XVI. The Pneumocystis carinii major surface glycoprotein (MSG): its potential involvement in the pathophysiology of pneumocystosis


*Corresponding author. Tel.: +45 36323015; Fax: +45 36473340; E-mail: tbenfield@inet.uni2.dk


The major antigens of rat Pneumocystis carinii are 45, 60 and 116 kDa, and those of human P. carinii 20, 40, 66, and 95 kDa [1]. The predominant band on acrylamide gels is a surface antigen, which varies in size (95–140 kDa) depending on the host and method of preparation. This antigen is referred to by a variety of names, including gpA, gp95, gp116, gp120 and MSG. In the following the latter term will be used for this antigen.

MSG is encoded by a family of approximately 100 genes, each of which is responsible for the expression of a different isoform of MSG [2, 3]. Human MSG is 95 kDa, slightly smaller than the corresponding rat antigen [4, 5]. Polyclonal and monoclonal antibodies directed against rat and human P. carinii MSG show that species-specific as well as antigenic determinants are shared and that individual isolates of human and rat P. carinii are antigenically different [4–8]. Biochemical studies indicate that MSG is an acidic glycoprotein composed of a protein core and a carbohydrate portion. N-Linked carbohydrates rich in mannose, glucose, and N-acetylglucosamine residues represent one tenth of the MSG mass [4, 9].

Expression of recombinant MSG has largely been unsuccessful. Rat P. carinii MSG has been recombinantly expressed using the Escherichia coli or baculovirus expression system, but differs in glycosylation from native MSG [10]. The difficulty of expressing human recombinant MSG is believed to rely on critical upstream conserved sequences not yet understood. Native MSG preparations from all species are obtained by solubilization of P. carinii organisms with zymolyase and purification utilizing high performance liquid chromatography [11]. Zymolyase hydrolyzes linear glucose polymers with β-1,3-linkages [12]. Studies of the cyst wall have demonstrated that zymolyase treatment removes the outer wall but leaves the plasma membrane unaffected [13]. MSG yields by solubilization are limited and this is basically related to the inability to reliably culture P. carinii organisms in vitro, and, therefore, lack of material.

Despite numerous problems in P. carinii research related to the inability to culture the organism, several investigators have studied the importance of MSG in host immune response, organism-host attachment, lymphocyte and macrophage recognition.

2Role of MSG in organism attachment

Upon inhalation P. carinii organisms pass the upper respiratory tract and reach the alveoli. Organisms establish infection by closely adhering to the type I pneumocyte without invasion of the epithelium [14]. In the immunocompetent host P. carinii is eliminated by host defense mechanisms, while infection will progress to pneumonia in the immunocompromised host.

Type I pneumocytes do not replicate in vitro and studies on the interaction between this cell type and P. carinii are not possible at present. Several in vitro studies using the human epithelial alveolar cell line A549 (with type II pneumocyte-like properties) have demonstrated that whole P. carinii organisms attach to A549 cells [15, 15, 16, 16]. Binding is mediated by at least two adhesive glycoproteins, fibronectin and vitronectin, that bind to both P. carinii and A549 cells, and, thereby, are presumed to facilitate attachment to alveolar epithelium [16, 17]. Further, MSG has been identified as a fibronectin binding protein required for P. carinii attachment [18]. Using A549 cells, the addition of free MSG to the culture medium or preincubation with polyclonal antibodies to MSG resulted in a decrease in attachment of organisms and decreased specific binding of radiolabelled fibronectin.

Cell surface fibronectin binding receptors of the integrin family have been identified. Pottratz et al. showed that antibodies towards the fibronectin binding αv and α5 integrin subunits significantly inhibited attachment of whole organisms [19]. Further, rat P. carinii organisms and MSG upregulated the expression of α5 integrin subunits on A549 cells, and anti-MSG antibody attenuated α5 integrin expression, suggesting that MSG is involved in upregulation and expression of molecules that mediates organism-cell attachment.

Components of lung surfactant, surfactant protein (SP) A and D, are present in enhanced amounts in the lower respiratory tract during P. carinii pneumonia (PCP) and bind to the organism [20, 21]. Zimmerman et al. found that SP-A binds to P. carinii in a specific and saturable manner, and by ligand immunoblot analysis they demonstrated that part of this binding is to MSG. Although not specifically studied, binding of SP-A may potentiate P. carinii attachment to human cells. O'Riordan et al. addressed this issue for SP-D [21]. They showed that SP-D binds to P. carinii through interactions with MSG, and that this augments binding of P. carinii to alveolar macrophages.

The macrophage is essential for clearance of pathogens, including P. carinii[22]. MSG per se is involved in binding to the macrophage mannose receptor [23]. O'Riordan et al. studied macrophage membrane proteins capable of binding MSG and identified these as the macrophage mannose receptor, and further showed that soluble MSG competitively inhibited P. carinii attachment to alveolar macrophages. This study demonstrated that P. carinii interacts directly with cellular receptors without the participation of other mediators.

In conclusion, MSG mediates attachment of P. carinii to host cells and different soluble host factors may enhance this process. However, it remains to be determined whether binding of surfactant and adhesive proteins to P. carinii is of net benefit to the host or the organism through immune evasion and facilitation of attachment.

3Humoral responses to MSG

Studies of IgG anti-P. carinii antibodies using the Western blot technique found high prevalences (>85%) towards any human P. carinii antigen (predominantly a 40-kDa antigen) among immunocompetent individuals [4, 24]. The seroprevalence of antibodies to human MSG by the same technique varied across the USA from 10 to 60%[24]. By ELISA 50–60% of adults without evidence of immunosuppression have IgG antibodies to human MSG [25]. In the same study seropositivity among children gradually developed with 30% of children aged 1–9 and 56% of children aged 10–19 having IgG antibodies to MSG. Only a minority of immunocompetent persons have IgM antibodies against MSG [24, 25].

Among immunosuppressed HIV-1 infected individuals, 12% of individuals with a single episode of P. carinii pneumonia developed an IgG response to the 116-kDa antigen as determined by Western blot [24]. Using purified human MSG in an ELISA, 66% of individuals with an episode of P. carinii pneumonia had detectable IgG antibodies compared to 34% of individuals without P. carinii pneumonia (P<0.05) [26]. Studies of serial samples from the same population showed that 43% of individuals with verified P. carinii pneumonia mounted an antibody response to the P. carinii infection.

The prevalence of IgM antibodies varies. Using Western blot, 10–20% of HIV-1 infected individuals with P. carinii had IgM antibodies to one or more of the antigens present on P. carinii[24], while only five of 128 HIV-1 infected individuals had IgM antibodies determined by MSG-ELISA [27]. Of these four had P. carinii pneumonia.

Several studies have investigated human antibody responses to P. carinii antigens derived from other species, in particular rat [1]. A recent study suggested that there are biologically significant antigenic differences between species, and although there is cross-reactivity between rat and human P. carinii epitopes, the use of xenoantigens may underestimate the true prevalence of species-specific antibodies [28].

The ability to mount a humoral response towards pathogens is an important host defense mechanism. Gigliotti and Hughes studied passive immunization of ferrets and mice with a monoclonal antibody (Mab) against P. carinii[29]. Animals were immunosuppressed with dexamethasone and administered either a specific Mab (M5E12) or an irrelevant, control IgG Mab throughout immunosuppression. Animals receiving the specific Mab were not protected against infection but had significantly fewer organisms in the lungs than animals receiving control IgG. This suggests that antibodies participate in host protection against infection with P. carinii.

4Cellular responses to MSG

Limited studies of cellular responses to MSG have been published. Theus et al. showed that rat T lymphocytes proliferate in response to rat MSG [30]. Expansion of T lymphocytes was mainly of the CD4+ T lymphocytes. Interestingly, only cells from rats with antibodies to P. carinii antigens responded to MSG, indicating that only animals previously exposed to P. carinii recognized the antigen. They further showed that T lymphocytes secrete significant amounts of the cytokines interleukin (IL) 1, IL-2 and tumor necrosis factor α (TNFα). Recently, the same group extended these investigations to include human T lymphocytes isolated from HIV-1 infected individuals [31]. They found that individuals with fewer than 200 CD4+ T lymphocytes per μl had a significantly lower proliferative response to MSG than did healthy controls. Interferon γ secretion in response to MSG was also significantly less, whereas IL-4 secretion was similar in all groups. Similar to the rat studies, individuals with a previous episode of P. carinii pneumonia had a significant increase in proliferative response compared to individuals who had not had a previous episode.

The role of T lymphocytes in host defense against P. carinii has been studied in the rat model [32]. Adoptive transfer of spleen cells and CD4+ cells isolated from rats and sensitized to MSG resulted in a significant reduction of P. carinii organism burden [32]. Previous exposure to P. carinii of the donor rats resulted in increased protection against P. carinii among recipients. These findings suggests that T lymphocytes participate in the host response to MSG and that P. carinii induces T cell immunity.

Human monocytes respond to human MSG stimulation by secreting IL-8 and TNFα in a dose and time dependent manner [33]. Cytokine secretion correlated to upregulation of mRNA steady-state levels. Competitive inhibition studies with mannan indicated that recognition of MSG by monocytes and secretion of IL-8 and TNFα involved a mannose mediated mechanism. This is in accordance with findings from Ezekowitz et al., who found that the macrophage mannose receptor is critical for uptake of P. carinii by macrophages and that uptake could be blocked by competitive inhibition with MSG [34].

Kandil et al. studied human HIV-1 uninfected and HIV1 infected macrophages [35]. In vitro both cells secreted IL-1β, IL-6 and TNFα in response to rat MSG with no significant difference.

The alveolar cell line A549 recognizes MSG and secretes significant amounts of IL-8 upon stimulation compared to control cultures [36]. Recognition is mannose dependent, although to a lesser extent than for monocytes, indicating that other mechanisms than mannose receptors are involved in this process (Figs. 1–2).

Figure 1.

Potential cellular effects of MSG stimulation. PC: Pneumocystis carinii; MSG: major surface glycoprotein; IL: interleukin; TNFα: tumor necrosis factor α; IFN-γ: interferon γ; Ig: immunoglobulin.

Figure 2.

Dose dependent secretion of interleukin-8 (IL-8) and tumor necrosis factor α (TNF) from human monocytes stimulated with human Pneumocystis carinii. Major surface glycoprotein (MSG) in concentrations ranging from 0.2 to 5 μg ml−1. Adapted from [33].

5MSG induced surfactant changes

Patients with HIV infection who develop P. carinii pneumonia have decreased levels of bronchoalveolar lavage (BAL) phospholipids compared to HIV infected patients without P. carinii pneumonia [37]. The reduced levels of surfactant lipid returned to normal levels with treatment [38]. In vitro isolated alveolar type II pneumocytes from rats with P. carinii pneumonia incorporate less radiolabeled choline into secreted phosphatidylcholine than do type II cells from rats without P. carinii pneumonia [39]. A recent study studied the effect of MSG on phosphatidylcholine secretion from rat type II pneumocytes and showed that MSG inhibited basal and ATP stimulated secretion [10]. Unglycosylated recombinant MSG had a similar effect, suggesting that functional activity is mediated by the protein core of MSG.

6Antigenic variation

Antigenic variation among protozoal and bacterial pathogens (e.g. Plasmodium, Trichomonas and others) is well documented and presumably plays a role in immune evasion. Although variability of a major surface antigen has not been described in the fungi, the identification of multiple genes encoding P. carinii MSG suggests that P. carinii has the potential for antigenic variation [3].

7Perspective and conclusion

Clinically, P. carinii pneumonia is characterized by a superimposed inflammatory reaction. Lung function is further impaired 3–5 days after initiation of antimicrobial therapy, unless adjuvant corticosteroids are administered at the same time [40]. Adjuvant corticosteroid therapy has reduced the overall mortality rate among moderately severe cases by 40%[41]. Several markers of disease severity are associated with outcome, including arterial oxygenation, serum lactate dehydrogenase, BAL neutrophilia, BAL IL-8, and PIIINP (a marker of activation and overall inflammation) [42–48].

Surfactant is an important component of alveolar homeostasis, and altered surfactant secretion from pneumocytes induced by MSG may contribute to impaired oxygenation during pneumonia. Cytokines are pleiotropic peptides with numerous diverse actions required in host defense against pathogens, but secretion of proinflammatory and chemotactic cytokines induced by MSG may have detrimental pulmonary effects through activation and recruitment of inflammatory cells, and some cytokines (e.g. TNFα) may cause direct damage to alveolar epithelium and endothelium per se.

Immunogenically, MSG appears to be a important epitope for the development of cellular and humoral immunity, however, the identification of multiple genes encoding the protein allows potential immune evasion through antigenic variation.

In conclusion, MSG remains an intriguing molecule capable of influencing several biological systems. However, more research is required to delineate the true significance of MSG in the pathophysiology of P. carinii pneumonia.