Keywords

11.1 Introduction

In vertebrates, two complementary vessel networks form the vascular system and achieve organ homeostasis: the blood and the lymphatic vasculature. The blood vasculature is a circulatory system that transports oxygenated blood from the heart to peripheral tissues via arteries, while veins return oxygen-poor blood to the heart. In a hierarchical tree-like system, blood passes from large arteries to smaller arterioles and infiltrates tissues and organs through an extensive network of capillaries. The exchange of oxygen, delivery and recycling of nutrients and waste between blood and the tissue are achieved at the capillary level, which subsequently drains into venules and large veins. For re-oxygenation, blood is transported via the pulmonary artery to the lungs. In contrast, the lymphatic vasculature is a blind-ended, unidirectional vessel network that developed evolutionary to transport dietary fats from the intestine to the liver (Kampmeier 1969). Later, the lymphatic system acquired additional functions to drain excessive interstitial fluid and to transport immune cells to lymph nodes for immune surveillance (Stritt et al. 2021).

Endothelial cells (ECs) are essential vessel components and line the inner vessel wall of both vessel networks to form a dynamic barrier between the circulating blood or lymph on their luminal side and the surrounding tissue on their abluminal side. Through their specialized, button-like cell–cell contacts (button-like junctions) (Baluk et al. 2007), lymphatic capillaries (also referred to as initial lymphatics) collect fluid, macromolecules and cells from the interstitial tissues. The lymph is then transported via pre-collecting lymphatic vessels to the larger collecting lymphatic vessels and returns to the bloodstream via lymphovenous valves at the intersection of the jugular and subclavian veins (Jeltsch et al. 2003; Tammela and Alitalo 2010; Geng et al. 2016). Lymphatic collecting vessels are equipped with luminal lymphatic valves (LVs), which support pumping and ensure unidirectional lymphatic transport without reflux (Potente and Mäkinen 2017). In contrast to the lymphatic capillaries, the lymphatic collecting vessels form continuous zipper-like cell–cell contacts and are therefore characterized by a reduced absorption of fluid, macromolecules, and cells from their interstitial environment (Baluk et al. 2007).

Due to their unique functions, each vessel subtype is exposed to a variety of mechanical forces. These vessel subtypes are therefore equipped of specialized EC subtypes with unique properties and genetic profiles that enable them to fulfill their specific function (Potente and Mäkinen 2017). Not only does each vessel subtype exhibit a unique endothelial genetic profile, but endothelial properties also differ in a tissue-specific manner. As an example, specialized lymphatic ECs (LECs) are found in the Schlemm’s canal vessels of the eyes (Petrova and Koh 2020), in the ascending vasa recta of the kidney (Kenig-Kozlovsky et al. 2018), and in meningeal lymphatic vessels in the brain (Aspelund et al. 2015; Louveau et al. 2015).

In addition, blood and lymphatic vessels are surrounded by unique tissue microenvironments with different mechanical and structural properties. ECs can be supported by various extracellular matrix (ECM) components, as well as by several mural cell types, like pericytes or smooth muscle cells (SMCs) (Gordon et al. 2020). The ECM is an essential part of the tissue environment and forms a complex three-dimensional scaffold consisting of the basement membrane (BM), which is mainly formed by the ECs themselves, and the interstitial matrix (IM), which fills the interstitial space between surrounding cells.

ECs recognize and respond to mechanical impacts, such as shear, stretch, and ECM stiffness, through their cell–cell contacts and cell–ECM adhesions, and translate physical stimuli into biological responses, in a process referred to as mechanotransduction. Shear mode and amplitude of fluid flow, as well as composition and mechanical properties of the ECM, differ across the vascular tree, in development and disease.

While the impact of changes in fluid shear stress (FSS) is well studied in blood ECs (BECs) and LECs (reviewed in Baeyens et al. 2016; Bálint and Jakus 2021; Campinho et al. 2020), in this chapter we will focus on changes of the lymphatic endothelial tissue microenvironment. We will briefly compare what is known about LEC and BEC tissue microenvironments and further discuss how tissue microenvironment alterations could contribute to (dys)regulation of LECs in development and disease.

11.2 The Tissue Microenvironment of Endothelial Cells

11.2.1 The Endothelial Basement Membrane and Extracellular Matrix Components

During cell migration and sprouting in embryonic and postnatal (lymph)angiogenesis, but also in pathological (lymph)angiogenic processes, the ECM surrounds individual ECs or EC clusters in three dimensions (3D). After vascular remodeling and maturation, ECs form the inner layer of lymphatic or blood vessels and adhere to the surrounding abluminal ECM environment in two dimensions (2D).

In the established blood vasculature, the blood vessel BM is mainly composed of laminin 411 (Thyboll et al. 2002; Stenzel et al. 2011) and 511 (Di Russo et al. 2017; Hallmann et al. 2005), collagen IV (Pöschl et al. 2004), fibronectin (Zhou et al. 2008; Van Obberghen-Schilling et al. 2011), and many other molecules, such as the proteoglycans perlecan (Zoeller et al. 2008; Lord et al. 2014; Douglass et al. 2015), agrin (Barber and Lieth 1997; Steiner et al. 2014), and nidogens (Bader et al. 2005). Besides collagen IV, additional collagens, such as collagens XVIII (Marneros et al. 2004), VIII (Sage and Iruela-Arispe 1990) and VI (Kuo et al. 1997; Groulx et al. 2011) have been identified to be part of the blood vessel BM. Genetic deletion or point mutations of genes encoding for these ECM components lead to severe blood vascular dysfunction (George et al. 1993; Thyboll et al. 2002; Van Obberghen-Schilling et al. 2011) and are often associated with embryonic or perinatal lethality (Costell et al. 1999; Pöschl et al. 2004; Bader et al. 2005; Coles et al. 2006). These findings support a pivotal function of BM components in blood vascular development. Additionally, it is not only important that ECM components are correctly expressed and assembled in the extracellular space, they also need to be remodeled and/or degraded to allow for functional blood vessel morphogenesis and homeostasis. This is achieved by matrix remodeling enzymes, such as matrix metalloproteinases (MMPs), reviewed in Sounni et al. (2011).

Although many studies have been performed focusing on the importance and contribution of the BM in the blood vasculature, less is known about the BM composition of the established lymphatic vasculature. Lymphatic vessel BM has been shown to be composed of laminin 421 (Saito et al. 2009), collagen IV (Lutter et al. 2012), fibronectin (Ou et al. 2010; Podgrabinska et al. 2002), hyaluronan (El-Chemaly et al. 2009), and Emilin1 (Danussi et al. 2008). For example, the elastic microfibril-associated protein emilin1 is a component of the anchoring filaments in lymphatic vessels. Emilin1 knockout mice display hyperplastic and disorganized lymphatic vessels with impaired drainage function, reduced number of anchoring filaments and dysfunctional cell-cell junctions (Danussi et al. 2008).

In contrast to the repertoire of remodeling MMPs expressed by BECs, reelin was initially considered a major lymphatic remodeling enzyme, which degrades fibronectin and laminin components (Lutter et al. 2012; Samama and Boehm 2005). However, important roles for MMPs during physiological and pathological lymphangiogenesis have also been discovered. For example, MMP14 has been shown to control lymphangiogenesis through regulation of lymphatic vessel endothelial hyaluronan receptor-1 (LYVE-1) and vascular endothelial growth factor C (VEGFC) in mice (Wong et al. 2016). MMP2 blockage affects lymphatic tube formation in cell culture spheroids and ex vivo lymphatic ring assays (Ingvarsen et al. 2013). Similarly, inhibition of MMP2 and MMP9 has been associated with lower expression of VEGFC and vascular endothelial growth factor receptor 3 (VEGFR3) followed by impaired corneal lymphangiogenesis (Du et al. 2017). Taken together, these studies highlight a more complex contribution of MMPs in lymphatic vessel growth than previously thought.

To mediate adhesion to the tissue microenvironment, ECs are associated to the ECM through so-called focal adhesions, which are comprised of ECM binding proteins known as integrins (Gordon et al. 2020; Stupack and Cheresh 2004). Integrins bind to their specific ECM ligands and are intracellularly coupled to the cytoskeleton through actin binding proteins, including the mechanosensors vinculin and talin (Bays and Demali 2017; Yan and Schwartz 2018), to sense extracellular mechanical information and transduce them into the cell (Gordon et al. 2020). For example, fibronectin, which is not only an integral component of the lymphatic BM but also of the LV matrix, binds to endothelial integrin α9 (Bazigou et al. 2009). EC-specific deletion of Itgα9 or mice lacking the EIIIA domain of fibronectin show disorganized fibronectin network in the valve matrix resulting in dysplastic LV leaflets and retrograde lymph flow (Bazigou et al. 2009). Consequently, mechanosensing and mechanotransduction not only depend on ECM ligand availability, but also on the proper endothelial integrin repertoire.

Besides regulating adhesion and providing structural support for the endothelium, ECM components participate in the regulation of additional signaling pathways involved in blood vessel development and homeostasis through sequestration of chemokines or enhancement of chemokine activation. For example, VEGF signaling (Jakobsson et al. 2006; Stenzel et al. 2011; Cecchi et al. 2012), transforming growth factor β (TGFβ) signaling (Fontana et al. 2005), platelet-derived growth factor (PDGF) signaling (Lindblom et al. 2003), and angiopoietin signaling (Xu and Yu 2001; Xu et al. 2004; Chomel et al. 2009) have been identified to be regulated via the ECM. Similar to what has been shown for BECs, ECM components influence lymphatic signaling. Through ligation of integrin α5ß1 and fibronectin, the transactivation and phosphorylation of VEGFR3 are enhanced (Zhang et al. 2005). The collagen and calcium-binding EGF domain 1 (CCBE1) protein, which is expressed by mesenchymal cells close to the nascent lymphatics (Facucho-Oliveira et al. 2011), activates the cleavage of VEGFC to its active form, allowing for lymphatic sprouting (Jeltsch et al. 2014; Bos et al. 2011; Hogan et al. 2009).

Within the blood vasculature, relative amounts of ECM components have been shown to correlate with variable stiffness of the EC microenvironment. Arteries are stiffer (50–150 kPa, Kohn et al. 2015) than veins (3–50 kPa, Xue et al. 2017) because they need to withstand high blood pressure. In an ageing-related arterial stiffening mouse model, collagen I and III depositions have been shown to increase the incremental stiffness value of the arterial walls, indicating arterial stiffening in old mice (Fleenor et al. 2010).

In the developing embryo, collagen I is the predominant collagen type of the ECM and correlates with tissue stiffness (Majkut et al. 2013; Chen et al. 2012). Interestingly, analysis of collagen I deposition in embryonic day (E) 11 mouse embryos revealed substantial differences in collagen I support of nascent blood vascular versus lymphatic structures (Frye et al. 2018). Arteries and blood capillaries exhibited the highest collagen I density, while the cardinal vein (CV) and the migrating LECs showed moderate or low collagen I deposition, respectively (Fig. 11.1a, b). Decreasing collagen I density also correlated with decreasing local tissue stiffness experienced by ECs, suggesting that already during early vascular development the density of collagen I reflects an essential hierarchy of ECM stiffness required to fulfill different vascular functions.

Fig. 11.1
4 illustrations. Two microscopic views of E 11 transverse vibratome are presented in a and d. Two graphs in b and c depict collagen 1 density and tissue stiffness. The collagen 1 density is high for artery and low for avascular.

Changes in matrix stiffness regulate early lymphatic development. (a) Immunofluorescence of transverse vibratome sections of E11 Prox1-GFP embryos using antibodies against PROX1 (magenta), GFP (green, Prox1 reporter), and collagen I (single channel images). Note, aorta (A) and blood capillaries (BC) show higher collagen I level compared to the CV (arrow) and migrating LECs (arrowheads). (b) Quantification of collagen I density in the respective vessel types and avascular tissue in E11 embryos. Data represent mean integrated density values of corrected total cell fluorescence ± s.e.m. (unpaired Student’s t-test) quantified from n = 10 images taken from two embryos. (c) Ex vivo atomic force microscopy (AFM) measurements on transverse vibratome sections (image on the right) of E11 Prox1-GFP embryos. Young’s Modulus (kPa) is a measure for the actual tissue stiffness. Horizontal lines represent mean (n = 10 measurements from one embryo (CV) and n = 114 measurements from three embryos (outside CV)). p value, unpaired Student’s t-test. Measurements were done on the dorsal side of the CV (arrow) and the area of LEC migration (boxed area, outside CV). Prox1-GFP+ spinal cord (SC), dorsal root ganglion (DRG), and nerve (N) were used for orientation. Scale bars: 100 μm. Modified from Frye M, et al. Nat Commun. 2018;9(1):1511 und licensed under a Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/)

Full size image

However, it is important to note that the BM underlying most ECs is only a thin ECM layer (30–500 nm) (Liliensiek et al. 2009). Interestingly, cells can “feel” up to several micrometers deep into a soft substrate. Consistently, induction of EC network formation on compliant substrate (0.4 kPa) was prevented on very thin compliant substrates (<20 mm) as ECs sensed the stiffness of the underlying coverslip (Davidson et al. 2019). These findings suggest that the absolute in vivo stiffness of the ECM microenvironment experienced by ECs might rather be generated by several different surrounding tissue structures, including the ECM and, presumably, several layers of adjacent cells.

11.2.2 Mural Cell Support of the Endothelium

The specific tissue microenvironment of ECs is established not only by various ECM components, discussed in the previous paragraph, but also by surrounding cells. These cells can align the endothelium and provide a passive stiffness scaffold or are closely associated and actively regulating the endothelium, like mural cells.

Mural cells can be distinguished in pericytes and SMCs mostly depending on their morphology and location, although a rigorous distinction is not always possible. Pericytes are usually solitarily associated with small caliber vessels, but their relative frequency to ECs is highly variable (between 1:100 in skeletal muscle to 1:1 in the retina) (Armulik et al. 2005). Instead, SMCs can form multiple concentric layers around large caliber vessel types and can contract in order to facilitate the movement of fluids (Armulik et al. 2005; Gaengel et al. 2009). Both cell types derive from various cellular sources and through the combined action of hemodynamic forces and signaling induction, are recruited soon after the blood flow and circulation has been initiated (Shen and Mccloskey 2017). A well-studied signaling pathway is the platelet-derived growth factor subunit B (PDGF-B)/PDGF receptor (PDGFR) axis. Binding of PDGF-B, expressed by ECs, to the PDGFR, expressed on mural cells, activates mural cell proliferation and initiates their migration toward the vessels (Shen and Mccloskey 2017; Armulik et al. 2005). Global knockout mice for Pdgfb have been associated with microvascular aneurisms, loss of pericyte coverage, and embryonic lethality (Lindahl et al. 1997), while a more recent EC-specific knockout model exhibited increased retinal leakage in adult mice (Park et al. 2017).

In the lymphatic vasculature, mural cell recruitment starts from E 17.5 and proceeds after birth (Norrmén et al. 2009). PDGF signaling has been shown to be regulated downstream of forkhead box protein C2 (FOXC2) transcription factor (Petrova et al. 2004). Global knockout mice for Foxc2 display upregulated Pdgfb expression, resulting in increased recruitment of pericyte/SMCs to the dermal lymphatic capillaries at E17.5, whereas control littermates did not show mural cell coverage of these vessels (Petrova et al. 2004). These studies highlight an essential role of mural cell recruitment to stabilize blood and lymphatic vessels during embryonic development and in vessel maintenance.

In the established vasculature, mural cell coverage and specific deposition of ECM components can distinguish different vessel subtypes by morphology and function. In the blood vasculature, capillaries are surrounded by pericytes and a continuous BM, although organ specific differences are possible; for example, in the liver sinusoids that exhibit discontinuous BM coverage (Potente and Mäkinen 2017). Large caliber vessels are instead surrounded by a continuous BM and an elastic lamina, consisting of numerous layers of SMCs, to withstand higher blood pressure (Potente and Mäkinen 2017). In the lymphatic vasculature, the evident morphological differences between initial and collecting lymphatics resemble the function of these two lymphatic vessel subtypes. Through a highly permeable barrier, which is devoid of a mural cell layer and a continuous BM, lymphatic capillaries drain fluids, macromolecules, and cells from the tissue. In contrast, collecting lymphatics transport the lymph back to the blood circulation and are therefore equipped with a continuous BM and SMCs that contract to allow the movement of the lymph (reviewed in Chen et al. 2014; Stritt et al. 2021).

Taken together, ECM composition and mural cell support can provide first hints about the local stiffness environment of ECs and their exposure to stretch. In the future, it will be necessary to analyze local in vivo stiffness experienced by ECs in more detail, to better understand the relationship between the different microenvironmental factors (ECM composition, mural cell support, and absolute tissue stiffness) and how the combined action of these factors could render ECs more susceptible to dysfunction. Particularly, the heterogeneity of the tissue microenvironment between BECs and LECs might point toward the idea that selected mechanoregulatory processes and their threshold to malfunction may be essentially different in LECs. In the next paragraph, we will highlight recent findings on lymphatic mechanoregulation during development.

11.3 Mechanoregulation of Lymphatic Development

Changes of the microenvironment have been shown to fundamentally regulate cellular processes during development. For example, ECM stiffness regulates the differentiation of multipotent mesenchymal stem cells. Stiff matrices mimicking bone were found to be osteogenic, while soft matrices mimicking brain were neurogenic (Engler et al. 2006). Methylcellulose hydrogels, mimicking external constraints in the bone marrow, positively influenced megakaryocyte differentiation and proplatelet formation (Aguilar et al. 2016). Besides regulating cell fate and behavior at the single cell level, ECM stiffness can regulate tissue morphogenesis (Majkut et al. 2013; Poh et al. 2014).

Interestingly, fate decision of endothelial lineages is also dependent on the mechanical environment of the stem cell niche of vascular progenitor cells, as endothelial lineages prefer softer substrates (10 kilopascal (kPa)) and SMC lineages stiffer substrates (plastic, gigapascal (GPa) range) (Wong et al. 2019). Differences in ECM stiffness experienced by endothelial progenitor cells (EPCs) have been furthermore suggested to regulate arterial-venous differentiation in vitro (Xue et al. 2017). In contrast to EPCs that have been cultured on venous substrate stiffness (7 kPa), EPCs cultured on arterial substrate stiffness (128 kPa) showed an increase in expression of the arterial marker EphrinB2 (Zhang et al. 2005; Xue et al. 2017).

The development of the mammalian lymphatic system is initiated in the CV through trans-differentiation of a subpopulation of venous ECs around E9.5. These lymphatic endothelial progenitors express the transcription factor Prospero homeobox protein 1 (PROX1) (Yang et al. 2012; Wigle and Oliver 1999), which is activated through the SRY-Box transcription factor 18 (SOX18) (François et al. 2008; Yang et al. 2012; Srinivasan et al. 2007) and nuclear receptor subfamily 2 (NR2F2, also known as COUP-TFII) (Srinivasan et al. 2010).

Around E10.5, PROX1+ LEC progenitors, expressing VEGFR3, start to delaminate from the CV and intersomitic vessels and migrate dorsolaterally toward a gradient of VEGFC to form the first lymphatic structures, the dorsal peripheral longitudinal lymphatic vessel (PLLV), and the ventral primordial thoracic duct (pTD) (commonly referred to as jugular lymph sacs (JLS)) (Yang et al. 2012; François et al. 2012; Hägerling et al. 2013).

The earliest evidence of lymphatic mechanoregulation via the tissue microenvironment has been demonstrated around E11, when lymphatic endothelial progenitors experience a decrease in ECM stiffness upon delamination from the CV (0.2 kPa outside the CV versus 4 kPa inside the CV) (Fig. 11.1c, d (Frye et al. 2018)). This decrease in matrix stiffness induces a GATA2-dependent transcriptional program, which is required to form the first lymphatic vessels. Transcriptome analysis showed that LECs grown on a soft matrix (0.2 kPa vs. 25 kPa) exhibit increased GATA2 expression and a GATA2-dependent upregulation of genes involved in cell migration and lymphangiogenesis, including VEGFR3. Analysis of endothelial-specific Gata2 deletion in mice demonstrated a cell-autonomous function of GATA2 in regulating LEC responsiveness to VEGFC, thereby controlling LEC migration and sprouting (Fig. 11.2). The study further compared the mechanosensitive transcriptional programs activated in LECs in response to different mechanical stimuli, such as increased matrix stiffness and oscillatory flow, and revealed that they appear remarkably different (Frye et al. 2018). Additionally, in contrast to increased GATA2 expression in LECs upon exposure to a soft matrix, GATA2 expression in BECs has been reported to increase in response to increased mechanical stimulus, such as matrix stiffening and induction of oscillatory flow (Mammoto et al. 2009; Kazenwadel et al. 2015). It has been described in several cell types, including ECs, that GATA2 interacts with other transcriptional regulators, including Etv2 (Shi et al. 2014) and Lmo2 (Coma et al. 2013), to form multimeric transcription complexes. An interesting question would be whether oscillatory flow- and soft matrix-induced differences in GATA2-mediated regulation of target genes and cellular responses can be explained by formation of different transcriptional complexes in different EC types.

Fig. 11.2
A 3-part illustration of the cellular responses in lymphatic development such as sprouting and migration, proliferation, and lymphatic valve formation and maturation.

Mechanosignaling during lymphatic development. Schematic of mechanoregulatory processes during lymphatic development. Initially, venous-derived LECs migrate out of the CV and are exposed to soft matrix, which induces GATA2-dependent increase in VEGFR3 expression and LEC migration. Next, LECs arrest and initiate proliferation in response to increased stretch, which is caused by an increase in interstitial fluid pressure. Last, initiation of lymphatic flow is necessary to induce lymphatic valve formation and maturation. Top regulated genes and signaling pathways (red, up; green, down) are indicated

Full size image

In parallel to increased responsiveness to VEGFC that ensures efficient dorsolateral migration, proliferation is downregulated in LECs that experience a softer microenvironment. For example, when LECs were cultured on softer 2D substrates (0.2 kPa vs. 25 kPa), these cells also show a reduced proliferation parallel to an induction of VEGFR3 expression (Frye et al. 2018). Similarly, proliferation of BECs is reduced on soft 2D substrates but induced on rigid 2D substrates. Subconfluent human umbilical vein ECs (HUVECs) which were cultured on stiffer substrates (10 kPa vs. 1 kPa) increase vascular endothelial growth factor receptor 2 (VEGFR2) internalization and thus VEGFR2 activation (Lavalley et al. 2017). This phenomenon of reciprocal control of cell proliferation and migration has been described for other cell types (De Donatis et al. 2010). In the context of lymphatic development, it could be speculated that LECs are firstly subjected to an activation of their motility to populate the area of the JLS and secondly, once they have arrested migration, they induce a strong proliferative response to successfully expand the JLS.

A strong induction of proliferation indeed is achieved from E11.5 onward during lymphatic development, once JLS have formed and vessel expansion is induced (Planas-PAZ et al. 2012). Initially, an increase in interstitial pressure between E11.5 and E12 results in swelling of the interstitium, which leads to a stretching of the tissue and the ECM surrounding the JLS (Planas-PAZ et al. 2012), demonstrating again the importance of lymphatic mechanoregulation via the abluminal tissue microenvironment (Fig. 11.2). Stretching of LECs is evident by an elongation of the LECs between E11.5 and E12 and results in activation of VEGFR3 signaling and a transient strong increase in LEC proliferation, which was shown to depend on β1 integrins (Planas-PAZ et al. 2012). Conversely, when interstitial fluid pressure decreases between E12.0 and E12.5, possibly due to efficient fluid drainage via the expanded lymphatic vasculature, the proliferation rate of LECs declines. In contrast to BECs, normal elongation of LECs in vitro is around 4%, whereas an increased elongation of 8% already leads to transcriptional induction of inflammatory cytokines, fibrotic markers, and lymphangiogenesis (Wang et al. 2017a). This could suggest that similar to EC type specific FSS set points (Baeyens et al. 2015), LECs have a lower stretch set point.

An indirect contribution of the immediate lymphatic microenvironment has also been identified during early lymphatic development. The CCBE1 protein is expressed by mesenchymal cells close to the nascent lymphatics (Facucho-Oliveira et al. 2011). CCBE1 does not have lymphangiogenic activity on its own, however it has been shown to activate a disintegrin and metalloproteinase with thrombospondin motifs 3 (ADAMTS3), which in turn cleaves VEGFC to its active form to allow for a proper VEGFC gradient and lymphatic sprouting, both in mouse (Jeltsch et al. 2014; Bos et al. 2011) and in zebrafish (Hogan et al. 2009; Wang et al. 2020).

Upon vessel expansion around E13.5, LECs of the pTD and the PLLV start to sprout toward the viscera and the superficial lymphatic plexus, respectively (Jafree et al. 2021). At the same time, non-venous LECs arise and start to coalesce to contribute to lymphatic vessel formation. In addition to the classical venous origin, it has been shown that LECs can also derive from other cell types, such as dermal blood capillary plexus (Pichol-Thievend et al. 2018), hemogenic endothelium (Stanczuk et al. 2015; Klotz et al. 2015), non-endothelial second heart field (Maruyama et al. 2019; Lioux et al. 2020), and paraxial mesoderm (Stone and Stainier 2019). These LECs appear as independent cell clusters that later fuse and incorporate into the nearby forming lymphatic vessels. This process is referred to as lymphvasculogenesis due to its similarity to the vasculogenesis process, where BECs differentiate from single precursors and then coalesce (Martinez-Corral et al. 2015). It seems likely that these additional LEC sources initially experience a different tissue microenvironment compared to the CV microenvironment with moderate stiffness (4 kPa). However, the potential impact of a differential microenvironment on non-venous derived LEC function has not been addressed yet and would be interesting to investigate in future studies.

Approximately at E14.5, an additional mechanical stimulus, the lymph FSS, is exerted on the LECs of the nascent lymphatic network. The onset of the intraluminal lymphatic flow has been shown to fundamentally contribute to the development of LVs (Fig. 11.2). LVs are predominantly formed at lymphatic branches, where the laminar flow pattern of the lymph is disturbed. At these valve initiation sites, oscillatory flow upregulates GATA2 that in turn maintains high PROX1 expression and transcriptionally activates FOXC2 transcription factor (Kazenwadel et al. 2015; Sabine et al. 2012). This transcriptional change activates the calcineurin/NFATC1 axis, connexin 37, and integrin α9 expression leading to valvulogenesis (Danussi et al. 2013; Sabine et al. 2012; Bazigou et al. 2009; Norrmén et al. 2009). Recently, an additional target of GATA2, the atypical cadherin FAT4, has been identified to control LEC polarity in response to flow and is required for lymphatic vessel morphogenesis, including valve formation (Betterman et al. 2020). By using mice that are deficient for the platelet-specific receptor C-type lectin-like receptor 2 (CLEC2), a complete block of lymphatic flow can be achieved as blood backfills the lymphatic network (Sweet et al. 2015). Similar to Foxc2-deficient vessels, Clec2-deficient lymphatic vessels are premature and excessive and fail to initiate valvulogenesis. Expression of PROX1, FOXC2, and VEGFR3 remains high in LECs of the mature valve but is downregulated in the mature lymphatic collecting vessels after LV formation (Norrmén et al. 2009). Presumably, disturbed flow patterns downstream of the LV maintain those expression patterns.

During postnatal lymphatic development, through secretion of PDGF-B, SMCs are recruited to lymphatic collecting vessels but not lymphatic capillaries (Sabine et al. 2012; Wang et al. 2017b). Another important phenomenon of collecting vessel maturation is the reduction of their diameter starting at E16.5 throughout postnatal stages (Norrmén et al. 2009). However, a potential direct impact of mechanical forces, such as matrix stiffening induced by mural cells or LEC constriction via the tissue microenvironment, has not been studied yet.

As development proceeds, collecting lymphatics and initial lymphatics remodel and mature. At the beginning of lymphatic development, all LECs are connected to each other through continuous zipper-like junctions (Yao et al. 2012). Remodeling of lymphatic capillary junctions to discontinuous button-like junctions increased from only 6% at E17.5 to 35% at birth, 50% at postnatal day (P)7 and 90% at P28 (Yao et al. 2012). This may be caused by mechanical forces generated through transmural lymph flow, which has been shown to induce delocalization and downregulation of vascular endothelial cadherin (VE-cadherin) and PECAM-1 in in vitro experiments (Miteva et al. 2010).

In addition to the mouse model system, lymphatic development has been extensively studied in zebrafish. Similar to mammals, in zebrafish LECs are distinguished from the rest of the endothelium by PROX1 expression and lymphangiogenesis is highly dependent on VEGFC/VEGFR3 signaling (Yaniv et al. 2006; Van Impel et al. 2014; Dunworth et al. 2014; Küchler et al. 2006; Le Guen et al. 2014; Shin et al. 2016). However, in contrast to mouse lymphatic development, a bipotent progenitor cell division underlies LEC specification in the CV of the zebrafish trunk (Koltowska et al. 2015; Nicenboim et al. 2015). From around 36 hours post-fertilization (hpf), sprouting of LEC progenitors and venous ECs occurs simultaneously, in a process called secondary sprouting (Yaniv et al. 2006; Küchler et al. 2006). About half of the sprouts form venous intersegmental vessels, while the remaining become parachordal lymphatic progenitors (PLs) in the horizontal myoseptum at around 52 hpf (Hogan et al. 2009). Subsequently, the PLs migrate ventrally and dorsally from the myoseptum to form the thoracic duct and the dorsal longitudinal lymphatic vessel, respectively (Yaniv et al. 2006). The network is then completed with the connection of these two main trunk lymphatic vessels through lymphatic intersegmental vessels. As in mouse, other non-venous sources of LECs have been discovered in zebrafish, such as the ventral aorta lymphangioblasts that contribute to facial lymphatics (Eng et al. 2019).

Formation of the proper lymphatic vascular network is highly dependent on the precise regulation of cell proliferation, which is achieved by the mitogenic VEGFC/VEGFR3/ERK signaling. Downstream of VEGFC/VEGFR3/ERK signaling, the RNA-helicase DDX21 ensures for proper RNA biogenesis and cell cycle progression (Koltowska et al. 2021). Furthermore, similar to the process of LEC progenitor delamination from the CV in mice, during secondary sprouting in zebrafish proliferation is decreased (Jerafi-Vider et al. 2021). The decrease in proliferation was shown to be regulated via a VEGFC/VEGFR3/ERK-controlled cell cycle arrest. However, if the LEC microenvironment also guides cell cycle dynamics in zebrafish remains to be uncovered.

The microenvironment also plays an important role to guide LEC migration in zebrafish. Multiple cellular sources have been shown to guide migrating LECs and secrete guiding cues, such as chemokines and growth factors; these include intersegmental arteries, neurons, and fibroblasts (Bussmann and Raz 2015; Cha et al. 2012; Wang et al. 2020). In addition, LECs migrate alongside notochord sheath cells, which have been found to secrete localized type II collagen (Col2α1) to support PL migration through cell-ECM guidance (Chaudhury et al. 2020). Defects in Col2α1 secretion result in impaired migration of PLs. Similarly, the ECM protein polydom (also called Svep1), a ligand for integrin α9β1, is expressed by mesenchymal cells in intimate proximity of remodeling venous ECs and LECs in zebrafish and mice (Morooka et al. 2017; Karpanen et al. 2017). Zebrafish polydom/svep1 mutants exhibit a decrease in secondary sprouting, which leads to an increased number of intersegmental arteries. Consequently, a reduced number of PLs in horizontal myoseptum fails to migrate dorsally or ventrally and from the TD (Karpanen et al. 2017). These studies underline the importance to investigate the microenvironment of the developing lymphatic system. Whether, in addition to the local presence of selected ECM proteins, specific mechanical forces, such as matrix stiffness or stretch, are also involved in zebrafish, remains to be studied.

Mechanosensing of FSS-induced extracellular mechanical information has been extensively studied. FSS-induced mechanosensory mechanisms that regulate gene expression and cellular function include the regulation of ion channels and endothelial junctional protein complexes (Bálint and Jakus 2021). For example, the Piezo-type mechanosensitive ion channel component 1 (PIEZO1) has been identified to mediate mechanotransduction in the development and maintenance of the LVs (Nonomura et al. 2018; Choi et al. 2019). Furthermore, the calcium release-activated calcium modulator 1 (ORAI1), a pore subunit of the calcium release-activated calcium (CRAC) channel, is activated upon FSS and mediates Ca2+-influx in LECs (Choi et al. 2017b) and induces Kruppel like factor 2 (KLF2) and KLF4 upregulation in LECs to promote VEGFC expression (Choi et al. 2017a).

Once extracellular mechanical information of the tissue microenvironment has been sensed through integrins and the FA complexes in BECs, the intracellular signals are propagated to the actin cytoskeleton. For example, with increasing matrix stiffness (3 kPa, 12 kPa, and 1.5 MegoPa), actin cytoskeleton remodeling becomes more organized, with an increasing amount of actin stress fibers (Jannatbabaei et al. 2019). The actin cytoskeleton is connected to VE-cadherin via its intracellularly associated proteins β- and α-catenin (Wessel et al. 2014). Tension or strain-induced actin remodeling tightly controls assembly and disassembly of VE-cadherin-based junctions (Oldenburg and De Rooij 2014). Additionally, cytoskeletal pulling at the VE-cadherin complex also recruits the tension sensor protein vinculin via α-catenin to reinforce endothelial junctions (Huveneers et al. 2012; Daneshjou et al. 2015).

In contrast to BECs, our knowledge of stiffness-induced mechanosensory mechanisms in LECs is limited. Besides a direct effect on endothelial junction stability, cytoskeletal mechanotransduction can result in structural modification of membrane-bound or cytoplasmic proteins and their subsequent shuttling to the nucleus. For example, an important class of nuclear shuttling proteins consists of Yes-associated protein (YAP) and WW Domain-Containing Transcription Regulator Protein 1 (WWTR1/TAZ), which are downstream effectors of the Hippo pathway (Zhong et al. 2018). YAP/TAZ are shuttled to the nucleus in lymphatic ECs grown on stiff substrates (25 kPa vs. 0.2 kPa) and induce their target genes connective tissue growth factor (CTGF) and ankyrin repeat domain 1 (ANKRD1) (Frye et al. 2018). YAP/TAZ function has been extensively studied in the development of the blood (Neto et al. 2018; Sivaraj et al. 2020) and lymphatic vasculature (Cho et al. 2019; Cha et al. 2020; Grimm et al. 2019). Precisely how alterations in ECM stiffness might regulate those processes in LECs remains to be investigated.

Taken together, lymphatic development is guided via luminal (e.g. FSS) and abluminal mechanical forces, such as ECM composition, mural cell support, tissue stiffness and stretch capacity. Besides FSS-induced mechanotransduction, it is now highly relevant to investigate stiffness- and stretch-modulated LEC signaling pathways (including the identification of specific sensors and transducers), not only during lymphatic development but also during the maintenance of the established lymphatic system, as those pathways are likely to be dysregulated in a variety of disease conditions.

11.4 Mechano-Dysregulation of Lymphatic Endothelial Cells in Disease

Tissue remodeling and growth require ECM remodeling; however, aberrant ECM alterations have been associated with a variety of diseases, e.g. central nervous system (CNS) injury (Gaudet and Popovich 2014), tumor development and metastasis (Girigoswami et al. 2021; Nicolas-Boluda et al. 2021), lymphedema (Kistenev et al. 2019) or inflammatory bowel disease (Petrey and De La Motte 2017; Gordon et al. 2014). Uncontrolled remodeling of the ECM may lead to either an excessive degradation of ECM (Zhen and Cao 2014) or abnormal deposition and ECM stiffening (Frantz et al. 2010).

Comprehensive analyses of local stiffness changes experienced by LECs in diseased tissues have not been addressed. The temporal sequence of stiffness changes and LEC dysfunction and their interplay during disease progression are not understood. Here, we review selected diseases that are associated with lymphatic dysfunction and aim to point out possible hints that matrix alterations and modulation of ECM stiffness do not only correlate with lymphatic dysfunction but might mutually define each other.

11.4.1 Lymphedema

Lymphedema is a chronic disease that can occur anywhere in the body, including extremities, face, thorax, and different body cavities. Lymphedema can be inherited (primary lymphedema), affecting 1 in 100,000 Americans (Smeltzer et al. 1985; Sleigh and Manna 2021), or caused by obstruction and injury of the lymphatic system (secondary lymphedema). Secondary lymphedema has a much higher incidence of 1 in 1000 people in developed countries, mainly due to malignant cancer treatment through lymph node (LN) dissection (Azhar et al. 2020). However, it is very likely that the incidence is underreported, especially in lower income countries (Torgbenu et al. 2020).

Several mutations in human genes have been identified to cause primary lymphedema, including GATA2, SOX18, FOXC2, FLT4, PTPN14, PIEZO1 and ITGA9, affecting mainly LEC specification and lymphatic development (Oliver et al. 2020). For example, mutations in FOXC2 cause Lymphedema-distichiasis (LD) syndrome with lymphatic vessels appearing normal but showing impaired lymphatic drainage due to valve dysfunction (Brice et al. 2002, Petrova et al. 2004, Brice et al. 2002). Foxc2-deficient mouse embryos and LD patients also display an ectopic mural cell and BM coverage (Petrova et al. 2004). Mutations in FLT4 cause congenital bilateral lower limb lymphedema (Nonne-Milroy disease Karkkainen et al. 2004, Gordon et al. 2013) and loss-of-function mutations in GATA2 lead to impaired development and maintenance of lymphovenous and lymphatic valves (Emberger Syndrome, Ostergaard et al. 2011, Kazenwadel et al. 2012).

Secondary lymphedema can additionally be caused by infection with the nematode Wuchereria bancrofti. The adult worm can obstruct lymphatic vessels and lymphatic transport, when lodging in the lymphatic system, and triggers inflammatory responses of the host (reviewed in Bennuru and Nutman 2009), which causes more than 60 million patients with lymphatic filariasis worldwide and about 25% of those patients suffering from lymphedema (Ramaiah and Ottesen 2014). A high prevalence of secondary lymphedema in the USA is also related to malignancy and tumor therapy. It has been described to accompany treatment routines of lymphoma, melanoma, urologic cancers and receives special attention for occurring after surgical and radiation therapy for breast cancer in women. The incidence of lymphedema after mastectomy ranges between 24% and 49%. Other studies report 4% to 28%, probably due to different measurement techniques and criteria (reviewed in Warren et al. 2007, Ly et al. 2017).

Due to ongoing fibrotic processes, it is a commonly accepted idea that lymphedema is associated with tissue stiffening, although lymphatic-proximal stiffness analysis has not been performed. Thus, it remains unclear how changes in ECM stiffness may affect LEC function and disease progression.

To study lymphedema in mice, several experimental lymphedema models, such as the mouse tail surgery model, the popliteal LN dissection and the related axillary LN dissection model, or an inducible transgenic lymphatic ablation model have been described (Ly et al. 2017). The mouse tail surgery model is the most commonly used lymphedema model. After carefully removing the 3–5 mm skin, the superficial and deep lymphatic vessels are ligated, while blood vessels are left intact. The resulting inflammation, lymphatic fluid stasis and vessel dilation, fibrosis and adipose deposition are mimicking the human post-surgical lymphedema. The observed edema increased interstitial pressure and fibrosis formation suggests stiffening of the tissue (Kashiwagi et al. 2011). Recently an advanced tail surgery model has been developed. One lymphatic collector is maintained intact, allowing the study of functional changes during disease progression in the intact vessel, while the common lymphedema phenotype can be observed in the disrupted lymphatic collector (Weiler et al. 2019).

In lymphedema, mainly collagen fibers are excessively deposited in both dermis and subcutaneous tissue (Gardenier et al. 2016). However, not much is known about the spatial organization of the collagen fiber network, although work by Wu et al. showed progressively less compacted and disorganized collagen, due to excessive fluid separating the collagen in the mouse tail model (Wu et al. 2011). Similarly, the development of limb lymphedema in patients with lymphadenectomy is convoyed by a thickening of the BM and an increase in collagen fibers (Mihara et al. 2012). The collagen increase was also confirmed in patients with stage II lymphedema by visualization of collagen fibers using second harmonic generation (Kistenev et al. 2019). More recently, a transgenic mouse model has been developed: a tamoxifen-inducible Cre-loxP system that expresses the human diphtheria toxin (DT) receptor under the control of the LEC-specific Flt4 promoter. Injection of DT into any of the limbs results in a local ablation of lymphatics (Gardenier et al. 2016), resulting in a histological representation of the human disease by displaying comparable radiographical and clinical symptoms, including progressive dermal fibrosis and deposition of subcutaneous fibroadipose tissue. The latter has also been reported in the mouse tail surgery model, with an increase in fat thickness and subcutaneous fat deposition (Aschen et al. 2012). This is in agreement with several patient studies reporting that tissue swelling in lymphedema, can be caused through fluid stasis and fat deposition; for example, in lymphedemic limbs of breast cancer patients (Schaverien et al. 2018; Azhar et al. 2020).

Changes in ECM deposition are accompanied by another classic hallmark of lymphedema: chronic inflammation of the dermis and its underlying tissue. Lymphatic fluid stasis leads to an inflammatory response triggered by CD4+ T cell infiltration into the surrounding tissue. Almost 70% of all inflammatory cells in lymphedema are CD4+ T cells and their infiltration positively correlates with disease severity (Dayan et al. 2018; Ly et al. 2017). Interestingly, mice lacking all types of T cells or the CD4+ subpopulation T cells fail to develop lymphedema in the tail surgery model. The infiltrating CD4+ cells show a bias toward a T helper (Th) 2 response (Wynn 2008). Th2 cells are the main drivers of the inflammatory response by secreting pro-inflammatory/pro-fibrotic cytokines IL4 and IL13. Together with upregulation of TGFβ1 signaling, this leads to fibrosis formation, which ultimately impairs lymphatic function (Dayan et al. 2018; Ly et al. 2017). Additionally, T cell-derived cytokines IL4, IL13, interferon gamma and TGFβ1 directly exhibit anti-lymphangiogenic function by decreasing LEC proliferation, migration, survival and responsiveness to VEGFC in human dermal LECs and in a mouse model of suture-induced corneal neovascularization (Savetsky et al. 2015). Besides CD4+ T cells, other immune cells are also involved. Depletion of macrophages, for instance, led to increased CD4+ T cell infiltration and Th2 differentiation and thus causing similar phenotypes with increased fibrosis and impaired lymphatic functions (Ghanta et al. 2015).

It can be speculated that, like chronic inflammation, severe changes in ECM deposition and stiffness are likely to directly contribute to lymphatic dysfunction and dysfunctional lymphangiogenesis or even prevent lymphatic re-growth in lymphedema. In agreement with this, in vitro experiments using human LECs demonstrate that decreasing matrix stiffness primes lymphatic tube formation, while increasing matrix prevented it (Alderfer et al. 2021). In human patients, stiffening of lymphedema tissue is seldomly assessed in a quantitative and objective way. The subjective view of the patients and medical staff (referring to tightening and stiffness) is the most common assessment (see, for example, Pekyavaş et al. 2014). The most common method for evaluating lymphedema severity is the circumference of the limb or volume determination, despite the knowledge that for some patients the affected limb softens due to treatment, but the circumference is not altered. Softening would indicate an improvement of disease (discussed and reviewed in Dayan et al. 2018, Hara and Mihara 2018). Additionally, macroscale indentation techniques, such as elastography and tonometry, have been employed to study tissue stiffness in lymphedema (Nowak and Kaczmarek 2018; Hara and Mihara 2018), but proper local stiffness values or parameters of early lymphedema stages are lacking. Although several studies try to identify risk parameters for lymphedema formation, like body mass index, age, and therapy approach, for example in women who underwent mastectomy for breast cancer treatment (Basta et al. 2017), it remains understudied whether stiffness increase, sensed by LECs, may serve as an early indicator of disease formation prior to clinical symptom development.

For secondary lymphedema, a potential LEC mechano-dysregulation has been proposed: due to surgical removal of a cancerous lymph node, the lymph flow between afferent (upstream) and efferent regions (downstream of the truncated LN) is disrupted. The lymph transport from the afferent region is abolished and leads to lymph accumulation in the lymphatic vessel, causing vessel dilation and a reduction in lymph drainage. It has been hypothesized that, similarly to the embryonic development, an increase in intestinal fluid and the swelling of the ECM is sensed by the LECs, resulting in an increase in β1 integrin signaling, phosphorylation of VEGFR3, and elevated LEC proliferation with hyperplasia and further dilation of the lymphatic vessel (Planas-Paz and Lammert 2014).

Functional lymphatic drainage requires correctly organized LEC cell–cell junctions. It has been widely reported that LEC junctions are altered during inflammation. Mature button-like junctions can reversibly transform to zipper-like junctions in inflammation (Yao et al. 2010). Zipper-like junctions reduce permeability and fluid uptake from the interstitial space is limited. This may be accompanied by reduced lymphatic flow (Zhou et al. 2010; Huggenberger et al. 2010; Cromer et al. 2015), although an initial increased flow has been reported in acute inflammation (Zhou et al. 2010). Both infections and inflammation are implicated in the development of secondary lymphedema (Yuan et al. 2019). Although little is known about how exactly the lymphatic system is altered in structure and function in lymphedema patients, Zhang et al. propose that LEC junction zippering may play a role in fluid retention and tissue swelling (Zhang et al. 2020). The authors suggest that the promotion of button formation could improve lymphatic drainage and subsequently reduce lymphedema. How changes of the tissue microenvironment could contribute to junctional alterations and how this might be addressed therapeutically needs to be further explored.

To date, lymphedema treatment is mainly limited to conservative therapies, such as manual drainage through physiotherapy and compression garments. Alternative treatments offer low-level laser therapy, stem cell therapy, and VEGFC treatment (Dayan et al. 2018; Oliver et al. 2020). However, clinical studies revealed that targeting these key regulators may increase the risk of metastasis and tumor recurrence in cancer patients (Skobe et al. 2001; Baker et al. 2010; Dayan et al. 2018).

Conclusively, identification and characterization of alternative lymphatic signaling pathways that are regulated via the tissue microenvironment could offer new possibilities to modulate and normalize lymphatic behavior in lymphedema disease conditions. At the same time, a comprehensive pre-symptomatic and symptomatic determination of ECM alterations in patients susceptible to lymphedema, such as cancer patients, could increase prevention of lymphedema development or improve treatment strategies.

11.4.2 Inflammatory Bowel Disease

Inflammatory bowel disease (IBD) is an umbrella term for multifactorial disorders of the digestive tract, leading to inflammation. Development and course of these diseases are not only determined via genetic susceptibility and immune dysregulation, but also via the microbial flora and environmental factors (reviewed in Lee and Chang 2021). Crohn’s Disease (CD) and ulcerative colitis (UC) are two frequently observed types of IBD. While CD can potentially affect any area of the gastrointestinal tract and frequently causes transmural inflammation (all layers of the intestinal mucosa), UC predominantly affects the colon, and patients present with superficial ulcerations of the intestinal mucosa and submucosa (Eichele and Kharbanda 2017).

To study IBD-like diseases in mice, several mouse models have been developed. Genetically induced IBD can be observed in IL10 deficient mice that spontaneously develop colitis (Schwager and Detmar 2019), probably due to lack of the anti-inflammatory properties of IL10 (Kühn et al. 1993; Spencer et al. 2002), and in TNFΔARE mice that develop ileitis as a consequence of deleted tumor necrosis factor (TNF) AU-rich elements (ARE) and dysregulated TNF biosynthesis (Kontoyiannis et al. 1999; Rehal and Von Der Weid 2017). Furthermore, murine IBD can be induced by administration of dextran sulfate sodium (DSS), at a concentration of 1–5% in the drinking water (Okayasu et al. 1990). DSS administration subsequently leads to a damage of the intestinal epithelium, thus compromising its barrier function so that luminal bacteria and associated antigens can enter the underlying tissue and release pro-inflammatory factors (Perše and Cerar 2012; Wirtz et al. 2007; Okayasu et al. 1990; Kiesler et al. 2015). If given for seven days, DSS induces an acute inflammation from which animals can recover, while for a chronic disease development, repeated administration cycles are required (Okayasu et al. 1990).

Although most studies have been conducted to investigate the contribution of immune cells, the breakdown of the epithelial barrier (Chidlow et al. 2007; Wei et al. 2020; Stürzl et al. 2021), or the role of the blood vasculature and VEGFA in IBD (Chidlow et al. 2011; Scaldaferri et al. 2009), it is important to note that the disease is also characterized by lymphatic vessel dilation, dysfunctional lymphangiogenesis and increased mesenteric lymphatic vessel leakage (Rehal et al. 2017), phenotypes also described for lymphedema.

Consistently, lymphatic vessel density is increased along with VEGFC expression in inflamed colons of IBD patients (D’Alessio et al. 2014). Furthermore, the study showed that in two murine disease models (DSS and IL10 deficient mice) systemic delivery of VEGFC can reduce disease severity by inducting proliferation to increase lymphatic vessel density. Consequently, lymphatic drainage was partly rescued with immune cells being mobilized from the inflamed intestine to the draining LNs. Together these findings suggest that functional lymphangiogenesis might be an important process for the resolution of intestinal inflammation (D’Alessio et al. 2014). In contrast to that, although DSS concentrations and animal age varied, another study showed that overexpression of VEGFC in older DSS treated mice led to a significant increase in clinical disease index, inflammatory edema, increased lymphatic vessel density and size, suggesting that at different IBD stages, lymphangiogenic processes may have pleiotropic effects (Wang et al. 2016). Similarly, dilated and leaking lymphatic vessels were observed in the ileal mucosa of TNFΔARE mice (Rehal et al. 2017).

BEC-derived MMPs have been shown to play fundamental roles in IBD (O'Shea and Smith 2014). Absence of the blood endothelial derived protease MT1-MMP from ECs impedes colitis progression, which is accompanied by limited deterioration of vascular perfusions and retained well-structured collagen fibers surrounding the colonic crypts (Esteban et al. 2020). Similarly, MT1-MMP expression and activity is elevated in ECs grown on stiffer glycan cross-linked ribose-collagen I substrates (0.5 kPa) compared to softer, more compliant collagen substrates (0.18 kPa) (Bordeleau et al. 2017). The angiogenic sprouting of EC spheroids was reduced in stiffer compared to the softer matrices. It can be speculated that similar signaling pathways might be involved in lymphatic dysregulation in IBD. This would be in line with observations by D’Alessio et al. (2014) that lower density of lymphatic vessels might be linked to an increased risk of the recurrence of CD.

The lymphatic abnormalities in IBD are accompanied by changes in the ECM, which affect the tissue stiffness (as reported for lymphedema). A study in UC patients measuring the colonic tissue stiffness using a microelastomer, revealed a sixfold stiffness increase of unfixed UC strictures (16.7 kPa) in comparison to unaffected margins of the resected bowel (Johnson et al. 2013). The latter showed no difference compared to healthy intestine with 2.6 kPa and 2.9 kPa, respectively. Using a multi-scale indentation system, another study in seven CD and three UC patients conformingly observed an increase of steady-state modulus in inflamed tissue compared to unaffected areas, both for the colon (0.698 ± 0.463 kPa vs. 1.143 ± 0.488 kPa) and for the ileum (0.641 ± 0.342 kPa vs. 0.991 ± 0.379 kPa) (Stewart et al. 2018). Furthermore, they report an increase in COL1A1 and MMP-1 but no alterations of collagen IV and fibronectin content in inflamed tissue compared to unaffected areas. Notably, the study also highlights that colon stiffness (measured as effective total modulus) observed in the mouse, is two-fold higher than in unaffected human biopsies (8.493 ± 5.365 kPa versus 3.985 ± 2.656 kPa, respectively).

Interestingly, a softening of diseased colon tissue was observed in acute DSS at day (d)10 (6 kPa) and in ill IL10 deficient mice (3 kPa) compared to healthy wild-type mice, pre-symptomatic DSS at d4, and healthy IL10 deficient animals (11 kPa) (Shimshoni et al. 2021). This study employed AFM-based microscale stiffness analysis to determine local stiffness values. A softening of diseased colon tissue was further supported by deterioration of collagen, with regional ECM degradation and deposition, as well as a heterogenous ultrastructure (Shimshoni et al. 2021). For the first time, the authors further identify collagen XVIII and fibrillin 1 as biomarkers for a pre-symptomatic state, which still is void of clinical symptoms like body weight loss and endoscopic or histological phenotypes.

More in-depth research is needed to elucidate whether pathological ECM stiffness changes are fundamentally different in murine IBD-like disease compared to human IBD. Macroscale techniques like elastography (Maksuti et al. 2016) or tonography (Nowak and Kaczmarek 2018) to study tissue stiffness in lymphedemic and inflamed tissue analyze the general tissue stiffness or the entire vessel structure. However, they do not take into account cell-scale differences within the tissue, like changes of the EC-proximal tissue microenvironment. Microscale techniques like AFM (Frye et al. 2018) or 4D displacement microscopy (Vaeyens et al. 2020) could be better suited to answer these questions.

11.4.3 Tumor Microenvironment and Tumor Metastasis

Besides lymphedema and IBD, primary tumor development and tumor metastasis are known as lymphatic-associated processes (Oliver et al. 2020). The tumor tissue microenvironment is highly complex. Not only is it composed of different, often dysregulated, cell types, such as ECs, fibroblasts, pericytes, or immune cells (reviewed in Labani-Motlagh et al. 2020), but the tumor microenvironment is often characterized by ECM stiffening (Trédan et al. 2007). For example, cancer cells can regulate collagen synthesis and in turn collagen may alter cancer cell behavior through integrin signaling (Levental et al. 2009). This observation is however not limited to collagen, but can be extended to fibronectin, laminin, and other ECM proteins (Baghban et al. 2020).

Solid tumors often induce the expansion of the surrounding lymphatic network, with matured lymphatic vessels being restricted to the tumor margin. As a consequence, the intestinal fluid accumulates in the tumor leading to an increase in interstitial pressure (Padera et al. 2016). Moreover, tumors are characterized by a remodeled ECM, including but not limited to collagen deposition and cross-linking (Northey et al. 2017) and stiffening, which strongly correlated with cancer progression and metastasis (reviewed in Emon et al. 2018).

For example, Wei and colleagues showed a positive correlation between an increase in collagen IV expression and human colorectal cancer progression (Wei et al. 2017). Another study, using HUVECs and bovine aortic ECs in spheroid assays showed that tumor angiogenesis (outgrowth, invasion, and vessel branching) is connected to an increased collagen I matrix cross-linking, which is linked to increased matrix stiffness, as assessed by measuring equilibrium compressive modulus. They were able to show that MT1-MMP activity is upregulated in stiffer collagen I matrix (100 mM ribose) compared to softer (0 nM ribose) matrix. The authors further showed that the increased matrix stiffness resulted in an impaired barrier function and mis-localized VE-Cadherin employing tunable polyacrylamide-based hydrogels to mimic stiff (10 kPa) and soft (0.2 kPa) environments (Bordeleau et al. 2017).

These stiffness changes can be aligned with studies that measured the stiffness of tumor tissue. For example, fibrotic colorectal cancer tissue had a median stiffness of 7.51 kPa, compared with about 0.936 kPa for healthy tissue when measured using a macroscale indentation device (Kawano et al. 2015). For mammary cancer (Levental et al. 2009), an almost ten-fold stiffer elastic modulus of around ~2 kPa has been reported compared to healthy tissue (~0.2 kPa), using unconfined compression analysis. How the alterations in stiffness may regulate or prevent (lymph)angiogenesis is not fully understood, although the increase in interstitial fluid pressure is likely to cause compression of vessels, causing poor perfusion and resulting in hypoxia (Stylianopoulos et al. 2013). Additionally, it has been shown that BECs cultured within a stiffer 3D matrix sprout less compared to BECs cultured in softer 3D matrix (Trappmann et al. 2017), implicating that an unphysiologically stiff matrix could present a physical barrier for growing vessels.

LNs are the most common sites of tumor metastases and are crucial predictors of the cancer prognosis for the patient. The presence of tumor cells in the LN either reflects the cancer probability to metastasize, with the disease within the LN being inconsequential, or reflects the ability of the cancer cells in the lymph node to leave and spread the disease (Padera et al. 2016). Induction of VEGFC-mediated intratumor lymphangiogenesis at the location of the primary tumor can enable metastatic breast cancer cells to enter the lymphatic vessels, allowing for increased metastasis in LNs and lung (Skobe et al. 2001). However, additional data have recently emerged. Using VEGFR3-blocking antibodies in mouse melanoma models, it has been shown that active VEGFC signaling enables a better immunotherapy response by recruiting naïve T cells through CCL21 induction, which are then locally activated (Fankhauser et al. 2017). In a glioblastoma mouse model, ectopic VEGFC expression led to an enhanced priming of CD8+ T cells in the cervical LNs and migration of these T cells into the tumor, resulting in a rapid clearance of glioblastoma tumor (Song et al. 2020).

Taken together, lymphatic vessels play a crucial role not only in tumor development and metastasis, but also in tumor regression. It is likely that they respond to tumor ECM stiffness changes, and stiffening may prevent functional lymphatics to enter the tumor tissue. In parallel to the frequently discussed concept of blood vessel normalization (to facilitate chemotherapy) versus blood vessel regression (to starve the tumor) (Augustin and Koh 2022), similar questions should be applied to tumor lymphatic vessels. In addition to modulation of well-studied (lymph)angiogenic signaling pathways, normalization of the tumor microenvironment and stiffness (for example via “ECM softening”) and, in particular, modulation of ECM stiffness-regulated lymphatic EC signaling pathways, might present a promising approach to tackle persisting hurdles in tumor therapy.

11.5 Summary

In contrast to blood endothelial mechanoregulation, the importance of lymphatic endothelial mechanoregulation via the tissue microenvironment has only recently been discovered. ECM composition and signaling capacity, as well as mural cell support of BECs, have been extensively studied. Identification of mechanoregulatory processes during lymphatic development and in vitro studies on LEC regulation via the matrix environment indicate that similar molecular mechanisms are likely to regulate LEC dysfunction in diseases associated with ECM alterations. Therefore, it appears particularly important now to extend our analysis with novel in vivo approaches of local, LEC-proximal tissue stiffness and stretch measurements to obtain a more holistic perspective. We believe it is necessary to analyze tissue stiffness and stretch capacity experienced by ECs, to understand (1) which microenvironmental factors (ECM composition, mural cell support, and absolute tissue stiffness) are key modulators of absolute tissue changes sensed by ECs, (2) how these factors mutually define each other, and (3) to identify thresholds of the combined action of these factors that render ECs more susceptible to dysfunction.