Axially decoupled patterned photo-stimulation and multiphoton imaging.

(a) (Top) Cartoon schematics of functional connectivity mapping of a neural circuit (e.g. olfactory bulb); photo-stimulation (2-D light mask from a digital micro-mirror device illuminated with a solid-state CW laser) and two photon imaging are confined to different optical z-planes that can be flexibly and independently adjusted by translating the diffuser and respectively the primary objective; (Bottom) Alternating (strobing) between photo-stimulation and imaging periods. Each red bar represents a single frame of multiphoton imaging. Photo-stimulation and imaging periods are interleaved. (b) Microscope schematics. DM, dichroic mirrors. DMD, digital micro-mirror device. I, iris diaphragm. L1-L12, lenses. O, primary objective. PMT, photomultiplier tube. PS, periscope. S, shutter. SM, scan mirrors. (c) (Top) Illustration of using a movable diffuser to decouple the patterned photo-stimulation and multiphoton imaging planes. The diffuser is imaged into the sample, in a 4f lens configuration; translating the diffuser along the optical path causes the corresponding projection plane to shift axially. OFP, objective focal plane. PSP, photo-stimulation plane. (d) Example projected grid pattern (red overlay) and ensuing fluorescence on a fluorescently coated slide; line width is 1 pixel (2.4 µm); distance between lines is 2 and 3 pixels respectively; marginal distributions represent average cross-section line profiles. (e) x-y FWHM (6.2 ± 1.6 µm) of a 4.8 µm nominal diameter spot. (f) x-z FWHM (34.0 ± 4.9 µm) of a 4.8 µm nominal diameter spot. (g) Axial confinement of light intensity of a 2-D projection pattern using a holographic diffuser conjugated to the focal plane of the principal objective (Left) vs. same optical configuration without a diffuser and the resultant light pencil (Right); spot diameter = 50 µm. (h) Measured spot x-y FWHM versus expected nominal spot size calculated by multiplying the pattern size in pixels by the expected size of a pixel given the DMD chip and the optical magnification of the system. (i) Measured axial (z) FWHM versus lateral (x-y) FWHM of the projected spots from (h) for three different types of diffusers; black circles = 5° holographic diffuser; red circles = 10° holographic diffuser; blue circles = grit ground diffuser. (j) DMD chip-to-CCD camera-to-2p microscope registration. (i) two photon micrograph of 10 µm fluorescence microbeads; arrows mark two microbeads, part of the larger DMD-modulated projection target pattern (8 microbeads), which were taken as fiduciary points; (ii) widefield fluorescence image (full field illumination) of a larger field of view including the target microbeads; (iii) ROIs selected from the 2p image were used to generate DMD-chip light masks; these were further projected at the primary objective focal plane and imaged using the primary CCD camera (CCD 1); (iv) overlay of DMD-generated photo-stimulation masks and widefield fluorescence image of 10 µm microbeads from (ii); note that fluorescence is restricted to the microbeads targeted by the DMD photo-stimulation masks with minimal spillover to adjacent (off-target) microbeads (see fiduciary marks). (k) Measured calibration curve of diffuser translation required to shift photo-stimulation plane by a given amount with respect to the focal plane of the objective. Each point represents Mean ± SEM across ∼6 months, showing the robustness of the setup and linear relationship between objective and compensatory diffuser displacements for the range of decoupling between the photo-stimulation and imaging planes sampled (500 µm).

Targeting individual glomeruli by photo-stimulation of olfactory sensory neuron terminals and monitoring glomerular responses of sister mitral and tufted cell dendrites.

(a) (i) Glomerular resting fluorescence in OMP-Cre x ReaChR-citrine x Thy1-GCaMP6s mice; (b) Fluorescence changes (two dimensional glomerular light activation maps, 2DLAMs) evoked by photo-stimulation of 2 example glomeruli (#3, #4) out of 15 glomeruli targeted in the example field of view (5.0 mW/mm2 for 50 ms). Red ellipses mark the location and contour of the projected light masks. Color scale in heat map indicates ΔF/F, comparing fluorescence after light stimulation to baseline; (c) Response matrix of fluorescence traces (individual repeats in gray, and mean signal as thicker lines) corresponding to 15 glomerular regions of interest versus the 15 (matching) glomerular light masks projected on the OB surface. Red traces mark responses of the targeted glomerulus in a given trial. (d) (Left) Fluorescence changes (two dimensional glomerular light activation maps, 2DLAMs; 5.0 mW/mm2) evoked by photo-stimulation of the 15 glomeruli targeted in the example field of view (a). (Right) Significant 2DLAMs. Non-significant responses were thresholded to gray background. Out of 15 targeted glomeruli, 5 glomeruli were considered non-responsive, given our signal threshold criteria (Methods). Of the 10 responsive light maps, 8 showed specific responses (pertaining to the target glomerulus), and 2 light maps were not specific (responses were observed across multiple glomeruli). (e) Example target glomeruli from a different example FOV across a range of three light intensities (2.5, 5.0 and 7.0 mW/mm2). Projected light masks match the red ellipses in (i) and (ii). Color scale in heat map indicates ΔF/F, comparing fluorescence triggered by light stimulation to baseline; (iii) Same as (ii) for statistically significant glomerular photo-stimulation responses (rectified 2DLAMs); (iv) Fluorescence traces (mean signal for each light intensity) corresponding to three regions of interest (ROI) in each heat map, whose position is indicated by red ellipses in i-iii. Different intensities are represented by different color traces (red - 2.5mW/mm2, green - 5.0mW/mm2, blue - 7.5mW/mm2). (f) Specificity of glomerular responses across light intensity. Blue trace: responsive pixels within the target glomerulus normalized by all responsive pixels located within glomerular boundaries across the FOV. Red trace: responsive pixels within the target glomerulus normalized by all responsive pixels within the FOV, including those spanning the juxtaglomerular space, surrounding glomeruli (Methods); N=7 mice, 21 FOVs; responsive vs. photo-stimulated glomeruli: 2.5mW/mm2, 12/74; 5.0mW/mm2, 48/122; 7.5mW/mm2, 27/57; 10.0mW/mm2, 48/98; 15.0mW/mm2, 48/77.

Identifying sister mitral and tufted cells via axially decoupled glomerular stimulation and two photon fluorescence imaging.

anaestherized (a-b); awake+anaesthetized (c-d);(a) (i) Resting glomerular fluorescence (ReaChR-mCitrine and GCaMP6s); (ii) Fluorescence changes (raw, Top and significant pixels glomerular 2DLAMs, Bottom) evoked by photo-stimulation of four target glomeruli (7.5 mW/mm2); (iii) Response matrix of fluorescence change traces (individual repeats in gray, and mean signal as thicker lines) corresponding to four glomerular regions of interest versus the four corresponding (matching) glomerular light masks projected on the OB surface. Red shaded areas mark responses of the targeted glomerulus in a given trial. (b) (i) Resting fluorescence (GCaMP6s) of mitral cell bodies at the same x-y coordinates, 210 µm below the glomerular field of view in (a); (ii) Significant cell body responses evoked by photo-stimulation of the four target glomeruli (a, 7.5 mW/mm2); (iii) Average fluorescence response traces corresponding to six example mitral cells in each heat map, whose positions are indicated by numbers in (ii); shaded area marks SEM. Two example pairs of sister cells (47, 7 with respect to parent glomerulus #1; and 37, 45 with respect to parent glomerulus #3) are shown. Note that cell #30 is non-specifically responding to photo-stimulation of multiple glomeruli (#2,#3,#4). In addition, glomerular photo-stimulation triggered sparse suppressed responses (e.g. cells 7,11) intermingled with the sister cells. (c) (i) (Top) Specificity of enhanced responding mitral and tufted cells as a function of the targeted glomerulus (N = 6 mice; 27 responsive and specific glomeruli / 46 photo-stimulated target glomeruli; 856 mitral & tufted cells). Note that most cells were non-responsive, or responded specifically to the photo-stimulation of the target glomerulus (13 % of cells had off-target responses to more than one glomerulus; also see Fig. S5a, Methods); (Bottom) Same for suppressed responses. (ii) (Top) Number of observed sister cells (enhanced responses) per glomerulus photo-stimulation across fields of view (Left; 6.9 ± 6.0, N = 27 FOVs); (Bottom) Number of cells per glomerulus whose baseline fluorescence was suppressed upon photo-stimulation of the same glomeruli (Right; 3.2 ± 2.3, N = 20 FOVs); (iii) (Top) Pairwise distance between sister cells vs. between any cells in the field of view (142.6 ± 82.1 µm vs. 137.7 ± 71.8 µm; 725 vs. 29,850 pairs). (Bottom) Pairwise distance between suppressed cells vs. any cells in the field of view (116.2 ± 58.6 µm vs. 138.2 ± 70.4 µm, 83 vs. 19,607 pairs). Unless specified Avg. ± SD values are shown. (d) Photo-stimulation of individual glomeruli triggered inhibitory responses in only a small subset of the sister cells of a given identified cohort (Avg. ± SEM, 14.0 ± 2.7 %, N=6).

Glomerular photo-stimulation of DAT+ interneurons triggers dense, but spatial heterogeneous inhibition of mitral cell activity.

(a) Schematic of in vivo experiment (Top) and baseline 2p fluorescence of three adjacent glomeruli in the field of view (Bottom); GABAergic/dopaminergic superficial short axon cells (DAT+) co-expressing GCaMP6f and ChR2-mCherry; circles and numbers indicate the location of projected light masks. (b) (i) Fluorescence (two-dimensional light activity maps) evoked by photo-stimulating DAT+ cells in glomeruli (1-3) and one (no apparent opsin expression) control site (4) marked in (a) using light masks equivalent to the dotted blue circles. Color scale in heatmap indicates ΔF/F, comparing fluorescence after light stimulation to baseline. (ii) Fluorescence traces (average, thick line, and four individual repeats, thin lines) corresponding to a region of interest (ROI) in each heatmap, whose position is indicated by matching color squares in (i). As in the case of OSN terminals photo-stimulation experiments, we flashed single 100 ms sub-glomerular light masks (20 µm diameter) between the imaging frames and recorded calcium transients in the DAT+ cells via multiphoton imaging; for analysis, the field of view was divided into a 20 by 20 grid (15 x 15 µm grid units), and the response was defined as the average change in fluorescence in each unit of the grid during the two frames (10 Hz) immediately following the light stimulus. (c) For each ROI (1-4), the evoked ΔF/F response is plotted against the distance of that ROI to the center of each stimulation pattern across repeats. For all stimuli except the control (4), the relationship was modeled as an exponential decay. (d) (i) Cartoon schematics: photo-stimulation of opsin expressing DAT+ cells while monitoring mitral cell activity (GCaMP6f) by axial decoupling of the photo-stimulation and imaging planes (awake). (ii) Example glomerular field of stimulation; mice express GCaMP6f in mitral and tufted cell glomerular dendritic tufts and ChrimsonR-tdtomato in DAT+ cells (not shown). (iii) Example field of view of mitral cells imaged while photo-stimulating DAT+ cells (200 ms, 7.5 mW/mm2) in the glomerular layer. (e) (i) (Top) Fluorescence changes (two dimensional light suppression maps) in mitral cells evoked by photo-stimulation of DAT+ cells across six glomeruli (G1-G6); (Bottom) Diversity of same mitral cell responses to six example odors; each cell in the field of view is displayed on a blue-to-red color scale indicating its response amplitude to light/odor stimulation; non-significant responses were thresholded to white (Methods). (f) (i) (Top) Dense, but heterogeneous inhibition of baseline fluorescence in example mitral cells across the field of view evoked by photo-stimulation of DAT+ cells across six different glomerular light masks (G1-G6); (Bottom) Odor responses of the same example mitral cells to six odors (isoamyl acetate, IAA; acetaldehyde, ADE; valeraldehyde, PEN; allyl tiglate, AT; heptanal, HEP; acetophenone, ACP). (g) Peak fluorescence changes (ΔF/F) of same mitral cells (49) across the field of view in response to glomerular DAT+ cells photo-stimulation (Left) and to a panel of eleven odors (Right). (h) (Left) Histogram of correlation coefficients of mitral cell ensemble responses across different pairs of DAT+ cell glomerular photo-stimulation masks (G1-G6). (Right) Same, across pairs of odors in the panel (1-11). Histograms of shuffled mitral cell index correlation distributions are shown in red; ‘self’ controls in blue (correlations between mitral ensemble responses to repeats of the same DAT+ photo-stimulation mask, Methods).

Photo-stimulation of DAT+ interneurons from different glomeruli on top of odor responses triggers decorrelates the mitral and tufted cell responses.

(a) Schematic of the experiment: in vivo photo-stimulation of ChrimsonR-tdtomato expressing DAT+ cells, while monitoring its impact on mitral and tufted cell odor responses (GCaMP6f). (b) Baseline 2p GCAMP6f fluorescence of an example field of DAT+ stimulation in the glomerular layer. (c) Fluorescence change (dF/F) elicited by odor presentation in one example tufted cell, and its subsequent modulation by projecting of a glomerular light mask to boost DAT+ cell activity (500 ms, 10mW/mm2; 2s from odor onset). (d) Odor response modulation (dF/F) via boosting DAT+ cells associated with different glomeruli. Two dimensional light suppression/activation maps) of tufted (Top) and mitral cell (Bottom) responses to one example odor (2-hexanone, 2-HEX) evoked by photo-stimulation of DAT+ cells across 2 example glomeruli (G1,G4) from (b). (e, g) Mitral (e) and tufted (g) cell peak odor responses (ADE, PEN, 2-HEX) and their modulation by DAT+ cell light masks associated with four different glomeruli (G1-G4) from the example field of view in (b); also see Fig. S8. (f, h) Decorrelation of MTC responses to same odor by boosting DAT+ inputs from different glomeruli. Histograms of pairwise correlations of modulating the tufted cell ensemble responses (f) and of mitral cell ensemble responses (h).