- Email: [email protected]
Co-release of neuropeptide Y and noradrenaline from pig spleen in vivo: Importance of subcellular storage, nerve impulse frequency and pattern, feedback regulation and resupply by axonal transport
Neuroscience Vol. 28, No. 2, pp. 475486, Printedin Great Britain
03044522/89 $3.00+ 0.00 PergamonPressplc
1989
IBRO
CO-RELEASE OF NEUROPEPTIDE Y AND NORADRENALINE FROM PIG SPLEEN IN’ VIF/o: IMPORTANCE OF SUBCELLULAR STORAGE, NERVE IMPULSE FREQUENCY AND PATTERN, FEEDBACK REGULATION AND RESUPPLY BY AXONAL TRANSPORT J. M. LUNDBERG,*? A. RUDEHILL,~ A. .%LLEVI,§ G. FRIELQ and G. WALLIN// Departments of *Pharmacology and $Physiology, Karolinska Institute and Department of SAnaesthesia, Karolinska Hospital, Stockholm, Sweden, and I/Department of Clinical Neurophysiology, Sahlgrenska Hospital, Gothenburg, Sweden
importance of subcellular storage, nerve impulse rate and pattern, and feedback regulation, and neuropeptide Y-like immu~oreactivity, was studied in the blood perfused pig spleen in viva. Vasoconstrictor responses were recorded as perfusion pressure changes. Subcellular fractionation experiments using sucrose density gradients showed a bimodal distribution of noradrenaline (peak concentrations at 0.8 and 1.1 M sucrose) while only one main peak of neuropeptide Y was present (at 1.1M sucrose). Overflow suggesting release of noradrenaline and neuropeptide Y-like immunoreactivity could be detected after 10 s stimulation at 10 Hz. The ratio for the output of noradmnal~ne and neuro~ptide Y upon continuous nerve stimulation in control animals decreased with frequency. After inhibition of noradrenaline reuptake by desipramine the vasoconstrictor response and noradrenaline output were enhanced while the corresponding overflow of neuropeptide Y was reduced by 50% at 0.5 Hz. Stimulation with the irregular or regular bursting patterns at high frequencies caused larger perfusion pressure increase and relative enhancement of neuropeptide Y output compared to noradrenaline than a continuous stimulation both before and after desipramine treatment. A similar fractional release per nerve impulse was calculated both for ~3H]noradrena~ne (5.6 + 1.0 x IO-‘) and neuropeptide Y (7.3 f 0.3 x lo-‘). After reserpine treatment combined with preganglionic denervation the vasoconstrictor responses were more long-lasting, neuropeptide Y release was enhanced while noradrenaline content and release were reduced by 99%. The difference in neuropeptide Y overflow between continuous and bursting types of stimulation was smaller after reserpine treatment. After prolonged intermittent stimulation with regular bursts (20 Hz) for I h the splenic content of neuropeptide Y was reduced by 580/o, while no change was observed for noradrenaline. The maximal perfusion pressure increase upon prolonged nerve stimulation after reserpine was simiiar in control and reserpine-treated animals, but after reserpine the vasoconstrictor response and neuropeptide Y release were subjected to fatigue. Ligation experiments of the splenic nerves revealed the splenic neuropeptide Y content was resupplied by axonal transport with a calculated total tissue turnover time of 11 days. In contrast, axonal transport contributed only to a marginal extent for the resupply of noradrenaline. In conclusion, stimulation with high frequency bursting preferentially enhanced overflow of neuropeptide Y compared to the co-existing classical transmitter noradrenaline, suggesting that exocytosis of material from large dense-cored vesicles mainly occurred under such stimulation conditions. The local biophase concentrations of noradrenaline seem to regulate neuropeptide Y release. Due to the limited resupply by axonal transport, excessive and prolonged release of neuropeptide Y cannot be maintained without depletion of terminal stores. Abstract-The
as well as resupply by axonal transport for the release of noradrenaline
Recordings of impulse activity in single sympathetic nerve fibres have indicated an average rate of firing up to 3 Hz during basal conditions.” The pattern of firing is highly irregular. 8*‘6.45 Thus, the nerve impulses occur in bursts separated by quiescent periods and the instantaneous discharge frequency (interval between two successive action potentials) could reach tTo whom correspondence should be addressed at the Department of Pharmacology, Karolinska Institutet, _ _ _ . ..~ _ P.U. Box bwoo, s-104I 01 Stockholm, Sweden. Abbreviations: DBH, dupaunuc-,,-rryu --^-A-- D ‘--+oxylase; DMI, desipramine; HPLC, high-performan tee liquid chromatography; Lf, like immunor~ctivity, ; NA, noradrena. line; NPY, neuropeptide Y; VIP, vasoactrve Intestinal polypeptide. NSC
28,2--H
475
35 Hz even at low average rates of 1 Hz.17 Recently, stimutation with the irregular bursting activity of human skin vasoconstrictor fibres was found to cause signi~cantly larger contractions of mesenteric arteries in vitro than a continuous frequency giving the same total number of impulses.” Furthermore, the atropine-resistant vasodilatory response in the cat submaxilIary gland upon parasympathetic nerve stimulation, which is likely to be related to release of vasoactive intestinal polypeptide (VIP),22,26 was enhanced by stimulation with bursts of high frequencies.’ Whether the co-reiease of the classical transmitter was also enhanced in the ,. , acetylcholine , , *. . _ sanvary grana wnen usmg tne burstmg type ot strmulation
is not known
although
functional
data sug-
476
.I. M. L~:YIWIK<~ <‘I i/i
gest that continuous seems to favour VIP choljne,?‘._‘?~‘”
high frequency stimulation release compared to acctyl-
Increasing evidence suggests that neuropeptide Y (NPY), which co-exist?’ and is co-released with noradrenaline (NA) from sympathetic nerves upon electrical stimillation~‘.~ or reflexogenic activation.~‘.~~‘~~ is of importance for sympathetic vascular control at both pre- and postjunctional levels.‘“~“~’ The present investigation evaluates the importance of impulse rate and pattern as well as local biophase concentrations of NA for the rcleasc of NA and NPY-like immunoreactivity (LI) from the pig spleen !rr i:it~. In addition, the subcellular storage and the dependence of axonal transport for resupply of terminal stores of NPY relative to NA is also investigated’ EXPERIMENTAL PROCEDURES
The coehac ganglion, splenic nerves along the main splenic vessels and the caudal portion of the spleen were dissected out from pigs under pentobarbitat anaesthesia, frozen on dry ice, weighed, extracted and the content of NA and NPY-LI was determined (see below). Subcellular jiuctionation experiments Subcelluiar fractionation combined with sucrose density gradients was performed (n = 4) on about log of the caudal pole of the spleen removed from anaesthetized pigs (see below). Details of the protocol have been published elsewhere.‘Z-‘4 Briefly a 10,OOOg supernatant from a spleen homogenate was placed on a sucrose density gradient (0.25- 1.4 M sucrose) and centrifuged at 280,OOOg max. The particulate and supematant fractions were then separated (see Ref. 14). The concentration of NA and NPY-LI in each fraction was determined by high-performance liquid chromatography (HPLC)‘” or radioimmunoassay,” respectively. Before the NA determinations, the sucrose fractions were extracted by perchloric acid (NA). Furthermore, before NPY radioiInmunoassay, the fractions were Passed through Sep Pak’” cartridges and the peptide content separately eluted.‘” Functional
in vivo smdies
Functional experiments were performed using the blood perfused pig spleen (for details of the ex~rimenta~ set up, see Refs 34 and 36). Briefly, the spleen from anaesthetized pigs was placed in saline at 37’C and constantly perfused with blood from a femoral artery. The perfusion pressure in the spleen (perfusion rate was adjusted to obtain a basal pressure of 8W?OmmHg) was recorded continuously and alterations were taken as indication of changes in vascular resistance. The blood flow in the splenic venous effluent was monitored by an electromagnetic flow probe (Nycotron”, Norway) and changes in outflow with a constant input are considered to mainly reflect alterations in splenic volume due to capsule contraction. The postganghonic sptenic nerves were stimulated electrically (see below) and output of NPY-LI and NA in the splenic venous eflluent was determined by calculating the venoarterial plasma concentration differences multiplied by plasma flow. NPY-LI in plasma and splenic tissue was determined by radioimmunoassay using antiserum Nl (see Ref. 44). The NPY-LI in splenic tissue and plasma detected by this antiserum was characterized by reversed phase HPLC in comparison to synthetic porcine NPY (l-36) (see Refs 34 and 44). The levels of catecholamines in plasma and splenic tissue were determined by using HPLC combined with electrochemical
detection (set Ref. IX). The ourput of NPY-LI and NA during (after 23 of the stimul~ti~~n) and 30 s. 2. 5. If) :tnd 15 mm after the stimulations was integrated and given as a total value of recovered material for each stimulation. In separate experiments the time courses for the overflow upon stimulation with 2 and IO Hz were followed in more detail in controls. after desipramine (DMi) and after reserpine treatment (see below).
Electrical stimulations (IO V. 5 ms) were performed on the transected perivascular nerve bundles accompanying the main splenic artery using a bipolar platinum electrode connected to a Grass 588 stimulator. First, stimuiations with regular frequencies at 0.5, 2 and 10 Hz giving a total of 240 impulses with 20 min intervals between stimulation periods (n = 5) were used. Second, the importance of impulse pattern for the functional response as well as overtlow of NPY-LI and NA was studied (n = 5). To this end, the responses obtained with a continuous frequency of 0.59 Hz for gmin or 2.3 Hz for 2min (a total of 285 and 276 impulses, respectively) were compared to those obtained with the irregular stimulation pattern of a few unit recordings of human vasoconstrictor nerve fibre activity. The neural recording was made under resting conditions from the muscle vasoconstrictor hbres in the peroneal nerve (0.59 Hz) and cutaneous vasoconstrictor fibres from the median nerve (2.3 Hz) at the elbow using tungsten electrodes (see Refs 3 and 16). The occurrence of sudomotor activity in the median nerve was detected in parallel using recordings of galvanic skin responses within the innervation zone in the hand. One segment of neural recording without sudomotor activity was chosen and the imp&es were transformed into standard pulses, stored on a magnetic tape and used for triggering the Grass stimulator. The average frequency of an 8 min period was 0.59 Hz and in a 2 min period 2.3 Hz. although the instantaneous frequency (i.e. rate cakntiated from the interval between two successive action potentials) was as high as 200 Hz. A stimulation sequence with regular bursting activity at a frequency of 23 Hz for 1s every IO s was also used for additional comparison. For each stimulation period the maximal perfusion pressure increase was used as a tneasure of the functional response. The order of the different types of stimulations was varied randomly. After a session with stimulations under control conditions. 100 PCi of [ ‘H]NA was infused into the splenic artery for IO min. After an additional 20 min. the monoamine uptake blocker DMI (desiprdmine hydrochloride, CIBA. Sweden 0.5 mg/kg) was given intravenously. Following two priming stimuiations 30 min later, fractional release of [‘H]NA was determined in alumina extracted plasma (see Ref. 19) in parallel to endogenous NA and NPY-LI in tnW%il and venous samples. In separate pigs (n = 5). the left splanchnic nerves were sectioned under pentobarbital anaesthesia to decentralize the postganglionic nerves to the spleen and reserpine (1 mg/kg) was given i.v. 24 h prior to the experiments to deplete NA with the NPY content intact (see Ref. 35).
To study the relative importance of axonal transport for resupply of NPY-LI compared to local synthesis of MA, long-term stimulations using regular 20 Hz burst stimulation for 1 h (see above) were performed in control (n = 5) and reserpine-treated (n = 5) pigs. Release during and after, as well as tissue content of NPY-LI and NA, were determined at the beginning and the end of the stimulations. Furthermore, one of the two main bundles of the splenic nerves along the remaining portion of the splenic artery and vein about 2 cm proximal to the transection site was ligated using a silk suture and a glass rod (see Ref. 24). Ten hours later two 5mm nerve segments. one just proximal to the
417
Co-release of NPY and NA from pig spleen
NPY-LI were lower in the coeliac ganglion (16O:l) and in the snlenic nerves (200: 1) compared to the _ spleen (625: i, Table 1). _
porcine NPY (l-36) (Fig. 2). When using stimulation with regular impulses the output of both NA and NPY-LI increased with frequency. Thus, the output of NA and NPY-LI was 360 4 38 vs 1.2 f 0.3 pmol at 0.5 Hz and 2233 + 201 vs 33 f 3 pmol at 10 Hz. The calculated ratio between NPY-LI/NA overflow was then around 1:300 at 0.5 Hz while at 10 Hz a ratio of about 1: 90 was observed (Fig. 3). To prevent the major removal process for NA, i.e. neuronal reuptake, DMI pretreatment was used. In the DMI pretreated animals the corresponding ratios between output of NPY-LI and NA were about 1: 1500 at 0.5 and 1: 200 at 10 Hz (Fig. 3). Compared to untreated animals this change in ratio seemed to be due both to an enhanced overflow of NA escaping reuptake as well as a significant reduction in output of NPY-LI. Thus, NPY overflow was reduced by 48 + 5% at 0.5 Hz (P < O.Ol), by 28 f 3% at 2Hz (P < 0.05), while only a small, non-significant change occurred at 10 Hz after DMI (n = 5).
Subcellular fractionation
Impulse pattern and release
In the subcellular fractionation experiments particulate NA had a bimodal distribution on the sucrose density gradient. The light density peak at 0.8 M sucrose had a concentration of 18 1 x 24 nM NA, while the peak at 1.1 M sucrose contained 266 f 53 nM (Fig. 1). NPY-LI, however, was present mainly in heavier fractions with a peak value of 393 f 44 pM at a sucrose density of 1.1 M (Fig. 1). In the supernatant the highest concentrations of NA and NPY-LI were present in the first fraction of the gradient (126 f 14 nM and 70 f 12 pM, respectively).
Electrical stimulation of the splenic nerves with regular frequency (2.3 Hz) caused a static increase in perfusion pressure (Fig. 4) of 74 f 4 mmHg (n = 7) as well as an increase in blood flow from the spleen. Furthermore, an overflow of both NA and NPY-LI occurred into the splenic venous effluent with the corresponding integrated outputs of 1075 f 147 and 7 + 2 pmol (n = 9), respectively. Stimulation with the recorded sympathetic nerve activity for 2 min caused a perfusion pressure increase which was also irregular in appearance (Fig. 4) and on average 41% larger than that caused by the continuous 2.3 Hz stimulation (Figs 4 and 5). The overflow of NA was only marginally increased compared to that obtained by continuous stimulation while the output of NPY-LI was significantly enhanced by 22% (P < 0.01, Fig. 5). Intermittent stimulation with bursts at 23 Hz for 1 s every 10 s for 2 min (23b) caused a transient increase in perfusion pressure for each burst (Fig. 4). The maximal perfusion pressure effect upon 23b stimulation was significantly (56%, P < 0.05) larger than
ligation and one from the non-ligated nerve, were removed, weighed, frozen on dry ice and subsequently analyzed for content of NPY-LI and NA. The increase of NPY-LI and
NA per hour of nerve ligation in the postganglionic nerve tissue just proximal to the ligation for both branches was calculated after subtracting the content in the non-ligated one. In addition the content of NPY-LI and NA in postganglionic spleen nerves from intact animals was determined for comparison.
Statistics Student’s t-test was used for comparing the functional responses and output of NA and NPY-LI for the continuous 0.59 and 2.3 Hz stimulations with the irregular human recordings of nerve activity or regular bursting patterns. A P-value 40.05 was considered significant. RESULTS
Tissue content The molar ratios between tissue content
of NA and
Frequency dependence of release
Electrical stimulation of the splenic nerves with continuous impulses caused increase in perfusion pressure and venous blood flow, as well as overflow of NA and NPY-LI into the venous effluent, suggesting release. Reversed phase HPLC analysis revealed that NPY-LI in splenic extracts and plasma in the splenic venous effluent collected during nerve stimulation eluted at the same position as synthetic
Table 1. Tissue content and molar ratio of noradrenaline and neuropeptide Y-like immunoreactivity various portions of postganglionic sympathetic neurons to pig spleen Tissue Coeliac ganglion Splenic nerves Splenic nerves 10 h ligation Spleen Spleen 1h stimulation
NA (nmol/g) control reserpine 31_+5 14 + 3 48& 7** 11.6+ 1.0 9.6 _+ 0.2
NPY-LI (pmol/g) control reserpine
in
NA/NPY-LI ratio control reserpine
-
195+39 68_f 15 1206 + 220**
235 + 25 51+9 1024 + 113**
160:1 200:1 4O:l
-
0.14 * 0.02*** 0.12 + 0.02***
18.6 + 1.4 7.8 & 1.3**
14.4 _+ 1.1 3.2 + 0.7**
625:l 123O:l
1O:l 38:1
(expressed as means f S.E.M., n = 4-10) of NA (nmol/g) and NPY-LI (pmol/g) in cell body region (coeliac ganglion), axons (splenic nerves) and terminal areas (spleen). The splenic nerves were ligated for 10 h under anaesthesia. Furthermore, the splenic nerves were stimulated for 1 h with regular bursts at 20 Hz for 1 s every 10 s. Controls were pretreated with DMI (0.5 mg/kg). Reserpine pretreatment (1 mg/kg i.v.) was performed 24 h prior to the experiments and combined with preganglionic decentralization. Asterisks denote significant differences compared to control (**P < 0.01 and ***p < 0.001)
Tissue content
using Student’s
?-test.
500
400
1
O--O .-.
-500
NPY NA
6
-400 -300 2 5 -200 24 -100 -0
Fig. I. Subcellular distribution of particulate NA and NPY-LI after centrifugation of a 10,000 g,,,,, supernatant derived from a pig spleen homogenate on a continuous sucrose density gradient of 280,000 g,,,., for 90min. The density of each fraction is given as M sucrose, and content of NA in nM and NPY-LI in pM, respectively. Data are given as means + S.E.M.
the response using continuous 2.3 Hz stimulation (Figs 4 and 5). Furthermore, the output of NPY-LI was enhanced more than three-fold upon 23b stimulation and the NA overflow was also significantly increased (by 80%, P < 0.05) (Fig. 5). When integrated during and up to 15 min after the start of the 2 min stimulation period, the ratio between the recovery of NA and NPY-LI in the splenic venous effluent was 115 : 1 at 23b and 160: 1 at 2.3 Hz, respectively. Regulation of release concentration
by biophase
more detail for continuous stimulations with 2 and 10 Hz (Pigs 6 and 7). A consistent detectable overflow of NA and NPY-LI was present after 30 s and 10 s upon stimulation with 2 and 10 Hz, respectively. After this the stimulation perfusion pressure declined to values below control. As described in Fig. 6, the decline in the plasma outflow of NA was much more rapid than for the NPY-LI. Pretreatment with DMI caused an enhancement of the maximal perfusion pressure increase and prolongation of the decay phase (especially at 2 Hz). After reserpine the overflow of NPY-LI was enhanced (especially at 2 Hz) while no detectable output of NA was present. The perfusion pressure response after reserpine treatment was slower in developing (at 2 Hz) and longer-lasting (at both 2 and 10 Hz). After reserpine treatment 2.3 Hz and 23b stimulations were associated with about three- and two-fold larger release of NPY-LI while NA overhow as well as splenic content of NA (Table 1) were reduced by 99%. The relative difference regarding NPY release upon continuous 2.3 Hz stimulations and bursting patterns was much smaller after reserpine (Fig. 5). The maximal per-
noradrenaline
After pretreatment with DMI, nerve stimulation with all three types of pattern caused a more rapid initial rise of the perfusion pressure compared to control conditions (Fig. 4). Furthermore, the irregular perfusion pressure responses upon stimulation with recorded sympathetic nerve activity as well as the effects of intermittent 23b stimulations were much less apparent (Fig. 4). A slight potentiation of the maximal perfusion pressure effects upon nerve stimulation occurred after DMI treatment (Fig. 4) (by 16% at 23b, 24% at irregular and 43% at continuous stimulation). The recovered output of NA in the splenic venous effluent was enhanced by DMI treatment using all three types of stimulations (by 24% at 23b, 45% at irregular and 87% at continuous stimulation). The relative differences between the responses in maximal perfusion pressure to the three stimulation patterns were smaller after DMI pretreatment while the overflow of NPY-LI was still much larger at 23b (Fig. 5). The NA output in the DMI-treated animals as well as the calculated fractional release of [3H]NA per nerve impulse was of comparable magnitude for all three types of stimulations (2.3 Hz: 4.5 + 1.7, human activity: 5.1 +_ 1.2 and 23b: 5.6 + 1.0 x 10V5 for fractional release, respectively). These values were of comparable magnitude to the calculated fractional release of NPY- at 23b (7.2 + 0.3 x IO-‘). The time course for the perfusion pressure increase and the outflow of NPY-LI and NA was followed in
PIG SPLEEN
NPY-LI hndesmactkm
ELUENT 8%
I
PIG
I
PLASMA
20
40
60
60
FRACTKM NUWER
Fig. 2. Reversed phase HPLC chromatogram of NPY-LI in an acetic acid extract from the pig spleen (middle panel) and plasma collected in the splenic venous effluent 2 min after nerve stimulation using intermittent bursts at 23 Hz (lower panel) in comparison with synthetic porcine NPY (l-36, top panel). An acetonitrile gradient with water was used and the content of NPY-LI in each fraction was analysed by RIA.
419
Co-release of NPY and NA from pig spleen
_ 10
2
0.5
FREQUENCY (Hz)
Fig. 3. Ratio between the totally recovered integrated output of NA and NPY-LI in the splenic venous effluent in
uivd upon nerve stimulation with c&tinuous frequencies at 0.5 Hz (for 8 min). 2 Hz (for 2 min) and 10 Hz (for 24 s) giving a‘total of 24d imp&s. Filled dircles represeAt control conditions and open circles the ratio after pretreatment with desipramine (DMI; 0.5 mg/kg i.v.). Data are given as means + S.E.M. (n = 7).
fusion pressure increase after reserpine treatment was slower in developing and longer-lasting (Figs 6 and 7) and only marginally reduced. Stimulation using the irregular impulses with an average frequency of 0.59 Hz also caused a rapidly fluctuating pattern in the perfusion pressure response (Fig. 8). Compared to a continuous 0.59 Hz stimulation the irregular pattern caused about a two-fold larger increase of NPY output and 40% larger perfusion pressure increase. After reserpine treatment the outflow of NPY-LI was enhanced about threefold at 0.59 Hz (from 2.3 + 0.2pmol to 6.0 f 1.5 pmol, P < 0.01). The perfusion pressure response was, however, smaller than in the controls, less fluctuating and longer-lasting (Fig. 8). Long-term stimulation
Electrical stimulation with 20b for 1 h in DMItreated animals (n = 5) caused a marked perfusion
1s
*s HZ
pressure increase, which slowly declined after 2-5 min to a steady-state level at about 60-70% of the maximal response which was maintained between 10 and 60 min of stimulation (Fig. 9). After the stimulation was turned off, the perfusion pressure response returned to baseline values within 5 min (Fig. 9). The stimulation-evoked maximal overflow of NPY-LI in the splenic venous effluent was delayed in comparison to NA (Fig. 9). Thus, whereas the maximal output of NPY-LI was present after about 20min the peak outflow of NA was already seen after about 5 min of stimulation (Fig. 9). The ratio between NA and NPY-LI in the splenic parenchyma (63O:l) may be compared with the ratio between the output of NA and NPY-LI upon nerve stimulation, which was 3000: 1 after 30 s and about 200: 1 after 20 min stimulation (Fig. 9). The maximal levels of NA and NPY-LI recovered in the venous effluent during the stimulation were 172 _+21 nmol/l and 692 f 178 pmol/l, respectively. After the stimulation was turned off, NA output rapidly decreased and was basal within 2 min while the NPY output was still elevated 5 min after cessation of the stimulation (Fig. 9). After reserpine treatment, the increase in perfusion was close to the maximal response in the control after about 5 min of stimulation. Then, the effect gradually declined (Fig. 9). A marked increase in NPY output occurred during the first phase of stimulation, but then the outflow of NPY-LI declined (Fig. 9). No clear-cut NA release was detected after reserpine. The maximal levels of NA and NPY-LI in the venous effluent upon stimulation after reserpine were 3 f I nmol/l and 1150 + 85 pmol/l respectively. The tissue concentration of NPY-LI in the spleen after 1 h stimulation with 20b in control pigs had decreased by 58% and by 78% after reserpine com-
HwIll
*cTI”ITY
ZSHZM
Fig. 4. Effects of electrical stimulation of the splenic nerves for 2 min with a total of 276 pulses using a continuous frequency of 2.3 Hz (left panel), an irregular pattern from recordings of human vasoconstrictor nerve activity (middle panel), and with regular bursts at 23 Hz for 1 s every 10 s (23b, right panel), on perfusion pressure (PP, mmHg) in the blood perfused pig spleen in ho. The control effects are compared with responses 20 min after iv. pretreatment with desipramine (DMI; 0.5 mg/kg). Note that after DMI a more rapid rise in the early phase of the perfusion pressure was present and most of the burst evoked variations of perfusion pressure were lost. Different time scales are used for the three upper (10 s) and the bottom (1 s) tracings of the stimulation pattern.
rather similar values were obtained. to 7.9 X 10 ‘, respectively.
i.e. 5.8 compared
Axrmai transport and turnnoer In the axonal transport experiments the accumulation of NPY-LI and NA proximal to the ligation of the whole splenic nerves was calculated to be 5 F I (n = 7) and 220 + 18 (n = 6) pmolih, respectively. This gives a calculated mean total turnover time in nerve endings of the total splenic content of NPY-LI (average splenic weight 75 g multiplied by tissue concentration; 18.6 pmol/g = 1395 pmol and divided by the accumulation via axonal transport per h; 5pmol) of about 11 days. To replace the 58% reduction in tissue content of NPY-LI seen after a 1 h stimulation with 20b would then take 740/5 h, i.e. about 6 days by axonal transport. When the corresponding calculations are made for NA, total replacement of the splenic NA content by axonal transport would need total content (858,000 pmol)/ 220 pmol/h = 162 days and the recovered overflow of NA during 1 h stimulation 102,~0/220 = 19 days. The content of NPY-Ll in the coeliac ganglion, unligated splenic nerves and the accumulation of NPY-LI above the ligature of the splenic nerves did not change 24 h after reserpine treatment (Table 1).
400 T
200
T
DISClJSSION
Vesicular storage in relation to release
Human activity
23 Hz bursts
Fig. 5. Summary of experiments using electrical stimulation of the splenic nerves for 2 min with the irregular bursting pattern and reguiar bursts at 23 Hz for I s every 10 s. The effects (means f S.E.M.) on perfusion pressure (PP, mmHg), as well as output of NPY-LI and NA, have been expressed as percentage of the responses obtained for a continuous stimulation at 2.3 Hz giving the same total number of impulses in the individual experiments. Open bars represent observations in control animals (n = 9). hatched bars observations after pretreatment with desipramine (DMI; 0.5mg/kg i.v.; n = 5) and filled bars responses 24 h after reserpine treatment (1 mg/kg iv.) combined with preganglionic denervation. Asterisks denote significant differences from the continuous stimulation (*P cc 0.05, ***P -=z0.01) using Student’s f-test.
pared to non-stimulated preparations (Table 1). A total output of 422 + 48 and 541 + 109 pmol NPY-LI was recovered in the splenic efffuent during and IOmin after the stimulation in control and reserpinized pigs, respectively. The splenic concentration of NA, on the other hand, was not significantly reduced after 1 h stimulation in the controls and 102 f IO nmol was recovered in the venous effluent (Table 1). When a fractional release per nerve impulse was calculated in control pigs for the whole stimulation period for [3H]NA and output of NPY-LI,
Adrenergic neurons in the spleen contain two types of storage vesicles as revealed by su~ellular fractionation experiments, the so-called *‘light” and “heavy” NA vesicles, which most likely correspond to small and large dense-cored vesicles, respectively (see Refs 5 and 11). In the present study the subcellular fractionation experiments showed that whereas NA had a bimodal distribution, NPY-LI was mainly present in the heavier peak of NA (see also Refs 12-14). The large vesicles release their content of amines, proteins such as dopamine-fi-hydroxylase (DBH) and chromatog~nin probably by exocytosis upon stimulation of the splenic nerves.42 Since the output ratio NA/NPY was not fixed but decreased with increasing stimulation frequencies, our data support the hypothesis that the classical transmitter (NA) is preferentially released from small storage vesicles at low frequencies while at higher frequencies an increasing proportion of large dense-cored vesicles release NPY (as well as NA). Not only frequency of stimulation but also impulse pattern may be of importance. Thus, the present data show that stimulation with either regular (23b) or irregular bursts of sympathetic impulses at high frequencies induced a larger vasoconstrictor response than a continuous frequency giving the same total number of impulses during the stimuIation period. This finding is in agreement with in vitro data using stimulation with human sympathetic activity regarding contractions of rat mesenteric arteries” while
G
481
of NPY and NA from pig spleen
Co-release
o
2
4
6
6
10
0
2
4
6
8
IO
60.
I
60.
l
4o 20.
-201
r.,.. 0 20
40
60
Time (set
V
60
100
I 120
Time (min)
)
2
Hz stimulation
V
Fig. 6. Time course for the release of NA and NPY-LI as well as perfusion pressure changes in the blood perfused pig spleen upon stimulation with 2 Hz for 2 min. The effects in control animals (filled circles, n = 5) have been compared with responses after pretreatment with desipramine (DMI; 0.5 mg/kg i.v.; open squares, n = 5) or reserpine (1 mg/kg i.v. combined with preganghonic denervation, open circles, n = 5). Data are given as means + S.E.M. Note the different time scales for during (s) and after (min) stimulation.
observations concerning sympathetic control of cat skeletal muscle blood flow in vivo did not show any enhanced functional response upon stimulation with bursts.2 In addition to increased functional responses we also found that irregular stimulation increased the output of NPY-LI considerably more than that of NA. This suggests that not only the frequency but also the impulse pattern is of crucial importance for the release of the peptide NPY compared to the coexisting classical transmitter NA. The background for such a “nerve impulse pattern coded release” of bioactive substances from autonomic nerves may be related to partly different subcellular storage sites for NPY and NA. Thus, whereas the established view is that NA is present in both small and large densecored vesicles in, for example, the spleen (see Refs 5 and 11) NPY-LI (Refs 12-14 and present results) seems to be preferentially stored in large dense-cored vesicles. Therefore, available data suggest that material from the large dense-cored vesicles is mainly
released upon high frequencies of stimulation. As discussed above, NA is also likely to be present in the large dense-cored vesicles. It has been argued that only l&15% of the large vesicles in a terminal could account for a significant amount of the NA released because of their relatively larger storage capacity compared to the small vesicles.” Furthermore, it has been reported that 25-30% of the vesicles in the terminals of the pig spleen are of the large type.20 From the present subcellular fractionation experiments it may be estimated that about half of the NA content may originate from the large vesicles. Furthermore, considering NPY overflow, this pool of large vesicles appears to contribute only to a moderate extent to the total NA output from the spleen upon stimulation with continuous low frequencies. An increase in the NA output in the absence of DMI was, however, seen upon stimulation with 23b compared to a continuous 2.3 Hz frequency when NPY release was markedly enhanced. When extrapolating from the NPY data, exocytosis of NA from large
J. M.
hNIX3ER(;
15
20
25
6-
2
4
6
8
10
10
(‘I ul.
01
0
5
10
loE6. 4.
*“>
2 0
-3o-
0
5
10
15
20
25
0
2
4
6
8
0
5
10
15
20
25
0
2
4
6
8
Time (see)
V
10 Hz stimulation
10
Time (min)
V
Fig. 7. Time course for the release of NA and NPY-LI as well as perfusion pressure changes in the blood perfused pig spleen upon stimulation with 10 Hz for 24 s. The effects in controls (filled circles, n = 5) have been compared with responses after pretreatment with desipramine (DMI; 0.5 mg/kg iv.; open squares, n = 5) or reserpine (1 mg/kg i.v. combined with preganglionic denervation, open circles, n = 5). Data are given as means & S.E.M. Note the different time scales for during (s) and after (min) stimulation.
A
corltml
1%
f.I59Hzinegularatim
Fig. 8. Recordings of perfusion pressure changes in the blood perfused pig spleen upon stimulation with an irregular pattern with an average frequency of 0.59 Hz for 8 min. The effects in controls have been compared with responses 24 h after reserpine treatment (1 mg/kg i.v. combined with preganglionic dener~tion).
Co-release of NPY and NA from pig spleen
Regulation of neuropeptide Y release by biophase concentrations of noradrenaline
zooo-1500. .E
~moo. E
,a
1
o- i_
.._o ,....__.. 0
,..,,,,,.
20
10
.@
30
2J.-...&*.
0
40
50
\
60
70
'51
6
:’ .._, ,:’
6’ j
...
‘..A
/TLk+_
/.*/’
‘. ”
‘0 . . . . .._..
c? 0
-kg>*
10
20
30
40
50
60
. 10
20
30
, 40
50
60
70
125-,
r” 1
loo75. 50.
g
25 04
l
0
Pret~atment with the monoamine uptake inDMI enhanced both functional responses and of NA evoked by nerve stimulation while overflow of NPY-LI at low frequency stimulation was reduced. The latter finding is probably related to prejunctional inhibition of NPY release due to increased local biophase concentrations of NA when the main inactivation pathway of the amine was inhibited by DMI (see Refs 21 and 43). Also nerve stimulation evoked release of another vesicular component DBH has been reported to be inhibited in the presence of DMI.6 Conversely, cr-adrenoceptor blockade has heen found to enhance nerve stimulation evoked output of the DBH (see Ref. 6) and NPY-LI from the spleen.**%” The present and earlieti8,” data therefore suggest that exocytosis of DBH, NPY-LI and NA from large dense-cored vesicles is regulated by the extraneuronal NA concentration at low to moderate stimulation frequencies. At 10 Hz, however, a clear-cut enhancement of NPY output relative to NA was observed in spite of DMI treatment (as revealed by the output ratio). This suggests that the relatively small NPY release at low frequencies may be related to prejunctional inhibitory a-adrenoceptor modulation. The common experimental procedure of using DMI treatment to obtain a better estimate of “true” NA release thus leads to an inhibition of NPY overflow, at least in the pig spleen using the low frequency types of stimulation. The enhanced nerve stimulation evoked NA output after DMI is therefore more likeiy to be related to a higher overtlow of transmitter that escapes local reuptake in the junctional cleft while the actual vesicular release may be depressed by the inhibitory feedback mechanism. Reserpine treatment (combined with p~gan~ioni~ denervation to prevent concomitant depletion of NPY-LI; see Refs 30 and 35) was associated with marked elevation of NPY overflow (see also Ref. 34) and prolongation of the perfusion pressure effect. This observation is also in line with the suggestion that the local biophase concentration of NA regulates NPY release. Furthermore, the difference between NPY overflow using continuous 2.3 Hz and 23b stimulations was smaller in reserpine-treated pigs than in controls. It has also been demonstrated that NPY can inhibit release of NA,3’,35 although it is not known whether this occurs under physiological conditions.
hibitor output
500~
0
,.-.
483
%3* 70
Time (min) B 20 Hz b stimulation
It
Fig. 9. Effects of eiectrical stimulation for 6Omin with regular bursts at 20 Hz for I s every 10 s on perfusion pressure (expressed as perfusion pressure increase, PP, mmHg), as well as output of NPY-LI @mol/min) and NA (pmoi/min) from the blood perfused pig spleen in uivo. Data (means f S.E.M.) represent observations from five control animals pretreated with desipramine (DMI; 0.5 mg/kg i.v.; filled circles) or five animals pretreated with reset-pine (1 mg/kg i.v.) combined with preganglionic denervation (open circles).
dense-cored vesicles may therefore also contribute to the total NA release upon stimulation with high frequency bursts. In man, the circtdating plasma levels of NPY-LI during resting conditions are 10w.~**~* Plasma concentrations of NA are already elevated upon moderate sympathetic activation in man while systemic NPY levels mainly increase upon severe stress.Zq,3R Furthermore, direct recordings of sympathetic nerve activity in man during graded hypotension have shown that systemic plasma NPY levels increase at a higher threshold of sympathetic activation than NA.* In the cat adrenal gland, however, where NPY is probably stored in the same ~hromaffin vesicles as adrenaline, high frequency burst stimulation enhanced both NPY and adrenaline outflow in paralIel.32
Fractional release, res~ppi~ and turnover
The fractional release of [ 3H]NA per nerve impulse was in parallel to the output of endogenous NA for all three types of stimulations. Since NPY is likely to be replaced to nerve endings exclusively by axonal transport at a rate of about 10 mm/h (see Ref. 12), the fractional release of NPY per nerve impulse upon a
long-term stimulation for 1 h was calculated from the recovered NPY-LI in the venous effluent. fnterestingly, a similar fractional release was obtained for both [‘HjNA and NPY-Ll. The release sites for sympathetic transmitter are known to be activated by only a few per cent of the stimuli (see Ref. 7) and therefore the present findings suggest that release of NPY ako occurs in a similar manner. Whether the release of NPY from large dense-cored vesicles is monoquantal or by partial exocytosis remains to be established. As discussed earlier with DBH as a marker (see Ref. 201, it cannot be excluded that the large dense-cored vesicles successively deplete their content of NPY-LI due to lack of local synthesis, and change into the small type of vesicles with maintained capacity to release NA. NPY is a larger molecule than NA and therefore our finding that NPY-LI appears in the venous effluent later and more slowly than NA may reflect different diffusion properties for the two molecules. Thus, in spite of the specialized fenestrated vessels present in the spleen, detection of maximal or total peptide release demands different sampling periods than for NA. The slow decline of plasma NPY-LI after release and the total recovery of released NPYLI in the splenic venous effluent during the I h stimulation compared to the reduction in tissue content suggests that the local degradation of peptide is relatively slow. In accordance, NPY has been demonstrated to have a considerably longer half-life in plasma (about 5 min in the pig” and man”“) compared to NA (l--2 min”‘). A delayed appearance and maintained levels of NPY-LI compared to NA in systemic blood after sympathetic activation have also been observed upon prolonged physical exercise in man.38 The maximal plasma levels of NPY-LI in the splenic venous effluent during stimulation with 20b for I h were in fact close to t nM which is in the same range as when exogenous NPY causes increase in perfusion pressure in the perfused spleen model?” or reduction in splenic blood flow in the intact animal.“” This clearly suggests that NPY could be of functional importance as a transmitter for sympathetic splenic vascular control since the concentration of locally released peptide close to the nerve terminals is likely to be much higher than that recovered in the venous eflluent. Long-term stimulation for I h in control pigs was associated with about 60% depletion of the total splenic content of NPY-Ll while the corresponding NA content was unchanged. This observation is likely to be directly related to differences in synthesis and resupply between these two agents. Thus, axonal
transport of NPY-LI was calculated to bc able to replace the released NPY-LI within a considerably shorter time period than the recovered NA in the splenic venous ehluent. Similar conclusions conccrning the relatively small importance of axonal transport versus local synthesis for resupply of NA after release have been obtained earlier in the spleen. Thus, only about I % of the total NA content of the spleen may be replaced by axonal transport in 24 h (see Ref. 15). The calculated total turnover time for the splenic NPY content considering the resupply by axonal transport was found to be around 1I days. This value may be compared with the dynamics for the vesicular enzyme DBH“ and the peptide VlPZ4 which have been used to derive estimates of about 5 days for the life span of noradrenergic granules and turnover for peptide containing vesicles, respectively. The decline in the output of NPY-LI upon prolonged stimulation under control conditions, and especially after the initial enhancement of release following pretreatment with reserpinc, suggests that limited amaunts of the total splenic content of peptide are available for release. It is thus likely that NPY present in axons represents a non-releasable pool (see also Ref. 30). Furthermore. the maximal depletion of NPY after reserpine treatment due to neurogenic activation is considerably smaller than that of NA.-‘U.3’No evidence for enhanced synthesis of NPY in the cell body region (coeliac ganglion) or increased axonal transport in the splenic nerves was observed 24 h after reserpine in the present expcriments. This apparent discrepancy from data obtained in the guinea-pig”’ may, however, be related both to the reserpine dose and the short survival period in the pig.
CONCLUSION
By comparing NPY and NA in the pig spleen it is possible to study the vesicular turnover and prejunctional regulation of transmitter release, as well as the possible cotransmitter role for a peptide in comparison with a classical transmitter. present studies have been supfrom the Swedish MRC (14X-07164, 14X-6554. 04X-03546-15Bt. The American Council for Tobacco Research Swedish ‘Tobacco Company, Petrus and Augusta Hedlunds Foundation, the Laerdai Foundation, Wiberes Stiftelse. The Swedish Societv of Medicine, Swedish Work ind Environmental Fund and Funds from the Karolinska Institute. For expert technical assistance we thank Miss Anette Hem&n and Miss Margareta Stensdotter and for secretarial help Mrs Karin Wiberg. Ackrtowled~ements-The
ported by‘grants
REFERENCES 1. 2.
Andersson P. O,, Bloom S. R. and Edwards A. V. (1982) Effects of stimulation of the chorda tympani in bursts on submaxillary responses in cat. J. Physiol., Land. 322, 469483. Andersson P. 0. (1983) Comparative vascular effects of stimulation continuously and in bursts of the sympathetic nerves to cat skeletal muscle. Aera physiof. scnnd. 118, 343-348.
Co-&ease
of NPY
and NA From pig spleen
485
3. Bini G., Hagbarth K. E., Hynninen P. and Wallin B. 0. (X980) Thermoregulatory and rhythm generating mechanisms governing the sudomotar and vasoconstrictor outflow in human cutaneous nerves. J. phy~iol., Lond. 306, 537-552. 4. Brimijoin S, (1972) Transport and turnover of dopamine-eta-hydroxyfase (EC 1.14.2.1) in sympathetic nerves of the rat. 3. [email protected], 2f83-2193. 5. Chubb J. W,, DePotter W. P. and DeSchaepdryver A, F. (1970) Evidence For two types of ~orad~nal~ne storage particles in dog spleen. Nature 288, 1203-1204. 6. Cubeddu L. X., Barnes E., Langer S. Z. and Weiner N. (1974) Release of norepinephrine and dopamine-betahydroxylase by nerve stimulation. I. Role of neuronal and extraneuronaf uptake and of alpha-presynaptic receptors. J. Phurmac. exp. Ther. 190, 431450. 7. Cunnane T, C. and StjHme L. (15184)Transmitter secretion from individual varicusities of guinea-pig and mouse vas deferens: highly jn~~ittent and MOnOquantaf. Meursrrience 13, l-20. 8. Delius W., Hagbarth K-E., Hongefl A. and Waflin B. G. (1972) General characteristics of sympathetic activity in human muscle nerves. Acta physiol. stand. 84, 65-81. 9. Eckberg D, L., Rea R., Andersson O., Hedner T., Pernow J., Lundberg J. M. and Wallin B, G. Baroreflex modulation of sympathetic activity and sympathetic neurotransmitters in man. Acta pkysiot. scund. (in press). f 0. Esfer M. (1982) Assessment of sympathetic nerves function in humans f’rom noradrenaline plasma kinetics.
E. and Pernow J. (1984) Guanethidine-sensitive release of neuropeptide Y-like immunoreactivity in the cat spleen by sympathetic nerve stimulation. Neurosci. Lett. 52, 175-180. 29. Lundberg J. M., Martinsson A., Hems&n A., Theodorsson-Norheim E., Svedenhag J., Ekblom E. and Hjemdahl P. (1985) Co-refeasz of neuropeptide Y and catechofamines during physical exercise in man. B&hem. bloph_ys. Res, Cummtm. 133, Z&36. 30. Lundberg J. M., Saria A., France-Cereceda
of reserpine and 6-hydroxydopamine
A., HBkfelt T., Terenius L. and Gofdstein M. (1985) Differential eff?cts on neuropeptide Y (NPY) and noradrenaline in peripheral neurons. Naunyn-
Schmiedeberg’s Arch. Pharmac. 328, 331-340. 3t. Lundberg J. M. and Hakfelt T. (1986) Multiple co-existence of peptides and classsical transmitters in peripheral autonomic and sensory neurons-functionaf and pharmacological implications. In Progress in Brain Research
(eds Hiikfelt T., Fuxe K. and Femow B.), Vol. 6g, pp. 243-262. Efsevier, Amsterdam. 32. Lundberg J. M., Fried G., Pernow J. and ~eodo~so~-Norheim E. (1986) Corelase of neuropeptide Y and catechoiamines upon adtcnaf activation in the cat. Acta physioi. stand. 126, 231-238. 33. Lundberg J. M., Hems&n A., Fried G., Theodorsson-Norheim E. and Lagercrantz H. (1986) High plasma levels of neuropeptide Y (NPY)-like immunoreactivity and catechofamines in newborn infants. Acru physiol. scund. 126, 471473. 34. Lundberg J. M., Rudehilf A., Soffevi A., The~or~o~-Norhejm E. and Hamberger B. (f986f Frequency- and ~rpi~e~~ndent chemical coding of sympathetic transmjas~on: difFerentiaf release of noradrenafine and neuropeptide Y from pig spleen. Neurosci. Lett. 63, 96-100.
486
J. M.
hNDHbXG
PI al.
35. Lundberg J. M., Pernow J., France-Cereceda A. and Rudehill A. (1987) Effects of antihvnertensive drws on sympathetic vascular control in relation to neuropeptide Y (NPY). J. Cardiouasc. Pharm. Sup$. 10, S51~ S58.36. Lundberg J. M., Rudehill A. and Sollevi A. Pharmacological characterization of neuropeptide Y and noradrenaline mechanisms in sympathetic control of pig spleen. Eur. J. Pharmac. (in press). 37. Nilsson H.. Ljung B., Sjoblom N. and Wallin B. G. (1985) The influence of the sympathetic impulse pattern on contractile response of rat mesenteric arteries and veins. Actu physiol. stand. 123, 303-309. _ 38. Pernow J.. Lundberg J. M.. Kaiiser L.. Hiemdahl P., Theodorsson-Norheim E.. Martinsson A. and Pernow B. (1986) Plasma neuropeptidk Y-like immunoreactivity and catecholamines during various degrees of sympathetic activation in man. C&I. Phys. 6, 561-578. 39. Pernow J., Lundberg J. M. and Kaijser L. (1987) Vasoconstrictor effects in uivo and plasma disappearance rate of neuropeptide Y in man. Life Sci. 40, 47754. 40. Rudehill A., Olcen M., Sollevi A., Hamberger B. and Lundberg J. M. (1987) Release of neuropeptide Y upon haemorrhagic hypovolemia in relation to vasoconstrictor effects in the pig. Acta physiol. scund. 131, 517 -523. 41. Rudehill A., Lundberg J. M., Sollevi A. and Hjemdahl P. (1987) Elevations of neuropeptide Y-like immunoreactivity and catecholamines in plasma upon increased intracranial pressure in the pig. Acfa anaesth. stand. 31, 131-I 38. 42. Smith A. D., DePotter W. P., Moerman E. J. and DeSchaepdryver A. F. (1970) Release of dopamine-beta-hydroxylase and chromogranin A upon stimulation of the splenic nerve. Tiss. Cell 2, 547-55 1. 43. Starke K. (1987) Presynaptic a-autoreceptors, Rev. Physioi. Biochem. Pharmac. 107, 74-146. 44. Theodorsson-Norheim E., Hemsen A. and Lundberg J. M. (1985) Radioimmunoassay for neuropeptide Y (NPY): chromatographic characterization of immunoreactivity in plasma and tissue extracts. Stand. J. Lab. Invest. 45,355-365. 45. Wallin B. G. (1981) New aspects of sympathetic function in man. In Butterworths Infernafional Medical Reviews Neurology I (eds Stalberg E. and Young R. R.), pp. 145-167. Butterworth, London. (Accepted
I I July 1988)
03044522/89 $3.00+ 0.00 PergamonPressplc
1989
IBRO
CO-RELEASE OF NEUROPEPTIDE Y AND NORADRENALINE FROM PIG SPLEEN IN’ VIF/o: IMPORTANCE OF SUBCELLULAR STORAGE, NERVE IMPULSE FREQUENCY AND PATTERN, FEEDBACK REGULATION AND RESUPPLY BY AXONAL TRANSPORT J. M. LUNDBERG,*? A. RUDEHILL,~ A. .%LLEVI,§ G. FRIELQ and G. WALLIN// Departments of *Pharmacology and $Physiology, Karolinska Institute and Department of SAnaesthesia, Karolinska Hospital, Stockholm, Sweden, and I/Department of Clinical Neurophysiology, Sahlgrenska Hospital, Gothenburg, Sweden
importance of subcellular storage, nerve impulse rate and pattern, and feedback regulation, and neuropeptide Y-like immu~oreactivity, was studied in the blood perfused pig spleen in viva. Vasoconstrictor responses were recorded as perfusion pressure changes. Subcellular fractionation experiments using sucrose density gradients showed a bimodal distribution of noradrenaline (peak concentrations at 0.8 and 1.1 M sucrose) while only one main peak of neuropeptide Y was present (at 1.1M sucrose). Overflow suggesting release of noradrenaline and neuropeptide Y-like immunoreactivity could be detected after 10 s stimulation at 10 Hz. The ratio for the output of noradmnal~ne and neuro~ptide Y upon continuous nerve stimulation in control animals decreased with frequency. After inhibition of noradrenaline reuptake by desipramine the vasoconstrictor response and noradrenaline output were enhanced while the corresponding overflow of neuropeptide Y was reduced by 50% at 0.5 Hz. Stimulation with the irregular or regular bursting patterns at high frequencies caused larger perfusion pressure increase and relative enhancement of neuropeptide Y output compared to noradrenaline than a continuous stimulation both before and after desipramine treatment. A similar fractional release per nerve impulse was calculated both for ~3H]noradrena~ne (5.6 + 1.0 x IO-‘) and neuropeptide Y (7.3 f 0.3 x lo-‘). After reserpine treatment combined with preganglionic denervation the vasoconstrictor responses were more long-lasting, neuropeptide Y release was enhanced while noradrenaline content and release were reduced by 99%. The difference in neuropeptide Y overflow between continuous and bursting types of stimulation was smaller after reserpine treatment. After prolonged intermittent stimulation with regular bursts (20 Hz) for I h the splenic content of neuropeptide Y was reduced by 580/o, while no change was observed for noradrenaline. The maximal perfusion pressure increase upon prolonged nerve stimulation after reserpine was simiiar in control and reserpine-treated animals, but after reserpine the vasoconstrictor response and neuropeptide Y release were subjected to fatigue. Ligation experiments of the splenic nerves revealed the splenic neuropeptide Y content was resupplied by axonal transport with a calculated total tissue turnover time of 11 days. In contrast, axonal transport contributed only to a marginal extent for the resupply of noradrenaline. In conclusion, stimulation with high frequency bursting preferentially enhanced overflow of neuropeptide Y compared to the co-existing classical transmitter noradrenaline, suggesting that exocytosis of material from large dense-cored vesicles mainly occurred under such stimulation conditions. The local biophase concentrations of noradrenaline seem to regulate neuropeptide Y release. Due to the limited resupply by axonal transport, excessive and prolonged release of neuropeptide Y cannot be maintained without depletion of terminal stores. Abstract-The
as well as resupply by axonal transport for the release of noradrenaline
Recordings of impulse activity in single sympathetic nerve fibres have indicated an average rate of firing up to 3 Hz during basal conditions.” The pattern of firing is highly irregular. 8*‘6.45 Thus, the nerve impulses occur in bursts separated by quiescent periods and the instantaneous discharge frequency (interval between two successive action potentials) could reach tTo whom correspondence should be addressed at the Department of Pharmacology, Karolinska Institutet, _ _ _ . ..~ _ P.U. Box bwoo, s-104I 01 Stockholm, Sweden. Abbreviations: DBH, dupaunuc-,,-rryu --^-A-- D ‘--+oxylase; DMI, desipramine; HPLC, high-performan tee liquid chromatography; Lf, like immunor~ctivity, ; NA, noradrena. line; NPY, neuropeptide Y; VIP, vasoactrve Intestinal polypeptide. NSC
28,2--H
475
35 Hz even at low average rates of 1 Hz.17 Recently, stimutation with the irregular bursting activity of human skin vasoconstrictor fibres was found to cause signi~cantly larger contractions of mesenteric arteries in vitro than a continuous frequency giving the same total number of impulses.” Furthermore, the atropine-resistant vasodilatory response in the cat submaxilIary gland upon parasympathetic nerve stimulation, which is likely to be related to release of vasoactive intestinal polypeptide (VIP),22,26 was enhanced by stimulation with bursts of high frequencies.’ Whether the co-reiease of the classical transmitter was also enhanced in the ,. , acetylcholine , , *. . _ sanvary grana wnen usmg tne burstmg type ot strmulation
is not known
although
functional
data sug-
476
.I. M. L~:YIWIK<~ <‘I i/i
gest that continuous seems to favour VIP choljne,?‘._‘?~‘”
high frequency stimulation release compared to acctyl-
Increasing evidence suggests that neuropeptide Y (NPY), which co-exist?’ and is co-released with noradrenaline (NA) from sympathetic nerves upon electrical stimillation~‘.~ or reflexogenic activation.~‘.~~‘~~ is of importance for sympathetic vascular control at both pre- and postjunctional levels.‘“~“~’ The present investigation evaluates the importance of impulse rate and pattern as well as local biophase concentrations of NA for the rcleasc of NA and NPY-like immunoreactivity (LI) from the pig spleen !rr i:it~. In addition, the subcellular storage and the dependence of axonal transport for resupply of terminal stores of NPY relative to NA is also investigated’ EXPERIMENTAL PROCEDURES
The coehac ganglion, splenic nerves along the main splenic vessels and the caudal portion of the spleen were dissected out from pigs under pentobarbitat anaesthesia, frozen on dry ice, weighed, extracted and the content of NA and NPY-LI was determined (see below). Subcellular jiuctionation experiments Subcelluiar fractionation combined with sucrose density gradients was performed (n = 4) on about log of the caudal pole of the spleen removed from anaesthetized pigs (see below). Details of the protocol have been published elsewhere.‘Z-‘4 Briefly a 10,OOOg supernatant from a spleen homogenate was placed on a sucrose density gradient (0.25- 1.4 M sucrose) and centrifuged at 280,OOOg max. The particulate and supematant fractions were then separated (see Ref. 14). The concentration of NA and NPY-LI in each fraction was determined by high-performance liquid chromatography (HPLC)‘” or radioimmunoassay,” respectively. Before the NA determinations, the sucrose fractions were extracted by perchloric acid (NA). Furthermore, before NPY radioiInmunoassay, the fractions were Passed through Sep Pak’” cartridges and the peptide content separately eluted.‘” Functional
in vivo smdies
Functional experiments were performed using the blood perfused pig spleen (for details of the ex~rimenta~ set up, see Refs 34 and 36). Briefly, the spleen from anaesthetized pigs was placed in saline at 37’C and constantly perfused with blood from a femoral artery. The perfusion pressure in the spleen (perfusion rate was adjusted to obtain a basal pressure of 8W?OmmHg) was recorded continuously and alterations were taken as indication of changes in vascular resistance. The blood flow in the splenic venous effluent was monitored by an electromagnetic flow probe (Nycotron”, Norway) and changes in outflow with a constant input are considered to mainly reflect alterations in splenic volume due to capsule contraction. The postganghonic sptenic nerves were stimulated electrically (see below) and output of NPY-LI and NA in the splenic venous eflluent was determined by calculating the venoarterial plasma concentration differences multiplied by plasma flow. NPY-LI in plasma and splenic tissue was determined by radioimmunoassay using antiserum Nl (see Ref. 44). The NPY-LI in splenic tissue and plasma detected by this antiserum was characterized by reversed phase HPLC in comparison to synthetic porcine NPY (l-36) (see Refs 34 and 44). The levels of catecholamines in plasma and splenic tissue were determined by using HPLC combined with electrochemical
detection (set Ref. IX). The ourput of NPY-LI and NA during (after 23 of the stimul~ti~~n) and 30 s. 2. 5. If) :tnd 15 mm after the stimulations was integrated and given as a total value of recovered material for each stimulation. In separate experiments the time courses for the overflow upon stimulation with 2 and IO Hz were followed in more detail in controls. after desipramine (DMi) and after reserpine treatment (see below).
Electrical stimulations (IO V. 5 ms) were performed on the transected perivascular nerve bundles accompanying the main splenic artery using a bipolar platinum electrode connected to a Grass 588 stimulator. First, stimuiations with regular frequencies at 0.5, 2 and 10 Hz giving a total of 240 impulses with 20 min intervals between stimulation periods (n = 5) were used. Second, the importance of impulse pattern for the functional response as well as overtlow of NPY-LI and NA was studied (n = 5). To this end, the responses obtained with a continuous frequency of 0.59 Hz for gmin or 2.3 Hz for 2min (a total of 285 and 276 impulses, respectively) were compared to those obtained with the irregular stimulation pattern of a few unit recordings of human vasoconstrictor nerve fibre activity. The neural recording was made under resting conditions from the muscle vasoconstrictor hbres in the peroneal nerve (0.59 Hz) and cutaneous vasoconstrictor fibres from the median nerve (2.3 Hz) at the elbow using tungsten electrodes (see Refs 3 and 16). The occurrence of sudomotor activity in the median nerve was detected in parallel using recordings of galvanic skin responses within the innervation zone in the hand. One segment of neural recording without sudomotor activity was chosen and the imp&es were transformed into standard pulses, stored on a magnetic tape and used for triggering the Grass stimulator. The average frequency of an 8 min period was 0.59 Hz and in a 2 min period 2.3 Hz. although the instantaneous frequency (i.e. rate cakntiated from the interval between two successive action potentials) was as high as 200 Hz. A stimulation sequence with regular bursting activity at a frequency of 23 Hz for 1s every IO s was also used for additional comparison. For each stimulation period the maximal perfusion pressure increase was used as a tneasure of the functional response. The order of the different types of stimulations was varied randomly. After a session with stimulations under control conditions. 100 PCi of [ ‘H]NA was infused into the splenic artery for IO min. After an additional 20 min. the monoamine uptake blocker DMI (desiprdmine hydrochloride, CIBA. Sweden 0.5 mg/kg) was given intravenously. Following two priming stimuiations 30 min later, fractional release of [‘H]NA was determined in alumina extracted plasma (see Ref. 19) in parallel to endogenous NA and NPY-LI in tnW%il and venous samples. In separate pigs (n = 5). the left splanchnic nerves were sectioned under pentobarbital anaesthesia to decentralize the postganglionic nerves to the spleen and reserpine (1 mg/kg) was given i.v. 24 h prior to the experiments to deplete NA with the NPY content intact (see Ref. 35).
To study the relative importance of axonal transport for resupply of NPY-LI compared to local synthesis of MA, long-term stimulations using regular 20 Hz burst stimulation for 1 h (see above) were performed in control (n = 5) and reserpine-treated (n = 5) pigs. Release during and after, as well as tissue content of NPY-LI and NA, were determined at the beginning and the end of the stimulations. Furthermore, one of the two main bundles of the splenic nerves along the remaining portion of the splenic artery and vein about 2 cm proximal to the transection site was ligated using a silk suture and a glass rod (see Ref. 24). Ten hours later two 5mm nerve segments. one just proximal to the
417
Co-release of NPY and NA from pig spleen
NPY-LI were lower in the coeliac ganglion (16O:l) and in the snlenic nerves (200: 1) compared to the _ spleen (625: i, Table 1). _
porcine NPY (l-36) (Fig. 2). When using stimulation with regular impulses the output of both NA and NPY-LI increased with frequency. Thus, the output of NA and NPY-LI was 360 4 38 vs 1.2 f 0.3 pmol at 0.5 Hz and 2233 + 201 vs 33 f 3 pmol at 10 Hz. The calculated ratio between NPY-LI/NA overflow was then around 1:300 at 0.5 Hz while at 10 Hz a ratio of about 1: 90 was observed (Fig. 3). To prevent the major removal process for NA, i.e. neuronal reuptake, DMI pretreatment was used. In the DMI pretreated animals the corresponding ratios between output of NPY-LI and NA were about 1: 1500 at 0.5 and 1: 200 at 10 Hz (Fig. 3). Compared to untreated animals this change in ratio seemed to be due both to an enhanced overflow of NA escaping reuptake as well as a significant reduction in output of NPY-LI. Thus, NPY overflow was reduced by 48 + 5% at 0.5 Hz (P < O.Ol), by 28 f 3% at 2Hz (P < 0.05), while only a small, non-significant change occurred at 10 Hz after DMI (n = 5).
Subcellular fractionation
Impulse pattern and release
In the subcellular fractionation experiments particulate NA had a bimodal distribution on the sucrose density gradient. The light density peak at 0.8 M sucrose had a concentration of 18 1 x 24 nM NA, while the peak at 1.1 M sucrose contained 266 f 53 nM (Fig. 1). NPY-LI, however, was present mainly in heavier fractions with a peak value of 393 f 44 pM at a sucrose density of 1.1 M (Fig. 1). In the supernatant the highest concentrations of NA and NPY-LI were present in the first fraction of the gradient (126 f 14 nM and 70 f 12 pM, respectively).
Electrical stimulation of the splenic nerves with regular frequency (2.3 Hz) caused a static increase in perfusion pressure (Fig. 4) of 74 f 4 mmHg (n = 7) as well as an increase in blood flow from the spleen. Furthermore, an overflow of both NA and NPY-LI occurred into the splenic venous effluent with the corresponding integrated outputs of 1075 f 147 and 7 + 2 pmol (n = 9), respectively. Stimulation with the recorded sympathetic nerve activity for 2 min caused a perfusion pressure increase which was also irregular in appearance (Fig. 4) and on average 41% larger than that caused by the continuous 2.3 Hz stimulation (Figs 4 and 5). The overflow of NA was only marginally increased compared to that obtained by continuous stimulation while the output of NPY-LI was significantly enhanced by 22% (P < 0.01, Fig. 5). Intermittent stimulation with bursts at 23 Hz for 1 s every 10 s for 2 min (23b) caused a transient increase in perfusion pressure for each burst (Fig. 4). The maximal perfusion pressure effect upon 23b stimulation was significantly (56%, P < 0.05) larger than
ligation and one from the non-ligated nerve, were removed, weighed, frozen on dry ice and subsequently analyzed for content of NPY-LI and NA. The increase of NPY-LI and
NA per hour of nerve ligation in the postganglionic nerve tissue just proximal to the ligation for both branches was calculated after subtracting the content in the non-ligated one. In addition the content of NPY-LI and NA in postganglionic spleen nerves from intact animals was determined for comparison.
Statistics Student’s t-test was used for comparing the functional responses and output of NA and NPY-LI for the continuous 0.59 and 2.3 Hz stimulations with the irregular human recordings of nerve activity or regular bursting patterns. A P-value 40.05 was considered significant. RESULTS
Tissue content The molar ratios between tissue content
of NA and
Frequency dependence of release
Electrical stimulation of the splenic nerves with continuous impulses caused increase in perfusion pressure and venous blood flow, as well as overflow of NA and NPY-LI into the venous effluent, suggesting release. Reversed phase HPLC analysis revealed that NPY-LI in splenic extracts and plasma in the splenic venous effluent collected during nerve stimulation eluted at the same position as synthetic
Table 1. Tissue content and molar ratio of noradrenaline and neuropeptide Y-like immunoreactivity various portions of postganglionic sympathetic neurons to pig spleen Tissue Coeliac ganglion Splenic nerves Splenic nerves 10 h ligation Spleen Spleen 1h stimulation
NA (nmol/g) control reserpine 31_+5 14 + 3 48& 7** 11.6+ 1.0 9.6 _+ 0.2
NPY-LI (pmol/g) control reserpine
in
NA/NPY-LI ratio control reserpine
-
195+39 68_f 15 1206 + 220**
235 + 25 51+9 1024 + 113**
160:1 200:1 4O:l
-
0.14 * 0.02*** 0.12 + 0.02***
18.6 + 1.4 7.8 & 1.3**
14.4 _+ 1.1 3.2 + 0.7**
625:l 123O:l
1O:l 38:1
(expressed as means f S.E.M., n = 4-10) of NA (nmol/g) and NPY-LI (pmol/g) in cell body region (coeliac ganglion), axons (splenic nerves) and terminal areas (spleen). The splenic nerves were ligated for 10 h under anaesthesia. Furthermore, the splenic nerves were stimulated for 1 h with regular bursts at 20 Hz for 1 s every 10 s. Controls were pretreated with DMI (0.5 mg/kg). Reserpine pretreatment (1 mg/kg i.v.) was performed 24 h prior to the experiments and combined with preganglionic decentralization. Asterisks denote significant differences compared to control (**P < 0.01 and ***p < 0.001)
Tissue content
using Student’s
?-test.
500
400
1
O--O .-.
-500
NPY NA
6
-400 -300 2 5 -200 24 -100 -0
Fig. I. Subcellular distribution of particulate NA and NPY-LI after centrifugation of a 10,000 g,,,,, supernatant derived from a pig spleen homogenate on a continuous sucrose density gradient of 280,000 g,,,., for 90min. The density of each fraction is given as M sucrose, and content of NA in nM and NPY-LI in pM, respectively. Data are given as means + S.E.M.
the response using continuous 2.3 Hz stimulation (Figs 4 and 5). Furthermore, the output of NPY-LI was enhanced more than three-fold upon 23b stimulation and the NA overflow was also significantly increased (by 80%, P < 0.05) (Fig. 5). When integrated during and up to 15 min after the start of the 2 min stimulation period, the ratio between the recovery of NA and NPY-LI in the splenic venous effluent was 115 : 1 at 23b and 160: 1 at 2.3 Hz, respectively. Regulation of release concentration
by biophase
more detail for continuous stimulations with 2 and 10 Hz (Pigs 6 and 7). A consistent detectable overflow of NA and NPY-LI was present after 30 s and 10 s upon stimulation with 2 and 10 Hz, respectively. After this the stimulation perfusion pressure declined to values below control. As described in Fig. 6, the decline in the plasma outflow of NA was much more rapid than for the NPY-LI. Pretreatment with DMI caused an enhancement of the maximal perfusion pressure increase and prolongation of the decay phase (especially at 2 Hz). After reserpine the overflow of NPY-LI was enhanced (especially at 2 Hz) while no detectable output of NA was present. The perfusion pressure response after reserpine treatment was slower in developing (at 2 Hz) and longer-lasting (at both 2 and 10 Hz). After reserpine treatment 2.3 Hz and 23b stimulations were associated with about three- and two-fold larger release of NPY-LI while NA overhow as well as splenic content of NA (Table 1) were reduced by 99%. The relative difference regarding NPY release upon continuous 2.3 Hz stimulations and bursting patterns was much smaller after reserpine (Fig. 5). The maximal per-
noradrenaline
After pretreatment with DMI, nerve stimulation with all three types of pattern caused a more rapid initial rise of the perfusion pressure compared to control conditions (Fig. 4). Furthermore, the irregular perfusion pressure responses upon stimulation with recorded sympathetic nerve activity as well as the effects of intermittent 23b stimulations were much less apparent (Fig. 4). A slight potentiation of the maximal perfusion pressure effects upon nerve stimulation occurred after DMI treatment (Fig. 4) (by 16% at 23b, 24% at irregular and 43% at continuous stimulation). The recovered output of NA in the splenic venous effluent was enhanced by DMI treatment using all three types of stimulations (by 24% at 23b, 45% at irregular and 87% at continuous stimulation). The relative differences between the responses in maximal perfusion pressure to the three stimulation patterns were smaller after DMI pretreatment while the overflow of NPY-LI was still much larger at 23b (Fig. 5). The NA output in the DMI-treated animals as well as the calculated fractional release of [3H]NA per nerve impulse was of comparable magnitude for all three types of stimulations (2.3 Hz: 4.5 + 1.7, human activity: 5.1 +_ 1.2 and 23b: 5.6 + 1.0 x 10V5 for fractional release, respectively). These values were of comparable magnitude to the calculated fractional release of NPY- at 23b (7.2 + 0.3 x IO-‘). The time course for the perfusion pressure increase and the outflow of NPY-LI and NA was followed in
PIG SPLEEN
NPY-LI hndesmactkm
ELUENT 8%
I
PIG
I
PLASMA
20
40
60
60
FRACTKM NUWER
Fig. 2. Reversed phase HPLC chromatogram of NPY-LI in an acetic acid extract from the pig spleen (middle panel) and plasma collected in the splenic venous effluent 2 min after nerve stimulation using intermittent bursts at 23 Hz (lower panel) in comparison with synthetic porcine NPY (l-36, top panel). An acetonitrile gradient with water was used and the content of NPY-LI in each fraction was analysed by RIA.
419
Co-release of NPY and NA from pig spleen
_ 10
2
0.5
FREQUENCY (Hz)
Fig. 3. Ratio between the totally recovered integrated output of NA and NPY-LI in the splenic venous effluent in
uivd upon nerve stimulation with c&tinuous frequencies at 0.5 Hz (for 8 min). 2 Hz (for 2 min) and 10 Hz (for 24 s) giving a‘total of 24d imp&s. Filled dircles represeAt control conditions and open circles the ratio after pretreatment with desipramine (DMI; 0.5 mg/kg i.v.). Data are given as means + S.E.M. (n = 7).
fusion pressure increase after reserpine treatment was slower in developing and longer-lasting (Figs 6 and 7) and only marginally reduced. Stimulation using the irregular impulses with an average frequency of 0.59 Hz also caused a rapidly fluctuating pattern in the perfusion pressure response (Fig. 8). Compared to a continuous 0.59 Hz stimulation the irregular pattern caused about a two-fold larger increase of NPY output and 40% larger perfusion pressure increase. After reserpine treatment the outflow of NPY-LI was enhanced about threefold at 0.59 Hz (from 2.3 + 0.2pmol to 6.0 f 1.5 pmol, P < 0.01). The perfusion pressure response was, however, smaller than in the controls, less fluctuating and longer-lasting (Fig. 8). Long-term stimulation
Electrical stimulation with 20b for 1 h in DMItreated animals (n = 5) caused a marked perfusion
1s
*s HZ
pressure increase, which slowly declined after 2-5 min to a steady-state level at about 60-70% of the maximal response which was maintained between 10 and 60 min of stimulation (Fig. 9). After the stimulation was turned off, the perfusion pressure response returned to baseline values within 5 min (Fig. 9). The stimulation-evoked maximal overflow of NPY-LI in the splenic venous effluent was delayed in comparison to NA (Fig. 9). Thus, whereas the maximal output of NPY-LI was present after about 20min the peak outflow of NA was already seen after about 5 min of stimulation (Fig. 9). The ratio between NA and NPY-LI in the splenic parenchyma (63O:l) may be compared with the ratio between the output of NA and NPY-LI upon nerve stimulation, which was 3000: 1 after 30 s and about 200: 1 after 20 min stimulation (Fig. 9). The maximal levels of NA and NPY-LI recovered in the venous effluent during the stimulation were 172 _+21 nmol/l and 692 f 178 pmol/l, respectively. After the stimulation was turned off, NA output rapidly decreased and was basal within 2 min while the NPY output was still elevated 5 min after cessation of the stimulation (Fig. 9). After reserpine treatment, the increase in perfusion was close to the maximal response in the control after about 5 min of stimulation. Then, the effect gradually declined (Fig. 9). A marked increase in NPY output occurred during the first phase of stimulation, but then the outflow of NPY-LI declined (Fig. 9). No clear-cut NA release was detected after reserpine. The maximal levels of NA and NPY-LI in the venous effluent upon stimulation after reserpine were 3 f I nmol/l and 1150 + 85 pmol/l respectively. The tissue concentration of NPY-LI in the spleen after 1 h stimulation with 20b in control pigs had decreased by 58% and by 78% after reserpine com-
HwIll
*cTI”ITY
ZSHZM
Fig. 4. Effects of electrical stimulation of the splenic nerves for 2 min with a total of 276 pulses using a continuous frequency of 2.3 Hz (left panel), an irregular pattern from recordings of human vasoconstrictor nerve activity (middle panel), and with regular bursts at 23 Hz for 1 s every 10 s (23b, right panel), on perfusion pressure (PP, mmHg) in the blood perfused pig spleen in ho. The control effects are compared with responses 20 min after iv. pretreatment with desipramine (DMI; 0.5 mg/kg). Note that after DMI a more rapid rise in the early phase of the perfusion pressure was present and most of the burst evoked variations of perfusion pressure were lost. Different time scales are used for the three upper (10 s) and the bottom (1 s) tracings of the stimulation pattern.
rather similar values were obtained. to 7.9 X 10 ‘, respectively.
i.e. 5.8 compared
Axrmai transport and turnnoer In the axonal transport experiments the accumulation of NPY-LI and NA proximal to the ligation of the whole splenic nerves was calculated to be 5 F I (n = 7) and 220 + 18 (n = 6) pmolih, respectively. This gives a calculated mean total turnover time in nerve endings of the total splenic content of NPY-LI (average splenic weight 75 g multiplied by tissue concentration; 18.6 pmol/g = 1395 pmol and divided by the accumulation via axonal transport per h; 5pmol) of about 11 days. To replace the 58% reduction in tissue content of NPY-LI seen after a 1 h stimulation with 20b would then take 740/5 h, i.e. about 6 days by axonal transport. When the corresponding calculations are made for NA, total replacement of the splenic NA content by axonal transport would need total content (858,000 pmol)/ 220 pmol/h = 162 days and the recovered overflow of NA during 1 h stimulation 102,~0/220 = 19 days. The content of NPY-Ll in the coeliac ganglion, unligated splenic nerves and the accumulation of NPY-LI above the ligature of the splenic nerves did not change 24 h after reserpine treatment (Table 1).
400 T
200
T
DISClJSSION
Vesicular storage in relation to release
Human activity
23 Hz bursts
Fig. 5. Summary of experiments using electrical stimulation of the splenic nerves for 2 min with the irregular bursting pattern and reguiar bursts at 23 Hz for I s every 10 s. The effects (means f S.E.M.) on perfusion pressure (PP, mmHg), as well as output of NPY-LI and NA, have been expressed as percentage of the responses obtained for a continuous stimulation at 2.3 Hz giving the same total number of impulses in the individual experiments. Open bars represent observations in control animals (n = 9). hatched bars observations after pretreatment with desipramine (DMI; 0.5mg/kg i.v.; n = 5) and filled bars responses 24 h after reserpine treatment (1 mg/kg iv.) combined with preganglionic denervation. Asterisks denote significant differences from the continuous stimulation (*P cc 0.05, ***P -=z0.01) using Student’s f-test.
pared to non-stimulated preparations (Table 1). A total output of 422 + 48 and 541 + 109 pmol NPY-LI was recovered in the splenic efffuent during and IOmin after the stimulation in control and reserpinized pigs, respectively. The splenic concentration of NA, on the other hand, was not significantly reduced after 1 h stimulation in the controls and 102 f IO nmol was recovered in the venous effluent (Table 1). When a fractional release per nerve impulse was calculated in control pigs for the whole stimulation period for [3H]NA and output of NPY-LI,
Adrenergic neurons in the spleen contain two types of storage vesicles as revealed by su~ellular fractionation experiments, the so-called *‘light” and “heavy” NA vesicles, which most likely correspond to small and large dense-cored vesicles, respectively (see Refs 5 and 11). In the present study the subcellular fractionation experiments showed that whereas NA had a bimodal distribution, NPY-LI was mainly present in the heavier peak of NA (see also Refs 12-14). The large vesicles release their content of amines, proteins such as dopamine-fi-hydroxylase (DBH) and chromatog~nin probably by exocytosis upon stimulation of the splenic nerves.42 Since the output ratio NA/NPY was not fixed but decreased with increasing stimulation frequencies, our data support the hypothesis that the classical transmitter (NA) is preferentially released from small storage vesicles at low frequencies while at higher frequencies an increasing proportion of large dense-cored vesicles release NPY (as well as NA). Not only frequency of stimulation but also impulse pattern may be of importance. Thus, the present data show that stimulation with either regular (23b) or irregular bursts of sympathetic impulses at high frequencies induced a larger vasoconstrictor response than a continuous frequency giving the same total number of impulses during the stimuIation period. This finding is in agreement with in vitro data using stimulation with human sympathetic activity regarding contractions of rat mesenteric arteries” while
G
481
of NPY and NA from pig spleen
Co-release
o
2
4
6
6
10
0
2
4
6
8
IO
60.
I
60.
l
4o 20.
-201
r.,.. 0 20
40
60
Time (set
V
60
100
I 120
Time (min)
)
2
Hz stimulation
V
Fig. 6. Time course for the release of NA and NPY-LI as well as perfusion pressure changes in the blood perfused pig spleen upon stimulation with 2 Hz for 2 min. The effects in control animals (filled circles, n = 5) have been compared with responses after pretreatment with desipramine (DMI; 0.5 mg/kg i.v.; open squares, n = 5) or reserpine (1 mg/kg i.v. combined with preganghonic denervation, open circles, n = 5). Data are given as means + S.E.M. Note the different time scales for during (s) and after (min) stimulation.
observations concerning sympathetic control of cat skeletal muscle blood flow in vivo did not show any enhanced functional response upon stimulation with bursts.2 In addition to increased functional responses we also found that irregular stimulation increased the output of NPY-LI considerably more than that of NA. This suggests that not only the frequency but also the impulse pattern is of crucial importance for the release of the peptide NPY compared to the coexisting classical transmitter NA. The background for such a “nerve impulse pattern coded release” of bioactive substances from autonomic nerves may be related to partly different subcellular storage sites for NPY and NA. Thus, whereas the established view is that NA is present in both small and large densecored vesicles in, for example, the spleen (see Refs 5 and 11) NPY-LI (Refs 12-14 and present results) seems to be preferentially stored in large dense-cored vesicles. Therefore, available data suggest that material from the large dense-cored vesicles is mainly
released upon high frequencies of stimulation. As discussed above, NA is also likely to be present in the large dense-cored vesicles. It has been argued that only l&15% of the large vesicles in a terminal could account for a significant amount of the NA released because of their relatively larger storage capacity compared to the small vesicles.” Furthermore, it has been reported that 25-30% of the vesicles in the terminals of the pig spleen are of the large type.20 From the present subcellular fractionation experiments it may be estimated that about half of the NA content may originate from the large vesicles. Furthermore, considering NPY overflow, this pool of large vesicles appears to contribute only to a moderate extent to the total NA output from the spleen upon stimulation with continuous low frequencies. An increase in the NA output in the absence of DMI was, however, seen upon stimulation with 23b compared to a continuous 2.3 Hz frequency when NPY release was markedly enhanced. When extrapolating from the NPY data, exocytosis of NA from large
J. M.
hNIX3ER(;
15
20
25
6-
2
4
6
8
10
10
(‘I ul.
01
0
5
10
loE6. 4.
*“>
2 0
-3o-
0
5
10
15
20
25
0
2
4
6
8
0
5
10
15
20
25
0
2
4
6
8
Time (see)
V
10 Hz stimulation
10
Time (min)
V
Fig. 7. Time course for the release of NA and NPY-LI as well as perfusion pressure changes in the blood perfused pig spleen upon stimulation with 10 Hz for 24 s. The effects in controls (filled circles, n = 5) have been compared with responses after pretreatment with desipramine (DMI; 0.5 mg/kg iv.; open squares, n = 5) or reserpine (1 mg/kg i.v. combined with preganglionic denervation, open circles, n = 5). Data are given as means & S.E.M. Note the different time scales for during (s) and after (min) stimulation.
A
corltml
1%
f.I59Hzinegularatim
Fig. 8. Recordings of perfusion pressure changes in the blood perfused pig spleen upon stimulation with an irregular pattern with an average frequency of 0.59 Hz for 8 min. The effects in controls have been compared with responses 24 h after reserpine treatment (1 mg/kg i.v. combined with preganglionic dener~tion).
Co-release of NPY and NA from pig spleen
Regulation of neuropeptide Y release by biophase concentrations of noradrenaline
zooo-1500. .E
~moo. E
,a
1
o- i_
.._o ,....__.. 0
,..,,,,,.
20
10
.@
30
2J.-...&*.
0
40
50
\
60
70
'51
6
:’ .._, ,:’
6’ j
...
‘..A
/TLk+_
/.*/’
‘. ”
‘0 . . . . .._..
c? 0
-kg>*
10
20
30
40
50
60
. 10
20
30
, 40
50
60
70
125-,
r” 1
loo75. 50.
g
25 04
l
0
Pret~atment with the monoamine uptake inDMI enhanced both functional responses and of NA evoked by nerve stimulation while overflow of NPY-LI at low frequency stimulation was reduced. The latter finding is probably related to prejunctional inhibition of NPY release due to increased local biophase concentrations of NA when the main inactivation pathway of the amine was inhibited by DMI (see Refs 21 and 43). Also nerve stimulation evoked release of another vesicular component DBH has been reported to be inhibited in the presence of DMI.6 Conversely, cr-adrenoceptor blockade has heen found to enhance nerve stimulation evoked output of the DBH (see Ref. 6) and NPY-LI from the spleen.**%” The present and earlieti8,” data therefore suggest that exocytosis of DBH, NPY-LI and NA from large dense-cored vesicles is regulated by the extraneuronal NA concentration at low to moderate stimulation frequencies. At 10 Hz, however, a clear-cut enhancement of NPY output relative to NA was observed in spite of DMI treatment (as revealed by the output ratio). This suggests that the relatively small NPY release at low frequencies may be related to prejunctional inhibitory a-adrenoceptor modulation. The common experimental procedure of using DMI treatment to obtain a better estimate of “true” NA release thus leads to an inhibition of NPY overflow, at least in the pig spleen using the low frequency types of stimulation. The enhanced nerve stimulation evoked NA output after DMI is therefore more likeiy to be related to a higher overtlow of transmitter that escapes local reuptake in the junctional cleft while the actual vesicular release may be depressed by the inhibitory feedback mechanism. Reserpine treatment (combined with p~gan~ioni~ denervation to prevent concomitant depletion of NPY-LI; see Refs 30 and 35) was associated with marked elevation of NPY overflow (see also Ref. 34) and prolongation of the perfusion pressure effect. This observation is also in line with the suggestion that the local biophase concentration of NA regulates NPY release. Furthermore, the difference between NPY overflow using continuous 2.3 Hz and 23b stimulations was smaller in reserpine-treated pigs than in controls. It has also been demonstrated that NPY can inhibit release of NA,3’,35 although it is not known whether this occurs under physiological conditions.
hibitor output
500~
0
,.-.
483
%3* 70
Time (min) B 20 Hz b stimulation
It
Fig. 9. Effects of eiectrical stimulation for 6Omin with regular bursts at 20 Hz for I s every 10 s on perfusion pressure (expressed as perfusion pressure increase, PP, mmHg), as well as output of NPY-LI @mol/min) and NA (pmoi/min) from the blood perfused pig spleen in uivo. Data (means f S.E.M.) represent observations from five control animals pretreated with desipramine (DMI; 0.5 mg/kg i.v.; filled circles) or five animals pretreated with reset-pine (1 mg/kg i.v.) combined with preganglionic denervation (open circles).
dense-cored vesicles may therefore also contribute to the total NA release upon stimulation with high frequency bursts. In man, the circtdating plasma levels of NPY-LI during resting conditions are 10w.~**~* Plasma concentrations of NA are already elevated upon moderate sympathetic activation in man while systemic NPY levels mainly increase upon severe stress.Zq,3R Furthermore, direct recordings of sympathetic nerve activity in man during graded hypotension have shown that systemic plasma NPY levels increase at a higher threshold of sympathetic activation than NA.* In the cat adrenal gland, however, where NPY is probably stored in the same ~hromaffin vesicles as adrenaline, high frequency burst stimulation enhanced both NPY and adrenaline outflow in paralIel.32
Fractional release, res~ppi~ and turnover
The fractional release of [ 3H]NA per nerve impulse was in parallel to the output of endogenous NA for all three types of stimulations. Since NPY is likely to be replaced to nerve endings exclusively by axonal transport at a rate of about 10 mm/h (see Ref. 12), the fractional release of NPY per nerve impulse upon a
long-term stimulation for 1 h was calculated from the recovered NPY-LI in the venous effluent. fnterestingly, a similar fractional release was obtained for both [‘HjNA and NPY-Ll. The release sites for sympathetic transmitter are known to be activated by only a few per cent of the stimuli (see Ref. 7) and therefore the present findings suggest that release of NPY ako occurs in a similar manner. Whether the release of NPY from large dense-cored vesicles is monoquantal or by partial exocytosis remains to be established. As discussed earlier with DBH as a marker (see Ref. 201, it cannot be excluded that the large dense-cored vesicles successively deplete their content of NPY-LI due to lack of local synthesis, and change into the small type of vesicles with maintained capacity to release NA. NPY is a larger molecule than NA and therefore our finding that NPY-LI appears in the venous effluent later and more slowly than NA may reflect different diffusion properties for the two molecules. Thus, in spite of the specialized fenestrated vessels present in the spleen, detection of maximal or total peptide release demands different sampling periods than for NA. The slow decline of plasma NPY-LI after release and the total recovery of released NPYLI in the splenic venous effluent during the I h stimulation compared to the reduction in tissue content suggests that the local degradation of peptide is relatively slow. In accordance, NPY has been demonstrated to have a considerably longer half-life in plasma (about 5 min in the pig” and man”“) compared to NA (l--2 min”‘). A delayed appearance and maintained levels of NPY-LI compared to NA in systemic blood after sympathetic activation have also been observed upon prolonged physical exercise in man.38 The maximal plasma levels of NPY-LI in the splenic venous effluent during stimulation with 20b for I h were in fact close to t nM which is in the same range as when exogenous NPY causes increase in perfusion pressure in the perfused spleen model?” or reduction in splenic blood flow in the intact animal.“” This clearly suggests that NPY could be of functional importance as a transmitter for sympathetic splenic vascular control since the concentration of locally released peptide close to the nerve terminals is likely to be much higher than that recovered in the venous eflluent. Long-term stimulation for I h in control pigs was associated with about 60% depletion of the total splenic content of NPY-Ll while the corresponding NA content was unchanged. This observation is likely to be directly related to differences in synthesis and resupply between these two agents. Thus, axonal
transport of NPY-LI was calculated to bc able to replace the released NPY-LI within a considerably shorter time period than the recovered NA in the splenic venous ehluent. Similar conclusions conccrning the relatively small importance of axonal transport versus local synthesis for resupply of NA after release have been obtained earlier in the spleen. Thus, only about I % of the total NA content of the spleen may be replaced by axonal transport in 24 h (see Ref. 15). The calculated total turnover time for the splenic NPY content considering the resupply by axonal transport was found to be around 1I days. This value may be compared with the dynamics for the vesicular enzyme DBH“ and the peptide VlPZ4 which have been used to derive estimates of about 5 days for the life span of noradrenergic granules and turnover for peptide containing vesicles, respectively. The decline in the output of NPY-LI upon prolonged stimulation under control conditions, and especially after the initial enhancement of release following pretreatment with reserpinc, suggests that limited amaunts of the total splenic content of peptide are available for release. It is thus likely that NPY present in axons represents a non-releasable pool (see also Ref. 30). Furthermore. the maximal depletion of NPY after reserpine treatment due to neurogenic activation is considerably smaller than that of NA.-‘U.3’No evidence for enhanced synthesis of NPY in the cell body region (coeliac ganglion) or increased axonal transport in the splenic nerves was observed 24 h after reserpine in the present expcriments. This apparent discrepancy from data obtained in the guinea-pig”’ may, however, be related both to the reserpine dose and the short survival period in the pig.
CONCLUSION
By comparing NPY and NA in the pig spleen it is possible to study the vesicular turnover and prejunctional regulation of transmitter release, as well as the possible cotransmitter role for a peptide in comparison with a classical transmitter. present studies have been supfrom the Swedish MRC (14X-07164, 14X-6554. 04X-03546-15Bt. The American Council for Tobacco Research Swedish ‘Tobacco Company, Petrus and Augusta Hedlunds Foundation, the Laerdai Foundation, Wiberes Stiftelse. The Swedish Societv of Medicine, Swedish Work ind Environmental Fund and Funds from the Karolinska Institute. For expert technical assistance we thank Miss Anette Hem&n and Miss Margareta Stensdotter and for secretarial help Mrs Karin Wiberg. Ackrtowled~ements-The
ported by‘grants
REFERENCES 1. 2.
Andersson P. O,, Bloom S. R. and Edwards A. V. (1982) Effects of stimulation of the chorda tympani in bursts on submaxillary responses in cat. J. Physiol., Land. 322, 469483. Andersson P. 0. (1983) Comparative vascular effects of stimulation continuously and in bursts of the sympathetic nerves to cat skeletal muscle. Aera physiof. scnnd. 118, 343-348.
Co-&ease
of NPY
and NA From pig spleen
485
3. Bini G., Hagbarth K. E., Hynninen P. and Wallin B. 0. (X980) Thermoregulatory and rhythm generating mechanisms governing the sudomotar and vasoconstrictor outflow in human cutaneous nerves. J. phy~iol., Lond. 306, 537-552. 4. Brimijoin S, (1972) Transport and turnover of dopamine-eta-hydroxyfase (EC 1.14.2.1) in sympathetic nerves of the rat. 3. [email protected], 2f83-2193. 5. Chubb J. W,, DePotter W. P. and DeSchaepdryver A, F. (1970) Evidence For two types of ~orad~nal~ne storage particles in dog spleen. Nature 288, 1203-1204. 6. Cubeddu L. X., Barnes E., Langer S. Z. and Weiner N. (1974) Release of norepinephrine and dopamine-betahydroxylase by nerve stimulation. I. Role of neuronal and extraneuronaf uptake and of alpha-presynaptic receptors. J. Phurmac. exp. Ther. 190, 431450. 7. Cunnane T, C. and StjHme L. (15184)Transmitter secretion from individual varicusities of guinea-pig and mouse vas deferens: highly jn~~ittent and MOnOquantaf. Meursrrience 13, l-20. 8. Delius W., Hagbarth K-E., Hongefl A. and Waflin B. G. (1972) General characteristics of sympathetic activity in human muscle nerves. Acta physiol. stand. 84, 65-81. 9. Eckberg D, L., Rea R., Andersson O., Hedner T., Pernow J., Lundberg J. M. and Wallin B, G. Baroreflex modulation of sympathetic activity and sympathetic neurotransmitters in man. Acta pkysiot. scund. (in press). f 0. Esfer M. (1982) Assessment of sympathetic nerves function in humans f’rom noradrenaline plasma kinetics.
E. and Pernow J. (1984) Guanethidine-sensitive release of neuropeptide Y-like immunoreactivity in the cat spleen by sympathetic nerve stimulation. Neurosci. Lett. 52, 175-180. 29. Lundberg J. M., Martinsson A., Hems&n A., Theodorsson-Norheim E., Svedenhag J., Ekblom E. and Hjemdahl P. (1985) Co-refeasz of neuropeptide Y and catechofamines during physical exercise in man. B&hem. bloph_ys. Res, Cummtm. 133, Z&36. 30. Lundberg J. M., Saria A., France-Cereceda
of reserpine and 6-hydroxydopamine
A., HBkfelt T., Terenius L. and Gofdstein M. (1985) Differential eff?cts on neuropeptide Y (NPY) and noradrenaline in peripheral neurons. Naunyn-
Schmiedeberg’s Arch. Pharmac. 328, 331-340. 3t. Lundberg J. M. and Hakfelt T. (1986) Multiple co-existence of peptides and classsical transmitters in peripheral autonomic and sensory neurons-functionaf and pharmacological implications. In Progress in Brain Research
(eds Hiikfelt T., Fuxe K. and Femow B.), Vol. 6g, pp. 243-262. Efsevier, Amsterdam. 32. Lundberg J. M., Fried G., Pernow J. and ~eodo~so~-Norheim E. (1986) Corelase of neuropeptide Y and catechoiamines upon adtcnaf activation in the cat. Acta physioi. stand. 126, 231-238. 33. Lundberg J. M., Hems&n A., Fried G., Theodorsson-Norheim E. and Lagercrantz H. (1986) High plasma levels of neuropeptide Y (NPY)-like immunoreactivity and catechofamines in newborn infants. Acru physiol. scund. 126, 471473. 34. Lundberg J. M., Rudehilf A., Soffevi A., The~or~o~-Norhejm E. and Hamberger B. (f986f Frequency- and ~rpi~e~~ndent chemical coding of sympathetic transmjas~on: difFerentiaf release of noradrenafine and neuropeptide Y from pig spleen. Neurosci. Lett. 63, 96-100.
486
J. M.
hNDHbXG
PI al.
35. Lundberg J. M., Pernow J., France-Cereceda A. and Rudehill A. (1987) Effects of antihvnertensive drws on sympathetic vascular control in relation to neuropeptide Y (NPY). J. Cardiouasc. Pharm. Sup$. 10, S51~ S58.36. Lundberg J. M., Rudehill A. and Sollevi A. Pharmacological characterization of neuropeptide Y and noradrenaline mechanisms in sympathetic control of pig spleen. Eur. J. Pharmac. (in press). 37. Nilsson H.. Ljung B., Sjoblom N. and Wallin B. G. (1985) The influence of the sympathetic impulse pattern on contractile response of rat mesenteric arteries and veins. Actu physiol. stand. 123, 303-309. _ 38. Pernow J.. Lundberg J. M.. Kaiiser L.. Hiemdahl P., Theodorsson-Norheim E.. Martinsson A. and Pernow B. (1986) Plasma neuropeptidk Y-like immunoreactivity and catecholamines during various degrees of sympathetic activation in man. C&I. Phys. 6, 561-578. 39. Pernow J., Lundberg J. M. and Kaijser L. (1987) Vasoconstrictor effects in uivo and plasma disappearance rate of neuropeptide Y in man. Life Sci. 40, 47754. 40. Rudehill A., Olcen M., Sollevi A., Hamberger B. and Lundberg J. M. (1987) Release of neuropeptide Y upon haemorrhagic hypovolemia in relation to vasoconstrictor effects in the pig. Acta physiol. scund. 131, 517 -523. 41. Rudehill A., Lundberg J. M., Sollevi A. and Hjemdahl P. (1987) Elevations of neuropeptide Y-like immunoreactivity and catecholamines in plasma upon increased intracranial pressure in the pig. Acfa anaesth. stand. 31, 131-I 38. 42. Smith A. D., DePotter W. P., Moerman E. J. and DeSchaepdryver A. F. (1970) Release of dopamine-beta-hydroxylase and chromogranin A upon stimulation of the splenic nerve. Tiss. Cell 2, 547-55 1. 43. Starke K. (1987) Presynaptic a-autoreceptors, Rev. Physioi. Biochem. Pharmac. 107, 74-146. 44. Theodorsson-Norheim E., Hemsen A. and Lundberg J. M. (1985) Radioimmunoassay for neuropeptide Y (NPY): chromatographic characterization of immunoreactivity in plasma and tissue extracts. Stand. J. Lab. Invest. 45,355-365. 45. Wallin B. G. (1981) New aspects of sympathetic function in man. In Butterworths Infernafional Medical Reviews Neurology I (eds Stalberg E. and Young R. R.), pp. 145-167. Butterworth, London. (Accepted
I I July 1988)