exp cell res- hansen
TRANSCRIPT
-
8/11/2019 Exp Cell Res- Hansen
1/10
EXPERIMENTAL CELL RESEARCH 192,587-596 (1991)
Differential Regulation of HSC70, HSP70, HSPSOcw, nd HSPSOP mRNA
Expression by Mitogen Activation and Heat Shock in Human Lymphocytes
LINDA K. HANSEN,~ J. P. HOUCHINS,? AND JAMES J. OLEARY*
Department of Laboratory Medicine and Pathology, and tlmmunobiology Research Center, Box 198 UMHC,
IJniuersity of Minnesota, Minneapolis, Minnesota 55455
A
subset
of heat shock proteins, HSPSOcr, HSPSO@,
and a member of the HSP70 family, HSC70, shows en-
hanced synthesis following mitogenic activation as well
as heat shock in human peripheral blood mononuclear
cells. In this study, we have examined expression of
mRNA for these proteins, including the major 70-kDa
heat shock protein, HSP70, in mononuclear cells follow-
ing either heat shock or mitogenic activation with phy-
tohemagglutinin (PHA), ionomycin, and the phorbol es-
ter, tetradecanoyl phorbol acetate. The results demon-
strate that the kinetics of mRNA expression of these
four genes generally parallel the kinetics of enhanced
protein synthesis seen following either heat shock or
mitogen activation and provide clear evidence that mi-
togen-induced synthesis of HSC70 and HSPSO is due to
increased mRNA levels and not simply to enhanced
translation of preexisting mRNA. Although most
previous studies have focused on cell cycle regulation of
HSP70 mRNA, we found that HSP70 mRNA was only
slightly and transiently induced by PHA activation,
while HSC70 is the predominant 70-kDa heat shock
protein homologue induced by mitogens. Similarly,
HSPSOa appears more inducible by heat shock than mi-
togens while the opposite is true for HSP908. These re-
sults suggest that, although HSP70 and HSC70 have
been shown to contain similar promoter regions, addi-
tional regulatory mechanisms which result in differen-
tial expression to a given stimulus must exist. They
clearly demonstrate that human lymphocytes are an
important model system for determining mechanisms
for regulation of heat shock protein synthesis in un-
stressed cells. Finally, based on kinetics of mRNA ex-
pression, the results are consistent with the hypothesis
that HSC70 and HSPSO gene expression are driven by
an IL-2/IL-2 receptor-dependent pathway in human T
cells.
@ 1991 Academic Press, Inc.
INTRODUCTION
Physiologic stress, such as heat shock, preferentially
enhances the synthesis of a limited number of intracel-
1 Current address: Department of Surgical Research, The Chil-
drens Hospital 300 Longwood Avenue, Boston, MA 02115.
To whom correspondence and reprint requests should be ad-
dressed.
lular proteins (heat shock proteins or HSPS).~ The re-
sponse has been observed in all cells so far tested and
some of the HSPs a re highly conserved across species
(reviewed in [l]). The most strongly heat-inducible and
conserved HSPs found in eukaryotic cells are proteins
of about 90 kDa (HSPSO) and 70 kDa (HSP70). How-
ever, HSPSO is also constitutively expressed in un-
stressed cells, and all cell types that have been exam-
ined have constitutively expressed homologues of
HSP70, including the glucose-reactive protein, GRP76
in human cells, and a structural and functional homo-
logue of HSP70, which is less strongly induced by heat
shock and has been designated HSC70 in human cells.
HSPSO and these homologues of HSP70 are abundant
intracellular proteins and it is clear that these proteins
must have important roles in normal cell function in
addition to whatever role they may play in cellular adap-
tation to stress.
However, while the regulation of HSPSO and HSP70
synthesis and gene expression by heat shock has been
well studied, little is known of the regulation of these
gene products in unstressed cells. We have previously
reported [2] that mitogen ac tivation of human periph-
eral blood mononuclear cells in culture results in strong
and sustained preferential enhancement in the synthe-
sis of HSPSO and HSC70 during the prereplicative in-
terval, which is a period of about 24 h following mitogen
addition but prior to entry of the activated cells into S
phase. Human HSPSO actually consists of at least two
proteins, Q and 0, each encoded by its own gene [3] and
appears to be involved in a number of importan t intra-
cellular processes in unstressed cells including serving
as a cytoplasmic shuttle protein for growth-related mac-
romolecules such as steroid hormone receptors [4] and
the protooncogene product src-tyrosine kinase [5].
Compared to human HSP70, HSC70 is only moderately
heat inducible, but is 83% homologous at the amino acid
level with HSP70 [6]. In eukaryotes HSC70 appears to
be involved in a number of intracellular processes im-
portant for cell growth, such as uncoating of clathrin
3 Abbreviations used: PHA, phytohemagglutinin; PBLs, peripheral
blood lymphocytes; HSP, heat-shock protein; IL-2, interleukin-2;
TPA, tetradecanoyl phorbol acetate.
587
0014.4827/91 3.00
c
opynght Q 1991 by Academic Press, Inc.
All rights o f reproduction in any form reserved.
-
8/11/2019 Exp Cell Res- Hansen
2/10
588
HANSEN, HOUCHINS, AND OLEARY
triskelions from endocytotic vesicles [ 71 and transloca-
tion of nascent polypeptide chains across organelle
membranes [8, 91.
The enhanced expression of HSC70 and HSPSO ob-
served after mitogen activation of human lymphocytes
shows marked differences from the enhanced expres-
sion of these gene products associated with the heat
shock response. Heat shock of human cells in culture
leads to enhanced expression of HSP70 and HSPSO
mRN A and enhanced synthesis of the proteins which
can be detected within 15-30 min of the heat shock [3,
lo]. However, after stimulation of human peripheral
blood lymphocytes with the mitogen phytohemagglu-
tinin (PHA), which activates primarily T cells, preferen-
tial enhancement in HSC70 and HSPSO synthesis does
not occur until about 8 h following PHA addition reach-
ing a stable maximum by about 12-18 h of culture with
PHA [ 111. Unlike the heat shock response, we observed
no detectab le increase in HSP70 synthesis during the
prereplicative interval or even as the activated cells be-
gin to enter S phase. By contrast, a few studies have
reported enhanced expression of HSP70 in serum-
stimulated HeLa cells [lo, 121, but the enhancement
appears less than that which we observed for HSC70
and HSPSO in lymphocytes and these findings may not
be relevant for nontransformed cells. Some enhance-
ment in HSP70 synthesis and mRNA expression has
been seen in purified T lymphocytes activated with
PHA [13], but this study did not examine the HSC70
gene product, and again the enhancement appears small
compared to that observed for HSPSO and HSC70 in
PHA-activated mononuclear cells. One problem with
such studies of mitogen activation in human lympho-
cytes is that purified T cells lack the accessory cells, i.e.,
monocytes, required for optimal proliferative response
[14] and which are present at optimal levels for mito-
gen-induced proliferation in the mononuclear cell prepa-
rations used in our previous work [a].
In fact, serum- or growth factor-induced expression
of HSC70 mRNA has not been previously examined in
human cells, although similar genes have been studied
in Drosophila [15] and rat cells [16] which, like human
HSC70, are expressed in the absence of heat shock [6,
111. Studies thus far find evidence only for serum induc-
tion of the HSPSOcr gene product in transformed human
cells [ 171. However, it appeared from two-dimensional
gel electrophoresis in our studies of T cells that both
HSPSO cu and /3 are mitogen inducible [II]. Thus, the
goal of the current study was to begin to clarify the
mechanisms responsible for the marked enhancement
in HSC70 and HSPSO protein synthesis observed follow-
ing PHA activation of human mononuclear cells. The
basic question was whether the enhanced synthesis of
these proteins reflects enhanced levels of mRNA ex-
pression, implying growth factor-modulated enhance-
ment in gene expression, or is primarily due to enhanced
translation without an increase in steady-s tate mRNA
levels. In addition, we sought to clarify whether mitogen
activation might result in some enhancement in HSP70
mRNA expression as implied by the study of purified T
cells [13]. The results are consistent with a slight, but
transient up-regulation of HSP70 mRNA abundance
during the prereplicative interval of human T cells.
However, they demonstrate clearly that HSC70 is the
predominant 70-kDa HSP induced following mitogen
activation both at the level of mRNA abundance and
protein synthesis. Because the kinetics of expression of
HSC70 mRNA are dramatically different from that of
HSP70 following mitogen activation and heat shock, it
seems clear that despite great functional similarities the
genes for these products must contain quite different
regulatory elements. The results also show some degree
of differential modulation of HSPSOcu and HSPSOP
mRNA abundance in heat shock compared to mitogen
response, but both forms are clearly enhanced in the
two responses, implying a greater degree of regulatory
homology at the gene level. More importantly, these re-
sults are perhaps the strongest support so far reported
for the hypothesis that HSPSO and HSC70 expression
can be modulated in unstressed cells by a growth factor-
mediated pathway [ 181.
MATERIALS AND METHODS
Cell isolation and culture. Peripheral blood mononuclear cells
were isolated from healthy donor whole blood by centrifugation on
H&opaque (Sigma Chemical Co.), washed three times with Hanks
Balanced Salt Solution (GIBCO, Grand Island, NY), and resus-
pended at 2-5 X 10s cells/ml in 10% pooled hu man serum and RPM1
1640 medium (GIBCO). Phytohemagglutinin (PHA) (Wellcome
Diagnostics) was added at 1 pg/m l where indicated. Ionomyc in was
used at 5 fig/ml alone or 1 pg/m l when added with tetradecanoyl phor-
bol acetate (TPA). TPA alone was used at 50 or 100 rig/ml when
added with ionomycin. Heat stress consisted of incubating cell cul-
tures in 50.ml tubes in a water bath at 42 or 45C for 30 or 15 min,
respectively, followed by a recovery period at 37C. For protein label-
ing, cells were cultured in the presence of [3H]leucine (ICN) at a spe-
cific activity of 50 Ci/mmol and a dose of 100 &i/ml and [35S]methi-
onine (ICN) at a specific activity of 344 mCi/mmol and a dose of 50
pCi/ml.
cDNA probes. The HSP70 probe was a full-length cDNA con-
tained in plasmid, pH 2.3, a gracious gift from Dr. Richard Morimoto.
The HSC70 probe was a 500.bp fragment isolated from X-phage 7,
also a gift from Dr. Richard Morimoto. This 500.bp fragment was
derived from the 5 end of the HSC70 coding region, identified by in
uitro translation of larger cDNA. The HSC70 and HSP70 cDNA
probes exhibit no cross-reactivity under normal stringency condi-
tions. The HSPSOtu and B probes were the kind gift of Dr. Eileen
Hickey and were 900. and 800.bp cDNA fragments, respectively.
These cDNA probes are also non-cross-reac ting under normal strin-
gency conditions. The major histocompatibility Class I cDNA probe
was specific for the human B7 gene.
Subcellular f ractionation for protein synthesis. After cell culture,
cells were scraped f rom culture dishes with a rubber policeman and
washed three times with cold Hanks. The cell pellet was resuspended
in low detergent lysis buffer (0.1% Triton X-100, 150 mM KCl, 8 mM
MgCl,, 20 mM Tris, 1 mM PMSF, pH 7.5) at a concentration of 10
&lo6 cells. After a 15.min incubation on ice, the solution was spun
briefly in a microfuge to pellet nuclei. The supernatant was saved as
the cytoplasmic fraction.
-
8/11/2019 Exp Cell Res- Hansen
3/10
MITOGEN- AND HEAT-INDUCED EXPRESSION OF HSP mRNA
589
RNA isolation. RNA isolation was performed using a modifica-
tion of the Chirgwin method 1191. Briefly, after incubation , cells were
washed two times in cold Hanks, then lysed in 3 ml guanidinium
thiocyanate solution (4 M guanidinium thiocyanate [Fluka], 25 mM
sodium citrate, pH 7, 0.5% sarcosyl, 0.1 M 2-merceptoethanol). The
lysate was passed through a 22-gauge needle 6-10 times and layered
on a 1.5.m15.7 M CsCl cushion in Beckman polyallomer tubes. This
preparation was spun at 35,000 rpm for 18 h. The supernatant was
carefully removed and 100 ~1 TES (10 mM Tris-HCl, pH 7.6, 1 mM
EDTA, 5% sarkosyl) was added to the pellet which was allowed to
dissolve for 30 min. The buffer was removed and another 100 ~1 was
added to the tube for 30 min. The resulting 200.~1 sample was read at
A,,, and A,, to determine concentration and sample purity.
SDS-polyacrylamide ge:elelectrophoresis. Following cell fraction-
ation, counts per minute of label incorporated into the cytoplasmic
fractions were determined by scintillation counting . Equal counts per
minute of cytoplasmic samples were added to each lane in a given gel
and separated by SDSPAGE using the method of Laemmli [20].
Gels were dried on cellophane (Bio-Rad) on plastic frames (Idea Sci-
entific) and placed on Kodak X-Omat film for autoradiography.
Probe preparation. The cDNA sequences were cut from the plas-
mids with the appropriate restriction endonuclease and then electro-
phoresed in a known quantity on low melting point agarose (BRL)
gels. The cDNA fragment was excised from the gel, weighed, and
boiled 10 min. The amount of fragment present was estimated and
the solution was aliquoted into lo-fig samples and frozen at ~20C.
To label, aliquots were removed and boiled 2 min. The Multiprime
labeling kit (Amersham) was used to label the cDNA with [32P]dCTP.
A specific activity of 1 X lo9 cpmlwg was generally achieved.
Northern analysis. Northern analysis was performed as described
by Thomas 1211. Briefly, RNA samples were denatured 15 min at
60C in a 1X Mops, 50% formamide, 12% formaldehyde solution, and
then applied to formaldehyde-agarose denaturing gels (16% formalde-
hyde, 1X Mops, 1.5% agarose [BRL]). RNA ladder (BRL) was used
for molecular weight determina tions. After electrophoresis, the
RNA-containing gel was incubated at room temperature in a 0.05 N
NaOH solution for 30 min, followed with 2~ SSC for 30 min. The
RNA was transferred from the gel to GeneScreen (DuPont) using the
capillary blotting technique at least 20 h. The blot was washed briefly
in 1X SSC and then baked in V~CUOat 80C for 2 h.
Hybridization was performed as follows: B lots were prehybridized
in hybridization solution (1 M NaCl, 50% formamide, 15% dextran
sulfate, 5X Denhardts, and 1% SDS) for at least 4 h at 42C. Labeled
probe was boiled for 10 min in the presence of depurinated salmon
sperm DNA and added to the blot at 5 X lo cpm/ml hybridization
buffer. Hybridization took place at 42C for at least 16 h, after which
blots were washed two times in 2X SSC at room temperatu re, two
times in 2X SSC, 1% SDS at 6OC, and two times in 0.2~ SSC, 0.1%
SDS at room tempera ture. Blots were then resealed in a heat-sealable
bag with 1X SSC-saturated 3-mm Whatman paper and exposed to
Kodak X-Omat film with DuPont intensifying screens.
Blots were stripped by boiling for 2 min in 1% SDS followed by
cooling for 15 min. This was repeated, and autoradiography was per-
formed to assure that all the radioactive probe was removed. Blots
were reprobed as described above.
RESULTS
Specific mRN A Induction following PHA Activation
of
Human Mononuclear Cells
In Fig. 1, mRNA levels were compared by Northern
analysis of human mononuclear cells cultured for
various times after PHA addition using the cDNA
probes for HSC70, HSP70, HSPSOa, and HSPSOP de-
scribed above. cDNA for the major histocompatibility
HSC70
HSP9Oa
Class I
,. .-
-2.6 kb
-2.5
-2.7
-1.6
FIG. 1. Kinetics of HSP mRNA expression following mitogen ic
activation of mononuclear cells. Peripheral blood mononuclear cells
were PHA-activated and cytoplasmic RNA was isolated at the indi-
cated times following PHA addition. Levels of cellular HSC70,
HSP70, HSPSOtu, and HSP90fi mRNA were determined using the
cDNA probes described under Materials and Methods and the results
shown are for the same blot which was stripped and reprobed with
each cDNA. The major histocompatibility complex Class I probe was
used as a control to compare RNA loading in each lane. Molecular
weights (kilobases) of the bands are indicated to the right.
Class I gene, which is not mitogen-inducible in human T
lymphocytes , was used as a control. As shown in Fig. 1,
HSC70, HSPSOcY, and HSPSOP mRNA are detectable in
freshly isolated (0 h) cells. Enhanced HSC70, HSP90@,
and HSPSOB mRNA levels were observed between 4 and
8 h after PHA addition, increasing out to 24 h. HSP70
mRNA levels showed a slight enhancement only at 12 h
of culture, but return to baseline (0 h) levels by 24 h of
culture.
mRNA and Protein Induction following Heat Shock
To determine the kinetics of mRN A induction follow-
ing heat stress, mononuclear cells were heat shocked at
42C for 30 min, then returned to 37C (Fig. 2). Samples
were removed from the 37C water bath at the indicated
times for total cellular RNA isolation. While HSP70
mRNA is not detectable in the control cells incubated
continuously at 37C, immediately following the 30-min
heat shock a substantial amount of HSP70 mRNA is
present. E levated, but decreasing, levels are maintained
through 2 h post heat shock, and by 4 h HSP70 mRNA is
-
8/11/2019 Exp Cell Res- Hansen
4/10
590
HANSEN, HOIJCHINS, AND OLEARY
HSC70
-2.5
HSP90a
-2.95
HSP90P
-2.7
FIG. 2. Kinetics of HSP mRNA expression following heat shock.
The cells were heat shocked at 42C for 30 min and returned to 37C,
and total cellular RNA was isolated at the indicated times following
return to 37C. Northern analysis was performed as in Fig. 1 and the
results shown are for the same blot which was stripped and reprobed
with each cDNA. Although the data are not shown, mRNA levels
detected with the Class I gene probe were not affected by heat shock
and, relative to the induction seen with the heat shock gene probes,
reflected only slight differences in loading.
barely detectable. HSC70 mRNA is detectable in the
control cells, and slightly enhanced HSC70 mRNA lev-
els are seen immediately following the 30.min heat
shock, peaking at 30 min of incubation at 37C, and
decreasing to resting levels by l-2 h. Like HSC70,
HSPSOol mRNA is detectable in resting cells and
slightly enhanced immediately following the heat shock.
Maximal levels are achieved within 2 h after stress and
by 4 h, levels have returned to control. HSPSOP mRNA
shows a similar pattern of response to heat shock, al-
though maximal levels are seen slightly earlier.
The kinetics of heat shock-induced cytoplasmic pro-
tein synthesis are shown in Fig. 3. Following a 30 min
42C heat shock, the mononuclear cells were returned
to 37C and pulse labeled with radiolabeled amino acids
for the indicated intervals. After the labeling period,
cells were lysed and counts per minute of incorporated
label determined with equal counts per minute added to
each lane in the gel, followed by separation on SDS-
PAGE and autoradiography. As expected with addition
of equal amounts of label per lane the increases in the
heat shock bands are accompanied by decreased inten-
sity of bands corresponding to non-heat shock proteins,
particularly the major band corresponding to actin at
about 42 kDa. The exposure time for the autoradiogram
was adjusted to permit optimal visualization of the rela-
tive enhancement in heat shock protein synthesis.
Thus, the control lane is somewhat underexposed (No
HS).; however, as shown in Fig. 7 and previously [2,11],
HSP70 synthesis is undetectable in control cells even
with greater relative exposure of the autorad iograms.
As shown in Fig. 3, HSP70 synthesis is enhanced within
the first 30 min following stress. Synthesis reaches a
peak at l-2 h and returns to control levels by 446 h. By
contrast, HSC70 synthesis is detectab le in resting con-
trol cells and is enhanced 30-60 min following stress
and returns to control levels by 2-3 h. As with HSC70,
the HSPSO band, composed of both HSPSOtr and
HSPSOP, is detectably synthesized in resting cells. En-
hanced synthesis is observed 30-60 min following stress
with maximal synthesis seen at l-2 h. Baseline levels
are achieved by 4-6 h. Thus, the heat-induced enhance-
ment in protein synthesis parallels increases in specific
mRN A levels, although as expected the kinetics are
somewhat delayed.
Comparison of mRNA Induction following Heat Shock
and Mitogen Activation
The degree of specific mR NA induction following
heat stress or mitogen activation is compared in Fig. 4,
HSPSO-
HSC70-
HSP70
FIG. 3. Kinetics of protein synthesis in mononuc lear cells follow-
ing heat shock. Cells were heat shocked at 42C for 30 min, then
returned to 37C, and incubated with [3H]leucine and [%]methio-
nine for the indicated intervals prior to cell fractionation. Equal
counts per minute of incorporated label were loaded in each lane with
proteins separated by SDSPAGE followed by autoradiography. The
autoradiograph is shown. Molecular weight markers (kDa) are indi-
cated to the right and the positions of HSPSO (LYand B forms are not
distinguishable), HSC70, and HSP70 as shown in previous studies [2,
111 are indicated to the left.
-
8/11/2019 Exp Cell Res- Hansen
5/10
MITOGEN- AND HEAT-INDUCED EXPRESSION OF HSP mRNA
591
HSP70
-2.6 kb
HSP90a
HSPJN)P
-2.7
FIG. 4. Comparison of heat shock-induced and mitogen-induced
levels of HSP mRNA in mononuc lear cells. The cells were heat
shocked at 42C for 30 min or 45C for 15 min, and then placed at
37C for 1 h at which time total cellular RNA was isolated. In addi-
tion, RNA was isolated from lymphocytes which were incubated with
PHA for 1 2, 24, or 48 h. The results show a single blot which was
stripped and reprobed with each cDNA. Although not shown, varia-
tions in loading detected with the Class I gene probe were insignifi-
cant compared to the differences seen with the other probes.
which shows mitogen induction of HSP mRNA and the
induction of HSP mRNA after a 42 and 45C heat shock
with RNA isolated 1 h after the heat stress. In addition,
mRNA expression was examined out to 48 h after PHA
activation to see if the enhanced message levels are sus-
tained later in the cell cycle or only transiently ex-
pressed during the prereplicative, GO-S phase, transi-
tion. As shown previously [22], by 48 h following PHA
addition, most responding T lymphocytes have com-
pleted at least one round of division and are progressing
through the cell cycle with increasing asynchrony.
HSP70 mRNA is, as expected from the previous re-
sults, undetectable in resting cells. Heat stress at 42C
enhances abundant HSP70 mRNA, with even greater
enhancement at 45C. Consistent with Fig. 1, a very
faint HSP70 band is detectable at 12 h following PHA
activation, with the mRNA diminishing to undetectable
levels by 48 h. Again consistent with the previous re-
sults, HSC70 m RNA is detectable in resting cells and
less strongly induced than HSP70 mRNA by 42C heat
shock; however, the mRNA is strongly enhanced at
45C. PHA induced HSC70 levels at 12 h are signifi-
cantly enhanced over that seen in control (37C) cells or
42C heat-stressed cells. This increase in HSC70
mRNA is maintained out to 24 h, but is diminished at 48
h. The heat shock induction of HSC70 mRNA is compa-
rable to mitogen-induced levels only at the higher heat
shock temperature.
HSPSOn mRNA is detectab le in a very low quantity in
resting cells, and shows a lo- and 20-fold increase with
42 and 45C heat stress, respectively. After PHA addi-
tion, mRNA expression continues to increase out to 48
h. The level of 42C heat shock-induced mRNA is about
equivalent to that seen at 24 h after PHA addition, but
the amount of mRNA induced by severe heat shock ex-
ceeds that seen at any time point following mitogen ac-
tivation. HSP908 mRNA is also low in resting cells and
increases in response to heat stress, but its extent of
induction by heat stress is lower relative to that induced
by mitogen. Like HSPSOcu, mitogen-enhanced mRNA
levels persist at least 48 h following stimulation. Thus,
HSPSOcu appears to be more inducible by heat stress
than by mitogen, whereas HSPSOP shows greater mito-
gen enhancement.
Cytoplasm ic protein synthesis induced by different
degrees of heat stress is shown in Fig. 5. As in Fig. 3,
equal counts of incorporated label were added to each
lane. Prior to cell fractionation, aliquots of cells were
heat shocked followed by incubation with radiolabeled
amino acids for 2 h at 37C or the radiolabeled amino
acids were added and the cells were left at the heat
shock temperature for the 2-h labeling period. The 30-
min 42C heat stress followed by 37C recovery shows
the expected pattern of protein synthesis, with HSP70
showing greatest induction, along with HSPSO, HSC70,
and other minor heat shock proteins. A 2-h continuous
incubation at 42C shows essentially the same pattern.
In contrast, the 15-min 45C stress induces HSP70 and
HSC70 as expected, but the quantity of HSPSO protein
is diminished relative to that seen with the heat shock at
the lower temperature. Continuous incubation for 2 h at
45C inhibited any detectab le incorporation of radiola-
beled amino acids into proteins.
The HSP protein synthesis observed following the
15-min 45C heat shock stands in contrast to the
marked enhancement in mRNA levels observed with
this stress compared to the mRNA levels observed with
heat shock at 42C. Thus, while HSPSO mRNA levels
were more strongly induced by the 15.min 45C heat
shock, HSPSO protein synthesis declined, and HSP70
and HSC70 show about the same degree of enhanced
protein synthesis at 45C despite a marked increase in
mRN A levels with this heat shock condition.
Effects of Phorbol Ester and Ionomycin on HSP
Induction in Mononuclear Cells
Certain agents in addition to PHA can be mitogenic
for lymphocytes. Ionomycin, a calcium ionophore, and
TPA, a phorbol ester, can induce a partial mitogenic
-
8/11/2019 Exp Cell Res- Hansen
6/10
592
HANSEN, HOUCHINS, AND OLEARY
205kd
-116
96
66
-45
-29
FIG. 5. Protein synthesis following heat shock at different tem-
peratures. Mononuclear cells were heat shocked at 42C for 30 min or
45C for 15 min, and [H]leucine and [?S]methionine were added to
the culture medium. The cells were then either returned to 37C or
maintained at the heat shock temperature for the labeling interval.
The cells were then lysed and equal counts of incorporated label were
added per lane and separated by SDSPAGE followed by autoradiog-
raphy. The autoradiograph is shown: No heat shock, 37C for 2 h;
42C for 30 min, 37 for 2 h; 42C for 2.5 h; 45C for 15 min, 37C for 2
h: 45C for 2.25 h.
response alone or in combination by increasing intra-
cellular free calcium and activating protein kinase C,
respectively. mRNA induction by these agents was ex-
amined and compared to levels in resting cells and cells
activated by PHA, using concentrations of TPA (50 ng/
ml) and ionomycin (5 pg/ml) that separately induced
the greatest proliferative response, which was about 20-
25% of that seen with PHA. The optimal concentrations
when used together were 100 rig/ml and 1 pg/ml, respec-
tively, and this combination induced a proliferative re-
sponse about
55%
of the optimal PHA-induced re-
sponse (data not shown).
In these experiments, the cells were incubated with
the same concentrations of mitogens for 24 h in the pres-
ence or absence of serum, a required cofactor for the
proliferative response. In the presence of serum, HSP70
mRNA appears to show some slight enhancement in the
TPA- and PHA-activated cells (Fig. 6). In contrast,
HSC70 shows a similar, 5- to lo-fold induction, by iono-
mycin, TPA, or PHA, and both HSPSOo( and HSP90/3
also show strong induction with all treatments. The
pattern of protein induction in the presence of serum is
also similar with each stimulus (Fig. 7). As in Figs. 3 and
5, equal counts per minute of label incorporated into the
cytoplasmic factions was added per lane. Consistent
with our previous reports [2,11] the induction of HSC70
and HSPSO by PHA is not as striking as the relative
induction seen following heat shock . However, the pref-
erential enhancement with PHA activation occurs
against a general increase in protein synthesis during
the prereplicative interval, and synthesis of these pro-
teins accounts for more than 5% of the total protein
synthesized during this interval [2]. PHA added in the
absence of serum, although nonmitogenic [2], is suffi-
cient to induce the synthesis of both HSC70 and
HSPSO, while serum alone elicits no induction. TPA,
like PHA, is able to induce this set of proteins in both
the presence and the absence of serum as does the com-
bination of ionomycin and TPA. By contrast, ionomycin
added in the absence of serum gives a very different
pattern of protein synthesis with strong induction of a
band with molecular weight similar to another member
of the HSP70 family, BiP, or GRP76, which has been
shown to be induced by calcium ionophores in another
cell type [23]. However, the preferentially enhanced syn-
E g
2
%
= c
2 .P c .,
f
HSP70
-2.6 kb
HSC70
-2.5
HSP90u
HSP90P
-2.7
FIG. 6. HSP mRNA induction by ionomycin and phorbol ester.
Mononuclear cells were incubated in medium and 10% serum alone
(lane marked serum) or with ionomycin (5 pg/ml), TPA (50 rig/ml),
ionomycin (1 rg/ml) plus TPA (100 rig/ml), or PHA (1 Kg/ml). After
incubation for 18 h, total cellular RNA was isolated and Northern
analysis was performed. One blot was stripped and reprobed with
each cDNA. The Class I probe (data not shown) again revealed only
small differences in gel loading.
-
8/11/2019 Exp Cell Res- Hansen
7/10
MITOGEN- AND HEAT-INDUCED EXPRESSION OF HSP mRNA
593
-205kd
HSPSO-
HSC70-
FIG. 7. Protein synthesis induced by ionomycin and phorhol es-
ter. Mononuclear cells were placed in medium alone or medium p lus
10% serum and incubated with the indicated agents at concentrations
which gave the optimal proliferative responses, as described in the
text. lonomycin (i) was added at 5 pg/ml; TPA was added at 50
rig/ml; and when added together, the concentrations were 1 pg/ml
and 100 rig/ml, respectively, and PHA was used at 1 pg/ml. The cul-
tures containing serum are noted by +s. and serum = control
with no mitogens. The remaining cultures contained medium alone
with no serum added. Af ter 18 h in culture, cells were lysed and equal
counts from the cytoplasmic fractions were run on SDS-PAGE, fol-
lowed by autoradiography. The autoradiograph is shown.
thesis seen in Fig. 7 is clearly associated with a general
decrease in synthesis of most of the major bands seen in
the control lanes or in the lane with ionomycin in the
presence of serum. More severe forms of stress, e.g., the
more severe examples of heat shock shown in Fig. 5, also
induce a general suppression in protein synthesis and
thus, it seems likely that this result represen ts a toxic
effect due to an excessive free concentration of ionomy-
tin in serum-free medium which lacks serum proteins
which can bind this hydrophobic compound.
DISCUSSION
This paper is the first report to our knowledge to ex-
amine mitogen-induced HSC70, HSPSOcu, and HSPSOP
mRNA expression in human lymphocytes presented
with optimal conditions for cell growth and prolifera-
tion. The results show that HSC70 and HSP90a and /3
mRN A are significantly enhanced by mitogen activa-
tion in human periphera l blood mononuclear cells (Fig.
1). These enhanced mRNA levels correlate with the ki-
netics of enhanced HSC70 and HSPSO synthesis ob-
served following PHA activation (Fig. 7 and [2, 111). As
noted under Results, the preferential enhancement in
synthesis of the heat shock proteins appears more strik-
ing following heat shock (Figs. 3 and 5), but in part this
is due to the lower general level of protein synthesis in
quiescent cells which forms the background against
which this enhancement is visualized. Nevertheless,
with the major exception of the results following a brief
45C heat shock (F igs. 4 and 5), where protein synthesis
is not further enhanced (HSC70, HSP70) or declines
(HSPSO) despite greatly increased mRNA abundance,
differences in kinetics of protein synthesis with heat
shock are also generally mirrored by differences in ex-
pression at the level of mRNA. Taken together, these
results indicate that the preferentially enhanced pro-
tein synthesis is not simply due to enhanced transla-
tional activity of preexisting mRNA in mitogen-acti-
vated or heat-shocked mononuclear cells.
For HSC70 and HSPSO, enhanced mRNA levels were
observed within 8 h following mitogenic activation and
were sustained out to 24 h, which is just prior to entry
into S phase [22], and the increased levels are main-
tained beyond this prereplicative interval. By contrast,
HSP70 mRNA is only transiently increased at about 12
h after PHA addition. As expected, heat shock-induced
mRNA expression was seen for all four genes examined,
but unlike the PHA response enhanced expression is
sustained only 2-4 h following heat shock (Fig. 2). Thus,
even though HSP70 and HSC70 show 75% homology at
the DNA level and contain similar promoter regions
(See below and [6]), they exhibit very different kinetics
of expression to either stimulus.
Peripheral blood mononuclear cells consist of approx-
imately 5% monocytes, 10% B lymphocytes, and 85% T
lymphocytes, and the majority of the lymphocytes are in
a natural GO, or resting, stage of the cell cycle. Activa-
tion of the T cells by a mitogen like PHA in the presence
of adequate numbers of accessory cells initiates a com-
plex sequence of events as the cells progress through the
prereplicative interval. Perhaps the most well-charac-
terized events involve synthesis of the T-cell growth
factor, IL-2, and expression of IL-2 receptors on T cells.
Within 2-6 h after activation, CD4+ helper T lympho-
cytes in the presence of accessory cells begin to synthe-
size and secrete IL-2 and both the CD4+ and the CD8+
(cytotoxic/suppressor) T cells begin to express IL-2 re-
ceptors [24]. IL-2 then acts as a growth factor required
for lymphocyte progression from Gl into S phase [24,
251. As shown here, expression of HSC70 and HSPSO
mRNA appears to be up-regulated subsequent to this
initial increase in IL-2 and IL-2 receptor expression
(Fig. 1). Thus, the kinetics of enhanced HSC70 and
HSPSO expression are consistent with enhancement fol-
lowing IL-2/IL-2 receptor interaction.
This conclusion is also supported by the results of
Ferris
et
al. [ 131 using purified human T cells. As noted
above, isolated T cells lack the accessory cells which are
-
8/11/2019 Exp Cell Res- Hansen
8/10
594
HANSEN, HOUCHINS. AND OLEARY
required for IL-2 synthesis [ 141, and it is not surprising
that neither sustained synthesis of HSPSO nor induc-
tion of HSP7O mRNA was seen in primary response to
PHA by these investigators. Some increase in HSP70
protein synthesis was observed in response to PHA, but
they found no induction of HSP70 mRNA by PHA or
phorbol12-myristate 13-acetate. HSPSOprotein synthe-
sis (the (Yand /3 orms were not distinguished) was found
to be rapidly and transiently induced by PHA with ki-
netics of expression different from HSP70. However,
while IL-2 synthesis requires accessory cells, PHA can
induce IL-2 receptor expression in T cells depleted of
monocytes [22]. When IL-2 was added subsequent to
PHA addition, Ferris et al. [ 131 observed some enhanced
HSP70 mRNA and protein expression and a sustained
induction of HSPSO synthesis, similar to the kinetics of
HSPSO synthesis we have observed in PHA activated
mononuclear cells [ 111. Although HSC70 induction was
not examined by Ferris et al. [ 131, these results clearly
support the hypothesis that HSC70 and HSPSO synthe-
sis in T cells is driven by the IL-2/IL-2 receptor interac-
tion.
In contrast to these results with purified T cells, we
have been unable to demonstrate any increases in
HSP70 protein synthesis following PHA activation of
the T cells present in human mononuclear cells [2, 111.
All that we have been able to observe for HSP70 per se is
the very transient increase in mRN A levels at about 12
h after PHA addition shown in the current study (Fig.
1). In fact, Kaczmarek et al. [26] and Ida and Yahara
[27] demonstrated diminished HSP70 mRNA and pro-
tein expression, respectively, in PHA-activated lym-
phocytes. The different results observed in these studies
may be due at least in part to the existence of up to 10
genes in the HSP70 family [28] and lack of standardized
nomenclature or use of different gene probes for HSP70
may explain some of these apparently conflicting re-
sults. However, taken together it seems clear that
HSP70 synthesis and mRNA expression is only weakly
inducible by T-lymphocyte mitogens, and the constuiti-
vely synthesized homologue, HSC70, is the predomi-
nant form induced in this growth factor response.
HSP70 and HSPSO expression have also been exam-
ined in the growth response of other cell types. HSP70
and HSPSOcY mRNA expression have been shown to be
enhanced in HeLa cells following serum addition [lo,
121. c-myc can also induce HSP70 expression in mam-
malian cells [29] and Ela adenov irus infection, which
leads to increased expression of cellular gene products
involved in cell growth, has been reported to induce ex-
pression of HSP70 and HSPSOa, but not HSPSOP, in
HeLa cells [3, 171. On the other hand, Kao et al. [30]
report an increase in HSP70 protein synthesis following
S phase in HeLa cells rather than in the Gl interval
following serum addition. Regard less of this apparent
conflict, most studies with other human-derived cells
have focused on expression in transformed cell lines.
Such cell lines exhibit proliferation that is relatively in-
dependent of growth factors and serum deprivation gen-
erally does not lead to growth arrest in a true GO state.
Thus, responses of growth-arrested transformed cells to
serum addition may not accurately reflect the effects of
serum or growth factor addition on cell cycle progres-
sion in nontrans formed cell types.
By contrast, it is clear from the current results that
up-regulation of HSC70 and the HSPSO genes is part of
the sequence of events involved in the prereplicative
interval of human T cells and these findings suggest
that similar modulation may be a common feature of the
growth factor response of other nontransformed human
cell types. Unlike the studies of transformed cells, again
there is not much evidence for modulation of HSP70
levels during the proliferative cycle of T cells and, while
some studies have reported serum induction of
HSPSOa, but not HSPSOP, in HeLa cells [3, 171, our
results indicate that both the (Y and the /3 forms are mi-
togen inducible and in fact, HSPSOP mRNA exhibits
greater PHA-stimulated induction, relative to induction
by heat shock, than HSPSOn mRNA (Fig. 6).
HSC70, HSP70, HSPSOcu, and HSPSOB mRNA were
also induced by ionomycin and TPA in the current
study, both of which are only partial mitogens for pe-
ripheral blood lymphocytes [31]. Calcium ionophores
such as ionomycin or A23187 induce the release of in-
tracellular calcium stores, one of the events in the mito-
genie pathway. On the other hand, the phorbol esters,
such as TPA, activate protein kinase C [25]. In the
current results, ionomycin at an optimal concentration
induced a proliferative response that was 26% of that
with PHA, while TPA gave a 20% response, and with
both reagents the response was only 55% of that seen
with optimal PHA dose (data not shown). Although
these agents elicited only a partial mitogenic response
compared to PHA, both ionomycin and TPA induced
expression of the same set of heat shock mRNA seen
with PHA and expression was enhanced to about the
same extent with each stimulus (Fig. 6). An intriguing
finding was that enhanced synthesis of these proteins
by PHA and TPA does not require the presence of
serum in the culture medium. As shown previously [ll],
serum alone does not affect HSP synthesis, and serum
supplies transferrin, a required cofactor for T-cell prolif-
erative response. Thus, these results demonstrate that
induction of the heat shock gene products by PHA or
TPA in the absence of serum is not by itself sufficient to
induce the cells to enter the cell cycle. By contrast, iono-
mycin in the absence of serum induced the synthesis of a
protein with a molecular weight similar to that expected
for the glucose reactive HSP70 homologue, BiP or
GRP76, which has been shown to be induced by iono-
mycin in another cell type [23]. As noted under Results,
however, the enhanced synthesis of this protein was ac-
companied by marked suppression of other synthesized
bands, suggesting that in the absence of serum binding
-
8/11/2019 Exp Cell Res- Hansen
9/10
MITOGEN- AND HEAT-INDUCED EXPRESSION OF HSP mRNA
595
proteins the cells were subjec t to excessive concentra-
tions or the drug. Thus, it seem s likely that the appar-
ently enhanced synthesis of GRP76 in this experiment
may be an artifact of cell injury rather than due to spe-
cific induction of this gene product.
There is growing evidence for at least two distinct
promoters in the 5 untranslated region of some HSP70
genes which is consistent with the hypothesis that gene
activity may be modulated by growth factor ac tivation.
All heat shock genes contain a highly conserved heat
shock consensus element (HSE) which is responsible
for heat shock induction, and an additional region has
been identified as the serum-responsive element in the
HSP70 gene of HeLa cells along with a CCAAT box
which has been observed in a number of eukaryotic
other gene promoters [32, 331. An HSE and a CCAAT
box have also been identified in the 5 untranslated re-
gion of the human HSC70 [6]. Despite these indications
that HSC70 and HSP70 may have similar promotor re-
gions, as shown in the current study the kinetics and
levels of expression differ dramatically in response to
heat shock or PHA in human lymphocytes. Thus, there
must exist additional regulatory elements which are re-
sponsible for this differential expression, such as, as yet
undefined differences in the promotor regions, trans-
acting enhancer elements, or differences in mRNA sta-
bility. The existence of possible growth factor-respon-
sive promotors in the HSPSO genes has yet to be deter-
mined, but based on the current results it seems clear
that similar non-HSE promotor regions must be pres-
ent to explain the very significant and sustained in-
creases in HSPSO mRNA expression we observe, and
again additional regulatory elements must be involved
to explain the differential enhancement in HSPSOa and
HSPSOP mRNA, following mitogen activation and heat
shock.
Only two instances were observed in this study where
protein expression does not parallel mRNA expression
and which might indicate control of HSP synthesis at
some level other than mRNA abundance. First, in com-
paring mild (42C) to severe (45C) heat stress, mRNA
expression increases dramatically for each gene at the
higher temperature (Fig. 4), but protein synthesis is not
enhanced or diminished (Fig. 5). Similar to these re-
sults, the 90-kDa heat shock protein expressed in HeLa
cells is synthesized at 42C but not 45C [34]. It has
been proposed that one of the toxic effects of severe
heat stress is inhibition of mRNA translation [35]. On
the other hand, it has been reported that at tempera-
tures greater than 42.5C, mammalian cells are unable
to splice introns from unprocessed messages [36]. Be-
cause the HSP70 gene contains no introns, and introns
are present in HSC70 and both HSPSO genes [6,12,37],
it is possible that the diminished HSC70 and HSPSO
synthesis we observed at the higher heat shock tempera-
ture could be accounted for by this mechanism. How-
ever, our Northern blots do not show evidence for un-
processed message after the 45C heat shock and in fact
show enhanced levels of what appears to be mature
mRNA. Alternatively, studies of RNA metabolism fol-
lowing heat stress indicate that the processing and
transport to the cytoplasm of non-heat shock message
appears to be inhibited, while the transcription of these
genes remains unaffected [38]. As in our study, how-
ever, the accumulation of heat shock mRNA is ob-
served, and it has been postulated that heat shock
mRNA may represent a class of mRNA which main-
tains normal p rocessing during stress [38]. Thus, our
results are not consistent with inhibition of HSP pro-
tein synthesis due to inhibition of splicing or mRNA
transport and processing following severe heat shock
and the most likely mechanism leading to diminished
protein synthesis following severe heat shock in lym-
phocytes is direct inhibition of mRNA translation.
The second instance in which protein synthesis does
not parallel mRNA expression is in the later stages of
mitogen activation. Previous work has shown sustained
synthesis of HSC70 and HSPSO gene products as late as
48 h following PHA addition [ll]. However, HSC70
mRNA is diminished by 48 h after PHA addition, while
HSPSOa and fi mRNA remain enhanced at this time
point (Fig. 4). Thus, enhanced HSC70 m RNA transla-
tion or some other regulatory process may be required
to sustain HSC70 synthesis at later times and this result
indicates that there may be a separate down-regulatory
signal which operates on HSC70, but not HSPSOol or p,
during S phase.
We thank Drs. Richard Morimoto and Eileen Hickey for providing
us with cDNA probes, and Dr. Helen Hallgren for critical review of
the manuscript. This work was supported
by
NIH Grants AG 02338
and CA 39692 to James J. OLeary and a grant from the University of
Minnesota Graduate School to Linda K. Hansen. Portions of this
manuscript will be a part of the doctoral thesis of Linda K. Hansen to
fulfill the requirements of the University of Minnesota Graduate
School.
REFERENCES
1.
2.
3.
4.
5.
6.
I.
8.
9.
Lindquist,S.
(1986)
Annu.
Reu.Biochem. 55,1151~1191.
Haire, R. N., Peterson, M. S., and OLeary, J. J. (1988a) J. Cell
Biol. 106,883-891.
Hickey, E., Brandon, S. E., Sadis, S. Smale, G., and Weber, L. A.
(1986)Gene 43,147-154.
Catelli, M. G., Binart, N., Jung-Testas, I., Renoir, J. M., Baulieu,
E. E., Feramisco, J. R., and Welch, W. J. (1985) EMBO J. 4,
3131-3135.
Opperman, H., Levinson, W., and Bishop, J. M. (1981) Proc.
Natl. Acad. Ski. 1JSA 78, 1067-1071.
Dworniczak, B., and Mirault, M-E. (1987) Nucleic Acids Res. 15,
5181-5197.
Ungewickel, E. (1985) EMBO J. 4,
3385-3391.
Deshaies, R. J., Koch, B. D., Werner-Washburne, M., Craig,
E. A., and Schekman, R. (1988) Nature (London,J 332,800-805.
Chirico, W. J., Waters, M. G., and Blobel, G. (1988) Nature (Low
don) 332.805~810.
-
8/11/2019 Exp Cell Res- Hansen
10/10
596
HANSEN, HOUCHINS, AND OLEARY
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Wu, B. J., and Morimoto, R. I. (1985) Iroc. N&l. Acad. Sci. USA
82, 6070-6074.
Haire, R. N., and OLeary, ,J. J. (1988h) Exp. Cell Rex 179 , 65-
78.
Hickey, E., Brandon, S. E., Smale, (i., I,loyd, D., and Weher,
L. A. (1989) Mol. Cell. Viol. 9, 2615-2626.
Ferris, D. K., Harel-Bellan, A., Morimoto, R. I., Welch, W. J..
and Farrar, W. L. (1988) Proc. Natl. Acad. Ski. 1lSA 85, 3850-
3854.
Chalita, T. A., Schwartz, D. H., Miller, R., and Geha, R. S. (1987)
Clin. Immunol. Immunopathol. 44, 235-247.
Ingolia, T. D., and Craig, E. A. (1982) Proc. Natl. Acad. Sci. lfSA
79, 5255529.
OMalley, K. O., Mauron, A., Barchas, J. D., and Kedes, I,.
(1985) Mol. Cell. Riol. 5, 3476-3483.
Simon, M. C., Kitchner, K., Kao, H-T., Hickey, E., Weller, I,.,
Voellmy, R., Heintz, N., and Nevins, J. R. (1987) Mol. Ccl/. Rio/.
7, 2884-2890.
Pledger, W. J., Stiles, C. D., Antoniades, H. N., and Scher, C. D.
(1978) Proc. Natl. Acad. Sci. liSA 75, 2839-2843.
Chirgwin, J. M., Przyhyla, A. E., MacDonald, R. J., and Rutter,
W. J. (1979) Biochemistry
18, 5294-5299.
Laemmli, 17. K. (1970) Nature (London) 277, 680-685.
Thomas, P. (1980) t-roc. Natl. Acad. Sci. 1iSA 77, 5201b5205.
OLeary, J. J., Hanrahan, 1,. R., Mehta, C., and Rosenherg, A.
(1980) Cell Tissue Kinet. 13, 41-51.
Resendez, E., Jr., Attenello, J. W., Grafsky, A., Chang, C. S., and
Lee, A. S. (1985) Mol. Cell. Biol. 5, 1212-1219.
24.
25.
26.
27.
28.
29.
30.
31.
32.
:x1.
:34.
35.
36.
37.
38.
Reed, J. C., Alpers, J. D., Nowell, P. C., and Hoover, R. G. (1986)
Proc. Natl. Acad. Sci. (ISA 83, 3982-3986.
Isakov, N., Scholz, W., and Altman, A. (1986) Immunol. Today
7, 271-277.
Kaczmarek, I,., Calabretta, B., Kao, H-T., Heintz, N., Nevins, J.,
and Baserga, R. (1987) J. Cell Biol.
10,4, 183-187.
Iida, H. and Yahara, I. (1984) J. Cell Biol. 99, 199-207.
Mues, (;. I., Munn, T. Z., and Raese, J. D. (1986) J. Biol. Chem.
261,874-877.
Kingston, R. E., Baldwin, A. S., and Sharp, P . A. (1984) Nature
(London) 3
12,280-282.
Kao, H-T., Capasso. O ., Heintz, N., and Nevins, J. R. (1985)
Mol. Cell. Biol. 5, 628-633.
Domenico, D., Greaves, M., Villa, S., and DeBraud, F. (1984)
Eur. J. Immunol. 14, 720-724.
Wu, B. ,J., Kingston, R. E., and Morimoto, R. I. (1986) Proc.
Nntl. Acad. Sci. [IsA 83, 629-633.
Morgan, W. D.. Williams, G. T., Morimoto, R. I., Greene, J.,
Kingston, R. E., and Tjian, R. (1987) Mol. (ell. Biol. 7, 1129-
1 1:%x
Hatayama, T., Honda, K-I., and Yukioka, M. (1986) Biochem.
Hiophys. RPS. Commun. 137, 957-963.
Welch, W. J.. and Mizzen, 1,. A. (1988) J. ((~11Hiol. 106, 1117-
11:io.
Yost, H. J., and I,indquist, S. (1986) (~11 45, 185-193.
Rebhe, N. F., Ware, ,J., Bertine, R. M., Modrick, P., and Stafford,
D. W. (1985) (;ene 53, 235-245.
Sadis, S.. Hickey, E., and Weher, L. A. (1988) J. C&I. Ph.y.siol.
135, 377-386.
Received July 2, 1990
Revised version received September 24, 1990