Neuropeptide Y: presence in perivascular noradrenergic neurons and vasoconstrictor effects on skeletal muscle blood vessels in experimental animals and man

Regulatory Peptides, 19 (1987) 313-324 Elsevier

313

RPT 00641

Neuropeptide Y: presence in perivascular noradrenergic neurons and vasoconstrictor effects on skeletal muscle blood vessels in experimental animals and man John Pernow ~, Anders t3hl6n 2, Tomas H6kfelt 3, Olle Nilsson 4 and Jan M. Lundberg I Departments of 1Pharmacology, ZPhysiology and 3Histology, Karolinska Institute, Stockholm, Sweden, and 4Department of Orthopedic Surgery, Karolinska Hospital, Stockholm, Sweden (Received 27 May 1987; revised version received 10 July 1987; accepted 28 July 1987)

Summary The presence of neuropeptide Y (NPY)-like immunoreactivity (-LI) in sympathetic perivascular nerves and the functional effects of NPY and noradrenaline (NA) on vascular tone were studied in skeletal muscle of various species. A dense network of NPY-LI was found around arteries and arterioles but not venules in the gluteus maximus muscle of man, gracilis muscle of dog, tenuissimus muscle of rabbit and quadriceps muscle of cat, rat, guinea pig and pig. The distribution of NPY-immunoreactive (-IR) nerves was closely correlated to the presence of tyrosine hydroxylase (TH) and dopamine-fl-hydroxylase (DBH)-positive fibers, two markers for noradrenergic neurons. Double-staining experiments revealed that NPY- and TH-IR as well as NPY- and DBH-IR nerve fibers around arteries and arterioles were identical. The veins and venules, however, lacked or had a very sparse innervation of NPY-, TH- and DBH-positive fibers. The NPY- and TH-IR nerves in quadriceps muscle of the guinea pig were absent after treatment with 6-hydroxydopamine. Lumbosacral sympathetic ganglia from the same species contained many NPY-positive cells which were also TH- and DBH-IR. NPY-LI was also detected by radioimmunoassay in extracts of skeletal muscle from guinea pig, rabbit, dog, pig and man as well as of lumbosacral sympathetic ganglia. The content of NPY-LI in skeletal muscle was relatively low (0.1-0.4 pmol/g), whereas lumbosacral sympathetic ganglia had a much

Correspondence. J. Pernow, Department of Pharmacology, Karolinska Institute, Box 60400, S-I04 01 Stockholm, Sweden. 0167-0115/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)

314 higher content (48-88 pmol/g). NPY (10-7 M) contracted arterioles in the tenuissimus muscle of the rabbit to a similar extent (by 65%) as NA (10 -6 M), as studied by intravital microscopy in vivo. NPY had no effect on the corresponding venules while NA caused a slight contraction of these vessels. In vitro studies of small human skeletal muscle arteries and veins revealed that NPY was more potent than NA in contracting the arteries, and the highest concentration of NPY (5 × 10-7 M) caused a contraction of a similar magnitude as NA I0 -s M. NA contracted veins from human skeletal muscle, while NPY had only small effects. It is suggested that NPY, together with NA, could be of importance for sympathetic control of skeletal muscle blood flow. Neuropeptide Y; Noradrenaline; Skeletal muscle blood vessel; Vasoconstriction; Coexistence

Introduction

Neuropeptide Y (NPY) is a 36-amino acid peptide [1] which is likely to be present together with noradrenaline (NA) in sympathetic perivascular nerves in a variety of organs [2-4]. NPY is coreleased with NA upon sympathetic nerve stimulation in experimental animals [5] and upon reflex sympathetic activation in man [6]. Furthermore, NPY causes potent vasoconstriction resistant to adrenoceptor blocking agents in vivo [7]. In contrast, NPY has been reported to exert variable contractile activity on larger isolated blood vessels in vitro [4,8-10]. The first aim of the present study was to determine in detail the presence of NPY-immunoreactivity (-IR) in perivascular nerve fibers of skeletal muscle in various species in relation to NAergic neurons. A second aim was to study the functional effects of NPY in relation to those of NA on small skeletal muscle arterioles and venules using intravital microscopy in vivo and on isolated small human skeletal muscle arteries and veins in vitro.

Materials and Methods Immunohistochemistr y

Sections of the tenuissimus muscle from urethane-anesthetized rabbit (see below), the gracilis muscle from sodium pentobarbitone anesthetized (30 mg/kg i.v.) dogs, the quadriceps muscle from pentobartitone anesthetized cats, pigs, guinea pigs and rats, the gluteus maximus muscle from humans undergoing surgery (see below) and lumbosacral sympathetic ganglia (L4-S1) from pig, guinea pig, rat and cat were immersion-fixed in an ice-cold mixture of formalin/parabenzoquinone [11,12]. In addition, human lumbosacral ganglia were obtained at autopsy. Cryostat sections of 14 #m thick where cut and processed for indirect fluorescence immunohistochemistry (see [13]), using (1) rabbit antiserum (diluted 1:400 or 1:800) to tyrosine hydroxylage

315 (TH) purified from rat pheochromocytoma [14] for tissues from dog, man, cat, pig, guinea pig and rat; (2) guinea pig antiserum (1:800) to dopamine-fl-hydroxylase (DBH) purified from rat adrenal glands (Goldstein, unpublished) which was used for rabbit and rat tissues; (3) rabbit antiserum (1:400) to synthetic porcine NPY (N1; see [ 15] and 102D; see [ 16]); and (4) sheep antiserum (1:200) to synthetic NPY [ 17]. After rinsing, the sections were incubated with secondary antibodies as follows: (1) fluorescein isothiocyanate (FITC)-conjugated donkey anti-goat lgG (Nordic, Tilburg, The Netherlands), (2) rhodamine-conjugated rabbit anti-guinea pig IgG (Cappel, Malvern, PA, U.S.A.) and (3) Texas red-conjugated donkey anti-rabbit IgG (Amersham, Dagenham, England), rinsed, mounted in a mixture of glycerol and phosphate-buffered saline containing 0.1% p-phenylenediamine [18,19], analyzed in a Zeiss fluorescence microscope equipped with an oil dark-field condensor and photographed with Tri-X black and white film (Kodak, Rochester, NY, U.S.A.). In double-labelling experiments, the sheep anti-NPY antiserum was mixed with rabbit anti-TH antiserum followed by a mixture of the FITC-conjugated donkey anti-goat IgG and Texas red-conjugated donkey anti-rabbit IgG as secondary antibodies. The sheep anti-NPY antiserum was also used with guinea pig anti-DBH antiserum. In this case incubations were made sequentially with FITC-conjugated donkey anti-goat IgG and with rhodamine-conjugated rabbit anti-guinea pig IgG in the second incubation step. For control purposes, the NPY antisera were preincubated with NPY peptide (10-6 M) (see section 'Drugs'), which caused a complete disappearance of the staining pattern described below. Normal rabbit and guinea pig sera were used as controls for TH and DBH antisera, respectively. Furthermore, sections were incubated first with one of the primary antisera followed by the two secondary antisera. No evidence for cross-reactivity between secondary antibodies was obtained. To further establish the nature of the perivascular nerve fibers in the skeletal muscle, 4 guinea pigs were treated with the sympathetic neurotoxin 6-hydroxydopamine (6-OHDA) according to Lundberg et al. [20]. One week after completed treatment, quadriceps muscle tissue was collected and subjected to immunohistochemistry as described above.

Radioimmunoassay ( RIA ) Skeletal muscle tissues from man (gluteus maximus), rabbit (tenuissimus), dog (gracilis), guinea pig and pig (quadriceps) and lumbosacral sympathetic ganglia from man, pig and guinea pig were weighed and frozen at -80°C. Samples were then extracted in 1 M acetic acid at 95°C, homogenized and lyophilized. The content of NPY-LI was determined with the NPY antiserum N1 which was raised in rabbits using synthetic porcine NPY. The N 1 antisera shows no (< 0.1%) cross-reactivity to structually related peptides such as pancreatic polypeptide or peptide YY (see [15]). The RIA has a detection limit of 7.8 fmol/g tissue (see also [15]). Intravital microscopy New Zealand white rabbits (0.8-1.2 kg b.wt.) were anesthetized with 20% urethane (1.5 g/kg b.wt., i.v.). The tenuissimus muscle in the left hind leg was prepared for intravital microscopy with nerve and vessel supply intact as previously described in

316

detail [21]. Observations were made with a Leitz Biomed Intravital microscope and water immersion lens (Leitz SW x 25, NA 0.60). In addition to direct observation in the microscope, the image was televised and stored on videotape. Arteriolar and venular diameters (control diameters 8-26 and 15-50/~m, respectively) were measured from the television images. NPY and NA were applied topically for 10 min with the buffer solution superfusing the muscle, and the effects on the vessel diameter are expressed as per cent of the control diameter. In vitro studies A piece of the gluteus maximus muscle was obtained from 7 (4 male, 3 female) patients (age 46--77) undergoing plastic surgery of the hip joint. The patients were premedicated with morphine scopolamine and they received epidural anesthesia with bupivacaine with addition of adrenaline or efedrine. Some of the patients were on regular medication with fl-adrenoceptor blocking agents or non-steroid, anti-inflammatory drugs. The muscle tissue was removed at the beginning of the operation and then immediately put in aerated chilled Krebs solution (for composition see below) and transported to the laboratory. Small arteries (0.2-0.4 mm in diameter) and veins (0.5-1.0 mm in diameter) were carefully dissected free from connective tissue and the vessels were cut into small segments (1-2 mm) which were mounted on two L-shaped metal holders, of which one was connected to a Grass force displacement transducer (model FTO3) and the other was used for applying a passive tension of 5 mN (see [22]). The tension of the vessel was registered on a Grass polygraph (model 7D). The mounted vessel was kept in a 2-ml organ bath containing Krebs solution at 37"C and constantly bubbled with 5% CO2 in 02. After an equilibration period of 30-45 min, the contractile activity of the vessel was tested by applying a Krebs solution containing a high concentration of K + (127 mM). After repeated rinsing with standard Krebs solution, NA and NPY were tested by adding 50 #1 of increasing concentrations of the respective agent into the organ bath to reach a final concentration of 10-5 and 5 × 10-7 M, respectively. The contractile effects are expressed as per cent of the contraction induced by NA 10 s M in each individual preparation. For each agent, the effect was studied on a number of vascular segments from at least 5 separate patients. In addition, segments of some of these vessels were immersion-fixed and studied by immunohistochemistry for the presence of TH- and NPY-IR nerves. Drugs NA (D,L-arterenol; Sigma, U.S.A.) and porcine NPY (CRB, U.K.) were dissolved in 0.9% NaC1. The buffer solution (Krebs-Henseleit) used in the intravital microscopy experiments contained (raM): NaC1 ll8, KCl 4.8, CaC12 2.5, MgSO4 1.20, NaHCO3 25.0 and KH2PO4 1.20. The Krebs solution used in the in vitro experiments had the following composition (raM): NaC1 122, KC1 4.7, CaCl2 2.5, MgC12 1.19, NaHCO3 15.5, KHEPO4 1.19 and glucose 11.5. K +-rich Krebs solution was obtained by exchanging the NaC1 with equimolar amounts of KC1. Statistical calculations were performed using Student's t-test for paired samples. The study was approved by the Ethics Committee of the Karolinska Hospital and the Karolinska Institute.

317

Results

Immunohistochernistry and RIA A perivascular network of NPY-IR

nerve fibers was present in the skeletal muscle

o f all s p e c i e s s t u d i e d i.e. m a n , d o g , r a b b i t , pig, g u i n e a pig, r a t a n d c a t . T h e N P Y - I R

TH

tl v

"

NPY

",.',,,i,,-.

,

o

Jr

g',s

d Fig. 1. Immunofluorescence micrographs of (a-c) human skeletal muscle (gluteus maximus) and of (d,e) two small arteries from human skeletal muscle after incubation with antiserum to neuropeptide Y (NPY) (a,c-e) and tyrosine hydroxylase (TH) (b). Double-labelling experiments (a and b) shows NPY- and TH-immunoreactivity (-IR) (arrow heads) in the same nerve fibers around an arteriole (asterisks) but no innervation of the adjacent venule (v). c: NPY-IR nerves can be seen around two arterioles (arrow heads) but not around the venules (v). d,e: two adjacent segments from the arteries which had been used for functional in vitro experiments are also innervated by NPY-IR nerves at the adventitiomedial junction (arrow heads). Double arrow heads indicate strong autofluorescence in the internal elastic membrane. Asterisks in a,d,e, arterial lumen; m, skeletal muscle tissue. Bar = 50 #m, in a for a--c and in d for d and e.

318

nerves were mainly associated with small arteries and arterioles in all species (Figs. 1 and 2). Double-staining experiments revealed that the NPY-IR nerves were identical with the TH- and DBH-positive fibers, and virtually all TH- and DBH-positive

\

d F

v

Fig. 2. Immunofluorescence mlcrographs using the double-labelling technique (a~l) of sections of (a,b) rabbit tenuissimus muscle, (c,d) dog gracilis muscle and (e,f) of adjacent sections of cat sympathetic ganglion ($1) after incubation with antiserum to NPY (a,c,e) and DBH (b) or TH (d,f). a,b: the same nerve fibers are NPY- (arrow heads) and DBH-IR around the arteriole of the tenuissimus muscle, while the venule (double arrows in a) shows no innervation but only some autofluorescence, c,d: an arteriole (asterisk in lumen) in the dog gracilis muscle is surrounded by varicose NPY- and TH-IR nerve fibers, m, indicates skeletal muscle tissue, e,f: The cat sympathetic ganglion shows many NPY- and TH-IR cells but some ganglionic cells are TH-positive but lack NPY-1R (arrows in e). Bars = 50 #m in a for a and b and in c for c-f.

319 fibers were also NPY-IR (Figs. 1 and 2). Around the veins and venules, however, no or very few nerve fibers were positive for TH, DBH or NPY in any of the muscles investigated (Figs. 1 and 2). Thus, immunohistochemistry of the rabbit tenuissimus muscle revealed a dense innervation of NPY- and DBH-positive nerves around the arterioles but no stained nerve fibers associated the venules (Fig. 2). Furthermore, staining of sections of the same human vessels used for functional experiments revealed that nerve fibers positive for T H and NPY were present around the artery (Fig. 1) but not the vein. In the dog, where the vascular tree of the gracilis muscle was followed from the main vessels in a peripheral direction, only single varicose NPY- and T H - I R nerves, but several axon bundles, were present around the main gracilis artery and vein. The density of positive varicose nerves around the arteries increased considerably when the vessels were below 0.5 mm in diameter and the arterioles were most densely innervated (Fig. 2). Only single or no varicose NPY- or T H - I R nerves were associated with large or small veins or venules (Fig. 2). After 6O H D A treatment of guinea-pigs, most NPY- and T H - I R perivascular fibers disappeared in the quadriceps muscle. Many cells in the lumbosacral sympathetic ganglia of all species studied were NPY-IR (Fig. 2). These ganglion cells also contained TH-IR. A separate population of T H positive cells, however, seemed to lack NPYIR (Fig. 2). The content of NPY-LI in skeletal muscle from man, rabbit, dog, guinea pig and pig varied from 0.1~0.4 pmol/g (Table I). The lumbosacral sympathetic ganglia from man, pig and guinea pig contained 200-500 times more NPY-LI than the corresponding skeletal muscle from the same species (72 + 6, 88 + 10 and 48 + 5 pmol/g, respectively). Intravital microscopy o f vascular effects in rabbit tenuissimus muscle

Topically applied NPY and N A both caused contractions of the arterioles (Fig. 3). The effects of the two substances were comparable in magnitude (65 + 3% and 64 + 7% reduction in diameter, respectively; n = 6), although the concentration of NA (10 - 6 M) was 10 times higher than that of N P Y (10 -7 M). In addition, NA caused a weak but clear-cut contraction of the venules, while no effect on venular diameter could be seen after administration of NPY (Fig. 3). The vasoconstrictor

TABLE I Content of NPY-LI in extracts of skeletal muscle from various species

Man Rabbit Dog Guinea pig Pig

n

NPY-LI

6 4 4 5 6

0.14 + 0.02 0.30 + 0.08 0.22 4- 0.05 0.17 ± 0.04 0.40 4- 0.05

Man, gluteus maximus; rabbit, tenuissimus; dog, gracilis; guinea pig and pig, quadriceps. The tissue content of NPY-LI is given as pmol/g, mean + S.E.M. The numbers (n) of observations are indicated.

320 [ ] Arteriole [--]Venule 100

,

m~6o ,~

/////, ~+4p

r/z/z,

20-

g/~ ,HH. NPY

CONTROL

10-7M

NA 10-6M

Fig. 3. Effects of NPY (10 -7 M) and NA (10 -6 M) on arteriolar and venular diameters in the rabbit tenuissimus muscle. The vessels were studied by intravital microscopy in vivo and the substances were applied topically in a buffer solution. The effects on the vessel diameter are expressed as per cent of control diameter before the drug was applied. Data are given as means ± S.E.M. Significant differences from control diameter were calculated using Student's t-test for paired samples. *P