Intravascular Glucocorticoid Metabolism during Inflammation and Injury in Mice

11β-Hydroxysteroid dehydrogenases (11βHSDs) catalyze interconversion of 11-hydroxy-glucocorticoids with inactive 11-keto metabolites. In blood vessel walls, loss of 11βHSD1 is thought to reduce local glucocorticoid concentrations, reducing the progression of atheroma and enhancing angiogenesis. Conversely, on the basis that 11βHSD1 is up-regulated approximately 5-fold by inflammatory cytokines in cultured human vascular smooth muscle cells, it has been proposed that increased 11βHSD1 during vascular inflammation provides negative feedback suppression of inflammation. We aimed to determine whether inflammation and injury selectively up-regulate 11βHSD1 reductase activity in vitro and in vivo in intact vascular tissue in mice. In isolated mouse aortae and femoral arteries, reductase activity (converting 11-dehydrocorticosterone to corticosterone) was approximately 10-fold higher than dehydrogenase activity and was entirely accounted for by 11βHSD1 because it was abolished in vessels from 11βHSD1^−/− mice. Although 11βHSD1 activity was up-regulated by proinflammatory cytokines in cultured murine aortic smooth muscle cells, no such effect was evident in intact aortic rings in vitro. Moreover, after systemic inflammation induced by ip lipopolysaccharide injection, there was only a modest (18%) increase in 11β-reductase activity in the aorta and no increase in the perfused hindlimb. Furthermore, in femoral arteries in which neointimal proliferation was induced by intraluminal injury, there was no change in basal 11βHSD1 activity or the sensitivity of 11βHSD1 to cytokine up-regulation. We conclude that increased generation of glucocorticoids by 11βHSD1 in the murine vessel wall is unlikely to contribute to feedback regulation of inflammation.

THE LINK BETWEEN glucocorticoids and cardiovascular disease is well established: systemic glucocorticoid excess (1) and exogenous glucocorticoid therapy (2, 3) are associated with an increase in cardiovascular risk. Epidemiological studies have associated higher cortisol secretion with risk factors for cardiovascular disease, including hypertension, insulin resistance, impaired glucose tolerance, and dyslipidemia (4, 5). Furthermore, the local action of glucocorticoids within the blood vessel wall may be important because glucocorticoids influence the regulation of vascular tone (6, 7) and the vascular response to inflammation and injury (8, 9). This may have therapeutic as well as pathophysiological significance: for example, glucocorticoid therapy has been shown to prevent neointimal proliferation after intraluminal vascular injury (10, 11).

Recent studies demonstrated the importance of tissue-specific regulation of glucocorticoid concentrations (12, 13). Interconversion of glucocorticoids (corticosterone in rodents) and their 11-keto-metabolites (11-dehydrocorticosterone) is catalyzed in target tissues by the 11β-hydroxysteroid dehydrogenases (11βHSDs). 11βHSD type 1 acts predominantly as a reductase in vivo, catalyzing the regeneration of active glucocorticoids and thereby promoting activation of glucocorticoid receptors in target tissues. 11βHSD type 2 is an exclusive dehydrogenase (14–16) and prevents illicit activation of mineralocorticoid receptors by glucocorticoids. Both isozymes are expressed in the vessel wall (17–20), as are both mineralocorticoid and glucocorticoid receptors (6, 21). In mouse, 11βHSD1 is predominantly localized to the smooth muscle, whereas 11βHSD2 is found in the endothelium (21). 11βHSD2 prevents glucocorticoids from inhibiting endothelium-dependent vasodilatation (22), but a role for 11βHSD1 in the vessel wall has only recently been elicited and appears to relate to vascular remodeling. By increasing local glucocorticoid concentrations, 11βHSD1 amplifies the angiostatic actions of glucocorticoids (9). Conversely, inhibition of 11βHSD1 protects against atherogenesis in ApoE^−/− mice; this effect is disproportionate to changes in serum lipid profile and may reflect alterations in local regeneration of glucocorticoids within the vessel wall (23).

It has been suggested that 11βHSD activity is regulated in diseased blood vessels, contributing to local feedback regulation of inflammation. This hypothesis is based on the observation that proinflammatory cytokines (e.g. TNFα) up-regulate 11βHSD1 activity and expression and down-regulate 11βHSD2 expression in cultured human aortic smooth muscle cells (24), favoring increased local glucocorticoid concentrations. However, whereas cytokines alter 11βHSD activity in vascular and other cultured cell types (24–31), the relevance of this observation during inflammation in intact blood vessels has not been established. Here we use models of systemic and local vascular inflammation in vitro and in vivo to address the hypothesis that exposure to proinflammatory cytokines selectively up-regulates 11βHSD1 activity in intact blood vessels.

Materials and Methods

Mice

Male, C57B6J (Charles River, Kent, UK) and 11βHSD 1^−/− mice on a C57B6J genetic background (Harlan Orlac, UK) (32) were maintained under controlled conditions of light (on 0800–2000 h) and temperature (21 C) with free access to chow (Special Diet Services, Witham, UK) and water. Animal experiments were carried out under Home Office license and conformed to standards defined in The Principals of Animal Care (National Institutes of Health publication 85–23, revised 1985).

Materials

Salts were obtained from BDH (Dorset, UK). 1,2,6,7-[³H]₄-corticosterone was obtained from Amersham Biosciences (Buckingham, UK). 1,2,6,7-[³H]₄-11-dehydrocorticosterone was synthesized in house from 1,2,6,7-[³H]₄-corticosterone using rat placental homogenate (33) and was more than 99% pure when assessed by HPLC. Murine recombinant TNFα, IL-1β, IL-4, and IL-13 (R&D Systems, Abingdon, UK) were stored in PBS containing 0.1% fetal calf serum (FCS) in aliquots at −20 C until required. Etanercept (Wyeth, UK) was dissolved in sterile water. Lipopolysaccharide (LPS), derived from Escherichia coli (serotype 0111:B4; Sigma, Poole, UK), was dissolved in sterile 0.9% saline and stored at −20 C. Bioactivity of TNFα was confirmed by means of a neutrophil apoptosis assay (data not shown) (34).

Effects of IL-1β on 11βHSD1 activity in cultured vascular smooth muscle cells

Mice were killed by cervical dislocation. The thoracoabdominal aortae and the iliofemoral vessels were immediately removed into ice-cold DMEM F12 (Invitrogen, Renfrewshire, UK) and cleaned of periadventitial tissue. Primary murine aortic (MA) smooth muscle cells (SMCs) were cultured using a modification of the method of Ray et al. (35). The cells were maintained in DMEM containing 20% FCS, 1% penicillin/streptomycin, and 1% l-glutamine 200 mm (Life Technologies, Inc., Paisley, UK) in a humidified oxygenated (95% O₂-5% CO₂) atmosphere.

To demonstrate that 11βHSD1 activity in murine-cultured SMCs is up-regulated after cytokine stimulation as in human SMCs (24), MA-SMCs (passage 2) were seeded onto 6-well plates at 1.75 × 10⁵ cells/well in 2 ml of assay medium. The following day, the medium was replaced with basal medium [containing 0.5% FCS plus IL-1β (20 ng/ml) or vehicle], and cells were incubated for 48 h. [³H]₄-11-dehydrocorticosterone (10 pmol) was then added to the appropriate wells and the cells incubated for a further 24 h. Steroids were extracted from the culture medium supernatant using C₁₈ Sep-pak columns (Waters Millipore, Watford, UK). [³H]₄-corticosterone and [³H]₄-11-dehydrocorticosterone were separated by HPLC and quantified by on-line liquid scintillation counting. Enzyme activity was expressed as conversion per 1.75 × 10⁵ cells after subtraction of apparent conversion in negative control wells.

Effects of cytokines on 11βHSD activity in intact vascular and hindlimb tissues in vitro

11βHSD activities were measured in murine aortic rings and iliofemoral vessels by adapting the method of Souness et al. (36). Briefly, vessels were incubated (24 h, 37 C) in DMEM-F12 (1 ml) containing [³H]₄-steroid and supplemented with streptomycin (100 μg/ml), penicillin (100 U/ml), and amphotericin (0.25 μg/ml). 11β-Reductase activity was determined in vessels from wild-type and 11βHSD1^−/− mice by adding 10 pmol [³H]₄-11-dehydrocorticosterone. 11β-Dehydrogenase activity was determined in vessels from wild-type mice by adding 10 pmol [³H]₄-corticosterone. 11β-Reductase activity was also determined using the method described for arteries in hindlimb tissues (dissected pieces of quadriceps muscle and skin with subcutaneous fat) from wild-type mice.

To assess the influence of cytokines in vitro on 11β-reductase activity, single aortic rings from C57B6J mice were preincubated with murine recombinant TNFα (10–1000 ng/ml); IL-1β (1–100 ng/ml); IL-4 (50 ng/ml); IL-13 (50 ng/ml); Etanercept [a fusion protein that antagonizes human and murine TNFα and that ameliorates the cardiovascular effects of murine TNFα (37), 0.1–10 μg/ml]; or vehicle. After 16 h incubation (24), 10 pmol [³H]₄-11-dehydrocorticosterone was added and incubation continued for a further 24 h. Control wells without tissue contained medium alone, [³H]₄-11-dehydrocorticosterone, or [³H]₄-11-dehydrocorticosterone with cytokines.

To observe the effect of cytokines in vitro on 11β-dehydrogenase activity, three to five aortic rings were preincubated for 16 h with murine recombinant TNFα, IL-1β, or vehicle. At 16 h, 10 pmol [³H]₄-corticosterone was added and incubation continued for a further 24 h. Control wells contained [³H]₄-corticosterone alone, medium alone, and [³H]₄-corticosterone with cytokines.

After incubation, steroids were extracted from the culture medium and assayed as described for cultured cells. Aortic ring tissue, which contains only 2–3% of the added radioactivity (36), was not included in the extraction.

Effects of LPS in vivo on vascular 11βHSD1 activity in aorta and perfused murine hindlimb

As described by Brandes et al. (38), the aorta at the thoracoabdominal transition was isolated, after cervical dislocation, and a cannula (24-gauge; Neoflon, Ohmeda, Sweden) introduced (distal to the renal arteries) and secured at the iliac bifurcation with a 3–0 prolene suture (Surgical Supplies Ltd., Cumbernauld, UK). The distal inferior vena cava was also cannulated (18-gauge cannula; Venflon, BD, UK) and secured with a 3–0 prolene suture. The hindlimb was constantly perfused, via the aortic cannula, with warmed oxygenated modified Krebs-Henseleit solution containing 2% BSA (Sigma) at flow rates of between 0.8 and 1.2 ml/min to achieve a perfusion pressure of approximately 40 mm Hg. Effluent was collected from the venous cannula. After a 10-min equilibration period, [³H]₄-11-dehydrocorticosterone (for determination of reductase activity) or [³H]₄-corticosterone (for determination of dehydrogenase activity) was added to the perfusion buffer (final concentration 5 nm). Perfusion was maintained for up to 60 min and aliquots of effluent collected at regular intervals. Steroids were extracted from the effluent and analyzed as described for aortic homogenates. To determine enzyme kinetics for 11β-reductase activity, the above procedure was replicated using concentrations of [³H]₄-11-dehydrocorticosterone between 5 and 500 nm.

The effects of inflammation on 11βHSD1 activity in the perfused hindlimb were assessed 6 h after ip administration of LPS (10 mg/kg) or vehicle (physiological saline; 20 ml/kg).

To assess the effects of in vivo LPS on aortic 11βHSD activities, aortic rings were obtained from C57B6J mice that had received either LPS (10 mg/kg ip) or vehicle 6 h before the animals were killed. 11β-Reductase and dehydrogenase activities were determined, ex vivo, by incubation with 10 pmol [³H]₄-11-dehydrocorticosterone or [³H]₄-corticosterone for 24 h, as described above.

Effects of vascular injury on 11βHSD1 enzyme activity in femoral arteries

Femoral artery injury was performed in anesthetized (halothane) C57B6J mice using the method of Sata et al. (39). Briefly, a guidewire (0.014 in. diameter; Cook Inc., UK) was advanced from the descending genual artery to more than 5 mm along the femoral artery, left in place for 1 min, and then removed. The genual artery was ligated, blood flow restored to the femoral artery, and the skin incision closed with 6–0 silk sutures. For histological evaluation, arteries were excised at various time points after injury, fixed in 10% neutral buffered formalin, and embedded in paraffin. Four-micrometer sections were stained with the United States trichrome stain (40).

To assess the effects of vascular injury on basal 11βHSD1 activity, injured and contralateral uninjured femoral arteries (from the descending genual artery to the bifurcation of the iliac artery, therefore containing the entire injured section of vessel) removed 7 d after surgery were incubated in vitro with vehicle for 16 h after which 11β-reductase activity was assessed as described above. In addition, to examine the effect of cytokines on 11βHSD1 activity in injured arteries, injured and contralateral uninjured femoral arteries were removed 7 d after injury and incubated in vitro with IL-1β (10 ng/ml) for 16 h after which 11β-reductase activity was assessed as described above.

Data analysis and statistics

In vitro experiments were performed in duplicate or triplicate and the mean in each experiment used in statistical analyses. Enzyme activity was expressed per 1.75 × 10⁵ cells or, in the case of intact tissue, per wet weight of tissue after subtraction of apparent conversion in negative control samples. Data are expressed as mean ± sem and analyzed by Student’s t test or ANOVA followed by post hoc tests where appropriate.

Results

Effects of IL-1β on 11βHSD activity in cultured mouse aortic smooth muscle cells

As previously reported in human SMCs (24), IL-1β increased 11β-reductase activity by approximately 40% (87 ± 2% conversion in 24 h), compared with controls (62 ± 3%; P < 0.05, n = 6).

Effects of cytokines on 11βHSD activity in intact vascular tissue in vitro

11β-Reductase and dehydrogenase activities were detected in aortic rings and femoral arteries from C57B6J mice (Fig. 1). 11β-Reductase activity was higher in aorta than femoral artery (P < 0.005), whereas 11β-dehydrogenase activity was similar in aorta and femoral artery (Fig. 1, P = 0.48; n = 6). 11β-Reduction was the predominant reaction direction, and was completely abolished in arteries from 11βHSD1^−/− mice (<0.1pmol/mg per 24 h, n = 6).

Fig. 1.

11βHSD activities in intact vascular tissues in vitro. 11β-Reductase (open bars) and dehydrogenase (closed bars) activities in intact arteries from C57B6J mice. Reductase and dehydrogenase activities are expressed as the amount of [³H]₄-corticosterone or [³H]₄-11-dehydrocorticosterone formed, respectively. Results are mean ± se, n = 6. 11β-Reductase activity was higher than dehydrogenase activity (***, P < 0.0001 for aorta; *, P < 0.05 for femoral artery). 11β-Reductase activity was higher in aorta than femoral artery (**, P < 0.005).

11β-Reductase activity in aortic rings from wild-type mice (8.30 ± 1.20 pmol product/mg per 24 h, n = 18) was unaffected after incubation with the proinflammatory cytokines IL-1β and TNFα or the antiinflammatory cytokines IL-4 and IL-13 (Fig. 2). To exclude the possibility that 11βHSD1 in aortic rings was already maximally induced by endogenous TNFα, rings were incubated with the TNFα antagonist, Etanercept, which had no effect on 11β-reductase activity (Fig. 2). 11β-Dehydrogenase activity in wild-type aortic rings (0.21 ± 0.08 pmol product/mg per 24 h) was unaffected by preincubation with either IL-1β (10 ng/ml; 0.33 ± 0.04 pmol product/mg per 24 h) or TNFα (100 ng/ml; 0.30 ± 0.01 pmol product/mg per 24 h, P = 0.25, n = 4).

Fig. 2.

Effect of cytokines on 11β-reductase activity in the mouse aorta. Aortic rings were incubated for 48 h with TNFα, IL-1β, IL-4, IL-13, Etanercept, or vehicle and then for a further 24 h in the presence of [³H]₄-11-dehydrocorticosterone. 11β-Reductase activity is expressed as [³H]₄-corticosterone formed relative to activity in control incubations without cytokine manipulation. Results are mean ± se. There were no differences between groups (n = 4–10).

Effect of LPS in vivo on 11βHSD1 activity in aorta and perfused murine hindlimb

In perfused murine hindlimb, both reductase and dehydrogenase activities were detected, with reductase predominating by approximately 10:1 (Fig. 3A). In 11βHSD1^−/− mice, reductase activity was abolished (Fig. 3A). The apparent Michaelis constant was 1.064 μm for 11β-reductase activity in the perfused hindlimb (Fig. 3B). Ex vivo assays of hindlimb tissues confirmed reductase activity in the hindlimb vasculature (11.4 ± 31.4 pmol/mg per 24 h), which was substantially greater than in skeletal muscle (0.14 ± 0.02 pmol/mg per 24 h; P < 0.0001) or skin with sc fat (0.27 ± 0.05 pmol/mg per 24 h; P < 0.0001, n = 4).

Fig. 3.

11βHSD activities in the perfused murine hindlimb. A, 11β-Reductase (closed squares) and dehydrogenase (closed triangles) activities in perfused hindlimbs of C57B6J mice and 11β-reductase activity in perfused hindlimbs of 11βHSD1^−/− mice (open circles). Enzyme activity is expressed as the percentage of [³H]₄-corticosterone or [³H]₄-11-dehydrocorticosterone formed. 11β-Reductase activity was abolished in 11βHSD1^−/− mice. Results are mean ± se, n = 3–4; **, P < 0.005, comparing reductase activity in 11βHSD1^−/− mice with controls. B, Kinetics of 11β-reductase activity in perfused hindlimbs of wild-type C57B6J mice. Using a time point of 11 min after the commencement of the perfusion, Michaelis constant was calculated from this Lineweaver-Burke plot as 1.064 μm and maximum velocity 313 pmol/min. Results are mean ± se, n = 3 at each concentration. A, 11-Dehydrocorticosterone.

Intraperitoneal LPS administration to wild-type mice induced weight loss (1.4 ± 0.2 vs. 0.7 ± 0.2 g, P < 0.05, n = 6) and splenomegaly (0.12 ± 0.01 vs. 0.08 ± 0.01 g, P < 0.05, n = 6), compared with vehicle. LPS induced a small increase in 11β-reductase (P < 0.05, n = 6) but not dehydrogenase (P = 0.16, n = 3) activity in aortic rings analyzed ex vivo (Fig. 4A) but did not significantly increase 11β-reductase activity in the perfused hindlimb (P = 0.12 by repeated measures ANOVA, n = 6; Fig. 4B).

Fig. 4.

Effects of in vivo LPS on 11βHSD activity in vascular tissue. A, Aortic rings from C57B6J mice were incubated for 24 h in the presence of either [³H]₄-11-dehydrocorticosterone or [³H]₄-corticosterone, 6 h after in vivo ip administration of LPS (10 mg/kg) or vehicle. 11β-Reductase and -dehydrogenase activities are expressed as the amount of [³H]₄-corticosterone or [³H]₄-11-dehydrocorticosterone formed, respectively. Results are mean ± se. 11β-Reductase activity was increased after LPS injection (*, P < 0.05, n = 6). 11β-Dehydrogenase activity was similar in both groups (P = 0.16, n = 3). B, 11β-Reductase activity in perfused hindlimbs of C57B6J mice 6 h after in vivo ip administration of LPS (10 mg/kg, closed circles) or vehicle (open squares). 11β-Reductase activity is expressed as the amount of [³H]₄-corticosterone formed. Results are mean ± se. There were no differences between groups (P = 0.12 by repeated-measures ANOVA, n = 6).

Influence of vascular injury on 11βHSD activity in femoral artery

Vascular injury produced extensive stretching of the vascular wall and atrophy of medial SMCs, with subsequent development of a proliferative, smooth-muscle rich neointima, evident at 7 d and peaking in size at 21 d (Fig. 5).

Fig. 5.

Development of neointimal lesions over time after intraluminal injury of the femoral artery. After wire injury of the femoral artery, arteries were excised at the time points indicated and stained with the United States trichrome stain. This indicated initial stretching of the vessel wall, followed by the time-dependent formation of neointimal lesions, which are first observed at 7 d and peak in size after 21 d. A control uninjured vessel is also shown. Scale bar, 100 μm. Arrowheads indicate internal elastic lamina.

11β-Reductase activity was not altered in femoral arteries after vascular injury in vivo (P = 0.33, n = 6; Fig. 6). Furthermore, vascular injury did not alter the effect of in vitro incubation with IL-1β on 11β-reductase activity (P = 0.20, n = 6; Fig. 6).

Fig. 6.

Influence of injury and inflammation on 11β-reductase activity in femoral arteries. Whole injured and uninjured femoral arteries obtained 7 d after surgery were incubated for 16 h with 10 ng/ml IL-1β or vehicle and then for a further 24 h in the presence of [³H]11-dehydrocorticosterone. 11β-Reductase activity is expressed as the amount of [³H]corticosterone formed. Results are mean ± se. There were no differences between groups (P = 0.33 for effect of injury, P = 0.20 for effect of cytokine, n = 6).

Discussion

We and others previously established that both isozymes of 11βHSD are present in the vascular wall (17–21), in which they may regulate local glucocorticoid availability and hence activation of glucocorticoid and mineralocorticoid receptors. Here we assessed whether vascular inflammation increases local generation of active glucocorticoid by selectively altering the activity of 11βHSD1. Our results indicate that, in contrast to cultured SMCs (24), 11βHSD1 in intact arteries is not up-regulated by proinflammatory cytokines in vitro or in vivo or the inflammatory response to injury in vivo. This suggests that endogenous glucocorticoid generation in the murine arterial wall does not provide a mechanism for feedback regulation of inflammation.

Although both isozymes of 11βHSD are expressed in murine vessels (21, 22), the net balance between active 11-OH and inactive 11-keto steroids was unpredictable. Whereas 11βHSD2 is an exclusive dehydrogenase and 11βHSD1 a predominant reductase, 11βHSD1 might operate as a dehydrogenase in the absence of cofactor generated by the neighboring hexose-6-phosphate dehydrogenase in the endoplasmic reticulum (41, 42). Here we show that the predominant reaction direction, by approximately 10-fold, in intact vessels is the reductase and that this can be accounted for entirely by 11βHSD1 because it is absent in 11βHSD1^−/− mice. Moreover, the kinetics of reductase activity in the perfused hindlimb are consistent with the 11βHSD1 isozyme (43). Furthermore, whereas dehydrogenase activity was similar in all vessels examined, regional differences in basal 11β-reductase activity (44) were suggested, with activity higher in aortae than femoral arteries. Our studies of regulation during inflammation were therefore focused on the reductase activity attributable to the 11βHSD1 isozyme.

Consistent with a previous report in human cultured SMCs (24), 11β-reductase activity in cultured murine vascular SMCs was up-regulated after exposure to the proinflammatory cytokine, IL-1β. The extent of up-regulation (40% increase in activity), however, was less dramatic than the 5-fold enhancement reported with human SMCs (24) and was difficult to replicate at later passages (data not shown). This may be indicative of a species difference in the regulation of 11βHSD1 activity by cytokines in cultured SMCs.

In contrast to cultured MA-SMCs, vascular 11βHSD1 reductase activity was not up-regulated by several different cytokines (IL-1β, IL-4, IL13, TNFα) in intact arteries in vitro. This lack of effect was not due to preexisting up-regulation of 11HSD1 activity because the TNFα antagonist, Etanercept, also had no effect on glucocorticoid generation. In addition, IL-1β and TNFα failed to produce the down-regulation of 11β-dehydrogenase activity reported in human SMCs (24).

In contrast to our in vitro findings, there was a small (18%), selective increase in 11β-reductase activity in aortic rings from mice that had received LPS in vivo. Thus, although individual cytokines were ineffective at enhancing 11βHSD1 activity in intact tissue, the result of in vivo LPS may be to produce an altered inflammatory milieu that favors a modest increase in 11β-reductase activity. Although this effect of LPS was not evident in perfused hindlimbs, there was a similar trend toward an increase in 11β-reductase activity, raising the possibility that there is a small effect of LPS on hindlimb vascular 11βHSD1 activity, which may be masked by the contributions from other tissue types within the regional perfused territory. Additionally, there may be regional differences in the inflammatory regulation of vascular 11βHSD1 activity. It is uncertain whether the resultant change in glucocorticoid availability after these modest effects would have physiologically relevant consequences, but importantly the magnitude of effect is substantially smaller than that observed in cultured cells (24).

The majority of reports detailing regulation of 11βHSD1 activity by inflammatory cytokines (in a variety of different cell types) demonstrate a selective increase in 11β-reductase activity and/or expression (24–31). This effect is not universal, however, because TNFα has no effect on 11β-reductase activity in cultured human hepatocytes (28). Furthermore, in circulating monocytes, 11βHSD1 expression is not up-regulated by TNFα or IL-1β but is induced during differentiation into macrophages and after exposure to IL-4 and IL-13 (25). All studies reported to date used cell culture systems, which undoubtedly alter the natural cell phenotype. The differences that we observed in the ability of cytokines to up-regulate 11βHSD1 activity in cell culture but not intact tissue preparations suggests that the regulation of 11βHSD1 by inflammatory stimuli may not only be tissue specific but may also depend on the degree of cell proliferation and/or differentiation. In vascular lesions, contractile SMCs dedifferentiate and take on a proliferative phenotype (45). Thus, the absence of inflammatory up-regulation of 11βHSD1 in healthy intact arteries cannot be extrapolated to arteries undergoing proliferative remodeling. This issue was addressed using the model of injury/proliferation in the mouse femoral artery.

Insertion of a wire in the femoral artery produces extensive stretching of the vascular wall followed by acute inflammation and the temporal development of a proliferative, smooth muscle-rich neointima (39). In our hands small lesions are first evident 7 d after injury and peak in size after 21 d. This is consistent with the pattern of remodeling reported by Sata et al. (39), in which significant medial and neointimal SMC proliferation was demonstrated at 7 d. Even under these circumstances, however, 11βHSD1 activity was not increased in the femoral artery wall. Furthermore, the remodeling process did not increase the sensitivity of 11βHSD1 to up-regulation by IL-1β. This supports the concept that inflammatory regulation of 11βHSDs in whole-tissue preparations, even during extensive inflammatory and proliferative remodeling, does not mirror that found in cell culture.

We conclude, therefore, that increased local generation of endogenous glucocorticoids by 11βHSD1 in vascular SMCs does not contribute to regulation of vascular lesion development by feedback inhibition of inflammation in mice.

We are grateful to Jonathan Seckl, John Mullins, and Janice Patterson for provision of 11βHSD1^−/− animals.

Disclosure Summary: A.R.D., L.J.M., E.M., and D.E.N. have nothing to declare. B.R.W. has received research funding from Biovitrum and Ipsen and lecture or consultancy fees from Abbott, Amgen, Biovitrum, Bristol Myers Squibb, Incyte, Johnson and Johnson, Merck, Syrrx, and Vitae. B.R.W. and P.W.F.H. are inventors on relevant patents owned by the University of Edinburgh, Edinburgh, United Kingdom.

Abbreviations:

 
• FCS,

  Fetal calf serum;

 
• 11βHSD,

  11β-hydroxysteroid dehydrogenase;

 
• LPS,

  lipopolysaccharide;

 
• MA,

  murine aortic;

 
• SMC,

  smooth muscle cell.