Regular Article
Comparative Proteomic Analysis of Endocytic Cell-Surface Proteins in Vascular and Lymphatic Endothelial Cells
2025 Volume 48 Issue 8 Pages 1224-1232
Details
2025 Volume 48 Issue 8 Pages 1224-1232
Vascular endothelial cells (VECs) and lymphatic endothelial cells (LECs) regulate the homeostasis of fluids, nutrients, and immune cells in the blood and lymphatic systems, respectively. These specialized functions depend on distinct cell-surface proteins and tightly regulated endocytic pathways, but the molecular determinants underlying the cell-type-specific endocytic profiles of these cells remain unclear. We sought to quantitatively characterize and compare the endocytic cell-surface proteomes of human umbilical vein endothelial cells (HUVECs) and human dermal LECs (HDLECs) using cell-surface biotinylation and internalization assays combined with sequential window acquisition of all theoretical fragment ion spectra-MS–based quantitative proteomics. HUVECs and HDLECs had a similar number of cell-surface and internalized proteins, with substantial overlap, but also included sets unique to each cell type. We identified 32 HUVEC- and HDLEC-enriched endocytic cell-surface proteins, respectively; these proteins were present in both enriched cell-surface and internalized fractions, representing cell-type-selective endocytic proteins. Functional enrichment analysis showed that HUVEC-enriched proteins were associated with angiogenesis, nutrient uptake, and metabolism, whereas HDLEC-enriched proteins were linked to immune regulation, extracellular matrix organization, and lymphangiogenesis. In conclusion, our present study demonstrates that VECs and LECs possess distinct endocytic cell-surface protein profiles that define their specialized functions and represent promising targets for endothelial-selective drug delivery and imaging.
The vascular system, comprising the blood and lymphatic vessels, is fundamental to human physiology, as it facilitates fluid, nutrient, and immune cell transport.1) Despite their shared mesenchymal origin, vascular endothelial cells (VECs) and lymphatic endothelial cells (LECs) exhibit distinct morphological and molecular characteristics. VECs line the interior surfaces of blood vessels and are crucial for regulating blood flow, vascular permeability, and hemostasis.2,3) LECs form the lining of lymphatic vessels and are significant in fluid homeostasis and immune cell trafficking.4,5)
In endothelial cells, cell-surface proteins are vital for maintaining the integrity of the vascular and lymphatic systems by serving as selective barriers that regulate molecular transport and permeability, while also facilitating essential signal transduction and intercellular communication. The internalization of the cell-surface proteins is a fundamental cellular process with diverse physiological roles, ranging from nutrient acquisition and signal regulation to maintaining cellular homeostasis and responding to environmental changes.6–9) Differences in the internalization of cell-surface proteins between VECs and LECs are thought to contribute to their specialized functions in the circulatory and lymphatic systems. In addition, a drug delivery system (DDS) can be used to improve vascular and lymphatic pathologies by selectively delivering drugs and therapeutic modalities specifically to VECs and LECs. Furthermore, the selective delivery of imaging agents for diagnostic purposes enables the early detection of anomalies in the vascular and lymphatic systems. Thus, understanding these endocytic cell-surface proteins is crucial, as it may reveal the molecular mechanisms underlying endothelial specialization and aid in identifying potential targets for selective drug delivery.
Only a few comparative studies have been conducted on the internalization of cell-surface proteins by human VECs and LECs. Human umbilical vein endothelial cells (HUVECs) are widely used for in vitro studies of vascular dynamics and angiogenesis.10,11) In contrast, human dermal LECs (HDLECs), isolated from lymphatic vessels in tissues such as the skin, serve as models for in vitro investigations into lymphatic-specific functions.12–14) Sequential window acquisition of all theoretical fragment ion spectra (SWATH) is a data-independent acquisition (DIA) proteomics technique that facilitates comprehensive and consistent protein quantification.15) We previously demonstrated that our integrated approach—combining cell-surface biotinylation, in vitro internalization assays, and SWATH-MS–based quantitative proteomics—can distinguish internalized protein profiles in hCMEC/D3 cells (i.e., a brain microvascular endothelial cell model) vs. HUVECs, thereby identifying proteins preferentially endocytosed at the blood–brain barrier.16)
In this study, we performed a comparative analysis of endocytic cell-surface proteins in HUVECs and HDLECs using our proteomic workflow coupled with functional enrichment analysis to elucidate the functions of VECs and LECs. These findings enhance our understanding of endothelial specialization and offer insights into the development of targeted therapies for vascular and lymphatic diseases.
HUVECs and HDLECs were purchased from PromoCell (Heidelberg, Germany). HUVECs and HDLECs were cultured using the Endothelial Cell Growth Medium Kit and the Endothelial Cell Growth Medium MV2 Kit (both from PromoCell), respectively. Cells were maintained at 37°C in a humidified atmosphere containing 5% CO2.
Internalization AssayAn internalization assay was performed as previously described.16) Briefly, HUVECs and HDLECs were seeded on 100-mm collagen I-coated dishes and were cultured until they reached subconfluence. Cells were then washed with ice-cold Hanks’ balanced salt solution (HBSS) and incubated with 0.4 mM EZ-Link Sulfo-NHS-SS-Biotin (Thermo Fisher Scientific, Waltham, MA, U.S.A.) in ice-cold HBSS for 30 min at 4°C; this was done to biotinylate cell-surface proteins. Unreacted EZ-Link Sulfo-NHS-SS-Biotin was then neutralized by incubating the cells with 100 mM glycine in ice-cold HBSS, yielding the cell-surface fraction. Cells treated only with ice-cold HBSS, without biotin labeling, served as the control fraction.
For endocytosis, the biotinylated cells were washed with warm HBSS and subsequently incubated with 20% human serum (Cosmo Bio Co., Ltd., Tokyo, Japan) in HBSS for 5 min at 37°C. The cells were then washed 3 times with ice-cold HBSS. To remove residual biotinylated proteins from the cell surface, the cells were incubated with 50 mM 2-mercaptoethane sulfonate (MESNA) buffer for 20 min at 4°C (i.e., this step was repeated 3 times), yielding the internalization fraction. Cell-surface biotinylated cells treated only with MESNA buffer served as the stripping fraction.
Cells from each fraction were harvested, collected, and lysed in an ice-cold radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitors (Sigma-Aldrich, St. Louis, MO, U.S.A.). The lysates were subjected to sonication and centrifugation, and the resulting supernatants were collected for subsequent analysis. Biotinylated proteins were isolated using streptavidin magnetic beads (Thermo Fisher Scientific). Following incubation at room temperature for 1 h, the beads were washed sequentially with RIPA buffer and RIPA buffer containing 0.5 M NaCl. Proteins were eluted by reducing disulfide bonds with 50 mM dithiothreitol (DTT) in a phase transfer surfactant (PTS) buffer at room temperature for 1 h. The eluted proteins were subsequently analyzed using Western blotting and quantitative proteomics.
Western BlottingProtein samples were mixed with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer without reducing agents and separated on a 5–20% gradient SDS-polyacrylamide gel. Subsequently, the proteins were transferred onto polyvinylidene difluoride membranes. The membranes were blocked with 3% bovine serum albumin in Tris-buffered saline with Tween-20 (TBST) for 45 min at room temperature. Next, the membranes were incubated with streptavidin–horseradish peroxidase (HRP) conjugate (1 : 10000 dilution) in Can Get Signal (TOYOBO, Osaka, Japan) for 1 h at room temperature. After washing with TBST, the proteins were detected using an enhanced chemiluminescence substrate (Western Blot Hyper HRP Substrate, TaKaRa Bio Inc., Shiga, Japan). The membrane was imaged using the Omega Lum G imaging system (Aplegen, San Francisco, CA, U.S.A.).
Silver StainingEluted protein samples from the streptavidin beads were mixed with SDS-PAGE sample buffer and separated on a 5–20% gradient SDS-polyacrylamide gel. The proteins were visualized using a Silver Staining Kit (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan), according to the manufacturer’s instructions. Images were captured using the Omega Lum G imaging system.
Quantitative Proteomic AnalysisThe proteins from each fraction were digested with trypsin using the PTS method.17) The digested peptides were analyzed using SWATH-MS on a TripleTOF 5600 instrument (SCIEX, Marlborough, MA, U.S.A.) coupled with a DIONEX Ultimate 3000 RSLCnano system (Thermo Fisher Scientific), as previously described.16) Protein identification and quantification were performed using the library-free search function in DIA-NN 1.818) with the UniProt human reference proteome database. Peptide intensities were normalized using retention time-dependent cross-run normalization, and protein concentrations were calculated from specific peptides using the MaxLFQ algorithm,19) which is integrated into DIA-NN. Peptides and proteins were filtered at a false discovery rate of less than 1% for identification and quantification.
Identification of Cell-Surface Proteins and Internalized ProteinsThe resultant endocytic proteins were classified using the criteria derived from the DIA-NN peptide data. First, cell-surface proteins were identified using the following criteria:
Among the cell-surface proteins, endocytic cell-surface proteins were classified based on the following criteria:
Proteins were identified from the selected peptides that satisfied the aforementioned criteria, and their abundances were determined using protein quantification data generated via DIA-NN.18)
Statistical AnalysisNumerical data were expressed as mean ± standard error of the mean. Inter-group comparisons were conducted using Welch’s t-test in Microsoft Excel (Microsoft Corp., Redmond, WA, U.S.A.). Functional enrichment analysis of the differentially expressed proteins was performed using Metascape.20) Graphical representations were generated using GraphPad PRISM9 (GraphPad Software, San Diego, CA, U.S.A.).
HUVECs and HDLECs were used as model cells to investigate the molecular characteristics of VECs and LECs (Figs. 1A and 1B). To identify endocytic cell membrane proteins in HUVECs and HDLECs, we performed cell-surface biotinylation, followed by internalization assays, where cells were incubated for 5 min at 37°C in 20% human serum and subsequently subjected to cell-surface biotin removal. This procedure yielded 4 distinct fractions, namely, Control, Cell-surface, Stripping, and Internalization (Fig. 1C). Western blotting of the cell lysates from each fraction of HUVECs and HDLECs confirmed that biotinylated proteins were predominantly detected in the cell-surface fraction and subsequently in the internalization fraction (Figs. 1D and 1E). No significant bands were observed in the control and stripping fractions (Figs. 1D and 1E). To isolate biotinylated proteins from each fraction, the proteins were 1st bound to streptavidin beads and eluted with DTT to cleave the disulfide bonds connecting biotin to the proteins. Silver staining revealed a higher abundance of protein bands in the cell-surface fractions than in the control fractions in HUVECs and HDLECs (Figs. 1F and 1G). In the internalization fraction, certain bands exhibited molecular weights identical to those of the cell-surface fraction, but not in the stripping fraction (Figs. 1F and 1G). These findings suggest that biotinylated cell-surface proteins internalized into HUVECs and HDLECs were successfully isolated. In contrast, silver staining revealed that some protein bands in the cell-surface, stripping, and internalization fractions had similar intensities (Figs. 1F and 1G). This likely reflects incomplete biotin removal by the stripping buffer, residual nonspecific biotin labeling of cell-surface proteins, and/or nonspecific binding of proteins to streptavidin beads (Figs. 1F and 1G). Consequently, proteomic analysis of cell-surface and internalized fractions alone is insufficient for accurate characterization of these protein populations.
(A, B) Representative images of HUVECs (A) and HDLECs (B). Scale bar: 200 μm. (C) Schematic of the cell-surface biotinylation/internalization assay. Four conditions are shown: control: no cell-surface biotinylation; cell-surface: biotinylation with EZ-Link Sulfo-NHS-SS-Biotin; stripping: MESNA treatment after cell-surface biotinylation to remove cell-surface biotin; internalization: after cell-surface biotinylation, cells were incubated at 37°C for 5 min in HBSS containing 20% human serum to facilitate the internalization of the cell-surface proteins, followed by MESNA treatment to remove residual cell-surface biotin. (D, E) Western blotting of biotinylated proteins in the cell lysates of HUVECs (D) and HDLECs (E). Equal amounts of cell lysate from each fraction were separated using SDS-PAGE and probed with HRP-labeled streptavidin. Arrowheads mark bands present in both the cell-surface and internalization fractions. (F, G) Silver staining of proteins eluted from the streptavidin beads in each fraction of HUVECs (F) and HDLECs (G). Biotinylated proteins were eluted with DTT, separated by SDS-PAGE, and visualized using silver staining. Black arrowheads indicate bands present in the cell-surface and internalization fractions but absent in the stripping fraction. White arrowheads indicate bands present in both the stripping and internalization fractions. DTT: dithiothreitol; HDLECs: human lymphatic endothelial cells; HUVEC: human vascular endothelial cells.
To identify cell-surface and internalized proteins in HUVECs and HDLECs, we performed quantitative proteomics on the control, cell-surface, stripping, and internalization fractions using SWATH-MS. Cell-surface proteins and internalized proteins were then selected based on peptide abundance using unique criteria (Fig. 2A). In HUVECs, 7888 and 1348 peptides were identified in the cell-surface and internalization fractions, respectively, with 1200 peptides common to both fractions (Fig. 2B). From these peptides, 862 and 254 proteins were identified as cell-surface and internalized proteins, respectively (Fig. 2B and Supplementary Table 1). Thus, 29.5% of the cell-surface proteins were internalized by the HUVECs. In HDLECs, 7073 and 1734 peptides were identified in the cell-surface and internalization fractions, respectively, with 1496 peptides common to both fractions (Fig. 2C). From these peptides, 821 and 290 proteins were identified as cell-surface and internalized proteins, respectively (Fig. 2C and Supplementary Table 1). Thus, 35.3% of the biotinylated cell-surface proteins were internalized into HDLECs. The abundance patterns of cell-surface and internalized proteins identified in HUVECs and HDLECs showed similar trends (Figs. 2D–2G), with high-intensity peaks for a small subset of proteins and a broad distribution of lower-abundance proteins. Overall, these results suggest that HUVECs and HDLECs have an approximately equal number of cell-surface and internalized proteins.
(A) Schematic diagram outlining the selection process for cell-surface and internalized proteins from the SWATH-based quantitative proteomics data of the control, cell surface, stripping, and internalization fractions. “Cell-surface proteins” were identified from the cell-surface fraction using predefined criteria. “Internalized proteins” were identified using a 2-step process: 1st, proteins were selected from the internalization fraction based on our defined criteria; these proteins were then verified for their presence in the “cell-surface proteins” list. (B, C) Identification of cell-surface and internalized proteins in HUVECs (B) and HDLECs (C). Venn diagrams illustrate the number of peptides and their corresponding proteins identified as cell-surface and internalized proteins in HUVECs (B) and HDLECs (C). (D, E) Rank-intensity plots of protein abundance, showing proteins identified as cell-surface proteins in the cell-surface fraction (blue, D) and as internalized proteins in the internalization fraction (red, E) for HUVECs. (F, G) Rank-intensity plots of protein abundance, displaying proteins identified as cell-surface proteins in the cell-surface fraction (blue, F) and as internalized proteins in the internalization fraction (red, G) for HDLECs.
To characterize HUVECs and HDLECs as VECs and LECs, respectively, we initially analyzed the expression of essential marker proteins in these cell populations. Established markers for VECs, including platelet endothelial cell adhesion molecule 1 (PECAM1/CD31), TEK receptor tyrosine kinase (TEK/TIE2), von Willebrand factor (vWF), and VE-cadherin (CDH5), were detected as cell-surface proteins in both HUVECs and HDLECs (Fig. 3A). In contrast, lymphatic-specific markers such as lymphatic vessel endothelial hyaluronan receptor-1 (LYVE-1), podoplanin (PDPN), and CD34, a marker of endothelial progenitor cells, were exclusive to HDLECs (Fig. 3A). Next, we examined the internalization of these proteins in HUVECs and HDLECs. PECAM1 and CDH5 were internalized in both cell types (Fig. 3B). In contrast, vWF was internalized only in HUVECs, whereas TEK, LYVE-1, and CD34 were internalized exclusively in HDLECs. Notably, PDPN remained at the cell surface in HDLECs (Fig. 3B). These findings highlight distinct cell-surface expression and internalization patterns of key vascular and lymphatic endothelial markers.
(A) Protein abundance in the cell-surface fractions of HUVECs and HDLECs. Data represent mean ± S.E.M. (n = 3). (B) Protein abundance in the internalized fractions of HUVECs and HDLECs. Data represent mean ± S.E.M. (n = 3). N.D.: not detected.
To investigate the similarities and differences in cell-surface and internalized proteins between VECs and LECs, we classified the cell-surface proteins quantified in HUVECs and HDLECs. We found 735 cell-surface proteins common to both cell types; however, 127 proteins were detected only in HUVECs, and 86 proteins only in HDLECs (Fig. 4A). Applying criteria of at least a 3-fold difference in abundance and a p-value below 0.05 among the shared proteins, we identified 29 proteins preferentially enriched in HUVECs and 61 in HDLECs (Fig. 4B). Overall, 156 and 148 cell-surface proteins were identified in HUVECs and HDLECs, respectively (Fig. 4C).
(A) Number of cell-surface proteins identified as common to or unique in HUVECs and HDLECs. (B) Volcano plot of cell-surface proteins identified as common to HUVECs and HDLECs. The x-axis represents the log2 fold change in protein abundance between HUVECs and HDLECs; the y-axis represents the –log10 (p-value, Welch’s t-test). Each data point represents a single protein; black data points indicate significantly enriched proteins (p < 0.05) and a fold-change of >3. (C) Total number of HUVEC- and HDLEC-selective cell-surface proteins. (D) Number of internalized proteins identified as common to or unique to HUVECs and HDLECs. (E) Volcano plot of internalized proteins identified as common to HUVECs and HDLECs. The x-axis represents the log2 fold change in protein abundance between HUVECs and HDLECs; the y-axis represents –log10 (p-value, Welch’s t-test). Each data point represents a single protein; black data points indicate significantly enriched proteins (p < 0.05) and a fold-change of >3. (F) Total number of HUVEC- and HDLEC-selective internalized proteins. Protein classification terms: unique protein: detected in only 1 cell type; common protein: detected in both cell types; preferential proteins: significantly enriched (p < 0.05 and a fold-change of >3) in 1 cell type; HDLEC-enriched proteins: sum of unique and preferential proteins in HDLECs; HUVEC-enriched proteins: sum of unique and preferential proteins in HUVECs.
Similarly, we classified the internalized proteins quantified in HUVECs and HDLECs (Fig. 4D). Consequently, 178 internalized proteins were commonly quantified in both cell types. In addition, 76 proteins were detected only in HUVECs, and 112 proteins were quantified only in HDLECs. Applying selection criteria of at least a 3-fold increase in protein abundance and a p-value below 0.05, we subsequently identified 11 proteins preferentially quantified in HUVECs and 14 proteins preferentially quantified in HDLECs among the 178 commonly quantified internalized proteins (Fig. 4E). In total, 87 and 126 internalized proteins were identified in HUVECs and HDLECs, respectively (Fig. 4F).
Biological Profiles of Endocytic Cell-Surface Proteins in HUVECs and HDLECsWe identified endocytic cell-surface proteins that are 1st expressed on the cell surface and subsequently internalized in a cell type–selective manner, reflecting their roles in the specialized functions of HUVECs and HDLECs. Specifically, 32 endocytic cell-surface proteins were identified in HUVECs by overlapping 156 enriched cell-surface proteins and 85 enriched internalized proteins (Fig. 5A and Supplementary Table 2). Similarly, 60 endocytic cell-surface proteins were identified in HDLECs by overlapping 145 enriched cell-surface proteins and 126 enriched internalized proteins (Fig. 5B and Supplementary Table 2). The top 10 endocytic cell-surface proteins enriched in HUVECs were transferrin receptor (TFRC), integrin subunit beta 6 (ITGB6), low-density lipoprotein receptor (LDLR), myosin heavy chain 9 (MYH9), C-type lectin domain family 14 member A (CLEC14A), LDL receptor-related protein 8 (LRP8), neuronal growth regulator 1 (NEGR1), neuronal cell adhesion molecule (NRCAM), major histocompatibility complex (MHC) class I polypeptide-related sequence B (MICB), and GLI pathogenesis-related 1 (GLIPR1) (Fig. 5C). The top 10 endocytic cell-surface proteins enriched in HDLECs were 5′-nucleotidase ecto (NT5E), CD99 molecule (CD99), beta-2-microglobulin (B2M), cadherin 13 (CDH13), jagged 1 (JAG1), prostaglandin F2 receptor negative regulator (PTGFRN), mannose receptor C-type 1 (MRC1), MHC class I, C (HLA-C), signal regulatory protein alfa (SIRPA), and fibronectin leucine-rich transmembrane protein 2 (FLRT2) (Fig. 5D).
(A, B) Endocytic cell-surface proteins in HUVECs (A) and HDLECs (B) based on the cell-selective cell-surface proteins and internalized proteins shown in Fig. 4. (C, D) The 10 most abundant endocytic cell-surface proteins in HUVECs (C) and HDLECs (D). (E, F) Functional enrichment analysis of endocytic cell-surface proteins in HUVECs (E) and HDLECs (F) using Metascape. The p-values are presented.
A functional enrichment analysis was performed to identify biological processes associated with endocytic cell-surface proteins in VECs and LECs (Figs. 5A and 5B). Specifically, this analysis was conducted using Metascape, a bioinformatics platform that integrates multiple databases to facilitate enrichment analysis, functional annotation, and visualization of biological pathways and networks.20) In HUVECs (32 proteins), the enriched terms included Cell–Cell Adhesion (GO: 0098609), Response to Wounding (GO: 0009611), ERBB2 Activates PTK6 Signaling (R-HSA-8847993), Cell Morphogenesis (GO: 0000902), Cell Activation (GO:0001775), and Blood Vessel Endothelial Cell Migration (GO: 0043534) (Fig. 5E). These terms are associated with processes such as tissue repair, angiogenesis, vascular remodeling, and responses to physical injury. In HDLECs (60 proteins), the enriched terms included Cell–Cell Adhesion (GO: 0098609), Extracellular Matrix Organization (R-HSA-1474244), Immunoregulatory Interactions Between a Lymphoid and a Non-Lymphoid Cell (R-HSA-198933), and Receptor-Mediated Endocytosis (GO: 0006898) (Fig. 5F). These terms are associated with processes, such as immune response and lymphatic system regulation.
This study presents a comparative proteomic profiling of endocytic cell-surface proteins in vascular and LEC, specifically aiming to elucidate mechanisms underlying their specialization. We found that over 85% of cell membrane proteins were expressed in both HUVECs and HDLECs, indicating a shared core essential for endothelial function; however, only 60–70% of these proteins were internalized in both cell types, revealing divergent endocytic pathways. Notably, vWF was internalized exclusively by HUVECs, whereas TEK and LYVE-1 were internalized only by HDLECs. Functional enrichment analyses revealed that HUVEC-associated endocytic proteins are linked to cell–cell adhesion, wound response, and endothelial migration, while HDLEC-associated proteins participate in extracellular matrix organization and immunoregulatory interactions. These findings highlight both the conserved endothelial core and the specialized endocytic mechanisms that distinguish VECs from LECs.
Significant differences were observed in the enriched biological processes and the proteins involved in each cell type. In HUVECs, we identified 32 endocytic cell-surface proteins (Fig. 5A). For example, TFRC and LDLR are critical for iron uptake and cholesterol metabolism, highlighting the increased metabolic demands of VECs exposed to circulating blood components.21,22) Their roles in these pathways underscore the robust mechanisms by which VECs meet energy demands and maintain homeostasis despite the stochasticity inherent in extrinsic and metabolic processes. Additionally, functional enrichment analysis revealed significant enrichment in biological processes related to cell–cell adhesion (GO: 0098609), enzyme-linked receptor protein signaling pathways (GO: 0007167), blood vessel development (GO: 0001568), and immune responses (Fig. 5E). Based on the profile of endocytic cell-surface proteins in HUVECs, VECs distinctly engage in nutrient uptake, lipid metabolism, cell adhesion, angiogenesis, and signal transduction compared to LECs.
In HDLECs, we identified 60 endocytic cell-surface proteins associated with lymphangiogenesis, immune regulation, and extracellular matrix interactions (Fig. 5B). For example, MRC1 and SIRPA are critical for immune regulation and antigen presentation, highlighting the specialized role of LECs in immune surveillance and modulation.23) These proteins underscore the robust mechanisms by which LECs transport immune cells and influence their activity, maintaining homeostasis in both innate and adaptive immunity despite extrinsic and intrinsic stochasticity.24–26) Additionally, functional enrichment analysis revealed significant enrichment in biological processes related to cell–cell adhesion (GO: 0098609), cell morphogenesis (GO: 0000902), receptor-mediated endocytosis (GO: 0006898), and basement membrane organization (GO: 0071711) (Fig. 5F). Thus, HDLEC-restricted endocytic cell-surface proteins indicate that LECs specialize in lymphangiogenesis, immune regulation, and extracellular matrix interactions, distinguishing them from VECs.
The cell-surface expression and internalization patterns of vital endothelial markers further elucidate the molecular distinctions between VECs and LECs (Fig. 3). CD31 (PECAM-1) is a crucial regulator of endothelial cell adhesion, immune cell migration, angiogenesis, and signal transduction, and is essential for maintaining vascular and immune system functions. CD31 is expressed and internalized in both cell types (Fig. 3B), underscoring its fundamental role in endothelial biology. In VECs, CD31 mediates cell–cell adhesion, maintains vascular integrity, and facilitates leukocyte transmigration while regulating vascular permeability.27–31) Although it exhibits lower expression levels in LECs, CD31 still contributes to lymphatic vessel formation and immune cell trafficking.32,33) Conversely, the lymphatic markers LYVE-1 and PDPN were identified only in HDLECs (Fig. 3). LYVE-1, primarily expressed in LECs, is a receptor for hyaluronan that regulates lymphatic fluid transport, immune cell trafficking, and lymphangiogenesis.34) It plays a significant role in maintaining tissue fluid balance and participates in immune response and tumor metastasis. Its internalization suggests its active involvement in hyaluronan uptake and turnover, which are essential for maintaining extracellular matrix interactions and facilitating immune cell trafficking within the lymphatic system. PDPN is a highly conserved mucin-type protein expressed on the cell membrane; however, it is not internalized and plays a critical role in the tumor microenvironment by regulating the immune system.35)
The identification of cell-enriched endocytic cell-surface proteins in HUVECs and HDLECs offers significant advances in the development of DDS targeting these distinct endothelial populations. For VECs, the crucial targets include TFRC, LDLR, and CLEC14A (Fig. 5B). TFRC is highly expressed in VECs and is critical for iron uptake, which is essential for cellular metabolism and angiogenesis.36) Its limited expression in non-endothelial tissues facilitates targeting by vascular-specific DDS, reducing unwanted interactions. LDLR, also enriched in VECs, participates in cholesterol metabolism and lipid homeostasis, which are essential for maintaining vascular integrity and function, making it a promising target for treating atherosclerosis and other lipid-related vascular diseases.37) CLEC14A, specifically expressed in angiogenic blood vessels, functions in endothelial cell adhesion and migration and represents a viable target for inhibiting pathological angiogenesis in cancer and diabetic retinopathy.38,39) Considering LECs, targets such as LYVE-1, NT5E, MRC1, and JAG1 are prominent (Fig. 5D). NT5E, predominantly expressed in LECs, is involved in adenosine production, immune response modulation, and lymphangiogenesis, which are beneficial for reducing tumor metastasis and treating chronic inflammatory conditions.40,41) MRC1, selectively expressed in HDLECs, is significant in antigen presentation and immune cell trafficking, making it suitable for DDS targeting autoimmune diseases and vaccine delivery.42) JAG1, which is enriched in LECs and involved in lymphatic vessel development, serves as a target to promote lymphangiogenesis in tissue repair or inhibit it in cancer metastasis. By focusing on these proteins, DDS can achieve high specificity and efficiency, improving therapeutic outcomes while ensuring safety through minimal off-target interactions.43,44) This strategic targeting exploits endothelial heterogeneity and facilitates precision medicine approaches for the treatment of vascular and lymphatic diseases.
We previously identified 34 proteins selectively internalized in hCMEC/D3 cells (i.e., a brain microvascular endothelial cell model) relative to HUVECs.16) In this study, we detected 60 proteins selectively internalized in HDLECs vs. HUVECs (Fig. 5B). Only 3 proteins (ITGA1, HLA-C, and FAS) were common to both groups (Supplementary Fig. 1), suggesting that approximately 90–95% of the proteins differ between hCMEC/D3 cells and HDLECs. Likewise, among the 34 proteins unique to HUVEC internalization, only these same 3 proteins (ITGA1, HLA-C, and FAS) overlapped with those in hCMEC/D3 cells (Supplementary Fig. 1), suggesting an 88–94% difference between hCMEC/D3 cells and HUVECs. This pronounced cell-type specificity likely reflects the physiological differences of each endothelial cell type. These findings underscore the need to accurately characterize the internalization profiles of target cell types, an approach crucial for the development of cell type-specific drug formulations and delivery strategies.
Our study has some limitations. Although HUVECs and HDLECs serve as representative models, they may not fully capture the heterogeneity of endothelial cells across different tissues, which exhibit organ-specific phenotypes influenced by their microenvironment. Additionally, our analyses were conducted under basal conditions; endothelial responses to physiological or pathological stimuli may alter protein expression and internalization profiles. While targeting cell-selective proteins enhances specificity, the expression of these proteins in other tissues may lead to unintended interactions, potentially affecting the safety and efficacy of therapeutics. Further in vivo validation and studies under various conditions are essential to confirm these findings and evaluate their clinical relevance.
In summary, we developed an integrated proteomic workflow that shifts endothelial profiling from static snapshots to dynamic analyses of endocytic behavior in VECs and LECs, thereby characterizing their physiological specialization. These cell type–specific endocytic signatures—undetectable by conventional expression analyses—establish a mechanistic framework for exploiting endothelial heterogeneity and lay the groundwork for precision-targeted therapeutics and diagnostics that harness the functional diversity of endothelial endocytosis.
This study received partial financial support in the form of a Grant-in-Aid for Scientific Research (B) (22H02786) from the Japan Society for the Promotion of Science, Japan, the Mochida Memorial Foundation for Medical and Pharmaceutical Research, and JST CREST, Japan (Grant Number JP171024167).
SI and MO contributed to the study design and conducted experiments. SI, MO, MY, TM, and SO performed the data analysis. SI and SO wrote the manuscript. All authors have provided final approval for the submitted manuscript.
The authors declare no conflict of interest.
The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.
This article contains supplementary materials.
