Regenerative Effects of Locally or Intra-Arterially Administered BMSCs on the Thin Endometrium

Abstract

Stem cell-based therapy plays a pivotal role in the regeneration of damaged endometrium. Previous studies have demonstrated the therapeutic potential of bone marrow mesenchymal stem cells (BMSCs) through diverse administration ways. However, the homing, survival, and differentiation potential of these differently administered BMSCs are poorly defined, and the best route of administration is not well-defined. Herein, we aim to compare the engraftment, retaining time, and therapeutic efficiency of differently administered BMSCs. To achieve this, GFP/Luc-labeled BMSCs administered in two modes were assessed in a thin endometrium rat model: either into the damaged horns directly or through the ipsilateral iliac artery. The retaining time and hemi-quantitative distribution were evaluated by in vivo bioluminescence imaging and immunohistological analysis. Locally administered BMSCs were strongly detected in the abdomen at the first 4 days post treatment but underwent a rapid decrease in luminescent signal afterward and were rarely found 28 days after treatment. In contrast, the retaining time of BMSCs injected through the iliac artery was longer, reflected by more GFP-positive cells detected in the uterine section 28 days post treatment. Differentiation toward endometrial stromal cells was observed. Both routes of administration contributed to the restoration of the damaged endometrium, showing a comparable increase in the endometrial thickness and a decrease in fibrosis. However, more importantly, higher expression of LIF and VEGF, better recruitment, and longer retainment were found in the intra-arterial administration, contributing to the establishment of the optimal administration mode in clinical practice.

Keywords: angiogenesis; fibrosis; regenerative; stem cells; thin endometrium.

FIGURE 1

Study design and experimental schematic. BMSCs were isolated from the femora of rats and were labeled with EGFP and luciferase. BMSCs were injected into thin endometrium models either through the damaged horns directly or through the ipsilateral iliac artery when the external iliac artery was clamped by nontraumatic artery clips for less than 15 min. The rats were imaged by in vivo imaging systems immediately after closing the abdomen to confirm ideal infusion. At 2, 4, and 8 days post treatment, BMSC recruitment was evaluated by IVIS. The uteri were excised 8, 12, and 28 days after therapy and subjected to H and E, Masson’s trichrome, IHC staining, and Western blotting. For each group and each time point of examination, a size of three animals was used.

FIGURE 2

Characterization of rat BMSCs. (A) Bright field image showed the typical fibroblast morphology of purified BMSCs. (B) EGFP expression of transfected BMSCs measured by a fluorescence microscope. (C) Immunophenotype of BMSCs was examined by flow cytometry; CD90 was the positive marker and CD45 was the negative marker. (D) Oil Red O and Alizarin Red S staining were carried out to determine the differentiating abilities of BMSCs.

FIGURE 3

In vivo and in vitro tracking of transplanted BMSCs. The recruitment and retention of BMSCs at the injured horn in two groups ((A): intra-artery, IA; (B): intrauterine, IU) were monitored by IVIS at different time intervals (immediately following transplantation, 2, 4, and 8 days post transplantation); (C) Luminescence signals were quantified and recorded as total flux (p/s). Uteri were collected at 8, 12, and 28 days post treatment; a confocal microscope was used to track transplanted BMSCs in IU (E) and IA (F) groups after staining with anti-EGFP antibody and DAPI. (D) GFP-positive cells were counted under a confocal microscope by taking the mean value of three randomly selected high-power fields (hpf).

FIGURE 4

Precise distribution and differentiation of GFP + BMSCs. The murine uterus tissue consists of endometrial (glandular and luminal) epithelium, stroma, and myometrium. (A) Nearly all the GFP-positive cells were localized in the stroma, but not in the epithelium or myometrium. The distribution of GFP-positive cells is concentrated around the vessels and glands, indicated by the white arrow. g: glands, V: vessels. (B) Some GFP-positive cells were located around the vessels in the surrounding ligaments of uterus in the intra-arterial infusion group. Myo: myometrium; E: endometrium; L: surrounding ligaments. The uterine sections were stained with anti-GFP and anti-vimentin (C) or anti-VEGF (D) GFP⁺ BMSCs co-localized with vimentin marker but not VEGF.

FIGURE 5

Morphological evaluation of the regenerative effect of BMSCs administered differently on the thin endometrium. The endometrium morphologies in the four groups (normal, model, BMSC IU-injection, and BMSC IA-injection) were examined by H&E staining (A) and Masson’s trichrome staining (B). H&E indicates hematoxylin and eosin. Immunohistochemical analysis of the cytokeratin (C) and vimentin (D) expression in the four groups.

FIGURE 6

Representative Western blot results of cytokeratin, vimentin, PCNA, TGF-β1, and VEGF. (A,B) Cytokeratin, vimentin, PCNA and TGF-β1 showed a comparable increase in the two BMSC-infused groups (group IA and IU) (C) Differential expression of integrin β3 in four groups at the same time point. (D) The semi-quantification analysis of the expression of Vimentin, TGF-β1, PCNA, Integrin β3 in four groups, achieve by Image J software. (E) Expression of VEGF in four groups at different time points post modeling. The expression levels analyzed by ImageJ quantification are shown on the right. (NC, normal; M, model; IU, BMSC IU-injection; IA, BMSC IA-injection).