Mononuclear phagocytic system and fibrosis: back to the future?

Bronchiolitis obliterans syndrome (BOS) is a chronic lung allograft dysfunction following lung transplantation [1–3], with histopathological characteristics of obliterative bronchiolitis (OB) that includes small airway epithelial disruption, submucosal inflammation and fibrosis, and obstruction of airway lumen. Clinical diagnosis is based on persistent decline of the allograft lung function measured by forced expiratory volume in 1 s and/or forced expiratory flow at 25–75% of forced vital capacity. 50% or more of lung transplantation recipients who survive beyond 5 years develop BOS, accounting for significant cases of lung transplantation failure and the leading cause of death for recipients who survived beyond 1 year after transplantation [4]. Current therapy is not effective in reversing BOS progression once that occurs. Although clinical studies have found several risk factors for BOS, which include infection and inflammation, alloimmune reaction and autoimmunity, graft airway injury and abnormal repair, the pathogenic mechanisms underlying BOS, in particular, progressive fibrotic obliteration of small airway lumen, remain unknown. Di Campli et al. [5] demonstrate that myeloid precursor-derived mononuclear phagocytes from allograft recipients are of clinical relevance and contribute to the mesenchymal cell population that invades the airways after transplantation. Targeting these cells may have clinical implications, and therapy has the potential to improve outcomes after lung transplantation.

The term “mononuclear phagocyte system” (MPS) was first introduced in 1972, by van Furth et al. [6]. This revised the earlier classifications (figure 1) of the “macrophage system”, described first by Metchnikoff in 1882, which included “macrophages” and “microphages” at the time, expanded by Aschoff in 1924 as the “reticuloendothelial system” by including endothelial cells and fibrocytes/fibroblasts. Volterra (1927) proposed a new classification and termed it “reticulo-histiocyte system” which later was reintroduced by Thomas in 1947. The revised MPS was based on the phagocytic capacity of cells and included “promonocytes and their precursors in the bone marrow, the monocytes in the peripheral blood, and the macrophages in the tissue” [6]. Reticular cells (collagen-producing cells), dendritic cells, endothelial cells and fibroblasts were excluded at that time as these cells are not highly phagocytic. The fascinating story of development of the MPS can be appreciated in the memoranda of van Furth et al. [6] and in many of the excellent reviews authored by D.A. Hume and co-workers, who elegantly describe the history and the evolvement of the MPS, and concluded in 2008 that “the concept of the MPS has perhaps outlived its usefulness”, and that novel systems of classifications might be warranted [7–9]. With the use of novel and sophisticated molecular technologies, such as cell lineage tracing and fate mapping strategies, including the publication by Di Campli et al. [5] in this issue of the European Respiratory Journal, it is agreed that the concept and classification of the MPS needs to be revised as it may have important physiological and pathological functions other than phagocytic activity.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1

Evolution of the mononuclear phagocyte system (MPS). As described in 1924 by Aschoff, the “reticuloendothelial cell system” (RES) involved various cell types including fibrocytes/fibroblasts and endothelial cells. In 1969, the system was revised to include only cells of phagocytic capacity. Modern day evidence suggests a shift in the understanding of the current MPS, to re-introduce cells such as the fibrocytes, fibroblasts, and endothelial cells.

Heterotopic and orthotopic tracheal transplant models in both small and large animals have been commonly used to study the molecular and cellular mechanisms of OB/BOS [10–12]. Furthermore, more complicated orthotopic lung transplant and intrapulmonary tracheal transplant animal models have also been developed [13, 14]. In the heterotopic transplant model, trachea from genetically discordant donor mice are transplanted into subcutaneous pouches in the upper back of a recipient mouse, resulting in delayed rejection of the allografts. The grafted airway rejection is preceded by substantial mononuclear cell infiltration in subepithelial and peri-airway regions (similar to lymphocytic bronchiolitis) that peaks between days 10 and 14, followed by lumen obliteration comparable to human OB by days 21 to 28. In contrast to the rejection pathology observed in tracheal airway allografts, genetically identical airway isografts have normal-appearing tracheal histology at day 28, with evidence of neovascularisation.

Myofibroblasts are the major cells to produce excessive collagen deposition as a key structural feature of fibrosis. In normal submucosa of airways, myofibroblasts are barely detected using anti-SMA immunohistochemistry, while they are enriched in OB. The origins of these myofibroblasts are not clear, and the data from different studies are controversial, varied from different studies using different allograft models. Several studies have suggested that donor origins of myofibroblasts are present in lung grafts with OB and may contribute to OB progression. By checking sex mismatched donor–recipient human grafts, mesenchymal stem cells isolated from BOS patients are about 90% positive for donor sex chromosome [15, 16]. Using an orthotopic left lung transplantation in mice, increased infiltration of recipient CD3⁺ T-cells and F4/80 macrophages is detected at all time points, with the peak at post-transplantation day 7–14 while alveolar macrophage is decreased. However, about 89% of collagen I⁺/E-cadherin⁻/CXCR4⁻ cells have donor-origin in the OB lesions at day 28 [17], suggesting a pathogenic role of donor resident mesenchymal cells. Interestingly, airway epithelial biopsies from lung transplantation patients with BOS demonstrates significant increased staining for mesenchymal markers (S100A4 and SMA) and reduced E-cadherin staining compared with the patients without BOS [18]. Mouse orthotopic tracheal transplantation with allograft showed similar reduced E-cadherin and increased co-staining of SMA and S100A4, as well as ZEB1 (a transcription factor involved in epithelial-mesenchymal transition) at post-transplantation day 7 [19], suggesting that donor epithelial–mesenchymal transition may also contribute to the fibrotic process of OB. However, circulating fibrocytes (CD45⁺/CollagenI⁺) of recipient origin were reported to be increased in BOS patients compared to non-BOS as measured by FACS, and such increase was associated with advancing BOS stage (III) [20, 21]. By microdissecting the OB lesions from 12 lung transplantation patients, about 32% of fibroblasts were recipient-derived [22]. Furthermore, in another two bone marrow transplantation patients who developed BOS, 6–16% cells in the OB lesions were from the transplanted bone marrow [22]. These suggest that recipient's circulating fibrocytes, probably bone marrow-derived, may also contribute to the fibrotic lesions in small airways of the allograft. Using both orthotopic and heterotopic tracheal transplantation mouse models with a ubiquitous green fluorescent protein cell reporter as the donor or the recipient, the majority of myofibroblasts (SMA+) were found to be of recipient origin [23]. Using a rat allograft model combining orthotopic lung transplantation and intrapulmonary tracheal transplantation, Sato et al. [24] showed that the myofibroblasts in the OB lesions were mainly extrapulmonary origin of the recipient. The study reported by Di Campli et al. [5] in this issue of the European Respiratory Journal shows for the first time that recipient's myeloid precursor-derived cells/mononuclear phagocyte system may contribute to the majority of myofibroblasts in OB lesions in a HTT mouse model, using a variety of lineage-specific reporter mice and a lineage-specific depletion mice. In this work, it is clearly demonstrated using elegant transplant technology and myeloid labelled strains, that most of the myofibroblasts in the occluded trachea do not originate from the transplanted trachea but are derived from the recipient host (figure 2). These myofibroblasts colocalise with CD68, a well-known marker of monocytic and macrophage origin, suggesting that these cells are derived from the myeloid lineage and transdifferentiated into collagen and extracellular matrix-producing myofibroblasts.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2

Haematopoetic cell mononuclear phagocyte system, as exhibited through experiments by Di Campli et al. [5]. Through lineage tracing of Cx3cR1⁺ cells, it has been shown that cells of monocytic origin contribute to the mesenchymal cell population. After tracheal transplant, the majority of the myofibroblasts in the trachea were from the recipient host, and expressed markers including smooth muscle α-actin (αSMA), CD45, CD68 and F4/80. GFP: green fluorescent protein.

Based on this work, several important questions needs to be addressed. 1) Do these findings exclude other competing hypotheses addressing the origin of myofibroblasts, including epithelial or local origins? 2) Do these findings provide additional rationale for novel antifibrotic therapeutic strategies, where targeting strategies directed toward the circulating precursors cells in the MPS system should be explored further? 3) Would therapeutic strategies targeting the extravasation process of circulating progenitors prevent their entry into the tissue? d) What are the key factors governing the transdifferentiation of progenitor cells into myofibroblasts, and could these be targeted therapeutically? Finally, and perhaps of importance, if a revised MPS classification will include fibroblasts and myofibroblasts, would we need to revise conclusions of a very large body of previous work depleting MPS precursors using a variety of elimination strategies, including clodronate pre-treatment? These are all valid questions that remain to be fully addressed. An increasing amount of data suggests that monocytes have the potential to become myofibroblasts [25–31]. In other models and organ systems, about 50% of myofibroblasts have been demonstrated to be colocalised with CD68, in concordance with the current manuscript, confirming the possibility of a myeloid origin [29]. With the avenue of publicly available single cell RNAseq datasets, we have here addressed the same question in patients diagnosed with idiopathic pulmonary fibrosis. The datasets were obtained, re-curated and processed. As shown in figure 3, 53.5% (dataset A) and 46.6% (dataset B) of ACTA2 positive cells (myofibroblasts) also expressed CD68, strengthening the argument in favour of re-including fibroblasts and myofibroblasts in the MPS. Although monocytes are short-lived, with a lifespan of approximately 24 h, macrophages and fibroblasts/myofibroblasts are known to be long-lived and can sometimes be shown to be proliferative, although the factors determining their long life are unclear. The key might be found in the microenvironment and the extracellular matrix, perhaps combined with a variety of factors (such as exposure to cigarette smoke and pollutants) that increase vascular leakage. Although these mechanisms remain to be fully understood, the story of the MPS is still to be written and harbours likely unique opportunities for therapeutic intervention and modulation of fibrotic diseases.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3

Co-expression of ACTA2 and CD68 in idiopathic pulmonary fibrosis (IPF) patients and healthy donors. Data were obtained from publicly available datasets containing samples from donors and IPF patients (GEO: GSE122960, “Dataset A”, and GSE135893, “Dataset B”). Only samples obtained from healthy donors and IPF patients were selected for the analysis. Dataset A contained samples obtained from eight lung biopsies from transplant donors and four lung explants from transplant recipients with IPF [32]. Dataset B contained samples obtained from peripheral lung tissue removed at the time of lung transplant surgery from 12 patients with IPF and from 10 non-fibrotic patients [33]. Processing, analysis and visualisations were performed using Seurat package [34] in R. Cell populations were defined using the markers from the source papers related to each of the datasets. After examination of the expression level distributions for CD68 and ACTA2 genes, all cells showing levels of expression >0 were defined as “positive” for that gene. a) tSNE plots showing cell populations (left panel) and expression of ACTA2 and CD68 separately and superimposed together in Dataset A. b) UMAP plots showing cell populations (left panel) and expression of ACTA2 and CD68 separately and superimposed together in Dataset B. c) Stratification of cells based on their levels of expression of ACTA2 and CD68 (Dataset A). d) Stratification of cells based on their levels of expression of ACTA2 and CD68 (Dataset B).

Shareable PDF