The adhesion of M. pneumoniae onto the respiratory epithelia is a precondition dictating the propagation and pathogenesis of M. pneumoniae (6). In addition to pseudo-stratified columnar ciliated epithelia, M. pneumoniae can also adhere to red blood cells, HeLa cells, fibroblasts, macrophages and tracheal organ cultures in vitro, and can adhere to the surfaces of glass or plastics (7). M. pneumoniae is asymmetric under electron microscopy (8). The cell membranes at one end can extend outside to form a proline-rich top structure, also termed the apical organ, and specifically adhere onto the neuraminic acid receptors on the membranes of target cells.

Adhesion is an intricate process, as the adhesion structure consists of an interactive adhesion network-like system and adhesion auxiliary proteins. Specifically, the 170 kDa P1 protein functions as a key ligand during adhesion (9). Pulse-tracking tag experiments have shown that 1 h following contact of M. pneumoniae with the target cells, the P1 precursor proteins, which are scattered in the cell membranes, rapidly shift to the apical organs, and the leading peptide on their amino terminal is hydrolyzed to mature P1 proteins (10). Due to its sole dependence on the key P1 protein, M. pneumoniae is unable to adhere to host cells, however, it can adhere with the assistance of several collaborative auxiliary proteins, including P30 adhesion factor-related protein A (72 KDa), B (85 KDa) and C (37 KDa), HMW 1–5 polypeptides, P40, P90 and P65; these components jointly constitute a characteristic high-electron-density ‘adhesion protein complex’ (Fig. 2) (11). This complex stabilizes the integrity of the M. pneumoniae apical structure by forming a cytoskeleton, anchoring the P1 protein into the cytoskeleton of the adhesive organs, and allowing the P1 proteins on the adhesion cell organs to adhere.

Marking experiments have shown that, in mutant strains with loss of adhesion auxiliary proteins, the P1 protein is chronically dispersed as a precursor in the cell membranes, however, it cannot aggregate to the apical organs or convert into mature P1 protein (9). Electron microscopy has demonstrated that the adhesion of a M. pneumoniae variant is concentrated in the adherend in the following order: HMW1, HMW3, Pl, P30, P90, P40 and P65, which indicates that these proteins have formed an interrelated adhesion network (12). Specifically, HMW1, HMW2 and HMW3 function as stable adherends and allow other adhesions to locate onto the adherend, and, they are involved in the adhesion onto the respiratory tract epithelia (13). As reported, the M. pneumoniae mutant strains, HMW1 and HMW2, can prevent the P1 protein from correctly locating onto the apical structure, which leads to irregular cell morphology, loss of toxicity and sliding ability, and loss of adhering function (14). P30 does not directly affect the positioning of the P1 protein onto the apical structure, however, it interferes with the binding between P1 and its receptor (15). The loss of P30 or enzymatic cleavage of the carboxyl terminal leads to the complete loss of adhesion function in M. pneumoniae, reduced sliding ability, and marked changes in morphology and structure (16). For example, a bifurcate structure appears in the apical tip, and numerous nucleoid-like substances appear in the cytoplasm. When transposon Tn4001 from the genes of an adhesion auxiliary protein C-mutant was used to transform M. pneumoniae, the mutant strain showed reduced cell adhesion ability. Following the loss of the P41 protein, the adherend in the sliding process of M. pneumoniae was separated from the cell (17). These adhesion auxiliary proteins and adhesion molecules jointly form adhesion protein complexes. The adsorption ability of host cells is decided by the positioning of adhesion proteins and the interaction between the components of the protein complexes.

M. pneumoniae can also utilize the MPN372 protein to combine with lung surfactant protein A (SP-A), pass through the host barrier and permanently adhere to target cells containing the SP-A receptor, including alveolar macrophages, alveolar epithelial type II cells, and other histiocytes inside and outside the lung (18). The pretreatment of M. pneumoniae with low-dose proenzyme reduces the binding between M. pneumoniae and SP-A by 80–90%, however, pretreatment with mannose does not inhibit the binding between M. pneumoniae and SP-A, indicating that M. pneumoniae protein components are involved in this process (19).

The cell membranes of the Mycoplasma genus are more durable, compared with those of other prokaryotes, and the cytoskeletal protein network-like structure functions as a cell walls in terms of maintaining cell integrity. Following M. pneumoniae infection, the lipid bilayer of cell membranes is susceptible to biomembrane fusion, and its structure involves the transcription of specific genes, cytoskeletal changes and changes in the nucleolus (20). Membrane fusion can also cause changes in receptor-identifying sites in the cell membranes, affecting the signal delivery between cells and the production of cellular factors (21).

The small-genome M. pneumoniae does not possess the ability to self-synthesize amino acids, fatty acids, cofactors or vitamins. Instead, following permanent adherence via the adherend to the respiratory tract epithelia, M. pneumoniae spreads microtubules and inserts them into host cells, enabling oxygen consumption, use of glucose, absorption of cholesterol, ingestion of amino acids and consumption of nutrients in host cells, causing injury to the host cells (22,23).

M. pneumoniae is usually regarded as an extracellular parasite, however, certain studies have shown it can also invade and damage cells. Studies have shown that M. pneumoniae can invade A549 lung cancer cells, evidenced by its detection in the cytoplasm and nucleus, and the invasive ability depends on the duration and temperature of infection (24). In cell culture in vitro, M. pneumoniae has been shown to invade non-phagocytes, survive for >6 months and synthesize DNA inside cells (7). When the clinically isolated RYC15989 strain was utilized to infect human Hep-G2 cells and rat N2A cells, intracellular Mycoplasma were observed under laser confocal microscopy, and the intracellular invasion damaging ability of M. pneumoniae was also confirmed (25). In addition, during invasion, certain enzymes inside M. pneumoniae, including hydrolase, nuclease and phosphoprotein phosphatase shift to the host cells. Nuclease degrades DNA in host cells, whereas phosphoprotein phosphatase interferes with the activity of serine/threonine and tyrosine protein kinase (26,27).

Adhesion provides conditions for M. pneumoniae to induce regional cytotoxic effects, and M. pneumoniae can directly induce damage via adhesion, auxiliary proteins, capsular and invasive enzymes. M. pneumoniae also exerts its toxin-like effects through its metabolites, exotoxin and exotoxin-like toxic substances, lipids, lipopolysaccharides and membrane lipoprotein (28). Following the adherence of M. pneumoniae onto the surface of bronchial cells, with the cytoskeleton rearrangement, M. pneumoniae penetrates through the bronchial mucous membranes and releases nuclease and H2O2, which result in swelling, necrosis and a binding of bronchial epithelial cells, slower microvilli movement, structural deformation, and the termination of swinging, thereby inducing the infiltration of lymphocytes, plasma cells and monocytes (22,29). With the lack of superoxide dismutase and catalase in M. pneumoniae, the H2O2 and superoxide groups synthesized by M. pneumoniae, and the endogenous toxic oxygen molecules produced by the host cells, increase the intracellular oxygen pressure in the epithelium, which leads to oxidative stress and subsequent cell death (Fig. 3) (30). The major virulence factors affecting the pathogenesis of M. pneumoniae include the accumulation of H2O2 inside host cells and the effects of superoxides on the ultrastructure of host cells (31). The ions of M. pneumoniae-produced superoxides inhibit the activity and degradation of catalases in the host cells, so that the host cells become more sensitive to the toxic oxygen, resulting in mitochondrial swelling, vascular degeneration, cilia destruction and weakened cilia movement in the epithelium (32,33). M. pneumoniae infection leads to the denaturation of red blood cell hemoglobin, loss of reduced glutathione and cytolysis (34,35).

M. pneumoniae is considered to be incapable of secreting cytotoxin. The N-terminal of the M. pneumoniae-associated pathogenic factor, MPN372, contains ADP-ribose transferase activity and its structure is similar to the S1 subunit of pertussis toxin, which induces extensive vascular degeneration and can cause the death of mammalian cells, thereby inducing chincough-like clinical symptoms; these are termed community acquired respiratory distress syndrome (CARDS) toxin (36). M. pneumoniae CARDS toxin is internalized via clathrin-mediated endocytosis (37), and the CARDS toxin induces pulmonary eosinophilic and lymphocytic inflammation (38,39). Cellular vacuoles induced by M. pneumoniae CARDS toxin originate from Rab9-associated compartments (40).

The M. pneumoniae membrane protein is associated with invasiveness, and its variation directly affects the toxicity of M. pneumoniae. The molecular weight of the M. pneumoniae membrane V-1 antigen can change and is associated with virulence. No toxicity or pathogenesis occurs when its molecular weight is 100–200 kDa, however, toxicity and pathogenesis are observed when its molecular weight is 30 kDa. The gene mutation in V-1 antigen occurs at a 17-amino-acid repetitive sequence at the C terminal of the 94.2 kDa antigen, whereas the 27.4 kDa antigen contains two adjacent, but discontinuous, nine-amino-acid repetitive sequences, and variation is induced by site-specific DNA inversion (63). In addition, the adsorption of RBCs by M. pneumoniae can alter the antigenicity of RBC membranes and induce autoantibody against RBC membrane I antigen, namely the cold agglutinin of RBCs, which induces autoimmune hemolytic anemia (64).

The Mycoplasma-induced viscous polysaccharide capsule, as with other bacteria, is readily formed inside the host, however, it disappears rapidly in vitro, indicating phagocytosis by the host cells. M. pneumoniae readily induces variation in surface membrane antigens, in order to evade attacks from the host immune system. M. pneumoniae may tightly adsorb onto the surface of the host cells, depending on the specific adhesion structure, to avoid phagocytosis prior to exact antibody adjustment (65). The polymorphism of M. pneumoniae adhesion antigens also weakens the effects of specific antibodies (66). The glycerophosphatide on M. pneumoniae cell membranes shares certain antigenic components with the host cells, and thus can also evade the host's immune surveillance. The invasion of intracellular parasitism assists in enabling M. pneumoniae to evade the host's immune clearance and drug effect. Therefore subjecting the patients to chronically infected persons or asymptomatic carriers. Thus, the various immune evasion mechanisms of M. pneumoniae constitute the predominant factor explaining why M. pneumoniae can survive chronically inside the host.

The M. pneumoniae membrane antigen is in antigen mimic of the RBC-membrane I antigen, and shares certain antigenic components with Streptococcus pneumoniae 23 or 32 and M. genitalium (67). As with several plants and bacteria, the membrane glycolipids of M. pneumoniae share a common antigen in the brain and lung tissues, which induce cross reaction. The carboxyl end of the P1 and P30 proteins in the adhesive organs of M. pneumoniae show high levels of homology to the cytoskeletal proteins, fibrinogen, keratin and troponin in eukaryotes (68). Thus, during infection, autoantibodies in the brain, lung, RBC-membrane, lymphocytes and myocardial cells commonly occur, which form immune complexes and magnify the autoimmune response, leading to multisystem immune damage.

M. pneumoniae membranes are full of Mycoplasma lipid-associated membrane proteins. At least three types of functional protein have been identified, including M. pneumoniae N602 (b subfamily of F0F1-ATPase), M. pneumoniae N162 and M. pneumoniae N611 (69). Specifically, the inflammatory capacity of M. pneumoniae N602 is higher (~100-fold), compared with that of M. pneumoniae N161 and M. pneumoniae N162, indicating that M. pneumoniae N602 is a potential superantigen component (70,71).

M. pneumoniae infection can induce immunosuppression in the body and cause maladjustment of T cell subgroups. Experiments have revealed that M. pneumoniae infection causes severe destruction of B cells and T cells (72). At 13–18 weeks in patients infected with M. pneumoniae, the serum level of IgG declines (73). Certain children infected with M. pneumoniae suffer from hypoglobulinemia, decreased chemoattraction in neutrophils, lower reactivity to phytohemagglutin phytolectin and reduced resistance against combined infections with other pathogens, including S. pneumoniae (72). These changes indicate that M. pneumoniae infection may induce immunosuppression.