Intravenous hMSCs improve myocardial infarction in mice because cells embolized in lung are activated to secrete the anti-inflammatory protein TSG-6

Abstract

Quantitative assays for human DNA and mRNA were used to examine the paradox that intravenously (i.v.) infused human multipotent stromal cells (hMSCs) can enhance tissue repair without significant engraftment. After 2 x 10(6) hMSCs were i.v. infused into mice, most of the cells were trapped as emboli in lung. The cells in lung disappeared with a half-life of about 24 hr, but <1000 cells appeared in six other tissues. The hMSCs in lung upregulated expression of multiple genes, with a large increase in the anti-inflammatory protein TSG-6. After myocardial infarction, i.v. hMSCs, but not hMSCs transduced with TSG-6 siRNA, decreased inflammatory responses, reduced infarct size, and improved cardiac function. I.v. administration of recombinant TSG-6 also reduced inflammatory responses and reduced infarct size. The results suggest that improvements in animal models and patients after i.v. infusions of MSCs are at least in part explained by activation of MSCs to secrete TSG-6.

Figures

Figure 1. Assays for Fate of hMSCs Infused into Mice

(A) Clearance of human Alu sequences from blood after i.v. infusion of about 2 × 106 hMSCs into mice. Values are means ± SD; n = 6. (B) Standard curves for real-time PCR assays of human Alu sequences in seven organs. Values indicate ΔΔCt for primers for mouse/human GAPDH genes and Alu sequences on same samples. (C) Tissue distribution of human Alu sequences 15 min after i.v. infusion of about 2 × 106 hMSCs into mice. Values are means ± SD; n = 6. (D) Standard curves for real-time RT-PCR assays of human mRNA for GAPDH. Values indicate ΔΔCt for primers for mouse/human GAPDH genes and cDNA for human-specific GAPDH on the same samples. (E) Kinetics of hMSCs in lung and six other tissues after i.v. infusion of about 2 × 106 hMSCs. Values are means ± SD; n = 6. (F) Appearance of hMSCs in heart after i.v. infusion of about 1 × 106 hMSCs 1 day after permanent ligation of the left anterior descending coronary artery.

Figure 2. Heat Map of Microarray Assays of Mouse Lungs after I.v. Infusion of hMSCs

About 2 × 106 hMSCs were infused i.v., and lung RNA was recovered 10 hr later for assays on both mouse-specific and human-specific microarrays (Affymetrix, Santa Clara, CA). Data were filtered for cross-hybridization (CV > 0.5 and call > 33%), analyzed with the Microarray Suite 5.0 program, and normalized to a value of 1 and variance of 3 SD (+3, red; 3, blue). Gene ontology categories of genes are indicated. The number of genes with expression differences is indicated in the boxes. (A) Assay on mouse-specific chip. (B) Assay of same RNA on human-specific chip. Symbols: con, lung from control mouse; hMSCs con, sample of hMSCs added to lung from control mouse before extraction of RNA; hMSCs i.v., sample from mouse lung 10 hr after i.v. infusion of hMSCs.

Figure 3. Activation of hMSCs to Express TSG-6

(A) Real-time RT-PCR assays for human-specific mRNAs in lung 10 hr after i.v. infusion of 2 × 106 hMSCs. Values are fold increase over values for cultured hMSCs, normalized by ΔΔCt for hGAPDH. Symbols: hMSCs con, sample of hMSCs added to lung from control mouse before extraction of RNA; hMSCs i.v. 1 and 2, samples from lungs of two mice 10 hr after i.v. infusion of hMSCs. (B) Real-time RT-PCR assays for human TSG-6 in mouse lung. About 2 × 106 hMSCs were infused i.v. into naive mice (IV-nor) or mice at 1 hr after MI (IV-MI), and lungs were recovered 0.25-24 hr after the infusions. Values are ± SD; n = 2 or 3 for normal mice; n = 6 for MI mice. (C) Real-time RT-PCR assays for TSG-6 in hMSCs and human fibroblasts from the same donor incubated in serum-free medium with 10 ng/ml TNF-α for 24 or 48 hr. Results with two passages of the same cells are shown. Values are ± SD; n = 3. (D) ELISA assays for TSG-6 in medium from hMSCs and human fibroblasts incubated in serum-free medium with 10 ng/ml TNF-α for 48 hr. Values are ± SD; n = 3. (E) Real-time RT-PCR assays TSG-6 of control hMSCs (Con), hMSCs treated with transfection reagents only (no siRNA), hMSCs transfected with a scrambled siRNA (scr siRNA), or hMSCs transduced with TSG-6 siRNA (TSG-6 siRNA). Cells were incubated with or without 10 ng/ml TNF-α for 6 hr. Values are ± SD; n = 3. (F) ELISA assays for TSG-6 in medium after incubation of cells with or without TNF-α for 48 hr. Symbols are as in (E). Values are ± SD; n = 3.

Figure 4. Assays of Serum and Heart

(A) Assay for cardiac troponin I in serum 48 hr after MI. Values are ± SD; **p 6 hMSCs infused i.v. 1 hr after MI; scr siRNA, 2 × 106 hMSCs transduced with scrambled siRNA infused i.v. 1 hr after MI; TSG-6 siRNA, 2 × 106 hMSCs transduced with TSG-6 siRNA infused i.v. 1 hr after MI; rhTSG-6, 30 μg rhTSG-6 protein infused i.v. 1 hr and again 24 hr after MI. Values are ± SD; **p < 0.01 with n = 3 mice per group. N.S., not significant. (C) Hearts assayed for pro- and active-matrix MMP9 on a gelatin zymogen gel 48 hr after MI. Image is reversed. Symbols are as in (B). (D and E) Granulocyte and monocyte infiltration in the heart 48 hr after MI. Sections stained with anti-Ly-6G and Ly-6C. Symbols are as in (B), except that 100 μg rhTSG-6 protein was infused i.v. 1 hr and again 24 hr after MI. Magnification ×4. Scale bars, 250 μm. Values are ± SD; n = 3 or 4 for each group. **p

Figure 5. Assays of Infarct Size

Each heart was cut from the apex through the base into over 400 sequential 5 μm sections and stained with Masson Trichrome. Every twentieth section is shown. Additional heart samples are shown in Figure S3. (A–E) Symbols are as in Figure 4B, except that 100 μg rhTSG-6 protein was infused i.v. 1 hr and again 24 hr after MI. (F) Infarct size measurements (%) obtained by midline length measurement from tenth section of the infarct area for a total of 20 sections per heart (Takagawa et al., 2007). Values are ± SD; n = 3 or 4 mice in each group; ***p

Figure 6. Echocardiographic Assays 3 Weeks after MI

(A) Representative M-mode echocardiograms. Symbols are as in Figure 4B. (B) Left ventricular fractional shortening (LVFS) and left ventricular ejection fraction (LVEF) from echocardiographic data. Values are ± SD; n = 5 or 6 for each group; *p