0022-1767/85/1346-4087~02.00/0
THEJOURNAL OF I M M U N O L ~ ~ ~ Y
Copyright 0 1985 by The American Associatlonof Immunologists
Vol. 134.No. 6, June 1985
Printed In U.S. A.
INCREASEDPHOSPHATIDYLINOSITOLMETABOLISM
IS AN IMPORTANTBUT
NOTAN OBLIGATORYEARLYEVENT
IN B LYMPHOCYTEACTIVATION'
STEPHAN A. GRUPP
AND
JUDITH A. K. HARMONY2
From the Department of Anatomy and Cell Biology, Divisionof Immunology in the Department of Medicine. Universityof
Cincinnati. College of Medicine, 231 Bethesda Avenue, Cincinnati,
OH 45267
The phosphatidylinositol (PI) response has been
implicated in membrane signaling and cell activation. The role of phospholipid metabolism among
the early events in B cell activation has not been
clear. We have treated murine B cells with anti-Ig
antibody and lipopolysaccharide (LPS)and have
found that, although anti-IgM induces the PI response, LPS does not. The increase in metabolic
labeling of PI is specific to PI, and not the phosphatidylinositols. Anti-IgM unresponsive B cells from
CBA/N mice, which may correspond to a specific
functional subpopulation of normal B cells, do not
increase PI metabolism in response to anti-IgM, nor
do they undergo blastogenesis or DNA synthesis.
Moreover, when these deficient B cells are given a
stimulus sufficient to drive them into S (LPS + antiIgM), there is stillno corresponding activation of PI
metabolism. These results are consistent with a
two-signal model of xid B cell activation by antiIgM. One very early signal primes the cells but does
not induce the PI response. A second early signal is
supplied by LPS. This signal sustains cells in the
activated state, allowing them to receive yet other
signals to proceed through GI and progress further
along the cell cycle. A similar sequence of events
may occur in the normal B cell, with the first signal
provided by priming with anti-IgM,and the second
signal, the PI response, supported by a sufficiently
high dose ofanti-IgM to induce PI turnover and
maintain the cell inGI.
and its phosphorylated derivatives, phosphatidylinositol4-phosphate (PIP)and phosphatidylinositol-4,5-bisphosphate (PIPP),to the inositol phosphates anddiacylglycerol
(DAG) is catalyzed by phospholipase C. PI resynthesis is
accomplished from DAG via the intermediate phosphatidic acid (PA). DAG has been implicated a s a second
messenger inthat it activates protein kinase C (2). which
itself is then further involved in catalyzing phosphorylation of membrane proteins. A role for PA, another possible second messenger, has been proposed in transmemSupport for this suggestion
brane flux of calcium ions (3).
has come from experimental evidence that PA can behave
a s a calcium ionophore (4). These observationsare consistent with the idea that PI itself and its breakdown
products are the most functionally significant components of this system. Other,more recent work, points to
a link between a breakdown product of PIPP, inositol
triphosphate, and mobilization of calcium ions from intracellular stores (5-7). This latter suggestion would accord the principai functional position in the PI system to
PIPP, the bisphosphorylated derivative. The role of PIPP
in the PI system is a central focusof the work currently
being done in the area of early membrane phospholipid
events. Of considerableinterest in this regard is the
demonstration that thetransforming proteinspp6Ov-src
(8)
and p68v-ros (9)
have PI kinase activity.
Early enhancementof PI metabolism upon stimulation
with phytohemagglutinin occurs in human lymphocytes
(10,11) and lymphocyte membrane fractions (12). This
PI response occurs in the T cell and not the B cell or the
The increase of phospholipid metabolism, specifically monocyte, because depleting the lymphocyte population
the metabolism of phosphatidylinositol (PI)3,is a n impor- of B cells and adherentcells enhances thePI response by
tant early event in stimulationand mitogenesis in a reducing the background contribution to lipid metabolic
number of cellular systems ( 1).PI is a central element in labeling (13). Finding no increase in PI metabolism in
a proposed activation mechanism of membrane events murine B cells after lipopolysaccharide (LPS)or pokeweed
mitogen stimulation led Betel et al. (14) to suggest that
and putative second messengers. The hydrolysis ofPI
the PI response is not a significant early event in the
activation of B lymphocytes. More recently, Chien and
Received for publication November 29, 1984.
Ashman (15) reported increases incholine incorporation
Accepted for publication February 5, 1985.
into phospholipids upon pokeweed mitogen stimulation
The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked of B cells, but this was characterized a s a much later
adoertisernentin accordance with 1 8 U.S.C. Section 1734 solely to indievent, reaching its maximum at 1 to 2 days.
cate this fact.
We have used enriched populationsof mouse splenic B
' This work was supportedby Grant HL27333 from theNational Institutes of Health and Grant BC-434 from the American Cancer Society. S.
cells
treated with LPS and anti-immunoglobulin antiA. G. is supported by NIH Molecular and Cellular Biology Training Grant
HL07527. J. A. K. H. is a n Established Investigatorof the American Heart bodies (anti-IgM) to explore PI metabolism a s a n early
Association.
event in B cell activation, a s both anti-IgM and LPS have
ZAddress correspondence to: Judith A. K. Harmony, Department of
been shown to be polyclonal activators of B cell proliferAnatomy and Cell Biology (ML 521). Universityof Cincinnati. College of
ation (16). Thereis good indication that LPS does not act
Medicine, 231 Bethesda Ave., Cincinnati. OH 45267.
Abbreviations used in this paper BCGF. B cell growth factor: DAG,
via a cell surface receptor. LPS and its smallest active
diacylglycerol: PA, phosphatidic acid: PC. phosphatidylcholine: PI, phosfragment, Lipid X, can insertthemselves directly into the
phatidylinositol: PIP. phosphatidylinositol 4-phosphate: PIPP, phosphatldylinositol4.5-bisphosphate;xid,X-linked immunodeficient.
membrane, thereto effect changes in membrane proteins
4087
4088
THE PHOSPHATIDYLINOSITOL RESPONSE AND B CELL ACTIVATION
[ 171. Anti-IgM. on the other hand, is thought to mimic using 2 ml of scintillation coctail 4a20 (Research Products lnternaA s well as determining cpm 32Piin each spot,total 32Piin all
antigen binding, inthe sense thatit too binds to surface tional).
phospholipids was totalled. Each experiment was done in triplicate
Ig and results in activation [ 18).In this report we examine and means were taken. Percentage of total phospholipid 32Pi found
the changes inphospholipid metabolism caused by stim- in a given phospholipid was calculated a s (mean of individual phosulation with LPS and anti-IgM polyclonal antibodies in pholipid cpm/mean of total phospholipid cpm) x 100. Phosphostandards added to theextracted lipids were visualized by developing
the murineB cell system, with particular attention to the
the plates in an
iodine tank for 10 min.
behavior of PI, its phosphorylated derivatives PIP and
Cell size characterization. Enriched B cellswere cultured in 24PIPP, PA,and phosphatidylcholine [PC).We also consider well cluster plates (Corning), 1 ml/well a t 1 x lo6 cells/ml.with
activators and with and without 15% EL-4 conditioned medium
the time courseof phospholipid changes inthe B cell.
containing BCGF but not IL 2 activity. The cells were then resusAdditionally, we examine another aspect of phospho- pended a t various times, and
200 pl of the cells were added to 9.8 ml
lipid behavior as it relates to B cell activation processes. Isoton for cell size measurements on a Coulter Counter/Channelizer.
data on cell size thus generated were passed to a n Apple IIe
It has been shown that B cells from X-linked-immuno- The
microcomputer for analysis andstorage.
deficient (xid)CBA/N mice are unresponsive to a variety
Long-term culturesfor DNA synthesis assays.Enriched B cells
of polyclonal activators, including type 1 antigens (19) were culturedin 96-well microtiter plates (Corning)in complete RPMt
and anti-IgM (20). This unresponsiveness is thought to plus 10%FBS. Cell concentrations rangingfrom 2.5 x lo5to 2 x lo6
cells/ml were achieved in 0.22 ml in each well, including 20-pl
be due to the lack
of a subpopulation of B cells, the samples of the varlous activators. Cultures
were incubated for 72 hr
subpopulation expressing the Lyb-5 surface antigen (2
1, a t 37°C with 1pCi (3H-methyl)thymidine (3H-TdR)added 10 hrbefore
fiber filters, and the amount
22). TheLyb-5+cells are proposed as theresponder cells harvest. Cells were harvested onto glass
incorporation was measuredin a scintillation counter as described
to anti-Ig and traditional T-independent antigens, and of
above.
are subject to regulation by soluble factorssuch as B cell
Polyclonal activators. Various Bcell activators were used, includgrowth factor [BCGF)and interleukin 1 [ 18).Lyb-5- cells ing E. coli LPS 055:b5. (W) (Difco, Detroit, MI] plus whole goat antimouse IgM ($-chainspecific) andits
F(ab’), fragment (Cooper
respond to T-dependent antigens and are considered to Biomedical,
Malvern, PA]. Fab fragments were obtained by papain
be under major histocompatibility complex-restricted digestion of the whole antibody followed by isolation of fragments
control (23, 24). We consider the question of whether xid on a Sephacryl S-200column (28).The Fc fragments were removed
unresponsiveness extends to early membrane phospho- from this mixture by dialysis against H 2 0 a t 4°C. The resulting Fab
preparation wasverified for absence of undigested Ig by SDS-PAGE:
lipid turnover events. In these ways, we hope to system- no bandsabove 5 0 K were observed. Cross-linking of the monovalent
atically explore the problem of phospholipid behavior in Fabfragmentswas accomplished by usingrabbit anti-goat 1gG
(heavy and light chain specific, Cooper]. All activators were in pfthe B cell and the implications the data have on early
HBSS.
events in B cell activation.
RESULTS
MATERIALSAND
METHODS
Mice. Male BALB/c mice were obtained from the Jackson Laboratory, Bar Harbor, ME. Male and femalemice, FI progeny of a CBA/
N x DBA/2 cross (CBD2FImice] were obtained from Dominion Laboratories, Wheeling, WV. All mice were between 6 and 12wk old.
B cell isolation and purtjication. Mice were treated with 0.15 ml
anti-thymocyte serum(MicrobiologicalAssociates, Walkersville. MD)
i.p. 2 days before sacrifice. The spleens were aseptically removed
and the spleen
cells isolated by gently abrading the spleens
between
two glass slides with groundglass surfaces. Erythrocyteswere lysed
by suspending thesplenocytes in 0.15 M NH4 for 2 min. Cells thus
obtained were washed, suspended a t 5 x 106/ml in complete RPMI
(RPMI 1640 plus 1% penicillin/streptomycin plus 300 maliter glutamine plus 300 pM 0-mercaptoethanol] containing 2% fetalbovine
serum (FBS), and treated with anti-Thy-1.2 monoclonal antibody
(New England Nuclear. Boston, MA) and 1% anti-thymocyte serum
for 45 min a t 4°C. The T cells werethen lysed by treatment with 5%
guinea pig complement (Flow Laboratories, Maclean. VA) a t 37°C for
1 hr. The resultingcells were washed, resuspended a t 2 X 106/ml in
complete RPMI plus 10% FBS, and incubated in glass petri dishes
for 1 hr at37°C to remove monocytes and other adherent
cells. Cells
thus treated were >90% positive for surface Ig and depleted of T
cells to the limits of flow cytometric analysis with fluoresceinated
anti-Thy-1.2, anti-Lyt-1, and anti-Lyt-2
monoclonal antibodies (Becton-Dickinson, Mountain View, CA).Viabilities were checked by the
method of fluorescein plus ethidium bromide staining (25).Typical
viabilities were approximately90%4 hr afterantibody plus complement treatment.
Analysis ofphospholipids. Enriched B cells weresuspended a t 2
x 1O6 cells/ml after a wash inPOr-free Hanks’ balanced salt
solution
(pfHBSS).in PO4-free complete RPMI plus 2%FBS (dialyzed against
pfHBSS) in 1-ml tube cultures with defined concentrations of the
various mitogens and cultured for differing times a t 37°C. The cells
were then pulsed for 2 hr or for the time indicated with 100 pCi of
32Pi.washed. then lysed with 1 ml 0.1 N HCl. and the lipids were
extracted by addition of another 1 ml of HCl and 2 ml CHC13:CH30H
(2:l v/v] (261. The lipids were then spotted on high-performance
thin-layer chromatography plates (MCB Reagents. Gibbstown. N J )
pretreated with 1% potassium oxalate in CH30H and Hz0 (2:3 v/v).
The plates were developed in a tank containing a solvent system of
CHC13,CH3COCH3,CH30H. CH3COOH, and H 2 0 (40:15:13:12:8 v/v)
for 3.5 hr (27).
Autoradiograms were obtainedand spots on the
silica
plates scraped off and counted by liquid scintillation spectrometry,
Effect of anti-IgM stimulation on the turnover of phospholipids, withspecial reference to PI. Stimulation with
anti-IgM resulted in a specific increase ofPI turnover in
a dose-dependent fashion and in cell entry into S phase
[activation of DNA synthesis) as measured by incorporation of 3H-TdR into DNA (Table I). The early effect of
increased PI turnover showed the same anti-IgM dose
dependency as the latereffect of entry intoS. The actual
values of n-fold increases over control are not comparable
between 3H-TdR incorporation into DNA and 32Piincorporation into PI, because the nature of the metabolic
events being assayed is different, but the trends in the
data areequivalent. As is seen in Figure 1 , 32Piincorporation intoPI during the first 2 hr was increased
threefold
over control at 5 pg/ml anti-IgM and was increased sevenfold at 100 pglml. When expressed as percentage of
total 32Pi incorporation into all phospholipids, the proportion in PI as well as absolute cpm increased on antiIgM stimulation. At 5 pg/ml anti-IgM in a 2-hr culture,
percentage in PI increased from the control value of 23%
to 45% of total phospholipid incorporation.
To test theeffect of the valency of the antibody stimulus, we obtained Fab fragments of the whole anti-IgM
antibody. As is shown in Figure 2, treatment with 100
pg/ml Fab did not result in a PI response. When 20 pg/
ml rabbit anti-goat IgG (heavy and light chain specific)
was added concurrently to allow the cross-linking of the
monovalent Fab fragments bound to surface Ig, the PI
response was restored. Both 30 @/ml anti-IgM [F[ab’)2)
and 100 pg/ml Fab plus 20 pg/ml anti-IgG produced a
sixfold increase over control in incorporationof 32Piinto
PI.
PI metabolism, increased well within the first hour
after stimulation by anti-IgM, continued to be increased
4089
THE PHOSPHATIDYLINOSITOL RESPONSE AND B CELL ACTIVATION
TABLE I
D N A and PI response to anti-lgM andLPS"
Activator (@/mi]
Data Type
None
(control)
Anti-IgM
Anti-IgM
1
10
Anti-IgM
100
LPS
LPS
LPS
0.5
5
50
cpm 3H-TdR in DNA (increase
937 k 46 3.942 f 5716.639 f 2,86768.130 k 5.16424,963 f 1,64949.378 f 2.332117.524 f 4,358
18
53
73
26
125
1
over control) 4
cpm 32Pi in PI (increase over 2.090 f 2402.582 f 3286,035
f 1,728 10.460k3372,226
? 126
2.382 2 3652,301
k 491
1
1
3
1
15
1.2
controll
a Uptake of 3H-TdR (DNA)and 32Pi (PI).mean of triplets f SEM at differing concentrationsof anti-IgM and LPS, in cpm and expressed as multiples
of the control values, DNA synthesis experiments were performed in microtiter wells at 1 X 10' cells/ml: 72-hr cultures, IO-hr 3H-TdR pulse. PI
experiments were performed in1-ml tube cultures at2 X 10' cells/ml: 2-hr cultures in phosphate-freemedium with 100 pCi 32Pi.
m
-
T
50
I
T
CI
c
0
40
7-
x 30
5
0
20
E
n
$
0
5
10
50
100
Anti- IgM Concentration (pg/ml)
s
10
50
Figure I. The effect of anti-IgM on PI metabolism depends on antiIgM concentration. Cells at 2 X 10' cells/ml in 1.0 ml phosphate free
RPMI were cultured with 0. 5, 10. 50, and 100 pglml anti-IgM (F(ab'),)
and 100 pCi 32Pifor 2 hr. Incorporationof 32Pi into PI (bars)and percent
of total 3aPi incorporation into phospholipids (PL) found in PI (line)were
determined. Mean + SEM of triplicates (bars)and percent of mean total
32Pi-PLfound in PI (line).
mI
0
0
F
X
I
I
1
I
I
1
2
4
6
8
I
I
1 2 2 4
Hours after Stimulation
E
a
Figure 3. The effectof anti-IgM on PI metabolism increasesover time.
Incorporation of 32Pi pulse added 1 hr (first point) or 2 hr (all others)
before indicated time after stimulation with
50 pglml anti-lgM at time 0.
A, cpm incorporationby control (0)and stimulated(EA] cells. B. percent of
total PL 32Piincorporation in PI control ( - E L ) and stimulated(&).
A.
mean + SEM of triplicates. B. percent of means over same triplicates.
0
U
ii
a
n
cu
m
U
Antibody Treatment
Figure 2. The effect of anti-1gM on PI metabolism depends on crosslinking. Incorporation of 32Pi after 2-hr cultures: control. 10 and 30 p g l
ml divalent anti-lgM (F(ab')p),20 pglml rabbit anti-goatIgG alone, 30 p g l
ml monovalent anti-lgM Fab alone, 30 pg/ml Fab + 20 pglml anti-IgG.
100 pgml Fabalone. and 100 pglml Fab + 20 pg/ml anti-IgG. In the dual
addition cultures. Fab was added to tube, followed by 1-ml cells, then
anti-IgG to recross-link the monovalent Fab fragments. 32Pi was added
immediately thereafter. Mean + SEM of triplicates.
over 24 hr (Fig. 3A). A 2-hr pulse of 100 pCi 32Pibefore
the indicated harvest timeof cultures stimulated with50
pg/ml anti-IgM revealed that PI turnover increased a s
much as 13-fold over control by 24 hr. A 160%increase
over control PI incorporation wasapparent a s early as 15
min after stimulation with 50 pg/ml anti-IgM (data not
shown). Percentage of total incorporation in PI was not
increased at thisearliest timepoint (7%unstimulated vs
6% stimulated). After 15 min, percentage of total 32Pi
incorporation into PI was increased, reaching its maximum increase in stimulatedcells compared with control
cells at 2 hr (Fig. 3B).Other phospholipids began to
incorporate higher levels of 32Piat 4 hr. As a n indication
of this change in the synthetic balance
of the various
phospholipids, the percentage incorporation into PI, still
increasedsignificantly
over control, began to drop
slightly at 4 hr (Fig. 3B). Because the PI effect extends
well into GI and the temporal window for soluble factor
control (29), welooked for the influence ofBCGF on
phospholipid turnover. Using BCGF activity from culture
supernatants of phorbol ester-stimulated EL-4 cells, we
saw no change in the metabolism ofPI
or any other
phospholipid in stimulated or control cells, at 2, 4,or 6
hr (data not shown). This same supernatant increased
3H-TdR incorporation into DNA twofold (72 hr) at 5 ~ g /
ml anti-IgM.
The phosphorylated derivatives ofPI.PIP
and PIPP,
responded differently to anti-IgM stimulus. The 32Piincorporation into PIP increased over control values, al-
4090
THE PHOSPHATIDYLINOSITOL RESPONSE AND B CELL ACTIVATION
though this increase was truly appreciable only at 4 hr
and not at the earlier times (Fig. 4A). PIPP showed no
repeatable tendency toward increasing metabolism at 2
hr and earlier,although after 4 hr it was increased over
control in absolute cpm [Fig. 48). Percentages of total
incorporation in PIP and PIPP were always small, ranging
from 4 to 16%in PIP and 1 to 8% in PIPP. There was no
tendency toward a n increase in percentage of total 32Pi
incorporation from control to stimulated populations in
either the caseof PIP or PIPP.
The only other phospholipid that showed a repeatable
and significant effect of increasing metabolism as a n
early eventwas PA. Incorporation into PA was increased
twofold at 2 hr after stimulation by 50 &ml anti-IgM
(PA was actually increased by 80%over control as early
as 15 min after treatmentwith this dose of anti-IgM)and
stayed significantly increased relative to control over 24
hr. In contrast to PI, however, percentages in PA remained a fairly constant fraction of total phospholipid
incorporation at anygiven time point (Fig. 5B).This fact
was particularly apparent in a dose-response curve at 2
hr (Fig. 5A).PC, the othermost metabolically active phospholipid, stayed constant from control cells to stimulated
cells at theearlier time points. It then began to increase
its 32Piincorporation at 6 to 8 hr (Fig. 6). Percentage of
total incorporation into PC actually decreased from control to stimulated, reflecting the much greater fractional
contribution ofPI to the total phospholipid turnover in
stimulated cells. This was true atall time points from 2
T
I
1.5
X
E
n
0
v
a
n
n
CL
N
mU
0
10
100
50
20
Hours after Stimulation
T
10
I
24
Hours after Stimulation
Figure 5. The dose-response relationship and time course of the PA
response to anti-IgM. A. 2-hr cultures:0. 10, 50. 100 pdml anti-IgM cpm
(bars) and percentage of total PL in PA (line). B, incorporation of "Pi
or 2 hr (all others) before harvest at indicated
pulse added 1 hr (first point)
time after stimulation with50 pg/ml anti-IgM at time 0: control (0)and
stimulated (H] cells. Inset. percent of total PL 32Pi incorporation in PA
and stimulated.I*(
Mean + SEM of triplicates. Inset.
control (-E-)
percent of means over same triplicates.
I
0
r
40
1;
I
1
X
E
Q
30
u
v
2
0'
Hours after Stirnulation
20
n
40
n
t
N
10
0
1 2 4 6 8 1 2 2 4
Hours after Stimulation
T
"
B
m-
T
5
Anti-lgM Concentration (pg/ml)
50
1 2 4 8 8 1224
Hours after Stimulation
T
2.0
CI
m
2
1
2
4
6
8
1
2
24
Hours after Stimulation
Figure 6. Time course of the PC response to anti-IgM. Incorporation
or 2 hr (all others] before the indicated
of 32Pipulse added1 hr (first point)
time after stimulation with50 pdml anti-lgM at time 0: control (0)and
stimulated (@) cells. Inset, percent of total PL 32Pi incorporation in PC
and)
stimulated (6).
Mean + SEM of triplicates. Inset.
control (n
percent of means over same triplicates.
1
2
4
6
8
1 2 2 4
Hours after Stimulation
Figure 4. Time courseof the PIP (A) and PlPP (€3) response to anti-IgM.
Incorporation of 32Pi pulse added 1 hr (first point) or 2 hr (all others)
before indicated time after stimulation with
50 pg/ml anti-IgM at time 0.
32Pi incorporation (cpm) by control (0)and stimulated (a)cells. Inset,
percent of total PL 32Piincorporation in PIP and PIPP control +()
and
Mean + SEM of triplicates. Inset, percent of means
stimulated (").
over same triplicates.
hr to 24 hr. Percentage of total incorporation in both PC
and PI were roughly equal in unstimulated cell cultures
up until thevery latest time points (12 to 24 hr).We were
interested inthe behavior of phosphatidylglycerol a s well,
because its phosphate source is the same as that of PI,
the cytidine triphosphate pool. We found no anti-IgM
activation of phosphatidylglycerol metabolism (data not
THE PHOSPHATIDYLINOSITOL RESPONSE
AND
4091
B CELL ACTIVATION
defect of unresponsiveness toanti-IgM. At 50 pg/ml antishown).
Effect of LPS, a potent polyclonal activator of DNA IgM, a n effective dose for normal B cell populations, xid
synthesis, on PI turnover. Although LPS is a very effi- B cells showed no DNA response. They responded only
to
cient activatorof B cells (TableI), inducing almosthalf of weakly to 25 pdmlLPS a s well (six times control, 4700
them to enter S (30),
this activation was not reflected in cpm in 72-hr cultures at 2 X lo5 cells/microtiter well),
proliferative response to treatincreased rates ofPI metabolism at 2 hr. Both at subop- but theydid show a strong
timal (0.5pg/ml) and optimal (5and 50 pg/ml) concentra- ment with both LPS and anti-IgM (64-fold over control,
tions of LPS, there was no trace of the same activation to 47,000 cpm at 50 pdml anti-IgM and 5 pg/ml LPS), a s
ofPI metabolism that was evident in these cells upon shown inFigure 7 A . In the early eventof the PI response,
these deficient B cellsdid not respond to anti-IgM. Intertreatment with anti-IgM. The lowest dose of LPS, 0.5
ml. was sufficient to cause significant (2040-fold
to over estingly, although theydid proliferate in response to the
control) increases in 3H-TdR uptake into DNA at 72 hr. combination of LPS and anti-IgM, there was no correNo dose of LPS was capable of producing a n increase of sponding PI response to this dual treatment (Fig. 7 4 .
In trying to understand therelationship of these obserPI metabolism (TableI). Values of 32Piincorporation into
PI stayed within 10 to 15%of control at each dose of LPS, vations to progress through the cell cycle, we decided to
compared with a fivefold increase with 100 pg/ml anti- examine the response of blastogenesis. Other investigators (30) have reported that xid B cells enlarge upon
IgM at 2 hr.
treatment with anti-1gM alone. We did not find this to be
PI effect in anti-IgM unresponsive cells fromxid mice.
B cells from female CBD2Fl mice, which are heterozygous the case. Figure 8 illustrates the change in Coulter volat thexid locus, responded toanti-IgM a s did the normal ume of B cells from male and female CBD2F1 mice culcells from the BALB/c mice, both in the late event of tured for 72 h r with anti-IgM, LPS. and thecombination
entry into S and the early event of the PI response (Fig. of the two. The normal B cells from the female mice
7). There was no synergism between LPS and anti-IgM increased in size in response to50 pg/ml anti-IgM. 25 p g /
when added in concert.Indeed, 3H-TdRincorporation into ml LPS, and the dual treatment. The xid B cells were
DNA decreased with the dual treatment compared with unresponsive in terms of size increase to anti-IgM, and
LPS alone (Fig. 7A). Male CBD2FI mice are hemizygous although increasing only slightly with LPS, showed a
blast response equalor superior tothat of normal B cells
at this X-linked locus andexpressthecharacteristic
T
too
A
80
60
40
20
0
10
8
6
T
4
2
0
0
50 antilgM
25 LPS
50: antiIgM
2 5 LPS
Activator (,ug/ml)
Figure 7. B cells from mice with xld defect do not turnover PI in
response to anti-IgM. Males express this defect.whereas females do not.
A. DNA synthesis response to 0. 50 pg/ml anti-IgM. 5 pg/ml LPS. 25 pg/
ml LPS. 5 0 pg/ml anti-IgM + 5 d m 1 LPS. 50 pglml anti-IgM + 25 pglml
LPS: males (0)and females (a).
B. PI response to0. 50 pg/ml antl-IgM, 25
pglrnl LPS. 50 p g m l anti-IgM + 25 pg/ml LPS in 2-hr cultures: males (0)
and females (a).Mean + SEM of triplicates.
O
7.53.5
52 1.63
7.989.75
Cell Volume(prn
Figure 8. The xfd cells do not enlarge upon treatment with anti-lgM.
Cells from male and female CBD2FI mice were cultured for 72 hr at 1 x
1Os/ml in 24-well cluster plates in the presence of 50 pg/ml anti-IgM (A),
25 pglml LPS (B).
and 50 pg/ml anti-IgM + 25 pg/ml LPS (C].
The cells
were resuspended and characterized as to size on a Coulter Counter/
Channelizer. Percent of total cell volume at each given cell volume is
plotted against that cell volume.
4092
T H E PHOSPHATIDYLINOSITOL
RESPONSE
AND
when provided both signals. This samerelationship held
at anearlier timepoint (36hr). At 18hr theblast response
was a s yet too small to draw anysignificant conclusions.
This same relationship alsoheld at higher cell densities
(2 X 106/ml).BCGF from EL-4 conditioned medium did
not alter the results (data
not shown).
DISCUSSION
In this report we examine early phospholipid metabolism in a polyclonally stimulated, enriched murineB cell
population. Three conclusions have emerged from our
work. 1) Increased metabolism ofPI is a n early event
upon stimulation with anti-IgM antibodies. The anti-IgM
depends onreceptor cross-linking for this effect, as demonstrated by a lack of the PI effect upon treatment with
monovalent anti-IgM and the restoration of that effect
with recross-linking of the antibody. The PI response
does not follow stimulation withLPS, although both antiIgM and LPS will drive the cells into GI and on to s. 2)
The metabolically significant phospholipid in our system
is PI, and not its phosphorylated derivatives. As early a s
15 min and as late a s 24 hr, turnover ofPI was preferentially activated as compared with the polyphosphoinositols PIP and PIPP. The turnoverof PA was enhanced
a s well. PC showed no response to stimulation until 6 to
8 hr after anti-IgM addition. 3) The inability of B cells
from xid mice to proliferate in response to anti-IgM is
reflected in a lackof activation of the PI response. These
cells also failed to enlargeupon anti-IgM treatment. When
given both the anti-IgM and the LPS signals, xid B cells
did undergo blastogenesis and entered S, but there was
still no evidence of the PI response in these dually treated
cells. This is the first report to explore the fact that,
although increasedPI metabolism is clearly a n important
early event in normal B cell activation, there are alternative activation pathways that can bypass the need for
the signals supplied by PI. Coggeshall and Cambier (31)
recently described the PI response in normal murine B
cells.
Uncertainty a s to the validity of including increased
turnover of PI among the early events inB cell activation
can be traced to two primary problems: lack of a sufficientlyenriched population of B cells, and lack of a
systematic appreciation of the activation mechanisms,
including the PI response, of the various polyclonal activators of B cells. A s to the firstproblem, the methods we
employedallowed preparation of a murine splenocyte
fractionwithless
than 10% surface Ig-negative cells.
When unfractionated splenocytes were tried in this experimental system, the levels of 32Pi incorporation into
phospholipids of unstimulated cells were high enough to
almost completely mask the phenomenon of increasing
PI metabolism. With each successive stage of T and adherent cell removal, this background level of 32Piincorporation dropped further, and the
PI signal emerged ever
more clearly from the morass of general phospholipid
metabolism (data not shown). To decrease this background even further, aneven more rigorous T and accessory cell depletion procedure should be coupled with some
effort such a s countercurrent centrifugation to remove
the residual, large, activatedB cell subpopulation.
A s to the second problem, different polyclonal activators clearly act by different mechanisms and may, a s
well, act on different cell subpopulations. The first fact
B ACTIVATION
CELL
is demonstrated by our observation that treatment of B
cells with anti-IgM antibodies resulted in a n increase in
PI metabolism, whereas treatmentwith LPS did not. Rosoff, Cantley, and Stein (32, 33) have studied LPS-induced differentiation of a pre-B cell line and demonstrate
that this induction is mediated by the activation of a n
amiloride-sensitive Na+/H+exchange system. These reports point out that phorbol esters have the sameeffects
a s LPS. These two inducers may therefore act through a
common mechanism, theactivation of protein kinase C.
Further work of Wightman and Raetz (34) supports the
conclusion that LPS activates protein kinase C. This
leads to a model of LPS activation wherein the need for
cell surface receptor-mediated production ofDAG from
PI would be bypassed, thereby obviating the PI response.
Therefore, the conclusion is that the PI response is a n
important early activation event only when the stimulus
is delivered via membrane receptor-activator binding.
That the ligand-receptor-induced PI response is important in complete cell activation is illustrated by the inability of the presumptively Lyb-5- xid B cell population
to react to anti-IgM treatment in termsof the PI response,
the blastogenic response, or the DNA synthetic response.
This lack of a PI response is puzzling, because IgM is
expressed on xid B cells at a higher density than on
normal B cells (35).The absenceof a PI response may be
due to inherent unresponsiveness of Lyb-5- B cells to
anti-IgM, requirement for a specific number of occupied
IgM receptors to trigger the PI response, deficiency in
processes connectinganti-IgM binding tothe PI response,
or some othercharacteristic of these cells. Anti-IgM
clearly primes xid B cells, as concurrent treatment with
anti-IgM and LPS is sufficient to cause theirblastogenesis
and proliferation, whereas either stimulus alone is not.
The lack of blastogenesis in xid B cells was a n unexpected result in the light of the work of Defranco et al.
(30). This discrepancy
may be due to experimental differences, a s we examined a period (18 to 72 hr) later than
in theirwork (1 to 24 hr).In any case,it is clear that antiIgM does not provide a sufficient signal to xid B cells to
move them through GI.
Two points are worth considering in the analysisof the
specifics of the PI response. First, PI metabolism was
increased in stimulated cells over the entire period (15
min to 24 hr) we examined. Considering that a rather
small fraction of the phospholipid in lymphocyte membrane is PI (36).the fact that over one-half of the total
phospholipid incorporation of 32Piwas found in PI 2 hr
after stimulation indicatesthat thecell is putting a great
deal of metabolic effort into the PI cycle. The other possible interpretation of this result is that the phosphate
precursor pool for PI (cytidine triphosphate) is more easily
labeled by our procedure than theadenosine triphosphate
pool. This possibility is eliminated by our observation
that metabolism of phosphatidylglycerol, alsosynthesized by using the cytidine triphosphate pool, was not
activated. Our resultthat metabolism of PI and the fractional turnover ofPI was still disproportionately high at
24 hrdovetails nicely with the observation that thesignal
provided by anti-IgM binding to surface IgM must continue over the entirety of this period (30). Second, we
found that the behavior of PI metabolism was significantly different from that of the otherphospholipids. PC
metabolism was not increased at 2 hr, and the later
THE PHOSPHATIDYLINOSITOL RESPONSE AND B CELL ACTIVATION
4093
increase most probably signals the start of new mem- and the“priming”by anti-IgMin xid B cells is analogous.
brane biogenesis. The activities of PIPand especially PIPP It is merely the special nature of the xid B cell that allows
were also not increased at the earlier, more functionally the priming signal to be dissected away from the activapriming signal could
interesting times, but
were increased after 2 to 4 hr. This tion maintenance signal. The initial,
increase may primarily reflect new membrane synthesis be mediated by one of a number of very early events,i.e.,
a low-dose dependent burst of PIPP metabolism or memas well. These results are consistent
withequivalent
observations in T cells(1 1,
13).The early increase inPA brane depolarization. It is consistent with these and other
turnover, on the other hand, is not seen in T cells (13). observations that cells given this signal will leave Go and
Although the fraction of total 32Pi incorporation in PA enter Go*. butcannot proceed into GI and along the
stayed relatively constant in unstimulated and stimu- proliferative pathway without sustainedactivation of
lated B cells, the absolute incorporation was increased. protein kinase C. In the absence of this further signal,
This may represent a difference inthe rate-limiting steps both normal and xid cell populations stop well before
blastogenesis. The Lyb-5-, anti-IgM responsive subpopof PI resynthesis between Tand B cells.
ulation of the normal B cells are stopped in waiting for a
The data presented here do not support the current
contention that PIPP is the species of principal impor- signal that canbe provided bya high dose of anti-IgM (50
&ml) and possible subsequent regulation by soluble
tance in this system. Furthermore, the
importance ofPI
starts the PI cycle, which allows for
metabolism itself as a membrane signalling event may factors (18,29). This
be called into question, because activation and prolifer- the activation of protein kinase C. The xid cells (and
ation are possible without the PI response. The assump- possibly their normal Lyb-5- counterparts) are stopped
tion that one mediator of the effectsof the PI response is for the want of some signal that LPS can provideby
DAG activation of protein kinase C, together with the activating protein kinase C, but this signal is effective
by anti-IgM.Those few xid B
observation that LPS can directly activate this kinase, only subsequent to priming
provide one possible unifying model for understanding PI cells that do proceed into S upon treatment with LPS
response-independent activation and proliferation in xid alone have probably already been given a priming signal
B cells. This in turn may reflect on the processes of in vivo. Because in Lyb-5’ cells both signals can be
normal, anti-IgM responsive, B cell activation. A s for the providedby binding of a high dose of anti-IgM, it is
procrole of PIPP, it may be that thewindow for activation of difficult to separate thetwo events in the signaling
PIPP metabolism by anti-IgM is outside the temporal res- ess. Why a low dose of anti-IgM willcause depolarization
olution of our system.Where work has focussed on PIPP but not movement through the cell cycle is not at all
breakdown, inositol triphosphate formation and inositol clear. It may be that a critical number of cell surface
triphosphate-mediatedCa++ mobilization. the burst of receptors must be occupied to initiate the PI response
PIPP metabolism occurs on the order of a few seconds and move the cells into and through GI. In this view,
(37)
to afew minutes (38).This may represent a signaling antigen in very low dose would prime the cells, readying
burst of PIPP that gives way to the later, sustained PI them for movement into G , when the antigen dose insignal to protein kinase C. Other very early events in creases tosome threshold, indicatingthat thereis enough
activation include membrane depolarization and subse- antigen to be worthy of an immune response. In any
quent hyper-I-A expression (39, 40). and have been put event, a high dose of anti-IgMwill in turn cause the
forth a s evidence of the movement of the cell from Go to stimulation of PI turnover and subsequentproduction of
Go*, an activated Go state. It is not clear whether these DAG. Regardless of whether this second signal is provery early events are interrelated.
If they arenot, the one vided by LPS or a receptor-mediated increase in PI turnamong them that is critical in moving the cell into the over, it is necessary for the cell to commit itself to G I .
This conclusion is supported by the lack of a blast reactivated Go state hasyet to be singled out.
Our data have lead us toaunifying
model of two sponse of xid B cells to anti-IgM and the restoration of
signals, one a n initial activating signal and the other a this response with the concurrent addition of LPS. This
signal that serves to maintain the activated state. This
is model must now be tested.
In summary, our results demonstrate the importance
supportable inthe light of the datapresented here on the
PI effects of LPS and anti-IgM, both in normal and xid of PI metabolism in B cells stimulated with anti-IgM, an
mice. It has been suggested that the xid proliferative activation process that shares many of the characterisIg. The lack of an effect
response to anti-IgM plus LPS is a result of “priming” (an tics of antigen binding to surface
early activation signal)by anti-IgM, which allows the cell of anti-IgM on phospholipid metabolism in xid B cells
to leave Go, followed by a further signal provided by LPS gives further credence to the speculation that there may
that allows the primed cell to continue into andthrough be two steps to the activation process, eachwith its own
GI. This priming signalcan be provided in normal B cells stimulus, acting sequentially to bring the cell along the
by a low dose of anti-IgM. The presence of this low dose activation pathway and ultimately on toS phase.
of anti-immunoglobulin will cause movement into Go*
(39, 40). but it is not sufficient to cause increases in PI
Acknowledgments. We thank Ms. Carole Stokes and
turnover or sustain thecell’s movement through the cell Ms. DeloresBolden forpreparing the manuscript, we
cycle. Without examining the putative markers for Go* greatly appreciate the advice of Ann L. Akeson through(membrane depolarization or hyper-I-A expression), it is out the investigation, and we thank Dr. Gunter Grupp for
still clear that anti-IgM must be priming the anti-IgM B his careful and critical readingof this manuscript.
cell in some manner thatallows it to be carried along the
cell cycle by LPS. Without both LPS and anti-IgM, moveREFERENCES
ment through GI and into S is not possible. It is our
1. Michell. R. H. 1982. Inositol lipid metabolism in dividing and differcontention that thepriming by anti-IgM in normal Bcells
entiating cells. Celt Calclurn 3:429.
4094
THE PHOSPHATIDYLINOSITOL RESPONSE AND B CELL ACTIVATION
2. Ku, Y.. A. Kishimoto, Y. Takai. Y. Ogawa, S. Kimura,and Y.
Nishizuka. 1981. A new possible regulatory system for protein phosporylation in human peripheral lymphocytes. 11. Possible relation to
phosphatidylinositol turnover induced by mitogens. J . Irnrnunol.
127:1375.
3. Farese. R.A. 1983. The phosphatidate-phosphoinositidecycle: a n
intracellular messenger system in the action of hormones and neurotransmitters. Metabolism 32:628.
4. Putney, J.. S. Weiss, C. Van De Walle, and R. Haddas. 1980. Is
phosphatidic acid a calcium ionophore under neurohumoral control?
Nature 284:345.
5. Kirk, C., J. Creba, C. Downes, and R. H. Michell. 1981. Hormone
stimulated metabolism of inositol lipids and its relationship to hepatic receptor function. Biochern. Soc. Trans. 9:377.
6. Berridge, M. 1983. Rapid accumulation of inositol triphosphate reveals that agonists hydrolyze polyphosphoinositides instead of phosphatidylinositol. Biochern. J . 21 2:849.
7. Creba. J.. P. Downes. P. Hawkins. G . Brewster, R. H. Michell. and
C. Kirk. 1983. Rapid breakdown of phosphatidylinositol 4-phosphate
and phosphatidylinositol 4.5-bisphosphate in rat hepatocytes stimulated by vasopressin and other calcium-mobilizing hormones. Biochern. J . 212:733.
8. Sugimoto, Y.,M. Whitman, L.C. Cantley, and R. Erikson. 1984.
Evidence that the Rous sarcoma virus transforming gene product
phosphorylates phosphatidylinositol and diacylglycerol. Proc. Natl.
Acad. Sci. USA 81:2117.
9. Macara, I., G . Marinetti. and P. Balduzzi. 1984. Transformingprotein of avian sarcoma virus UR-2 is associated with phosphatidylinositol kinase activity: possible role in tumorigenesis. Proc. Natt.
Acad. Sci. USA 81:2728.
10. Fisher, D.B., and G. C. Mueller. 1968. An early alteration in the
phospholipid metabolism of lymphocytes by phytohemagglutinin.
Proc. Natl. Acad. Sci. USA 60:1396.
11. Hui, D. Y.,and J. A. K. Harmony. 1980. Phosphatidylinositol turnover in mitogen-activated lymphocytes. Suppression by low-density
lipoproteins. Biochem. J . 192:91.
12. Fisher, D. B., and G . C. Mueller. 1971. Studies on the mechanism
by which phytohemagglutinin rapidly stimulates phospholipid metabolism of human lymphocytes. Biochirn. Biophys. Acta 248:434.
13. Akeson, A. L., D. W.Scupham, and J. A. K.Harmony. 1984. The
phosphatidylinositol response and proliferation of oxidative enzyme
activated human T lymphocytes: suppression by plasma lipoproteins.
J . Lipid Res. 25: 1195.
14. Betel, I.. J. Martijnse, and K. van den Berg. 1974. Absence of a n
increase of phospholipid-phosphate turnover in mitogen-stimulated
B lymphocytes. Cell Zmrnunol. 14:429.
15. Chien, M. M., and R. F. Ashman. 1983. Phospholipid synthesis by
activated human B lymphocytes. J. Irnrnunol. 130:2568.
16. Melchers, F.. and J. Andersson. 1984. B cell activation: three steps
and their variations. Cell37:715.
17. Raetz, C. R. H., S. Purcell, and K. Takayama. 1983. Molecular
requirements for B-lymphocyte activation by Escherichia coli lipopolysaccharide. Proc. Natl. Acad. Sci. USA 80:4624.
18. Howard, M., and W. E. Paul. 1983. Regulation of B-cell growth and
differentiation by soluble factors. Annu. Rev. Irnrnunot. 1:307.
19. Amsbaugh, D., C. Hansen, B. Prescott, P. Stashak, D. Barthold,
and P. Baker. 1972. Genetic control of the proliferative response to
type Ill pneumococcal polysaccharide in mice. 1. Evidence that anXlinked gene plays a decisive role in determining responsiveness. J .
Exp. Med. 136:93 I .
20. Sieckmann. D. G., R. Asofsky. D. E. Mosier, I. M. Zitron. and W.E.
Paul. 1978. Activation of mouse lymphocytes by anti-immunoglobulin. 1. Parameters of the proliferative response. J . Exp. Med.
147:814.
21. Sieckmann, D. G., I. Scher, R. Asofsky, D.E. Mosier, and W. E.
Paul. 1978. Activation of mouse lymphocytes by anti-immunoglobulin. 11. A thymus-dependentresponse by a mature subset of B
lymphocytes. J . Exp. Med. 148: 1628.
22. Mosier, D. E., I. M. Zitron, J. J. Mond, A. Ahmed, I. Scher, and W.
E. Paul. 1977. Surface immunoglobulin D as a functional receptor
for a subclass of B lymphocytes. Immunol. Rev. 37:89.
23. Asano, Y., A. Singer, and R. J. Hodes. 1981. Role of the major
histocompatability complex in T cell activation of B cell subpopulations. J . Exp. Med. 154: 110.
24. Singer, A., P. Momssey, K. Hathcock, A. Ahmed, I. Scher, and R.
J. Hodes. 1981. Role of the major histocompatibility complex in T
cell activation of B cell subpopulations. J. Exp. Med. 154501.
25. Ugai, K.. M. Ziff, and P. E. Lipsky. 1979. Gold-induced changes in
the morphology and functional capabilities of human monocytes.
Arthritis Rheum.22:1342.
26. Folch, J., M. Lees, and G . Stanley. 1957. A simple method for the
isolation and purification of total lipides from animal tissue. J . Biol.
Chern. 226:497.
27. Jolles, J., H. Zwiers, A. Dekker, K.W.A. Wirtz, and W. Gispen.
1981. Corticotropin-(1-24)-tetracosapeptideaffects protein phosphorylation and polyphosphoinositide metabolism in the rat brain.Biochern. J . 194:283.
28. Shirai, IC.,D. Wisner. D. Johnson, L. Srivastava, andR. L. Jackson.
1982. Immunological studies on bovine milk lipoprotein lipase: effects of Fab fragmentson enzyme activity. Biochirn. Biophys. Acta
712:lO.
29.Howard, M., J. Farrar. M. Hilfiker. B. Johnson, K. Takatsu, T.
Hamakoa, and W. E. Paul. 1982. ldentification of a T cell-derived B
cell growth factordlstinct from interleukin 2. J . Exp. Med. 155914.
30. DeFranco, A. L., J. T. Kung, and W. E. Paul. 1982. Regulation of
growth and proliferation in B cell subpopulations. Zrnrnunol. Rev.
64:161.
31. Coggeshall. K. M., and J. C. Cambier. 1984. B cell activation VU.
Membrane immunoglobulins transduce signals via activation of
phosphatidylinositol hydrolysis. J . Zrnrnunol. 133:3382.
32. Rosoff. P. M., and L. C. Cantley. 1983. Increasing the intracellular
Na+ concentration induces differentiation in a pre-B lymphocyte
line. Proc. Natl. Acad. Sci. USA 80:7547.
33. Rosoff. P. M., L. F. Stein, and L.C. Cantley. 1984. Phorbol esters
induce differentiation in a pre-B-lymphocyte cell line by enhancing
Na+/H+ exchange. J . Biol. Chern. 259:7056.
34. Wightman. P. D., and C. R. H. Raetz. 1984. Theactivation of protein
kinase C by biologically active lipid moieties of lipopolysaccharide.
J. Biol. Chern. 259: 10048.
35. Whitlock, C. A., and J. D. Watson. 1979. B-cell differentiation in
the CBA/N mouse. J . Exp. Med. 150:1483.
36. Majerus, P., E. Neufeld, and D. Wison. 1984. Production of phosphoinositide-derived messengers. Cell 37~701.
37. Prentki, M., T. Biden, D. Janjic, R. Imine, M. Berridge. and C.
Woolheim. 1984. Rapid mobilization of Ca2+from rat insulinoma
Nature 309:562.
microsomes by inositol-1,4,5-triphosphate.
38. Hasegawa-Sasaki, H., and T. Sasaki.1983. Phytohemagglutinin
induces rapid degradation of phosphatidylinositol 4.5-bisphosphate
and transient accumulation of phosphatidic acid and diacylglycerol
in a human T lymphoblastoid cell line, CCRF-CEM. Biochirn. Biophys. Acta 754~305.
39. Monroe, J. G., J. Niedel, and J. C. Cambier. 1984. B cell activation.
IV. Induction of cell membrane depolarization and hyper-I-A expression by phorbol diesters suggests a role for protein kinase C in murine
B lymphocyte activation. J . Zrnrnunol. 132:1472.
40. Cambier, J. C., and J. G . Monroe. 1984. B cell activation. V. Differentiation signaling of B cell membrane depolarization, increased I-A
expression, Go to G I transition and thymidine uptake by anti-IgM
and anti-IgD antibodies. J . Irnrnunol. 133:576.