Download Characterization of serine/threonine protein phosphatases in

Bioscience Reports, Vol. 13, No. 6, 1993
Characterization of Serine/Threonine
Protein Phosphatases in RINm5F Insulinoma
Cells
~ k e Sjiiholm, 1'2'3's Richard E. Honkanen, 1'4 and
Per-Oiof Berggren z
Received September 29, 1993; accepted October 22, 1993.
This study investigates the occurrence and regulation of serine/threonine protein phosphatases
(PPases) in insulin-secreting RINm5F insulinoma cells. PPases types 1 and 2A were identified in crude
RINm5F cell homogenates by both enzymatic assay and Western blot analysis. We then characterized
and compared the inhibitory actions of several compounds isolated from cyanobacteria, marine
dinoflagellates and marine sponges, (viz. okadaic acid, microcystin-LR, calyculin-A and nodularin)
cation-independent PPase activities in RINm5F cell homogenates. It was found that okadaic acid was
the least potent inhibitor (IC50 ~ 10 - ' M, ICto o ~ 10-6 M), while the other compounds exhibited ICs0
values of = 5 - 1 0 - t ~
and IC~co~5 9 10 9M. The findings indicate that the inhibitory substances
employed in this study may be used pharmacologically to investigate the role of serine/threonine
PPases in RINm5F cell insulin secretion, a process that is likely to be regulated to a major extent by
protein phosphorylation.
KEY WORDS: insulin secretion; insulinoma; okadaic acid; protein phosphatase.
INTRODUCTION
Reversible phosphorylation of certain intracellular proteins is believed to be an
important mechanism for regulating their biological activity which, in turn,
controls a variety of cellular functions. For instance, significant changes in protein
kinase activities and in protein phosphorylation patterns occur after the stimulation of insulin release in response to certain secretagogues [1-3]. Therefore, the
molecular mechanisms regulating phosphorylation of proteins involved in the
University of Hawaii at Manoa, Cancer Research Center of Hawaii, Molecular Oncology Program,
1236 Lauhala Street, Honolulu, HI 96813, U.S.A.
ZDepartment of Endocrinology, The Rolf Luft Center for Diabetes Research, Karolinska Institute,
Karolinska Hospital, S-171 76 Stockholm, Sweden.
3 Department of Internal Medicine, L6wenstr6mska Hospital, S-194 89 Upplands V~isby, Sweden.
4Department of Biochemistry (MSB 21981, College of Medicine, University of South Alabama,
Mobile AL 36688, U.S.A.
5 To whom correspondence should be addressed (at a).
349
0144-8463/93/1200-0349507.00/0 9 1993 Plenum Publishing Corporation
Sj6holm, Honkanenand Berggren
350
insulin secretory process have been extensively investigated. However, far less is
known about the role and regulation of protein dephosphorylation by various
protein phosphatases (PPases). Biochemically, most serine/threonine PPases can
be placed in one of two major classes (PPase-1 and PPase-2) based on their
sensitivity to two endogenous heat and acid stable inhibitory proteins (I-1 and I-2)
and their ability to dephosphorylate either the c~ or fi subunit of phosphorylase
kinase [for rev; see 4-7]. The catalytic subunit of PPase-1 preferentially
dephosphorylates the fi-subunit of phosphorylase kinase and is inhibited by
nanomolar concentrations of I-1 and I-2. PPase-2 preferentially dephosphorylates
the o:-subunit of phosphorylase kinase and is not sensitive to nanomolar
concentrations of I-1 or I-2. PPase-2 can be further divided into three subtypes
based on their requirements for divalent cations; PPase-2A, like PPase-1, does
not require divalent cations, while PPase-2B (calcineurin) and PPase-2C have an
absolute requirement for CaZ+/calmodulin and Mg 2+, respectively. Other PPases,
(e.g., PPase-X, PPase-Y, PPase-Z [8] and rdgC [9]) have been identified by
cloning studies, and a minor PPase (PPase-3) has recently been purified from
bovine brain [10]; however, the abundance and tissue distribution of these PPases
have yet to be determined [5, 8-10].
This paper investigates the occurrence and regulation of serine/threonine
PPases in insulin-secreting RINm5F insulinoma cells. Parts of this study has been
published in abstract form [11, 12].
MATERIALS AND METHODS
Materials
Okadaic acid was obtained from Moanoa Bioproducts (Honolulu, HI).
Calyculin-A was from LC Services Corp., Woburn, MA, while nodularin was
obtained from Calbiochem. Microcystin-LR was isolated from Microcystis aeruginosa (strain 7820) as described [13] and was identified from its fast atom
bombardment spectrum (m/z 995 [MH+]) and [13C]NMR spectrum in MeOH-d4.
Amounts of microcystin-LR were calculated by UV analysis [14]. Cyclic
adenosine-3',5'-monophosphate (cAMP), protein kinase A (cAMP-dependent),
crude histone (type 2-AS) were from Sigma. Ammonium molybdate was obtained
from Mallinckrodt, while [y-32p]ATP (350Ci/mol) was from New England
Nuclear.
Immunoblotting
Purified PPase-1 and PPase-2A from rabbit muscle and whole RINm5F cells
homogenates were utilized for immunoblotting using type-specific rat polyclonal
antibodies generously provided by Dr. D. Alexander [15]. Aliquots of RINm5F
cell homogenates and PPase-1 or PPase-2A, as controls, were subjected to
SDS-PAGE on 10% polyacrylamide gels. Gels were then electrophoretically
transferred to Immobilon-P and immunoreactivity of the primary antibody was
visualized by fluorography with an enhanced chemiluminescence system (ECL,
Amersham) essentially according to the methods of the manufacturer.
Protein Phosphatesin RINm5FCells
351
Preparation of Phosphohistone and Determination of Phosphatase Activity
Histone type 2-AS was phosphorylated with rabbit muscle type I protein kinase
A (cAMP-dependent) as described previously [16]. Incubation mixtures (4 ml)
consisted of 20 mg histone, 1 mg protein kinase A , 20 mM Tris-HCl (pH 7.5),
5mM MgC12, 0.5mM dithiothreitol (DTT), 0.7 gM cAMP and 150gM [732p]ATP. The reaction mixture was incubated at 30~ for 3 h and terminated by
addition of trichloroacetic acid (20% final concentration). The phosphorylated
histone was recovered according to Meisler and Langan [17]. RINm5F cells [18]
were cultured to subconfluence for 2 days in 60 mm plastic Falcon culture dishes
in medium BME supplemented with 10% fetal bovine serum, 100 U/ml benzylpenicillin and 0.1 mg/ml streptomycin. The medium was quickly removed, cells
washed three times in ice-cold PBS, scraped off plates in 1 ml Tris buffer (20 mM
Tris-HCl, 1 mM EDTA and 2 mM DTT, pH 7.4) and disrupted in a Polytron
homogenizer. After pelleting debris, homogenates were swiftly transferred to
Eppendorf tubes which were immediately plunged into liquid nitrogen and stored
at -80~ pending analysis. Phosphatase activity against phosphohistone was
determined by measuring the liberation of [32p]. Assays (80/~1 total volume)
containing 50mM Tris-HCl (pH7.4), 0.5mM DTT, l mM EDTA and
[32p]histone (2/~M PO4) plus the appropriate test substance (or equal amounts of
its solvent, dimethyl formamide) were conducted at 30~ for 10 min [10, 14, 16].
Reactions were stopped by addition of 100/~1 1 M H2SO4 containing 1 mM
potassium phosphate. Substrate dephosphorylation was kept to less than 10% of
total phosphorylated substrate, and preliminary experiments were performed to
determine the titration end point [14, 16] ensuring that the reactions were linear
with respect to time and enzyme concentration. [32p] released was extracted as a
phosphomolybdate complex and measured according to Kiililea et al. [19].
RESULTS AND DISCUSSION
In the last decade, a great deal of interest has been focused on how
reversible phosphorylation of proteins is involved in regulation of cellular
functions [4-7, 20]. Reversible changes in levels of phosphoserine and phosphothreonine at specific residues are a means by which the activity of many key
proteins is regulated and a way through which cells convey extracellular signals
into such diverse biological responses as mitogenesis, ion channel activity,
substrate uptake and metabolism, and hormone secretion [4-7, 21-27]. This
phosphorylation/dephosphorylation cycle is now known to be a dynamic process
with cellular levels of protein phosphorylation being determined by the combined
actions of numerous protein kinases and PPases. However, compared with
protein kinases, serine/threonine PPases have received far less attention, partly
because the lack of specific inhibitors, and little is known about their role in the
insulin-producing pancreatic/3-cell.
Studies designed to determine the biological functions of PPases have
recently been aided by the discovery of several potent and specific PPase
352
Sj6holm, Honkanen and Berggren
OH
O
H3CO
CH3/
o
~
o
o
Ho
,
"'
r~
6H
~CH~
H
,
' .....
"
,""
,.
0
_
!
\
OH
Okadaic acid
H
CO~H
CH3
HN
FI ~ S
O
N
O
O
NH
O
H
CH
/ 3
H
~
OO
~n 3
CO2H O
NH
HN~NH 2
Microcystin-LR
Fig. 1.
C%H
\
II - " " a
f
O
()CH 3
Calyculin A
H3C.
H OCH3
OH
NH
HN~
NH2
Nodularin
Structure of the PPase inhibitors okadaic acid, microcystin-LR, calyculin-A and nodularin.
inhibitors (Fig. 1). Okadaic acid, a complex polyether fatty acid produced by
certain strains of marine dinoflagellates, calyculin-A, a novel spiro-ketal isolated
from a marine sponge, and two cyclic peptides produced by cyanobacteria,
microcystin-LR and nodularin, have all been shown to specifically inhibit different
types of PPases, and in some cases, exert tumorigenic activity [6, 7, 14, 16, 28,
29]. The inhibitory potency of the most extensively investigated compound,
okadaic acid, has been carefully characterized on type 1, 2A, 2B, 2C and 3
PPases [6, 10, 16, 30]. These studies revealed that, in a dilute whole cell
homogenate, 1 nM of okadaic acid inhibits essentially all PPase-2A activity, while
having no appreciable effect on PPase-1, -2B or -2C [6, 10, 16, 31]. One
nanomolar okadaic acid also partially inhibit the activity of PPase-3; however,
compared to PPase-1 and -2A, PPase-3 represents a minor amount (<3%) of the
total okadaic acid sensitive and divalent cation independent PPase activity in a
whole cell homogenate, and PPase-3 is generally not in an active state prior to
solubilization with cholate [10]. Furthermore, no appreciable amount of PPase-3
activity could be detected in RINm5F cell homogenates or upon Western analysis
with polyclonal antisera produced against PPase-3 [A.S., R.E.H. & P.-O.B.,
(unpublished observations). Okadaic acid, at any concentraion, is without
apparent effect on PPase-2C, tyrosine phosphatases, inositol trisphosphatase,
mitochondrial pyruvate dehydrogenase, or acid/alkaline phosphatases, as well as
numerous kinases (e.g. cAMP-dependent, CaZ+/calmodulin-dependent, and
Ca2+/phospholipid-dependent protein kinases [5, 6, 14, 16]. At concentrations
>5/~M okadaic acid will partially inhibit the activity of PPase-2B [16, 30].
Protein Phosphates in RINm5F Cells
353
Fig. 2. Immunochemicalanalysis of RINm5F cell homogenates identifying the presence of PPase-1 and PPase-2A.
Purified PPase-1 (lane 1) RINm5F cell homogenate (lane 2
and 4), and purified PPase-2A (lane 3) were analyzed by
immunolabeling using polyclonal antibodies specific for
PPase-1 (1-2) or PPase-2A (3-4) as detailed in the Materials
and Methods section.
In Fig. 2 it is demonstrated by Western analyses that RINm5F cells contain
both PPase-1 and PPase-2A protein. The CaZ+/calmodulin-dependent PPase-2B
(calcineurin) has also recently been localized by Western blotting in rat pancreatic
islets [32]; however, its functional role in the/3-cell has not yet been studied. In
addition, since the PPase assays in this paper were performed in the presence of
the cation chelator E G T A , the CaZ+/calmodulin-dependent PPase-2B and the
MgZ+-requiring PPase-2C do no contribute to the results obtained. As seen in
Fig. 3, the divalent cation independent PPase activity in the dilute RINm5F cell
homogenate is completely inhibited by okadaic acid at a concentration of 1 ~M,
with partial inhibition occurring at a concentration of less than 0.1 nM. This
indicates that both PPase-1 and PPase-2A are active in RINm5F cells (i.e. if only
PPase-1 was active then in appropriately diluted samples inhibition should not
occur at concentrations below 1 nM [16, 31]). Similarly, if only PPase-2A was
active then complete inhibition would be observed at a concentration of 1-2 nM
[6, 14, 16, 31]). Of course, other as of yet biochemically uncharacterized PPases
(i.e. PPase-Y, PPase-Z etc. [8]) could also be sensitive to okadaic acid and, thus,
comprise a portion of the activity ascribed here to PPase-1 or PPase-2A.
Comparison of the inhibitory activity of okadaic acid with that of other PPase
inhibitors demonstrates that, in dilute RINm5F cell homogenates, okadaic acid is
the least potent inhibitor, exhibiting IC50 at --~10-9M and ICloo at ~ 1 0 - 6 M . As
mentioned above, these values agree well with previously published results
employing a combination of pure PPases types 1 and 2A and other cell
homogenates containing PPase-1 and -2A [8, 10, 16, 31]. The inflection point in
PPase activity noted at 10-9M okadaic acid (Fig. 3) likely reflects the concentra-
354
Sj6holm, Honkanen and Berggren
i
f
i
I
I
i
t
i
i
100
80
4,g
S 60
0
0
#%
0
m N 40
20
0
I
1E-131E-12
1E-11 1 E - I O
I
I
1E-9
1E-8
[Okadaic
i
i
I
acid]
I
(r
1E-7
1E-6
1E-5
1E-4
(M)
i
i
1E-8
1E-7
100
80
4.J - -
9~ o
~E
U
60
0
0
9
4O
20
0
1E-15 1E-121E-11
I
1E-10 1E-9
[Microcystin-LR]
1E-6
I
1E-5
1E-4-
(M)
Fig. 3. Different inhibitory profiles of okadaic acid,, microcystin-LR, calyculin-A and
nodularin on RINm5F PPase activities. Cell homogenates were incubated for 10 min at 30~
with the indicated agent using [32p]histone as a substrate. PPase activity was assayed as
detailed in the Materials and Methods section. Values are mean percent of controls
(incubated with solvent only) +S.E.M, for 4 separate experiments.
Protein Phosphates in RINm5F Cells
355
100
80
4"
D
9~ o
~
~~
E~
60
O
O
(D
a_N 40
n
..j
20
0
1E-131E-121E-111E-10
I
1E-9
1E-8
[Calyculin-A]
i
i
i
i
1E-7
I
1E-6
1E-5
1E-4
(M)
i
i
i
i
100
80
"~
O
.---
E 60
if)
O
a_ N
40
20
1E-131E-121E-111E-101E-9
1E-8
[Nodularin]
Fig. 3.
(Continued)
1E-71E-6
(M)
1E-51E-4
356
Sj6holm, Honkanen and Berggren
tion where all PPase-2A activity is inhibited and type 1 activity remains
unaffected [14, 16]. Again, this is in agreement with studies with purified enzymes
[8, 10, 14, 16], and suggests that PPase-1 and PPase-2A represent the quantitatively most important cation-independent serine/threonine PPases in RINm5F
cells.
In comparison with okadaic acid, the other compounds tested, i.e.,
microcystin-LR, calyculin-A and nodularin, were more potent inhibitors of
RINm5F cell PPases, exhibiting IC50 values of ~ 5 - 1 0 - 1 ~
and IC1oo of
~ 5 . 1 0 - 9 M . These findings indicate that such inhibitors can be used pharmacologically to evaluate a possible role of different PPases in the regulation of
the stimulus-secretion coupling in insulin-producing cells, which is believed to be
controlled to a great extent by protein phosphorylation. More specifically, under
the conditions utilized in this study, microcystin-LR, calyculin-A and nodularin all
completely inhibit PPases 1 and 2A activity at a concentration of ~ 1 - 5 nM,
essentially the same concentration at which okadaic acid inhibits all PPase-2A
activity without affecting the activity of PPase-1.
In insulin-secreting cells, previous studies have established that stimulation of
protein phosphorylation by direct activation of Ca2+/phospholipid-dependent
protein kinase (protein kinase C) with phorbol esters or pharmacological
elevation of cAMP levels, activating protein kinase A, stimulates insulin secretion
[1-3, 33-38]. Further, a number of physiological stimuli of insulin secretion
increase fl-cell phosphorylation state [1-3, 33, 34]. In a preliminary report [27],
acute exposure of RINm5F cells to the cell-permeant okadaic acid was found to
enhance Ca 2+ entry and to stimulate insulin secretion. It should be possible to
design experiments distinguishing the functional consequences of inhibiting either
PPase-1, PPase-2A or both. For example, if similar results are obtained with
~ l . 0 n M okadaic acid and ~ l . 0 n M calyculin-A, then the observed change
produced by these inhibitors is likely to be due to the loss of PPase-2A activity.
Alternatively, if a change is observed in the presence of 1.0 nM calyculin-A while
1.0 nM okadaic acid is without effect and 1.0/~M okadaic acid produces an effect
similar to that observed with calyculin-A, then PPase-1 is likely to be the PPase
involved. However, because of differences in cellular uptake kinetics between the
drugs, the utility of such experimental approaches may be confined to studies in
permeabilized or homogenized cells to which the different inhibitors or purified
PPases may be directly introduced. An equally interesting approach may be to
microinject the inhibitors or purified PPases or protein kinases into cells or to use
them in patch-clamp studies to determine whether these enzymes alter ion
channel activity and thereby the insulin secretory response.
Thus in conclusion, we have demonstrated that insulin-secreting cells
contain, in addition to previously reported PPase-2B/calcineurin [32], PPases
types 1 and 2A. Moreover, the presently reported characterization of doseresponse relationships between the four PPase inhibitors, okadaic acid,
microcystin-LR, calyculin-A and nodularin, indicate that such compounds may be
useful in probing the roles of different PPases in the regulation of the
stimulus-secretion coupling in insulin-producing cells.
Protein Phosphates in RINm5F Cells
357
ACKNOWLEDGEMENTS
We thank Dr. D. Alexander for generously supplying us with type specific
antiserum to PPase-1 and to PPase-2A. Financial support was received from the
Swedish Medical Research Council (03X-09890, 19X-00034), the Cancer Foundation, the Sandoz Corporation, the R6nnow Fund, the Pharmacia Fund for
Biotechnological Research, the Wallenberg Fund, the Anna Cederberg Fund, the
Wenner-Gren Foundation, Funds of the Karolinska Institute, the Bank of
Sweden Tercentenary Foundation, the Swedish Diabetes Association, the Swedish Society of Medicine, the Nordic Insulin Fund, Fredrik and Inger Thuring's
Foundation and Syskonen Svensson's Fund.
REFERENCES
1.
2.
3.
4.
5,
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Hedeskov, C. J. (1980) Physiol. Rev. 60:442-508.
Efendic', S., Kindmark, H. and Berggren, P.-O. (1991) J. Int. Med. 229 (Suppl. 2):9-22.
Turk, J., Wolf, B, A. and McDaniel, M. L. (1987) Prog. Lipid Res. 26:125-181.
Cohen, P. (1989) Annu. Rev. Biochem. 58:453-508.
Shenolikar, S. and Nairn, A. C. (1991)Adv. Sec. Mess, Phosphoprot. Res. 23:1-121.
Cohen, P., Holmes, C. F. B. and Tsukitani, Y. (1990) Trends Biochem. Sci. 15:98-102.
Bollen, M. and Stalmans, W. (1992) Crit. Rev. Biochem. Mol. Biol. 27:227-281.
Cohen, P. T. W., Brewis, N. D., Hughes, V. and Mann, D. J. (1990) FEBS Lett. 268:355-358.
Steele, F. R., Washburn, T., Rieger, R. and O'Tousa, J. E. (1992) Cell 69:669-676.
Honkanen, R. E., Zwiller, J., Daily, S., Khatra, B. S., Dukelow, M. and Boynton, A. L. (1991)
J. Biol. Chem. 266:6614-6619.
Sj6holm, A., Honkanen, R. E. and Berggren, P.-O. (1992) Diabetologia 35 (Suppl. 1):A100
(abstract).~
Sj6holm, A., Honkanen, R. and Berggren, P.-O. (1993) J. Cell. Biochem. Suppl. 17A:315
(abstract).
Krishnamurthy, T., Carmichael, W. W. and Sarver, E. W. (1986) Toxicon 24:865-873.
Honkanen, R. E., Zwiller, J., Moore, R. E., Daily, S. L., Khatra, B. S., Dukelow, M. and
Boynton, A. L. (1990)J. Biol. Chem. 265:19,401-19,404.
Alexander, D. R., Hexham, J. M. and Crumpton, M. J. (1988) Biochem. J. 256:885-892.
Honkanen, R. E., Dukelow, M., Moore, R. E., Zwiller, J., Khatra, B. S. and Boynton, A. L.
(1991) Mol. Pharmacol. 40:577-583.
Meisler, M. H. and Langan, T. A. (1969) J. Biol. Chem. 244:4961-4968.
Sj6holm, ,~., Welsh, N., Bankston, P. W., Hoftiezer, V. and Hellerstr6m, C. (1991) Biochem. J.
277: 533-540.
Killilea, S. D., Aylward, J. H., Mellgren, R. L. and Lee, E. Y. C. (1978) Arch. Biochem.
Biophys. 191: 638-646.
Nishizuka, Y. (1988) Nature 334:661-665.
Chiavaroli, C., Vacher, P. and Schlegel, W. (1992) Eur. J. Pharmacol. 227: 173-180.
Wagner, P. D. and Vu, N.-D. (1990) J. Biol. Chem. 265"10,352-10,357.
Reinhart, P. H., Chung, S., Martin, B. L., Brautigan, D. L. and Levitan, I. B. (1991) J.
Neurosci. 11 91627-1635.
Wu, Y. N. and Wagner, P. D. (1991) Biochim. Biophys. Acta 1092:384-390.
White, R. E., Schonbrunn, A. and Armstrong, D. L. (1991) Nature 351:570-573.
ZwiUer, J., Honkanen, R. and Boynton, A. L. (1990) Exp. Cell Res. 187:193-202.
Larsson, O., Haby, C., Islam, M. S., Zwiller, J. and Berggren, P.-O. (1993) J. Cell. Biochem.
Suppl. 17A: 287 (abstract).
Meriluoto, J. A., Eriksson, J. E., Havada, K., Dahlem, A. M., Sivonen, K. and Carmichael, W.
W. (1990)J. Chromatogr. 509:390-395.
Ishihara, H., Martin, B. L., Brautigan, D. L., Karaki, H., Ozaki, H., Kato, Y., Fusetani, N.,
Watabe, S., Hashimoto, K. and Uemura, D. (1989) Biochem. Biophys. Res. Commun.
159: 871-877.
358
Sj6holm, Honkanen and Berggren
30. Bialojan, C. and Takai, A. (1988) Biochem. J. 250:283-290.
31. Cohen, P. (1991) Meth. Enzymol. 201:389-397.
32. Gagliardino, J. J., Krinks, M. H. and Gagliardino, E. E. (1991) Biochim. Biophys. Acta
1091: 370-373.
33. Wollheim, C. B. and Biden, T. J. (1986) Ann. N Y A c a d . Sci. 488:317-333.
34. Amm/il~i, C., Ashcroft, F. M. and Rorsman, P. (1993) Nature 363:356-358.
35. Sj6holm, A. (1991) FEBS Lett. 294:257-260.
36. Sj6holm, oA. (1992) Am. J. Physiol. 263:Cl14-C120.
37. Sjfholm, A. (1992) FEBS Len. 311:85-90.
38. Sj6holm, fi,. (1993) Am. J. Physiol. 264:C501-C518.