|
|
||||||||
University of Arizona College of Medicine, Tucson Arizona, USA; and
* INSERM U430, Université Pierre et Marie Curie, Paris, France
1Correspondence: Microbiology and Immunology, College of Medicine, AHSC, University of Arizona, 1501 N. Campbell Ave., Tucson, AZ 85724, USA. E-mail: marchjj{at}email.arizona.edu
ABSTRACT
The primordial combinatorial immune recognition repertoire arose in the evolution of jawed vertebrates
450 million years ago as a rapid genetic process independent of antigenic selection. We propose that it encompassed the entire repertoire of innate immunity involving molecules that had evolved over billions of years. The antigen-driven compartment involving invasive pathogens operates in real time showing inducibility and increases in affinity. Individuals within a species differ in their repertoires because of distinct antigenic challenges, genetics, or local environmental effects. The homeostatic compartment that recognizes invariant cell and serum components should be conserved in all individuals of a species. The potential to recapitulate the entire recognition spectrum must be regenerated during the formation of new species. Evidence for the capacity of the combinatorial response to encompass the entire preexisting repertoire was obtained in studies of natural human IgG antibodies present in intravenous immunoglobulin. Since essential cellular recognition and regulatory elements are conserved throughout evolution, we propose that the natural antibodies of sharks, the most anciently emerged vertebrates to possess the combinatorial immune response, will resemble those of mammals in showing specificity for the conserved recognition/regulatory molecules. If verified, this hypothesis will establish the fundamental importance of natural antibodies not only in defense, but in regulation and functional homeostasis of the individual.Marchalonis, J. J., Kaveri, S., Lacroix-Desmazes, S., Kazatchkine, M. D. Natural recognition repertoire and the evolutionary emergence of the combinatorial immune system.
Key Words: immune response natural antibody repertoire antigenic selection
OUR 'IMMUNE RESPONSE' can be considered to consist of three separate but interacting systems. The first is the ancient innate system that matured over long evolutionary time incorporating elements of cellular differentiation and protection against oxidative damage (1)
into a defense mechanism against microbial infection (2)
. The second is the combinatorial system, which emerged in a rapid blink of the evolutionary eye of
10 million years as a stochastic event (3)
, probably catalyzed by the horizontal transfer of genes enabling site-specific recombination from microbes to an ancestral jawed vertebrate (4
5
6)
. This accidental system had the potential to recognize all possible antigens, whether of self or non-self origin. In us, its expressions are the natural antibody and T cell receptor repertoires. The third is an advanced system driven by clonal selection (7)
that has acted at both individual somatic and evolutionary natural selection levels to shape expression of the combinatorial system. This last addition has evolved over 450 million years and culminates in mammalian antibodies showing somatic hypermutation and affinity maturation (8
9
10)
. Here we consider the interplay of these with respect to their evolutionary origins and with respect to the recognition and functional antibody repertoires retained by the immunoglobulin (Ig) pool of normal humans (11)
.
Constitutive and inducible defense mechanisms against invading pathogenic organisms are ancient in evolution and widely dispersed throughout living plants and animals (12
13
14
15)
. Common elements involved in these processes of recognition and defense have been grouped together under the rubric of innate immunity encompassing certain circulating molecules, phagocytic cells, cellular receptors for microbial products, pathways of cellular activation such as that involving NF
B and molecular chaperones, and antimicrobial and antifungal peptides (16)
. The innate system was developed over billions of years of evolutionary time, based on selection for resistance to infection with microbes having carbohydrate and lipopolysaccharide type surface determinants that did not occur in eukaryotes. It appears that anciently arisen mechanisms of differentiation and defense were co-opted by the innate immune system and that these subsequently became integral parts of the cellular interactions involved in the activation of the combinatorial system.
The adaptive (17)
immune system dates from jawed vertebrates. We suggest it should be termed combinatorial because this name denotes both the genetic mechanism and the potential of the system to act in real time. That is, the recombination mechanism operates during somatic differentiation of the individual to generate 1), a comprehensive recognition repertoire that encompasses the innate capacity and 2) an inducible system capable of varying because of ongoing somatic mutation after induction to keep up with or surpass variation in the microbial invaders. The elements uniquely defining the combinatorial immune system (CIR) are antigen-recognizing lymphocytes, Igs (including antibodies and Ig family T cell receptors, or TCR), products of the major histocompatability complex, and recombinase activator genes 1 and 2 (RAG1 and RAG2). The complete ensemble is present in all jawed vertebrates (8
9
10)
. The system emerged rapidly and accidentally in evolution (3
, 14
, 18)
probably as the result of horizontal transfer of microbial genes for site-specific recombination (4
5
6
, 18
, 19)
. Ab initio, the CIR was neither a defense mechanism nor part of an internal regulatory system (20)
, but its unparalleled capacity to generate recognition molecules in real time gave it the potential to function effectively in both contexts. The CIR co-opted preexisting cellular and humoral mechanisms for activation, differentiation, and destruction of bound targets including the NF
B pathway, the IL-1/TOLL receptor complex, and interaction with more ancient complement components to initiate cytolysis, killing, and trigger inflammation (12
13
14
15)
.
Because all complete Ig VH/VL and many partial structures can bind to other molecules (21
22
23)
, the combinatorial system was able to encompass the entire spectrum recognized by the anciently arisen noncombinatorial regulatory and defense molecules. Whether in serum or on lymphocytes as receptors for antigen, naturally occurring antibodies (NAbs) are essential to the selective basis of the combinatorial/vertebrate immune response. Comparisons between NAb repertoires of sharks and humans have significantly enhanced our understanding of the operation of the CIR: even though these species had an ancestral divergence more than 450 million years ago and the germline organization of their Ig gene segments is substantially different (27)
, their antigen binding sites are constructed on the same principles (3
, 8
9
10
, 24)
and they express similar profiles in recognition of molecules implicated in regulatory or defense functions (25
, 26)
.
We will analyze fundamental features of the combinatorial antigen recognition system shared between antibodies of sharks and humans and provide evidence supporting the hypothesis that the combinatorial system in one step enveloped the entire recognition potential of the roughly 3 billion years of evolution of the innate system. The focus will be on IgMs because this is the ancient and universal isotype common to all vertebrates. It constitutes the predominant if not the only circulating antibody class in sharks and bony fish (9
, 10
, 24)
. We will show evidence from recent studies of intravenous immunoglobulins (IVIG) that are therapeutic preparations of IgG pooled from plasma of thousands of normal individuals that possess the spectrum of natural human antibodies (11)
.
CHARACTERISTICS OF RECOGNITION MOLECULES OF THE COMBINATORIAL IMMUNE SYSTEM
All vertebrate antibodies and
/ß TCR show epitope promiscuity, which is the capacity to bind specifically to peptide epitopes that share little or no sequence identity. This has been found not only for natural IgM and IgG antibodies of humans (28
, 29)
and IgM antibodies of sharks (8
, 25
, 26
, 30)
, but for also induced murine monoclonals, those raised against HIV P41 (31)
and monoclonal human IgM (32)
. Epitope promiscuity is well established for mammalian
/ß T cell receptors (33
34
35)
but has not been studied in TCR of sharks, although cartilaginous fishes possess
/ß and
/
TCR showing considerable variable domain sequence variability (9)
. Polyreactivity is a more general property in which affinity-purified polyclonal antibodies or monoclonal antibodies can bind unrelated large complex antigens (36
37
38
39
40
41)
. IgM anti-DNA antibodies of all species tend to be polyreactive in that they bind proteins and unrelated molecules as well as DNA used in affinity purification or immunization and the selection process. In our experience, NAbs to DNA are always polyspecific, but NAbs to thyroglobulins, viral glycoproteins, and T cell receptors can show relative specificity for the cognate selecting antigens. Nonetheless, even affinity-purified (26
, 41
, 42)
and monoclonal autoantibodies to T cell receptors (29
, 43)
show varying degrees of epitope recognition promiscuity characteristic of the particular monoclonal antibody.
Recognition by receptors of the combinatorial immune system is degenerate. This property has two characteristic features. First, an individual antibody, either a natural antibody or an induced and selected one (as well as individual T cell receptors) can recognize distinct antigenic epitopes (28
, 31
, 44
45
46)
. Usually this occurs with different affinities, but can involve comparable binding affinities and competition for the same portions of the combining site (31
, 46)
. Second, the same epitope can be recognized by a variety of distinct antibodies or T cell receptors (22
, 28
, 29
, 43
, 44
, 47
, 48)
. This is illustrated by the high degree of variability especially in the third hypervariable region of the variable domains of heavy chains binding DNA (36
, 49
, 50)
or T cell receptors (43
, 51)
. Similarly,
/ß TCR binding the same peptide epitope such as the melanoma MAGE1 peptide show different V
and Vß usage as well as distinct CDR3 usage (22)
. As above, differences in binding can reflect comparable or distinct affinities.
In principle, any complete VH/VL, V
, Vß, or V
/V
structure will bind a set of epitopes, specificity and degree of epitope promiscuity being characteristic properties of the individual molecule. In some cases, incomplete structures such as the heavy chains of cameloids, which cannot form heterodimers with light chains, can still bind numerous antigens (23)
or variable domains of TCR
chains containing extensive deletions (22)
can still be effective parts of heterodimers involved in the binding of peptide epitopes. Heavy chains incapable of dimerizing with light chains also occur in the nurse shark, where they are termed new antigen receptors (NAR) (21)
. Single-domain NARs have been generated as recombinant libraries and found to bind some protein antigens (52)
. Thus, the capacity for recognition by molecules generated by the combinatorial system is extremely rich. Although this review does not discuss tolerance to self, it is readily apparent that tolerance or specific suppression mechanisms are necessary to prevent deleterious autorecognition, because the original combinatorial process in the evolution of and individual ontogeny is not based on antigenic selection but on the need to recapitulate the entire recognition repertoire.
Like IgM antibodies, TCR have low intrinsic affinities that are compensated for by aggregation to increase relative avidity (53)
. Thus, the usual form of IgM in serum is a pentamer expressing 10 identical combining sites and is extremely efficient in fixing complement. Tetrameric forms in teleost fish and monomers of the form (light/µ)2 expressing two combining sites occur in the serum of sharks and as surface receptors of B cells in all species (9
, 10
, 24)
. These membrane-associated monomers are likewise of low affinity and compensate by aggregation when multivalent antigens bind to the cell (54)
.
THE NATURAL RECOGNITION REPERTOIRE
The original combinatorial immune repertoire in evolution was independent of antigenic selection. We propose that it encompassed the entire repertoire of noncombinatorial recognition and regulatory molecules. The antigen-driven compartment involving invasive pathogens and mutating or released sequestered self-antigens operated in real time showing inducibility and either small changes in affinity such as illustrated by IgM and TCR or great increases in IgG as a result of clonal selection after immunization. Individuals within a species would be expected to differ in their repertoires generated under the influence of distinct antigenic challenges by pathogens and possibly individual source conditions by genetics or local environmental effects. On the other hand, the homeostatic compartment that recognizes invariant cell and serum components and functions in cellular regulation or removal of damaged element, such as the recognition of the senescent cell antigen (55)
, should be fundamentally conserved in all individuals of a species. Consistent with fundamental conservation of the combinatorial response, the recognition repertoires of germ-free mice are essentially identical to those of conventional animals, although Ig levels are lower (56)
. Since the combining sites of antibodies and TCRs vary continually as part of the somatic combinatorial process, showing mutations in the V gene and in the combinatorially generated CDR3 segments, anti-antibody and TCR idiotypes arise continuously within the antibody and TCR components. The idiotype profiles of antibodies and TCRs would be a function of the antigen/pathogen exposure history of the individual and should vary within individuals of a species.
The potential to recapitulate the entire recognition spectrum was not only generated with the emergence of the combinatorial system in evolution, but is regenerated during the formation of new species in evolution. This has been accomplished by various means because of the genetic organizational plasticity of the combinatorial system (9
, 24)
. The combinatorial immune system of teleosts and other vertebrates (higher than sharks) occurred by inheritance of a small number of light and heavy chain gene cassettes of the hundreds or thousands in the germlines of ancestral sharks and amplification of these by tandem duplication of the variable domains and sometimes the diversity (D) and joining (J) minigenes. Other mechanisms such as functional diversification can contribute when recombination occurs. These events serve to maintain the high recognition potential of the system, which is also buttressed by the fundamental property of epitope recognition promiscuity that gives individual molecules the potential to recognize diverse sets of epitopes specifically. Memory in the IgM and TCR compartments arises predominantly from an increase in the number of specifically reactive cells after antigenic challenge. Although continued appropriate antigenic challenge can result in selection for higher affinities, these fall far short of the magnitudes generated in mammalian IgG where the process involves hypermutation correlated with class switching (54)
. Although IgM of all species is generally considered polyspecific and IgG monospecific, recent studies show that epitope promiscuity occurs in both types of antibodies, even in induced monoclonal antibodies (31
, 46)
.
HOW COMPREHENSIVE IS THE HUMAN NATURAL ANTIBODY REPERTOIRE?
Persuasive evidence for the capacity of the combinatorial response to encompass the entire preexisting repertoire is obtained in studies of natural human IgG antibodies present in IVIG (11
, 57
58
59
60
61
62
63
64)
. These consist of highly purified IgGs isolated from pools of serum collected from more than 10,000 normal individuals. Extensive studies have been carried out with natural IgM antibodies of humans (65)
, mice (66)
, bony fishes (67)
, and sharks (25
, 30
, 68)
. We consider the repertoire of antibodies termed natural those that arose in the absence of deliberate immunization and apparently independent of exposure to foreign antigens. A great variety of specificities of natural antibodies have been identified; in some cases, molecules were isolated and characterized from sera of humans and sharks, two species representing the extremes of the combinatorial system. With humans and mice, which express five classes of serum Igs, natural antibodies of comparable recognition specificity and degree of promiscuity occur in the IgM and IgG compartments (25
, 28
, 31
, 46)
. Although we focus on natural antibodies, monoclonal IgG antibodies induced in response to standard immunization procedures can also show pronounced epitope recognition promiscuity in that they bind equally well to sets of peptides showing little or no identity in amino acid sequence. Normal human Igs like those of the shark contain subsets of molecules binding to foreign antigens as well as those considered to be autoantigens. Phosphorylcholine, a membrane component of pathogenic bacteria and other infectious organisms, and recombinant envelope proteins of retroviruses (25
, 40)
are bona fide epitopes expressed by infectious agents to which NAbs have been found in humans and sharks. NAbs against many types of antigens are considered to be autologous and are widely conserved in vertebrate evolution in both species. These include antibodies directed against surface components such as the senescent cell antigen, which is a marker of defective red cells that initiates removal by antibody-dependent phagocytosis (55)
, variable (69
, 70)
, and constant (50
, 71)
region epitopes of Igs and T cell receptors, and autoantigens recognized in autoimmune diseases, including thyroglobulins (25
, 72)
and denatured and native DNA. Shark NAbs bind to the human Igs, TCRs, and mammalian thyroglobulins (human, porcine, bovine) probably because of the strong homologies of these molecules among vertebrates. Moreover, the human pool contains NABs to regulatory molecules including cytokines and their receptors (61)
, markers of cell differentiation such as CD4 and CD5 (59
60
61)
, HLA class I molecule (62)
, FAS (63)
, and coreceptors involved in binding of HIV (64)
. Natural antibodies of many species react with internal cellular proteins, including actin and myosin (47)
.
THE PRIMORDIAL REPERTOIRE GENERATED INDEPENDENTLY OF ANTIGENIC SELECTION
Figure 1
illustrates in a Venn diagram the universe of antigens recognized by the primordial combinatorial antibody repertoire as would be generated by the stochastic system in the absence of selection. This set of all epitopes includes both self and non-self because, in principle, there is no difference between them. All proteins can be antigenic. The ability of individuals to accept their own antigens is learned through mechanisms for the generation of tolerance. Overlap between the endogenous or self markers and the non-self or exogenous antigens exists; for example, cross reactions occur between retroviral glycoproteins and lymphocyte surface receptors (73
74
75)
. Even in the case of microbial antigens not found in vertebrates, such as lipopolysaccharides, peptide epitopes can mimic nonproteinacious antigens, including bacterial carbohydrates (76)
and DNA (77)
. The self compartment can be further broken down into an extracellular component and an intracellular one. Intracellular self-antigens would normally be excluded from susceptibility to immune surveillance but would be released after cell damage or death. Such molecules include DNA, ribonucleic acid/protein complexes, heat shock proteins, and proteins involved in regulatory mechanisms including those of the ancient NF
B pathway for response to stress. Extracellular self-antigens include surface receptors for cytokines and antigens, differentiation markers, and serum molecules of all types. Overlap in determinants occurs not only because of limited or extensive sharing of stretches of amino acid sequence, but because of epitope recognition promiscuity and polyreactivity since monoclonal antibodies and TCR can recognize molecules having little or no sequence identity (25)
. The subsequent inducible repertoire is modified and shaped by exposure of the individual to antigen challenge such as the autologous generation of neoantigens or exposure of sequestered antigens and contact with exogenous molecules via infection or deliberate immunization. In sharks, the predominant antibody is IgM, which is of low affinity and does not show class switching or affinity maturation after multiple immunizations (9
, 10
, 24)
. Mammals switch predominantly to IgG molecules, which show increased affinity after repeated immunizations due to somatic hypermutation and selection by antigen. Nevertheless, IgG from normal individuals as represented by IVIG express essentially the same repertoire as that of the primordial IgM system.
|
CONCLUSIONS
The combinatorial immune response arose rapidly in evolution and emerged in complete form in the jawed vertebrates. One consequence of the acquisition of the genetic recombination mechanism in evolution was the capacity for all individuals of a vertebrate species to respond to the vast theoretical panoply of antigens of foreign and potentially invading organisms as well as the ability to recognize internal regulatory and homeostatic molecules in order to provide additional control mechanisms for processes of the internal milieu. Studies suggest that the recognition capacities of the most advanced species to possess the system as well as the phylogenetically most primitive are comparable. The framework segments of the variable domains of light chains, heavy chains, and TCRs of sharks and humans are clearly homologous; in some cases, orthologous relationships have been found (8
, 78)
. However, the properties of epitope promiscuity and degeneracy are universal within the antibody system of jawed vertebrates, such that great heterogeneity in the third hypervariable segments of heavy chains occurs for antibodies recognizing the same epitopes, even in humans (36
, 40
, 43)
and mice (51)
. Although more than 60 complete shark VH sequences have been determined (8)
, no identities have been found between shark HCDR3s and CDR3s of human or murine antibodies of known specificity. Extensive studies with natural antibodies of humans show that naturally occurring antibodies against self-surface markers of differentiation and regulation are part of the repertoire. These findings support the conclusion that the natural combinatorial antibody repertoire can recapitulate the entire recognition spectrum of the innate system that was developed over 3 billion years of evolutionary time. Functional studies with these natural antibodies establish that they can modulate the regulatory systems toward which they are directed (11)
and can play a significant role in protection against infection (79
80
81
82)
and in selection of autoreactive B cells (83)
. Since many cellular recognition and regulatory elements contain molecules that are clearly conserved throughout evolution (12
13
14
15
, 26)
, we predict that natural antibodies of sharks will likewise have specificities for the conserved recognition/regulatory molecules. If verified, this hypothesis will establish the importance of natural antibodies not only in defense, but in regulation and functional homeostasis of the individual vertebrate.
ACKNOWLEDGMENTS
Original contributions by J.J.M. were supported in part by grants from the U.S. National Science Foundation and the Arizona Disease Control Research Commission. The French group is supported by Institut National de la Santé et de la Recherché Médicale (INSERM), France. We also acknowledge the support of the ZLB Bioplasma AG, Bern, Switzerland.
Received for publication December 7, 2001. Revision received January 28, 2002. REFERENCES
light chain mapped by comprehensive peptide synthesis. J. Prot. Chem. 12,659-665[CrossRef][Medline]
ß T-cell receptors in human immunodeficiency virus (HIV) infection: dysregulation and mimicry. Proc. Natl. Acad. Sci. USA 91,10849-10853This article has been cited by other articles:
![]() |
M. Gonthier, R. Llobera, J. Arnaud, and B. Rubin Self-Reactive T Cell Receptor-Reactive CD8+ T Cells Inhibit T Cell Lymphoma Growth In Vivo J. Immunol., December 1, 2004; 173(11): 7062 - 7069. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. H. STRAUB and H. O. BESEDOVSKY Integrated evolutionary, immunological, and neuroendocrine framework for the pathogenesis of chronic disabling inflammatory diseases FASEB J, December 1, 2003; 17(15): 2176 - 2183. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. J. Marchalonis, G. K. Whitfield, and S. F. Schluter Rapid Evolutionary Emergence of the Combinatorial Recognition Repertoire Integr. Comp. Biol., April 1, 2003; 43(2): 347 - 359. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |