Three-Dimensional Visualization of Lymphatic Vessels in Human Skin and Their Structural Changes During Lymphedema

There is increasing interest in the function of lymphatic vessels since the recognition of their role in maintaining skin homeostasis, beyond their well-established involvement in cancer metastasis^1,2. Lymphatic vessels are structurally divided into capillaries and collecting vessels, each possessing distinct characteristics. Lymphatic capillaries consist of a single-layered structure formed by lymphatic endothelial cells and serve as initial collection sites for interstitial fluid, waste products, and immune cells. In contrast, collecting lymphatic vessels have a bilayered structure with smooth muscle cells surrounding the lymphatic endothelial layer, enabling active transport of lymph fluid collected by the capillary network toward the central circulation³. Mouse models have been utilized to elucidate the structural features, particularly of bronchial lymphatic vessels, and have revealed how inflammatory conditions induce structural remodeling⁴. Previous work demonstrated that ultraviolet-induced inflammation causes abnormal dilation of lymphatic capillaries, which compromises their fluid collection function. The molecular mechanisms underlying this dysfunction were identified, specifically showing that increased VEGF-A expression coupled with decreased VEGF-C expression drives these structural alterations^5,6,7.

While three-dimensional visualization techniques have significantly advanced understanding of the lymphatic network in mouse models^8,9, comparable structural analysis of human lymphatic vessels has remained challenging. This analytical gap arises from the difficulties of working with human tissues, including limited sample availability and the technical issues involved in achieving comprehensive three-dimensional imaging. In particular, the skin, characterized by a three-layer structure comprising the epidermis, dermis, and subcutaneous fat, and functioning as a barrier to the external environment, presents greater challenges than other soft organs. Furthermore, human skin presents distinct analytical challenges relative to mouse models, including greater tissue thickness, hampering deep imaging penetration. In addition, the abundant extracellular matrix components, particularly collagen, along with their age-related structural alterations, create optical barriers that reduce the efficacy of emerging tissue-clearing techniques that have revolutionized lymphatic imaging in mice¹⁰.

Various tissue-clearing agents were previously tested to visualize human skin capillaries, demonstrating their tissue compatibility¹¹. Following that work, tissue-clearing reagents were successfully used to visualize human skin lymphatic vessels in three dimensions¹². By focusing on the adherens junctions of lymphatic endothelial cells, structural changes in lymphatic vessels associated with the progression of lymphedema were identified¹³. The purpose of this article is to describe in detail the procedures for three-dimensional visualization of human skin lymphatic vessels using combinations of three skin-clearing techniques, along with confocal microscopy and light-sheet microscopy, and immunohistochemical staining tailored to specific research objectives. In this study, although we focus solely on fresh human skin tissues that were fixed on the day of excision and have not undergone freeze-thaw cycles, these techniques provide insights into not only the structural changes of lymphatic vessels during disease and aging, but also the associated functional alterations.

Skin samples of human subjects were acquired from the Chiba University Hospital (Chiba, Japan). All procedures involving human subjects were approved by the Institutional Review Board of the Chiba University Hospital and the Shiseido Global Innovation Center, and all subjects provided written informed consent. The reagents and the equipment used are listed in the Table of Materials.

1. Pretreatment of human skin samples

1. Remove the subcutaneous tissue of excised specimens upon acquisition.
   NOTE: The interval between skin excision and acquisition was approximately h. During transport, the tissue was immersed in DMEM and kept on ice packs.
2. Place the samples with the epidermal side down on a 60 °C hot plate for 45 s to separate the epidermis and dermis.
3. Immediately immerse the specimens in ice-cold PBS.
4. Separate the epidermis and dermis using tweezers.
5. Fix the dermal parts in 4% paraformaldehyde at 4°C for 16 h, and subsequently wash with PBS prior to use.
   NOTE: After fixation, the samples were stored in PBS at 4 °C, and the immunostaining steps were initiated within months.
   CAUTION PFA solution is toxic. All preparation and handling must be performed in a fume hood while wearing appropriate personal protective equipment.

2. Whole-mount immunostaining combined with clearing reagents

1. iDISCO technique
   1. Trim the specimens to approximately 5 mm × 5 mm × 5 mm in size.
   2. Transfer the specimens into 5% Triton X-100 and 2.5% Tween 20 in PBS on an orbital shaker set to 50 rpm at 24 °C for 16 h for permeabilization.
   3. Wash the specimens with PBS at 24 °C for 1 h, repeating the process three times.
   4. Immerse the specimens in freshly prepared blocking buffer A (1% BSA, 0.1% Tween 20, 0.03% sodium azide in PBS) on an orbital shaker set to 50 rpm at 24 °C for 72 h.
   5. Incubate the specimens with primary antibody against podoplanin (1:100) in a blocking buffer A on an orbital shaker set to 50 rpm at 37 °C for 72 h.
   6. Wash the specimens with washing buffer A (0.1% Tween-20 in PBS) at 24 °C for 1 h, repeating the process three times. Then, maintain the specimens in washing buffer A on an orbital shaker set to 50 rpm at 4 °C for 72 h.
   7. Incubate the specimens with secondary antibody against mouse IgG conjugated to Alexa Fluor 647 (1:1000) in a blocking buffer A on an orbital shaker set to 50 rpm at 37 °C for 72 h.
   8. Repeat step 1.6.
   9. Incubate the specimens with 50% MeOH at 24 °C for 3 h.
   10. Incubate the specimens with 70% MeOH at 24 °C for 3 h.
   11. Incubate the specimens with 100% MeOH at 24 °C for 16 h.
   12. Incubate the specimens with dichloromethane at 24 °C for 15 min, repeating the process twice.
   13. Clear the specimens with dibenzyl ether at 24 °C for 16 h.
       NOTE: If the tissue has not become transparent at this point, further waiting will result in no significant improvement.
   14. Capture images of the specimens with a light-sheet microscope.
       NOTE: Use a Lightsheet with an EC Plan-Neofluar 5×/0.16 objective; set illumination (excitation) at 63 nm and detection at 648 nm; acquire a Z-stack with a step size of 5.8 µm.
       CAUTION: Methanol, dichloromethane, and dibenzyl ether are hazardous. All preparation and handling must be performed in a fume hood while wearing appropriate personal protective equipment.
2. Rapiclear 1.52 reagent
   1. Trim the specimens to approximately 5 mm × 5 mm × 2 mm in size.
   2. Transfer the specimens into 2% Triton-X 100 in PBS on an orbital shaker set to 50 rpm at 24 °C for 16 h for permeabilization.
   3. Immerse the specimens in freshly prepared blocking buffer B (5% donkey serum and 1% Triton-X 100 in PBS) on an orbital shaker set to 50 rpm at 4 °C for 16 h.
   4. Incubate the specimens with primary antibodies against podoplanin and VE-cadherin (both at 1:100) in a blocking buffer B on an orbital shaker set to 50 rpm at 4 °C for 72 h.
   5. Wash the specimens with washing buffer B (3% NaCl and 0.2% Triton-X 100 in PBS) at 24 °C for 1 h, repeating the process three times. Then, maintain the specimens in washing buffer B on an orbital shaker set to 50 rpm at 4 °C for 16 h.
   6. Wash the specimens with PBS at 24 °C for 1 h.
   7. Incubate the specimens with secondary antibodies against mouse IgG and rabbit IgG, conjugated to Alexa Fluor 647 and 568, respectively (both at 1:1000), in blocking buffer B on an orbital shaker set to 50 rpm at 4 °C for 16 h.
   8. Repeat steps 1.5. and 1.6.
   9. Clear the specimens with at least 5 times the sample volume of Rapiclear 1.52 at 24 °C for 16 h.
      NOTE: The clearing of the peripheral regions of the tissue was observed in less than 1 h.
   10. Embed the cleared specimens in a polydimethylsiloxane gel chamber placed on a coverslip with fresh Rapiclear 1.52 reagent.
   11. Image the specimens with a confocal microscope.
       NOTE: Use a confocal microscope with a Plan-Apochromat 10×/0.45 M27 objective; set excitation at 633 nm and 561 nm, and detection range 641-754 nm and 566-628 nm, with a GaAsP detector; acquire a Z-stack with a step size of 3.15 µm.
3. CUBIC technique
   1. Trim the specimens to approximately 5 mm × 5 mm × 3 mm in size.
   2. Immerse the specimens in 50% CUBIC-L at 24 °C for 24 h as a pretreatment for delipidation.
   3. Transfer the specimens into 100% CUBIC-L on an orbital shaker set to 50 rpm at 37 °C for 7 days for delipidation.
   4. Wash the specimens with PBS at room temperature for 24 h.
   5. Immerse the specimens in freshly prepared blocking buffer A on an orbital shaker set to 50 rpm at 24 °C for 72 h.
   6. Incubate the specimens with primary antibody against VE-cadherin, podoplanin, and CD31 (1:100, 1:100, and 1:200, respectively) in blocking buffer A on an orbital shaker set to 50 rpm at 37 °C for 72 h.
   7. Wash the specimens with washing buffer A at 24 °C for 1 h. Repeat the process three times, then maintain the specimens in washing buffer A on an orbital shaker set to 50 rpm at 4 °C for 72 h.
   8. Incubate the specimens with secondary antibodies against mouse IgG, rabbit IgG, and sheep IgG, conjugated to Alexa Fluor 647, 568, and 488, respectively (all at 1:1000), in blocking buffer A on an orbital shaker set to 50 rpm at 37 °C for 72 h.
   9. Repeat step 3.7.
   10. Pre-clear the specimens with 50% CUBIC-R+(M) at 24 °C for 16 h.
   11. Clear the specimens with 100% CUBIC-R+(M) at 24 °C for 16 h.
       NOTE: If the tissue has not become transparent at this point, further waiting will result in no significant improvement.
   12. Image the specimens with a confocal microscope.
       ​NOTE: Use a confocal microscope with a Plan-Apochromat 20×/0.8 M27 objective; set excitation at 633 nm, 561 nm, and 488 nm, and detection range 641-754 nm, 566-628 nm, and 479-550 nm, with a GaAsP detector; acquire a Z-stack with a step size of 0.78 µm.

Difference in light transmittance of cleared human skin tissue with and without epidermis
Figure 1 shows representative photographs of human skin specimens with and without epidermis, following permeabilization and clearing with Rapiclear 1.52 reagent. Epidermal removal enhances tissue transparency by eliminating melanin, a potent light absorber, as confirmed by the enhanced visibility of the black line beneath the samples.

Three-dimensional macro-scale visualization of lymphatic vessels in human skin
By combining tissue clearing using the iDISCO technique with light sheet microscopy, three-dimensional macroscopic visualization of lymphatic vessels in human skin tissue was achieved through immunostaining of podoplanin, a major lymphatic vessel marker14 (Figure 2). This method provides comprehensive information on the overall lymphatic network structure, rather than detailed structural features. The images reveal dense lymphatic capillary networks beneath the epidermis that converge into larger pre-collecting lymphatic vessels in the deeper dermis, consistent with established anatomical descriptions15.

Three-dimensional micro-scale visualization of lymphatic vessels in human skin
To examine the structure of lymphatic capillaries in human skin in fine detail, the combination of podoplanin immunostaining with the tissue clearing reagent Rapiclear and confocal laser microscopy was employed (Figure 3). In the current study, skin samples from secondary lymphedema patients with disease severity ranging from type 1 (subclinical lymphedema) through type 2 (mild lymphedema) and type 3 (moderate lymphedema) to type 4 (severe lymphedema) were used. Despite the tissue alterations, including fibrosis associated with the progression of lymphedema, this approach enables high-resolution three-dimensional reconstruction suitable for quantitative morphometric analysis of lymphatic vessel volume, branching number, and diameter13,16.

Visualization of adherens junction structures in human skin lymphatic vessels
To investigate the cell-cell adhesion structures formed by lymphatic endothelial cells, immunostaining for VE-cadherin, a transmembrane protein of adherens junctions17, was employed. The CUBIC tissue clearing technique combined with confocal laser microscopy enabled visualization of VE-cadherin in lymphatic capillaries from lymphedema patients (Figure 4)13. These results show that the expression pattern of VE-cadherin in lymphatic capillary endothelium is transformed from the characteristic discontinuous button-like junctions to a continuous zipper-like structure with increasing severity of lymphedema. Consequently, this visualization makes it possible to predict potential impacts on drainage, a critical function of lymphatic vessels. It should be noted that the Rapiclear reagent was less suitable than the CUBIC technique for VE-cadherin visualization under the experimental conditions described in this protocol (Figure 5). In contrast, the immunostaining of podoplanin, used for visualizing lymphatic vessel structures in Figure 2 and Figure 3, showed lower compatibility with the CUBIC technique as compared to Rapiclear (Figure 6). When the CUBIC technique was used, it was necessary to confirm the authenticity of the low-resolution signals of podoplanin through colocalization with CD31, a general endothelial cell marker18 (Figure 6). These observations underscore the importance of matching clearing protocols to specific molecular targets and experimental objectives.

Figure 1: Cleared human skin tissue with and without epidermis. Representative photographs of human skin specimens with (left) and without (right) epidermis, taken with the epidermal side facing upward before (upper) and after (lower) clearing treatment with Rapiclear 1.52 reagent following permeabilization. Scale bar = 5 mm. Please click here to view a larger version of this figure.

Figure 2: Three-dimensional macro-scale visualization of lymphatic vessels in human skin. Representative image of podoplanin-positive lymphatic vessels (white) in healthy human skin, cleared using the iDISCO technique and visualized with light-sheet microscopy. Scale bar = 300 µm. Please click here to view a larger version of this figure.

Figure 3: Three-dimensional micro-scale visualization of lymphatic vessels in human skin. Representative three-dimensional images of podoplanin-positive lymphatic capillaries (white) in lymphedematous skin (Types 1-4) and in healthy subjects (Normal), cleared using the Rapiclear 1.52 reagent and visualized with confocal microscopy. Type 1, subclinical lymphedema; Type 2, mild lymphedema; Type 3, moderate lymphedema; Type 4, severe lymphedema. Scale bars = 400 µm. This figure is adapted from Itai et al.13. Please click here to view a larger version of this figure.

Figure 4: Visualization of lymphatic vessel adherens junction structures in human skin. Representative images showing the distribution of VE-cadherin (red) on lymphatic capillaries (indicated by dotted line) in lymphedematous skin (Types 2 and 4) and in healthy subjects (Normal), cleared using the CUBIC technique and visualized with confocal microscopy. Type 2, mild lymphedema; Type 4, severe lymphedema. Scale bars = 50 µm. This figure is adapted from Itai et al.13. Please click here to view a larger version of this figure.

Figure 5: Poor VE-cadherin staining using the Rapiclear reagent. Representative image of immunostained VE-cadherin (red) in human skin, cleared using the Rapiclear 1.52 reagent. A lymphatic capillary is indicated by dotted lines. Scale bar = 30 µm. Please click here to view a larger version of this figure.

Figure 6: Poor podoplanin staining using the CUBIC technique. Representative images of immunostained podoplanin (white) and CD31 (green), a pan-endothelial marker, in human skin, cleared with the CUBIC technique. Scale bars = 100 µm. Please click here to view a larger version of this figure.

Three-dimensional visualization of lymphatic networks in biological tissues is critically important for understanding their structural characteristics and physiological functions, as well as pathological changes. This protocol presents three distinct tissue clearing methods optimized for human skin tissue, each tailored for specific observational objectives.

Epidermal removal is considered a critical step for eliminating melanin, which reduces tissue transparency and interferes with histological assessment¹⁹. Consistent with this, immunostaining of human skin samples with the epidermis intact using the iDISCO method showed a weak signal, particularly just beneath the epidermis, suggesting that antibody penetration and laser transmission from the epidermal side were hindered. Additionally, a brief heat stimulation was utilized to remove the epidermis in this protocol. This approach is thought to cause less damage to skin tissue structure and cell membrane proteins than enzymatic methods, such as those employing dispase²⁰.

In the present study, three distinct tissue clearing techniques, including Rapiclear, CUBIC, and iDISCO, were utilized. Regarding the use of Rapiclear reagents, the protocol is primarily based on the manufacturer's recommendations; however, the following modifications were implemented: (1) the tissue thickness is reduced to approximately 2 mm to ensure efficient reagent permeation, (2) the dehydration and rehydration steps are omitted, (3) antibody incubation is performed at 4°C for approximately half the duration specified in the product datasheet. Regarding the use of the CUBIC technique^21,22, CUBIC-L reagent was selected for delipidation of human skin tissue, as the more potent delipidation agent, CUBIC-HL, was found to compromise tissue structural integrity. This method achieved higher transparency than the Rapiclear method, likely due to the inclusion of optimal chemicals for delipidation and decoloring^23,24. However, CUBIC showed low compatibility with immunostaining for podoplanin, a major lymphatic vessel marker¹⁴, under the experimental conditions described in this protocol. Although lymphatic vessel endothelial hyaluronan receptor 1 (LYVE1) and prospero homeobox protein 1 (PROX1) are also commonly used as lymphatic vessel markers^25,26, the expression of LYVE1 in lymphatic endothelial cells decreases in response to inflammatory factors and in lymphedema skin tissue^13,27, and PROX1 did not yield clear staining results in this protocol. Therefore, podoplanin was chosen as the marker for lymphatic vessels in this study. The iDISCO technique, in contrast to the two aforementioned methods employing water-soluble reagents, utilizes hydrophobic solutions for tissue clearing, leading to rapid fluorescence fading²⁸. Therefore, samples cleared using this method are better suited for observation with light-sheet microscopy, which minimizes fluorescence fading by illuminating the sample parallel to the detection plane, rather than confocal laser microscopy, which illuminates the sample vertically to the detection plane¹¹. In all methods, incomplete transparency may occur, particularly in the central regions of the tissue. Reducing tissue thickness as much as possible could serve as a potential solution to this issue. Furthermore, it is crucial to ensure that the tissue is fully immersed in the solution throughout each incubation step. Particular attention should be given to the potential changes in concentration and volume due to evaporation, especially during incubation in methanol in the iDISCO protocol.

Considering the features of these clearing methods, the iDISCO method provided optimal results for whole-tissue macro-scale visualization when combined with light-sheet microscopy, and it can be performed with relatively inexpensive reagents. In contrast, while Rapiclear is more limited in terms of specimen size compared to iDISCO, it demonstrated superior antibody compatibility and maintained reasonably high-resolution imaging capabilities. The CUBIC method demonstrated limited compatibility with podoplanin immunostaining; however, it achieved the highest transparency levels and allowed for observation with confocal laser microscopy, making it suitable for structural adhesion studies. Although the three-dimensional observation of lymphatic vessels in human skin, including diseased skin, has been reported using a benzyl alcohol/benzyl benzoate solution (BABB) combined with light-sheet microscopy²⁹, the detailed visualization of lymphatic vessel adhesion structures achieved in this study using the CUBIC method and confocal laser microscopy represents a key differentiating point.

The main limitation of this protocol is the limited variety of human skin tissue types examined. The effectiveness of this protocol has been verified on skin tissues from different anatomical sites and age groups, including UV-exposed cheek skin and UV-nonexposed buttock skin, as well as pathological skin samples derived from lymphedema patients with fibrosis, regardless of severity^12,13. However, it remains unclear whether the same protocol can be applied to cases that may present greater challenges for three-dimensional visualization, such as scleroderma or melanoma. Additionally, all skin tissues used in this protocol were derived from Asian and Caucasian donors, and tissues from donors of other ethnic backgrounds with different skin characteristics have not been examined. Depending on variations in skin conditions, additional treatments, such as hydrogen peroxide processing or adjustments to incubation conditions, may be required.

In conclusion, the current protocol enables the visualization of lymphatic networks in human skin by using tissue-clearing reagents combined with confocal or light-sheet microscopy. The three methods presented here are applicable to human skin and can be used in combination with immunostaining of lymphatic markers such as podoplanin, as well as adherens junction molecules like VE-cadherin, thereby advancing the understanding of lymphatic vessel structure and function in human tissues. The most appropriate method depends upon the experimental target. This work also provides a foundation for future modifications to address other molecules of interest and other experimental objectives.