AAOS Comprehensive Orthopaedic Review

Section 1 - Basic Science

Chapter 5. Bone and Joint Biology

I. Bone




1. Functions of bone—The unique composition and structure of bone enables this tissue to accomplish the following:


a. Provide mechanical support


b. Regulate mineral homeostasis


c. House the marrow elements


2. Types of bones—long, short, and flat


3. Formation of bones


a. Long bones are formed via endochondral ossification, which is formation of bone from a cartilage model.


b. Flat bones are formed by intramembranous bone formation, which is the formation of bone through loose condensations of mesenchymal tissue.





1. Long bones are composed of three anatomic regions: the diaphysis, the metaphysis, and the epiphysis (

Figure 1).


a. Diaphysis—The shaft of a long bone, consisting of a tube of thick cortical bone surrounding a thin central canal of trabecular bone (the intramedullary canal).


i. The inner aspect of the cortical bone is called the endosteal surface.


ii. The outer region is called the periosteal surface. This surface is covered by the periosteal membrane, which is composed of an outer layer of fibrous connective tissue and an inner layer of undifferentiated, osteogenic progenitor cells.


   *John C. Clohisy, MD, or the department with which he is affiliated has received research or institutional support from Wright Medical Technology and Zimmer and is a consultant or employee for Zimmer. Dieter Lindskog, MD, is a consultant or employee for Arthrocare.


b. Metaphysis—Transition zone from epiphysis to diaphysis; composed of loose trabecular bone surrounded by a thin layer of cortical bone.


c. Epiphysis—Specialized end of bone that forms the joint articulation.


i. The growth plate (physis or physeal scar) divides the epiphysis from the metaphysis.


ii. The epiphysis is composed of loose trabecular bone surrounded by a thin layer of cortical bone.


iii. The articular portion of the bone has a specialized subchondral region underlying the articular cartilage.


2. Flat bones


a. Flat bones include the pelvis, scapula, skull, and mandible.


b. The composition of these bones varies from purely cortical in some regions to cortical bone with a thin inner region of trabecular bone.


3. Neurovascular anatomy of bone


a. Innervation—The nerves that innervate bone derive from the periosteum and enter bone in tandem with blood vessels. Nerves are found in the haversian canals and Volkmann canals (

Figure 2).


b. Blood supply


i. Nutrient arteries pass through the diaphyseal cortex and enter the intramedullary canal. These vessels provide blood supply to the inner two thirds of the cortical bone and are at risk during intramedullary reaming.


ii. The outer one third of the cortical bone derives its blood supply from the periosteal membrane vessels. These vessels are at risk with periosteal stripping during surgical procedures.


B. Structure

1. Macroscopic level


a. Cortical bone—A dense, compact bone with low porosity and no macroscopic spaces.


[Figure 1. Schematic diagram of cortical and trabecular bone showing the different structures and cell types. 1 = osteoclasts, 2 = osteoblasts, 3 = bone lining cells, 4 = osteocytes, 5 = marrow space.]

i. In the diaphyseal region, cortical bone is load bearing.


ii. In the metaphysis and epiphysis, cortical bone serves as a border to trabecular bone. It supports only a portion of the load, which is primarily carried by the trabecular bone in these regions.


b. Trabecular bone—Bone composed of a loose network of bony struts (rods and plates). These struts have a maximum thickness of approximately 200 μm.


i. Trabecular bone is porous, with a macroscopic porosity ranging from 30% to 90%, and houses the bone marrow contents.


ii. In osteoporosis, the macroscopic porosity is increased because of thinning of the trabecular struts.


2. Microscopic level


a. Woven bone is primary bone that is characterized by random orientation of collagen and mineral.


b. Lamellar bone is secondary bone that results from the remodeling of woven bone into an organized bone tissue.


c. Lacunae are ellipsoidal spaces in bone that are occupied by osteocytes. Small channels through the bone called canaliculi connect the lacunae and contain osteocyte cell processes that interact with other cells.



Composition of the extracellular matrix (ECM)—The ECM is composed of 60% to 70% mineral matrix and 20% to 25% organic matrix.


[Figure 2. Diagram of the structure of cortical bone, showing the types of cortical lamellar bone: the internal circumferential system, interstitial system, osteonal lamellae, and outer circumferential system. The diagram also shows the intraosseous vascular system that serves the osteocytes and connects the periosteal and medullary blood vessels. The haversian canals run primarily longitudinally through the cortex, whereas the Volkmann canals create oblique connections between the haversian canals. Cement lines separate each osteon from the surrounding bone. Periosteum covers the external surface of the bone and consists of two layers: an osteogenic (inner) cellular layer and a fibrous (outer) layer.]

1. Mineral matrix


a. Responsible for the compressive strength of bone


b. Composed primarily of calcium and phosphate (but also some sodium, magnesium, and carbonate) in the form of hydroxyapatite and tricalcium phosphate


c. The mineral component of bone is closely associated with collagen fibrils.


d. Tropocollagen helices in the fibrils are organized in a quarter-staggered arrangement, with empty regions (hole zones) between the ends and pores running lengthwise between collagen fibrils (

Figure 3).


e. Mineral crystals form in the hole zones and pores.


2. Organic matrix—90% type I collagen; 5% other collagen types (III and IV), noncollagenous proteins, and growth factors; the remaining tissue volume is occupied by water.


a. Collagen


i. Type I collagen is the primary ECM protein of bone.


ii. Type I collagen is fibril-forming, with a triple helical structure (three α chains) that contributes tensile strength to the ECM.


iii. Fibrils are intrinsically stable because of noncovalent interconnections and covalent cross-links between lysine residues.


[Figure 3. Diagram describing mineral accretion.]

iv. Small amounts of types III and IV collagen also are present in bone.


b. Noncollagenous ECM proteins


i. Vitamin K-dependent proteins—Osteocalcin is the most common vitamin K-dependent, noncollagenous protein in bone and is a marker of osteoblast differentiation. Osteocalcin undergoes carboxylation in a vitamin K-dependent manner.


ii. Adhesive proteins—Facilitate the interaction of cells (attachment and detachment) with the ECM via cell surface receptors called integrins. Fibronectin and vitronectin are common adhesive proteins of bone.


iii. Matricellular proteins—Proteins that mediate cell-matrix interactions by modulating signaling from the matrix to the cell.


iv. Phosphoproteins—Phosphorylated (negatively charged) extracellular proteins that interact with calcium and are thought to play a role in mineralization.


v. Growth factors and cytokines—Biologically active proteins that are potent regulators of differentiation and activation. These include bone morphogenetic proteins (BMPs), transforming growth factor-β (TGF-β), basic fibroblast growth factor (bFGF), insulin growth factors (IGFs), and interleukins (ILs).


vi. Proteoglycans—Molecules composed of protein core and glycosaminoglycan side chains. Proteoglycans provide tissue structure, bind to growth factors, regulate proliferation, and act as cell surface receptors.



Composition of bone cells—The cells associated with the bone ECM include osteoblasts, osteocytes, and osteoclasts. Cells of the marrow and periosteum also contribute significantly to the process of bone remodeling.


1. Osteoblasts—Bone surface cells that form bone matrix and regulate osteoclast activity.


a. Marker proteins include alkaline phosphatase, osteocalcin, osteonectin, and osteopontin.


b. Osteoblasts have parathyroid hormone (PTH) receptors and secrete type I collagen.


c. Differentiation


i. Osteoblasts arise from marrow stromal cells and periosteal membrane cells. A series of cellular regulators serve as differentiation cues for osteoblast development from stem cell to mature osteoblast/osteocyte (

Figure 4).


ii. Cells committed to osteoblastic differentiation are called osteoprogenitor cells.


iii. Each stage of differentiation has characteristic molecular markers, transcription factors, and secreted proteins.


iv. Runx2 and osterix are essential transcription factors required for osteoblast cell function.


v. The mature osteoblast has a lifespan of 100 days. It can then become a bone lining cell or an osteocyte, or it can undergo apoptosis. Bone lining cells are relatively inactive cells that cover the surfaces of bone. These cells likely have the ability to become reactivated as functional osteoblasts.


2. Osteocytes


a. Active osteoblasts become embedded in the mineralized matrix and become osteocytes.


b. Osteocytes reside in the lacunar spaces of trabecular and cortical bone. They are nonmitotic and are not highly synthetic.


c. Distinct from the osteoblast, they do not express alkaline phosphatase.


d. Osteocytes have numerous cell processes that communicate with other cells via the canaliculi.


e. Signaling between osteocytes is mediated by protein complexes called gap junctions.


f. Osteocytes contribute to regulation of bone homeostasis.


3. Osteoclasts—Multinucleated bone-resorbing cells.



Marker proteins include tartrate-resistant acid


[Figure 4. Osteoblast differentiation. This idealized depiction of the osteoblast developmental lineage illustrates the key concepts of early proliferation versus terminal phenotypic differentiation, the temporal onset of molecular markers, and important regulators of this process as well as the different fates possible for cells of the osteoblastic lineage. CBFA1 = core binding factor alpha 1, ALP = alkaline phosphatase, BSP = bone sialoprotein, PGE2 = prostaglandin E2.]


phosphatase (TRAP), calcitonin receptor, and cathepsin-K.


Differentiation—Osteoclasts are hematopoietic cells, members of the monocyte/macrophage lineage. The multinuclear osteoclast polykaryons form by fusion of mononuclear precursors, a process that requires receptor activator for nuclear factor κ B ligand (RANKL) and macrophage-colony stimulating factor (MCSF).


Activity and important features—Mature osteoclasts attach to bone/mineral surfaces and form a sealing zone underneath the cells. The plasma membrane underneath the cell forms the resorptive domain of the cell, which features a highly convoluted ruffled border. Proteases and ions are secreted through this domain to dissolve both organic and nonorganic material.


Regulation—Differentiation and activity of osteoclasts is regulated primarily by RANKL and osteoprotegerin (OPG). RANKL binds to its cognate receptor, RANK, on the membrane of monocyte/macrophage. OPG is a decoy receptor (member of the tumor necrosis factor [TNF] receptor family) that binds to and sequesters RANKL, thus inhibiting osteoclast differentiation and activity (

Figure 5).



Bone homeostasis—Balanced bone formation and resorption.


1. Remodeling


a. Bone is a dynamic tissue that is constantly undergoing remodeling, primarily through osteoblasts (bone-forming cells) and osteoclasts (resorptive cells) (Figure 5).


[Figure 5. Schematic representation of osteoclast differentiation and function regulated by RANKL and M-CSF. Osteoclast progenitors and mature osteoclasts express RANK, the receptor for RANKL. Osteotropic factors such as 1α,25(OH)2D3, PTH, and IL-1 stimulate expression of RANKL in osteoblasts/stromal cells. Membrane- or matrix-associated forms of both M-CSF and RANKL expressed by osteoblasts/stromal cells are responsible for the induction of osteoclast differentiation in the co-culture. RANKL also directly stimulates fusion and activation of osteoclasts. Mainly osteoblasts/stromal cells produce OPG, a soluble decoy receptor of RANKL. OPG strongly inhibits the entire differentiation, fusion, and activation processes of osteoclast induced by RANKL.]


Figure 6. An illustration of a bone marrow unit showing the various stages of cellular activity that it passes through temporally from the resorption of old bone by osteoclasts and the subsequent formation of new bone by osteoblasts. For simplicity, the illustration shows remodeling in only two dimensions, whereas in vivo it occurs in three dimensions, with osteoclasts continuing to enlarge the cavity at one end and osteoblasts beginning to fill it in at the other end. OC = osteoclast, OB = osteoblast.]

b. The regulatory mechanisms of remodeling are critical to the understanding of bone homeostasis and disease states.


c. An individual's bone mass is "turned over" completely every 4 to 20 years, depending on age. At adulthood, the rate of turnover is 5% per year. This process of bone turnover replaces potentially compromised bone with structurally sound bone.


2. Trabecular bone remodeling (Figure 6)


a. Osteoclastic activation leads to development of a resorption pit called a Howship lacuna.


b. After pit formation, osteoclasts are replaced by osteoblasts that form new bone matrix.


c. The cement line is the region where bone resorption stopped and new bone formation begins.


d. After new bone formation is completed, bone lining cells cover the surface.


3. Cortical bone remodeling (

Figure 7)


a. Osteoclasts tunnel through bone to form a cutting cone of resorption.


b. Blood vessel formation occurs in the cutting cone.


c. Osteoblast recruitment and new bone formation occur in the resorbed space of the cutting cone.


d. This results in circumferential new bone formation around a blood vessel. This structure is called an osteon, and the vessel space is the haversian canal (

Figure 8).


4. Mechanisms of osteoblast/osteoclast coupling


a. The biologic activity of osteoblasts is closely associated with that of osteoclasts, and intercellular signaling mechanisms are being studied.


b. Osteoblastic regulation of osteoclast function has been well documented. PTH is a pro-osteoclastogenic cytokine that acts through osteoblast cell-surface receptors. These receptors stimulate the synthesis of factors, including RANKL and M-CSF, that are critical to osteoclast development.


c. In addition to secreting pro-osteoclastogenic RANKL, osteoblasts also can produce OPG, a potent anit-osteoclastogenic protein. Therefore, osteoblasts have positive and negative regulatory effects on osteoclast activity.


d. Osteoclast activity is also regulated by systemic factors like serum calcium levels and circulating hormones.


i. Vitamin D and PTH stimulate osteoclastic activity.


ii. Calcitonin decreases osteoclastic activity.


e. Osteoclast regulation of osteoblast differentiation and activity is less understood. One working


[Figure 7. Diagram showing a longitudinal section through a cortical remodeling unit with corresponding transverse sections below. A—Multinucleated osteoclasts in Howship lacuna advancing longitudinally from right to left and radially to enlarge a resorption cavity. B—Perivascular spindle-shaped precursor cells. C—Capillary loop delivering osteoclast precursors and pericytes. D—Mononuclear cells (osteoblast progenitors) lining reversal zone. E—Osteoblasts apposing bone centripetally in radial closure and its perivascular precursor cells. F—Flattened cells lining the haversian canal of completed haversian system or osteon. Transverse sections at different stages of development: (I) resorptive cavities lined with osteoclasts; (II) completed resorption cavities lined by mononuclear cells, the reversal zone; (III) forming haversian sytem or osteons lined with osteoblasts that had recently apposed three lamellae; and (IV) completed haversian sytem or osteon with flattened bone cells lining canal. Cement line (G); osteoid (stippled) between osteoblast (O) and mineralized bone.]

[Figure 8. Electron photomicrographs of cortical bone. A, A thin-ground cross section of human cortical bone in which osteocyte lacunae (arrows) and canaliculi have been stained with India ink. Osteocytes are arranged around a central vascular channel to constitute haversian systems. Active haversian systems (1, 2, and 3) have concentric lamellae in this plane. Older haversian systems (4, 5, and 6) have had parts of their original territories invaded and remodeled. This is seen most clearly where 2 and 3 have invaded the territory originally occupied by 5. (Original magnification: X185.) B, Higher magnification of part of a haversian system showing the successive layering (numbers) of osteocytes (large arrows) from the central core (H) that contains the vasculature. Small arrows identify the canaliculi that connect osteocyte lacunae in different layers. (Original magnification: X718.)]




Table 1. Characteristics of Various Disease States]

   hypothesis is that osteoclastic bone resorption releases bioactive factors (BMP, TGF-β, IGF-1) that stimulate osteoblast differentiation and new bone formation.


f. The process of bone remodeling is abnormal in disease states (eg, osteoporosis and osteopetrosis), and therapies are directed at correcting the remodeling abnormalities.



Disease states


1. Characteristics (Table 1)


2. Therapies


a. Bisphosphonates—Inhibit osteoclastic bone resorption; used to treat osteoporosis, bone metastasis, and Paget disease.


b. Intermittent PTH dosing—Used to stimulate bone formation (continuous dosing stimulates bone resorption).


c. OPG and anti-RANKL antibodies—Potential use as antiresorptive agents for various bone loss disorders (not available for clinical use at press time).


d. Corticosteroids—Decrease bone formation and increase bone resorption. Osteopenia is a common side effect of chronic steroid use.



Injury and repair (fracture)


1. Injury


a. Bone injury can be caused by trauma or surgical osteotomy.


b. Injury disrupts the vascular supply to the affected tissue, leading to mechanical instability, hypoxia, depletion of nutrients, and elevated inflammatory response.


2. Repair


a. Unlike tissues that repair by the development of scar tissue, bone heals with the formation of new bone that is indistinguishable from the original tissue.


b. Motion at the fracture site (cast, external fixator, intramedullary rod) results in healing, primarily through endochondral ossification, whereas rigidity at the fracture site (plate fixation) enables direct intramembranous ossification. Most fractures heal with a combination of these bone repair processes.


3. Repair stages


a. Hematoma and inflammatory response—Macrophages and degranulating platelets infiltrate the fracture site and secrete various inflammatory cytokines, including platelet-derived growth factor (PDGF), TGF-β, IL-1 and IL-6, prostaglandin E2 (PGE2), and TNF-α. These factors impact a variety of cells in the fracture hematoma microenvironment.


b. Early postfracture


i. Periosteal preosteoblasts and local osteoblasts form new bone.


ii. Mesenchymal cells and fibroblasts proliferate and are associated with the expression of basic and acidic fibroblast growth factors. Primitive mesenchymal and osteoprogenitor cells are associated with expression of the BMPs and TGF-β family of proteins.


c. Fracture hematoma maturation


i. The fracture hematoma produces a collagenous matrix and network of new blood vessels. Neovascularization provides progenitor cells and growth factors for mesenchymal cell differentiation.


ii. Cartilage formation (endochondral ossification), identified by expression of collagen types I and II, stabilizes the fracture site. Chondrocytes proliferate, hypertrophy, and express factors that stimulate ossification.


d. Conversion of hypertrophic cartilage to bone—A complex process in which hypertrophic chondrocytes undergo terminal differentiation, cartilage calcifies, and new woven bone is formed.


i. A variety of factors are expressed as hypertrophic cartilage is replaced by bone. These include BMPs, TGF-β, IGFs, osteocalcin, and collagen I, V, and XI.


ii. Hypertrophic chondrocyte apoptosis and vascular invasion ensue.


e. Bone remodeling


i. The newly formed woven bone is remodeled through coordinated osteoblast and osteoclast functions.


ii. Mature bone is eventually established and is not distinguishable from the surrounding bone. Mature bone contains a host of growth factors, including TGF-β, BMPs, and IGFs.

II. Synovial Joints

A. Overview


1. Synovial joints are specialized structures that allow movement at bony articulations.


a. The structure is composed of a joint cavity lined by synovium and containing bones lined with articular cartilage.


b. Joints are stabilized by ligaments and motored by tendon attachments from adjacent musculature (

Figure 9).


B. Formation—The formation and development of synovial joints is a poorly understood process.


1. Limb skeletogenesis starts with long uninterrupted condensations of mesenchymal tissue.


2. Condensations of mesenchymal cells form at specific locations. This appears to be controlled by the HOX family of genes.


3. Apoptosis then occurs within the so-called inter-zone, and the tissues separate through cavitation.


4. Joint-specific development then ensues through a control mechanism not yet understood.


C. Structure


1. Anatomy—The particular anatomy of each joint varies according to the location and demands of motion placed on the joint. Joint structure ranges from highly matched bony surfaces, such as the ball-and-socket hip joint, to the less congruent shoulder joint, which allows greater range of motion but provides less stability.


2. Structural components


a. Articular cartilage—Highly specialized tissue allowing low-friction movement.


b. Ligament—Collagenous structure connecting articulating bones; provides stability and restraint to nonphysiologic motion.


c. Joint capsule—Tough, fibrous tissue surrounding the joint cavity.


[Figure 9. Schematic illustration of a synovial joint.]

d. Synovium—This tissue, which lines the noncartilaginous portions of the joint cavity, is composed of two distinct layers, the intimal lining and the connective tissue sublining.


i. The intimal lining, which is only a few cells thick and is in direct contact with the joint cavity, produces the synovial fluid. The intimal layer functions as a porous barrier and lacks tight junctions between cells; it has no true basement membrane. This layer is composed of type A and type B cells. Type A cells, which make up only 10% to 20% of the synovial cells, derive from bone marrow precursors and function as tissue macrophages. Type B cells are from the fibroblast lineage. These cells produce hyaluronin and contain a unique enzyme, uridine diphosphoglucose dehydrogenase, which is critical in the pathway for synthesis of hyaluronin.


ii. The sublining is relatively acellular and is composed of fibroblasts, fat, blood vessels, and lymphoid cells. A rich vascular network supplies the sublining and allows for the high solute and gas exchange that is needed to supply the cartilage with nutrition.


e. Synovial fluid


i. Produced and regulated by the synovium


ii. This is an ultrafiltrate of plasma with a low albumin concentration (45% compared to plasma) and high concentration of hyaluronic acid and lubricin.


D. Sensory innervation—Composed of two systems.


1. Fast-conducting myelinated type A fibers, found in the joint capsule and surrounding musculature, produce information on joint positions and motion.


2. Slow-conducting unmyelinated type C fibers are found along blood vessels in the synovium and transmit diffuse pain sensations.


E. Function


1. The synovial joint allows extremely low friction motion between articulating bones.


2. Its function depends on the specific nature of the anatomic makeup of the joints, as well as the specific tissue characteristics of the tissue.

III. Nonsynovial Joints

A. Nonsynovial joints lack a synovial lining bordering the joint cavity and do not allow for low-friction or large-range movements. Different kinds of nonsynovial joints are found throughout the body, including symphyses, syndchondroses, and syndesmoses.


B. Symphyses


1. In this type of joint, bone ends are separated by a fibrocartilaginous disk and are attached with well-developed ligamentous structures that control movement.


2. Intervertebral disks form a symphysis between vertebral bodies.


3. The pubic symphysis occurs at the anterior articulation between each hemipelvis and is composed of articular cartilage-covered rami separated by a fibrocartilage disk with firm ligamentous support. This joint is optimized for stability and load transmission but allows only limited motion.


C. Synchondroses


1. In this type of joint, bone ends are covered with articular cartilage but no synovium is present and no significant motion occurs.


2. Examples include the sternomanubrial joint, rib costal cartilage, and several articulations within the skull base.


D. Syndesmoses


1. This type of joint consists of two bones that articulate without a cartilaginous interface and have strong ligamentous restraints that allow limited motion.


2. The distal tibia-fibula syndesmosis is the only extracranial syndesmosis.


Gamble JG, Simmons SC, Freedman M: The symphysis pubis. Clin Orthop Relat Res 1986;203:261-272.

Miller JD, McCreadie BR, Alford AI, Hankenson KD, Goldstein SA: Form and function of bone, in Einhorn TA, O'Keefe RJ, Buckwalter JA (eds): Orthopaedic Basic Science, ed 3. Rosemont, IL, American Academy of Orthopaedic Surgeons, 2007, pp 129-160.

Pacifici M, Koyama E, Iwamoto M: Mechanisms of synovial joint and articular cartilage formation: Recent advances, but many lingering mysteries. Birth Defects Research 2005;75: 237-248.


Top Testing Facts

1. Endochondral bone formation (long and short bones) occurs through a cartilage model; intramembranous bone formation (flat bones) results from condensations of mesenchymal tissue.


2. The inner two thirds of cortical bone is vascularized by nutrient arteries that pass through the diaphyseal cortex and enter the intramedullary canal and are at risk during intramedullary reaming. The outer one third of the cortical bone derives blood supply from the periosteal membrane vessels. These vessels are at risk with periosteal stripping during surgical procedures.


3. The extracellular matrix of bone is composed of 60% to 70% mineral components and 20% to 25% organic components. The organic matrix is 90% type I collagen and 5% noncollagenous proteins.


4. Type I collagen is fibril-forming and has a triple helical structure (three α chains). The fibrils are intrinsically stable because of noncovalent interconnections and covalent cross-links between lysine residues.


5. Mature osteoblast marker proteins include alkaline phosphatase, osteocalcin, osteonectin, and osteopontin. The potential fates of a mature osteoblast include differentiation into an osteocyte or bone lining cell, or apoptosis.


6. The marker proteins for osteoclasts include TRAP, calcitonin receptor, and cathepsin-K. Osteoclast differentiation and activity are regulated in large part by the bioactive factors RANKL (positive regulator) and OPG (negative regulator).


7. Osteoblast and osteoclast functions are coupled via various systemic and local factors. Regulatory proteins (RANKL and OPG) secreted by osteoblasts provide direct coupling in bone remodeling.


8. Fractures commonly heal with a combination of endochondral and intramembranous bone formation. Motion at the fracture site results in healing primarily through endochondral ossification, whereas stability at the fracture site enables direct intramembranous ossification.


9. Fracture healing includes a sequence of biologic stages including injury, inflammation, hematoma maturation, hypertrophic cartilage formation, new bone formation, and remodeling to mature bone.


10. Articular joint synovium is composed of two layers—the intimal lining, which contains tissue macrophage-like cells and fibroblast-like cells that produce hyaluronin, and the connective tissue sublining.