https://doi.org/10.5061/dryad.hqbzkh1rk

TH_CD31_1R through TH_CD31_6R LIF files. The dataset contains image files obtained on a Leica DMi8 CS DLS confocal laser scanning microscope. Fluorescence labels were excited with an argon laser (488nm) and a helium neon laser (633nm), respectively. Whole iWAT samples were scanned using photomultiplier tube-detectors, using a HC FLUOTAR L 25x objective (NA 0.95, WD 2.4 mm, water) at 1024 x1024 pixel resolution. Z-stacks were bidirectionally recorded across the full tissue depth of approximately 2mm with 15µm step size.

Linear Regression evaluation graph pad file. Linear regression analysis of manual neuron length quantification plotted against automated neuron length quantification by our pipeline.

Multiple Annotator Test graph pad file. Accuracy was tested by comparing the results obtained from 6 independent manual annotators with the results of the automated quantification by our pipeline.

Vessel density graph pad file. C57BL/6J mice where either single-housed at 30°C or at 4°C for 7 days. Mice were sacrificed, iWAT excised, labeled for CD31 and cleared. Quantification of iWAT vascular density across approximately 850 regions and three biological replicates per group. A Mann-Whitney U test was performed to analyze the non-normally distributed data.

Neuron density graph pad file. C57BL/6J mice where either single-housed at 30°C or at 4°C for 7 days. Mice were sacrificed, iWAT excised, labeled for Tyrosin hydroxylase and cleared. Quantification of iWAT vascular density across approximately 850 regions and three biological replicates per group. A Mann-Whitney U test was performed to analyze the non-normally distributed data.

Neuron density Heatmap graph pad file. Heatmap showing measured neuron density (neuron length per regional lobe volume) of the same sample.

The image analysis pipeline was implemented in Fiji (NIH, Bethesda, MA, USA). The following plugins were used within the pipeline: Bio-formats importer was used for data import, Labkit was used for image segmentation, Skeletonize3D was used for 3D skeletonization.  Heatmap generation was executed in Python 3.11, utilizing the data analysis library Pandas 1.5.3 for data preparation and GraphPad Prism 9.0.0 for visualizing the heatmap. 3D volume rendering was performed using Blender (Foundation).

Statistical analysis was performed in GraphPad Prism 9.0.0. To compare neuron and vascular density, 876 and 803 regions from three biological replicates per group were analyzed. ROUT outlier detection (Q=0.2%) was performed to remove statistical outliers within the dataset. Normal distribution was tested using a Shapiro-Wilk test. Subsequently a Mann-Whitney U test was performed to analyze the non-normally distributed data. 

Analyzing clearness of samples using Adipoclear (Clearness_Adipoclear_TH.tif) or our new methanol (Clearness_MeOH_TH.tif) and THF (Clearness_THF_TH) protocol.

Tif files from Analyzing labeling quality using Adipoclear (Labeling_quality_AC) or our new methanol (Labeling quality_MeOH) and THF (Labeling_quality_THF) protocol.

Tif files from Analyzing labeling stability using our Methanol or THF protocol at day 0, day, 7, and day 14 after labeling. (Labeling_stability_MeOH and labeling_stability_THF, respectively)

Testing methanol sensitivity of antibodies for labeling is shown in the MeOH_sensitivity_AC/MeOH/THF_ gain_850 tif files, respectively

Network density examples for neurons and vessels at 4°C and 30°C are shown in tif files: Network_density_examples_neurons and network density_examples_vessels, respectively

Adipocyte diameters after clearing have been determined. Raw data and calculation is depicetd in the graph pad file: Adipocyte_diameter.pzfx

Raw image files of Tissue morphology of uncleared (preclearing) and cleared samples is shown in the files: Tissue_morpholology_pre_clearing and Tissue_morphology_post clearing. The first number in the filename indicates the biological replicate, the second the selected region.