C H A P T E R

1
Autoimmune Disease: Reflections
and Projections
Noel R. Rose
Department of Pathology, Brigham and Women’s Hospital/Harvard Medical School,
Boston, MA, United States

O U T L I N E

Foreword 3 Genetics and Exposures 6

Personal Introduction 3 Epidemiology and Prediction 7
Autoimmunity and Autoimmune Disease 4 Acknowledgment 8
Clonal Balance and Regulation 5

FOREWORD

The overriding message of this book is that autoimmunity is a common phenomenon in normal individuals,
but that it may be a signal of disease, and play a primary or secondary role in the disease process. Autoimmune
diseases are diverse because they occur in differing anatomical sites, but they share many key pathogenetic fea-
tures. In human medicine, they represent a category of disease and benefit from considering them as a family,
sharing a genetic pool and experiencing many similar exposures. Most of this book, therefore, is devoted to
descriptions of the human diseases that are most firmly established to be the consequence of an autoimmune
response. Other chapters of the book detail the common features, their genetic, molecular, and pathologic signifi-
cance, and their application to diagnoses and treatments.

PERSONAL INTRODUCTION

I began my career in immunology when I was a student at the University of Pennsylvania School of Medicine in
the years 1948 51. I was fortunate to receive a fellowship at the University’s Center for the Study of Venereal
Diseases. Syphilis and other sexually transmitted diseases had undergone the expected increase during World War II
and remained a major public health problem in the immediate postwar years. Most of my time was devoted to my
basic research on Treponema pallidum My sole clinical responsibility at the center was to assist in a study of the efficacy
of the newly available drug, penicillin, in patients with primary syphilis. We were provided with samples of chancre
fluid on admission to the clinic. Following penicillin therapy, samples were taken daily until all the spirochetes were
inactive. My job was to enumerate the proportion of viable T. pallidum under the darkfield microscope. With a little
practice, it was quite easy to recognize the living T. pallidum by its well-named “queenly motion,” a dignified profile
worthy of royalty. Other dermal spirochetes showed jerky movements suggestive of plebeians.

The Autoimmune Diseases, 6th.

DOI: https://doi.org/10.1016/B978-0-12-812102-3.00001-4 3 Copyright © 2020 Elsevier Inc. All rights reserved.
4 1. AUTOIMMUNE DISEASE: REFLECTIONS AND PROJECTIONS

At this time, the field of syphilis serology was undergoing a major resolution, occasioned by the introduction
of the treponemal immobilization test (TPI) by Robert Nelson and Manfred Mayer at the Johns Hopkins
University. As the name suggested, the assay was based on the loss of motion of viable T. pallidum produced by
antibodies (and complement) from patients. This test represented the first diagnostic procedure for demonstrat-
ing antibodies specific for the pathogen rather than the standard “Wassermann test” used in various formats in
the diagnosis of syphilis. Intrigued by the fact that the TPI depended upon a loss of motility, I wondered if the
antibody reacted with flagella of the organism. It was not known whether T. pallidum had flagella or moved by
its spiral motion. My first publication was an electron microscope study of spirochetes showing that they do
have flagella-like structures. I never had an opportunity to go on with the hunch that syphilitic patients produce
antibodies to flagellar antigen.

AUTOIMMUNITY AND AUTOIMMUNE DISEASE

The Wassermann test mentioned above was devised by Wassermann et al. at the Robert Koch Institute in
Berlin in the early 1900s (Chapter 2, Autoimmunity: A History of the Early Struggle for Recognition). In its many
modifications, it quickly became the standard serologic test to assist in diagnosis and in following the treatment
of patients with syphilis (Chapter 69). The antigen to which the Wassermann antibody is directed was later char-
acterized by Pangborn at the New York State Health Department as a phospholipid and named cardiolipin
because of its relative abundance in heart tissues. Cardiolipin, a component of cell wall, is found in differing
amounts in virtually any organ of the body and in any vertebrate. Despite a great deal of speculation over the
years, cardiolipin has never been shown to play a pathogenic role in syphilis even though its specific antibody is
still of great value in recognizing and following the infection. It represents the first autoantibody widely used as
a diagnostic reagent in the clinical laboratory. One might speculate that had T. pallidum not already been identi-
fied as the cause of syphilis, cardiolipin, or a replica might have been considered the causative agent of an “auto-
immune” disease, syphilis.
The gradual acceptance of autoimmunity as a frequent consequence of a normal immune response and of dis-
ease as an occasional consequence of autoimmunity is described in Chapter 2. Briefly, studies of human immu-
nology began near the end of the 19th century with experiments on infection by Roux and Yersin at the Pasteur
Institute showing that several major human diseases such as tetanus and diphtheria are attributable to produc-
tion by the respective pathogen of a specific toxin. Behring and Kitasato produced a disease-specific antibody to
the toxin could treat or even prevent the disease. Even in other human diseases in which there is no special toxin,
an immune response could provide protection.
Soon afterwards, Jules Bordet showed that an immune response was generated by parenteral injection of even
harmless substances, such as red blood cells from another species. Thus immunology became the physiological
science not only of infection but of recognition of substances foreign to the host.
A natural follow-on question raised at the time is whether the immune response can be generated by injection
of molecules of the host itself. Early experiments clearly showed that antibodies could be induced quite regularly
to isolated sites such as the eye or the testis. Production of these antibodies were sometimes associated with dis-
ease (Chapter 2).
Paul Ehrlich found that he could produce antibodies to red blood cells of most other goats but never to the
immunized donor goat itself. He suggested that there were natural barriers to such autoimmune responses,
because they could result in harm. These experiments were widely interpreted by most immunologists to suggest
that immune responses to the host itself were not possible. Reports that autoantibodies could be the agents of dis-
ease were greeted with the greatest skepticism. A change in thinking followed research on inflammation as a con-
sequence of immunization.
The concept of inflammation as the body’s response infection or other cell injury goes back to the earliest days
of medicine. Metchnikoff pointed out that, although normally a method of protection, inflammation that exceeds
its normal boundaries can itself cause disease. In contrast to immunity, inflammation is nonspecific; that is, it
does not target to the specific inducing molecule. The major cells initiating the inflammatory process such as
polymorphonuclear and mononuclear cells broadly respond to nonself or altered-self substances. They are not
capable of distinguishing fine differences among molecules as can the specialized lymphocytes of the immune
system (Chapters 4, 10 and 11).

I. GENERAL CONSIDERATIONS
 CLONAL BALANCE AND REGULATION 5
By the 1950s, immunology crossed the line and recognized that uncontrolled inflammation induced by an
autoimmune response can cause disease in the host. This dramatic change in thinking can be related to the intro-
duction of a number of new technical advances in immunology that opened previously impossible opportunities.
They included novel methods for separating mixtures of antibodies or antigens in gels, localizing antibodies in
tissues by labeling them and providing more sensitive procedures for identifying and quantitating circulating
antibodies. These methods showed that antibodies capable of reacting with antigens of the host are commonly
found normal animals and in healthy individuals. When present in heightened amounts, beyond a population-
based “normal” range autoantibodies were often associated with, and predictive of, disease (Chapters 69 and 70).
On occasion autoantibodies can be acknowledged directly as a causative agent of pathology based on accidental
or deliberate transfer of the autoantibodies between human subjects. Most of the advancement in the field, how-
ever, has depended on investigating genetically defined animals in which cellular manipulation and transfer can
prove that autoantibodies or selected lymphocytes, even in the absence of antibody, can cause of similar disease.
Animal experiments were particularly convincing in instances where the human disease involves inflammation
in a specific organ site such as the brain or the thyroid gland [Chapters 40 and 51]. Other diseases more often sys-
temic can be modeled in animals by selection or by genetic manipulation [Chapters 30 and 31]. By the early
1980s, the list of diseases reputed to be caused or significantly aggravated by autoimmunity, based on firm exper-
imental evidence, increased several folds.
Concurrently, the immune response itself was dissected with greater precision. The clones of B cells as the
source of specific autoantibodies (Chapters 8 and 9) provide the basis for clonal-based therapies (Chapters 71
and 72). Clonal definition of T cells based on antigen specificity is a bit more difficult because additional limita-
tions are placed on T-cell recognition. It requires help from cells bearing the corresponding major transplantation
markers which serve as antigen-presenting cells, to be discussed later. Despite their clear association with
autoimmune-induced inflammatory diseases, autoantigen-specific T cells are more often more conveniently iden-
tified and measured through cell surface markers or by the cytokine products generated by living, activated cells
rather than by direct antigen binding (Chapters 5 and 6).
The realization that self-reactive B cells and T cells are plentiful in normal individuals raised the question that
Ehrlich had considered: Will they do harm? Disease due to autoimmunity is relatively uncommon compared to
the frequency of autoimmune lymphocytes because evolution has provided the host with devises to tolerate
them. Much of the knowledge of tolerance came from studies on transplantation which proceeded in tandem
with research on autoimmune disease in the 1950s and 60s.

CLONAL BALANCE AND REGULATION

The experimental basis of specific immune tolerance was established by Medawar et al. Their first experiments
were inspired by the original observation of Ray Owen. He reported that cattle exposed in utero to nonidentical
blood cells can tolerate these cells without the expected immune response. Medawar supported these observa-
tions by transferring spleen cells into genetically different new-born mice. He found that the recipient mouse tol-
erated skin grafts from donor mice.
MacFarlane Burnet realized that these experiments provided an alternative to the then-prevalent instructive
theories of antibody production. Those earlier theories suggested that the broad, almost infinite, recognition
capacity of the cells of the immune system was achieved by instructions given by the antigen itself. Burnet pro-
posed that the role of antigen may be to select its counterpart “immunocyte” (lymphocyte) from a large repertoire
of T cells and B cells. The selected ones then multiplied and produced a clone of identical antigen-specific lym-
phocytes. Burnet further envisioned that an antigen-driven negative selective process takes place during the
embryonic development of immunologically active lymphocytes. In that way the immune repertoire encompasses
the universe of nonself-antigens but eliminates reaction to self-antigens.
The antigen-selection model became the prevalent explanation for self-tolerance but left open the question of
how autoimmunity could ever occur. Burnet’s first suggestion was that autoimmunity may represent a chance
mutation, resulting in production of a “forbidden clone.” Experimental evidence supporting this view, however,
has been meager because autoimmune responses are rarely mono- or pauciclonal. More commonly negative
selection is incomplete, so that a small population, mainly of low affinity T cells or B cells, escapes complete elim-
ination. Small numbers of self-reactive lymphocytes are not uncommon in the blood. Based on their low fre-
quency and low affinity, they may never inflict pathologic effects. They do, however, represent a potential risk
for pathogenic autoimmunity.

I. GENERAL CONSIDERATIONS
6 1. AUTOIMMUNE DISEASE: REFLECTIONS AND PROJECTIONS

If the self-antigen is presented in a particularly potent manner even low affinity T cells and B cells can be stim-
ulated into action. These conditions prevail when an antigen is given with a powerful adjuvant to potentiate the
response. In inducing experimental models of autoimmune disease, the appropriate antigen is commonly admin-
istered with complete Freund adjuvant containing both a mycobacterial peptide and selected mineral oil. This
method is used in producing models of organ-specific autoimmune disease in experimental animals.
Autoimmune diseases have often been associated with a prior infection that provides an adjuvant effect
(Chapters 21 and 70). In other instances the stimulus from a viral infection or environmental agent may serve as
an adjuvant (Chapters 20 and 64).
Good health of the immune system depends upon maintaining homeostasis through a proper balance between
the signals that may stimulate clones of self-reactive lymphocytes and those that tend to suppress them. I sug-
gested several years ago that the term “clonal balance” is more appropriate than clonal selection for describing a
healthy immune system. It is likely that multiple mechanisms are in place in most individuals most of the time
that maintain favorable clonal balance. Some of these regulatory agents may be found within the immune system
itself such as suppressor B cells, T cells, and monocytes. Additional signals come from other regulatory systems
such as the endocrine system. The sex bias, usually favoring females, and age-related susceptibilities represent
the close interaction between endocrine and immune systems (Chapter 24). Recent studies of neurologic and psy-
chological disorders and autoimmune disease exemplify comparable interactions, particularly through the gener-
ation of mediators that affect both immunologic and neurologic responses. The availability of multiple,
overlapping regulatory mechanisms is the most effective mechanism for assuring a well-balanced immune
system.
Strong evidence supporting the importance of clonal balance in human subjects has come from recent investi-
gations on immunologic approaches on cancer therapy. The oncologist may administer drugs designed to inhibit
critical steps (checkpoints) that regulate self-directed immune responses, including antigens on cancer cells.
Checkpoint inhibitors are drugs that facilitate the development of a cancer-specific immune response. In some
instances, however, they can also trigger an immune response to normal self-antigens and cause autoimmune
disease. Two diseases that were considered rare are relatively frequent outcomes of efforts to alter the normal
immune homeostasis. Well-studied examples are hypophysitis (Chapter 43) and myocarditis (Chapter 64).
Why does the body nurture and maintain self-reactive immune cells? One possibility is based on the
well-described phenomenon of molecular mimicry; that is, antigenic determinants (epitopes) found in other
organisms, even in plants, may closely resemble epitopes present in the body of the host. In some instances,
the mimicry may be involved in inducing a pathogenic autoimmune response (Chapter 21). The complete
elimination of these cross-reacting lymphocytes could significantly deplete the extensive immunologic repertoire
that is the basis of clonal selection theory.

GENETICS AND EXPOSURES

The next theme for research on autoimmune disease reached prominence in the 1970s based on understanding
that immune system function follows the principles of Darwinian selection. The growing acceptance of autoim-
mune diseases as a family of relats diseases reflected itself in a growing number of clinical reports of clustering
of different autoimmune disorders. A patient with one autoimmune disease tends to have a greater likelihood of
a second or even third autoimmune disease suggesting shared heredity and/or environment. Some of the
reported cooccurrences were quite striking; for example, an individual with autoimmune thyroid disease a
heightened risk of autoimmune diabetes due to shared genes or common exposures (Chapters 40 and 70).
A concurrent pathway of research supporting a genetic component of autoimmune disease came from long-
standing investigations of genetically defined rodents, especially rats and mice. These investigations, dating back
to the 1930s by Peter Gorer in London and George Snell at the Jackson Laboratory, recognized that transplanta-
tion of tumors between experimental animals depended mainly upon the genetic constitution of the animals
rather than on tumor-specific antigens. The studies led to development of inbred strains of rodents in which
histocompatibility genes were defined. By the1960s, it was recognized that there is a cluster of genes that are pre-
eminent in controlling acceptance of tumor and tissue transplants. The major histocompatibility complex (MHC)
dominated acceptance of allografts in all of the animals investigated, including humans (Chapters 23 and 26).
Hugh McDevitt and Michael Sela, taking advantage of these genetically defined animals, showed that the MHC
regulated the immune response to simple synthetic or small natural peptides.

I. GENERAL CONSIDERATIONS
 EPIDEMIOLOGY AND PREDICTION 7
Investigations on autoimmune disease began in the late 1960s when Vladutiu and I were able to induce experi-
mental autoimmune thyroiditis in inbred mice, taking advantage of the many defined strains provided by the
Jackson Laboratory. We were able to discern that the autoimmune response to a large protein molecule, thyro-
globulin, was determined genetically by the mouse MHC. Later we were able to report that a similar MHC-
related predominance was present in spontaneously developing thyroiditis in partially inbred chickens. Parallel
studies of other autoimmune diseases including both experimentally induced models and in spontaneous exam-
ples rapidly followed and genetic studies in humans became the focus of extensive investigation.
In addition to the MHC as the dominant inherited determinant of resistance to autoimmune disease in experi-
mental animals and humans, many other genes make small but significant contributions. Frequently these genes
turn out to be the controlling factors already known to be important in regulating the normal immune response
as well as controls of cancer immunity.
These genes, acting collectively, contribute to the broad susceptibility to autoimmune diseases. Yet a few
examples have been described in which a single gene locus dictates susceptibility to a group of autoimmune dis-
eases. This turned out to be particularly striking in the polyendocrine syndromes where autoimmune regulatory
genes predominate by preventing expression in the thymus of certain organ-specific antigens the endocrines dur-
ing negative selection (Chapter 39).
Despite the importance of genetic regulation, susceptibility to autoimmune disease is not attributable mainly
to heredity. Among even highly inbred rodents and identical human twins, there are still differences in the occur-
rences of autoimmune disease. Monozygotic twins may show a statistically significant cooccurrence but rarely
develop the same autoimmune disease at the same time. These differences may be attributable to the well-known
postgermline changes. There is already solid evidence that epigenetics plays an important role in determining
autoimmune susceptibility (Chapter 25). Genetically identical animals differing only in the sex-related genes can
be dramatically different in susceptibility (Chapter 24). Although in most instances, the bias favors females, few
autoimmune diseases are more prevalent and sometimes more severe in males.
In addition to their own genome, humans carry a second population with its own genetic constitution, the
microbiome (Chapter 19). The interaction between a human host and its bacterial inhabitants calls attention to
the importance of nutrition in immune response. The microbial population, once established during infancy,
tends to be quite stable in health but can change remarkably with illness. Working out the cause-and-effect rela-
tionship remains an intriguing problem for research.
One of the great voids in research on autoimmune diseases is the lack of firm knowledge of when and how
noninfectious agents in the environment may instigate a pathogenic autoimmune response (Chapter 2,). Studies
of experimental animals have made it clear that exposure to nonviable agents such as mercury have vastly differ-
ent effects on autoimmune responses. Studies on humans have been difficult because it is rare that a single envi-
ronmental agent can be identified and studied in isolation. Generally, large populations must be investigated
over long periods of time to minimize the effects of background confounders. Other than governmental agencies,
few institutions are able to provide the infrastructure for such large-scale, long-term studies. Yet it is likely that
such exposures represent the reason that many of the autoimmune diseases appear to be increasing in preva-
lence, particularly in the more industrialized societies.
One example of reductionist experiments carried out on human subjects is the administration of vaccines
(Chapter 22). Because of their importance in maintaining a high standard of public health, vaccine safety remains
a matter of great concern. Despite the hundreds of millions of vaccinations administered around the world, rarely
is there is a statistically valid increase in a well-recognized disease. A small cohort of individuals receiving sea-
sonal influenza vaccine in 1967 developed Guillain Barré syndrome. The basic biologic for this unique event has
not been elucidated. Ongoing studies of narcolepsy may provide a second example of influenza vaccine related
autoimmune disease. Here the potential antigen must be better defined.
Vaccination is immunology’s greatest success. Vigilance must be maintained for public health purposes but
also for opportunities to identify critical early steps in the induction of pathogenic autoimmune responses in
humans.

EPIDEMIOLOGY AND PREDICTION

As indicated previously, remarkable progress has been made in detailed genetic studies of susceptibility to
autoimmune disease in experimental animals. In contrast, epidemiologic investigations in humans have too often
been indecisive. Large populations must be studied over time to track all but the most prominent genes.

I. GENERAL CONSIDERATIONS
8 1. AUTOIMMUNE DISEASE: REFLECTIONS AND PROJECTIONS

Future volumes of “The Autoimmune Diseases” will certainly need to provide a greater understanding of the
epidemiology of autoimmune disease. Only in that way can we determine who is most likely to develop one of
these diseases, what the outcome may be, and what methods could be instituted to treat or prevent them. Such
epidemiologic study is emerging through technologic advances in handling large amounts of information. They
permit us for the first time to study enough individuals over long periods of time to look for correlations and
associations with environmental and dietary exposures.
At the same time, the availability of clinical data collected over the life-span allows us to focus with more pre-
cision on individuals and permit a rational definition on a personal basis of normality and departure from nor-
mality. The clinical laboratory will not have to depend upon population-based normal ranges but may also look
more precisely at departures from personal norms, representing an individual’s established homeostasis.
At this time, the best predictor of autoimmune disease remains the antibody. It is well established that in
many autoimmune disease elevated levels of autoantibodies appear well before clinical evidence. Combined with
family history or genetic analysis, they open the possibility to earlier intervention.
The most advanced studies have been carried out with autoimmune diabetes (Chapter 70). A combination of
genetic data obtained either from family history or actual genome studies have been coupled with appearance or
rise of multiple relevant autoantibodies. These early warning signs convey with a high degree of probability that
a child will develop diabetes. The next goal will be the design of benign interventions; that is, procedures that
could be instituted in a clinically well child that will prevent the occurrence of a disease without itself producing
injury.
New, more targeted treatments appear almost daily and promising opportunities for the specific inhibitors of
key steps of pathogenesis are on the horizon (Chapters 70 and 72). The future management of the autoimmune
diseases will certainly be intervention before irreversible damage has occurred. Ian Mackay and I hope that this
book will help to achieve that goal.

Acknowledgment
I thank Arthur Silverstein and Jorge Kalil for their corrections, suggestions, and wise advice.

I. GENERAL CONSIDERATIONS