Cell Biology & Histology

The "three faces" of epithelium are not casual names but territories demarcated by polarity proteins. Step one: the epithelial cell uses three complexes, Par, Crumbs and Scribble, to partition itself into three domains; step two: each face has a different job, so the structures mounted on it differ; step three: the free (apical) face fronts the lumen and handles absorption (microvilli), beating (cilia) and sensation (stereocilia); step four: the lateral face is in charge of "binding to the neighbors", and every cell junction hangs here; step five: the basal face is in charge of "rooting downward", with hemidesmosomes and the basement membrane anchoring the epithelium to connective tissue. So whenever a question asks on which face the microvilli, the hemidesmosomes or the tight junctions sit, think of the three words "outside, neighbor, foundation" and the answer leaps out.

Telling the four junctions apart relies not on rote memorization of names but on looking first at "what work it does (sealing / adhesion / communication)" and then at "which cytoskeleton it connects to". ① The tight junction (zonula occludens) seals the intercellular gap shut with the proteins claudin and occludin, maintaining polarity and the barrier — it connects to no cytoskeleton, because its job is to plug the seam, not to resist pulling; ② the adherens junction (zonula adherens, the adhesion "belt") encircles the cell like a belt and connects to microfilaments (actin filaments), since actin itself runs in a belt-like distribution; ③ the desmosome (macula adherens, the adhesion "spot") resembles individual point-like buttons and connects to intermediate filaments (IF, chiefly keratin), bearing mechanical tension — which is why the epidermis, a tissue constantly tugged and stretched, has the most desmosomes; ④ the gap junction is built from connexin proteins into channels that allow direct cell-to-cell communication, through which ions and small molecules pass; ⑤ the hemidesmosome, on the basal face, also connects to intermediate filaments and anchors the epithelium to the basement membrane. Memory hook: D for desmosome brings to mind a nail Driven Deep (the original mnemonic pairs D with 釘, "nail" — intermediate filaments resist pulling like steel nails), and A for adherens brings to mind Actin (microfilaments). Asked which junction connects to intermediate filaments, answer desmosome and hemidesmosome; asked about the adhesion spot, it is the desmosome; asked about the adhesion belt, it is the adherens junction (connecting to actin).

The epidermis is a keratinized stratified squamous epithelium, and from bottom to top its layers run stratum basale, stratum spinosum, stratum granulosum, (stratum lucidum) and stratum corneum. The stratum lucidum is seen only in thick skin (palms and soles) and is absent from thin skin, because the keratinization route in thick skin is longer and needs the lucidum as a transitional stage. Each of the four cell types of the epidermis has its own address, and the address is dictated by function: ① melanocytes live in the stratum basale, because they must deliver melanosomes through their dendrites to every keratinocyte above them, and dispatch from the bottom layer is the most efficient (neural crest origin); ② Merkel cells also live in the stratum basale, because they are tactile mechanoreceptors that must connect to sensory nerve endings in the dermis, and the lowest layer puts them closest to the nerves; ③ Langerhans cells live in the stratum spinosum, because they are patrolling immune dendritic cells that must intercept, in the middle of the epidermis, antigens that have crossed the stratum corneum (bone marrow origin); ④ keratinocytes fill every layer and form the bulk of the tissue. Memory hook: Merkel minds "touch" (the original pairs Merkel with 摸 mō, "to touch" — think M for Mechanoreceptor), at the very bottom, wired to the nerves; Langerhans minds "immunity" (Langer → Lymph), patrolling the stratum spinosum in the middle.

The difference between apoptosis and necrosis goes beyond "active vs passive" — they are two entirely different causal chains. The chain of apoptosis: ① the program is triggered (the intrinsic mitochondrial pathway or the extrinsic death-receptor pathway) → ② the caspase cascade cleaves specific proteins → ③ a nuclear endonuclease cuts the DNA neatly between nucleosomes → producing a ladder pattern → ④ the cell shrinks and packages its organelles in membrane as apoptotic bodies → ⑤ macrophages engulf them cleanly, and no inflammation is provoked. The entire process requires ATP (because proteins must be actively cleaved and the remains packaged). The chain of necrosis runs the other way: ① an irreversible insult (ischemia, toxin) cuts off the ATP supply → ② the membrane Na⁺/K⁺ pump stalls, and sodium and water flood in → ③ the cell swells and the membrane ruptures → ④ DNA is randomly degraded into a smear → ⑤ intracellular enzymes and DAMPs (damage-associated molecular patterns) leak out → provoking inflammation. Hence apoptosis = ladder DNA + no inflammation + ATP required, and necrosis = swelling + membrane rupture + inflammation + no ATP required. Do not memorize the two directions in reverse.

How cartilage grows is decided entirely by a single causal chain: whether or not it possesses a perichondrium. ① The perichondrium is a thin layer wrapped around the outside of the cartilage, composed of dense connective tissue together with undifferentiated chondrogenic precursor cells → ② these precursor cells differentiate into chondrocytes at the surface and lay down new matrix, which constitutes appositional growth → ③ the chondrocytes within the lacunae divide on their own and expand the tissue from within, which constitutes interstitial growth → ④ hence cartilage "without a perichondrium" has only one route left: interstitial growth. Hyaline cartilage (type II collagen: articular surfaces, trachea, costal cartilage, epiphyseal plate) has a perichondrium, so both modes are available; elastic cartilage (elastic fibers plus type II collagen: auricle, epiglottis) has a perichondrium, so both modes are available; fibrocartilage (predominantly type I collagen: intervertebral discs, pubic symphysis, menisci) has no perichondrium, and therefore can grow only interstitially — this is the fundamental reason it repairs slowly and a torn meniscus does not readily grow back.

Appositional growth is carried out by the cells of the inner layer of the perichondrium adding new layers at the surface, so it requires a perichondrium; interstitial growth is carried out by the chondrocytes within the lacunae dividing and expanding the tissue from within. Think this causal chain through, and when a question asks "which cartilages are capable of appositional growth," you will rule out fibrocartilage on your own — it has no perichondrium and cannot add layers. Another frequently examined detail: fibrocartilage has lacunae, whereas dense regular connective tissue does not — because the fibroblasts of dense connective tissue are not trapped in lacunae and can migrate between the fiber bundles; chondrocytes, by contrast, are "locked in" by the matrix they themselves secrete, and so they reside in lacunae.

Bone is a dynamic tissue: ① the bone matrix consists of an organic component (type I collagen + osteoid) and an inorganic component (hydroxyapatite, a calcium phosphate salt), the organic part providing toughness and the inorganic part providing hardness → ② three cell types divide the labor into a three-shift system of "build – maintain – demolish" → ③ osteoblasts differentiate from mesenchymal stem cells and are responsible for building bone (secreting osteoid); → ④ osteocytes are osteoblasts trapped by the matrix they themselves secreted, living in lacunae, responsible for maintaining the matrix and for mechanosensation (mechanical strain sensors), and they have the longest lifespan; → ⑤ osteoclasts come from the monocyte/macrophage lineage (hematopoietic origin), fuse into multinucleated cells, and are responsible for resorbing bone (secreting hydrogen ions H⁺ and the protease cathepsin K to dissolve it), and because their workload is heavy, they have the shortest lifespan. Be careful not to memorize the directions backwards: osteocytes live the longest, osteoclasts the shortest; osteoclasts come from the hematopoietic system, not from mesenchyme — this is because they must fuse into multinucleated giant cells, which requires the fusion capacity of the hematopoietic lineage.

Intramembranous ossification: mesenchyme forms bone directly, with no cartilage template — the representatives are flat bones such as the skull and the clavicle. Endochondral ossification: a hyaline cartilage template forms first and is then ossified — the representatives are the long bones, the limbs, and the vertebral column. A long bone has two ossification centers, the primary ossification center in the diaphysis and the secondary ossification center in the epiphysis. Why is it that bone itself cannot grow interstitially? A three-step causal chain: ① the bone matrix has already been calcified by hydroxyapatite and is as hard as stone → ② the osteocytes are trapped in their lacunae and cannot divide → ③ therefore bone can only "add layers" from the surface (appositional growth) and cannot expand from within. A long bone "lengthens" through interstitial growth of the cartilage of the epiphyseal plate followed by ossification (growth in height ceases once the plate closes at puberty), and a long bone "thickens" through appositional (circumferential) bone formation by the periosteum. Mnemonic: intramembranous ossification is like laying bricks directly (skull, clavicle); endochondral ossification is like building a wooden mold first and then pouring in the concrete (long bones).

White adipose tissue: a single large lipid droplet (unilocular), the nucleus squeezed to the periphery, few mitochondria, responsible for energy storage, insulation, and protection; brown adipose tissue: multiple small lipid droplets (multilocular), abundant mitochondria — hence brown to the naked eye — responsible for non-shivering thermogenesis. The thermogenic mechanism is a four-step chain: ① the sympathetic nerves release norepinephrine, which activates brown fat → ② lipolysis yields fatty acids that enter the mitochondria for β-oxidation, pumping protons into the intermembrane space → ③ normally the protons should return to the matrix through ATP synthase and drive ATP production; but the inner mitochondrial membrane of brown fat carries UCP-1 (uncoupling protein-1, thermogenin), which opens a bypass channel → ④ the protons leak straight back into the matrix without passing through ATP synthase, and the energy of oxidation is converted into heat rather than ATP. A neonate cannot yet generate heat by muscular shivering the way an adult can, and so relies on the furnace of brown fat to maintain body temperature, which is why neonates have abundant and widely distributed brown fat (interscapular, cervical), whereas it gradually diminishes in adults.

Distinguishing the three muscle types takes only four questions, and behind every answer lies a developmental or functional reason. ① Are there striations? Striations arise from the orderly arrangement of sarcomeres: skeletal and cardiac muscle are arranged in register and are therefore striated, whereas the filaments of smooth muscle run obliquely and therefore show none. ② How many nuclei, and where? Skeletal muscle forms by the fusion of many myoblasts, so it is multinucleated with the nuclei pushed to the periphery; cardiac and smooth muscle each develop from a single cell, so each is mononucleated with a central nucleus. ③ Voluntary or involuntary? Skeletal muscle is controlled by somatic nerves and is therefore voluntary; cardiac and smooth muscle are under autonomic control and are therefore involuntary. Follow the chain through: skeletal muscle = striated, multinucleated, peripheral nuclei, voluntary; cardiac muscle = striated, mononucleated (occasionally binucleated), central nucleus, involuntary; smooth muscle = non-striated, mononucleated, central nucleus, involuntary. So "smooth muscle is multinucleated" is wrong — multinucleation belongs to skeletal muscle, and examiners love to flip this direction on you.

The sarcomere is the segment between two Z lines and is the basic unit of contraction. Each of the three bands has its own meaning: the I band contains only thin filaments (actin), with a Z line at its center; the A band spans the full length of the thick filaments (myosin), including the region of overlap with actin; the H zone is the central part of the A band that contains myosin alone. The causal chain of the sliding filament theory runs in three steps: ① the filaments themselves never change length from start to finish (myosin stays just as long, and so does actin) → ② it is actin that slides inward along myosin, enlarging the zone of overlap → ③ therefore the segments where actin and myosin do not overlap (the I band and H zone) shorten, while the A band, being precisely the full length of myosin, stays the same. The classic test point: the A band stays constant while the I band and H zone shorten.

The cross-bridge cycle requires two things, each playing a different role. Ca²⁺ binds to troponin C, pushing aside the tropomyosin that blocks the actin and exposing the myosin-binding sites — this is the "unlocking" step. ATP in fact has two roles: ① before hydrolysis, ATP binds to the myosin head and detaches it from actin (release); ② after hydrolysis, the myosin head is reset (re-cocked) into its high-energy state, ready for the next power stroke. Hence without ATP the myosin head stays clamped to actin and cannot let go — this is rigor mortis: after death ATP is exhausted, the bridges cannot release, and the muscles stiffen. A question that says only that ATP "supplies the energy for the power stroke" has missed half the story — its principal function is actually to let the two filaments separate.

Excitation-contraction coupling in skeletal muscle is a five-step chain: ① the action potential propagates along the sarcolemma → ② it dives into the T-tubule → ③ the dihydropyridine receptor (DHPR) on the T-tubule membrane acts like a mechanical switch, directly tugging on the ryanodine receptor (RyR) of the adjacent sarcoplasmic reticulum → ④ RyR opens and Ca²⁺ floods out of the sarcoplasmic reticulum (SR) into the cytosol → ⑤ Ca²⁺ binds troponin and sets the cross-bridge cycle in motion. In skeletal muscle the T-tubule and the two flanking SR terminal cisternae form a triad, located at the A-I junction; in cardiac muscle the arrangement is a dyad, with the T-tubule at the Z line. Why does smooth muscle have no T-tubules? A three-step causal chain: ① smooth muscle cells are small, with a high surface-area-to-volume ratio → ② the invaginated pits called caveolae, together with a modest amount of SR, suffice to spread the Ca²⁺ signal throughout → ③ moreover, a large share of smooth muscle's Ca²⁺ enters from outside the cell through voltage- or receptor-gated channels, so there is simply no need for T-tubules reaching deep into the cell. Hence "smooth muscle lacks T-tubules and relies on caveolae and vesicles to transport Ca²⁺" is a correct statement.

Normal transmission across the neuromuscular junction (NMJ) is a five-step chain: ① an action potential reaches the motor neuron terminal → ② presynaptic voltage-gated Ca²⁺ channels open → ③ Ca²⁺ influx triggers the release of acetylcholine (ACh) vesicles → ④ ACh diffuses to the nicotinic ACh receptors (nicotinic AChR) on the postsynaptic sarcolemma → ⑤ the receptors open, sodium flows in, the muscle depolarizes, and contraction follows; acetylcholinesterase (AChE) immediately breaks down ACh to terminate the signal. In the receptor zone the sarcolemma folds inward to form junctional folds that increase receptor density, and these folds are formed by the infolding of the sarcolemma itself — not of the sarcoplasmic reticulum. Four frequently tested clinical conditions each sever the chain at a different point: in myasthenia gravis (MG), antibodies attack the nAChR at step ⑤, the receptors are occupied and the muscle side cannot hear the signal, so the patient grows weaker with use, often in association with thymoma; in Lambert-Eaton myasthenic syndrome (LEMS), antibodies attack the presynaptic Ca²⁺ channels at step ②, so ACh cannot be released at first, but with repetitive stimulation presynaptic Ca²⁺ accumulates and ACh is gradually released after all, so the patient grows stronger with use — a paraneoplastic phenomenon commonly seen with small cell lung cancer; botulinum toxin cleaves the SNARE proteins at step ③ and inhibits ACh release, producing flaccid paralysis; organophosphate poisoning inhibits AChE so that ACh accumulates, producing a cholinergic crisis.