NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Alberts B, Johnson A, Lewis J, et al. Molecular Biology of the Cell. 4th edition. New York: Garland Science; 2002.

  • By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.
Cover of Molecular Biology of the Cell

Molecular Biology of the Cell. 4th edition.

Show details

Helper T Cells and Lymphocyte Activation

Helper T cells are arguably the most important cells in adaptive immunity, as they are required for almost all adaptive immune responses. They not only help activate B cells to secrete antibodies and macrophages to destroy ingested microbes, but they also help activate cytotoxic T cells to kill infected target cells. As dramatically demonstrated in AIDS patients, without helper T cells we cannot defend ourselves even against many microbes that are normally harmless.

Helper T cells themselves, however, can only function when activated to become effector cells. They are activated on the surface of antigen-presenting cells, which mature during the innate immune responses triggered by an infection. The innate responses also dictate what kind of effector cell a helper T cell will develop into and thereby determine the nature of the adaptive immune response elicited.

In this final section, we discuss the multiple signals that help activate a T cell and how a helper T cell, once activated to become an effector cell, helps activate other cells. We also consider how innate immune responses determine the nature of adaptive responses by stimulating helper T cells to differentiate into either TH1 or TH2 effector cells.

Costimulatory Proteins on Antigen-Presenting Cells Help Activate T Cells

To activate a cytotoxic or helper T cell to proliferate and differentiate into an effector cell, an antigen-presenting cell provides two kinds of signals. Signal 1 is provided by a foreign peptide bound to an MHC protein on the surface of the presenting cell. This peptide-MHC complex signals through the T cell receptor and its associated proteins. Signal 2 is provided by costimulatory proteins, especially the B7 proteins (CD80 and CD86), which are recognized by the co-receptor protein CD28 on the surface of the T cell. The expression of B7 proteins on an antigen-presenting cell is induced by pathogens during the innate response to an infection. Effector T cells act back to promote the expression of B7 proteins on antigen-presenting cells, creating a positive feedback loop that amplifies the T cell response.

Signal 2 is thought to amplify the intracellular signaling process triggered by signal 1. If a T cell receives signal 1 without signal 2, it may undergo apoptosis or become altered so that it can no longer be activated, even if it later receives both signals (Figure 24-62). This is one mechanism by which a T cell can become tolerant to self antigens.

Figure 24-62. The two signals that activate a helper T cell.

Figure 24-62

The two signals that activate a helper T cell. (A) A mature antigen-presenting cell can deliver both signal 1 and 2 and thereby activate the T cell. (B) An immature antigen-presenting cell delivers signal 1 without signal 2, which can kill or inactivate (more...)

The T cell receptor does not act on its own to transmit signal 1 into the cell. It is associated with a complex of invariant transmembrane proteins called CD3, which transduces the binding of the peptide-MHC complex into intracellular signals (Figure 24-63). In addition, the CD4 and CD8 co-receptors play important parts in the signaling process, as illustrated in Figure 24-64.

Figure 24-63. The T cell receptor and its associated CD3 complex.

Figure 24-63

The T cell receptor and its associated CD3 complex. All of the CD3 polypeptide chains (shown in green), except for the ζ (zeta) chains, have extracellular Ig-like domains and are therefore members of the Ig superfamily.

Figure 24-64. The signaling events initiated by the binding of peptide-MHC complexes to T cell receptors (signal 1).

Figure 24-64

The signaling events initiated by the binding of peptide-MHC complexes to T cell receptors (signal 1). When T cell receptors are clustered by binding to peptide-MHC complexes on an antigen-presenting cell, CD4 molecules on helper cells or CD8 molecules (more...)

The combined actions of signal 1 and signal 2 stimulate the T cell to proliferate and begin to differentiate into an effector cell by a curiously indirect mechanism. In culture, they cause the T cells to stimulate their own proliferation and differentiation by inducing the cells to secrete a cytokine called interleukin-2 (IL-2) and simultaneously to synthesize high affinity cell-surface receptors that bind it. The binding of IL-2 to the IL-2 receptors activates intracellular signaling pathways that turn on genes that help the T cells to proliferate and differentiate into effector cells (Figure 24-65). As discussed in Chapter 15, there are advantages to such an autocrine mechanism. It helps ensure that T cells differentiate into effector cells only when substantial numbers of them respond to antigen simultaneously in the same location, such as in a lymph node during an infection. Only then do IL-2 levels rise high enough to be effective.

Figure 24-65. The stimulation of T cells by IL-2 in culture.

Figure 24-65

The stimulation of T cells by IL-2 in culture. Signals 1 and 2 activate T cells to make high affinity IL-2 receptors and to secrete IL-2. The binding of IL-2 to its receptors helps stimulate the cell to proliferate and differentiate into effector cells. (more...)

Once bound to the surface of an antigen-presenting cell, a T cell increases the strength of the binding by activating an integrin adhesion protein called lymphocyte-function-associated protein 1 (LFA-1). Activated LFA-1 now binds more strongly to its Ig-like ligand, intracellular adhesion molecule 1 (ICAM-1), on the surface of the presenting cell. This increased adhesion enables the T cell to remain bound to the antigen-presenting cell long enough for the T cell to become activated.

The activation of a T cell is controlled by negative feedback. During the activation process, the cell starts to express another cell-surface protein called CTLA-4, which acts to inhibit intracellular signaling. It resembles CD28, but it binds to B7 proteins on the surface of the antigen-presenting cell with much higher affinity than does CD28, and, when it does, it holds the activation process in check. Mice with a disrupted CTLA-4 gene die from a massive accumulation of activated T cells.

Most of the T (and B) effector cells produced during an immune response must be eliminated after they have done their job. As antigen levels fall and the response subsides, effector cells are deprived of the antigen and cytokine stimulation that they need to survive, and the majority die by apoptosis. Only memory cells and some long-lived effector cells survive.

Table 24-3 summarizes some of the co-receptors and other accessory proteins found on the surface of T cells.

Table 24-3. Some Accessory Proteins on the Surface of T Cells.

Table 24-3

Some Accessory Proteins on the Surface of T Cells.

Before considering how effector helper T cells help activate macrophages and B cells, we need to discuss the two functionally distinct subclasses of effector helper T cells, TH1 and TH2 cells, and how they are generated.

The Subclass of Effector Helper T Cell Determines the Nature of the Adaptive Immune Response

When a an antigen-presenting cell activates a naïve helper T cell in a peripheral lymphoid tissue, the T cell can differentiate into either a TH1 or TH2 effector helper cell. These two types of functionally distinct subclasses of effector helper T cells can be distinguished by the cytokines they secrete. If the cell differentiates into a TH1 cell, it will secrete interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α) and will activate macrophages to kill microbes located within the macrophages' phagosomes. It will also activate cytotoxic T cells to kill infected cells. Although, in these ways, TH1 cells mainly defend an animal against intracellular pathogens, they may also stimulate B cells to secrete specific subclasses of IgG antibodies that can coat extracellular microbes and activate complement.

If the naïve T helper cell differentiates into a TH2 cell, by contrast, it will secrete interleukins 4, 5, 10, and 13 (IL-4, IL-5, IL-10, and IL-13) and will mainly defend the animal against extracellular pathogens. A TH2 cell can stimulate B cells to make most classes of antibodies, including IgE and some subclasses of IgG antibodies that bind to mast cells, basophils, and eosinophils. These cells release local mediators that cause sneezing, coughing, or diarrhea and help expel extracellular microbes and larger parasites from epithelial surfaces of the body.

Thus, the decision of naïve helper T cells to differentiate into TH1 or TH2 effector cells influences the type of adaptive immune response that will be mounted against the pathogen—whether it will be dominated by macrophage activation or by antibody production. The specific cytokines present during the process of helper T cell activation influence the type of effector cell produced. Microbes at a site of infection not only stimulate dendritic cells to make cell-surface B7 costimulatory proteins; they also stimulate them to produce cytokines. The dendritic cells then migrate to a peripheral lymphoid organ and activate naïve helper T cells to differentiate into either TH1 or TH2 effector cells, depending on the cytokines the dendritic cells produce. Some intracellular bacteria, for example, stimulate dendritic cells to produce IL-12, which encourages TH1 development, and thereby macrophage activation. As expected, mice that are deficient in either IL-12 or its receptor are much more susceptible to these bacterial infections than are normal mice. Many parasitic protozoa and worms, by contrast, stimulate the production of cytokines that encourage TH2 development, and thereby antibody production and eosinophil activation, leading to parasite expulsion (Figure 24-66).

Figure 24-66. The activation of TH1 and TH2 cells.

Figure 24-66

The activation of TH1 and TH2 cells. The differentiation of helper T cells into either TH1 or TH2 effector cells determines the nature of the subsequent adaptive immune responses that the effector cells activate. Whether a naïve helper T cell (more...)

Once a TH1 or TH2 effector cell develops, it inhibits the differentiation of the other type of helper T cell. IFN-γ produced by TH1 cells inhibits the development of TH2 cells, while IL-4 and IL-10 produced by TH2 cells inhibit the development of TH1 cells. Thus, the initial choice of response is reinforced as the response proceeds.

The importance of the TH1/TH2 decision is illustrated by individuals infected with Mycobacterium leprae, the bacterium that causes leprosy. The bacterium replicates mainly within macrophages and causes either of two forms of disease, depending mainly on the genetic make-up of the infected individual. In some patients, the tuberculoid form of the disease occurs. TH1 cells develop and stimulate the infected macrophages to kill the bacteria. This produces a local inflammatory response, which damages skin and nerves. The result is a chronic disease that progresses slowly but does not kill the host. In other patients, by contrast, the lepromatous form of the disease occurs. TH2 cells develop and stimulate the production of antibodies. As the antibodies cannot get through the plasma membrane to attack the intracellular bacteria, the bacteria proliferate unchecked and eventually kill the host.

TH1 Cells Help Activate Macrophages at Sites of Infection

TH1 cells are preferentially induced by antigen-presenting cells that harbor microbes in intracellular vesicles. The bacteria that cause tuberculosis for example, replicate mainly in phagosomes inside macrophages, where they are protected from antibodies. They are also not readily attacked by cytotoxic T cells, which mainly recognize foreign antigens that are produced in the cytosol (see Figure 24-58). The bacteria can survive in phagosomes because they inhibit both the fusion of the phagosomes with lysosomes and the acidification of the phagosomes that is necessary to activate lysosomal hydrolases. Infected dendritic cells recruit helper T cells to assist in the killing of such microbes. The dendritic cells migrate to peripheral lymphoid organs, where they stimulate the production of TH1 cells, which then migrate to sites of infection to help activate infected macrophages to kill the microbes harboring in their phagosomes (see Figure 24-66).

TH1 effector cells use two signals to activate a macrophage. They secrete IFN-γ, which binds to IFN-γ receptors on the macrophage surface, and they display the costimulatory protein CD40 ligand, which binds to CD40 on the macrophage (Figure 24-67). (We see later that CD40 ligand is also used by helper T cells to activate B cells.) Once activated, the macrophage can kill the microbes it contains: lysosomes can now fuse more readily with the phagosomes, unleashing a hydrolytic attack, and the activated macrophage makes oxygen radicals and nitric oxide, both of which are highly toxic to the microbes (discussed in Chapter 25). Because dendritic cells also express CD40, the TH1 cells at sites of infection can also help activate them. As a result, the dendritic cells increase their production of class II MHC proteins, B7 costimulatory proteins, and various cytokines, especially IL-12. This makes them more effective at stimulating helper T cells to differentiate into TH1 effector cells in peripheral lymphoid organs, providing a positive feedback loop that increases the production of TH1 cells and, thereby, the activation of macrophages.

Figure 24-67. The differentiation of TH1 cells and their activation of macrophages.

Figure 24-67

The differentiation of TH1 cells and their activation of macrophages. (A) An infected dendritic cell that has migrated from a site of infection to a peripheral lymphoid organ activates a naïve helper T cell to differentiate into a TH1 effector (more...)

TH1 effector cells stimulate an inflammatory response by recruiting more phagocytic cells into the infected site. They do so in three ways:

1.

They secrete cytokines that act on the bone marrow to increase the production of monocytes (macrophage precursors that circulate in the blood) and neutrophils.

2.

They secrete other cytokines that activate endothelial cells lining local blood vessels to express cell adhesion molecules that cause monocytes and neutrophils in the blood to adhere there.

3.

They secrete chemokines that direct the migration of the adherent monocytes and neutrophils out of the bloodstream into the site of infection.

TH1 cells can also help activate cytotoxic T cells in peripheral lymphoid organs by stimulating dendritic cells to produce more costimulatory proteins. In addition, they can help effector cytotoxic T cells kill virus-infected target cells, by secreting IFN-γ, which increases the efficiency with which target cells process viral antigens for presentation to cytotoxic T cells (see Figure 24-59). An effector TH1 cell can also directly kill some cells itself, including effector lymphocytes: by expressing Fas ligand on its surface, it can induce effector T or B cells that express cell-surface Fas to undergo apoptosis (see Figure 24-46B).

Both TH1 and TH2 cells can help stimulate B cells to proliferate and differentiate into either antibody-secreting effector cells or memory cells. They can also stimulate B cells to switch the class of antibody they make, from IgM (and IgD) to one of the secondary classes of antibody. Before considering how helper T cells do this, we need to discuss the role of the B cell antigen receptor in the activation of B cells.

Antigen Binding Provides Signal 1 to B Cells

Like T cells, B cells require two types of extracellular signals to become activated. Signal 1 is provided by antigen binding to the antigen receptor, which is a membrane-bound antibody molecule. Signal 2 is usually provided by a helper T cell. Like a T cell, if a B cell receives the first signal only, it is usually eliminated or functionally inactivated, which is one way in which B cells become tolerant to self antigens.

Signaling through the B cell antigen receptor works in much the same way as signaling through the T cell receptor (see Figure 24-64). The receptor is associated with two invariant protein chains, Igα and Igβ, which help convert antigen binding to the receptor into intracellular signals. When antigen cross-links its receptors on the surface of a B cell, it causes the receptors and its associated invariant chains to cluster into small aggregates. This aggregation leads to the assembly of an intracellular signaling complex at the site of the clustered receptors and to the initiation of a phosphorylation cascade (Figure 24-68).

Figure 24-68. Signaling events activated by the binding of antigen to B cell receptors (signal I).

Figure 24-68

Signaling events activated by the binding of antigen to B cell receptors (signal I). The antigen cross-links adjacent receptor proteins, which are transmembrane antibody molecules, causing the receptors and their associated invariant chains (Igα (more...)

Just as the CD4 and CD8 co-receptors on T cells enhance the efficiency of signaling through the T cell receptor, so a co-receptor complex that binds complement proteins greatly enhances the efficiency of signaling through the B cell antigen receptor and its associated invariant chains. If a microbe activates the complement system (discussed in Chapter 25), complement proteins are often deposited on the microbe surface, greatly increasing the B cell response to the microbe. Now, when the microbe clusters antigen receptors on a B cell, the complement-binding co-receptor complexes are brought into the cluster, increasing the strength of signaling (Figure 24-69A). As expected, antibody responses are greatly reduced in mice lacking either one of the required complement components or complement receptors on B cells.

Figure 24-69. The influence of B cell co-receptors on the effectiveness of signal I.

Figure 24-69

The influence of B cell co-receptors on the effectiveness of signal I. (A) The binding of microbe-complement complexes to a B cell cross-links the antigen receptors to complement-binding, co-receptor complexes. The cytosolic tail of one component of the (more...)

Later in the immune response, by contrast, when IgG antibodies decorate the surface of the microbe, a different co-receptor comes into play to dampen down the B cell response. These are Fc receptors, which bind the tails of the IgG antibodies. They recruit phosphatase enzymes into the signaling complex that decrease the strength of signaling (Figure 24-69B). In this way the Fc receptors on B cells act as inhibitory co-receptors, just as the CTLA-4 proteins do on T cells. Thus, the co-receptors on a T cell or B cell allow the cell to gain additional information about the antigen bound to its receptors and thereby make a more informed decision as to how to respond.

Unlike T cell receptors, the antigen receptors on B cells do more than just bind antigen and transmit signal 1. They deliver the antigen to an endosomal compartment where the antigen is degraded to peptides, which are returned to the B cell surface bound to class II MHC proteins (see Figure 24-60). The peptide-class-II-MHC complexes are then recognized by effector helper T cells, which can now deliver signal 2. Signal 1 prepares the B cell for its interaction with a helper T cell by increasing the expression of both class II MHC proteins and receptors for signal 2.

Helper T Cells Provide Signal 2 to B Cells

Whereas antigen-presenting cells such as dendritic cells and macrophages are omnivorous and ingest and present antigens nonspecifically, a B cell generally presents only an antigen that it specifically recognizes. In a primary antibody response, naïve helper T cells are activated in a peripheral lymphoid organ by binding to a foreign peptide bound to a class II MHC protein on the surface of a dendritic cell. Once activated, the effector helper T cell can then activate a B cell that specifically displays the same complex of foreign peptide and class II MHC protein on its surface (see Figure 24-66).

The display of antigen on the B cell surface reflects the selectivity with which it takes up foreign proteins from the extracellular fluid. These foreign proteins are selected by the antigen receptors on the surface of the B cell and are ingested by receptor-mediated endocytosis. They are then degraded and recycled to the cell surface in the form of peptides bound to class II MHC proteins. Thus, the helper T cell activates those B cells with receptors that specifically recognize the antigen that initially activated the T cell, although the T and B cells usually recognize distinct antigenic determinants on the antigen (see Figure 24-70). In secondary antibody responses, memory B cells themselves can act as antigen-presenting cells and activate helper T cells, as well as being the subsequent targets of the effector helper T cells. The mutually reinforcing actions of helper T cells and B cells lead to an immune response that is both intense and highly specific.

Figure 24-70. Comparison of the signals required to activate a helper T cell and a B cell.

Figure 24-70

Comparison of the signals required to activate a helper T cell and a B cell. Note that in both cases secreted and membrane-bound molecules can cooperate to provide signal 2. Although not shown, CD40 is also expressed on the surface of mature dendritic (more...)

Once a helper T cell has been activated to become an effector cell and contacts a B cell, the contact initiates an internal rearrangement of the helper cell cytoplasm. The T cell orients its centrosome and Golgi apparatus toward the B cell, as described previously for an effector cytotoxic T cell contacting its target cell (see Figure 24-45). In this case, however, the orientation is thought to enable the effector helper T cell to provide signal 2 by directing both membrane-bound and secreted signal molecules onto the B cell surface. The membrane-bound signal molecule is the transmembrane protein CD40 ligand, which we encountered earlier and is expressed on the surface of effector helper T cell, but not on nonactivated naïve or memory helper T cells. It is recognized by the CD40 protein on the B cell surface. The interaction between CD40 ligand and CD40 is required for helper T cells to activate B cells to proliferate and differentiate into memory or antibody-secreting effector cells. Individuals that lack CD40 ligand are severely immunodeficient. They are susceptible to the same infections that affect AIDS patients, whose helper T cells have been destroyed.

Secreted signals from helper T cells also help B cells to proliferate and differentiate and, in some cases, to switch the class of antibody they produce. Interleukin-4 (IL-4) is one such signal. Produced by TH2 cells, it collaborates with CD40 ligand in stimulating B cell proliferation and differentiation, and it promotes switching to IgE antibody production. Mice deficient in IL-4 production are severely impaired in their ability to make IgE.

The signals required for T and B cell activation are compared in Figure 24-70, and some of the cytokines discussed in this chapter are listed in Table 24-4.

Table 24-4. Properties of Some Interleukins.

Table 24-4

Properties of Some Interleukins.

Some antigens can stimulate B cells to proliferate and differentiate into antibody-secreting effector cells without help from T cells. Most of these T-cell-independent antigens are microbial polysaccharides that do not activate helper T cells. Some activate B cells directly by providing both signal 1 and signal 2. Others are large polymers with repeating, identical antigenic determinants (see Figure 24-29B); their multipoint binding to B cell antigen receptors can generate a strong enough signal 1 to activate the B cell directly, without signal 2. Because T-cell-independent antigens do not activate helper T cells, they fail to induce B cell memory, affinity maturation, or class switching, all of which require help from T cells. They therefore mainly stimulate the production of low-affinity (but high-avidity) IgM antibodies. Most B cells that make antibodies without T cell help belong to a distinct B cell lineage. They are called B1 cells to distinguish them from B2 cells, which require T cell help. B1 cells seem to be especially important in defense against intestinal pathogens.

Immune Recognition Molecules Belong to an Ancient Superfamily

Most of the proteins that mediate cell-cell recognition or antigen recognition in the immune system contain Ig or Ig-like domains, suggesting that they have a common evolutionary history. Included in this Ig superfamily are antibodies, T cell receptors, MHC proteins, the CD4, CD8, and CD28 co-receptors, and most of the invariant polypeptide chains associated with B and T cell receptors, as well as the various Fc receptors on lymphocytes and other white blood cells. All of these proteins contain one or more Ig or Ig-like domains. In fact, about 40% of the 150 or so polypeptides that have been characterized on the surface of white blood cells belong to this superfamily. Many of these molecules are dimers or higher oligomers in which Ig or Ig-like domains of one chain interact with those in another (Figure 24-71).

Figure 24-71. Some of the membrane proteins belonging to the Ig superfamily.

Figure 24-71

Some of the membrane proteins belonging to the Ig superfamily. The Ig and Ig-like domains are shaded in gray, except for the antigen-binding domains (not all of which are Ig domains), which are shaded in blue. The function of Thy-1 is unknown, but it (more...)

The amino acids in each Ig-like domain are usually encoded by a separate exon. It seems likely that the entire gene superfamily evolved from a gene coding for a single Ig-like domain—similar to that encoding β2-microglobulin (see Figure 24-50A) or the Thy-1 protein (see Figure 24-71)—that may have mediated cell-cell interactions. There is evidence that such a primordial gene arose before vertebrates diverged from their invertebrate ancestors about 400 million years ago. New family members presumably arose by exon and gene duplications.

The multiple gene segments that encode antibodies and T cell receptors may have arisen when a transposable element, or transposon (discussed in Chapter 5), inserted into an exon of a gene encoding an Ig family member in an ancestral lymphocyte-like cell. The transposon may have contained the ancestors of the rag genes, which, as discussed earlier, encode the proteins that initiate V(D)J joining; the finding that the RAG proteins can act as transposons in a test tube strongly supports this view. Once the transposon had inserted into the exon, the gene could be expressed only if the transposon was excised by the RAG proteins and the two ends of the exon were rejoined, much as occurs when the the V and J gene segments of an Ig light chain gene are assembled (see Figure 24-37). A second insertion of the transposon into the same exon may then have divided the gene into three segments, equivalent to the present-day V, D, and J gene segments. Subsequent duplication of either the individual gene segments or the entire split gene may have generated the arrangements of gene segments that characterize the adaptive immune systems of present-day vertebrates.

Adaptive immune systems evolved to defend vertebrates against infection by pathogens. Pathogens, however, evolve more quickly, and they have acquired remarkably sophisticated strategies to counter these defenses, as we discuss in Chapter 25.

Summary

Naïve T cells require at least two signals for activation. Both are provided by an antigen-presenting cell, which is usually a dendritic cell: signal 1 is provided by MHC-peptide complexes binding to T cell receptors, while signal 2 is mainly provided by B7 costimulatory proteins binding to CD28 on the T cell surface. If the T cell receives only signal 1, it is usually deleted or inactivated. When helper T cells are initially activated on a dendritic cell, they can differentiate into either TH1 or TH2 effector cells, depending on the cytokines in their environment: TH1 cells activate macrophages, cytotoxic T cells, and B cells, while TH2 cells mainly activate B cells. In both cases, the effector helper T cells recognize the same complex of foreign peptide and class II MHC protein on the target cell surface as they initially recognized on the dendritic cell that activated them. They activate their target cells by a combination of membrane-bound and secreted signal proteins. The membrane-bound signal is CD40 ligand. Like T cells, B cells require two simultaneous signals for activation. Antigen binding to the B cell antigen receptors provides signal 1, while effector helper T cells provide signal 2 in the form of CD40 ligand and various cytokines.

Most of the proteins involved in cell-cell recognition and antigen recognition in the immune system, including antibodies, T cell receptors, and MHC proteins, as well as the various co-receptors discussed in this chapter, belong to the ancient Ig superfamily. This superfamily is thought to have evolved from a primordial gene encoding a single Ig-like domain.

By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.

Copyright © 2002, Bruce Alberts, Alexander Johnson, Julian Lewis, Martin Raff, Keith Roberts, and Peter Walter; Copyright © 1983, 1989, 1994, Bruce Alberts, Dennis Bray, Julian Lewis, Martin Raff, Keith Roberts, and James D. Watson .
Bookshelf ID: NBK26827