Mesenchymal stem cells promote lymphangiogenic properties of lymphatic endothelial cells

Abstract Lymphatic metastasis is one of the main prognostic factors concerning long‐term survival of cancer patients. In this regard, the molecular mechanisms of lymphangiogenesis are still rarely explored. Also, the interactions between stem cells and lymphatic endothelial cells (LEC) in humans have not been well examined. Therefore, the main objective of this study was to assess the interactions between mesenchymal stem cells (MSC) and LEC using in vitro angiogenesis assays. Juvenile LEC were stimulated with VEGF‐C, bFGF, MSC‐conditioned medium (MSC‐CM) or by co‐culture with MSC. LEC proliferation was assessed using a MTT assay. Migration of the cells was determined with a wound healing assay and a transmigration assay. To measure the formation of lymphatic sprouts, LEC spheroids were embedded in collagen or fibrin gels. The LEC's capacity to form capillary‐like structures was assessed by a tube formation assay on Matrigel®. The proliferation, migration and tube formation of LEC could be significantly enhanced by MSC‐CM and by co‐culture with MSC. The effect of stimulation with MSC‐CM was stronger compared to stimulation with the growth factors VEGF‐C and bFGF in proliferation and transmigration assays. Sprouting was stimulated by VEGF‐C, bFGF and by MSC‐CM. With this study, we demonstrate the potent stimulating effect of the MSC secretome on proliferation, migration and tube formation of LEC. This indicates an important role of MSC in lymphangiogenesis in pathological as well as physiological processes.

facilitate the spontaneous formation of tumours. Furthermore, they also exhibit immunosuppressive properties upon transplantation. 3 It has already been described that MSC positively influence angiogenesis of blood vessels and the revascularization of ischemic tissue through the secretion of blood endothelial cell (BEC) stimulating factors. 4 MSC also contribute to the formation of tumour blood vessels via integration as atypical vascular endothelial growth factor A (VEGF-A) secreting cells. 5 In contrast to the well-characterized relationship between BEC and MSC, however, little is known about lymphatic endothelial cell (LEC)-MSC interaction.
The survival rate associated with a certain type of cancer is mainly determined by the tumour cell's ability to form distant metastases. Cancer cells can disseminate from the primary tumour site via haematogenic and lymphatic routes. 6 Starting from the sentinel lymph node, they spread to other lymph nodes and distant organs. 7 Lymphatic vessels participate in tumour metastasis providing channels for tumour cells to leave lymph nodes 8 and play a complex role in metastatic tumour spread. 9 While the molecular mechanisms of lymphangiogenesis are still rarely explored, some of the involved growth factors and molecular signalling pathways have already been discovered. 10 One of the most studied group of pro-lymphangiogenic growth factors are VEGFs. 11 VEGFs are highly specific mitogens for vascular endothelial cells. They induce endothelial cell proliferation, promote cell migration and inhibit apoptosis. It is known so far that VEGF-C is the main lymphangiogenic growth factor in both physiological und pathological settings. 12 After processing, VEGF-C develops a higher affinity for VEGFR-3, which is exclusively expressed on LEC. 13 The expression of VEGF-C first occurs during embryogenesis, but remains high in adult lymph nodes. 14 The VEGF-C/VEGFR-3 signalling pathway is essential for tumour-associated lymphangiogenesis. 15 VEGFs not only influence lymphangiogenesis directly but also interact with other factors both directly and indirectly. One of these factors is the basic fibroblast growth factor (bFGF), which is also known as FGF2. Together with VEGF-C, it synergistically promotes lymphangiogenesis in the tumour microenvironment. 12 Furthermore, bFGF directly induces LEC proliferation and migration via activation of FGFR-1. Another important lymphangiogenic growth factor is hepatocyte growth factor (HGF), which also promotes proliferation, migration and tube formation of LEC via its receptor HGF-R. 16 HGF directly affects lymphangiogenesis and is not dependent on VEGFR3 activation. 17 In order to make further advances in the fields of tissue engineering and regenerative medicine as well as to address questions related to the lymphatic spread of tumour cells, a better understanding of the underlying mechanisms of lymphangiogenesis and the interactions between LEC, other cells, and in particular stem cells is needed. In contrast to the well-characterized interactions between MSC and BEC, to the best of our knowledge, the paracrine interactions between MSC and LEC have not been studied in detail using primary human cells until now. 18,19 Therefore, the aim of this study was to evaluate the in vitro interactions of LEC and MSC as a basis for further lymphangiogenesis and metastasis research.

| Cell culture
Human dermal LEC derived from juvenile foreskin (HDLEC) were purchased from PromoCell GmbH (Heidelberg, Germany) and cultured in endothelial cell growth medium MV (ECGM MV, PromoCell) with the corresponding supplement mix (see C-22020 PromoCell). LEC in passages 6 and 7 were used for all experiments.
Human MSC derived from bone marrow (hMSC-BM) were purchased from PromoCell and cultured in MSC growth medium (MSC medium, MSC-GM, PromoCell) with the corresponding supplement mix (see C-28010 PromoCell). MSC in passages 6 and 7 were used for all experiments.
Culture medium was changed 3 times a week, and the cells were passaged 1:3 after reaching a confluence of 80%. All cells were cultured at 37°C in an atmosphere of 5% CO 2 .

| MSC-conditioned medium
MSC were seeded in T75 flasks (Greiner Bio-One, Frickenhausen, Germany). After the MSC reached confluence, the medium was removed and the cells were washed once with PBS (Biochrom GmbH, Berlin, Germany). The MSC were incubated for 48 hours with 10 mL of endothelial cell basal medium (ECBM, PromoCell) containing 0.5% FBS (foetal bovine serum, FBS superior; Biochrom GmbH). After 48 hours, the culture medium was collected and used for experiments.

| Proliferation assay
LEC were seeded in 96-well plates at a density of 2.

| Migration assay
LEC migration was assessed with a scratch assay and a transmigration (Boyden chamber) assay. For the scratch assay, 4 x 10 5 cells/ well were seeded in a 6-well plate. As soon as the cells reached confluence, a lesion was generated in a standardized fashion using a 1000 lL pipette tip and the cells were cultivated with 2 mL ECBM containing 0.5% FBS as a negative control; ECBM containing 0.5 % ECBM supplemented with 0.5% FBS was used as negative control.
After incubation for 16 hours at 37°C in a 5% CO 2 atmosphere, the cells on the upper surface of the membrane were removed using a cotton swab and the cells on the lower surface of the filter were fixed with 100% ice-cold methanol and stained with DAPI (Roche Diagnostics GmbH). Images of the transmigrated LEC were captured at 10-fold magnification in 4 random fields, and the migrated cells were counted with the Olympus cellSens imaging software (version 1.12).

| Tube formation assay
The formation of three-dimensional capillary-like structures was examined by performing a Matrigel â -based tube formation assay.

| Statistical analysis
Data from all experiments are displayed as the mean of all independent experiments AE standard deviation (SD). The statistical analysis was performed with SPSS 21 by IBM. As negative control, ECBM with 0.5% FBS was used. Levene 0 s test was performed to test for homogeneity of variance. The test was always non-significant, except for the scratch, transmigration and sprouting assay. One-way ANOVA was performed for comparing multiple samples. Tukey 0 s test was conducted as a post hoc test. Student's t-test was performed for pairwise comparisons (ELISA measurements). Differences were considered statistically significant at P ≤ .05 and highly significant at P ≤ .01.

| Effect of different concentrations of MSC-CM on LEC
After 72 hours, stimulation with every dilution of MSC-CM resulted in a highly significant increase of LEC proliferation in comparison to the negative control ( Figure 1A). At 72 hours, cultivation with 30%, 50% and 70% MSC-CM did not show any significant differences between the concentrations. However, 10% MSC-CM stimulated cell proliferation significantly less compared to higher concentrations like 50% or 100%. Therefore, 100% MSC-CM was used in all following experiments.

| MSC-CM stimulated LEC migration to the same extent as VEGF-C and bFGF
In the scratch assay, LEC migration was significantly increased by VEGF-C + bFGF and MSC-CM compared to the negative control group after 12 hours (Figure 2A,B). MSC-CM induced LEC migration to the same extent as the combination of growth factors VEGF-C and bFGF and was significantly increased after 12 hours compared to the positive control PMA. After 24 hours, migration was significantly increased in all groups compared to negative control.

| Formation of capillary-like structures was stimulated by MSC-CM and co-cultivation with MSC
In comparison with the negative control, the formation of vessel-like structures could be enhanced by adding MSC-CM or through co-cultivation with MSC ( Figure 4A-C). In the co-culture group, capillarylike structures were formed both by LEC and MSC ( Figure 4A). In each group, longer tubes were measured compared to the negative control. Concerning the total tube length, there was no difference between MSC-CM, co-cultivation with MSC, the positive control or the combined growth factors VEGF-C and bFGF ( Figure 4B). The covered area was slightly increased in the positive control, the growth factor group and the MSC-CM group compared to the negative control. In the MSC group with 2,000 cells, a significantly larger ROBERING ET AL. | 3743 area was covered while there was no difference between the group with 1000 MSC and the negative control ( Figure 4C). In every group, except the MSC groups, more total branching points were measured compared to the negative control without differences between the groups ( Figure 4C). MSC alone are also able to form tubes but to a lesser extend than LEC + MSC together. In LEC-MSC co-cultures, both cell types contributed to tube formation demonstrated by podoplanin staining (Figure S1).

| Sprouting of LEC in fibrin gels was equally induced by stimulation with MSC-CM or the combination of VEGF-C and bFGF
Sprouting of LEC spheroids embedded in collagen and fibrin gels could be enhanced by stimulation with the positive control (Figure 5). In terms of VEGF-C + bFGF, LEC spheroid sprouting was enhanced in the fibrin gels compared to the collagen gels. In both gel types, MSC-CM stimulated less LEC sprouting than PMA. MSC-CM stimulated LEC sprouting in fibrin gels to a similar extent as a combination of VEGF-C and bFGF ( Figure 5C).  14,21,22 It is well-known that MSC contribute to the formation of new blood vessels. This effect is based on a combination of the direct and indirect influences of MSC on BEC. On the one hand, they secrete several factors directly implicated in angiogenesis such as VEGF-A, angiopoietin-1 and bFGF. 23,24 On the other hand, MSC secrete cytokines such as interleukin-6, which induce endothelin-1 production in cancer cells and thereby enhance endothelial cell recruitment and activation in an indirect manner. 25 Their contribution to lymphangiogenesis has not been investigated in detail yet, especially compared to blood vessel angiogenesis, but it can be assumed that cytokines secreted by MSC (e.g. VEGF, angiopoietin-2, bFGF and HGF) play a crucial role. 26 Moreover, MSC can contribute to lymphangiogenesis by  Because of their tissue-like mechanical properties and immunologic integrity, we used fibrin gels for the sprouting assay. 41 Fibrin matrices for implantations can be autologously harvested from the graft-recipient themselves and allow the adaptation of the polymerization and degradation rate by varying the concentration of aprotinin. 42,43 As second hydrogel, we chose collagen because of its abundancy in mammalian tissue 44 and its successful and common usage in other angiogenesis studies. 45,46 Compared to all other assays in the present study, the MSC-CM effect on LEC sprouting was not as high as expected.

| Concentration of VEGF-C, HGF and bFGF in the MSC-CM
Although it is believed that the process of lymphangiogenesis is composed of several single steps (invasion, capillary organization, tubular branching, network formation, maturation), the precise mechanisms are still not fully understood. In contrast to the blood vascular endothelium, which is in direct contact with the basement membrane components, lymphatic capillaries lack a basal lamina. 47 As a result, LEC have to penetrate an interstitial collagen barrier in the extracellular matrix (ECM), for example by matrix  The results from our in vitro experiments will provide the basis for the in vivo part to follow. To do this, the lymphangiogenic effect of MSC-secreted factors should be evaluated in the rat arteriovenous (AV) loop model. 49 This model will subsequently be used for lymphangiogenesis, anti-lymphangiogenesis and metastasis research in future in vivo studies. 50,51

| CONCLUSION
In the present study, the interaction between lymphatic endothelial cells (LEC) and mesenchymal stem cells (MSC) was evaluated using several in vitro angiogenesis assays. This study demonstrates the positive influence of a conditioned medium of primary human MSC on the lymphangiogenic response of primary human LEC. The lymphangiogenic growth factors secreted by the MSC enhanced proliferation and transmigration of LEC to a higher extent than the combination of VEGF-C and bFGF. In the scratch assay, the stimulative effect was similar to the combination of the growth factors VEGF-C and bFGF, but higher compared to the negative control.
Understanding the mechanisms of lymphangiogenesis and the role of the involved growth factors could help to gain deeper insights into the mechanisms of lymphangiogenesis in pathological processes as well as lymphatic metastasis. Furthermore, understanding the mechanism behind the MSC's stimulating effect on endothelial cells is a crucial requirement for the transition of novel MSC-based therapies from bench to bedside.