8.2 Cell Death Pathways: Necrosis, Apoptosis, Autophagy & Mitochondrial Dysfunction

Why cell-death classification is a Domain II.5 skill

Handbook II.5 asks you to distinguish direct versus indirect action, primary versus secondary effects, and on-target versus off-target effects, and to follow a toxicant from a molecular interaction to downstream effects in individuals and populations (II.5 C). Independent OpenExamPrep teaching in this section uses necrosis, apoptosis, necroptosis, autophagy, and mitochondrial permeability transition as the cellular language for those distinctions. The section is original study material. It is not an ABT product and does not claim official approval, review, or partnership with ABT.

If you cannot say whether a cell swelled and burst or packaged itself into apoptotic bodies, you will mis-read inflammation, mis-time a sampling window, and mis-attribute a fetal finding to a receptor that was never engaged.

Oncotic necrosis versus apoptosis

Oncotic necrosis (accidental, lytic necrosis) is what happens when ATP falls below the level needed to run Na+/K+-ATPase. Sodium and water enter, the cell swells (oncosis), blebs rupture, and damage-associated molecular patterns (DAMPs) such as HMGB1 recruit inflammation. The nucleus typically undergoes karyolysis (fading of chromatin) in a dirty, inflammatory field. Plasma-membrane failure is early. Leakage enzymes in chapter 5 (ALT, GLDH, CK) are the clinical-pathology face of that membrane failure.

Apoptosis is an ATP-consuming, caspase-driven dismantling program. The cell shrinks, chromatin condenses (pyknosis), the nucleus fragments (karyorrhexis) into apoptotic bodies, and phosphatidylserine exposure invites phagocytosis with little neutrophilic inflammation if clearance is efficient. Initiator caspases are caspase-9 (intrinsic pathway: cytochrome c + Apaf-1 apoptosome) and caspase-8 (extrinsic pathway: death receptors Fas, TRAIL, TNFR1). Executioner caspases (caspase-3 and caspase-7) cleave ICAD, structural proteins, and PARP (the 89 kDa fragment is a practical blot marker). Cleaved caspase-3 immunohistochemistry supports apoptosis when the nucleus looks pyknotic, not when the field is a sheet of swollen, ruptured cells.

TUNEL (terminal deoxynucleotidyl transferase dUTP nick-end labeling) tags free 3′-OH DNA ends. Apoptotic internucleosomal cleavage gives a strong TUNEL signal, but necrotic random DNA degradation is also TUNEL-positive. TUNEL without pyknosis or apoptotic bodies, and without caspase cleavage, is not an apoptosis diagnosis. A DNA ladder on a gel is similarly more persuasive with morphology than alone.

Pyknosis versus karyolysis is the nuclear contrast the handbook-level item is after. Pyknosis is condensed, hyperchromatic chromatin typical of apoptosis (and of some early necrotic nuclei before they fade). Karyolysis is loss of nuclear basophilia as DNases digest chromatin in a lytic cell—the necrotic stamp. Do not memorize the words as decorations; pair them with ATP and inflammation.

Necroptosis and autophagy flux

Necroptosis is programmed necrosis. When caspase-8 is inhibited or absent, RIPK1 and RIPK3 form the necrosome and phosphorylate MLKL, which oligomerizes at the plasma membrane and lyses the cell. Morphology is necrotic (swell, rupture, inflammation); biochemistry is a pathway you can interrupt with necrostatins or MLKL loss. Death-receptor ligands can therefore produce apoptosis or necroptosis depending on caspase-8 competence—an on-pathway fork, not two unrelated toxicants.

Macroautophagy sequesters cytoplasm in autophagosomes that fuse with lysosomes. LC3-I lipidation to LC3-II marks autophagosome membranes. p62/SQSTM1 is degraded when cargo actually reaches the lysosome. An LC3-II increase can mean more flux or blocked degradation. The discriminating experiment is a flux assay: bafilomycin A1 or chloroquine blocks lysosomal acidification or fusion; if LC3-II rises further from a low-p62 baseline, flux was running; if LC3-II was already high and p62 accumulated before the block, flux was stalled. Autophagy is usually cytoprotective (including mitophagy of damaged mitochondria). Calling every LC3-II blot “autophagic cell death” is a category error. True autophagic death is uncommon and needs flux-plus-rescue evidence (blocking autophagy saves the cell).

Starvation, mTOR inhibition, and ER stress induce autophagy as physiology. A high-dose dam that stops eating will show hepatic autophagy that is secondary to reduced food consumption, not proof the test article’s primary receptor is an autophagy protein—the same II.5 logic you will use for fetal ossification.

Mitochondria: MPT pore, ATP, and calcium

The mitochondrial permeability transition (MPT) pore opens when matrix Ca2+ is high, ROS are high, and adenine nucleotides are low. Cyclophilin D is the genetically best-supported regulator; cyclosporin A can delay opening in some models. Open MPT collapses Δψm, uncouples oxidative phosphorylation, swells the matrix, and dumps cytochrome c and NAD+. If residual ATP still supports apoptosome assembly, the cell may die by apoptosis. If ATP collapses completely, the same mitochondrion produces oncotic necrosis. That is why one hepatotoxicant can look apoptotic at the lesion rim and necrotic in the core.

Calcium is both signal and wrecking tool. ER Ca2+ release and plasma-membrane leak load mitochondria through the mitochondrial calcium uniporter. Ca2+-dependent phospholipases, calpains, and endonucleases digest membranes and chromatin. Chelation or MCU modulation can shift death mode without changing the parent toxicant’s primary receptor. Section 8.1’s lipid peroxidation and iron-catalyzed hydroxyl radical are common openers of this pore; they are not a second, unrelated chapter.

Direct, on-target, off-target, and secondary to maternal toxicity (II.5)

Direct / on-target toxicity is the primary molecular interaction occurring where you think it occurs: a glycine-receptor antagonist causing spinal disinhibition; a kinase inhibitor occupying its intended kinase in a proliferating epithelium; doxorubicin engaging TOP2B in cardiomyocytes. Off-target is binding or inhibiting a different protein at a relevant concentration: the same kinase inhibitor blocking a cardiac kinase it was not designed to hit; an hERG block that is not the therapeutic target. Indirect / secondary toxicity is downstream of a different organ or physiologic state. The handbook’s own examples are delayed ossification due to maternal toxicity, reduced fetal body weight, decreased maternal food consumption, and stress. Those fetal findings can be real, dose-related, and statistically significant and still not prove a direct fetal osteoblast or growth-plate receptor interaction.

Sort them on a study with three questions:

1. Is the molecular target expressed and engaged in the tissue that shows the lesion (occupancy, pathway marker, right cell type)?
2. Does the finding persist when the primary exaggerated pharmacology or the maternal stressor is removed (metabolite block, pair-feeding, anti-stress husbandry, a maternally tolerated dose)?
3. Is there a coherent dose–response at the receptor, or only a coherent dose–response of maternal wasting?

ICH S5 and OECD 414 interpretation practice the same split: delayed ossification and reduced fetal weight that appear only beside marked maternal body-weight loss are often secondary. They still get tabulated. They do not automatically become a fetal on-target MOA.

Receptor-level events versus downstream individual and population effects

II.5 C is the translation sentence. A receptor event is measured in a synapse, hepatocyte, or binding assay. The individual adverse outcome is the clinical or anatomic result in that animal (convulsions, liver failure, fetal death). Population effects are incidence, herd productivity, or wildlife recruitment—the same molecular interaction after ecology, behavior, and demography have acted. Do not jump from a Ki to a population decline without the intermediate key events section 8.3 names.

Strychnine is taught here as an original receptor-level example, not as a recycled board stem. The alkaloid is a competitive antagonist at the inhibitory glycine receptor (GlyR), a ligand-gated chloride channel on spinal and brainstem motor circuits, including Renshaw-cell synapses. Blocking glycine removes postsynaptic inhibition of motor neurons. The individual outcome is tetanic extensor spasms, opisthotonos, and death from respiratory-muscle tetany or exhaustion. That is direct, on-target, receptor-level toxicity. Secondary rhabdomyolysis, hyperthermia, and lactic acidosis are downstream of the convulsions, not a second primary receptor. A developmental cohort that showed reduced fetal weight only in dams that convulsed and stopped eating, with pair-fed non-convulsing controls matching the fetal-weight loss, would be secondary to maternal toxicity, not evidence that fetal growth-plate glycine receptors were the target.

Off-target can sit on the same molecule. A drug intended for a peripheral kinase that also antagonizes GlyR at the Cmax you actually achieve is off-target convulsant activity. The convulsions are still receptor-level and direct on the nervous system; they are off-target relative to the intended pharmacology.

Putting death mode next to II.5

On-target mitochondrial uncoupling that drains ATP produces necrosis in the target organ—direct. Off-target hERG block produces arrhythmia; if the animal collapses and the liver shows ischemic necrosis, the hepatic necrosis is secondary. Maternal stress glucocorticoids can produce thymic apoptosis that looks immunotoxic until you check body weight, food, and recovery. Autophagy induction in a starved dam is physiology. Name the initiating interaction and the death mode as two sentences, not one blur.

Scenario

A kinase inhibitor causes fetal delayed ossification at doses that drop maternal body weight 15% and food intake 30%, with no ossification delay at a maternally tolerated dose that still occupies the primary kinase in fetal tissue. That package is secondary to maternal toxicity until a direct fetal MOA is shown. A second compound occupies an off-target retinoic-acid receptor in the fetus at maternally tolerated doses and produces the same bone delay—off-target, direct on the fetus. A third compound is a glycine-receptor antagonist: dams convulse, fetuses are small, and pair-fed non-convulsing controls have the same fetal-weight loss—weight is secondary; the convulsions are on-target.

High-dose liver: cores of karyolytic, swollen hepatocytes with ALT in the thousands; lesion rims with pyknosis and cleaved caspase-3. That is an ATP gradient across one lesion, not two unrelated mechanisms.

Traps

• Diagnosing apoptosis from TUNEL alone.
• Equating LC3-II increase with completed autophagy or with death.
• Calling every fetal ossification delay a direct developmental toxicant.
• Treating secondary ischemic necrosis as proof the liver receptor was the drug target.
• Skipping from a binding assay to a population claim without individual key events.
• Treating strychnine’s rhabdomyolysis as a second primary muscle receptor rather than convulsion downstream injury.