Hi Dr. Baudry,
Thank you for your question! Just to reiterate, the main factor effects are determined by comparing the average of the metric (in this case, MSC secretion of PGE2) when the factor was present in the MSC activating cocktail to that when it was not present. Two-factor (possiblly synergistic) interactions are determined by comparing the levels the metric for MSC activating cocktails where neither factor was present, only one factor was present, or both factors were present. Any trends deduced from this collection of data may be convoluted with the presence of other factors in the cocktail (i.e. the presence of IFN-b when calculating results for LPS).

To control for this and to confirm the results of the high throughput experiment, we included smaller scale experiments using activation cocktails including only one factor or specific combinations of factors of interest. For your specific question about Figure 5, an experiment was performed using monolayer MSC cultured in medium or medium with LPS, IL-1b, or LPS + IL-1b. PGE2 levels were analyzed in the supernatants. Here we observed, again an apparent synergistic effect of exposure to both factors on MSC PGE2 secretion. The trend toward increased potency of IL-1b compared to LPS was also sustained, although more pronounced in this small scale control experiment. Therefore analysis of interactions using this smaller scale approach may reveal effects that the high throughput screen is not sensitive enough to detect.

Because the goal of this high throughput screen was to identify MSC-activating factors that would induce MSC-mediated modulation of macrophages toward an M2 phenotype (i.e. simultaneous attenuation of macrophage TNF-a and promotion of IL-10), only the effects of LPS and IL-1b were further explored in these smaller scale control experiments. However, factors such as IFN-b and IFN-g did produce interesting effects that may warrant their further investigation using this smaller scale approach.

Hi Dr. Seeliger,
Thank you for your question and interest in the work! For the experiments presented here, the MSCs were in monolayer cultures, not encapsulated in alginate. The alginate-encapsulated MSCs (eMSCs) are microcapsules, about 500 um in diameter, and roll around freely in culture. Monolayer cultures were used to ease the implementation of the high throughput fractional factorial design, as these cells are adherent, which allows for easier handling of conditioned media (i.e. using a multichannel pipette).

In previous work done in our lab, we have observed that the MSC secretome is enhanced when the cells are encapsulated, constitutively and after activation with TNF-? and IFN-? [Barminko et al. (2011)]. Therefore it reasonable to expect that we may see different or enhanced modulation of macrophages by eMSC activated with IL-1? and/or LPS as compared to monolayer MSCs.

To test this, we investigated the effect of IL-1? and/or LPS activation of eMSCs on macrophage modulation, using the same conditioned media approach outlined in Schematic 2. We observed that attenuation of macrophage TNF-? secretion was comparable between monolayer MSCs and eMSCs that were unactivated or activated with LPS. For IL-1 ? activation, however, eMSCs exhibited an enhanced ability to attenuate macrophage TNF-? secretion. Future work will investigate how this difference may be connected to MSC-alginate interactions in the microcapsules.

Hi Dr. Ramesh,
Thank you for your question! We, too, have begun to ponder over the mechanism underlying the results of this analysis. The factors used in this panel are ligands of different cell surface receptors which activate different intracellular signaling pathways in the MSCs. These different pathways have different effects on gene expression, cellular function, and secretion. Herein lies the motivation and logic for this high throughput screen of factors: we hypothesize that the specific activation of one or more of these signaling pathways will result in changes in MSC secretion that will have specific effects on target cells (i.e. modulation of macrophage phenotype).
Interestingly, we found that factors that activate the same intracellular pathways induced similar effects on MSC immunomodulation. LPS and IL-1? activate TLR4 and IL1R, respectively. These receptors share intracellular accessory proteins that stimulate MyD-88-dependent activation of NF-kB [Lindstrom et al. 2005)]. Translocation of this transcription factor to the nucleus results in the upregulated expression of many genes, including cyclooxygenase-2, an enzyme essential in the production of PGE2. This may explain why synergistic effects were seen by simultaneous exposure of MSCs to these factors. IFN-? and IFN-? also had similar effects, seemingly activating the MSCs to upregulate both TNF-? and IL-10 from macrophages. These factors both initiate JAK/STAT signaling pathways, which induces cytokine production and anti-viral activity in immune cells [Kerr et al. (1991)]. It is well established that IFN-? is required for MSC-mediated suppression of T-cell activity [Ren et al. (2008)], but to the best of my knowledge, the effect of interferons on MSC-mediated macrophage modulation has not been explored. Therefore, future experiments will look at further elucidating the mechanisms by which these factors activate MSCs.

Hi Dr. Hager,
Thank you for your very interesting question! It is true that a potential benefit of pre-activation will be to quicken the onset of response once the cells are delivered in vivo, as they will be already primed for action. As of yet, we have not begun to investigate this aspect of the work. In previous studies, however, in which we have delivered MSCs in a rat model of contusive spinal cord injury by lumbar puncture, we began to see therapeutic effects as early as 7 days after injection. Therefore, if pre-activation will ameliorate a delay in onset of action of the MSCs in vivo, we may expect to see therapeutic effects at earlier time points than 7 days after injection.
Related to this, we have begun to investigate how restrictive the window for use is after the MSCs are activated. By this I mean that we have started conducting experiments that look at the dynamics of MSC activation after the activating factor is removed. In these preliminary in vitro studies, we exposed the MSCs to LPS or IL-1b for a time, removed the activating factors, and continued to culture the cells in basal medium. We observed that initially elevated MSC secretion of IL-6 decreases over time after removal of the activator, approaching basal levels. We also observed the IL-1b was a more potent upregulator of MSC IL-6. And, despite the drop-off in secretion, IL-1b-activated MSCs still secreted more IL-6 than unactivated or LPS-activated MSCs, even 48 hours after removal of the activating factor. This suggests that the MSCs have changed in some way after pre-activation. Therefore if MSCs are injected directly after their pre-activation, they may still be in a state of elevated secretion that will allow them to start imparting beneficial effects more or less immediately. Subsequent studies will investigate whether these pre-activated MSCs are more sensitive to secondary stimuli, such as what they may encounter in vivo.

Hi Dr. Yang,
Thanks for your question! In response to your first question, several MSC-secreted factors have been implicated in the promotion of anti-inflammatory macrophage phenotypes. These include PGE2 [Laskin et al. (2011)], which is discussed in the poster, IL1RA [Ortiz et al (2007)], TGF-b [Gong et al. (2012), and TSG6 [Hosoon et al. (2011)]. We plan to probe our MSC conditioned media for these other molecules in an effort to further understand the mechanism underlying activated MSC modulation of macrophages.
Regarding your second question, studies in literature report that MSCs do not secrete TNF-a [van den Berk et al. (2012)]. Some studies, however, have reported MSC secretion of IL-10 [Aggarwal et al. (2005) ]. When we probed our activated MSC conditioned media using ELISA, IL-10 was not detectable. Therefore, all TNF-a and IL-10 measured in the macrophage supernatants were produced by the macrophages themselves.

Hi Myisha,
Thanks for your interest in the project! Yes, it is definitely true that the microenvironment to which cells are exposed is a major determinant of their function. This paradigm applies to multiple facets of my research. First, the panel of factors included in the high throughput screen contains molecules that MSCs may encounter in vivo in inflammatory responses to a bacterial or viral infection, injury, or an autoimmune reaction. This is appropriate due to the fact that one of our metrics of activation involves the modulation of innate immune cells. Secondly, in our encapsulated platform, the cells are also exposed to a microenvironment comprised of alginate crosslinked by divalent ions. Several properties of this capsule microenvironment have the potential to affect MSC function, including material stiffness, changes in cell morphology, diffusion of nutrients and waste, and direct cell-alginate interactions. Because material stiffness has been demonstrated to affect MSC differentiation [Engler et al. (2006)], we have optimized our encapsulation parameters to generate microcapsules with mechanical properties that retain MSCs in their undifferentiated state. We have previously seen, however, that encapsulation of the MSCs in alginate enhances the MSCs secretion of several factors, both constitutively and in response to activating molecules [Barminko et al. (2011)]. Future studies will aim to determine the mechanism underlying this observation.

Hi Dr. Moghe,
I’m glad you enjoyed the video and poster! To answer your first question, as of yet, our analysis of the activated MSC secretome has been restricted to PGE2. However, other MSC-secreted molecules have been implicated in the modulation of macrophage phenotype, including IL1RA [Ortiz et al (2007)], TGF-b [Gong et al. (2012), and TSG6 [Hosoon et al. (2011)]. Additionally, many other MSC secreted molecules are associated with other MSC functions, such as promoting angiogenesis (VEGF, bFGF, Ang-1), modulating other immune cells (IDO, iNOS, HLA-G5), and promoting neurogenesis (BDNF, NGF, HGF) [da Silva Meirelles et al. (2009); Grataitong et al. (2011)]. Therefore the analysis of the activated MSC secretome using our approach will be expanded greatly to further elucidate how MSC activation leads to different MSC responses. This analysis can be expedited using a multiplex immunoassay on activated MSC supernatants.
Regarding your second question, our envisioned approach to utilizing activated MSCs clinically is to expose the cells to an activation factor(s) for a certain amount of time, remove that stimulus, and subsequently administer the pre-activated cells to the patient. As you pointed out, this raises questions regarding the dynamics of MSC activation after the activating factor is removed. In our preliminary studies, in which we have done this in vitro using LPS and IL-1b as the activating factors, we have seen that initially elevated MSC secretion of IL-6 decreases over time after removal of the activator, approaching basal levels. We also observed the IL-1b was a more potent upregulator of MSC IL-6. And, despite the drop-off in secretion, IL-1b-activated MSCs still secreted more IL-6 than unactivated or LPS-activated MSCs, even 48 hours after removal of the activating factor. This suggests that the MSCs have changed in some way after pre-activation. Subsequent studies will investigate whether these pre-activated MSCs are more sensitive to secondary stimuli, such as what they may encounter in vivo.

Thanks Nir!