On-Slide Clearing and Imaging of 100-µm-Thick Histological Sections Using Ethyl Cinnamate and Epifluorescence

. On-slide


INTRODUCTION
The inhomogeneity of light scattered by the tissue is a fundamental problem that leads to undesirable effects when histological tissue samples are observed. [1]The light emit-ted from the microscope slows down as it passes through the sample, which is due to cellular constituents such as lipids, carbohydrates and pigments which all have different refractive indices. [2]As a result, the light emitted by the sample is scattered in many directions, thus creating an opaque appearance of the tissue [3] , particularly in thick sections.To counteract this problem, methods commonly known as tissue clearing have been introduced.Tissue clearing is now commonly used to match the refractive indices of a tissue sample to the refractive index of the clearing medium. [4]his is achieved by removing mostly the lipids from the cell with chemical compounds and replacing the lipids with the clearing medium, while preserving tissue integrity, and in some protocols also the antigenicity of the tissue.The latter is a major factor when the experiments require immunostainings. [5]There are a variety of ways to clear a tissue.As we have previously summarized, there are non-aqueous/ solvent-based clearing approaches and aqueous-based clearing techniques. [6]Solvent-based clearing concentrates on dehydrating a tissue, followed by incubating it in the clearing medium, which matches the refractive indices of the surrounding medium and the tissue; macroscopically, the tissue becomes (nearly) transparent. [7]However, lipids and water are lost during the dehydration process which might restrict the useability of immunofluorescent staining.Furthermore, tissue shrinkage can be observed, and the toxicity of organic solvents can be problematic.Another approach is the aqueous-based clearing.The main techniques here are either simple immersion or hyperhydration, respectively. [1]Simple immersion matches the refractive index of the tissue by incubating it in a high-refractive index aqueous solution, such as formamide.It does not remove lipids, and thus the slight milky appearance of the tissue is preserved.Hyperhydration focuses on lowering the refractive index of the tissue itself by removing lipids.For instance, the CUBIC method [8] utilizes high levels of basic amino alcohols to effectively remove lipids, but as a possible problem, a change in the morphology can also be observed when the tissue sample gains size. [8]n this study, we focus on the application of a non-aqueous-based clearing method which utilizes ethyl cinnamate (ECi) as the main clearing agent.Ethyl cinnamate is a low-toxic, high-refractive index liquid providing optimal transparency even after short incubation periods.It allows the preservation of antigenicity and offers exceptional compatibility with immunofluorescent labeling.The method was first introduced by W. Masselink et.al. [7] and used for clearing human organoids, Drosophila melanogaster, zebrafish, axolotl, and Xenopus laevis in the form of larger pieces of tissue.
Our approach focuses on clearing slide glass-attached histological specimens of up to 300 µm in thickness, i.e., significantly thicker than the 10-25 µm thick frozen sections typically used for immunostaining experiments.Imaging of blood vessels and nerve fibers in 3D might be improved when larger proportions of these structures are captured within a single section.Ethyl cinnamate-based clearing is a simple and quick approach when compared to other tissue clearing methods, taking only about 4 days to fully prepare the tissue sample for imaging versus 10 or more days for other approaches.
As a proof of principle, in our study we used human colorectal cancer tissue, which, due to its dense structure, is challenging as a target of tissue clearing.

AIM
To demonstrate a simple and fast on-slide clearing of thick tissue sections using ethyl cinnamate, preserving signals of immunofluorescent labeling and tissue morphology.

Sample acquisition
Samples of human colorectal adenocarcinomas were acquired at St. Marina University Hospital, Varna.Surgeries were done open or minimally invasive robotic-assisted laparotomy using the Da Vinci Xi System (Intuitive Surgical).Tissue samples were collected following permission from the Ethics Committee of Prof. Dr. Paraskev Stoyanov Medical University in Varna (No. 39/07.07.2014).Fresh surgical material was immediately fixed, brought to the pathology department, and thoroughly examined (Table 1).

Tissue processing
Tissue samples were processed by cutting 10×10×5 mm slices.These slices were cryoprotected in sucrose solution and subsequently submerged in Tissue-Tek O.C.T. for 5 min and standard blocks for frozen section were prepared.Sectioning was done at 100 µm, 200 µm, and 300 µm thickness on a cryostat.Sections were kept on Superfrost microscopic slides (Thermo Scientific) and dried for 12 hours at 37°C to achieve proper adhesion to the glass surface (bonding).The tissue was then rehydrated in PBS for 2 hours, blocking was done for 8 hours using antibody diluent (100 mg Gelatine, 50 mg NaN 3 , 0.5 ml Triton X-100 in 100 ml PBS) before we applied primary antibodies against von Willebrand factor (DAKO, vWF polyclonal rabbit anti-human antibody, A0082) for 24 hours at room temperature (dilution 1:100 in antibody diluent).The tissue was then thoroughly washed in PBS, and we incubated a mixture of the secondary antibody (Invitrogen Alexa Fluor 488 goat anti-rabbit IgG (H+L), A11034, diluted 1:200) and DAPI (Invitrogen D1306, 5 µg/ml).We then post-fixed the tissue using a 4% formaldehyde solution for 1 hour at room temperature.
After another washing step in PBS we dehydrated the tissue in an increasing ethyl-alcohol (EtOH) chain ending in pure isopropyl alcohol.
The clearing started by application of several drops of ECi to fully cover the tissue on the slide.We incubated it for 10 minutes and replaced the ECi by blotting off the excess liquid and reapplying ECi to cover the tissue.After the second incubation of 10 minutes, we again replaced the ECi as described and applied a smaller amount of ECi (estimated to fill the gap between the glass slide and the cover slip) and coverslipped the tissue without any pressure.All steps are listed in Table 2.

Image acquisition and processing
Image acquisition of the samples was done with an Axio-Imager Z.2 (Zeiss, Germany) using standard epifluores- cence with a 20× magnification objective (EC Plan-NEO-FLUAR, NA=0.5).We acquired z-stacks of different regions covering the whole tissue thickness.We chose the optimal z-step value fulfilling the Nyquist criterion giving the final voxel size of 0.322×0.322×1.23 µm.The same settings were used for the cleared and non-cleared (control) samples.Image data was exported in the TIFF format and loaded in Fiji [9] for 3D visualization.We measured (n=15) the depth of visible and "usable" signals along the z-axis with equal settings for both images of control and cleared tissue and performed a two-tailed t-test to confirm statistically significant increase in transparency.

RESULTS
Within minutes of incubation in ECi, the tissue section's macroscopic appearance changed from white and opaque to transparent (Fig. 1).The second ECi incubation made the tissue nearly fully transparent (Fig. 1).Moreover, we were able to show that the visible depth into the tissue samples using standard epifluorescence microscopy was significantly (p<0.001)increased compared to a non-cleared tissue sample (Fig. 2).Non-cleared tissue samples had a detectable immunofluorescent signal up to a depth of 32.7±1.98 µm along the z-axis, while in the cleared tissue we were able to measure signals up to 81.1±16.30µm of depth.This equals approximately a 2.5-fold increase in   observable tissue thickness (Fig. 3).
The tissue integrity was well preserved.We did not observe macroscopic tissue shrinkage or expansion and delicate structures, for example blood vessels were clearly identifiable.

DISCUSSION
We propose our protocol for investigation of tissue samples in laboratories with available standard epifluorescence microscopy.Additional workload and effort may be further reduced, if larger quantities of ECi are available and the microscopic slides could be dipped (instead of repetitive application of drops onto the slide).ECi-on-slide-clearing is a straightforward, cheap and effective way to improve imaging.
We chose samples of colorectal cancer, because this type of tissue is very opaque, and clearing is critical if images of more than 40 µm thickness need to be acquired.We were able to show that clearing tissue sections on microscopic slides with ECi as the main reagent is achievable.In fact, the ECi incubation times in the range of minutes rendered the samples macroscopically transparent.Our clearing protocol is compatible with immunofluorescent labeling and results in a significant increase in observable tissue thickness when compared to non-cleared samples.
Further improvement is to be expected if the proposed clearing approach is combined with more advanced imaging modalities such as confocal or light-sheet microscopy.

CONCLUSION
We have shown that On-Slide-ECi-clearing is a simple and effective extension for standard immunofluorescent protocols.Future experiments setups targeting specific research questions on different tissues such as colorectal cancer will benefit from the ECi-clearing protocol.

Figure 2 .
Figure 2. Microscopic visualization of cleared and non-cleared specimens; 3D representation of microscopic images of non-cleared (A, B) or cleared (C, D) samples in two orthogonal projections.It can be seen that the detectable signal in non-cleared tissue (A) reaches a depth of approximately 33 µm, while the signal in cleared tissue (C) can be easily detected at a depth of more than 80 µm from the surface (in the orthogonal projections A and C the upper surface is here on the left).Scale bar 100 µm.

Figure 1 .
Figure 1.Macroscopic visualization of cleared and non-cleared specimens.Macroscopic appearance of 300 µm thick sections of human colorectal cancer in traditional non-cleared (A) versus cleared (B) mounted sections.

Figure 3 .
Figure 3. Observable tissue thickness.Statistically significant increase (p<0.001) in the observable tissue thickness after clearing showing an ~2.5-fold increase.

Table 1 .
Data of cases of colorectal cancer patients used in this study

Table 2 .
Detailed clearing protocol with immunofluorescent labeling steps