| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Medicine, University of Arizona Medical College, Tucson, Arizona 85724-5099
Address all correspondence and requests for reprints to: Dr. Seymour Reichlin, Department of Medicine, University of Arizona Medical College, Tucson, Arizona 85724-5099.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-adrenergic receptors
(phentolamine) did not prevent the IL-6 response. That the peripheral
response was due to passage of the injected IL-1ß into blood from the
brain was supported by several observations. Immunoreactive IL-1ß
appeared in peripheral blood by 10 min after icv injection and remained
constant between 10–100 min after injection; values after icv
injection were virtually identical to those after iv injection at 60
and 80 min. Radioiodine-labeled IL-1ß appeared in blood as early as 5
min, and by phamacokinetic analysis was found to be transferred from
the brain at a rate greater than 2% of brain content per
min-1. IL-1ß infused iv in a pattern mimicking brain to
blood transfer induced IL-6 levels that were more than double the
values induced by a single bolus injection and were not significantly
different from the values observed after icv injection. Sustained
levels of IL-1ß in blood over time contribute to the high peripheral
IL-6 response. This was shown by administering the same total dose iv
as a single bolus of 100 ng or in two doses of 50 ng 1 h apart.
Rats given a divided dose had 6–10 times higher blood IL-6 levels at
2 h than those given a single injection. The high levels of IL-6
in blood after icv injection of IL-1ß are best explained by the
reservoir function of the brain IL-1ß pool and the self-priming
effect of IL-1ß in peripheral tissues. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(TNF) response to icv injection of
Escherichia coli endotoxin [lipopolysaccharide (LPS)], in
some ways an analogous stimulus to cytokine release, was blocked by
pretreatment with phentolamine, a nonselective -adrenergic blocker
(16). However, other findings suggest that the release of IL-6 into peripheral blood after central administration of IL-1 is not mediated by peripheral catecholamines. In a previous study from this laboratory chlorisondamine did not inhibit the peripheral IL-6 elevation induced by icv IL-1, and icv injections of CRF in doses known to activate the sympathetic nervous system did not induce elevation of peripheral IL-6 levels (3). Chemical sympathectomy by peripheral injections of 6-hydroxydopamine did not block the blood IL-6 response to icv injection of LPS (17); indeed, sympathetic nerve denervation appeared to enhance the LPS-induced response.
Alternative mechanisms to catecholamine mediation have been considered. The brain was shown to be a possible source of the elevated peripheral IL-6 levels observed after icv injection of IL-1ß by the finding that blood in the superior sagittal sinus (the main route of venous drainage from the brain) had higher concentrations of bioactive IL-6 than aortic blood (3). In addition, pharmacokinetic studies using radioiodinated IL-6 demonstrated that more than a third of the dose of IL-6 injected icv entered the blood from the brain by the mechanism of bulk flow (18). However, brain secretion of induced IL-6 cannot alone explain the early peripheral IL-6 response, because the peripheral changes are maximal when the central response is just detectable, and the peripheral response has virtually ended at a time when the central response is still increasing (5, 6).
A third hypothesis to explain the early acute rise in peripheral IL-6 after centrally administered IL-1ß is that the injected cytokine itself enters circulating blood and induces an early activation of IL-6 in peripheral tissues. IL-6 is synthesized and released into the blood after IL-1 injection by many tissues, including spleen, liver, lymphocytes, monocytes, and vascular endothelia. This idea was initially considered and then discarded by DiSimoni and colleagues when they showed that the magnitude of the peripheral IL-6 response after icv injection of IL-1ß was much greater than the response to iv injection (1), a finding that has been reproduced in mice (19) and the rhesus monkey (7). Furthermore, the response in mice to icv injection of IL-1ß was greater than the response to the same dose given ip (4). On the other hand, more recent work by Di Santoa and colleagues suggests that the intraventricular route may be more effective than the iv route in delivering IL-1ß into the circulation, because they found that blood levels of immunoreactive hIL-1ß after icv injection in mice are higher than levels achieved after iv injection (20).
In an attempt to resolve the continuing uncertainty about the roles of
the sympathetic nervous system and of the brain to blood transfer of
IL-1ß in bringing about peripheral IL-6 activation after icv IL-1ß
administration in the rat, we have repeated our previous studies of
chlorisondamine using an improved method for icv injection that
minimizes leakage of IL-1ß from the needle tract and have extended
the range of adrenergic blockers to determine the effects on the
peripheral IL-6 response to IL-1ß of sympathetic blockade by
propanalol, a nonselective ß-adrenergic blocker; butoxamine, a
selective ß2-blocker; and phentolamine, a nonselective
-adrenergic blocker.
To elucidate the role of brain to blood transfer of icv injected IL-1ß in inducing the peripheral IL-6 response, the time course of appearance of immunoreactive human IL-1ß in blood after icv and iv injection was determined, and a pharmacokinetic analysis of brain to blood passage of radioiodinated hIL-1ß was made. Having determined the rate at which IL-1ß leaves the brain after icv injection, peripheral levels of IL-6 after icv injection of IL-1ß were then compared with levels observed after an iv infusion of IL-1ß administered so as to duplicate the rate at which IL-1ß leaves the brain after icv injection. This study showed that prolonged exposure to IL-1ß induced much higher peripheral IL-6 levels than the same dose administered as a single bolus injection. To test the hypothesis that the greater effectiveness of prolonged infusion of IL-1ß compared with bolus injection was due to sensitization by the initial exposure to the cytokine, the IL-6 response to a single bolus iv injection was compared with the response to the same amount of IL-1ß administered in two doses 1 h apart. All protocols were approved by institutional animal care committee.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Agents used to block sympathetic discharge or to block peripheral
sympathetic receptors and their dosages were chosen on the basis of
previously reported studies in rats, which are cited below.
Chlorisondamine was injected sc immediately before or 1 h before
icv injection of IL-1ß or saline (as shown in Results).
This agent (a gift from Ciba-Geigy, Summitt, NJ; now
Novartis), which blocks sympathetic ganglion synaptic
neurotransmission, was administered in a normal saline solution
containing 3.0 mg/ml (dose, 3.0 or 5.0 mg/kg BW). Phentolamine
(Sigma Chemical Co., St. Louis, MO), a nonspecific
-adrenergic blocker, was administered ip in a saline solution (2.5
mg/ml, 5.0/kg BW) 30 min before icv injection of IL-1ß. The
nonselective ß-blocker propanalol (Sigma Chemical Co.;
saline solution containing 10 mg/ml; dose, 10 mg/kg BW) and the
selective ß2 blocker butoxamine (Sigma Chemical Co.; saline solution containing 12.5 mg/ml; dose, 25 mg/kg BW)
were administered immediately before icv injection of IL-1-ß. In one
experiment (indicated below) icv injections of IL-1ß were given while
the animals were anesthetized with ketamine-xylazine (5 and 1 mg/100 g
BW, respectively), whereas in all of the others, injections were made
under pentobarbital anesthesia.
After 2 or 4 h, as indicated in Results, groups of rats were reanesthetized with pentobarbital, and samples were taken from the cysterna magna, superior sagittal sinus, inferior vena cava, and aorta as previously described. Samples of blood were put in heparinized tubes, kept in crushed ice until centrifugation, and then stored at -80 C until assayed for IL-6 by the Aarden B9 assay (22) as modified (3, 5). Samples were assayed in duplicate at eight dilutions; log dose regressions were calculated for each individual sample from the linear portion of the log dose-response curve and were converted to absolute concentrations of recombinant IL-6 standard (Genzyme Corp., Cambridge, MA) by comparison with the straight line portion of a standard curve made up in 11 dilutions.
Transfer of radioiodinated hIL-1ß from the blood after icv and iv
injection
Clearance of radioiodinated hIL-1ß from the blood and brain
was measured as previously reported for studies of clearance of IL-6
(18) and TNF (21). It should be emphasized that outer guide cannulas
were placed at least 5 days before the experiment, which was carried
out under pentobarbital anesthesia. Injections were made under
stereotaxic control using previously determined coordinates that
permitted free flow into the lateral ventricle. The radiolabeled
hIL-1ß (NEX 232, NEN Life Science Products,
Boston, MA) was injected in a volume of 10 µl over a 2-min period,
and the amount of isotope injected was determined by counting the
loaded inner cannula before injection and the inner and outer cannula
together after injection. The connecting tubing was compression sealed
immediately after injection to prevent back flow from the brain.
Intravenous injections were made into the left external jugular vein
through a previously inserted polyethylene cannula, and samples were
removed from the right atrium through an indwelling external jugular
catheter. The catheter was heparinized to avoid clotting, and 0.2-ml
volumes of blood were removed at 5-, 10-, 20-, and 20-min intervals up
to 120 min. Blood volume was replaced with 5% BSA in saline after each
sample of blood had been removed, and animals were kept on a hot pad
throughout the procedure. Intravenous injections were made of the same
amount of radioactive material dissolved in 0.5 ml 5% BSA in normal
saline. The blood samples were precipitated by equal volumes of cold
10% trichloroacetic acid (to separate protein-bound from free iodine),
and total and acid-precipitable counts were determined in a
scintillation well counter.
Estimates of the amount of tracer that entered the blood after icv injection were made by measuring the area under the curve (AUC) of radioactivity by the trapezoidal method and comparing it with the AUC after iv injection. Estimates of apparent pool size and fractional turnover of the IL-1ß from blood after iv injection were made as in previous studies (18, 21) by fitting a two-compartment model to the blood clearance data using the WINNONLIN Nonlinear Estimation Program VO 1.1, Core Version (Scientific Consulting, Apex, NC). Estimates of the rate of transfer of tracer from brain to blood in individual rats were made using an empirically derived best-fit model using SIMUSOLV EXECUTIVE version 3.0 software (Dow Corning Corp., Midland, MI). Blood radioactivity appearance curves were consistent with a simple first order absorption model.
Transfer of immunoreactive hIL-1ß from brain to blood
Intraventricular injection and serial collections of blood were
carried out as described above for brain to blood clearance of
radioiodinated hIL-1ß, except that the injected material was hIL-1ß
(100 ng in 10 µl), and measurements of immunoreactive hIL-1ß were
made using an enzyme-linked immunosorbent sandwich assay set up in this
laboratory using mouse monoclonal anti-hIL-1ß antibody (catalogue no.
MAB 601, R & D Laboratories, Minneapolis, MN) as the capture antibody
and rabbit anti-hIL-1ß (Sigma catalogue no. I-4893) as
the detecting antibody. Peroxidase-labeled antirabbit goat IgG
(Sigma A4914) was used to quantitate the concentration of
detecting antibody. The minimum detectable concentration of hIL-1ß
was 156 pg/ml. All samples were run in the same assay to avoid
between-assay variability.
Programmed infusion of IL-1ß
The rate of transfer of brain to blood of icv injected
radioiodinated IL-1ß was determined to be 0.02 of the brain
content/min (see Results). This parameter was used to
program a Genie infusion pump (Kent Scientific Corp., Litchfield, CT)
to deliver the cytokine at a geometrically decreasing rate
corresponding to the falling residual brain content. Following the
manufacturer’s instructions, this was accomplished by setting the
delivery rate to decrease every 3 min in 40 steps over the 120-min
experimental period.
Priming effect of IL-1ß
Intraatrial cannulas were placed in three groups of rats under
pentobarbital anesthesia. One group was given 100 ng IL-1ß (in 0.1 ml
saline/5% BSA diluent) as a bolus injection, one group was given 50 ng
at time zero and 50 ng at 1 h (in 0.1 ml diluent), and a third
group was given diluent, 0.1 ml at time zero and 0.1 ml at 1 h. At
2 h after injection, aortic and venacaval blood was removed and
prepared for IL-6 assay as in other experiments.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
)
|
The chlorisondamine blocking experiment was repeated a third time. A
dose of 5 mg/kg BW was administered 60 min before the icv injection of
hIL-1ß (Table 1). Mean venous blood
levels of IL-6 were virtually identical when chlorisondamine had been
given and with IL-1ß alone (3,554 ± 358 vs.
3,580 ± 890 pg/ml); mean aortic blood levels were lower in
chlorisondamine-treated animals (3,550 ± 900 vs.
2,344 ± 488 pg/ml), but the difference did not reach the level of
statistical significance (P = 0.18). As in previous
experiments, IL-6 levels in CSF were markedly increased by IL-1ß
injection (76 ± 10 vs. 22,900 ± 600 pg/ml).
|
-adrenergic blocker. Neither propanalol nor butoxamine
significantly modified aortic or superior sagittal sinus blood levels
of IL-6 (Fig. 2). Mean levels of IL-6 in
aortic blood after phentolamine injection were approximately twice
those in control animals given IL-1ß icv, but the differences were
not significant by statistical test, probably due to the small number
of animals in the phentolamine group. Phentolamine was reevaluated in a
larger group of rats under slightly different conditions; the drug was
administered ip 30 min before injection of the intraventricular
cytokine, and the animals were anesthetized with ketamine-xylazine
instead of pentobarbital for the injection (Fig. 3). As before, all animals receiving
IL-1ß icv had significantly elevated levels of IL-6 in both aortic
and SSS blood, and the SSS values were significantly and consistently
higher than peripheral blood levels in all groups. Mean values for
aortic IL-6 in animals treated with phentolamine were almost double
those in animals treated with saline ip, and the difference was
significant by statistical test (P = 0.04).
|
|
)
|
). Pharmacokinetic analysis (Table 2) showed that over the 4-h period,
86 ± 8.0% of the injected dose had entered the blood from the
brain. Brain to blood passage was rapid, being more than 0.02 of the
total content leaving the brain per min-1, a value that
was approximately twice previously measured rates of brain to blood
transfer of labeled IL-6 (18) and TNF (21).
|
|
)
|
). At 1 h, rats receiving 50 ng
IL-1ß also showed a significant increase in IL-6 in atrial blood, but
the response to 100 ng was significantly greater (1,442 ± 157
vs. 3,033 ± 614 pg/ml; P = 0.02). Most
striking was the 6- to 10-fold difference at 2 h in both venous
and arterial blood in rats that had received 100 ng IL-1ß in two
divided doses instead of a single bolus injection of the same amount:
atrial blood, 35,494 ± 16,581 vs. 3,528 ± 543
pg/ml (P = 0.0008); aortic blood, 40,325 ± 10,196
vs. 7,254 ± 1,368 pg/ml (P = 0.0008;
by Mann-Whitney test).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study treatment with a blocker of preganglionic cholinergic
neurotransmission, chlorisondamine, and with ß-adrenergic antagonists
propanalol (a selective ß-adrenergic antagonist), butoxamine (a
selective ß2-adrenergic antagonist), and phentolamine (a
nonselective -adrenergic antagonist) tested individually did not
block peripheral IL-6 responses to centrally administered IL-1ß.
Treatment with phentolamine, a nonselective -adrenergic antagonist
in fact appeared to potentiate the IL-6-elevating effects of icv
injected IL-1ß. These findings are compatible with our earlier
observation that chlorisondamine did not block peripheral IL-6
responses to central IL-1ß (3) and with the report of De Luigi
et al., who found that chemical sympathectomy with
-methyl-dopa did not block the IL-6 response to central IL-1
injection (17). They are at variance with the report of Kitamura
et al. (4), who found that phentolamine blocked the
elevation of plasma and tissue IL-6 that followed icv injection of
IL-1ß. Our findings are also at variance with the somewhat comparable
study of Finck et al. (16), who reported that phenotolamine
blocked the elevation in peripheral IL-6 that followed icv injection of
LPS.
Of first concern before one can conclude that the early IL-6 response is not due to sympathetic nervous system activation is whether the doses of drugs used, their route of administration, and the timing of the sequence of adrenergic blockade and icv injection of IL-1ß were adequate. With respect to chlorisondamine, two dosage levels were administered (3.5 and 5 mg/kg BW), and two time intervals were chosen for the larger dose, either 1 h or immediately before icv injection of IL-1ß. This dose is to be compared with earlier studies in rats in which peripheral catecholamine levels or catecholamine-dependent responses were followed as an indicator of chlorisondamine action. These include the demonstration that 3 mg/kg blocked the rise in serum epinephrine and norepinephrine after icv injection of CRF (24), the suppressive effects of ventromedial nuclear electrical stimulation on suppression of splenic proliferation (25), the suppression of natural killer cell activity observed after icv injection of IL-1ß in rats (26, 27), and the serum and tissue IL-6 response to icv injection of IL-1ß (4).
Prior studies of phentolamine have reported that 5 mg/kg blocked hypertriglyceridemia induced by endotoxin (28), the IL-6 response to icv LPS (16), and the inhibition of gastric acid secretion induced by icv injection of IL-1 (29).
Prior studies of propanalol have shown that 7.5 mg/kg blocks the peripheral IL-6 response to open field stress (8), 5 mg/kg blocked inhibition of gastric activity induced by icv injection of IL-1 (28), and 10 mg/kg blocked CRF-induced suppression of natural killer cell activity (30).
Prior studies of butoxamine indicated that 25 mg/kg blocked natural killer cell suppression after icv injection of CRF (30).
Sympathetic activation of IL-6 production by the liver can be invoked as the likely explanation for the elevation of IL-6 that follows physical (11) and emotional (8, 9, 10) stress, central administration of PG (11), and peripheral administration of catecholamines (11, 12) in rats. On the other hand, in a previous study from this laboratory, icv injection of CRF in amounts known to stimulate peripheral sympathetic nervous system activity did not raise blood IL-6 levels (3). It is possible that stress-induced changes are smaller than those induced by relatively large amounts of circulating proinflammatory substances and would be obscured in a test system in which potent cytokines or toxins entered the peripheral blood from the brain.
Our findings and those of De Luigi et al. (17) indicate that sympathetic outflow is not the pathway by which a central inflammatory stimulus activates peripheral IL-6 secretion, but the studies of Kitamura et al. (4) (who studied the effects of IL-1ß) and of Finck et al. (16) (who studied effects of LPS) provide evidence to the contrary. There is not an obvious explanation for these contradictory findings.
The second hypothesis tested in this study was that sufficient amounts of injected IL-1ß were transported at a sufficiently early time interval to induce peripheral IL-6 production. This hypothesis would explain the failure of sympathetic ganglionic blockade in our studies and those of De Luigi et al. (17) to prevent peripheral IL-6 responses after central IL-1 administration and is compatible with reports of the relatively early appearance of IL-6 in peripheral blood and the early expression of IL-6 mRNA in the spleen compared with brain induction of IL-6 expression after LPS injection (23).
That macromolecules leave the brain by bulk transport after icv
injection has been amply shown by a number of workers (31, 32), and this
phenomenon has been extended to brain-blood transport of cytokines and
LPS. Radioiodinated TNF appears in peripheral blood after icv
injection (21, 33), is first detected at 5 min, and reaches peak levels
by 2 h, with 70% having appeared within 4 h (21). Similar
brain to blood clearance has been documented for passage of IL-6 from
brain to blood (18), for radioiodinated LPS (6), and for leptin, a
molecule whose size is in the same range as IL-6, IL-1ß, and TNF
(34).
Passage from brain to blood of IL-1ß appeared to be higher than
clearance rates of IL-6 (18) and TNF (21). As the transfer rates of
IL-6 and TNF are the same as the rates of exit of inert molecules
such as dextrans and albumin, thus corresponding to the rate of bulk
flow, it is likely that IL-1ß is secreted from the brain by an active
process. Indeed, Banks and collaborators have shown that IL-1ß is
transported out of the mouse brain by an active, saturable process that
is more rapid than bulk flow (35).
The phenomenon of brain to blood transfer of IL-1ß, its rapidity, and the relatively large proportion of injected material that finds its way into the blood within the 2-h period of study (86 ± 8%) are all compatible with the hypothesis that peripheral IL-6 responses are due to the IL-1ß that has entered peripheral blood.
The pattern of blood levels of IL-1ß after icv injection and the pattern of peripheral response to IL-1ß have an important quantitative effect on IL-6 levels. Intravenous infusion of IL-1ß at a rate approximating that of entry from brain to blood induced much higher levels of IL-6 than when given by bolus injection and in the same range as found after icv injection. Although mean values after programmed iv infusion are higher than those after icv injection, the differences did not reach statistical significance.
The most likely explanation for the finding that prolonged elevation of IL-1ß is more effective in stimulating IL-6 than a single bolus iv injection of the same dose of cytokine is that the initial phase of IL-1ß primes the IL-6 response to IL-1ß present later in the course of infusion. This was shown by administering IL-1ß in two iv bolus doses of 50 ng each 1 h apart instead of as a single large bolus injection. The IL-6 response in venous and arterial blood was 6–10 times greater in rats injected with two doses rather than one dose. In previous studies comparing peripheral IL-6 responses to icv injection with responses to iv injection, the iv injections were not administered in a pattern that sensitizes IL-6 responses as does icv injection.
IL-1ß sensitizes peripheral tissues to later exposure to IL-1ß by several possible mechanisms. Expression of IL-1 receptor mRNA is rapidly induced by IL-1 in a number of tissues and cell lines, including endometrial stroma (36), fibroblasts (37), insulinoma Rinm5F cells (38), pituitary AtT-20 cells (39), and hepatocytes (40). Furthermore, IL-1 is capable of inducing IL-1 secretion in several peripheral tissues (41, 42); the IL-1 so induced may stimulate IL-6 secretion by binding to the up-regulated IL-1 receptor.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received June 14, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-adrenoreceptor-mediated mechanism. Am J
Physiol 272:R1880–R1887
from the brain into the blood following
intracerebroventricular injection in rats. Neuroimmunomodulation 5:221–276[CrossRef][Medline]
-Adrenergic receptors mediate the hypertriglyceridemia induced by
endotoxin, but not tumor necrosis factor, in rats. Endocrinology 135:2644–2650[Abstract]
across the blood-brain barrier. Brain Res
Bull 23:433–437[CrossRef][Medline]
stimulate the mRNA expression of
interleukin-1 receptors in mouse anterior pituitary AtT-20 cells.
Neurosci Lett 187:53–56[CrossRef][Medline]
. J Immunol 140:4238–4244[Abstract]
suppresses IL-1 but not
lipopolysaccharide-induced transcription of IL-1. J Immunol 144:2216–2222[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  M. J. Kenney, F. Blecha, R. J. Fels, and D. A. Morgan Altered frequency responses of sympathetic nerve discharge bursts after IL-1beta and mild hypothermia J Appl Physiol, July 1, 2002; 93(1): 280 - 288. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Reichlin, G. Chen, and M. Nicolson Blood to Brain Transfer of Leptin in Normal and Interleukin-1{beta}-Treated Male Rats Endocrinology, June 1, 2000; 141(6): 1951 - 1954. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
