In the present study, we demonstrate that exposure of primary human monocytes to CSN1S1 in vitro consistently skews cellular differentiation towards macrophages, including morphological changes, distinct surface marker expression, and functional properties such as increased phagocytic activity. Additionally, CSN1S1 induces the expression of proinflammatory cytokines. Besides these functions, the most obvious role of human CSN1S1 is to provide an amino acid source to the new-born. However, the acquisition of additional functionality in an evolutionary context is an increasingly recognized phenomenon, also referred to as protein promiscuity[23, 24]. In accordance with this notion, caseins are considered to have arisen from innate immune genes, and that their nutritive functions are a consequence of a more recent evolutionary development. This assumption is based on the conserved organization of the casein genes in a cluster of innate immune genes that also includes the histatin/statherin-family. The hypothesis that CSN1S1 is a multifunctional protein is further supported by its state as a disordered protein with multiple potential tertiary conformations. This last point has to be regarded with caution however, since – to the best of our knowledge – crystallographic analyses of the human CSN1S1 structures currently do not exist. Since CSN1S1 is not only an endogenous produced protein, but is also a component of milk, the question arises, which potential functions CSN1S1-induced IL-1β expression could have in the offspring. Intestinal exposure to antigens and milk constitutes an important trigger for the development of a competent immune system in the new-born. It is therefore interesting to speculate that CSN1S1 in mother milk may contribute to the development of a patent immune system by triggering immune responses to potential pathogens by activation of innate immune responses like for instance IL-1β secretion. Moreover, CSN1S1 by itself gives rise to sustained specific IgG antibody production in nursed individuals. Early infantile autoantibody production in turn is speculated to confer protection to pathogens. On the other hand, there are several mechanisms which could potentially prevent overwhelming inflammation triggered by exposure to CSN1S1 in milk: CSN1S1 is only a minor component of human milk and constitutes approximately 5% of the casein-fraction. Moreover, CSN1S1 may be degraded by proteases in the healthy gut, thereby preventing IL-1β induction. Further research is clearly warranted to clarify these exciting new hypotheses and to explore, if variations in CSN1S1 exposure or extra-mammary expression may contribute to defective immune reactions. The recent findings of CSN1S1 overexpression in the autoimmune diseases multiple sclerosis and rheumatoid arthritis[9, 10, 12] may be considered supportive of this hypothesis.
In the present experiments, the effect on all aspects of cellular differentiation, i.e. change of morphology, surface marker expression and increased phagocytosis, were observed rapidly, within 24 h of stimulation. Furthermore, CSN1S1 was able to reverse early GM-CSF-induced monocyte differentiation into DC, resulting in a macrophage like phenotype. In vitro differentiation of monocytes towards macrophages or DC is most commonly carried out over 5 days, although more rapid differentiation in the course of several hours is recognized depending on the stimulus used. In accordance with this notion, characteristic differences between in vitro differentiation towards macrophages (using M-CSF/IFNγ) or DC (using GM-CSF/IL-4) were observed after 120, but not 24 h. Of note, surface markers were strikingly similar between M-CSF/IFNγ and CSN1S1 treated cells. However, CSN1S1 failed to reverse in vitro generation of early DC by a combination of GM-CSF and IL-4. This may be due to the more potent effect on in vitro DC generation by the combined cytokines compared to GM-CSF alone[15, 21, 31].
We were consequently interested to explore potential mechanisms employed by CSN1S1 to induce monocyte differentiation and cytokine expression. It was previously reported that primary human monocytes secrete GM-CSF in response to CSN1S1. This was somewhat puzzling, because GM-CSF is known to influence the differentiation of monocytes towards a DC phenotype. On the other hand, according to the present data, CSN1S1 does also increase the secretion of M-CSF into culture supernatants. However, addition of a neutralizing M-CSF antibody to stimulated monocytes did not abrogate CSN1S1-effects. Importantly, there were also no changes in expression of the GM-CSF- or M-CSF-receptors (CD115 or CD116, respectively). Thus, CSN1S1 likely induces its effects on monocyte differentiation by a mechanism independent from M-CSF signalling. Concerning intracellular messengers, CSN1S1, like other proinflammatory cytokines such as IL-32 for example, employs p38 MAPK to induce proinflammatory cytokine expression[13, 19]. Inhibition of another member of the MAPK family, ERK1/2, a well-known regulator of cellular differentiation, but not p38 or JNK led to a decrease in CSN1S1 induced upregulation of CD14 in the present experiments. This effect may be specific for CSN1S1 rather than attributable to the process of differentiation of monocytes towards macrophages in general, because M-CSF induced upregulation of CD14 was inhibited by JNK exclusively. Furthermore, in contrast to differentiation, the secretion of proinflammatory cytokines (i.e. IL-6 and IL-1b) was influenced by the inhibition of JNK and/or p38, but not by ERK1/2. It cannot be excluded that other second messengers are employed for CSN1S1 induced cellular differentiation as well, especially because CD64 was not significantly affected by ERK1/2 inhibition. In conclusion, the data suggest that MAPK may be differentially involved in mediating CSN1S1 induced effects on cellular differentiation or cytokine expression. Further research in this direction is warranted however, before firm conclusions can be drawn.
A limitation to the study consists in the fact that the concentration of CSN1S1 in potentially relevant tissues for monocyte differentiation such as e.g. inflamed nerves, joints, or even the gastrointestinal tract is unknown. In order to simulate physiologic conditions, the concentrations of CSN1S1 used in the present experiments was determined based on previous observations: While in vitro-experiments suggest that ectopic CSN1S1 secretion by monocytes is in the range of ng/ml[9, 13], human milk contains 2.4 mg/ml total casein, approximately 5% of which is made up of CSN1S1. This results in a concentration of 120 μg/ml CSN1S1. Although proteins contained in milk are exposed to proteases within the digestive tract, they may be absorbed in an intact form which is even favoured by immature digestive functions of infants and protease inhibitors within milk. Thus, concentrations used in the present experiments may reflect local conditions in vivo.