21. The Immune System

Innate and Adaptive Defense Against Disease

Learning Objectives

21.1 The Three Lines of Defense, and What "Immunity" Means

You are, from a microbe's point of view, an unusually warm, wet, nutrient-rich, and stable habitat. Roughly 38 trillion bacteria already live on and in you. The remarkable fact is not that people occasionally get infected; it is that they usually do not.

That outcome is produced by three lines of defense, and the conventional split between them is real and worth holding onto.

The first two lines together constitute innate (nonspecific) immunity: present from birth, fast, treating broad classes of threat identically, and forgetful — the thousandth exposure is handled exactly like the first. The third is adaptive (specific) immunity: slow to start, exquisitely specific to one molecular target, systemic in reach, and equipped with memory.

Why two systems, and not one

Ask why a body would carry both, since each is expensive.

Innate immunity is fast because its receptors are germline-encoded: about a hundred kinds of pattern recognition receptor, inherited, already made, already deployed, recognizing molecular signatures that whole classes of microbes cannot easily abandon — bacterial lipopolysaccharide, peptidoglycan, flagellin, unmethylated CpG DNA, fungal glucans, viral double-stranded RNA. There is no waiting because there is no learning. The cost is that a system with a hundred receptors cannot distinguish two strains of the same organism, and cannot improve.

Adaptive immunity is slow because its receptors are somatically generated: assembled at random in each individual lymphocyte, producing on the order of 10¹¹ distinct specificities across the population, of which the few that fit any given threat must be found and multiplied. That takes about a week. What it buys is recognition of targets that have never existed before — a virus that mutated last month, a molecule no ancestor ever met — plus a memory that makes the second encounter unrecognizably easier.

Against an organism that divides every twenty minutes and mutates as it goes, fast-and-general is not enough. Against an organism you will meet once and never again, slow-and-specific arrives too late. Carrying both is not redundancy. It is coverage of two different failure modes.

Innate immunity instructs adaptive immunity

The two systems are not parallel and independent. This is the single most important structural idea in the chapter, and it will recur in every section:

Innate immunity decides whether an adaptive response happens at all, and what kind.

Three mechanisms enforce this, and you will meet each in turn:

• Antigen presentation. A dendritic cell that has engaged its pattern recognition receptors migrates to a lymph node and presents antigen. One that has not, does not. The tissue's innate alarm is what puts the antigen in front of a lymphocyte at all (§21.8).
• Co-stimulation. A naive T cell requires two signals: antigen, and co-stimulation from a presenting cell that has itself been innately activated. Antigen without co-stimulation produces anergy — unresponsiveness — not activation. This is a safety interlock against responding to harmless material (§21.8).
• Cytokine context. The cytokines the innate system releases determine which kind of helper T cell response develops, and therefore whether the body mounts an antibody response, a macrophage-activating response, or a barrier-defending one.

The traffic runs the other way too. Antibody made by the adaptive system activates the classical complement pathway, an innate mechanism; interferon gamma from T cells is the principal activator of macrophages. Each system amplifies the other. But the decision to escalate is innate, and that is why a purified protein vaccine does not work without an adjuvant, and why a sterile injury — Amara's infarct — can produce a full inflammatory response with nothing foreign anywhere.

     INNATE vs ADAPTIVE — SPEED, SPECIFICITY, AND MEMORY

  response
  magnitude
     ▲
     │                                    ╭──────╮  ADAPTIVE
     │                                   ╱        ╲  (2nd exposure:
     │                                  ╱          ╲  see Fig 21.5)
     │                    ADAPTIVE     ╱            ╲
     │                   (1st exp.)   ╱              ╲
     │                        ╭──────╯                ╲
     │   INNATE              ╱                         ╲
     │  ╭────────────╮      ╱                           ╲
     │ ╱              ╲____╱                             ╲___
     │╱                                                       
     └──┬────┬────┬────┬────┬────┬────┬────┬────┬────┬────► time
        0    4h   12h  1d   2d   4d   6d   8d  10d  14d

     ├──── INNATE ────┤├──────── ADAPTIVE ────────────┤
     │  barriers      ││  antigen presentation (d1-2) │
     │  complement    ││  clonal selection    (d2-4)  │
     │  phagocytes    ││  proliferation       (d3-5)  │
     │  NK cells      ││  effector output     (d5-10) │
     │  inflammation  ││  MEMORY               (life)  │
     │  fever         │└──────────────────────────────┘
     └────────────────┘

  ┌──────────────┬────────────────────┬─────────────────────┐
  │              │  INNATE            │  ADAPTIVE           │
  ├──────────────┼────────────────────┼─────────────────────┤
  │ Onset        │  seconds–hours     │  6–10 days (1st)    │
  │ Specificity  │  classes (PAMPs)   │  ONE epitope        │
  │ Receptors    │  germline-encoded, │  somatically        │
  │              │  ~100 kinds        │  rearranged, ~10^11 │
  │ Memory       │  none (trained     │  YES — decades      │
  │              │  immunity aside)   │                     │
  │ Distribution │  local + systemic  │  systemic           │
  │ Key cells    │  neutrophil, macro-│  B cell, T cell     │
  │              │  phage, NK, mast   │                     │
  │ Instructs?   │  INSTRUCTS adaptive│  amplifies innate   │
  │              │  (Ag presentation) │  (antibody→C1, IFNγ)│
  └──────────────┴────────────────────┴─────────────────────┘

Where this chapter happens

Almost every event described below takes place on the architecture built in Chapter 20. Dendritic cells capture antigen in tissue and enter a lymphatic capillary through its minivalves. Lymph carries them to a lymph node, where flow is deliberately decelerated so that capture probability rises, and where antigen arriving in lymph meets lymphocytes arriving from blood through high endothelial venules. B cells respond in follicles and form germinal centres in the cortex; T cells are engaged in the paracortex; plasma cells accumulate in the medullary cords and secrete antibody into efferent lymph. Blood-borne antigen is handled in splenic white pulp instead, and mucosal antigen in MALT.

If any of that is unfamiliar, read §20.6 before continuing. This chapter is the physiology of that anatomy.

21.2 Innate Defenses I — The Surface Barriers

The skin and mucous membranes are the only line of defense that most pathogens ever encounter, and their success is invisible precisely because it is complete. Barriers do not generate symptoms, produce lab values, or appear in case reports. They simply prevent.

Physical barriers

• Keratinized stratified squamous epithelium of the epidermis (Chapter 5) is a dry, many-layered, continuously shed sheet of dead, keratin-filled cells. Very few organisms can penetrate intact skin. Note the three properties doing the work: dry (few pathogens grow without water), layered (penetration requires crossing many cells), and shedding (about a million cells an hour, taking adherent organisms with them). A barrier that renews itself cannot be permanently colonized.
• Mucous membranes line every cavity open to the exterior — respiratory, digestive, urinary, reproductive. They are not keratinized and must stay moist to function, so they are inherently more vulnerable, and they compensate chemically and mechanically.

Mechanical clearance

• Mucus traps particles; the mucociliary escalator of the respiratory tract sweeps the trapped load upward toward the pharynx at about 1 cm/min (Chapter 22).
• Flushing: tears, saliva, and urine mechanically wash surfaces. The 1–2 L of daily urine flow is a major reason the urinary tract stays sterile, and interruption of it — a catheter, an obstructing stone, an enlarged prostate — is one of the most reliable ways to produce infection.
• Coughing, sneezing, vomiting, diarrhoea are all expulsive reflexes, and suppressing them entirely is not always the right move.

Chemical defenses

This is where the barrier stops being passive and becomes actively hostile.

Normal flora deserves emphasis, because it is a defense mechanism made of other organisms. Roughly 38 trillion bacteria live on and in you, and they defend their territory against newcomers by occupying attachment sites, consuming available nutrients, and secreting inhibitory compounds. Broad-spectrum antibiotics eliminate this defense along with the target, which is why Clostridioides difficile colitis and vaginal candidiasis follow antibiotic courses. You did not acquire a new pathogen; you removed the incumbents.

21.3 Innate Defenses II — Phagocytes and Natural Killer Cells

Once an organism crosses the barrier, the second line engages. It has four components — phagocytes, natural killer cells, inflammation, and antimicrobial proteins — plus fever, which is systemic. This section covers the cells; §21.4 covers inflammation, and §21.5 the proteins and fever.

Phagocytes

Neutrophils are the most abundant white cell (40–60% of circulating leukocytes) and the first responders. They are short-lived — hours to a few days — arrive in enormous numbers, kill aggressively, and die at the site, becoming the principal cellular component of pus. They cannot divide and cannot present antigen. They are ammunition.

Macrophages are monocytes that have left the blood and matured in tissue. They are long-lived, can divide, phagocytose repeatedly, secrete more than a hundred signalling molecules, and — critically — present antigen to T cells, linking innate to adaptive immunity. Tissue macrophages carry local names: Kupffer cells in liver, microglia in the CNS, alveolar macrophages in lung, osteoclasts (a related lineage) in bone.

The division of labour is worth stating plainly, because it explains the timeline of every inflammatory response you will ever see: neutrophils kill and die; macrophages kill, decide, and rebuild.

Phagocytosis proceeds in five steps:

1. Adherence. The phagocyte must bind the particle. Its receptors recognize broad molecular patterns — PAMPs, pathogen-associated molecular patterns such as bacterial lipopolysaccharide, peptidoglycan, flagellin, and unmethylated CpG DNA — using pattern recognition receptors such as the Toll-like receptors. These patterns are shared by whole classes of microbes and absent from host cells, which is how a system with no specificity manages not to eat you.
2. Ingestion. Pseudopods extend around the particle and fuse, enclosing it in a phagosome.
3. Fusion. The phagosome fuses with a lysosome to form a phagolysosome.
4. Killing and digestion. Two mechanisms operate together. The respiratory burst uses NADPH oxidase to generate superoxide, hydrogen peroxide, and — in neutrophils, via myeloperoxidase — hypochlorite, the active ingredient of household bleach, manufactured on site. Alongside it, lysosomal enzymes, defensins, and lactoferrin (which sequesters the iron bacteria need) do chemical work.
5. Exocytosis of indigestible residue.

Opsonization transforms step 1. Some organisms — the encapsulated ones of §20.6, and anything with a slippery surface — adhere poorly. Coating them with antibody or with the complement fragment C3b provides handles: phagocytes carry receptors specifically for the Fc region of antibody and for C3b. Opsonization can increase phagocytic efficiency by more than a thousandfold. The word means "to prepare for eating," which is exactly right.

Natural killer cells

NK cells are large granular lymphocytes that belong to the innate system. They are not antigen-specific, require no prior exposure, and are ready immediately.

Their targeting logic is the elegant part, and it is the opposite of everything you will learn in §21.8. A cytotoxic T cell kills when it recognizes a specific foreign peptide presented on MHC class I. An NK cell kills when MHC class I is missing — the "missing self" hypothesis. NK cells carry inhibitory receptors that are engaged by normal levels of self MHC class I, and engagement says do not kill. Downregulate MHC class I and that restraint is released.

Why does this matter? Because the commonest way for a virus or a tumour to evade cytotoxic T cells is to stop displaying MHC class I. Doing so makes the cell invisible to T cells and, in the same stroke, maximally visible to NK cells. The two systems close each other's escape hatch. It is one of the most satisfying pieces of design in immunology, and it is worth noticing that neither system had to be told about the other: they simply key on opposite states of the same molecule.

NK cells kill using perforin and granzymes (§21.8), and they also secrete large amounts of interferon gamma, which activates macrophages — an innate cell amplifying other innate cells. They also carry Fc receptors, which lets antibody direct them onto specific targets (antibody-dependent cell-mediated cytotoxicity, §21.7) — an adaptive molecule aiming an innate weapon.

21.4 Inflammation in Full

Inflammation is the innate response of vascularized tissue to injury of any kind: infection, trauma, burn, chemical, radiation, foreign body, ischaemia, or immune attack.

Read that sentence again, because it contains the answer to Amara's first question. It is triggered by tissue damage, not by infection specifically. Nothing in the initiating machinery asks whether a microbe is present.

Inflammation accomplishes four things: it prevents the spread of damaging agents, it disposes of debris and pathogens, it alerts the adaptive immune system, and it sets the stage for repair. The last of those four is the one that generates Amara's scar, and it is the subject of §21.10.

The four cardinal signs, each derived

The cardinal signs were described by Celsus around 30 CE. They have survived two thousand years because each is the visible consequence of one specific vascular change.

• Redness (rubor) and heat (calor). Chemical mediators cause arteriolar vasodilation, increasing blood flow into the injured area — hyperaemia. More blood in dilated superficial vessels means more red colour and more delivered core heat. Both signs are the same mechanism seen two ways. The increased flow is also functional: it delivers more leukocytes, more complement, more antibody, more oxygen, and more clotting factors per minute.
• Swelling (tumor). Mediators make the postcapillary venules leaky by causing endothelial cells to contract and pull apart at their junctions. Two things then follow. Fluid filtration rises, because the increased flow raises capillary hydrostatic pressure. And protein, including fibrinogen and antibody, escapes into the interstitium, raising OP_if and pulling still more fluid out. The result is an exudate — protein-rich, unlike the protein-poor transudate of heart failure (§20.8). The exudate is not a side effect: it dilutes toxins, delivers antibody and complement into the tissue, and brings fibrinogen, which clots to form a physical scaffold that walls off the area and gives repair cells something to crawl on.
• Pain (dolor). Three contributions. Bradykinin and prostaglandins directly stimulate and sensitize nociceptors; prostaglandin E₂ in particular lowers the firing threshold so that normally innocuous stimuli hurt. Released potassium and hydrogen ions from damaged cells depolarize nerve endings. And pressure from the accumulating exudate mechanically distorts them. Pain enforces rest of the injured part, which is genuinely protective.
• Loss of function (functio laesa), the fifth sign added later, follows from pain and swelling.

The mediators

Cyclooxygenase is the target of aspirin and the NSAIDs; blocking it removes prostaglandin synthesis and therefore reduces pain, fever, and some vasodilation, which is why one class of drug does all three. Antihistamines block only histamine, which is why they work well for allergy — where mast cell histamine is the dominant mediator — and poorly for the pain of a sprained ankle, where kinins and prostaglandins dominate. Matching the drug to the dominant mediator is the whole of anti-inflammatory pharmacology, and it requires knowing which mediator is doing what.

Leukocyte recruitment, in four steps

This is the mechanism by which a cell in the bloodstream ends up in the tissue, and it is one of the most beautifully staged processes in physiology — a sequence of progressively stronger adhesions, each of which must succeed before the next is attempted.

1. Leukocytosis. Within hours, IL-1 and TNF-α induce release of neutrophils from bone marrow reserves. The circulating count rises — often to 11,000–20,000/µL. This is the number on Amara's chart.
2. Margination. Inflamed endothelium expresses selectins, which bind carbohydrate ligands on neutrophils weakly and transiently. The neutrophil stops tumbling in midstream and begins rolling along the vessel wall — the bonds form and break faster than the cell can be swept past. Chemokines displayed on the endothelial surface then activate neutrophil integrins, which bind endothelial ICAM firmly, and the cell arrests.
3. Diapedesis. The arrested neutrophil flattens and squeezes between endothelial cells, through the junction, and crosses the basement membrane into the tissue.
4. Chemotaxis. Now in the interstitium, it migrates up a concentration gradient of chemical attractants — C5a, bacterial formylated peptides, chemokines, leukotriene B₄ — to the source.

The staging matters because each step is a checkpoint. A cell that rolls but never receives a chemokine signal continues on its way; only tissue that is actively signalling gets cells delivered. Genetic loss of the integrin subunit CD18 produces leukocyte adhesion deficiency: neutrophils are present in the blood in large numbers, roll normally, and cannot arrest — so pus never forms, wounds do not heal, and the umbilical cord separates late. The disease isolates step 2 exactly as chronic granulomatous disease isolates step 4.

Neutrophils arrive first, within an hour, and dominate the first day. Monocytes follow over 8–24 hours, mature into macrophages in the tissue, and become dominant after roughly 48 hours. Macrophages are the cells that clean up, and they are also the cells that decide what happens next — which is where §21.10 goes.

        THE INFLAMMATORY RESPONSE — INJURY TO RESOLUTION

  ①  INJURY / PATHOGEN                     time 0
     tissue cells damaged → release DAMPs (ATP, DNA, HMGB1, uric acid)
     pathogens present → PAMPs (LPS, peptidoglycan, flagellin)
              │
              ▼  both bind PATTERN RECOGNITION RECEPTORS
  ②  SENTINEL CELLS ACTIVATE                seconds–minutes
     mast cells degranulate  →  HISTAMINE
     resident macrophages    →  IL-1, TNF-α, IL-6, chemokines
     plasma cascades fire    →  kinins, complement (C3a, C5a)
              │
      ┌───────┴────────┐
      ▼                ▼
  ③ VASODILATION   ④ ↑ PERMEABILITY          minutes
    (arterioles)      (venules; endothelial cells contract)
    blood flow ↑↑     fluid + PROTEIN out  →  EXUDATE
      │                    │
      ▼                    ▼
   REDNESS + HEAT       SWELLING ──► + kinins/PGs ──► PAIN
      │                    │                            │
      └──────┬─────────────┴────────────────────────────┘
             ▼
  ⑤  LEUKOCYTE RECRUITMENT                   1–24 hours
     LEUKOCYTOSIS  → MARGINATION (selectins: rolling)
                   → firm adhesion (integrins/ICAM)
                   → DIAPEDESIS (squeeze through junction)
                   → CHEMOTAXIS (up the C5a / chemokine gradient)
             │
             ▼
  ⑥  PHAGOCYTOSIS + KILLING                  hours–days
     neutrophils first (peak day 1) → macrophages (peak day 2–3)
     debris + dead neutrophils + fluid = PUS
             │
             ▼
  ⑦  THE FORK — determined by macrophage phenotype
       ┌─────────────────────────┬──────────────────────────┐
       ▼                         ▼                          ▼
   RESOLUTION              REPAIR by FIBROSIS        CHRONIC
   stimulus gone,          parenchyma cannot         inflammation
   debris cleared,         regenerate, or damage     stimulus persists
   architecture            was extensive             → granuloma,
   RESTORED                → COLLAGEN SCAR             ongoing damage
   (skin abrasion)         (MYOCARDIAL INFARCT)        (atherosclerosis)

Resolution is an active program

Resolution was long assumed to be passive — the mediators simply run out. It is not.

Resolution is an active, programmed shutdown. Macrophages switch phenotype from pro-inflammatory to reparative; they clear apoptotic neutrophils, a process called efferocytosis that itself signals the switch; and a family of specialized pro-resolving mediators — resolvins, protectins, lipoxins, derived from omega-3 and omega-6 fatty acids — actively terminates the response. There is even a deliberate lipid mediator class switch partway through an inflammatory episode, in which the same arachidonic acid pathway stops making pro-inflammatory prostaglandins and starts making pro-resolving lipoxins.

Inflammation has an off switch. That single fact reframes chronic inflammation: it is often the off switch failing rather than the on switch sticking, and it is why "anti-inflammatory" and "pro-resolving" are not synonyms.

21.5 Antimicrobial Proteins and Fever

Complement

Complement is a system of about 30 plasma proteins, made mostly by the liver, circulating in inactive form. Activation is a cascade: each protein cleaves the next, so a handful of initiating events produces an enormous, rapid, amplified response — the same design principle as the clotting cascade of Chapter 17, and for the same reason. A cascade converts a small, localized signal into a large, fast, spatially confined response, and it can be regulated at every step.

Three activation pathways converge on a single step: the cleavage of C3 into C3a and C3b.

The alternative pathway deserves a second look, because it answers a question that ought to be difficult: how does a system with no specificity avoid attacking you?

C3 hydrolyses spontaneously in plasma at a low rate, all the time, and the resulting C3b sticks to whatever surface is nearby. On your cells, membrane regulatory proteins — CD59, decay-accelerating factor, factor H — immediately inactivate it. On a bacterial surface, which lacks these regulators, C3b persists and initiates the cascade.

Discrimination is achieved not by recognizing the enemy but by protecting the self. That is a completely different strategy, and a robust one: it requires no knowledge of what the pathogen looks like, and it cannot be defeated by a pathogen changing its surface — only by one that acquires host regulators, which some, tellingly, have evolved to do.

Three outcomes follow from activation:

1. Cell lysis. C3b drives assembly of C5b through C9 into the membrane attack complex (MAC) — a ring of C9 molecules that inserts into the membrane and forms an open pore about 10 nm across. Water and ions rush in down their gradients; the cell swells and bursts. This matters particularly against Neisseria species, which is why people deficient in the late complement components (C5–C9) get recurrent meningococcal disease and essentially nothing else.
2. Opsonization. C3b coats the microbial surface and is bound by complement receptors on phagocytes. Quantitatively this is complement's most important function — more so than lysis.
3. Inflammation. C3a and C5a are anaphylatoxins: they trigger mast cell degranulation, causing vasodilation and increased permeability. C5a is additionally the most potent chemotactic factor known for neutrophils.

Interferons

Interferons are the antiviral arm, and their logic is unusual: a cell that is already lost spends its last resources warning its neighbours.

A virus-infected cell detecting viral double-stranded RNA secretes interferon alpha and beta. These bind receptors on neighbouring cells and induce them to synthesize antiviral proteins — notably protein kinase R and 2'-5' oligoadenylate synthetase — that lie dormant until viral RNA appears, at which point they shut down protein synthesis and degrade viral RNA. The neighbours are placed in an antiviral state before infection reaches them.

Two features make this powerful. It is a warning shout, so protection spreads faster than the virus can. And because the interferon response is not virus-specific, it works against viruses the body has never met — innate immunity's characteristic trade of precision for breadth.

Interferon gamma is a different molecule with a different job: made by T cells and NK cells, it is chiefly a macrophage activator, dramatically increasing killing capacity. It also increases MHC expression, which raises the visibility of infected cells to T cells.

Fever

Fever is an abnormally high body temperature, and it is not a malfunction. It is a deliberately relocated set point, exactly as described in Chapter 1 and Chapter 25.

The mechanism. Macrophages exposed to PAMPs or DAMPs secrete pyrogenic cytokines — IL-1, IL-6, TNF-α — which reach the hypothalamus and induce prostaglandin E₂ synthesis in the vascular organ of the lamina terminalis. PGE₂ raises the hypothalamic thermostat setting from 37.0 °C to, say, 38.5 °C. Every subsequent event is normal negative feedback against the new target: cutaneous vasoconstriction, shivering, seeking warmth, and the subjective sensation of cold. The patient shivers while febrile because relative to a 38.5 °C set point, a 37.5 °C body is too cold.

Why raise it at all? Moderate fever appears to be adaptive on several counts: many bacterial and viral pathogens replicate less efficiently above 37 °C; the liver and spleen sequester iron and zinc, starving bacteria of trace metals they cannot synthesize; and the rate of phagocytosis, lymphocyte proliferation, and essentially the whole immune response accelerates with temperature. Very high fever — above about 41 °C — becomes dangerous because host enzymes and membranes fail too, and above 43 °C it is generally fatal. The set point is normally moved a little, not a lot, which is itself evidence of regulation rather than of failure.

Antipyretics work by inhibiting cyclooxygenase, blocking PGE₂ synthesis and returning the set point to 37 °C. Note what happens next: the body, now at 38.5 °C with a 37 °C target, treats itself as overheated and responds with vasodilation and profuse sweating. That is the drenching sweat of a breaking fever — not the infection resolving, but the thermostat moving back.

Amara's 37.9 °C is a low-grade fever produced by IL-1, IL-6, and TNF-α released by macrophages clearing her infarct. There is no organism. The macrophages cannot tell the difference, and that is the entire point of question 1.

21.6 Adaptive Immunity — The Framework

Innate immunity is fast, broad, and forgetful. Against an organism that replicates every twenty minutes and mutates as it goes, that is often not enough. Adaptive immunity is the answer, and it buys its power with three properties the innate system lacks.

• Specificity. An adaptive response is directed at one molecular target, not at a class.
• Systemic action. It is not confined to the site of infection; antibodies and effector cells circulate everywhere.
• Memory. A second exposure produces a faster, stronger, and qualitatively better response — the property on which all vaccination depends.

The cost is time. A primary adaptive response takes 6–10 days to become effective, which is why the innate system must hold the line in the meantime, and why the two are best understood as one integrated defense operating on different timescales (Figure 21.1).

Antigens, epitopes, and haptens

An antigen is any substance capable of provoking an immune response and of binding to the products of that response. Most are large (above ~10 kDa), complex, and foreign: proteins are the strongest antigens, followed by large polysaccharides, then nucleic acids and lipids.

A lymphocyte does not recognize a whole antigen. It recognizes an epitope (antigenic determinant) — a small surface patch, typically 6–15 amino acids or sugar residues. A single protein carries many distinct epitopes, which is why one antigen elicits many different antibodies, and why a polyclonal response is the normal outcome.

A hapten is a small molecule too small to be antigenic alone that becomes antigenic when it binds covalently to a body protein. The immune system then responds to the hapten–protein combination — and, unhelpfully, to the hapten wherever it appears. This is the mechanism behind poison ivy (urushiol), nickel allergy, and penicillin allergy: penicillin itself is 334 daltons and invisible to the immune system until it haptenizes serum proteins.

Self-antigens are the molecules on your own cells. The most important set is the MHC proteins — the major histocompatibility complex, called HLA (human leukocyte antigen) in humans. MHC genes are the most polymorphic in the genome: hundreds to thousands of alleles exist at each locus, so the probability that two unrelated people carry identical MHC is vanishingly small. Your MHC molecules are, functionally, a molecular signature of you, and they are why transplants are rejected (§21.8).

Lymphocyte development, and the two arms

All lymphocytes originate from haematopoietic stem cells in red bone marrow. Where they mature determines what they become.

• B cells mature in the bone marrow. They become the humoral arm.
• T cells migrate to the thymus and mature there (§20.5). They become the cellular arm.

Two things happen during maturation, in this order.

1 · Receptor generation. Each developing lymphocyte randomly rearranges its receptor gene segments — the V, D, and J segments — splicing them together in an enormous number of combinations, with additional nucleotides inserted at the joins. The result is that each lymphocyte ends up with one receptor specificity, chosen at random, before it has ever seen an antigen. Across the population, the repertoire covers on the order of 10¹¹ distinct specificities. The system does not design receptors to fit threats; it generates every receptor it can, and waits.

2 · Selection. A randomly generated repertoire will contain receptors that are useless and receptors that are dangerous. Both must be removed, which is what the thymic cortex and medulla are for (§20.5).

• Positive selection (thymic cortex). Thymocytes are presented with self-MHC molecules. One whose receptor can bind self-MHC with some affinity receives a survival signal; one that cannot bind self-MHC at all dies by neglect, because every antigen a T cell will ever see is presented on self-MHC. This step selects for MHC restriction, and roughly 90% of thymocytes fail it.
• Negative selection (thymic medulla, and cortex). Survivors are presented with self-antigens, including tissue-specific proteins expressed ectopically under the control of the AIRE transcription factor. A thymocyte binding a self-antigen strongly is deleted by apoptosis, or in some cases diverted into becoming a regulatory T cell. This establishes self-tolerance.

The net yield is brutal: only about 2% of thymocytes survive both selections. The other 98% die in the thymus, which is why the thymic cortex is full of macrophages busy with apoptotic debris.

The two failure modes are worth naming, because they are two categories of disease. Negative selection that is too permissive leaves self-reactive clones in circulation — the substrate of autoimmunity (§21.9). Selection that is too aggressive or too stringent leaves holes in the repertoire, so that particular antigens cannot be responded to at all.

Clonal selection — the central idea

Everything above converges here. Clonal selection is the single organizing idea of adaptive immunity, and if you retain one mechanism from this chapter, make it this one.

The repertoire is generated before antigen exposure and at random. Each naive lymphocyte carries many copies of one receptor. The vast majority will never meet a matching antigen and will die unused within weeks.

When an antigen does enter the body, it encounters, by chance, the few lymphocytes whose pre-existing receptor happens to fit it — perhaps one cell in 10⁵ to 10⁶. Binding, together with the required second signal, activates that cell. It then proliferates, generating a clone of thousands of identical daughter cells, all specific for that same epitope. The clone differentiates into two populations: effector cells, which fight the current infection and mostly die within days, and memory cells, which persist for years to decades.

The antigen does not instruct the lymphocyte what to recognize. It selects a cell that already recognized it. This is a Darwinian mechanism operating inside your body on a timescale of days, and the parallel is exact: random variation, differential survival, expansion of the successful.

Three consequences follow directly, and each is testable:

• The response is specific, because only fitting clones expand.
• The response is slow the first time, because a clone of ~1 cell must divide roughly 12–14 times to reach an effective size, at 6–12 hours per division. That is the 6–10 day lag.
• The response is fast the second time, because the clone is no longer 1 cell but 10³–10⁴ memory cells, already expanded and pre-differentiated.

            CLONAL SELECTION — ONE ANTIGEN, ONE CLONE

  BEFORE ANTIGEN: a repertoire generated at random, ~10^11 specificities

     ⟨A⟩   ⟨B⟩   ⟨C⟩   ⟨D⟩   ⟨E⟩   ⟨F⟩   ⟨G⟩  ...  naive lymphocytes
      │     │     │     │     │     │     │        each with ONE
      │     │     │     │     │     │     │        receptor specificity
      ▼     ▼     ▼     ▼     ▼     ▼     ▼

  ANTIGEN "D" ENTERS  ───────────────►  binds ONLY cell ⟨D⟩
                                              │
     ⟨A⟩   ⟨B⟩   ⟨C⟩         ⟨E⟩ ...          │  + second signal
      ·     ·     ·           ·               │    (helper T cell
   (no match — remain naive, die unused)      │     or co-stimulation)
                                              ▼
                                       ACTIVATION + PROLIFERATION
                                       12–14 divisions, 6–12 h each
                                              │
                        ┌─────────────────────┴──────────────────┐
                        ▼                                        ▼
              ╔══════════════════╗                    ╔═════════════════╗
              ║  EFFECTOR CELLS  ║                    ║  MEMORY CELLS   ║
              ║  PLASMA CELLS    ║                    ║  ~10^3–10^4     ║
              ║  ~2000 antibody  ║                    ║  long-lived     ║
              ║  molecules/sec   ║                    ║  (years–decades)║
              ║  live 4–5 days   ║                    ║  do NOT secrete ║
              ║  do NOT divide   ║                    ║  CAN divide     ║
              ╚════════╤═════════╝                    ╚════════╤════════╝
                       │                                       │
                       ▼                                       ▼
              fights THIS infection              waits for the NEXT
              (dies when it is over)             exposure to "D"
                                                 → SECONDARY RESPONSE
                                                   (Fig. 21.5)

  KEY: the antigen does not TEACH the cell. It SELECTS a cell that
       already, by chance, matched it. Random variation → selection →
       clonal expansion. Darwinian, inside you, over days.

21.7 Humoral Immunity — Antibodies

Humoral immunity — from humor, the old word for body fluid — is mediated by antibodies dissolved in blood and lymph. It is the arm that deals with pathogens outside cells: bacteria in the interstitium, viruses in transit between cells, toxins in the blood.

B cell activation

A naive B cell displays thousands of copies of its receptor, which is a membrane-bound antibody of the IgM and IgD classes. Activation requires two signals.

Signal 1 is antigen binding and cross-linking of the B cell receptor. The B cell then internalizes the antigen, digests it, and displays fragments on MHC class II.

Signal 2 is provided by a helper T cell whose receptor recognizes that same displayed fragment. The helper cell binds and secretes cytokines that license the B cell to proliferate and differentiate. This is why the two arms are not independent, and why losing helper T cells destroys antibody responses as well as cellular ones (§21.9).

A minority of antigens — repetitive polysaccharides, such as bacterial capsules — can activate B cells without T cell help. These T-independent responses are fast but produce only IgM, no class switching, and little or no memory, and they work poorly in children under two. That single fact explains an entire vaccine design strategy: the pneumococcal and Hib conjugate vaccines chemically link the bacterial polysaccharide to a carrier protein, converting a T-independent antigen into a T-dependent one. That recruits helper T cells, class switching, and durable memory — and it is why these vaccines work in infants when the older polysaccharide-only vaccines did not.

Activated B cells proliferate and differentiate into:

• Plasma cells — the effector. A plasma cell is a factory: it has an enormously expanded rough endoplasmic reticulum, an eccentric "clock-face" nucleus, and secretes roughly 2,000 antibody molecules per second. It no longer divides and lives 4–5 days, though a subset homes to the bone marrow and survives for years, maintaining serum antibody titres decades after exposure.
• Memory B cells — long-lived, non-secreting, and carrying higher-affinity receptors as a result of affinity maturation in germinal centres.

Antibody structure

An antibody (immunoglobulin) is a Y-shaped glycoprotein built from four polypeptide chains: two identical heavy chains and two identical light chains, joined by disulfide bonds.

Each chain has a variable (V) region at the amino terminus and a constant (C) region. The variable regions of one heavy and one light chain together form one antigen-binding site, so every antibody monomer has two identical binding sites, at the tips of the Y's arms. The variable region determines what the antibody binds; the constant region of the heavy chain — the stem of the Y, called the Fc region — determines what happens next: which class the antibody belongs to, which cells can bind it, whether it fixes complement, whether it crosses the placenta, and where in the body it goes.

That division of labour is the whole design. Variable region: specificity. Constant region: function. Two antibodies with identical binding sites but different Fc regions bind the same target and do completely different things with it — and class switching, in which an activated B cell swaps its heavy chain constant region while keeping its variable region unchanged, exists precisely to exploit that.

The arms are joined to the stem at a flexible hinge region, which lets the two binding sites splay from about 60° to 180°. Flexibility is functional: it lets one antibody grip two epitopes on two separate particles at different spacings, which is the basis of agglutination.

              ANTIBODY STRUCTURE AND THE FIVE CLASSES

           antigen-binding site            antigen-binding site
            (V_H + V_L together)                    │
                    │                               │
                    ▼                               ▼
              ╔═════════╗                     ╔═════════╗
              ║  V_L    ║                     ║   V_L   ║   LIGHT CHAIN
              ╠═════════╣  ╔═════════╗ ╔══════╣═════════╣   (κ or λ)
              ║  C_L    ║  ║  V_H    ║ ║ V_H  ║   C_L   ║
              ╚═══╤═════╝  ╠═════════╣ ╠══════╣═════╤═══╝
                  ╰──S─S───╢  C_H1   ║ ║ C_H1 ╟──S─S╯
                           ╚════╤════╝ ╚═╤════╝
                          ╭─────┴──HINGE─┴─────╮  ← flexible: arms
                          │      S─S  S─S      │    splay 60°–180°
                          ╞════════════════════╡
                          │       C_H2         │  ← COMPLEMENT
                          │                    │    (C1) binds here
                          ╞════════════════════╡      = Fc REGION
                          │       C_H3         │  ← PHAGOCYTE Fc
                          │                    │    receptors bind here
                          ╘════════════════════╛
                        HEAVY CHAINS (2, identical)

   VARIABLE region  → WHAT it binds  (specificity)
   CONSTANT region  → WHAT HAPPENS NEXT (class, effector function)

  ┌──────┬────────┬─────────┬──────────────────────────────────────┐
  │Class │Structure│% serum  │ Where it is, and what it does        │
  ├──────┼────────┼─────────┼──────────────────────────────────────┤
  │ IgG  │monomer │ 75–85%  │ Blood, lymph, TISSUE. The workhorse. │
  │      │        │         │ Only class crossing the PLACENTA.    │
  │      │        │         │ Fixes complement; best opsonin;      │
  │      │        │         │ dominant in SECONDARY response.      │
  ├──────┼────────┼─────────┼──────────────────────────────────────┤
  │ IgA  │monomer │ 10–15%  │ SECRETIONS: saliva, tears, mucus,    │
  │      │(serum) │         │ bile, BREAST MILK. Blocks pathogen   │
  │      │DIMER   │         │ attachment to epithelium. Does NOT   │
  │      │(secret)│         │ fix complement — non-inflammatory     │
  │      │        │         │ by design (mucosa must stay quiet).  │
  ├──────┼────────┼─────────┼──────────────────────────────────────┤
  │ IgM  │PENTAMER│ 5–10%   │ Blood. FIRST antibody made in a      │
  │      │(10 arms)│        │ primary response; also the B cell    │
  │      │        │         │ receptor (as monomer). BEST at       │
  │      │        │         │ complement fixation + agglutination  │
  │      │        │         │ (10 arms). Too big to leave vessels. │
  ├──────┼────────┼─────────┼──────────────────────────────────────┤
  │ IgE  │monomer │ <0.01%  │ Bound to MAST CELLS and BASOPHILS by │
  │      │        │         │ its Fc. Cross-linking → degranulation│
  │      │        │         │ → ALLERGY. Also anti-parasitic.      │
  ├──────┼────────┼─────────┼──────────────────────────────────────┤
  │ IgD  │monomer │ <1%     │ B cell surface receptor with IgM.    │
  │      │        │         │ Function largely as an activation    │
  │      │        │         │ receptor; little secreted.           │
  └──────┴────────┴─────────┴──────────────────────────────────────┘

   MNEMONIC for the order of appearance:  IgM first, IgG later and larger.

The four antibody mechanisms

Antibodies do not kill anything. They bind, and binding produces four consequences.

1. Neutralization — the simplest and, for viruses and toxins, the most important. The antibody covers the business end of the molecule: the receptor-binding site of a virus, the active site of a toxin. The target cannot dock, so it cannot act. Tetanus antitoxin and antivenom work this way, and so do the neutralizing antibodies measured after vaccination.
2. Agglutination — multivalent antibody cross-links whole cells into clumps. Because each IgG has two arms and IgM has ten, a single antibody can bridge two particles. Clumps are immobilized and far easier for phagocytes to ingest. Blood typing (Chapter 17) is agglutination made visible.
3. Precipitation — the same cross-linking applied to soluble molecules, which fall out of solution as insoluble lattices and are phagocytosed. When these immune complexes are not cleared efficiently, they deposit in vessels and glomeruli — the mechanism of type III hypersensitivity (§21.9).
4. Complement fixation — antigen-bound IgM or IgG exposes complement-binding sites on their Fc regions, and C1 binds, initiating the classical pathway. This is quantitatively the most destructive antibody mechanism: it produces MAC lysis, C3b opsonization, and C3a/C5a inflammation all at once. Note the safety design: an antibody in free solution does not fix complement; only an antibody that has already bound its target does, because C1 binding requires two adjacent Fc regions, which only happens when antibodies are clustered on a surface. The cascade cannot fire in the bloodstream by accident.

A fifth, half-way mechanism deserves naming: antibody-dependent cell-mediated cytotoxicity. NK cells carry Fc receptors, bind antibody-coated target cells, and kill them. Antibody here acts as a targeting system for an innate cell.

Active and passive immunity

The distinction is not academic — it determines both onset and duration.

Active immunity is slow (days to weeks), because the recipient must actually mount a clonal response, but it is durable, because it generates memory cells. Passive immunity is immediate, because the antibodies are already made, but it is temporary: IgG has a serum half-life of about 23 days, so transferred antibody is gone within months, and there is no memory because the recipient's own lymphocytes were never activated.

This is why a person bitten by a rabid animal receives both: immune globulin for protection now, and vaccine for protection in two weeks, because the incubation period is long enough that active immunity can arrive in time.

Primary and secondary responses

The graph below is arguably the most consequential curve in medicine, because it is the reason vaccination works.

       PRIMARY vs SECONDARY ANTIBODY RESPONSE

  serum
  antibody
  titre
  (log)
   10^4 ┤                                              ╭──────╮
        │                                             ╱ IgG    ╲
        │                                            ╱          ╲___
   10^3 ┤                                           ╱                ╲___
        │                                          ╱  SECONDARY
        │                                         ╱   ~2–3 days lag
   10^2 ┤            ╭────╮                      ╱    100–1000× peak
        │       IgG ╱      ╲___                 ╱     HIGH AFFINITY
        │          ╱           ╲___            ╱      mostly IgG
   10^1 ┤    ╭─╮  ╱                 ╲___      ╱       lasts months–yrs
        │   ╱IgM╲╱                       ╲___╱
        │  ╱     ╲  PRIMARY: 5–10 d lag
   10^0 ┤─╯       ╲ low peak · IgM then IgG
        │          ╲ low affinity · fades in weeks
        └──┬───┬───┬───┬───┬────────────┬───┬───┬───┬───┬───► time
           0   7   14  21  28          ▲    7   14  21  28  (days)
           ▲                           │
        1st EXPOSURE            2nd EXPOSURE (weeks–decades later)
        (or 1st vaccine dose)   (or booster dose)

  WHY THE DIFFERENCE — three things changed, none of them magic:
    1. STARTING POPULATION  1 cell  →  10^3–10^4 memory cells
    2. CELL STATE           naive, must differentiate → memory cells are
                            pre-primed; some are pre-switched to IgG
    3. RECEPTOR AFFINITY    germline → affinity-matured in germinal
                            centres by somatic hypermutation + selection

  VACCINATION = deliberately buying the SECOND curve without paying
                for the disease that normally produces the first.

Two design consequences follow, and both are visible on any immunization schedule. Booster doses exist because each re-exposure drives another round of affinity maturation and expands the memory pool further, so titres both rise higher and decay more slowly. And an adjuvant — alum, or a lipid nanoparticle — is included in most non-live vaccines for a reason that belongs to §21.1 and §21.4 rather than to this section: it produces local innate activation, without which purified antigen alone signals "harmless" and is largely ignored. Innate immunity decides whether adaptive immunity should bother. A vaccine must therefore contain both the specificity and the alarm.

21.8 Cellular Immunity — T Cells

Antibodies patrol fluid. They cannot reach a virus that has already entered a cell, a bacterium living inside a macrophage, or a cancer cell whose abnormality is a mutant protein made in its own cytoplasm. Something has to inspect the inside of cells, and that is the cellular arm.

Antigen presentation and the two MHC classes

T cells cannot recognize free antigen. A T cell receptor binds only a peptide fragment held in the groove of an MHC molecule on another cell's surface. This is MHC restriction, and it is the constraint that positive selection in the thymus was testing for (§21.6).

There are two classes, and the difference between them is the most elegant logic in the chapter.

Now derive the distribution from the message, rather than memorizing it.

Class I is on every nucleated cell because every nucleated cell can be subverted. Any cell can be infected by a virus or become cancerous, and when it does, the abnormality is a protein being synthesized in its cytoplasm. So every cell continuously samples its own cytosolic proteins, degrades a fraction of them in the proteasome, loads the fragments onto class I, and displays them. This is self-surveillance: a running public declaration of internal state. A cell displaying only normal self-peptides is passed over; a cell displaying a viral or mutant peptide is destroyed by a CD8 T cell. Red blood cells are exempt because they have no nucleus and make no protein — and, usefully, that is exactly why the malaria parasite hides in them.

Class II is restricted to antigen-presenting cells because its message is different. It does not say "I am infected." It says "I have sampled the environment, and here is what is in it." That is an alerting function, and it should be performed only by cells whose job is surveillance — dendritic cells that patrol tissue and migrate to nodes, macrophages that eat, and B cells that capture antigen with their receptor. If every cell in the body could display engulfed material on class II, every cell could summon helper T cells, and the coordination signal would be noise.

The two-class system therefore separates "kill me" from "help me organize", and it assigns each message to cells competent to send it. Both are needed for the same reason: the intracellular world and the extracellular world are different problems.

          THE TWO MHC CLASSES — TWO MESSAGES, TWO AUDIENCES

  ══ CLASS I ══ "HERE IS WHAT I AM MAKING"  ══ SELF-SURVEILLANCE ══

   ANY NUCLEATED CELL (muscle, neuron, hepatocyte, tumour cell…)
   ┌──────────────────────────────────────────────────────┐
   │  cytosolic proteins (self · viral · mutant)          │
   │            │                                         │
   │            ▼  PROTEASOME chops a sample              │
   │        peptides ──► TAP ──► ER ──► loaded on MHC I   │
   └────────────────────────────────┬─────────────────────┘
                                    ▼  displayed on surface
                         ╔══════════════════════╗
                         ║  MHC I  +  peptide   ║
                         ╚══════════╤═══════════╝
                                    │  read by
                                    ▼
                        ┌───────────────────────┐
                        │  CD8 CYTOTOXIC T CELL │
                        └───────────┬───────────┘
                 self peptide ──────┴────── viral/mutant peptide
                    │                              │
                 PASS OVER                    ► KILL THE CELL
                                               perforin + granzyme
                                               → APOPTOSIS

   AND IF THE CELL HIDES BY REMOVING MHC I  ─────────┐
                                                     ▼
                                        ┌────────────────────┐
                                        │  NK CELL: "missing │
                                        │  self" ► KILLS IT  │
                                        └────────────────────┘

  ══ CLASS II ══ "HERE IS WHAT I FOUND OUT THERE" ══ ALERTING ══

   ONLY dendritic cell · macrophage · B cell
   ┌──────────────────────────────────────────────────────┐
   │  ENGULFED material from outside                      │
   │            │                                         │
   │            ▼  PHAGOLYSOSOME digests it               │
   │        peptides ──────────────► loaded on MHC II     │
   └────────────────────────────────┬─────────────────────┘
                                    ▼
                         ╔══════════════════════╗
                         ║  MHC II +  peptide   ║  + B7 (only if the
                         ╚══════════╤═══════════╝  cell's PRRs fired)
                                    │  read by
                                    ▼
                        ┌───────────────────────┐
                        │  CD4 HELPER T CELL    │
                        └───────────┬───────────┘
        signal 1 only (no B7) ──────┴────── signal 1 + signal 2
                    │                              │
                 ANERGY                       ► ACTIVATION
              (safety interlock)          licenses B cells, arms CD8,
                                          activates macrophages

  THE RULE:   endogenous → class I → CD8 → kill
              exogenous  → class II → CD4 → coordinate

Helper T cells — the coordinating hub

Helper T cells (CD4) do not kill anything and do not secrete antibody. They secrete cytokines, and almost everything else in adaptive immunity depends on them.

A naive CD4 cell is activated when its receptor binds peptide on MHC class II and it receives a co-stimulatory signal (B7 on the presenting cell binding CD28 on the T cell). The requirement for two signals is a safety interlock: antigen recognition without co-stimulation induces anergy — functional unresponsiveness — rather than activation. Since only an innate-activated presenting cell upregulates B7, the design means adaptive immunity fires only when innate immunity has already declared a threat. Once again: innate instructs adaptive.

Activated helper cells then:

• License B cells to proliferate, class switch, and form germinal centres (§21.7). Without this, antibody responses are limited to low-affinity IgM.
• Activate cytotoxic T cells by providing IL-2, the principal T cell growth factor. Most CD8 responses are inefficient without CD4 help.
• Activate macrophages via interferon gamma, dramatically increasing their killing power — which is how intracellular bacteria such as Mycobacterium tuberculosis are controlled.
• Recruit and direct other cells through chemokines, and shape the type of response through differentiation into subsets: Th1 (interferon gamma; intracellular pathogens), Th2 (IL-4, IL-5, IL-13; parasites, and the driver of allergy), Th17 (IL-17; extracellular bacteria and fungi at mucosal surfaces), and T follicular helper cells, which do the B cell licensing in germinal centres.

Remove CD4 cells and both arms fail. Hold on to that; it is the answer to a question in §21.9.

Cytotoxic T cells and how they kill

Cytotoxic T cells (CD8) recognize peptide on MHC class I and kill the presenting cell. The killing is deliberate, contact-dependent, and precise.

1. The T cell binds and forms an immunological synapse with the target — a tight, sealed junction that confines everything released to the space between the two cells, so neighbouring cells are not damaged.
2. Granules containing perforin and granzymes are polarized toward the synapse and released.
3. Perforin polymerizes in the target membrane to form pores.
4. Granzymes — serine proteases — enter through those pores and activate the target cell's own caspase cascade, triggering apoptosis.
5. Alternatively, the T cell's FasL engages Fas on the target, delivering the same death signal through a different receptor.

The choice of apoptosis over necrosis is not incidental. Apoptosis packages the cell's contents — including intact viral genomes and enzymes — into membrane-bound bodies that are phagocytosed without spilling. Necrosis would release live virus and inflammatory contents into the tissue. The immune system kills quietly on purpose.

The cytotoxic T cell then detaches and moves on, killing repeatedly. One cell can destroy many targets, which is why cytotoxic responses scale.

Regulatory T cells

Regulatory T cells (Tregs), marked by CD4, CD25, and the transcription factor FoxP3, suppress other lymphocytes by secreting IL-10 and TGF-β, by consuming IL-2, and by direct contact. They are the brake. They enforce peripheral tolerance — the second line of defense against self-reactivity, catching self-reactive clones that escaped thymic negative selection — and they terminate responses once a pathogen is cleared. Mutations in FoxP3 cause IPEX syndrome, a catastrophic multi-organ autoimmune disease of infancy that demonstrates exactly what the brake was doing.

Transplantation and rejection

Graft rejection is cellular immunity working correctly on the wrong target.

Because MHC is extraordinarily polymorphic, an allograft's MHC molecules are themselves foreign, and an unusually large fraction of the recipient's T cells — 1–10%, versus 1 in 10⁵–10⁶ for a typical antigen — can recognize them directly. That is why rejection is so vigorous and why matching MHC (HLA) between donor and recipient improves survival so much.

Graft-versus-host disease is the mirror image, seen after bone marrow transplantation: the graft contains competent donor T cells, and they attack the recipient's tissues.

21.9 When Immunity Fails

Three categories, and they are three different failure modes of the same machine: too little (immunodeficiency), too much (hypersensitivity), and aimed at the wrong target (autoimmunity).

Immunodeficiency

Severe combined immunodeficiency (SCID) is a group of inherited disorders in which both T and B cell function fail — most commonly through mutations in the common gamma chain shared by several cytokine receptors, or in adenosine deaminase, whose deficiency lets a toxic metabolite accumulate and kill lymphocytes. Infants appear well for the first months, protected by maternal IgG (§21.7), and then present with persistent thrush, chronic diarrhoea, failure to thrive, and infections with organisms that never trouble a normal child. Untreated it is fatal in the first year. Haematopoietic stem cell transplantation is curative, and SCID was the first disease successfully treated with gene therapy.

Note the timing, because it is diagnostic: the illness begins as maternal IgG decays, at around three to six months. The infant was never protected by its own immunity; it was living on borrowed antibody, and the presentation is the moment the loan expires.

Acquired immunodeficiency syndrome (AIDS) is the paradigm acquired form, and its mechanism makes the point of §21.8 unforgettable.

Hypersensitivity

Hypersensitivity is an immune response that damages the host — an appropriate mechanism aimed at an inappropriate or harmless target, or deployed at excessive scale. Four types.

              THE FOUR TYPES OF HYPERSENSITIVITY

  ┌──────┬──────────┬──────────┬──────────────┬──────────────────────┐
  │ TYPE │ MEDIATOR │  TIMING  │  MECHANISM   │  EXAMPLES            │
  ├──────┼──────────┼──────────┼──────────────┼──────────────────────┤
  │  I   │   IgE    │ SECONDS  │ Allergen     │ Hay fever, allergic  │
  │IMMED-│          │ to ~30   │ cross-links  │ asthma, food allergy,│
  │ IATE │ on MAST  │ minutes  │ IgE on mast  │ ANAPHYLAXIS,         │
  │      │ CELLS    │          │ cells →      │ atopic dermatitis    │
  │      │          │          │ DEGRANULATION│                      │
  │      │          │          │ (histamine,  │ Local vs systemic is │
  │      │          │          │ leukotrienes,│ purely a question of │
  │      │          │          │ tryptase)    │ WHERE the mast cells │
  │      │          │          │              │ degranulate.         │
  ├──────┼──────────┼──────────┼──────────────┼──────────────────────┤
  │  II  │ IgG/IgM  │ minutes  │ Antibody     │ Transfusion reaction,│
  │ CYTO-│ vs a CELL│ to hours │ binds a      │ hemolytic disease of │
  │ TOXIC│ SURFACE  │          │ FIXED cell   │ the newborn (Rh),    │
  │      │ antigen  │          │ antigen →    │ Graves disease,      │
  │      │          │          │ complement + │ myasthenia gravis,   │
  │      │          │          │ phagocytosis │ autoimmune hemolysis │
  │      │          │          │ → THAT cell  │                      │
  │      │          │          │ dies         │ Damage is CELL-      │
  │      │          │          │              │ SPECIFIC.            │
  ├──────┼──────────┼──────────┼──────────────┼──────────────────────┤
  │ III  │ IgG +    │ hours to │ Ag + Ab form │ Serum sickness, SLE, │
  │IMMUNE│ SOLUBLE  │ ~3 weeks │ SOLUBLE      │ post-strep glomerulo-│
  │COMPL-│ antigen  │          │ complexes →  │ nephritis, farmer's  │
  │  EX  │          │          │ DEPOSIT in   │ lung                 │
  │      │          │          │ vessel walls,│                      │
  │      │          │          │ glomeruli,   │ Damage is at the     │
  │      │          │          │ joints, skin │ DEPOSITION SITE, not │
  │      │          │          │ → complement │ where the antigen    │
  │      │          │          │ → neutrophils│ came from.           │
  ├──────┼──────────┼──────────┼──────────────┼──────────────────────┤
  │  IV  │  T CELLS │ 12–72 h  │ Sensitized   │ TB skin test (PPD),  │
  │DELAY-│ (no anti-│ (hence   │ Th1 cells →  │ poison ivy / contact │
  │  ED  │  body at │ DELAYED) │ cytokines →  │ dermatitis (nickel,  │
  │      │  all)    │          │ MACROPHAGE   │ latex), type 1       │
  │      │          │          │ activation;  │ diabetes, MS, celiac,│
  │      │          │          │ or CD8 kills │ graft rejection      │
  │      │          │          │ directly     │                      │
  └──────┴──────────┴──────────┴──────────────┴──────────────────────┘

  THE ORGANIZING SPLIT:
       Types I, II, III  =  ANTIBODY-mediated  →  fast (minutes–hours)
                            transferable by SERUM
       Type IV           =  T CELL-mediated    →  slow (12–72 h)
                            transferable ONLY by CELLS

  Mnemonic: A-C-I-D  —  Anaphylactic · Cytotoxic · Immune complex · Delayed

Autoimmunity

Autoimmune disease is a failure of self-tolerance: the immune system attacks self-antigens. About 5% of adults in developed countries are affected, and the female-to-male ratio is roughly 2–3 to 1 for most such diseases.

Self-tolerance is maintained at two levels — central tolerance (thymic and marrow negative selection) and peripheral tolerance (regulatory T cells, anergy from missing co-stimulation, and physical sequestration of some antigens). Autoimmunity requires a failure of both, and the recognized mechanisms are:

1. Escape from central tolerance. Not every self-antigen is presented in the thymus, and selection thresholds are imperfect. Some self-reactive clones always reach the periphery.
2. Failure of regulatory T cells — as in IPEX syndrome, in extreme form.
3. Molecular mimicry. A microbial epitope closely resembles a self-epitope, so an appropriate anti-microbial response cross-reacts with tissue. Rheumatic fever is the classic case: antibodies to streptococcal M protein cross-react with cardiac myosin and valve glycoproteins, and the resulting valve damage is a leading cause of mitral stenosis worldwide.
4. Release of sequestered antigens. Some tissues — the lens and uveal tract of the eye, sperm, parts of the CNS and thyroid — are normally hidden behind barriers and were never presented during thymic selection, so no tolerance exists to them. Trauma that releases them can provoke a genuine immune response, as in sympathetic ophthalmia after penetrating eye injury.
5. Novel or altered self-antigens. A drug binding a self-protein as a hapten, or a post-translational modification (citrullination of proteins in rheumatoid arthritis), creates a target the immune system was never tolerized to.
6. Inappropriate MHC class II expression. Cells that normally never present antigen may be induced by inflammation to express class II and present their own peptides, converting a bystander into a target.

Genetic susceptibility is real and largely HLA-linked — HLA-B27 with ankylosing spondylitis, HLA-DQ2/DQ8 with coeliac disease — which makes sense, since HLA alleles determine which self-peptides can be presented at all. But concordance in identical twins is typically only 20–50%, so environment matters as much as genotype.

21.10 Advanced Topic · Inflammation as a Double-Edged Sword

Everything in §21.4 described inflammation as a defense. It is. It is also, in the aggregate, one of the largest causes of death and disability in the developed world — not through infection, but through inflammation directed at sterile targets, sustained beyond its usefulness, or resolving into scar where scar cannot be afforded.

This section resolves Amara's case, and it does so by taking the question apart rather than by answering it as posed.

Acute versus chronic

The critical row is the tissue outcome. In acute inflammation, destruction happens and then repair happens. In chronic inflammation the two occur at the same time and in the same place, which is why chronically inflamed tissue is simultaneously being eaten and being fibrosed, and why the end result is a distorted organ rather than either a healed one or a destroyed one.

Chronic inflammation arises three ways: a persistent stimulus the immune system cannot clear (Mycobacterium tuberculosis, silica, a suture, oxidized LDL in an artery wall); an autoimmune process where the stimulus is the body itself and therefore never runs out; or failure of resolution, in which the acute response is never actively terminated (§21.4).

The fork: resolution or fibrosis

When inflammation ends, tissue takes one of two paths, and which one depends on two things.

Regeneration replaces destroyed cells with the same cell type, restoring both architecture and function. It requires that the tissue's parenchymal cells can divide and that the connective tissue framework — especially the basement membrane — survived intact. Epidermis, gut epithelium, liver, and bone regenerate well.

Fibrosis replaces destroyed tissue with collagen: fibroblasts proliferate, deposit extracellular matrix, and the defect becomes scar. Scar restores mechanical continuity and strength. It restores no function whatsoever.

The determining factors are exactly the ones from Chapter 4:

1. Can the parenchymal cells divide? Labile tissues (epithelium, marrow) divide constantly. Stable tissues (liver, kidney tubule, fibroblasts) can divide when stimulated. Permanent tissues — cardiac muscle, skeletal muscle, neurons — essentially cannot. For permanent tissue, fibrosis is the only option available.
2. Is the framework intact, and how large is the defect? Even a regenerating tissue scars if the basement membrane and stroma are destroyed, because the surviving cells have nothing to crawl along and no template to rebuild against — and because a large defect fills with granulation tissue before regeneration can bridge it.

Cardiac muscle fails both tests after an infarct. This is the mechanistic core of Amara's case, and it was flagged in Chapter 4 and again in Chapter 9: scar is not muscle.

Atherosclerosis: chronic inflammation in an artery wall

Chapter 19 described atherosclerotic plaque anatomically. Here is what it actually is.

Atherosclerosis begins with endothelial dysfunction — from hypertension's shear stress, hyperglycaemia, tobacco smoke, or oxidized LDL. The injured endothelium does what injured endothelium does anywhere: it expresses adhesion molecules. Monocytes marginate, undergo diapedesis into the intima, and become macrophages. There they engulf oxidized LDL through scavenger receptors that, unlike the normal LDL receptor, are not down-regulated by intracellular cholesterol. The macrophage therefore keeps eating until it is engorged with lipid droplets — a foam cell — and eventually dies, spilling its lipid into the intima and forming the plaque's necrotic core.

Meanwhile T cells arrive, secrete interferon gamma, and activate more macrophages; smooth muscle cells migrate from the media and lay down a fibrous cap of collagen over the core. That cap is the plaque's stability, and its integrity is under continuous inflammatory control: macrophage matrix metalloproteinases degrade collagen while smooth muscle cells synthesize it.

Plaque rupture — the event that caused Amara's infarct — is an inflammatory event. A plaque with a thin cap, a large lipid core, and dense macrophage infiltration ruptures; a thick-capped, fibrous, calcified plaque generally does not. Degree of stenosis is a poor predictor of which plaque will rupture; inflammatory activity is a better one. This is why CRP carries independent cardiovascular risk information beyond LDL, and why trials that reduced cardiovascular events by targeting inflammation without changing lipids were conceptually important: they demonstrated that the inflammation is not a marker of the disease but part of its mechanism.

Amara's CRP of 38 mg/L on day 2 is dominated by her infarct. But her pre-existing baseline — elevated by visceral adiposity, insulin resistance, twenty years of circadian disruption, and hypertension — is part of why she had a plaque that ruptured at 45.

   risk of
   upper resp.  ▲
   infection    │                                       ╱  OVERTRAINING
                │╲                                    ╱    heavy chronic
     above      │ ╲                                  ╱     load, elite
     average    │  ╲                               ╱       endurance
                │   ╲                            ╱
     average  ──┼────╲──────────────────────────╱──────────
                │     ╲                       ╱
     below      │      ╲___________________╱
     average    │        MODERATE REGULAR EXERCISE
                └────────────────────────────────────────►  training load
                sedentary        moderate           very heavy

The phases of infarct healing

Amara's infarct heals on a schedule, and the schedule is inflammation's schedule.

Phase 1 · Inflammatory (days 1–3). Cardiomyocytes deprived of oxygen die by necrosis — they swell and rupture rather than undergoing orderly apoptosis, because apoptosis is ATP-dependent and there is none. Rupture spills the cell's contents into the interstitium: ATP, mitochondrial DNA, HMGB1, heat shock proteins, uric acid, and the whole intracellular ion composition. These are DAMPs — damage-associated molecular patterns.

DAMPs bind the same pattern recognition receptors that bind bacterial PAMPs. The receptors do not distinguish them, and there is a reason: from an evolutionary standpoint, cell rupture and bacterial invasion co-occurred so consistently that a receptor tuned to either was tuned to both. Complement is activated, mast cells degranulate, IL-1 and TNF-α and IL-6 pour out, and neutrophils marginate and invade in enormous numbers, peaking at 24–72 hours.

This is sterile inflammation: a full inflammatory response with no organism anywhere. It is the answer to question 1.

Phase 2 · Proliferative (days 3–14). Neutrophils undergo apoptosis and are eaten by macrophages, and this act of efferocytosis is one of the signals that switches macrophage phenotype from pro-inflammatory to reparative. The reparative macrophage secretes TGF-β, VEGF, and growth factors: it recruits fibroblasts and drives angiogenesis. Granulation tissue — new capillaries, fibroblasts, and loose type III collagen — fills the defect.

The wall here is at its weakest around days 4–7, because the dead muscle has been removed but the collagen replacing it is immature. Free wall rupture, papillary muscle rupture, and ventricular septal rupture cluster in exactly this window, which is a mechanical consequence of an immunological timetable.

Phase 3 · Maturation (weeks 2–8). Fibroblasts become myofibroblasts, type III collagen is replaced by cross-linked type I, the scar contracts, and capillary density falls. By six to eight weeks the region is a dense, pale, avascular, relatively acellular collagenous scar: strong in tension, thin, electrically inert, and non-contractile.

That is what Amara's six-week echocardiogram shows.

Resolving the double edge

Now the second question can be answered properly, because the two processes are not two processes.

The macrophage that phagocytoses a dead cardiomyocyte on day 2 is the same lineage of cell that, on day 6, secretes TGF-β and recruits the fibroblast that lays down the collagen. Clearance and scarring are sequential phases of one program, executed by one cell population changing phenotype. There is no molecular switch that removes debris without also initiating repair, because in every tissue where this program evolved, removing debris without repairing the defect would leave a hole.

So: is inflammation helping her or hurting her? The question as posed has no answer, and recognizing that is the point. What can be said precisely is:

• Debris clearance is mandatory. Necrotic myocardium left in place would continue releasing DAMPs, sustain inflammation indefinitely, and provide no mechanical strength. Blocking phase 1 outright — which has been tried, with high-dose corticosteroids in the 1970s and with NSAIDs — produces thinner scars, more infarct expansion, and more free-wall rupture. Less inflammation produced worse outcomes.
• Scar formation is mandatory. Cardiac muscle is permanent tissue. Without fibrosis the ventricular wall would have no tensile strength at systolic pressure and would aneurysm or rupture. The scar is what keeps her ventricle intact.
• The scar is permanently costly. It cannot contract, so stroke volume falls; it is stiff, so diastolic filling worsens — the mechanism of her heart failure with preserved ejection fraction; and it conducts abnormally, creating a substrate for re-entrant arrhythmia.
• What is modifiable is the magnitude and duration, not the existence. Reperfusing the artery early limits how much muscle dies and therefore how much scar forms — which is why door-to-balloon time is measured in minutes. ACE inhibitors and beta-blockers limit adverse remodelling of the surrounding healthy myocardium. Cardiac rehabilitation lowers baseline inflammation.

Inflammation is not good or bad. It is a program with an obligatory cost, and clinical medicine's leverage is over its size and its timing, not over whether it happens.

Chapter Summary

§21.1 Defense has three lines: surface barriers, internal innate defenses, and adaptive immunity. The first two are innate — germline-encoded receptors, about a hundred kinds, recognizing conserved PAMPs, acting in seconds to hours, with no memory. The third is adaptive — somatically rearranged receptors covering ~10¹¹ specificities, one epitope each, taking 6–10 days on first exposure, acting systemically, and remembering for decades. The organizing relationship is that innate immunity instructs adaptive immunity, through antigen presentation, through co-stimulation (antigen without it produces anergy), and through cytokine context. Adaptive immunity in turn amplifies innate mechanisms via antibody and interferon gamma.

§21.2 Surface barriers are the only defense most pathogens meet: keratinized, layered, shedding epidermis; mucous membranes with mucus, the mucociliary escalator, and flushing; and chemical defenses — acid mantle, gastric acid, lysozyme, defensins, vaginal acidity, and competitive exclusion by normal flora. Every medical device that crosses a barrier predicts its own infection, and removing the device is usually a more effective intervention than adding a drug.

§21.3 Neutrophils are numerous, short-lived, and cannot present antigen: ammunition. Macrophages are long-lived, present antigen, and decide what happens next. Phagocytosis proceeds through adherence (via pattern recognition receptors), ingestion, phagolysosome fusion, killing by the respiratory burst and lysosomal enzymes, and exocytosis; opsonization with antibody or C3b can raise efficiency a thousandfold. NK cells kill by the "missing self" rule, closing the escape hatch used by viruses and tumours that downregulate MHC class I.

§21.4 Inflammation is the innate response of vascularized tissue to injury of any kind, not to infection specifically. Redness and heat come from arteriolar vasodilation; swelling from venular permeability, producing a protein-rich exudate whose fibrinogen builds a scaffold; pain from bradykinin, prostaglandins, released ions, and pressure. Leukocytes reach tissue by leukocytosis, margination (selectins for rolling, integrins and ICAM for arrest), diapedesis, and chemotaxis. Neutrophils peak on day 1, macrophages on days 2–3. Resolution is an active program driven by efferocytosis and specialized pro-resolving mediators, not a passive running-out.

§21.5 Complement is a cascade of ~30 plasma proteins with three activation pathways — classical (antibody), lectin, alternative — converging on C3, and three outcomes: MAC lysis, opsonization by C3b (quantitatively the most important), and inflammation via the anaphylatoxins C3a and C5a. The alternative pathway distinguishes self from non-self by protecting the self with membrane regulators rather than by recognizing the enemy. Interferons alpha and beta place uninfected neighbours in an antiviral state; interferon gamma activates macrophages. Fever is a deliberate hypothalamic set-point shift driven by IL-1, IL-6, TNF-α, and PGE₂: shivering means the temperature is still rising toward the new target, sweating means the target has come back down.

§21.6 Adaptive immunity is specific, systemic, and remembers, at the cost of a 6–10 day lag. Lymphocytes generate one random receptor each by combinatorial V(D)J rearrangement, then face positive selection (must bind self-MHC) and negative selection (must not bind self-antigen strongly); only ~2% survive. Clonal selection is the organizing idea: antigen does not instruct, it selects a pre-existing clone, which expands into effector and memory cells. Slow first responses and fast second ones are arithmetic consequences of starting population size, not a separate memory mechanism.

§21.7 B cell activation needs antigen plus helper T cell licensing; T-independent polysaccharide antigens produce IgM without memory, which is why conjugate vaccines exist. Plasma cells secrete ~2,000 antibodies per second; memory B cells persist. An antibody's variable region determines specificity and its constant (Fc) region determines function. IgG is the tissue workhorse and the only placental class; secretory IgA guards mucosae without fixing complement; pentameric IgM is first and best at agglutination and complement; IgE arms mast cells; IgD is a B cell receptor. Antibodies act by neutralization, agglutination, precipitation, and complement fixation. Active immunity is slow and durable; passive is immediate and temporary. Vaccination buys the secondary curve without the disease.

§21.8 T cells see peptide only in an MHC groove. Class I is on every nucleated cell and reports internal state to CD8 cytotoxic cells — self-surveillance. Class II is on antigen-presenting cells only and reports the sampled environment to CD4 helper cells — alerting. Helper T cells are the coordinating hub for B cells, cytotoxic T cells, and macrophages, and require co-stimulation as a safety interlock. Cytotoxic T cells kill by perforin and granzymes, inducing apoptosis inside a sealed synapse so that contents are contained. Regulatory T cells enforce peripheral tolerance. Allograft rejection is this machinery aimed at foreign MHC, and immunosuppression demonstrates that immune surveillance of tumours is real.

§21.9 Immunodeficiency: SCID removes both arms congenitally and presents as maternal IgG decays; HIV removes CD4 helper cells and therefore collapses humoral, cellular, and macrophage-dependent defense at once. Hypersensitivity types I–III are antibody-mediated and fast and transferable by serum; type IV is T cell-mediated, delayed 12–72 hours, and transferable only by cells. Anaphylaxis is systemic mast cell degranulation, and epinephrine reverses every feature through α₁, β₁, and β₂ receptors. Autoimmunity follows failure of central and peripheral tolerance through escape, Treg failure, molecular mimicry, sequestered antigen release, altered self, or inappropriate class II expression. Immunosenescence and inflammaging together make the old less able to respond to new antigens and more inflamed at rest — and often afebrile when infected.

§21.10 Acute inflammation resolves or scars; chronic inflammation destroys and repairs simultaneously. Regeneration requires dividing parenchyma and an intact framework; fibrosis is the only option for permanent tissue such as cardiac muscle. Atherosclerosis is a chronic inflammatory disease of the artery wall, and plaque rupture is an inflammatory event. Sterile inflammation driven by DAMPs explains the response to an infarct, which heals in inflammatory (days 1–3), proliferative (days 3–14, with the wall weakest at days 4–7), and maturation (weeks 2–8) phases. Clearance and scarring are one program executed by one cell population, so inflammation cannot be "helping" or "hurting" — only its magnitude and timing are modifiable.

Case File 21 · Resolution

Question 1 — Why did the immune system respond to tissue that was never infected? What is it actually detecting?

Because inflammation is triggered by tissue damage, not by infection. The receptors that initiate it — Toll-like receptors and the other pattern recognition receptors of §21.3 — do not detect "foreignness." They detect molecular patterns that indicate something has gone badly wrong.

Two classes of pattern bind those receptors. PAMPs (pathogen-associated molecular patterns) are molecules characteristic of microbes: lipopolysaccharide, peptidoglycan, flagellin. DAMPs (damage-associated molecular patterns) are normal intracellular molecules that appear where they should never be — outside a cell.

When Amara's cardiomyocytes died of ischaemia, they died by necrosis rather than apoptosis, because apoptosis is ATP-dependent and there was no ATP. A necrotic cell swells and ruptures, spilling ATP, mitochondrial DNA, HMGB1, heat shock proteins, uric acid, and its whole cytosolic composition into the interstitium.

Every one of those is a DAMP. Extracellular ATP, mitochondrial DNA that still resembles bacterial DNA in its unmethylated CpG motifs, and HMGB1 all bind pattern recognition receptors on resident macrophages and mast cells. The receptor cannot tell the difference and, evolutionarily, had no reason to develop one: for most of the history of multicellular life, cells that burst open and bacteria that got in occurred together. A receptor tuned to either was tuned to both.

The cascade then proceeded exactly as §21.4 describes. Complement was activated; mast cells degranulated; macrophages secreted IL-1, IL-6, and TNF-α; endothelium expressed selectins and ICAM; neutrophils marginated, underwent diapedesis, and followed chemotactic gradients into the infarct. Her lab values are that cascade, measured:

• WBC 11.4 K/µL with 78% neutrophils — IL-1 and TNF-α mobilizing the bone marrow reserve pool. The neutrophil predominance places this in the first 72 hours of an acute response.
• CRP 38 mg/L — an acute-phase protein synthesized by hepatocytes in response to IL-6. The liver is participating in an inflammatory response occurring in the heart, via a cytokine travelling in blood. CRP is itself functional: it binds phosphocholine on damaged membranes and acts as an opsonin, activating the classical complement pathway without antibody.
• ESR 44 mm/h — a downstream consequence of the same acute-phase response. Elevated fibrinogen neutralizes the negative surface charge that normally keeps erythrocytes apart, so they stack into rouleaux, and stacked cells sediment faster. ESR is an indirect, slow measure of the same hepatic response CRP measures directly and quickly.
• Lymphocytes 14% — the relative lymphopenia of an acute stress response, driven partly by cortisol-induced redistribution of lymphocytes out of the circulation.

The attending was right to decline further cultures and antibiotics. This is sterile inflammation, and it is not merely permissible — it is the necessary first phase of healing.

Question 2 — The inflammation that removes the dead muscle is the same process that lays down the scar. Is it helping or hurting, and can they be separated?

The question cannot be answered as posed, and understanding why is more valuable than an answer would be. Clearance and repair are not two processes that happen to be linked. They are sequential phases of one program executed by one cell population.

Follow the macrophage. On day 2 it is pro-inflammatory: it phagocytoses necrotic myocytes, secretes IL-1 and TNF-α, and generates reactive oxygen species. Around days 3–5 it begins eating the apoptotic neutrophils that arrived before it, and that act — efferocytosis — is itself one of the signals that flips its phenotype. The same cell now secretes TGF-β, VEGF, and growth factors, recruits fibroblasts, and drives angiogenesis. It has not been replaced by a different cell type with a different agenda. It has changed jobs.

So there is no version of the program that removes the necrotic tissue and then stops. The sequence exists because, in every tissue where it evolved, a cleared defect that was not filled would be a hole.

What can be said precisely is this:

• Phase 1 is mandatory. Necrotic myocardium left in place would keep releasing DAMPs and drive chronic inflammation indefinitely, while providing no mechanical strength. When phase 1 has been pharmacologically suppressed — high-dose corticosteroids in the 1970s, NSAIDs since — the result is thinner scar, more infarct expansion, and more free-wall rupture. Less inflammation gave worse outcomes.
• Phase 3 is mandatory. Cardiac muscle is a permanent tissue: cardiomyocytes essentially cannot divide, and the framework was destroyed, so regeneration fails both of the two tests in §21.10. Without a collagen scar the ventricular wall could not withstand systolic pressure and would aneurysm or rupture. The scar is the reason Amara has an intact ventricle at six weeks.
• The cost is permanent and real. The scar cannot contract, reducing stroke volume. It is stiff, impairing diastolic filling — the mechanism of her heart failure with preserved ejection fraction. It conducts abnormally, creating a substrate for re-entrant ventricular arrhythmia. And it is thinner than the wall it replaced, so by Laplace's law it bears higher wall stress, which drives adverse remodelling of the surrounding healthy myocardium.
• What medicine can change is magnitude and timing, not existence. Reperfusing the artery early limits how much muscle dies, and therefore how much scar forms; this is the entire rationale for measuring door-to-balloon time in minutes. ACE inhibitors and beta-blockers limit remodelling of the surviving myocardium. Cardiac rehabilitation lowers her baseline inflammatory tone by 20–40% over three to six months.

Amara's inflammation saved her ventricle's structural integrity and cost her its contractile function, and both were the same event. This is the most important idea in the chapter, and it generalizes: the immune system is not a friendly system. It is a necessary one, with an obligatory bill.

Question 3 — Why is a low-grade fever expected here, and what would change your mind?

Why it is expected. Fever is not a response to infection. It is a response to pyrogenic cytokines, and pyrogenic cytokines are produced by activated macrophages regardless of what activated them.

The chain is short and entirely accounted for. Macrophages clearing Amara's infarct engaged their pattern recognition receptors on DAMPs and secreted IL-1, IL-6, and TNF-α — the same three molecules that produced her CRP of 38 mg/L. Those cytokines reached the vascular organ of the lamina terminalis and induced PGE₂ synthesis, which raised the hypothalamic set point from about 37.0 °C to about 38.0 °C. Everything after that is normal thermoregulation defending the new target: mild cutaneous vasoconstriction, a slight rise in metabolic heat production, and a measured temperature of 37.9 °C.

Note that her fever and her CRP are the same event measured in two organs. IL-6 reaching the liver produced an acute-phase protein; IL-6 reaching the hypothalamus produced a set-point shift. One cytokine, two effectors. If you accept the CRP as sterile in origin, you must accept the fever as sterile too — they cannot have different causes.

The magnitude and timing also fit. Post-infarct fever is characteristically low grade (37.5–38.5 °C), begins within 24–48 hours as neutrophil infiltration peaks, and settles over about three to seven days as the inflammatory phase gives way to the proliferative phase. Amara is on day 2, at 37.9 °C, with a neutrophil-predominant leukocytosis of 11.4 K/µL and negative cultures. Every number is where the infarct timetable says it should be.

What would change your mind. This reasoning holds only because the pattern fits, and a clinician who cannot say what would falsify it is not reasoning. Any of the following would break the timetable and demand a different explanation:

The clinical skill on display when the attending declined antibiotics was not "knowing that infarcts cause fever." It was holding a predicted trajectory and committing to it, while naming in advance the observations that would overturn it. That is what distinguishes a diagnosis from a guess.

Systems Integration Case File · Entry 21

Review

Level 1 · Recall

21.1 Redness and heat in acute inflammation are produced by:

a) increased capillary permeability    b) arteriolar vasodilation    c) bradykinin acting on nociceptors    d) neutrophil diapedesis

21.2 All three complement activation pathways converge on the cleavage of:

a) C1    b) C3    c) C5    d) C9

21.3 A natural killer cell is triggered to kill when the target cell:

a) displays a foreign peptide on MHC class II    b) has lost MHC class I    c) is coated in IgE    d) expresses high levels of B7

21.4 MHC class II molecules are found on:

a) all nucleated cells    b) red blood cells    c) dendritic cells, macrophages, and B cells    d) cytotoxic T cells only

21.5 Which antibody class crosses the placenta?

a) IgA    b) IgM    c) IgE    d) IgG

21.6 A tuberculin skin test read at 48 hours is an example of:

a) type I hypersensitivity    b) type II hypersensitivity    c) type III hypersensitivity    d) type IV hypersensitivity

21.7 Fever is best described as:

a) failure of hypothalamic thermoregulation    b) an upward shift of the hypothalamic set point    c) heat generated directly by bacterial metabolism    d) a side effect of vasodilation in inflammation

Level 2 · Comprehension

21.8 A vaccine contains purified protein antigen and an adjuvant. Explain why the adjuvant is necessary, using the two-signal requirement for lymphocyte activation.

21.9 A cytotoxic T cell kills by inducing apoptosis rather than necrosis. Explain why that choice matters for a virus-infected cell.

21.10 Compare active and passive immunity on onset, duration, and mechanism, and explain why a person bitten by a potentially rabid animal receives both immune globulin and vaccine.

Level 3 · Clinical Application

21.11 A 19-year-old college student with a known peanut allergy eats a cookie at a party and within four minutes develops flushing, hives, a hoarse voice, wheeze, and a blood pressure of 78/40. A bystander asks whether he should give the antihistamine in her bag. Explain what is happening at the cellular level and why the answer is no.

21.12 A 34-year-old man is admitted with Pneumocystis jirovecii pneumonia. His CD4 count is 90 cells/µL. His total serum immunoglobulin is elevated, and his B cell count is normal. Explain why he is nonetheless profoundly immunodeficient, and predict two other infections he is at risk of.

21.13 On day 6 after her infarct, Amara spikes a temperature of 38.9 °C with rigors. Her WBC is 19.8 K/µL with 12% band forms, and her CRP, which had fallen to 21 mg/L on day 4, is now 64 mg/L. Explain why this cannot be attributed to her healing infarct, and outline what has changed physiologically.

Level 4 · Integration and Synthesis

21.14 Amara is 45, obese by BMI, insulin-resistant, hypertensive, and has worked night shift for twenty years. Construct the full causal chain from those risk factors to her plaque rupture, identifying at every step whether inflammation is acting as cause or as consequence. Then explain why her CRP carries prognostic information that her LDL does not.

21.15 Design an argument, using only material from this chapter and Chapters 1, 4, 17, 19, and 20, for why the same anti-inflammatory drug can be correct treatment in one patient and harmful in another. Use rheumatoid arthritis and acute myocardial infarction as your two cases.

21.16 Predict, with mechanism, what would happen to a person engineered to have no regulatory T cells but an otherwise normal immune system — and separately, to a person with normal Tregs but a non-functional AIRE gene. Why are these two conditions similar, and how do they differ?

Concept Map to Complete

Copy this onto blank paper and fill in every bracket from memory before checking the chapter. Then, in a different colour, add what you missed.

                          IMMUNE DEFENSE
                                │
              ┌─────────────────┴──────────────────┐
       [ __________ ]                       [ __________ ]
       fast, no memory                      slow first, HAS memory
              │                                     │
      ┌───────┴────────┐              ┌─────────────┴─────────────┐
  1st LINE          2nd LINE      [ ________ ] arm         [ ________ ] arm
  [ _______ ]    ┌──┼──┬───┬───┐   cell: [ ______ ]        cell: [ ______ ]
  [ _______ ]    │  │  │   │   │   effector: [ _____ ]     kills via
  chemicals:  phago NK infl  anti-  product: [ ______ ]    [ _______ ] +
  [ _______ ]  cytes    amm  micro-                         [ _______ ]
  [ _______ ]             │  bial      │                         │
                          │   │        │                    reads MHC class
              4 cardinal signs│    5 classes:                [ __ ] on
              [ ___ ] ← vaso- │    [__][__][__][__][__]      [ __________ ]
              [ ___ ] ← dilat.│                              cells
              [ ___ ] ← perme-│    4 mechanisms:
              [ ___ ] ← kinins│    [ ________ ]         helper T reads MHC
                              │    [ ________ ]         class [ __ ] on
                    COMPLEMENT│    [ ________ ]         [ __________ ] only
                    3 pathways:    [ ________ ]
                    [ _______ ]                    CENTRAL IDEA =
                    [ _______ ]                    [ ______________ ]
                    [ _______ ]
                    3 outcomes:              FAILURE MODES
                    [ _______ ]        ┌──────────┼──────────┐
                    [ _______ ]    too little  too much   wrong target
                    [ _______ ]    [ ______ ]  [ ______ ]  [ ______ ]
                                   example:    4 types:    mechanisms:
                                   [ ______ ]  I [ ____ ]  [ __________ ]
                                               II [ ___ ]  [ __________ ]
                                               III[ ___ ]
                                               IV [ ___ ]

  ═══ AND THE ONE-WAY RULE THAT ORGANIZES BOTH HALVES ═══
      [ __________ ] immunity INSTRUCTS [ __________ ] immunity
      via antigen presentation, [ ______________ ], and cytokines
      ⇒ a vaccine must contain BOTH the antigen AND the [ ________ ]

  ═══ THE FORK AFTER INFLAMMATION ═══
      can the parenchyma divide?  ─── YES ──► [ ____________ ]
                                  ─── NO ───► [ ____________ ]
      cardiac muscle is a [ __________ ] tissue ⇒ only one option

Lab / Self-Exploration

1. Watch inflammation happen, with a stopwatch. The next time you get a small cut, splinter, or insect bite, record the four cardinal signs in order with times: when did redness appear, when heat, when swelling, when pain, and when did each resolve? Compare your timeline against the numbered sequence in Figure 21.2. Most people are surprised by how fast steps 2 to 4 complete, and by how much later the swelling peaks than the redness.
2. Track a fever properly, if you get one. Record temperature every two hours along with whether you feel cold or hot and whether you are shivering or sweating. Plot the two together. Shivering should precede the peak (set point above current temperature) and sweating should follow it (set point below current temperature). You will have measured a set-point shift in yourself, which is the Chapter 1 concept made personal — and you will never again read shivering as "getting worse."
3. Read your own vaccination record against Figure 21.5. For each vaccine, identify how many doses were in the primary series and how many boosters followed. Then explain, for three of them, why that particular schedule was chosen: a live vaccine deferred to 12 months, a conjugate vaccine begun at 2 months, a tetanus booster every 10 years. Each answer is a statement about maternal antibody decay, T-independent antigens, or memory pool size.
4. Find the two-signal rule in a drug label. Look up a transplant immunosuppressant (tacrolimus, mycophenolate, belatacept) and identify which step of §21.6–21.8 it blocks: IL-2 production, lymphocyte proliferation, or co-stimulation. Then predict, from that step alone, which infections that drug should predispose to. Compare your prediction with the label's warnings.
5. Test the delayed-versus-immediate distinction. If you have ever had a reaction to nickel jewellery, a plaster, or poison ivy, recall how long it took to appear. Anything appearing in minutes is antibody-mediated (type I); anything appearing over one to three days and peaking at 48–72 hours is T cell-mediated (type IV). The clock alone classifies the mechanism, which is the single most useful practical consequence of Figure 21.7.
6. Map your own barriers. Over one day, list every action that maintains a surface barrier — handwashing, blinking, coughing, urinating, the acid in your stomach doing its work after a meal, the flora on your skin. Then list anything that breached one: a paper cut, a course of antibiotics, an antacid, a piercing. The exercise takes five minutes and makes §21.2 concrete in a way that reading it does not.

Key Terms

Next: Chapter 22 · The Respiratory System — where the fluid backing up behind Amara's newly scarred left ventricle reaches the pulmonary capillaries, and where the same Starling forces that swelled her ankles produce the finding that makes her unable to lie flat.