HGF Expressing Stem Cells in Usual Interstitial Pneumonia Originate from the Bone Marrow and Are Antifibrotic

Abstract

Background: Pulmonary fibrosis may result from abnormal alveolar wound repair after injury. Hepatocyte growth factor (HGF) improves alveolar epithelial wound repair in the lung. Stem cells were shown to play a major role in lung injury, repair and fibrosis. We studied the presence, origin and antifibrotic properties of HGF-expressing stem cells in usual interstitial pneumonia.

Methods: Immunohistochemistry was performed in lung tissue sections and primary alveolar epithelial cells obtained from patients with usual interstitial pneumonia (UIP, n = 7). Bone marrow derived stromal cells (BMSC) from adult male rats were transfected with HGF, instilled intratracheally into bleomycin injured rat lungs and analyzed 7 and 14 days later.

Results: In UIP, HGF was expressed in specific cells mainly located in fibrotic areas close to the hyperplastic alveolar epithelium. HGF-positive cells showed strong co-staining for the mesenchymal stem cell markers CD44, CD29, CD105 and CD90, indicating stem cell origin. HGF-positive cells also co-stained for CXCR4 (HGF+/CXCR4+) indicating that they originate from the bone marrow. The stem cell characteristics were confirmed in HGF secreting cells isolated from UIP lung biopsies. In vivo experiments showed that HGF-expressing BMSC attenuated bleomycin induced pulmonary fibrosis in the rat, indicating a beneficial role of bone marrow derived, HGF secreting stem cells in lung fibrosis.

Conclusions: HGF-positive stem cells are present in human fibrotic lung tissue (UIP) and originate from the bone marrow. Since HGF-transfected BMSC reduce bleomycin induced lung fibrosis in the bleomycin lung injury and fibrosis model, we assume that HGF-expressing, bone-marrow derived stem cells in UIP have antifibrotic properties.

Figure 1. HGF-positive cells are present in the fibrotic lung.

Single mononuclear HGF-positive cells were detected in the broadened alveolar septum of UIP, many of them in close proximity of alveolar epithelium (a) Furthermore, alveolar epithelial cells stained positive for c-met, the receptor for HGF (b). magnification (×20) inset higher magnification (×40).

Figure 2. HGF-positive cells do not stain for markers of fibroblasts, dendritic cells, macrophages, myofibroblasts or neutrophils.

To further characterize the HGF-positive cells in UIP, the lung tissues were co-stained with markers of the alveolar epithelium, fibroblasts, myofibroblasts, dendritic cells and macrophages using specific antibodies. a) Co staining revealed that the HGF-positive cells (green) did not stain for vimentin (fibroblast marker (red)). (magnification ×20 (inset higher magnification ×40), Figure 2a) b) Co staining with SP-C (pink) shows that the HGF signal (brown) is not present in alveolar epithelial cells (Figure 2b). c) Moreover, no co staining was seen either with CD 68 (macrophage marker) (pink); (Figure 2c) or with DC sign (dendritic cell marker) (pink)(Figure 2d). (magnification ×20). Also the co-stainings with alpha smooth muscle actin to detect myofibroblast (pink) (Figure 2e). (magnification×20 (inset ×100), and anti-myeloperoxidase (pink) to detect neutrophils (Figure 2f). (magnification ×20 (inset ×100), did not show any co staining's. (In Figure 2b-2f HGF is stained brown) (Arrows indicating positive cells).

Figure 3. HGF-positive cells stain positive for mesenchymal stromal cell markers.

To identify the HGF-positive cells in UIP, the sections were further stained with the mesenchymal stromal cell markers CD44 and CD105. Cells positive for CD 44 (Figure 3a) were seen in varying numbers and homogenously distributed in the tissue section. The cells positive for CD 44 (red) co stained for HGF (brown) (Figure 3b) (magnification ×40). More information with different markers in the supplement. To further localize these cells in relation to alveolar epithelial cells, the triple immunofluroscence staining with HGF (green) (Figure 3c), SPc (red) (Figure 3d) and CD 105 (purple) (Figure 3e) was performed. The cells co expressing CD 105 and HGF, were seen in close proximity to the SP-C positive cells (Figure 3f), (magnification ×40 oil) for better visualization a 3D image rendering was done which clearly demonstrates the co expression of CD105 and HGF and also the close proximity of HGF positive mesenchymal cell to the SPc positive cells (3 g).

Figure 4. HGF positive cells also stain for CXCR4.

Triple immunoflurosence was performed to determine if the CXCR4-positive cells have any correlation with the HGF positive mesenchymal stromal cells, revealed that the cells positive for HGF and CD 105 were also positive for CXCR4. HGF (green),CD 105 (purple) and CXCR4 (red). (Figure 4a), (magnification ×60 oil) the right panel is the 3D construction after image rendering for clarity of the image (Figure 4b), enlarged image for a clear view the colored arrows indicate different staining in the same cell.

Figure 5. Similar staining pattern was observed in cells isolated from lung biopsies.

To further confirm the presence of HGF-secreting pluripotent stem cells of mesenchymal origin, lung biopsies obtained from patients with UIP were grown in culture and the outgrowing cells were stained with markers as in the formalin fixed tissues. The isolated cells stained positive for HGF (red) (Figure 5a), similar as seen in the tissue sections. To further confirm the origin of these isolated cells, co-staining with HGF (red) and CXCR4 (purple) was performed and the similar pattern of costaining was observed (Figure 5b). Furthermore, a triple immunofluroscence staining with HGF (green), CD105 (purple) and CXCR4 (red) was also observed (Figure 5c) in the cells derived from lung biopsies of IPF patients. 3D reconstruction for better visualization (Figure 5d). Moreover, HGF (green) and CXCR4 (red) positive cells also co stained for Oct4 (purple) (Figure 5e), 3D reconstruction for better visualization (Figure 5f) (magnificationx40 oil).

Figure 6. HGF modified BMSC instilled in the bleomycin induced rat lung were seen in the lung parenchyma.

The HGF-positive cells were distinctly visible in the rat lung parenchyma by immunohistochemistry (brown) (Figure 6a) (magnification ×60), triple immunofluroscence showing BMSC (DiO labeled, green), SP-C (red), HGF (pale yellow) (6b), 3D image for better visualization (6c). (magnification ×60 oil)

Figure 7. HGF modified BMSC attenuate bleomycin induced fibrosis in rat lung.

To further study the possible therapeutic effect of HGF-modified BMSC, the cells were administered intratracheally into bleomycin injured lungs of the adult male Fischer (F344) rats. Seven days after instillation of HGF modified BMSC the lungs were assessed and showed marked improvement in the histological pattern at day 7 (Figure 7c), compared to control animals (Figure 7a) and the group treated with BMSC only (Figure 7b) furthermore when the animals were sacrificed 14 days after HGF modified BMSC instillation there was slight improvement (Figure 7f) compared to control (Figure 7d) and BMSC only group (Figure 7e) in the histological grading which was confirmed with improved Aschroft score (Figure 7h) and reduced collagen content as measured by hydroxyproline assay (Figure 7 g).

Figure 8. Increased number of open alveoli and reduced volume of destructed lung parenchyma was observed by unbiased stereological analysis.

Stereological data of total numbers of alveoli per lung (N(alv,lung)) (Figure 8a) and the total volume of destructed lung parenchyma per lung (V(des,lung)) (Figure 8b). A negative correlation between the number of alveoli and the volume of destructed lung parenchyma can be established (Figure 8c) p<0.01, r = −0.59.