Results from the pilot experiment demonstrated that TNF provided significant regulation of BM-derived MCs. I can also conclude that TNF established a persistent effect by changing the epigenetic marker of CpG3 methylation in BM-MCs. I studied a region around CpG3 for binding sites of transcription factors according to the online HOCOMOCO database. CpG3 appeared to be a part of or in immediate proximity to binding sites of NF-kB transcription factors (Fig. 4).
NF-kB factors are among downstream targets of the TNF signal pathway, but they also can be activated by other signals. TNF and NF-κB signaling induce proinflammatory reactions and are associated with accelerated aging [17, 32, 33]. TNF is used as an inflammatory marker for age-related studies [34]. I suggest that demethylation of CpG3 in response to TNF could serve as a persistent marker of inflammatory stresses endured or sustained by a human.
Since IL1B expression was elevated by more than 100 times in response to TNF in the pilot study, it was taken as one of the factors for studying BM-MCs characteristics in the main experiment, which was part of a broader study attempting to establish criteria of efficiency of BM-MCs samples in prevention of GVHD [22]. Low number of ineffective samples prevented us from a straightforward conclusion and the data were not included in the publication.
Results demonstrating low response of BM-MCs to IL1β and IFNγ were unexpected, since these factors are among the main activators of BM-MCs by the immune system [25, 35, 36]. A response to TNF, which manifested as a substantial increase in IL1B expression on the 4th day, was in line with other studies conducted on analogous or similar cell types [25]. Co-cultivation of BM-MCs with lymphocytes and especially with PHA-activated lymphocytes resulted in upregulation of IL1B expression, which can serve as a marker of BM-MCs activation. It is worth to note, that most reports dealing with BM-MCs recognize their activation by different proinflammatory stimuli as activation of their immunomodulation, and specifically as induction of immunosuppressive rather than pro-inflammatory signals [35, 36]. The presented findings align with a recent report demonstrating that IL1B expression is an indicator of pro-inflammatory BM-MCs [17]. It was also suggested that pathogen-associated molecules might reverse the suppressive effects of MSCs on T cells, thereby restoring efficient T-cell responses to pathogens [37, 38]
Initial attempts to demonstrate correlations between methylation of IL1B promoter and its expression under different conditions did not provide required significance (data not shown). The reason seems to be that high values of demethylation represent semi-homogenic group in general, yet with significant fluctuations. So, I split all samples into two groups according to the estimated level of CpG3 demethylation with a gap intended to account for measurement errors and divide samples into two groups with higher confidence. This approach resulted in detection of significant differences between the two groups. It implied that I found the right perspective to study data and that CpG3 methylation level was indeed related to IL1B expression. Changes in CpG3 methylation under different conditions were not studied, due to the estimation that 4 days could be not enough to shift methylation remarkably.
The results demonstrated higher IL1B expression in group2, if the cells were cultured in standard conditions (untreated). The same difference remained after one and four days of TNF action and after one day of IL1β action. So, I demonstrated that the difference in initial CpG3 methylation marks a phenomenon, which is not just a response to TNF, but also has some inertia or memory. As I changed the perspective and considered IL1B expression under conditions relative to control samples, I could find that group2, in which CpG3 was initially more demethylated, demonstrated significantly lower activation of IL1B expression in response to lymphocytes at days one and four and to PHA-activated lymphocytes at day one. At the same time, I did not see differences in IL1B expression level. This can be because the activation has occurred to a certain level and it has overridden the existing initial differences. The presented facts lead me to conclude that the mechanism of activation of IL1B expression during co-culture with lymphocytes differs from the mechanism of activation by TNF.
The presented results demonstrate a connection between methylation of the CpG in the promoter and the expression for IL1B. It also exemplifies the usage of previously developed procedures to determine CpG methylation [26]. As IL1B expression is used as a marker in clinics [21, 39–41], there is a particular significance of an approach studing DNA associated marks in addition to RNA expression. Since DNA is more stable, used approach to studying DNA methylation simplifies procedures and allows to analyse stored samples with potential RNA degradation [42–44]. Extrapolation of demonstrated connections for other tissues or cell subpopulations requires direct experimental confirmation.
Demonstrated changes of methylation are associated with DNA region presented as one per cell. In contrast with data of RNA or protein levels, where significant changes in a small number of cells can be responsible for changes in a sample value, a value of a CpG methylation is associated with a sample composition on cellular level. I cannot be sure whether the observed methylation differences in the data are due to changes in the methylation level in particular cells or to changes in the numbers of subpopulations with different but fixed methylation. To resolve such a discrepancy, homogeneous cells should be studied, excluding any implicit divergence. Activation or imprinting of epigenetic markers for a prolonged time can occur not only in differentiated BM-MCs, but also in mesenchymal stem cells (MSCs). I cannot discriminate the impact of MSCs in the current study, since other techniques are needed to study this small subpopulation, which comprises 10− 4 – 10− 5 of bone marrow cells [45, 46]. To do this, single-cell level approaches or some other targeted schemes of experiment are required [46–48]. I intentionally call the object of the study as BM-MCs rather than multipotent mesenchymal stromal cells, which are commonly turned into an acronym “MSCs”. I would like to emphasize that “MSC” should be reserved for the true mesenchymal stem cell responsible for maintenance of the stroma of bone marrow and other connective tissues. This will help to put the term MSCs in line with designation of other stem cell (SC) types in the adult mammalian organisms, such as muscle stem cells (MuSCs), fair follicle stem cells (HFSCs), neuronal stem cells (NSCs) and so on. This problem of naming arises from multiple usage of “MSC” both for stem and for non-stem cells, which actually should be distinguished [49, 50], as it has also been stated before [3, 46].