Identification of a novel inositol phosphate recognition site: Specific [3H]inositol hexakisphosphate binding to brain regions and cerebellar membranes

Vol. March

167,

No.

16,

2, 1990

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Pages

1990

Identification [3H]Inositol

of a Novel Inositol Phosphate Recognition Hexakisphosphate Binding to Brain Regions Membranes

P. T. Hawkins,

D.J.M. Reynolds*,

D. R. Poyner,

819-827

Site : Specific and Cerebellar

and M. R. Hanley

MRC Molecular Neurobiology Unit MRC Centre, Hills Road CambridgeCB2 2QH, England *MRC Unit and University Department of Clinical Pharmacology Radcliffe Infiiary Oxford OX2 6HE, England Received

January

26,

1990

[3H]Inositol hexakisphosphate (InsPs) binds with a heterogeneousdistribution to frozen sectionsof unfixed rat brain and is displacedby unlabelled InsP6. The pattern of binding correlates with binding to neuronal cell bodies. [sH]InsPe binding to cerebellar membraneshas been further character&d, is reversible, and saturable,and exhibits high specificity for inositol polyphosphates.The IC50 for competition by unlabelled InsP6 is approximately lOOnM, whereas inositol 1,3,4,5,6 pentakisphosphate (Ins(13456)Ps), inositol 1,3,4,5 tetrakisphosphate(Ins(l345)P4), and inositol 1,4,5 trisphosphate(Ins(145)Ps)bind with an affinity at leastone order of magnitude lower. [3H]InsPb binding is clearly distinct from previously characterised Ins( 145)Ps (ref. 1, 2) and Ins(1345)P4 (ref. 3) binding, both in terms of pharmacologyand brain distribution. o KGIO Academic PES, W.

Inositol hexakisphosphate (InsP6) has been known as a major source of organic phosphorusin plants for many years (14), but hasonly recently beenfound in a variety of mammaliancells and tissues(6,7, 17, 18, 19, 20,22), including rat brain (4). The functions of InsP6are unknown, but its metabolismis complex and is linked indirectly, if at all, to the inositol lipid signal transduction pathway (8, 9, 13, 16, 21, 26). Furthermore,

InsP6 and inositol pentakisphosphate (InsPs) possessa novel

extracellular pharmacology,sinceboth can activate specific neuronalpopulationsin rat brain stem(4, 1l), hippocampus(lo), the dorsalhorn of the spinal cord (5) and alsoin primary cultures of cerebellargranulecells (12). The searchfor physiological functions of InsPs and InsP6,in particular as extracellular signals,would be greatly assistedby the identification of specific recognition sites.We have identified specific [3H]InsP6 binding sitesthat exhibit a distinctive regionaldistribution in rat brain. 0006-291W90

819

$1.50

Copyright 0 1990 by Academic Press. Inc. All rights of reproduction in any form reserved.

Vol.

167,

No.

2, 1990

BIOCHEMICAL

160 fraction

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

numba

-Bi

Bii -

FIG. 1 HPLC SEPARATION OF INOSITOL PHOSPHATES Analysis of the products of the chemical phosphorylation of [3H]inositol. An aliquot of the (4 reaction mixture (see methods) was mixed with an internal standard of [32P]InsPe (prepared from [32P]-labelled mung beans, (25)) and separated by HPLC. A partisil IO-SAX column (25cm. Jones Chromatography, U.K.) was used and eluted with a gradient of A (H20) versus B (3SM ammonium formate, pH adjusted to 3.7 with orthophosphoric acid) as follows; 0 min, 0% B: 5 min, 0% B: lOmin, 21.4%B: 12min, 21.4%B: 18 min, 28.5% B: 23 min, 28.6% B: 30 min, 40.0% B: 40 min, 42.0% B: 60 min, 99.9% B: 65 min, 99.9% B: 67 min, 0% B. The flow rate was 1.25mllmin and fractions (0.3 min) were collected from t=O. The fractions were solubilised by the addition of methanol/H20 (1: 1 v/v, l.Oml) and scintillant (4.0ml; Quickszint 401, Zinsser, U.K.) and their radioactivity determined by dual-label scintillation counting. (o), r3H] radioactivity (19,608 cpm applied to column, 19,247 cpm recovered): (o), [=P] radioactivity ( 1,200 cpm applied to column, 1,100 cpm recovered). Analysis of [3H]InsPtj after incubation with cerebellar membranes. [3H]hSP6 was 03 incubated with cerebellar membranes for 90 minutes at 4”C, as described in the methods. The incubation was terminated by centrifugation and the resulting membrane pellet and supematant each 820

Vol.

167,

No.

2, 1990

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

METHODS Prenaration of [ZHlInsPg The method used for the chemical phosphorylation of [3H]inositol was essentially that described in reference 15. An aliquot of [3H]inositol (approximately SOCi/mmol: Amersham International, U.K.) was dried in vacua in a glass “quick-fit” tube. An aliquot of polyphosphoric acid (0.18g : Aldrich) was added and the sample heated at 15o”C, under vacuum, for 6 hours. The vacuum was released, an aliquot of H20 (OSml) added and the sample heated at 100°C for a further 3 hours. The sample was then allowed to cool, diluted to lOOm1 with Hz0 and the pH adjusted to 7.0 with NaOH. An aliquot (1 .Oml) of the sample at this stage was analysed directly by HPLC (Figure 1A). For the routine preparation of large quantities of [3H]InsPg, the sample (1OOml) was applied to a small (0.2ml) Ag-1 (X8, 200-400 mesh) formate-form resin column (Bio-rad, U.K.). The column was eluted with 0.2 M ammonium formate/O.lM formic acid (10ml) to remove Pi and then 2.OM ammonium formate/O.lM formic acid (2.0ml) to elute [3H]inositol polyphosphates, including [3H]InSP6. The sample (2.0ml) was freeze-dried (to remove the ammonium formate by sublimation), redissolved in Hz0 (l.Oml), and applied to a Partisil lo-SAX HPLC column (see legend to Fig. 1A). The fractions corresponding to [3H]InsPg were combined (only the peak fractions amounting to approximately 75% of the total peak area were collected) and desalted (17). The sample of [3H]InSP6WaS stored in Hz0 at -20-C. l.OmCi of [3H]inositol was routinely phosphorylated to yield approximately 2OOpCi of HPLCpurified [3H]InSP6. Autoradioaranhic localisation of [3=6 binding to brain sections Bovine brain stem was acquired at a local abattoir, removed within 5 minutes of slaughter and frozen in dry ice. Rat brains were removed immediately after decapitation and frozen in dry ice. Tissue was stored for no more than 4 weeks at -70°C and frozen sections taken at 12pm and stored at -20°C for no more than 5 days before use. Sections were preincubated for 30 mins at 4°C in binding buffer (20mM Tris buffer pH 7.7 at 4°C containing 20mM NaCl, 1OOmM KCI, 5mM EDTA and 0.1% BSA). They were then incubated in 5nM [3H]InsPg in incubation buffer at 4°C for 30 mins with or without the addition of 25pM unlabelled InsPg to define nonspecific binding. All sections were then washed twice for 2 mins each in ice cold incubation buffer and dipped in distilled H20 at 4°C before being air dried and apposed to Amersham Hyperfilm for 4-6 weeks. Non specific binding accounted for less than 5% of total binding. 1zHlInsPcbinding to cerebellar membranes Cerebella were removed from rats, homogenised in 20 volumes of 20mM NaCl, 1OOmM KCl, 5mM EDTA, 20mM Tris pH 7.7, and spun at 35000g for 30 mins. The pellet was washed once and finally resuspended at approx 0.2mg protein/ml in the above buffer. Assays were carried out in lml volumes with radioligand and competing drugs added in 10~~1aliquots. Incubations containing [3H]InsPg were carried out for 90 mins.. (unless indicated otherwise) whereas for [JH]Ins( 145)P3, incubations were for 10 mins. Experiments were terminated by spinning in a microfuge for 5 mins at 17000g. All manipulations were carried out at 4°C. The pellets were washed twice with H20, dissolved in Soluene and radioactivity measured by liquid scintillation counting. reconstituted to a final volume of 1.Oml, at a final concentration of 3.5% perchloric acid (v/v). The samples were vortexed vigorously, the protein precipitate removed by centrifugation and the supematent neutralised with 2.OM KOH/O.lM MES (approximately 0.35ml). The samples were left on ice for 30 min., the resulting potassium perchlorate precipitates removed by centrifugation and the supematants analysed by HPLC. together with an internal standard of [32p]hSP6. HPLC was performed as described above, except that a shortened gradient was used: 0 min. 55% B: 6 min, 55% B: 20 min. 100% B: 26 min, 100% B. (i) Extract prepared from membrane pellet. [sH]-radioactivity: approximately 450 cpm applied to column, 543 cpm recovered. (ii) Extract prepared from supematant. [3H]-radioactivity: 24,000 cpm applied to column 19.649 cpm recovered.

821

Vol.

167,

No.

BIOCHEMICAL

2, 1990

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

RESULTS of high specific radioactivity [approx. 80Ci/mmol] was prepared by chemical phosphorylation of [3H]inositol and subsequently purified by HPLC (see Fig. [3H]hSPg

1A). [3H]InsPg prepared in this manner was incubated with unfixed rat and bovine brain sections under conditions previously used to study [3H]Ins( 145)PJ binding (2). Autoradiographic analysis indicates that the pattern of binding of [3H]InsP,j is similar to that seen with a Nissl stain (Fig. 2 and data not shown). The highest levels of binding are seen in the purkinje cell layer of the cerebellum, the pyramidal cell layer of the hippocampus, the granule cells of the dentate gyrus, the choroid plexus, the pineal gland and the dorsomedial hypothalamic nucleus. In the brainstem, high levels of binding are apparent in the nucleus of the solitary tract, the nucleus of the spinal trigeminal tract and the inferior olivary nucleus. The distribution [3H]Ins(l45)P3

of displaceable binding of [3H]InsPg is in contrast to that seen with (2) where the highest levels of binding are seen in the cerebellar

molecular layer which is rich in purkinje cell dendrites, and to both cell bodies and dendrites of intrinsic neurones of the hippocampus.

Binding of [3H]Ins( 145)Ps was

observed in the nucleus of the spinal trigeminal tract as with [3H]InsP6.

[3H]InsP6,

but not [3H]Ins( 145)P3, binds to the nucleus of the solitary tract and the overlying area postrema in the caudal medulla oblongata. Autoradiography

indicated that the cerebellum was enriched in high affinity [3H]InsPg

binding sites. Therefore, membranes from this region were chosen for the detailed characterisation of [3H]InsPg binding. [3H]InsPg binding to crude cerebellar membranes was rapid at 4°C clearly reaching a plateau by 30 minutes (Fig 3A). This binding was reversible,

since addition of 1.0 mM unlabelled InsPg after 90 mins

displaced 280% of the bound radioactivity (Fig. 3A). [3H]InsP6 binding was directly proportional to the amount of tissue in the assay up to at least 2mg protein/ml (data not shown). Under the conditions used in the binding assay [3H]InsP6 was metabolically stable (Fig 1Bii) and the radioactivity bound to the membranes extraction to be authentic [3H]hSP6 (Pig 1Bi).

was shown

by

The binding of [3H]InsPg to cerebellar membranes was examined by competition with unlabelled inositol phosphates and other compounds to establish the affinity and specificity of the site (see Fig 3B). The IC50 for competition with unlabelled InsPg was approximately 100 nM whereas that for competition with Ins(1345QP5 was closer to 5l.rM. The inhibition curves for both h&j and Ins (13456)Ps had Hill coefficients significantly less than unity, suggesting that the ligands might be recognising multiple binding sites. Both curves could be resolved into two components: after correction for ligand occupancy, 88% of the InsPg bound to a site with a Kd of 60nM, whilst the remainder

had a Kd of over 50pM.

By contrast, 822

the two sites recognised

by

Vol.

167,

No.

2, 1990

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

FIG. 2 AUTORADIOGRAMS SHOWING TOTAL BINDING OF [3HlINSP6-GENERATED OVER 12uM FROZEN SECTIONS OF RAT AND BOVINE BRAIN Shows the total binding of [3H]InsPg in a section of rat brain taken at the level of the 6% cerebellum. A high density of binding sites is seen in the purkinje cell layer of the cerebellar cortex (CX), the choroid plexus (CP) and the nucleus of the solitary tract (NTS). Scale bar = 1.2mm Shows the total binding at the level of the hippocampus in the rat. Dense binding is seen (B) over the granule cells of the dentate gyms (D), and over the pyramidal cells of the hippocampal formation (H). Lower levels of binding are also apparent in the dorsal hypothalamic nuclei (DH). Scale bar = 1.75mm. Shows the total binding of [3H]InsPh in the bovine medulla at the level of the obex. ((3 Binding is seen in the nucleus of the solitary tract (NTS) particularly in its most dorsomedial region, in the area postrema (AP) and in the dorsal motor nucleus of the vagus nerve (DMX). In contrast, there is little binding in the underlying nucleus of the hypoglossal nerve (XII). TS = solitary tract and IV = fourth ventricle. Scale bar = 1.25mm.

Ins( 13456)P5

were

of approximately

equal

proportions,

with Kds of 280nM and

22pM. We have insufficient evidence to decide whether there are discrete binding sites, or whether there is a singlereceptorcapableof exhibiting multiple, convertible binding states. The fact that the other inositol phosphateswere much lesseffective competitors 823

Vol.

167,

No.

2, 1990

0

BIOCHEMICAL

I

40

AND

BIOPHYSICAL

60 Time win)

RESEARCH

120

COMMUNICATIONS

160

FIG. 3 BINDING OF [3HIlNSPc AND 13HIINS(1.4.5)P3 TO CEREBELLAR MEMBRANE5 Kinetics of binding of 0.5nM [3H]lnsP6 (approx. 90,OOO dpm/incubation): total binding (4 (0); binding in the presence of ImM unlabelled h&j(A). After 90 mins, 1.OmM unlabelled LnsPe was added to a portion of the incubation which had been prelabelled with OSnM [3H]hSP6, and the total binding (0) measured after further 5 and 60 minutes incubation. Maximal binding was approximately 3000 dpm/incubation (approx 80 fmol/mg protein). Curves were drawn by eye. Each point represents the mean + standard deviation of three determinations. 824

Vol.

167,

No.

BIOCHEMICAL

2, 1990

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

of [3H]InsPg binding than InsP6 itself suggests that the site has a unique specificity for InsP6 and, perhaps, predicts a separate site for InsPg binding. The relative abilities of InsPg and Ins(145)P3 to compete with [3H]InSPg and [3H]Ins(145)P3 binding to cerebellar membranes (Fig 3B and .K). suggests that the [3H]InsPg and [3H]Ins( 145)P3 binding sites are distinct.

This result is in agreement

with competition curves previously published for [3H]Ins( 145>P3 binding (1) and the different distribution (see above).

of [3H]InsP3 and [3H]InsPg binding sites in brain sections (2)

The high affinity

InsPg binding site in cerebellar membranes had a

capacity of about lOpmol/mg protein, about half that seen for the Ins(145)P3 binding site (approx 2OpmoVmg protein; Fig. 3C and ref. 1) in the same preparation. DISCUSSION InsP6 appears to be a ubiquitous component of animal cells where it can accumulate to concentrations in the range lo-1OOpM (6, 22). The details of its biosynthesis are still unclear (9, 16, 21,26, 27,30), but it is likely that InsP6 will be established as the endproduct of a new, complex metabolic pathway. To date, the only functions ascribed to InsPs and InsP6 in animal cells are as allosteric regulators of soluble proteins, such as haemoglobin in avian erythrocytes (28) and aldolase in mammalian tissues (29). Thus, the identification in washed membranes of a specific, high-affinity recognition site for InsP6 introduces an important new approach to understanding InsP6 biology, emphasising the possibility that InsP6 has additional membrane-associated functions. Furthermore, the purification and subsequent cloning of this site represents a logical route towards elucidating these functions. In this regard, the discovery of an enriched source of Ins( 145)Px binding sites in cerebellum was pivotal to the recent purification and cloning of the Ins( 145)Ps controlled intracellular Ca*+-channel(23,24). One obvious membrane function for InsP6 is our earlier suggestion that this compound may have an extracellular

role as a neuronal excitant (4).

The observation

that

Competition of OSnM [3H]InsPg binding by unlabelled hSP6 (a), Ins(13456)Ps (A), (B) Ins( 1345)P~ (0), Ins( 145)P3 (w), inositol hexakissulphate (0) and inorganic phosphate (A). Each point represents the mean + standard deviation of 2-15 determinations. Maximal binding was as above. Non-specific binding was defined in the presence of 1mM unlabelled hSP6 (approx. 500 dpm). The competition curve for InsP6 had a Hill coefficient of 0.77 + 0.04, and was fitted to a two site model where 88 + 1.6% of the sites showed an IC50 of 61 + 4.4nM: and the remainder had an IC50 of 53 + 42pM. The displacement curve for Ins( 13456)Ps also had a Hill coefficient significantly less than unity (0.51 + 0.074) and was fitted to a two site model, where 44 + 6.7% of the sites had an IC50 of 280 + 82nM and the remainder had an IC50 of 22 + 11.3pM. The data for the remaining ligands was fitted by eye. Competition of 0.51&I [3H]Ins( 145)P3 binding (approx. 24,ooO dpm/incubation) by (C) unlabelled Ins( 145)P3 (e) and InsP&) in a representative experiment that was repeated twice. E&h pint represents the mean + standard deviation of triplicate determinations. Maximal binding was approximately 1700 dpmlincubation (approx. 150 fmol/mg protein). Both curves were fitted to Langmuir adsorption isotherms where the IC50 for Ins( 145)P3 was 57 + 3.4nM, and that for hS&jWaS 18.5 + 4pM. 825

Vol.

167,

No.

BIOCHEMICAL

2, 1990

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

[3H]InsPh binding is markedly enriched in each of the three brain regions where InsP6 has been shown to have activity, namely the brain stem nucleus tractus solitarius (4, 1 l), cerebellar granule cells (12) and hippocampus (lo), is entirely consistent with the notion that these sites represent the sites mediating the pharmacological activity of InsPe. In view of this possibility, it will clearly be of interest to determine whether any of a variety of neurally

active drugs or toxins recognise this site. However,

the

recognition selectivity for InsP6 over other naturally-occurring inositol phosphates has an immediate application in providing the basis for a simple and sensitive radioreceptor assay for InsP6. Although

we have emphasised the relationship

between the binding sites and the

extracellular effects of InsP6, the distribution of specific [3H]InsP6 binding to brain sections is widespread, suggesting that InsP6 may have additional actions in the central nervous system. Indeed, as InsP6 seems to be found in many, perhaps all, animal cells, part or all of the binding described here may represent binding to a more general ‘house-keeping’

site. This situation is analogous to that already accepted for ATP,

which has general intracellular functions, but also has cell surface receptors in a number of specialised tissues (eg. 31). Recent work has shown that InsP6 causes calcium entry in neurones (12). Therefore, it is interesting to speculate that some portion of the [3H]InsP6 binding described here may be to a cell surface version of the intracellular Ins( 145)P3 gated calcium channel (32). ACKNOWLEDGMENTS We thank Len Stephens, Andy Letcher Dave Lander and Robin Irvine for the gift of pure samples of h&j, Ins( 13456)P5, Ins(1345)Pq and Ins( 145)Ps. We also thank Phil Godfrey for helpful discussions and Ron Leslie for his advice and enthusiasm for the project. Mario VaIlejo and Rosario Moratalla made a significant contribution in the preliminary stages of this work with experiments on [3H]Ins(13456)Pg binding. PTH. is a Lister Institute Research Fellow. DRP is funded by a grant from CellTech plc, U.K. MRH is a recipient of a Research Award from the ILSI Research Foundation. REFERENCES 1.

2. 3. 4. 5. 6. 7.

Worley P.F., Baraban J.M., Supattapone S., Wilson V.S. & Snyder S.H. (1987) J. Biol. Chem. 262, 12132-12136. Worley P.F., Baraban J.M. & Snyder S. H. (1989) J. Neuroscience 9, 339346. Theibert A.B., Supattapone S., Worley P.F., Baraban J.M., Meek J.L. & Snyder S.H. (1987) B&hem. Biophys. Res. Comm. 148, 1283-1289. Vallejo M., Jackson T., Lightman S. & Hanley M.R. (1987) Nature (London) 330, 656-658. Hanley M.R, Jackson T.R., Vallejo M., Patterson S.I., Thastrup O., Lightman S., Rogers J., Henderson G. & Pini A (1988) Phil. Trans. R. Sot. Lond. B 320, 381-398. Szwergold B.S., Graham R.A. & Brown T.R. (1987) B&hem. Biophys. Res. Comm. m,874-881. Heslop J.P., Irvine R. F., Tashjian A.G. & Berridge M.J.(1985) J. Exp. Biol. 119. 395-401. 826

Vol.

167,

8. 9. 10. 11. 12. 13. ::: 16. 17. 18. :;: z:: 23. 24. 2 27. 28. 29. 30. ;::

No.

2, 1990

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Carpenter D., Hanley M.R., Hawkins P.T., Jackson T.R., Stephens L.R. & Vallejo M. (1988) Biochem. Sot. Trans. l7,3-5. Stephens L., Hawkins P.T., Morris A.J. & Downes C.P. (1988) Biochem. J. 249, 283-292. Spanswick D., Hawkins P.T., Logan S. & Stephens L. manuscript in preparation. Barraco R.A., Phyllis J.W. & Simpson L.L. (1989) Eur.J. Pharm. m,75-84. Nicoletti F., Bruno V., Fiore L., Cavallaro S. & Canonico P.L. (1989) J. Neurochem. 3, 1026-1030. Jackson T.R., Hallam T.J., Downes C.P. & Hanley M.R. (1987) EMBO J. 6, 49-54. Cosgrove D.J. (1966) Rev. Pure Applied Chem. l6,209-224. Cosgrove D.J. (1980) Inositol Phosphates, their Chemistry, Biochemistry and Physiology, Elsevier, Amsterdam. Stephens L.R., Hawkins P.T., Barker C.J. & Downes C.P. (1988) Biochem. J. 253, 721-733. Hawkins P.T., Stephens L.R. & Downes C.P. (1986) B&hem. J. 238, 507516. Tilly B.C., van Paridon P.A., Verlaan I., Wirtz K.W.A., de Laat S.W. 8z Moolenaar W.M. (1987) Biochem. J. 244, 129-135. Morgan R.O., Chang J.P. & Catt K.J. (1987) J.Biol. Chem. 262, 1166-1171. Dean N.M. & Moyer J.D. (1988) B&hem. J. 250,493-500. Shears. S.B. (1989) B&hem. J. 260, 313-324. Mayr G.W. (1988) B&hem. J. =,585-591. Furuichi T., Yoshikawa S., Miyawaki A., Wada K., Maeda N. & Mikoshiba K.(1989) Nature (London) 342, 32-38. Ferris C.D., Huganir R.L., Supattapone S. & Snyder S.H. (1989) Nature (London) 342. 87-89. Graf E. (1983) Anal. B&hem. U,351-355. Shears S.B. (1989) J. Biol. Chem. 264, 19879-19886. Stephens L R., & Downes C.P. (1990) Biochem. J. =,435-452. Isaacks R.E., & Harkness D.R. (1980) Amer. Zoo. 20, 115-129. Koppitz B., Vogel F. & Mayr G.W. (1986) Eur. J. B&hem. m,421-433. Hunyady L., Baukal A.J., Guillemette G., Balla T. & Catt K.J. (1988) Biochem. Biophys. Res. Comm. 157, 1247-1252. Benham C.D. & Tsien R.W. (1987) Nature (London) 32&275-278. Fink L.A. & Kaczmarek L. (1988) Trends. Neurosci. 11, 338-339.

827