EUROPEAN  TISSUE  REPAIR  SOCIETY

FIBROCYTES: CIRCULATING FIBROBLASTS THAT MEDIATE TISSUE REPAIR
Richard Bucala, The Picower Institute for Medical Research, Manhasset, NY 11030, USA

Wound Repair is a carefully co-ordinated process that results from a complex interplay of cellular, humoral, and connective tissue elements. The host response to wounding has generally been consi­dered to encompass three distinct but overlapping phases, a ‘clotting’ or haemostatic phase in which coagulation proteins and blood platelets act to prevent blood loss, an ‘inflammatory’ phase in which peripheral blood leukocytes combat infection and debride necrotic tissue, and a ‘repair/remodelling’ phase in which connective tissue cells, in concert with inflammatory cells and their products, act to restore damaged tissue to its original integrity and architecture.^1,2

Among the connective tissue cells which are critical for the remodelling and restoration of injured tissue is the fibroblast. The fibroblast is the primary cellular component of connective tissue and a major synthetic source of matrix proteins such as collagen. Connective tissue fibro-blasts are normally quiescent and remain sparsely-distributed throughout the extracellular matrix.³ Beginning with the inflammatory phase of the wound healing response however, fibroblasts begin to play an increasingly dominant role in tissue repair. These cells become activated by the cytokine and growth factor rich milieu of the wound and contribute importantly to the repair response by entering the cell cycle, increasing their synthesis of autocrine and paracrine acting growth factors, and releasing-in a carefully co-ordinated fashion, both matrix proteins and matrix metalloproteases. Over time these fibroblast products effect a repair and remodelling response that is critical for the proper healing of injured tissue.^1-3

Where Do Wound Fibroblasts Come From?

The precise origin of the fibroblasts that appear within wounds has been the subject of considerable interest, beginning with the first microscopic studies of developing connective tissue described by Paget in 1863.⁴ It is generally accepted that the connective tissue fibro­blasts present in and around an injured area can be induced to undergo proliferation and migration as part of the repair/remodelling response, and this idea has been supported by various microscopic studies pointing to an apparent entry of fibro-blasts from adjacent areas. Nevertheless, beginning with the observations of Conheim, and then of Metchnikoff, which also were performed over 100 years ago, the hypothesis that certain of the fibroblast-like cells that are present in wound arise from the influx of circulating cells has been re-considered from time to time.⁵ This concept has been kept alive by independent observations which have supported a model of leukocyte differentiation into fibroblasts or one involving the production of collagen by peripheral blood cells within subcutaneous diffusion chambers.^6,7 Observations that India Ink-tagged monocytes fail to develop into tissue fibroblasts in vivo also have not formally excluded the potential blood source of fibroblast or fibroblast-like precursor cells.⁸Nevertheless, the model of fibroblast ingrowth from subjacent tissue has become so dominant over the last twenty-five years that the alterna-tive notion that fibroblasts might arise from the circulation and contribute significantly to effect the repair of injured areas has until recently not received careful attention.

Discovery of Fibrocytes

A number of years ago, investigations by our laboratory into the phenotype of cells present within experimentally-implanted, subcutaneous wound chambers led to the discovery of an apparently novel, fibroblast-like cell that enters these chambers from the circulation.⁹ The wound chambers we employed comprised short segments of silastic tubing filled with polyvinyl alcohol sponge. These chambers serve as a nidus for an exudative reaction that is to say their implantation into the subcutaneous tissue of mice produces a rapid infiltration of inflammatory cells from peripheral blood.¹⁰ We observed, however, that the in vitro cultivation of the cells aspirated within the first day of wound chamber implantation resulted in large numbers of adherent, spindle-shaped cells that closely-resembled fibroblasts by morphology. The rapid appearance of these cells together with the unlikely possibility that they could have arisen by the dislodgement and migration of subjacent connective tissue cells led us to propose that these fibroblast-like cells arose form a population present in the inflammatory exudate. While we recog­nized at the time that circulating monocytes can sometimes be observed to assume an extreme spindle-shaped morphology, this possibility was negated by an analysis of the surface phenotype of these cells. Flow cytometric analysis of these spindle-shaped cells showed them to stain positively for Type I collagen and for CD349. At the time, the cell surface marker CD34 was believed to be expressed exclusively by hematopoietic stem cells or endothelial cell precursors, and the combination of CD34 and collagen expression had only previously been described on mouse embryonic brain fibroblasts or in the 3T6 fibroblast cell line.¹¹ Additional markers expressed by these cells included the pan-myeloid antigen CD13, the pan-leukocyte marker CD45RO, and HLA class II. Fibrocytes do not synthesize epithelial (cyto-keratin), endothelial (von Willebrand factor VIII-related protein), or smooth muscle (a-actin) cell markers and are negative for non-specific esterases as well as the monocyte/macrophage-specific markers CD169. These studies provided molecular support for the concept of a blood-borne cell population displaying both fibroblast-like properties as well as the markers of bone marrow-derived cells. Follow-up immunohistochemical analyses of wound chambers that had been implanted in mice confirmed the presence of CD34+ spindle-shaped cells in areas of collagen matrix deposition. These cells also were identified to be present in mouse or human cutaneous scars, particularly early in the cellular phases of the repair/remodeling response.⁹ We termed these circulating fibroblasts ‘fibro-cytes’, akin to other blood-borne cells such as erythrocytes, leukocytes, and thrombocytes.

More detailed investigations have since demonstrated that peripheral blood fibrocytes comprise ~0.5% of circulating leukocytes. Scanning electron microscopy has shown these cells to be morphologically distinct from blood-borne leukocytes and to display unique cytoplasmic extensions intermediate in size between microvilli and pseudopodia (Fig 1). Interestingly, a recent report has implicated these projections in the adhesion of the Lyme disease pathogen Borrelia burgdorferi to the fibrocyte.¹²

Figure 1. Scanning electron micrograph of a human fibrocyte (from ref. 1).

The presence on fibrocytes of the hematopoietic stem cell antigen CD34 led us to investigate the potential bone marrow origin of these cells. In a study utilizing lethally-irradiated female mice re-constituted with bone marrow from male mice, we tracked by DNA amplification the male-specific SRY-gene in fibrocytes that had been recruited to subcutaneously-implanted wound chambers. The conclusion of these experiments was that fibrocytes do not arise from a radiosensitive bone marrow population but are derived instead from either a radioresistant, bone marrow progenitor or from other tissue sources. Conceivably, fibro-cytes may arise from bone marrow stroma, which comprises radioresistant, fibroblast-like cells that express CD3413. However, this possibility awaits experimental validation.