Visualization of Organ-Specific Lymphatic Growth: An Efficient Approach to Labeling Molecular Markers in Cleared Tissues

Organ-specific lymphatics are essential for the maintenance of healthy organ function and lymphatic dysfunction can lead to the development of various diseases. However, the precise role of those lymphatic structures remains unknown, mainly due to inefficient visualization techniques. Here, we present an efficient approach to visualizing organ-specific lymphatic growth. We used a modified CUBIC protocol to clear mouse organs and combined it with whole-mount immunostaining to visualize lymphatic structures. We acquired images using upright, stereo and confocal microscopy and quantified them with AngioTool, a tool for the quantification of vascular networks. Using our approach, we then characterized the organ-specific lymphatic vasculature of the Flt4kd/+ mouse model, showing symptoms of lymphatic dysfunction. Our approach enabled us to visualize the lymphatic vasculature of organs and to analyze and quantify structural changes. We detected morphologically altered lymphatic vessels in all investigated organs of Flt4kd/+ mice, including the lungs, small intestine, heart and uterus, but no lymphatic structures in the skin. Quantifications showed that these mice have fewer and dilated lymphatic vessels in the small intestine and the lungs. Our results demonstrate that our approach can be used to investigate the importance of organ-specific lymphatics under both physiological and pathophysiological conditions.


Introduction
Lymphatic vessels have long been known to play a role in fluid homeostasis, immune cell trafficking and dietary lipid uptake. Recent studies have shown that lymphatics have additional organ-specific functions and that lymphatic dysfunction can lead to the development of disorders such as obesity, atherosclerosis, lung diseases, lymphedema and further metabolic diseases [1][2][3][4][5][6][7][8][9][10][11].
Lymphatic dysfunction can be caused by mutations of important regulators of lymphangiogenesis, such as Flt4 [12,13]. A mouse model showing symptoms of lymphatic dysfunction is the Flt4 kd/+ strain, carrying a mutated Flt4 allele which leads to the inactivation of the tyrosine kinase activity of vascular endothelial growth factor receptor-3 (VEGFR-3) [14]. These mice develop chylous ascites, show enlarged lymphatics in the intestinal tissue, are missing lymphatic vessels in the skin and suffer from swollen feet [14], but little is known about the lymphatic vasculature of other organs, mainly due to inefficient visualization techniques.
Despite obvious limitations, lymphatic vessels are still widely visualized by immunostaining of sections, although tissue sectioning is time-consuming and sections represent only small parts of a complex organ. Morphological changes of the lymphatic vasculature are difficult to capture and no information can be obtained about the length of lymphatic vessels, the number of lymphatic junctions or the total lymphatic area. Alternative visualization techniques are whole-mount immunostaining or lymphatic reporter mouse models,

The Lymphatic Network Cannot Be Sufficiently Visualized with Conventional Visualization Methods
To test whether conventional visualization techniques can be used to visualize the organ-specific lymphatic vasculature, we stained tissue sections and used a lymphatic reporter mouse model (Figure 1). We first detected the native fluorescence in various organs of Prox1 GFP lymphatic reporter mice expressing the green fluorescent protein (GFP) in lymphatic endothelial cells (Figure 1a). We could detect lymphatic vessels in the ear skin, small intestine, lungs and aorta, but due to the opacity of native tissue we could only visualize superficial lymphatic vessels.
We then performed paraffin-based histology on various organs of adult C57BL/6 mice and stained sections with the lymphatic marker anti-LYVE1 (Figure 1b). Anti-LYVE1staining showed that lymphatic vessels are present in the villi intestinales and in the submucosal layer of the intestinal tissue. We could detect lymphatic vessels around the bronchi of the lungs, in the dermis of the ear skin and around the aorta. Since sections only represent a limited part of an organ, it was impossible to assess the total length or area of lymphatic vessels or to efficiently quantify the organ-specific lymphatic vasculature of those organs.

Optimized Tissue-Clearing Protocol Clears Mouse Organs Quickly and Effectively
Our tissue-clearing protocol was 15 days long and consisted of the following main elements: a transcardial perfusion, tissue collection and post-fixation on day 0, decolorization for 5 days with Reagent 1, whole-mount immunostaining with the primary and secondary

Optimized Tissue-Clearing Protocol Clears Mouse Organs Quickly and Effectively
Our tissue-clearing protocol was 15 days long and consisted of the following main elements: a transcardial perfusion, tissue collection and post-fixation on day 0, decolorization for 5 days with Reagent 1, whole-mount immunostaining with the primary and secondary antibodies in Staining Solutions 1 and 2 and an RI-adjustment step in which the refraction indices of the cellular compartments were matched overnight in Reagent 2 ( Figure 2).  With this approach, we could successfully clear various mouse organs with everyday lab equipment. We tested our protocol with the ear skin, small intestine, lungs, heart with aorta, uterus with ovaries and testicles with epididymis and achieved a high level of transparency in all cases ( Figure 3).
This tissue-clearing protocol is based on the previously described CUBIC protocol With this approach, we could successfully clear various mouse organs with everyday lab equipment. We tested our protocol with the ear skin, small intestine, lungs, heart with aorta, uterus with ovaries and testicles with epididymis and achieved a high level of transparency in all cases ( Figure 3).

Figure 2.
Optimized tissue-clearing protocol. Our optimized tissue-clearing protocol consists of an overnight fixation of the tissue sample, decolorization, whole-mount immunostaining with chosen antibodies and RI-matching before the sample is ready for imaging on day 15. Samples can be stored long-term at 4 °C in Reagent 2. With this approach, we could successfully clear various mouse organs with everyday lab equipment. We tested our protocol with the ear skin, small intestine, lungs, heart with aorta, uterus with ovaries and testicles with epididymis and achieved a high level of transparency in all cases ( Figure 3).
This tissue-clearing protocol is based on the previously described CUBIC protocol [23][24][25]. We modified the original protocol by omitting several steps, adjusting the incubation times and adding whole-mount immunostaining after the decolorization step.

Our Approach Enables Us to Visualize Organ-Specific Lymphatic Growth in Various Mouse Organs
To visualize organ-specific lymphatic growth in healthy animals, we tissue-cleared and stained organs of C57BL/6 wild-type and Prox1 GFP lymphatic reporter mice with various immunostainings ( Figure 4). This tissue-clearing protocol is based on the previously described CUBIC protocol [23][24][25]. We modified the original protocol by omitting several steps, adjusting the incubation times and adding whole-mount immunostaining after the decolorization step.

Our Approach Enables Us to Visualize Organ-Specific Lymphatic Growth in Various Mouse Organs
To visualize organ-specific lymphatic growth in healthy animals, we tissue-cleared and stained organs of C57BL/6 wild-type and Prox1 GFP lymphatic reporter mice with various immunostainings (Figure 4).
We were able to visualize the organ-specific lymphatic vasculature of the ear skin of C57BL/6 mice with an anti-LYVE1-staining ( Figure 4a). Lymphatics are clearly visible in the images at higher magnifications.
We were also able to visualize the organ-specific lymphatic vasculature of the intestine of adult C57BL/6 mice with an anti-LYVE1-staining ( Figure 4b). Images at different magnifications show that the intestinal lymphatic vasculature can be visualized in great detail, even with a stereo microscope. Our approach allows us to visualize the morphology of the lymphatic network. Images of the lumen of the small intestine show the lacteals in the villi intestinales and the lymphatics in the submucosal region.
The organ-specific lymphatic vasculature of the lungs of Prox1 GFP mice could be visualized with GFP staining (Figure 4c). The lymphatic network of the lungs is visible in different magnifications and the specific morphology of the lymphatic vessels is clearly visible. We then visualized the organ-specific lymphatic vasculature of the femoral arteries of C57BL/6 mice with an anti-LYVE1 antibody ( Figure 4d). Images show that the arteries are surrounded by a network of lymphatic vessels.
The lymphatic vasculature of the uteri of C57BL/6 mice was visualized by an anti-LYVE1 staining (Figure 4e). Lymphatic vessels form an organized network surrounding the uterine horns and the cervix.
Visualization of cardiac lymphatics with an anti-GFP antibody and visualization of blood vessels with an anti-vWF ( Figure 4f) or anti-αSMA antibody (Figure 4g) show the great potential of our tissue-clearing protocol. It enables us to label molecular markers with various antibodies in cleared tissues. Confocal microscopy enabled us to visualize the lymphatic vessels surrounding a coronary artery in great detail. We then visualized the lymphatic vasculature of the uteri of Prox1 GFP mice with anti-GFP and anti-vWF antibodies with confocal microscopy (Figure 4h). The image shows the lymphatic vessels and blood vessels of the uterine horn in higher magnification.
Anti-GFP and anti-LYVE1 staining helped us visualize the lymphatic vasculature of the testicle and epididymis of Prox1 GFP mice (Figure 4i).
Furthermore, we found that long-term storage of stained samples in Reagent 2 at 4 • C did not quench fluorescence signals. A strong, preserved fluorescent signal was still detectable 18 months after staining ( Figure 5).

Lymphatic Structures Can Be Quantified with AngioTool, a User-Friendly Tool for the Analysis of Vascular Networks
To simplify the quantification process, we tested AngioTool, a free computational tool which was developed for the quantitative analysis of vascular networks [29]. We quantified the lymphatic junctions and lymphatic end points of four different confocal

Lymphatic Structures Can Be Quantified with AngioTool, a User-Friendly Tool for the Analysis of Vascular Networks
To simplify the quantification process, we tested AngioTool, a free computational tool which was developed for the quantitative analysis of vascular networks [29]. We quantified the lymphatic junctions and lymphatic end points of four different confocal images of the small intestines of Prox1 GFP mice with AngioTool ( Figure 6a). Images were quantified with AngioTool, a user-friendly tool for the quantification of vasculature [29]. Every sample is marked with a specific color to analyze the results of the individual samples. Magnification of the images shows the quantification of AngioTool in detail: blue dots represent lymphatic junctions and red lines mark the length and the end points of lymphatic vessels. The Images were quantified with AngioTool, a user-friendly tool for the quantification of vasculature [29]. Every sample is marked with a specific color to analyze the results of the individual samples. To evaluate whether the quantifications with AngioTool met our requirements, we manually quantified the lymphatic junctions and end points of the same images and compared the results (Figure 6b). To analyze the results of the individual images, we marked all images with a different color. We could not detect a significant difference between the manually quantified images and the images that were quantified with AngioTool.

Flt4 kd/+ Mice Show Morphologically Altered Lymphatic Structures in All Investigated Organs
To visualize the organ-specific lymphatic growth in Flt4 kd/+ mice, we tissue-cleared and stained various organs of Flt4 +/+ and Flt4 kd/+ mice crossed with the Prox1 GFP transgenic reporter strain (Figure 7).
First, we visualized the organ-specific lymphatic growth in the skin of Flt4 +/+ ; Prox1 GFP and Flt4 kd/+ ; Prox1 GFP mice with an anti-GFP antibody and anti-von Willebrand Factor (vWF) (Figure 7a). Flt4 +/+ ; Prox1 GFP mice show a normal blood and lymphatic vessel network in the skin. Flt4 kd/+ ; Prox1 GFP mice show normal blood vessels in the skin, but lymphatic vessels could not be detected.
We then visualized the lymphatic vasculature of the intestines of Flt4 +/+ ; Prox1 GFP and Flt4 kd/+ ; Prox1 GFP mice with an anti-GFP antibody and anti-LYVE1 (Figure 7b). Flt4 +/+ ; Prox1 GFP mice show normal and normal lymphatic vasculature in the small intestine. The GFP-positive nuclei of the lymphatic endothelial cells are clearly visible in the lymphatic structures. In comparison to their control littermates, Flt4 kd/+ ; Prox1 GFP mice show greatly enlarged and dilated lymphatic vessels in the intestine.
Next, we visualized the lymphatic vasculature of the lungs of Flt4 +/+ ; Prox1 GFP and Flt4 kd/+ ; Prox1 GFP mice with an anti-GFP antibody (Figure 7c). Flt4 kd/+ ; Prox1 GFP mice show fewer but dilated lymphatic structures in the lungs in comparison to their littermate control mice. The lymphatic vasculature did not reach the periphery of the lungs.
We then visualized the organ-specific lymphatic vasculature of the heart of Flt4 +/+ ; Prox1 GFP and Flt4 kd/+ ; Prox1 GFP mice with an anti-GFP antibody and the lymphatic marker anti-LYVE1 (Figure 7d). Flt4 kd/+ ; Prox1 GFP mice show extremely dilated lymphatic vessels in the heart.
Finally, we visualized the lymphatic vasculature of the uteri of Flt4 +/+ ; Prox1 GFP and Flt4 kd/+ ; Prox1 GFP mice with an anti-GFP antibody and the smooth muscle marker antialpha smooth muscle actin (αSMA) (Figure 7e) and found that Flt4 kd/+ ; Prox1 GFP mice show extremely dilated and morphologically altered lymphatic vessels in the uterus.

Morphological Changes of the Lymphatic Vasculature Can Be Detected with Our Approach but Not with Conventional Visualization Techniques
To highlight the advantages of our protocol, we quantified images of lung and small intestine sections and images of tissue-cleared lungs and intestines (Figure 8). We performed paraffin-based histology of the lungs and the small intestines of Flt4 +/+ ; Prox1 GFP and Flt4 kd/+ ; Prox1 GFP mice and stained them with an anti-GFP antibody and anti-LYVE1 (Figure 8a). We quantified the images of the lung sections manually with NIS-Elements imaging software (Figure 8b). Quantifications showed no significant differences regarding the number of lymphatics, the total lymphatic lumen and the average lymphatic lumen in the lungs between Flt4 kd/+ ; Prox1 GFP mice and their littermate control mice. We then quantified the images of the small intestine sections manually with NIS-Elements ( Figure 8c). Quantifications showed a significant decrease in the number of intestinal lymphatics in Flt4 kd/+ ; Prox1 GFP mice. Total lymphatic lumen and the average lymphatic lumen were significantly increased in the small intestines of Flt4 kd/+ ; Prox1 GFP mice in comparison to the intestines of their littermate control mice.
To highlight the advantages of our approach, we tissue-cleared and stained lungs and small intestines of Flt4 +/+ ; Prox1 GFP and Flt4 kd/+ ; Prox1 GFP mice and acquired confocal images of the samples (Figure 8d).
Quantifications of the tissue-cleared lungs showed a decreased total lymphatic area, decreased total length of lymphatic vessels, fewer lymphatic junctions and fewer lymphatic end points in the lungs of Flt4 kd/+ ; Prox1 GFP mice in comparison to their controls, but a significant increase in the lymphatic diameter (Figure 8e). Our results suggest that Flt4 kd/+ ; Prox1 GFP mice have significantly fewer but dilated lymphatic vessels in the lungs.
Quantifications of the tissue-cleared small intestines showed no difference in the total lymphatic area but a significant decrease in the total length of lymphatic vessels, fewer lymphatic junctions and fewer lymphatic end points in the small intestines of Flt4 kd/+ ; Prox1 GFP mice in comparison to their controls, but a significant increase in the lymphatic diameter (Figure 8f). These results suggest that Flt4 kd/+ ; Prox1 GFP mice have fewer but dilated lymphatic vessels in the small intestine.
The results show that morphological changes in the lymphatic vasculature, especially changes in lymphatic parameters such as total lymphatic area, total length of lymphatic vessels, lymphatic junctions and end points can be detected effectively with our approach.  Images were acquired by confocal microscopy, scale bar 500 µm. Quantifications of the tissue-cleared lungs (e) and small intestines (f) were quantified with AngioTool and include the total lymphatic area, the total vessel length, lymphatic junctions and lymphatic end points per field of view. Lymphatic diameters were quantified manually using NIS-Elements. Data are represented as mean ± SEM. Samples were statistically analyzed by a two-tailed, unpaired t-test; p = 0.0256 for the total lymphatic area of tissue-cleared lungs, p = 0.0202 for the total vessel length of tissue-cleared lungs, p = 0.0068 for the average lymphatic diameter of tissue-cleared lungs, p = 0.0427 for the lymphatic junctions of tissue-cleared lungs, p = 0.0173 for the lymphatic end points of tissue-cleared lungs, p = 0.9156 for the total lymphatic area of tissue-cleared small intestines, p = 0.0377 for the total vessel length of tissue-cleared intestines, p = 0.0008 for the average lymphatic diameter of tissue-cleared small intestines, p = 0.0181 for the lymphatic junctions of tissue-cleared small intestines and p = 0.0631 for the lymphatic end points of tissue-cleared small intestines; * p < 0.05; ** p < 0.01; *** p < 0.001; n = 3-6 lungs or small intestines of 3-6 mice per group. Representative images of the experiments are shown.

Discussion
In this study, we presented an efficient approach to visualize, analyze and quantify the organ-specific growth of lymphatic vessels. Conventional visualization techniques such as paraffin sections, reporter mouse models and whole-mount staining are not sufficient to visualize the lymphatic vasculature of organs due to the opacity of native tissue (Figure 1) [15]. Tissue-clearing techniques address this problem by clearing pigments and matching RI-indices, creating optically transparent tissue samples that can be used for imaging, but many of the protocols have limitations that make them unsuitable for the visualization of lymphatic vessels [16,18,19,21,22,26,[30][31][32][33][34]. We decided to base our protocol on a non-toxic and efficient tissue-clearing protocol named CUBIC [23][24][25]. We modified the original protocol by omitting steps, changing incubation times and combining it with whole-mount immunostaining to be independent from reporter mouse models (Figure 2).
Our protocol enabled us to sufficiently clear various mouse organs in a reasonable time and without any visible shrinkage or damage to the tissue (Figure 3).
We showed that our protocol works with different fluorescent markers that allow us to visualize organ-specific lymphatic structures (Figure 4). We demonstrated that it can be used for double-staining with different antibody combinations and is suitable for the investigation of other structures of interest, such as blood vasculature (Figure 4f-h). With our approach, we then aimed to visualize and quantify the organ-specific lymphatic vasculature of the Flt4 kd/+ mouse, a mouse model showing symptoms of lymphatic dysfunction. The model was described nearly 20 years ago, but, aside from the lack of lymphatics in the skin and dilated intestinal lymphatics, little is known about the organ-specific lymphatic vasculature of this model [14].
We could confirm that these mice lack lymphatic vessels in the skin and that they have enlarged lymphatic vessels in the small intestine, as previously described [14]. Furthermore, we found that these mice have morphologically altered lymphatic vessels in the lungs, heart and uterus that have not yet been described (Figure 7).
The fluorescence signals of our samples remain stable for more than 18 months. Longterm stability of staining allows flexibility regarding the imaging of samples and even allows re-imaging of a sample after a long period of time if the focus of a study has changed ( Figure 5).
In comparison to other CUBIC protocols [23][24][25][26], we could acquire images of great quality by stereo and confocal microscopy and did not need an expensive and complex LSFM (Figures 4-8).
We focused on simplifying analysis and quantified our images using AngioTool (Figures 6 and 8). AngioTool is a free and user-friendly computational tool specifically developed for the quantification of vascular structures [29]. It quantifies images in seconds by simply uploading them into the program. It allowed us to adjust parameters such as vessel diameter, intensity and particle size and to quantify vessel-relevant parameters such as total vessel area, total vessel length, number of vessel junctions and end points. Results generated with AngioTool showed no significant differences when compared to manual quantifications; therefore, we propose that AngioTool represents a suitable tool for quantifying images of lymphatic vasculature ( Figure 6).
We then quantified sections of the lungs and small intestines of Flt4 kd/+ mice and compared them with quantifications of tissue-cleared lungs and small intestines ( Figure 8). Based on the sections, we could quantify the number of lymphatic vessels, the total lymphatic lumen and the average lymphatic lumen. Results of the small intestine sections showed that lymphatic vessels are dilated in Flt4 kd/+ animals; results of the lung sections on the other hand were inconclusive. Based on the tissue-cleared samples, we were able to quantify more parameters, including lymphatic junctions, lymphatic end points and the total length of lymphatic vessels. Quantifications revealed that lungs and small intestines of Flt4 kd/+ animals have fewer lymphatic junctions and end points and that the total lymphatic area and length of lymphatic vessels is greatly reduced. The significantly increased lymphatic diameter shows that lymphatic vessels in the lungs and the small intestine are dilated, a sign of lymphatic malfunction. We showed that tissue-cleared samples allow a much better assessment of the lymphatic network of an organ than immunostaining of sections. Our protocol enabled us to visualize and quantify the differences between the organ-specific lymphatic vasculature of Flt4 kd/+ mice and their healthy littermate controls.
In this study, we demonstrated that our efficient protocol enabled us to visualize the lymphatic network of the studied organs, to detect differences and morphological alterations of lymphatic structures in mouse models and to quantify those differences. Many tissue-clearing protocols aim for high-resolution 3D-images of tissue-cleared organs but require expensive LSFM and previous knowledge of analyzing 3D-image data that are not readily available in every laboratory and therefore prevent the majority of researchers from using tissue-clearing techniques [18,21,26,30]. Here, we could provide an efficient protocol without the need for any special equipment or knowledge, making tissue-clearing possible for anyone. We showed that Flt4 kd/+ mice have morphologically altered organspecific lymphatic vessels, which suggests that lymphatic vessels are important for the maintenance of healthy organ function. Further functional studies could help clarify the role of organ-specific lymphatic vessels and the impact of morphologically altered lymphatic vessels in the Flt4 kd/+ organs on the lymphedema phenotype.

Mouse Models
In this study, we used 3-to 5-month-old male and female C57BL/6 and Prox1 GFP BAC transgenic lymphatic reporter mice to visualize lymphatic vasculature [35]. Prox1 GFP mice were maintained on a C57BL/6 genetic background in heterozygous form and genotyped by a transgene specific PCR using 5 -GAT GTG CCA TAA ATC CCA GAG CCT AT−3 forward and 5 -GGT CGG GGT AGC GGC TGA A−3 reverse primers. Additionally, we used 3-to 5month-old Flt4 kd/+ mice, a previously described model for primary lymphedema, and their littermate control mice [14]. Mice carrying the kinase-dead

Microscopic Imaging and Processing
On day 15, samples were imaged with a stereo microscope connected to a camera or a confocal microscope (Nikon Instruments, A1 HD25, Tokyo, Japan) connected to a confocal scanner unit (Yokogawa Electric Corporation, CSU-W1, Tokyo, Japan). Images were processed and analyzed using NIS-Elements imaging software (Nikon Instruments, version BR 4.60.00).

Manual Quantification of Vascular Structures
Manual quantifications were performed using NIS-Elements. For the quantification of sections, lymphatic number, total lymphatic lumen and average lymphatic lumen of 5 fields of view (20× magnification) were quantified per mouse. For the quantification of the lymphatic diameter of tissue-cleared lungs and intestines, a minimum of 40 lymphatic vessels (10× magnification) were measured per mouse.

Quantification of Vascular Structures with AngioTool
Total lymphatic area, total length of lymphatic vessels, lymphatic junctions and lymphatic end points were quantified with AngioTool, a free computational tool that has been developed for the quantitative analysis of vascular networks [29]. After uploading the images to AngioTool, we adjusted parameters such as the vessel diameter and intensity and removed small particles from the calculation. All images were quantified using the same parameters.

Data Presentation and Statistical Analysis
Representative images of the experiments are shown. Data were processed and statistically analyzed using GraphPad Prism (version 7.03) and Excel (Microsoft, version 2018).  Data Availability Statement: Datasets for this study are available from the corresponding author upon reasonable request.