www.elsevier.com / locate / bres
Short communication
Differential transport of rat and human interleukin-1
a
across the
blood–brain barrier and blood–testis barrier in rats
a ,
*
b c dScott R. Plotkin
, William A. Banks , Lawrence M. Maness , Abba J. Kastin
a
Partners Neurology Program, Harvard Medical School, Boston, MA 02115, USA
b
GRECC, Veterans Affairs Medical Center-St. Louis and St. Louis University School of Medicine, Division of Geriatrics, Department of Internal Medicine, St. Louis, MO 63106, USA
c
Division of Neurovascular Biology, Center for Aging and Developmental Biology, University of Rochester Medical Center, Rochester, NY 14642, USA
d
Veterans Affairs Medical Center and Tulane University School of Medicine, New Orleans, LA 70112-1262, USA
Accepted 25 July 2000
Abstract
Human interleukin-1a is transported across the murine blood–brain barrier (BBB) and blood–testis barrier (BTB) by a saturable transport system. Differences in the biological activity and binding of human IL-1 in mouse and rat brain raise the possibility of species
125
differences in the transport of IL-1 across the BBB and BTB. We measured the transport of recombinant human I-IL-1a(I-huIL-1a)
125
and rat I-IL-1a (I-ratIL-1a) across the rat BBB and BTB after intravenous injection using a sensitive in vivo technique and film autoradiography. I-ratIL-1a was found to cross the rat BBB and rat BTB at rates comparable to those reported previously for murine IL-1a in mice. Passage across the BBB was inhibited by the addition of unlabeled rat IL-1a, demonstrating saturable transport. In contrast, I-huIL-1aentered the brain of the rat much more slowly, and its entry was not inhibited by the addition of unlabeled human IL-1a. These results show that the rat interleukin-1 transporter, unlike the murine transporter, does not transport human IL-1a. This difference highlights the importance of species specificity in IL-1atransport and may partly explain the different physiological responses to exogenous human IL-1a among rodent species. 2000 Elsevier Science B.V. All rights reserved.
Theme: Cellular and molecular biology
Topic: Blood–brain barrier
Keywords: Inflammation; Aluminum; Transport; Peptide
1. Introduction and IL-1 receptor antagonist cross the murine BBB through shared transport systems that are inhibited by The peripheral administration of interleukin-1 (IL-1) aluminum and by excess amounts of unlabeled IL-1s produces many central effects including initiation of fever, [3,5,10]. The murine blood–testis barrier (BTB) is also induction of slow-wave sleep, and activation of the hypo- permeable to human IL-1a [2].
thalamic–pituitary–adrenal axis [7]. One mechanism by The binding of IL-1 to brain and endocrine tissues varies which blood-borne IL-1 can affect the central nervous across species. In mice, recombinant human IL-1 binds system (CNS) is by direct passage across the blood–brain with high affinity to the hippocampus and choroid plexus barrier (BBB). Murine IL-1a, murine IL-1b, human IL-1a, in brain, and to the epididymis and interstitial areas in testis [1,18,20]. In rats, however, results are inconclusive. Some studies demonstrate binding of human IL-1 to rat *Corresponding author. Veterans Affairs Medical Center, Research
hippocampus, cerebellum, and hypothalamus [8,12], (151), 915 N. Grand Street, St. Louis, MO 63106, USA. Fax:1
1-114-whereas other studies fail to demonstrate binding in the rat 289-6374.
E-mail address: [email protected] (S.R. Plotkin). brain [19,21]. The species specificity of IL-1 binding in
brain raises the possibility that rat and human IL-1a are al influx constant (Ki ) and distribution volume (Vi ) can be transported differently across the rat BBB. Therefore, the derived from the equation:
objectives of this study were to compare the entry of
Am5Ki(expt)1Vi
human IL-1a and rat IL-1a across the BBB and BTB in
rats. where Am is the tissue / blood ratio (ml / g) and exposure time (expt) is calculated using the equation:
2. Materials and methods expt5
E
Cp(t) dt/Cpt2.1. Labeling of cytokines where t is time and Cpt is concentration in serum at time t.
Recombinant rat IL-1a and recombinant human IL-1a 2.5. Film autoradiography
125
were labeled with I (I-ratIL-1a and I-huIL-1a, respec-tively) by the enzymobead method (Biorad, Richmond,
As previously described [14], rats were injected
in-99m
CA). Human serum albumin was labeled with Tc (T- 7
travenously with about 4310 cpm of I-huIL-1a with or Alb) with a kit from Amersham Health Care (San Antonio,
without 10 mg / kg of unlabeled human IL-1a. After 30 TX). The labeled cytokines were separated from free
min, the brains were perfused with 20 ml of ice-cold iodine on a column of Sephadex G-10, as previously
lactated Ringer’s solution to clear the vascular space. Rats described [5].
were decapitated and the brains removed carefully to avoid damage. The intact brains were flash frozen and sliced 2.2. Intravenous (iv) injections coronally into 20 mm sections. The sections were de-siccated for 24 h at 48C and placed on Kodak film for 24 Male Sprague–Dawley rats (Harlan–Sprague Dawley, weeks at room temperature. The film was developed with Indianapolis, IN) weighing 250–350 g were anesthetized D-19 developer and analyzed using the MCID image with an intramuscular injection of ketamine (80 mg / ml) analyzer (Imaging Research, St Catharines, Ontario, and xylazine (8 mg / ml) prior to surgery, and the left Canada).
jugular vein and the right carotid artery exposed. At time
7 7
0, 1310 cpm of iodinated IL-1a and 1310 cpm of 2.6. Statistics T-Alb were injected into the jugular vein. At time-points
between 2 and 30 min later, blood was collected from the Linear regressions of tissue / blood ratios vs. expt were carotid artery, and the rat was decapitated. Whole brain, performed and compared by the Prism 2.0 program excluding pineal and pituitary, and testes were collected. (GraphPad Software, Inc., San Diego, CA). Values for Ki All tissues were rinsed in saline to remove excess blood and Vi are reported with their standard error of the estimate and weighed. Samples of arterial blood were centrifuged at and with their correlation coefficient (r). Statistical signifi-45003g for 10 min and 50 ml of serum removed for cance was set at P,0.05.
analysis. All tissues were counted in a gamma-counter.
2.3. Saturation and inhibition studies 3. Results
Saturation studies were performed as described above 3.1. Entry rate, saturation, and competition for I-ratIL-except that 10 mg / kg of unlabeled rat IL-1a and human 1a and I-huIL-1a into rat brain
IL-1a were included in the injections of I-ratIL-1a and
I-huIL-1a, respectively. Inhibition studies involved the The clearance of I-ratIL-1a from the serum was in-pretreatment of rats with 20 mg elemental aluminum hibited by the addition of unlabeled rat IL-1a. The relation (AlCl3?6H 0, 10 mg / ml) administered into the peritoneum2 between the brain / blood ratio and expt was statistically
at least 30 min prior to anesthesia. significant in rats injected with I-ratIL-1aand with I-ratIL-1a plus unlabeled rat IL-1a (Table 1A). Entry of I-ratIL-2.4. Calculation of unidirectional transport rate 1ainto brain was inhibited by the addition of unlabeled rat
IL-1a (F1,8518.05, P,0.005, n512).
Table 1
Comparison of unidirectional influx rate (Ki in units of ml / g / min) and the volume of distribution (Vi in units ml / g) for IL-1ainto brain and testes of rats
Compound Ki Vi R P N *Significantly decreased compared with entry of I-ratIL-1aalone into brain (P,0.005). ** Significantly decreased compared with entry of I-ratIL-1ainto testes (P,0.001). ***Significantly decreased compared with entry of I-huIL-1ainto testes (P,0.001).
into brain was not inhibited by the addition of unlabeled groups were combined to calculate an aggregate Ki (Table human IL-1a or by the addition of elemental aluminum. 1A and B, respectively).
3.2. Entry rate, saturation, and competition for I-ratIL- 3.4. Film autoradiography 1a and I-huIL-1a into rat testis
Systemic injection of I-huIL-1a and I-huIL-1a plus The relation between the testis / blood ratio and expt was unlabeled human IL-1a resulted in negligible labeling statistically significant in rats injected with I-ratIL-1a and throughout the brain, including the cortex, choroid plexus, with I-ratIL-1a plus unlabeled rat IL-1a (Table 1A). The or deeper nuclei.
entry of I-ratIL-1a into testis was significantly faster than the entry of T-Alb (F1,10558.21, P,0.001, n514). The
entry of I-ratIL-1a into testis was not inhibited by the 4. Discussion
addition of unlabeled rat IL-1a.
The relation between the testis / blood ratio and expt was The widespread availability of recombinant mouse, rat, statistically significant in rats injected with I-huIL-1a and and human IL-1a has led to the use of these reagents in with I-huIL-1a plus unlabeled human IL-1a (Table 1B). laboratory studies without consideration of possible species The entry of I-huIL-1a into testis was significantly faster differences. For example, a controversial issue has been than the entry of T-Alb (F1,15556.79, P,0.001, n519). whether any of the behavioral effects of IL-1 seen after its The entry of I-huIL-1ainto testis was not inhibited by the systemic administration are due to its transport across the addition of unlabeled human IL-1a. BBB. Many of the studies examining human IL-1a have used a murine model to examine transport phenomena and 3.3. Entry rate of Tc-Alb into rat brain and testis a rat model to examine behavioral effects.
In this study, both I-ratIL-1a and I-huIL-1a accumu-In brain, the relation between the brain / blood ratio and lated in rat brain more rapidly than the vascular marker expt was not statistically significant for T-Alb. Therefore, albumin. The rates of entry approached values seen in the the results were averaged to give a mean brain / blood ratio transport of other cytokines (Table 2). This phenomenon
23
Table 2
Comparison of the rates of entry (Ki in units of ml / g / min) for various cytokines across the BBB and BTB
24 24
Species Cytokine Brain (Ki310 ) Saturable Testis (Ki310 ) Saturable Reference
Mouse Murine IL-1a 13.5 Yes – – [5]
Mouse Murine IL-1b 4.73 Yes – – [5]
Mouse Human IL-1a 2.48 Yes 18.1 Yes [2,5]
Mouse Human IL-1ra 5.19 Yes – – [10]
Rat Rat IL-1a 6.75 Yes 19.9 No Present study
Rat Human IL-1a 1.42 No 25.1 No Present study
The absence of labeling by I-huIL-1a as determined by demonstrates the independence of blood–tissue barriers within a single species.
autoradiography confirmed that little, if any, I-huIL-1a crossed the BBB to reach the parenchyma. These au-toradiograpic findings were in marked contrast to a
previ-References
ous study in which systemic I-huIL-1a localized to the cortex, choroid plexus, and the posterior division of the
[1] E. Ban, G. Milon, N. Prudhomme, G. Fillion, F. Haour, Receptors septum in mouse brain [14]. Thus, the accumulation of
for interleukin-1 (aandb) in mouse brain: mapping and neuronal human IL-1a in rat brain likely represented non-specific
localization in hippocampus, Neuroscience 43 (1991) 21–30. binding to the vascular endothelium. [2] W.A. Banks, A.J. Kastin, Human interleukin-1a crosses the blood–
In testis, the entry rates of I-ratIL-1a and I-huIL-1a testis barriers of the mouse, J. Androl. 13 (1992) 254–259. were comparable to the entry rate previously reported for [3] W.A. Banks, A.J. Kastin, D.A. Durham, Bidirectional transport of
interleukin-1 alpha across the blood–brain barrier, Brain Res. Bull. human IL-1a into the murine testis (Table 2). Neither
23 (1989) 433–437. I-ratIL-1anor I-huIL-1awere transported saturably across
[4] W.A. Banks, R.N. McLay, A.J. Kastin, U. Sarmiento, S. Scully, the BTB, as unlabeled material did not inhibit their entry. Passage of leptin across th blood–testis barrier, Am. J. Physiol. 276 These results indicate that, unlike transport of IL-1aacross (1999) E1099–E1104.
[5] W.A. Banks, L. Ortiz, S.R. Plotkin, A.J. Kastin, Human interleukin the BTB in mice [2], IL-1a is not actively transported
(IL) 1a, murine IL-1aand murine IL-1bare transported from blood across the BTB in rats. A similar phenomenon has been
to brain in the mouse by a shared saturable mechanism, J. Phar-observed for leptin previously [4]. macol. Exp. Ther. 259 (1991) 988–996.
Results from many fields suggest that the rodent and [6] R.G. Blasberg, J.D. Fenstermacher, C.S. Patlak, Transport of a -human forms of interleukin-1 are not equivalent. Based on aminoisobutyric acid across brain capillary and cellular membranes,
J. Cerebr. Blood Flow Metab. 3 (1983) 8–32. amino acid sequence, the homology of rat IL-1ato rabbit,
[7] C.A. Dinarello, Biology of interleukin 1, FASEB J. 2 (1988) human, and murine IL-1a is 64%, 65%, and 83%,
respec-108–115.
tively [15]. Phylogenetic trees of amino acid sequence [8] W.L. Farrar, P.L. Kilian, M.R. Ruff, J.M. Hill, C.B. Pert, Visualiza-show that rat and murine IL-1 diverged early from human tion and characterization of interleukin 1 receptors in brain, J. IL-1 in the course of evolution, whereas rabbit, pig, and Immunol. 139 (1987) 459–463.
[9] E.G. Gutierrez, W.A. Banks, A.J. Kastin, Murine tumor necrosis bovine IL-1 diverged more recently [11].
factor alpha is transported from blood to brain in the mouse, J. In physiological studies, rat IL-1b selectively binds the
Neuroimmunol. 47 (1993) 169–176.
rat IL-1 receptor in vitro but not the monkey cell line IL-1 [10] E.G. Gutierrez, W.A. Banks, A.J. Kastin, Blood-borne interleukin-1 receptor [13] and is more active than human IL-1bin a rat receptor antagonist crosses the blood–brain barrier, J. Neuroim-culture model [13]. In vivo, murine TNF-a, but not human munol. 55 (1994) 153–160.
[11] A.L. Hughes, Evolution of the interleukin-1 gene family in mam-TNF-a, produces fever in rats when injected peripherally
mals, J. Mol. Evol. 39 (1994) 6–12. [17]. In the mouse, IL-1s are transported across the BBB at
[12] G. Katsuura, P.E. Gottschall, A. Arimura, Identification of a high-different rates depending on the form (avs.b) and species affinity receptor for interleukin-1 beta in rat brain, Biochem. (murine vs. human) – relative rate: murine IL-1a.murine Biophys. Res. Commun. 156 (1988) 61–67.
IL-1b.human IL-1a [3,5]. [13] C. Liu, Y. Bai, D. Ganea, R.P. Hart, Species-specific activity of rat recombinant interleukin-1b, J. Interferon Cytokine Res. 15 (1995) In summary, rat, but not human, IL-1a is transported
985–992. across the rat BBB. We postulate that structural differences
[14] L.M. Maness, W.A. Banks, J.E. Zadina, A.J. Kastin, Selective between rat and human IL-1a result in altered binding of transport of blood-borne interleukin-1a into the posterior division of these molecules to the rat IL-1 transporter. A similar the septum of the mouse brain, Brain Res. 700 (1995) 83–88. hypothesis was used in the past to explain the difficulty in [15] T. Nishido, N. Nishino, M. Takano, Y. Sekiguchi, K. Kawai, K.
Mizuno, S. Nakai, Y. Masui, Y. Hirai, Molecular cloning and measuring rat, human, and murine IL-1 binding in brain
expression of rat interleukin-1a cDNA, J. Biochem. 105 (1989) [19,21]. This phenomenon may explain, in part, the
351–357.
125
[17] A. Stefferl, S.J. Hopkins, N.J. Rothwell, G.N. Luheshi, The role of [ I]interleukin-1 binding in brain, endocrine and immune tissues, TNF-ain fever: opposing actions of human and murine TNF-aand Brain Res. 623 (1993) 172–176.
interactions with IL-b in the rat, Br. J. Pharmacol. 118 (1996) [20] T. Takao, D.E. Tracey, W.M. Mitchell, E.B. De Souza, Interleukin-1 1919–1924. receptors in mouse brain: characterization and neuronal localization, [18] T. Takao, W.M. Mitchell, D.E. Tracey, E.B. De Souza, Identification Endocrinology 127 (1990) 3070–3078.
of interleukin-1 receptors in mouse testis, Endocrinology 127 (1990) [21] M. Wong, J. Licinio, Localization of interleukin 1 type 1 receptor
251–258. mRNA in rat brain, Neuroimmunomodulation 1 (1993) 110–115.
