| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
-induced microcirculatory dysfunction and apoptotic cell death
Institute for Clinical and Experimental Surgery, University of Saarland, D-66421 Homburg/Saar, Germany
1Department of Experimental Surgery, University of Rostock, D-18055 Rostock, Germany. E-mail: brigitte.vollmar{at}med.uni-rostock.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-induced injury by attenuating inflammation-associated microcirculatory dysfunction and leukocytic response. Mechanisms of protection, however, are not fully understood. We studied whether the metabolites of the HO and NOS pathway, exerting potent vasodilatory, antioxidant, and anti-apoptotic properties, are involved in tissue cryoprotection. In animals pretreated with L-NAME or SnPP-IX, cooling-associated abrogation of TNF--induced microcirculatory dysfunction was abolished. Combined L-NAME/SnPP-IX pretreatment did not cause greater blunting than seen when each mediator system was inhibited separately. In SnPP-IX- but not L-NAME-pretreated animals, transient hypothermia failed to reduce TNF--mediated leukocyte adherence. Vice versa, treatment of TNF--exposed animals with either the NO donor L-arginine or the HO-1 inductor hemin mimicked cooling-associated tissue protection except for failure of L-arginine to abrogate the inflammatory leukocyte response. The efficiency of cooling to inhibit TNF--induced apoptotic cell death was blunted in SnPP-IX-, L-NAME-, and SnPP-IX/L-NAME-pretreated animals. Coadministration of Trolox in SnPP-IX-treated animals partly attenuated leukocyte adherence and cell apoptosis, implying that the HO pathway metabolite biliverdin contributes to the salutary effects of cooling. Thus, our study provides evidence that metabolites of the HO and the NOS pathway mediate the cooling-associated protection of inflamed tissue. Biliverdin rather than CO and NO mediates the anti-inflammatory action, whereas a coordinated function of the gaseous monoxides prevents microcirculatory dysfunction and apoptotic cell death.—Amon, M., Menger, M. D., Vollmar, B. Heme oxygenase and nitric oxide synthase mediate cooling-associated protection against TNF--induced microcirculatory dysfunction and apoptotic cell death.
Key Words: cooling • inflammation • apoptosis • TNF- • nitric oxide • carbon monoxide • mice • microcirculation • leukocyte–endothelial cell interaction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3000–3500 BC) recommended the use of cold applications on wounds of the head and on ulcerated breast (1)
. Perhaps the first application in trauma was by Patroklos in Homer’s Iliad, who dressed his wounded leg with cold water packs (2)
. Snow and ice have always been used to relieve pain, stanch bleeding, improve performance, and hasten healing (2)
. In recent years, cold therapy has been introduced in sports traumatology as an essential part of the first acute treatment phase of soft tissue injuries (3
, 4)
. Treatment is inexpensive, easily available, humane, and harmless; however, like any other therapeutic measure, cold must be administered with understanding and skillful management.
The pathophysiology of soft tissue injury comprises elevated cellular metabolism, hemorrhage, hyperemia, edema, and leukocyte recruitment. It is reasonable to speculate that lowering skin temperature decreases cutaneous blood supply but also reduces metabolic demands of the tissue, thereby preventing edema formation and preserving tissue viability. Despite transient tissue hypoxia, cooling does not elicit an inflammatory cell response, as recognized from hypoxia in conditions of warm ischemia and reperfusion (5)
. We recently confirmed the beneficial effects of cooling by demonstrating that in inflamed TNF--exposed tissue, transient local hypothermia reduces nutritive capillary perfusion failure, limits leukocyte response, macromolecular leakage, and abolishes apoptotic cell death (6)
. Although the local cooling procedure evoked a clear cut arteriolar vasoconstriction with reduced microcirculatory blood perfusion during hypothermia, the termination of cooling with rewarming of tissue was associated with complete restoration of the microcirculation, vasodilation, and reactive hyperemia (6)
. Mechanisms underlying the cryoprotection of inflamed tissue, however, have not been addressed in detail and are incompletely understood.
There is increasing evidence indicating that the gaseous monoxides nitric oxide (NO) and carbon monoxide (CO) play a substantial role as regulators of vascular tone (7
, 8)
. NO is enzymatically produced from L-arginine by nitric oxide synthase (NOS), an enzyme present as various inducible and constitutive isoforms in smooth muscle, endothelial cells, macrophages, and other cell types (7)
. NO mediates many of its biological actions by activation of the soluble guanylate cyclase (sGC) pathway and production of cyclic guanosine monophosphate (cGMP). cGMP in turn inhibits intracellular calcium release and causes smooth muscle relaxation (7)
.
The main endogenous source of CO is heme oxygenase (HO), which exists in constitutive (HO-2 and HO-3) and inducible (HO-1) isoforms (8)
. HO catalyzes the rate-limiting step in the conversion of heme to iron, biliverdin-IXa, a potent antioxidant, and CO, a putative vasodilator (8)
. Many properties pertaining to CO have strong analogies with the well-established biological activities elicited by NO such as control of vascular tone and microvascular blood flow, inhibition of cell adhesion and aggregation, and modulation of apoptotic cell death (7
, 8)
.
Though the physiological and pathophysiological functions of both enzyme systems have been well defined in a variety of acute and chronic diseases, their obvious role in tissue cryoprotection has never been addressed. To investigate the role of mediator systems in prevention of tissue injury by cooling, the pathways were inhibited with either the specific blocker of the HO pathway SnPP-IX or the false L-arginine analog NG-nitro L-arginine methyl ester (L-NAME). To further confirm the specificity of SnPP-IX and L-NAME in cooling-associated tissue protection, supplementation of NO donor and induction of HO-1 by hemin were tested as to whether they can elicit specific activities in TNF--inflamed tissue reminiscent of those mediated by cooling. In other experiments, the antioxidant Trolox (vitamin E analog) was administered to better distinguish the role of CO and biliverdin in this model of tissue cryoprotection.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental model
The dorsal skinfold chamber in mice was used for intravital microscopy, as described (9)
. Mice were anesthesized intraperitoneally with a mixture of ketamine (90 mg/kg bw) and xylazine (25 mg/kg bw), and two symmetric titanium frames were implanted to sandwich the extended double layer of the skin. One layer was removed in a 15 mm diameter circular area. The remaining layer consisted of epidermis, subcutaneous tissue, and striated skin muscle and was covered with a glass coverslip incorporated in one of the titanium frames. Animals tolerated the chamber well and showed no signs of discomfort or changes in sleeping and feeding habits. A recovery period of at least 4 days was allowed before intravital observation.
Intravital fluorescence microscopy
By use of a modified fluorescence microscope with a 100-W HBO mercury lamp (Axiotech, Zeiss, Jena, Germany) attached to an UV (330–380/>415 nm excitation/emission wavelength), blue (450–490/>515 nm), and green (525–555/>580 nm) filter system, striated muscle microcirculation was analyzed in an epi-illumination technique. Microscopic images were recorded by a CCD video camera (FK 6990, COHU, Prospective Measurements Inc., San Diego, CA, USA), transferred to a video system (S-VHS Panasonic AG 7350, Matsushita, Tokyo, Japan), and recorded on videotape for subsequent off-line evaluation. Animals received a tail vein injection of 0.05 mL FITC-dextran (50 mg/mL saline; Sigma Chemical Co., St. Louis, MO) and 0.05 mL rhodamine 6G (0.1 mg/mL saline; Sigma) for vascular contrast enhancement and leukocyte staining in vivo. This allowed us to quantitatively analyze arteriolar, capillary and venular perfusion as well as leukocyte flow and the leukocyte–endothelial cell interaction. Nuclei of tissue cells were visualized in vivo by topical application of 0.1 mL bisbenzimide H33342 (1 mg/mL saline; Sigma). Using water immersion objectives (W 20x/numerical aperture (NA) 0.5; W 40x/NA 0.75; Zeiss, Jena, Germany), magnifications of x750 and x1500 were achieved on the video screen (PVM-2130 QM, Sony, Munich, Germany).
Quantitative analysis of striated muscle microcirculation
At the first observation, the chamber was scanned with a x2.5/NA 0.08 objective (Leitz, Wetzlar, Germany) for random selection of distinct observation areas, which included second- and/or third-order arterioles, nutritive capillaries, and draining postcapillary venules. Video printouts were made during videography and initially marked to indicate the exact localization for measurements of vessel diameter, red blood cell velocity, volumetric blood flow, and functional capillary density. Using a computer-assisted image analysis system (Capimage, Zeintl Software, Heidelberg, Germany), functional capillary density was assessed at x750 magnification as the length of red blood cell-perfused capillaries per observation area (cm/cm2). Centerline red blood cell velocity was analyzed by computer-assistance using the line-shift method (CapImage). Volumetric blood flow (VQ) was calculated in arterioles and venules at x750 magnification from red blood cell velocity (V) and vessel cross-sectional area (
x r2) according to the equation VQ = V x x r2, assuming a cylindrical vessel geometry. The number of permanent adherent leukocytes (defined as cells that adhered to the microvascular endothelium over a period of 20 s) was evaluated at x750 magnification as number of cells per mm2 venular endothelial surface (calculated from diameter and length of the vessel segment studied, assuming cylindrical geometry). Topical application of the vital dye bisbenzimide H33342 allowed us to assess in vivo cell nuclear morphology with apoptosis-associated condensation, fragmentation, and crescent-shaped formation of chromatin (10)
. Using the x40 objective for recording (magnification x1500), cells exhibiting these apoptotic characteristics were counted and are given as number per observation field.
Chemicals
The false L-arginine analog L-NAME (Sigma), an inhibitor of NO synthesis, was dissolved in sterile saline at room temperature to a final concentration of 5 mg/mL. Tin-protoporphyrin-IX (SnPP-IX; Frontier Scientific, Lancashire, UK), an inhibitor of HO and thus CO and biliverdin release, was dissolved in 8.4% sodium-bicarbonate and phosphate-buffered saline (PBS) to achieve a final concentration of 5 µmol/mL. The solution was stored at a maximal temperature of 8°C in the dark and used within the next hour. The NO donor L-arginine (Sigma) and the water-soluble vitamin E analog Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carbonic acid; Sigma-Aldrich, Steinheim, Germany) were dissolved in sterile saline and in PBS to a final concentration of 10 and 20 mg/mL, respectively. Hemin, an HO-1 inductor (Fluka, Steinheim, Germany), was dissolved in DMSO to a final concentration of 5 µmol/mL. All solutions were freshly prepared at the day of the experiment according to the manufacturers’ directions. 2000 units of mouse TNF- (Roche, Mannheim, Germany) were dissolved in 100 µL sterile PBS and stored at -20°C until use.
Experimental protocol
After baseline microscopy, chambers were exposed to TNF- to induce a sustained and marked inflammatory response (6)
. Microscopic recordings with microcirculatory analysis were repeated 30, 60, 90, 120, and 180 min after TNF- exposure. The number of apoptotic cells was evaluated 24 h later. Local cooling of the chamber to 8–10°C was performed by immersion with ice-cold saline directly after TNF- exposure of the chamber and terminated after 30 min. During cooling, surface temperature was continuously monitored using a LICOX type K thermocouple probe (GMS, Kiel-Mielkendorf, Germany).
Animals were assigned to the following treatment regimens: 1) TNF- and local cooling (n=6; cooling); 2) L-NAME pretreatment, followed by TNF- and local cooling (n=6; L-NAME); 3) SnPP-IX pretreatment, followed by TNF- and local cooling (n=6; SnPP-IX); and 4) L-NAME and SnPP-IX pretreatment, followed by TNF- and local cooling (n=6; L-NAME/SnPP-IX). SnPP-IX (50 µmol/kg bw i.p.) and L-NAME (50 mg/kg bw i.v.) were applied 24 h and 15 min before TNF- exposure, respectively. TNF--exposed animals without cooling served as controls (n=6; control).
To further address the specificity of L-NAME and SnPP-IX to block the respective enzyme systems, additional animals underwent the following treatment: 5) L-arginine pretreatment, followed by TNF- exposure without local cooling (n=6; L-arginine); 6) hemin pretreatment, followed by TNF- exposure without local cooling (n=6; hemin). The role and effect of reactive oxygen species in TNF--induced tissue injury was addressed by 7) pretreatment of animals with Trolox and subsequent TNF- exposure without cooling (n=6; Trolox) as well as by 8) coadministration of Trolox in SnPP-IX-treated animals, followed by TNF- exposure and local cooling (n=6; SnPP-IX/Trolox). Hemin (50 µM/kg bw i.p.) was applied 18 h before TNF- exposure whereas L-arginine (100 mg/kg bw i.p.) and Trolox (20 mg/kg bw i.p.) were administered 15 min before TNF- exposure.
Statistical analysis
Data are given as mean ± SE. After proving the assumption of normality, comparisons between the experimental groups were performed by one-way ANOVA, followed by the appropriate post hoc comparison. For inner group comparison, data were analyzed by one-way ANOVA for repeated measurements, followed by the Student-Newman-Keuls test for paired samples (SigmaStat, Jandel, San Rafael, CA). Statistical significance was set at a P value of < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
exposure exposure caused a decrease of volumetric blood flow in arterioles and venules to 60–70% of baseline values (Fig. 1A
and Fig. 2A
). This was primarily due to a reduction in red blood cell velocity (Table 1
). In contrast, arteriolar and venular diameters were almost unchanged (Table 2
). As a consequence, nutritive tissue perfusion as given by the functional capillary density was markedly deteriorated, demonstrating an average value of only 60% of baseline at the end of the 180 min observation period (Fig. 3A
). TNF- exposure progressively increased leukocyte adherence to the venular endothelial lining, indicating a pronounced inflammatory response (Fig. 4A
). Finally, tissue injury could be documented by a 10-fold and >20-fold increase of apoptotic cell death 180 min and 24 h after TNF- exposure (Fig. 5A
). TNF--associated apoptotic tissue injury was indicated by a high number of detached cells with nuclear condensation and fragmentation (anoikis). Those cells were primarily observed at the end of the 24 h observation period (Fig. 6A
).
|
|
|
|
|
|
|
|
TNF- exposure and cooling
Reduction of tissue temperature to 8–10°C for the first 30 min after TNF- exposure induced an initial arteriolar vasoconstriction (Table 2)
, which was associated with a decrease of arteriolar and venular volumetric blood flow and a diminution of functional capillary density (Table 1
, Figs. 1A
2
3A
). Upon termination of cooling and passive rewarming of the tissue (30–180 min), however, there was a complete restoration of microhemodynamics and nutritive perfusion to baseline levels (Figs. 1A
2
3A
), indicating successful abrogation of the TNF--induced microcirculatory disorders. Cooling also dampened the TNF--associated venular leukocyte–endothelial cell interaction (Fig. 4A
) and almost completely abolished apoptotic tissue injury (Figs. 5A
and 6A
).
TNF- exposure and cooling in L-NAME-, SnPP-IX-, and L-NAME/SnPP-IX-pretreated animals
Animals pretreated with either L-NAME, SnPP-IX, or L-NAME and SnPP-IX showed a marked arteriolar vasospasm (Table 2)
and a restriction of arteriolar, capillary, and venular perfusion after 30 min of TNF- exposure and cooling. However, this did not recover upon tissue rewarming (Table 1
, Fig. 1B-D
, Fig. 2B—D
, and Fig. 3B-D
). Despite some minor additive effects seen in the animals pretreated with L-NAME and SnPP-IX, the L-NAME, SnPP-IX, and L-NAME/SnPP-IX groups did not differ markedly with respect to lack of vasoreactivity and microcirculatory recovery. These findings imply that HO and NOS are needed to mediate the cooling-associated protection against TNF--induced microcirculatory dysfunction.
The cooling-associated anti-inflammatory effect was almost unaffected by pretreatment with L-NAME (Fig. 4B
). In contrast, HO blockade by SnPP-IX abolished the anti-inflammatory effect of cooling (Fig. 4C
), as indicated by some adherent leukocytes even higher than observed in TNF--exposed animals without cooling (Fig. 4A
). Leukocyte adherence in L-NAME- and SnPP-IX-pretreated animals was enhanced (Fig. 4D
); the enhancement was slightly less pronounced than that in solely SnPP-IX-pretreated animals (Fig. 4C
) but corresponded well with the extent of leukocyte adherence observed in TNF--exposed animals without cooling (Fig. 4A
).
Finally, in vivo analysis of apoptosis revealed that pretreatment with either L-NAME, SnPP-IX, or both markedly counteracted the cooling-associated protection against TNF--mediated apoptotic cell death. At 24 h after TNF- exposure, chamber tissue showed a 8- to 14-fold increase of cells that exhibited nuclear condensation and/or fragmentation (Fig. 5B-D
). As a consequence, pretreatment with either L-NAME, SnPP-IX, or both abolished the cooling-associated reduction of apoptotic cells with loss of cell–cell contact and detachment from tissue (Fig. 6B-D
).
TNF- exposure in L-arginine and hemin-pretreated animals
Cooling-induced restoration of microcirculation could be mimicked in noncooled TNF--exposed animals by pretreatment with L-arginine or hemin (Fig. 7A, B
, Fig. 8A, B
). However, L-arginine supplementation in TNF--exposed animals without cooling revealed comparably high numbers of adherent leukocytes as observed in control animals, underlining the lack of anti-adhesive properties of NO in this experimental setting (Fig. 7C
). In contrast to NO donor supplementation, HO-1 induction by hemin could sufficiently block TNF--induced leukocyte activation (Fig. 8C
), similar to that observed upon cooling of tissue. Apoptotic cell death (Figs. 7D
and 8D
) and anoikis (data not shown) upon TNF- exposure were found to be partly limited by hemin and almost completely prevented by L-arginine pretreatment, respectively.
|
|
TNF- exposure and cooling in SnPP-IX/Trolox-pretreated animals
To differentiate between the contribution of CO and biliverdin in tissue cryoprotection, other animals were pretreated with SnPP-IX and Trolox. Coadministration of Trolox and SnPP-IX was not able to restore microcirculatory dysfunction, i.e., arteriolar and venular blood flow as well as functional capillary density (Fig. 9A, B
), but reduced leukocyte adherence and apoptotic cell death associated with the blockade of the HO pathway in animals upon TNF- exposure and cooling (Fig. 9C, D
). In parallel, treatment of noncooled TNF--exposed animals with Trolox failed to normalize microcirculatory dysfunction but could limit leukocyte adherence and cell apoptosis (data not shown), as observed after SnPP-IX and Trolox treatment in those animals exposed to TNF- and cooling (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-induced tissue injury. Although the inflammatory microvascular leukocyte–endothelial cell interaction is down-regulated predominately by the action of the antioxidative biliverdin, preservation of the nutritive microcirculation and prevention of apoptotic cell death seem to require the function of diatomic gases CO and NO.
Methodological considerations
Using the skinfold chamber preparation in mice, we were able to directly monitor the diverse effects of cooling on inflamed tissue microcirculation. The delay of at least 4 days between chamber preparation and the start of the experiment allowed the tissue to recover from surgical trauma and thus avoided interference with microvascular dysfunction mediated by trauma-induced TNF--release (11)
. Intravital fluorescence microscopy of chamber tissue in PBS-exposed animals revealed that repeated manipulation, e.g., removal of the glass coverslip for topical application of TNF- and bisbenzimide, does not alter striated muscle microcirculation (data not shown). By studying the microvascular response to short periods of ischemia, we earlier demonstrated that older animals reveal a restricted capacity of the microcirculation to produce adequate reactive hyperemia (12)
. Therefore, the age of the animals used in the present study was restricted to between 6 and 8 wk. As all animal measurements wereconfined to identical vessels throughout the experimental period, we could further reduce potential variation of data due to intraindividual biological heterogeneity to a minimum.
In contrast to whole body hypothermia used by others in previous studies to characterize cold shock-induced systemic cellular stress response (13
, 14)
, we aimed at studying local cooling of tissue to a temperature of 8–10°C without affecting the core body temperature. Thereby, systemic influences such as hypotension, rigidity, and stiffness of circulating leukocytes (15)
caused by whole body hypothermia could be avoided and allowed us to strictly evaluate tissue-confined local response to cold therapy.
Cooling-associated protection against TNF--induced inflammatory response and microcirculatory dysfunction
Inflammation is a complex cellular and biochemical response to injurious stimuli and is regulated by an extensive network of cytokines. TNF- is best known to be released immediately after a local injury to tissue and to play multiple roles in acute phase responses, including leukocytic recruitment and endothelial adhesiveness (16
, 17)
. TNF- also affects the capability of vascular endothelium to release various mediators. Using the TNF--treated striated muscle as a well-characterized model of acute cytokine-dependent inflammation (6
, 18
, 19)
, we could demonstrate microvascular and tissue protection by local hypothermia. Although research in this area has been conducted for many decades, the mechanisms of protection are unclear and still a matter of controversy. In line with previous reports by our group (5
, 6)
, the cooling-associated protective effects have been ascribed to a decrease in metabolic rate of tissue, restoration of blood flow, preservation of endothelial barrier function, inhibition of apoptosis, and attenuation of inflammatory response. The present study is significant because it provides a direct investigation on whether the cooling-associated beneficial effects involve the HO and NOS pathways, both releasing endothelial mediators with potent action on vasomotor control, vascular reactivity, and endothelial function.
Role of HO and NOS in cooling-associated protection against TNF--induced microcirculatory dysfunction
Cooling is associated with a transient but strong vasoconstriction, causing tissue hypoxia. Most tissues are able to respond to a short period of hypoxia by vasodilation with a transient rise in blood flow, termed reactive hyperemia. Although the present experimental setting is not directly comparable to established models for the study of reactive hyperemia (e.g., the human forearm), we demonstrate here that systemic administration of the NO synthesis inhibitor L-NAME is able to blunt the change in vascular reactivity upon cooling and rewarming of TNF--exposed tissue, similar to what has been described for reactive hyperemia in canine diaphragm (20)
and hindlimb circulation (21)
. It has been shown that the constitutive isoform of NOS is sensitive to varying oxygen tensions with increased production of NO and formation of cGMP within 30 min after exposure to hypoxia (22)
. Thus, in L-NAME-pretreated animals, the cooling-associated tissue hypoxia may not have been translated into increased NO production to protect against the TNF--induced microcirculatory dysfunction.
Apart from hypoxia, the myogenic response, involving NO function, may be responsible for the cooling-associated protection of the microcirculation. The myogenic response is defined as contraction of a blood vessel, probably by endothelial-derived constricting factor, that occurs when intravascular pressure is elevated and, conversely, the vasodilation (probably by NO) that follows a reduction in pressure (23)
. Thus, the low flow conditions during hypothermia associated with low intravascular pressure may provoke a NO-mediated vasodilation with recovery of the microcirculation, which is abrogated when the animals are pretreated with L-NAME.
Like NO, exogenously applied CO relaxes isolated blood vessels (24)
. A more recent report demonstrated constitutive HO-2 activity in vascular endothelium and found that CO release contributes to endothelium-dependent vasodilation (25)
. The present observations that cooling failed from restoration of tissue perfusion in SnPP-IX- and SnPP-IX/Trolox-pretreated animals indicate that vessel wall-derived CO serves (along with NO) as an endogenous regulator of vascular tone (26)
. Though the HO pathway metabolite biliverdin might indirectly protect microvascular perfusion by its antioxidative property, coadministration of SnPP-IX and Trolox disproved a relevant role of the antioxidative action of biliverdin in mediating cryoprotection against microcirculatory dysfunction.
Despite marked similarities of NOS and HO-2 localization and function in blood vessels (25)
, NO seems unable to take over CO function (and vice versa) in protecting against TNF--induced microcirculatory dysfunction, because blockade of synthesis of either of the diatomic gases abolished the cooling-induced microvascular recovery. Because SnPP-IX can block sGC irrespective of its action on HO (27)
, the above mentioned effects as well as the absence of synergistic actions by SnPP-IX/L-NAME pretreatment might be explained simply by involvement of the NO-cGMP system. However, it has been shown that inhibition of HO activity by SnPP-IX reverses the component of endothelial-derived relaxation of porcine distal pulmonary arteries not reversed by an inhibitor of NOS (25)
. In the present study, not only L-arginine, but also hemin, was able to mimic cryoprotection against TNF--induced microcirculatory dysfunction, disproving a solitary role of the NO-cGMP system in mediating tissue cryoprotection. Thus, these findings underscore the view that NO and CO may play a coordinated role rather than additive or synergistic action in physiological vasomotor control and protection against tissue injury by cooling.
Role of HO and NOS in cooling-associated protection against TNF--induced inflammatory response
With respect to the cooling-associated restriction of microvascular inflammatory response, we could demonstrate that metabolites of the HO rather than of the NOS pathway mediate the anti-adhesive effects upon cooling of TNF--exposed tissue. This is supported by a recent study demonstrating that in rat mesenteric venules, CO via HO-1 induction contributes to the modulation of sevoflurane-induced leukocyte and platelet adhesive interactions, although this study indicated that NO rather than CO may dominate maintenance of local microvascular hemodynamics (28)
.
Although previous studies have shown that L-NAME promotes vascular leukocyte adhesion, indicating anti-adhesive properties of NO, there is an apparent dichotomous role of NO in inflammation, because many studies have indicated promotion of inflammation by NO probably through peroxynitrite production (29
, 30)
. In fact, in the present study the inhibition of NOS did not blunt the cooling-associated protection against leukocyte adherence, indicating that NO may play a minor role in the anti-inflammatory action of cooling. To further clarify this issue, we studied the effect of NO in TNF--inflamed tissue that did not undergo the cooling procedure. In these experiments, application of L-arginine was not able to reduce the TNF--mediated leukocytic response, which also supports the view that, in the experimental setting used in our study, NO may not be anti-inflammatory.
Without denying potential anti-adhesive properties of NO (31)
, the ultimate biologic effects of NO may be due to where NO is produced, the level of NO production, and the environmental milieu into which it is released. Such factors modulate its redox status and determine its nature of action (32)
. Thus, under certain conditions, such as cytokine-induced tissue inflammation and cooling, NO, although sufficient to regulate basal vascular tone, may not be sufficient to inhibit cell activation and adherence. Based on the fact that NO derived from constitutively active eNOS may be involved primarily in the physiological regulation of vascular tone, whereas NO derived from inducible NOS may participate in the inflammatory response (33)
, it might be speculated that the present findings relate more to eNOS than to iNOS.
The predominant role of the HO pathway in the present experimental setting might be due to an abundant induction of HO-1/heat shock protein (HSP)-32 gene upon cold stress, as this has been shown in a model of local cooling under transient spinal cord ischemia (34)
. NO is able to induce HO-1 expression and CO production in vascular smooth muscle cells (35
, 36)
. HO-1 catalyses the oxidative cleavage of heme molecules to biliverdin, CO, and iron, thereby exerting potential protective effects in inflammation (8)
. Biliverdin has both antioxidant and anti-complement effects. CO up-regulates cGMP, induces vasodilation, and inhibits blood cell aggregation and adhesion. Heme degradation by HO leads to ferritin synthesis with sequestration of iron, thus preventing its participation in subsequent oxidative injury (8)
. This multifaceted profile of anti-inflammatory actions of HO-1 might account for the superiority of HO over the NOS pathway to mediate anti-adhesive properties of cooling (37)
. Our data even indirectly indicate some proadhesive potential of NO, which became apparent by 1) increased leukocyte adherence over controls when HO was concurrently blocked, and 2) some amelioration of leukocyte adherence when the NOS pathway as well was inhibited. Upon exposure to low temperature, HO seems to predominately mediate the anti-adhesive protection.
To further distinguish the roles of CO and biliverdin in this model, the effect of cooling was studied in TNF--exposed tissue of animals with coadministration of SnPP-IX and Trolox. This regimen significantly attenuated inflammatory leukocyte response associated with the blockade of the HO pathway, implying that biliverdin contributes to the anti-inflammatory property of the HO pathway in cryoprotection of inflamed tissue.
It is difficult to exclude the contribution of various physical and mechanical factors to the cooling-induced attenuation of leukocyte adherence. Indeed, the function of adhesion molecules, e.g., endothelial E-selectin transcription (38)
and binding of integrin receptors to their ligands (15
, 39)
, has been shown to be temperature dependent and inefficient at low temperatures. Restored capillary perfusion observed after cold therapy may have prevented sustained tissue hypoxia and potential hypoxia-associated leukocyte adherence (5)
. Vasodilation and enhanced flow rate might counteract adhesive interactions with the endothelium. However, the distinct effect of pretreatment with L-NAME, SnPP-IX, SnPP-IX/L-NAME, and SnPP-IX/Trolox underscores the view that cooling-associated tissue protection results from a complex interplay of NO, CO, and biliverdin rather than being the mere net result from changes in rheology and temperature.
Role of HO and NOS in cooling-associated protection against TNF--induced apoptotic cell death
In some models of cerebral ischemia, mild hypothermia has been shown to confer protection against apoptotic cell death (40
, 41)
, perhaps by interrupting endonuclease activity (42)
or activating transcription factors (43)
. We now demonstrate that the anti-apoptotic effect of local hypothermia, including nuclear condensation, fragmentation, and cell detachment from tissue, is significantly counteracted in SnPP-IX- and L-NAME-pretreated animals. This implies that hypothermia mediates its anti-apoptotic action via HO and NOS pathway metabolites. Whereas high toxic levels of NO and CO might promote manifestation of apoptosis (44
, 45)
, it becomes more and more apparent that endogenous synthesis of NO and CO is associated with inhibition of apoptosis. Indeed, NO was demonstrated to inhibit lipopolysaccharide-induced apoptosis in pulmonary artery endothelial cells (46)
. Similarly, TNF--induced apoptosis in cultured fibroblasts was shown to be inhibited by HO-1-dependent CO release (47)
. Considering the pivotal role of reactive oxygen species in mediating apoptosis, one may suggest that the antioxidant biliverdin partly mediates the anti-apoptotic action of the HO pathway, as underscored by the results obtained in animals with SnPP-IX/Trolox coadministration. Finally, the anti-apoptotic effect of shear rate-induced NO (48)
may be another mechanism by which cooling prevents from cell death, because 1) hypothermia-induced vasoconstriction, followed by vasodilation upon tissue rewarming, causes marked shear stress and thus NO release, and 2) L-arginine supplementation was indeed able to almost completely abrogate the TNF--induced apoptotic cell death.
In summary, this study unravels the important role of HO and NOS in cooling-associated tissue protection against TNF--induced inflammation. Protection includes potent vasoactive, anti-adhesive, and anti-apoptotic properties of CO, biliverdin, and NO, which may act in parallel or complementary to each other, but may also function in a coordinated system.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 18, 2002. Accepted for publication October 28, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-induced microvascular perfusion failure, leukocyte adhesion, and apoptosis in the striated muscle. Surgery 126,881-889[Medline]
induced leukocyte recruitment by different mechanisms in vivo and in vitro. J. Immunol. 158,3391-3400[Abstract]
This article has been cited by other articles:
|
|
 |
|
  K. Knobloch, M. Spies, K. H. Busch, P. M. Vogt, K.-D. Schaser, A. C. Disch, and T. Mittlmeier Letter to the Editor * Authors' Response Am. J. Sports Med., September 1, 2007; 35(9): e1 - e3. [Full Text] [PDF]
|
|
|
|
 |
|
 C. Contaldo, C. Meier, A. Elsherbiny, Y. Harder, O. Trentz, M. D. Menger, and G. A. Wanner Human recombinant erythropoietin protects the striated muscle microcirculation of the dorsal skinfold from postischemic injury in mice Am J Physiol Heart Circ Physiol, July 1, 2007; 293(1): H274 - H283. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K.-D. Schaser, A. C. Disch, J. F. Stover, A. Lauffer, H. J. Bail, and T. Mittlmeier Prolonged Superficial Local Cryotherapy Attenuates Microcirculatory Impairment, Regional Inflammation, and Muscle Necrosis After Closed Soft Tissue Injury in Rats Am. J. Sports Med., January 1, 2007; 35(1): 93 - 102. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Kubulus, F. Roesken, M. Amon, M. Rucker, M. Bauer, I. Bauer, and M. D. Menger Mechanism of the delay phenomenon: tissue protec |
P < 0.05 vs. SnPP-IX; n = 6 per group.