The absence of SP-A in KO mice has been shown to increase susceptibility to infection, compromise bacterial clearance and survival from pneumonia, and adversely affect the response to various other types of lung injury [26, 28, 33, 36–42, 120]. However, virtually nothing is known about the impact of SP-A on the alveolar macrophage proteome under basal or unstimulated conditions. We employed a discovery proteomics approach to investigate, in vivo, the impact of SP-A on global protein expression by the alveolar macrophage as a way of gaining insight into molecules and mechanisms affected by SP-A under basal conditions. We thus studied and compared the intracellular proteome of the alveolar macrophage from WT mice and SP-A KO mice, and SP-A KO mice "rescued" with exogenous SP-A and sacrificed at 6 hr, a time point we have previously shown to be sufficient for the same dose of exogenous SP-A to restore tubular myelin formation in KO mice , and 18 hr after treatment. In the present study under basal conditions we observed: 1) differences in individual protein content between macrophages from KO and WT mice; 2) that a single in vivo treatment of KO mice with SP-A exerted a significant impact on the expression of proteins by alveolar macrophages compared to untreated KO mice; 3) as the time after SP-A treatment of KO mice increased, the proteomic profile more closely approximated that of WT mice; 4) a diverse set of SP-A-dependent functions in the alveolar macrophage, some of which have not been identified in vivo previously; including proteins involved in the regulation of the cytoskeleton; and 5) changes in F-actin and cell size of the alveolar macrophage in response to SP-A.
In previous studies of BAL proteome analysis, SP-A was found, under both resting conditions and after various perturbations (infection, ozone exposure), to have a major impact on BAL protein expression, particularly with respect to proteins involved in host defense or the regulation of redox balance [26, 42]. Proteins secreted by the alveolar macrophage, and by other lung cells (particularly epithelial cells), together with proteins from plasma, are the principal sources of BAL protein. The observations made here indicate that the alveolar macrophage may also be a significant contributor to the observed BAL changes. Furthermore, the observed alveolar macrophage proteome changes in response to SP-A may contribute to the SP-A-dependent differences in the phagocytic function of these cells, which may in turn contribute to differences in disease susceptibility. In fact, it has been shown that both the phagocytic activity of the alveolar macrophages from SP-A KO mice , and the survival of SP-A KO mice were compromised compared to WT mice following infection with K. pneumoniae in the presence or absence of ozone-induced oxidative stress [27, 28]. The present findings, as assessed by the multitude of molecules and pathways affected by SP-A, support the notion that the alveolar macrophage from the KO mice is not biologically "ready" to successfully combat injurious agents, and hence the poorer clinical outcome shown in studies of the SP-A KO mice.
The SP-A dose and time points, as noted above, were based on a study where using the same conditions we were able to restore tubular myelin to the lungs of KO mice . Moreover, based on measurements of SP-A levels in the BAL of lungs of C57BL/6 mice, the dose used was within the physiological range. A similar exogenous SP-A treatment has been reported [3, 122], although a number of other studies have used much higher doses [20, 123–125] that are clearly supraphysiological.
The "restoration" of the global pattern of protein synthesis in KO macrophages to that resembling WT macrophages, when exogenous SP-A was given, was remarkable. In 59 out of the 76 (78%) identified proteins the change seen, compared to KO at 18 hr after exogenous SP-A, was consistent with the differences between KO and WT. In other words, when WT levels were lower than KO, the 18 hr SP-A value was lower than KO and the opposite trend was apparent when WT levels were greater than KO. However, this transition from KO towards WT in response to SP-A treatment was complex, because levels of some proteins (~1/3) increased and others (~2/3) decreased in WT macrophages. This raised the possibility that with different proteins SP-A may influence protein synthesis, metabolism, and/or secretion by macrophages to achieve the intracellular levels appropriate for baseline levels in the macrophage.
Changes in the proteome of the treated versus untreated mice could be the result of several different regulatory mechanisms. SP-A may act directly on the alveolar macrophage, through different receptors to which it is known to bind [13, 126–128], to directly influence the levels of expression of specific proteins. A second potential mechanism would involve indirect effects of SP-A on macrophage protein expression that result from the direct influence of SP-A on other cells (such as the type II cell), causing the synthesis and release of mediators that, in turn, regulate the production of specific proteins by the alveolar macrophages in a paracrine fashion. A third possibility is that SP-A exerts its influence on the alveolar macrophage proteome by regulating the release of regulatory molecules from one set of alveolar macrophages that act on themselves or another set of alveolar macrophages via either autocrine or paracrine mechanisms, respectively. Regardless of whether one or more of these possibilities contribute to the proteome changes observed after SP-A treatment, what is clear is that SP-A, either via its direct or indirect effects, has a major impact on the protein expression pattern of the alveolar macrophage. This impact includes significant changes in more than half of the identified proteins.
An important indirect factor in terms of SP-A regulation of gene expression becomes apparent when we consider one of our previous studies of the BAL proteome of KO mice in which we postulated increased oxidative stress in the absence of SP-A . In the current study, this would likely be manifested by changes in the expression of Nrf2-regulated proteins. Indeed, the Nrf2-mediated oxidative stress response pathway was identified as a significant pathway in the present study. Consistent with this notion are the observations made here where nearly 1/3 of the identified proteins are Nrf2-regulated, and are expressed at higher levels in KO than in WT mice. In the WT and with SP-A treatment, levels of expression of many of these proteins are significantly reduced, suggesting a lessening of oxidative stress. Glucose-6-phosphate dehydrogenase, a regulator of oxidative stress  stands out as an example of this. This enzyme is expressed at significantly higher levels in KO than in WT mice and the 18 hr SP-A treatment reduces it to levels significantly lower than KO baseline.
The most common phenotypic trait that has been examined with the SP-A KO mice is increased susceptibility to pneumonia. The simplest explanation for this phenotype is that host defense against specific pathogens is compromised in the absence of SP-A and its opsonic function. Indeed, replacement of SP-A in these models has afforded some restoration of normal host defense function by apparently increasing pathogen clearance [39, 130]. Additional studies in which exogenous SP-A has been administered to SP-A KO mice have shown a diversity of functions or regulatory mechanisms being restored. These include SP-A effects on the regulation of allergic or inflammatory responses [3, 120, 122, 131, 132] and other host defense molecules [20, 125], as well as the ability of exogenous SP-A to restore tubular myelin, an extracellular form of surfactant . This diversity of effects, along with our BAL proteomic studies [26, 42], indicate that SP-A exerts a pleiotropic effect in lung biology and/or health. The observations from the present study not only provide support for an SP-A pleiotropic effect on lung biology, but identify many protein molecules and pathways affected by the interaction of SP-A with the alveolar macrophage. Together these put forward the notion that the interaction of SP-A with the alveolar macrophage under basal conditions is critical, in terms of preparing and maintaining the alveolar macrophage in a state of "readiness" so that it can potentially successfully combat injurious agents. Furthermore, based on the findings of the present study where SP-A exhibits a broad and varied impact on the alveolar macrophage proteome and previous studies of the BAL proteome where the baseline expression profile of SP-A KO and WT mice differed significantly [26, 42], we postulate that these differences are key to underlying mechanisms that are responsible for the less favorable clinical outcome of the SP-A KO mice with pneumonia [28, 39, 40, 131] or bleomycin-induced lung injury  where SP-A may be dysfunctional due to its being oxidized. Thus, the collective evidence of the role of SP-A in the first line defense of the alveolar surface depicts a diverse array of SP-A-mediated functions that help maintain lung health, or attenuate injury when the lung is challenged with various injurious agents, irritants, pathogens, and toxins present in the external environment. Given the constant exposure of the distal lung surface to all these external harmful agents, it is important that mechanisms be in place to remove these potentially damaging agents and maintain normal lung structure. The alveolar macrophage and SP-A, both important components of innate immunity, appear to play key roles in insuring that this occurs.
SP-A has previously been shown to exert regulatory effects on the production of several macrophage proteins [7, 14, 16–21, 125]. The present study shows for the first time that an acute in vivo treatment with SP-A under unstimulated, baseline conditions affected the levels of significantly more and diverse macrophage proteins than previously known. These include proteins involved in macrophage motility, phagocytosis, neutralization of reactive oxidant species and other toxins, removal of damaged tissue/proteins from the alveolus, and other functions. These effects begin to be evident by as early as 6 hr after exogenous SP-A treatment. By 18 hr after intratracheal SP-A treatment, the alveolar macrophage proteome of the KO mouse becomes strikingly similar to that of WT animals that have been continually and naturally exposed to SP-A. The diversity of changed proteins broadly indicates that SP-A, under basal conditions, plays a major role in alveolar macrophage "health" in terms of its potential ability to move about and its "readiness" for pathogen elimination and tissue repair.
The responses identified exhibited several consistent patterns of protein expression. In one pattern the KO mice and WT mice differed significantly from one another (both increases and decreases) and intermediate responses proceeding in the same direction were observed at the two SP-A replacement time points. This pattern was observed in 17 of the 20 proteins in which KO and WT differed significantly (see Additional File 4). The obvious conclusion from this subset of responses is that lack of continually available SP-A in the KO mice alters protein expression and that acute administration of exogenous SP-A leads toward a restoration of the WT phenotype. Of the total of 76 proteins identified in the present study, 37 were proteins with functions broadly related to actin and the cytoskeleton, and of these 37 proteins, 11 differed significantly between KO and WT and followed the pattern described above where SP-A treatment changes expression levels toward those seen in WT. Many other proteins in the actin-related/cytoskeletal group (n = 21 of 37) exhibited the same trend without achieving statistical significance. In most of the 32 responses made up by these two groups (significant, n = 11 + not significant but with a similar trend, n = 21), the time points after SP-A replacement showed a logical progression from KO to WT, indicating a trend toward restoration of the WT phenotype by exogenous SP-A treatment. The inclusion of so many of the significantly changing proteins (n = 11) in the actin/cytoskeleton-related protein group out of 20 significant changes between WT and KO indicates that SP-A has a profound influence on actin-related/cytoskeletal processes such as macrophage motility and phagocytosis, potentially explaining the susceptibility of the KO mice to injury or infection. In vitro studies have shown that an intact actin cytoskeleton, as assessed by the use of cytochalasin D, an actin depolymerizing agent, is required for certain SP-A-mediated processes  and in extrapulmonary tissues SP-A regulates F-actin filament organization . Although given the importance of actin in cell motility, phagocytosis, and endocytosis, it is not surprising that changes in these proteins do occur, but the extent to which SP-A affects the content of so many actin/cytoskeletal-related proteins in vivo is remarkable. Significant changes in protein content with a similar pattern were also seen in several proteins involved in the inflammatory regulation (n = 4) and protease/chaperone (n = 5) functional groups.
The SP-A-dependent changes in actin-related/cytoskeletal proteins may also be the nexus for the differences in the other functional groups. As mentioned earlier, many of the changing proteins are either known to be regulated by Nrf2 or belong to functional classes known to be Nrf2-regulated . Keap 1, the repressor of Nrf2, has been shown to regulate Nrf2 activity through its binding to the actin cytoskeleton . Similarly, a role for the actin cytoskeleton has been described for NF-κB , a mechanism via which SP-A exerts some of its functions [44, 135, 136]. These two transcriptional regulators in addition to each interacting with the actin cytoskeleton, respond in concert to some stimuli in various cells or tissues and a number of genes regulated by both of them . However, the interaction between NF-κB and Nrf2 is complex and may vary in different systems or under different circumstances .
In addition, a number of proteins (n = 13) differed significantly between KO mice and mice that had received SP-A 18 hr earlier, even though no significant differences were observed between KO and WT mice. For those proteins however, in most cases the trend in the WT usually resembled that of the effect seen 18 hr after SP-A. This subset of responses supports a scenario where a robust acute response to SP-A is observed, with some of the proteins exhibiting an overcompensation of expression. For example, a number of protein spots (Figure 5) are increased at 6 hr but decreased at 18 hr after treatment, as in the WT, although they may not have not reached significance yet, indicating that some protein changes need a longer exposure to SP-A and/or multiple doses of SP-A to reach the WT phenotype. These responses included 6 proteins from the actin group and 3 from the protease/chaperone group, indicating that the acute SP-A treatment (like the continuous presence of SP-A in WT) has a significant potential to contribute to alterations in macrophage function.
Although statistical techniques such as gene set enrichment analysis or hypergeometric distribution calculations have been used with microarray data to determine whether changes in subsets of genes differ significantly from a reference group or the overall population in a specific group, these cannot be used here. This is because of the relatively small data set examined by gel-based proteomics and the comparatively large numbers of proteins in our subsets that make up a substantial percentage of our overall population or reference group. For example 50% of the identified proteins in our reference group were in our actin-related group and 28% were included in the Nrf2-related proteins, thereby violating the assumption that the analyzed groups be independent of one another and precluding the use of such methods in this study [139, 140].
The large number of actin-related proteins with altered levels of expression after SP-A treatment led us to speculate that we would be able to validate the proteomic findings with a preliminary examination of parameters related to actin cytoskeletal dynamics. In fact, macrophages from SP-A KO mice treated with SP-A in vitro were significantly larger than comparable control cells and had lower amounts of F-actin, both at the 6 hr time point used for this proteomic study (although total gamma-actin levels were unchanged at this time point) and at 1 hr, a time point where we have previously demonstrated an SP-A-dependent enhancement of phagocytosis by alveolar macrophages . As SP-A stimulation increased cell size, redistribution of the cortical actin over a larger area likely contributes to the decrease in the actin intensity per pixel. We speculate that decreases in F-actin reflect cells that are more actively remodeling their cytoskeletons, resulting in cells of increased size. The larger size may indicate the availability of adequate amounts of excess membrane for phagocytic or endocytic events to occur and the changes in F-actin distribution likely reflect more dynamic actin filaments with altered rates of actin polymerization/depolymerization and branching. Thus, in the absence of SP-A, macrophages appear to be small and "stiff", limiting their motility and phagocytic ability. However, when SP-A is made available, in as little as one hour a more dynamic actin cytoskeleton allows for enhanced motility and ability to successfully locate and clear particles and pathogens. Our previously published observations on in vivo comparisons of SP-A KO and WT mice in a model of pneumonia induced by Klebsiella pneumoniae
 and on phagocytosis by alveolar macrophages in vitro support this scenario .
This study was designed to allow us to study restoration of the WT phenotype from KO alveolar macrophages after giving a single dose of SP-A and studying its effect 6 hr and 18 hr later. Remarkably, the dose and sampling times chosen restored levels of many of the proteins we studied. It is likely that additional doses and time points will further restore the phenotype by regulating the expression of additional proteins with different kinetics of translation and secretion. It is also highly likely that SP-A regulation by continually present SP-A (as in the WT state) or by multiple doses of SP-A will differ somewhat from the pattern resulting from a single acute dose, because in those cases a steady-state equilibrium of the SP-A effects on the alveolar macrophage proteome may be attained. Moreover, we had previously proposed that SP-A "primes" the alveolar macrophage  to prepare it to effectively respond to infectious or toxic agents. The proteomic data presented here support a role for SP-A under baseline conditions to drive an ill-prepared "dysfunctional" alveolar macrophage from an SP-A KO mouse towards readiness and effective host defense as afforded by WT alveolar macrophages. Moreover, these data demonstrate the potential for the therapeutic use of a single dose of exogenous SP-A to counter challenges due to injury or infection.