Immunity
J R Rodgers, Baylor College of Medicine, Houston, TX, USA
ª 2009 Elsevier Inc. All rights reserved.
Defining Statement
Introduction
Definitions
Immunocytes (Figure 1)
Trafficking
Fundamentals of Innate Immunity
Acquired Immunity
Antigen Processing and Antigen Presentation
Glossary
acquired immunity Immunity elicited during a primary
response by clonal expansion of immunocytes bearing
antigen-specific receptors and characterized by the
ability to establish immune memory, from which a robust
and rapid secondary response can be recalled.
adjuvant A compound that when formulated with
antigens nonspecifically promotes an antigen-specific
response. Adjuvants have both immune-stimulatory and
antigen storage (depot) properties.
anamnesis (1) The restoration of memory from a state
of (apparent) memory loss. (2) Recall responses of
acquired immunity.
antigen (1) Any molecule recognized by an antigen
receptor. (2) Any molecule or compound that induces
the formation of antigen receptors specific for it.
antigen receptor (1) A molecule, generated by somatic
mutation, that binds its cognate antigen specifically. (2)
The receptor for antigen on T cells (TCR) or on B cells
(also known as antibody).
autoimmunity Pathological conditions in which
antigen-specific receptors interact with self-antigens
and promote immune destruction of healthy tissues.
B cell An immunocyte expressing a functional B cell
antigen receptor (antibody/ immunoglobulin).
complement A proteolytic cascade of proteins whose
sequence, initiated by antigen–antibody complexes or
by innate mechanisms, results in acute immune
responses, including chemoattraction of immunocytes,
and cell lysis.
costimulation A process of activating T cells
nonspecifically through certain receptors, notably
CD28, that provides the ‘second signal’ to complement
the antigen-specific primary signaling through the T cell
receptor.
cytokine A soluble protein hormone bioactive on cells
of the immune system.
MHC Restriction
B Cell Responses
Complement
Patterns of Regulation in Immunology
Allergy and Autoimmunity
Vaccines
Further Reading
immune deviation (1) The shifting or skewing of a T cell
response from Th1 to Th2, or vice versa. (2) Any similar
skewing.
immunoglobulin A family of antigen-binding proteins
produced by B cells.
inflammation A state of cells or tissues characterized
by the release of inflammatory compounds, typically
resulting in the clinical signs of swelling, redness, and
pain, but mediated by a variety of soluble factors and
cell types.
innate immunity Immunity to potential pathogens
based on constitutive mechanisms, such as the skin,
or inducible mechanisms, mediated by pathogen
recognition receptors.
memory (1) Formally, the property of a system by which
a change in the system results in a long-term or
permanent change in the system. (2) The formation of
memory T and B cells that allow subsequent stronger
and more rapid responses.
opsonization The process by which bacteria and other
target particles are tagged by antibodies and
complement components for subsequent phagocytosis
by macrophages.
T cell A leukocyte expressing a functional T cell
receptor (TCR) and recognized generically by the
presence of the CD3 molecule at the cell surface.
tolerance The state of nonresponsiveness to self- or
foreign antigens. Central tolerance is defined by clonal
deletion or inactivation during immunocyte development
in primary lymphoid organs. Peripheral tolerance
requires deletion or inactivation of clones in peripheral
tissues and/or active suppression by suppressive
mechanisms.
vaccine A formulation of antigen and adjuvant that
elicits a protective acquired immune response.
481
482 Pathogenesis | Immunity

Abbreviations LRR leucine-rich repeat

ADCC antigen-dependent cellular MAC membrane-attack complex
cytotoxicity MAMP microbe-associated molecular
AID activation-induced deaminase pattern
AIRE autoimmune regulator MBP mannose-binding protein
APC Antigen-presenting cells MHC major histocompatibility complex
CMV cytomegalovirus molecules
CTL cytolytic T cells NK natural killer
DAF decay-accelerating factor NOD Non-obese diabetic
DC dendritic cells PAF platelet-activating factor
dsRNA double-stranded RNA PAMP pathogen-associated molecular
ER Endoplasmic reticulum patterns
FDC follicular dendritic cell PRR pattern-recognition receptor
GALT gut-associated lymphoid tissues RAG recombination-activating gene
IFN interferon RAGE/AGER advanced glycation end-product
IKDC interferon-producing killer DC receptor
IL interleukin RSS recombination signal sequences
ITAM immunoreceptor tyrosine-based TAP transporter associated with antigen
activating motif processing
ITIM immunoreceptor tyrosine-based TCR T cell receptor
inhibitory motif TLR toll-like receptor
KIR killer-inhibitory receptors TNF tumor necrosis factor
LFA low-affinity integrin ULBP cytomegalovirus UL16-binding
LGL large granular lymphocytes protein
LPS lipopolysaccharide

Defining Statement
This article summarizes key concepts and features of
innate and acquired immunity, with an emphasis on
mammalian systems. It elucidates terminology and meta-
phors that permeate the literature of immunology, and
outlines the major mechanisms allowing for pattern and
antigen recognition and effector function by immuno-
cytes and their soluble products.
Introduction
Immunity is an extensive topic, worthy of an encyclope-
dia of its own. Here we cannot summarize the field in
detail, but will identify key concepts. These concepts
include (1) the difference between innate and acquired
immunity and how they relate to each other; (2) the
notions of specificity and immune memory; (3) the some-
times antagonistic concepts of self and danger; and (4) the
mutually defined ideas of an antigen and its receptor. This
article will arm the microbiologist not with a storehouse
of information, the classic goal of an encyclopedia, but
with a fundament of understanding with which to read the
larger literature of immunity.
The word ‘immunity’ derives from the Latin immunitas,
the legal status of Roman city-states granted immunity
from paying tributes to Rome or to individuals freed from
municipal duties; the root munis referring to change and
(ex)changeable goods. This is the direct origin of the legal
meaning of ‘immunity from prosecution’, but , in the first
century, Lucan (De Bello Civile) had already used the word
metaphorically to describe the Psylli of North Africa as
immune to the bites of venomous snakes. Biological
immunity can refer to constitutive physical innate
mechanisms, such as the physical protection afforded
against infection by skin, the activity of natural killer
(NK) cells against virus-infected cells, or the natural
resistance of mice to diphtheria toxin because of the
absence of a receptor for that toxin. Immunity can also
be innate but inducible, as in the antiviral state induced by
exposure to double-stranded RNA (dsRNA). Finally,
immunity to specific microbes can be acquired during
the lifetime of the individual by infection or vaccination.
The origins of immunology as a science are lost in
antiquity but have always been fundamentally connected
with microbiology. It was certainly known before the begin-
ning of the Common Era that survivors of certain plagues
(perhaps smallpox) were immune to its recurrence.
Observations such as these were rendered uncertain by
imprecise diagnoses of the illness, but advanced sufficiently

so that by the end of the first millennium, Common Era,
Chinese and Hindu healers were aware of the efficacy of
the homeopathic practice of insufflation, in which pow-
dered scabs of the afflicted were blown through straws
into the lungs of healthy individuals. This observation
drew these ancient doctors to a fundamental insight on
acquired immunity – some property of the diseased could
induce long-standing and specific protection in naive indi-
viduals. Centuries of observations and reconceptualizations
about the specificity of this protection led to Fracastoro’s
fourteenth century germ theory of infectious disease, which
held that infectious diseases were caused by disease-specific
agents. This core concept was dealt a minor blow in the late
1700s, when Jenner found that vaccination with cowpox
protected against the different though closely related dis-
ease smallpox. However, the successes of both variolation
and vaccination spurred the deliberate experiments of
Pasteur in the following century to develop attenuated
vaccines and a modern version of the Specific Germ
Theory of infectious disease. The modern concepts of
acquired immunity maintain that induced protection is
specific for each infectious agent, but recognize specificity
at the level of molecules rather than microbes, which
allows cross-reactivity and self-reactivity (autoimmunity).
Definitions
The various mechanisms of immunity are classified under
the rubrics of innate and acquired immunity. This division
reflects the frequent use of analogies between immunolo-
gical and psychological phenomena including ‘memory’
and ‘the self’. In these metaphors, innate and acquired
immunity correspond to inborn (instinctive) and learned
animal behaviors.
Innate immunity refers both to mechanisms of resistance
to infection existing constitutively in the organism and to
mechanisms inducible by stimulus-response systems resem-
bling the homeostatic mechanisms of general physiology. In
this regard, the terms ‘innate’ and ‘natural’ are used inter-
changeably, both referring to traits essential to the ‘nature’
of the organism. The concept of a specific nature for each
species, traceable to the Greek philosophers, is interpreted
by the modern biologist as referring to traits encoded in the
germ line and subject to neo-Darwinian evolution. The
term innate, preferable as more precise and less imbued
with metaphysical overtones, is gradually replacing the
term natural in the immunological literature. The latter is
retained in phrases such as ‘natural killer cell’, ‘natural
antibodies’, and ‘natural Treg cells’.
For the microbiologist, the chief scientific problems asso-
ciated with the category of innate immunity are to
understand how the host distinguishes between pathogenic
and nonpathogenic microbes, how innate responses inhibit
virulence, how innate immunity contributes to microbial
Pathogenesis | Immunity 483
pathogenicity, how these mechanisms might be modified in
vaccines and therapies, and how microbes coevolve with
their hosts.
Acquired immunity refers to mechanisms of resistance
not encoded in the germ line but evolving during the
lifetime of each organism. Such mechanisms are fre-
quently called ‘adaptive’. However, the phrase ‘adaptive
immunity’ can be misleading and imprecise. It is mislead-
ing because acquired immunity is often maladaptive,
causing allergic and autoimmune disease or increasing
rather than decreasing the pathogenicity of microbes. It
is imprecise because innate immunity can be also ‘adap-
tive’ in both Darwinian and homeostatic senses.
The critical concepts pertaining to acquired immunity
are those of memory, specificity, antigen, the immune self,
and self-tolerance, all intimately related to mechanisms of
acquiring immunity. The chief scientific problems for the
microbiologist include how the host differentiates between
pathogenic and nonpathogenic microbes and between
microbes and its own tissues, the effector mechanisms that
suppress microbial virulence, how acquired immunity can
be enhanced therapeutically or preemptively in the form of
vaccines, and how acquired immunity contributes to allergy
and autoimmunity.
The mechanisms of innate and acquired immunity are
closely integrated, sharing the same or closely related
pathways for inter- and intracellular communication and
effector mechanisms. Moreover, some of the receptors
mediating acquired immunity function in a manner that
closely resembles the receptors of innate immunity. It is
likely, for example, that the acquired antigen receptors of
B and T cells evolved from innate receptors. Innate and
acquired responses are functionally integrated in the
sense that innate responses profoundly influence acquired
responses. Recent years have witnessed a flood of descrip-
tions of ‘the interface’ between innate and acquired
immunity. Such an interface is probably more conceptual
than physiological. Thus, innate and acquired mechan-
isms are thoroughly interdigitating.
Immunocytes (Figure 1)
The history of immunology was long contested by humor-
alists and cellularists. The term humoral immunity refers to
the presence of soluble proteins called antibodies, which
circulate in the blood, one of the four humors of medieval
physiology. The humoral immunologists, typified by Paul
Ehrlich (Nobel 1908) and Karl Landsteiner (Nobel 1930),
had the training and interests of chemists, and were able to
make rapid progress in deciphering the first critical para-
digm of acquired immunity: the quantifiable chemical
specificities and mechanisms of interactions between solu-
ble immunoglobulins (antibodies) and their ligands,
antigens.
484 Pathogenesis | Immunity

T17
Th1 Tfh
Plasmacyte
Th2

Plasmacytoid Thymus
VE-cadherin+ dendritic B cell
CD45+ CD34+ cell DP Treg
endothelial precursor thymocyte CD4 T cell
CD8 T cell
CD34+
γδ T cell
Pluripotent
Lymphoid
stem cell
progenitor
NK cell
Myeloid
progenitor
IKDC

Erythrocyte Monocyte

Megakaryocyte
Neutrophil

Platelets Eosinophil Myeloid dendritic cell

Basophil
Macrophage
Mast cell

Figure 1 Leukocyte development. A working model of the differentiation pathway of mammalian hematopoietic cells. Solid lines
connected with cell symbols reflect well-substantiated pathways. Arrows not anchored in cell symbols, or with dashed lines, indicate
less certain pathways. Earlier analyses suggested a bifurcation of the pathway leading from the pluripotent stem cell to distinct myeloid
and lymphoid lineages. Accumulating evidence suggests that a single lineage may first progressively loose the potential to make
erythrocytes, then myeloid cell types, then B cell lineages.

In contrast, cellular immunology languished in the
wings for more than half a century. The first cellularist
was Ilya Metchnikoff (Nobel 1908) who, inspired by
Darwin’s writings, compared the mechanisms of self-
protection of vertebrates and invertebrates. Primarily a
zoologist, he discovered and named the macrophage. He
may also be regarded as the forefather of innate immu-
nity, though this category was not available in his time.
Although Metchnikoff and the humoralists debated bit-
terly over the relative importance of cellular and humoral
immunity, these two threads were more often discon-
nected than at odds.
Influenced by the early work of the neo-Darwinist
Julian Huxley on the biological concept of the individual,
and aware of the role of cells as producers of antibodies,
Macfarlane Burnet (Nobel, 1960) wove these two threads
into the single fabric of modern immunology, in which
the core mechanism of acquired immunity is based on the
Clonal Selection Theory, in which antigen receptors
expressed by clones of immunocytes evolve during the
lifetime of the organism.
Also under the influence of Huxley’s views on biolo-
gical individuality, Burnet promoted the concept of the
immune self. From this viewpoint, which dominated the
landscape of immunology for half a century, acquired
immune reactions with self-molecules during embryo-
genesis lead to clonal deletion (central tolerance); later
reactions with nonself molecules lead to immune reaction.
The concept of the immune self has been modified
somewhat in recent decades to include observations that
self-tolerance includes peripheral mechanisms as well as
central tolerance. Moreover, the distinction between self
and nonself has been blurred in recent decades by the
understanding that nonself molecules by themselves are
not sufficient to induce an immune reaction, while local
inflammation can stimulate autoimmunity. In the modi-
fied landscape, the distinction between self and nonself is
accompanied by the understanding that recognition of
foreign antigens requires the activation of innate immune
receptors for microbial products and/or self-molecules
that signal the presence of local danger or trauma.
The central theme of cellular immunology is that
immunity is mediated by immunocytes even if the immune
factors they produce are secreted in soluble form. Although
most cells of the body are active in innate immunity, the
cells principally involved are blood cells. Immunocytes
derive from the same hematopoietic stem cell that gives
rise to the red blood cells during vertebrate ontogeny.

The chief functions of the primitive immunocytes are
irritability (allowing response to foreign substances and
cells), mobility (giving access to sites of wounding, infec-
tion, or digestion), and phagocytosis, the ability to ingest
substances and particles including other cells.
The predominant blood cell in the simple circulatory
system of the insect is the plasmatocyte (not to be con-
fused with the B cell-derived plasmacyte of vertebrates), a
macrophage derived from endothelial cells. The hemato-
poietic stem cells of mammals, which reside in the adult
bone marrow, also derive from an endothelial-like cell;
they still express the mucin-like adhesion protein CD34,
a protein otherwise characteristic of endothelial cells.
The CD34þ bone marrow-derived stem cell gives rise
through mitosis to new stem cells and the progenitors of all
blood cells. Differentiation steps driven by members of the
Notch protein family successively give rise to so-called
myeloid and lymphoid progenitors. The former in turn
give rise to red blood cells among others but apparently
have lost the ability to become lymphocytes. In contrast,
lymphoid progenitors loose the ability to form red blood
cells but may be able to generate some myeloid lineages.
Myeloid progenitors generate erythrocytes (and their
end products, red blood cells) and the myeloid series of
leukocytes (white blood cells): monocytes and their dif-
ferentiation products, macrophages and myeloid
dendritic cells (DCs); granulocytes (eosinophils, neutro-
phils, basophils, and mast cells); and platelets. All of the
myeloid leukocytes play important but distinctive roles as
effector and regulatory cells of both innate and acquired
immunity. The lymphoid progenitors give rise to B cells,
T cells, and NK cells. These are also leukocytes but are
also called lymphocytes because of their high concentra-
tion in the lymph fluid.
The principal cells of immunity – the leukocytes – are
of hematopoietic origin; the principal organs are a mix-
ture of leukocytes and cells of other lineages. The so-
called primary lymphoid organs are those that produce
lymphocytes: the fetal liver; the bone marrow, also the site
of adult hematopoiesis; the thymus, training ground for T
cells, located in the superior mediastinum of the chest and
derived from the equivalents of embryonic gill slits. The
secondary lymphoid organs are the sites in which T and B
cells mature, where acquired immunity is initiated and
immune memory maintained. These are the spleen, also
the graveyard of red blood cells, and lymph nodes. Lymph
nodes develop as encapsulated outgrowths of lymphatic
vessels provided with nerve and blood supply. There are
two categories of lymph nodes. The superficial nodes
drain skin and outer tissues; the deep nodes drain the
gut and respiratory mucosa and other deep organs such
as the aorta.
Pathogenesis | Immunity 485
Trafficking
Most white blood cells enter the circulation in mature
form from the bone marrow. However, the T progenitors
first migrate to the thymus where they develop from
thymocytes into T cells. Leukocytes circulate rapidly
through the blood but quickly enter tissues under the
influence of locally produced chemoattractants called
chemokines. These are generated by local cells including
resident macrophages and DCs in response to local infec-
tion or tissue irritation and trauma. Most myelocytes
remain in the circulation but the vast majority of lym-
phocytes remain in the peripheral tissues or lymphoid
organs including the lymph vessels. Steady-state and
induced changes in homing to different tissues is
mediated by changes in the avidity of leukocytes for
adhesion molecules in those tissues. These changes are
induced chiefly by chemoattractant protein hormones
called chemokines secreted by many different cell types,
or by the bacterial tripeptide, N-formyl-methionine–leu-
cine-phenylalanine (fMetLeuPhe).
Although the details differ slightly for different leuko-
cytes, T cells illustrate the general principles of leukocyte
trafficking. Naive T cells circulating in the bloodstream
tumble through capillary beds, making weak transient
interactions with the endothelial wall through contacts
between CD62L (L-selectin) and CD34 on endothelial
cells. Resting T cells express high levels of low-affinity
integrin LFA-1, but exposure to chemokines causes a
rapid increase in affinity and avidity of LFA-1 for its
ligand, ICAM-1, found on many cell types including the
lumenal surface of the capillary endothelium. These
interactions permit the tumbling T cell to form tight
adhesions with the endothelium of high endothelial
venules within lymph nodes. Subsequent burrowing
(called diapedesis or extravasation) of the T cell through
the endothelium gives it access to the tissue space where
it proceeds through amoeboid movement to engage other
cells. Except for tissue macrophages and DCs, such as the
Langerhans cells of the skin, most myelocytes circulate in
the blood and enter inflamed tissues only. Their local
efforts are typically followed by apoptosis within a day;
their cell corpses are engulfed by macrophages. In con-
trast, T cells and other lymphocytes spend most of their
time in lymph nodes and other peripheral tissues, but
occasionally leave those tissues through draining lympha-
tic vessels, tributaries of the two lymphatic systems that
empty into large veins in the chest cavity.
Apart from a few exceptions (notably, the brain), the
lymphatic vessels drain every organ. Like the vessels for
blood circulation, lymphatic vessels are walled by
endothelial cells. However, they lack smooth muscle
486 Pathogenesis | Immunity
cells; movement of cells through the lymphatics requires
peristaltic motion mediated by surrounding body move-
ments. A series of lymphatic valves retards retrograde
motion, allowing lymphocytes to move slowly from tissue
to lymph nodes or spleen, and from lymph nodes or
spleen toward the (left and right) thoracic ducts. The
former is the major route by which lymphoid cells reenter
the blood circulation.
In addition to returning lymphocytes to the blood,
lymph vessels serve to drain fluid from tissues and are
the chief mechanism for transporting digested fat from the
gut. Finally, lymphatic vessels draining infected tissues
provide a mechanism for lymphocytes and DCs in per-
ipheral tissues to find each other in draining lymph
follicles and nodes.
Lymph nodes are specialized organs for initiating and
regulating acquired immunity, and for immune memory.
They are formed during embryogenesis through interac-
tions of blood vessels, which later provide T and B cells,
and lymphatics, which provide DCs. Nodes remain very
small until an acquired immune response is initiated, after
which they grow in size by accumulating macrophages,
DCs, and T and B cells; the T and B cells increase in
number through rapid proliferation.
Fundamentals of Innate Immunity
Immunity starts at the interface of the organism with its
environment, whether that is viewed as the external
milieu, which includes the physical and chemical, or the
internal milieu, which also includes the genetic. At this
level, the distinction between the functions of immunity
and the other life functions of feeding and reproducing is
blurred and every cell exhibits immune functions. Thus,
both the bacterium in restricting bacteriophage DNA and
the neuron, whose impulses can affect the outbreak of
Herpes simplex, exhibit immunity.
Appreciation of the importance of innate immunity his-
torically lagged behind that of acquired immunity, but
recent years have led to tremendous interest in this area.
Innate immunity overlaps with general mechanisms of
inflammation. Inflammation results not only from classi-
cally defined immunity but also from wound healing,
chemical irritations (including those resulting from excess
serum glucose and dietary fat), hypoxia, cellular starvation,
heat shock, and cell death (including cell death due to
normal cellular differentiation). The hallmarks of inflam-
mation are classically defined by three clinical terms: rubor
(redness), tumor (swelling), and dolor (pain). The under-
lying mechanisms, however, are manifold and mediated by
a variety of different cytokines, chemokines, and cell types.
Innate responses are typically rapid, induced by
microbe-derived ligands or self-derived danger signals,
and are memory-free (Figure 2). Innate responses to
Innate Acquired
Response
Dose
Response
Time
Figure 2 Comparative kinetics of innate and acquired immune
response. Both innate and adaptive immune responses are
induced by specific interactions between receptors and their
stimulating ligands (arrows). Innate responses (left panel) are
immediate and usually short-lived (bottom curves); subsequent
responses to the same stimulus exhibit similar kinetics and dose
responsiveness (top curves). In contrast, primary adaptive
immune responses exhibit an apparent delay in response while
undetectable clones expand. Secondary responses are faster
and stronger and respond to lower doses of ligand. Acquired
immunity exhibits system memory; the critical change due to
memory is that the responsiveness of the system has changed.
microbes resemble the homeostatic mechanisms by
which the organism responds to many other variations
in its external and internal environment. That is, temporal
variation in the concentration of chemical ligands is
detected through protein receptors specific for those
ligands and transmitted rapidly to affect the cytoskeleton,
cell motility, and gene expression in responding cells.
The resulting signal is typically rapid and often regulated
by feed-forward activation, feedback inhibition, and mod-
ulation from other response pathways.
Like other homeostatic mechanisms, the hallmarks of
the innate immune response are its immediacy and the
operation of homeostatic controls that restore the system
to its basal state. A true basal state has no record of not
being in the basal state; ideal homeostatic mechanisms
thus erase the memory of system perturbations.
Homeostasis thus explains the memory-free aspect of
innate memory. Thus, an organism exposed to an innate
immune insult, such as exposure to a bacterial lipopoly-
saccharide (LPS), will respond within seconds to that
exposure, but within hours will begin to revert to basality.
Unless excessive exposure leads to irreversible damage,
shock, or death, the wounded organism is healed comple-
tely; after a few days, secondary exposure to LPS induces a
response of similar kinetics, magnitude, and consequence.
Innate immunity is initiated when host cells detect
‘‘stranger’’ signals (biochemicals produced by microbes
but not their eukaryotic hosts) or danger signals (pro-
duced by injured or stressed host cells). At this level,
innate recognition of pathogens closely resembles

receptor-based recognition of foodstuffs and mating
types. The former were originally termed pathogen-asso-
ciated molecular patterns (PAMPs) by Charles Janeway
(a term that stresses the correlation, established by evolu-
tion, between microbes and the need for host defense) but
also called microbe-associated molecular patterns
(MAMPs). The cellular receptors for MAMPs are called
pattern recognition receptors (PRRs) and are the basis for
Janeway’s ‘stranger hypothesis’. This hypothesis states
that innate immune receptors should be sensitive to
local accumulation of microbial products. In contrast,
Polly Matzinger stressed the concept that innate recep-
tors have evolved not to recognize microbial pathogens
per se but to recognize forms of stress characteristic of
infection, including physical trauma associated with
wounding. In her view, innate receptors recognize immu-
nological ‘danger’. A broader organismic view is that a
variety of homeostatic mechanisms respond to a variety of
stresses, different innate mechanisms may respond differ-
ently to different kinds of stress, and a variety of response
pathways are used to integrate those responses appropri-
ately to the kind of stress involved. The stress-responsive
ligands are called ‘alarmins’.
The type specimen of PRR is TLR4, toll-like recep-
tor-4, named for its homology to the innate immune
receptor Toll of Drosophila. TLR4 is a cell-surface recep-
tor present on many cell types, which, especially in the
presence of its coreceptors CD14 and MD-2 (proteins
expressed by macrophages), has a high affinity for LPS,
a product of Gram-negative bacteria. Breach of the skin or
gut by bacteria exposes macrophages to LPS, initiating a
signaling cascade leading to macrophage activation and a
variety of associated innate immune functions, including
phagocytosis, increased motility and gene activation, and
secretion of cytokines, including the inflammatory cyto-
kines interleukin-1 (IL-1), TNF- and interferons, and
chemokines. At low concentrations, it can upregulate its
own expression and the expression of other PRR, includ-
ing other TLR molecules. This leads to a transitory
hypersensitivity of the system to LPS and other MAMP.
At high concentrations, as in septicemia, LPS can induce
toxic shock and death. This results from the release of the
chromatin-binding protein HMBG1 by activated macro-
phages. Necrotic but not apoptotic cells also release
HMBG1, accounting in part for the greater inflammatory
activity of necrosis over apoptosis. As a soluble molecule,
HMGB1 acts as an alarmin by activating the advanced
glycation end product receptor (AGER, also known as
RAGE), which drives late-stage shock. Thus, PRRs,
though encoded by the germ line, need not be constitutive
of the basal state. The immune system, by propagating
states of hyperresponsiveness and waves of signaling, may
find it difficult to return to the basal state, leading to
irreversible pathology, including death.
Pathogenesis | Immunity 487
The theme of induced hypersensitivity is common in
innate reactions. Thus, in most cell types, interferon
induces the expression of PRR that increases cellular
sensitivity to dsRNA, a MAMP associated with replica-
tive intermediates of RNA viruses. As a result, prior
exposure of uninfected cells to interferon will arrest pro-
tein synthesis upon infection. In this sense, innate
immunity can generate a variety of nonbasal states that
therefore exhibit forms of short-term system memory
(Figure 3). In addition, some PRR can be induced by
interferon- (IFN-) produced by T cells during an
acquired immune response.
Another mechanism for response to pattern-associated
molecules is the so-called inflammasome, an intracellular
assembly that couples activation of intracellular PRRs,
Death
MAMP2
More
Heightened effector
responsiveness functions
Effector
MAMP1 functions
Basal state Death
Clonal expansion Effector cell
Signal Two
Heightened
responsiveness Memory cells
Signal One
antigen i Anergy,
ARi cell death
Basal state
Figure 3 The two-signal model applied to innate and acquired
immunity. The top figure illustrates how independent activation of
different pattern recognition receptors by different microbe-
associated molecular pattern (MAMP) molecules can induce
both responsiveness and effector functions during an innate
immune response. Excessive signaling leads to death, while
cessation of signaling allows an immunocyte to recover the basal
state. Note that by removing the immunocyte, death also can
restore the basal state. The bottom figure illustrates a two-signal
model adapted from that of Bretscher and Cohn. Antigenic signal
One is insufficient to drive T or B cells to proliferate and exhibit
effector functions, but can lead to cell death or anergy. Signal
Two from activated antigen-presenting cells (APC) (for T cells) or
activated T cells (for B cells) permits clonal expansion, memory
cell formation, and effector function. Excessive signaling leads to
clonal exhaustion and death. T cells rarely if ever return to a basal
state. The two models are aligned to maximize a hypothesized
evolutionary derivation of acquired immune activation from a
two-signal innate activation model.
488 Pathogenesis | Immunity
such as Non-obese diabetic (NOD), to an enzyme,
capase-1, that activates certain proinflammatory cyto-
kines, including IL-1. An emerging theme is that high
production of IL-1 requires two signals. TLRs signaling
through MyD88 activate transcription from the IL-1 gene,
while inflammasome activation is needed to generate the
active form of IL-1. This limitation may prevent inap-
propriate responses to benign microbes such as gut flora.
Initial innate insult in a peripheral tissue causes the
release of chemokines, especially IL-8, and inflammatory
cytokines, especially IL-6, from local cells. Within min-
utes, the former attract neutrophils, which extravasate
into the infected tissues and phagocytose microbes. The
rapidly accumulating corpses of neutrophils are in turn
digested along with bacteria, by a second wave of extra-
vasating monocytes. Under the influence of inflammatory
cytokines including IFN-, these become ‘angry’ macro-
phages that actively phagocytose particles and secrete the
inflammatory cytokines IFN-, tumor necrosis factor-
(TNF-), IL-1, and IL-12. Among other effects, the last
two cytokines help drive proliferation of NK cells and
differentiation of activated T cells into the T-helper 1
(Th1) or inflammatory T cell pathway.
A third phagocyte activated by innate responses is the
DC – so called because of its many projecting cell pro-
cesses. These cells reside in an immature but highly
phagocytic state in peripheral tissues but rapidly mature
during an innate response. This mobilizes them to cease
phagocytosis and traffic via the nearest lymphatic capil-
lary to the draining lymph node, and upregulate
expression of antigen-processing functions, antigen pre-
sentation molecules, and T cell costimulatory factors.
Within the draining lymph nodes, these DCs potently
present their antigen to T cells in the form of peptides
bound by class I and class II major histocompatibility
complex (MHC) molecules.
The cellular taxonomy of DCs is poorly understood.
The so-called T cell-associated DC can be derived from
both myeloid and lymphoid precursors. The myeloid-
derived DC is apparently the source of most tissue-
resident DCs including the Langerhans cells of the skin
and the hepatic DCs; these might be the major mechanism
of presenting antigens to T cells during primary immune
responses. Through differential expression of costimula-
tory molecules and cytokines, these DCs can also regulate
the differentiation path of T cells responding to antigen.
The chief lymphoid-derived DC is the plasmacytoid DC.
This shares certain cell-surface markers with B cells, and
produces large quantities of type I ( and ) interferons.
By virtue of the latter, plasmacytoid cells are important
effector cells during responses to viral infection. A newly
identified DC is the IFN-producing killer DC (IKDC),
generated from a subset of NK cells when these are
exposed to MAMPs. A large subset of T cells in the gut-
associated lymphoid tissues (GALT) expresses DC
markers, and might be a T cell equivalent to the IKDC.
Finally, the follicular dendritic cell (FDC) of germinal
centers is unrelated to the DC described above. It has
surface receptors for immunoglobulin, and presents anti-
gens to B cells in the germinal centers.
Myelocytes serve both innate and acquired immunity
by helping to stimulate T and B cells or by providing
effector functions for soluble antibodies. Platelets aggre-
gate to assist in blood clotting and wound healing. When
activated, they also release granules containing factors
that are stimulatory for both innate and acquired immu-
nity. Neutrophils, macrophages, and DCs are active
phagocytes; the latter also present antigens to T cells.
Eosinophils and macrophages recognize antibodies – pro-
duced by the acquired immune system – bound to
pathogens, directing antigen-dependent cellular cytotoxi-
city (ADCC) toward those targets. Basophils and mast
cells bear receptors for two classes of antibodies, IgG
and IgE, respectively, and when triggered by exposure
to antigen, release platelet-activating factor (PAF) and
histamine, respectively, to trigger smooth muscle contrac-
tion in the gut and lungs, vascular dilation, and, in
extreme cases, asthma and anaphylaxis. In addition, mast
cells particularly are important in regulating immune
activation and immune tolerance.
Natural Killer Cells
Though recent evidence suggests that they also express a
novel kind of acquired antigen receptor, NK cells are cells
of the lymphoid lineage with predominately innate
immune functions. They reside in tissues throughout the
body and circulate in the peripheral blood as large gran-
ular lymphocytes (LGL); their granules contain the
cytotoxic proteins granzyme B and perforin with which
they attack and kill their target cells: cancer cells, virus-
infected cells, and cells exposed to a variety of other
stresses. NK cells are required immediately after infection
for resistance to many viruses and bacteria and are major
producers of IFN- and IL-2, needed for subsequent
T cell responses.
Metchnikoff described the macrophage as a policeman
charged with maintaining the order of metazoan life. In
vertebrates, NK cells seem to be specialized for the police
function. They are not phagocytic but are heavily armed
with cytocidal weapons. Like macrophages, they recog-
nize and kill target cells decorated with antibodies. In
addition, they bear a large variety of receptors by which
they scan the surface of other cells for evidence of truancy
or other malfeasance. These receptors activate or inhibit
the NK cells, and it is the balance of their signals that
decides whether the NK cell attacks. Inhibitory receptors
scan the surface of target cells for molecules consistent
with a healthy cell. Although this activity was described

originally in terms of a ‘missing self’ hypothesis, it can be
seen more broadly in terms of ‘missing health’.
Among such markers of health are the class I MHC
molecules, present on healthy cells but frequently down-
regulated in cancer or virus-infected cells. Human NK
cells express killer-inhibitory receptors (KIR) that recog-
nize class I MHC molecules; in their absence NK cells are
triggered to kill. Activating receptors scan for positive
signs of stress. For example, the molecules MIC
and cytomegalovirus UL16-binding Protein (ULBP) are
induced by cell stress and frequently appear on tumor
cells; they activate the NK receptor NKG2D.
Receptors for some molecules come in both activating
and inhibitory forms and one of these may serve to illus-
trate the entire paradigm. The receptor NKG2A is
inhibitory due to an immunoreceptor tyrosine-based
inhibitory motif (ITIM) motif in its cytoplasmic tail,
while its paralogous receptor NKG2C is activating due
to an immunoreceptor tyrosine-based activating motif
(ITAM) motif. Both recognize the class I MHC molecule
HLA-E bound with a peptide derived from other MHC
molecules. When human cytomegalovirus (CMV) infects
cells, it inhibits the expression of most class I MHC
molecules, including delivery of peptides derived from
them to HLA-E. As a result, both KIR and NKG2A
receptors lose their signal, inviting NK attack. However,
CMV provides a MHC surrogate peptide to restore
HLA-E function and suppress NK cells through
NKG2A, a suppression countered in part by NKG2C.
MHC Molecules
MHC molecules belong to the immunoglobulin super-
family, of which immunoglobulins – antigen receptors of
acquired immunity – are the type specimens. An impor-
tant subset of MHC molecules plays a role in acquired
immunity by presenting peptide antigens to T cells; it was
this function that led to their initial description as trans-
plantation antigens. The family of MHC molecules,
however, has a variety of other functions, and includes
olfactory receptors, transporters for iron or immunoglo-
bulin, and stress indicators for NK cells.
Inasmuch as NK receptors can differentiate among
class I molecules on the basis of bound peptides, and
given the ability of pathogens to manipulate the peptides
bound to MHC molecules, it is plausible that T cell
receptors (TCRs) evolved from activating receptors on
NK cells. An evolutionary arms race might have driven
the sophistication of viruses to subvert these responses
and of the immune system to detect subversion.
Class I MHC molecules have several critical features
ideal for such an arms race. Their chief characteristic is
the ability to bind tightly to short peptides found in the
endoplasmic reticulum (ER) and transport them to the
cell surface where they can be scanned by lymphocytes.
Pathogenesis | Immunity 489
The peptides themselves are derived from endogenous
proteins partially digested by proteasomes and pumped
into the ER by transporter associated with antigen pro-
cessing (TAP). As a result, class I MHC molecules can
report changes in the population of peptides displayed by
a cell, whether affected by cell stress per se or by the
presence of intracellular pathogens.
NK Licensing
The inhibitory receptors of NK cells recognize the MHC
alleles of their own immune self; NK cells reject cells that
do not express those alleles. This is the ‘missing self’
phenomenon, defined by Klaus Kärre, which predicted
inhibitory receptors and an educative process for NK cells
to learn ‘self’. The latter is provided by two mechanisms:
random expression of NK receptors and licensing. Most
immature NK cells randomly and stably express one or
two MHC-reactive NK receptors from a large germ line
repertoire. As a result, many NK cells do not express any
self-reactive receptors and, if allowed to become cytoly-
tic, would attack healthy cells. Through a mechanism
currently not understood, only self-reactive NK cells
become licensed; the others survive but remain anergic.
Licensed NK cells mature and, thereafter, kill unless
inhibited by the presence of self-MHC molecules.
The Role of Innate Immune Cells in Immune
Activation and Immune Tolerance
It has long been thought that macrophages and NK cells
are inflammatory part of host defenses against microbes
and cancer. However, it has recently been appreciated
that macrophages and NK cells, and perhaps all leuko-
cytes, can play both protective and tolerant roles. In their
protective mode, both cell types are cytocidal. In the
pregnant uterus, and in some tumors, both macrophages
and NK cells can promote angiogenesis and are
tolerogenic.
Acquired Immunity
The linchpin of acquired immunity is the antigen recep-
tor. The antigen receptor is that which acquires new
recognitive specificities, is coupled to effector mechan-
isms, and is coupled to a mechanism ensuring immune
memory. It is the proliferative expansion of rare immu-
nocyte clones expressing antigen-specific receptors that
explains the three chief characteristics of acquired immu-
nity: the relatively slow response of acquired immunity to
initial insults, the establishment of memory, and the more
rapid and robust memory response to secondary infection
or experimental immune challenge.
490 Pathogenesis | Immunity
The antigen receptor of vertebrates shares some of
these features with a parallel mechanism found in the
nematode Caenorhabditis elegans. The vertebrate receptor
is a protein; the receptor in C. elegans is a ribonucleopro-
tein called Slicer (the RNA-induced silencing complex).
In this organism, as in vertebrates, infection by RNA
viruses generates MAMPs in the form of dsRNA longer
than 20 bp. One strand, the ‘guide’ strand of a 20 bp frag-
ment is loaded into Slicer, which can then attack and
destroy many copies of viral RNA. The RNA interference
machinery itself is a PRR, part of innate immunity, while
the addition of the virus-derived guide RNA converts
Slicer into a specific receptor for the pathogen.
This mechanism, also operating in vertebrates, generates
new recognitive specificities coupled to effector mechan-
ism, but no memory is generated. To achieve memory, an
RNA-dependent RNA polymerase in C. elegans replicates
the guide RNAs. The new copies diffuse through the worm,
providing advance protection for noninfected cells, and can
be cytoplasmically inherited by oocytes to engender a sort
of Lamarckian inheritance of acquired immunity.
Vertebrates apparently do not have this polymerase.
There are three critical differences between the
acquired immune mechanism of C. elegans and that of
vertebrates. These concern the mechanism of generating
the new specificity, the mechanism of storing it as memory,
and the consequent locus of storage. In the worm, the
molecular structure of the pathogen instructs the formation
of new recognitive specificity; it imposes its molecular
form on the plastic Slicer. Memory is stored in RNA-
diffusible molecules and maintained by RNA polymerases
activated when the guide encounters its target. In contrast,
the vertebrate receptor is generated in advance by random
mutation of DNA genes encoding antigen receptors. The
clones of cells expressing a mutant gene with new specifi-
city for the pathogen are selectively activated to proliferate
and differentiate: the Clonal Selection Theory of Burnet.
Immune memory is maintained by memory cells that
persist in the body long after the pathogen has left. The
worm acquires immunity by instruction; the locus of mem-
ory is diffusible among cells; memory is maintained by
persistent exposure to the pathogen. In contrast, the verte-
brate acquires immunity through selection; the locus is the
cell; and memory is maintained chiefly by memory cells
even in the absence of pathogen.
Antigens and Antigen Receptors
The antigen receptors of vertebrates are proteins that
bind selectively to chemical moieties associated with the
pathogen. The ligands for these receptors, called antigens,
can be proteins as well, but lipids, sugars, nucleic acids,
and other chemical compounds can also be antigens. The
term antigen historically referred to any chemical able to
generate or induce antibodies (antikorpor), one of two
main categories of antigen receptors. The generative
meaning has since been transferred to the term immuno-
gen, leaving antigen to refer to its role as ligand for
antigen receptors. Unlike the molecular patterns recog-
nized by innate receptors, there is often no pattern to
antigens, no intrinsic foreignness.
The antigen receptors of jawed vertebrates are
expressed by two kinds of cells: membrane-bound anti-
bodies (also called immunoglobulins) by B cells and the
TCR by T cells. These two receptors are closely homo-
logous in sequence and structure and belong to the
immunoglobulin gene superfamily. The existence of a
third kind of antigen receptor in mouse NK cells has
been intimated by studies of NK-mediated hypersensi-
tivity responses to chemical haptens. NK cells from these
mice exhibit properties of hapten-specific transferable
memory. Whether this phenomenon actually represents
a new antigen receptor remains conjectural. Such a recep-
tor might, for example, resemble the gene-rearranging
antigen receptor expressed by immunocytes of jawless
fish (agnathans, sea lampreys, and hagfish), which belong
to the leucine-rich repeat (LRR) gene family.
Antibodies were the first discovered and remain the
paradigm of antigen receptors. With a few exceptions
(camels have antibodies of just a heavy chain), antibodies
are tetramers of two heavy and two light chains, each of
which consists of an N-terminal variable domain, which
binds a specific antigen, and one or more constant
domains, which mediate a variety of functions.
Immunoglobulins consist of a protein family, which in
mammals consists of a heavy chain (IgM, IgD, and several
isoforms of IgG, IgA, and IgE) and a light chain (, ).
The corresponding heavy chain genes go by the Greek
appellations , , , , and ".
The variable regions of a heavy and a light chain
combine to form a single binding site for antigens; the
full heterotetramer then has two full binding sites. In
resting B cells, the heavy chains of immunoglobulins are
expressed as type I transmembrane proteins with a
cleaved signal peptide at the N-terminus of the prepro-
tein, an extracellular multidomain globular protein with
several N-linked glycans, a transmembrane region, and a
short cytoplasmic domain that has no known function
(Figure 4). The membrane-bound immunoglobulin is
complexed with two other proteins,  and , whose
long cytoplasmic domains mediate signaling from the B
cell receptor complex.
B cell activation stimulates alternative splicing at the
39 end of the primary transcript of the heavy chain, lead-
ing to a secreted form lacking the transmembrane and
cytoplasmic domains: the soluble antibody. When
secreted, different isoforms multimerize differently: IgG
and IgE are secreted as singleton tetramers, IgA as multi-
mers of one, two, or three singletons, and IgM as
pentamers, with a total of ten antigen-binding sites.

MHC
TCR
Antibody
Figure 4 Antigen-binding receptors. The antigen-binding
receptors of T cells (left) are composed of two chains, each with
one antigen-binding domain, one structural domain, a
transmembrane (TM) domain, and a short cytoplasmic tail. Each
chain of the T cell receptor (TCR) resembles the light chain of
immunoglobulin, except the latter is not membrane-bound. The
antigen-binding domain of the TCR recognizes a small peptide
presented by an MHC molecule on the antigen-presenting cell.
Antibodies recognize ligands independently of a presenting
molecule. The heavy chain of the immunoglobulin contains
several additional structural domains and can be either
membrane-bound or secreted, depending on the site of
polyadenylation and splicing of the primary RNA transcript. When
cell bound, both receptors are bound tightly by other proteins
that together form a signaling complex (not shown).
Like the antibody, the TCR is a heterodimer of two
chains,  and , or, in a minor set of T cells, the para-
logous chains  and . Each of these has a variable and a
single constant domain; the heterodimer has a single
antigen-binding site. Both chains are membrane-bound
with very short cytoplasmic tails; there is no secreted
form of the TCR. The single heterodimer associates
obligatorily with CD3, itself a complex of several mem-
brane-spanning signaling proteins.
The Antigen-Combining Site of the Antigen
Receptor
Each antigen-binding domain of an antibody or TCR
contains a binding surface, called a paratope, which
binds stereospecifically to its molecular complement on
the antigen, called the epitope. Antigens can thus be much
larger than an epitope, and many antigens contain multi-
ple epitopes.
Epitopes recognized by antibodies can be either linear
or conformational. Linear epitopes are composed of short
polymers of molecules such as polysaccharides, peptides,
or glycolipids. Conformational epitopes are formed by the
juxtaposition of noncontiguous molecules, as in a folded
protein. This distinction has a practical outcome:
Conformational epitopes are available for binding only
when the antigen is properly folded and are usually found
on the surface of proteins. Well-conformed antigens are
needed to elicit antibodies for conformational antigens by
immunization. In contrast, linear protein epitopes are
Pathogenesis | Immunity 491
available for binding when a protein is denatured, as in
denaturing gel electrophoresis and Western blots; they
may be located anywhere in a large protein, and their
corresponding antibodies can be generated in animals
immunized with synthetic peptides.
The molecular size of linear epitopes recognized by
antibodies is limited by the size of the corresponding
paratopes and ranges from a minimum on the order of a
single residue of a sugar molecule or amino acid to a
maximum of about eight residues. At one extreme, both
antibodies and TCRs can bind minimal epitopes, called
haptens, as small as fluorescein. At the other extreme,
antibodies recognize linear polysaccharides as long as
six sugars. The epitopes recognized by TCRs are more
complicated, formed by the folding together of a linear
peptide epitope with an antigen-presenting molecule.
They are conformational in the sense that the TCR
must recognize a well-conformed MHC molecule pre-
senting the peptide. The peptide itself is linear.
Antigen Processing and Antigen
Presentation
The peptide epitopes recognized by the TCR are not
available in the native or even the denatured form of
protein antigen, but must undergo antigen processing.
Most antigens processed for presentation by class I mole-
cules are partially digested by proteasomes within the
cytosol, and are pumped into the ER by TAP. Within
the ER, newly synthesized class I MHC molecules
achieve a mature conformation by folding with those
peptides for which they have binding specificity.
Successfully folded molecules are then transported to
the cell surface for presentation to T cells; class I mole-
cules that do not find peptides fail to fold and are
degraded. Class I molecules are expressed on almost
every cell type; this allows cytolytic T cells (CTLs) to
recognize and lyse any of these cells. Most antigens pro-
cessed for presentation by class I molecules are
synthesized within the presenting cell itself. As a result,
detection by T cells of viral peptides presented by the
class I molecules by most cells is a sign that the latter are
actively infected.
There is one exception to this rule: DCs are able to
transfer protein antigens from their phagosomes to the
cytoplasm where they enter the proteasomes pathway. By
this mechanism, sometimes called cross-priming, DCs in
draining lymph nodes activate T cells specific for virus or
tumor antigens even if the DCs themselves do not express
those antigens. (Cross-priming originally referred to the
ability of antigen-presenting cells (APC) of the host to
present antigens from transferred heterologous cells; the
term is also used to refer to the general ability of APC to
492 Pathogenesis | Immunity
acquire and present antigens from other cells, regardless
of the kind of MHC molecule involved.)
Peptides presented by class II molecules follow a dif-
ferent path. Most antigens enter the pathway of antigen
processing through some form of endocytosis: receptor-
mediated endocytosis and phagocytosis of exogenous
antigens, and autophagocytosis of internal proteins. The
presence on a presenting cell of antigens presented by
class II MHC molecules is therefore no indicator of
infection. Most cells do not express class II MHC mole-
cules; those that do are called professional APC. Of these,
DCs are relatively rare, but extremely efficient at acquir-
ing, processing, and presenting antigens. Macrophages are
less effective on a per cell basis, but very common at the
site of active infection. B cells are not phagocytic, but
because their surface antigen receptors (antibodies) have
very high affinity for antigen, they are very effective at
acquiring (through endocytosis of their own surface anti-
bodies) cognate antigen.
MHC Restriction
The epitopes recognized by antibodies are usually formed –
even when conformational – by the chemical residues of
a single molecule or a single subunit of a multimeric
protein. In some cases, the juxtaposition of residues from
two subunits of a protein can produce a conformational
epitope accessible to the antibody. Such antibodies
recognize the residues of one subunit only in the mole-
cular context of the residues of the second subunit. This
situation is relatively rare for antibodies, but is the rule
for TCRs. Under unusual situations it is possible to
obtain TCRs that recognize small haptens or that recog-
nize antigens in the absence of MHC molecules.
However, most TCRs recognize short peptides only
when presented by particular MHC molecules. This
phenomenon is called MHC restriction; the ability of
the TCR to recognize the peptide epitope is restricted to
the particular allele of the MHC molecule. Indeed, ana-
lyses of the X-ray crystal structures of TCRs engaged
with peptide–MHC complexes revealed that about 90%
of the contact surface is composed of MHC residues.
Class I MHC molecules bind peptides 8–10 residues in
length tightly within a groove. Two or three residues of the
peptide make intimate contact with the MHC molecule,
two or three are available for direct contact with the TCR.
The former property entails that an MHC molecule binds
only a subset of peptides with high affinity – the subset
defined by the two or three ‘anchor residues’; and potential
epitopes may be predicted by their possession of this
‘binding motif’. In terms of peptide binding, class I and II
MHC molecules differ chiefly in that the former enclose
the termini of peptides, the latter allow the termini of
undefined length to extend into the water phase.
The dependence on two to three anchor residues
entails that a given MHC molecule recognizes fewer
than 1 in 400 random peptides, and an immune system
relying on a single MHC molecule would provide ample
opportunities for pathogens to evolve resistance. Most
species of vertebrates express two or three loci each of
class I and class II MHC molecules devoted to presenting
a wide sampling of peptides to T cells. Rather than
increasing the number of loci, which would probably
create problems of excessive self-tolerance, class I-a and
class II MHC genes have become the most polymorphic
of all genes, with up to hundreds of alleles in natural
populations. This enormous diversity has two important
consequences for the organism responding to infection
and three for vaccinologists. First, assuming random mat-
ing in the population, most individuals are heterozygous
at each locus, doubling the number of peptides recognized
by MHC molecules. Second, a large-enough population is
likely to contain some individuals that can generate
acquired immunity to a new pathogen. Third, vaccines
designed to target a peptide that is ‘immunodominant’ for
individuals bearing one set of MHC alleles are not likely
to be effective in the general populace.
Fourth, because of the huge polymorphism of MHC
molecules T cells of each individual must be ‘educated’ to
recognize the particular MHC molecules of their ‘self’. It
is this feature, discussed below under the rubric of thymic
education, that gave rise to the notion that the MHC
defines ‘self’ and that accounts for the fifth consequence
of MHC polymorphism – the violent rejection of organ
transplants from donors not matched for MHC
polymorphism.
The Generation of Antigen Receptor Diversity
The variability of the antigen-binding sites of antigen
receptors depends not on a plasticity of the sites them-
selves, but on the somatic mutation of the receptor genes.
This involves a variety of mechanisms. For example, in
germinal center B cells, the enzyme activation-induced
deaminase (AID) mutates DNA sequences encoding the
antigen-binding site of expressed immunoglobulins. This
receptor editing can increase or decrease the affinity of
the antibody for its antigen, or change the specificity
altogether. AID belongs to a family of enzymes that edit
RNA or DNA, and it is plausible that this or a similar
mechanism of somatic mutation was the primordial means
of antigen receptor diversification. Receptor editing does
not operate on TCR genes, at least in the mouse.
A second mechanism of somatic mutation uses gene
conversion to replace the variable region of an expressible
immunoglobulin locus with variable region segments
from a library of alternatives. This mechanism operates
in chickens and to some extent in rabbits as a primary
mechanism of generating antibody genes, but it can also
 Pathogenesis | Immunity 493

contribute to receptor editing in germinal centers. Lymphocyte Selection

Notably, in both chickens and rabbits, the main organ
According to Burnet’s Clonal Selection Theory, clones of
for generating B cells is not the bone marrow, as in
lymphocytes are selected negatively if they react with
humans and mice, but intestinal lymphoid tissue: the
self-antigens, positively if they react with foreign anti-
Bursa of Fabricius in the chicken and the appendix in
gens. Both forms of selection operate chiefly at the time of
the rabbit. Thus, gene conversion might be considered as
first encounter – negative selection during embryogenesis
a mechanism operating primarily in secondary lymphoid
or early stages of lymphocyte differentiation in the adult,
tissues.
positive selection after encounter with a pathogen. These
A third form of somatic mutation operates as the chief
essential features remain in the more recent models of
means of receptor diversification in B and T cells in the
acquired immunity, but have been modified to some
bone marrow and thymus, respectively. This mechanism
extent to account for the mechanisms of diversification,
requires the site-specific DNA recombination-activating
and to account for MHC restriction.
gene (RAG). In contrast to gene conversion, which trans-
Survival of newly formed lymphocytes in each lineage
fers sequence information though patches of homology,
requires the formation of a functional receptor; thereafter,
RAG mediates rearrangement through short recombina-
the negative and positive clonal selection mechanisms
tion signal sequences (RSS) that allow nearly perfect
differ for B and T cells. Nevertheless, the end result is
placement of variable region cassettes into the expression
similar in both cases: mature lymphocytes rarely express
locus. Although RAG expression is limited to lympho-
autoreactive receptors (self-tolerance) but have the
cytes, there is evidence in sharks that RAG can also
potential for high-affinity reactions with foreign antigens.
rearrange immunoglobulin genes in the germ line itself.
B cells undergo negative selection in their primary
This recombinase is distantly related to site-specific
tissues, followed by positive selection in the periphery.
recombinases of bacterial transposons and to the DNA
Perhaps one-half of the newly formed B cells express
integrases of retroviruses. RAG has been found in verte-
autoreactive receptors. Most of these are destroyed by
brates but not in jawless fishes or invertebrates and might
negative selection, some in the bone marrow and the
have been introduced to vertebrates by lateral transmis-
remaining in the periphery. Self-reactive B cells escaping
sion from bacteria. Thus it appears likely that the
negative selection circulate without being activated in
evolution of vertebrate antigen receptors using RAG
most cases. However, in some individuals, these can be
and site-specific recombination was superimposed on a
activated by autoreactive T cells to release pathogenic
preexisting mechanism mediated by AID and homology-
autoimmune antibodies.
based recombination.
Positive selection of B cells occurs initially in lymph
In mice and humans, RAG operates during the early
nodes or the spleen and, after they form during an
stages of B and T cell ontogeny to generate a single
immune reaction, in the germinal centers of those organs.
functional expressed locus from two or three variable
Under the control of T cells, exposure to cognate antigen
region elements (V, D, and J) upstream of the single
drives the proliferation and differentiation of B cells.
constant region. The large number of different germ line
Selection also operates following the reactivation of
copies of V, D, and J elements generates a large combi-
memory B cells. In this case, clonal variants expressing
natorial diversity in the final product; additional diversity
receptors modified by hypermutation or receptor editing
is generated by imprecision when the segments are joined.
are positively selected on the basis of high affinity for
A final combinatorial diversification is obtained by the
antigen. As a result, the antibodies produced by memory
random selection of heavy and light chains. Through a
responses have higher affinity than those produced by
variety of mechanisms, nearly all nascent B and T cells
primary responses – a phenomenon called affinity
successfully rearrange and express a functional antigen
maturation.
receptor from a single allele only, a phenomenon termed
allelic exclusion. This ‘one cell–one receptor’ mechanism
ensures that the selective fate of each such clone depends
Thymic Education
on the specificity of a single antigen receptor character-
istic of that clone. Selection of the receptor repertoire of T cells differs
Nonvertebrate chordates display an additional strikingly from that of B cells, chiefly to account for
mechanism for generating antigen receptor diversity. MHC restriction. In particular, T cells – called thymo-
The immunocytes of jawless fish clonally express a single cytes during thymic education – first undergo positive
functional copy of a rearranged antigen receptor belong- selection to achieve MHC restriction, followed by nega-
ing to the LRR family. As in the case of the germline copy tive selection to remove autoreactive clones. Later, in the
of chicken immunoglobulin, the germline copy is non- periphery, clones of T cells may expand in response to
functional; an AID-dependent gene conversion mechan- antigen, a process equivalent to peripheral positive selec-
ism has been implicated in making it functional. tion of B cell clones.
494 Pathogenesis | Immunity
As is the case for B cells, generation of the unselected
repertoire occurs in the primary lymphoid organ for T
cells – the thymus. The activity of two kinds of molecules
in T cells ensures that most MHC molecules that survive
the initial round of selection are MHC-restricted. The
first is the coreceptor, CD8 for class I MHC molecules,
and CD4 for class II molecules. Most thymocytes differ-
entiate through a double-positive stage in which they
express both CD4 and CD8; it is the double-positive
thymocyte that undergoes the initial round of positive
selection. The extracellular domains of coreceptors
engage the conserved but class-specific residues of the
MHC molecules. The cytoplasmic domains of CD4 and
CD8 sequester the signaling molecule Lck, a tyrosine
kinase required for TCR signaling. TCRs that do not
engage MHC molecules do not have access to Lck; a
thymocyte bearing an unrestricted TCR therefore dies
from neglect. In contrast, thymocytes whose receptors
engage class I or class II MHC molecules activate Lck
and are positively selected to progress to the next stage.
Only those TCRs that bind with sufficient affinity to
MHC molecules and their embedded peptides are posi-
tively selected. As a result, the thymocytes that emerge
from the thymus are restricted by self-MHC.
Double-positive cells undergo another educative pro-
cess in which they become single-positive thymocytes,
expressing only the coreceptor appropriate to their MHC
restriction. This mechanism is instructive rather than
selective. Double-positive thymocytes that have been
positively selected transiently downregulate CD8. MHC
class II-restricted thymocytes continue to engage their
MHC class II through CD4 and maintain signaling
through Lck. This causes the CD8 locus to shut down
permanently, allowing the thymocytes to mature into
CD4þ thymocytes. In contrast, class I-restricted thymo-
cytes transiently lacking CD8 expression lose signaling
through Lck. Transient loss of signaling causes them to
lose expression of CD4 permanently, regain expression of
CD8, and become CD8þ thymocytes.
During thymic negative selection, thymocytes with
excessive affinity for self-peptides suffer one of two
fates: deletion or conversion to so-called regulatory T
cells (Treg). The first removes most overtly autoreactive
T cells, the second generates a cell that suppresses the
autoreactivity of other cells. Only about 5% of thymo-
cytes survive thymic education; of these, about 5%
become Treg. Commitment to the thymic or ‘natural’
Treg lineage involves expression of the transcription fac-
tor Foxp3 and constitutive expression of the high-affinity
subunit of the IL-2 receptor, CD25. Peripheral T cells
that lack Foxp3 expression can also be induced to express
Foxp3, but the function and stability of these cells in vivo
remains controversial.
Self-antigens present in the thymus include those
expressed by thymic cells themselves or brought into
the thymus through blood vessels. In addition, stromal
cells and cells of the thymus express a protein, autoim-
mune regulator (AIRE), that induces local expression of a
large number of self-proteins, apparently for the purpose
of tolerizing T cells. This phenomenon underscores the
basic correctness of Burnet’s conception of self as defined
by central tolerance. Indeed, deletion of AIRE in mice
and humans causes autoimmune disease.
A small fraction of thymocytes proceed through alter-
native routes of education. Thymocytes expressing
functional  TCRs are positively selected but may
never express coreceptors, and never go through negative
selection. They home from the thymus to the mucosa-
associated lymphoid tissues where they make up about
half the resident T cells. The functions of these cells are
poorly understood, but may have NK-like activity, pro-
ducing cytokines when activated, and may suppress
acquired immunity to food antigens. Another small subset
of thymocytes is restricted by the MHC class I-like mole-
cule, CD1, which binds neither CD4 nor CD8. CD1 binds
peptides, but also binds endogenous and mycobacteria-
derived waxy compounds. Such T cells express a very
limited repertoire of variable region segments that are
preadapted for binding CD1. Their functional similarity
to NK cells and invariant use of a single TCR- chain
have garnered these CD1-restricted T cells the appella-
tion of invariant NKT (iNKT) cells.
It should be apparent that positive selection of thymo-
cytes shares at least superficial features of NK cell licensing.
In both cases, a random mechanism generates lymphocytes
with receptors, some of which are self-reactive. Licensing
and positive selection then induce differentiation of self-
reactive cells, and functional inactivation of the others. For
NK cells, NKT, , and Treg cells, positive selection is the
last selective step. There is an important difference: docu-
mented licensed NK receptors are inhibitory, TCRs are
activating.
T Cell Activation and Differentiation
The predominant T cell expresses the  TCR. On
leaving the thymus as a naive T cell, it circulates rapidly
through the blood, entering lymphoid tissues through
high endothelial venules and, if it does not encounter
antigens, it slowly returns to the blood through the
lymphatics.
To detect antigens, amoeboid T cells crawl through
lymphoid tissue, transiently engaging DCs with nonspe-
cific adhesion molecules. Previously activated in
peripheral infected tissues, these DCs have already pro-
cessed antigens and present them in the form of peptides
on MHC molecules. Most naive T cells are nonreactive
and quickly disengage. The rare reactive lymphocyte,
however, detects cognate antigen on the DCs; its TCR
begins signaling to initiate a profound series of cellular

changes. Within seconds the affinity of adhesion mole-
cules is increased, strengthening the adhesion between T
cell and DC, and the T cell stops crawling. Membrane
flow continues, allowing TCRs and the costimulatory
receptor CD28 to accumulate at the center of the interface
with the DC. These form an ‘immunological synapse’ –
similar to the focal adhesion complex of other cells. As
more TCRs and costimulatory are engaged, the strength
of signaling increases and multiple signaling cascades are
activated. These induce expression of additional adhesion
molecules, CD25, the high-affinity receptor for the mito-
genic cytokine IL-2, production of cytokines including
IL-2, very rapid proliferation (doubling every 6 h), and T
cell differentiation.
It is at this time that the T cells differentiate into
various subsets. The chief subsets of effector T cells are
characterized by the coreceptors CD8 and CD4.
CD8þ T cells become CTL. Textbooks have for years
referred to CD8 as ‘suppressors’. This appellation might
merely reflect a different functional assay rather than a
differentiated subset, but one can frequently detect CD8
T cells that express IFN- but not perforin, lending
strength to the notion of different states of differentiation.
CTLs carry large cytolytic granules and express the
cytolytic Fas ligand at their surface. They leave the lym-
phoid tissues, scouring the body for virus-infected cells.
During acute virus infection, the frequency of CTL reac-
tions with an immunodominant epitope may increase
within a few days from a naive level of ten per million
to 10%. The bulk of these cells die, leaving behind a
residue of memory cells.
Many CD4 T cells proliferate and remain resident in
the lymphoid tissues in which they are activated. Here
they contribute to the formation of secondary lymphoid
follicles as part of the germinal center reaction, a complex
of dividing B and T cells, T cell-associated DCs, and
follicular DCs. Within the germinal center, CD4 T cells
subset differentiation is driven chiefly by the balance of
cytokines. Five subtypes or lineages of CD4 T cells are
currently recognized.
Inflammatory innate responses drive IL-12 production
by DCs. Th1 cells are induced especially by the cytokines
IL-12 and IFN- and themselves express the lineage-
critical transcription factor T-bet, the cytokines IL-2,
TNF-, and IFN-. They promote the activation of
cytolytic cells (T cells and NK cells) and cytolytic anti-
bodies (subclasses IgG2a and IgG3); their function is
critical for resistance to viruses and bacteria. In addition,
IFN- inhibits the proliferation of Th2 cells.
In the absence of IL-12 or IFN-, CD4 cells tend to
develop through the Th2 pathway. Th2 cells are induced
by the cytokines IL-4 and IL-10, express the lineage-
critical transcription factor GATA-3, and produce the
cytokines IL-4 and IL-10. IL-4 is mitogenic for Th2
cells but suppresses proliferation of Th1 cells, while
Pathogenesis | Immunity 495
IL-10 suppresses IL-12 production by DCs. In addition,
IL-4 promotes the development of B cells secreting IgG1
and IgE. Th2 cells are essential for resistance to extra-
cellular pathogens including fungi and nematodes, and to
the pathological development of many forms of allergy
and of asthma.
The differentiation of Th1 and Th2 cells tends to be
mutually exclusive because the cytokines produced by
each type of cell are self-promotional and mutually sup-
pressive. As a result, immune response to vaccines and
pathogens tend to be ‘skewed’ or exhibit ‘immune devia-
tion’. Indeed, many of the T cell suppressor reactions
described in the 1980s were probably examples of
immune deviation.
The cytokine TGF- induces expression of Foxp3 in
the otherwise uncommitted CD4þ T cells and, in some
conditions, these cells functionally resemble thymus-
derived Treg cells. Both thymus-derived and induced
Treg cells secrete TGF- andIL-10, both of which sup-
press Th1 cells. In addition, Treg cells are able to suppress
the activation of other T cells through a cell contact-
dependent process, the mechanism of which is otherwise
unknown. The stability of the functional state of induced
Treg cells and their importance in vivo or in therapeutic
applications remain controversial.
The inflammatory cytokine IL-6, produced by many
cell types, drives CD4þ cells to express another lineage-
critical transcription factor, RORT, and become T17
cells. T17 cells secrete the inflammatory cytokine IL-17
that further promotes the T17 pathway, recruits neutro-
phils, and stimulates the proliferation of keratinocytes.
T17 cells promote the development of autoimmunity.
A fifth subtype of CD4þ T cell has been proposed and
characterized as a follicular helper T (TFH) cell.
Generation of these cells appears to require IL-21, and
they themselves secrete the same cytokine. They home to
the follicles of germinal centers where they appear to
cooperate with B cells and generally to promote antibody
production. The transcription factor Bcl-6 might be
important for formation and/or maintenance of this cell
type.
T cell memory
After an acute response to vaccination or infection, most
effector T cells die. A residue of cells serves as memory
cells, which may persist for decades in humans in the
apparent absence of antigen. They proliferate very slowly
to maintain a steady state. There are two kinds of memory
cells. Central memory cells closely resemble naive T
cells, replicate rapidly upon secondary exposure to anti-
gen, but replicate before expressing effector functions. As
a result, secondary memory responses from central mem-
ory cells are strong but delayed. In contrast, so-called
memory-effector cells often lack the costimulatory recep-
tor CD28 and proliferate very poorly, but maintain the
496 Pathogenesis | Immunity
effector function. For example, like NK cells, memory-
effector CTL can lyse targets within minutes without the
need to proliferate. Indeed, many memory-effector cells
exhibit surface markers and other functions characteristic
of NK cells and have been described as another kind of
NKT cell.
The mechanisms of immune memory are difficult to
study partly because of the timescale of relevant experi-
ments. It is critical to establish experimentally that a
primary state of acquired immunity has lapsed, to permit
tests of the ability of the immune system, or of parts of it
under study, to restore immunity. The recall of immune
memory from a state of apparent immune amnesia is
called the anamnestic response.
B Cell Responses
The naive B cell expresses its clonotypic immunoglobulin
chiefly in the form of IgM. The paradigmatic response of
B cells to antigen occurs in the germinal center and
requires engagement with Th1 or Th2 and perhaps TFH
cells. This process is apparently initiated by cross-linking
the surface IgM by polyvalent antigen – such as the sur-
face of a virus. Such cross-linking is sufficient for
internalizing, processing, and presenting peptides from
the antigen, and for increasing the avidity of B cell adhe-
sion molecules, allowing the B cells to engage the T cells
they encounter as they traffic through germinal centers. If
the antigen peptides are recognized by T cells, the T cells
also respond by forming an immune synapse with the B
cell and by upregulating expression of CD40 ligand
(CD40L), which activates the costimulatory receptor
CD40 on B cells. In turn, the B cell upregulates expression
of the costimulatory molecules B7 (reactive with CD28
on most T cells) and ICOS ligand (reactive with ICOS on
TFH cells).
Activated by the T cell, the B cell can now proliferate
and differentiate, giving rise to terminally differentiated
plasma cells; most of these migrate to the bone marrow
where they secrete large quantities of soluble antibody
into tissue space and bloodstream. As with T cells, a few B
cells remain to become memory cells. Unlike T cells, but
requiring their help, memory B cells undergo affinity
maturation and receptor editing. If antigen concentrations
remain high, or if there is a secondary exposure to anti-
gen, the B cell undergoes immunoglobulin gene class
switching.
This form of DNA rearrangement is unique to immu-
noglobulins. The heavy chain locus contains a series of
heavy chain constant regions, each encoding a complete
constant region with a class-specific DNA switch region
at its 59 end. The 59 most constant region is , and this is
the configuration first expressed after VDJ recombination,
allowing initial expression of IgM. Reactivation of B cells
induces an AID-dependent recombination event that
deletes the  region and connects the expressed variable
region with one of the downstream constant regions.
Switching is regulated by cytokines. For example, IL-4
induces switching to the 1 or " constant regions, gener-
ating B cells expressing IgG1 or IgE, respectively. In
contrast, IFN inhibits switching to IgG1 or IgE, but
instead induces switching to the 3 or 2a constant
regions, leading to production of IgG3 or IgG2a,
respectively.
Through the mechanism of heavy chain class switch-
ing, the single functional variable region present in the
naive B cell can be expressed with functionally distinct
constant regions in different subclones. The antibodies of
different classes differ in their half-life in the serum, their
ability to fix complement, their ability to be transcytosed
from the placenta into the fetus or in the mammary gland
into milk, across the gut epithelium into the intestine, and
their ability to bind receptors (called Fc receptors) for
immunoglobulins found on different immunocytes. For
example, IgM is moderately active in neutralizing virus, is
extremely effective at activating complement, but does
not cross the placenta or enter milk and is not active in
antibody-dependent cellular cytotoxicity (ADCC)
mediated by NK cells. In contrast, IgG1 is better at
virus neutralization, ADCC, moderately activates com-
plement, is efficiently transported across the placenta, and
triggers release of PAF from basophils. IgA is secreted
efficiently into the gut lumen where it is effective at virus
neutralization. IgE binds to specific Fc receptors on mast
cells; cross-linking with antigen triggers mast cell degra-
nulation and histamine release.
The activities of antibodies depend on two aspects of
their structure. On the one hand, antibody binding to
antigens can neutralize a virus by allosteric changes in
capsid molecules, or by blocking access to the cellular
receptor for the virus. In some cases, however, antivirus
antibodies can increase infectivity, as with IgG antibodies
against HIV, most of which merely increase its ability to
infect monocytes. On the other hand, the constant region
of an antibody couples the receptor to effector functions.
The effect may be to increase the avidity of interaction (as
with pentameric IgM), to target the antibody to specific
cellular receptors, or to bind soluble effector molecules
such as complement.
Complement
Complement represents a series of soluble proteins, pre-
sent constitutively in the blood and tissue spaces that can
be triggered through an explosive cascade of protein
cleavage reactions to activate potent innate immune
mechanisms. The effector mechanisms include the for-
mation of lethal membrane pores in bacteria or cells,

activation of phagocytosis by macrophages, and recruit-
ment of neutrophils.
Activation (fixing) of complement by antibodies was
discovered first and hence is called the classical comple-
ment activation pathway; in this respect it is an important
effector arm of acquired immunity. In this pathway, anti-
gen–antibody complexes bind the complement factor
C1q, initiating its self-cleavage to release activated sub-
units, C1r and C1s; these quickly lead to the cleavage and
activation of C3 convertase. The complement cascade is
also activated by two other mechanisms, which also lead
to activation of C3 convertase. In the ‘alternative’ path-
way, C3 binds to bacterial surface and is induced to
activate itself, generating the C3 convertase. In the lec-
tin-mediated pathway, carbohydrate structures on
microbial surfaces are recognized by the lectin man-
nose-binding protein (MBP) (Table 1). Once cross-
linked, MBP triggers the release of homologues of C1r
and C1s, which activate the C3 convertase.
The common link for all three triggering arms of the
complement pathway is C3 convertase, which triggers
three principal downstream pathways. In one pathway,
C3a and C5a are produced; these are potent mediators of
inflammation and recruit phagocytes. A second product of
activated C3 convertase, C3b, binds to phagocyte recep-
tors, triggering phagocytosis and clearance of antigen–
antibody complexes. This process, by which bacteria
and other particles can be targeted for destruction, is
called opsonization. Finally C3b also initiates the forma-
tion of a membrane-attack complex (MAC) involving the
pore-forming components of complement.
The essential features of complement may be summar-
ized as follows: an initial trigger event initiates the
sequential deposition of complement factors on the trig-
gering surface, targeted either by a MAMP or by an
antigen–antibody complex. Activation of the C3 conver-
tase then triggers formation of the MAC on the same
surface, leading to lysis of the microbe or cell, secretion
Table 1 Representative receptors of innate and acquired immunity
Innate receptors Example
TLR TLR-2
TRL-3
TRL-4
TLR-9
NOD NOD-2
Chemokine receptor Formyl-peptide receptor
Lectin Mannose receptor
NK receptor NKG2Aa
NKG2C
NKG2D
T cell receptor V14 TCR
Alarmin receptor AGER
a
NKG2A is an inhibitory NK receptor; all the other receptors in this table are activating.
Pathogenesis | Immunity 497
of soluble factors that recruit phagocytes, and the target-
ing of phagocytes to the activating surface.
Uncontrolled activation of the complement cascade is
lethal to the host organism. Two general mechanisms
limit its effect. In the first, most triggers target the effects
to a limited surface – a bacterium or infected cell. In the
second, a variety of soluble and cell-associated inhibitory
factors inhibit the complement cascade. For example, the
membrane protein CD55, also known as decay-accelerat-
ing factor (DAF), inhibits activation of C3 convertase.
Patterns of Regulation in Immunology
The mechanisms for triggering and limiting complement
activation illustrate a general feature of both innate and
acquired immunity: an uneasy balance between positive
feed-forward and feedback inhibition mechanisms. Both
innate immunity, in the form of complement, and
acquired immunity, in the form of clonal expansion, can
generate exponential feed-forward reactions to control
the otherwise explosive growth of microbial pathogens.
The effector functions of both forms of immunity are
coupled especially through the classical complement
pathway.
The two arms are also coupled in terms of activation;
innate immunity provides a form of ‘second signal’ to
permit an immune response. In the original formulation
of the Clonal Selection Theory by Burnet and modified by
Joshua Lederberg, clones of B cells would pass through a
tolerizable state en route to maturity. Any clone reactive
with self-antigens would be tolerized; thereafter, exposure
to antigen alone would be sufficient to trigger immuno-
cyte activity. However, Paul Bretscher and Melvin Cohn
advanced arguments that in a system in which
immunocytes could be activated by cognate antigen
alone, autoimmune reactions would occur too readily.
Translated into contemporary terms, they proposed that
Ligand Source
Zyomasan Yeast
Derma RNA viruses
LPS Gram-negative bacteria
Nonmethylated CpG DNA Bacterial DNA
Muramyl dipeptide Gram-positive bacteria
Formyl-peptides Bacteria
N-mannosylated proteins Bacterial glycoproteins
HLA-E with MHC signal Healthy cells
peptides or heat shock protein peptides Stressed cells
MIC Stressed cells
CD1 þ waxy lipids Self þ Mycobacteria
Advanced glycation end products High serum glucose
HMGB1 Necrotic cells
498 Pathogenesis | Immunity
T and B cell responses could only occur if each was
independently activated by cognate antigen and provided
help from another, independently responding, immuno-
cyte. In their formulation, a cognate antigen provides the
first signal to an immunocyte, while the cooperating
immunocyte provides a second, permissive, signal. This
is essentially the model described above for the germinal
center reaction, in which T and B cells provide help to
each other while responding to distinct epitopes on a
single antigen.
While this model explained the germinal center reac-
tion, it did not adequately explain a variety of T cell
responses, including those of CTL, in which B cells are
not involved. In such situations, there was no second anti-
gen-specific immunocyte to provide a permissive signal to
T cells. In such cases, it then appeared that professional
APC would provide the second signal to T cells at the time
of antigen presentation. The general finding that T cell
responses required a variety of soluble (e.g., IL-1) and cell-
bound (e.g., B7) costimulatory signals was seen as consis-
tent with this model. However, if professional APC such as
macrophages and DCs constitutively present antigen, T
cells would always receive a second signal during exposure
to antigen. As discussed by Janeway and by Matzinger, T
cells do not respond to every foreign antigen they encoun-
ter. In particular, there are many foreign antigens absorbed
through the gut or respiratory tracts, to which an immune
response would be unhealthy.
Considerations such as these led to the suggestion that
immunostimulatory signals might be obtained through
innate receptors for ‘stranger’ or ‘danger’ ligands. These
would indicate the actual presence of pathogens or actual
danger to the organism. For a time it seemed that the
concept of ‘danger’ represented a profound challenge to
Burnet’s concept of self and nonself. In particular instead
of requiring a special mechanism for a tolerizable state, it
would be sufficient to assert in the general case that
detecting antigen in the absence of danger signals (alar-
mins, MAMP) should tolerize a T cell. For the present,
however, it appears that both theories will prevail. The
notion of self-antigens seems to readily encapsulate the
function of AIRE in the thymus. Moreover, it is manifest
that thymocytes and many B cells in the bone marrow do
go through a tolerizable state for which there seems no
need to invoke the concept of relative danger. The syn-
thetic view in which concepts of the immune self coexist
with concepts of danger as inheritor of the general notion
of two signals reflects the current thinking of many
immunologists.
Allergy and Autoimmunity
Despite the requirement for a second signal, allergic and
autoimmune responses still occur in many individuals. As
with complement fixation, additional inhibitory pathways
are important to suppress autoimmunity. These include
the activity of Treg cells and other suppressive T cells that
inhibit antigen-specific T cell responses through cell–cell
contact and soluble factors. Innate immune mechanisms
also suppress T cell responses. For example, NK cells can
kill activated DCs.
A long-held model for autoimmunity is that of mole-
cular mimicry, in which a little exposure to microbial
antigens triggers a cross-reactive reaction with self-
antigens; this sterile autoimmunity persists in the absence
of the original infection. One extension of this model has
been the hypothesis that the increase in childhood vacci-
nation explains an increase in frequency of autoimmunity
and other pathologies, such as autism, conjectured to be
autoimmune manifestations. While the twin models of
molecular mimicry and sterile autoimmunity have
experimental support in mice, their applicability in
humans remains conjectural. A somewhat opposite view
is taken by those who conjecture that an increase in
allergy and autoimmunity is caused by excessive cleanli-
ness in early childhood, including the prevention of
childhood infections by vaccines.
An alternative hypothesis is that autoimmunity is
caused not by exposure to antigens, but by innate insults
inappropriately signaling danger to the immune system.
A clinical model for this concept is psoriasis, in which an
acute blow to tissue can initiate local self-perpetuating
pathology.
Vaccines
The steady march of successful vaccines including those
for smallpox and polio led to the hubristic view that
vaccine development was conceptually simple. A series
of vaccine failures, including vaccines for Mycobacterium
tuberculosis, malaria, and HIV, are impressive especially by
underscoring how little we understand about why they
fail. Moreover, HIV induces a robust immune response in
most patients, initially.
The profound challenges facing the microbiologist in
immunology are both fundamental and practical. Among
the former can be included the problems of how the many
mechanisms of innate recognition are integrated to guide
the differentiation of T and B cells, how memory cells are
formed and maintained, and how the different T cell
subtypes interact with each other and other compart-
ments of the immune system. In addition, the functions
of NK cells and several myelocytic cell types remain
poorly understood. Finally, the mechanisms by which
some viruses and bacteria manipulate the immune system
to their own ends are barely understood.
On a practical level, the problem of testing vaccines
for their ability to generate immune memory is nearly

untouched. Pharmacological agents, long used to inhibit
immune functions, have been barely explored for the pur-
pose of enhancing adjuvants. In this respect, new studies of
innate immunity may have elucidated a hitherto unknown
mechanism for adjuvants. Previously, it was thought that
adjuvants work through two mechanisms. First, physical
properties of some adjuvants, such as the emulsion-forming
oils of Freund’s adjuvant, provide a depot effect, allowing
slow and steady release of antigen. Second, the presence of
microbial by-products in some adjuvants (again, the myco-
bacteria-derived waxy compounds present in Freund’s
adjuvant) were thought to activate innate responses. These
explanations failed to account for the properties of alum.
This composite contains aluminum hydroxide and lacks
both oils and bacterial products and has been the only
adjuvant allowed for routine human use. Recent studies
suggest that aluminum hydroxide crystals resemble natu-
rally occurring alarmins (danger signals) and activate
inflammasomes.
It is not clear if the problem with developing vaccines
for malaria, tuberculosis, and AIDS is merely practical or
represents a major deficit in our understanding. After a
century of testing Ehrlich’s ‘magic bullets’, the clinician’s
armamentarium of pharmaceuticals for promoting immu-
nity to infectious disease is nearly empty. The one
exception is the convincing efficacy of interferon to
treat patients with hepatitis C, effective in only half the
cases. Despite the urgent need for treating autoimmunity
and the high hopes of treating cancer with immunother-
apy, the greatest problem of immunity remains that for
which immunity evolved: the control of infection.
The now-classic work on the history of immunology is
Silverstein’s. Highly readable introductory textbooks
might refer to the latest editions of Abbas’, Janeway’s, or
Pathogenesis | Immunity 499
Kuby’s textbooks. Advanced workers should consult
Paul’s Fundamental Immunology, which contains a core of
material difficult to find elsewhere, or Rich’s Clinical
Immunology. Readers interested in the latest reviews on
research topics should consult The Annual Reviews of
Immunology for longer papers and Nature Reviews
Immunology or Trends in Immunology for highly readable
short reviews. High-profile research papers and minire-
views in immunology are published in Immunity, Nature,
Nature Immunology, Science, Blood, and The Journal of
Experimental Medicine. The Journal of Immunology and The
European Journal of Immunology are high-quality journals of
professional immunology. One of the earliest presenta-
tions of the Two Signal model is from 1970.
See also: Immune Suppression; Vaccine Development:
The Development of Avian Influenza Vaccines for Human
Use; Vaccines, Viral
Further Reading
Abbas AK, Lichtman AH, and Pillai S (2006) Cellular and Molecular
Immunology, 6th edn. New York: Saunders.
Bretscher P and Cohn M (1970) A theory of self–nonself discrimination.
Science 169: 1042–1049.
Kindt TJJ, Goldsby RA, Osborne BA, and Osborne RA (2006) Kuby
Immunology, 6th edn. New York: W. H. Freeman Company.
Murphy K, Travers PT, and Walport M (2007) Janeway’s
Immunobiology, 7th edn. New York: Taylor & Francis, Inc.
Paul WE (ed.) (2003) Fundamental Immunology, 5th edn. New York:
Lippincott Williams & Wilkins.
Rich RR, Fleischer TA, Shearer WT, Schroeder HW, Frew AJ, and
Weyand CM (eds.) (2008) Clinical Immunology, 3rd edn. New York:
Mosby.
Silverstein AM (1989) A History of Immunology, 1st edn. San Diego, CA:
Academic Press.