Macrophages have heterogeneous phenotypes and complex functions within both innate and adaptive immune responses . To date, most experimental studies have been performed on BMs, isolated SPMs and PMs . However, differences among macrophages from these particular sources remain unclear. In this study, the features of macrophages from spleen, bone marrow and peritoneal cavity were compared. We found that PMs appear to be more mature than SPMs and BMs, based on their morphology and surface molecular characterizatics. BMs showed the strongest capacity in both proliferation and phagocytosis among the three populations of macrophage; under resting conditions, SPMs maintained high level pro-inflammatory cytokine expression (IL-6, IL-12 and TNF-α), whereas, BMs had high level expression of suppressive cytokines (IL-10 and TGF-β); after LPS activation, SPMs expressed relatively high levels of all those cytokines.
In macrophage studies, macrophage cell lines including J774A.1, RAW264.7, P388D1 and U937 [20, 21] can be used, however, continuous subculture of these cell lines may cause gene loss and impair macrophage immune functions. Therefore, macrophages from bone marrow, spleen and peritoneum in primary culture are more commonly used. To date, macrophage studies have been performed and validated extensively using BMs [22–24], but less so with SPMs and PMs. Unlike macrophages obtained directly from spleen and peritoneum, BMs can be fully differentiated in vitro from macrophage dendritic cell precursors . Although there are many advantages in using BMs in immunological studies, such as their high yield, homogeneity and long lifespan , the features of BM macrophages are not fully characterized. Morphological changes of macrophages from three sources were examined to compare their maturation. Consistent with the previous studies , there are some similarities among SPMs, BMs and PMs with regard to their sphere and deeply stained nuclei, but SPMs and PMs contained much more cytoplasm than BMs, suggesting that BMs may be less mature then SPMs and PMs. When comparing cytoplasm of SPMs with PMs, PMs exhibited a larger size and lysosomal content than SPMs, suggesting that PMs may be more mature than SPMs. In addition to morphological analysis, surface molecular expression could also be used, at least in part, to indicate the maturity of the three populations. A study from Alatery showed that both SPMs and BMs were not fully mature and needed to undergo a further maturation in vitro in culture . Our study detected surface molecular expression that related to macrophage maturation and function. PMs had high level MHC II and CD86 expression, whereas BMs had high level CD115 and GR-1 expression. MHC II and CD86 are expressed highly on fully functional macrophages, which also indicates their maturity [27, 28]. CD115 and Gr1 are usually expressed on precursors of monocytes and macrophages, indicating that the cells are less differentiated and more immature . Therefore, our study showed that PMs appear to be the most mature macrophage, followed by SPMs, then BMs. These differences are likely important considerations in the experimental use of macrophages from different sources.
Following great technical improvements in the in vitro generation of macrophages, they are now considered as candidates for cell therapy [17, 30–32]. Currently, there is a much variation in the preparation of macrophages from different sources for therapeutic use. A recent study of muscle regeneration demonstrated the therapeutic potential of macrophages derived from bone marrow . However, both the experimental and clinical use of regulatory macrophages (M2) for treating central nervous system injury relied on generation of macrophages from peripheral blood. Previously we have demonstrated the therapeutic efficacy of M2 macrophages derived from spleen, but not bone marrow, to resolve inflammation and repair the kidney injury [34–37]. We have shown a similar efficacy of M2 macrophages derived from peritoneum as from spleen (unpublished data). This demonstrates the importance of the origin of macrophages used for treating disease. In this present study, the proliferative, phagocytotic and antigen presenting ability of BMs, SPMs, and PMs were assessed. It was found that BMs exhibited the strongest proliferative capability among the three populations, with SPMs demonstrating slight and PMs no proliferative capability, suggesting that macrophages derived from spleen and peritoneum might be more functionally and phenotypically stable. This observation is consistent with our previous report that M2 macrophage generated from bone marrow rather than spleen showed strong proliferation in vivo and failed to protect against renal disease, apparently due to the loss of function and phenotype of macrophages linked to their proliferation ability . In addition to proliferative ability, phagocytotic capacity of macrophages was assessed. BMs have been shown to maintain the highest capability of phagocytosis [38, 39], which was confirmed in our study and may be an important consideration in regards to their therapeutic efficacy.
T-cell activation and proliferation is associated with many chronic inflammatory diseases, including chronic kidney disease, rheumatoid arthritis and atherosclerosis [17, 40, 41]. Inhibition of T-cell activation is important in effectively suppressing inflammatory responses. A previous study showed that B7-H1 binding to its receptor, PD-1, results in inhibition of antigen-induced T-cell activation . High expression of B7-H1 on PMs suggests PMs might inhibit T cell activation more effectively than SPMs or BMs. Such a property of PMs indicates a greater potential for treating chronic inflammatory diseases.
Although SPMs, BMs and PMs exhibited different levels of expression of molecules involved in antigen presentation, such as MHCII, CD80 and CD86, they showed similar antigen presenting ability. Many PMs are recruited into peritoneal cavity in response to bacterial infection, in greater amount than other cell types [43, 44]. In spleen, several subpopulations of macrophage have been characterized in vivo, including F4/80+ red pulp macrophages, MOMA-1+ marginal metallophilic macrophages, ER-TR9+ marginal zone macrophages and MOMA-2+ white pulp macrophages in mice . F4/80 is prodominantly expressed on red pulp macrophages, but not on others such as marginal metallophilic macrophages, marginal zone macrophages and white pulp macrophages. Therefore, F4/80 stained cells might be less diverse and could be considered as a relative uniform population. However, other subpopulations of splenic macrophages require further study.
Comparison of cytokine expression profile of SPMs, BMs and PMs might contribute to the understanding of their distinct properties and provide a valuable reference for further macrophage related studies. The significantly higher expression of TGF-β and IL-10 by resting BMs in comparison to SPMs and PMs suggests that in vitro generated BMs might be potentially more likely to have a M2 phenotype. M-CSF has been shown to induce differentiation of BMs from bone marrow progenitors  and also to induce human macrophages into a M2 phenotype . Compared to an only 1 day in vitro incubation time of SPMs and PMs, the requirement of 7 days stimulation of bone marrow cells with M-CSF may push them towards M2 differentiation. Combined with high proliferation and phagocytosis ability of BM, thus suggests that BMs might be less mature and phenotypically stable than SPMs and PMs, giving caution to the use of BMs in cell therapy. Alternatively, pro-inflammatory cytokines including IL-6, IL-12 and TNF-α were significantly more highly expressed on SPMs with and without activation than BMs or PMs, which may be relevant to the specific microenvironment of spleen. In spleen, SPMs play an important role in removal of red cells, which may require SPMs to produce abundant cytotoxicity-associated cytokines such as IL-12, TNF-α and IL-6 [47, 48]. Therefore, cytokine expression of BMs, SPMs and PMs reflect their biological function.