Evidence for endothelium-dependent and endothelium-independent vasodilation in human skin flaps HAEL EL L, KREIDSTEIN,~ H y. O '.ANG,' L ~ Y N. D CARISEN, AND NINGXU Research Institute, T%eHospifal for Sick Children, 555 University Avenue, Toronto, O~zt.,Canada M5G 1x8 hltkd
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
Departments
a$
Surgery and Physiobogy, University of Toronto, Toronto, Ont., Canada Received January 22, 1992
KREIDSTEIN, M. L., PANG,C. Y. , CARLSEN, L. N., and Xu, N. 1992. Evidence for endothelium-dependent and endotheliumindependent vasodilation in human skin flaps. Can. J. Physiol. Pharmacol. 70: 1208- 12 16. Acetylcholine (ACh) and nitroglycerin (NTG) were used as probes to study endothelium-dependent and endothelinmindependent vascular relaxation in isolated perfused transverse paraun~bilicalhuman skin flaps. Ht was observed that ACh (18-") significantly ( p < 0.05) decreased the vascular resistance and increased dermal capillary perfusion (assessed by surface fluorometry) in norepinephrine (NE, M) precsnstricted skin flaps, despite the presence of a cyclooxygemse inhibitor (indornethacin, 3 x H8-%) and a fi-adrenergic receptor antagonist (propranolol, 1 0 - w ) . The ability of ACh to induce vascular relaxation in NE-preconstricted skin flaps was lost after damaging the vascular eraelothelial lining with saponin perfusion (100 rng * E-I, 5 min). In contrast, NTG (1W6 M) induced vascular relaxation to a similar extent before and after saponin treatment. In a separate study, ACh was seen to induce vascular relaxation in a concentration-dependent This vascular relaxation effect of ACh over the dose range of manner in sfin flaps preconstrieted with NE (10-%). lo-% was significantly ( p < 0.08) inhibited in the presence of Nu-nitro-L-arginine (los5 M), a nitric oxide (NO) synthesis inhibitor. These observations were taken to indicate the presence of endotkelium-dependent and endotheliumindependent vascular relaxation in human skin flaps and that the ACh-induced endothelium-dependent relaxation is probably mediated by NO. The importance of impairment of endothelium-dependent relaxation in the pathogenesis of skin flap ischemia, and the potential use s f topical anitrovasodilators or NO donors for prevention and (or) treatment sf skin flap ischemia were d s o discussed. Key words: acetylcholine, nitroglycerin, Nu-nitro-L-arginine, vascular relaxation, human skin. KREIDSTEIN,M. IL., PANG,C. Ye, CAWWEN, E. N., et Xu, N. 1992. Evidence for endothelium-dependent and endotheliurnindependent vasodilation in human skin flaps. Can. J. Bhysiol. Bharmacol. 70 : 1208- 1216. On a utilisC 17acCtylcholine(ACh) et la nitroglycCrine (NTG) c o m e sondes pour examiner la relaxation vasculaire dCpendank et indkpendante de 17endothCliumdans des Iamkaux eutanis para-ombilicanx transverses, p e r h d s et isolCs d9humains. L'ACh (IOP6 M) a diminart significativement ( p < 0,05) la resistance vasculaire et augment6 la perhsion capillaire dermique (Cvalute par fluoromCtrie de surface) dans les lambeaux cutanCs prk-comprimCs par la nor6pintphrine (NE, M), malgrt la prCsence d9un irabaibiteur de cyclooxygtnase (indomtthacine, 3 x 10-"M) et d'un antagoniste des rkcepteurs @adrknergiques (propanolol, 10s6 M). L'ACh a perdu sa capacitt d'induire une relaxation vasculaire dans les lambeaux cutanks prd-comprim6s par Ia NE, aprks l'alttration de I'endothtliurn vasculaire par la perfusion de sapnine (100 mg - L-', M) a induit une relaxation vasenlaire un degrC similaire, avant et aprks le traitement 5 min). A 170pposC,la NTG B la saponine. Dans une autre Cmde, I'ACh a induit une relaxation vasculaire, dCpeadante de la concentration, dans les lambeaux cutanCs prd-comprimCs avec la NE M). Cet effet de 17ACh sup la relaxation vasculaire dans la plage de lO-'- lo-" a at6 significativement inhibt ( y < 0,Ol) en prCsence Be Nd-nitro-L-arginine (10-%), un inhibiteur de la synth&sede l'oxyde nitrique (ON). Ces observations ont permis d'indiquer la presence d'une relaxation vasculaire dCpendante et indipendante de 19endothCliumdans les lambeaux cutanCs humains, la relaxation dCpndante de I'endothClium induite par 19AChCtant probablement mkdi6e par 1'8N. On a aussi discutk de l'importance d9unealtiration de la relaxation dtpendante de I9endoh6lium dans la pathogenkse de 19ischCmiede lambeaux cutanCs et de l'utilisation potentielle de nitrovasodiIatateurs locaux oar de donnmeurs 8 N pour la prevention et (ou) le traitement de 17ischCmiede lambeaux cutanis. Mots c&&s: acktylcholine, nitroglycirine, Nu-nitro-L-arginine, relaxation vasculaire, peau humine. [Traduit par la rCdactisn]
Introduction Using acetylcholine (ACh), Furchgott and Zawadzki ((1980) demonstrated endsthelium-dependent relaxation of the rabbit aorta in the presence of cyelooxygenase inhibitors. They introduced the term non-prostansid endsthelibam-derived relaxing factor (EDRF) to describe the substance that caused this vascular relaxation. Subsequently, many advances have been made in this field, including the identification of EDRF as nitric oxide (NO) or a labile nitrosothisl compound capable of spontaneously releasing NO (Pdmer et al. 1987; Ignarro et al. 1987; Myers et a!. 1990), L-arginine as the precursor of NO (Pdmer et al. 1987; Moncada et al. 19891, and the stimulation 'Correspsndence may be sent to the author at the Hospital for Sick Children address. Printed in Canada / Imprimt au Canada
of the soluble guanylate cyelase and generation of guanosine 3'3'-cyclic monophosphate (cGMP) as the cellular mechanism of NO-induced vascular relaxation (Holzmann 1982; Rapsport and Murad 1983; FCrstermann et al. 1986; Hgnarrs et ab. 1988). However, there are inter- and intra-species differences in endsthelibam-dependent vascular responses. For example, ACh causes an endothelium-dependent relaxation of isolated coronary arteries s f the dog but not the pig (Grgser et al. 1986). ACh induces endothelium-dependent vascular relaxation of isolated canine coronaq , femoral, pulmonary , saphensus, and splenic arteries but not the basilar artery where it induces an endshelium-independent vascular contraction (Debfey and Vadsutte 1982; Katusic et a&. 198%; Griser et a&.1986; Kananam et al. 1987). Isolated segments from
1209
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
#REIDSTEIN ET AL.
porcine aorta relax t o ACh, but not those from the porcine coronary artery (Gordon and Mar-tin 1983; Gr5ser et a&.1986; Shirnohwa et al. 1987). At the present time, the vascular responsiveness of human arteries to ACh is still unclear (Bossaller et al. 1987; Toda and Cbkanura 1989). Of particular interest to us is the responsiveness s f human skin flap vasculature to ACh stimulation. Therefore, the aim of this experiment was to investigate the endothelium-dependent and endothelium-independent vas-activity of isolated prksegf human skin flaps, using ACh and nitroglycerin (NTG) as probes. This information may provide important insight into the pathogenesis s%skin ischemia, which is one of the most common complications in sfin flap surgery (Stranc and Stranc 19'75; Wray et a!. 1982).
1987)- Recently, we have modified this tahanique for in vitm perhsion of the transverse paraumbilical human skin flap model (Kreidstein et al. 1991). The same skin flap model and perfusion technique were used in the present study. Briefly described, the commercially available TwoITen Perfuser (MX International, Aurora, Colo.) equipped with two reservoirs was used as a perfusion apparatus. The arterial catheter of the flap was connected by polyethylene tubing (PE- 190) to a variable-speed peristaltic pump (Masterflex. Cole-Parrner Instrument Company) fitted with a roller pump head (model 7821- 14.) and Ioaded with size I6 Tygon tubing. A three-way connector that linked the tubing from the peristaltic pump to the arterial angiocatheter permitted a parallel tubing to be connected to a pressure transducer (AB High Performance Pressure Transducer, Data Instruments, Inc., Lexington, Mass.). The transducer output was displayed continuously on a digital monitor (Trendicator I1 6218 digital strain gauge, Doric Scientific, San Diego, Calif.) and a chart recorder (Lineacorder WR 3 101, Graphtec, Japan).
Source of human skip2 Demolipectomy is routinely performed on patients with redundant abdominal skin. This surgical procedure involves removal of a large skin pannus from the abdomen. The excised skin pannus serves no purpose to the patient m d is normally disposed by incineration. A clinical protocol was approved by a hospital ethics committee for in vitro perfusion of h u m n skin. The procurement and disposal of the human skin specimens were in accordance with the policy of the responsible Pathology Department of St. Joseph's Medical Centre. All skin specimens accepted for the present study fulfilled the following criteria: (i) the skin was free of previous surgical scar, lesion, or infection; and (ii) the patients were not known to smoke or suffer any systemic diseases or take anti-inflammatory or cardiovascular medications.
Pe@sion medium A modified Krebs -Henseleit buffer of the following composition (mM) was used as pefisate: NaC1, 110; KC1, 4.60; KH,PO,, 1.36; MgSO,, 1.20; CaCl,, 2.75; NaHCO,, 31.0; glucose, 11; and sucrose, 1 1. Purified water (Milli-Q water system, Millipore Ltd., Bedford, Mass.) was used to make fresh buffer for each study. Bovine serum albumin (Cohn fraction V) was dissolved in the buffer (65 g L-I). The perksate was filtered (Whatman No. 44, Whatman International, Maidstone, England), placed in reservoirs, and equilibrated with 95% 00,and 5 % CO, at 37"C, The reservoirs were equipped with rotating screens and a filter-bubble trap was placed distal to the reservoirs to prevent air ernbolizatisn of the skin flap.
Design and cannulation of the skin pap The location, surgical technique, and design of the human paraumbilical skin flap model used in the present experiment have been described in detail (Kreidstein ef al. 1891). Briefly described, the skin pannus excised from the abdomen was placed on a grounded Mayo stand. An 8 x 20 cm skin flap was excised from this skin pamas, using unipolar electrocautery. Upon release of this 8 x 20 cm skin flap by the Pathology Department, the flap was immediately transpomaed to our laboratory in a plastic bag at ambient temperature. After completisn of an isolated perfusion experiment as described below, the skin flap was returned to the Pathology Department for disposal in the usual manner. The proximal end of the transverse paraumbilical skin flap possessed a paired paraumbilical artery and vein constituting the vascular pedicle (base) of the skin flap. The vasculature of this flap model has been studied by angiography (Kreidstein 199I). The arterial and venous stumps of the vascular pedicle were cannuHated with 22- and 1 $-gauge angiocatheters, respectively. Heparinized perfusate (100 U . mL-') was gently perfused into the artery with a syringe until venous outflow was observed. The skin flap was then placed on a fiberglass effluent collection tray. Thermistor probes (YSI series 400, Yellow Springs Instruments, Yellow Springs, Ohio) connected to a thermometer (microcomputer thermometer series 884202, Cole-Pamer Instruments, Chicago, Ill .) were positioned on the surface of the mid-pint of the skin flap and at the hub of the arterial catheter to allow csntinugsus monitoring of sfin surface and perfusate temperature, respectively. After 30 min of stabilization, the perfusate and skin surface temperatures were 36.0 -36.5 and 33-34"C, respectively. At the beginning of each perfusion, any identifiable leaking vessels were ligated with small or large size ligak r e clips (ligaclips, Ethicon, Peterbsrough, Ont.). In addition, pressure clips (Double Clips, Grand & Toy Co., Toronto, Ont.) were applied around the edges of the flap to minimize leakage from the dermal - subdermal plexus. Skin flap pefision kchniqtde An in vitro perfusion technique for the pig skin flap has been described previously (Riviere et (21. 1986; Monteirs Riviere et al.
Measurement of skin flaps pe@sion pressure The digital monitor and chart recorder that recorded perhsion pressure were set at zero after the initial connection of the pump tubing to the arterial catheter of the skin flap. The flow rate was adjusted (about 8- 10 mL . minsl) to achieve a perhsion pressure sf 50 mmHg (1 mmHg = 133 Pa), and the flap was allowed to stabilize for 30 mine A baseline perfusion pressure of 50 m d g was chosen because preliminary observations revealed it to provide good tissue perfusion with minimal leakage and edema ( < 10%). During this initial 30-min stabilization period, flaps with embolism and (or) thrombosis displayed abnormally high perfusion pressure and these flaps were not further studied. Skin flap vascular resistance was calculated as the ratio between arterial pressure and arterial flow rate during flap perhsion. In the present experiment, the vascular response to drug treatment was studied under "constant flow" as we11 as 66con~tant pressure9' conditions. Under the "constant flow" condition, the pump rate was kept constant and the pressure was allowed to change. Under the "constant pressure" condition, the perfusion pressure was kept constant by adjusting the pump rate.
Sugace fgusrometry technique f i r assessment of skin perjksion Fluorometry has been used for in viva assessment of dermal perfusion in pig skin flaps (Thsmson and Kerrigan 1989). This technique was modified for in vitro assessment of dermal perfusion. Beginning at the base of the skin flap where the arterial pedicle entered the skin flap, 1-cm diameter circles were marked out in a confluent line along the longitudinal midline of the skin flap surface. A 1 skin surface fluorometry readings were taken within these sample areas (Fig. I). Before taking fluorometry readings, the surface fluorometer (Fluoroscan, Santa Barbara Technology, Inc., Santa Barbara, Calif,) was calibrated against the fluorescent standard provided by the manufacturer. Background skin surface fluorometry readings were taken by placing the fluorometer probe (5 mm diameter) in each of the marked circles along the midline of the flap. 8 digital reading was displayed every second by the flaorometer, and the reading displayed after 4 s was recorded, as r~ommendedby the manufacturer. After taking background surface fluorsmetry readings, fluorescein dye (Fluorescite, Alcon Canada hnc., Mississauga, Ont.) was added to the perhsate in the reservoirs (I0 mg L-j), and after 4 min of fluorescein
-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
CAN. J. PHYSIBL. PHARMACOE. V 8 L . 70. 1992
with Fluorescein
'
I
Effluent
FIG.1. Simplified diagrammatic illustration s f surface fluorometry technique. The probe s f the surface fluorometer was placed against the skin surface to measure skin surface fluorescence. Readings were taken at each circular skin sample area along the longitudinal midline of the skin flap.
perfusion, fluorometry readings were taken again from each circle along the midline of the flap. The difference between the background and post-fluorescein perfusion fluorometry reading for each circle was defined as the net surface fluorometry reading for that site. The total net surface fluorometry of the skin flap was the sum of all net fluorometry readings taken from all circles marked along the midline of the flap. Fluorescein perfusions were separated by 10 min of perfusion with fluorescein-free perfusate to permit fluorescein washout and the maintenance of relatively stable and low background fluorometry readings. His~oksgy Saponin was used to damage vascular endothelial cells in the study of endothelium-dependent and endothelium-independent vasoacthity . The presence and absence of vascular endothelial cells before and after saponin treatment was qualitatively assessed by histologic examination. Skin samples (1 cm2) were taken from well-perfused areas in the proximal aspects of the skin flaps before and after saponin treatment. Samples of arterial pedicle (1.0 cm long) were taken from beyond the distal end of the angiocatheter at the end of the experiment. All samples were immediately placed in 18% formaldehyde fixative (pH 7.4). Permanent sections stained with hematoxylin and eosin or imunoperoxidase were made from skin and blood vessel samples. Immunohistochemical technique was used for staining for factor VIII on the endothelial cell surface (Mukai et ah. 1980; Haniuda et ah. 1989). All slides were prepared by the Research Service Laboratory of the Department of Pathology at The Hospital for Sick Children. Stained sections were viewed by light microscopy and photographed (Polyvar Widefield Photo-microscope, type 622 1-82 Micrometer, Reichert-Jung, Austria). A cardiovascular pathologist (Dr. Greg Wilson) was consulted for determination of vascular endohelium pathology. Experimental protocol Study 1: Endothekium-depenknt and en&thekium-independent skin vasodilwtion Owing to the limited numbers of human skin specimens available, each flap was used for the study of both endothelium-dependent and endothelium-independent vasodilation. Each skin flap underwent three stages of perfusion as described in the following protocol, and the total perfusion time for each flap was less than 4 h. We have demonstrated previously that the metabolic activity, basal perfusion pressure, and vascular reactivity of this human skin flap model remained intact and stable afaer 4 -5 h of isolated perfusion (Kreidstein et ah. 1991). In our preliminary experiment, we repeatedly induced vasodilation in isolated perfused human skin flaps with ACh and NTG with a 15-min washout and stabilization period between each application of drug, and we observed that the order of drug application did not affect the vassdilating action s f these two drugs.
stage I: To study the efiect sf NE on skin &p vascubar resistance and su$ace f l u o m m t ~ ~ The skin flaps were perfused with modified Krebs buffer csntaining propranolol (lQs%) and indsmethacin (3 x 1W5M). The perfusion rate was adjusted to produce a basal perfusion pressure s f 50 m H g , which provided good tissue perfusion in this flap model (Kreidstein et al. 1991). At the end of a 30-min stabilization period, the perfaasion rate was recorded, and net skin surface fluorometry was determined. NE (loP6 M) was added to the perfusate, and the perfusion rate was lowered to maintain a constant perfusion pressure of 50 mmHg. After 10 min s f NE challenge, vascular resistance was stable and the perhsion rate and net skin surface fluorometry were determined. Stage Ik To study the vascular eflects of ACh and NTG on precsnstrictsd skin flaps with kntwct vascular en&the/ium The skin flaps were now perfused for the remainder of the experiment with modified Krebs buffer containing indomethacin (3 x 10-5 M), prspranolol (1W6 M), and NE (1W6 M). This perfusate was perfused into the flap at 100 m H g for 10 min, at which time vascular resistance was stable. Both indomethacin and propranolol did not have any effects on the basal perhsion pressure. Net skin surface fluorometry and vascular resistance were determined, ACh (lo-") was added to the perfusate, and after 10 min of perfusion (constant flow condition), vascular resistance and skin surface fluorometry were reassessed. Perfusion with fluorescein-free perfusate containing A@h (1W6 M) was then resumed for 10 min at a flow rate sufficient to maintain perfusion pressure at 100 m H g (constant pressure condition), and then net surface fluorornetq readings were again determined. The flap was then perfused with ACh-free perfusate for 15 min; with vascular resistance returned to 100 m d g , NTG (1W6 M), an endothelium-independent vasodilator, was added to the perfusate and its effect on vascular resistance after 10 min of perfusion was determined (constant Wow condition). A skin specimen (1 cm2) from the proximal skin flap was taken for histologic examination of the vascular endothelium. Stage HHHs To study the vascular eflect sf ACh and NTG on prempestrr'cted skin Paps with damaged vascular endo~helium The skin flap was perfused for 5 rnin with perfaasate containing saponin (100 mg . E-'1, an endsthelium damaging detergent (Samata et al. 1986; Graser et ak. 1988). Perfusion with saponin-free perfusate was then resumed for 30 min until vascular resistance had M) stabilized with a perfusion pressure of 100 mmHg. ACh was then added to the perfaasate, and its effect on vascular resistance was assessed (constant flow condition). Perfusion with ACh-free perhsate was performed for 10 rnin; once the vascular resistance had stabilized, NTG (loPBM) was added to the perfusate, and its effect s f vascular resistance was assessed (constant flow condition). A skin sample (1 cm2) from the proximal skin flap and a segment of the vascular pedicle were taken from the skin flap at the end of the experiment for histologic study of the blood vessels after saponin treatment. St@& 2s Inhibition of ACh-i~zducedskin vascujar relaxation Concentration-dependemat vascular relaxation effect of A@h was studied in the absence or presence of Nu-nitro-L-arginine (LNA, a NO synthesis inhibitor) in isolated perfused skin flaps preconstricted with NE. Six skin flaps were available. After 30 min of stabilization of perfusisn pressure at 50 m H g , NE (1W6 M) was added to the reservoir to raise the perfaasion pressure at constant flow rate. Maximum perfusion pressure was achieved within 10 min. Increasing doses of ACh (10-80- 10-%) were infused through a side-am into the perfusate immediately before entering the angiocatheter in the flap. When the presence of NO synthesis inhibitor was required, LNA was added to the reservoir (los5 M) 15-20 min before ACh infusion.
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
KRBHDSTBHN ET AL.
FIG. 2. Vascular resistance in isolated p e h s e d skin flaps before M) treatment. (pre-NE) and after (post-NE) norepinephrine Values are means f SEM; n = 7. Mean values s f vascular resistance for pre-NB and post-NE treatment groups are significantly different; paired t-test (r = 2.93, *p < 0.05).
FIG.3. Total net surface fluorometry in skin flaps before (pre-NE) and after (psst-NB) norepinephrine (1W6 M) treatment. Values are means f SEM; ;n = 7. Mean values of tot$ flap fluorometry for pre-NE and post-NE treatment groups are significantly different; paired t-test (P -- 2.76; *p < 0.05).
Drugs and chemiculs All drugs and chemicals were obtained from Sigma Chemical Company (St. Louis, Miss.) except the following: norepinephrine bitartrate (Winthrop Laboratories, Aurora, Ont ,) , fluorescein sodium (Alcon Canada Inc., Mississauga, Ont .), and nitroglycerin (Parke Davis, Searborough, Ont .). Data ana&ysis All values are expressed as means f standard error of the mean (SEM). The paired t-test was used for comparison of two means. P values reflect the Bonferroni correction for multiple outcomes (Sokal 1987). One-way analysis of variance (ANOVA) with repeated measures followed by the least significant difference test (Fisher's protected least significant difference test) were used in the compario n s of multiple 8ependent means (Milliken 1984). Probabilities with p I0.05 were. accepted as significant.
Results Study b :Endothebium-depedent and enhthebium-independeat skin vasodilation Skin flaps with intact vascular endothelium (i. e. , befire saponin pe@sion) Addition of NE ( I W g M) to the perfusate significantly ( p < 0.05; n = 7) increased the vascular resistance of the isolated perfused human skin Waps (Fig. 2). This NE treatment also significmdy ( p < 0.05) decreased the total net surface fluorometry in these skin Waps (Fig. 3). The distance of perfusion along the skin flap detected by fluorometry was reduced to 69 +- 12% of baseline following the addition of NE. ACh (lo-$ M) significantly ( p < 8.01; n = 7) attenuated the vascular resistance of skin flaps preconstricted with NE (lo-$ M) despite the presence of a cyclooxygenase inhibitor M) and a 0-adrenergic receptor (indomethacin, 3 x (Fig. 4). The total surface blocker (propranolol, lo-") fluorometry in these flaps was significantly ( p < 0.05) increased under conditions of constant pressure perfusion, but not when Wow rate was maintained constant (Fig. 5). This observation indicated that perfusion pressure must be sustained for local vascular relaxation to result in increased flow rate and hence tissue perfusion. The distances of perfusion as detected by fluorornetry under constant pressure or constant flow conditions were 156 &- 49 and $5 f 6.4 % of baseline,
FIG.4. Effect of acetylcholine (ACh) on the vascular resistance of skin flaps preconstrieted with norepinephrine (NE). The concentrations of NE and ACh were f W6M. Values are means f SEM; n = 7. Mean values of vascular resistance for NE and NE ACh treatment groups are significantly different; paired t-test (P = 5.20; *p < 0.05).
+
respectively. Similar to ACh, NTG treatment significantly ( p < 0.01; n = 7) reduced the vascular resistance of skin flaps preconstricted with NE (Fig. 6). Skin paps with damaged vascular endothelium (i. e., aper saponin pefision) Histobogy-The endothelid lining in skin and arterial pedicle specimens was studied under light microscopy. Before saponin perfusion, the endothelial lining of the arterial pedicle was fully intact, as discerned by the presence of endothelial cell nuclei and positive staining for factor %PHI.Conversely, the arterial pedicle was denuded of emdothelial cells following saponin perfusion. Examination of arteries within the dermal subdermal plexus revealed an intact endothelium prior to saponin perfusion (Fig. 71, but total or partial removal of endohelid cell lining following saponin perfusion (Fig. 8). Vascular r e a e t i v i - h e perfusion pressure increased over 200 m H g during the 5 min of saponin infusion. However, after 25 -40 min of saponin-free perfusion, the perfusion pres-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
1212
CAN. d. PWYSIOL. PHARMACOL. VOL. 70, 1992
NE
NE+ACh
NE+ACh
NE
ME + NTG
FIG. 5. Effect of acetylcholine (ACh) on total flap Wuorometq in skin flaps preconstricted with norepinephrine (NE). Values are means f SEM; n = 7, Means without a common letter are significantly ( p < 64.05) different. Statistics: one-way ANOVA with repeated measures (F = 8,482) and Fisher's protected least significant difference test.
FIG.6. Effect of nitroglycerin (NTG) on the vascular resistance of skin Waps preconstricted with norepinephrine (NE). The concentrations of NE and NTG were 10s6M. Values are means f SEM; n = 7. Mean values s f vascular resistance for NE and NE + NTG treatment groups are significantly different; paired t-test (t = 4.60; *p < 0.05).
sure stabilized at 25-30 mmHg above the basal perfusion pressure. Since perfusion pressure is proportional to vascular resistance under constant flow condition, the vascular relaxation effect sf ACh and NTG can be expressed as a percentage of increased p r h s i o n pressure (tone) induced by NE (Ebeigbe eb a&.1890). Following saponin treatment, the ability for ACh (18-%M) to produce vascular relaxation in skin Waps preconstricted with NE was significantly ( p < 0.01) impaired (Fig. 9). In contrast, the ability of NTG (lo-$ M) to induce vascular relaxation in isolated perfused skin flaps remained unchanged following saponin treatment (Fig. 9).
period of about 4 h (Kreidstein et a&. 1891). Furthermore, in study 2 of the present experiment, a distinct concentrationdependent vascular relaxation effect of acetylcholine was observed (Fig. 10) and each perfusion study lasted only for about 2 h. Therefore, it is most unlikely that a significant endotoxin conmination could have been present in our preparations in studies 1 and 2 to confound the vascular relaxation effect of acetylcholine.
S t d y 2: lahibition of ACh-induced skin vascular relaation by a NO synthesis inhibitor The bas& perfusion pressure stabilized at 50-52 mmHg within 30 min (a = 6). Addition of NE (1W6 M) into the reservoir caused skin flap vasoconstriction, raising the perfusion pressure to 129 f 8 mmHg at constant flow condition. The net increase in perfhasion pressure was 79 f 8 mmMg. The vascular relaxation effect sf ACh in the absence and presence of ENA was studied in NE-preconstricted flaps under consBnt flow condition. Again, the vascular relaxation effect of ACh is expressed as a percentage of increased perhsion pressure (tone) induced by NE (Ebeigbe et al. 1990). Infusion of ACh in skin flags preconstrictd with NE (18-") caused vascular relaxation in a concentration-dependent manner (Fig. 10). However, h i s vascular relaxation effect of ACh at the dose range of BO-9-10-5 M was significantly ( p < 0.01) inhibited in the presence of 1W5 M LNA, a NO synthesis inhibitor. LNA alone did not have any effect on the basal perfusion pressure of the skin flaps. Possibility of endotmin-induced vascular relaxation due to csn~amination In the present experiment, fresh Milli-Q water was used to prepare Krebs buffer 2 h before each perhsion study. The Milli-Q water system was fed by distilled water. Using the same human sfin flap model, the same source of Milli-Q water, and the same batch of bovine serum albumin, we demonstrated previously that the perfusion pressure of control flaps did not change significantly (
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
Departments
a$
Surgery and Physiobogy, University of Toronto, Toronto, Ont., Canada Received January 22, 1992
KREIDSTEIN, M. L., PANG,C. Y. , CARLSEN, L. N., and Xu, N. 1992. Evidence for endothelium-dependent and endotheliumindependent vasodilation in human skin flaps. Can. J. Physiol. Pharmacol. 70: 1208- 12 16. Acetylcholine (ACh) and nitroglycerin (NTG) were used as probes to study endothelium-dependent and endothelinmindependent vascular relaxation in isolated perfused transverse paraun~bilicalhuman skin flaps. Ht was observed that ACh (18-") significantly ( p < 0.05) decreased the vascular resistance and increased dermal capillary perfusion (assessed by surface fluorometry) in norepinephrine (NE, M) precsnstricted skin flaps, despite the presence of a cyclooxygemse inhibitor (indornethacin, 3 x H8-%) and a fi-adrenergic receptor antagonist (propranolol, 1 0 - w ) . The ability of ACh to induce vascular relaxation in NE-preconstricted skin flaps was lost after damaging the vascular eraelothelial lining with saponin perfusion (100 rng * E-I, 5 min). In contrast, NTG (1W6 M) induced vascular relaxation to a similar extent before and after saponin treatment. In a separate study, ACh was seen to induce vascular relaxation in a concentration-dependent This vascular relaxation effect of ACh over the dose range of manner in sfin flaps preconstrieted with NE (10-%). lo-% was significantly ( p < 0.08) inhibited in the presence of Nu-nitro-L-arginine (los5 M), a nitric oxide (NO) synthesis inhibitor. These observations were taken to indicate the presence of endotkelium-dependent and endotheliumindependent vascular relaxation in human skin flaps and that the ACh-induced endothelium-dependent relaxation is probably mediated by NO. The importance of impairment of endothelium-dependent relaxation in the pathogenesis of skin flap ischemia, and the potential use s f topical anitrovasodilators or NO donors for prevention and (or) treatment sf skin flap ischemia were d s o discussed. Key words: acetylcholine, nitroglycerin, Nu-nitro-L-arginine, vascular relaxation, human skin. KREIDSTEIN,M. IL., PANG,C. Ye, CAWWEN, E. N., et Xu, N. 1992. Evidence for endothelium-dependent and endotheliurnindependent vasodilation in human skin flaps. Can. J. Bhysiol. Bharmacol. 70 : 1208- 1216. On a utilisC 17acCtylcholine(ACh) et la nitroglycCrine (NTG) c o m e sondes pour examiner la relaxation vasculaire dCpendank et indkpendante de 17endothCliumdans des Iamkaux eutanis para-ombilicanx transverses, p e r h d s et isolCs d9humains. L'ACh (IOP6 M) a diminart significativement ( p < 0,05) la resistance vasculaire et augment6 la perhsion capillaire dermique (Cvalute par fluoromCtrie de surface) dans les lambeaux cutanCs prk-comprimCs par la nor6pintphrine (NE, M), malgrt la prCsence d9un irabaibiteur de cyclooxygtnase (indomtthacine, 3 x 10-"M) et d'un antagoniste des rkcepteurs @adrknergiques (propanolol, 10s6 M). L'ACh a perdu sa capacitt d'induire une relaxation vasculaire dans les lambeaux cutanks prd-comprim6s par Ia NE, aprks l'alttration de I'endothtliurn vasculaire par la perfusion de sapnine (100 mg - L-', M) a induit une relaxation vasenlaire un degrC similaire, avant et aprks le traitement 5 min). A 170pposC,la NTG B la saponine. Dans une autre Cmde, I'ACh a induit une relaxation vasculaire, dCpeadante de la concentration, dans les lambeaux cutanCs prd-comprimCs avec la NE M). Cet effet de 17ACh sup la relaxation vasculaire dans la plage de lO-'- lo-" a at6 significativement inhibt ( y < 0,Ol) en prCsence Be Nd-nitro-L-arginine (10-%), un inhibiteur de la synth&sede l'oxyde nitrique (ON). Ces observations ont permis d'indiquer la presence d'une relaxation vasculaire dCpendante et indipendante de 19endothCliumdans les lambeaux cutanCs humains, la relaxation dCpndante de I'endothClium induite par 19AChCtant probablement mkdi6e par 1'8N. On a aussi discutk de l'importance d9unealtiration de la relaxation dtpendante de I9endoh6lium dans la pathogenkse de 19ischCmiede lambeaux cutanCs et de l'utilisation potentielle de nitrovasodiIatateurs locaux oar de donnmeurs 8 N pour la prevention et (ou) le traitement de 17ischCmiede lambeaux cutanis. Mots c&&s: acktylcholine, nitroglycirine, Nu-nitro-L-arginine, relaxation vasculaire, peau humine. [Traduit par la rCdactisn]
Introduction Using acetylcholine (ACh), Furchgott and Zawadzki ((1980) demonstrated endsthelium-dependent relaxation of the rabbit aorta in the presence of cyelooxygenase inhibitors. They introduced the term non-prostansid endsthelibam-derived relaxing factor (EDRF) to describe the substance that caused this vascular relaxation. Subsequently, many advances have been made in this field, including the identification of EDRF as nitric oxide (NO) or a labile nitrosothisl compound capable of spontaneously releasing NO (Pdmer et al. 1987; Ignarro et al. 1987; Myers et a!. 1990), L-arginine as the precursor of NO (Pdmer et al. 1987; Moncada et al. 19891, and the stimulation 'Correspsndence may be sent to the author at the Hospital for Sick Children address. Printed in Canada / Imprimt au Canada
of the soluble guanylate cyelase and generation of guanosine 3'3'-cyclic monophosphate (cGMP) as the cellular mechanism of NO-induced vascular relaxation (Holzmann 1982; Rapsport and Murad 1983; FCrstermann et al. 1986; Hgnarrs et ab. 1988). However, there are inter- and intra-species differences in endsthelibam-dependent vascular responses. For example, ACh causes an endothelium-dependent relaxation of isolated coronary arteries s f the dog but not the pig (Grgser et al. 1986). ACh induces endothelium-dependent vascular relaxation of isolated canine coronaq , femoral, pulmonary , saphensus, and splenic arteries but not the basilar artery where it induces an endshelium-independent vascular contraction (Debfey and Vadsutte 1982; Katusic et a&. 198%; Griser et a&.1986; Kananam et al. 1987). Isolated segments from
1209
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
#REIDSTEIN ET AL.
porcine aorta relax t o ACh, but not those from the porcine coronary artery (Gordon and Mar-tin 1983; Gr5ser et a&.1986; Shirnohwa et al. 1987). At the present time, the vascular responsiveness of human arteries to ACh is still unclear (Bossaller et al. 1987; Toda and Cbkanura 1989). Of particular interest to us is the responsiveness s f human skin flap vasculature to ACh stimulation. Therefore, the aim of this experiment was to investigate the endothelium-dependent and endothelium-independent vas-activity of isolated prksegf human skin flaps, using ACh and nitroglycerin (NTG) as probes. This information may provide important insight into the pathogenesis s%skin ischemia, which is one of the most common complications in sfin flap surgery (Stranc and Stranc 19'75; Wray et a!. 1982).
1987)- Recently, we have modified this tahanique for in vitm perhsion of the transverse paraumbilical human skin flap model (Kreidstein et al. 1991). The same skin flap model and perfusion technique were used in the present study. Briefly described, the commercially available TwoITen Perfuser (MX International, Aurora, Colo.) equipped with two reservoirs was used as a perfusion apparatus. The arterial catheter of the flap was connected by polyethylene tubing (PE- 190) to a variable-speed peristaltic pump (Masterflex. Cole-Parrner Instrument Company) fitted with a roller pump head (model 7821- 14.) and Ioaded with size I6 Tygon tubing. A three-way connector that linked the tubing from the peristaltic pump to the arterial angiocatheter permitted a parallel tubing to be connected to a pressure transducer (AB High Performance Pressure Transducer, Data Instruments, Inc., Lexington, Mass.). The transducer output was displayed continuously on a digital monitor (Trendicator I1 6218 digital strain gauge, Doric Scientific, San Diego, Calif.) and a chart recorder (Lineacorder WR 3 101, Graphtec, Japan).
Source of human skip2 Demolipectomy is routinely performed on patients with redundant abdominal skin. This surgical procedure involves removal of a large skin pannus from the abdomen. The excised skin pannus serves no purpose to the patient m d is normally disposed by incineration. A clinical protocol was approved by a hospital ethics committee for in vitro perfusion of h u m n skin. The procurement and disposal of the human skin specimens were in accordance with the policy of the responsible Pathology Department of St. Joseph's Medical Centre. All skin specimens accepted for the present study fulfilled the following criteria: (i) the skin was free of previous surgical scar, lesion, or infection; and (ii) the patients were not known to smoke or suffer any systemic diseases or take anti-inflammatory or cardiovascular medications.
Pe@sion medium A modified Krebs -Henseleit buffer of the following composition (mM) was used as pefisate: NaC1, 110; KC1, 4.60; KH,PO,, 1.36; MgSO,, 1.20; CaCl,, 2.75; NaHCO,, 31.0; glucose, 11; and sucrose, 1 1. Purified water (Milli-Q water system, Millipore Ltd., Bedford, Mass.) was used to make fresh buffer for each study. Bovine serum albumin (Cohn fraction V) was dissolved in the buffer (65 g L-I). The perksate was filtered (Whatman No. 44, Whatman International, Maidstone, England), placed in reservoirs, and equilibrated with 95% 00,and 5 % CO, at 37"C, The reservoirs were equipped with rotating screens and a filter-bubble trap was placed distal to the reservoirs to prevent air ernbolizatisn of the skin flap.
Design and cannulation of the skin pap The location, surgical technique, and design of the human paraumbilical skin flap model used in the present experiment have been described in detail (Kreidstein ef al. 1891). Briefly described, the skin pannus excised from the abdomen was placed on a grounded Mayo stand. An 8 x 20 cm skin flap was excised from this skin pamas, using unipolar electrocautery. Upon release of this 8 x 20 cm skin flap by the Pathology Department, the flap was immediately transpomaed to our laboratory in a plastic bag at ambient temperature. After completisn of an isolated perfusion experiment as described below, the skin flap was returned to the Pathology Department for disposal in the usual manner. The proximal end of the transverse paraumbilical skin flap possessed a paired paraumbilical artery and vein constituting the vascular pedicle (base) of the skin flap. The vasculature of this flap model has been studied by angiography (Kreidstein 199I). The arterial and venous stumps of the vascular pedicle were cannuHated with 22- and 1 $-gauge angiocatheters, respectively. Heparinized perfusate (100 U . mL-') was gently perfused into the artery with a syringe until venous outflow was observed. The skin flap was then placed on a fiberglass effluent collection tray. Thermistor probes (YSI series 400, Yellow Springs Instruments, Yellow Springs, Ohio) connected to a thermometer (microcomputer thermometer series 884202, Cole-Pamer Instruments, Chicago, Ill .) were positioned on the surface of the mid-pint of the skin flap and at the hub of the arterial catheter to allow csntinugsus monitoring of sfin surface and perfusate temperature, respectively. After 30 min of stabilization, the perfusate and skin surface temperatures were 36.0 -36.5 and 33-34"C, respectively. At the beginning of each perfusion, any identifiable leaking vessels were ligated with small or large size ligak r e clips (ligaclips, Ethicon, Peterbsrough, Ont.). In addition, pressure clips (Double Clips, Grand & Toy Co., Toronto, Ont.) were applied around the edges of the flap to minimize leakage from the dermal - subdermal plexus. Skin flap pefision kchniqtde An in vitro perfusion technique for the pig skin flap has been described previously (Riviere et (21. 1986; Monteirs Riviere et al.
Measurement of skin flaps pe@sion pressure The digital monitor and chart recorder that recorded perhsion pressure were set at zero after the initial connection of the pump tubing to the arterial catheter of the skin flap. The flow rate was adjusted (about 8- 10 mL . minsl) to achieve a perhsion pressure sf 50 mmHg (1 mmHg = 133 Pa), and the flap was allowed to stabilize for 30 mine A baseline perfusion pressure of 50 m d g was chosen because preliminary observations revealed it to provide good tissue perfusion with minimal leakage and edema ( < 10%). During this initial 30-min stabilization period, flaps with embolism and (or) thrombosis displayed abnormally high perfusion pressure and these flaps were not further studied. Skin flap vascular resistance was calculated as the ratio between arterial pressure and arterial flow rate during flap perhsion. In the present experiment, the vascular response to drug treatment was studied under "constant flow" as we11 as 66con~tant pressure9' conditions. Under the "constant flow" condition, the pump rate was kept constant and the pressure was allowed to change. Under the "constant pressure" condition, the perfusion pressure was kept constant by adjusting the pump rate.
Sugace fgusrometry technique f i r assessment of skin perjksion Fluorometry has been used for in viva assessment of dermal perfusion in pig skin flaps (Thsmson and Kerrigan 1989). This technique was modified for in vitro assessment of dermal perfusion. Beginning at the base of the skin flap where the arterial pedicle entered the skin flap, 1-cm diameter circles were marked out in a confluent line along the longitudinal midline of the skin flap surface. A 1 skin surface fluorometry readings were taken within these sample areas (Fig. I). Before taking fluorometry readings, the surface fluorometer (Fluoroscan, Santa Barbara Technology, Inc., Santa Barbara, Calif,) was calibrated against the fluorescent standard provided by the manufacturer. Background skin surface fluorometry readings were taken by placing the fluorometer probe (5 mm diameter) in each of the marked circles along the midline of the flap. 8 digital reading was displayed every second by the flaorometer, and the reading displayed after 4 s was recorded, as r~ommendedby the manufacturer. After taking background surface fluorsmetry readings, fluorescein dye (Fluorescite, Alcon Canada hnc., Mississauga, Ont.) was added to the perhsate in the reservoirs (I0 mg L-j), and after 4 min of fluorescein
-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
CAN. J. PHYSIBL. PHARMACOE. V 8 L . 70. 1992
with Fluorescein
'
I
Effluent
FIG.1. Simplified diagrammatic illustration s f surface fluorometry technique. The probe s f the surface fluorometer was placed against the skin surface to measure skin surface fluorescence. Readings were taken at each circular skin sample area along the longitudinal midline of the skin flap.
perfusion, fluorometry readings were taken again from each circle along the midline of the flap. The difference between the background and post-fluorescein perfusion fluorometry reading for each circle was defined as the net surface fluorometry reading for that site. The total net surface fluorometry of the skin flap was the sum of all net fluorometry readings taken from all circles marked along the midline of the flap. Fluorescein perfusions were separated by 10 min of perfusion with fluorescein-free perfusate to permit fluorescein washout and the maintenance of relatively stable and low background fluorometry readings. His~oksgy Saponin was used to damage vascular endothelial cells in the study of endothelium-dependent and endothelium-independent vasoacthity . The presence and absence of vascular endothelial cells before and after saponin treatment was qualitatively assessed by histologic examination. Skin samples (1 cm2) were taken from well-perfused areas in the proximal aspects of the skin flaps before and after saponin treatment. Samples of arterial pedicle (1.0 cm long) were taken from beyond the distal end of the angiocatheter at the end of the experiment. All samples were immediately placed in 18% formaldehyde fixative (pH 7.4). Permanent sections stained with hematoxylin and eosin or imunoperoxidase were made from skin and blood vessel samples. Immunohistochemical technique was used for staining for factor VIII on the endothelial cell surface (Mukai et ah. 1980; Haniuda et ah. 1989). All slides were prepared by the Research Service Laboratory of the Department of Pathology at The Hospital for Sick Children. Stained sections were viewed by light microscopy and photographed (Polyvar Widefield Photo-microscope, type 622 1-82 Micrometer, Reichert-Jung, Austria). A cardiovascular pathologist (Dr. Greg Wilson) was consulted for determination of vascular endohelium pathology. Experimental protocol Study 1: Endothekium-depenknt and en&thekium-independent skin vasodilwtion Owing to the limited numbers of human skin specimens available, each flap was used for the study of both endothelium-dependent and endothelium-independent vasodilation. Each skin flap underwent three stages of perfusion as described in the following protocol, and the total perfusion time for each flap was less than 4 h. We have demonstrated previously that the metabolic activity, basal perfusion pressure, and vascular reactivity of this human skin flap model remained intact and stable afaer 4 -5 h of isolated perfusion (Kreidstein et ah. 1991). In our preliminary experiment, we repeatedly induced vasodilation in isolated perfused human skin flaps with ACh and NTG with a 15-min washout and stabilization period between each application of drug, and we observed that the order of drug application did not affect the vassdilating action s f these two drugs.
stage I: To study the efiect sf NE on skin &p vascubar resistance and su$ace f l u o m m t ~ ~ The skin flaps were perfused with modified Krebs buffer csntaining propranolol (lQs%) and indsmethacin (3 x 1W5M). The perfusion rate was adjusted to produce a basal perfusion pressure s f 50 m H g , which provided good tissue perfusion in this flap model (Kreidstein et al. 1991). At the end of a 30-min stabilization period, the perfaasion rate was recorded, and net skin surface fluorometry was determined. NE (loP6 M) was added to the perfusate, and the perfusion rate was lowered to maintain a constant perfusion pressure of 50 mmHg. After 10 min s f NE challenge, vascular resistance was stable and the perhsion rate and net skin surface fluorometry were determined. Stage Ik To study the vascular eflects of ACh and NTG on precsnstrictsd skin flaps with kntwct vascular en&the/ium The skin flaps were now perfused for the remainder of the experiment with modified Krebs buffer containing indomethacin (3 x 10-5 M), prspranolol (1W6 M), and NE (1W6 M). This perfusate was perfused into the flap at 100 m H g for 10 min, at which time vascular resistance was stable. Both indomethacin and propranolol did not have any effects on the basal perhsion pressure. Net skin surface fluorometry and vascular resistance were determined, ACh (lo-") was added to the perfusate, and after 10 min of perfusion (constant flow condition), vascular resistance and skin surface fluorometry were reassessed. Perfusion with fluorescein-free perfusate containing A@h (1W6 M) was then resumed for 10 min at a flow rate sufficient to maintain perfusion pressure at 100 m H g (constant pressure condition), and then net surface fluorornetq readings were again determined. The flap was then perfused with ACh-free perfusate for 15 min; with vascular resistance returned to 100 m d g , NTG (1W6 M), an endothelium-independent vasodilator, was added to the perfusate and its effect on vascular resistance after 10 min of perfusion was determined (constant Wow condition). A skin specimen (1 cm2) from the proximal skin flap was taken for histologic examination of the vascular endothelium. Stage HHHs To study the vascular eflect sf ACh and NTG on prempestrr'cted skin Paps with damaged vascular endo~helium The skin flap was perfused for 5 rnin with perfaasate containing saponin (100 mg . E-'1, an endsthelium damaging detergent (Samata et al. 1986; Graser et ak. 1988). Perfusion with saponin-free perfusate was then resumed for 30 min until vascular resistance had M) stabilized with a perfusion pressure of 100 mmHg. ACh was then added to the perfaasate, and its effect on vascular resistance was assessed (constant flow condition). Perfusion with ACh-free perhsate was performed for 10 rnin; once the vascular resistance had stabilized, NTG (loPBM) was added to the perfusate, and its effect s f vascular resistance was assessed (constant flow condition). A skin sample (1 cm2) from the proximal skin flap and a segment of the vascular pedicle were taken from the skin flap at the end of the experiment for histologic study of the blood vessels after saponin treatment. St@& 2s Inhibition of ACh-i~zducedskin vascujar relaxation Concentration-dependemat vascular relaxation effect of A@h was studied in the absence or presence of Nu-nitro-L-arginine (LNA, a NO synthesis inhibitor) in isolated perfused skin flaps preconstricted with NE. Six skin flaps were available. After 30 min of stabilization of perfusisn pressure at 50 m H g , NE (1W6 M) was added to the reservoir to raise the perfaasion pressure at constant flow rate. Maximum perfusion pressure was achieved within 10 min. Increasing doses of ACh (10-80- 10-%) were infused through a side-am into the perfusate immediately before entering the angiocatheter in the flap. When the presence of NO synthesis inhibitor was required, LNA was added to the reservoir (los5 M) 15-20 min before ACh infusion.
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
KRBHDSTBHN ET AL.
FIG. 2. Vascular resistance in isolated p e h s e d skin flaps before M) treatment. (pre-NE) and after (post-NE) norepinephrine Values are means f SEM; n = 7. Mean values s f vascular resistance for pre-NB and post-NE treatment groups are significantly different; paired t-test (r = 2.93, *p < 0.05).
FIG.3. Total net surface fluorometry in skin flaps before (pre-NE) and after (psst-NB) norepinephrine (1W6 M) treatment. Values are means f SEM; ;n = 7. Mean values of tot$ flap fluorometry for pre-NE and post-NE treatment groups are significantly different; paired t-test (P -- 2.76; *p < 0.05).
Drugs and chemiculs All drugs and chemicals were obtained from Sigma Chemical Company (St. Louis, Miss.) except the following: norepinephrine bitartrate (Winthrop Laboratories, Aurora, Ont ,) , fluorescein sodium (Alcon Canada Inc., Mississauga, Ont .), and nitroglycerin (Parke Davis, Searborough, Ont .). Data ana&ysis All values are expressed as means f standard error of the mean (SEM). The paired t-test was used for comparison of two means. P values reflect the Bonferroni correction for multiple outcomes (Sokal 1987). One-way analysis of variance (ANOVA) with repeated measures followed by the least significant difference test (Fisher's protected least significant difference test) were used in the compario n s of multiple 8ependent means (Milliken 1984). Probabilities with p I0.05 were. accepted as significant.
Results Study b :Endothebium-depedent and enhthebium-independeat skin vasodilation Skin flaps with intact vascular endothelium (i. e. , befire saponin pe@sion) Addition of NE ( I W g M) to the perfusate significantly ( p < 0.05; n = 7) increased the vascular resistance of the isolated perfused human skin Waps (Fig. 2). This NE treatment also significmdy ( p < 0.05) decreased the total net surface fluorometry in these skin Waps (Fig. 3). The distance of perfusion along the skin flap detected by fluorometry was reduced to 69 +- 12% of baseline following the addition of NE. ACh (lo-$ M) significantly ( p < 8.01; n = 7) attenuated the vascular resistance of skin flaps preconstricted with NE (lo-$ M) despite the presence of a cyclooxygenase inhibitor M) and a 0-adrenergic receptor (indomethacin, 3 x (Fig. 4). The total surface blocker (propranolol, lo-") fluorometry in these flaps was significantly ( p < 0.05) increased under conditions of constant pressure perfusion, but not when Wow rate was maintained constant (Fig. 5). This observation indicated that perfusion pressure must be sustained for local vascular relaxation to result in increased flow rate and hence tissue perfusion. The distances of perfusion as detected by fluorornetry under constant pressure or constant flow conditions were 156 &- 49 and $5 f 6.4 % of baseline,
FIG.4. Effect of acetylcholine (ACh) on the vascular resistance of skin flaps preconstrieted with norepinephrine (NE). The concentrations of NE and ACh were f W6M. Values are means f SEM; n = 7. Mean values of vascular resistance for NE and NE ACh treatment groups are significantly different; paired t-test (P = 5.20; *p < 0.05).
+
respectively. Similar to ACh, NTG treatment significantly ( p < 0.01; n = 7) reduced the vascular resistance of skin flaps preconstricted with NE (Fig. 6). Skin paps with damaged vascular endothelium (i. e., aper saponin pefision) Histobogy-The endothelid lining in skin and arterial pedicle specimens was studied under light microscopy. Before saponin perfusion, the endothelial lining of the arterial pedicle was fully intact, as discerned by the presence of endothelial cell nuclei and positive staining for factor %PHI.Conversely, the arterial pedicle was denuded of emdothelial cells following saponin perfusion. Examination of arteries within the dermal subdermal plexus revealed an intact endothelium prior to saponin perfusion (Fig. 71, but total or partial removal of endohelid cell lining following saponin perfusion (Fig. 8). Vascular r e a e t i v i - h e perfusion pressure increased over 200 m H g during the 5 min of saponin infusion. However, after 25 -40 min of saponin-free perfusion, the perfusion pres-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by WA STATE UNIV LIBRARIES on 11/27/14 For personal use only.
1212
CAN. d. PWYSIOL. PHARMACOL. VOL. 70, 1992
NE
NE+ACh
NE+ACh
NE
ME + NTG
FIG. 5. Effect of acetylcholine (ACh) on total flap Wuorometq in skin flaps preconstricted with norepinephrine (NE). Values are means f SEM; n = 7, Means without a common letter are significantly ( p < 64.05) different. Statistics: one-way ANOVA with repeated measures (F = 8,482) and Fisher's protected least significant difference test.
FIG.6. Effect of nitroglycerin (NTG) on the vascular resistance of skin Waps preconstricted with norepinephrine (NE). The concentrations of NE and NTG were 10s6M. Values are means f SEM; n = 7. Mean values s f vascular resistance for NE and NE + NTG treatment groups are significantly different; paired t-test (t = 4.60; *p < 0.05).
sure stabilized at 25-30 mmHg above the basal perfusion pressure. Since perfusion pressure is proportional to vascular resistance under constant flow condition, the vascular relaxation effect sf ACh and NTG can be expressed as a percentage of increased p r h s i o n pressure (tone) induced by NE (Ebeigbe eb a&.1890). Following saponin treatment, the ability for ACh (18-%M) to produce vascular relaxation in skin Waps preconstricted with NE was significantly ( p < 0.01) impaired (Fig. 9). In contrast, the ability of NTG (lo-$ M) to induce vascular relaxation in isolated perfused skin flaps remained unchanged following saponin treatment (Fig. 9).
period of about 4 h (Kreidstein et a&. 1891). Furthermore, in study 2 of the present experiment, a distinct concentrationdependent vascular relaxation effect of acetylcholine was observed (Fig. 10) and each perfusion study lasted only for about 2 h. Therefore, it is most unlikely that a significant endotoxin conmination could have been present in our preparations in studies 1 and 2 to confound the vascular relaxation effect of acetylcholine.
S t d y 2: lahibition of ACh-induced skin vascular relaation by a NO synthesis inhibitor The bas& perfusion pressure stabilized at 50-52 mmHg within 30 min (a = 6). Addition of NE (1W6 M) into the reservoir caused skin flap vasoconstriction, raising the perfusion pressure to 129 f 8 mmHg at constant flow condition. The net increase in perfhasion pressure was 79 f 8 mmMg. The vascular relaxation effect sf ACh in the absence and presence of ENA was studied in NE-preconstricted flaps under consBnt flow condition. Again, the vascular relaxation effect of ACh is expressed as a percentage of increased perhsion pressure (tone) induced by NE (Ebeigbe et al. 1990). Infusion of ACh in skin flags preconstrictd with NE (18-") caused vascular relaxation in a concentration-dependent manner (Fig. 10). However, h i s vascular relaxation effect of ACh at the dose range of BO-9-10-5 M was significantly ( p < 0.01) inhibited in the presence of 1W5 M LNA, a NO synthesis inhibitor. LNA alone did not have any effect on the basal perfusion pressure of the skin flaps. Possibility of endotmin-induced vascular relaxation due to csn~amination In the present experiment, fresh Milli-Q water was used to prepare Krebs buffer 2 h before each perhsion study. The Milli-Q water system was fed by distilled water. Using the same human sfin flap model, the same source of Milli-Q water, and the same batch of bovine serum albumin, we demonstrated previously that the perfusion pressure of control flaps did not change significantly (
