Modulation of the secretome of MSCs is an essential step towards achievement of the full therapeutic potential of MSCs. Despite the enhancement in the secretion of cytokines and trophic factors that has been previously described in 3D-MSCs [18, 29, 30], there are only a few studies in vitro that tested this potential in macrophages [20] and cancer cells [21]. The efficacy of 3D-MSCs in vivo has not been fully proven, with just a few studies reporting their pro-angiogenic effects [15, 16], and only a couple of secretome modulation approaches have already been proven to be successful in in-vivo models of cerebral ischaemia [17, 19].

Here, we show that 3D culture modifies the secretome of human MSCs, leading to an enhanced secretion of G-CSF and IL-1Ra, key molecules involved in tissue repair and modulation of inflammation. As we have recently demonstrated that CM from 2D-MSCs exerts potent anti-inflammatory effects on LPS-treated BV2 microglial cell cultures [13], in this present study we hypothesised that an enhanced anti-inflammatory phenotype of 3D-MSCs could exert more potent anti-inflammatory effects on LPS-treated BV2 cells. Specifically, we tested whether the secretion of inflammation markers in BV2 cells could be modified by treating them with 3D-CM or co-culturing with whole MSC spheroids placed in inserts.

Only IL-1-primed spheroid CM resulted in decreased TNF-α secretion from LPS-treated BV2s, indicating the importance of priming treatments for modulating responses to inflammation in other cell types. This effect was lost when the whole spheroids were co-cultured with BV2s, as TNF-α secretion was not reduced and an increase in IL-6 was observed instead, particularly in the presence of IL-1β-primed spheroids.

To explain these results, more cytokines were included in a secretome analysis, and it was revealed that IL-1 priming also increased the secretion of several proteins involved in the inflammatory and immune response and the recruitment of immune cells (including GRO-α, MCSF, CCL22, and CCL7), and that this increase was more marked in 2D cells than in spheroids. This effect is achieved by secreting a wide variety of cytokines [9] that would act in other cells, mostly immune cells [31, 32], but not necessarily microglia (or BV2 cells in this case). In other words, IL-1 priming might be enhancing the ability of MSCs to induce a microenvironment that permits the repair of injured and inflamed tissues. Despite the upregulation of some proteins under both 2D and 3D conditions, the magnitude of the change was smaller in 3D conditions. This attenuated effect of IL-1 priming in 3D-MSCs may be due to the fact that the IL-1 signalling cascade is already upregulated due to the spheroid formation process itself [28], or to the fact that only the cells on the outer surface of the spheroids are exposed to these treatments. Besides, 3D-MSCs not only modified their secretome towards the release of more anti-inflammatory mediators but also induced other changes that may be detrimental in our co-culture model, such as the increase in IL-6 and the decrease in IL-10 and CCL22. We cannot discard here the dual role of certain cytokines such as IL-6 or G-CSF [3335].

We also assessed the secretion of matrix metalloproteinases (MMPs) involved in tissue remodelling and repair, and which are believed to also have a dual role in neurological conditions such as stroke. While they mediate tissue injury in the acute phase of ischaemic stroke, their actions may well be beneficial during the recovery phase [36, 37]. In fact, MMP-9 and MMP-13 are involved in lesion growth [37], and MMP9 is also associated with increases in lesion volume and neurological deficits [38]. Priming of 3D-MSC spheroids leads to increased levels of MMP-13, particularly after treatment with IL-1β, and so may not have a therapeutic benefit if given in the acute phase after stroke. On the other hand, MMPs may have a valuable role in the recovery phase by contributing to neurovascular remodelling [36, 39] and by increasing of the bioavailability of VEGF [40]. Furthermore, TIMP-1 has anti-apoptotic effects and may be neuroprotective when given after stroke [41, 42].

Despite being promising in preclinical and in-vitro preparations, MSC spheroids are no more effective than their 2D counterparts [43, 44], confirmed by the fact that there are few published reports showing positive effects of MSC spheroids.

Nevertheless, MSC spheroids still represent a potential as a therapeutic option since they become less entrapped in the lungs and the capillaries when transplanted, and thus can survive for longer [28, 45], and they offer the chance of obtaining a unique combination of anti-inflammatory and immunomodulatory factors [31]. However, despite presenting all these desirable characteristics, the efficacy of spheroid MSCs still remains unproven. Indeed, treatment with CM (instead of the whole spheroid) may be a more effective treatment, maintaining anti-inflammatory and immune modulatory effects but mitigating the risks associated with a cell transplantation. Future studies should focus on assessing the efficacy of these treatments on in-vivo models of inflammation.

However, it is crucial to continue to increase our knowledge of the basic biology of MSCs, as priming treatments and changes in culture conditions can have great effects on the MSC secretome [13, 31, 45] and may increase their efficacy in the treatment of CNS conditions.