Alginate micro-encapsulation of mesenchymal stromal cells enhances modulation of the neuro-inflammatory response

Abstract

Background aims: Modulation of inflammation after brain trauma is a key therapeutic goal aimed at limiting the consequences of the subsequent injury cascade. Mesenchymal stromal cells (MSCs) have been demonstrated to dynamically regulate the inflammatory environment in several tissue systems, including the central nervous system. There has been limited success, however, with the use of direct implantation of cells in the brain caused by low viability and engraftment at the injury site. To circumvent this, we encapsulated MSCs in alginate microspheres and evaluated the ability of these encapsulated MSCs to attenuate inflammation in rat organotypic hippocampal slice cultures (OHSC).

Methods: OHSC were administered lipopolysaccharide to induce inflammation and immediately co-cultured with encapsulated or monolayer human MSCs. After 24 h, culture media was assayed for the pro-inflammatory cytokine tumor necrosis factor-alpha (TNF-α) produced by OHSC, as well as MSC-produced trophic mediators.

Results: Encapsulated MSCs reduced TNF-α more effectively than did monolayer MSCs. Additionally, there was a strong correlation between increased prostaglandin E2 (PGE2) and reduction of TNF-α. In contrast to monolayer MSCs, inflammatory signals were not required to stimulate PGE2 production by encapsulated MSCs. Further encapsulation-stimulated changes were revealed in a multiplex panel analyzing 27 MSC-produced cytokines and growth factors, from which additional mediators with strong correlations to TNF-α levels were identified.

Conclusions: These results suggest that alginate encapsulation of MSCs may not only provide an improved delivery vehicle for transplantation but may also enhance MSC therapeutic benefit for treating neuro-inflammation.

Keywords: co-culture; inflammation mediators; mesenchymal stromal cells; organ culture; traumatic brain injury.

Figure 1

Rat TNF-α ELISA of cell culture media supernatant collected after 24 h of LPS stimulation ± MSC treatment in OHSC. Data are normalized to untreated LPS-stimulated OHSC and represented as mean ± SE from five experiments, each with n = 2–3 cultures per condition. Encapsulated MSC treatment significantly reduced TNF-α levels in a dose-dependent manner that was more effective than monolayer MSC treatment. Empty capsule treatment had no significant effect on TNF-α reduction. *P < 0.02, **P < 0.002, ***P < 0.0001 compared with LPS + no treatment; ^#P < 0.01, ^##P < 0.002 compared with treatment with equivalent number of free MSCs.

Figure 2

Total PGE₂ concentration in cell culture media supernatant collected after 24 h of culture ± MSC treatment (1 × 10⁵ MSCs/well, with or without LPS), as measured by EIA. Data are represented as mean ± SE from three experiments, each with n = 2–3 cultures per condition. Encapsulated MSC conditions resulted in a significant increase of total PGE₂ concentration as compared with untreated LPS-stimulated OHSC, regardless of LPS presence. Only monolayer MSC treatment with LPS stimulation resulted in a significant increase of PGE₂ as compared with untreated LPS-stimulated OHSC. Monolayer MSC treatment without LPS stimulation, as well as treatment with empty capsules, produced negligible amounts of PGE₂. *P < 0.05, **P < 0.0005 compared with OHSC + LPS + no MSC treatment.

Figure 3

Correlation between total PGE₂ and rat TNF-α measured in OHSC culture supernatants for MSC treatment conditions (1 × 10⁵ MSCs/well). Linear regression and Pearson’s co-efficient (r) were derived from z-scores of standardized data. There is a significant (P = 0.0008) negative correlation—increasing PGE₂ correlates with decreasing TNF-α—and a clear grouping of treatment conditions: encapsulated MSC treatment clusters at high levels of PGE₂ and low levels of TNF-α, and monolayer MSC treatment clusters at low levels of PGE₂ and higher TNF-α.

Figure 4

(A) Rat TNF-α ELISA of cell culture media supernatant collected from organotypic cultures after 24 h of LPS stimulation ± human PGE₂. Data are normalized to untreated LPS-stimulated OHSC and represented as mean ± SE from three experiments, each with n = 3 cultures per condition. Addition of exogenous human PGE₂ significantly reduced TNF-α levels in a dose-dependent manner. *P < 0.01, **P < 0.005, ***P < 0.0001 compared with LPS + no treatment. ^#P < 0.01, ^##P < 0.001, ^###P < 0.0001 between dose groups. (B) Polynomial curve fit of data presented in A, overlaid with corresponding mean levels of PGE₂ production and TNF-α reduction by monolayer and encapsulated MSCs.

Figure 5

(A) Heat map representation of multiplex (human) secretome data. Secretion by encapsulated MSCs (1 × 10⁵ cells/well) co-cultured with LPS-stimulated OHSC is normalized to secretion by monolayer MSCs (1 × 10⁵ cells/well) co-cultured with LPS-stimulated OHSC. Increased levels of secretion are represented in shades of red and decreased levels in shades of green. (B) Multiplex analysis of cell culture media collected after 24 h of MSC co-culture with LPS-stimulated hippocampal slices. Data are represented as mean ± SE from one experiment, with n = 3 cultures per condition. Of 17 detectable analytes, 10 were identified as exhibiting significantly different levels of secretion by encapsulated MSCs as compared with monolayer MSCs. *P < 0.05, **P < 0.005, ***P < 0.0001.

Figure 6

Correlation between rat TNF-α and MSC-secreted factors measured by means of multiplex bead assay for monolayer and encapsulated MSC treatment conditions. Only analytes that exhibited significant differences between treatment groups (monolayer versus encapsulated MSC) are depicted. Linear regression and Pearson’s coefficient (r) were derived from z-scores of standardized data. Co-efficients of correlation (r) and significance for each analyte can be found in Table II. *P < 0.05.

Figure 7

Total PGE₂ concentration in MSC cell culture media supernatant (1 × 10⁵ cells/well) collected after 24 h ± 1 μg/mL LPS, as measured by EIA. Data are represented as mean ± SE from three experiments, each with n = 2–3 cultures per condition. Encapsulated MSCs produced a significantly greater amount of PGE₂ compared with monolayer MSC, regardless of LPS presence. *P < 0.0001 compared with monolayer MSC counterpart.

Figure 8

Heat-map representation of hierarchical cluster analysis on secretome data of (A) MSC treatment conditions (+OHSC, ±LPS) normalized to baseline monolayer MSC-only condition (no LPS, no co-culture) and (B) monolayer (+LPS) and encapsulated MSC (±LPS) culture alone normalized to baseline monolayer MSC only condition (no LPS). For all conditions, 1 × 10⁵ MSCs/well were used. Increased levels of secretion are represented in shades of red and decreased levels in shades of green.