Lecture List

Vertebrate Immunity. II. Adaptive.

The view that I Metchnikoff and his pupils vigorously sustained in the early part of the 20th century was that phagocytic cell was the key to immunity (antibodies being helpers of phagocytosis). This view, however, ignored the dramatic demonstrations of humoral immunity and resulted in a lack of interest for the cells of immunity that lasted approximately half a century. Only when it became clear that delayed hypersensitivity (1945) and graft rejection (1954) could be transferred by blood peripheral cells, interest re-emerged on the nature of these cells and the mechanisms by which they produced cellular immunity.

Graft versus host disease

In the mid-1950s there was a general view that adaptve humoral immunity was due to antibodies and complement and that the cells responsible for immunity were the macrophage (the phagocytic cells) and the plasma cells (the antibody-producing cells). The latter had been shown in 1948 to be able to produce antibodies by A Fagraeus (using a fluorescent-conjugated secondary antibody) but the origin of these cells met with little interest.

In the late 1950s and early 1960s, however, evidence emerged that lymphocytes were involved in a number of immune reactions. First, it became apparent that lymphocytes were responsible for the graft versus host disease, (GVH) a reaction caused by a graft of a tissue containing white blood cells in an immune incompetent host (M Simonsen, 1957). There are three forms of GVH that have been extensively studied in mouse.

The first is the GVH in newborn mice. It takes advantage of the fact that newborn mice are immunologically incompetent for 24-48 hours after birth. If these mice are injected with allogeneic adult lymphocytes, these will not be rejected. However the grafted cells reject host tissues as foreign. This produces splenomegaly, diarrhea and failure to develop a healthy fur and gain weight (runt disease). If the number of cells grafted is sufficient, the host dies in 2-3 weeks. A second form of GHV can be produced in F1 hybrid mice. A cross of two homozygous parental strains (a/a x b/b) will result in heterozygous F1 (a/b) that will accept grafts from either parents. A parental graft containing lymphocytes, however, induces GVH in F1 recipients. Finally, a third form of GVH disease is observed in irradiated adults. If irradiated adults are injected with allogeneic lymphocytes, these cause GVH in the host.

Gowans experiments

Another line of experiments conducted in the early 1960s extended the role of lymphocytes in cell-mediated immunity and demonstrated a role for these cells in humoral immunity. Experiments by J Gowans and his colleagues in the late fifties and early sixties demonstrated that selective lymphocyte depletion caused by chronic drainage of the thoracic duct in rats decreased both antibody responses as well as graft rejection. Early in the 1960s, therefore, evidence was established that a single cell type, the lymphocyte, was responsible for both types of adaptive immunity: one mediated by antibodies and one mediated directly by the cells.

Lymphocytes are found in the body in lymphoid tissues. These include bone marrow, spleen, lymph nodes and several tissues associated with the gut: tonsils, appendix, Peyer's patches (referred collectively as gut-associate lymphoid tissue). In birds two further lymphoid organs are found: the so-called bursa of Fabricius, located near the cloaca and the thymus. The bursa (the bone marrow in mammals) and the thymus are primary lymphoid organs because lymphocyte maturation occurs in these. All other lymphoid tissue/organs constitute secondary lypmphoid organs.

Let us now consider three facts discussed above: (i) there are two main types of immunity (humoral and cell-mediated), (ii) the small lymphocyte appears responsible for both types of immunity (Gowans experiment) and, (iii) the small lymphocytes in different tissues/organs have generally similar morphology. Does this mean that the same cells is capable of both cell-mediated immunity and antibody production or that there exist at least two different subpopulations of lymphocytes that are morphologically undistinguishable but functionally different ?

Birds have a distinct lymphoid organ, located near the cloaca and known as the bursa. It was described in 1621 by the Italian anatomist Girolamo Fabrizi, hence is generally referred as the bursa of Fabricius. Evidence supporting the latter hypothesis (ie the heterogeneity of the lymphocyte population) emerged from physiologic experiments reported in 1954 by B Slick and colleagues who demonstrated that the effects of bursectomy in adult chickens had no noticeable morphologic or functional effects on the immune system. In contrast, bursectomy before or immediately after hatching produced major immunological impairments: specific antibody responses were abolished, serum antibody levels were very low, plasma cells and germinal centers were absent. Rejection of allografts, however, was normal. Different results were obtained in thymectomy experiments in newly hatched chicks. Thymectomy resulted in impaired allograft rejection and delayed-type hypersensitivity responses and had variable effects on antibody responses (for example the response to SRBC was reduced but response to other antigens was unaffected). The level of circulating antibodies was normal, and so were plasma cells and germinal centres. Similar conclusions were reached from experiments of neonatal thymectomy in the mouse by J Miller in 1962 (see Table Effects of bursectomy and thymectomy in newly born chicks and mice) .

Effects of neonatal bursechtomyand thymectomi

Thus the results of the bursectomy and thymectomy experiments pointed to the presence of two distinct (yet morphologically indistinguishable) populations of lymphocytes: one population, under the control of the bursa accounts for the ability to mount specific antibody responses. The second population, under the control of the thymus is responsible for graft rejection. These two populations of cells are defined as B lymphocytes and T lymphocytes respectively.


The variable effects on antibody production of early thymectomy experiments led to further experiments that clarified the role of the B and T cells as well as other cell types in the antibody response.

Jerne's plaque assay

These studies took advantage of two essential experimental tools: the earlier development of inbred strains of mice and the development of simple, reliable assays for quantitation of antibody-producing and hence of the antibody response. Inbred strains have two critical features: (i) they are homozygous at each locus and, (ii) they are syngeneic, ie each mouse of an inbred strain is genetically identical to every other mouse of the same strain. The most sensitive and useful assays for antibody producing cells was Jerne's haemolytic plaque assay. The assay exploits the fact that anti sheep red blood cells (SRBC) antibodies typically lyse SRBC in the presence of complement (see Fig Jerne's plaque assay). Thus if antibody-producing cells are plated in a semisolid medium and covered with a layer of SRBC, when complement (C') is added an area of SRBC lysis will appear around the cells that produce anti-SRBC antibodies. This assay is sensitive and quantitative in the sense that allows counting of antibody producing cells.

Claman's experiment

The first experiment that addressed directly the issue of cell cooperation in antibody production is the so called Claman's experiment (1966) in which the antibody response of irradiated mice to SRBC was investigated after transfer of syngeneic cells from the spleeen, the bone marrow, the thymus or a mixture of marrow and thymic cells. The experiment showed that spleen cells transferred the antibody response, that marrow or thymic cells did not but the mixture of the two cell types did. Claman interpreted the results postulating that one population of cells produced antibodies but required the second population in order to do so, an interpretation later proved to be correct.

A subsequent study (the so-called Mosier's experiment, 1967) implied a third cell type in the antibody response. Mosier used an in vitro antibody response to show that whole spleen cells produced an effective response but that non-adherent spleen cells alone, ie the lymphocytes did not unless macrophages were added. This indicated that a full antibody response required further cooperation from a third cell types, namely the macrophage.

Mitchell and Miller's experiment

Which cell type was actually responsible for antibody production was established by the Mitchell and Miller experiment (1968). In this experiment adult CBA mice were irradiated and reconstituted with syngeneic bone marrow. This was followed two weeks later by injection of SRBC and CBA x C57Bl/6 F1 T cells (from the thoracic duct) and a week later by the analysis of the antibody response in spleen in the presence of anti-CBA or anti-C57Bl/6 antibodies. The anti-C57Bl/6 antibody had no effect on the response but the anti-CBA antibody dramatically reduced the number of PFCs.

Thus the antibody producing cells had no C57Bl/6 antigens on their cell surface (ie they were not the CBA x C57Bl/6 F1). The conclusion of these experiments therefore is that antibody producing cells are bone marrow-derived and not thymus-derived but that these cells require both T cells and macrophages in order to produce antibodies.

T lymphocytes are involved in two major activities: cooperation in the antibody response on the one hand and graft rejection and delayed hypersensitivity (the Koch's phenomenon) on the other. That the role of T lymphocytes in graft rejection and delayed type hypersensitivity was not due to their helper function in the antibody response was evident from the fact that graft rejection and delayed-type hypersensitivity could not be transferred with serum. Further, in 1961 lymphocytes with cytotoxic activity were discovered and these cells appeared the effectors for phenomena such as graft rejection. These finding, nevertheless, did not clarify whether the same T cell is capable of these two activities or whether different subpopulations of T cells exist, one with helper activity and the other with cytotoxic function.


In the early 1970s it had become clear that helper T lymphocytes and T lymphocytes involved in graft rejection shared certain surface antigens but not others. These antigens are known as CD antigens (cluster of differentiation) in humans or Ly antigens (lymphocytes) in mouse. Antibodies directed to these molecules are powerful tools for studying the B and T cell lineages and for defining cell subpopulations. In particular, graft-type T cells appeared to express high levels of the Ly-2,3 antigens while helper-type T cells did not (see Table Surface markers on lymphocyte subpopulations).

Cantor adn Boyse's experiment

On the strength of this information H Cantor and EA Boyse (1975) constructed an experiment that proved beyond reasonable doubt the presence of T cell subpopulations. They enriched T cells from B6 (H-2b) mouse spleen with nylon wool (a procedure that leaves behind the bulk of B cells and macrophages) and treated aliquots of the cells suspension with anti-Ly-1 Ab + C' (complement) or anti Ly-3 Ab + C'. These two cell populations were subsequently injected either in irradiated B6 mice along with B cells and SRBC (in order to measure anti-SRBC antibodies) or in B6 (H-2b) x C (H-2k) F1 (in order to measure anti-C cytotoxic cells).

The cells of adaptive immunity

The results of the experiment indicated that there exists a population of T cells (Lyt-1+) that could provide B cell help but could not become effector cytotoxic T cells and another population of T cells (Lyt-3+) that could provide effector cytotoxic functions but not B cell help. Thus at least two supopulations of T cells pre-exist antigen exposure, one with helper activity and one with cytotoxic activity.

This work constituted a further significant advance in our uderstanding of the cells of immunity and the Fig The cells of the adaptive immunity of higher vertebrates summarises the lymphocytes subpopulations discussesd in this part of the lecture.


The proposers of clonal selection theories of antibody formation

The most important conceptual advance towards our understanding of the cellular basis of the antibody response was provided by the clonal selection theory proposed by MacFarlane Burnet in 1957, which improved and extended an earlier theory put forward by Niels Jerne in 1955. It should also be emphasised that Paul Erlich's side chain theory is a precursor of all modern theories. According to MacFarlane Burnet's theory there is a repertoire of antibody specificities in the body that equals the number of B lymphocytes (immunocytes in Jerne's language), ie each different B lymphocyte expresses a single antibody specificity.

A schematic of the way in which the clonal selection theory may work

Antigen binding triggers proliferation of the B lymphocyte(s) that recognize antigen and leads to clonal amplification of such cell(s). Clonal amplification is accompanied by the differentiation of the small lymphocyte into plasmacells, the cells that produce soluble antibodies (Fig 2). This remarkable theory made a number of important predictions: (i) that the small lymphocyte expresses an antigen receptor on the cell surface, (ii) that only one type of receptor is expressed on each small lymphocyte and, (iii) that the small lymphocyte that expresses a specific antigen receptor is the cell that will subsequently proceed to differentiate in plasma cells and produce soluble antibodies All these predictions have been supported by subsequent experimentation and the clonal selection theory has become the key foundation of modern cellular immunology.

The B cell receptor (BCR).

Let us now introduce the molecules that enable B cells and T cells to respond to antigen and cooperate in the antibody response. Thirteen years after MacFarlane Burnet's clonal selection theory, M Raff , M Sternberg and RB Taylor demonstrated the presence of surface immunoglobulin on a subpopulation of spleen cells. Other experiments demonstrated that antigen-binding in B cells could be blocked by anti-immunoglobulin reagents. Thus in the early 1970s it became established that B cells expressed an antigen receptor, and that this was indistiguishable from soluble antibody.

An elegant experiment conducted in 1969 by GL Ada and P Byrt addressed another key point anticipated by the clonal selection theory, namely that those cells binding a particular antigen are the same that secrete antibody to it. Ada and Byrt incubated mouse spleen cells with high-specific activity 125-I antigen (polymerized flagellin). After incubation for 16-20 hours, the cells were transferred into irradiated syngeneic mice and challenged with a dose of either the same or a distinct flagellin. The results of this experiment clearly demonstrated that the cells that bind antigen are the ones that would subsequently produce antibody to it (Ada and Byrt experiment, 1969). Thus, the conclusion of these and other experiments are that: (i) B cells bind antigen directly through the B cell receptor, (ii) B cells recognise proteins as well as other antigens such as carbohydrate, complex lipids, nucleic acids and, (iii) upon antigen-binding B cells undergo clonal proliferation as well as differentiation into antibody-secreting plasma cells. These studies were paralleled by a large body of work over half a century on the nature and properties of soluble antibodies and these two lines of experiments together have provided the basis for the current understanding of the relationship between membrane-bound and soluble antibodies, a topic discussed in subsequent lectures.

T cell receptor (TCR).

Given the features of T cell responses to antigen and the nature of the antigen receptor on B cells, early views about the nature of the T cell receptor were informed by the thought that the T cell receptor too was a surface immunoglobulin. This view, however, proved not to be correct. In 1983 P Marrack and colleagues showed that the receptor consisted of two polypeptide chains of ~ 45 KDa and a year later M Davis and colleagues cloned cDNAs for the T cell receptor using a novel approach (subtractive hybridization) and the fundamental assumption that T cell receptor genes (as shown earlier on by S Tonegawa for antibody genes) underwent somatic rearrangement in T cells. Subtractive hybridization (or later modifications of this technique) has since become a major tool for isolating gene differentially expressed in different cell types or in differnent stages of differentiation of a single cell type.

A striking feature that emerged early on from the work on the T cell receptor and confirmed by transfection experiments after the T cell receptor had been cloned is that the T cell receptor, unlike the B cell receptor only binds protein antigens and does not bind antigen directly but only after antigen has been internalised by other cell types and re-expressed on the cell surface in association with molecules that were first discovered on the basis of an important role in graft rejection and therefere designated the antigens of the major histompatibility complex (MHC). This point will be discussed below.

There are two classes of MHC proteins. MHC class I molecules are expressed ubiquitously and present endogenous antigens, ie antigens expressed in the cytoplasmic compartment, for example as a result of a viral infection, to the subpopulation of T cells with cytotoxic activity (Lyt-3+ in mice, CD8+ in humans). Viruses are obligate intracellular parasites that subvert and utilise the DNA, RNA and protein synthetic machinery of host cells in order to secure survival and replication. As a result, cells infected by viruses accumulate proteins encoded by viral genes in their cytoplasm (both regulatory and structural). Some of these proteins, like the protein of the host cell, are targeted for degradation by the ubiquitin - proteasome pathway during viral infection. Ubiquitin is a small protein that is covalently attached to proteins targeted for degradation in an ATP-dependent reaction and the resulting ubiquitin-target protein complex is recognised by the proteasome for degradation.

The proteasome is a large multiprotein complex with a cylindrical shape and a central hole that enables entry, transit and degradation of the proteins targeted for degradation. A number of proteases are associates with the proteasome but a few of them, specifically LMP2, LMP7 and LMP10 (the first two of which are encoded by genes at the MHC locus) cleave proteins at basic or hydrophobic residues and generate peptides that bind well to MHC class I proteins. These proteases, therefore, are specially important in antigen processing. Cytoplasmic peptides generated by the proteasome are next transported into the endoplasmic reticulum by the so-called Transporters associated with Antigen Processing (TAP1 and 2), members of the large (ATP-binding cassette) ABC transporter superfamily. Once translocated, peptides can bind to native MHC class I molecules assembled first as a alpha chain - calnexin complex and, subsequently to the association of beta2 microglobulin to the alpha chain, as a MHC-calreticulin-tapasin complex. Upon peptide binding, MHC class I are targeted for secretion and membrane insertion. Class I proteins failing to bind peptide are targeted for degradation. Antigen recognition by cytotoxic T cell in association with MHC class I protein leads to the killing of the cell presenting the antigen.

Antigen presentation by MHC class I proteins

MHC class II proteins are expressed only in certain cell lineages (such as macrophages and B cells) and present antigen to the other major subpopulation of T cells (Lyt-1+ in mice, CD4+ in humans) that provides helper function in antibody production. MHC class II proteins are involved in the presentation of exogenous antigens either as a result of paghocytic activity (as in macrophages or dendritic cells) or as a result of endocytosis (either receptor-mediated or fluid-phase pinocytosis). Antigens internalised via receptor-mediated endocytosis travel from early to late endosomes and finally lysosomes and are degraded in the process. Fusion with vesicles in the trans-Golgi compartment brings partially degraded protein antigens in contact with native class II molecules and their loading prior to expression on the cell surface for presentation to Th cells displaying TCR and CD4 co-receptors.

Class II have an effective safeguard mechanism preventing them from loading endogenous peptides translocated by TAPs. The safeguard mechanism is a protein, known as the invariant chain (Ii, CD74) that engages three class II molecules and obstruct their antigen-binding site. During transport through the Golgi and the trans-Golgi, the invariant chain is degraded leaving only a fragment of it, known as class II invariant chain peptide (CLIP), in the antigen-binding pocket. CLIP is displaced next by antigenic peptides in a reaction that involves the non-classical class II molecule HLA-DM. The result of antigen presentation to helper T cells is proliferation of the T cells engaged by antigen (through induction of expression of lymphokines such as inteleukin 2 that act in an autocrine manner) and release of other lymphokines (such as interleukin 4 and 5) that induce proliferation of antigen-specific B cell clones.

Although the interactions described above involve primarily MHC proteins (in one cell) and the T cell receptor (on the other cell) the presence of CD8 molecules on cytotoxic T cells and CD4 molecule on T helper cells is essential for the formation of stable MHC-TCR complexes. CD8 and CD4 thus function as co-receptors. One further point should be outlined. MHC antigens (both class I and class II) are the product of highly polymorphic genes. Thus different individuals have distinct sets (haplotypes) of MHC alleles as will be discussed in detail in a subsequent lecture. However, it is important to note from start that T-cytotoxic and T-helper cells only recognize antigen presented by antigen presenting cells of the same MHC haplotype. This phenomenon (MHC restriction, see lecture on: B and T cell development) is the basis for the development of self vs non-self discrimination and immunological tolerance.

The interface between innate and adaptive immunity

From the topics discussed above it is now possible to draw a first summary of the cells of innate and adaptive immunity and their functional relationships (see Fig The interfacing of adaptive and innate immunity). Macrophages and NK lymphocytes are key effectors of innate immunity because of their phagocytic and natural killing activity respectively towards foreign agents. On the other hand, Th and Tc lymphocytes and B lymphocytes and plasmacells constitute the key components of adaptive (specific) immunity. One further point should be outlined. MHC antigens (both class I and class II) are the product of highly polymorphic genes. Thus different individuals have distinct sets (haplotypes) of MHC alleles as will be discussed in detail in a subsequent lecture. T-cytotoxic and T-helper cells only recognise antigen presented by antigen presenting cells of the same MHC haplotype. This 'MHC restriction' has important implications for immune tolerance and T cell function and will be discussed in a subsequent lecture.

  • The role of lymphocytes in immunity
  • Lymphocyte subpopulations
  • Cell cooperation in immunity
  • Subpopulations of T lymphocytes
  • General properties of adaptive vs innate immune response
  • The clonal selection theory
  • Lymphocyte antigen receptors (BCR and TCR)
  • Antigen presentation
  • The interface between native and adaptive immunity

[1 ] The experimental foundations of modern immunology. Clark W.R. J. Wiley Inc., (1991)

[2] Snell GD. Studies in histocompatibility. Nobel Lecture (1980)

[3] Dausset J. The major histocompatibility complex in man -- past, present, and future concepts. Nobel Lecture (1980)

[4] Benacerraf B. The role of MHC gene products in immune regulation and its relevance to alloreactivity. Nobel Lecture (1980)

[5] Burnet MF. Immunological recognition of self. Nobel Lecture (1960)

[6] Jerne NK. The generative grammar of the immune system. Nobel Lecture (1984)