Cell Junctions and the Extracellular Matrix

Introduction

In multicellular organisms, cells do not exist in isolation. They are organized into tissues — cooperative assemblies in which cells adhere to one another, communicate through direct contacts, and are embedded in a complex meshwork of secreted macromolecules called the extracellular matrix (ECM). The architecture of every organ, from the elastic walls of an artery to the transparent layers of the cornea, depends on the precise arrangement of cell-cell junctions, cell-matrix adhesions, and ECM components.

The molecular machinery that holds tissues together also transmits information. Adhesion receptors are not passive glue: they sense mechanical forces, relay signals to the cell interior, and influence whether a cell proliferates, differentiates, migrates, or dies. Disruption of cell adhesion and ECM signaling is a hallmark of cancer invasion and metastasis, while inherited defects in adhesion molecules or ECM proteins cause a wide spectrum of diseases — from blistering skin disorders to vascular fragility syndromes.

This lesson covers three interconnected topics: the junctions that link cells to each other (19.1), the extracellular matrix that fills the spaces between cells (19.2), and the integrin receptors that connect cells to their matrix and transduce mechanical signals (19.3). We also introduce bioinformatics resources for analyzing adhesion molecules, ECM composition, and integrin signaling networks.

19.1 — Cell-Cell Junctions

Animal cells connect to their neighbors through several types of junctions, each built from a distinct set of transmembrane adhesion proteins linked to intracellular scaffolds and cytoskeletal elements.

Cadherins Form a Diverse Family of Adhesion Molecules

Cadherins are a large superfamily of calcium-dependent transmembrane adhesion proteins. The human genome encodes more than 100 cadherins, but the best understood are the classical cadherins — including E-cadherin (epithelial), N-cadherin (neural and mesenchymal), P-cadherin (placental), and VE-cadherin (vascular endothelial). Classical cadherins have five extracellular cadherin repeats (EC1–EC5), each adopting a β-sandwich immunoglobulin-like fold. Ca²⁺ ions bind at the interfaces between repeats, rigidifying the extracellular domain into a curved, rod-like structure. Removal of calcium causes the extracellular domain to collapse, which is why cadherins are exquisitely calcium-dependent. Beyond the classical cadherins, the superfamily includes desmosomal cadherins (desmogleins and desmocollins), the large protocadherins (important in neuronal wiring), and atypical cadherins such as Fat and Dachsous that regulate tissue growth and planar cell polarity.

Cadherins Mediate Homophilic Adhesion

Cadherins exhibit homophilic binding — they preferentially bind the same type of cadherin on an adjacent cell. E-cadherin on one epithelial cell binds E-cadherin on its neighbor; N-cadherin binds N-cadherin. This selectivity was demonstrated in classic cell-sorting experiments: when cells expressing different cadherins are mixed, they self-segregate into clusters of like cells, recapitulating aspects of tissue formation during embryonic development. The adhesive interaction occurs through a “strand-swap” mechanism in the EC1 domain, where a conserved tryptophan residue on one cadherin inserts into a hydrophobic pocket on the partner cadherin. Homophilic binding is crucial for tissue organization: during gastrulation, cells switching from E-cadherin to N-cadherin undergo the epithelial-to-mesenchymal transition (EMT), enabling them to leave epithelial sheets and migrate as individual cells.

Cadherins Link to the Cytoskeleton Through Catenins

On their cytoplasmic side, classical cadherins connect to the actin cytoskeleton through a set of adaptor proteins called catenins. The cadherin cytoplasmic tail binds β-catenin (or its relative γ-catenin/plakoglobin), which in turn binds α-catenin. α-catenin links the complex to actin filaments either directly or through actin-binding proteins such as vinculin, α-actinin, and EPLIN. A third catenin, p120-catenin, binds the juxtamembrane region of the cadherin tail and stabilizes the cadherin at the cell surface by preventing its endocytosis. Notably, β-catenin has a dual role: beyond its structural function in adherens junctions, it is a key effector of the Wnt signaling pathway. When Wnt signals are absent, cytoplasmic β-catenin is phosphorylated by a destruction complex (APC, Axin, GSK3β) and targeted for proteasomal degradation. When Wnt ligands bind their receptors, the destruction complex is inactivated, β-catenin accumulates, and it enters the nucleus to activate target genes. This dual adhesion-signaling function means that changes in cell adhesion can directly affect gene expression and vice versa.

Adherens Junctions Respond to Forces Generated by the Actin Cytoskeleton

Adherens junctions are not static structures — they are mechanosensitive, actively responding to forces generated by the actin cytoskeleton. When actomyosin contraction pulls on the cadherin-catenin complex, α-catenin undergoes a conformational change that exposes a binding site for vinculin, recruiting additional actin filaments and reinforcing the junction. This force-dependent strengthening is an example of mechanotransduction at cell-cell contacts. In epithelial sheets, adherens junctions form a continuous adhesion belt (zonula adherens) encircling each cell just below the apical surface. The actin filaments associated with these junctions are linked into a contractile network that can change the shape of the entire epithelial sheet — a process fundamental to morphogenesis.

Desmosomes Give Epithelia Mechanical Strength

Desmosomes (maculae adherentes) are button-like junctions that provide tissues with exceptional mechanical resilience. They are especially abundant in tissues subject to intense mechanical stress, such as skin, cardiac muscle, and the uterine cervix. Desmosomes use desmosomal cadherins — desmogleins and desmocollins — as their transmembrane adhesion molecules. On the cytoplasmic side, these cadherins bind to plakoglobin and plakophilins, which connect through desmoplakin to the intermediate filament network (keratins in epithelia, desmin in cardiomyocytes). By anchoring to intermediate filaments rather than actin, desmosomes create a tissue-wide network of mechanical connections that distributes tensile forces across many cells, preventing individual cells from being torn apart. Autoimmune destruction of desmogleins causes pemphigus vulgaris, a severe blistering skin disease that demonstrates the critical role of desmosomes in epithelial integrity.

Tight Junctions Form a Seal Between Cells and a Fence Between Membrane Domains

Tight junctions (zonulae occludentes) form a continuous belt around epithelial and endothelial cells near the apical surface, serving two essential functions:

• Paracellular barrier (seal) — tight junctions seal the space between adjacent cells, controlling the passage of ions and small molecules through the paracellular pathway. The tightness of this seal varies by tissue: the blood-brain barrier is extremely tight, while kidney proximal tubule epithelia have leakier tight junctions to allow selective reabsorption
• Fence function — tight junctions prevent the lateral diffusion of membrane proteins and lipids between the apical and basolateral membrane domains, maintaining the polarized distribution of transporters, channels, and receptors that is essential for directed epithelial transport

Tight Junctions Contain Strands of Transmembrane Adhesion Proteins

The sealing function of tight junctions depends on strands of transmembrane proteins that form a meshwork at the contact between adjacent cells. The major components are:

• Claudins — a family of ~27 members in humans, each with four transmembrane domains. Claudins are the primary determinants of tight junction permeability; different claudin combinations create junctions with different ion selectivities and tightness. For example, claudin-2 forms a cation-selective pore, while claudin-4 creates a tight, charge-selective barrier
• Occludin — a four-transmembrane protein important for barrier regulation, though mice lacking occludin still form tight junctions (with subtle barrier defects)
• Junctional adhesion molecules (JAMs) — single-pass transmembrane proteins that contribute to junction assembly and leukocyte transmigration

Scaffold Proteins Organize Junctional Complexes

Scaffold proteins on the cytoplasmic face of tight junctions organize the adhesion molecules and link them to the actin cytoskeleton and intracellular signaling pathways. The ZO proteins (ZO-1, ZO-2, ZO-3), members of the MAGUK (membrane-associated guanylate kinase) family, are central organizers. They contain multiple protein-interaction domains (PDZ, SH3, guanylate kinase-like) that bind claudins, occludin, JAMs, actin filaments, and signaling molecules. Scaffold proteins at adherens junctions and desmosomes serve analogous organizing roles, creating modular signaling platforms at sites of cell-cell contact. The tight junction scaffold also recruits polarity proteins such as Par3, Par6, and aPKC, linking junction formation to the establishment of epithelial apical-basal polarity.

Gap Junctions Couple Cells Both Electrically and Metabolically

Gap junctions provide direct channels between the cytoplasms of adjacent cells, allowing ions, metabolites, second messengers (cAMP, IP₃, Ca²⁺), and other small molecules (up to ~1 kDa) to pass freely between coupled cells. This intercellular communication enables:

• Electrical coupling — in cardiac muscle, gap junctions allow action potentials to spread rapidly from cell to cell, coordinating the rhythmic contraction of the heart
• Metabolic cooperation — gap junctions allow cells to share metabolites and small signaling molecules, so that even cells without direct access to nutrients can be sustained by their neighbors
• Coordinated developmental signaling — morphogens and second messengers can spread through gap junctions to coordinate cell behavior during embryogenesis

A Gap-Junction Channel Is Made of Connexins

Each gap-junction channel is formed by the docking of two connexons (hemichannels), one contributed by each of the two coupled cells. Each connexon is a hexamer of connexin subunits. The human genome encodes 21 connexin genes, and different connexins form channels with different conductances, permeability properties, and gating responses. Connexons can be homomeric (six identical connexins) or heteromeric (a mix of connexin types), and gap-junction channels can be homotypic (identical connexons on both sides) or heterotypic (different connexons docking together). This combinatorial diversity allows fine-tuning of intercellular communication in different tissues.

Gap Junctions in an Epithelium Have Diverse Functions

In epithelia, gap junctions serve varied roles beyond simple metabolic coupling. In the lens of the eye, gap junctions composed of connexin46 and connexin50 are essential for maintaining lens transparency — mutations in these connexins cause congenital cataracts. In the cochlea, gap junctions between supporting cells recirculate K⁺ ions that enter hair cells during sound transduction; mutations in connexin26 (GJB2) are the most common cause of hereditary nonsyndromic deafness. In the liver, gap junctions help synchronize metabolic responses across hepatocyte populations. The regulation of gap-junction permeability — by pH, Ca²⁺, phosphorylation, and voltage — allows cells to control when and how much they communicate with their neighbors.

In Plants, Plasmodesmata Perform Many of the Same Functions as Gap Junctions

Plant cells, surrounded by rigid cell walls, use plasmodesmata rather than gap junctions for intercellular communication. Plasmodesmata are membrane-lined channels (~50–60 nm diameter) that traverse the cell wall, directly connecting the cytoplasms of adjacent plant cells. Unlike gap junctions, plasmodesmata contain a central strand of endoplasmic reticulum (the desmotubule) that threads through the channel. The functional pore — the cytoplasmic sleeve between the desmotubule and the plasma membrane — allows passage of ions, metabolites, and even some proteins and RNA molecules. Plasmodesmata can be gated by the deposition of callose (β-1,3-glucan), which narrows the channel aperture. Viruses exploit plasmodesmata to spread from cell to cell, using movement proteins that dilate the channels.

Bioinformatics: Cell Adhesion Bioinformatics

The diversity of adhesion molecules demands computational approaches for classification and analysis.

Cadherin and cell adhesion molecule (CAM) family classification uses sequence and structural databases to organize the superfamily. The Cadherin Resource and annotations in UniProt and InterPro catalog cadherin repeats, domain architectures, and subfamily assignments. Phylogenetic analysis of cadherin domains reveals evolutionary relationships among classical cadherins, protocadherins, and atypical cadherins across species.

Cell junction protein interaction databases such as STRING, BioGRID, and the Adhesome database catalog experimentally validated and predicted interactions among junction components. These databases enable construction of protein interaction networks at adherens junctions, desmosomes, and tight junctions, revealing how scaffold proteins integrate adhesion with signaling.

Adhesion molecule expression analysis in tissues leverages single-cell RNA-seq datasets (e.g., from the Human Cell Atlas) to map which adhesion molecules are expressed in each cell type and tissue. This information is critical for understanding tissue-specific junction composition — why claudin-2 predominates in the proximal tubule while claudin-4 dominates in the collecting duct, for example.

Mechanotransduction pathway modeling uses computational approaches to simulate how forces at cell-cell junctions propagate through the cadherin-catenin complex to activate intracellular signaling. Finite-element models of adherens junctions combined with molecular dynamics simulations of α-catenin unfolding under tension reveal how mechanical force is converted into biochemical signal.

Let us compare the protein properties of two major ECM structural proteins — collagen and elastin — to see how their amino acid compositions reflect their distinct mechanical roles.

let collagen = "GPPGPAGPKGERGDAGAPGPQGPA"
let elastin = "VPGVGVPGVGVPGVGVPGVG"
print("Collagen Gly-X-Y repeat properties:")
print(Struct.protein_props(collagen))
print("Elastin VPGVG repeat properties:")
print(Struct.protein_props(elastin))

Notice the high glycine and proline content characteristic of collagen — glycine occupies every third position in the Gly-X-Y repeat, and proline (often hydroxyproline in the mature protein) is enriched in the X and Y positions. Elastin, by contrast, is rich in hydrophobic residues (valine, glycine, proline) that give it rubber-like elasticity.

19.2 — The Extracellular Matrix of Animals

The Extracellular Matrix Is Made and Oriented by the Cells Within It

The extracellular matrix (ECM) is a complex and dynamic network of macromolecules that fills the spaces between cells in animal tissues. Far from being an inert scaffold, the ECM actively influences cell behavior — regulating proliferation, differentiation, survival, and migration. Cells both produce the ECM and respond to it, creating a reciprocal relationship: fibroblasts secrete collagen and fibronectin, then remodel these molecules by exerting mechanical forces, and the resulting matrix stiffness in turn affects fibroblast gene expression and behavior. The composition and organization of the ECM vary enormously between tissues: bone matrix is calcified and rigid, cartilage matrix is resilient and compressive, and the corneal matrix is transparent and precisely ordered.

Glycosaminoglycan (GAG) Chains Occupy Large Amounts of Space and Form Hydrated Gels

Glycosaminoglycans (GAGs) are long, unbranched polysaccharide chains composed of repeating disaccharide units. Their high density of negative charges (from sulfate and carboxyl groups) attracts cations and water, causing GAG chains to swell enormously and occupy a volume far greater than their mass would suggest. This creates a hydrated, gel-like matrix that resists compressive forces — a property essential in cartilage and other load-bearing tissues.

The four main classes of GAGs are:

Hyaluronan Acts as a Space Filler During Tissue Morphogenesis and Repair

Hyaluronan (hyaluronic acid) is unique among GAGs: it is not sulfated, not linked to a protein core, and is synthesized at the plasma membrane rather than in the Golgi. Hyaluronan chains can be enormous — up to 25,000 disaccharide repeats — and they form viscous, space-filling hydrated gels. During embryonic development, hyaluronan-rich matrices create swollen spaces into which cells can migrate, as in the formation of heart valves and limb buds. In wound repair, hyaluronan is among the first matrix components to be produced. Hyaluronan binds to the cell-surface receptor CD44, which mediates cell migration through hyaluronan-rich matrices. Hyaluronan is also a major component of synovial fluid, where it acts as a lubricant in joints.

Proteoglycans Are Composed of GAG Chains Covalently Linked to a Core Protein

Proteoglycans consist of one or more GAG chains covalently attached to a core protein. The number, type, and length of GAG chains determine the proteoglycan’s properties. Some proteoglycans are enormous: aggrecan, the major proteoglycan of cartilage, carries ~100 chondroitin sulfate chains and ~30 keratan sulfate chains on a single core protein. Multiple aggrecan molecules bind to a hyaluronan backbone (via link proteins) to form massive aggregates that give cartilage its resilience. Cell-surface proteoglycans such as syndecans (transmembrane) and glypicans (GPI-anchored) carry heparan sulfate chains that bind and concentrate growth factors (FGFs, Wnts, Hedgehog) at the cell surface, modulating signaling.

Collagens Are the Major Proteins of the Extracellular Matrix

Collagen is the most abundant protein in the animal body, constituting approximately 30% of total protein mass. The collagen superfamily includes 28 types in humans (types I–XXVIII), all sharing a characteristic triple-helical structure: three polypeptide chains (α-chains) wind around each other in a right-handed superhelix. The triple helix requires a repeating Gly-X-Y tripeptide sequence (where X is often proline and Y is often hydroxyproline), since only glycine is small enough to fit in the crowded interior of the helix. The hydroxylation of proline and lysine residues, catalyzed by enzymes that require vitamin C (ascorbic acid) as a cofactor, is essential for triple-helix stability — vitamin C deficiency causes scurvy, in which collagen is structurally defective and connective tissues break down.

Collagen Secretion and Assembly

Collagen biosynthesis is an elaborate process. The α-chains are synthesized as precursors called procollagens, with N- and C-terminal propeptides that prevent premature fibril assembly inside the cell. Procollagen molecules are secreted into the extracellular space, where specific procollagen proteinases cleave the propeptides to yield tropocollagen molecules. These tropocollagen molecules then spontaneously assemble into collagen fibrils in a staggered arrangement, with each molecule offset by ~67 nm (the D-period) relative to its neighbors. Covalent crosslinks formed by the enzyme lysyl oxidase between lysine and hydroxylysine residues on adjacent molecules stabilize the fibrils and give them enormous tensile strength. Collagen fibrils are further organized into collagen fibers and higher-order structures (tendons, ligaments, bone matrix).

Cells Help Organize the Collagen Fibrils They Secrete by Exerting Tension on the Matrix

Collagen fibril organization is not random — cells actively orient their matrix. Fibroblasts secrete collagen and then pull on the surrounding matrix through integrin-mediated adhesions, aligning collagen fibrils along lines of mechanical stress. This reciprocal interaction between cells and matrix means that the ECM architecture reflects the mechanical history of the tissue. In tendons, collagen fibrils are aligned in parallel, providing enormous tensile strength along a single axis. In skin, fibrils are organized in a woven basketweave pattern that resists forces from multiple directions.

Elastin Gives Tissues Their Elasticity

Elastin is a highly hydrophobic protein that forms elastic fibers, providing tissues with the ability to stretch and recoil. Elastic fibers are particularly abundant in the walls of arteries (which must expand and contract with each heartbeat), the lungs (which inflate and deflate with each breath), and the skin. Elastin molecules are crosslinked into a random-coil network by lysyl oxidase (the same enzyme that crosslinks collagen). This cross-linked network behaves like a rubber band: when stretched, the disordered elastin chains are pulled into a more extended conformation; when released, they snap back to their relaxed, disordered state. Elastic fibers contain a core of cross-linked elastin surrounded by a sheath of fibrillin microfibrils, which provide structural support and guide elastic fiber assembly. Mutations in the gene encoding fibrillin-1 cause Marfan syndrome, characterized by defective elastic fibers and resulting in tall stature, arachnodactyly, and aortic aneurysms.

Fibronectin and Other Multidomain Glycoproteins Help Organize the Matrix

Fibronectin is a large, multidomain glycoprotein that serves as a major organizer of the ECM. It exists as a disulfide-bonded dimer, with each monomer containing repeating domains (types I, II, and III) that bind collagen, heparin, fibrin, and cell-surface integrins. The key cell-binding site is the RGD (Arg-Gly-Asp) motif in the type III₁₀ repeat, recognized by several integrins (notably α5β1). Fibronectin circulates in blood as a soluble, compact molecule and is assembled into insoluble fibrils at cell surfaces in a process that requires integrin-mediated tension — cells pull on bound fibronectin molecules, exposing cryptic self-association sites that promote fibril assembly. Other ECM glycoproteins include tenascin (modulates cell adhesion during development and wound healing) and thrombospondins (regulate angiogenesis and matrix assembly).

The Basal Lamina Is a Specialized Form of Extracellular Matrix

The basal lamina (basement membrane) is a thin, specialized sheet of ECM (~40–120 nm thick) that underlies all epithelial and endothelial cell layers and surrounds individual muscle cells, fat cells, and Schwann cells. It serves as a structural foundation, a selective permeability barrier (e.g., the glomerular basal lamina in the kidney filters blood), and a source of signals that influence cell behavior.

Laminins and Type IV Collagen Are Major Components of the Basal Lamina

The basal lamina is built on a framework of two independent polymer networks: type IV collagen and laminin. Laminins are large cross-shaped heterotrimeric glycoproteins (composed of α, β, and γ chains) that self-polymerize into a sheet-like network and bind to cell-surface receptors, including integrins and dystroglycan. Type IV collagen does not form fibrils; instead, its triple-helical molecules associate at their ends and along their length to form a two-dimensional mesh. The two networks are linked by nidogen (entactin) and the heparan sulfate proteoglycan perlecan, creating a composite material with both strength and signaling capacity.

Basal Laminae Have Diverse Functions

Basal laminae serve remarkably diverse roles depending on tissue context. In the kidney glomerulus, the basal lamina acts as a molecular filter, controlling which blood components pass into the urine. In muscle, the basal lamina organizes the neuromuscular junction, clustering acetylcholine receptors through the proteoglycan agrin. In the nervous system, basal laminae guide axon growth during development and regeneration. During development, basal laminae pattern epithelial morphogenesis — they can serve as barriers to cell migration or as tracks along which cells move. The composition of the basal lamina (which laminin and collagen IV isoforms are present) varies by tissue and developmental stage, encoding tissue-specific information.

Cells Have to Be Able to Degrade Extracellular Matrix as Well as Make It

ECM remodeling requires controlled proteolysis. The major ECM-degrading enzymes are the matrix metalloproteinases (MMPs) — a family of ~23 zinc-dependent endopeptidases in humans that collectively can degrade virtually all ECM components. MMPs are secreted as inactive zymogens (proMMPs) and activated by proteolytic cleavage in the extracellular space. Their activity is regulated at multiple levels: transcriptional control, proenzyme activation, and inhibition by tissue inhibitors of metalloproteinases (TIMPs). Other ECM-degrading enzymes include ADAMs (a disintegrin and metalloproteinase) and ADAMTS (ADAMs with thrombospondin motifs). Controlled ECM degradation is essential for tissue remodeling during development, wound healing, and immune cell migration, while excessive or inappropriate MMP activity contributes to cancer invasion, arthritis, and vascular disease.

Matrix Proteoglycans and Glycoproteins Regulate the Activities of Secreted Proteins

The ECM is not merely structural — it acts as a reservoir and regulator of signaling molecules. Heparan sulfate proteoglycans bind and concentrate growth factors such as FGFs, BMPs, Wnts, and VEGF in the matrix, controlling their local concentration, diffusion range, and presentation to cell-surface receptors. Proteolytic release of matrix-bound growth factors by MMPs can rapidly mobilize stored signals during wound healing or tissue remodeling. The ECM also sequesters TGF-β in a latent form; mechanical tension on the matrix (transmitted through integrins) can unfold the latent complex and activate TGF-β signaling without any protease activity — a striking example of mechanotransduction.

Bioinformatics: Matrisome Analysis

The complexity of the ECM has driven the development of specialized databases and computational tools.

The matrisome database (MatrisomeDB) defines the matrisome — the complete set of ECM and ECM-associated proteins in an organism. The human matrisome comprises ~1,000 genes, divided into the core matrisome (collagens, proteoglycans, ECM glycoproteins) and matrisome-associated proteins (ECM regulators, secreted factors, ECM-affiliated proteins). MatrisomeDB integrates proteomics, transcriptomics, and interaction data to catalog ECM composition across tissues and diseases.

Extracellular matrix proteomics analysis uses mass spectrometry to identify and quantify ECM proteins in tissues. Because the ECM is heavily cross-linked and insoluble, specialized extraction protocols (sequential detergent extraction followed by guanidine solubilization) are required. ECM proteomics has revealed dramatic differences in matrix composition between normal and tumor tissues.

Glycosaminoglycan and proteoglycan databases such as GlyConnect, GlyTouCan, and the UniCarbKB catalog GAG structures, biosynthetic enzymes, and binding partners. These resources support the emerging field of glycomics — the systematic study of glycan structures and their biological functions.

Matrix metalloproteinase (MMP) substrate prediction uses machine learning trained on known MMP cleavage sites to predict novel substrates from protein sequences. Tools like CleavPredict and PROSPER score candidate cleavage sites using position-specific scoring matrices and structural features.

ECM stiffness and mechanobiology modeling combines finite-element analysis of tissue mechanics with molecular-scale models of collagen fiber mechanics and cell-matrix force transmission. These models predict how changes in ECM composition (e.g., increased collagen crosslinking in fibrosis) alter tissue stiffness and cellular mechanotransduction.

The ECM composition varies dramatically between connective tissues. Let us visualize the major components of a typical connective tissue ECM.

let ecm_data = '[{"label": "Collagen I", "value": 60}, {"label": "Collagen IV", "value": 15}, {"label": "Fibronectin", "value": 10}, {"label": "Laminin", "value": 8}, {"label": "Proteoglycans", "value": 7}]'
let chart = Viz.bar(ecm_data, '{"title": "Connective Tissue ECM Composition (%)", "color": "#F59E0B"}')
print(chart)

Collagen I dominates the ECM of most connective tissues, reflecting its role as the primary structural protein. The relative proportions of these components vary between tissues — the basal lamina, for instance, is enriched in collagen IV and laminin rather than fibrillar collagen I.

19.3 — Integrins and Cell-Matrix Adhesion

Integrins Are Transmembrane Heterodimers

Integrins are the principal receptors by which animal cells attach to the ECM. Each integrin is a heterodimer of one α subunit and one β subunit, both of which are single-pass transmembrane glycoproteins. The human genome encodes 18 α and 8 β subunits, which combine to form 24 distinct integrin heterodimers. Different integrins recognize different ECM ligands: α5β1 binds fibronectin (via the RGD motif), α2β1 binds collagen, α6β4 binds laminin, and the αVβ3 integrin binds vitronectin and other RGD-containing proteins. The extracellular domains of both subunits contribute to ligand binding, while the short cytoplasmic tails connect to intracellular signaling and cytoskeletal proteins.

Integrins Can Switch Between Active and Inactive Conformations

Integrins exist in at least three conformational states: a bent/inactive conformation (low affinity for ligand), an extended/primed conformation (intermediate affinity), and an extended/open conformation (high affinity). The transition between these states is regulated bidirectionally:

• Inside-out signaling — intracellular signals (e.g., the binding of talin and kindlin to the β subunit cytoplasmic tail) trigger conformational changes that propagate through the transmembrane domains to the extracellular headpiece, switching the integrin to its active, high-affinity state
• Outside-in signaling — ligand binding to the extracellular domain triggers conformational changes that propagate inward, activating intracellular signaling pathways

This conformational switching means that cells can rapidly regulate their adhesiveness — critical for processes such as platelet aggregation (where integrin αIIbβ3 must be activated rapidly upon vascular injury) and leukocyte extravasation (where integrins must switch on to enable firm adhesion to endothelium).

Integrins Cluster to Form Strong Adhesions

Individual integrin-ligand interactions are weak (K_d in the millimolar range), but cells achieve strong adhesion by clustering integrins into multivalent assemblies. At focal adhesions — large, elongated adhesion structures on the ventral surface of cultured cells — hundreds of integrins cluster together, each connected through adaptor proteins to bundles of actin stress fibers. Focal adhesions contain over 150 different proteins, including talin, vinculin, paxillin, focal adhesion kinase (FAK), and zyxin. The assembly of focal adhesions is itself force-dependent: actomyosin-generated tension promotes the maturation of small, nascent adhesions (focal complexes) into large, stable focal adhesions. Other types of integrin-based adhesions include hemidesmosomes (which link α6β4 integrin to intermediate filaments via plectin, providing strong anchorage of epithelial cells to the basal lamina) and podosomes/invadopodia (actin-rich, MMP-secreting structures used by macrophages and invasive cancer cells to degrade and penetrate matrix).

Extracellular Matrix Attachments Act Through Integrins to Control Cell Proliferation, Survival, and Cell Shape

Integrin-mediated adhesion to the ECM does far more than anchor cells in place — it controls fundamental cell fate decisions. Most normal cells require attachment to a suitable ECM substrate in order to proliferate; without it, they arrest in G1 or undergo a form of apoptosis called anoikis (“homelessness”). The mechanical properties of the matrix matter as well: cells cultured on stiff substrates (mimicking fibrotic tissue or bone) spread, generate strong traction forces, and tend to proliferate, while the same cells on soft substrates (mimicking brain or fat tissue) remain rounded and quiescent. Mesenchymal stem cells can be directed toward different lineages based solely on substrate stiffness — soft matrices promote neuronal differentiation, intermediate stiffness promotes muscle, and stiff matrices promote bone. This phenomenon, called mechanotransduction, demonstrates that physical forces and ECM mechanics are as important as soluble growth factors in controlling cell behavior.

Integrins Recruit Intracellular Signaling Proteins at Sites of Cell-Matrix Adhesion

Integrin cytoplasmic tails lack intrinsic enzymatic activity, but they recruit a diverse array of signaling proteins to focal adhesions. The key mediators include:

• Focal adhesion kinase (FAK) — a tyrosine kinase that autophosphorylates upon integrin clustering, creating a docking site for Src family kinases and activating downstream pathways (Ras–MAPK, PI3K–Akt) that promote cell survival and proliferation
• Integrin-linked kinase (ILK) — binds the β-integrin tail and connects to actin through parvin and α-actinin
• Rho GTPases (Rho, Rac, Cdc42) — activated at adhesion sites to reorganize the actin cytoskeleton, linking matrix adhesion to cell shape changes and migration
• YAP/TAZ — transcriptional co-activators in the Hippo pathway that are regulated by cell-matrix adhesion and cytoskeletal tension; on stiff matrices, YAP/TAZ enter the nucleus and activate genes promoting proliferation and survival

The convergence of integrin signaling with growth factor signaling (through shared downstream pathways such as MAPK and PI3K) explains why adhesion and mitogenic signals are jointly required for cell proliferation — and why loss of adhesion-dependence is a hallmark of cancer cells.

Bioinformatics: Integrin Interaction Networks

The complexity of integrin signaling has motivated large-scale proteomics and computational network approaches.

Integrin-ligand interaction databases catalog the specificities of all 24 human integrin heterodimers for their ECM ligands. Resources such as the IUPHAR Guide to Pharmacology and curated reviews provide ligand-binding profiles, affinities, and tissue expression patterns for each integrin.

Focal adhesion proteomics (Adhesome database) provides a curated inventory of all proteins identified at integrin adhesion sites. The Adhesome (adhesome.org) catalogs over 230 components of the consensus adhesome and organizes them into functional modules (actin regulators, adaptors, kinases, phosphatases). Network analysis of the adhesome reveals hub proteins (talin, vinculin, FAK, paxillin) that serve as critical nodes in adhesion signaling.

Integrin signaling pathway analysis uses tools such as Reactome, KEGG, and Gene Set Enrichment Analysis (GSEA) to analyze integrin-downstream signaling in transcriptomic and proteomic datasets. Enrichment of integrin/focal adhesion gene sets in expression data can reveal whether integrin signaling is activated or suppressed in a disease context.

Mechanosensing simulation models combine molecular dynamics of integrin conformational changes with continuum-mechanics models of cell-matrix force transmission to predict how cells sense and respond to matrix stiffness. These models incorporate the force-dependent unfolding of talin (which exposes vinculin-binding sites under tension) and the catch-bond behavior of integrins (whose bond lifetime increases under moderate force).

Lysyl oxidase is the key enzyme that crosslinks both collagen and elastin, converting lysine residues to the reactive aldehyde allysine. Let us examine the chemical properties of this crosslinking intermediate.

let lysyl_oxidase_product = "OC(=O)C(N)CCCC=O"
print("Allysine (lysyl oxidase product) - crosslinks collagen:")
print(Chem.properties(lysyl_oxidase_product))

The aldehyde group of allysine reacts spontaneously with other allysine residues or with unmodified lysine residues on adjacent collagen molecules, forming covalent crosslinks that give collagen fibrils their enormous tensile strength. The same enzyme crosslinks elastin into its elastic network.

Exercises

Exercise: Collagen Gly-X-Y Repeat

Collagen’s triple helix requires glycine at every third position (Gly-X-Y pattern). Analyze why glycine is essential — it’s the smallest amino acid and fits in the crowded center of the triple helix:

let normal_collagen = "GPPGPAGPKGERGDAGAPGPQ"
let mutant_collagen = "APPGPAGPKGERGDAGAPGPQ"
print("Normal collagen (Gly at every 3rd):")
print(Struct.protein_props(normal_collagen))
print("Mutant collagen (Gly→Ala at position 1):")
print(Struct.protein_props(mutant_collagen))
// Which forms a proper triple helix?
let answer = "normal"
print(answer)

Exercise: Compare Tissue ECM Compositions

Different tissues have dramatically different ECM compositions that reflect their mechanical requirements. Compare bone, cartilage, and tendon:

let bone = '[{"label": "Collagen I", "value": 30}, {"label": "Hydroxyapatite", "value": 60}, {"label": "Proteoglycans", "value": 5}]'
let cartilage = '[{"label": "Collagen II", "value": 25}, {"label": "Aggrecan", "value": 50}, {"label": "Water", "value": 20}]'
let tendon = '[{"label": "Collagen I", "value": 80}, {"label": "Elastin", "value": 10}, {"label": "Proteoglycans", "value": 5}]'
print(Viz.bar(bone, '{"title": "Bone ECM (%)", "color": "#6B7280"}'))
print(Viz.bar(cartilage, '{"title": "Cartilage ECM (%)", "color": "#3B82F6"}'))
print(Viz.bar(tendon, '{"title": "Tendon ECM (%)", "color": "#F59E0B"}'))
// Which tissue has the highest proportion of collagen I?
let answer = "tendon"
print(answer)

Exercise: Cell Junction Protein Properties

Different cell junctions use different transmembrane proteins. Compare the extracellular domains of cadherins (adherens junctions) and claudins (tight junctions):

let cadherin_ec = "DWVIPPINLPESDSGTSVTLSIQAT"
let claudin_loop = "WMKGLWMSCVSQSTG"
print("E-cadherin extracellular domain:")
print(Struct.protein_props(cadherin_ec))
print("Claudin-1 extracellular loop:")
print(Struct.protein_props(claudin_loop))
// Which protein has the larger extracellular domain?
let answer = "cadherin"
print(answer)

Knowledge Check

Summary

In this lesson you covered how cells adhere to each other, interact with the extracellular matrix, and transduce mechanical signals:

• Cadherins are calcium-dependent adhesion molecules that mediate homophilic binding through the strand-swap mechanism in their EC1 domain; they connect to the actin cytoskeleton through β-catenin, α-catenin, and p120-catenin
• Adherens junctions are mechanosensitive, reinforcing under tension as α-catenin unfolds to recruit vinculin; β-catenin also functions in Wnt signaling
• Desmosomes use desmosomal cadherins (desmogleins, desmocollins) linked to intermediate filaments through desmoplakin, providing exceptional mechanical strength to epithelia and cardiac muscle
• Tight junctions contain claudins, occludin, and JAMs organized by ZO scaffold proteins; they create a paracellular seal and a fence that maintains membrane polarity
• Gap junctions are channels of six connexin subunits (connexons) that allow ions and small molecules to pass directly between cells, enabling electrical and metabolic coupling
• Plasmodesmata in plants serve analogous functions to gap junctions, providing cytoplasmic continuity through the cell wall
• The ECM is a dynamic network of collagens, elastin, proteoglycans (with GAG chains), and glycoproteins (fibronectin, laminin) that provides structural support and regulates cell behavior
• Collagen, the most abundant animal protein, forms triple-helical fibrils cross-linked by lysyl oxidase; its assembly requires vitamin C–dependent hydroxylation
• The basal lamina is a specialized ECM sheet composed of laminin, type IV collagen, nidogen, and perlecan that underlies epithelia and performs tissue-specific functions
• MMPs degrade ECM components and are regulated by TIMPs; controlled proteolysis is essential for tissue remodeling but contributes to cancer invasion when dysregulated
• Integrins are αβ heterodimers that switch between inactive and active conformations via inside-out and outside-in signaling
• Focal adhesions cluster integrins and recruit over 150 proteins (FAK, talin, vinculin, paxillin) to link the ECM to actin stress fibers and intracellular signaling pathways
• Mechanotransduction through integrins controls proliferation, survival, differentiation, and cell shape; substrate stiffness alone can direct stem cell fate
• Bioinformatics tools — MatrisomeDB, the Adhesome database, MMP substrate predictors, and mechanosensing simulation models — enable systematic analysis of ECM composition, adhesion networks, and mechanical signaling