DNA and the Immune System

  1. David S. Pisetsky, MD, PhD
  1. Durham Veterans Administration Medical Center, Duke University Medical Center, Durham, NC 27705 Grant Support: In part by a Veterans Administration Merit Review grant; by the Veterans Administration Research Center on AIDS and HIV Infection; and by grant AR-39162 from the National Institutes of Health. Requests for Reprints: David S. Pisetsky, MD, PhD, Durham Veterans Administration Medical Center, Box 151G, 508 Fulton Street, Durham, NC 27705.

    Deoxyribonucleic acid (DNA) conveys an enormous amount of information because of the myriad ways in which its four bases can be arranged. As a genetic blueprint, this information provides elaborate instructions for the structure of proteins and the regulation of their expression. Encoded in DNA, however, is information of a fundamentally different kind. As provocative new data indicate, DNA from bacteria has sequences that can instruct the immune system to distinguish “foreign” from “self.” These sequences, which bear characteristic motifs, can trigger innate immunity and are important not only in host defense but in the burgeoning use of DNA to prevent and treat disease [1].

    The categorization of DNA as an immune activator contrasts with previous portrayals of DNA as immunologically uniform and inert (Figure 1). Only in systemic lupus erythematosus did DNA attract interest as an immunologic player. In this prototypic autoimmune disease, antibodies to DNA are prominent and serve as markers for diagnosis and prognosis. The diagnostic usefulness of antibodies to DNA has suggested that immune responses to DNA reflect events unique to the autoimmune state. Indeed, immunization of normal animals with DNA fails to induce systemic lupus erythematosus; this result enforces the view that DNA differs from other macromolecules in its ability to generate responses [2, 3].

    Figure 1. The immunologic properties of bacterial DNA result from CpG motifs and may influence the human immune system in several settings. A. In normal immunity, bacterial DNA can nonspecifically stimulate the immune system by activating cytokine production and B-cell immunoglobulin secretion. B. Normal humans produce antibodies to DNA that target specific sequences exclusive to bacterial DNA. In persons with systemic lupus erythematosus, antibodies to DNA bind conserved backbone determinants present on both bacterial and mammalian DNA. C. Antisense agents are small oligonucleotides that can block specific gene expression. Because of their sequence and backbone modifications (shown schematically by the zig-zag line), some antisense compounds can also mimic bacterial DNA and nonspecifically promote B-cell activation (see panel A). D. The DNA vaccines are plasmid constructs encoding a foreign protein that is then released into the immune system of the host, where it serves as an antigen that stimulates targeted protective immunity. These plasmids are taken up by cells and lead to effective B-cell and T-cell responses. The potency of these vaccines may be nonspecifically enhanced by the presence of CpG motifs through mechanisms similar to those shown in panel A. Ag = antigen; IFN = interferon; Ig = immune globulin; IL = interleukin; NK = natural killer; SLE = systemic lupus erythematosus. Immunologic activities of bacterial DNA.

    Revision of these ideas has come gradually and was spurred initially by findings in the field of tumor immunology. In studies done to identify bacterial products with antitumor activity, Tokunaga and colleagues [4] showed that DNA extracted from Mycobacterium bovis bacille Calmette–Guérin organisms caused the regression of transplantable tumors in mice. As shown in vivo and in vitro, the antitumor activity resulted not from direct cytotoxicity but from the enhancement of natural killer cells. The increase in natural killer cell activity was found to be a consequence of the DNA-induced production of interferon-α, interferon-β, and interferon-γ. The activation of natural killer cells by the DNA extracts also occurred with human peripheral blood cells [5, 6].

    Subsequent studies have shown that DNA from many microorganisms is immunomodulatory, an effect that is attributable to short sequences found in bacterial DNA. These sequences contain a six-base motif with an unmethylated cytosine-guanosine core flanked by two 5′ purines and 3′ pyrimidines [7]. These sequences are active as small oligonucleotides and can induce interferon and natural killer cell activity in both humans and mice. In mice, these sequences (as well as intact bacterial DNA) can also stimulate the production of interleukin-6 and interleukin-12. They can also directly activate B cells for proliferation and antibody production [8-10].

    The six-base motif has been called an immunostimulatory sequence and occurs at a dramatically higher frequency in bacterial DNA than in mammalian DNA. In mammalian DNA, cytosine and guanosine occur together much less often than predicted by chance, a phenomenon known as “CpG suppression.” Furthermore, cytosine is often methylated in mammalian DNA, perhaps to regulate transcriptional activity; cytosine is not methylated in bacterial DNA [11]. The origin of these structural differences is not understood, but they may have evolved as a system for distinguishing foreign DNA.

    In its immunologic activities, bacterial DNA resembles endotoxin. This suggests that foreign DNA can trigger innate immunity and can function synergistically with endotoxin in promoting immune responses during infection. The ability of foreign nucleic acids to induce immunity has precedent because double-stranded RNA, which occurs in viruses and their replication intermediates, is also a potent inducer of interferon-α, interferon-β, and interleukin-12 [12]. These findings suggest that, in general, foreign nucleic acids can stimulate immune responses because of structural microheterogeneity.

    Demonstration of the immunologic activity of bacterial DNA has led to a reevaluation of the serology of systemic lupus erythematosus. Although most studies indicate that antibodies to DNA occur only in patients with systemic lupus erythematosus, the expression of antibodies to DNA is widespread among normal persons [13]. These responses were previously missed because of a failure to test an adequate number of bacterial DNA as antigens. Thus, antibodies to DNA from certain bacteria, including Micrococcus lysodeikticus and Staphylococcus epidermidis, occur abundantly in the serum of most normal humans. However, these antibodies differ in isotype from the antibodies to DNA seen in persons with systemic lupus erythematosus. Like responses to bacterial carbohydrates, they are predominantly IgG2. In contrast, antibodies to DNA in persons with systemic lupus erythematosus are predominantly IgG1 [14]. Furthermore, unlike antibodies to DNA in persons with systemic lupus erythematosus, which bind conserved backbone determinants on all DNA, antibodies to DNA in normal serum bind to the sequential determinants present on some DNA.

    Serologic findings suggest that bacterial DNA is immunogenic and can drive production of antibodies (including autoantibodies to DNA) during ordinary encounters with bacteria. This possibility is supported by the results of immunization experiments done in animals. Thus, under conditions in which mammalian DNA is inactive in normal animals, DNA from Escherichia coli can induce antibodies that specifically target bacterial antigen. Furthermore, in autoimmune NZB/NZW mice, DNA from E. coli can elicit autoantibodies to DNA that, like antibodies to DNA in persons with systemic lupus erythematosus, can bind to DNA from all species. Preliminary evidence suggests that abnormalities in the B-cell repertoire of autoimmune persons determines whether contact with foreign DNA leads to a protective response to nonconserved sequences on bacterial DNA or to a deleterious response to the conserved conformational determinants on all DNA [15].

    In addition to having a role in host defense and autoimmunity, the immunomodulatory properties of bacterial DNA may influence therapeutic approaches, such as antisense therapy and DNA vaccination, that use free or “naked” DNA. Antisense agents are short oligonucleotides that can inhibit the production of specific proteins through interactions with their messenger RNA (sense). These agents have the potential for broad applicability in medicine; projected uses include therapy for viral infections and tumors. As recent data indicate, some antisense agents-because of their sequence or because of chemical modifications in their backbone-show immunostimulatory activities that resemble bacterial DNA. These effects, which include polyclonal B-cell activation, could confound the use of these agents unless the agents are engineered judiciously [16].

    One of the most exciting developments in the field of infectious diseases has been the discovery of DNA vaccines. This radically new approach to immunization involves the administration (through intramuscular or intradermal injection) of a DNA plasmid encoding a foreign protein. Through an unknown mechanism, this plasmid DNA enters cells, where it can be stably expressed and can lead to the induction of protective humoral and cellular immune responses to the encoded protein [17, 18].

    Because the plasmid vectors for vaccination are propagated in bacteria, they bear bacterial sequences and patterns of base methylation. These sequences can induce cytokines that can influence both the magnitude of the response and the type of helper T cells. Most DNA vaccinations lead to a Th1 response, which would be expected from the induction of interleukin-12 and interferon-γ by immunostimulatory sequences on the plasmid [19, 20]. Although the immunostimulatory sequences may provoke inflammation or production of autoantibodies, these events have not complicated the use of DNA vaccines in animal models thus far.

    The immune properties of bacterial DNA suggest new pathways of host defense as well as a novel paradigm for gene expression in which exogenous DNA serves as a regulator. The next few years should prove fascinating as these issues are explored in the context of clinical and basic research and in the development of powerful new pharmaceutical agents.

    David S. Pietsky, MD, PhD

    Durham Veterans Administration Medical Center; Duke University Medical Center; Durham, NC 27705

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.
    « Previous | Next Article »Table of Contents

    Navigate This Article