Transforming growth factor beta: the good, the bad, and the ugly.

Abstract

A is often the case with peptide growth factors and cytokines, the original activity discovered and identified as TGF-3 merely represented the tip of the iceberg. Not only does it regulate growth, as evident by its name, but TGF-B mediates far-ranging biologic processes including inflammation and host defense, in addition to development, tissue repair, and tumorigenesis. Released locally from platelet stores early in an inflammatory response, TGF-3 is then generated by inflammatory cells themselves as part of the cytokine network. Initially, in its defensive role, TGF-~ mediates the egress of undifferentiated leukocytes and, subsequently, it facilitates resolution of inflammation and promotes tissue repair (1, 2). But this carefully choreographed series of events, which represents the "good" side of TGF-3, is dependent upon a critical balance of the growth factor. It is becoming increasingly evident that too much of a good thing can be bad; excess TGF-B within a lesion has been associated with unresolved inflammation and fibrotic events (1-3). Now, it appears, based on the study by Lowrance et al. (4) in the MRL/Ipr autoimmune murine model, that TGF-3 has an even uglier side, in that its accumulation in the circulation may predispose the host to serious and recurrent infections. Moreover, in humans, genetic and acquired diseases such as systemic lupus erythematosis (SLE) may mirror the MRL/lpr mouse. Under the best of circumstances, TGF-3 provides a link between each of the processes by which cells and/or tissues respond to infection or injury and initiate repair. By increasing adhesion molecules, generating a potent chemotactic gradient, and inducing itself and an array of other factors as part of the cytokine network (5-7), TGF-~ orchestrates leukocyte recruitment and activation. TGF-13 then downregulates these processes by inhibiting the functions of inflammatory cells once they are activated (2, 8) and facilitates healing by promoting fibroblast recruitment and matrix synthesis (3). As recently reviewed (2), this apparent contradictory influence of TGF-3 on cells of the immune system, both stimulatory and inhibitory, is accounted for, in part by the differential effects of TGF-3 on resting and activated cells. As a general, but by no means exclusive rule, resting, immature ceils are stimulated by TGF-3, whereas activated representatives of the same cell populations may be inhibited by TGF-3. Thus, TGF-3 serves as a conversion factor, converting an active inflammatory site into one dominated by resolution and repair. Upsetting the delicate balance of TGF-3 that dictates these events may have pathologic consequences. Indeed, persistent stimulation of immune/inflammatory events by internal or external challenges can overload the balance with chronic secretion of TGF-B. Not only might this local excess lead to unresolved inflammation (9), but also, TGF-~/may reach the circulation. When elevated systemic levels of TGF-13 occur, either endogenously during disease states (4, 10), or after exogenous administration (11-14), a plethora of additional cellular targets and consequently, diverse signaling pathways with different outcomes, may become operative. Although the source(s) of the endogenous levels of blood TGF-/3 is uncertain in the autoimmune MRL/lpr mice, the spleen may contribute to the circulating pool in this and perhaps other models characterized by immune malfunction (4, 15). Why TGF-/~ accumulates is unknown, but TGF-/~ autoinduces its own synthesis (7). In addition, dysregulation of its activity because of persistence of the enzymatic machinery essential for its activation or insu~cient levels of inhibitors may be contributory. Nonetheless, whereas the tightly controlled, localized production of TGF-B occurs as a necessary component of an active inflammatory process, the aftermath of excess TGF-B, now found in the circulation, is to inhibit the same immune and inflammatory pathways. It is not inconceivable that this built-in negative feedback loop, occurring in TGF-B excess, evolved for the protection of the host during bouts of rampant inflammation. The dichotomy between local and systemically disseminated TGF-B was first documented in an experimental arthritis model in which local secretion or administration of TGF-13 was found to drive the inflammatory response, whereas systemic inoculation inhibited the same response (9, 13, 16) (Fig. 1). By what mechanism(s) does circulating TGF-3 promote immune suppression and can these pathways be used to clinical advantage? Recent evidence suggests that systemic TGF-13 targets endothelial ceUs, where it inhibits E-selectin expression to block adhesion and targeting of leukocytes to the site of inflammation (17) (Fig. 2). Moreover, since leukocytes are normally sensitive to a concentration gradient of chemotactic signals emanating from the site of inflammation (5, 9), the presence of elevated TGF-3 in the blood would destroy such an outward gradient (Fig. 2). Evidently, the effects of secreted or exogenous TGF-3 extend far beyond the control of cell growth and differentiation to include a myriad of other regulatory activities converging on phagocytic cells and activated lymphoid cells of all lineages (reviewed in 1, 2). While it