TISSUE PROLIFERATIVE ACTIVITY

The tissues of the body are divided into three groups on the basis of the proliferative activity of their cells: continuously dividing (labile tissues), quiescent (stable tissues), and nondividing (permanent tissues). This time-honored classification should be interpreted in the light of recent findings on stem cells and the reprogramming of cell differentiation.

STEM CELLS

Research on stem cells is at the forefront of modern-day biomedical investigation and stands at the core of a new field called regenerative medicine. The enthusiasm created by stem cell research derives from findings that challenge established views about cell differentiation, and from the hope that stem cells may one day be used to repair damaged human tissues, such as heart, brain, liver, and skeletal muscle.^3,6,7

Stem cells are characterized by their self-renewal properties and by their capacity to generate differentiated cell lineages (Fig. 3-4). To give rise to these lineages, stem cells need to be maintained during the life of the organism. Such maintenance is achieved by two mechanisms⁸: (a) obligatory asymmetric replication, in which with each stem cell division, one of the daughter cells retains its self-renewing capacity while the other enters a differentiation pathway, and (b) stochastic differentiation, in which a stem cell population is maintained by the balance between stem cell divisions that generate either two self-renewing stem cells or two cells that will differentiate. In early stages of embryonic development, stem cells, known as embryonic stem cells or ES cells, are pluripotent, that is, they can generate all tissues of the body (Fig. 3-4). Pluripotent stem cells give rise to multipotent stem cells, which have more restricted developmental potential, and eventually produce differentiated cells from the three embryonic layers. The term transdifferentiation (discussed later) indicates a change in the lineage commitment of a stem cell.

FIGURE 3-4 Stem cell generation and differentiation. The zygote, formed by the union of sperm and egg, divides to form blastocysts, and the inner cell mass of the blastocyst generates the embryo. The cells of the inner cell mass, known as embryonic stem (ES) cells, maintained in culture, can be induced to differentiate into cells of multiple lineages. In the embryo, pluripotent stem cells divide, but the pool of these cells is maintained (see text). As pluripotent cells differentiate, they give rise to cells with more restricted developmental capacity, and finally generate stem cells that are committed to specific lineages.

In adults, stem cells (often referred to as adult stem cells or somatic stem cells) with a more restricted capacity to generate different cell types have been identified in many tissues. They have been studied in detail in the skin, the lining of the gut, the cornea, and particularly in the hematopoietic tissue. An unexpected finding has been the discovery of stem cells and neurogenesis in areas of the central nervous system of adult animals and humans.⁹ Somatic stem cells for the most part reside in special microenvironments called niches (Fig. 3-5), composed of mesenchymal, endothelial, and other cell types.^10,11 It is believed that niche cells generate or transmit stimuli that regulate stem cell self-renewal and the generation of progeny cells. Recent groundbreaking research has now demonstrated that differentiated cells of rodents and humans can be reprogrammed into pluripotent cells, similar to ES cells, by the transduction of genes encoding ES cell transcription factors.^12,13 These reprogrammed cells have been named induced pluripotent stem cells (iPS cells). Their discovery has opened open an exciting new era in stem cell research and its applications.

FIGURE 3-5 Stem cell niches in various tissues. A, Skin stem cells are located in the bulge area of the hair follicle, in sebaceous glands, and in the lower layer of the epidermis. B, Small intestine stem cells located near the base of a crypt, above Paneth cells (stem cells in the small intestine may also be located at the bottom of the crypt²⁵). C, Liver stem (progenitor) cells, known as oval cells, are located in the canals of Hering (thick arrow), structures that connect bile ductules (thin arrow) with parenchymal hepatocytes (bile duct and Hering canals are stained for cytokeratin 7). D, Corneal stem cells are located in the limbus region, between the conjunctiva and the cornea.

(C, courtesy of Tania Roskams, MD, University of Leuven, Leuven, Belgium; D, Courtesy of T-T Sun, MD, New York University, New York, NY.)

We start our discussion of stem cells with a brief consideration of ES cells and the newly identified iPS cells. This is followed by a presentation on adult stem cells from a few selected tissues and the role they play in regeneration and repair.

Reprogramming of Differentiated Cells: Induced Pluripotent Stem Cells

Differentiated cells of adult tissues can be reprogrammed to become pluripotent by transferring their nucleus to an enucleated oocyte. The oocytes implanted into a surrogate mother can generate cloned embryos that develop into complete animals. This procedure, known as reproductive cloning, was successfully demonstrated in 1997 by the cloning of Dolly the sheep.¹⁷ There has been great hope that the technique of nuclear transfer to oocytes may be used for therapeutic cloning in the treatment of human diseases (illustrated in Fig. 3-6). In this technique the nucleus of a skin fibroblast from a patient is introduced into an enucleated human oocyte to generate ES cells, which are kept in culture, and then induced to differentiate into various cell types. In principle, these cells can then be transplanted into the patient to repopulate damaged organs.¹⁸ In addition to the ethical issues associated with this technique, therapeutic as well as reproductive cloning are inefficient and often inaccurate. One of the main reasons for the inaccuracy is the deficiency in histone methylation in reprogrammed ES cells, which results in improper gene expression.

FIGURE 3-6 Steps involved in stem cell therapy, using embryonic stem (ES) cells or induced pluripotent stem (iPS) cells. Left side, Therapeutic cloning using ES cells. The diploid nucleus of an adult cell from a patient is introduced into an enucleated oocyte. The oocyte is activated, and the zygote divides to become a blastocyst that contains the donor DNA. The blastocyst is dissociated to obtain ES cells. Right side, Stem cell therapy using iPS cells. The cells of a patient are placed in culture and transduced with genes encoding transcription factors, to generate iPS cells. Both ES and iPS cells are capable of differentiating into various cell types. The goal of stem cell therapy is to repopulate damaged organs of a patient or to correct a genetic defect, using the cells of the same patient to avoid immunological rejection.

(Modified from Hochedlinger K, Jaenisch R: Nuclear transplantation, embryonic stem cells, and the potential for cell therapy. N Engl J Med 349:275–286, 2003.)

Until recently there were no clues about the mechanisms that maintain pluripotency of ES cells. A series of landmark experiments have now demonstrated that the pluripotency of mouse ES cells depends on the expression of four transcription factors, Oct3/4, Sox2, c-myc, and Klf4, while the homeobox protein Nanog (named after Tir na n’Og, the Celtic land of the ever-young) acts to prevent differentiation.^19-22 Human fibroblasts from adults and newborns have been reprogrammed into pluripotent cells by the transduction of four genes encoding transcription factors (Oct3/4, Sox2, c-myc and Kfl4 in one laboratory; Oct3/4, Sox2, Nanog, and Lin28 in experiments from another laboratory).^12,13 The reprogrammed cells, known as iPS cells, are able to generate cells from endodermal, mesodermal, and ectodermal origin. They have also been used to rescue mice with a model of sickle cell anemia, proving that they function in vivo even after genetic manipulation and transplantation.²³ More recently, pluripotent iPS cells were generated by transfecting mouse hepatocytes, gastric cells, and terminally differentiated mature B lymphocytes with genes for the same four transcription factors.^24,25 Thus, iPS cells may become a source of cells for patient-specific stem cell therapy, without the involvement of nuclear transfer into oocytes (see Fig. 3-6). To make the dream of using iPS cells in human regenerative medicine a reality (iPS cells have been called ES cells without embryos), much additional work is required, including the development of new methods for gene delivery and the replacement of c-myc and Kfl4, which are oncogenes.²⁶ In any event, exciting new achievements can be expected in the near future from work on ES cells, iPS cells, and cell reprogramming.

GROWTH FACTORS

The proliferation of many cell types is driven by polypeptides known as growth factors. These factors, which can have restricted or multiple cell targets, may also promote cell survival, locomotion, contractility, differentiation, and angiogenesis, activities that may be as important as their growth-promoting effects. All growth factors function as ligands that bind to specific receptors, which deliver signals to the target cells. These signals stimulate the transcription of genes that may be silent in resting cells, including genes that control cell cycle entry and progression. Table 3-1 lists some of the most important growth factors involved in tissue regeneration and repair. Here we review only those that have major roles in these processes. Other growth factors are alluded to in various sections of the book.

TABLE 3-1 Growth Factors and Cytokines Involved in Regeneration and Wound Healing

SIGNALING MECHANISMS IN CELL GROWTH

In this section we review the process of receptor-mediated signal transduction, which is activated by the binding of ligands such as growth factors, and cytokines to specific receptors. Different classes of receptor molecules and pathways initiate a cascade of events by which receptor activation leads to expression of specific genes. Here we focus on the biochemical pathways and transcriptional regulation that mediate growth factor activity.

According to the source of the ligand and the location of its receptors (i.e., in the same, adjacent, or distant cells), three general modes of signaling, named autocrine, paracrine, and endocrine, can be distinguished (Fig. 3-8).

FIGURE 3-8 General patterns of intercellular signaling demonstrating autocrine, paracrine, and endocrine signaling (see text).

(Modified from Lodish H et al [eds]: Molecular Cell Biology, 3rd ed. New York, WH Freeman, 1995, p 855. © 1995 by Scientific American Books. Used with permission of WH Freeman and Company.)

Receptors and Signal Transduction Pathways

The binding of a ligand to its receptor triggers a series of events by which extracellular signals are transduced into the cell resulting in changes in gene expression. Although single receptor molecules can transduce some signals upon ligand binding, signaling typically involves clustering of two or more receptor molecules by the ligand. Receptors are generally located on the surface of the target cell but can also be found in the cytoplasm or nucleus.

It is useful at this point to summarize the properties of the major types of receptors and how they deliver signals to the cell interior (Fig. 3-9). This is pertinent to an understanding of normal and unregulated (neoplastic) cell growth (discussed in Chapter 7).

• Receptors with intrinsic tyrosine kinase activity. The ligands for receptors with tyrosine kinase activity include most growth factors such as EGF, TGF-α, HGF, PDGF, VEGF, FGF, c-KIT ligand, and insulin. Receptors belonging to this family have an extracellular ligand-binding domain, a transmembrane region, and a cytoplasmic tail that has intrinsic tyrosine kinase activity.⁶¹ Binding of the ligand induces dimerization of the receptor, tyrosine phosphorylation, and activation of the receptor tyrosine kinase (Fig. 3-10). The active kinase then phosphorylates, and thereby activates, many downstream effector molecules (molecules that mediate the effects of receptor engagement with a ligand). Activation of effector molecules can be direct or through the involvement of adapter proteins. A prototypical adapter protein is GRB-2, which binds a guanosine triphosphate–guanosine diphosphate (GTP-GDP) exchange factor called SOS. SOS acts on the GTP-binding (G) protein RAS and catalyzes the formation of RAS-GTP, which triggers the mitogen-activated protein kinase (MAP kinase) cascade (see Fig. 3-10). Active MAP kinases stimulate the synthesis and phosphorylation of transcription factors, such as FOS and JUN. The transcription factors activated by these various signaling cascades in turn stimulate the production of growth factors, receptors for growth factors, and proteins that directly control the entry of cells into the cell cycle. Other effector molecules activated by receptors with intrinsic tyrosine kinase activity include phospholipase Cγ (PLCγ) and phosphatidyl inositol-3 kinase (PI3K) (see Fig. 3-9). PLCγ catalyzes the breakdown of membrane inositol phospholipids into inositol 1,4,5-triphosphate (IP₃), which functions to increase concentrations of calcium, an important effector molecule, and diacylglycerol, which activates the serine-threonine kinase protein kinase C that in turn activates various transcription factors. PI3K phosphorylates a membrane phospholipid, generating products that activate the kinase Akt (also referred to as protein kinase B), which is involved in cell proliferation and cell survival through inhibition of apoptosis. Alterations in tyrosine kinase activity and receptor mutations have been detected in many forms of cancer and are important targets for therapy (Chapter 7).

• Receptors lacking intrinsic tyrosine kinase activity that recruit kinases. Ligands for these receptors include many cytokines, such as IL-2, IL-3, and other interleukins; interferons α, β, and γ; erythropoietin; granulocyte colony-stimulating factor; growth hormone; and prolactin. These receptors transmit extracellular signals to the nucleus by activating members of the JAK (Janus kinase) family of proteins (see Fig. 3-9). The JAKs link the receptors with and activate cytoplasmic transcription factors called STATs (signal transducers and activation of transcription), which directly shuttle into the nucleus and activate gene transcription.⁶² Cytokine receptors can also activate other signaling pathways, such as the MAP kinase pathways already mentioned.

• G protein–coupled receptors. These receptors transmit signals into the cell through trimeric GTP-binding proteins (G proteins). They contain seven transmembrane α-helices (see Fig. 3-9) and constitute the largest family of plasma membrane receptors, with nonodorant G protein–coupled receptors accounting for about 1% of the human genome. A large number of ligands signal through this type of receptor, including chemokines, vasopressin, serotonin, histamine, epinephrine and norepinephrine, calcitonin, glucagon, parathyroid hormone, corticotropin, and rhodopsin. An enormous number of common pharmaceutical drugs targets such receptors.⁶³ Binding of the ligand induces changes in the conformation of the receptors, causing their activation and allowing their interaction with many different G proteins. Activation of G proteins occurs by the exchange of GDP, present in the inactive protein, with GTP, which activates the protein. Among the many branches of this signal transduction pathway are those involving calcium and 3′,5′-cyclic adenosine monophosphate (cAMP) as second messengers. Activation of G protein–coupled receptors (as well as of tyrosine kinase receptors, discussed above) can produce inositol triphosphate (IP₃), which releases calcium from the endoplasmic reticulum. Calcium signals, which are generally oscillatory, have multiple targets, including cytoskeletal proteins, chloride- and potassium-activated ion pumps, enzymes such as calpain, and calcium-binding proteins such as calmodulin. cAMP activates a more restricted set of targets that include protein kinase A and cAMP-gated ion channels, important in vision and olfactory sensing. Inherited defects involving G protein–coupled receptor signal transduction are associated with retinitis pigmentosa, corticotropin deficiencies, and hyperparathyroidism.

• Steroid hormone receptors. These receptors are generally located in the nucleus and function as ligand-dependent transcription factors. The ligands diffuse through the cell membrane and bind the inactive receptors, causing their activation. The activated receptor then binds to specific DNA sequences known as hormone response elements within target genes, or they can bind to other transcription factors. In addition to steroid hormones, other ligands that bind to members of this receptor family include thyroid hormone, vitamin D, and retinoids. A group of receptors belonging to this family are called peroxisome proliferator-activated receptors.⁶⁴ They are nuclear receptors that are involved in a broad range of responses that include adipogenesis (Chapter 24), inflammation, and atherosclerosis.

FIGURE 3-9 Overview of the main types of cell surface receptors and their principal signal transduction pathways (see text). Shown are receptors with intrinsic tyrosine kinase activity, seven transmembrane G protein–coupled receptors, and receptors without intrinsic tyrosine kinase activity. cAMP, cyclic adenosine monophosphate: IP₃, inositol triphosphate; JAK, Janus kinase; MAP kinase, mitogen-activated protein kinase; PI3 kinase, phosphatidylinositol 3-kinase; PKB, protein kinase B, also known as Akt; PLC-γ, phospholipase C gamma; STATs, signal transducers and activators of transcription.

FIGURE 3-10 Signaling from tyrosine kinase receptors. Binding of the growth factor (ligand) causes receptor dimerization and autophosphorylation of tyrosine residues. Attachment of adapter (or bridging) proteins (e.g., GRB2 and SOS) couples the receptor to inactive RAS. Cycling of RAS between its inactive and active forms is regulated by GAP. Activated RAS interacts with and activates RAF (also known as MAP kinase kinase kinase). This kinase then phosphorylates a component of the MAP kinase signaling pathway, MEK (also known as MAP kinase kinase or MKK), which then phosphorylates ERK (MAP kinase or MK). Activated MAP kinase phosphorylates other cytoplasmic proteins and nuclear transcription factors, generating cellular responses. The phosphorylated tyrosine kinase receptor can also bind other components, such as phosphatidyl 3-kinase (PI3 kinase), which activates other signaling systems.

Transcription Factors

Many of the signal transduction systems used by growth factors transfer information to the nucleus and modulate gene transcription through the activity of transcription factors. Among the transcription factors that regulate cell proliferation are products of several growth-promoting genes, such as c-MYC and c-JUN, and of cell cycle–inhibiting genes, such as p53. Transcription factors have a modular design and contain domains for DNA binding and for transcriptional regulation. The DNA-binding domain permits binding to short sequence motifs of DNA, which may be unique to a particular target gene or may be present in many genes. The transactivating domain stimulates transcription of the adjacent gene.

Growth factors induce the synthesis or activity of transcription factors. Cellular events requiring rapid responses do not rely on new synthesis of transcription factors but depend on post-translational modifications that lead to their activation. These modifications include (a) heterodimerization, as for instance, the dimerization of the products of the protooncogenes c-FOS and c-JUN to form the transcription factor activator protein-1 (AP-1), which is activated by MAP kinase signaling pathways, (b) phosphorylation, as for STATs in the JAK/STAT pathway, (c) release of inhibition to permit migration into the nucleus, as for NF-κB, and (d) release from membranes by proteolytic cleavage, as for Notch receptors (see Fig. 3-16).

FIGURE 3-16 Notch signaling and angiogenesis. A, The Notch receptor binds a ligand (a delta-like ligand, Dll, is shown in the figure) located in an adjacent cell, and undergoes two proteolytic cleavages (the first cleavage by ADAM protease, the second by δ-secretase), releasing a C-terminal fragment known as Notch intracellular domain (Notch-ICD). B, Notch signaling in endothelial cells during angiogenesis, triggered by the binding of the Dll4 ligand in a tip cell to a Notch receptor in a stalk cell. Notch-ICD migrates into the nucleus and activates the transcription of target genes. C, Sprouting angiogenesis, showing a migrating tip cell and stalk cells connected to the endothelial cells of the main vessel.

(A, modified from Weinberg RA: The Biology of Cancer. New York, Garland Science, 2007, Fig. 5.22; B, modified from Kerbel RS: Tumor angiogenesis. N Engl J Med 358:2039, 2008.)

LIVER REGENERATION

The human liver has a remarkable capacity to regenerate, as demonstrated by its growth after partial hepatectomy, which may be performed for tumor resection or for living-donor hepatic transplantation (Fig. 3-11). The popular image of liver regeneration is the daily regrowth of the liver of Prometheus, which was eaten every day by an eagle sent by Zeus (Zeus was angry at Prometheus for stealing the secret of fire, but did he know that Prometheus’s liver would regenerate?). The reality, although less dramatic, is still quite impressive. In humans, resection of approximately 60% of the liver in living donors results in the doubling of the liver remnant in about one month. The portions of the liver that remain after partial hepatectomy constitute an intact “mini-liver” that rapidly expands and reaches the mass of the original liver (see Fig. 3-11). Restoration of liver mass is achieved without the regrowth of the lobes that were resected at the operation. Instead, growth occurs by enlargement of the lobes that remain after the operation, a process known as compensatory growth or compensatory hyperplasia. In both humans and rodents, the end point of liver regeneration after partial hepatectomy is the restitution of functional mass rather than the reconstitution of the original form.⁶⁹

FIGURE 3-11 Liver regeneration after partial hepatectomy. A, The lobes of the liver of a rat (M, median; RL and LL, right and left lateral lobes; C, caudate lobe). Partial hepatectomy removes two thirds of the liver (median and left lateral lobes). After 3 weeks the right lateral and caudate lobes enlarge to reach a mass equivalent to that of the original liver without regrowth of the median and left lateral lobes. B, Entry and progression of hepatocytes in the cell cycle (see text for details). C, Regeneration of the human liver in living-donor transplantation. Computed tomography scans of the donor liver in living-donor hepatic transplantation. Upper panel is a scan of the liver of the donor before the operation. The right lobe, to be used as a transplant, is outlined. Lower panel is a scan of the liver 1 week after performance of partial hepatectomy. Note the great enlargement of the left lobe (outlined in the panel) without regrowth of the right lobe.

(A, From Goss RJ: Regeneration versus repair. In Cohen IK et al [eds]: Wound Healing. Biochemical and Clinical Aspects. Philadelphia, WB Saunders, 1992, pp 20–39; C, courtesy of R. Troisi, MD, Ghent University, Ghent, Belgium; reproduced in part from Fausto N: Liver regeneration. In Arias I, et al: The Liver: Biology and Pathobiology, 4th ed. Philadelphia, Lippincott Williams & Wilkins, 2001.)

Almost all hepatocytes replicate during liver regeneration after partial hepatectomy. Because hepatocytes are quiescent cells, it takes them several hours to enter the cell cycle, progress through G₁, and reach the S phase of DNA replication. The wave of hepatocyte replication is synchronized and is followed by synchronous replication of nonparenchymal cells (Kupffer cells, endothelial cells, and stellate cells).

There is substantial evidence that hepatocyte proliferation in the regenerating liver is triggered by the combined actions of cytokines and polypeptide growth factors. With the exception of the autocrine activity of TGF-α, hepatocyte replication is strictly dependent on paracrine effects of growth factors and cytokines such as HGF and IL-6 produced by hepatic nonparenchymal cells. There are two major restriction points for hepatocyte replication: the G₀/G₁ transition that bring quiescent hepatocytes into the cell cycle, and the G₁/S transition needed for passage through the late G₁ restriction point. Gene expression in the regenerating liver proceeds in phases, starting with the immediate early gene response, which is a transient response that corresponds to the G₀/G₁ transition. More than 70 genes are activated during this response, including the proto-oncogenes c-FOS and c-JUN, whose products dimerize to form the transcription factor AP-1; c-MYC, which encodes a transcription factor that activates many different genes; and other transcription factors, such as NF-κB, STAT-3, and C/EBP.⁷⁰ The immediate early gene response sets the stage for the sequential activation of multiple genes, as hepatocytes progress into the G₁ phase. The G₁ to S transition occurs as previously described (see Fig. 3-7).

Quiescent hepatocytes become competent to enter the cell cycle through a priming phase that is mostly mediated by the cytokines TNF and IL-6, and components of the complement system. Priming signals activate several signal transduction pathways as a necessary prelude to cell proliferation. Under the stimulation of HGF, TGFα, and HB-EGF, primed hepatocytes enter the cell cycle and undergo DNA replication (Fig. 3-11). Norepinephrine, serotonin, insulin, thyroid and growth hormone, act as adjuvants for liver regeneration, facilitating the entry of hepatocytes into the cell cycle.

Individual hepatocytes replicate once or twice during regeneration and then return to quiescence in a strictly regulated sequence of events, but the mechanisms of growth cessation have not been established. Growth inhibitors, such as TGF-β and activins, may be involved in terminating hepatocyte replication, but there is no clear understanding of their mode of action. Intrahepatic stem or progenitor cells do not play a role in the compensatory growth that occurs after partial hepatectomy, and there is no evidence for hepatocyte generation from bone marrow–derived cells during this process.^28,37 However, endothelial cells and other nonparenchymal cells in the regenerating liver may originate from bone marrow precursors.

COLLAGEN

Collagen is the most common protein in the animal world, providing the extracellular framework for all multicellular organisms. Without collagen, a human being would be reduced to a clump of cells, like the “Blob” (the “gelatinous horror from outer space” of 1950s movie fame), interconnected by a few neurons. Currently, 27 different types of collagens encoded by 41 genes dispersed on at least 14 chromosomes are known⁷² (Table 3-2). Each collagen is composed of three chains that form a trimer in the shape of a triple helix. The polypeptide is characterized by a repeating sequence in which glycine is in every third position (Gly-X-Y, in which X and Y can be any amino acid other than cysteine or tryptophan), and it contains

TABLE 3-2 Main Types of Collagens, Tissue Distribution, and Genetic Disorders

Courtesy of Dr. Peter H. Byers, Department of Pathology, University of Washington, Seattle, WA.

FIGURE 3-12 Main components of the extracellular matrix (ECM), including collagens, proteoglycans, and adhesive glycoproteins. Both epithelial and mesenchymal cells (e.g., fibroblasts) interact with ECM via integrins. Basement membranes and interstitial ECM have different architecture and general composition, although there is some overlap in their constituents. For the sake of simplification, many ECM components (e.g., elastin, fibrillin, hyaluronan, and syndecan) are not included.

the specialized amino acids 4-hydroxyproline and hydroxylysine. Prolyl residues in the Y-position are characteristically hydroxylated to produce hydroxyproline, which serves to stabilize the triple helix. Types I, II, III and V, and XI are the fibrillar collagens, in which the triple-helical domain is uninterrupted for more than 1000 residues; these proteins are found in extracellular fibrillar structures. Type IV collagens have long but interrupted triple-helical domains and form sheets instead of fibrils; they are the main components of the basement membrane, together with laminin. Another collagen with a long interrupted triple-helical domain (type VII) forms the anchoring fibrils between some epithelial and mesenchymal structures, such as epidermis and dermis. Still other collagens are transmembrane and may also help to anchor epidermal and dermal structures.

The messenger RNAs transcribed from fibrillar collagen genes are translated into pre-pro-α chains that assemble in a type-specific manner into trimers. Hydroxylation of proline and lysine residues and lysine glycosylation occur during translation. Three chains of a particular collagen type assemble to form the triple helix (Fig. 3-15). Procollagen is secreted from the cell and cleaved by proteases to form the basic unit of the fibrils. Collagen fibril formation is associated with the oxidation of lysine and hydroxylysine residues by the extracellular enzyme lysyl oxidase. This results in cross-linking between the chains of adjacent molecules, which stabilizes the array, and is a major contributor to the tensile strength of collagen. Vitamin C is required for the hydroxylation of procollagen, a requirement that explains the inadequate wound healing in scurvy (Chapter 9). Genetic defects in collagen production (see Table 3-2) cause many inherited syndromes, including various forms of the Ehlers-Danlos syndrome and osteogenesis imperfecta⁷³ (Chapters 5 and 26.

FIGURE 3-15 Angiogenesis by mobilization of endothelial precursor cells (EPCs) from the bone marrow and from preexisting vessels (capillary growth). A, In angiogenesis from preexisting vessels, endothelial cells from these vessels become motile and proliferate to form capillary sprouts. Regardless of the initiating mechanism, vessel maturation (stabilization) involves the recruitment of pericytes and smooth muscle cells to form the periendothelial layer. B, EPCs are mobilized from the bone marrow and may migrate to a site of injury or tumor growth. At these sites, EPCs differentiate and form a mature network by linking to existing vessels.

(Modified from Conway EM et al: Molecular mechanisms of blood vessel growth. Cardiovasc Res 49:507, 2001.)

ELASTIN, FIBRILLIN, AND ELASTIC FIBERS

Tissues such as blood vessels, skin, uterus, and lung require elasticity for their function. Proteins of the collagen family provide tensile strength, but the ability of these tissues to expand and recoil (compliance) depends on the elastic fibers. These fibers can stretch and then return to their original size after release of the tension. Morphologically, elastic fibers consist of a central core made of elastin, surrounded by a peripheral network of microfibrils. Substantial amounts of elastin are found in the walls of large blood vessels, such as the aorta, and in the uterus, skin, and ligaments. The peripheral microfibrillar network that surrounds the core consists largely of fibrillin, a 350-kD secreted glycoprotein, which associates either with itself or with other components of the ECM. The microfibrils serve, in part, as scaffolding for deposition of elastin and the assembly of elastic fibers. They also influence the availability of active TGFβ in the ECM. As already mentioned, inherited defects in fibrillin result in formation of abnormal elastic fibers in Marfan syndrome, manifested by changes in the cardiovascular system (aortic dissection) and the skeleton⁷⁴ (Chapter 5).

CELL ADHESION PROTEINS

Most adhesion proteins, also called CAMs (cell adhesion molecules), can be classified into four main families: immunoglobulin family CAMs, cadherins, integrins, and selectins. These proteins function as transmembrane receptors but are sometimes stored in the cytoplasm.⁷⁵ As receptors, CAMs can bind to similar or different molecules in other cells, providing for interaction between the same cells (homotypic interaction) or different cell types (heterotypic interaction). Selectins have been discussed in Chapter 2 in the context of leukocyte/endothelial interactions. Selected aspects of other cell adhesion proteins are described here. Integrins bind to ECM proteins such as fibronectin, laminin, and osteopontin providing a connection between cells and ECM, and also to adhesive proteins in other cells, establishing cell-to-cell contact. Fibronectin is a large protein that binds to many molecules, such as collagen, fibrin, proteoglycans, and cell surface receptors. It consists of two glycoprotein chains, held together by disulfide bonds. Fibronectin messenger RNA has two splice forms, giving rise to tissue fibronectin and plasma fibronectin. The plasma form binds to fibrin, helping to stabilize the blood clot that fills the gaps created by wounds, and serves as a substratum for ECM deposition and formation of the provisional matrix during wound healing (discussed later). Laminin is the most abundant glycoprotein in the basement membrane and has binding domains for both ECM and cell surface receptors. In the basement membrane, polymers of laminin and collagen type IV form tightly bound networks. Laminin can also mediate the attachment of cells to connective tissue substrates.

Cadherins and integrins link the cell surface with the cytoskeleton through binding to actin and intermediate filaments. These linkages, particularly for the integrins, provide a mechanism for the transmission of mechanical force and the activation of intracellular signal transduction pathways that respond to these forces. Ligand binding to integrins causes clustering of the receptors in the cell membrane and formation of focal adhesion complexes. The cytoskeletal proteins that co-localize with integrins at the cell focal adhesion complex include talin, vinculin, and paxillin. The integrin-cytoskeleton complexes function as activated receptors and trigger a number of signal transduction pathways, including the MAP kinase, PKC, and PI3K pathways, which are also activated by growth factors. Not only is there a functional overlap between integrin and growth factor receptors, but integrins and growth factor receptors interact (“cross-talk”) to transmit environmental signals to the cell that regulate proliferation, apoptosis, and differentiation (Fig. 3-13).

FIGURE 3-13 Mechanisms by which ECM components and growth factors interact and activate signaling pathways. Integrins bind ECM components and interact with the cytoskeleton at focal adhesion complexes (protein aggregates that include vinculin, α-actin, and talin). This can initiate the production of intracellular messengers or can directly mediate nuclear signals. Cell surface receptors for growth factors may activate signal transduction pathways that overlap with those activated by integrins. Signaling from ECM components and growth factors is integrated by the cell to produce various responses, including changes in cell proliferation, locomotion, and differentiation.

The name cadherin is derived from the term “calcium-dependent adherence protein.” This family contains almost 90 members, which participate in interactions between cells of the same type. These interactions connect the plasma membrane of adjacent cells, forming two types of cell junctions called (1) zonula adherens, small, spotlike junctions located near the apical surface of epithelial cells, and (2) desmosomes, stronger and more extensive junctions, present in epithelial and muscle cells. Migration of keratinocytes in the re-epithelialization of skin wounds is dependent on the formation of dermosomal junctions. Linkage of cadherins with the cytoskeleton occurs through two classes of catenins. β-catenin links cadherins with α-catenin, which, in turn, connects to actin, thus completing the connection with the cytoskeleton. Cell-to-cell interactions mediated by cadherins and catenins play a major role in regulating cell motility, proliferation, and differentiation and account for the inhibition of cell proliferation that occurs when cultured normal cells contact each other (“contact inhibition”). Diminished function of E-cadherin contributes to certain forms of breast and gastric cancer. As already mentioned, free β-catenin acts independently of cadherins in the Wnt signaling pathway, which participates in stem cell homeostasis and regeneration. Mutation and altered expression of the Wnt/β-catenin pathway is implicated in cancer development, particularly in gastrointestinal and liver cancers (Chapter 7).

In addition to the main families of adhesive proteins described earlier, some other secreted adhesion molecules are mentioned because of their potential role in disease processes: (1) SPARC (secreted protein acidic and rich in cysteine), also known as osteonectin, contributes to tissue remodeling in response to injury and functions as an angiogenesis inhibitor; (2) the thrombospondins, a family of large multifunctional proteins, some of which, similar to SPARC, also inhibit angiogenesis; (3) osteopontin (OPN) is a glycoprotein that regulates calcification, is a mediator of leukocyte migration involved in inflammation, vascular remodeling, and fibrosis in various organs^76,77 (discussed later in this chapter); and (4) the tenascin family, which consist of large multimeric proteins involved in morphogenesis and cell adhesion.

GLYCOSAMINOGLYCANS (GAGS) AND PROTEOGLYCANS

GAGs make up the third type of component in the ECM, besides the fibrous structural proteins and cell adhesion proteins. GAGs consist of long repeating polymers of specific disaccharides. With the exception of hyaluronan (discussed later), GAGs are linked to a core protein, forming molecules called proteoglycans.⁷⁸ Proteoglycans are remarkable in their diversity. At most sites, ECM may contain several different core proteins, each containing different GAGs. Proteoglycans were originally described as ground substances or mucopolysaccharides, whose main function was to organize the ECM, but it is now recognized that these molecules have diverse roles in regulating connective tissue structure and permeability (Fig. 3-14). Proteoglycans can be integral membrane proteins and, through their binding to other proteins and the activation of growth factors and chemokines, act as modulators of inflammation, immune responses, and cell growth and differentiation.

FIGURE 3-14 Proteoglycans, glycosaminoglycans (GAGs), and hyaluronan. A, Regulation of FGF-2 activity by ECM and cellular proteoglycans. Heparan sulfate binds FGF-2 (basic FGF) secreted into the ECM. Syndecan is a cell surface proteoglycan with a transmembrane core protein, extracellular GAG side chains that can bind FGF-2, and a cytoplasmic tail that binds to the actin cytoskeleton. Syndecan side chains bind FGF-2 released by damage to the ECM and facilitate the interaction with cell surface receptors. B, Synthesis of hyaluronan at the inner surface of the plasma membrane. The molecule extends to the extracellular space, while still attached to hyaluronan synthase. C, Hyaluronan chains in the extracellular space are bound to the plasma membrane through the CD44 receptor. Multiple proteoglycans may attach to hyaluronan chains in the ECM.

(B and C, modified from Toole KR: Hyaluronan: from extracellular glue to pericellular cue. Nat Rev Cancer 4:528, 2004.)

There are four structurally distinct families of GAGs: heparan sulfate, chondroitin/dermatan sulfate, keratan sulfate, and hyaluronan (HA). The first three of these families are synthesized and assembled in the Golgi apparatus and rough endoplasmic reticulum as proteoglycans. By contrast, HA is produced at the plasma membrane by enzymes called hyaluronan synthases and is not linked to a protein backbone.

HA is a polysaccharide of the GAG family found in the ECM of many tissues and is abundant in heart valves, skin and skeletal tissues, synovial fluid, the vitreous of the eye, and the umbilical cord.⁷⁹ It is a huge molecule that consists of many repeats of a simple disaccharide stretched end-to-end. It binds a large amount of water (about 1000-fold its own weight), forming a viscous hydrated gel that gives connective tissue the ability to resist compression forces. HA helps provide resilience and lubrication to many types of connective tissue, notably for the cartilage in joints. Its concentration increases in inflammatory diseases such as rheumatoid arthritis, scleroderma, psoriasis, and osteoarthritis. Enzymes called hyaluronidases fragment HA into lower molecular weight molecules (LMW HA) that have different functions than the parent molecule. LMW HA produced by endothelial cells binds to the CD44 receptor on leukocytes, promoting the recruitment of leukocytes to the sites of inflammation. In addition, LMW HA molecules stimulate the production of inflammatory cytokines and chemokines by white cells recruited to the sites of injury. The leukocyte recruitment process and the production of pro-inflammatory molecules by LMW HA are strictly regulated processes; these activities are beneficial if short-lived, but their persistence may lead to prolonged inflammation.

MECHANISMS OF ANGIOGENESIS

Angiogenesis is a fundamental process that affects physiologic reactions (e.g. wound healing, regeneration, the vascularization of ischemic tissues, and menstruation), and pathologic processes, such as tumor development and metastasis, diabetic retinopathy, and chronic inflammation. Therefore great efforts have been made to understand the mechanisms of angiogenesis, and to develop agents that have pro- or anti-angiogenic activity.

Around 4000 BC, Egyptian physicians believed that there were “vessels for every part of the body, which are hollow, having a mouth which opens to absorb medications and eliminate waste”.⁸⁰ Fortunately, our understanding of blood vessels has clearly improved since then.^81,82 We know now that blood vessels are assembled during embryonic development by vasculogenesis, in which a primitive vascular network is established from endothelial cell precursors (angioblasts), or from dual hemopoietic/endothelial cell precursors called hemangioblasts. Blood vessel formation in adults, known as angiogenesis or neovascularization, involves the branching and extension of adjacent pre-existing vessels, but it can also occur by recruitment of endothelial progenitor cells (EPCs) from the bone marrow (Fig. 3-15).⁸¹

Growth Factors and Receptors Involved in Angiogenesis

Despite the diversity of factors that participate in angiogenesis, VEGF is the most important growth factor in adult tissues undergoing physiologic angiogenesis (e.g., proliferating endometrium) as well as angiogenesis occurring in chronic inflammation, wound healing, tumors, and diabetic retinopathy.^81,82

As mentioned earlier,⁵⁸ VEGF is secreted by many mesenchymal and stromal cells. Of the various receptors for VEGF, VEGFR-2, a tyrosine kinase receptor, is the most important in angiogenesis. It is expressed by endothelial cells and their precursors, by other cell types, and by many tumor cells. VEGF, or more specifically, its circulating isoforms VEGF₁₂₁ and VEGF₁₆₅, signal through VEGFR-2 (also known as KDR in humans and flk-1 in mice). VEGF induces the migration of EPCs in the bone marrow, and enhances the proliferation and differentiation of these cells at sites of angiogenesis. In angiogenesis originating from preexisting local vessels, VEGF signaling stimulates the survival of endothelial cells, their proliferation and their motility, initiating the sprouting of new capillaries. The main components of the VEGF/VEGFR system and their main actions are listed in Table 3-3. Endothelial cell proliferation, differentiation, and migration can also be stimulated by FGF-2.

Given the multiplicity of VEGF effects and the diverse mechanisms that regulate its expression, how do endothelial cells develop into a perfect pattern of vessels during angiogenesis? One recently identified mechanism for the modulation of vasculogenesis is the Notch pathway, which promotes the proper branching of new vessels and prevents excessive angiogenesis by decreasing the responsiveness to VEGF.^83-85 Notch ligands and receptors are membrane-bound molecules conserved between species. In mammals there are five Notch ligands (Jagged 1 and 2, and Delta-like ligand [Dll] 1, 3, and 4) and four transmembrane receptors (Notch 1–4). The receptors contain EGF-like repeats on their extracellular surface that serve as ligand-binding sites (Fig. 3-16). The Delta-like ligand 4 (Dll4) ligand is endothelial cell–specific and is expressed in arteries and capillaries but not in veins; the importance of this ligand is demonstrated by the embryonic lethality of mice that lack a single Delta-like ligand 4 allele. During angiogenesis the leading cell, known as a tip cell, undergoes proliferation and migration, but stalk cells maintain their connection with the existing vessel. VEGF induces Delta-like ligand 4 in tip cells, while Notch1 and Notch4 are expressed in stalk cells (see Fig. 3-16C). The interaction between Delta-like ligand 4 and Notch receptors in adjacent tip and stalk cells leads to a two-step proteolytic cleavage of the receptor, releasing the Notch intracellular domain, which translocates to the nucleus and activates genes that dampen responsiveness to VEGF. Blockade of Delta-like ligand 4 causes increased proliferation of endothelial cells and capillary sprouting; VEGF blockade has the opposite effects and also decreases the survival of endothelial cells (Fig. 3-17).

FIGURE 3-17 Interactions between Notch and VEGF during angiogenesis. VEGF stimulates delta-like ligand 4 (Dll4)/Notch, which inhibits VEGFR signaling. Compared with unperturbed angiogenesis, Dll4 blockade causes an increase in capillary sprouting and endothelial cell (EC) proliferation, creating vessels that are disorganized and have a small lumen size. VEGF blockade decreases capillary sprouting, and the proliferation and survival of ECs.

(Courtesy of Minhong Yan, Genentech, San Francisco, CA).

Regardless of the process that leads to capillary formation, newly formed vessels are fragile and need to become “stabilized.” Stabilization requires the recruitment of pericytes and smooth muscle cells (periendothelial cells) and the deposition of ECM proteins. Angiopoietins 1 and 2 (Ang1 and Ang2), PDGF, and TGF-β participate in the stabilization process. Ang1 interacts with a receptor on endothelial cells called Tie2 to recruit periendothelial cells. PDGF participates in the recruitment of smooth muscle cells, while TGF-β stabilizes newly formed vessels by enhancing the production of ECM proteins.⁵⁸ The Ang1-Tie2 interaction mediates vessel maturation from simple endothelial tubes into more elaborate vascular structures and helps maintain endothelial quiescence. Ang2, in contrast, which also interacts with Tie2, has the opposite effect, making endothelial cells become either more responsive to stimulation by growth factors such as VEGF or, in the absence of VEGF, more responsive to inhibitors of angiogenesis. A telling proof of the importance of these molecules is the existence of a genetic disorder caused by mutations in Tie2 that is characterized by venous malformations. Agents or conditions that stimulate VEGF expression, such as certain cytokines and growth factors (e.g., TGF-β, PDGF, TGF-α), and notably, tissue hypoxia, can influence physiologic and pathologic angiogenesis. VEGF transcription is regulated by the transcription factor HIF, which is induced by hypoxia.

CUTANEOUS WOUND HEALING

Cutaneous wound healing is divided into three phases: inflammation, proliferation, and maturation⁸⁶ (Fig. 3-18). These phases overlap, and their separation is somewhat arbitrary, but they help to understand the sequence of events that take place in the healing of skin wounds. The initial injury causes platelet adhesion and aggregation and the formation of a clot in the surface of the wound, leading to inflammation. In the proliferative phase there is formation of granulation tissue, proliferation and migration of connective tissue cells, and re-epithelialization of the wound surface. Maturation involves ECM deposition, tissue remodeling, and wound contraction.

FIGURE 3-18 Phases of cutaneous wound healing: inflammation, proliferation, and maturation (see text for details).

(Modified from Broughton G et al: The basic science of wound healing. Plast Reconstr Surg 117:12S–34S, 2006.)

The simplest type of cutaneous wound repair is the healing of a clean, uninfected surgical incision approximated by surgical sutures (Fig. 3-19). Such healing is referred to as healing by primary union or by first intention.^86-88 The incision causes death of a limited number of epithelial and connective tissue cells and disruption of epithelial basement membrane continuity. Re-epithelialization to close the wound occurs with formation of a relatively thin scar. The repair process is more complicated in excisional wounds that create large defects on the skin surface, causing extensive loss of cells and tissue. The healing of these wounds involves a more intense inflammatory reaction, the formation of abundant granulation tissue (described below), and extensive collagen deposition, leading to the formation of a substantial scar, which generally contracts. This form of healing is referred to as healing by secondary union or by second intention (see Figs. 3-19 and 3-20). Despite these differences, the basic mechanisms of healing by primary (first intention) and secondary (second intention) union are similar. They are described together and the differences will be indicated where appropriate.

FIGURE 3-19 Wound healing and scar formation. A, Healing of wound that caused little loss of tissue: note the small amount of granulation tissue, and formation of a thin scar with minimal contraction. B, Healing of large wound: note large amounts of granulation tissue and scar tissue, and wound contraction.

FIGURE 3-20 Healing of skin ulcers. A, Pressure ulcer of the skin, commonly found in diabetic patients. The histologic slides show: B, a skin ulcer with a large gap between the edges of the lesion; C, a thin layer of epidermal re-epithelialization and extensive granulation tissue formation in the dermis; and D, continuing re-epithelialization of the epidermis and wound contraction.

(Courtesy of Z. Argenyi, MD, University of Washington, Seattle, WA.)

A large number of growth factors and cytokines are involved in cutaneous wound healing.⁸⁹ The main agents, and the steps at which they participate in the repair process, are listed in Table 3-4. We next discuss the sequence of events in wound healing.

TABLE 3-4 Growth Factors and Cytokines Affecting Various Steps in Wound Healing

HB-EGF, heparin-binding EGF; IL-1, interleukin 1; TNF, tumor necrosis factor; other abbreviations as given in Table 3-1.

Formation of Blood Clot.

Wounding causes the rapid activation of coagulation pathways, which results in the formation of a blood clot on the wound surface (Chapter 4). In addition to entrapped red cells, the clot contains fibrin, fibronectin, and complement components. The clot serves to stop bleeding and also as a scaffold for migrating cells, which are attracted by growth factors, cytokines and chemokines released into the area.⁸⁹ Release of VEGF leads to increased vessel permeability and edema. However, dehydration occurs at the external surface of the clot, forming a scab that covers the wound. In wounds causing large tissue deficits, the fibrin clot is larger, and there is more exudate and necrotic debris in the wounded area. Within 24 hours, neutrophils appear at the margins of the incision, and use the scaffold provided by the fibrin clot to march in. They release proteolytic enzymes that clean out debris and invading bacteria.

Formation of Granulation Tissue.

Fibroblasts and vascular endothelial cells proliferate in the first 24 to 72 hours of the repair process to form a specialized type of tissue called granulation tissue, which is a hallmark of tissue repair. The term derives from its pink, soft, granular appearance on the surface of wounds. Its characteristic histologic feature is the presence of new small blood vessels (angiogenesis) and the proliferation of fibroblasts (Fig. 3-21). These new vessels are leaky, allowing the passage of plasma proteins and fluid into the extravascular space. Thus, new granulation tissue is often edematous. Granulation tissue progressively invades the incision space; the amount of granulation tissue that is formed depends on the size of the tissue deficit created by the wound and the intensity of inflammation. Hence, it is much more prominent in healing by secondary union. By 5 to 7 days, granulation tissue fills the wound area and neovascularization is maximal. The mechanisms of angiogenesis in repair processes were discussed earlier in this chapter.

FIGURE 3-21 A, Granulation tissue showing numerous blood vessels, edema, and a loose ECM containing occasional inflammatory cells. Collagen is stained blue by the trichrome stain; minimal mature collagen can be seen at this point. B, Trichrome stain of mature scar, showing dense collagen, with only scattered vascular channels.

Cell Proliferation and Collagen Deposition.

Neutrophils are largely replaced by macrophages by 48 to 96 hours. Macrophages are key cellular constituents of tissue repair, clearing extracellular debris, fibrin, and other foreign material at the site of repair, and promoting angiogenesis and ECM deposition (Fig. 3-22).

FIGURE 3-22 Multiple roles of macrophages in wound healing. Macrophages participate in wound debridement, have antimicrobial activity, stimulate chemotaxis and the activation of inflammatory cells and fibroblasts, promote angiogenesis, and stimulate matrix remodeling and synthesis. ROS, reactive oxygen species.

Migration of fibroblasts to the site of injury is driven by chemokines, TNF, PDGF, TGF-β, and FGF. Their subsequent proliferation is triggered by multiple growth factors, including PDGF, EGF, TGF-β, FGF, and the cytokines IL-1 and TNF (see Table 3-4). Macrophages are the main source for these factors, although other inflammatory cells and platelets may also produce them. Collagen fibers are now present at the margins of the incision, but at first these are vertically oriented and do not bridge the incision. In 24 to 48 hours, spurs of epithelial cells move from the wound edge (initially with little cell proliferation) along the cut margins of the dermis, depositing basement membrane components as they move. They fuse in the midline beneath the surface scab, producing a thin, continuous epithelial layer that closes the wound. Full epithelialization of the wound surface is much slower in healing by secondary union because the gap to be bridged is much greater. Subsequent epithelial cell proliferation thickens the epidermal layer. Macrophages stimulate fibroblasts to produce FGF-7 (keratinocyte growth factor) and IL-6, which enhance keratinocyte migration and proliferation. Other mediators of re-epithelialization are HGF and HB-EGF.⁸⁹ Signaling through the chemokine receptor CXCR 3 also promotes skin re-epithelialization.

Concurrently with epithelialization, collagen fibrils become more abundant and begin to bridge the incision. At first a provisional matrix containing fibrin, plasma fibronectin, and type III collagen is formed, but this is replaced by a matrix composed primarily of type I collagen. TGF-β is the most important fibrogenic agent (see Table 3-4). It is produced by most of the cells in granulation tissue and causes fibroblast migration and proliferation, increased synthesis of collagen and fibronectin, and decreased degradation of ECM by metalloproteinases. The epidermis recovers its normal thickness and architecture, and surface keratinization.

Scar Formation.

The leukocytic infiltrate, edema, and increased vascularity largely disappear during the second week. Blanching begins, accomplished by the increased accumulation of collagen within the wound area and regression of vascular channels. Ultimately, the original granulation tissue scaffolding is converted into a pale, avascular scar, composed of spindle-shaped fibroblasts, dense collagen, fragments of elastic tissue, and other ECM components. The dermal appendages that have been destroyed in the line of the incision are permanently lost, although in rats new hair follicles may develop in large healing wounds under Wnt stimulation.⁹⁰ This result suggests that, with appropriate treatment procedures, regrowth of skin appendages during wound healing might be achieved in humans. By the end of the first month, the scar is made up of acellular connective tissue devoid of inflammatory infiltrate, covered by intact epidermis.

Wound Contraction.

Wound contraction generally occurs in large surface wounds. The contraction helps to close the wound by decreasing the gap between its dermal edges and by reducing the wound surface area. Hence, it is an important feature in healing by secondary union. The initial steps of wound contraction involve the formation, at the edge of the wound, of a network of myofibroblasts that express smooth muscle α-actin and vimentin. These cells have ultrastructural characteristics of smooth muscle cells, contract in the wound tissue, and may produce large amounts of ECM components (see discussion of hypertrophic scars in this chapter), such as type I collagen, tenascin-C, SPARC, and extra-domain fibronectin.⁹¹ Myofibroblasts are formed from tissue fibroblasts through the effects of PDGF, TGF-β, and FGF-2 released by macrophages at the wound site, but they can also originate from bone marrow precursors known as fibrocytes, or from epithelial cells, through the process of epithelial-to-mesenchymal transition.

Connective Tissue Remodeling.

The replacement of granulation tissue with a scar involves changes in the composition of the ECM. The balance between ECM synthesis and degradation results in remodeling of the connective tissue framework – an important feature of tissue repair. Some of the growth factors that stimulate synthesis of collagen and other connective tissue molecules also modulate the synthesis and activation of metalloproteinases, enzymes that degrade these ECM components.

Degradation of collagen and other ECM proteins is achieved by matrix metalloproteinases (MMPs), a family of enzymes that includes more than 20 members that have in common a 180-residue zinc-protease domain (MMPs should be distinguished from neutrophil elastase, cathepsin G, kinins, plasmin, and other important proteolytic enzymes, which also degrade EMC components and which are serine proteinases, not metalloenzymes). Matrix metalloproteinases include interstitial collagenases (MMP-1, -2, and -3), which cleave the fibrillar collagen types I, II, and III; gelatinases (MMP-2 and 9), which degrade amorphous collagen as well as fibronectin; stromelysins (MMP-3, 10, and 11), which act on a variety of ECM components, including proteoglycans, laminin, fibronectin, and amorphous collagens; and the family of membrane-bound MMPs (ADAMs) described below. MMPs are produced by fibroblasts, macrophages, neutrophils, synovial cells, and some epithelial cells. Their secretion is induced by growth factors (PDGF, FGF), cytokines (IL-1, TNF), and phagocytosis in macrophages, and is inhibited by TGF-β and steroids. Collagenases cleave collagen under physiologic conditions. They are synthesized as a latent precursor (procollagenase) that is activated by chemicals, such as free radicals produced during the oxidative burst of leukocytes, and proteinases (plasmin). Once formed, activated collagenases are rapidly inhibited by a family of specific tissue inhibitors of metalloproteinases, which are produced by most mesenchymal cells, thus preventing uncontrolled action of these proteases. Collagenases and their inhibitors are essential in the debridement of injured sites and in the remodeling of connective tissue necessary to repair the defect.

A large and important family of enzymes related to MMPs is called ADAM (disintegrin and metalloproteinase-domain family). Most ADAMs are anchored by a single transmembrane domain to cell surface. ADAM-17 (also known as TACE, for TNF-converting enzyme) cleaves the membrane-bound precursor forms of TNF and TGF-α, releasing the active molecules. ADAM-17 deficiency in mice causes embryonic or neonatal lethality associated with pulmonary hypoplasia. Members of the ADAM family are also involved in the pathogenesis of bronchial asthma (Chapter 15), and thrombotic microangiopathies (Chapter 13).

Recovery of Tensile Strength.

Fibrillar collagens (mostly type I collagen) form a major portion of the connective tissue in repair sites and are essential for the development of strength in healing wounds. Net collagen accumulation, however, depends not only on increased collagen synthesis but also on decreased degradation. How long does it take for a skin wound to achieve its maximal strength? When sutures are removed from an incisional surgical wound, usually at the end of the first week, wound strength is approximately 10% that of unwounded skin. Wound strength increases rapidly over the next 4 weeks, slows down at approximately the third month after the original incision, and reaches a plateau at about 70% to 80% of the tensile strength of unwounded skin. Lower tensile strength in the healed wound area may persist for life. The recovery of tensile strength results from the excess of collagen synthesis over collagen degradation during the first 2 months of healing, and, at later times, from structural modifications of collagen fibers (cross-linking, increased fiber size) after collagen synthesis ceases.

LOCAL AND SYSTEMIC FACTORS THAT INFLUENCE WOUND HEALING

The adequacy of wound repair may be impaired by systemic and local host factors.

Systemic factors include those listed below:

Local factors that influence healing include those listed here:

PATHOLOGIC ASPECTS OF REPAIR

Complications in wound healing can arise from abnormalities in any of the basic components of the repair process. These aberrations can be grouped into three general categories: (1) deficient scar formation, (2) excessive formation of the repair components, and (3) formation of contractures.

• Excessive formation of the components of the repair process can give rise to hypertrophic scars and keloids. The accumulation of excessive amounts of collagen may give rise to a raised scar known as a hypertrophic scar; if the scar tissue grows beyond the boundaries of the original wound and does not regress, it is called a keloid (Fig. 3-23). Keloid formation seems to be an individual predisposition, and for unknown reasons this aberration is somewhat more common in African Americans. Hyperthrophic scars generally develop after thermal or traumatic injury that involves the deep layers of the dermis. Collagen is produced by myofibroblasts, which persist in the lesion through the autocrine production of TGF-β, and the establishment of focal adhesions.⁹²

FIGURE 3-23 Keloid. A, Excess collagen deposition in the skin forming a raised scar known as keloid. B, Note the thick connective tissue deposition in the dermis.

(A, from Murphy GF, Herzberg AJ: Atlas of Dermatopathology. Philadelphia, WB Saunders, 1996, p 219; B, courtesy of Z. Argenyi, MD, University of Washington, Seattle, WA.)

Exuberant granulation is another deviation in wound healing consisting of the formation of excessive amounts of granulation tissue, which protrudes above the level of the surrounding skin and blocks re-epithelialization (this process has been called, with more literary fervor, proud flesh). Excessive granulation must be removed by cautery or surgical excision to permit restoration of the continuity of the epithelium. Finally (fortunately rarely), incisional scars or traumatic injuries may be followed by exuberant proliferation of fibroblasts and other connective tissue elements that may, in fact, recur after excision. Called desmoids, or aggressive fibromatoses, these lie in the interface between benign proliferations and malignant (though low-grade) tumors. The line between the benign hyperplasias characteristic of repair and neoplasia is frequently finely drawn (Chapter 7).

FIGURE 3-24 Wound contracture. Severe contracture of a wound after deep burn injury.

(From Aarabi S et al: Hypertrophic scar formation following burns and trauma: new approaches to treatment. PLOS Med 4:e234, 2007.)

Fibrosis

Deposition of collagen is part of normal wound healing. However, the term fibrosis is used more broadly to denote the excessive deposition of collagen and other ECM components in a tissue. As already mentioned, the terms scar and fibrosis are used interchangeably, but fibrosis most often indicates the deposition of collagen in chronic diseases. The basic mechanisms that occur in the development of fibrosis associated with chronic inflammatory diseases are generally similar to the mechanisms of skin wound healing that have been discussed in this chapter. However, in contrast to the short-lived stimulus that triggers the orderly steps of wound healing in the skin, the injurious stimulus caused by infections, autoimmune reactions, trauma, and other types of tissue injury persists in chronic diseases, causing organ dysfunction and often organ failure.

The persistence of injury leads to chronic inflammation, which is associated with the proliferation and activation of macrophages and lymphocytes, and the production of a plethora of inflammatory and fibrogenic growth factors and cytokines, mentioned earlier and summarized in Figure 3-25.

FIGURE 3-25 Development of fibrosis in chronic inflammation. The persistent stimulus of chronic inflammation activates macrophages and lymphocytes, leading to the production of growth factors and cytokines, which increase the synthesis of collagen. Deposition of collagen is enhanced by decreased activity of metalloproteinases.

The host response to harmful stimuli is orchestrated to first clear these stimuli and then heal the damage. As we discussed in Chapter 2 (see Fig. 2-10), the initial wave of the host response to external invaders and tissue injury generates “classically activated macrophages”, which are effective in ingesting and destroying microbes and dead tissues. This is followed by the accumulation of “alternatively activated macrophages”, which suppress the microbicidal activities and instead function to remodel tissues and promote angiogenesis and scar formation.⁹³ The cytokines that induce classical macrophage activation are those produced by T_H1 cells, notably IFN-γ and TNF, whereas alternative macrophage activation is best induced by IL-4 and IL-13, cytokines produced by T_H2 cells and other cells including mast cells and eosinophils. Alternatively activated macrophages produce TGF-β and other growth factors that are involved in the repair process.

TGF-β is practically always involved as an important fibrogenic agent (see Table 3-4) in these diseases, regardless of the original cause. It is produced by most of the cells in granulation tissue and causes fibroblast migration and proliferation, increased synthesis of collagen and fibronectin, and decreased degradation of ECM due to inhibition of metalloproteinases. The levels of TGF-β in tissues are not primarily regulated by the transcription of the gene but depend on the post-transcriptional activation of latent TGF-β, the rate of secretion of the active molecule, and factors in the ECM that enhance or diminish TGF-β activity.

The mechanisms that lead to the activation of TGF-β in fibrosis are not precisely known, but cell death by necrosis or apoptosis and the production of reactive oxygen species seem to be important triggers of the activation, regardless of the tissue. Similarly, the cells that produce collagen under TGF-β stimulation may vary depending on the tissue. In most cases, such as in lung and kidney fibrosis, myofibroblasts (already discussed in this chapter) are the main source of collagen, but stellate cells are the major collagen producers in liver cirrhosis.

Recent studies provide evidence for an important role for osteopontin (OPN) in would healing and fibrosis.⁷⁷ OPN is strongly expressed in fibrosis of the heart, lung, liver, kidney and some other tissues. In animal experiments, blockage of OPN expression during wound healing decreases the formation of granulation tissue and scarring.⁹⁴ Although the mechanisms by which OPN promotes fibrosis are not completely understood, recent data show that OPN is a mediator of the differentiation of myofibroblasts induced by TGF-β.

In marked contrast to wounds in adults, fetal cutaneous wounds heal without scar formation.^95,96 Several factors have been proposed to promote scarless healing, including the secretion of non-fibrogenic forms of TGF-β, the lack of osteopontin, and the absence of a T_H2 response, but no definitive results have been obtained. Given the serious organ dysfunction produced by fibrosis, there is a major effort to develop clinically useful antifibrotic agents. Among agents being tested are inhibitors of TGF-β binding or signaling, angiogenesis inhibitors, Toll-like receptors antagonists, and the decoy receptor IL-13Rα2 which blocks IL-13.

Fibrotic disorders include diverse diseases such as liver cirrhosis, systemic sclerosis, fibrosing diseases of the lung (idiopathic pulmonary fibrosis, pneumoconioses, and drug-, radiation-induced pulmonay fibrosis), chronic pancreatitis, glomerulonephritis, and constrictive pericarditis. These conditions are discussed in the appropriate chapters throughout the book.

This concludes the discussion, begun in Chapter 1, of cellular and tissue injury, the inflammatory reaction to such injury (Chapter 2), and tissue healing by regeneration and repair. An overview of the relationships between these processes is illustrated in Fig. 3-26.

FIGURE 3-26 Repair, regeneration, and fibrosis after injury and inflammation.