The introduction of mesenchymal stromal cells induces different immunological responses in the lungs of healthy and M. tuberculosis infected mice

Abstract

Mesenchymal stromal cells (MSC) have strong immunomodulatory properties and therefore can be used to control inflammation and tissue damage. It was suggested recently that MSC injections can be used to treat multi-drug resistant tuberculosis (TB). However, MSC trafficking and immunomodulatory effects of MSC injections during Mycobacterium tuberculosis (Mtb) infection have not been studied. To address this issue we have analyzed MSC distribution in tissues and local immunological effects of MSC injections in Mtb infected and uninfected mice. After intravenous injection, MSC accumulated preferentially in the lungs where they were located as cell aggregates in the alveolar walls. Immunological analysis of MSC effects included detection of activated, IFN-γ and IL-4 producing CD4+ lymphocytes, the frequency analysis of dendritic cells (CD11c+F4/80) and macrophages (CD11c-F4/80+) located in the lungs, the expression of IA/IE and CD11b molecules by these cells, and evaluation of 23 cytokines/chemokines in lung lysates. In the lungs of uninfected mice, MSC transfer markedly increased the percentage of IFN-γ+ CD4+ lymphocytes and dendritic cells, elevated levels of IA/IE expression by dendritic cells and macrophages, augmented local production of type 2 cytokines (IL-4, IL-5, IL-10) and chemokines (CCL2, CCL3, CCL4, CCL5, CXCL1), and downregulated type 1 and hematopoietic cytokines (IL-12p70, IFN-γ, IL-3, IL-6, GM-CSF). Compared to uninfected mice, Mtb infected mice had statistically higher "background" frequency of activated CD69+ and IFN-γ+ CD4+ lymphocytes and dendritic cells, and higher levels of cytokines in the lungs. The injections of MSC to Mtb infected mice did not show statistically significant effects on CD4+ lymphocytes, dendritic cells and macrophages, only slightly shifted cytokine profile, and did not change pathogen load or slow down TB progression. Lung section analysis showed that in Mtb infected mice, MSC could not be found in the proximity of the inflammatory foci. Thus, in healthy recipients, MSC administration dramatically changed T-cell function and cytokine/chemokine milieu in the lungs, most likely, due to capillary blockade. But, during Mtb infection, i.e., in the highly-inflammatory conditions, MSC did not affect T-cell function and the level of inflammation. The findings emphasize the importance of the evaluation of MSC effects locally at the site of their predominant post-injection localization and question MSC usefulness as anti-TB treatment.

Fig 1. Following intravenous injection, MSC accumulate preferentially in the lungs.

MSC were labeled with CFSE and injected i.v. into uninfected and Mtb infected mice. Blood, lung, spleen, liver, lymph node and bone marrow cells were prepared 3 hours (day 0) and 1, 3 and 10–12 days post-transfer. CFSE⁺ cells were identified by flow cytometry. A, The scheme of experiments. B, Identification of CFSE⁺ MSC in the lungs of recipient mice (example of flow cytometry data). C,E, Percent of CFSE⁺ cells out of all cells (frequency) in the lungs (C) or spleen (E). D, F, Percent of CFSE⁺ cells out of all injected cells (proportion) found in the lungs (D) or spleen (F). Summarized data of 7 (days 0 and 1) and 2 (days 3 and 10–12) independent experiments (n = 6–19 per time-point).

Fig 2. MSC stimulate lung CD4⁺ cell activation and IFN-γ production in uninfected mice and slightly suppress cell activation during Mtb infection.

Uninfected mice and mice challenged with Mtb (600 CFU/mouse given aerogenically) were transferred with MSC (two injections with 3 days interval performed on days 46–55 post-challenge) or received injections of PBS. Three days after the last injection lung cell suspensions were analyzed for the expression of the indicated markers and IFN-γ. A, Scheme of experiments. B, Co-expression of CD62L and CD69 by lung CD4⁺ cells (representative example of flow cytometry data). C, D, summarized data showing percentages of CD69⁺ cells out of all CD4⁺ (C) or effector CD62L^-CD4⁺ (D) lymphocytes. E, Identification of IFN-γ producing CD4⁺ cells in the lungs (representative example of flow cytometry data). F, G, Summarized data showing percentages of IFN-γ producing (F) and CD25⁺CD127^low (G) CD4⁺ cells. Data are summarized from 3 (Mtb infected mice) or 4 (uninfected mice) independent experiments (n = 11-19/group).

Fig 3. MSC promote the accumulation of DC and the activation of DC and macrophages in the lungs of uninfected mice.

Uninfected mice and mice challenged with Mtb (600 CFU/mouse given aerogenically) were transferred with MSC as described in the legend to Fig 2 and lung cell suspensions were analyzed by flow cytometry. A, Identification of CD11c⁺F4/80^- (DC) and CD11c^-F4/80⁺ (macrophages) populations by flow cytometry. B, Representative example showing the expression of IA/IE and CD11b by lung DC. C, E, Summarized data showing percentages of DC (C) and macrophages (E). D, F, Summarized data showing the expression of MHC class II by DC (D) and macrophages (F).

Fig 4. MSC effects on cytokine and chemokine levels in the lungs of uninfected and Mtb infected mice transferred with MSC or injected with PBS.

Mice were infected and transferred with MSC as described in the legend to Fig 2. Cytokine and chemokine levels were determined in lung cell homogenates of uninfected (A) and Mtb infected mice (B) using 23-plex assay. Checked bars, mice transferred with MSC; open bars, mice injected with PBS. Only those factors that differed significantly between MSC transferred and control mice are shown (summarized data of 3 independent experiments, n = 8-14/group). For full list of data see S3 Fig.

Fig 5

Localization of MSC in lung tissue of uninfected (A, B) and Mtb infected TB⁺ (C, D) mice, 1 h after the transference. MSC were labeled with PKH26 (red), and the slices were counterstained with ActinGreen 488 (green) and Hoechst 33342 (blue). A, C, An overview showing MSC in the tissue (arrows), epifluorescence images were obtained by automatic stitching of 36 intermediate images (Keyence Image Joint Function). B, D, High magnification confocal images of MSC aggregations (x63, each image is a maximum intensity projection of 5x0.6 μm optical slices). Scale bars: A, C = 200 μm; B, D = 10 μm. Similar MSC localization was observed 1 and 2 days after the transfer.

Fig 6. MSC transfer does not affect the course of Mtb infection in mice.

A-D. Mice were challenged with Mtb (600 CFUs/mouse), transferred with MSC or PBS (twice) and analyzed at 3 days after the last transfer (see legend to Fig 2). A, Representative sections of HE staining of lungs from TB^- and TB⁺ mice injected with MSC (MSC⁺) or PBS (MSC^-, magnification x65) and the inset showing higher-magnification view of the inflammatory focus from Mtb infected mouse. B, Area (percent) of lung tissue affected by inflammation (mean + SE). C, Mtb load in the lungs of TB⁺ mice injected with MSC or PBS. D, Weight dynamics (100%—weight on the day of Mtb challenge; the results of one representative out of 3 independent experiments are shown). E, Mice challenged with Mtb received 4 MSC transfers. Weight dynamics is shown (one experiment). All differences between the groups were insignificant.