JOURNAL
OF SURGICAL
RESEARCH
Modulation
52,359-363
(19%)
of Lymphatic Smooth Muscle Contractile by the Endothelium
Responses
MARK K. FERGUSON, M.D. Department Presented
at the Annual
Meeting
of Surgery, The University
of the Association
of Chicago, Chicago, Illinois 60637
for Academic
The endothelium regulates smooth muscle tone in blood vessels through the release of endothelium-derived relaxing factor (EDRF). We hypothesized that the lymphatics possess endothelium-dependent relaxant properties analogous to those in blood vessels. Fresh porcine tracheobronchial lymphatic vessel rings were mounted in organ baths and connected to forcedisplacement transducers. Rings were submaximally precontracted with 10e6 M histamine and exposed to X cumulative addition of acetylcholine (ACH; lo-‘-3 10d4 M) or bradykinin (BRD; lo-*-3 X 10m6 M), both of which are known to promote release of EDRF. ACH caused concentration-dependent relaxation (maximum effect = -2.3 + 2.6% of initial histamine-induced active tension), while a similar but less pronounced effect was seen with BRD (39.6 + 5.4%). The relaxant effects of ACH and BRD were inhibited by collagenase pretreatment and mechanical endothelial denudation. The results confirm our hypothesis that lymphatic vessels possess endothelium-dependent relaxant properties and suggest that lymph vessels can regulate lymph flow through processes similar to those found in blood 0 1992 Academic Press, Inc. vessels.
INTRODUCTION
Recently it has been hypothesized that the lymphatics are active in regulating lymph flow from the lung and other mediastinal structures [l, 21. The mechanisms of these regulatory activities are poorly understood, although one possible route is through alterations in lymphatic smooth muscle tone. Tracheobronchial lymphatics are known to be affected by substances that promote contraction of lymphatic smooth muscle [3, 41. These findings emphasize the potential contribution of resistance effects to regulation of lymph flow by the tracheobronchial lymphatics. The endothelium regulates smooth muscle tone in blood vessels through a variety of mechanisms. Recent experimental work has focused on endothelium-derived relaxing factor, or EDRF, which is now known to be nitric oxide in the case of mammalian vascular smooth
Surgery, Colorado
Springs, Colorado,
November
20-23, 1991
muscle [5, 61. EDRF is released in response to several different stimuli, including increased arterial wall shear stress and administration of endothelium-dependent vasodilators such as acetylcholine and bradykinin [ 7-91. Release of EDRF, or nitric oxide, causes relaxation of smooth muscle in resistance arteries [lo] and has similar but less pronounced effects on veins [ll, 121. Recognizing that analogous methods of vascular smooth muscle tone regulation may exist in the lymphatic system, we postulated that tracheobronchial lymphatics possess endothelium-dependent relaxant properties. We studied the effects of acetylcholine- and bradykinin-stimulated EDRF release on precontracted porcine tracheobronchial lymph vessels with and without intact endothelium. We report in this study that lymph vessels release EDRF. These data suggest that tracheobronchial lymph vessels can regulate lymph flow through sophisticated processes similar to those found in blood vessels. MATERIALS
AND
METHODS
Tissue Preparation Blocks of mediastinal tissue from freshly slaughtered pigs (125-225 kg; male or female) were immersed in saline at 37°C. A 1% solution of Evans blue in saline was injected into lymph nodes at the tracheobronchial junction, and the tissue blocks were allowed to incubate for 30 min to permit staining of the efferent lymphatics. Vessels measuring 2-5 mm in diameter were ligated downstream and dissected sharply from surrounding tissue. The vessels were cut into 5-mm-width rings, each of which was suspended in a water-jacketed bath (10 ml) containing buffered Krebs solution (NaCl, 118 mM, NaHCO,, 24 mM; KCl, 4.7 mM, kH,PO,, 1.2 mM; CaCl,, 1.6 mM, MgCl,, 0.4 mM; dextrose, 5.5 mM, pH 7.4) at 37°C continuously aerated with 95% 0, and 5% CO,. Each vessel ring was mounted between two rigid stainless steel wires. One wire was attached to a glass rod within the bath. The other wire was fixed with a loop of 0000 silk ligature to a Grass FT-03 force-displacement transducer that was rack and pinion mounted to permit
359
OOZZ-4804/92 54.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
360
JOURNAL
OF SURGICAL
RESEARCH:
VOL. 52, NO. 4, APRIL
1992
125
stretching the tissue preparation to desired lengths. Vessel tensions were recorded on a Grass polygraph. Following a 1-hr equilibration period, resting tension was set at 600 mg and muscle viability was established by bathing the vessels in 65 mM KCl-substituted Krebs perfusate. Vessel rings were then submaximally precontracted with lo-’ M histamine. After the addition of KCl, histamine, or experimental drugs, total tension was measured, and active tension was calculated as the difference between it and the resting tension. At the conclusion of each experiment the length of each vessel ring was measured with a micrometer, and the rings were gently blotted and weighed.
100
75 50 25 0 -25 -7
-8 Acetylcholine and Bradykinin Intact Tissue
Effects in Endothelium-
Precontracted vessel rings were randomly assigned to exposure to cumulative doses of acetylcholine (1O-‘-1O-5 M), to bradykinin (10L8-3 X 10L6 M), or to additional doses of Krebs buffer as a control. The concentration at half maximum effect, or EC,,, was calculated for each experimental drug. Acetylcholine
Effects in Collugenase-Treuted
-6
Log Drug Concentration
-5 (M)
FIG. 1. Effects of acetylcholine and bradykinin on precontracted tracheobronchial lymphatic vessel rings with intact endothelium. Results shown are means + SEM of active tension expressed as a percentage of initial response to 10e5M histamine. Control vessel rings (circles; n = ‘7) were exposed only to histamine and additional doses of Krebs buffer. Other rings were exposed to cumulative addition of bradykinin (IO-“-3 X lOmeM, n = 16; solid triangles) or acetylcholine (lo-‘-lo-’ M, n = 11; open triangles). Asterisks indicate P < 0.05 compared to control vessel rings.
Tissue
Uncut vessels were cannulated, ligated distally, and distended at a pressure of 10 cm H,O for 5 min with either collagenase (2 mg/ml Krebs buffer) to remove the endothelium or Krebs buffer as a control solution. Vessels were then cut into rings, mounted, and submaximally precontracted with histamine as described above. Rings were then exposed to ‘cumulative addition of acetylcholine (lo-‘-3 X lop4 M).
between means in the experiment employing two groups were calculated using paired t tests. In the other experiments, differences among the experimental groups were examined using analysis of variance (ANOVA), and subsequent t tests using the Bonferonni correction were performed if statistically significant differences were present [ 131. Statistical significance was declared when
Acetylcholine
Drug Preparation
Effects in Endothelium-Stripped’Tissue
Vessel rings were randomly assigned to one of three treatments: (1) eversion, gentle rubbing of the endothelial surface with a moist cotton swab to remove the endothelium, followed by inversion (endothelium-stripped group); (2) eversion and inversion only (shamgroup); or (3) no manipulation (control group). Rings were then mounted, submaximally precontracted with histamine, and exposed to cumulative addition of acetylcholine as described above for collagenase-treated tissue. Bradykinin
EfiFects in Endothelium-Stripped
Tissue
Vessel rings randomly underwent one of three treatments as described above for endothelium-stripped tissue. They were then mounted, precontracted with histamine, and exposed to cumulative addition of bradykinin (10-8-3
Analysis
x 1o-g M).
of Data
Active tension was expressed as a percentage of initial vessel ring active tension generated in response to histamine. Data are expressed as means k SEM. Differences
P < 0.05.
The drugs used were histamine, acetylcholine, bradykinin, and collagenase (Sigma). All were dissolved in Krebs buffer and added to the baths to yield the concentrations mentioned in the text. RESULTS Data from 118 lymph vessel rings were used to tabulate the results. The mean vessel ring length at 500 mg resting tension was 4.05 + 0.09 mm and the average weight was 8.83 -t 0.21 mg. Initial active tensions developed in response to 65 mM KC1 averaged 980 f 69 mg, while those developed in response to 10e5M histamine averaged 1195 + 78 mg. The histamine response expressed as a percentage of the response to 65 mM KC1 was 128.0 + 3.5%. Acetylcholine and Bradykinin Intact Tissue
Effects in Endothelium-
Tissue rings precontracted with histamine maintained their generated active tensions through the duration of the experiment in the absence of exposure to acetylcholine or bradykinin (n = 7; Fig. 1). Cumulative ad-
MARK
K. FERGUSON:
LYMPHATIC
VASCULAR
ENDOTHELIAL
RELAXING
361
PROPERTIES
lOA M (-8.1 f 5.0% vs 6.8 f 4.4% of initial histamine response, respectively; P = NS). Endothelial-stripped vessel rings (n = 8) demonstrated attenuation of the usual response to acetylcholine, yielding an active tension of 54.3 + 5.9% of the initial histamine response at 10e5 M. Significant differences in active tension between the endothelium-stripped vessels and the sham and control groups were found at all concentrations of acetylcholine tested. Bradykinin
-7
-6
Log Acetylcholine
-5 Concentration
-4 (M)
FIG. 2. Effects of acetylcholine (10m7-3 X 10m4 M) on precontracted tracheobronchial lymphatic vessel rings collagenase-treated (n = 20; triangles) or with intact (n = 17; circles) endothelium. Results shown are means + SEM of active tension expressed as a percentage of initial response to 10m5M histamine. Asterisks indicate P < 0.05 compared to vessels with intact endothelium.
dition of acetylcholine caused a dose-dependent relaxation in the vessel rings (rz = ll), with a maximum effect of -2.3 + 2.6% of the initial histamine response at a concentration of 10e5 M. Bradykinin similarly relaxed precontracted vessel rings, with a maximum effect of 30.8 f 4.3% of the initial histamine response at a concentration of 3 X 10e6 M.(n = 16; P < 0.0001 vs acetylcholine). The calculated EC,, for acetylcholine was 3.4 X lo-’ M. The EC,, for bradykinin was 1.7 X lop7 M (P = NS). Acetylcholine
Effects in Collagenase-Treated
Effects in Endothelium-Stripped
Tissue
Control vessels displayed a typical dose-dependent relaxation response to bradykinin that reached a maximum at a concentration of 3 x 10e6 M (19.2 +- 7.4% of initial histamine response; n = 6). Sham-treated vessels showed a similar relaxation response curve that was not significantly different from that seen in the control vessel rings (n = 9; Fig. 4). Endothelium-stripped vessel rings (n = 9) demonstrated attenuation of the usual relaxation response to bradykinin that was significantly different from control rings but not sham-treated rings at the two highest concentrations tested (P = 0.011 and P = 0.007, respectively). DISCUSSION Maintenance of blood vessel smooth muscle tone is dependent in part on substances released from the vascular endothelium in response to local or circulating factors. The most comprehensively studied of these factors, EDRF, causes relaxation of smooth muscle in both arteries and veins. Recent information suggesting that many similarities exist between blood vessels and lymph ves-
Tissue
Collagenase treatment reduced the initial vessel ring contractile response to both 65 mM KC1 (737 + 88 in 20 treated rings vs 1644 + 217 mg in 17 control rings; P = 0.0007) and histamine (1100 + 170 vs 1957 +- 205 mg in controls; P = 0.0029). The histamine response expressed as a percentage of the KC1 response remained unchanged (128.9 + 6.3 vs 153.8 +- 12% in controls; P = NS). Collagenase treatment eliminated the relaxant effect produced by acetylcholine to a statistically significant extent at all concentrations of acetylcholine tested (Fig. 2). The maximum absolute difference between the collagenase-treated and control groups occurred at the highest acetylcholine concentration tested (3 X 10d4 M, 18.2 + 3.9 in controls vs 41.9 -+ 6.3% of initial histamine response in treated rings; P = 0.0038). Acetylcholine
Effects in Endothelium-Stripped
Tissue
Control (n = 6) and sham-treated (n = 9) vessel rings showed similar and normal concentration-dependent relaxation responses to acetylcholine (Fig. 3), with maximum effects reaching a plateau at a concentration of
100
80 60
-7
-6
Log Acetylcholine
-5 Concentration
-4 (M)
FIG. 3. Effects of acetylcholine (10e7-3 X 10d4 M) on precontracted tracheobronchial lymphatic vessel rings, comparing control rings (n = 6; circles) to sham-treated (n = 9; open triangles) and deendothelialized rings (n = 8; solid triangles). Results shown are means f SEM of active tension expressed as a percentage of initial response to 10e5 M histamine. Asterisks indicate P < 0.05 compared to both control and sham-treated rings.
362
JOURNAL
OF SURGICAL
RESEARCH:
80 60
i
t a
*O t I 0'
-8 Log Bradykinin
-7 Concentration
-6 (M)
FIG. 4. Effects of bradykinin (10-s-3 X 10m6M) on precontracted tracbeobronchial lymphatic vessel rings, comparing control rings (n = 6; circles) to sham-treated (n = 9; open triangles) and deendothelialized rings (n = 9; solid triangles). Results shown are means + SEM of active tension expressed as a percentage of initial response to 10M5M histamine. Asterisks indicate P < 0.05 compared to control rings.
sels prompted the hypothesis that lymph vessels also may be subject to regulation by endothelial factors. The findings in this report support the existence of an endothelium-dependent factor released by lymph vessels that causes smooth muscle relaxation. Lymphatics are muscular-type vessels that are primarily under myogenic control [14], but which are also affected by neural and pharmacologic stimuli [15-171. Thus, in addition to their traditional role as passive conduits for the passage of lymph fluid, lymph vessels are now thought to be actively involved in the regulation of lymph flow [2]. The means by which these regulatory activities work are not well understood, although one likely mechanism is through alterations in lymphatic smooth muscle tone. Lymphatic smooth muscle is known to respond to a variety of locally released and circulating substances. For example, mediators of inflammation such as histamine and 5-hydroxytryptamine are contractile agonists [3], as are epinephrine and prostaglandins A,, B,, and F,, 117,181. These findings support the possible role of resistance effects in the regulation of lymph flow. Contractions in resistance blood vessels are opposed by the influences of EDRF and other relaxant substances released locally by the endothelium. EDRF release may be stimulated in uiuo by increased arterial blood flow or by local release of bradykinin [7, 91, or experimentally by the application of acetylcholine [8]. Evidence that establishes a number of similarities between blood and lymphatic vessels is emerging. Based on information currently available, vasoconstrictive effects appear to predominate in lymph vessel reactivity to circulating substances [3,16]. This prompted the hypothe-
VOL.
52, NO. 4, APRIL
1992
sis that tracheobronchial lymph vessels also possess endothelium-dependent relaxant properties. Our data from normal vessels demonstrate that administration of either acetylcholine or bradykinin in vitro produces concentration-dependent relaxation of lymphatic tissue submaximally precontracted with histamine. These findings are similar to those previously reported for both arteries and veins [8-121. Removal of the lymphatic endothelium using collagenase resulted in a significant diminution of this relaxant effect, suggesting that it was endothelium dependent. Collagenase did cause smooth muscle dysfunction in the tissues, evidenced by a decrease in contractile activity in response to both KC1 and histamine. This prompted the use of mechanical techniques to deendothelialize the vessels in subsequent experiments, the data from which confirm the dependence of the relaxation phenomenon on the presence of normal endothelium. Histamine was used to increase the resting tension in the vessel rings because it is a known powerful contractile agonist of tracheobronchial lymphatic smooth muscle [ 3,4]. In one report, histamine was shown to produce relaxant responses in rat thoracic aorta smooth muscle through the activation of H, receptors located on the endothelium [19]. In the present model, endothelial denudation had no apparent effect on lymphatic vessel ring responses to histamine. These findings support the concept that histamine-induced contractile activity in lymph vessels is independent of the endothelial layer, suggesting that very different phenomena are operative in these two systems. Relaxation responses to acetylcholine and bradykinin differed in these experiments. Lymphatic tissue was sensitive to the same drug concentrations that promote relaxation in arteries and veins [8, 91, with bradykinin producing effects at concentrations half as great as those of acetylcholine. However, the amplitude of the bradykinin relaxation response was much less than that seen with acetylcholine. Acetylcholine appears to cause EDRF release through stimulation of muscarinic receptors located on the endothelium, while the mechanisms of EDRF release by bradykinin are less well understood [20]. In addition, bradykinin is known to stimulate release of arachidonic acid metabolites via both the cyclooxygenase and the lipoxygenase pathways [21], some of which may produce concurrent contractile responses in lymphatic smooth muscle. These additional effects may contribute to the blunted relaxation response seen with bradykinin. The identity of the EDRF released by porcine tracheobronchial lymphatic smooth muscle is unknown. EDRF in mammalian blood vessels has been identified as nitric oxide [5,6], a potent vasodilator with a half-life of only 6-50 set, making attempts to study it somewhat complicated. It is reasonable to suppose that EDRF in mammalian lymph vessels is also nitric oxide, although additional experiments are necessary to prove this.
MARK
K. FERGUSON:
LYMPHATIC
VASCULAR
Our data confirm the hypothesis that endothelial-dependent mechanisms causing smooth muscle relaxation exist in tracheobronchial lymphatic smooth muscle. These findings suggest that lymph vessels can regulate flow through sophisticated processes similar to those found in blood vessels.
ENDOTHELIAL
1.
Staub, M. C. Pulmonary edema. Physiol. Reu. 54: 678,1974. A. E. The lymphatic edema safety factor: The role of edema-dependent lymphatic factors (EDLF). Lymphology 23: 111,199o. 3. Ferguson, M. K., Williams, U., Leff, A. R., and Mitchell, R. W. Heterogeneity of tracheobronchial lymphatic smooth muscle responses to histamine and 5-hydroxytryptamine. Am. Rev. Resp. Dis. 143: A768, 1991. [Abstract] 4. Ferguson, M. K., and Tzeng, E. Attenuation of histamine induced lymphatic smooth muscle contractility by arachidonic acid. J. Surg. Res. 51: 500, 1991. 5. Palmer, R. M. J., Ferrige, A. G., and Moncada, S. Nitric oxide release accounts for the biological activity of endotheliumderived relaxing factor. Nature 327: 524, 1987. 6. Ignarro, L. J., Buga, G. M., Wood, K. S., Byrns, R. E., and Chaudhuri, G. Endothelium-derived relaxing factor produced and released from the artery and vein is nitric oxide. Proc. N&l. Acad. Sci. USA 84: 9265, 1987. I. Rubanyi, G. M., Romero, J. C., and Vanhoutte, P. M. Flow-induced release of endothelium-derived relaxing factor. Am. J. Physiol. 250: H1145, 1986. 8. Furchgott, R. F., and Zawadzki, J. V. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 288: 373, 1980. 9. Cherry, P. D., Furchgott, R. F., Zawadzki, J. V., and Jothianandan, D. Role of endothelial cells in relaxation of isolated arteries by bradykinin. Proc. Natl. Acad. Sci. USA 79: 2106, 1982.
PROPERTIES
363
lo. Furchgott,
R. F., Carvalho, M. H., Khan, M. T., and Matsunaga, K. Evidence for endothelium-dependent vasodilation of resistance vessels by acetylcholine. Blood Vessels 24: 145, 1987.
ll. De May, J. G., and Vanhoutte, P. M. Heterogeneous behavior of the canine arterial and venous wall. Circ. Res. 51: 439, 1982. 12. Luscher, T. F., Diederich, D., Siebenmann, R., Lehmann, K., Stulz, P., von Segesser, L., Yang, Z., Turina, M., Gradel, E., Weber, E., and Buhler, F. R. Difference between endotheliumdependent relaxation in arterial and venous coronary bypass grafts. N. Engl. J. Med. 319: 462, 1988.
REFERENCES
2. Taylor,
RELAXING
13.
Wallenstein, S., Zucker, C. L., and Fleiss, J. L. Some statistical methods useful in circulation research. Circ. Res. 47: 1, 1980.
14. Zweifach,
B. W., and Prather, J. W. Micromanipulation of pressure in terminal lymphatics in the mesentery. Am. J. Physiol.
228:1326,1975. 15. Allen, J. M., McHale,
N. G., and Rooney, B. M. Effect of norepinephrine on contractility of isolated mesenteric lymphatics. Am. J. Physiol. 244: H479, 1983.
16. Ferguson,
M. K., Shahinian, H. K., and Michelassi, F. Lymphatic smooth muscle responses to leukotrienes, histamine, and platelet activating factor. J. Surg. Res. 44: 172, 1988.
17. Ohhashi,
T., Kawai, Y., and Azuma, T. The response of lymphatic smooth muscles to vasoactive substances. Pfluegers Arch.
375:183,1978. 18.
Ohhashi, T., and Azuma, T. Variegated effects of prostaglandins on spontaneous activity in bovine mesenteric lymphatics. Microvast. Res. 27: 71, 1984.
19.
Van de Voorde, J., and Leusen, I. Role of the endothelium in the vasodilator response of rat thoracic aorta to histamine. Eur. J. Pharmacol. 87: 113,1983.
20. Vanhoutte,
P. M., Rubanyi, G. M., Miller, V. M., and Houston, D. S. Modulation of vascular smooth muscle contraction by the endothelium. Annu. Rev. Physiol. 48: 307, 1986.
21. Revtyak,
G. E., Hughes, M. J., Johnson, A. R., and Campbell, W. B. Histamine stimulation ofprostaglandin and HETE synthesis in human endothelial cells. Am. J. Physiol. 255: C214, 1988.
OF SURGICAL
RESEARCH
Modulation
52,359-363
(19%)
of Lymphatic Smooth Muscle Contractile by the Endothelium
Responses
MARK K. FERGUSON, M.D. Department Presented
at the Annual
Meeting
of Surgery, The University
of the Association
of Chicago, Chicago, Illinois 60637
for Academic
The endothelium regulates smooth muscle tone in blood vessels through the release of endothelium-derived relaxing factor (EDRF). We hypothesized that the lymphatics possess endothelium-dependent relaxant properties analogous to those in blood vessels. Fresh porcine tracheobronchial lymphatic vessel rings were mounted in organ baths and connected to forcedisplacement transducers. Rings were submaximally precontracted with 10e6 M histamine and exposed to X cumulative addition of acetylcholine (ACH; lo-‘-3 10d4 M) or bradykinin (BRD; lo-*-3 X 10m6 M), both of which are known to promote release of EDRF. ACH caused concentration-dependent relaxation (maximum effect = -2.3 + 2.6% of initial histamine-induced active tension), while a similar but less pronounced effect was seen with BRD (39.6 + 5.4%). The relaxant effects of ACH and BRD were inhibited by collagenase pretreatment and mechanical endothelial denudation. The results confirm our hypothesis that lymphatic vessels possess endothelium-dependent relaxant properties and suggest that lymph vessels can regulate lymph flow through processes similar to those found in blood 0 1992 Academic Press, Inc. vessels.
INTRODUCTION
Recently it has been hypothesized that the lymphatics are active in regulating lymph flow from the lung and other mediastinal structures [l, 21. The mechanisms of these regulatory activities are poorly understood, although one possible route is through alterations in lymphatic smooth muscle tone. Tracheobronchial lymphatics are known to be affected by substances that promote contraction of lymphatic smooth muscle [3, 41. These findings emphasize the potential contribution of resistance effects to regulation of lymph flow by the tracheobronchial lymphatics. The endothelium regulates smooth muscle tone in blood vessels through a variety of mechanisms. Recent experimental work has focused on endothelium-derived relaxing factor, or EDRF, which is now known to be nitric oxide in the case of mammalian vascular smooth
Surgery, Colorado
Springs, Colorado,
November
20-23, 1991
muscle [5, 61. EDRF is released in response to several different stimuli, including increased arterial wall shear stress and administration of endothelium-dependent vasodilators such as acetylcholine and bradykinin [ 7-91. Release of EDRF, or nitric oxide, causes relaxation of smooth muscle in resistance arteries [lo] and has similar but less pronounced effects on veins [ll, 121. Recognizing that analogous methods of vascular smooth muscle tone regulation may exist in the lymphatic system, we postulated that tracheobronchial lymphatics possess endothelium-dependent relaxant properties. We studied the effects of acetylcholine- and bradykinin-stimulated EDRF release on precontracted porcine tracheobronchial lymph vessels with and without intact endothelium. We report in this study that lymph vessels release EDRF. These data suggest that tracheobronchial lymph vessels can regulate lymph flow through sophisticated processes similar to those found in blood vessels. MATERIALS
AND
METHODS
Tissue Preparation Blocks of mediastinal tissue from freshly slaughtered pigs (125-225 kg; male or female) were immersed in saline at 37°C. A 1% solution of Evans blue in saline was injected into lymph nodes at the tracheobronchial junction, and the tissue blocks were allowed to incubate for 30 min to permit staining of the efferent lymphatics. Vessels measuring 2-5 mm in diameter were ligated downstream and dissected sharply from surrounding tissue. The vessels were cut into 5-mm-width rings, each of which was suspended in a water-jacketed bath (10 ml) containing buffered Krebs solution (NaCl, 118 mM, NaHCO,, 24 mM; KCl, 4.7 mM, kH,PO,, 1.2 mM; CaCl,, 1.6 mM, MgCl,, 0.4 mM; dextrose, 5.5 mM, pH 7.4) at 37°C continuously aerated with 95% 0, and 5% CO,. Each vessel ring was mounted between two rigid stainless steel wires. One wire was attached to a glass rod within the bath. The other wire was fixed with a loop of 0000 silk ligature to a Grass FT-03 force-displacement transducer that was rack and pinion mounted to permit
359
OOZZ-4804/92 54.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
360
JOURNAL
OF SURGICAL
RESEARCH:
VOL. 52, NO. 4, APRIL
1992
125
stretching the tissue preparation to desired lengths. Vessel tensions were recorded on a Grass polygraph. Following a 1-hr equilibration period, resting tension was set at 600 mg and muscle viability was established by bathing the vessels in 65 mM KCl-substituted Krebs perfusate. Vessel rings were then submaximally precontracted with lo-’ M histamine. After the addition of KCl, histamine, or experimental drugs, total tension was measured, and active tension was calculated as the difference between it and the resting tension. At the conclusion of each experiment the length of each vessel ring was measured with a micrometer, and the rings were gently blotted and weighed.
100
75 50 25 0 -25 -7
-8 Acetylcholine and Bradykinin Intact Tissue
Effects in Endothelium-
Precontracted vessel rings were randomly assigned to exposure to cumulative doses of acetylcholine (1O-‘-1O-5 M), to bradykinin (10L8-3 X 10L6 M), or to additional doses of Krebs buffer as a control. The concentration at half maximum effect, or EC,,, was calculated for each experimental drug. Acetylcholine
Effects in Collugenase-Treuted
-6
Log Drug Concentration
-5 (M)
FIG. 1. Effects of acetylcholine and bradykinin on precontracted tracheobronchial lymphatic vessel rings with intact endothelium. Results shown are means + SEM of active tension expressed as a percentage of initial response to 10e5M histamine. Control vessel rings (circles; n = ‘7) were exposed only to histamine and additional doses of Krebs buffer. Other rings were exposed to cumulative addition of bradykinin (IO-“-3 X lOmeM, n = 16; solid triangles) or acetylcholine (lo-‘-lo-’ M, n = 11; open triangles). Asterisks indicate P < 0.05 compared to control vessel rings.
Tissue
Uncut vessels were cannulated, ligated distally, and distended at a pressure of 10 cm H,O for 5 min with either collagenase (2 mg/ml Krebs buffer) to remove the endothelium or Krebs buffer as a control solution. Vessels were then cut into rings, mounted, and submaximally precontracted with histamine as described above. Rings were then exposed to ‘cumulative addition of acetylcholine (lo-‘-3 X lop4 M).
between means in the experiment employing two groups were calculated using paired t tests. In the other experiments, differences among the experimental groups were examined using analysis of variance (ANOVA), and subsequent t tests using the Bonferonni correction were performed if statistically significant differences were present [ 131. Statistical significance was declared when
Acetylcholine
Drug Preparation
Effects in Endothelium-Stripped’Tissue
Vessel rings were randomly assigned to one of three treatments: (1) eversion, gentle rubbing of the endothelial surface with a moist cotton swab to remove the endothelium, followed by inversion (endothelium-stripped group); (2) eversion and inversion only (shamgroup); or (3) no manipulation (control group). Rings were then mounted, submaximally precontracted with histamine, and exposed to cumulative addition of acetylcholine as described above for collagenase-treated tissue. Bradykinin
EfiFects in Endothelium-Stripped
Tissue
Vessel rings randomly underwent one of three treatments as described above for endothelium-stripped tissue. They were then mounted, precontracted with histamine, and exposed to cumulative addition of bradykinin (10-8-3
Analysis
x 1o-g M).
of Data
Active tension was expressed as a percentage of initial vessel ring active tension generated in response to histamine. Data are expressed as means k SEM. Differences
P < 0.05.
The drugs used were histamine, acetylcholine, bradykinin, and collagenase (Sigma). All were dissolved in Krebs buffer and added to the baths to yield the concentrations mentioned in the text. RESULTS Data from 118 lymph vessel rings were used to tabulate the results. The mean vessel ring length at 500 mg resting tension was 4.05 + 0.09 mm and the average weight was 8.83 -t 0.21 mg. Initial active tensions developed in response to 65 mM KC1 averaged 980 f 69 mg, while those developed in response to 10e5M histamine averaged 1195 + 78 mg. The histamine response expressed as a percentage of the response to 65 mM KC1 was 128.0 + 3.5%. Acetylcholine and Bradykinin Intact Tissue
Effects in Endothelium-
Tissue rings precontracted with histamine maintained their generated active tensions through the duration of the experiment in the absence of exposure to acetylcholine or bradykinin (n = 7; Fig. 1). Cumulative ad-
MARK
K. FERGUSON:
LYMPHATIC
VASCULAR
ENDOTHELIAL
RELAXING
361
PROPERTIES
lOA M (-8.1 f 5.0% vs 6.8 f 4.4% of initial histamine response, respectively; P = NS). Endothelial-stripped vessel rings (n = 8) demonstrated attenuation of the usual response to acetylcholine, yielding an active tension of 54.3 + 5.9% of the initial histamine response at 10e5 M. Significant differences in active tension between the endothelium-stripped vessels and the sham and control groups were found at all concentrations of acetylcholine tested. Bradykinin
-7
-6
Log Acetylcholine
-5 Concentration
-4 (M)
FIG. 2. Effects of acetylcholine (10m7-3 X 10m4 M) on precontracted tracheobronchial lymphatic vessel rings collagenase-treated (n = 20; triangles) or with intact (n = 17; circles) endothelium. Results shown are means + SEM of active tension expressed as a percentage of initial response to 10m5M histamine. Asterisks indicate P < 0.05 compared to vessels with intact endothelium.
dition of acetylcholine caused a dose-dependent relaxation in the vessel rings (rz = ll), with a maximum effect of -2.3 + 2.6% of the initial histamine response at a concentration of 10e5 M. Bradykinin similarly relaxed precontracted vessel rings, with a maximum effect of 30.8 f 4.3% of the initial histamine response at a concentration of 3 X 10e6 M.(n = 16; P < 0.0001 vs acetylcholine). The calculated EC,, for acetylcholine was 3.4 X lo-’ M. The EC,, for bradykinin was 1.7 X lop7 M (P = NS). Acetylcholine
Effects in Collagenase-Treated
Effects in Endothelium-Stripped
Tissue
Control vessels displayed a typical dose-dependent relaxation response to bradykinin that reached a maximum at a concentration of 3 x 10e6 M (19.2 +- 7.4% of initial histamine response; n = 6). Sham-treated vessels showed a similar relaxation response curve that was not significantly different from that seen in the control vessel rings (n = 9; Fig. 4). Endothelium-stripped vessel rings (n = 9) demonstrated attenuation of the usual relaxation response to bradykinin that was significantly different from control rings but not sham-treated rings at the two highest concentrations tested (P = 0.011 and P = 0.007, respectively). DISCUSSION Maintenance of blood vessel smooth muscle tone is dependent in part on substances released from the vascular endothelium in response to local or circulating factors. The most comprehensively studied of these factors, EDRF, causes relaxation of smooth muscle in both arteries and veins. Recent information suggesting that many similarities exist between blood vessels and lymph ves-
Tissue
Collagenase treatment reduced the initial vessel ring contractile response to both 65 mM KC1 (737 + 88 in 20 treated rings vs 1644 + 217 mg in 17 control rings; P = 0.0007) and histamine (1100 + 170 vs 1957 +- 205 mg in controls; P = 0.0029). The histamine response expressed as a percentage of the KC1 response remained unchanged (128.9 + 6.3 vs 153.8 +- 12% in controls; P = NS). Collagenase treatment eliminated the relaxant effect produced by acetylcholine to a statistically significant extent at all concentrations of acetylcholine tested (Fig. 2). The maximum absolute difference between the collagenase-treated and control groups occurred at the highest acetylcholine concentration tested (3 X 10d4 M, 18.2 + 3.9 in controls vs 41.9 -+ 6.3% of initial histamine response in treated rings; P = 0.0038). Acetylcholine
Effects in Endothelium-Stripped
Tissue
Control (n = 6) and sham-treated (n = 9) vessel rings showed similar and normal concentration-dependent relaxation responses to acetylcholine (Fig. 3), with maximum effects reaching a plateau at a concentration of
100
80 60
-7
-6
Log Acetylcholine
-5 Concentration
-4 (M)
FIG. 3. Effects of acetylcholine (10e7-3 X 10d4 M) on precontracted tracheobronchial lymphatic vessel rings, comparing control rings (n = 6; circles) to sham-treated (n = 9; open triangles) and deendothelialized rings (n = 8; solid triangles). Results shown are means f SEM of active tension expressed as a percentage of initial response to 10e5 M histamine. Asterisks indicate P < 0.05 compared to both control and sham-treated rings.
362
JOURNAL
OF SURGICAL
RESEARCH:
80 60
i
t a
*O t I 0'
-8 Log Bradykinin
-7 Concentration
-6 (M)
FIG. 4. Effects of bradykinin (10-s-3 X 10m6M) on precontracted tracbeobronchial lymphatic vessel rings, comparing control rings (n = 6; circles) to sham-treated (n = 9; open triangles) and deendothelialized rings (n = 9; solid triangles). Results shown are means + SEM of active tension expressed as a percentage of initial response to 10M5M histamine. Asterisks indicate P < 0.05 compared to control rings.
sels prompted the hypothesis that lymph vessels also may be subject to regulation by endothelial factors. The findings in this report support the existence of an endothelium-dependent factor released by lymph vessels that causes smooth muscle relaxation. Lymphatics are muscular-type vessels that are primarily under myogenic control [14], but which are also affected by neural and pharmacologic stimuli [15-171. Thus, in addition to their traditional role as passive conduits for the passage of lymph fluid, lymph vessels are now thought to be actively involved in the regulation of lymph flow [2]. The means by which these regulatory activities work are not well understood, although one likely mechanism is through alterations in lymphatic smooth muscle tone. Lymphatic smooth muscle is known to respond to a variety of locally released and circulating substances. For example, mediators of inflammation such as histamine and 5-hydroxytryptamine are contractile agonists [3], as are epinephrine and prostaglandins A,, B,, and F,, 117,181. These findings support the possible role of resistance effects in the regulation of lymph flow. Contractions in resistance blood vessels are opposed by the influences of EDRF and other relaxant substances released locally by the endothelium. EDRF release may be stimulated in uiuo by increased arterial blood flow or by local release of bradykinin [7, 91, or experimentally by the application of acetylcholine [8]. Evidence that establishes a number of similarities between blood and lymphatic vessels is emerging. Based on information currently available, vasoconstrictive effects appear to predominate in lymph vessel reactivity to circulating substances [3,16]. This prompted the hypothe-
VOL.
52, NO. 4, APRIL
1992
sis that tracheobronchial lymph vessels also possess endothelium-dependent relaxant properties. Our data from normal vessels demonstrate that administration of either acetylcholine or bradykinin in vitro produces concentration-dependent relaxation of lymphatic tissue submaximally precontracted with histamine. These findings are similar to those previously reported for both arteries and veins [8-121. Removal of the lymphatic endothelium using collagenase resulted in a significant diminution of this relaxant effect, suggesting that it was endothelium dependent. Collagenase did cause smooth muscle dysfunction in the tissues, evidenced by a decrease in contractile activity in response to both KC1 and histamine. This prompted the use of mechanical techniques to deendothelialize the vessels in subsequent experiments, the data from which confirm the dependence of the relaxation phenomenon on the presence of normal endothelium. Histamine was used to increase the resting tension in the vessel rings because it is a known powerful contractile agonist of tracheobronchial lymphatic smooth muscle [ 3,4]. In one report, histamine was shown to produce relaxant responses in rat thoracic aorta smooth muscle through the activation of H, receptors located on the endothelium [19]. In the present model, endothelial denudation had no apparent effect on lymphatic vessel ring responses to histamine. These findings support the concept that histamine-induced contractile activity in lymph vessels is independent of the endothelial layer, suggesting that very different phenomena are operative in these two systems. Relaxation responses to acetylcholine and bradykinin differed in these experiments. Lymphatic tissue was sensitive to the same drug concentrations that promote relaxation in arteries and veins [8, 91, with bradykinin producing effects at concentrations half as great as those of acetylcholine. However, the amplitude of the bradykinin relaxation response was much less than that seen with acetylcholine. Acetylcholine appears to cause EDRF release through stimulation of muscarinic receptors located on the endothelium, while the mechanisms of EDRF release by bradykinin are less well understood [20]. In addition, bradykinin is known to stimulate release of arachidonic acid metabolites via both the cyclooxygenase and the lipoxygenase pathways [21], some of which may produce concurrent contractile responses in lymphatic smooth muscle. These additional effects may contribute to the blunted relaxation response seen with bradykinin. The identity of the EDRF released by porcine tracheobronchial lymphatic smooth muscle is unknown. EDRF in mammalian blood vessels has been identified as nitric oxide [5,6], a potent vasodilator with a half-life of only 6-50 set, making attempts to study it somewhat complicated. It is reasonable to suppose that EDRF in mammalian lymph vessels is also nitric oxide, although additional experiments are necessary to prove this.
MARK
K. FERGUSON:
LYMPHATIC
VASCULAR
Our data confirm the hypothesis that endothelial-dependent mechanisms causing smooth muscle relaxation exist in tracheobronchial lymphatic smooth muscle. These findings suggest that lymph vessels can regulate flow through sophisticated processes similar to those found in blood vessels.
ENDOTHELIAL
1.
Staub, M. C. Pulmonary edema. Physiol. Reu. 54: 678,1974. A. E. The lymphatic edema safety factor: The role of edema-dependent lymphatic factors (EDLF). Lymphology 23: 111,199o. 3. Ferguson, M. K., Williams, U., Leff, A. R., and Mitchell, R. W. Heterogeneity of tracheobronchial lymphatic smooth muscle responses to histamine and 5-hydroxytryptamine. Am. Rev. Resp. Dis. 143: A768, 1991. [Abstract] 4. Ferguson, M. K., and Tzeng, E. Attenuation of histamine induced lymphatic smooth muscle contractility by arachidonic acid. J. Surg. Res. 51: 500, 1991. 5. Palmer, R. M. J., Ferrige, A. G., and Moncada, S. Nitric oxide release accounts for the biological activity of endotheliumderived relaxing factor. Nature 327: 524, 1987. 6. Ignarro, L. J., Buga, G. M., Wood, K. S., Byrns, R. E., and Chaudhuri, G. Endothelium-derived relaxing factor produced and released from the artery and vein is nitric oxide. Proc. N&l. Acad. Sci. USA 84: 9265, 1987. I. Rubanyi, G. M., Romero, J. C., and Vanhoutte, P. M. Flow-induced release of endothelium-derived relaxing factor. Am. J. Physiol. 250: H1145, 1986. 8. Furchgott, R. F., and Zawadzki, J. V. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 288: 373, 1980. 9. Cherry, P. D., Furchgott, R. F., Zawadzki, J. V., and Jothianandan, D. Role of endothelial cells in relaxation of isolated arteries by bradykinin. Proc. Natl. Acad. Sci. USA 79: 2106, 1982.
PROPERTIES
363
lo. Furchgott,
R. F., Carvalho, M. H., Khan, M. T., and Matsunaga, K. Evidence for endothelium-dependent vasodilation of resistance vessels by acetylcholine. Blood Vessels 24: 145, 1987.
ll. De May, J. G., and Vanhoutte, P. M. Heterogeneous behavior of the canine arterial and venous wall. Circ. Res. 51: 439, 1982. 12. Luscher, T. F., Diederich, D., Siebenmann, R., Lehmann, K., Stulz, P., von Segesser, L., Yang, Z., Turina, M., Gradel, E., Weber, E., and Buhler, F. R. Difference between endotheliumdependent relaxation in arterial and venous coronary bypass grafts. N. Engl. J. Med. 319: 462, 1988.
REFERENCES
2. Taylor,
RELAXING
13.
Wallenstein, S., Zucker, C. L., and Fleiss, J. L. Some statistical methods useful in circulation research. Circ. Res. 47: 1, 1980.
14. Zweifach,
B. W., and Prather, J. W. Micromanipulation of pressure in terminal lymphatics in the mesentery. Am. J. Physiol.
228:1326,1975. 15. Allen, J. M., McHale,
N. G., and Rooney, B. M. Effect of norepinephrine on contractility of isolated mesenteric lymphatics. Am. J. Physiol. 244: H479, 1983.
16. Ferguson,
M. K., Shahinian, H. K., and Michelassi, F. Lymphatic smooth muscle responses to leukotrienes, histamine, and platelet activating factor. J. Surg. Res. 44: 172, 1988.
17. Ohhashi,
T., Kawai, Y., and Azuma, T. The response of lymphatic smooth muscles to vasoactive substances. Pfluegers Arch.
375:183,1978. 18.
Ohhashi, T., and Azuma, T. Variegated effects of prostaglandins on spontaneous activity in bovine mesenteric lymphatics. Microvast. Res. 27: 71, 1984.
19.
Van de Voorde, J., and Leusen, I. Role of the endothelium in the vasodilator response of rat thoracic aorta to histamine. Eur. J. Pharmacol. 87: 113,1983.
20. Vanhoutte,
P. M., Rubanyi, G. M., Miller, V. M., and Houston, D. S. Modulation of vascular smooth muscle contraction by the endothelium. Annu. Rev. Physiol. 48: 307, 1986.
21. Revtyak,
G. E., Hughes, M. J., Johnson, A. R., and Campbell, W. B. Histamine stimulation ofprostaglandin and HETE synthesis in human endothelial cells. Am. J. Physiol. 255: C214, 1988.
