Application of Microscale Culture Technologies for Studying Lymphatic Vessel Biology

Abstract

Immense progress in microscale engineering technologies have significantly expanded the capabilities of in vitro cell culture systems for reconstituting physiological microenvironments that are mediated by biomolecular gradients, fluid transport, and mechanical forces. Here, we examine the innovative approaches based on microfabricated vessels for studying lymphatic biology. To help understand the necessary design requirements for microfluidic models, we first summarize lymphatic vessel structure and function. Next, we provide an overview of the molecular and biomechanical mediators of lymphatic vessel function. Then we discuss the past achievements and new opportunities for microfluidic culture models to a broad range of applications pertaining to lymphatic vessel physiology. We emphasize the unique attributes of microfluidic systems that enable the recapitulation of multiple physicochemical cues in vitro for studying lymphatic pathophysiology. Current challenges and future outlooks of microscale technology for studying lymphatics are also discussed. Collectively, we make the assertion that further progress in the development of microscale models will continue to enrich our mechanistic understanding of lymphatic biology and physiology to help realize the promise of the lymphatic vasculature as a therapeutic target for a broad spectrum of diseases.

1. Introduction

A better understanding of how lymphatic vessels develop, grow, and function may be accommodated through the use of controllable experimental systems that mimic living tissue. Along these lines, it is well-appreciated by vascular biologists and physiologists that research in blood vessel angiogenesis and vascular function has benefitted immensely from the development of an impressive breadth of dedicated ex vivo, in vivo, and in vitro bioassays and techniques [1]. However, one can argue that comparable advances have not yet been made for studying lymphatic vessel biology, physiology, and lymphangiogenesis. One major constraint that had impeded progress in lymphatic culture models specifically was that the molecular markers for distinguishing lymphatic endothelial cells (LECs) from blood endothelial cells (BECs), such as lymphatic vessel endothelial hyaluronan receptor 1 (LYVE-1), prospero-related homeodomain transcription factor 1 (PROX-1), and podoplanin, were only recently discovered within the past 20 years [2-5]. Consequently, there is an opportunity to help improve both the quality and availability of in vitro experimental model systems for the application of dissecting important mediators of lymphatic vessel growth and function.

One of the most recent advances in in vitro culture models has been through the design and implementation of microscale culture technologies (e.g. microfluidics) as biomimetic platforms that reconstitute tissue level function in vitro [6]. Microfluidic devices are most commonly fabricated by rapid prototyping of poly(dimethylsiloxane) (PDMS), an optically transparent and gas permeable elastomer [7]. These devices contain networks of micron-scale fluid-filled channels that are similar in size and architecture to lymphatic vessels in vivo [8]. Moreover, microfluidic techniques enable controlled application of fluid flow and wall shear stress to cultured endothelial cells in a platform that can be readily scaled up for high-throughput screening studies [9]. These unique attributes of microfluidic models make them especially versatile for various in vitro, flow-based cell culture experiments [10,11]. Other enabling properties of microfluidics include controllable biomolecular gradients across 3-D scaffolds [12] and spatially-defined cell patterning [13]. These capabilities can be integrated into a single in vitro system that faithfully reconstitutes the mechanical, chemical, cellular, and matrix environments of in vivo physiology.

Here we describe the unique opportunities afforded by microscale culture models for advancing our understanding of lymphatic vessel biology. These microscale models, which are functionally analogous to living tissue, have already been widely adopted for studying blood vessel function and angiogenesis [9,14-19]. However, we argue that comparable progress has not yet been made in the applications of these microscale models for studying the lymphatic microcirculation, although advances to this field are occurring rapidly. To put forth a balanced perspective, we first give a brief introduction on the physiology of lymphatic vessels. Next, we give an overview of the mediators of lymphatic vessel formation and maintenance, with an emphasis on the biomechanical determinants. Subsequently, we provide a critical assessment of the current state of microscale culture models for studying the lymphatic vasculature. The scope of this review is the application of microfluidics for studying: 1) lymphatic vessel barrier function, 2) lymphangiogenesis and 3) interstitial mass transport mediated by lymphatics. Technical challenges and future directions for microfluidic approaches are also discussed. Our intention for this review is to serve not only biologists and physiologists interested in latest technological developments, but also help enable engineers and microtechnologists towards bridging their research with studying lymphatics as it pertains to human health and disease.

2. Lymphatic vessel physiology, function, and lymphangiogenesis

The lymphatic vasculature is comprised of two main elements: 1) the initial lymphatic vessels and 2) the collecting lymphatic vessels (Fig. 1). These two vessel types have distinct and complementary roles towards accomplishing the primary function of lymphatic vasculature, which is to prevent the accumulation of fluid pressure and subsequent edema in tissue [20]. The initial lymphatic vessels (also known as lymphatic capillaries) are small blind-ended vessels of 30 – 80 μm diameter. These vessels are composed of a single layer of lymphatic endothelial cells (LECs), a discontinuous basement membrane, and are not associated with pericytes [21]. The initial lymphatics are also characterized by unique, button-like, and overlapping cell-cell junctions, which function as primary one-way valves that facilitate absorption of interstitial fluid [22,23]. Fluid entry into the lymphatics is driven largely by pressure gradients and tissue distention that cause the primary valves of the initial lymphatics to be stretched open such that fluid flows along its pressure gradient and into lymphatics [21,22,24]. The interstitial fluid absorbed by the initial lymphatic plexus is converted to lymph and moved towards the collecting lymphatic vessels. Unlike the initial lymphatics, the collecting lymphatic vessels have an intact basement membrane, zipper-like interendothelial junctions, and perivascular smooth muscle cells (SMCs) [25]. Importantly, the intraluminal valves of the collecting lymphatic vessels prevent back flow, and the contractions by the SMCs propagate lymph fluid towards the lymph nodes where it is eventually returned to blood circulation [26,27].

Figure 1.

Lymphatic vessel anatomy. The initial lymphatic vessels absorb interstitial fluid while the collecting lymphatic vessels lined with smooth muscle cells (SMCs) pump fluid towards the lymph node. Expansion of the lymphatic network occurs by lymphangiogenesis or vessel sprouting.

Next, we will discuss the known molecular and biomechanical (Fig. 2) mediators of the lymphatic vasculature that have been studied with various ex vivo, in vivo, and in vitro models.

Figure 2.

Lymphatic mechanobiology. (A) Interstitial flow is driven by the pressure gradient oriented from the interstitium and towards lymphatic vessels. In addition to interstitial flow, lymphatic vessels in vivo are subjected to (B) intravascular shear stress, and (C) circumferential cyclic strain.

2.1. Molecular mediators of lymphatic vessel biology

The expansion of the lymphatic capillary plexus occurs by lymphangiogenesis, which analogous to the angiogenic growth of blood vessel capillaries, entails the extension of endothelial sprouts from pre-existing vessels [28], perfusable lumen formation [29], and associated remodeling of the extracellular matrix (ECM) [30,31]. The predominant signaling molecule that governs LEC biology and lymphangiogenesis is vascular endothelial growth factor (VEGF), particularly the VEGF-C and VEGF-D variants [32-35]. Activation of the receptor tyrosine kinase (RTK) VEGFR-3 expressed by LECs with VEGF-C/D drives lymphangiogenesis and the maintenance of new lymphatic vessels [20,36]. Loss of VEGFR-3 function results in the regression of lymphatic vessels [34] with mutated VEGFR-3 being an important predictor for diseases associated with lymphatic dysfunction [36-38]. Moreover, while both VEGF-C and VEGF-D signal through VEGFR-3, distinct receptor-binding specificities and biological activities are emerging between these two VEGF variants [39,40]. Presumably, these findings will have important implications for therapeutic targeting of VEGF-C and/or VEGF-D for controlling lymphangiogenesis.

While the VEGF-A/VEGFR-2 signaling axis is most commonly associated with blood vessel angiogenesis [41], it also plays a role in lymphangiogenesis [36,42]. For instance, inhibiting VEGFR-2 in adult mice resulted in decreased LEC sprouting, although overall lymphatic function remained intact [43]. Other signaling molecules that play an important role in mediating lymphangiogenesis include angiopoietins, which bind angiopoietin receptors (TIE-1 and TIE-2) and are known stimulators of lymphatic growth [36,42,44]. Sphingosine 1-phosphate (S1P) helps regulate endothelial permeability and vascular morphogenesis and is required for normal lymphatic patterning [36,45]. In addition, basic fibroblast growth factor (bFGF) has been shown to induce LEC proliferation, migration, and contributes to lymphangiogenesis [46,47]. Other molecules that that have been identified to regulate lymphangiogenesis include collagen and calcium binding EGF domain-containing protein 1 (CCBE1) [48,49] and platelet-derived growth factor B (PDGF-B) [50].

A number of in vitro studies have focused on probing for the specific functions of signaling molecules. For example, Knezevic et al. formed lymphatic and blood vessel networks to interrogate the effects of VEGF-C and adipose-derived stromal cells (ASC), demonstrating the necessity of cell-cell contact between LECs and ASC to form microvascular networks [51]. In addition, Gibot et al. successfully developed a stable 3-D lumen-containing lymphatic capillary networks by co-culturing LECs with fibroblasts on sheets of tissue-engineered skin substitutes [52]. Using this model, they showed that fibroblast-derived hepatocyte growth factor (HGF) and VEGF-C are necessary for enhanced proliferation and tube elongation by lymphatic vessels.

2.2. Lymphatic vessel mechanobiology

Mechanical stimulation due to interstitial fluid flow or slowly moving convective flow through the tissue interstitium has emerged as a potent mediator of vascular morphogenesis [53,54]. The driving forces for interstitial flow are hydrostatic and osmotic pressure differences between blood vessels, the interstitial space, and lymphatic vessels [55,56] (Fig. 3). These forces (also known as the Starling forces [57]) determine the rate of transendothelial filtration (Fig. 3). The level of interstitial flow is highly dependent on the state of the tissue. For example, under normal conditions, interstitial flow velocities are ~1 μm/s [58] but these levels elevate greatly during inflammatory responses [54]. It is important to note that unlike blood flow, which is driven by a central pump, interstitial flow comprises an open flow loop. Thus, there are numerous challenges to manipulating and accurately measuring interstitial flow in vivo [59]. To this point, much of our current understanding of lymphatic vessel responses to interstitial flow have been obtained with macroscale in vitro 3-D flow chambers, such as radial flow chambers or modified 3-D Boyden chambers (i.e. Transwell filter systems), that were pioneered by the group of Swartz [54,60,61]. These chambers enabled the assessment of the biophysical effects of interstitial fluid flow in 3-D with integration of cellular and molecular analysis in situ.

Figure 3.

Interstitial and transvascular fluid transport. The determinants of interstitial and transvascular transport is governed by Starling’s equation [31]. The hydrostatic pressure gradient (ΔP = Pc − Pi) causes fluid to leave blood vessels (filtration) whereas the oncotic pressure gradient (LΔ = Δc − Δi) causes fluid to enter blood vessels (reabsorption). The level of transvascular volume flux (Jv) scales with the hydraulic conductivity (Lp), which is a function of vessel permeability. Fluid outside of blood vessels moves through the interstitium as interstitial flow that is collected by lymphatic vessels.

In addition to interstitial flow, another important fluid mechanical force in the lymphatic vessel network is intravascular shear stress. LECs are subjected to a wide range of shear stress levels, with typical levels ranging between 4–10 dyn/cm² [62]. Recently, shear stress has been shown to orchestrate the molecular events that control lymphatic valve formation and stabilization [63-65]. Cyclic strain, or the circumferential expansion of lymphatic vessels, is another mechanical force that is prominent in the lymphatic vasculature [62,66]. The initial lymphatics are subjected to circumferential expansion or contraction due to local tissue stresses, which regulate entry of interstitial fluid through their primary valves [21]. The collecting lymphatics also experience significant cyclic strain levels during the opening and closure of the intraluminal lymphatic valves [62]. Interestingly, it was shown using mathematical modeling that the divergent responses to shear stress and vessel stretch produced two complementary mechanobiological oscillators that are sufficient to control lymphatic pumping and lymph transport [67]. Shear stress induces the release of the vasodilator nitric oxide (NO) by LECs while Ca²⁺-mediated contractions can be triggered by vessel stretch. Impressively, this simple mechanism demonstrated adaptability and autoregulation in lymphatic pumping responses to tissue-level changes in pressure [67]. Nonetheless, further experimental validation may be warranted, for example through the perturbation of the relevant signaling pathways using genetic manipulation or pharmacological inhibitors [68,69].

In accordance with this preceding statement, intravascular shear stress due to lymph flow has also been shown to help drive the genetic program that mediates valve morphogenesis and maturation in the collecting lymphatics [70]. The transcription factors PROX1 and FOXC2 are known to be upregulated and required from the onset of lymphatic valve formation [71,72]. However, a new perspective was put forth by the study by Sabine et al. which elucidated the interplay of PROX1 and FOXC2 with fluid flow in regulating the expression of the gap-junction protein connexin 37 (Cx37) and the activation of calcineurin [63]. Cx37 and calcineurin-dependent signaling were previously shown to be required for lymphatic valve formation [71,73]. Yet, especially intriguing from a mechanotransduction perspective was that time-varying oscillatory shear stress (OSS) and not constant laminar shear stress (LSS) enhanced the expression of FOXC2, Cx37, and calcineurin activation [63]. More recently, OSS was shown to promote lymphatic valve development mediated by canonical Wnt/β-catenin and GATA2 signaling in LECs [65,74,75]. In addition, the role of FOXC2 was further elucidated to promote overlapping intercellular junctions and induce growth arrest in LECs in response to OSS to help promote valve self-organization and stabilization [64]. Collectively, these emerging results suggest that OSS and transient retrograde flow encode specific responses to LECs at the future location of valve leaflets that lead to valve initiation, formation, and maintenance. Despite these significant advancements, a challenge towards a more complete understanding of the mechanisms of force-sensing in LECs is that OSS often regulates molecules that also have flow-independent roles [70]. However, recently the mechanically activated cation channel PIEZO1 [76] was shown to be required for lymphatic valve formation [77], thereby identifying a putative mechanosensor that may be targeted to help enhance lymphatic valve regeneration.

Unlike most ex vivo preparations, many in vitro systems can readily apply controlled levels of dynamic fluid flow. For example, 2-D in vitro flow chambers were used to show that cultured LECs in response to OSS reproduced features of lymphatic valve forming cells such as upregulation of lymphatic valve makers [63-65,75]. In addition, 2-D in vitro microchannel systems are widely used to study LEC monolayer responses to shear stress. Jafarnejad et al. showed that shear-mediated calcium signaling is dependent of the magnitude of applied shear stress, suggesting the importance of dynamic calcium flux in lymphatic mechanotransduction [78]. In a similar experimental setup, Kassis et al. further demonstrated different responses of shear-mediated intracellular calcium signaling in LECs upon simulation with lipoproteins, implicating the negative effect of lipids on lymphatic functions [79]. Other flow chamber models have been developed to study LEC response to shear stress. For example, Breslin et al. used an Electrical Cell-Substrate Impedance Sensor (ECIS) where it was found that LECs dynamically alter their morphology and barrier function in response to changes in shear stress in a Rac1-mediated manner [80]. More recently, Ostrowski et al. developed an impinging flow device capable of exposing LECs to gradient levels of shear stress that varied from 0 to a maximum range of 9–210 dyn/cm² [81]. In this study, LECs migrated against the flow direction and concentrated at regions of maximum shear stress when at high confluence. However, these same cells at lower density exhibited the opposite behavior by moving with the direction of flow. Using a similar in vitro system, Surya et al. demonstrated that S1P signaling is involved with the collective migration of LECs [82].

Although there is a wealth of evidence that fluid forces affect LEC phenotype, to our knowledge, no study has simultaneously examined the interplay of interstitial flow, luminal shear, and cyclic strain in mediating lymphangiogenesis from an intact vessel with facile experimental manipulation. Microfluidic systems have previously been applied for systematically dissecting the contributions of multiple instructive biophysical and biochemical (or physicochemical) cues on blood vessel angiogenesis (e.g. combined shear stress, interstitial flow, and biomolecular gradients) [16,83,84]. Therefore, we anticipate that future applications of microfluidic models will make comparable advances for studies pertaining to lymphatic vessel mechanobiology. In the next section, we elaborate on the enabling attributes of microfluidic technology.

3. Microfluidic approaches for studying lymphatic vasculature

Many of the key regulators of lymphangiogenesis have been identified in in vivo model systems [85-87]. However, one challenge for studies conducted in vivo is the involvement of confounding inflammatory reactions that are difficult to discriminate from the direct physicochemical effects on LECs [88]. In contrast, microfluidic models are comprised of the necessary 3-D matrix and cell components that are assembled using a microscale “bottom-up” approach to enable exquisite control over the chemical and physical environment such as tuning of ECM properties, spatial patterning of cellular constituents, and specification of biomolecular gradients [6,8]. Therefore, 3-D microfluidic devices permit visualization of lymphatic sprouting, quantification of vessel permeability, and ECM remodeling since they are readily compatible with labeling and imaging techniques such as immunofluorescence, confocal reflectance microscopy, and second harmonic generation imaging. Below we highlight the application of microfluidic technologies in studying three important physiological functions of lymphatics: 1) lymphatic vessel permeability (or barrier function), 2) lymphangiogenesis, and 3) mass transport mediated by lymphatics.

3.1. Quantifying lymphatic vessel barrier function

Regulation of lymphatic barrier function is crucial for optimal lymph formation and transport [80]. Disruptions to lymphatic barrier function is linked to various diseases [89]. For example, malabsorption of dietary fats or lipids by lacteals, or the lymphatic capillaries of the small intestine, can lead to severe clinical pathologies such as intestinal lymphangiectasia [90,91]. In addition, impaired lymphatic barrier function is a risk factor for type 2 diabetes, obesity-associated metabolic disorders, and cardiovascular diseases [20,92,93]. Harvey et al. reported that inactivation of a single allele of Prox1 gene in mice caused adult-onset obesity due to ruptured lymphatic vessels [94]. Moreover, chronic lymphedema, characterized by local accumulation of interstitial fluids due to impaired lymphatic drainage, commonly leads to an abnormal buildup of adipose tissue [95,96]. Lymphatic dysfunction is also involved with the pathogenesis of Crohn’s disease, which is a chronic inflammatory disease of the gastrointestinal tract [97].

Ex vivo approaches using isolated lymphatic vessel segments have been widely used to report values for lymphatic vessel permeability [69,98,99]. However, these measurements have predominantly been in collecting lymphatics. Quantitative permeability values of lymphatic capillaries are not widely reported due to technical difficulties in cannulating this vessel type [100]. Yet, lymphatic capillaries are especially permeable due to their overlapping and discontinuous interendothelial junction structure [21], which also renders these vessels more susceptible to paracellular intravasation by immune and cancer cells [101]. Therefore, microfluidic technology may offer engineering-based approaches for reconstituting lymphatic capillary analogues in vitro and for quantifying vessel permeability in response to external stimuli.

One microfluidic approach for quantifying vessel permeability is to use microchannels that are templated within a semi-porous ECM scaffold such as collagen. This technique was pioneered by the group of Tien where a microchannel is formed by casting collagen housed within a PDMS chamber around a cylindrical needle or rod that is ~100 μm in diameter [15]. Once the scaffold has polymerized, the needle or rod is removed leaving an open cylindrical microchannel that is embedded within the 3-D scaffold. Subsequently, the circular microchannels are seeded with endothelial cells that are allowed to attach and spread on the internal surface to yield a fully endothelialized circular lumen. Using this approach, Tien and colleagues measured changes in the permeability responses of blood endothelium to inflammatory cytokines [15] and cyclic AMP [102]. Moreover, this approach was adapted for the lymphatic endothelium, where unlike the microengineered blood microvessels, many focal leaks were observed in the microengineered lymphatic microvessels [103]. However, treatment of the lymphatic microvessels with cyclic AMP significantly improved barrier function and increased VE-cadherin expression levels at cell-cell borders [103]. Another approach for measuring vessel permeability was demonstrated by Sato et al. where a two-layer microfluidic model with upper and lower channels was separated by a semi-porous membrane [104]. Using this device setup, the vascular permeabilities between co-cultured LECs and BECs was assessed. Moreover, this study assayed for the effects of vascular damage incurred by habu snake venom on blood endothelium permeability and lymphatic return rate.

In addition to probing the effects of biomolecular signaling molecules, microfluidic models have been used to study the role of fluid mechanical forces, such as shear stress, on vessel permeability. However, the findings by the research field have so far focused on blood and not lymphatic vessels. For example, in another study by Tien’s group using the previously described cylindrical microchannel model, it was shown that barrier function of blood endothelium was a function of shear stress levels [105]. Buchanan et al. also used a cylindrical microchannel model that contained a BEC-lined microvessel that was surrounded by MDA-MB-231 breast cancer cells [106]. This study demonstrated that increasing levels of shear stress ranging from 1–10 dyn/cm² lowered vessel permeability by counteracting the effects of the pro-angiogenic molecules secreted by the MDA-MB-231 cells. More recently, it was demonstrated in a microfluidic blood vessel bifurcation vessel model that the local flow dynamics due to the branching vessel architecture controlled endothelial permeability in a NO-dependent manner [14]. Nonetheless, these aforementioned microfluidic systems for studying blood microvessels can be readily adapted for the lymphatic vasculature for assessing shear-mediated changes in lymphatic vessel permeability, which may be further warranted due to prior work implicating shear-induced NO release and bioavailability in regulating lymphatic permeability in type 2 diabetic mice [107].

3.2. Investigating the mediators of lymphangiogenesis

Lymphangiogenesis plays an important role in a wide range physiological conditions and pathological complications [108,109]. For example, inflammation induced by bacterial lipopolysaccharides (LPS) has been shown to trigger lymphangiogenesis by upregulating VEGF-C by LECs [110]. Increased lymphangiogenesis has been reported in the joints of mice with inflammatory arthritis [111]. The resulting increase in lymphatic vessel density augments macrophage activation and mobilization during immune responses [112]. Patient autopsies revealed that increased lymphangiogenesis often occurs near arterial walls that are subject to inflammation or plaque buildup [113]. These newly formed lymphatic vessels serve to remove the deposited cholesterols and assist in alleviating atherosclerosis [20]. Moreover, recent studies using conditional mouse models have provided novel insights on the therapeutic implications of enhancing lymphatic vessel growth. For instance, it was shown that stimulating cardiac lymphangiogenesis in vivo with VEGF-C improves immune cell clearance and the resolution of inflammation following myocardial infarction (MI) [114]. Similarly, kidney specific expression of VEGF-D increased renal lymphangiogenesis, reduced renal immune cell accumulation, and prevented hypertension [115]. Furthermore, lymphangiogenesis is required to mitigate impaired wound healing which is a common complication in diabetic patients [116].

While lymphangiogenesis is essential to coordinate immune responses, increased lymphatic vessel growth may have unfavorable outcomes in certain settings [117]. One such example is increased lymphangiogenesis enhanced immune responses in kidney and corneal transplant rejection [118,119]. In addition, aberrant levels of lymphangiogenesis is common in numerous primary human cancers including melanoma [120], breast [121], lung [122], and head and neck [123,124]. Furthermore, increased lymphangiogenesis, particularly at the margins of primary tumors, is strongly implicated in metastasis through the lymphatic vasculature [125,126]. Elevated expression of VEGF-C and VEGF-D by cancer cells is also correlated with poor prognosis, invasion, and metastasis [32,127]. Additionally, the CCL21/CCR7 chemokine axis induces secretion of VEGF-C by breast cancer cells to promote lymphangiogenesis [128].

Systematic exploration of the mediators of lymphangiogenesis under controlled culture conditions may help improve our fundamental understanding of the etiology of lymphatic pathologies. To this end, a significant advance in 3-D culture models of lymphangiogenesis was previously reported by Bruyere et al. who developed an ex vivo lymphatic ring assay [88], which was an adaptation of the widely used aortic ring assay for angiogenesis [1]. In addition, the rat mesentery has effectively been used as an ex vivo model for comparing the effects of VEGF-C on lymphatic and blood vessel sprouting density [129], and understanding the role of VEGF-A in increasing the frequency of blood-lymphatic misconnections [130]. However, both the aforementioned lymphatic ring and rat mesentery are static culture models [131]. Consequently, they are not conducive for dynamic control of interstitial pressure gradients, transmural pressure, and intravascular shear stress. Therefore, a better understanding of the complex interplay between molecular, cellular, and fluid mechanical factors during lymphangiogenesis may be achieved with 3-D microfluidic models. One recent advancement was demonstrated by Kim et al. who developed a 5-parallel channel microfluidic model to investigate lymphangiogenesis in vitro when co-cultured with fibroblasts (Fig. 4). This report demonstrated that a cocktail of pro-angiogenic factors that included VEGF-A, VEGF-C, bFGF, and S1P was the most potent of the combinations tested for triggering lymphatic vessel sprouting [132]. Moreover, this study demonstrated that interstitial flow (0.5 – 4 μm/s) selectively augments lymphangiogenesis when applied opposite or against the direction of sprouting. Together, these results suggest that biomolecular factors and interstitial flow cooperate in guiding lymphangiogenesis. It is noteworthy that the outcome described by Kim et al. is similar to what was previously reported with the application of microfluidic systems on the directional bias of blood vessel angiogenesis when interstitial flow is oriented against the direction of sprouting [17,84,133].

Figure 4.

Microfluidic approaches for studying lymphatic vasculature. (A) Schematic of a representative configuration of a microfluidic vessel and corresponding example [128]. The multiple parallel microfluidic channels fully-lined with LECs and laterally adjacent to a localized ECM for which the LECs can undergo sprouting. (B) Application of biochemical gradients and corresponding example [142]. Microfluidic models can be configured such that a sustained biochemical gradient of a signaling molecule (e.g. VEGF) forms from the source channel, across the ECM, and into the sink channel. (C) Controlled application of fluid flow and corresponding example [135]. Microfluidic systems are highly versatile in the nature of fluid flow that can be applied. Therefore, these systems are conducive for studying LEC mechanobiology in response to interstitial flow and laminar shear stress. Microfluidic systems can also be applied to study the effects of drug transport combined with interstitial flow that is representative of the tumor microenvironment.

In a separate study, Chung et al. created a biomimetic tumor microenvironment (TME) model that allowed for simultaneous observation of angiogenesis and lymphangiogenesis while co-cultured with cancer cells and fibroblasts [134]. Choi et al. used a cylindrical 3-D vessel assay that demonstrated that laminar shear stress (2 dyn/cm²) selectively suppresses Notch activation in LECs but not BECs, resulting in an increase in lymphatic sprouting [135]. In a previous microfluidic study, it was shown that laminar shear stress (3 dyn/cm²) suppressed VEGF-A induced blood vessel sprouting [84]. Therefore, the results reported by Choi et al. provide a molecular basis for the divergent shear stress-mediated responses by LECs and BECs as it pertains to lymphangiogenesis and angiogenesis [135]. Finally, a recent study by Kamm and colleagues developed a dual cylindrical vessel model to form fully-perfusable and parallel lymphatic and blood microvessels [136]. While the engineered blood microvessel displayed zipper-like junctions, the lymphatic microvessel exhibited both zipper-like and button-like junctions when treated with the anti-inflammatory drug dexamethasome. In addition, it was shown that application of a VEGFR-3 inhibitor (SAR131675) suppressed sprouting from the lymphatic vessel when grown as a monoculture. While this aforementioned result was expected, a somewhat surprising outcome was that the effects of inhibiting VEGFR-3 were abrogated when the lymphatic vessel was co-cultured with a blood vessel. The authors attributed this result to MMP secretion by BECs during angiogenesis, which concurrently facilitated lymphangiogenesis. Therefore, this study provided interesting insights on the molecular cross-talk mechanisms that arise during simultaneous angiogenesis and lymphangiogenesis.

3.3. Studying the effects of lymphatics on interstitial mass transport

Since the lymphatic circulation mediates tissue fluid balance [20], it is a critical determinant of convective fluid and solute transport in interstitial tissue that is necessary for the distribution of nutrients, waste products, and signaling molecules involved in regulating cell function [55]. In addition, the lymphatic system also has an important role in regulating lipid transport [137]. The attributes of microfluidic models make them readily capable for precise quantitative analysis of the mass transport properties of the interstitium [138]. This principle was recently demonstrated by Thompson et al. where the fluid drainage properties of microengineered blind-ended lymphatic vessels embedded within collagen scaffolds were characterized [139]. Interestingly and somewhat surprisingly, the solute drainage rates were greater in collagen gels that contained lymphatic vessels than in those that had bare channels by preventing solute transport back into the scaffold. Therefore, this study represents a significant step in successfully engineering the fluid absorption functions of the initial lymphatics in vitro.

Other in vitro approaches, such as Transwell filter models, have also paved a way for studying interstitial transport properties of lymphatic endothelium in the settings of cancer metastasis and immune systems [140-142]. Similarly, microfluidic systems have been applied to understand how interstitial mass transport is influenced by functional lymphatics. One example in the context of tumor interstitial transport is a tumor-microenvironment-on-a-chip (T-MOC) model that features distinct blood capillary, interstitial, and lymphatic compartments with independent control of fluid pressure [143]. This system demonstrated that the transport of nanoparticles is significantly hindered when IFP is higher than the capillary pressure, thereby replicating a physiological barrier posed by solid tumors that prevents uniform delivery and diminished efficacy of cancer therapeutics [144]. In another study from the same group, the T-MOC model was used to study the differential responses of breast cancer subtypes to doxorubicin-loaded nanoparticles [145] (Fig. 4). While these studies demonstrated the influence of cancer drug transport due to passive drainage to lymphatics, recapitulating the active transvascular mechanisms into these systems would both be an impressive technical achievement and augment the physiological relevance of the described microfluidic systems. Future studies that integrate the capabilities of microfluidics to reconstitute distinct tissue compartments while also incorporating independent control of biomechanical and chemical mediators will help further develop strategies for drug delivery in the lymphatic vasculature.

3.4. Limitations

PDMS is the material of choice for microfluidic systems due to its desirable qualities that include ease of fabrication and optical clarity. One advantage of PDMS is the ability to tune the hydrophobic surface properties to become more hydrophilic to improve biocompatibility and biofunctionality using oxygen plasma treatment, surface coating, and silanization [146,147]. However, one important limitation of PDMS is that this material is prone to non-specific absorption of proteins and hydrophobic drug molecules, thereby reducing the effective concentrations delivered to cells [148]. Researchers are addressing this concern using chemical surface modifications [149] or alternative materials [150,151]. While the elastomeric properties of PDMS make it suitable for long-term experimental usage without significant mechanical deformation, the Young’s modulus (a measure of mechanical stiffness) of PDMS is much higher than physiological tissue. The stiffness of PDMS can be modified by adjusting the ratio of basic and curing agent of PDMS which has been shown to regulate cell responses of muscle and nerves [152]. Nonetheless, incorporating natural scaffolds or hydrogels such as collagen or fibrin gels into microfluidic systems can help ensure that the in vitro microtissues that are formed matches the mechanical properties of physiological tissue [153,154]. Recently, 3-D printing of biocompatible polymers has enabled an additive manufacturing approach for constructing in vitro biomimetic scaffolds [155]. Successful integration of established microfabrication techniques with 3-D printing should enable the biofabrication of increasingly more complex lymphatic microenvironments in vitro.

4. Future outlooks

From the lymphatic biologist perspective, the ideal microfluidic device should mimic conditions in living organisms and be easy to assemble. It should also enable the tight control and manipulation of all parameters relevant to lymphatic biology, be scalable to high-throughput analysis, and facilitate the application of standard methods of molecular analysis and live imaging [156]. The immense progress in microfabrication and biological integration has helped bridge this technology gap for organotypic cell-culture studies [6]. Nonetheless, numerous opportunities for technological advancement remain. Microscale culture models are readily amenable for incorporating novel molecular and genetic approaches, force probes, and imaging and detection techniques, which can be cultivated through collaborations with biomedical scientists and physical scientists and engineers. For example, with CRISPR engineered knockout cells becoming commercially available, one can imagine that microfluidic models will help foster the development of more biologically sophisticated disease models for studying lymphatic vessel function. In addition, integration of nanobiosensors into living endothelial cells [19,157] will enable the detection of early-stage molecular signaling and biomechanical mechanisms that mediate lymphatic vessels, which have not yet been fully elucidated, especially at the sub-cellular length scales.

The majority of the studies using microfluidic culture models have so far focused on studying lymphangiogenesis and lymphatic vessel permeability in the context of the initial lymphatic capillaries, with limited insights on the lymphatic pumping mechanisms of the collecting lymphatics invested with perivascular SMCs. This aspect is a notable limitation in lieu of compelling results with isolated lymphatic vessels that demonstrate autoregulatory mechanisms by which fluid is propelled against an adverse pressure gradient [69,158]. Therefore, one could argue that the full pattern of lymphatic vascular remodeling is still beyond the reach of existing microfluidic devices. This outcome is likely a consequence of the “bottom-up” or reductionist approaches that are commonly employed by microfluidic systems that prioritize controlled microenvironments at the expense of biological complexity. However, a possible pathway for microfluidic models to study vascular remodeling of the collecting lymphatic vessels effectively is to adapt a previously described “artery-on-a-chip” platform for probing the role of spatiotemporal heterogeneities in the regulation of small artery tone and function [159]. This platform was developed in response to the wire-myograph method commonly employed by physiologists on isolated lymphatic vessels [160], and allows for loading, precise placement, fixation as well as controlled perfusion and superfusion of a small artery segment. As a result, there may be an opportunity for adapting this microfluidic-based technique for studying structure and function responses of collecting lymphatic vessels in a platform that is scalable for high-throughput with potential for automation and standardization [159]. Moreover, it is known that lymphatic valves frequently form at the bifurcation points of collecting lymphatic vessels [70,161], which correspond to regions of flow separations and disturbances [162]. With the emergence of 3-D microfluidic models featuring branched vessel geometry [14,163,164], these systems may be capable of interrogating the mechanisms of lymphatic valve formation in vitro that is coordinated by local flow dynamics, cell-cell interactions, and ECM remodeling.

In addition, it was previously believed that the brain and central nervous system (CNS) lacked a functional lymphatic system that could aid in immunosurveillance and fluid drainage [165]. However, recent work has revealed that lymphatic vessels do in fact exist in the meninges of the brain [166-168]. These findings have prompted a reassessment of the role of the lymphatic vasculature in the CNS and its involvement in a variety of neuroimmunological disorders. For instance, Kipnis and colleagues showed that meningeal lymphatic vessels play an essential role in draining amyloid-β peptides from cerebrospinal fluids and into cervical lymph node to maintain brain homeostasis [169]. Accumulation of amyloid-β in the brain parenchyma is a hallmark of Alzheimer’s disease [170]. Therefore, dysfunctional lymphatics in the brain may contribute to Alzheimer’s and other neurodegenerative disorders. A number of microfluidic blood brain barrier (BBB) models have been developed [171-173]. Thus, with the recent discovery of lymphatic vasculature in the CNS, it is foreseeable that microscale technologies could provide capable and versatile lymphatic-brain analogue systems to serve as neurological disease models.

5. Conclusion

In this review, we, highlight that microfluidic systems are uniquely capable experimental models that are amenable to faithfully recapitulating physicochemical cues of lymphatic microenvironments for quantitatively studying lymphatic barrier function, lymphangiogenesis, and effects of lymphatics on interstitial mass transport. We anticipate that further advancements in microscale culture models will make significant contributions to our mechanistic understanding of lymphatic biology and function to help realize its promise as a therapeutic target for cardiovascular medicine, wound healing, tissue engineering, and oncology.

Abbreviations used

ASC

adipose-derived stromal cell

bFGF

basic fibroblast growth factor

BBB

blood-brain barrier

BEC

blood endothelial cell

CCBE1

calcium binding EGF domain-containing protein 1

CNS

central nervous system

Cx37

connexin 37

ECIS

Electrical Cell-Substrate Impedance Sensor

ECM

extracellular matrix

FOXC2

forkhead box protein C2

HGF

hepatocyte growth factor

IFP

interstitial fluid pressure

LEC

lymphatic endothelial cell

LPS

lipopolysaccharides

LSS

laminar shear stress

LYVE-1

lymphatic vessel endothelial hyaluronan receptor 1

MI

myocardial infarction

MMP

matrix metalloproteinase

NO

nitric oxide

OSS

oscillatory shear stress

PDGF-B

platelet derived growth factor B

PDMS

poly(dimethylsiloxane)

PROX-1

prospero-related homeodomain transcription factor

RTK

receptor tyrosine kinase

S1P

spingosine-1-phosphate

SMCs

smooth muscle cells

T-MOC

tumor-microenvironment-on-a-chip

TIE

tyrosine kinase with immunoglobulin-like and EGF-like domains 1

TME

tumor microenvironment

VEGF

vascular endothelial growth factor

VEGFR

vascular endothelial growth factor receptor