In vivo brain imaging demands high spatiotemporal resolution microscopy. Sub-micron spatial resolution is needed to resolve synapses, while sub-second temporal resolution is crucial to track neuronal activity. Existing methods often lack the capability to image synapses at high spatiotemporal resolution in 3D at depth. Two-photon fluorescence microscopy (2PFM) is a popular technique for imaging opaque tissues, but achieving high spatial resolution at depth requires adaptive optics (AO) to correct for sample-induced aberrations. While Bessel beams offer increased temporal resolution through volumetric imaging, their susceptibility to aberrations at depth remains a challenge. This study addresses this challenge by developing a novel AO method that combines the advantages of both 2PFM and Bessel beams for high-resolution volumetric imaging of synaptic activity.

Two-photon fluorescence microscopy (2PFM) has become a standard for in vivo imaging, particularly when combined with adaptive optics (AO) to overcome the limitations of light scattering and aberration in deep tissue. Studies have shown the potential of Bessel beams in improving temporal resolution in 2PFM by enabling simultaneous imaging of structures along the extended axial length of the beam. However, previous work has demonstrated that Bessel beams, despite their self-healing properties, are still affected by sample-induced aberrations, particularly non-circularly symmetric ones. This limitation has hindered the development of high-resolution volumetric imaging methods capable of simultaneously measuring neuronal activity at multiple locations within a three-dimensional space.

The researchers constructed an adaptive optics (AO) Bessel focus scanning two-photon fluorescence microscope. The system uses two spatial light modulators (SLMs): SLM1, conjugated to the objective focal plane, and SLM2, conjugated to the objective pupil plane. For Gaussian focus, SLM2 is used for wavefront correction via a pupil-segmentation-based AO method. For Bessel focus generation, a binary phase pattern is displayed on SLM1, while an annular mask spatially filters the excitation light. A crucial innovation is the development of a high-efficiency focal-plane aberration correction method for Bessel foci. Instead of correcting aberrations at the pupil plane (SLM2), the method calculates and applies a corrective phase pattern to SLM1, compensating for both phase and amplitude distortions caused by aberrations. This is achieved by computationally determining the optimal phase pattern on SLM1 based on the corrective pattern obtained from the pupil plane. This pattern compensates for both phase and amplitude distortions, significantly improving signal recovery compared to pupil-plane-only correction. The system's performance was characterized by imaging fluorescent beads under various aberration conditions, demonstrating the superior effectiveness of focal-plane correction for Bessel beams. The researchers further investigated how different aberration modes differentially affected Gaussian and Bessel foci, both experimentally and through numerical simulations. In vivo experiments were conducted on zebrafish larvae and mice. Cranial windows were implanted in mice to allow imaging of the primary visual cortex. Neurons expressing fluorescent proteins (GFP, GCaMP6s, GCaMP7s, iGluSnFR-A184S) were imaged, with aberrations measured using fluorescent beads or neuron cell bodies. The focal-plane AO method was then used to correct sample-induced aberrations. Volumetric imaging was performed using both Gaussian and Bessel focus scanning. Data analysis included image registration, deconvolution, and manual ROI selection for quantifying fluorescence signals. For functional imaging, drifting grating stimuli were used to evoke neuronal activity, with orientation selectivity assessed using ANOVA and the global orientation selectivity index (gOSI).

The focal-plane aberration correction method significantly improved the quality of Bessel focus imaging. Aberration correction substantially improved the sensitivity and resolution of both structural and functional measurements of synapses in vivo. In mouse primary visual cortex, AO correction of Bessel focus resulted in a 4x increase in the signal of dendritic spines compared to the 3x increase observed with Gaussian focus. The improved resolution led to the detection of more dendritic spines and better characterization of their morphology. AO correction enabled more accurate characterization of the orientation tuning properties of dendritic spines in the awake mouse visual cortex during calcium imaging. More calcium transients were detected in dendrites and spines after aberration correction, leading to a significant increase in the measured orientation selectivity. The study also demonstrated the simultaneous volumetric imaging of glutamate release at both apical and basal dendritic spines with high sensitivity. AO correction dramatically improved the detection of glutamate transients, enabling accurate measurement of orientation tuning properties for these spines. A significant 81° shift in the dominant orientations of apical versus basal dendritic spines was observed, indicating distinct input preferences.

This study successfully demonstrated a highly efficient method for correcting sample-induced aberrations in Bessel focus scanning 2PFM. The focal-plane aberration correction method proved superior to conventional pupil-plane correction for Bessel beams, enabling high-resolution volumetric imaging of neuronal structures and activity at unprecedented depths. The significant improvements in signal, resolution, and sensitivity were validated through in vivo experiments in both zebrafish and mice. The ability to accurately characterize the orientation tuning properties of dendritic spines, particularly with glutamate imaging, highlights the importance of this advance in neuroscience research. The findings provide strong evidence for the practical advantages of the combined use of Bessel beams and focal-plane AO for resolving detailed functional properties of neurons at the synaptic level, and its implications for understanding neuronal circuits.

The development of an efficient focal-plane aberration correction method for Bessel focus scanning 2PFM represents a significant advancement in in vivo brain imaging. This method enables high-resolution volumetric imaging with both high spatial and temporal resolution, revealing previously inaccessible details of neuronal structure and function. Future research directions could focus on extending this technique to other types of genetically encoded sensors or developing algorithms for improved demixing of overlapping signals in densely populated neuronal regions.

The study primarily focused on the primary visual cortex of mice, which may limit the generalizability of findings to other brain regions or species. The focal-plane AO method requires computational calculation of the corrective phase pattern, which could be computationally intensive for very large datasets. While the study improved the sensitivity and resolution of glutamate detection, the low brightness and temporal dynamics of glutamate sensors relative to calcium sensors still pose challenges.