BIOCHEMICAL SOCIETY TRANSACTIONS
I250
SDSlpolyacrylamide-gel electrophoresis. After either deglycosylation procedure, kidney and striatal ACE migrated
identically as a single polypeptide of M , 150000. Thus, the
size differences between pig kidney and striatal ACE are
apparently due to differential glycosylation occurring in
different cell types. EP-24.1 I has also been shown to differ
in size between tissues as a result of differential glycosylation
(Relton et ul., 1983) and two differentially glycosylated
forms of EP-24.1 I have been observed in human placental
membranes (Johnson et ul., 1984).
Our results therefore lead us to question the conclusions
o f Strittmatter et ul., (1985) that brain and kidney contain
distinct isoenzymes of ACE with different specificities for
certain amidated peptides. Our present data support the
hypothesis that endopeptidase-24.11 is the major cellsurface peptidase involved in the metabolism of the mammalian tachykinins in the brain. The endogenous peptide
substrate(s) for brain ACE have yet to be identified.
We thank the M.R.C. for financial support. N.M.H. is in receipt of
,in M.R.C. Research Studentship.
Hull, H. G., Thornberry, N . A. & Cordes, E. H. (1985) J . B i d . Chem.
260.2963 2972
Cascieri, M. A., Bull. H. G.. Mumlbrd. R. A.. Patchett. A. A,.
Thornberry. N . A. & Liang, T. (1984) Mol. Phurmacol. 25, 287 293
Hooper, N. M. & Turner. A. J. (1985) FEBS L P I I .190. 133 136
Hooper. N. M.. Kenny, A. J. & Turner. A. J. (1985) Bioc,hcwi. J . 231.
357-361
Johnson, A. R., Skidgel, R. A., Gafford. J. T. & Erdiis. E. G. (1984)
Peptides 5. 789 796
Matsas. R., Fulcher. I . S.. Kenny, A. J. & Turner. A. J. (1983) Pro(,.
Null. Arud. Sci. U.S.A. 80, 31 I 1 31 15
Matsas, R.. Kenny, A. J. & Turner, A. J. (1986) Nruroscwnc~c 18.
991 1012
Relton. J . M., Gee, N . S.. Matsas. R.. Turner. A. J. & Kenny. A. J.
(1983) Bioc,hem. J . 215, 519 523
Skidgel. R. A. & Erdiis, E. G. (1985) Proc. N u l l . A u d . Sci. 1I.S.A. 82.
1025 -1029
Skidgel, R. A., Engelbrecht. S.. Johnson, A. R. & Erdiis. E. G. ( 1984)
Peplides 5, 769-776
Soffer, R. L. (1976) Annu. Rev. Biochem. 45. 73 94
Stewart, J. R., Chaplin, M. F. & Kenny, A . J. (1984) Biochcwi. J. 221.
919 922
Strittmatter. S. M.. Lo, M . M . S.. Javitch. J. A. & Snyder. S. H.
(1984) Proc,. " 4 1 1 . ,4CUd. Sci. U . S . A .81. 1599 1603
Strittmatter. S. M.. Thiele. E. A.. Kapiloff, M . S. & Snyder. S. H.
(1985) J . BicJl. Chem. 260. 9825 9832
Thiele, E. A., Strittmatter. S. M. & Snyder. S. H. (1985) B i ( ~ d i c ~ i .
Bioph,ys. Rrrs. Crmmun. 128, 317 37-4
Yokosawa. H.. Endo. S., Ogura. Y. & Ishii. S.-I. (19x3) Biocheni.
Biophys. Rc>s. Commun. 116. 735 142
~
~
Received I3 June I986
ATP and adenosine release from the cholinergic nerve terminal
PETER J . RICHARDSON and SUSAN J. BROWN
Dtputment 01' Clinicul Biochemistrv, University of
Cumbridge, AddenhrookP's Hospital, Hills Roud,
Cumbridge C'B2 2 Q R , U.K.
I t has been proposed that ATP and adenosine function as
neuromodulators in the central nervous system (reviewed by
Stone, 1981). However, because of the ubiquitous nature of
these two metabolites it is unclear as to the source of extracellular punnes. As far as the cholinergic system is concerned,
ATP is present in cholinergic synaptic vesicles (Zimmerman,
1978), and there is evidence for the co-release of ATP with
acetylcholine at the neuromuscular junction (Silinsky,
1975). It has however proved difficult to determined whether
ATP is released with acetylcholine from nerve terminals
derived from the central nervous system (Potter & White,
1980; White et ul.. 1980). This report describes the uptake of
adenosine. and the release of both ATP and adenosine, by
immunoaffinity cholinergic nerve terminals derived from rat
caudate nucleus. The fate of the released ATP is described
here, and the possible role of ATP and adenosine in the
feedback control of these terminals is described in the
accompanying paper (Brown & Richardson, 1986).
Cholinergic nerve terminals were affinity purified using
sheep anti-(Chol- I ) serum as previously described (Richardson et ul., 1984). The depolarization-induced release of ATP
was detected using the luciferin/luciferase reaction described
41 (n = 6) pmol of ATP
by Potter & White (1980). 248
were released per mg of nerve terminal protein upon stimulation bv 25 mM-KCI. This can be compared with 1.9nmol
ofacetyicholine per mg measured
the method
Israel
Lesbats (1982). The molar ratio of acetylcholine/ATP
release was 9.2 f 0.7 : 1 with 25 mM-KC1 and 1 1.2 f 1.5 : 1
with 50p~-veratrIdine.This can be compared with an
acetylcholineiATP ratio of 6.7 + 1 .0 : 1 found in synaptic
vesicles purified from these terminals.
In order to investigate the probable fate of the release
ATP, ''C-labelled ATP was added to the outside of the
terminals during a 2 min depolarization with veratridine.
Separation of the metabolites by h.p.1.c. indicated a greater
than 90% hydrolysis of ATP. resulting in the extracellular
accumulation of AMP. adenosine and ionsine. Approx.
75% ol' the initial label was taken up by the terminals.
presumably via the high-affinity adenosine-uptake system
(Fig. I). Inhibition of the ecto 5'-nucleotidase using a specific
antibody reduced the production of adenosine and ionsine
as well as the uptake of the label.
Prelabelling the nerve terminals with [' Hladenosine
resulted in the labelling of the intraterminal ATP pool.
Subsequent stimulation with 50 PM-veratridine resulted in
the release of labelled nucleotides and nucleosides. The
-0.25
0
0 25
05
0.75
1.0
I/S ( I I P M )
Fig, 1 . uptake c?fadenosjneby c,holinergic
termjn&
Affinity-purified nerve terminals were incubated with
pH]adenosine (1-50pCi/nmol) for 5 min. The terminals were
collected on glass-fibre filters, washed, and the radioactivity
determined. The results are shown as a Lineweaver-Burke plot
of a representative experiment. rC, = 1 6 . 6 ~ ~V,,,
; =
600pmol/min per mg of protein.
1986
619th MEETING, CAMBRIDGE
Ca'+ -dependent release of labelled adenosine and ionsine
was abolished by SO pwdipyridamole, whereas inhibition of
5'-nucleotidase resulted in an increase in the total label
released. This last effect was mimicked by the adenosine
receptor antagonist theophylline. It is therefore possible
that the increase in label release observed on inhibition of
5'-nucleotidase, was due to the relief of an adenosinereceptor-mediated inhibition of the terminal.
Depolarization of the terminals in the presence of dipyridamole (to inhibit nucleoside release) and anti-(5'-nucleotidase) serum (to prevent nucleotide breakdown) indicated
that the Ca'+-dependent release of 'H label consisted of
approx. 25% nucleotide (presumably ATP) and 75%
nucleoside release. Taking into account the specific activities
o f the labelled nucleoside and ATP pools, this suggested
that approx. 30 times as much ATP was released as
adenosine.
I t is concluded that central cholinergic terminals release
both ATP and adenosine upon depolarization and that the
released ATP is rapidly broken down by ecto nucleotidases.
1251
The resulting adenosine is generated in sufficient amounts to
inhibit further ATP release.
We are grateful to Drs. J. P. Luzio and E. M. Bailyes for advice and
provision of anti-(5'-nucleotidase) serum. S.J.B. is the recipient of an
M.R.C. Training Award. This work was supported by grants from the
M.R.C. and the Muscular Dystrophy Group.
Brown. S. J. & Richardson, P. J. (1986) Biochem. Soc. Truns. 14,
000 000
Israel. M . & Lesbats, B. (1982) J . Neurochrm. 39. 2 4 8 ~ 2 5 0
Potter. P. & White, T. D. (1980) Neuroscience 5, 1351-1356
Richardson. P. J., Siddle, K. & Luzio, J . P. (1984) Biochem. J . 210,
647 .654
Silinsky, E. M . (1975) J . Phvsiol. (London) 247, 145-162
Stone. T. W. (1981) Neurosci. 6. 523-555
White, T.. Potter, P. & Wonnacott, S. (1980) J . Nmrochem. 34.
I109-~1112
Zimmermann, H. (1978) Nrumsciunce 3. 827-836
Received I I June 1986
ATP and adenosine modulation of the cholinergic nerve terminal
SUSAN J . BROWN and PETER J . RICHARDSON
Dcpurtmmt of' c'linicul Biochemistry, University af
C'umhridgc). A ddenhrooke 's Hospital, Hills Road.
C'umhridgt) ('B2 2QR. U . K .
Recent work has focused much attention on ATP and
adenosine as possible neuromodulators in the mammalian
nervous system (Stone, I98 I ) . Adenosine and its analogues
depress nerve cell firing in the neural axis (Phillis & Wu,
1981) as well as inhibiting presynaptic release of several
different transmitters (Fredholm & Hedqvist, 1980). The
concept of purinergic modulation in the central nervous
system (CNS) is supported by work on brain slices
(Dunwiddie & Hoffer, 1980) and on heterogeneous synaptosome preparations (Pedata et ul., 1986). The action of these
compounds is though to be via extracellular adenosine (A,
and A,) and ATP (P2) receptors. proposed by Burnstock
(1983). Although it is difficult to determine the source of
purine release in the CNS, there is evidence for the corelease of ATP and acetylcholine (ACh) from cholinergic
nerve terminals (Richardson & Brown, 1986). In the present
study a possible role for ATP and adenosine in feedback
control of ACh release via presynaptic extracellular receptors is examined.
Cholinergic nerve terminals were affinity purified from rat
caudate nucleus, as previously described (Richardson et al.,
1984). The immunoadsorbent terminals were incubated for
5 min at 37°C in Krebs-Ringer-Hepes buffer (KRH), then
with [3H]choline(10 pCi/ml, 1.2 p ~ for
) 5 min, washed and
placed in a 200 pI perifusion chamber over a glass-fibre filter
(Whatman GF/C). The terminals were perifused at 37°C
with KRH at a rate of 1 ml/min for 12min and then stimulated with 5 x 10 'M-veratridine for 2min (SI). After a
further IOmin perifusion with KRH, the terminals were
again stimulated with 5 x 10 5M-veratridine for 2min
(S2). Fractions ( 2 ml) were collected and subjected to liquid
scintillation counting. Adenosine deaminase ( I m-unit/ml)
was included in KRH up to the end of SI and modulators
were included in the medium after SI. In experiments investigation the effects of endogenous ATP release adensoine
deaminase was omitted from SI. Changes in the ratio of
S2/S1 were compared with control perifusions (KRH and
Abbreviations used: CNS, central nervous system; ACh, acetylcholine;
KRH, KRH, Krebs-Ringer-Hepes buffer.
Vol. 14
h
"
I1
-7
-12
-11
-9
-10
-8
log Concn. (M)
Fig. 1. Inhibition
-6
of' ACh release by adenosine and 2-chloro-
adenosine
Nerve terminals labelled with [3H]choline were placed in a
perifusion apparatus and given two 2 min stimuli (SI and S2) 10
minutes apart, with SO pM-veratridine. Adenosine deaminase
was included up to the end of SI and 50pM-dipyridamole after
S 1. Adenosine and 2-choloroadenosine were included after S I ,
and the control contained adenosine deaminase throughout.
Reductions in S2/S1 ratios were used as indicators of inhibition
by (m) adenosine and ( 0 ) 2-chloroadenosine. Arrowhead
indicates addition of 10- M-theophylline.
adenosine deaminase throughout) and used as indicators of
modulation.
The effect of extraterminal adenosine and its analogue
2-chloroadenosine to inhibit ACh release was optimal at
IO-'M and 1 0 - I " ~respectively (Fig. I). Here the maximal
degree of inhibition was approx. 30% and this was relieved
by theophylline (
M) and isobutylmethylxanthine