protein kinase c activities are increased in rat thyroid epithelial cells expressing v-ras genes
TRANSCRIPT
Vol. 157, No. 3, 1988
December 30, 1988
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Pages ]093-]]03
PROTEIN KINASE C ACTIVITIES ARE INCREASED IN RAT THYROID
EPITHELIAL CELLS EXPRESSING V-RASGI~ES
Spina, A., Chiosi, E., and Illiano, G.
3acattedra di Chimica Biologica, Dipartimento di Biochimica e
Biofisica
Uhiversit~ di Napoli, Italy
** Berlingieri, M.T., Fusco, A., and Grieco, M.
Centro di Fmdocrinologia ed Oncologia Sperimentale del C.N.R. 2 a Facolt~ di Medicina e Chirurgia
Universit~ di Napoli, Italy
Received November 2, 1988
SUMS: Both cytoplasmic and membrane-bound protein kinase C activities are increased in: Harvey-Sarcoma Virus, infected
thyroid epithelial cells. The cytoplasmic kinase C increase is
found in the chromatographic fraction eluted at lower salt
concentration (i00 mM NaCI-SI00), while the more acidic protein
fraction eluted at higher salt concentration (350 mM NaCI-S350) is virtually absent. Although the cytoplasndc SI00 fraction from the
control and ras-virus infected cells display a+~o~oarable PBt 2
binding activity ,they are different in the Ca -dependence ~d
the TPA down regulation. In addition, the membranes from the
control and ras-virus infected cells are different phosphate
acceptors in place of the HI histories. ©19s8 Academic Press, lnc.
~11tured rat thyroid epithelial cells (PC C13) retain the
typical markers of thyroid differentiation (iodide uptake,
thyreoglobulin synthesis and secretion) and depend on TSH for
growth. Upon infection by Harvey murine sarcoma virus, they become
To whom all correslmondence concerning the biochemical studies should be addressed.
To when all correspondence concerning the cellular should be addressed.
Abbreviations: TPA, 12-0-tetradecanoylphorbol 13-acetate; PBt2, phorbol-12,13 dibutyrate; TSH, thyroid stimulating hormone; PBS, phosphate saline buffer; PKC, protein kinase C; Ptd-Ser, phosphatidyl serine.
studies
1093
0006-291X/88 $1.50 Copyright © 1988 by Academic Press, Inc.
All rights of reproduction in any form reserved.
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
transformed, loosing the markers of thyroid differentiation,
becoming independent from TSH for growth, although they are not
tumorigenic (i). The v-ras infected cells, express a pointly
mutated 21.000 daltons protein, belonging to the p21 family, which
shares some structural and functional properties with the
alfa-subunit of the hetero -trimeric G-proteins (2). In fact, it
binds and hydrolizes GTP and, like other G-proteins, it has been
proposed to play a central role on transducing external signals
toward internal amplifying systems. In the same cellular model, we
have already shown that activated p21 proteins cannot be
considered directly involved in the regulation of the adenylate
cyclase activity in v-ras-transformed cells (3) Their putative
role on regulating the hydrolysis of inositol-phospholipids has
been investigated with controversial results (4-6). On the ot~er
hand strong experimental evidences (7) have shown that
intracellular p21 protein microinjection results in a sudden
increase of diacyl-glycerol levels and prompted us to investigate
on the PKC activities in the v-ras infected cells.
Materials and Methods
Cells: PC C13 is a cell line derived from Fisher rat thyroid
glands as described by Ambesi-Impiombato et al. (8); PC C13 HaMSV
are PC C13 infected with HaMSV. The cell lines were grown in
r~dified FI2 medium supplemented with 5% calf serum (Flow Laboratories, Inc., McLean, Va.) and six growth factors (Thyrotropic hormone, hydrocortisone, insulin, transferrin,
somatostatin and glycil-histidyl-lysine, all from Sigma Chemical
Co., St.Louis, Mo.). Control experiments have been carried out
growing the infected cells in the absence of the above mentionated
six growth factors. The cells wore harvested from confluent cell
monolayers and detached with a rubber-policeman and washed two
times in phosphate buffered saline, PBS (137 mMNaCI, 2.7 mMKCI,
1.47 mM KH2P04, 8.1 mMNa2P04). Cell ~mbrane Crude Preparatzon: The cells, suspended in PBS, wore
washed three times at 5,000 rpm x 5 min in J-21 Beckman Centrifuge
with isotonic buffer, containing 10 mM ~CI 2 . F~zymic source preparation for the Kinase C activity: The cells
suspended in PBS wore washed three times by centrifugation at
5.000 rpm per 5 min. in J-21-Beckman centrifuge with the sucrose
saline buffer: 0.33 M sucrose, 20 ~4 --Tris HCI, 2 mM EDTA, 0.5 mM
EGTA, 2 ~M PMSF, 5 ug/ml leupeptin, pH 7,5. The pellet was added to the s~T~ buffer and the cells suspension was homogenized with
1094
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
a loose-fitting glass-glass dounce bo~)geniger (30 strokes). The
soluble fraction, obtained after centrifugation of the <:ell hor,mogenate at 15,000 x g i~er 12 rain. was retained m~d the mei~0rane
pellet was resus[~ended w~th 5 nJ of the same buffer without
sucrose mld treated for 30 lltin, with ]% triton X I00 in a ice slowly rotation. The detergent treated sus[~?nsion ~as brought to a
hor,~)geneous resus[x~.nsion by a dounce hcm~x3enization (I0 strokes)
and then recentrifuged at 12,000 x g per I0 ~tin. to obtain the
detergent-solubi]Jzed protein kinase C from the nmnff]rmle fraction.
DF2EE ce]lulose chromatograph[ (DE 52 ) : The cytosolic axed
solubilized preparation containing about i0 r~] of protein were applied on the DE 52 cromatographic col~x~Is (0.9 x 2 cm)
eqmilibrated with 20 l~,l Tris HCI pH 7.5, 0.5 Ii~d EGTA, 2 r~,l EDTA,
washed with t~) vol~u~s of the equilibrium louffer m~d eluted
step-wise ~¢ith ]00 and 350 NaCI in 20 ~I~M Tris HCI pH 7.5. All the
buffer contained 2 ~M P[,~F and 5 ug/ml leupeptin, q~le I00 and 350
~;~,] fraction containing the protein kinase C activity, were assayed
on the stone day. (9)
Assay of the kinase C: ~e incubation mixture contained in a final
vol~re of 250 ul: 25-50 ug of enz~nic source, 2@~M Tris IlCl pH
7.5, 0.75 mM calci~ml chloride, I0 ~4 magnesi~ acetate, 0.i r~M A~, 0.2 mg/~id histone H] (Boehringer), 0.05 n~]/m] ]eupept~n,
(O~p) ATP (New ~hgland Nuclear; specific activity: 500 Ci/[mDl). The kinase C activity was triggered by adding a lipid ntixttwe of
1.6 ug 1,2 diolein and 24 ug phosphatidylserine in 20 mM Tris HCl
pH 7.5. The lipid mixture was prepared with 300 ul of l0 mg
phosphatffdylserine/ml chloroform solution and i0 ul of 500 ug
dioleine/ml chlorofornl, evaporated under nitrogen gas, diluted with ffml 20 mM Tris HCI pH 7.5 and sonicated with three l-rain.
bursts with a Kontes ~iicro Cell Disrupter at high power. Lipids
are added to the reaction ntixture just prior to initiation of the reaction. The reaction was stopped by adding 1 nt] of ice-cold 25%
TCA and collected by filtration over Millii:~re filter (0.45 t~n) after 2 h at 4 ° C and washed five times with 2 ml 5% TCA. q]~e
filters were dissolved in Beckman Ready Solv ~ and couted for
radJ oactivity.
Prot~n kinase C activity is determined by subtracting the an~m]t
of P incorporated into 3~istone HI in the absence of lipid mixture from the anount of P incorporated in the presence. To~91
protein kinase C activity ~s expressed as nn~oles of P
incorporate [x~r min. per mg of protein both of cytosolic and
membrane preparation.
Protein concentration assay- The protein concentration was
deterntined according to Bradford (I0).
Results
The protein kinase activities have been investigated L~oth in tJ~e
cytoplasm and in the membrane. As indicated in .~ethods, the
soluble cytoplasmic protein kinase activities have been assayed
either in the crude supernatant or after partial purification on a
1095
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TAHLE I
Cytosolic and membrane-bound protein kinase activities in Control (PC C13) and Harvey MSV infected PC C13 cells (PC C13 Ha MSV)
~hzymic source nmoles 32p incorporated/mg enzyme x m in.
Addition Calculated PKC activity ++
none Ca ,diacyl-glycerol phosphatidyl-serine
Crude cytosol* PC C13 0.36+.02 0.41+.021 0.05
m
PC C13 Ha MSV 0.83+.020 1.43+.027 0.60 Chromatograph.** Purified Cytosol: S i00:
PC C13 0.40+.01 0.42+.01 0.02 PC C13 Ha~IgV 0.67~,O4 1.28~.06 0.61
S 350: PC C13 0.30+.02 0.42+.02 0.12 PC C13 HaMSV I. 38~. 12 I. 38~. 03 0
C~rcmatograph. ** purified solubilized membrane proteins: M i00:
PC C13 0.67+.02 0.90+.03 0.23 PC C13 Ha~i~V 1.14~.01 1.79~.04 0.65
M 350: PC C13 1.15+.092 1.45+.08 0.30 PC C13 Ha MSV 1.26+.071 2.64+.11 1.38
The results reported, obtained fran a typical experiment carried out in triplicate, have been normalized for the cellular protein content. *,**: see Materials and Methods. S-100, S-350: Cytosolic DE52 chromatography fractions eluted respectively by 100ram and 350ram NaCI. M-100, M-350: Solubilized membrane proteins eluted through DE52 column, respectively by 100mM and 350n~ NaCI.
DE 52 column eluted by a step gradient of i00 and 350 mMNaCI. As
of the men~rane-bound kinase activities are concerned, they have
been solubilized by treating the membranes, at room temperature
for 15 min. in the presence of 0.1% Triton X I00 and partially
purified through chromatography on the anion exchange resin DE52,
as above indicated. The results are shown in the Table 1. They
indicate that a substantially higher protein kinase activity is
present in the cytosol of the v-ras infected cells, compared to
the one present in the control cells. Such activities, are of
course coni0rehensive of c-AS~ independent and dependent protein
1096
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
kinases, as well as of the ones which can be activated by ++
Ca -phospholipid.
As far as the PKC activity is concerned, the results reported in
Table I indicate that in the cytoplasm of the infected cells the ++
Ca -phospholipid protein kinase activities are activated in the
presence of the corresponding modulators to a larger extent in the
transformed than in the control cells.
The chromatography both of the crude supernatant and of the 1%
Triton solubilized membranes on an anion exchange resin (DE 52),
allow to collect in separate fractions, two populations of PKC
activities. The first one is eluted by 100 i~I NaCI and is more ++
sensitive to Ca -phospholipids than the second one, eluted at
the higher salt concentration (350 mH NaCI), likely because the
last one is already surrounded by a phospholipid environment. The
first kinase population is increased (about thirty times) in the
cytoplasm of the infected cells; on ~e contrary the second
population, eluted at 350 ~I NaCI concentration, is actually
undetectable in the cytoplasm of the infected cells. Finally, the
results reported in the Table I indicate that also the membrane
bound PKC activities are higher in the virus-infected cells than
in the control ones; the increase appears to be concerning both
the population eluted at i00 and 350 mMNa CI.
Ca2 + activation and the phorbol ester (PBt2) binding of the PKC
SI00 fraction isolated from the cytosol of normal and ras-virus
infected epithelial thyroid cells.
To test whether the kinase C activities associated to the
cytosolic fraction eluted by 100 mM NaCI may be regulated
differently in the two cell types, we have determinated the Ca 2+
requirements for the activation and the binding activity for the
PBt 2 The results reported in the Fig. 1 show the Ca 2+
concentration dependence of the PKC activities assayed in the
presence of standard Ptd-Ser and diacyl-glycerol concentrations,
indicating marked differences between the enzymic activities
present in the eytosol of control and transformed cells, mainly at
the extreme Ca2+ concentrations tested. In fact, at sublimiting
1097
V o l . 1 5 7 , N o . 3 , 1 9 8 8 B I O C H E M I C A L A N D B I O P H Y S I C A L RESEARCH C O M M U N I C A T I O N S
o-oPC-CI3 o--ePC-Hor v ey
c 10C - - .o_
>~ 8C
3 o
6C o .E
E
~ 2 c
"[~ CaCI 2 0.50 1.0 1.50 CoCl2
(EGTA+EDTA)(I:4)
4-t- Fig. i. Ca stimulation dose effect relationship of the PKC soluble activity eluted, with the crhomatographyc fraction S-100, from control (PC C13) and v-ras infected (PC C13 Ha 5kqV) thyroid
epithelial cells. The enzimic activity essay was carried out in presence of diacyl glycerol and phosphatidyl serine used at concentration described in Methods. ++ On the X-axis has been indicated both the absolute Ca ++ concentration and the ratio between Ca and his chelants (EGTA:EDTA, 1:4) concentration.
2+ Ca2+ Ca concentrations ( 2.5 mM corresponding to a / BGTA + EDTA
ratio of 0.5 ) the PKC activities fran the infected cell cytosol
get already a 50 % of the maximum increase and reache the maximum
at 5 ~ (Ca 2+ /EGTA = i), while a decrease from the maximum is
observe(] at satLmating Ca 2+ concentrations ( 7.5 mM, Ca 2+ /EGTA +
EDTA= 1.5). The Ca 2+ dependence of the PKC activities purified
from the cytosol of the control cells is depicted by a sigmoidal
curve, showing that the 50 % of the maximum increase is reached at
4.5 ~N4 Ca 2+, and the plateau at saturating Ca 2+ concentrations
(7.5 ~M,Ca 2+ /EGTA 1.5). The differences on the Ca 2+ dependence
could reflect structural differences between the regulatory
domains of the two enzymic populations.
As of the phorbol-dibutyrate binding activity is concerned, the
two enz~c populations (data not reported) do not show
sJgnificative differences .
Solubilized m~mbrane-bound PKC activities assayed with crude
m~mbrane preparations as phosphate acceptors.
1098
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TA~,E II 32p incorporation, in crude membrane preparations, by solubilized
membrane-bound kinase C fraction, partially purified through DE52 chrcmatography and eluted with 100 or 350 ~MNaCl
Specific activity (32p ~noles incorporated/mg protein/min)
I~zymic preparation Substrates
i00 r~,,1NaCI fraction: PC C13 membranes PC Cl3HaMSVmembranes MI00 PC C13 1.48 + 0.08 3.39 + 0.23 PC CI3Ha ~v~V 2.67 + 0.018 7.28 + 0.41
350 mH NaCI fraction: M350 PC C13 1.30 + 0.28 2.20 + 0.21 PC CI31~ r.~V 17.22 + 0.58 3.46 + 0.26
The incubation mixture was performed as indicated in Material and b~thods, except that cell membranes (50-70/protein), prepared as previously indicated, replaced histone HI as phosphate acceptors.
We have carried out experiments using crude n~mbrane
preparations as phosphate acceptors, instead of Histone HI with
the purpose of testing the kinase C activities on substrates
which could be more physiological. The results are reported in
Table II. ~ley confirm that PKC activities solubilized from
transformed cell membranes are more effective than the ones from
control men,fanes on phosphorylating either type of membranes
(control or trasformed).
b~reover, the two irembrane types behave differently as phosphate
acceptors. In fact, the PKC activities solubilized from control
cell membranes, independently of which eluted fraction they are
associated with, can transfer a higher amount of -phosphates to
the membranes from transformed cells than to the ones from control
cells.
The PKC activities solubilized from infected cell membranes de
not display a univocal behaviour: the ones associated with the
fraction eluted by 100 mMNaCI work much better on the transformed
membranes too; on the contrary, the ones associated with the
fraction eluted by 350 mM NaCI work much better on the membrane
fractions prepared from the control than from the transformed
cells. Membrane protein phosphorylation depends also on several
1099
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
membrane factors: n~nber, exposure and accessibility of the serin
and threonin residues, c~nposition of die membrane lipids
m~choring the enz~ne and may be nmny o~ers. The combination of
all these factors must act more favor~)ly on the transforr~d
membranes with the exception of the solubi]ized PKC fraction
eluted at 350 mM NaCl. It has been described that PKC's
solubilized from the membrane have a richer lipid environn~nt
which also explains their relative independence from the
phospholipid-diaeyl glycerol activation; we suggest that the lipid
environment of the trasformated ~n~ranes, which contaminates
this enz~ic fraction, allows a more favorable interaction with
the normal rather than with the transformed membranes.
PKC activation and intracellular translocation in normal and
v-ras infected cells treated in vivo by TPA.
The well-known t~nor promoter TPA has l~en described to
activate in vitro PKC and to pronDte in vivo an intracellular
translocation of the enz~nic proteins from the cytosol to the
membrane (ii). This substance is believed to mimic diacyl glycerol
binding at the enzyme regulatory site, because of the structural
similarities between the natural and the succedaneous activator
(12). The membrane traslocation and activation of PKC lasts a
short time, variable according to the cell type investigated,
since it is shortly followed by activation of proteolytic enzymes
which destroy and down- regulate the kinase activity. Therefore,
because of the TPA regulation, the PKC decrease in the cytosol
does not parallel a corresponding increase on the membrane. This
is what we observe in the control cell, treated in vivo by i0 uM
TPA for 10 min. (Table III), while in virus- infected cells
exposed for 10m in. to TPA, the membrane bound PKC still appears
activated.
Our data do not elucidate in detail the molecular mechanisms of
differential regulation of PKC in virus-infected cells, whether
this is due to the lack of protease activation or to the
involvement of different kinase isoenzymes, which are more
resistant to protease action, or to a combination of both. In any
1100
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TA~.E III
Distribution of the protein kinase C activity between Cytosol and Membrane, both in Control and Harvey MSV infected cells, in vivo treated
by 10ul TPA for 10min
% of kinase C activity after i0 mM TPA i0 min presence versus ~O as vehicle
Crude Cytosol O/tosolic fractions DE52
i0~I NaCI 350mM NaCI
PC C13 PC-Ha PC C13 PC-}~ PC C13 PC-[~
~,~mbrane fractions DE52
i00~4 NaCI 35@~M NaCl
PC C13 PC-Ha PC C13 PC-Ha
60% 76% 66% 68% 68% 65% 111% 201% 68% 151% The results are expressed as the percentage of the kinase C activity present in the cytosol and in men~ranes of the corresponding control cells. For other experimental procedures see 5k~terials and Methods.
case the consequences must be physiologically relevant for the
cells.
Discussion
Our results indicate that higher levels of PKC are present "in
toto" in the v-ras transformed rat thyroid epithelial cells. These
could be related to the expression of the mutated p21 proteins .In
fact it has been shown that p21 proteins are involved in the
diacyl-glycerol production (7), which is known to be anDng the
"physiological activators" of these enzymic activities.
Our results also indicate qualitative differences concerning the
PKC of the normal and v-ras-transformed cells which are difficult
to directly correlate with the v-ras gene expressed proteins. In
fact the results shown in the Table I indicate that the more
acidic enzymic population, associated to the fraction eluted by
350 mMNaCl,is almost undetectable on the cytoplasm of transfonmed
cells. Several possibilities could explain this observation: in
the cytoplasm of transforrmed cells a same_ protein could he present
in a less phosphorylated form and this, in turn, because both of
less active auto or hetero-phosphorylating activities and of
more active phosphatase ones. ~]t we have to take to account that
1101
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
at least five isoenzyme forms of PKC have been described and
cloned from several cell systems (13). The differential gene
expression of one isoenzyme compared to others might also explain
this result.
The PKC in the v-ras transformed cells displays a higher
sensitivity to Ca 2+ activation (Fig.l). This suggests some
structural differences between the regulatory d~nains of the
enzymic proteins. Moreover, other differences that we think cannot
be directly correlated to the viral p21 proteins,are indicated by
the results of the experiments in vivo with TPA: t~y indicate
that PKC's from treated cells are less promptly down regulated,
after the TPA activation and translocation, likely through the
activation of proteolytic processes. Whatever the molecular
mechanisms underlying this phenomenon in the transformed cells
(less active proteases,structural differences in the enz~nic
proteins which make them more resistant to proteolytic digestion
or a combination of both) they cannot be directly related to the
viral p21 proteins.
It has been already described that ras-virus infection and
transformation is characterized both by the overexpression of
point mutated p21 proteins and by the activation of phospholipase
C (and than PKC) (14). Our finding of an increased PKC activity in
ras-virus transformed epithelial cells is to us very interesting
in consideration of the biological peculiarities of t~
investigated model (PC C13 Ha MSV): the ras-virus infected thyroid
epithelial cells do not respond any more to TSH (8) and have lost
all the biological responses (iodin uptake, thyroglobulin
synthesis...) triggered by TSH except the growing capacity; this
might be related to the fact that PC C13 Ha MSV cells have the
Protein Kinase C system in the "on" state too. On the other hand,
it is well known that TSH interaction with sensitive epithelial
thyroid cells results in the activation of a variety of amplifying
systems, one of them leading to P}~ activation (15-16): our
finding reinforces the hypothesis that PKC activation is a
critical moment in the TSH action on thyroidal cell growth.
1102
Vol. 157, No. 3, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Acknowledgment
This research was supported by grants from C.N.R.N. 87.01491.44
("Progetto Finalizzato Oncologia").
References
i. Fusco, A., Berlingieri, M.T., Di Fiore, P.P., Portella, G., Grieco M., and Vecchio, G. (1987) ~I. Cell Biology, 7,
3365-3370. 2. Barbacid, M. (1987) Ann. Rev. Biochem. 56, 779-827. 3. Spina, A., Di Donato, A., Colella, G., and Illiano, G.
(1987) Biochem. and Biophys. Research Ccm~., 142,
527-535. 4. Fleishman, L.F., Ch~wala, S.B., Contley (1986) Science,
231, 407-410. 5. Wakelam, M.J.O., Davies, S.A., Honoloy, M.D., ~.~lioy, I.,
Marshall, C.J., Hall, A. (1986) Nature, 323, 173-176.
6. Bor-Sagi, P., Feramisco, J.R. (1986) Science, 233,
1061-1068. 7. De La Pena, P., Moscat, J., Garcia-Barreno, P., Anderson,
P.S., Aaronson, S.A., and Lacal, J.C. (1987) 8. Ambesi-Inpi~nbato, F.S., Parks, L.A.M., and Coon, H.G.,
(1980) Proc. Natl. Acad. Sci. USA 77, 3455-3459. 9. Thomas P. Thcm~s, Gopalakrishna, R., and Anderson,
W.B.(1987) ~thod in Enzimology, 141, 397-411. i0. Bradford, M.M., (1976) Anal. Biochem. 72, 248-254 ii. Kraft A.S., Anderson, W.B. (1983) Nature, 307, 621-623. 12. Nishizuka, Y. (1984) Nature, 308, 693-698. 13. Ohno, S., Kawasaki, H., Imajho, S., and Suzuki, K.
Inagaki, M., Yokokura, H., Sakoh, T., Hidaka, H., Nature Voi.325, 161, 1987
14. C]liarugi, V.P., Porciotti, F., Pasquali, F., and Brm~i, P. (1985), Biochem. Biophys. Research. C~n., 132, 900-907.
15. Kohn, L.D., Aloj, S.M., Tombaccini, D., Rotella, C.M., Toccafondi, R., Marcocci, C., Corda, D., and Grollman, E.F. (1985), Biochemical Action of Hor~mm~s, Vol.12, 457-512, Academic Press, Orlando, FL.
16. Exton, J.H. (1985) Am. J. Physiol. 248, E633-E647.
1103