
Mycoplasma Interactions with the Immune System: Implications for Disease Pathology (Barry C Cole)
Members of the class Mollicutes, mycoplasmas and ureaplasmas, may play a more significant role in human disease than previously realized. Although Mycoplasma pneurnoniae is a well-recognized pulmonary pathogen, it can cause other less common, but important, complications, such as erythematous maculopapular and vesicular exanthems, ulcerative stomatitis and conjunctivitis, hemolytic anemia, central nervous system disease, invasive pericarditis, and arthritis. Mycoplasma hominis and Ureaplasma urealyticum contribute to a variety of diseases of the human genitourinary tract, including nongonococcal urethritis, low birth weights. and pulmonary infections of the newborn. They also may contribute to prostatitis, Reiter's disease, and vaginitis. U. urealyticum is recovered from a high percentage of vaginal and urine cultures of women with systemic lupus erythematosus. In addition, Mycoplasma fermentans, Mycoplasma penetrans and Mycoplasma pirium are frequently isolated from AIDS patients, although their role in this disease remains to be defined. M. fermentans causes lethal, multiorgan fulminant infections. Immunocompromised individuals are particularly susceptible to damaging joint infections by mycoplasmas previously considered to bc commensaIs.
Mycoplasmas exert profound effects on immune cells in vitro and on the immune systems of the hosts which they infect. The various genera and species of the class Mollicutes are quite diverse in their genomic composition, their metabolic pathways, and the specific properties that lead to disease pathology. Much research is now focusing on the pathways by which immunomodulatory effects are mediated. The superantigens (SAg) represent a new class of microbial proteins that interact with the immune system in a unique manner. One of the prototype examples is the T-cell mitogen produced by Mycoplasma arthritidis, an agent of acute and chronic murine arthritis. SAg exhibit potent effects on lymphocytes, macrophages and natural killer (NK) cells and may play a role in diseases such as diabetes mellitus, multiple sclerosis, rheumatold arthritis (RA), and AIDS.
Protective Pathways against M. pneumoniae Infection
M pneumoniae is the major cause of human primary atypical pneumonia. which most often affects individuals between 5 and 20 years of age. Symptoms range from a relitively mild upper respiratory tract infection with flu-like symptoms to more serious conditions pneumonia, disseminated intravascular clotting, and adult respiratory distress syndrome. M. pneumoniae apparently exerts multiple effects on the immune systems, including polyclonal activation of T and B cells and the associated secretion of cytokines.
Whereas the binding of' the tip organelle of' M. pneumoniae to respiratory epithelial cells plays a major role in the pathogenic properties of' the organism, there is now evidence that a distinct binding site on mononclear (MNC) may evoke a protective response according to Brent Limbaugh working with Marianna Egan at the Birmingham Veterans Affairs Medical Center and the University of Alabama in Birmingham. A low dose of M. pneumoniae (multiplicity of infection of 1:1) rapidly increases the activity, in mixed MNC cultures, of NK cells, important in first line defenses against many microorganisms. Following M. pneumoniae-induced NK cell activation, expression of' the cytotoxicity-related antigens CD69 and CD25 on CD56+ K cells is greatly increased. However, purified CD56+ NK cells could not be directly activated by the organisms.
Depletion of monocytes (CD14+), T cells (CD3+) and B cells (CD19+) from MNC fails to affect NK cell activation. However, adding a subpopulation of MNC exhibiting a CD14-, CD3-, CD19-, CD56-, and DR+ phenotype restores the activation of purified CD56+ NK cells by M. pneumonie. Exposure to M. pneumoniae visibly activates this subpopulation of NK cell activator (NKCA) cells. Since monoclonal antibody (MAb) to CD18 inhibits M. pneumoniae augmentation of NK activity, Limbaugh and Egan speculate that NKCA cells activate NK cells through the receptor ligand interactions of LFA-1 on NK cells and ICAM-1 on NKCA cells.
NKCA cells do not selectively bind M. pneumoniae cells; instead, all subsets of MNC bind M. pneumoniae equally. The binding site for M. pneumoniae on MNC is different from that on respiratory epithelial cells. Thus, M. pneumoniae nonreverting mutants, which lack a functional binding tip organelle and do not bind epithelial cells, bind equally well to MNC and activate NK cells in the same way as wild-type organisms. Thus, the receptor for M. pneumoniae on MNC, which is distinct from the receptor on epithelial cells, activates a pathway that protects against infection by M. pneumoniae.
Role of Cytokines in the Pathogenesis of MRM
Athough immunopathology is a major component of many mycoplasma respiratory diseases, little is known about the underlying immune and inflammatory mechanisms. However, recent studies in patients and in animal models suggest that T-cell responses contribute to the severity of M. pneumoniae disease. For instance, Jerry Simecka of the University of North Texas Health Sciences Center in Fort Worth in collaboration with Gail Cassell, Samuel Cartner, Chad Faulkner, and J. Russell Lindsey at the University of Alabama in Birmingham is studying murine respiratory mycoplasmosis (MRM ) to identify protective and immunopathologic pathways.
The early events following infection with Mycoplasma pulmonis play a major role in the subsequent development of the disease. Instead of' antibody or T-Cell activation, alveolar macrophages clear this pathogen from the lungs. According to Simecka and his colleagues, macrophages and cytokines also contribute to pulmonary immunopathology. Thus, as early as 1 h after infection, the levels of tumour necrosis factor (TNF-a)
more to follow....
and interleukin-let (IL-Ici) mRNA increase in the~ .~ing. t)f infectt'-~l (-tnim~ils ~ TNF-L~ appeat's in ui i~ .'~ L~C tlutti. ."""Li~).~L'qLiL'Iit'.y . 4 h tlL~r ir~fet-tiui,
IL-I l'~ an(l ll~6 £'ir{'. dett'cted, l)ttL IFN-y mRNA does not incre('tse until at least '24 h after infection. Increased TNF-cy, IL-i, and IL-6 mRNA expression is confined to alve(il -ii' macrophages, suggesting that stepwise activation of' niacrophages is initiated by mycoplasmas and medi('tted by the various cytokines produced. TNF-cx is also important in subsequent in '-ivo production of IL-i and IL-6 and can stimulate NK cells to produce gamma interferon (IFN-y). These cytokines may~thus promote antimicrobial activities within the lung, with TNF-~ playing a direct role in pulmonary clearance. Cytokines are also implicated in the pathophysiology of lung injury. By comparing cytokine production in mouse strains which differ in susceptibility to virulent and avirulent strains of M. pulmonis, Simecka and his colleagues showed that higher levels of TNF-~ and IL-6 production correspond to increased severity of acute alveolar disease early after infection. TNF-a, an important mediator of neutrophil recruitment, is elevated in lungs of infected animals, and large numbers of neutrophils are also present. IL-i, which also enhances neutrophil recruitment, acts synergistically with TNF-~. IL-6, a mediator of acute-phase proteins, has both local and systemic effects on the host. ~ but not TNF-~ or IL-i, shows a prolonged elevation in samples from mice with severe respiratory disease. In addition, its increase of IL-6 in the lungs and sera is prolonged. In chronic respiratory mycoplasmosis, mouse strain differences significantly influence the outcome of disease. For example, lymphoid infiltration in the susceptible strain DBAI2N is less pronounced and the disease has a higher fatality than in the susceptible C3H mouse. Differences in cheniotactic cytokines, the ~ chemokines, might be responsible. Lymphocytes also play a major role in chronic disease. Whereas antibody to M. puIrno~iis is unable to clear organisms once infection is established, high levels of antibody are associated with more-sevei'e disease. To determine the role of B and T cells in this disease, Simecka and colleagues also are monitoring this infection in mice with severe combined immunodeficiency (SCID). In one set of experiments, immune serum from M. pulmonis-infccted mice was intravenously injected into SCID mice 7 days after infection with Al. pu1moni.~. These mice did not develop severe lung disease, nor was arthritis evident. To examine lyniphocyte function in the pathogenesis of' MRM, splenic lymphocytes from uninfected mice were iiijected into SCID mice, and 1 week later, the mice weic infected. ~Vithiii 2 weeks, the lymphocytereconsLiLuted SCID nuce developed severe lung disease. similar to that iii immunoconipetent mice. In contrast, untreated SCID mice (controls) developed less-severe lung disease after .V!~ piLl??) Uni.'; infectitin. 'l'li us, lynipht~cy Li' I'esl)()nses hive 1)0th CIctri inent~il aud beuclicial ell'~cts on the pd'thogenesis of' Al. pulI7li)' nis disease in mice. The investi~attirs ci)nclude that, depeuding on cii'- cums~n~es, cytokine production and lymphocyte activ,~', i~)t~ i'ith~i' mi!iimizi ~ i)r ','unttIbtit~ ti) di.~L'~'1~L'
~-chemoki nes ~ rE as~ocia ted with earlv l~.~i()n dcvQloI)I1)Lnt, but they md'y also pri)!fl()te ho.~t defenses d'gainst mycoplasma in fecti on. Moreover1 lymphocytes ci)ntribute to the development of pulmonary I nflammation in chronic M. pu1nzoni.~ disease, ~'ind T-cell responses are likely to be involved, because antibody alone does not exacerbate pulmonary lesions.
Ureaplasma Urease: Biologic Mimicry with HLAB27
Specific class I and class II major histocompatibility complex (MHC) alleles predispose individuals to a number of diseases in which dysfunction of the immune system leads to inflammatory responses. The strongest such association occurs in patients having the class I antigen HLA-B27. The incidence of this cell membrane protein in patients with ankylosing spondylitis, Reiter's syndrome, postinfection or sexually acquired reactive arthritis, acute anterior uveitis, or juvenile RA, is 90, 80, 60 to 80, 50, and 25%, respectively. Healthy individuals range from 9 to 15% positive for HLA-B27. Ankylosing spondylitis, Reiter's syndrome, and reactive arthritis develop after infection with certain microorganisms. For these HLA-B27-as-sociated diseases a microbial antigen may act as a molecular mimic of B27, inducing host antibodies that cross-react with host B27 molecules, according to John Davis and R. Sydney Alozie of Bronx Community College, City Uiiiversity of New York, and Kustina Williams of the Food and Drug Administration, Bethesda, Md. For ii'istance, MAb to HLA-B27 cross-re~'tcts with a number of microbial ureases, including those from Kiebsiella p~leumofliae, Yersinia enterocolitica, Heli')bacter pylon, and U. urealyticum, each of which has been implicated in HLA-B27-associated diseases. Consistent with these obsei-vations, there is sequence homology between the Ureaplasnia urease large (66- to 72-kDa) subunit (Ut-cC) and the synthetic immunogen that was used to raise the MAb to HLA-B27. They find that three hydrophilic urease peptides representing tl~ree epitopes cross-react with the NIAb, suggesting that microbial ureases mimic MHC antigens and account for the HLA-B27 component of inflammatory disea-es associated with these organisms. Biological mimicry between bacterial or viral agents and heat shock proteins or joint constituei~ts may account for a number of other human rheumatic diseases, including i'l~eun~atoid arthritis. MQanwhile. i'eactive peptides enhance binding of t~is MAb to B27-positive cells but not to B27~ncgative c~ll.'-', whereas nonre'~ctive pe1)ti(1es have no effect on MAb l)inding. These dat-i suggest that a mechanism otl'er th~-in mimicry participates. '['he investigators are ex~mlning the potential arthritugenic effect~ of Ureaq ill-ease and u i-ease-derived peptides. Mycoplasmas and SAg: Models for the Triggering of Autoimniune t)isease
The term SAg designates a fascin'iting group of inicl-()l)ial products with a profound impact on the immuTle system. SAg which trigger a selective Tandlor B-cell mitogenic response include murine mainmary tumor virus gene products that encode what were previously known as minor lymphocyte-stimulating antigens, an expanding group of enterotoxins produced by common bacteria (including Staphylococcus aureus), the Streptococcus M protein, and the gene product of at least one species of mycoplasma, M. arthritidis. The T-cell SAg share a number of features which distinguish them from conventional peptide antigens. In the absence of antigen processing, SAg bind avidly to MHC class II molecules at a site distinct from the conventional antigen-binding cleft. The SAg-MHC class II complex is recognized by and activates a subset of T cells, based on the T-cell receptor (TCR) V~ chain gene family. In some cases, binding to V~ chains may precede the MHC interaction (see figure). Mapping studies suggest that the SAg interact with an exposed polymorphic region of the ~-pleated sheet of the TCR ~ chain, termed the CDR4 region, which is distinct from the highly polymorphic antigen-binding CDR3 region. Taken together, these properties render SAg potent V selective T-cell mitogens, activating from 5 to 30% o~the T-cell pool. My colleagues Curtis Atkin, Kevin Knudtson, Allen Sawitzke, Arnold Oliphant, and Muniraj Manohar and I are studying one such SAg, a mitogen derived from M. arthritidis (MAM). This SAg induces arthritis in rodents, leading to chronic disease that is associated with periods of exacerbation and resolution as well as production of rheumatoid factor (RF). Since a large number of mycoplasm1- species cause acute or chronic arthritis in different animal species, the mycoplasmas are considered potential etiologic agents for human rheumatic diseases. MAM is pi-eferentially presented to T cells by the c~ chain of the niurine H-2E molecule or the human HLA-DR molecule. Murine T cells expressing V~36 or V 8 chain~ or their human equivalents, are those w~ich mob"t' commonly respond to MAM. In recent studies MANI was purified to homogeneity and the encoding getie was identified, sul)cloned, and sequenced. Analysis of the amino acid sequence shows th1-it MAM is not phylogenetically related to any of the other bacterial or retroviral SAg. However, short i.egions of honiology in functionally active regions of the molecule are shared with regions on otl~er SAg that interreact witl~ MHC £-ind TCR molecules. Although uiii-elated, difl'ci'ent SAg may share functional sequences that could be cross-reactive in ii~imune modul~'ition and iii disease inaiiifestation. MAM also pussess~'s the lectin legume mE)tif~, which helps contigure the carbohydi-ite binding site fUL' l~ctin initugens. By studying the niycoplasma SAg ~ we are leurning more about how SAg influence the immune system in ViV(J md trigger autoimmune disease. Inject- mg ~ into i'espon~tve ii~uuse .~trains leuds to lymphadenop3thy and splenomegaly, ~long with clonal expansion of the i\IAM-reactive T cells that bear V~ 6 and V ~8 c".~ain segments of the TCR. Although induces Thi-type cytokii~es in vitro, cytokine profiles in vivo change fi'om a Thi- to ~ Th2-like pattern resulting in lncre('ised immunoglobulin synthesis and increased specific responses to injected antigells. When MA~i is injected into mice that have recovered fi-om autoimmune collagen-induced arthritis, which in BIO.RIII mice is driven by T cells expressing the V~6 and V~~8 TCRs, there is a marked increase in ai'thritis. No increase is seen when mice are injected with SAg that activate T cells bearing MAM-nonreactive V~ chains. Most iniportantly, mice which are suboptimally immunized with type II articular collagen and which never develop arthritis suffer severe arthritis when injected with MAiM 130 to 200 days later. Since MkM is not itself arthritogenic (except when injected in the joint), it is likely that MAM activates autoimmune arthritis by expanding preexisting autoimmune clones.
Role of SAg Being Explored in Human Disease
Can MAM, MAiM-like molecules, or SAg possessing potentially functional cross-reactive epitopes trigger human autoimmune arthritis? Recently, SAg activity was detected in patients with juvenile diabetes. providing some evid('?lc(' that this pathogenic mechanism applies to humans. Steven Friedman at the Hospital for Speci~l Surgery, New York, N.Y., and his colleague Mary Crow find that MAM may induce autoimmunity and SAg might play a role in human RA. Several mechanisms might mediate inflammatory di~ease that leads to autoimmunity. First, SAg may trigger a massive cytokine release. as In toxic shock syndrome induced by an S. aureus SAg, toxic shock syndrome toxin 1. Second, SAg may activate normally quiescent autoreactive T cells. Third, the dual aftinity of SAg for both MHC class II aud '1'CR V13 gene products promotes a form of cognate Th-B -cell interaction. accord i !ig to Friedman and Crow. They find that hutuan or murine irradiated CD£~ Th cells in the p~'e5cnce ot' MAM help purified B cells t~' pi-olite i-ate and synt.I1L'size immuni)globulin and I~F. ~l~)rei)vei', specific ,liitigei~s such as lieterologous ervtlii'iicvte iii Th-13-cell ~)l'..Is ~I.-\~l cultures promote ~ tic a ntibod,v pr()&ILI('tiun Of' particul~-ir interest . liunian unsepai'~iti-d pi'I~ pi-al bl~)od lyniphocvt~s .~ciiv'tted by MAM t)U 'itlier l)actel'i'il SAg beh~'tve iii d~st Oct ~v.ivs. Thus, fi'i' iiis~a'.ice. the staphylococcal enterotoxin SAg elicit a vigorous T-cell mitogenic response with no immunoglobulin production, whereas MAM is a less potent mitogen for peripheral human T cells but results in strong polyclonal immunoglobulin M and G responses. Although SAg can mediate apoptotic death of T cells, Th-B-c~ell activation could also result in B-cell apoptosis, a process which could down regulate humoral immunity, accol-ding to l?riedman and colleagues. Signalling of B cells is mediated by CD40, a member of the TNl'(FGF receptor family. When CD4OL combines with its lig~'ind on activated CD4 T cells, B-cell ,'ictivati()ii is pi)tent, le-Iding ti) a survival signal. Cugii~ite CI)4 'lli-B-cell intci'd'ctioii also leads to upregul~ition of B-cell las expressii)n and, in the absence of the CL)40L receptor, B-cell ipoptosis. ~'Ii~ tl~i l'd pI'opi)~cd Iflc~h1ni5n~ ut' ~'Iut()i in mu nity m(diat.i'd by SAg iflvi)1ve5 ~elective expansion ofpreexisting autoimmune T-cell clones. In a series of studies to L'h~l'~'~cteri.ze the T-cell repertoire in the joint tissues of individuals with RA, Friedman and Crow observed an increase in T cells bearing V~17, which is the human equivalent of murine V~6 that is used by MAM. It may not be coincidental that the specific V~ chain-l)eanng T cells identified in rheum~'~toid joints arc those that are activated by MAM and are the murine equivalents of those that drive the autoimmune response to type [I collagen in the mi)use model of autoimmune ai~hritis. Thus MAM, MAM-like molecules, or MAM-cross-reactive epitopes may play a role in human RA.
AIDS-Associated Mycoplasmas
Several years ago Shyh-Ching Lo at the Armed Forces Institute of Pathology in Bethesda, Md., reported that a high proportion of AIDS patients harbor M. fer,,ieii tens and M. pe~ietrans. In fact, M. penet-an~, a recently discovered mycoplasma species, was isolated only from AIDS patients. Unlike most mycoplasma species, the isolated organisms grow intracellularly. M. pefletra7Ls also seems to be associated with the development of Kaposi's sarcoma in U.S. human immunodeficiency virus-infected male homosexuals. Since many mycoplasma species are known to p05sess mitogenic membrane constituents, Lo's laboratory has been investigating the mitogenic properties of lipid-associated membrane preparations (L~IPS) of AL [ermentens and M. pei~etrans. According to I-,o's colleague Shaw-Hucy Feng, both mycoplasma preparations contained a B-cell mitogen that activates mouse splenocytes. Unlike the B-cell-activating properties of SAg, the activity of LAMPS is T cell independent. B-cell proliferation is followed by massive secrLtion of polyclonal immunoglobulin M, which could help account for the autoimmune reactions and polyclonal gammopathy which occur in some AIDS patients. In addition, LAMPS are efficient inducers of IL-i, IL-6, TNF-a, and IL-12 in inurine peritoneal macrophage cultures at both the protein and RNA levels. The mycoplasma preparations also act as a costimulator of niti'ic oxide in the presence of IFN-y. Murine splenocytes produce IFN-y and IL-b in the presence of LA~MPS, whereas human peripheral monocytes(macrophages produce high levels of IL-i, IL-6, and TNF-~. Such fiudings suggest that work to investigate the role of mycoplasmas and their products as cofactors in AIDS should be pursued.
Update
J\~~cently Joel B~senian Univc'i'.';itv of Texd's) and othcis hive ti)u ntl that ~1. pIleufl~ol1iaL', A!. geIliI(IliLLnL, and oilier mycoplasmas can also become intracellular. which provided a unique opportunity for resisting host defenses and establishing chronic infections. In addition. mycoplasmas have been detected by PCR and by culture in the joints of patients with inflammatory arthritis but not in controls, according ti) Christianna Beber (Univei'sity of Bordeaux in France) 'md David Taylor-Robinson (St. Mary's Hospital, London, United Kingdom). In my own laboratory, Sawitzke and l~udtson with the help of David Joyner are finding that the sera from patients with RA but not those with lupus erythematosis contain elevated levels of antibodies to MAM compared with the levels seen in healthy individuals. These combined observations should promote a reexamination of the role of mycoplasmas iii chronic and autoimmune disease. El
Acknowledgments I thank Marianne Egan, Jerry Simecka, John Davis, Steven Friedman. and Shaw-ilucy Feng for their contribuiijins and critique of thi~ article. The symposium was sponsored in part by the Respiratory Disease Branch of the National Institutes of Allergy and Infectious Diseases, Bethesda, Md.
Suggested Reading Clinical Infectiou',' l)i.ieases. 1993. The changing role of mycoplasmas in respiratory di.~ease and AIDS. Report of a symposium held at Scottsdale, Ari7.ona. Clin. Infect. Dis. 17(Suppl. 1). Cole. B.C., K. L. Knudtson, A. Oliphant, A. D. Sawitzke, A. Pole. M. Manohar, 1, 5, Benson, E. Ahmed, and C. L. Atkin. 1996. The "'equence of the Mycoplasma arthritidis superantigen MkM: identification of functional domains and comparison with microbial superantigens and plant iectin mitogens. J. Exp. Med. 183:1105-1110. ~ Ci~le, B. C., and A. I). .','awitzke. 1995. Mycoplasmas. superantig('.ns d'nd autoimmune arthritis, p.47-66. In B. Henderson, J. C. W. Edwards. and h. R. Pettipher (ed.), Mechani.~m~ and models in rheumatoid irtiritis. Academic Press, San Diegti, Cd lif. Faulkner. C. B., J. W. Simecka, M. K. Davidson. J. K. Davis, T. k. Schiieb, J. R. Lindsey, and M. P. Everson. 1995. Gene expression and production uf TNF-~, IL-i, IL-6, and IFN-y in mice infected with Mycop1as~na pulmoiiis. Infect. Immun. 63:4138-4142. Herman, A., J. W. Kappler, P. Marrack, and A. M Pullen. 1991. Superantigens: mechanism of T cell stimulation md role in imm une responses. Annu. Rev. Immunol. 9:745-772.
A Manilofi, J., R. N. McElhaney, L. R. Finch, and J. B. Baseman (ed.). 1992. Mycoplasmas: molecular biology and pathogenesi.~. American Society for Microbiology. Washington, D.C. Mobley, H. T., M. D. l~la'nd, and R. P. Hausinger. 1995. Molecular biology of microbia' urea~es. Microbiol. Rev. 59:451-480. Schaitner, E.. K. B. Elkon. J. R. Tumang, M. K. Crow. and S. M. Friedman. 1995. Signaling through the B L'ell antigen Cl)40 induces APO-1 expre.~sion on human B lymphocy'ei and ficiltati's malignant B cell diatli. J. Exp. Med. 182:1557-1565. X.Tsuchiya. N., and Ii. C. Willi-ims. 1992. Melecul-ir miiiiicry-llyp~ithesis or realitv? ~Vest. .1 Med. 157:l.3J'-138. Zagon. C., J. R. Tuniatig. I). N. P()~fl~~tt~ S. M. Fridman, and -'ii. K. Crow. 19q.4. Incri'asi'd lre~uency i)f V(')17.pii.-Itivi' T cells in piL~Ciit with rheitildiuld LrLl'ritis. Artlii.iii~ !~hi.~iiii lO 14:: 1-i 4'Iii
