Following tissue injury, fibroblast cells activate, divide and play key roles in both tissue repair and pathological scarring – fibrosis – that can drive organ failure.
Vanderbilt investigators have now discovered that, in contrast to prevailing dogma, fibroblasts are not all alike; instead, they have distinctive functions following tissue injury.
“Our work offers a new perspective over the currently held thinking that fibroblasts are a single population of cells working in the same manner to coordinate wound repair,” said Pampee Young, MD, Ph.D., adjunct professor of Pathology, Microbiology and Immunology.
The findings, reported in Nature Communications, suggest that it might be possible to prevent the pathological scarring effects of fibroblasts without impairing the functions that are necessary for wound healing.
Young, who is the Chief Medical Officer of the American Red Cross, and Sarika Saraswati, Ph.D., research assistant professor of Pathology, Microbiology and Immunology, have focused on understanding how organs repair themselves after injury.
They knew that although activated fibroblasts appear to be heterogenous, most studies of functional roles have treated the cells as a homogenous entity.
“Our goal was to identify and understand functional differences in major post-injury fibroblast subtypes,” Saraswati said.
The researchers evaluated the expression patterns of two marker proteins expressed in injury-activated fibroblasts: fibroblast specific protein 1 (FSP1) and alpha smooth muscle actin (alphaSMA) in mouse models of heart, skin and kidney injury, and in human heart tissue collected after heart attacks.
They found that FSP1 and alphaSMA were expressed by distinct fibroblast cells after tissue injury and that FSP1 fibroblasts were present at wound sites earlier than alphaSMA fibroblasts.
Previous studies have characterized alphaSMA as a hallmark of pathological fibrosis.
To explore molecular and functional features of these fibroblast subtypes, the researchers isolated FSP1 and alphaSMA cells from mouse models of heart injury.
They found that the FSP1 fibroblasts had pro-angiogenic (blood vessel promoting) gene expression and protein profiles. In an in vivo wound healing assay, the FSP1 cells promoted blood vessel development.
Together, the gene signature and functional findings support a role for FSP1 fibroblasts in immune cell recruitment to wound sites, cellular proliferation and angiogenesis.
These characteristics distinguish the FSP1 fibrolast subtype from the pro-fibrotic alphaSMA fibroblast, the researchers noted.
“This study clarifies the molecular and functional uniqueness of two distinct and prevalent post-injury fibroblast subtypes and begins to fill a critical gap in our knowledge about the roles of fibroblasts in healing and fibrosis,” Saraswati said.
“We hope that greater understanding of fibroblast subtypes will provide a new paradigm for treating organ fibrosis.”
Fibrosis is the typical response to injury that leads to distorted tissue architecture, pathologic signaling and ultimately organ dysfunction1.
In response to tissue injury, a population of fibroblasts activates, proliferates, and plays a vital part in the entire repair process: from the generation of granulation tissue, revascularization and restoration of cardiac architecture, and function to tissue remodeling and eventually scar formation2.
Alternatively, continued and unchecked fibroblast activity results in fibrosis, adverse cardiac remodeling, and ultimately heart failure4.
It is increasingly recognized that fibroblasts are a phenotypically and functionally diverse population of cells that must be studied in a temporal and tissue/injury-specific context5–7. In an uninjured heart, for example, fibroblasts are quiescent and express surface markers such as vimentin, Tcf21, and PDGFRα8.
Two markers expressed in the post-injury (activated) fibroblasts are fibroblast specific protein 1 (FSP1; S100A4) and α smooth muscle actin (αSMA); however, neither marker is exclusively specific to fibroblasts10–13.
Less is known about the significance of FSP1, a member of the S100 family of intracellular calcium-binding proteins. Several studies have confirmed FSP1 as a marker of injury-induced fibroblasts12,13,15.
Interestingly, FSP1+ and αSMA+ fibroblasts have been reported to be non-overlapping fibroblast populations in the injured heart, skin, liver, and kidney5,6,11–13,16, but their population dynamics as well as their functional and molecular differences in the context of tissue repair are lacking.
We hypothesize that functional and molecular characterization of the major post-injury fibroblast subtypes is necessary to target the pathological scarring/fibrotic effects of fibroblasts without impeding their pro-reparative functions.
In this study, we used transgenic mice in which FSP1 and αSMA-expressing cell populations were genetically tagged with GFP. We isolated these fibroblast subpopulations from the injured murine heart, taking specific care to exclude non-fibroblast populations, which are also known to express the FSP1 protein. Detailed molecular and functional studies are done to identify the unique roles of FSP1 and αSMA-expressing fibroblasts in injury repair.
FSP1 and αSMA are present in distinct cells post injury
We evaluated the expression pattern of FSP1 and αSMA proteins in uninjured organs and at various time points following tissue injury.
Human cardiac tissue was collected postmortem from cadavers at least 2 weeks post infarct (Fig. 1d).
For skin injury, excisional (1 cm) full-thickness wounds (Fig. 1b) were generated on the dorsum18. Kidney injury (Fig. 1c) was induced by unilateral ureteral obstruction by ligating the ureter just distal to the renal pelvis19.
FSP1 and αSMA protein expression and localization with respect to injury were evaluated by co-immunofluorescence (IF) using confocal microscopy. Consistent with published studies, uninjured heart, skin, and kidney tissues were negative for anti-FSP1 staining11–13. Anti-αSMA staining in uninjured and at early time points after injury was restricted to vascular structures (such as pericytes and vascular smooth muscle cells).
FSP1+ stromal cells were detectable within 48 h of injury in heart, at day 4 in skin, and at day 7 in the injured kidney parenchyma, whereas significant αSMA expression (excluding perivascular cells) appeared between 7 and 8 days after injury in all three models.
In the cardiac infarct model the abundance of cells expressing both markers declined significantly (or were completely absent) by day 30 after the generation of stable scar. No co-localization or overlap between these two cell populations was identified in the heart, kidney, or skin at any post-injury time point evaluated. FSP1 has been noted by multiple groups to also mark hematopoietic (myeloid) cells10,16 and endothelial cells8.
We confirmed that a population of both hematopoietic and endothelial cells expressing FSP1 were present in injured mouse left ventricles (LV) at post-MI d8 (Supplementary Fig. 1A). However, around 15% of the GFP+/FSP1+ cell population did not stain with CD45 or CD31 as identified through flow cytometry analysis and confocal microscopy, indicating the presence of a non-hematopoietic and non-endothelial FSP1-expressing cell subset (Supplementary Fig. 1, A, B).
To further study the degree to which FSP1+ cells originated from the bone marrow, FSP1-GFP chimeric mice were generated that had undergone ablative bone marrow transplantation (BMT) from syngeneic C57BL/6 mice.
The FSP1-GFP reporter mice (gift from Dr. Eric Neilson) express GFP under the FSP1 promoter. However, the FSP1-GFP chimeric animals, following engraftment with C57BL/6 marrow, would cease to express GFP from any bone marrow-derived cells.
These chimeric animals enabled us to evaluate the degree to which the number of GFP+cells present in the injured site were the result of inflammatory cells migrating into the injury site from circulation.
As expected, there was a significant reduction in GFP+ hematopoietic cells in the injured myocardium as identified by CD45 expression alone (from 42.9 to 2.4%) or by CD45/CD31 co-staining (from 34.8 to 7.75%) in BMT GFP-FSP1 mice in comparison to non-BMT GFP-FSP1 mice (Supplementary Fig. 2).
These results indicated that a substantial number of GFP-expressing FSP1+ cells in the injured heart were indeed hematopoietic cells originated from bone marrow. After exclusion of the BM-derived FSP1+ population, we observed a relative increase in non-hematopoietic GFP+/FSP1+ cells, lacking hematopoietic markers from 14.7 to 65.8% in BMT GFP-FSP1 chimeric mice, demonstrating that a majority of the FSP1+ fibroblasts originated at the site of injury (Supplementary Fig. 2). T
hese data confirmed that the injured heart contained substantial numbers of both FSP1+ fibroblasts and non-fibroblasts (hematopoietic and endothelial cells) post injury. In addition, most of the non-hematopoietic cells originate at the site of injury and are not recruited from the bone marrow.
More information: Sarika Saraswati et al. Identification of a pro-angiogenic functional role for FSP1-positive fibroblast subtype in wound healing, Nature Communications (2019). DOI: 10.1038/s41467-019-10965-9
Journal information: Nature Communications
Provided by Vanderbilt University