Extracellular Matrix Dynamics in Tubulogenesis

Abstract

Biological tubes form in a variety of shapes and sizes. Tubular topology of cells and tissues is a widely recognizable histological feature of multicellular life. Fluid secretion, storage, transport, absorption, exchange, and elimination—processes central to metazoans—hinge on the exquisite tubular architectures of cells, tissues, and organs. In general, the apparent structural and functional complexity of tubular tissues and organs parallels the architectural and biophysical properties of their constitution, i.e., cells and the extracellular matrix (ECM). Together, cellular and ECM dynamics determine the developmental trajectory, topological characteristics, and functional efficacy of biological tubes. In this review of tubulogenesis, we highlight the multifarious roles of ECM dynamics—the less recognized and poorly understood morphogenetic counterpart of cellular dynamics. The ECM is a dynamic, tripartite composite spanning the luminal, abluminal, and interstitial space within the tubulogenic realm. The critical role of ECM dynamics in the determination of shape, size, and function of tubes is evinced by developmental studies across multiple levels—from morphological through molecular—in model tubular organs.

Introduction

Tubulogenesis—the assembly of cells, tissues, and organs into tubes—is a hallmark of metazoan development. The physiological framework of both diploblastic (with the exception of Trichoplax adhaerens) and triploblastic animals critically depends on the tubular organization of cells and tissues [1]. Tubes, therefore, are integral to metazoan organ systems; and the intricacies of their assembly are realized through the physical, chemical, and biological processes that direct organogenesis. Animals with increased overall size evolved organ systems with tubes as their anatomical facet [2]. Processes essential to sustain multicellular life, e.g., digestion, respiration, circulation, excretion, and reproduction, depend on the tubular architecture of organs. Therefore, developmental programs that direct organogenesis have evolved regulatory mechanisms governing the two factors, which impact tubular organ assembly, viz., the behaviors of cells and the attributes of the extracellular milieu.

Model organ systems from nematodes, arthropods, and chordates are the frequent focus of developmental analyses aimed at investigating the molecular and morphological processes underlying tubulogenesis [3–5]. By employing diverse methodologies across multiple model systems of organ development, tubulogenesis research draws on their collective advantages and strengths to not only reveal the morphogenetic mechanisms that are unique to each model system, but also those that are shared across phyla. Despite the diversity in their final shapes and sizes, biological tubes in these systems feature a fundamental characteristic—namely, of having luminal and abluminal surfaces. These two surfaces, which define tube topology, are consequent to apicobasal cell polarity. Therefore, strategies for tubulogenesis in the three major tube types—multicellular, unicellular, and subcellular (Figure 1A)—have been categorized, broadly, into those assembling tubes via cells with predetermined apicobasal polarity, and those assembling tubes via cells that acquire apicobasal polarity de novo [6]. Both regulated morphogenetic processes (e.g., cell division, cell rearrangement, cell shape changes) and molecular mechanisms (e.g., membrane dynamics, cell junction remodeling, cytoskeletal network polarization, growth/guidance factor signaling) transform individual cells and cell collectives into functional tubes during development. Absent in most considerations of tubular organ morphogenesis, however, is the role of extracellular matrix (ECM) dynamics.

Figure 1: Common tubular forms in the metazoan; and the tripartite extracellular matrix in the tubulogenic realm.

(A) Cross-sectional topologies of common metazoan tube types, viz., multicellular, unicellular, and subcellular tubes. The en face plane of multicellular and unicellular tubes highlight intercellular and autocellular adherens junctions (orange), respectively. Subcellular tubes occur as cytoplasmic extensions of unicellular tubes, and they lack adherens junctions; thus, they are also referred to as seamless tubes.

(B) A generic monolayered epithelial tube depicting the tripartite ECM—luminal ECM (red), abluminal ECM (sepia), and the interstitial ECM (khaki).

ECM dynamics is an indispensable aspect of animal development [7–9]. Observations from the perspective of tubulogenesis reveal the specialized and quintessential roles of ECM in organ morphogenesis. ECM is an integral component of the cellular basal lamina [10, 11]. This basal ECM, therefore, marks the abluminal surface of tubular organ systems. ECM is present, furthermore, in the apical (luminal) compartment of tubular organs as apical ECM. With its presence, therefore, in the architecturally and mechanically critical regions of biological tubes, ECM is a singular candidate for interactions—chemical and mechanical—with both the basal and apical membrane compartments of cells. Thus, the ECM is ideally situated to significantly impact tubulogenesis. Here, we highlight the tubulogenic trajectory of select representative organs from the perspective of ECM dynamics—an underappreciated aspect of tubular organ morphogenesis—in widely studied model organisms.

Tubulogenic ECM—a tripartite composite

The topologically unspecific, umbrella-term “ECM,” although sufficient for many treatments of the topic, is inadequate in capturing its dynamic roles in tubulogenesis. The apicobasal polarity of epithelial tubes imposes specific location-dependent biological characteristics to the ECM; therefore, ECM in the tubulogenic realm may be classified into three kinds in order to gain a broad perspective of its impact: (1) luminal (apical) ECM—the component in association with, or in the vicinity of the apical cell membrane compartment; (2) abluminal (basal) ECM—the component in association with, or in the vicinity of the basal cell membrane compartment (note that, in the tubulogenic realm, this would also encompass the basal lamina, or the basement membrane ECM); and (3) interstitial (stromal) ECM—the component that spans the interstitial space between tissues. Hence, viewed from this perspective, the ECM reveals itself as a dynamic, tripartite composite that co-determines, with cells, both the architectural and functional outcomes in the tubulogenic realm (Figure 1B).

Luminal ECM Dynamics

The lumen is a defining characteristic of all tubes, and it is also the principal site of physiological activity in tubular organ systems, e.g., secretion, absorption, storage, and transport of fluids. Studies from both invertebrate and vertebrate systems have provided compelling evidence for the dynamic role of luminal ECM in shaping biological tubes.

In Drosophila, which is a frequently used organism for tubulogenesis studies, its respiratory tube network—the trachea (Figure 2A)—has been studied extensively as a premier model organ for branching morphogenesis [4]. Studies of the tracheal luminal material—its luminal matrix—have, hitherto, provided strong evidence for the requirement of luminal ECM in tubulogenesis [12]. A critical step during tracheal tubulogenesis is the optimization of tube size, and it involves regulation of both the luminal length and diameter (Figure 2B). Tube size adjustments are essential to maintain the cross-sectional area of the parent tubules in agreement with the combined cross-sectional areas of its branches for efficient gas flow [13]. Lumen size control mechanisms are coordinated and are mediated by: (1) factors involved in polarized membrane growth, (2) molecules involved in luminal ECM synthesis and remodeling, and (3) protein complexes involved in the organization and function of septate junctions. The central component of the morphogenetic program that regulates tube size is the luminal ECM. Morphogenetic transformations that follow tracheal placode invagination and tubular network assembly, moreover, occur under eutelic conditions, i.e., constant cell number in the absence of cell division or death. Tube size optimization, therefore, is primarily determined by ECM dynamics, namely the transient chitinous luminal matrix [14–16].

Figure 2: Luminal ECM dynamics and tubulogenesis.

(A) The Drosophila trachea spans the entire three dimensional embryonic space. All ten metameric tube units from one side of the embryo are shown. Inset: A hemisegment from the ninth metameric unit composed of one unit of the dorsal trunk and the dorsal branch highlights the various tubular architectures—a multicellular tubule with intercellular junctions, a unicellular tubule with autocellular junctions, and a subcellular (seamless) tubule devoid of cell junctions.

(B) The chitin-rich luminal matrix regulates the size along both the axial and diametric (circumferential) dimensions of tracheal tubules.

(C) Drosophila embryonic salivary gland morphology is determined by the luminal ECM. The secreted luminal ECM contains a fibrillar component that keeps the apical membranes of opposing cells from contacting each other thus preventing luminal space constriction. The secreted metalloprotease ADAMTS-A cleaves the apical/luminal ECM, hence, untethering it from the apical membrane and forestalling impedance to collective cell migration.

(D) Drosophila hindgut is composed of an expansive small intestinal tubular wall and a narrow large intestinal tubular wall. The difference in tubular architecture manifested by small and large intestinal segments is a consequence of hydrostatic pressure differentials between the two tubular structures (walls)—which are, in turn, the result of concentration differences in the secreted luminal glycoprotein, Tenectin; High Tenectin concentration allows expansion of the small intestine, and its low concentration limits the free expansion of large intestinal wall.

Chitin is a polymer of beta-1,4-N-Acetylglucosamine (GlcNAc), and is produced by tracheal transmembrane chitin synthases that convert cytoplasmic UDP-GlcNAc to long polysaccharide chains that are extruded into the luminal space [12]. Thus, chitin filaments are the major component of luminal ECM in tracheal tubules. Tracheal tubes lacking the function of genes involved in chitin biosynthesis, e.g., krotzkopf verkehrt (kkv) and mummy (mmy), show widespread luminal abnormalities [17–20]. Loss-of-function of genes involved in chitin deacetylation, e.g., vermiform (verm) and serpentine (serp), results in over-elongated tubes [21, 22]. Meanwhile, both secreted and luminal membrane proteins involved in luminal chitin matrix organization, e.g., Gasp, Obstructor-A (Obst-A), Knickkopf (Knk), and Retroactive (Rtv), determine lumen diameter [23, 24]. Thus, the chemical and mechanical processes associated with chitin synthesis, secretion, its interaction with chitin-binding proteins, its structural modifications, and its organization into fibrils—processes grouped by the overarching label of luminal matrix synthesis, secretion, and remodeling—determine the lumen size and consequently, tubule morphology.

But how does the luminal chitin matrix affect tubule size? One hypothesis is that the regulated luminal accumulation of the chitin matrix acts like a mandril to control tube size by virtue of its form and rigidity [21]. It is thus conceivable that by uniformly lining the luminal surface along the tubules, the chitin matrix propagates long-range mechanical forces across the epithelium and maintains tubule integrity and size. The interaction of luminal chitin matrix to the epithelial (apical) membrane is mediated by zona pellucida (ZP) domain-containing proteins like Piopio (Pio) and Dumpy (Dpy), which not only interact with each other in the lumen to organize the luminal matrix but also anchor the matrix to the apical membrane [25–27]. The localization and secretion of these luminal matrix proteins is determined by various factors. In particular, Dpy-mediated epithelial cell-luminal matrix interactions are dependent on its proper localization/anchorage at the luminal membrane, which is determined by O-GlcNAcylation by EGF-domain O-GlcNAc transferase, EOGT [28]. The secretion of luminal matrix components, e.g., Pio and Gasp, moreover, depends on PI(4,5)P2 and Rho-1 GTPase-mediated Diaphanous (Formin actin-nucleating factor) localization at the apical membrane. Hence, Dia nucleates actin filaments at the apical membrane to support polarized transport of vesicles containing matrix components for luminal secretion [29, 30].

The most striking abnormalities in tube morphogenesis due to the loss-of-function of ZP-domain containing proteins Pio and Dpy occur in the unicellular tubes of embryonic trachea that deform into cysts [26]. The unicellular tubes are normally assembled by intercalation of cells within a multicellular branch outgrowth. To intercalate within the branch outgrowth, cells glide past each other while simultaneously wrapping themselves around the luminal matrix in a morphogenetic transformation that gradually replaces their intercellular junctions with autocellular junctions. In the pio and dpy mutants, however, the dynamic epithelial interaction with the luminal matrix is compromised, and therefore, the multicellular branch outgrowth deforms into a cyst instead of transforming into a unicellular tube. These observations underscore the importance of luminal matrix even in the largely cell-driven morphogenetic transformations during tubulogenesis.

Determination of tube morphogenesis by luminal ECM dynamics also occurs in unbranched tubes. The Drosophila embryonic salivary glands are a pair of jalapeno-shaped, mono-layered epithelial tubes that contain a luminal fibrillar matrix [31] (Figure 2C). Defects in their luminal matrix content result in gross abnormalities of tube shape and size. Loss-of-function of SG1 and SG2, two subunits of the ER-resident prolyl 4-hydroxylase that functions in hydroxylation of prolines in secreted and transmembrane proteins, results in numerous constrictions and dilations along the length of the lumen. This indicates the requirement for a normal fibrillary luminal matrix content to maintain a uniform tube lumen shape [32]. More recently, a critical role for secreted matrix metalloproteases in salivary gland lumen shape determination and in tube elongation was demonstrated with AdamTS-A (a disintegrin and metalloprotease with thrombospondin motifs). Morphological analysis of AdamTS-A mutant cells indicated luminal surface irregularities consequent to failed apical membrane detachment from the luminal matrix, therefore, resulting in abnormal tube shape [33].

Morphogenesis of the candy cane-shaped Drosophila hindgut tube is yet another example of the dynamic roles of luminal matrix. The hind gut tube incorporates both small and large intestinal regions with the tip of the candy cane handle corresponding to the small intestine and the shaft corresponding to the large intestine (Figure 2D). Hind gut tube morphogenesis, like that of the trachea and the salivary gland, is eutelic; and therefore, tube growth occurs via cell growth, shape changes, and rearrangements. The non-chitinous luminal matrix plays a major role in hind gut tube size determination. Evidence for luminal ECM dynamics-dependent hind gut tube morphogenesis was shown by the analysis of the large glycoprotein Tenectin (Tnc), which forms a dense striated luminal matrix and directs tube diameter expansion [34]. The manner in which Tnc drives dose-dependent tube size changes, and the possibility that it can form a gel-like matrix via hydration suggests that it generates luminal physical force in the form of hydrostatic pressure at its site of accumulation. Thus, Tnc exemplifies the capacity for non-chitinous luminal ECM to regulate tube size, a scenario pertinent to tube morphogenesis by vertebrate cells, which do not secrete chitin.

In addition to the above-mentioned roles in which the mechanical and chemical signaling roles of the luminal ECM were paramount for lumen size and shape regulation, studies have demonstrated the role for apical membrane, or luminal components, in the generation of the lumen itself. Temporal uncoupling of polarity acquisition and lumen formation by ectodermally derived epithelial cells occurs in the Drosophila compound eye as demonstrated by the role of a secreted proteoglycan Eyes shut (Eys) in lumen formation [35, 36]. Each ommatidium of the fly contains a luminal space, the interrhabdomeral space, which separates the rhabdomeres, i.e., the light receptor-housing compartment, of the photoreceptor cells. Secretion of glycosylated Eys is essential for the creation of interrhabdomeral space (Figure 3A), thus, highlighting the role of a luminal ECM component in the formation of the lumen itself [37].

Figure 3: Luminal ECM dynamics and tubulogenesis (continued...).

(A) The ommatidium of the Drosophila compound eye consists of a lumen called the interrhabdomeric space. Its luminal expansion, and the consequent rhabdomeric separation motion depend on Eyes shut, a proteoglycan, which is a component of the luminal extracellular matrix, secreted by the surrounding epithelial cells.

(B) Heart lumen expansion is facilitated by multiplexin, the Drosophila ortholog of mammalian collagen-XV/XVIII, by its synergistic action with Slit-Robo signaling at the cardioblast apical membrane.

(C) Left: The unicellular tubule components of the C. elegans excretory system: The H-shaped canal cell; the duct cell; and the pore cells. Right: Pore cell swapping, anchored/mediated by the luminal ECM (red), occurs during excretory organ development. To highlight the tubular cell connectivity in the cartoon, the canal cell arms have been truncated, and the excretory gland cell connection at the canal-duct junction is not shown.

ECM components mediate lumen generation and expansion in non-ectodermal derivatives as well. Drosophila heart is a case in point (Figure 3B). Lumen formation, and its subsequent expansion in the transfer-pipet-shaped Drosophila dorsal vessel—the linear myoendothelial tube compartmentalizing both the posterior heart and the anterior aorta—is dependent on membrane repulsion signals transduced by Slit-Robo signaling. It is regulated, furthermore, by the luminal membrane protein dystroglycan [38, 39]. Integrins and the transmembrane proteoglycan syndecan—both localized to the apical membrane of crescent-shaped cardioblasts—also play a role [40, 41]. Multiple ECM components are observed in the embryonic heart lumen including Multiplexin (Mp), the Drosophila orthologue of mammalian Collagen-XV/XVIII. Luminal Mp, expressed at the stage of heart tube closure, forms a complex with Slit and affects its distribution and stability. Mp-mediated Slit/Robo activity at the luminal membrane lowers the membrane-proximal F-actin levels to promote heart tube expansion [42]. Intriguingly, tubulogenesis by endothelial deadhesion in the mouse aorta has been shown to also depend on luminal components. The CD34 sialomucin podocalyxin/gp135 (PODXL) mediates lumen formation in the aortic endothelium by an electrostatic cell repulsion mechanism that is dependent on its negatively charged extracellular sialic acid residues [43].

Yet another illuminating example highlighting the role of luminal matrix in tube assembly and morphogenetic transformations (in this case, of unicellular tubes) could be drawn from the C. elegans excretory system (Figure 3C). The C. elegans excretory system is composed of three unicellular tubes: the catamaran-shaped canal cell spans the length of the animal, and is connected with the duct cell; which in turn, is connected with the pore cell that opens to the exterior [44]. Two more cell types, the syncytial (binucleate) gland cell and a pair of neurosecretory cells, complement the tricellular tubular ensemble to furnish the functional organ. Distinctive lumen-generating mechanisms are utilized by the three unicellular tube cells [3]. Lumen formation in the canal cell occurs by cell hollowing—a mechanism that utilizes polarized vesicle trafficking to assemble an intracellular apical membrane compartment. Meanwhile, lumen formation in the duct cell occurs by cell wrapping followed by autofusion; whereas the pore cell becomes tubular by wrapping itself around. Of these, both the duct and pore cell tubulogenesis are dependent on luminal matrix dynamics. During embryogenesis, the doughnut-shaped duct cell undergoes a morphogenetic transformation to become an elongated tubular loop that connects proximally with the canal cell and distally with the pore cell. Duct cell elongation is dependent on the ZP-glycoprotein LET-653. In let-653 mutants, the duct cell lumen is fragmented and shows dilations, thus, underscoring the critical role of this transient precuticular luminal matrix component in tubulogenesis [45]. Other luminal matrix components such as LET-4 and EGG-6, which belong to the family of leucine-rich repeat only proteins, also function in epithelial-matrix remodeling interactions, thus, affecting tube morphogenesis [46]. Luminal matrix integrity and membrane dynamics during tubulogenesis are also dependent on LPR-1, a secreted lipocalin that binds lipophilic molecules [47]. The last major morphogenetic transformation of C. elegans excretory system tubulogenesis is pore cell swapping. This occurs during the larval stage to replace the departing (delaminating) G1 pore cell with the G2 epidermal cell, and is facilitated by the cuticular luminal ECM; the latter anchors the wrapping G2 cell to form the replacement pore [48].

Collectively, evidence from diverse model systems of tubulogenesis suggest a critical role for luminal matrix dynamics in tubulogenesis. The dynamic attributes of luminal matrix components derive from the biochemical processes that govern their synthesis, assembly, degradation, reassembly, and chemical modification. The biophysical aspects of luminal ECM dynamics are also evident in some of the well-studied models of tubulogenesis. For example, the primary mode of tubulogenic action is mechanical in the Drosophila trachea; the material properties of luminal matrix impose mechanical constraints, which counteract the forces generated by cell cortical cytoskeleton-mediated luminal membrane expansion [49]. Thus, a balance is struck between the mechanical forces generated by both cells and the ECM to regulate lumen shape and size. It is also possible that the luminal matrix acts as a dynamic scaffold that enables long-range coordination of cellular behavior across the entire tubular network [15]. Thus, based on these observations, which demonstrate its salient composition and dynamic behavior across various model systems, the luminal matrix presumably plays a singular role in mediating tube morphogenesis.

Abluminal ECM Dynamics

The abluminal ECM of epithelial tubes is specialized in the form of a basal lamina or basement membrane, which engages in receptor mediated cell signaling and remodeling processes. It is also critical for tubular organ branching, elongation, migration, and positioning. The salient material and biological properties of abluminal ECM components, e.g., collagen IV, laminins, heparan sulfate proteoglycans (perlecan and syndecan), nidogen, and entactin, direct several aspects of tubulogenesis. During tube morphogenesis, the impact of mechanical signaling from the luminal ECM is complemented by the myriad biochemical signaling processes mediated by the abluminal ECM. The effects of abluminal/basal ECM on the morphogenesis of epithelia—both tubular and non-tubular—have been studied extensively [50]. Therefore, we highlight only a few select examples to illustrate the crucial roles of abluminal ECM in tubulogenesis.

The critical role played by abluminal ECM dynamics in shaping unicellular tubes is demonstrated by the canal cell (Figure 3C), which is the major component of the C. elegans excretory system, spanning the entire length of the adult animal. In becoming the largest cell of C. elegans, its elongating tubular processes—canals—extend longitudinally along the lateral epidermal ridge nestled between the hypodermis and a basement membrane [51]. Whether the basement membrane is synthesized by the canal cell or the hypodermis is unknown. The tube elongation program, however, is regulated by abluminal ECM components—namely, collagen, laminin, and matrix metalloproteinases (MMPs) which act at the growing canal tip [3].

A crucial process mediated by abluminal ECM dynamics is tubular cell/organ migration. Integrin-mediated tubulogenic processes, which depend on the ability of cells to migrate, fail in Drosophila laminin βchain (LanB1) mutant embryos [52]. Tube elongation and migration defects are observed in association with the loss of abluminal ECM in the LanB1 mutant Drosophila embryonic salivary gland (Figure 2C) and tracheal tubes (Figure 2A). Early steps of tubulogenesis—invagination and tube elongation—are unaffected by αPS1 (Mew) integrin loss-of-function. Tube positioning, which is the late step of salivary gland tubulogenesis requiring collective cell migration, however, is abnormal in Mew mutants [53]. In contrast to the normal lumen dimensions of the salivary gland tube in the Mew loss-of-function mutants, integrin-talin adhesion complexes are indispensable for proper luminal dimensions and branching morphogenesis of the tracheal terminal cells that form subcellular tubes [54]. Interestingly, the cells of one particular tracheal branch, the visceral branch, also express Mew while migrating over a substrate—the cells of visceral mesoderm—that express αPS2 (if) integrin [55]. Collectively, these results suggest that for collective cell motion, the degree of dependency on abluminal ECM dynamics is tissue-specific, and in certain cases (as with the tracheal visceral branch tubules), it is branch-specific and intricately tied to the cell specialization program.

Even within the same tubular organ, moreover, abluminal ECM dynamics could elicit diverse morphogenetic responses as seen with the renal tubules of Drosophila. Drosophila renal (Malpighian) tubules serve as a vivid example for the distinct roles of abluminal ECM in tubulogenesis [56] (Figure 4A). Two morphogenetic processes underlying Malpighian tubule elongation in the embryo—cell intercalation and dynamic anchoring—present highly specific, yet distinct requirements for abluminal ECM dynamics. The Malpighian tubules (two anterior and two posterior) stem from the embryonic hindgut. The distal anchoring of all tubules is with the hindgut, thus maintaining their continuity with the ureter. The tip cell-mediated proximal anchoring, however, varies—anterior tubules anchor to the alary muscles and posterior tubules anchor to the hindgut visceral nerve, respectively. The precise positioning of the tubules via anchoring is essential because insects have an open circulatory system from which to clear excretory material; hence, mispositioning of the tubules could potentially jeopardize excretory function.

Figure 4: Abluminal ECM dynamics and tubulogenesis.

(A) Top: Drosophila renal (Malpighian) tubules are anchored distally to the hindgut, and proximally to the A3/A4 segment-alary muscle (anterior tubule) and the hindgut visceral nerve (posterior tubule); only two out of four tubules are shown for clarity. Bottom: Tube elongation by cell intercalation occurs over the hemocyte-deposited abluminal ECM. Inset: Tube pathfinding and positioning by the tip cell is facilitated by selective denudement of its abluminal ECM. Note the absence of abluminal ECM around the tip cell basal membrane.

(B) Xenopus intestine metamorphosis depends on abluminal ECM dynamics (remodeling). The replacement of the larval epithelial layer by the stem cell-derived nascent adult cells is determined by thyroid hormone (T3) and MMP-dependent thickening of abluminal ECM.

(C) Adult Drosophila midgut homeostatic epithelial remodeling occurs via a basal-to-apical polarity acquisition by nascent enteric epithelial cells—a process that relies on continuous cellular contact with the abluminal ECM. The new epithelial cells traverse towards the lumen to replace the damaged enteric epithelial cells while acquiring basement membrane adhesion-dependent apicobasal polarity.

Malpighian tubule elongation occurs by cell intercalation, and it transforms a short tubule with several circumferential cells (ca. 8-12) into a long tubule with just two circumferential cells. Cell intercalation is critical for Malpighian tubule elongation, and is achieved over the abluminal (basement membrane) ECM. The abluminal ECM, which acts as the substrate for tube elongation, is itself a product primarily of circulating hemocytes—a circulating macrophage-like cell type within the Drosophila hemolymph not structurally associated with the Malpighian tubules [57]. The abluminal ECM assembled over the tubule cells by the circulating hemocytes, hence, acts as a substrate upon which cell rearrangement (intercalation) ensues, promoting tube elongation.

In sharp contrast to the requirement for the presence of abluminal ECM in Malpighian tubule elongation, however, is the requirement for the absence of abluminal ECM in the tip cell-mediated proximal anchoring and organ positioning. The tip cell lineage, besides acting as the master regulator of renal tubule cell division [58], also has a major role in tube elongation and positioning—processes that have been best examined in the anterior tubules [59]. To enable tube elongation, the tip cell makes transient contacts with the alary muscles—using them as waypoints while progressing sequentially from the posterior to the anterior abdominal segment boundaries. Its final contact with the alary muscle at the A3/A4 segment boundary anchors and positions the renal tubule. To accommodate the morphogenetic demands, the tubule tip cell adopts a specialized morphology, with a small apical domain and a relatively large basolateral domain (Figure 4A). These specializations allow the tip cell to reconcile its selective requirement for the absence of abluminal ECM concordant with the requirement in the remaining cells for the presence of abluminal ECM (to promote intercalation). Since the presence of abluminal ECM at the stalk-like basolateral region of the tip cell might impede its exploratory activity—required to make and break sequential alary muscle contacts—there is only a low level of secretion of the abluminal ECM components around the tip cell. Any ECM component deposited over its surface, furthermore, is actively denuded by the tip cell via MMP1-mediated degradation and Rab-5 mediated transcytosis. Tip cell anchoring of the renal tubule is stabilized, moreover, by its adhesive properties that are dependent on dynamically assembled integrin complexes. Thus, the Drosophila embryonic renal tubule assembly is an exquisite morphogenetic waltz between the epithelia and the hemocytes, performed over the dynamic abluminal ECM.

The influence of abluminal dynamics during tubulogenesis extends beyond its role in the shaping of cells; it also ensures their survival in certain cases. The importance of abluminal ECM remodeling in cell survival and niche enrichment during tubular organ morphogenesis is exemplified by the anuran intestine—a derivative of the embryonic endoderm—which undergoes extensive morphological changes during development. The larval intestine of Xenopus laevis undergoes a metamorphosis-associated morphogenetic transformation that molds a nearly circular lumen into a characteristic tube with stellate luminal folding patterns reminiscent of the mammalian intestine with crypts and villi [60]. Key to this luminal shape switching is the thyroid hormone (T3)-induced, stromal MMP11-dependent, abluminal ECM remodeling [61]. The luminal shape transformation is marked by apoptosis of the larval intestinal epithelial cells that are replaced by stem cell-derived adult (secretory/absorptive) epithelial cells, and it is associated with abluminal ECM thickening (Figure 4B). Abluminal ECM remodeling is critical for the tube morphogenetic transformation because in vitro addition of laminin, collagen, and fibronectin inhibits thyroid hormone-induced intestinal epithelial apoptosis [62]. Moreover, other MMPs—MMP2, MMP9, and MMP14—are also upregulated during intestinal tube morphogenesis [63]; and the formation of the stem cell niche—the cornerstone of intestinal cell renewal—is hyaluronan-dependent [64]. Thus, abluminal ECM remodeling, functioning downstream of T3 signaling, determines cell survival during tube morphogenesis in the Xenopus intestinal epithelium with the roles of ECM relatively more direct for epithelial apoptosis than it is for the refurbishment of the adult stem cell niche.

That the dynamics of the abluminal ECM are critical for the maintenance of tubular architecture beyond early developmental stages is demonstrated by the digestive tract in yet another model tubular organ, again, from Drosophila—the midgut. The Drosophila midgut is the only endodermally-derived epithelium, and is topologically continuous with the ectodermal derivatives—the foregut and hindgut. During post-larval stages, the midgut undergoes autophagic histolysis, and in its place, the adult midgut is assembled via the proliferation of undifferentiated adult midgut progenitor (imaginal) cells [65]. The adult midgut tube is an epithelial monolayer that contains four cell types: intestinal stem cells, absorptive enterocytes, secretory enteroendocrine cells, and postmitotic, bipotent enteroblasts. Unlike the rest of the Drosophila gut epithelium, the midgut epithelium features a reversed arrangement of polarity determinants that is reminiscent of the vertebrate epithelia, i.e., the septate (occluding) junction is localized apical to the adherens junction (Figure 4C). This reversal is attributed to the lack of luminal cuticle—a mode of protection afforded by the foregut and hindgut epithelia but is an impediment to the primarily absorptive function of the midgut epithelium. Thus, the reversed polarity of midgut cells is crucial to maintain luminal integrity in the absence of a robust protective barrier. The polarity of these cells, is determined—beginning with the embryonic stages—by their integrin-mediated adhesions with the abluminal ECM [66]. Maintenance of this polarity by integrin-mediated adhesions continues through the adult stages. The adult midgut epithelial tube is under homeostatic renovation as the functionally worn enteric epithelial cells die away, and are replaced by the nascent midgut epithelial cells originating from the intestinal stem cell divisions. These divisions occur at the basal surface of the midgut epithelium, and therefore, the nascent cells squeeze their way up to gain a free apical surface and integrate with the preexisting epithelial cells. Throughout their positional ascendance towards the lumen, the new enterocytes maintain their basal adhesions to the abluminal ECM since the assembly of their unique (basal to apical) polarity system, which is critical for the maintenance of midgut tubular morphology, is dependent on these basal adhesions [67]. Adhesion to the abluminal ECM is required for the formation of septate junctions, and the septate junctions are required for the formation of the apical domain. Thus, the abluminal ECM serves as the indispensable substrate upon which cells of the midgut organize their polarity determinants to maintain luminal function. The abluminal ECM-dependent adhesions in the midgut are also critical, moreover, in the maintenance of the intestinal stem cell niche and proliferation—processes vital to gut tube homeostasis [68–70].

The abovementioned examples highlight the key roles of abluminal ECM in determining both the generic (migration, elongation) and salient (tissue anchoring, polarity determination) morphogenetic features of tubulogenesis in a diverse assortment of organs. The requirement for abluminal ECM dynamics in tubulogenesis, highlighted by these cases, is indeed superposed to the myriad requirements imposed by epithelia on the basal ECM for the former’s stability as well as mobility and maintenance in the tubulogenic realm.

Interstitial ECM Dynamics

Of the three types of ECM in the tubulogenic realm, the interstitial ECM casts the most widespread and diverse effects. It is host not only to the constituent matrix molecules, but also to a constellation of cell types and signaling factors. The interstitial ECM, hence, acts as a flush substrate upon which the mesoscale—tens through hundreds of microns—morphodynamic processes that direct tubulogenesis are revealed. Accordingly, we consider three aspects of interstitial ECM dynamics among an eclectic array of morphogenetic processes that highlight its contributions in shaping tubular tissues and organs.

(i). Signaling within the interstitial ECM

Several branched tubular networks form in close association with the mesenchyme—an interstitial ECM-rich tissue composite. Tubes originating from every germ layer are shaped by the epithelial-mesenchymal interactions characteristic of branching morphogenesis: ectoderm (e.g., mammary gland), endoderm (e.g., mouse submandibular {salivary} gland and lung), and mesoderm (e.g., kidney). Despite their disparate germ layer origins, these tubular organs share a common morphogenetic trajectory along which incipient tubules are transformed into intricately woven tubular networks within an interstitial (stromal) ECM niche. The ECM niche facilitates receptor tyrosine kinase-mediated growth factor (RTK) signaling—the common denominator of branching morphogenesis in these tubular organs. Thus, the niche acts as a warehouse within which sequestration, storage, and presentation of growth factors facilitate cell signaling [71, 72]. In certain cases, select ECM components of the stroma directly signal the branching tubular parenchyma (cell-matrix signaling) that shapes network topology as demonstrated by the mammalian salivary gland assembly.

Ex vivo studies of the mouse salivary gland tubulogenesis under conditions of component isolation have revealed that the mesenchyme is requisite for branching morphogenesis. Importantly, the sufficiency of mesenchyme-derived growth factors for rescuing branching morphogenesis of isolated submandibular gland epithelium, cultured in 3D Matrigel or laminin ECM, demonstrates the critical roles of interstitial (and abluminal) ECM dynamics in salivary gland formation [73, 74]. Clefting is the hallmark of salivary gland branching morphogenesis. In contrast to other systems, e.g., mammary gland (see below), in which epithelial buds branch via bifurcation, trifurcation or lateral/side branching within the interstitial matrix, salivary gland end-bud clefting is characterized by multiple micro-indentations, i.e., preclefts, out of which only a few stabilize and enlarge to form mature clefts. Stabilization of clefts, as well as their progression, is an ECM-dependent process driven primarily by the accumulation—within the preclefts—of collagens, laminins, and fibronectin along the abluminal ECM-interstitial ECM interface (Figure 5A) [75–77]. Notably, fibronectin accumulation is associated with a decrease in E-cadherin expression, and consequently, the emergence of cell-matrix contacts—in the wake of dwindling numbers of cell-cell contacts—that promote cleft stabilization via a regulatory program mediated by BTBD7, a BTB (POZ) domain-containing transcription factor [78, 79]. Thus, interstitial ECM dynamics play a critical role on par with cytoskeleton-driven cellular dynamics in promoting submandibular gland clefting—a major morphogenetic feature of salivary gland branching morphogenesis [80, 81].

Figure 5: Interstitial ECM dynamics (signaling) and tubulogenesis.

(A) Mouse submandibular gland form is sculpted by fibronectin-dependent regulation of epithelial morphology. Insets: Cleft formation is orchestrated by the local enrichment of fibronectin, which leads to the downregulation of E-cadherin. Low E-cadherin levels allow the labile epithelial cells to accommodate clefting.

(B) Mouse lung alveolarization occurs following the growth factor-dependent early branching morphogenesis mediated by the interstitial ECM. Inset: Sacculation, which marks the development of alveolar sacs in association with the pulmonary vasculature, is shaped largely by an elastin-rich interstitial matrix crest that forms an organized network at the alveolar ridges.

(C) Branching morphogenesis of the ureteric bud leads to mouse kidney assembly, and is driven by glial cell line-derived neurotrophic factor (GDNF) signaling; The interstitial extracellular matrix hosts remodeling factors and facilitates cell-cell and cell-matrix morphogenetic signaling.

(D) Mammary gland duct branching morphogenesis is mediated by the collagen-rich interstitial ECM—the major component of the fat pad mesenchyme.

While salivary gland tubulogenesis is in progress, two other organ-level tubular networks are also shaped by interstitial ECM dynamics via growth factor signaling in the mouse embryo—the lung and the kidney. Tubular network formation in both the lung and the kidney is driven by reciprocal signaling interactions between the respective epithelium and the corresponding mesenchyme, i.e., the interaction of respiratory endoderm with the pulmonary mesenchyme in lung tubulogenesis, and the interaction of the ureteric bud with the metanephric mesenchyme in kidney tubulogenesis [82, 83].

Lung branching morphogenesis is composed of an early phase (E12.5–16.5) characterized by stereotypical branching subroutines following lung bud outgrowth; and a late phase (E16.5, extending postnatally, to P3–5) characterized by the co-development of terminal alveolar sacs (sacculation) with the pulmonary vasculature. The molecular mechanisms mediating sacculation are unclear whereas the early stages of stereotypical lung bud outgrowth and branching are driven primarily by Fgf10 signaling via its receptor Fgfr2 with contributions also from retinoic acid (RA), Hedgehog (HH), Wnt, and BMP/Tgf- β signaling [84–90]. The pulmonary interstitial ECM network—woven by a mixture of various types of collagens, fibrillins (1&2), elastin, laminin, and fibronectin—acts as the dynamic medium shaping the molecular signaling cross-talk between pulmonary epithelial and mesenchymal cells with a preponderance of contribution from elastin-based networks during sacculation (Figure 5B) [91–99]. The pulmonary interstitial ECM, furthermore, undergoes structural changes mediated by the action of proteases, which also, in turn, impact lung branching morphogenesis [100–102].

Branching morphogenesis of the mouse kidney, meanwhile, occurs as the final stage in the succession of morphogenetic transformations—from the evolutionarily early pronephros to the late mesonephros and metanephros—within the intermediate mesoderm. The elaboration of the ureteric bud, an initial epithelial outgrowth of the mesonephric duct, by reciprocal interactions with the metanephric cap mesenchyme is the hallmark of kidney branching morphogenesis, and it transforms the ureteric bud into the entire ureteric collecting duct system during a 5-day period (E11.5-E16.5) [83]. Branching morphogenesis of the ureteric bud occurs in concomitance with nephrogenesis, the latter also encompassing glomerulogenesis. Ureteric bud branching is driven by a conglomerate of cell-cell and cell-matrix signaling interactions, involving several signals (including Fgf10), all centered, however, around glial cell line-derived neurotrophic factor (GDNF) signaling mediated via its receptor tyrosine kinase RET and a coreceptor (GFRα1) (Figure 5C) [103–106]. ECM is integral to the branching morphogenesis of the kidney as evidenced by a plethora of data for the roles of basement membrane ECM components, i.e., abluminal ECM [107], and glycosaminoglycans (GAGs) [108]. The interstitial ECM, moreover, plays a dominant role as well, in a manner similar to its role in shaping lung branching morphogenesis, by mediating cell-cell signaling and cell-matrix signaling cross-talk by hosting and facilitating the action of MMPs and tissue inhibitors of metalloproteinases (TIMPs). For example, MMP9 protects mesenchymal cells from apoptosis while stimulating ureteric bud branching morphogenesis by the likely release of stem cell factor [109]. A proposed model, based on in vitro studies, posits a requirement for MMPs and TIMPs that is dependent on stage-specific ureteric branching morphogenesis [110].

Despite the striking parallels in overall tubule patterning and morphogenetic signaling mechanisms that are observed between lung and ureteric bud branching morphogenesis, there remain subtle, yet crucial, contrasts in pattering parameters between the two: e.g., increased branching asymmetries in the kidney compared to the lung, and higher epithelial tip numbers in the lung compared to the kidney. These crucial differences in branch patterning are not inherent properties of the branching epithelium, i.e., the lung bud or the ureteric bud; but they are instigated by the growing tubules in response to the signals provided by the interstitial ECM-laden mesenchyme. Hence, the ureteric bud could be reprogrammed to undergo lung-type branching morphogenesis in a heterotypic tissue recombinant culture: i.e., by culturing the ureteric bud within the pulmonary mesenchyme [111]. Alternatively, the same interstitial matrix constituent may mediate different branch patterning outcomes depending on its specific role in the developing lung or the kidney. For example, loss of Adamts18, a secreted metalloprotease, results in markedly reduced lung branching morphogenesis, whereas, the kidney presents with prolific branching and double ureters [102]. Thus, interstitial/stromal ECM dynamics play a defining role in the tubulogenesis of the lung and the kidney by directly impacting their branching patterns.

Yet another striking example of the critical role of interstitial ECM morphodynamics in tubulogenesis is the ovarian hormone-triggered elaboration of the rudimentary mammary gland ductal architecture—which is laid down during embryogenesis—into an arborized tubular network during puberty. Branching morphogenesis, marked by iterative duct terminal end-bud elongation, bifurcation, and lateral branching, occurs within the mammary fat pad-associated mesenchyme [112]. In striking contrast to the stereotypical branching topology of other tubular organs, e.g., the lung, mammary gland tube morphology is highly variable. The individual variability in mammary duct form underscores the importance of stochastic dynamics within microenvironmental factors—of which the interstitial ECM is a key component—over predetermined genetic control factors in directing branching morphogenesis [112, 113].

Key to the elaboration of mouse mammary gland ductal architecture is the communication between epithelial and stromal cells that directs a series of coordinated cell divisions, rearrangements, and shape changes. The major mediators of this communication are growth factors—hepatocyte growth factor, epidermal growth factor, insulin-like growth factor, and fibroblast growth factor [114–119]—and MMP signaling [120–122], both of which occur in close spatial and temporal concordance with interstitial ECM remodeling processes. The mammary gland-associated interstitial ECM is collagen-rich, and it hosts a wide variety of cell types including fibroblasts, adipocytes, endothelial cells, and innate immune cells [123]. Hence, the myriad roles of the interstitial ECM in growth factor signaling and MMP remodeling along with its effects in channeling the epithelial-stromal cell communications within this pullulating microenvironment, is crucial for the emergence of local dynamic patterning, i.e., stochastic branching forms (Figure 5D). Indeed, MMPs within the mammary stromal matrix perform non-proteolytic actions, in addition to their canonical proteolytic roles (e.g., MMP11 [124]. MMP3, and MMP 2 [121]) to promote tubulogenesis. An example is provided by the functions of MMP3, which binds WNT5B to increase mammary stem cell number and promote duct arborization [125].

Therefore, interstitial ECM dynamics, as the aforementioned examples illustrate, is an indispensable component of growth factor signaling-driven branching morphogenesis in multiple tubular organs. The role of interstitial ECM dynamics via its contribution to cell-cell and cell-matrix signaling, however, is only the most recognized among its roles that also include two other, less recognized and poorly understood phenomena associated with tubulogenesis: motion and cavitation.

(ii). Motion within the interstitial ECM

Convective motion of ECM is a poorly understood, yet important, contributor to tubulogenesis. The earliest experimental evidence suggesting the possibility of interstitial ECM motion came from the observation that non-motile retinal pigment epithelial cells were translocated by convective tissue-scale flow upon their introduction into neural crest migratory pathways in the avian embryo [126]. Contrary to the notion that ECM is a passive substrate for cellular dynamics, plausible physical scenarios—owing to the capacity of matrix molecules for spontaneous assembly, and formation of percolating networks with dynamic physical properties such as surface tension and viscosity—allow the ECM to exhibit in vitro matrix-driven translocation—a prerequisite for in vivo motion [127, 128]. Convective tissue movements, identified as temporally persistent interstitial ECM motion, are extracted from morphogenetic data by low-pass temporal filtering of the fluorescent-label-conjugated ECM displacements, obtained over time-scales of hours; this is in contrast to the active cell movements, which are defined as displacements relative to the local ECM scaffold, calculated as the difference between local cellular and global ECM movements [129]. The contributions of interstitial ECM motion to tubulogenesis are highlighted by the data obtained from two organogenetic processes in the avian embryo—cardiogenesis and primary vasculogenesis.

During avian cardiogenesis, the right and left primary heart fields, which are initially planar structures within the anterior lateral plate mesoderm, are transformed into a bilayered tubular heart (tube-within-a-tube) separated by the ECM, i.e., cardiac jelly [130, 131]. This tubular transformation, requiring the congregation of primordial cardiac cells at the midline, is also dependent on nearby ECM displacements to the midline (Figure 6A) [132]. Movements of both the fibronectin and fibrillin-2 cardiac cell-associated networks are predominantly driven by shortening of the endoderm, which also transports the endocardial and myocardial precursors to the midline for cardiac tube assembly [133, 134]. Thus, large-scale tissue movements transport the avian cardiac progenitor cells and presumptive cardiac jelly to the midline, thus setting the stage for local cell-autonomous myocyte progenitor motility, ‘vasculogenesis’ of the endocardium, and expansion of the cardiac jelly space to accommodate the assembly of primordial heart valves.

Figure 6: Interstitial ECM dynamics (motion & cavitation) and tubulogenesis.

(A) Avian heart tube is shaped by the convective tissue-scale motion of fibronectin and fibrillin-rich interstitial ECM, coordinated with the movement of cardiac precursor cells, within the primary heart fields towards the midline.

(B) Convective tissue-scale motion involving interstitial ECM and the endothelial cells, within the expanding splanchnopleural ECM, is critical for the establishment of the primary vasculature during avian embryogenesis. Insets: Tissue (interstitial ECM plus endothelial cell) motion convects endothelial cords during the initial stages of vasculogenesis (no lumen). Vascular drift, concomitant with the condensation of the primary polygonal network (lumen present), occurs during the final stages. Arrows indicate the direction of tissue motion, within the region of interest beneath the embryonic heart, to establish connectivity between the omphalomesenteric vein and the sinus venosus.

C) Interstitial fluid pressure-induced cavitation may be critical for the development of primary body cavities circumscribed by the interstitial ECM.

Primary vasculogenesis is the formation of endothelial tubes from naive mesoderm. Morphogenesis of a primary vascular network requires coordinated motion across various length scales (from μm to mm). In amniotes, the most fundamental vascular pattern is a planar network of interconnected polygons [135]. Five major motion patterns are implicated in primary vasculogenesis: (1) tissue deformations that convect primordial endothelial cells; (2) ‘vascular drift’, observed as a composite medial displacement of the entire vascular bed; (3) condensation of the primary polygonal network; (4) cell-autonomous motion of primordial cells along existing vascular cords; and (5) subcellular extensions/retractions across avascular zones that form/remove connections within the polygonal network [136, 137]. Interstitial ECM-associated, large-scale tissue motion is the driver of vascular fusions in which smaller vessels coalesce into larger vessels (Figure 6B). The composite tissue motion is also implicated in the repositioning of extraembryonic vasculature and enable its ‘docking’ with the intraembryonic vasculature [138]. Tissue-scale ECM motion also positions the dorsal aortae in the avian embryo [137].

Thus, interstitial ECM dynamics, in the form of convective motion of matrix molecular networks and tissue-scale deformations, are major contributors to the organ-level tubulogenic processes in amniote embryos.

(iii). Cavitation within the interstitial ECM

A vital but poorly understood role of interstitial ECM is in shaping primary body cavities, i.e., pseudocoel, which are histologically distinct from secondary body cavities, i.e., coelom [1]. The central lumen of the latter is lined by epithelium whereas the primary body cavity is lined by the interstitial ECM. Body cavities, owing to their primary functions—similar to cell-lined tubes—in fluid storage and transport, are apt for a discussion highlighting the role of interstitial ECM dynamics.

Primary body cavities perform critical functions in several taxa, e.g., the heart of mollusks and the circulatory system of phoronids, bryozoans, and brachiopods [139]. Primary body cavities, moreover, occur in association with coelomic compartments as mixocoels in several taxa, e.g., onychophorans and euarthropods [1]. The morphogenetic processes by which primary body cavities develop from cracks and fissures within the interstitial ECM (lamina fibroreticularis) is poorly understood. Pressure exerted by the accumulation of fluids within the interstitial space may play a dominant role (Figure 6C). Hence, a crucial role for the dynamic material properties of the ECM to accommodate physiological fluctuations in interstitial fluid pressure, and to allow the cavities to be carved out of the matrix-laden interstices could be postulated. Thus, primary body cavities are tubes indeed that are shaped by the coupled dynamics of interstitial ECM and the associated fluids. In the case of mixocoel formation, morphogenetic contributions from cell/tissue-level physical forces are also, presumably, involved.

Outlook

From the perspective of tubulogenesis, defining the ECM as a triad in terms of its luminal, abluminal, and interstitial dynamics reveals the subtleties of its role in shaping tissues and organs. The select examples presented in our discussion—for luminal ECM dynamics: Drosophila trachea, salivary gland, hindgut, eye, and heart, and the C. elegans excretory system; for abluminal ECM dynamics: Drosophila trachea, salivary gland, Malpighian tubule, and midgut, the C. elegans excretory system, and the Xenopus intestine; for interstitial ECM dynamics: mouse salivary gland, lung, kidney, and mammary gland, and the avian heart and vasculature, and the formation of primary body cavities in evolutionarily ancient taxa—highlight the overarching role of ECM dynamics in making and shaping tubular organs. Broadly gleaned, the general findings from this collage of model systems, i.e., tube size coordination by luminal ECM, tube migration and cell survival determination by abluminal ECM, and facilitation of tube branching morphogenesis by interstitial ECM, exemplify the active and versatile roles of ECM—on par with the contributions from cellular behavior—in tubulogenesis.

Whether the diverse ECMs in the tubulogenic realm perform distinct functions owing to their distinct compositions or composite material properties is not only an open question but also a fertile ground for future research pursuits in this area. A speculative synthesis could be formulated with the caveat that almost all known mechanical properties of ECM have been measured on fabricated biomaterials, e.g., reconstituted collagen hydrogels (or) cross-linked synthetic elastin, and furthermore, data on the physical properties of embryonic ECM are scarce. In the case of luminal ECM, a computational model postulating the ability of ECM to generate mechanical tension, and to balance the forces of tube elongation has been tested [16, 49]. In this scenario, the apical ECM acts like a viscoelastic material coupled to the apical membrane. Thus, the elasticity of apical ECM balances the lumen expansion force generated by apical membrane growth. The balancing interaction of apical ECM and tube luminal membrane, enabled by the mechanical coupling of membrane-anchored matrix proteins like Dpy, is a critical determinant of lumen stability, and consequently, tube size control. While the luminal ECM could be modeled as a mobile viscoelastic material, the abluminal ECM is construed as a provider of structural stability. The basement membrane organization has been reinterpreted to be bifunctional to meet the distinct requirements from the epithelium and stromal matrix emplaced on either of its faces [140]. The laminin polymer lattice provides the substrate for cell adhesion while the collagen IV network lattice is indispensable for maintaining tissue stability. In particular, during embryogenesis, the expansion and directional growth requirements of various tubular tissues and organs are accommodated by the differential compositions of abluminal matrix (Laminins, Collagen IV, Nidogens, Proteoglycans, Perlecan HSPGs, Agrin, Collagen XV/XVIII—Multiplexin, etc.). In certain cases, moreover, it is the absence or weakening of abluminal ECM that is a requirement for tube morphogenesis, e.g., Drosophila Malpighian tubule positioning. Beyond tissue stability, the individual abluminal ECM components serve as mechanisms for instructing tube polarity [141], architecture [142–144], and bilayered-adhesions with highly specialized functions (e.g., adhesion between abluminal ECMs of (1) lung alveolar epithelium and capillary endothelium; (2) kidney glomerular podocyte and capillary endothelium). Finally, in the case of interstitial or stromal ECM, its abundant (local) compositional complexity makes it challenging to draw common principles applicable to all of the tubulogenic realm although some general properties of the ECM composite may be postulated to mediate specialized biomechanical states conducive to tubulogenesis. These properties—fiber thickness, orientation/alignment, density, stiffness, porosity, texture, and viscosity—could collectively determine the material properties required for various roles of stromal matrix in tubulogenesis, viz., signaling & branching (Fibronectin-rich ECM for salivary acinar branching, Elastin-rich ECM for lung alveolar branching, and Collagen-rich ECM for mammary ductal branching), motion, and cavitation.

Although our choice of examples was circumscribed by the in vivo tubulogenic realms of select model organisms, the equally important insights offered by in vitro systems, e.g., cell, tissue, and organoid cultures, could not be overlooked [145–147]; these systems provide outstanding opportunities for unraveling otherwise intractable processes in the complex in vivo scenarios, e.g., material property-dependent ECM network dynamics in tubulogenesis. The increasing availability of matrisome-linked high throughput data for multiple tubulogenic systems, moreover, allows the construction of comprehensive systems-based models for testing the roles of ECM dynamics [148, 149]. Approaches based on mathematical and computational modeling have, furthermore, rendered the ECM dynamics toolkit in the tubulogenic realm more powerful, and hence, have empowered students of matrix biology to investigate long-standing problems, e.g., ECM motion and tissue-scale deformations [150–152]. Therefore, the integration of findings from cellular and molecular biological studies of tubulogenesis with those obtained from (1) in vitro (bio-engineered) systems, (2) genome-scale approaches, and (3) computational modeling, could be expected to provide a better foundation for our understanding of the roles of ECM dynamics in making and shaping biological tubes.

Highlights.

• Tubulogenesis is an essential facet of metazoan development

• Extracellular matrix dynamics is a critical morphogenetic component of tubulogenesis

• Matrix dynamics in the tubulogenic realm encompasses the luminal, abluminal, and interstitial spaces