Ang-2 modulates VEGFR3-induced lymphangiogenesis in peritumoral adipose tissue of metastatic head and neck cancer

Abstract

Experimental studies have highlighted the connection between tumor-associated lymphatic vessel formation and adipose tissue. However, peritumoral lymphangiogenesis in metastatic head and neck cancer (HNC) remains insufficiently explored. This study contrasts lymphangiogenesis and regulatory factor expression in peritumoral adipose tissue (PTAT) surrounding metastatic head and neck tumors versus primary tumors. Histological, histochemical, and real-time PCR analyses were conducted on perioperatively collected adipose tissue samples from primary and metastatic tumors. Lymphatic Vessel Endothelial Hyaluronan Receptor-1 (LYVE-1) positive lymphatic vessel density significantly increased in peritumoral adipose tissue around metastatic tumors (M-PTAT). In M-PTAT, lymphatic vessels expressing LYVE-1 or VEGFD significantly increased, correlating with inflammatory cell count, elevated angiopoietin-2 (Ang-2) expression, and fibrosis factors. Expression levels of VEGF-D, VEGF receptor-3, and other lymphangiogenesis regulators in M-PTAT significantly surpassed those in peritumoral adipose tissue around primary tumors (P-PTAT). Notably, Ang-2 expression, confirmed histologically and at the mRNA level, was higher in M-PTAT than in P-PTAT. Ang-2 and VEGF-D stimulation promoted sprout formation in human lymphatic endothelial cells. This study underscores that M-PTAT, compared to primary tumors, exhibits more extensive lymphangiogenesis, inflammation, and fibrosis due to overexpression of Ang-2, LYVE-1, and VEGF-D.

Introduction

Head and Neck Cancer (HNC) remains a significant global health challenge due to its high prevalence, aggressive nature, and poor prognosis in advanced stages. A defining feature of HNC is its propensity for lymphatic metastasis, which serves as a major determinant of disease progression and patient outcomes. Previous clinical and experimental studies have provided evidence indicating a beneficial role of the adventitia in the process of lymphatic vessel wall remodeling, which encompasses neointima formation [1]. The pathophysiological correlation between the expansion of thin-walled, permeable lymphatic vessels, and the emergence of micro lymphatic vessels in the vicinity of tumors lies in the structural instability of lymphatic vessels, which results in an augmented infiltration of inflammatory cells [2]. Lymphatic vessels are commonly found in proximity to arteries and veins within most organs, although they are notably absent from the walls of these blood vessels themselves [3]. Despite our comprehensive understanding of the lymphatic system's extent for over a century, its significance in the mechanism of tumor lymph node metastasis has not garnered the attention it rightfully deserves [4, 5]. The primary function of lymphatic vessels is to regulate the efflux of interstitial fluid, a process essential for the maintenance of tissue homeostasis [6, 7]. Lymphangiogenesis plays a critical role in tumor progression by facilitating the spread of cancer cells to regional lymph nodes. This process is regulated by key molecular markers such as vascular endothelial growth factor-C (VEGF-C) and VEGF-D, which stimulate lymphatic endothelial cell proliferation and migration. Other markers, including Prospero homeobox-1 (Prox-1) and LYVE-1, are essential for lymphatic vessel maturation, remodeling, and function, further influencing tumor dissemination and metastatic capacity. Dysregulated expression of these markers has been closely linked to the structural instability of lymphatic vessels, increased vascular permeability, and heightened infiltration of inflammatory cells, all of which contribute to the aggressive behavior of HNC. Furthermore, lymphatic vessels play a pivotal role in establishing systemic immunity by guiding immune cells to draining lymph nodes and facilitating the removal of inflammatory mediators, thereby contributing to the reduction of inflammation [8, 9].

Studies conducted in animal models, focused on the genetic regulation of experimentally induced lymphatic vessel formation, have unequivocally demonstrated the pivotal role of the lymphatic system in the intricate process of tumor metastasis [10, 11]. Subsequent experimental findings in mice have further substantiated the causal relationship between lymphatic dysfunction and the mechanism of tumor lymph node metastasis. Specifically, the excessive accumulation of cholesterol has been associated with alterations in the structure and function of lymphatic vessels. Notably, this dysfunction manifests prior to the onset of tumor lymph node metastasis [12, 13]. The ambiguity lies in whether these experimental findings accurately depict the correlation between lymphatic dysfunction or the lymphatic network surrounding the tumor and the occurrence of local lymph node metastasis. Furthermore, it remains uncertain if there are systematic distinctions between individuals at risk and those in control groups [14].

Considering the anatomical proximity of adipose tissue surrounding tumors to lymphatic vessels and the well-established paracrine functions of the underlying adipose tissue, we postulated the existence of potential distinctions in the extent of the lymphatic system and the expression of lymphopoietic growth factors between the adipose tissue around metastatic tumors and that around primary tumors.

Methods and materials

Clinical cases

This research included 28 patients (54% male; mean age 57.2 ± 8.3 years) with head and neck squamous cell carcinoma (SCC) diagnosed and resected in head and neck surgery in two hospitals. In our study, we analyzed a total of 28 primary tumor samples and metastatic tissue samples. All patients were diagnosed with head and neck squamous cell carcinoma by pathological section staining. All patients had signed the informed consent form before being included in this research. Due to the limited amount of sample taken from patients during the operation, not all samples could be analyzed completely, so the samples had to be divided into multiple groups. Patients with carcinoma in situ, distant metastasis, and a history of malignancy other than head and neck cancer at the time of diagnosis, as well as patients who had received neoadjuvant chemotherapy or radiotherapy before surgery, were excluded.

Tissue sample collection

Peritumoral adipose tissue (PTAT) around the tumor and PTAT around metastatic HNC were collected during the perioperative period and transferred to cold sterile saline solution immediately. Sterile saline was used to remove residual blood and loosely attached connective tissues. PTAT samples were then processed under sterile conditions and either cultured in conditioned medium or stored in liquid nitrogen for protein separation. The PTAT samples for immunohistochemistry experiments were fixed with formalin and cut into 5 μM thick paraffin sections.

Preparation of conditioned medium for adipose tissue around tumor

M-PTAT and P-PTAT samples were fully dissociated under aseptic conditions and then transferred to Dulbecco modified Eagle medium containing 1% streptomycin and fetal bovine serum (0.5%). To elucidate, it represents the adipose tissue surrounding metastatic lesions rather than the metastases themselves. Then, the tissue samples are cultured in the medium (37 ℃, 5% CO₂) for 24 h. The culture medium alone, without any treatment, was used as the parallel control. After removing the PTAT fragments, the conditioned medium was centrifuged at 10,000 g for 15 min, and the supernatant was stored in liquid nitrogen for analysis.

Protein membrane array analysis

The secreted proteins of M-PTAT and P-PTAT from 10 HNC patients were examined with customized antibody arrays for 9 target proteins according to the test kit instructions. A ChemiDoc imaging system was used for chemiluminescence detection. Gilles Carpentier's ImageJ macro protein array analyzer was utilized to conduct semi-quantitative analysis by densitometry. Quantitative data were presented as fold changes relative to P-PTAT, calculated by normalizing the signal intensity of each target protein to its corresponding value in the P-PTAT group. For the growth factor analysis, a total of 21 samples with sufficient volume and integrity were included. For the protein array, a subset of 10 samples with high protein concentration and no evidence of degradation or hemolysis were selected to ensure assay accuracy and sensitivity. Sample selection was based on predefined quality control parameters, including minimum protein yield and optical density.

RT-qPCR

Total RNA was extracted from the experimental samples using a commercially available RNA extraction kit following the manufacturer's instructions. The quality and quantity of the extracted RNA were assessed using spectrophotometric analysis or gel electrophoresis. The synthesized cDNA was amplified using real-time PCR technology. GAPDH served as the internal control. All reactions were run in technical triplicates. The primer sequences are listed in Table 1. Statistical analysis was performed to assess the significance of any observed differences.

Table 1.

Primer sequences used for quantitative real-time PCR (qRT-PCR)

Gene	Forward primer (5′−3′)	Reverse primer (5′−3′)
VEGF-D	GCTCCTGGCCAGAAGGTG	TGTGTTCAGGCATCTCCT
VEGF-C	AGCTGGGACTCCCTGAGA	CAGTGGCTTCTCCACGTT
VEGF-B	TGGCACATGGAGAGAGCTTT	GGAGTTGACAGGGCTTGG
VEGF-A	CTTGCCTTGCTGCTCTACC	TGTGCTGGCTTTGGTGAG
VEGFR-3	GGACAGTCCGTGGACACAT	GCTGCTGGAGGTTGAGTTGT
VEGFR-2	GCTGTTGAAGCCAGGAACTC	TGCTGTCCAGCAGAGTGTTC
VEGFR-1	GGCAGGCTGCTGTTTCAT	GAGCTGGCAGTGGATCTGTT
PROX-1	CTCTGGCGATGTCAGTGAG	CACCGCTTCCGTTGTTCT

The table lists the forward and reverse primer sequences designed for the amplification of target genes, including VEGF family members (VEGF-A, VEGF-B, VEGF-C, VEGF-D), VEGF receptors (VEGFR-1, VEGFR-2, VEGFR-3), and the lymphatic endothelial marker PROX-1.

Histology and immunohistochemistry

Histological examination of paraffin-embedded thick tissue sections was conducted, as well as double-color staining for immunohistochemistry to examine specimens. This method involved deparaffinization and rehydration of tissue sections followed by staining with specific dyes to highlight cellular and tissue structures. The use of standardized staining protocols ensured reproducibility and consistency across experiments. In this study, double-color staining (LYVE-1 and DAPI) was employed to simultaneously detect the expression of two different proteins within the same tissue section. One of the antibodies used in the immunohistochemistry experiments targeted LYVE-1, a marker for lymphatic endothelial cells. LYVE-1 staining allowed for the selective visualization of lymphatic vessels within the tissue samples. Sections were then incubated with a primary antibody against CD31 at 4 °C overnight. After washing, sections were incubated with a horseradish peroxidase (HRP)-conjugated secondary antibody. By specifically labeling lymphatic endothelial cells, the distribution and morphology of lymphatic vessels were studied in the context of the research.

Xenograft tumor models (in vivo)

Approval for all animal experiments was duly obtained from the Institutional Animal Care and Use Committee of the Second Military Medical University, reflecting a steadfast commitment to ethical research practices. For the xenotransplantation experiments, 8 mice were included in each group. Housing arrangements encompassed four animals per cage, maintaining a consistent 12:12 light: dark cycle and ensuring ad libitum access to nourishment and hydration, thereby upholding the well-being of the experimental subjects. Adherence to approved protocols remained paramount throughout the experimental procedures. In line with these rigorous standards, female athymic nude mice (nu/nu), aged between 3 and 4 weeks, were procured from the Shanghai Laboratory Animal Co., Ltd. These mice were meticulously maintained and housed under specific pathogen-free conditions at the Animal Experimental Center of Second Military Medical University, fostering an environment conducive to reliable experimental outcomes. The xenograft model was established through the subcutaneous inoculation of FaDu cells (hypopharyngeal squamous cell carcinoma), totaling 5 × 10⁶ cells, into the right rear flank of the nude mice. Subsequent monitoring of tumor progression entailed regular assessments every 2 days, employing a vernier caliper to measure tumor volume. The calculation formula utilized was V = L × W²/2, where V denotes volume, L signifies length, and W represents width. This meticulous approach to data acquisition underpins the robustness and accuracy of the experimental findings.

Western blotting

Total protein was extracted from the experimental samples using a protein extraction buffer or lysis buffer containing protease inhibitors to prevent protein degradation. The protein samples were mixed with a loading buffer containing reducing agents and heated to denature the proteins. Antibody Dilution Buffer: PBS (Phosphate-Buffered Saline): pH 7.4, Bovine Serum Albumin (BSA): 1%, Tween 20: 0.05%. Transfer Buffer (for Western Blotting): Tris base: 25 mM, Glycine: 192 mM, Methanol: 20%. The samples were then loaded onto the SDS-PAGE gel along with protein size markers. Electrophoresis was carried out under constant voltage, allowing the proteins to migrate through the gel matrix based on their size. GAPDH was used as the housekeeping protein (loading control) in the western blotting experiments. The intensity of the protein bands was quantified using image analysis software. Statistical analysis was performed to assess the significance of any observed differences in protein expression levels.

Spheroid lymphangiogenesis assay

Human dermal lymphatic endothelial cells were cultured in Endothelial Cell Growth Medium MV2 up to passage 5 for cellular integrity. Spheroids of HDLECs were prepared on day one using a standardized protocol. Cultured HDLECs were detached using 0.05% trypsin–EDTA and resuspended at 400 cells per 100 µL in Endothelial Cell Growth Medium MV2 with 20% methylcellulose in Medium 199. The cell suspension was transferred to a round-bottom 96-well suspension culture plate and incubated for 24 h at 37 °C and 5% CO₂ for spheroid formation. On day two, spheroids were collected, centrifuged at 390 g for three minutes, and resuspended in a 1:1 mixture of 50% rat tail collagen I, 0.05% acetic acid (in Medium 199), and 5% fetal bovine serum in methylcellulose/Medium 199 solution. The suspension was transferred to a 24-well plate and incubated for 30 min before adding Endothelial Cell Growth Medium MV2. Spheroids were stimulated with recombinant human angiopoietin-2 (ANGPT2; 100 ng/mL) and/or recombinant human vascular endothelial growth factor C (VEGF-C; 100 ng/mL) for 24 h to induce lymphangiogenesis. Data in the control group were normalized to a baseline value of 1.0 to facilitate comparison across groups.

Statistical analysis

If the data is not normal distribution, it is presented with the median (25% and 75% interquartile interval); if it is normal distribution, it is presented as the mean ± standard deviation. Pearson Comprehensive Normal Test for normal distribution data. Student’s paired t-test for comparison and analysis of samples with normal distribution data. If it was not normal distribution, Wilcoxon paired pair signed rank test was used. Multiple comparisons using Turkey’s multiple comparison tests or one-way ANOVA. All statistical analysis was completed in GraphPad PRISM 8.0 software.

Results

Primary p-PTAT contained fewer lymphatic vessels than metastatic m-PTAT

The research findings indicated that M-PTAT in HNC patients exhibited a higher degree of lymphangiogenesis and a greater number of lymphatics compared to P-PTAT (Fig. 1A–F). This observation was further confirmed by immunohistochemical analysis of CD31 in paired PTAT tissue samples from 21 HNC patients. The density of CD31-immunoreactive lymphatic vessels in M-PTAT was significantly higher than that in P-PTAT (P < 0.0001; Fig. 1A, B). To distinguish the lymphatic network, specific markers of lymphatic endothelium, namely D2-40 and LYVE-1, were employed in the experiment. D2-40 was expressed in lymphatic collectors, while LYVE-1 was predominantly expressed in lymphatic capillaries. Experimental studies revealed a significant increase in the number of LYVE-1 (P < 0.0001; Fig. 1C, D) and D2-40 (P < 0.0001; Fig. 1E, F) immunopositive lymphatic vessels in M-PTAT compared to P-PTAT. These findings highlighted the enhanced lymphatic vessel formation and lymphangiogenesis in M-PTAT, indicating a potential role in tumor metastasis and progression.

Fig. 1.

Lymphangiogenesis in peritumoral adipose tissue (PTAT). Immunohistochemical levels of CD31 (A, B), anti-LYVE-1 (lymphatic endothelial hyaluronic acid receptor 1) (C, D) and anti D2-40 (E, F) were detected in metastatic HNC M-PTAT and primary HNC P-PTAT samples. Detect the count of CD31, LYVE1 and D2-40 immune positive lymphatic vessel per square millimeter in 21 patients with head and neck cancer (HNC). The paired M-PTAT and P-PTAT values of a single patient are connected by a single line. Dimension bar represents 100 μm. Statistical analysis using Student’s paired t-test. The difference is considered statistically significant if P < 0.05

M-PTAT exhibits increased levels of lymphangiogenic growth factors and their receptors relative to P-PTAT

To determine whether the lymphoid network in M-PTAT was synced with the expression of growth factors that induce lymphangiogenesis, the expression levels of VEGF-D, VEGF-C, VEGF-B, and VEGF-A in M-PTAT and P-PTAT from 21 patients were analyzed by qPCR. The analyses showed that compared with P-PTAT, the expression of VEGF-D (P = 0.001; Fig. 2A) and VEGF-C (P = 0.0154; Fig. 2B) in M-PTAT was dramatically enhanced. However, compared with P-PTAT, the mRNA level of VEGF-B in M-PTAT was obviously decreased (P < 0.0001; Fig. 2C), while the mRNA level of VEGF-A was not significantly different (P = 0.2598; Fig. 2D). The mRNA expression of VEGFR-3, VEGFR-2, and VEGFR-1 was then detected. Compared with P-PTAT, the mRNA level of VEGFR3 in M-PTAT was obviously enhanced (P = 0.0142; Fig. 2E), while there was no difference in mRNA level of VEGFR2 (P = 0.7382; Fig. 2F) and VEGFR1 (P = 0.6143; Fig. 2G) between the two PTAT repositories. These results indicated that VEGF-D/VEGF-C mediated VEGFR-3 signal transduction was in a high state, which might be the reason for the observed difference in lymphoendothelial cell density between M-PTAT and P-PTAT. The experimental analysis of the project on the transcription factor PROX-1 (prospero-related homeobox-1), the main regulatory factors of embryonic lymphoid development, showed that the levels of PROX-1 mRNA in M-PTAT and P-PTAT were similar (P = 0.1491; Fig. 2H). This result showed that the difference in lymphatic structure was not the result of the difference in the lymphopoietic potential defined by the embryo in the two adipose tissue banks.

Fig. 2.

Analysis of growth factor in lymphangiogenesis. Quantitative level of VEGF-D (A), VEGF-C (B), VEGF-B (C), VEGF-A (D), VEGFR-3 (E), VEGFR-2 (F), VEGFR-1 (G) and PROX-1 mRNA in M-PTAT and P-PTAT of 21 HNC patients. Statistical analysis using Wilcoxon paired sign rank test. If P < 0.05, the difference is considered statistically significant

Protein array analysis showed that Ang-2 and VEGF-D levels in M-PTAT were significantly higher than those in P-PTAT

Previous studies had confirmed that PTAT could secrete and express a variety of cytokines, which might locally act on the adjacent lymphatic vessel wall in a paracrine manner. In view of our findings, our next goal was to study a wider range of PTAT secreting factors, with particular attention to potential candidate factors involved in lymphangiogenesis, fibrosis, and inflammation. Semi-quantitative analysis of 9 candidate factors in M-PTAT and P-PTAT samples from 10 HNC patients using membrane-based antibody arrays (Fig. 3A) was conducted. Among them, interleukin-2, interleukin-8 (inflammatory factor), matrix metalloproteinase (MMP1), and its inhibitors TIMP1 and growth factors. Ang-2 were detected in two adipose tissue samples (heat map in Fig. 3A). Protein array analysis showed that Ang-2 and VEGF-D levels in M-PTAT were significantly higher than those in P-PTAT. Quantitative analysis indicated that there was a statistically significant difference in the expression of Ang-2 and VEGF-D (Fig. 3B).

Fig. 3.

Protein array analysis of factors related to peritumoral lymphangiogenesis. A Using membrane-based protein array analysis, the expression of 9 proteins in M-PTAT and P-PTAT samples from 10 HNC patients was quantitatively heat map. B Quantitative analysis indicated that there was a statistically significant in the expression of Ang-2 and VEGF-D. The bar graph represents the fold change of average protein expression compared with P-PTAT. The error bar represents the standard deviation

Exogenous Ang-2 induced neovascularization of human lymphoendothelial cells

To further study the importance of the above results on the lymphangiogenesis of adipose tissue around the tumor, a three-dimensional spherical lymphangiogenesis experiment was used in HDLECs to detect the role of exogenous Ang-2. Compared with the VEGF-D treated spheroids, stimulation of HDLEC spheroids with Ang-2 (50 ng/mL) for 12 h did not change the cumulative length of sprouts or the number of sprouts (Fig. 4A–C). HDLEC spheroids were stimulated with Ang-2 and VEGF-D (80 ng/mL), both of which were overexpressed in M-PTAT (Ang-2 at the level of mRNA and protein, VEGF-D at the level of mRNA), significantly increasing the cumulative sprout length (P = 0.032) and the number of sprouts (P = 0.014). Given the hypoxic character of the solid tumor microenvironment, the possible role of perivascular hypoxia in the observed overexpression of Ang-2 and VEGF-D was examined. Induction of chemical hypoxia by stimulating cultured HDLEC with 1 mM CoCl₂ for 5 h was conducted. Compared with control groups, the expression of Ang-2 mRNA in HDLEC treated with CoCl₂ was significantly increased (4.8-fold increase compared with the control group; P < 0.001; Fig. 4D). VEGF-D also increased significantly at the mRNA level (8.6-fold increase compared with control groups; P < 0.001; Fig. 4E).

Fig. 4.

The Effect of VEGF-D and Ang-2 on Human Lymphatic Vessel sprouting. Representative image of HDLEC (human lymphatic endothelial cell) globules (A) and the number of sprouts (B) and cumulative sprouts length (C) stimulated by recombinant Ang-2 (μm) and VEGF-D (90 ng/mL) or control. Dimension bar represents 200 μm. D, E In 10 independent experiments, Ang-2 (D) and VEGF-D (E) mRNA expression in HDLEC was quantified after 4 h of chemical hypoxia induced by 1 mM CoCl₂. The bar graphs illustrate the multiple changes of average mRNA level compared to control group. Statistical analysis using Turkey’s multiple comparison test and one-way ANOVA (B, C) and student’s paired t-test (D, E). The values for the control group were normalized to 1.0, and all experimental group data were expressed relative to the control baseline

Dynamics of lymphangiogenesis in the early stage after hypoxia

Previous studies had shown that chemical hypoxia could induce endogenous lymphangiogenesis. However, in the early stage after hypoxia, the reaction kinetics of lymphangiogenesis had not been clarified. Therefore, this issue was discussed here. In this experiment, the HDLEC spheroid was subjected to chemical hypoxia and was observed for 4 h. This study evaluated the protein expression of VEGFR3 and VEGF-D in HDLEC during different hypoxia periods (Figs. 5A, B). One hour after hypoxia, the level of VEGF-D was dramatically enhanced. The level of VEGFR3 increased after 1 to 4 h of hypoxia (Fig. 5C, D).

Fig. 5.

Dynamics of lymphatic angiogenesis in early stage after hypoxia. The expression levels of VEGF-D (A) and VEGF-R3 (B) proteins in HDLEC undergoing different hypoxia periods were analyzed by representative immunoblotting (C, D). Values are means ± SEM. H indicates hours. *P < 0.05 and **P < 0.01 vs Control

Evaluation of P-PTAT lymphatic vessels by focusing on lymphatic density

The experiment showed that the number of LYVE1 positive cells increased obviously from the 8th day of xenotransplantation and continued to the 14th day of xenotransplantation (Fig. 6A, B). To determine whether the increase of LYVE1 positive cells indicates lymphocyte proliferation, treatment of cancer subgroups with 5-bromo-2'-deoxyuridine after hypoxia began. Experimental studies showed that from the 8th day of xenotransplantation, the number of LYVE1⁺ cells stained with 5-bromo-2'-deoxyuridine increased significantly (Fig. 6C, D). Finally, the size of the lymphatic lumen in M-PTAT was measured to assess lymphatic drainage capacity. The analysis of the research results showed that the lumen area increased significantly from the 8th day of xenotransplantation to the 14th day after xenotransplantation (Fig. 6E).

Fig. 6.

Evaluate P-PTAT lymphatic vessels by focusing on lymphatic density. A Image of M-PTAT stained with LYVE-1 antibody shows LYVE-1 or BrdU positive cell and lymphatic lumen area (B). C LYVE-1 positive cell count. D Summary of 5-bromo-20-deoxyuridine (BrdU) labeled LYVE1 positive cells. E Lymphatic lumen area. Sample was collected from M-PTAT in different hypoxic areas. Bar: 100 μm (A); Values are means ± SEM; *P < 0.05 and **P < 0.01 vs Control

Schematic drawing depicting the main findings of this study

A schematic drawing was included to illustrate the key findings of this study. This diagram served as a visual representation of the main results and provided a concise summary of the key concepts and relationships explored in our research. Compared with primary tumors, peritumoral adipose tissue around metastatic head and neck cancer showed more extensive lymphangiogenesis, inflammation, and fibrosis, which may be due to the overexpression of Ang-2, LYVE-1, and VEGF-D (Fig. 7).

Fig. 7.

Peritumoral adipose tissue around metastatic head and neck cancer. Compared with primary tumors, peritumoral adipose tissue around metastatic head and neck cancer shows more extensive lymphangiogenesis, inflammation and fibrosis, which may be due to the overexpression of Ang-2, LYVE-1 and VEGF-D

Discussion

The principal findings of this study encompass the following: 1, In comparison to peritumoral adipose tissue (PTAT) surrounding primary tumors, PTAT around metastatic tumors exhibits a notably more abundant lymphatic network. 2, mRNA and protein expression levels of lymphopoietic growth factors (VEGFD and VEGFC) and their corresponding receptors (VEGF-R3) are significantly elevated in PTAT around metastatic tumors when compared to PTAT around primary tumors. 3, This elevation in lymphangiogenesis is concurrent with heightened levels of Ang-2 within PTAT around metastatic tumors. 4, Furthermore, the heightened levels of Ang-2 in PTAT around primary tumors may represent a potential contributing factor to the observed increased lymphangiogenesis around the tumor.

Furthermore, the marked heterogeneity observed in gene expression between peritumoral adipose tissue (PTAT) and primary tumors in the vicinity of head and neck metastatic tumors underscores the significance of variations in peripheral fat composition as a determinant of lymphatic neogenesis. The potential involvement of adipose tissue surrounding tumors in the pathophysiology of lymphangiogenesis has garnered increased attention in recent research [15–17]. Prior investigations have ascertained that factors released within the adipose tissue surrounding the tumor constitute pivotal local mediators in the intricate process of lymphangiogenesis [18–20]. In contrast to the perivascular network formed by vascular endothelial cells, research on lymphangiogenesis within the adipose tissue in the context of tumor lymph node metastasis remains relatively limited [6, 21]. During the progression of chronic diseases, lymphatic vessels in the vicinity of the tumor undergo a reactive phase aimed at enhancing lymphatic drainage. However, in the later stages, they tend to become dysfunctional and exhausted due to sustained overactivation [22, 23]. While this study does not definitively resolve whether the observed lymphangiogenesis around metastatic tumors is a consequence or a contributing factor to the initiation of peritumoral adipose tissue, the augmented expression of Angiopoietin-2 (Ang-2), macrophage infiltration, and fibrosis observed within the peritumoral adipose tissue (PTAT) of metastatic tumors appears to align with a reactive response. Specifically, it suggests that adventitial lymphangiogenesis may be induced as a reaction to inflammation or fibrosis. In a broader context, both this study and previous research underscore the intricate and multifaceted relationship between inflammation, fibrosis, and lymphangiogenesis around tumors. Protein array analysis revealed relatively higher levels of inflammatory cytokines, such as IL-2 and IL-8, in M-PTAT compared to P-PTAT, suggesting a more pro-inflammatory profile. This finding supports the interpretation of enhanced inflammatory activity in M-PTAT. It is essential to acknowledge that the observed pathophysiological pathways may not follow a unidirectional course [24, 25]. The available data suggest the existence of a circulatory mechanism that encompasses multiple pathways across various biological processes [26, 27]. Among these pathways, it is noteworthy that inflammation and fibrosis in the later stages of tumor development promote the generation of peripheral lymphatic vessels [23, 28].

Regarding the signaling pathways involved in lymphangiogenesis around the tumor, it is observed that increased lymphangiogenesis in peritumoral adipose tissue (PTAT) of metastatic tumors is concomitant with elevated mRNA expression of vascular endothelial growth factor D (VEGFD), vascular endothelial growth factor C (VEGFC), and vascular endothelial growth factor receptor 3 (VEGFR3) when compared to PTAT in primary tumors [22, 29]. The pivotal role of VEGF-D in the regulation of lymphatic vessel growth has been comprehensively demonstrated [30]. Other investigations have ascertained that the overexpression of VEGF-D selectively stimulates lymphangiogenesis while leaving accompanying angiogenesis unaffected. Ang-2 is associated with lymphangiogenesis, which may indirectly impact the peritumoral vasculature within adipose tissue [31]. Vascular endothelial growth factor receptor-3 (VEGFR3) and VEGF-D are involved in lymphangiogenesis and may influence the lymphatic drainage of adipose tissue surrounding tumors. LYVE-1 is a lymphatic endothelial marker, potentially indicating a role in lymphatic vessel changes in peritumoral adipose tissue. Given the complex interplay between tumor cells, the microenvironment, and these genes' roles in lymphangiogenesis, further research is needed to elucidate the precise relationship between these genes and peritumoral adipose tissue in metastatic head and neck cancer. Such investigations could provide valuable insights into tumor progression and potential therapeutic targets [32]. The influence of VEGF-D on lymphangiogenesis may be contingent on the expression levels of VEGFR3 and VEGFR2 receptors. This inference is supported by findings in mice expressing specific VEGFR3 mutants, which demonstrated altered lymphangiogenic responses [33]. Nevertheless, our study did not reveal a significant disparity in the levels of VEGFA between peritumoral adipose tissue around primary tumors (p-PTAT) and peritumoral adipose tissue around metastatic tumors (m-PTAT). However, notably, VEGFB exhibited a significant reduction in our investigation.

This study is subject to certain limitations. Due to the limited availability of tissue samples for each patient, our research had to be subdivided into various test groups, potentially impacting the precision of our observations. Moreover, it is important to acknowledge that differences in the local microenvironment between the two peritumoral adipose tissue (PTAT) samples cannot be entirely ruled out. Given the experimental setting utilizing murine models, the researchers were mindful of the plausible occurrence of LYVE1 expression within a specific subset of murine macrophages, known for their propensity to infiltrate transplanted tissue or stroma. In summary, our findings in patients with metastatic head and neck tumors reveal an increased lymphangiogenic activity within PTAT surrounding the tumor when compared to PTAT around the primary tumor. This heightened activity is attributed to the increased expression of Ang-2, activation of the VEGFR3 signaling pathway mediated by VEGFD and VEGFC. These factors, in conjunction with the local hypoxic environment and proinflammatory conditions, likely underpin the observed pathophysiological changes. In our study, we utilized human dermal lymphatic endothelial cells (HDLECs) as a widely available and well-characterized model to investigate lymphangiogenesis. While HDLECs provide valuable insights, we acknowledge that they do not fully recapitulate the specific microenvironment or developmental context of lymphatic endothelial cells in the head and neck region.

Author contributions

A.H. and L.-Y.Y. conceived and designed the project and prepared the manuscript. J.-W.Z. performed a part of the experiments in slices, analyzed results, prepared some of the figures. A.H. edited and commented on the manuscript. All authors read and approved the manuscript.

Funding

This study was funded by grants from the Shanghai Healthy Youth Project (Grant No.20234Y0055).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Ethics approval and consent to participate Experiments were performed in strict accordance with the ethics approval and were approved by the Ethical Committee at Second Military Medical University.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.