imaging from a large network composed of ∼16,000 neurons at 7.5 Hz from a 9 mm² contiguous image plane, including more than 10 sensory-motor and higher-order areas of the cerebral cortex in awake mice. Network analysis based on single-cell activities revealed that the brain exhibits small-world rather than scale-free behavior. The FASHIO-2PM is expected to enable studies on biological dynamics by simultaneously monitoring macroscopic activities and their compositional elements.

Cortical ensembles selective for context

Neural processing of sensory information is strongly influenced by context. For instance, cortical responses are reduced to predictable stimuli, while responses are increased to novel stimuli that deviate from contextual regularities. Such bidirectional modulation based on preceding sensory context is likely a critical component or manifestation of attention, learning, and behavior, yet how it arises in cortical circuits remains unclear. Using volumetric two-photon calcium imaging and local field potentials in primary visual cortex (V1) from awake mice presented with visual "oddball" paradigms, we identify both reductions and augmentations of stimulus-evoked responses depending, on whether the stimulus was redundant or deviant, respectively. Interestingly, deviance-augmented responses were limited to a specific subset of neurons mostly in supragranular layers. These deviance-detecting cells were spatially intermixed with other visually responsive neurons and were functionally correlated, forming a neuronal ensemble. Optogenetic suppression of prefrontal inputs to V1 reduced the contextual selectivity of deviance-detecting ensembles, demonstrating a causal role for top-down inputs. The presence of specialized context-selective ensembles in primary sensory cortex, modulated by higher cortical areas, provides a circuit substrate for the brain's construction and selection of prediction errors, computations which are key for survival and deficient in many psychiatric disorders.

Multiplexing viral approaches to the study of the neuronal circuits

Neural circuits are composed of multitudes of elaborately interconnected cell types. Understanding neural circuit function requires not only cell-specific knowledge of connectivity, but the ability to record and manipulate distinct cell types independently. Recent advances in viral vectors promise the requisite specificity to perform true "circuit-breaking" experiments. However, such new avenues of multiplexed, cell-specific investigation raise new technical issues: one must ensure that both the viral vectors and their transgene payloads do not overlap with each other in both an anatomical and a functional sense. This review describes benefits and issues regarding the use of viral vectors to analyse the function of neural circuits and provides a resource for the design and implementation of such multiplexing experiments.

Temporally multiplexed dual-plane imaging of neural activity with four-dimensional precision

Spatiotemporal patterns of neural activity generate brain functions, such as perception, memory, and behavior. Four-dimensional (4-D: x, y, z, t) analyses of such neural activity will facilitate understanding of brain functions. However, conventional two-photon microscope systems observe single-plane brain tissue alone at a time with cellular resolution. It faces a trade-off between the spatial resolution in the x-, y-, and z-axes and the temporal resolution by a limited point-by-point scan speed. To overcome this trade-off in 4-D imaging, we developed a holographic two-photon microscope for dual-plane imaging. A spatial light modulator (SLM) provided an additional focal plane at a different depth. Temporal multiplexing of split lasers with an optical chopper allowed fast imaging of two different focal planes. We simultaneously recorded the activities of neurons on layers 2/3 and 5 of the cerebral cortex in awake mice in vivo. The present study demonstrated the proof-of-concept of dual-plane two-photon imaging of neural circuits by using the temporally multiplexed SLM-based microscope. The temporally multiplexed holographic microscope, combined with in vivo labeling with genetically encoded probes, enabled 4-D imaging and analysis of neural activities at cellular resolution and physiological timescales. Large-scale 4-D imaging and analysis will facilitate studies of not only the nervous system but also of various biological systems.

Light-field microscopy for fast volumetric brain imaging

Recording neural activities over large populations is critical for a better understanding of the functional mechanisms of animal brains. Traditional optical imaging technologies for in vivo neural activity recording are usually limited in throughput and cannot cover a large imaging volume at high speed. Light-field microscopy features a highly parallelized imaging collection mechanism and can simultaneously record optical signals from different depths. Therefore, it can potentially increase the imaging throughput substantially. Furthermore, its unique instantaneous volumetric imaging capability enables the capture of highly dynamic processes, such as recording whole-animal neural activities in freely moving Caenorhabditis elegans and whole-brain neural activity in freely swimming larval zebrafish during prey capture. Here, we summarize the principles of and considerations in the practical implementation of light-field microscopy as currently applied in biological imaging experiments. We also discuss the strategies that light-field microscopy can employ when imaging thick tissues in the presence of scattering and background interference. Finally, we present a few examples of applying light-field microscopy in neuroscientific studies in several important animal models.

Identification and quantification of neuronal ensembles in optical imaging experiments

Recent technical advances in molecular biology and optical imaging have made it possible to record from up to thousands of densely packed neurons in superficial and deep brain regions in vivo, with cellular subtype specificity and high spatiotemporal fidelity. Such optical neurotechnologies are enabling increasingly fine-scaled studies of neuronal circuits and reliably co-active groups of neurons, so-called ensembles. Neuronal ensembles are thought to constitute the basic functional building blocks of brain systems, potentially exhibiting collective computational properties. While the technical framework of in vivo optical imaging and quantification of neuronal activity follows certain widely held standards, analytical methods for study of neuronal co-activity and ensembles lack consensus and are highly varied across the field. Here we provide a comprehensive step-by-step overview of theoretical, experimental, and analytical considerations for the identification and quantification of neuronal ensemble dynamics in high-resolution in vivo optical imaging studies.

Fully Affine Invariant Methods for Cross-Session Registration of Calcium Imaging Data

In vivo two-photon microscopy permits simultaneous recording of the activity of the same neuronal population across multiple sessions in days or weeks, which is crucial for addressing many fundamental questions of neuroscience. The field-of-view (FOV) alignment is a necessary step for identifying the same neurons across multiple imaging sessions. Accurate FOV alignment becomes challenging in the situations of image blurring, insufficient common neurons, or uneven background brightness. The existing methods largely fail to align FOV pairs in these situations. The fully affine invariant approach has been applied in computer vision to register real scene images with different backgrounds. However, its performance in calcium imaging data is unknown. We explored the feasibility of using the fully affine invariant approach to align calcium FOV images across multiple sessions by examining the performance of five methods. Further, we compared their performance with common feature-based methods as well as some classical methods with or without adaptive contrast enhancement. Using cellular resolution calcium imaging data recorded from two areas of the mouse motor cortex over weeks, we show that all fully affine invariant methods provide more accurate FOV alignment results than other methods in general and in the case of a few common neurons identified, uneven background brightness or image blurring. This study demonstrated the feasibility and reliability of the fully affine invariant methods in cross-session FOV alignment. These methods could be useful for neuroscience research, especially on questions that involve experience-dependent plasticity spanning over days or weeks.

Axonal Tree Morphology and Signal Propagation Dynamics Improve Interneuron Classification

Neurons are diverse and can be differentiated by their morphological, electrophysiological, and molecular properties. Current morphology-based classification approaches largely rely on the dendritic tree structure or on the overall axonal projection layout. Here, we use data from public databases of neuronal reconstructions and membrane properties to study the characteristics of the axonal and dendritic trees for interneuron classification. We show that combining signal propagation patterns observed by biophysical simulations of the activity along ramified axonal trees with morphological parameters of the axonal and dendritic trees, significantly improve classification results compared to previous approaches. The classification schemes introduced here can be utilized for robust neuronal classification. Our work paves the way for understanding and utilizing form-function principles in realistic neuronal reconstructions.

Converting lateral scanning into axial focusing to speed up three-dimensional microscopy

In optical microscopy, the slow axial scanning rate of the objective or the sample has traditionally limited the speed of volumetric imaging. Recently, by conjugating either a movable mirror to the image plane in a remote-focusing geometry or an electrically tuneable lens (ETL) to the back focal plane, rapid axial scanning has been achieved. However, mechanical actuation of a mirror limits the axial scanning rate (usually only 10-100 Hz for piezoelectric or voice coil-based actuators), while ETLs introduce spherical and higher-order aberrations that prevent high-resolution imaging. In an effort to overcome these limitations, we introduce a novel optical design that transforms a lateral-scan motion into a spherical aberration-free axial scan that can be used for high-resolution imaging. Using a galvanometric mirror, we scan a laser beam laterally in a remote-focusing arm, which is then back-reflected from different heights of a mirror in the image space. We characterize the optical performance of this remote-focusing technique and use it to accelerate axially swept light-sheet microscopy by an order of magnitude, allowing the quantification of rapid vesicular dynamics in three dimensions. We also demonstrate resonant remote focusing at 12 kHz with a two-photon raster-scanning microscope, which allows rapid imaging of brain tissues and zebrafish cardiac dynamics with diffraction-limited resolution.

Understanding the circuit basis of cognitive functions using mouse models

Understanding how cognitive functions arise from computations occurring in the brain requires the ability to measure and perturb neural activity while the relevant circuits are engaged for specific cognitive processes. Rapid technical advances have led to the development of new approaches to transiently activate and suppress neuronal activity as well as to record simultaneously from hundreds to thousands of neurons across multiple brain regions during behavior. To realize the full potential of these approaches for understanding cognition, however, it is critical that behavioral conditions and stimuli are effectively designed to engage the relevant brain networks. Here, we highlight recent innovations that enable this combined approach. In particular, we focus on how to design behavioral experiments that leverage the ever-growing arsenal of technologies for controlling and measuring neural activity in order to understand cognitive functions.

Dual-plane 3-photon microscopy with remote focusing

3-photon excitation enables in vivo fluorescence microscopy deep in densely labeled and highly scattering samples. To date, 3-photon excitation has been restricted to scanning a single focus, limiting the speed of volume acquisition. Here, for the first time to our knowledge, we implemented and characterized dual-plane 3-photon microscopy with temporal multiplexing and remote focusing, and performed simultaneous in vivo calcium imaging of two planes beyond 600 µm deep in the cortex of a pan-excitatory GCaMP6s transgenic mouse with a per-plane framerate of 7 Hz and an effective 2 MHz laser repetition rate. This method is a straightforward and generalizable modification to single-focus 3PE systems, doubling the rate of volume (column) imaging with off-the-shelf components and minimal technical constraints.