Introduction
Streptococcus pneumoniae is a major human pathogen responsible for respiratory tract infections, septicaemia and meningitis. The pneumococcus is particularly well adapted to colonising the mucosal surfaces of the nasopharynx and the combination of bacterial virulence factors and the manipulation of host tissue components allow the pneumococcus to spread from the nasopharynx to sterile regions of the lower respiratory tract, leading to infections such as pneumonia. In the early stages after infection, natural pulmonary defence mechanisms are required for efficient clearance of the pneumococcus. Recent studies have drawn attention to the important role of lung surfactant protein D (SP-D) as the first line of defence in natural innate immunity to microbial invasion of the respiratory tract, involved in the binding, aggregation, and phagocytic uptake of invading micro-organisms [
1‐
4]. In addition, SP-D has also been shown to be involved in binding to apoptotic polymorphonuclear leukocytes and alveolar macrophages to enhance their clearance by healthy resident macrophages [
5].
SP-D, is a member of the collectin family that also includes mannose binding lectin (MBL), conglutinin, collectin-43 and surfactant protein A (SP-A). It is predominantly found in the respiratory tract, but is also detected at other non-pulmonary mucosal surfaces such as the salivary and lachrymal gland, ovary, uterus, oesophagus, stomach, testes, thyroid, heart and kidney [
4,
6,
7]. In the lung, SP-D is secreted by alveolar type II cells and by non-ciliated Clara cells as dodecamers consisting of four collagenous trimers cross-linked by disulphide bonds, to create a cruciform structure. Each trimer of the molecule consists of three polypeptide chains and each subunit consists of four domains: a short amino acid terminal end, a collagen-like region followed by a short α-helical region and a C-type carbohydrate recognition domain (CRD) responsible for its lectin activity [
1,
2,
8,
9].
A number of pulmonary pathogens, including
Streptococcus pneumoniae, have been reported to be agglutinated by lung surfactant protein D
in vitro [
10‐
13]. In one such study using SP-D knockout mice (SP-D-/-), the
in vivo requirement for SP-D in the early pulmonary clearance and modulation of the inflammatory response to bacterial pathogens was shown. Although increased inflammation, oxidant production and decreased macrophage phagocytosis were associated with SP-D deficiency in the lungs of mice, killing of Gram-negative (
Haemophilus influenzae) and Gram-positive (group B streptococcus) bacteria was unaltered [
14]. In another study, a decrease in viral clearance and an increase in production of inflammatory cytokines were detected in response to viral challenge in SP-D-deficient mice when compared to control mice [
15]. Furthermore, treatment of wild-type mice with native full length SP-D or recombinant SP-D substantially increased their survival rate in mice challenged intranasally with
Aspergillus fumigatus spores [
16] and recombinant SP-D promoted the clearance of fungal spores from the mouse lung (Howard Clark et al., unpublished).
Another study reported that highly multimerised SP-D molecules bound to strains of serotype 4, 19 and 23
S. pneumoniae, causing their agglutination and enhancing their uptake by neutrophils [
17]. More recently, we showed that recombinant human SP-D, expressed in
Escherichia coli, consisting of the head and neck regions of the native molecule, bound to all strains of
S. pneumoniae that were tested, but the extent of binding varied between strains. Full-length native SP-D aggregated pneumococci in a calcium-dependent manner
in vitro, but the aggregation of pneumococci varied not only between strains of the same multilocus sequence type (but different serotypes), but also between strains of the same serotype. Neither recombinant truncated SP-D nor native full-length SP-D enhanced killing of pneumococci by human neutrophils in the absence of serum however [
11].
Given the above findings, we hypothesise that SP-D has an important role to play in the innate immune defence of the upper and lower respiratory tract against pneumococcal infection in vivo, by promoting the agglutination and subsequent clearance of S. pneumoniae. This would prevent the colonisation of the nasopharynx and subsequently limit the spread of pneumococci from the upper to the lower respiratory tract by enhancing clearance via the mucocilliary system, thus allowing enough time for other components of both the innate and adaptive immune system to come into play. In the present study we assessed the in vivo contribution of SP-D to host defence by intranasally infecting SP-D-deficient and sufficient mice with S. pneumoniae. Bacterial growth kinetics in the nasopharynx, trachea, lungs and blood, development of lung pathology and host inflammatory leukocyte infiltration into lungs was compared in both strains of mice following infection.
Discussion
Previous evidence has shown that SP-D interacts with
S. pneumoniae in vitro [
11,
17]. The results of the current study are the first to demonstrate
in vivo, that SP-D has an important role to play in pneumococcal clearance. Pneumococcal colonisation of the upper and lower respiratory tract, and infiltration patterns of leukocytes into the lungs of infected mice were affected by the absence of SP-D. Pulmonary clearance of intranasally administered
S. pneumoniae was significantly reduced in SP-D deficient mice as compared to SP-D sufficient controls. Furthermore, our results clearly demonstrate that lack of SP-D allows persistent pneumococcal colonisation of the nasopharynx and trachea and early onset and increased levels of bacteraemia in colonised mice. Our results also indicate that SP-D influences the accumulation of T cells within the vicinity of inflamed bronchioles, whereby increased levels of T cell infiltration into SP-D deficient lungs was observed. This is the first report to demonstrate
in vivo, that SP-D deficiency leads to increased pneumococcal colonisation of the nasopharynx and trachea, hastens the onset and development of bacteraemia, and affects leukocyte infiltration patterns into infected lungs.
SP-D is synthesised and secreted not only by pulmonary epithelial cells but also by epithelial cells and submucosal glands of the trachea of the normal adult mouse [
20] and has been detected at low concentration (56 ng/ml) in nasopharyngeal washings of normal mice [
21]. Based on our results in the nasopharynx and trachea it is clear that SP-D has a crucial role to play in these sites during pneumococcal infection. Consequently, it is clear therefore that SP-D prevents persistent upper airway colonisation by pneumococci and helps protect against invasion of the lower airways. However, it is also conceivable that lack of SP-D may affect resident leukocyte populations involved in host response or alters host tissue sites as to make them more suitable for pneumococcal adherence and colonisation. We are currently investigating these possibilities.
Our results also indicate that of lack of SP-D contributes to the early onset and increased levels of bacteraemia during pneumococcal pneumonia. It is important to note that SP-D+/+ mice cleared bacteria from their blood by 48 hrs post infection and that the numbers of pneumococci in the blood of both strains of mice reflected their levels in the lung. These results strongly suggest that lung surfactant protein D plays an important role in delaying the appearance of pneumococci in the blood and in limiting their numbers in the bloodstream.
SP-D binds and agglutinates
S. pneumoniae in the presence of calcium and is thought to enhance mucociliary and phagocytic clearance [
11,
17]. In addition, binding of SP-D to lipoteichoic acid and peptidoglycan [
22] may suggest a role for SP-D in the prevention of bacterial colonisation of the alveolar epithelium. Elimination of these SP-D functions could explain the colonisation of the trachea and nasopharynx, the decreased pneumococcal clearance from lungs and the early onset of pneumococcal bacteraemia observed in SP-D deficient mice in our study.
As reported for other strains of mice [
19,
23,
24], pneumococcal infection was coupled with an influx of neutrophils into the lung tissue of both SP-D+/+ and SP-D-/- mice. This is consistent with the data of LeVine and colleagues [
14,
15] who also showed that neutrophil accumulation was similar in the lungs of SP-D-/- and SP-D+/+ mice after
H. influenzae and group B streptococcal infection. In our study, the recruitment of neutrophils in the first 24 hrs post-infection was not affected by the absence of SP-D. However, our results also indicate that the neutrophil response in SP-D deficient mice was not maintained for as long as in wild-type mice. SP-D has been reported as a chemotactic factor for neutrophils
in vitro [
25], and although our data demonstrates that the lack of SP-D does not effect early neutrophil infiltration into lungs, it does clearly affect the longer-term influx of neutrophils as demonstrated by the significant drop in neutrophil infiltration by 48 hrs in SP-D-/- mice. This is not a simple reflection of lung pneumococcal numbers either, as by 24 hrs although there is a significant difference in bacterial CFUs in mice (see figure-
3, SP-D+/+ compared to SP-D-/-), the neutrophil numbers in these mice at 24 hrs is not significantly different. Although a similar accumulation of neutrophils was observed in the lungs of both SP-D+/+ and SP-D-/- mice by 24 h after infection, there were significantly greater numbers of pneumococci in the lungs of SP-D-/- mice at this timepoint. This could have resulted in decreased levels of phagocytosis due to the deficiency in the binding and opsonisation of the pneumococcus due to the lack of SP-D, but also could be due to other factors affecting neutrophil activity. For example, as others and we have previously shown, SP-D deficient mice, despite their healthy appearance, develop progressive alveolar proteinosis and have increased numbers of foamy alveolar macrophages [
5,
18,
26]. Thus, it is possible that the excess lipid in SP-D-/- lungs may inhibit the neutrophil respiratory burst, as previously demonstrated
in vitro [
27].
Together with others we have also previously shown that SP-D deficient mice have a 5- to 10-fold increase in the number of apoptotic and necrotic alveolar macrophages compared to wild-type mice, suggesting a contribution of SP-D to immune homeostasis by recognising and promoting removal of apoptotic cells
in vivo [
28,
29]. It will be of value to assess the clearance of infected apoptotic neutrophils during pneumococcal infection in SP-D deficient and sufficient mice. We are currently in the process of examining this.
Previous studies have also reported that SP-D inhibits T lymphocyte proliferation and local T cell responses
in vitro [
30,
31]. It is therefore noteworthy that we found a heavy infiltration of T lymphocytes in the vicinity of inflamed bronchioles in SP-D deficient lungs at 24 hrs post pneumococcal infection, in contrast to infected SP-D+/+ mice, which exhibited minimal numbers of T cell infiltration. Thus, it appears that SP-D influences T cell infiltration patterns in lungs during pneumococcal infection. It is unclear however, whether SP-D influences T lymphocyte recruitment directly or whether the enhanced T cell infiltration is a consequence of the stimulus of bacteria persisting longer in the respiratory tract of the SP-D deficient mouse. Our previous studies would indicate however, that T cell infiltration is not directly dependant upon pneumococcal numbers as similar colony forming units of pneumococci in lungs and blood of mice can result in totally different T cell infiltration patterns [
32]. In addition, in this study we have shown that significantly different pneumococcal numbers can result in significantly different leukocyte infiltration patterns and vice versa.
It has been suggested that SP-D might provide an important link between innate and adaptive immunity, by modulation of antigen presenting cells and T cell function [
33] whereby SP-D would enhance the uptake of respiratory pathogens in the alveolar space by recruited antigen presenting cells, whilst suppressing T cell activation in the alveolar space in order to prevent an inflammatory cascade that could damage the local lung airspaces and impair gas exchange [
4,
33]. Our findings also support an important anti-inflammatory role for SP-D in pneumoccocal infection
in vivo. Indeed, previous studies have also shown increased pulmonary inflammation, cellular recruitment, oxidant production and decreased macrophage phagocytosis in SP-D deficient mice infected with
Haemophilus influenzae and group B streptococcus. No decrease in bacterial killing in the lungs of these mice were observed in this study [
14], suggesting that other aspects of immunity compensated for the lack of SP-D and cleared the infection effectively. However, after intranasal infection with influenza A virus, SP-D deficient mice showed decreased viral clearance and uptake by alveolar macrophages and increased production of inflammatory cytokines in response to viral challenge [
15]. Additional studies are clearly required to further elucidate the role of SP-D in regulating adaptive immune responses
in vivo.
The potential of truncated recombinant forms of SP-D as a new therapy for infectious and inflammatory diseases has recently been investigated [[
34]-35]. Treatment by intranasal administration of SP-D and a 60-KDa recombinant fragment of human SP-D (rSP-D) had a protective effect in a murine model of fungal infection and allergy caused by
Aspergillus fumigatus [
16]. The survival rate of mice increased to 60 and 80% after treatment with SP-D and rSP-D, respectively [
16]. In addition, intrapulmonary administration of rSP-D reduced the number of apoptotic and necrotic alveolar macrophages and partially corrected lipid accumulation in SP-D-/- mice [
28]. Thus, it would be of a great interest to investigate whether the co-administration of SP-D or truncated rSP-D with
S. pneumoniae would correct the defects observed in SP-D deficient mice during pneumococcal bronchopneumonia. Administration of SP-D at intervals after infection may also indicate at what stages in the disease process, the protein is most heavily involved. We are currently in the process of investigating these questions.
In summary, the absence of lung surfactant protein D increases the persistence of pneumococcal colonisation and infection in the upper and lower respiratory tract, as well as leading to earlier onset and increased levels of bacteraemia. In addition, the pattern of cellular infiltration into the lungs of SP-D-/- mice following pneumococcal infection is different from SP-D+/+ mice, as characterised by shorter-term neutrophil influx and increased levels of T cell infiltration. SP-D clearly has an important function in the early stages of infection as part of the host immune response to pneumococcal invasion and warrants further study.