Fibroblasts and endothelial cells are able to influence immune responses through expression of specific cell adhesion molecules [27
], cytokines [28
] and chemokines [29
] that collectively define the parameters required for entry, proliferation, survival and exit of leukocytes in a particular tissue. This stromal-derived 'address code' plays a key role in the initiation and resolution of inflammation but it is also an important factor in the disordered leukocyte trafficking associated with chronic persistent inflammation [30
]. In previous reports we identified components of the stromal address code that are associated with specific tissues [31
] and inflammatory responses [29
]. Glucocorticoids are powerful endogenous modulators of inflammatory signal transduction [3
], and because their effects vary between tissues we hypothesized that components of glucocorticoid action may also contribute to the stromal address code. Data presented here suggest that one of these factors, namely the glucocorticoid-metabolizing enzyme 11β-HSD1, is likely to play an important role in tissue-specific and inflammation-specific regulation of fibroblast function.
The predominant GR isoform, GRα, is ubiquitously expressed and mediates a wide range of responses to endogenous and exogenous glucocorticoids. Thus, dysregulated GR signalling is likely to be an integral component of persistent inflammation, and previous studies have associated glucocorticoid resistance with inflammatory disease [3
]. In the present study we could detect no difference in GR expression between tissues or in response to cytokine treatment. Instead, striking variations in expression were observed for the enzyme 11β-HSD1, which exhibited cell-specific and cytokine-specific variations in expression and activity. Cytokine induction of stromal cell 11β-HSD1 has previously been reported in adipocytes [32
], osteoblasts [33
] and amnion-derived fibroblasts [20
]. However, our data indicate that 11β-HSD1 expression and function are also subject to significant tissue-specific variations. Differential expression of 11β-HSD1 did not appear to be due to underlying inflammatory disease in donor RA patients, because similar observations were also made with cells from patients with OA. Furthermore, because fibroblasts from different sites were of matched donor origin, the effects of different exposure to drugs and disease duration can be excluded as confounding factors. It is possible that the method of fibroblast isolation could affect cell phenotype, but fibroblasts isolated in this way appear to maintain a pattern of gene expression to that seen in vivo
There was a strong correlation between 11β-HSD1 and the transcription factors C/EBPα and C/EBPβ, which have been associated with hepatic expression of 11β-HSD1 [26
]. C/EBPβ has been shown to be constitutively activated in RA synovial tissue but it does not appear to be involved in IL-1-induced expression of IL-6 or IL-8, both of which are more closely linked to induction of nuclear factor-κB [35
]. We also observed no significant alteration in C/EBP expression in cells treated with proinflammatory cytokines, but our data suggest that tissue-specific variations in 11β-HSD1 may be dependent on the relative levels of C/EBP isoforms.
Although 11β-HSD1 is classically associated with hepatic glucocorticoid responses, its effects on stromal cells appear to be equally important as exemplified by the recent characterization of the adipocyte-specific transgenic mouse for 11β-HSD1, the phenotype of which is profound obesity [36
]. In addition to this effect on mature cell function, the enzyme also appears to be intimately associated with stromal cell proliferation [15
] and differentiation [16
]. Differential expression of fibroblastic 11β-HSD1 may thus fulfil several functions. For example, the induction of the enzyme by IL-1 and TNF-α may be part of an autocrine feedback mechanism regulating inflammatory signalling. Autocrine metabolism of glucocorticoids has been identified in several cell types associated with immune responses, including macrophages [37
], dendritic cells [38
], synoviocytes [39
], and lymphocytes [40
]. Each of these different cell types appears to utilize 11β-HSD1 to attenuate distinct immune responses in an autocrine fashion [37
], but the manner in which they are likely to interact in specific tissues remains unclear.
One recent study [39
] suggested that synthesis of cortisol is reduced in cells from RA patients when compared with their OA counterparts, suggesting that a decreased availability of anti-inflammatory cortisol may contribute to the development or persistence of RA. This study reported expression and activity of 11β-HSD1 and suggested that this increased in response to inflammation, but the major difference between RA and OA tissues was the apparent expression of the glucocorticoid-inactivating enzyme 11β-HSD2 in nonfibroblastic cells. Although the nature and origin of the 11β-HSD2-expressing cells in RA is clearly of importance, it still appears that the normal response of synovial fibroblasts is to generate active glucocorticoids.
Exogenous glucocorticoids have a dramatic effect on synovial inflammation but their use is limited by systemic side effects. The data presented here indicate that there is an endogenous local counterpart that may play an important role in regulating synovial inflammation. It is possible that this system is needed to counteract endogenous inflammatory regulators (such as macrophage migration inhibitory factor [41
]) that are more highly expressed in synovium than in other tissues. It is also possible that impairment of local glucocorticoid generation will adversely affect the inflammatory response within the joint. Although we saw no difference in the capacity of synovial fibroblasts from patients with RA or OA to generate active glucocorticoids, it is possible that variations in the timing of this response may contribute to the pathogenesis or severity of inflammatory arthritis. For example, local metabolism of prednisone/prednisolone may partly explain the early dramatic response to therapy that occurs as a result of increased generation of active glucocorticoids in affected tissues. Future studies of 11β-HSD1 in vivo
will help to clarify its importance for the resolution, persistence and treatment of inflammation.