Development of the HIF-Clear pipeline.

(A) Comparison of DiD signal intensity in 2-mm-thick slices of undelipidated formaldehyde-fixed mouse brain, PFA-SDS mouse brain, and FFPE mouse brain treated under various detergent conditions.

(B) Multi-point confocal fluorescence images of neuronal nuclei in the cortex (top row) and axons in the striatum (bottom row) from 2-mm-thick FFPE mouse brain slices that had been treated under various detergent conditions. Images of optical sections captured at a depth of 1 mm are shown. Detergent types and concentrations are indicated. NeuN, a neuronal nuclear marker; SMI312, a pan-axonal marker.

(C) Gross views of 2-mm-thick slices of FFPE mouse brain treated under various detergent conditions.

(D) Schematic of the HIF-Clear pipeline.

(E) Images of coherent anti-Stokes Raman scattering (CARS) of dewaxed FFPE mouse brain, HIF-Clear-processed FFPE mouse brain, and PFA-SDS mouse brain. The images were taken at a Raman shift of 2850 cm-1 to address the CH2 vibration frequency.

(F) Statistical analysis of the gray values for the three treatment groups in (E). Gray values for 18 z-continuous images of each group were analyzed.

Statistical analysis: (A) n = 10; (E) n = 18. Mean ± SD, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, one-way ANOVA with Tukey’s multiple comparison test.

Validation of the HIF-Clear pipeline.

(A) Brain-wide tyrosine hydroxylase (TH) expression (left panel) and registration (right panel) in FFPE-HIF-Clear C57BL/6 mouse brains. Whole mouse brain images were acquired using light-sheet microscopy, and the 3D-rendered images of the horizontal (XY) and sagittal (YZ) views are shown. Major dopaminergic regions are indicated: SN (substantia nigra), VTA (ventral tegmental area), LRN (lateral reticular nucleus), and PRN (pontine reticular nucleus). The results of brain registration for these regions (marked with green dotted lines and numbered in the XY view) are displayed from the coronal (XZ) view in the right panel.

(B) Comparison of brain region volumes between the left and right hemispheres of an FFPE-HIF-Clear mouse brain. All abbreviations are listed in Supplementary Table 2.

(C to E) Application of HIF-Clear to a 15-year-old FFPE block with TH-immunolabeling and multi-point confocal imaging. Specimen thickness: 2 mm.

(C) The projection image.

(D) Magnification of the cyan-lined region in (C). Dopaminergic regions are indicated. PVT, paraventricular nucleus of thalamus; PH, posterior hypothalamic nucleus; PVp, periventricular hypothalamic nucleus.

(E) Magnification of the magenta-lined region in (D). Scale bars are indicated in each panel.

Application of HIF-Clear to long-term fixed mouse brain hemispheres.

Mouse brain hemispheres were fixed under various fixation conditions and then subjected to FFPE processing and HIF-Clear. The specimens were stained with TH antibodies and imaged with light-sheet microscopy.

(A) Projection images of the horizontal (XY) view. The magenta-lined regions indicating Nigrostriatal fiber tract, VTA, and LRN were magnified in (B), (C), and (D), respectively. The yellow arrowheads indicate PRN, which was clearly immunolabeled in the 4% PFA 24-hour-fixed specimen but not in the 3-month-fixed specimens.

Multi-round immunostaining of FFPE-HIF-Clear whole mouse brain (HIF-Clear+).

Light-sheet microscopy imaging datasets of six-round immunolabeling performed on one whole FFPE mouse brain. Projection images of the horizontal (XY), sagittal (YZ), and transverse (XZ) views are shown. Projection images (200-μm thick) of magenta-lined and cyan-lined regions marked in the XY and XZ views are magnified and displayed within correspondingly colored frames. Scale bars are indicated in the first row.

Merged multichannel neuron circuitry image generated by HIF-Clear+.

(A) Merged multichannel image generated using the datasets from Fig. 3. Optical sections from the horizontal (XY) and sagittal (YZ) perspectives are presented. For the transverse (XZ) views, projection images of the regions encompassed by the dotted lines in the XY view are displayed.

(B) The spatial relationship of the Hbn-IPN circuitry, dopaminergic (TH), serotoninergic (TPH2), and calbindin+ GABAergic neurons. Fluorescence (right) and segmented (left) images are shown. mHb, medial habenula; FR, fasciculus retroflexus; IPN, interpeduncular nucleus; SuM, medial mammillary area of the hypothalamus; DR, dorsal raphe nuclei.

Multi-round immunostaining of a FFPE-HIF-Clear human brain specimen.

Three-round immunolabeling was performed on a FFPE human brain specimen collected from a patient with cerebral hemorrhage. Images were acquired using multi-point confocal microscopy.

(A) 3D rendering of the entire specimen.

(B) Images depicting merged and single channels of an optical section from the selected area (white frame) in (A) at depths of 100 μm, 500 μm, and 900 μm. GFAP: Glial fibrillary acidic protein; SMI312: pan-axonal marker; MAP2: microtubule-associated protein 2 (a dendritic marker).

Applications of HIF-Clear in disease models.

(A to F) HIF-Clear reveals cell-tumor relationships in an astrocytoma model.

(A) Projection light-sheet image of a FFPE-HIF-Clear mouse brain with a GFP-expressing astrocytoma xenograft (ALTS1C1 cells, see Materials and Methods) stained for GFP (yellow) and GFAP (cyan).

(B) Segmented tumor (white) and surrounding astrocytes (colored). The astrocytes are colored according to the color-coded scale of cell-to-tumor distance.

(C) Magnification of the region marked in (A), showing astrocyte morphology.

(D) Sagittal view of the segmented tumor and color-coded astrocytes.

(E) Magnification of the clipping plane marked in (D, yellow dashed line), showing astrocytes inside (dark blue, purple) and outside (red, orange, yellow, green, cyan, blue) the tumor.

(F) Quantification of detected astrocytes and their classification according to distance to the tumor surface.

(G to M) HIF-Clear reveals brain damage and the therapeutic effect of alternating magnetic field-responsive NO-release octahedrons (aNORO) administered using a paracetamol-coated silk-based microneedle (sMN) in a TBI mouse model.

(G to J) Segmentation of dopaminergic regions in FFPE-HIF-Clear TBI mouse brains.

(G) Whole-brain projection light-sheet image focusing on TH-positive regions.

(H) Frontal view of the segmented striatum (dashed white line in (G)). Magenta: the injured side; cyan: the contralateral side.

(I) Horizontal views of the nigrostriatal fiber tract (paired white boxes in (G)); fluorescence (top) and segmented (bottom) images are shown. Magenta: the injured side; cyan: the contralateral side.

(J) Detection of dopaminergic cells in the substantia nigra (SN). Magnification of the marked area in (G), including the TH fluorescence signal (yellow) and segmented cells (cyan dots).

(K) Comparison of the therapeutic effects of different treatments. n = 3 (three indicators: striatum volume, nigrostriatal fiber tract volume, SN cell number), mean ± SD, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, one-way ANOVA with Tukey’s multiple comparison test.

(L) Reconstruction of blood vessels in TBI brains with or without treatments. The regions of interest (ROIs) selected to assess angiogenesis are indicated in the top row (magenta boxes, directly below the injury site). 3D reconstructions of blood vessels are shown in the bottom row.

(M) Quantification of blood vessel volume, blood vessel surface area, number of blood vessel branches, and blood vessel length. n = 3, mean ± SD have been plotted.

Expansion microscopy on a FFPE-HIF-Clear human brain specimen.

(A) Gross views of expansion of a 1-mm FFPE specimen.

(B) Fluorescence images of a FFPE human brain pre-(top row) and post-expansion (bottom row). Left: merged images of multiplexed staining; right: fluorescence images of SMI312 staining.

(C) Profiles of axon signal intensity taken along the white lines shown in the images of the right panel in (B). The scale bar has been divided by the expansion factor.