**4. Discussion**

Bone and muscle tissues constantly interact with each other by secreting cytokines and other soluble factors [69]. Analysis of cytokine expressions during bone regeneration is very important for developing new therapeutic approaches. We aimed to analyze the different macrophage-specific cytokine expressions during bone regeneration in the sections of the entire femur surrounded by muscle. Before sectioning, the tissues must be fixed and embedded in paraffin. In our case, paraffin embedding was not possible because the muscle significantly slowed down (around 5 weeks) the decalcifying solution (ethylene diamine tetra acetic acid (EDTA)) from penetration and did not allow the use of ISH techniques in these sections (data not shown) due to the mRNA degradation. To avoid this problem, an improved version of the CryoJane tape transfer system was developed in our laboratory to obtain the cryosections of the entire femur surrounded by muscle, which was suitable for ISH techniques [49].

The role of macrophages in tissue development, homeostasis, and wound healing is essential [17,19,21,22,37–39]. Bone regeneration is a very complex process. The continuous interaction between bone-forming cells and macrophages is essential for successful bone healing [1,70,71]. However, their role in bone regeneration remains poorly understood. Subsequently, a deeper understanding of their involvement after an injury is essential to developing new therapeutic strategies for bone repair. A broad spectrum of biomaterials, differing in their structure, porosity, composition, and chemistry, is used as therapeutic strategies for tissue repair. It is known from in vitro studies that macrophage polarization remains biomaterial dependent. Data in the literature sugges<sup>t</sup> that the host response in the in vivo condition is more complex which limits the interpretation of in vivo results compared to in vitro observations [72–74]. Previously, it was speculated that M1 and M2 macrophages subsequently participate in a different stage of bone regeneration [1,75–77]. Three weeks after lesion/surgery, both M1-like and M2-like types of macrophages were detected in the injured bone tissue. Mostly M2-like macrophages were identified, while few but detectable M1-like macrophages were observed. Our results sugges<sup>t</sup> that bone regeneration is a complex process, and in contrast to our expectation, it requires the continuous presence of two types of macrophages. Several publications suggested that macrophages exert their effects on bone regeneration through secreted cytokines [1,8,36,38]. Vi and collaborators (2015) demonstrated that even the conditioned medium of the macrophages was able to activate the "mineral deposition and bone formations" [8].

Macrophage subtypes produce diverse cytokines at different stages of regeneration. Identification of cytokine-expressing cells by immunolabeling is quite complicated because these proteins and peptides are secreted, making it very difficult to determine which cells have produced them. In this work, we demonstrated the possibility to identify macrophages in bone after the detection of cytokines expressed by macrophages with in situ hybridization coupled with in situ HCR detection methods. This tool allows in situ characterization of macrophages during bone regeneration, which was not possible earlier. Fluorescence in situ hybridization (FISH) appeared in the 1980s [78–80]. The advantage of this method compared with in situ DIG was that the expression of several mRNAs could be analyzed in the same section [47,81]. However, it was only suitable for detecting highly expressing mRNA. Unfortunately, the expression of cytokines is usually low. The in situ HCR [49–53] method is a promising new method to study gene expression in bone tissues. This technique allows the detection of multiple mRNAs also in the same section and is suitable for the visualization of mRNA with low expression.

Other groups exhibited that at the bone damage location, M1 macrophages secreted pro-inflammatory cytokines (TNF, IL-1β, and IL6) and induced the recruitment of the mesenchymal stem cells (MSC) and other progenitor cells at the local site. Cytokine expression was investigated in the operated and non-operated tissues by qPCR and in situ DIG methods. Interestingly, very high expression of IL-6 and IL-1β was detected by qPCR, and in situ DIG showed that these two cytokines were localized around the biomaterial. M1-like and M2-like macrophages were present in regenerating bone although

the M2-like macrophages were more abundant. The M2b macrophages are the only M2-like macrophages that co-secrete these two cytokines. We found that the in situ HCR method is a good compromise in macrophage identification and allows in situ determination of the presence of the M2b-like macrophages.
