XV. Pneumocystosis pathophysiology

Although immunodeficiency represents the main etiopathogenetic factor [1], the basic mechanism of Pneumocystis carinii pneumonia (PCP) pathogenesis remains poorly understood. When growing into alveoli, P. carinii interacts with cells and other components; the initial stages of PCP involve parasite attachment to alveolar epithelium, preferentially to type I pneumocytes [2–5], and produce an increase in alveolo-capillary permeability [5, 6]; this last event leads to an influx of plasma proteins into the alveolar space and thereby may alter the composition of epithelial alveolar lining fluid. Pulmonary surfactant, which covers the alveolar surface, plays an important role in the homeostasis of the alveolar environment and in the defense of the lung [7–11]. Since P. carinii comes into close contact with surfactant, it seems logical to suppose, on the one hand, that the parasite could interact with surfactant components and, on the other hand, that pulmonary surfactant may have an impact on the pathophysiology of PCP. Pulmonary surfactant, which is synthesized by alveolar type II cells, is a complex material containing lipids and proteins (Table 1) [7, 8]. Its synthesis is developmentally regulated in fetal lung and can be accelerated by glucocorticoids and other hormones [12]. Lipids are the main components (90%) of pulmonary surfactant [7, 8]. Phospholipids represent 80–90% of the lipids and phosphatidylcholines, which are mainly dipalmitoylphosphatidylcholines, constitute the major part (70–80%) of phospholipids. Cholesterol comprises the largest amount of neutral lipids. In addition to lipids, four surfactant-specific proteins (SP) with different structural properties have been identified [13, 14]: SP-B and SP-C, which represent only 1% of total surfactant, are two extremely hydrophobic proteins; they are involved in keeping surface active properties of the surfactant [14, 15]. SP-A and SP-D are hydrophilic proteins; SP-A is the major surfactant protein constituting about 50% of total proteins. Both proteins are members of the collectin family; they contain an amino-terminal collagen-like domain and a carboxy-terminal lectin domain; this is a globular heavily glycosylated domain that can bind to carbohydrate moieties on the membranes of pathogens. Because of structural similarities between SP-A, the complement component C1q and the mannose binding protein [14], it has been suggested that surfactant proteins could play a role in lung defense mechanisms. Composition of pulmonary surfactant Composition of pulmonary surfactant Surfactant spreads as a monolayer at the air-liquid interface. Its best characterized function is the ability to reduce tension at the alveolar air-liquid interface, thereby preventing alveolar collapse at low lung volumes [7, 8], and maintaining lung permeability [16, 17]. However, several lines of evidence suggest that surfactant may also have ‘non-surfactant’ functions [9–11]. First, surfactant components may have physico-chemical protective effects, by facilitating mucociliary transport [18] and also by inhibiting activity on free radical formation [19]. This property appears to be due, on the one hand, to the presence of unsaturated fatty acyl residues in surfactant phospholipids through a protective effect by oxidation of the double bound and on the other hand to a catalase activity of SP-A allowing the protection against peroxides. Second, surfactant may affect the behavior of the host's immune cells [9–11]. Predominant lipids have antiinflammatory effects; thus, exposure of lymphocytes to natural surfactant results in a suppression of proliferative responses to mitogens, but this effect is not the same for all lipid fractions. The main surfactant lipids, phosphatidylcholine and phosphatidylglycerol, inhibit the proliferation, whereas some of the other lipids either have no effect or, in the case of phosphatidylethanolamine, cholesterol or sphingomyelin, stimulate lymphocyte proliferation. Surfactant lipids may also modulate the expression of proinflammatory cytokines by the alveolar macrophage. They suppress the stimulation of tumor necrosis factor a (TNFα), interleukin (IL)-1 and IL-6 expression, which is observed after macrophage stimulation with endotoxin. Many in vitro effects of lipids are opposed to those exerted by proteins, which increase the degree of stimulation of immune cells. SP-A enhances the proliferative response of lymphocytes to mitogens. Moreover, SP-A produces a significant increase in TNFα, IL-1 and IL-6 production in human peripheral blood mononuclear cells. Similarly, in alveolar macrophages, peripheral blood mononuclear cells and splenocytes from the rat, SP-A causes an increase in TNFα production [10]. Third, surfactant proteins may enhance the uptake and the phagocytosis by the alveolar macrophages of pulmonary pathogens [20–22], Bacillus Calmette-Guerin [23], Aspergillus fumigatus[24] and influenza A virus [25]. In short, the ultimate effect exerted by surfactant depends on the relative amounts of its components. As the relative proportion of SP-A and SP-D to lipids increases, the degree of stimulation increases as well. There is no evidence that hydrophobic proteins play any role in the regulation of host defense function [14]. During PCP, alterations of both quantity and quality of pulmonary surfactant have been reported (Table 2). Surfactant analysis during Pneumocystis carinii pneumonia Surfactant analysis during Pneumocystis carinii pneumonia Several studies have demonstrated surfactant abnormalities in bronchoalveolar lavage (BAL) fluids from AIDS patients with PCP. Phospholipid amounts are decreased and lipid composition is altered [26–29]: a reduction in diacylglycerophospholipid [27], a decrease in phosphatidylcholine percentage and an increase in lysophosphatidylcholine have been reported [26, 28]. These changes might be related to enhanced activity of a phospholipase A2[27]. Furthermore, we have shown that surfactant changes occurred early even when P. carinii was not already detectable in BAL fluid, i.e. prior to evident pneumonia [26]. Pulmonary surfactant proteins are also modified during PCP. Phelps et al. [30] detected elevated levels of SP-A; the SP-A amount was greater in the case of severe PCP. Furthermore, SP-A content in BAL fluid correlates significantly with the number of P. carinii organisms [31]. Recently, Limper et al. have reported similar changes for SP-D [32]. The findings in humans are in agreement with data from experimental PCP in rats [32–35]. However, as corticosteroids cause an increase in surfactant lipids and proteins [36, 37], previous results obtained using corticosteroid-treated hosts may be controversial. Nevertheless, studies of non-corticosteroid immunosuppressed PCP models such as the rabbit, which spontaneously develops benign PCP at weaning [38], or SCID mice intranasally inoculated with the parasite [39], have clearly shown that surfactant changes were linked to P. carinii growth; moreover, these changes occurred early before a noticeable proliferation of the parasites [40]. In spite of these findings, the relationship between P. carinii growth and early modifications of surfactant metabolism remains hypothetical. Surfactant abnormalities observed during PCP are similar to those encountered in several other lung diseases [10]. This raises the question whether the surfactant changes are the consequence of the pneumonia or whether they are partly responsible for the development of the parasite into the alveolar space and, thereby, play a role in the pathogenesis of the disease. When the parasite rate is high, surfactant abnormalities probably result from transudation of plasma proteins into the alveoli. The attachment of P. carinii to alveolar epithelium and its proliferation produce an increase in alveolo-capillary membrane permeability, which has been well documented [5, 6]. However, a low number of parasites can early modify the composition of the alveolar lining fluid [40], and this suggests that the parasite could influence the turnover of surfactant by type II alveolar cells. Indeed, P. carinii organisms may alter the amount and the type of surfactant produced, by inhibiting the synthesis and the secretion of phosphatidylcholine from rat alveolar type II cells [41]. Moreover, the synthesis of surfactant proteins is increased as demonstrated by the enhanced expression of SP-A and SP-D mRNAs [32, 35]. Therefore, the surfactant changes that are observed seem to be partly due to a direct alteration of surfactant metabolism by the parasite. Another main finding is that surfactant proteins may interact specifically both with P. carinii organisms and with alveolar macrophages. SP-A and SP-D function as nonimmune opsonins, several studies have shown that they can bind P. carinii both in vivo and in vitro [42, 43] through its major surface glycoprotein (MSG). This binding is time- and concentration-dependent and is significantly inhibited by glucose and mannose. This is consistent with an interaction between the carbohydrate recognition domain of these collectins and the mannose residues of the major surface protein of P. carinii[42, 43]. SP-A and SP-D also enhance P. carinii attachment to alveolar macrophages in a dose-dependent manner [32, 44], the precise mechanism of this interaction remains indefinite. Recent studies have suggested a possible role for the collagen-like domain of the SP-A molecule interacting with the surface of the macrophage [44]. The in vivo role of pulmonary surfactant in PCP appears unclear: changes in surfactant components, especially increases in protein levels, are not efficient to enhance the clearance of the parasite and to avoid the onset of the pneumonia [30, 31]. Even if attachment is increased, the phagocytosis of P. carinii organisms by the alveolar macrophage is not significantly enhanced by surfactant proteins [43, 45]; the alveolar macrophage, which remains the key for effective clearance of P. carinii, presents a dysfunction partly due to inadequate secretion of cytokines commonly observed in immunosuppressed patients [44]. Nevertheless, there is a good possibility that surfactant changes are physiologically significant because SP-A levels are several times higher than normal prior to infection, while at the same time surfactant lipids are decreased. It is not clear whether surfactant alterations are of benefit for the host or the parasite. For example, P. carinii can directly interact with the alveolar macrophages by binding the mannose receptor on the cell surface [46]. Therefore, the binding of surfactant proteins with the MSG of P. carinii might protect the parasite from its recognition and uptake by the macrophage. Some authors have also suggested that SP-A may enhance adherence of P. carinii to alveolar epithelium, thereby facilitating infection [47]. Changes in surfactant may have a contributory effect on the course of PCP, thus abnormal surfactant produces disturbances in lung compliance and gas exchange and, consequently, may affect the clearance of the parasite from alveoli. In this way, Eijking et al. [48] obtained a neat improvement of hypoxemia in P. carinii-infected rats after surfactant replacement. Moreover, exogenous surfactants may inhibit the parasite growth in vitro in short-term co-cultures [40]; this allows us to suggest that abnormal surfactant might have a permissive effect on P. carinii proliferation in vivo. Although the relationship between P. carinii development and pulmonary surfactant is not fully understood, some lines of evidence suggest that surfactant could play a role in the pathogenesis of pneumocystosis. Further studies should be conducted in animal models without corticosteroids to determine how P. carinii induces early changes in surfactant composition and to assess the real impact of surfactant changes on the parasite growth in vivo. This work was developed in the framework of the BIOMED-1 European Concerted Action on Pneumocystis.