In the present study, a mouse strain carrying a disrupted gene encoding SP-D was used to investigate the nature of SP-D modulation of body weight and metabolism. The main results derived from the explorations of the SP-D deficient phenotype include hyperphagia with no apparent effect evoked by short-term SP-D administration and no indication for involvement of hypothalamic appetite regulation. Moreover, the SP-D deficient phenotype was characterized by redistribution of body fat, insulin resistance, and increased systemic bacterial endotoxin levels, indicative of propensity for metabolic endotoxemia.
Causality between energy intake and expenditure and the obese phenotype has previously been addressed in various genetically modified strains including the extremely obese genotypes such as the CH4r−/− and the ob/ob mice 
. Energy restricted Spd−/− mice did not have an increased gain of body mass, fat/lean mass ratio, or a decreased metabolic rate. Thus, the obesity seen in the Spd−/− mice is primarily dependent upon an increased energy intake and there was no apparent alteration in gastrointestinal assimilation. Moreover, in young mice, where only small and insignificant differences in body mass had developed, oxygen consumption did not differ significantly. In contrast, hyperphagia seemed to predominate in young mice as increased energy intake was evident in Spd−/− mice from 6–7 weeks of age, but normalized in >18 week old mice.
We chose to examine hypothalamic POMC and AGRP, due to their major function as part of the melanocortin system involving in appetite regulation 
. In our experiments no effect of the short term SP-D treatment on hypothalamic POMC and AGRP gene expression could be measured after peripheral and/or central administration. This type of negative observation is reported previously e.g. in fasted GM-CSF deficient and hyperphagic rats 
. Likewise analysis of c-fos induction after icv or iv treatment did not show any difference between treatment groups, neither within the genotypes receiving different compounds nor between the genotypes treated with the same compound. The lack of effect of short term exogenous SP-D on c-fos expression in hypothalamus, however, leaves open the possibility that the primary action of SP-D on energy balance is located in the extrahypothalamic area. Expression of SP-D in human brain tissue has been reported 
and immunostaining has localized SP-D to human brain endothelium (GL Sorensen and U Holmskov, unpublished) 
. On the basis of the performed experiments, we cannot rule out that alternative routes of administration or prolonged treatment periods with exogenous SP-D would affect the energy intake.
Given that pulmonary derangement per se may be involved in the development of the metabolic syndrome, mouse strains with pulmonary phenotypes may be predisposed for an obese or diabetic phenotype. SP-D deficient mice and Granulocyte Macrophage-Colony Stimulating Factor (GM-CSF) deficient mice may represent such strains. These strains are previously reported with resemblance between emphysematic phenotypes, although the resulting pathology may be evoked through distinct mechanisms 
. The GM-CSF deficient mice also represent a model of mild obesity. GM-CSF, however, is reported to act through receptor mediated CNS signalling 
. The emphysematic Spd−/− phenotype was not quantified in this study, but as the mice neither had altered energy expenditure from possible failure to thrive nor emphysema related cachexia, this aspect of the SP-D deficient phenotype was apparently not involved in the induction of the metabolic phenotype.
Development of the metabolic syndrome in humans is thought promoted by high fat diet. To investigate the effect of fat on the metabolic syndrome in Spd−/− mice, we fed the mice a diet with an increased proportion of fat. The low and high fat diets used in our study were isoenergetically exchanged to evaluate the effect on body mass development without differences in protein, vitamins and minerals to influence the results. Although the body mass of mice on a high fat diet was elevated by 9% compared to mice on a low fat diet the genotype dependent increase in body weight, was not significantly different between the different diets and thus did not determine the major outcomes of the study.
Previously low serum SP-D has been demonstrated to be associated with BMI in human studies 
. In the setting of obesity, the ability of the adipose tissue to produce antiinflammatory mediators such as adiponectin decreases, whereas proinflammatory cytokines such as IL-6, TNF-α, and acute phase reactants, and other innate immune mediators increase 
. SP-D deviates somewhat from this picture and is found to be decreased not only in subjects with obesity, but also type 2 diabetes and is negatively associated with fasting serum glucose. Moreover, SP-D is found positively associated with insulin sensitivity, suggesting that decreased SP-D expression could be behind the association of lung function with impaired insulin actions in type II diabetes 
. In contrast, a recent study has demonstrated that high systemic SP-D is predictive for cardiovascular mortality 
underscoring that the interpretation about causality behind the clinical associations must come from animal studies. The data presented here support that a lack of SP-D induces hyperphagia, and fat deposits, and reduces insulin sensitivity. On the other hand, previous data indicated that SP-D induced atherosclerosis 
. Thus, the complex clinical associations are so far supported by a similarly complex SP-D deficient phenotype.
We further observed that Spd−/− mice had increased subcutaneous fat and decrements in retroperitoneal and gonadal fat fad sizes according to feeding regime. The SP-D dependent redistribution of fat deposits may be particularly interesting in the context of atheromata formation with visceral fat being associated to disease 
. In the present investigations adipocyte number, size, or the fraction of adipose tissue macrophages did not indicate that SP-D deficiency would affect the inflammatory state of the fat tissue. Moreover, no alterations in mean arterial pressure or in endothelial function assessed by the responsiveness towards L-name infusion were found to potentially explain the role of SP-D in atheroma formation.
A possible role for the gut microbiota in obesity and additional aspects of the metabolic syndrome was recently proposed 
. In both humans and mice, the development of obesity was dependent on the relative abundance of specific bacterial phyla in the gut 
, and transfer of the gut microbiota from obese mice to germ-free wild-type mice led to an increase in fat mass in the recipients indicating increased capacity for host energy harvest induced by the microbiome 
. A recently appreciated characteristic of obesity is an aberrant intestinal microbiota composition in obese individuals or animals, which appears to be linked to the obese state itself and susceptible to dietary modulation. However, whether this aberrant microbiota composition plays an etiological role in obesity or is a consequence of the diet in obesity remains to be determined as reviewed by Conterno L., et al., 2011 
Lately the involvement of innate immune receptors was documented. Toll-like receptor 5 (TLR5) is expressed in the gut mucosa and it was demonstrated that TLR5−/− mice exhibited hyperphagia and developed metabolic disturbances, including insulin resistance, and increased adiposity. The changes correlated with changes in the composition of the gut microbiota 
. Likewise, mice with loss-of-function mutation in TLR4 are protected against the development of diet-induced obesity and insulin resistance 
. It is further demonstrated that high fat diets could induce metabolic endotoxemia, which can dysregulate the inflammatory tone and trigger body weight gain and diabetes 
Whereas SP-D is present in the human circulation in relatively high levels 
there is no or very low measurable SP-D in the normal mouse blood, but in this species, age or pulmonary inflammation will induce serum SP-D to measurable levels 
. SP-D is expressed in the gut epithelium 
. SP-D may thus contribute to scavenging or neutralizing systemic endotoxin derived from gut microbiota at these sites. SP-D has been described to bind to a variety of endotoxin subtypes and to interact directly with the TLR4/CD14 complex affecting endotoxin signalling 
. In this context, systemic endotoxin levels were tested in SP-D deficient mice and found significantly increased. A potentially altered composition of gut microbiata in the SP-D deficient mice remains to be explored as a consequence of prolonged SP-D treatment and of obesity per se. Nevertheless, the present data indicate that intestinal or systemic SP-D neutralization of metabolic endotoxemia may contribute to the hyperphagic and metabolic phenotype in the SP-D deficient mice.
In conclusion, we have shown that SP-D modulates energy intake, and that hyperphagia is the primary cause of the obese phenotype of SP-D deficient mice. Moreover, we have found that Spd−/− mice have altered fat distribution, a characteristic not dependent on hyperphagia, that increased insulin resistance and hyperleptinaemia were developed secondary to hyperphagia in Spd−/− mice, and that low-grade endotoxemia may contribute to the metabolic phenotype. The results are in line with association between low serum SP-D and BMI and type II diabetes in clinical stuides. The present experiments did not provide definitive mechanistic insight into the hyperphagic phenotype of the KO mice. However, SP-D mediated clearance or neutralization of endotoxin or SP-D dependent signaling in relation to hormonal control of energy intake may directly drive hyperphagia and other mechanisms of the metabolic syndrome. Further studies are needed in order to clarify the mode of action of SP-D in this context.