Airway Blood Flow Modifies Allergic Airway Smooth Muscle Contraction 1- 3

MARC E. CSETE, ALEJANDRO D. CHEDIAK, WILLIAM M. ABRAHAM, ,and ADAM WANNER

Introduction

In patients with severe bronchial asthma, the airway wall is characterized by inflammatory changes, including vascular congestion and edema (1). This suggests that the airway circulation participates in the physiologic manifestations of bronchial asthma. Bronchial blood flow and microvascular permeability have not been measured in humans because of the invasiveness of the required methodology. However, in animal models of airway anaphylaxis, increases in bronchial blood flow and microvascular hyperpermeability for macromolecules have been demonstrated (2-4). These vascular responses could contribute to airway narrowing during allergic bronchoconstriction by thickening the mucosa, an effect that would be most noticeable in small airways (5, 6). Wereasoned that the increase in airway blood flow might also blunt the magnitude and shorten the duration of antigen-induced airway smooth muscle contraction by enhancing the clearance of locally released spasmogenic mediators. With the present investigation, we therefore wished to determine if the pattern of antigeninduced airway smooth muscle contraction in the trachea of allergic sheep can be modified by manipulating local blood flow. Methods In Vivo Studies Experimental techniques. Thirteen adult ewes (mean weight, 32 kg; range, 19to 42 kg; mean age, 38months; range, 18to 108months) with natural cutaneous hypersensitivity to Ascaris suum extract wereselectedon the basis of their previously demonstrated airway responsiveness to the same antigen (7, 8). All animals exhibited repeatedly either an immediate or a dual increase in pulmonary airflow resistance (RL)after inhalation challenge withA. suum extract. They were used in more than one experiment; at least 2 wk were allowed to elapse between antigen challenges. The sheep were secured, with their body resting on a sling suspended on a mobile cart. After nasotracheal intubation (vide infra), their

SUMMARY We te.ted the hypothe.l. that airway perfu.lon modlfle. the contractile ....pon.. of airway .mooth muscle to allergen challenge by Influencing the clearance of locally released apamogen•• In .Ix Intact, lightly sedated, .heep allergic to Ascaris suum, we measured tracheal mucoul blood flow (Otr) with a soluble gas uptake method and tracheal dead .pace (VIr), an Index of airway .mooth muscle tone, by helium dilution before and aerially after local aerosol challenge with A. suum extract or ragweed extract (control). The former challenge was...peated during contlnuou.lntraveneue Infu.lon of either vaaop.....ln or nitroglycerin, which by them..lves had no effect on VIr and decreased and Increased Otr, re.pectlvely. Ragweed had no effect on Otr and VIr, whe...a. A. suum Increased mean (± SE) Otr by 111 ± 31% (p < 0.05) and dec ...ased mean VIr by 15 ± 2~ (p < 0.05) Immediately after challenge, with Otr returning to baseline by 40 min and VIr by 80 min. V8eopressln Infu.lon prevented the A. suum-Induced Increase In atr and prolonged the decre_ In mean Vir (p < 0.05). During nitroglycerin Infu.lon, A. suum failed to alter atr or VIr. Vuopre..ln and nitroglycerin had no effect on the contractile re.ponse. of tracheal .mooth muscle to A. suum In rlfro. These re.ults Indicate that the effects of vaaopreaaln and nitroglycerin on antigen-Induced airway .mooth muscle contraction In rlro were due to alteration. In airway blood flow rather than to alteration. In the release of or airway .mooth muscle re.pon.lvene.. to chemical mediators. All REV RESPIR DIS 1111; 144:51-13

heads were immobilized with adhesive tape on a cushioned head rest extending over the front of the cart. The sheep had been previously surgically prepared with an externalized carotid loop covered by a skin flap to allow placement of an indwelling needle under local anesthesia with 2010 xylocaine solution (9). The needle was connected via polyethelene tubing to a strain gauge (Model P23XL; Gould Instruments, Oxnard, CA). The gauge was leveled with the arterial needle to measure mean systemic arterial pressure (Psa), which was recorded continuously on a physiologic recorder (Model 7DAG; Grass Instruments, Quincy, MA). Arterial Pol' Pcos, and pH were determined at 37° C (Model ABL 30; Radiometer, Rotterdam, Holland) using blood samples obtained anaerobically from the carotid loop. Another indwelling catheter was placed in either a leg vein or an internal jugular vein and used for the administration of pharmacologic agents. Heparinized physiologic saline flushes were used to keep the lines patent. A plastic tube (outer diameter, 7 mm) with two inflatable cuffs 10em apart was then introduced transnasally into the trachea under topical anesthesia with 2010 xylocaine solution applied to the nasal passage. Under direct visualization with a fiberoptic bronchoscope, the tube was positioned such that the proximal cuff was situated below the larynx and the distal cuff above the orifice of the right upper lobe bronchus, which arises from the trachea. Inflation of the two cuffs created a

tracheal chamber with wallsthat weredefined by the outside of the endotracheal tube, the two cuffs, and the tracheal mucosa between the two cuffs. Incorporated into the chamber weretwo venting catheters (ID, 4 mm) that extended from the proximal and distal ends of the chamber to the outside of the animal. The catheters were not occluded; therefore, the tracheal chamber contained room air except during physiologic measurements (vide infra). Through the animal's other nostril, a balloon catheter was introduced and placed in the distal esophagus between 5 and 10 em from the gastroesophageal junction to estimate pleural pressure; in this location, cardiac oscillations were clearly discernible on the pressure tracings, and inspiration created a negative deflection. The animals were then sedated intravenously with sodium pentobarbital (2 mg/kg); this dose was repeated every 30 to 60 min to maintain light sedation throughout the experiment. Once sedated, the (Received in original form April 24, 1990 and in revised form November 21, 1990) 1 From the Division of Pulmonary Disease, University of Miami at Mount Sinai Medical Center, Miami Beach, Florida. 2 Supported by Grant HL-20989 from the National Heart, Lung, and Blood Institute. 3 Correspondence and requests for reprints should be addressed to Adam Wanner, M.D., Division of Pulmonary Disease, Mount Sinai Medical Center, 4300 Alton Road, Miami Beach, FL 33140.

59

60

animals were mechanically ventilated with air using a Harvard animal respirator at 10 breaths/min (Harvard Apparatus Co., South Natik, MA). Arterial blood gas composition was monitored periodically. Tidal volume was chosen so as to maintain an arterial PC02 between 32 and 40 mm Hg, In all experiments, the arterial Po2 remained above 70 mm Hg, thus obviating the need for supplemental oxygen. For the measurement of RL,a side-hole catheter was introduced through the endotracheal tube and positioned in the distal trachea, and transpulmonary pressure was determined by referencing tracheal pressure to pleural pressure using a differential pressure transducer (Model DP-45; Validyne Corp., Northridge, CAl. The flow signal at the airway opening and the transpulmonary pressure signal were fed into a digital computer for on-line calculation of RL (7); the average value of five to 10 breaths free of swallowing artifact was determined. Thoracic gas volume was measured by briefly disconnecting the sheep from the ventilator and placing them in a constant-volume body plethysmograph (7). Specific pulmonary airflow resistance (sRL)was obtained by multiplying RLby thoracic gas volume. Tracheal dead space (Vtr), an index of airway smooth muscle tone, was measured by helium dilution, and tracheal mucosal blood flow (Qtr) was measured by determining the uptake of dimethyl ether as previously described (10). The known volume of the endotracheal tube between the two cuffs was added to the dead space determined by helium dilution to obtain Vtr. Through the catheter leading to the distal part of the tracheal chamber, 50ml of dimethyl ether wereinjected at fractional concentrations between 0.85 and 0.95 as determined with a mass spectrometer (Perkin-Elmer, Pomona, CAl prior to injection. The syringe used for injection was heated by a coil perfused with water at 380 C (sheep body temperature). At the end of injection, the empty syringe was left in place to create a seal. The proximal catheter venting the tracheal chamber was connected via a l00-ml dead space to a no. 00 Fleisch pneumotachograph. The pressure drop across the pneumotachograph was sensed by a Validyne differential gauge (Model CD-12). The signal was amplified, filtered (to eliminate cardiac oscillations), and recorded on a Grass polygraph (Model 78-0). After completion of dimethyl ether injection, the pneumotachograph flow reversedtoward the tracheal chamber as dimethyl ether was absorbed, became transiently high as dimethyl ether equilibrated with tracheal tissue, and then decreased to a steady state that was entirely dependent on Qtr. Qtr was calculated from the steadystate uptake of dimethyl ether, the dimethyl ether concentration in the tracheal chamber (the same as the injected dimethyl ether concentration), and the solubility coefficient of dimethyl ether in tissue. The pneumotachograph signal was fed into a ~igital computer for on-line calculation of Qtr. Because Qtr is dependent on perfusion pressure, Qtr was

CSETE, CHEDIAK, ABRAHAM, AND WANNER

also expressed as Qtr normalized for Psa (normalized Qtr, Qtrn = Qtr Psa- l·l02). All measurements were recorded in triplicate, and the mean was used for data analysis. Experimentalprotocols. The effects of antigen challenge on Vtr, Qtr, and RL were determined as follows. A solution of crude A. suum extract (Greer Diagnostics, Lenoir, NC) prepared by diluting a stock solution with phosphate-buffered saline to a final concentration of 82,000 PNU/ml was aerosolized by an ultrasonic nebulizer (Monahan Model 670; Monahan, Plattsburg, NY) and directed into the tracheal chamber through the distal catheter, with the proximal catheter serving as a vent. Thus, only the trachea was exposed to antigen. The aerosol (mass median aerodynamic diameter [MMAD), 3 JJ.m; geometric SD, 1.4 JJ.m as determined by an Anderson cascade particle size analyzer) was delivered at an airflow of 3 L'min? over a period of 5 min. One milliliter of the solution was nebulized over this period of time. The physiologic measurements (which took about 10 min to complete) were made before and immediately after challenge, and then 40 and 80 min later. 1\vo control experiments were carried out. In the first, ragweed extract (Greer Diagnostics) in phosphate-buffered saline at a concentration of 20,000 PNU/ml was nebulized for tracheal challenge (sheep have no sensitivity to this antigen). In the second, A. mum extract in phosphate-buffered saline (82,000 PNU/ml) was nebulized by a side-arm jet nebulizer (MMAD, 3.1 JJ.m; SD, 2.0 JJ.m) inserted into the inspiratory line of the Harvard respirator (20 min) for bronchopulmonary challenge without tracheal challenge. The vasodilator nitroglycerin (NTG) and the vasopressor vasopressin (VP) were used to assess the role of Qtr in the response of Vtr to antigen challenge. A continuous intravenous infusion of NTG (160JJ.g·min- l ) or VP (0.12 units-min") in physiologic saline was

started. 1\venty minutes later, tracheal A. suum challenge was carried out as described above. Measurements of Qtr and Vtr were made before starting the infusion and before, immediately after, and 20 and 40 min after antigen challenge. The same measurements were made in control experiments in which the NTG or VP infusions were not combined with antigen challenge. In Vitro Studies Experimental techniques. Fourteen adult ewes (weighing 17 to 54 kg) with natural cutaneous hypersensitivity to A. suum extract were used for these experiments. The animals were killed with a rapid intravenous injection of sodium pentobarbital (25 mg-kg"), and the tracheas were immediately removed and placed in cold oxygenated Krebs-Henseleit solution (composition in mM: Na", 143.9; K+, 5.9; Mg++, 1.2; Ca++, 2.5; CI-, 126.0; HCO-3, 2.50; H 2PO-", 1.3; SO..-2, 1.2; glucose, 5.6). The tracheas were used within 24 h of death. Tracheal strips were obtained by cutting single rings from the middle of the trachea, removing the connective tissue and fat, and then transecting the ventral portion of the cartilage. Each strip was suspended by sutures in a 4O-ml organ bath containing KrebsHenseleit solution (pH, 7.50) maintained at 380 C and bubbled with a gas mixture of 950/0 oxygen and 50/0 carbon dioxide. One suture was connected to a force displacement transducer (Grass Model FT03) that was linked to a Grass polygraph (Model 7D) to record isometric tension. A resting tension of 2 g was placed on the muscle, and the muscle was allowed to equilibrate for at least 1 h. During the period of equilibration, the Krebs-Henseleit bathing solution was changed four times. Experimentalprotocol. Paired tissues (i.e., tissues from the same animal) were either untreated (controls) or pretreated with VP (18 mU·ml- l ) or NTG (1,3, 10 nM, 0.1 JJ.M), 15

TABLE 1 BASELINE PHYSIOLOGIC PARAMETERS·

Protocol

Animals Studied (n)

Otrn (mfmin- 1·mm Hg-1'1CJ2)

Vtr (ml)

Vtrn (ml·kg-')

SRL (cm H2 O s)

Intratracheal A. suurri

6

0.95 (0.08)

38.7 (1.3)

1.04 (0.13)

1.3 (0.2)

Intratracheal ragweed

6

1.43 (0.27)

39.7 (1.7)

1.04 (0.07)

NO

Bronchopulmonary A. suum

6

1.11 (0.45)

39.6 " (2.1)

1.05 (0.12)

1.0 (0.1)

Nitroglycerin and intratracheal A. suum

5

1.17 (0.16)

19.6 (0.6)

0192 (0.13)

NO

Vasopressin and intratracheal A. suum

5

1.51 (0.07)

27.4 (3.1)

0.98 (0.08)

NO

Nitroglycerin

5

1.18 (0.09)

19.2 (0.6)

0.96 (0.13)

NO

Vasopressin

5

1.29 (0.07)

19.0 (0.6)

0.98 (0.08)

NO

Definitionof abbreviations: Otrn • trachealmucosalbloodflow normalized for meansystemicarterialpressure; Vtr • tracheal deadspace;Vtrn • Vtr normalized for body weight; SRL • specificbronchopulmonary airflow resistance; NO • not determined. * Values are meanwith SE shownIn parentheses.

81

AIRWAY BLOOD FLOW MODIFIES ALLERGIC AIRWAY SMOOTH MUSCLE CONTRACTION

min prior to the addition of purified A. suum antigen (10to 100tJ.I; 164,000PNU/tJ.I). Contractile responses to A. suum were expressed as a percentage of the contractile response to 10mM acetylcholine. Doses of VP and NTG used had no effect on resting smooth muscle tone nor did the doses of NTG relax sheep tracheal smooth muscle precontracted with 1 JJM acetylcholine.

Statistical Analyses Data were expressed as mean (± SEM). Changes over time wereassessed in the different treatment groups using an analysis ofvariance. When appropriate, multiple comparisons weremade using the Newman-Keulstest. A p value of less than 0.05 using a twotailed analysis was considered statistically significant. Results

In Vivo Studies Mean baseline Qtrn and RL (when applicable) were similar under the sevenexperimental conditions (table 1). In contrast, there was a considerable variation in mean baseline Vtr among the different groups because of differences in animal weight and hence tracheal size.However, the variability in weight and Vtr within groups was small. Because wewere interested in both tracheal vasomotion and perfusion, the data were expressed as Qbr and Qtrn (index of vasomotion). The two parameters showed similar relative changes (table 2 and figures 1and 2). Immediately after intratracheal A. suum challenge, mean Qtrn increased by 111 ± 31010 (p < 0.05) and returned towards and to baseline by 40 and 80 min postchallenge, respectively (figure 1). Mean Vtr decreased by 15 ± 2010 (p < 0.05) immediately after challenge, by 6 ± 1010 (p < 0.05) 40 min later, and returned to baseline 80 min later. Mean sRL remained unchanged after A. suum challenge for 40 min and showed a tendency towards increasing at 80 min after challenge. Control challenge \Vith ragweed had no effect on mean Qtrn and Vtr. Bronchopulmonary A. suum challenge produced an immediate increase in mean sRLby 317 ± 73070 (P < 0.05), with a gradual return toward baseline over the ensuing 80 min while having no effect on mean Qtrn and Vtr. The increase in sRL resulted mainly from an increase in RL, although thoracic gas volume also increased by 10to 20070 in most animals. The intravenous infusion of NTG decreased mean Psa from 105 ± 6 to 92 ± 10 mm Hg (P < 0.05) and caused an increase in mean Qtrn between 30 and 40070, an effect that was maintained throughout an 80-min observation peri-

TABLE 2 EFFECTS OF A. SUUM AND VASOACTIVE DRUGS ON TRACHEAL MUCOSAL BLOOD FLOW (Otr)*

Ctr (mi·min-f)

Baseline

Protocol A. suum Nitroglycerin Vasopressin Nitroglycerin + A. suum Vasopressin + A. suum

1.07 1.12 1.39 1.28 1.61

After Beginning of Vasoactive Drug Infusion

(0.11) (0.05) (0.09) (0.15) (0.06)

Immediately After A. suum

T1t

2.01 (0.28) 1.39 1.04 1.68 1.16

(0.15) (0.09) (0.16) (0.11)

1.70 1.41 1.16 1.50 1.30

1.55 (0.13) 1.25 (0.12)

T2t

(0.27) (0.12) (0.10) (0.17) (0.12)

1.13 1.55 1.09 1.35 1.34

(0.18) (0.14) (0.05) (0.12) (0.10)

* Values are mean with SE shown in parentheses.

t T1 and T2 reflect measurements at 40 and 80 min after A. suumchallenge(or correspondingtime in the absenceof vasoactive drug infusion).

od (figure 2). The intravenous infusion of VP caused a decrease in mean Qtrn between 15and 25070; this was also maintained for 80 min. Neither NTG nor VP had an effect on mean Vtr. When mean Qtrn was first reduced by VP by 20 ± 14070 (P < 0.05), the subsequent A. suum challenge failed to increase Qtrn (figure 2), but it decreased mean Vtr immediately by 17 ± 2070 (P < 0.05); mean Vtr remained decreased by 17 ± 2070 (P < 0.05) at 40 min and by 15 ± 1070 (P < 0.05) at 80 min. The last two values were different from the corresponding values after A. suum chal-

lenge alone. After an increase in mean Qtrn by 35 ± 9070 (P < 0.05) with NTG, A. suum challenge failed to further increase mean Qtrn, and it blunted the decrease in mean Vtr (figure 2); the values immediately and 40 and 80 min postchallenge were different from the corresponding values after A. suum alone (P < 0.05). In Vitro Studies Purified A. suum antigen produced dosedependent increases in tracheal smooth muscle tension in vitro, with a mean increase in two antigen control experiments BRONCHOPULMONARY ASCARIS

TRACHEAl.. ASCARIS 400

...J

Vl

eo ~

250

E w

o ~ r

u

~

100

-50-+-----...--------.-------..ASC Tl BSl T2

-50----...-----..--------.-

TIME

TIME

ASC

BSL

T1

T2

TRACHEAl RAGWEED

100

.--===-- ----i===-

UJ

o ~ r

-=~

~ -50 -100+----~----.--------r­

BSL

T1

RW

T2

TIME

Fig. 1. Effects of tracheal and bronchopulmonary A. suum challenge on normalized tracheal mucosal blood flow (Otrn), tracheal dead space (Vtr), and speclnc bronchopulmonary airflow resistance (sRL) in sheep with natural A. suum allergy. Mean of six sheep; brackets reflect SEM. BSL - baseline; ABC :II immediately after A. suum challenge, FfN :II immediately after ragweed challenge; T1 • 40 min after challenge; 1'2 • 80 min after challenge. Asterisks indicate significantly different from baseline and corresponding control (tracheal Ascaris versus bronchopulmonary Ascaris; tracheal Ascaris versus tracheal ragweed) (p < 0.05). OPen circles - SRL; closed circles :II Qtrn; open triangles Vtr.

62

CSETE, CHEDIAK, ABRAHAM, AND WANNER

TRACHEAl ASCARIS

TRACHEAL ASCARIS

1~

10

0 - 0 ASC

•

I

.-. ASC+NTC

0 - 0 ASC

f 10 /·~t~:_6A5C+VP

O ~ I

u50 ~ c

.cf

0

6

6

•

•

• - . ASC+NTC 6 - 6 ASC+VP

_

0

! o~" ~

>

-10

"

1./ ~

\../

'6

•

-20

.,?

~

ASC

T1

~

~

£:,.-£:,.

e-e

e-e

T e - - - eT

25/.1.

1

.

•

50 w

~

--------=a----~---·

:I: U

>

O~.

6---------6

-25

DRUC

BSL

T1

VP

NTG

C)

!!.

E

00'

T2

VP/NTG CONTROL

100

6 - 6 VP NTC

50

~

TIME

VP/NTC CONTROL

~

.

.

---

BSl

ASC

TIWE

75

i~;

e--/T

!!.

-50.1-----+-----..----.......T1 T2 BSL

,

-50

•

6

-100 DRUG

BSL

T2

Tl

T2

TIME

TIME

Fig. 2. EffectsoftrachealA. suum(ASC) challengeon normalizedtrachealmucosal bloodflow(atrn) andtracheal dead space (Vtr) in sheep allergic to A. suum, in the presence or absence of continuous vasopressin (VP) or nitroglycerin (NTG) infusion. Mean(SEMin brackets) of five sheep.The effectsof VP and NTGwithout A. suum challenge(timecontrol)are shownon thebottom.BSL = baseline;ASC = immediately after A. suumchallenge, T1 = 40 min after challenge;T2 = 80 min after challenge. Asterisks indicatesignificantlydifferent from baseline (p < 0.05). Thechangesin Otrn after A. suum combinedwith VP or NTG infusionare referenced to newbaselines resultingfromVP(decrease in Otrn)and NTG (increase in atrn) infusion,whichstartedbeforeA. suumchallenge.

Discussion to 16 ± 60/0 (p < 0.05) and 32 ± 100/0 (p < 0.05)of acetylcholine-induced max- The results of this study showed that traimal tension at the highest A. suum dose cheal antigen challenge constricts the traof 100 IJ,I (164,000 PNU/IJ,I) (figure 3). chea of allergic sheep via a locally mediVasopressin and all concentrations of ated pathway and suggestedthat the magNTG per se had no effect on resting ten- nitude and duration of the constriction sion nor did they modify the contractile are influenced bytracheal blood flow. Beresponses to A. suum antigen in matched cause we have previously found that tissues. There was a tendency (p = NS) , chemical mediators releasedin the airtoward a protective effect of NTG at the way are responsible for allergic airway O.1-IJ,M concentration. NTG, at the con- narrowing in sheep (7, 8), we concluded centrations used, did not reverse acetyl- that local blood flow influenced allergencholine-induced (1 IJ,M) contractions. induced tracheal narrowing by control-

50

50

Cl CONTROL

40

_

30

20

20

10

10

JO ul

ASCARIS SUUW ANTIGEN

NG (0.1 utA)

40

JO

10 ul

Cl CONTROL

B!8 NG (10 nM)

VP ~'8 mU/ml) 1 nW) 122 NG J nW) _

sss NG

100 ul

0'"'---.......... 10 ul

30 u l l O a ul

ASCARIS SUUM ANTIGEN

Fig. 3. Effectofvasopressin (VP)andnitroglycerin (NTG) on antigen-induced contractions of sheeptrachealsmooth muscle in vitro. The antigen controls (differentfor V~ 1 nM, 3 nM NTGand for 10 nM, 0.1 J,IM NTG) increased tension significantlyat the 1QO-J.I,I dose (p < 0.05). In the presenceof VP and NTG the response to antigen was not significantlydifferentfrom that of antigenalone. At thesedoses,VP did not contracttrachealsmoothmuscle, and NTG had no relaxant effect on tracheal smooth muscle precontracted with acetylcholine (10-8 M).

ling the clearance of antigen or, more likely, chemical mediators from the airway. Several conditions had to be met for this conclusion to be valid. First, it had to be established that the antigen-induced airway smooth muscle contraction was indeed locally mediated. Bronchopulmonary challenge with antigen distal to the trachea produced bronchoconstriction as reflected by an increase in airflow resistance between the carina and the pleural surface (sRL)but no tracheal narrowing and no increase in Qtrn. The latter is consistent with the previously reported observation that allergic bronchoconstriction in sheep fails to increase cardiac output (11). Conversely, tracheal antigen challenge narrowed the trachea but failed to increase sRL. Thus, neurally or humorally mediated reflexes were not of major importance in these measurements, supporting the idea that the antigen elicited airway narrowing through a local mechanism. Second, the assumption had to be made that a decrease in Vtr was caused by airway smooth muscle contraction rather than by mucosal hyperemia or edema. In the trachea, the radius of the airway lumen is much greater than the depth of the mucosa, and therefore thickening of the mucosa would be expected to have a minimal effect on Vtr. This has been supported by our previous observations that significant increases in Qtr and mucosal tissue volume do not influence Vtr (10). Therefore, the observed changes in Vtr in the present study are likely to have reflected contraction of tracheal smooth muscle. Tracheal shortening, if also present, did not influence Vtr because the length of the tracheal chamber was determined by the dual-cuffed tracheal tube and was, consequently, constant. Third, we had to ensure that the vasoactive drugs had no direct effect on tracheal smooth muscle contraction. In the in vitro study, where the influence of blood flow was eliminated, VP (18 J.1M·~I-l) and NTG (1 nm to 0.1 J.1M) had no effect on the resting tension of tracheal smooth muscle and failed to modify the contractile response to antigen (figure 3). At these doses, NTG also failed to relax tracheal smooth muscle precontracted with acetylcholine. The concentrations of VP and NTG in vitro were of the same order as or lower than the highest possible tissue concentrations in vivo assuming uniform intra/extravascular drug distribution and no drug metabolism or excretion during the infusion period. Thus, the possibility that

AIRWAY BLOOD FLOW MODIFIES ALLERGIC AIRWAY SMOOTH MUSCLE CONTRACTION

NTG and VP inhibited or promoted mediator secretion or the contractile response of the airway smooth muscle to the mediators is remote. The in vitro study also argues against the possibility that NTG and VP altered the metabolism of the locallyreleased mediators (12). Wehave shown in other experiments that intravenously administered VP and NTG do not alter Vtr (5, 10), and NTG and VP have been reported to have negligible effects on airway smooth muscle tone (13, 14). Finally, we had to assume that tracheal smooth muscle contraction per se had no mechanical effect on Qtrn. We have previously shown that aerosolized methacholine and carbachol fail to alter Qtr and bronchial blood flow while producing tracheal narrowing and bronchoconstriction in sheep (15, 16). Those findings support our premise that there was no mechanically mediated interaction between Qtrn and Vtr in the present study. The endotracheal tube may have limited the degree of tracheal narrowing. For example, a baseline Vtr of 20 ml translated into a tracheal radius of 0.8 em, whereas the radius of the endotracheal tube was 0.35 em. Assuming uniform tracheal narrowing, a greater than 50070 reduction in tracheal radius would have been required to bring the tracheal wall in proximity of the endotracheal tube. This was not the case given the observed decreases in Vtr. However, localized narrowing could have brought the tracheal wall in contact with the endotracheal tube, which could have resisted further constriction. The soluble gas technique measures blood flow in the mucosa to a depth of < 200 J.1m (10). Although the major fraction of total airway blood flow is distributed to the mucosa (> 70070), the observed changes in Qtrn in our study may not have been accompanied by concomitant changes in blood flow to the airway smooth muscle layer (17). The presumed sources of chemical mediators during antigen challenge are mast cells and other leukocytes located in the airway lumen, the mucosa, and deeper in the airway wall, including perivascular sites (1, 18). The chemical mediators released after tracheal aerosol challenge with antigen could therefore be significantly cleared by the microcirculation before reaching their targets on the airway smooth mus-

83

cle. This could explain the prolonged du- Asthma. 2nd 00. London: Chapman and Hall, 1983; ration of allergic tracheal narrowing 79-98. Long WM, Yerger LD, Martinez H, et ale when antigen was aerosolized during VP- 2. Modification of bronchial blood flow during alinduced hypoperfusion of the trachea. lergic airway responses. J Appl Physiol 1988; The reasons why tracheal narrowing was 65:272-82. not enhanced by VP are not clear, but 3. Persson CGA, Erjefalt I, Andersson P. Leakthey include the possibility that antigen- age of macromolecules from guinea-pig tracheobronchial microcirculation. Effects of allergen, leuinduced airway smooth muscle contrac- kotrienes, tachykinins, and anti-asthma drugs. Acta tion was maximal without VP adminis- Physiol Scand 1986; 127:95-105. tration, possibly because of the endotra- 4. Evans TW, Rogers DF, Aursudkij B, Chung KF, cheal tube. NTG blunted the magnitude Barnes PJ. Inflammatory mediators involved in antigen-induced airway microvascular leakage in of allergic tracheal constriction, consis- guinea pigs. Am Rev Respir Dis 1988; 138:395-9. tent with the idea that the augmented air- 5. Csete ME, Abraham WM, Wanner A. Vasoway perfusion at the time of mediator motion influences airflow in peripheral airways. secretion cleared the mediators from the Am Rev Respir Dis 1990; 141:1409-13. Laitinen LA, Robinson NP, Laitinen A, Widtissue more rapidly. The enhanced clear- 6. dicombe JO. Relationship between tracheal mucosal ance could have reduced their concen- thickness and vascular resistance in dogs. J Appl tration and the duration of their presence Physiol 1986; 61:2186-93. at airway smooth muscle receptor sites. 7. Wanner A, Mezey RJ, Reinhart ME, Eyre P. The same mechanism operating at vas- Antigen-induced bronchospasm in conscious sheep. J Appl Physiol 1979; 47:917-22. cular smooth muscle sites could explain 8. Abraham WM, Delehunt JC, Yerger L, Marwhy antigen failed to further increase chette B. Characterization of a late phase pulmonary response following antigen challenge in allerQtrn during NTG infusion (figure 2). We have no explanation for our find- gic sheep. Am Rev Respir Dis 1983; 128:839-44. Bone JF, Metcalf J, Parer JT. Surgical prepaing that during pharmacologic vasocon- 9. ration of a carotid loop in sheep. Am J Vet Res striction (15 to 25070 decrease in Qtrn), 1962; 23:1113-6. antigen challenge failed to increase Qtrn 10. Wanner A, Barker JA, Long WM, Mariassy while increasing it by about 100070 in the AT, Chediak AD. Measurement of airway mucosal absence of pharmacologic manipulation perfusion and water volume with an inert soluble gas method. J Appl Physiol 1988; 65:264-71. (figure 2). Possibly, the site of mediator 11. Kung M, Abraham WM, Greenblatt DW, et clearance (capillary bed) and mediator- ale Modification of hypoxic pulmonary vasoconinduced vasoconstriction (arterioles) are striction by antigen challenge in sensitized sheep. sufficiently removed from each other to J Appl Physiol 1980; 49:22-7. 12. Westcott JY, McDonnell TJ, Voelkel NF. Alallow blood flow to control tissue con- veolar transfer and metabolism of eicosanoids in centrations of vasoactive mediators be- the rat. Am Rev Respir Dis 1989; 139:80-7. fore they reach their receptors on vascu- 13. Byrick RJ, Hobbs EG, Martineau R, Noble lar smooth muscle. Our findings in a sys- WHo Nitroglycerin relaxes large airways. Anesth tem involving endogenously released Analg 1983; 62:421-5. 14. Gruetter CA, Childers CE, Bosserman MK, inflammatory mediators are in keeping Lemke SM, Ball JG, Valentovic MA. Comparison with the previously reported observation of relaxation induced by glyceryl trinitrate, isosorthat the airway response to exogenous bide dinitrate, and sodium nitroprusside in bovine histamine in the lung periphery of dogs airways. Am Rev Respir Dis 1989; 139:1192-7. IS. Barker JA, Chediak AD, Baier HJ, Wanner can be altered by manipulating bronchi- A. Tracheal mucosal blood flow responses to aual and pulmonary blood flow (19). tonomic agonists. J Appl Physiol1988; 65:829-34. The physiologicsignificanceof our ob- 16. Baier H, Rodriguez JL, Chediak AD, Wanservation is that microvascular hyperfu- ner A. 'Iracheal narrowing during histamine induced sion, a typical feature of inflammation, bronchoconstriction. J Appl Physiol 1988; 64: 1223-8. may influence the magnitude and dura- 17. Parsons GH, Kramer GC, Link DP, et at. tion of tissue target responses to the in- Studies of reactivity and distribution of bronchial flammatory process. The prevention of blood flow in sheep. Chest 1985;87(Suppl:18OS-4S). this vascular control function by phar- 18. Schulmon ES, MacGlashan DW Jr, Peters SP, Schleimer RP, Newball HH, Lichtenstein LM. Humacologic or other therapeutic measures man lung mast cells: purification and charactermay have undesired effects on the clini- ization. J Immunol 1982; 129:2662-7. cal manifestations of asthma and other 19. Kelly L, Kolbe J, Mitzner W, Spannhake EW, Bromberger-Barnea B, Menkes H. Bronchial blood inflammatory diseases. References 1. Lopez-Vidriero MT, Reid L. Pathological changes in asthma. In: Clark TJH, Godfrey S, eds.

flow affects recovery from constriction in dog lung periphery. J Appl Physiol 1986; 60:1954-9.