Journal of Leukocyte Biology Volume 61, May 1997 637

Preliminary characterization of diacyiglycerol generation in

human basophils: temporal relationship to histamine

release and resolution of degranulationAlfonso Oriente, Thomas Hundley, Walter C. Hubbard, and Donald W. MacGlashan, Jr.

Johns Hopkins Asthma & Allergy Center, Baltimore, Maryland

Abstract: Purified human basophils were examined

for changes in diacylglycerollevels to determine whether

the transient nature of a N-formyl-methionyl-leucyl-

phenylalanine (fMLP) -stimulated elevation in mem-

brane protein kinase C (PKC) activity could be ex-

plained by the transient production of diacyiglycerol

(DAG). In preliminary experiments total 1MG levels

were measured by the DAG kinase assay. AltJ�ough ele-

vations followed stimulation with 1 tM fMLP (basal

levels of 15 pmol/106 basophils vs. 45 pmolIlO6 baso-

phils at the 3-mm time point), there were no detect-able changes in the first 60 s ofthe reaction. Histamine

release is typically complete by 30-45 s. Measurement

ofinositol trisphosphate indicated a rapid increase by

5 5 of2.5 pmolIlO6 basophils. IfDAG were produced

at similar levels, the DAG kinase assay would not havedetected the elevation. Consequently, fMLP-sthnulated

basophils were examined for changes in 1-stearoyl, 2-

arachidonoyl, 3-sn-glycerol (SA-DAG) and 1-oleoyl,2-arachidonoyl, 3-sn-glycerol by GC-NICIMS (nega-

five ion chemical ionization mass spectroscopy). A 5-s

elevation in these two species averaged 2 pmol/106basophils, consistent with the inositol trisphosphatelevels and occurring during the period of histaminerelease. However, a much more pronounced secondphase to the SA-DAG response also occurred, mirroring

the total DAG levels. This second phase of the DAG

response, either total or SA-DAG, was transient on a

time scale temporally coincident with the appearance

and resolution of degranulation sacs as measured byfluorescence microscopy. These data suggest that thereis selective generation of DAG species in the early re-

action and the later appearance ofDAG may be relatedto the formation and resolution of granule structures

that follow the secretion ofhistamine. J. Leukoc. Biol.

61: 637-642; 1997.

Key Words: signal transduction . lipids . molecular species

INTRODUCTION

As in most secretory cells, protein kinase C (PKC) appears

to participate in the early signal transduction of several re-

ceptor systems in human basophils [1, 2]. In studies per-

formed several years ago, we found that a kinase activity

that was dependent on Ca2� and phosphatidylserine in-

creased in the membranes of basophils stimulated with either

anti-IgE antibody or N-formyl-methionyl-leucyl-phenylalanine

(IMLP) 11, 2]. During fMLP stimulation the activity ap-

peared to translocate from the cytosol to the membrane in a

manner consistent with many other studies ofPKC 121. AfterIgE-mediated stimulation, membrane activity increased in

the absence of a loss in the cytosolic fraction �1, 2]. In both

cases, however, this enzymatic activity, which appeared to

be PKC, was transient. For IgE-mediated release, PKC ac-

tivity returned to resting levels after approximately 30 mm,

at a time when histamine release had ceased. For IMLP

the reaction resolved faster (< 5 mm), on a time scale con-

sistent with the faster histamine release (and its cessation)

and the faster generation of a cytosolic calcium elevation for

this secretagogue. We therefore became interested in which

factors were controlling the PKC reaction and whether known

PKC activators displayed a similar transient behavior.

Enzymatic activity of the classical PKC isozymes, both

calcium dependent and independent (type I and II), de-

pends on the presence of diacylglycerols f3, 4]. However,

many studies of diacylglycerol (DAG) generation find that

elevations of this second messenger persist for relatively

long periods of time. For example, in rat peritoneal mast

cells, histamine release is complete within the first 2-3 mmwhereas DAG remains elevated for at least 30 mm t�, 61.

If a similar pattern were to occur during either fMLP or

anti-IgE antibody-mediated stimulation of human baso-

phils, the transient nature of PKC activity would not result

from regulation of DAG but some other factor or process.

This study was begun to determine the pattern of DAG

generation in basophils stimulated with fMLP. The goal

was to determine whether DAG generation was transient on

a time scale that would appropriately explain the transient

activity of PKC.

Abbreviations: IMLP, N-formyl-methionyl-leucyl-phenylalanine; PKC.

protein kinase C; DAG. diacylglycerol; IP3, inositol trisphosphatase;

HPLC. high-performance liquid chromatography.

Correspondence: Donald W. MacGlashan, Jr. . Johns Hopkins

Asthma & Allergy Center. 5501 Hopkins Bayview Circle, Baltimore, MD

21224.Received November 25. 1996; revised February 5. 1997; accepted

February 6, 1997.

638 Journal of Leukocyte Biology Volume 61, May 1997

METHODS

Reagents

The suppliers for the following reagents were as follows: high-performance

liquid chromatography (HPLC) grade chloroform and HPLC grade water

(Burdick & Jackson, Muskegon, IL), HPLC grade hexane (Curtis & Mat-

theson, Houston, TX), HPLC grade methanol (J. T. Baker, Philippsburg,

NJ), hydrochloric acid (Fisher Scientific, Fair Lawn, NJ), anhydrous di-

ethyl ether (Mallinckrodt, Paris, KY), [32PJATP (DuPont Medical Prod-

ucts, Wilmington, DE), silica gel (Silicar type 60 A; Maffinckrodt), sn-

1,2-diacylglycerol assay (Amersham, Amersham, UK). glasswool (Pesti-

cide Grade, Supelco, Bellefonte, PA); and inositol trisphosphate assay

kit (Amersham).

Buffers

PAG buffer consisted of the following: 25 mM PIPES (Sigma Chemical

Co., St. Louis, MO), 140 mM NaC1, 5 mM KC1, 0.003% human serum

albumin (Miles Laboratories, Elkhart, IN), and 0.1% glucose. PAGCM

consisted of PAG with 1 mM CaC12 and 1 mM MgCI2.

Cell preparation

Basophils were purified from residual leukocytes, obtained from normal

donors undergoing plateletpheresis, using counter-current elutriation

and Percoll density gradients as previously described [7J. Basophils

were stained with alcian blue 181 and counted in a Spiers-Levy cy-

tometer. In the experiments presented, basophil purities ranged from

84 to 90% with an average of 87%. For both the assays of DAGs, a com-

parison was made between the responses observed in these purified

basophil preparations and cell preparations containing only the contam-

inant cells typically present. For these assays, changes in DAG in the

contaminant cells were found to range from 100 to 150% of the re-

sponse observed in the purified basophils. These results indicated that

only a small component of the observed responses in the purified baso-

phil preparations could be attributed to the contaminant cells.

Histamine release measurements

Histamine release was determined by the automated method of Siragan-

ian 191 and histamine release was calculated as the ratio of released his-

tamine to total histamine after subtracting the spontaneous release from

each value.

TotalDAG analysis

Cellular reactions were stopped by transferring the cells to a borosilicate

glass tube containing a mixture of chloroform/methanol/0.01 N hydro-

chloric acid (v/v; 0.5/1/0.4), which is a modification of the procedure

of Bligh and Dyer 1101. The chloroform was removed under a stream

of nitrogen, the lipids were resuspended in 5% etherfhexane, and the

samples passed through a silicic acid minicolumn ( � 75 �tL of silicic

acid in cut-off glass pipettes). Triglycerides were removed by eluting with

0.6 mL of 15% ether/hexane, and the DAGs removed with 0.6 mL of

35% ether/hexane. Phospholiplids were retained on the silicic acid.

These percentages were predetermined by testing the column elution

with radiolabeled triglycerides, 1-stearoyl, 2-arachidonoyl glycerol, or

1-stearoyl, 2-arachidonoyl phosphatidyl choline (New England Nuclear)

and are similar to those reported by Kennerly � The 35% fractions

were recovered in prewashed (methanol and ethyl ether/hexane 35%)

polypropylene tubes and evaporated under a stream of nitrogen. DAG

was measured according to the manufacturer’s protocol (Amersham).

The basis for this assay is the labeling of sample DAG with 32PO4 cat-

alyzed by Escherichia coli-derived DAG kinase �11, 121. 32P04-DAG

is separated from the ‘y-32P04 donor, ATP, by extraction. The assay is

linear for approximately 3 logs of sample or standard DAG. DAG recov-

eries were determined by including within each experiment cell samples

to which a small amount of DAG was added (two to three times more than

a typical stimulation level) to cells just before the Bligh-Dyer extraction.

Recoveries averaged 71 ± 16% (mean ± SD). The l32Plphosphatidic

acid generated was determined by liquid scintillation counting and the

amount of DAG generated was calculated against a standard curve. The

values plotted do not reflect the average 70% recoveries.

Inositol trisphosphate levels (1P3)

1P3 levels in basophils were measured with the commercially available

receptor-based assay from Amersham. The basis for this assay is the

competitive binding of radiolabeled IP3 with the unlabeled sample IP3

to a preparation of the 1P3 receptor. With the exception that the IP3 re-

ceptor was used to capture 1P3. rather than an antibody, this assay was

analyzed like a radioimmunoassay. There were a few modifications nec-

essary. It was previously noted 1131 that the granule proteoglycans pres-

ent in RBL cells would interfere with the measurement of 1P3 by this

method. The solution for the RBL cell studies, which we used, was to

filter the lysed basophil supernatants with Millipore 10,000 NMWL

Ultrafree#{174}-MC filter units. The procedure was to challenge 75 �tL of cell

suspension with 75 i�tL of stimulus, stop the reaction with 30 p1 of tn-

chloroacetic acid, and extract with 340 p1 of the recommended

TCE/tnioctylamine (see manufacturer’s protocol). This recovered enough

solution to filter with the Millipore units to recover a sample for the 1P3

assay. Centrifugation of the Millipore units was carried out at 2500 rpm

( � 1000 g) for 2 h at 4#{176}C.To determine the recovery of 1P3 in the con-

text of the entire assay, each experiment included resting cell samples

(resting cells only), which were spiked with levels of IP3 consistent with

the stimulated levels expected (1 and 3 pmol) just before lysis with TCA.

Recovery averaged 88 ± 25% (n = 7, mean ± SD).

Extraction, derivatization, and GC-NICIMS analysis ofthe 1-stearoyl-2-arachidonoyl-sn-3-glycerol

The procedures for the extraction, denivatization, and GC-NICIMS analy-

sis of the 1-stearoyl-2-arachidonoyl-sn-3-glycerol were performed as re-

cently described 1141. Briefly, the cells were initially processed as de-

scnibed above for the total DAG assay. However, a standard amount of

2H3-1-stearoyl-3-arachidonoyl-sn-2-glycerol (deuterated standard) was

added to the initial buffer-CHC13-methanol mixture of the Bligh-Dyer

extraction before phase separation. The mini-columns were run only to

remove phospholipids, therefore, the samples were resuspended in hex-

ane, run into the columns, and the columns eluted with 35% ether!

hexane. These samples were dried under nitrogen and derivatized with

pentafluorobenzoyl chloride for analysis by GC-NICIMS.

Morphometrics of basophil degranulation bymicroscopy

The procedures for determining the appearance of fluorescent-free re-

gions within basophils undergoing degranulation are complex and the

algorithms used are fully described in a prior publication from this lab-

oratory 1151. In these previous studies, the method to determine the size

of holes within the cell was presented but those studies did not spe-

cifically present the data concerning this parameter, focusing instead on

measurements of cell perimeter. However, the same algorithm gives a

pixel count that is the sum of all hole perimeters for a particular cell.

Briefly, as a stimulated basophil degranulates, fluorescent-free regions

appear within the cell. Similar regions appear in electron microscopic

images and are shown to be large degranulation sacs, the expansion of

normal granules during the resolution of the response f16J. To obtain

measurement of these regions with the use of light microscopy, basophils

were labeled with Fura-2!AM and changes in morphology (and calcium)

were monitored by fluorescence digital videomicroscopy. Applying the

aforementioned algorithms to the fluorescent images extracts the size of

the perimeters ofthese regions for each cell under observation. The data

for the m 50 cells under observation during a typical challenge was

averaged at each of 50-100 time points and is presented as the net

change in the hole perimeter above resting levels (although, in fact the

resting cells had no measurable holes).

U)

�30

2 25

� 20

�15

(�10

z0

Time After fMLP Addition, S Time After fMLP Addition, s

RESULTS noted in Methods, recovery of 1P3 was not statistically dif-

ferent than 100%, so this value should be approximately

Kinetics of total DAG generation correct. Resting levels averaged less than 0.25 pmol/106

Total DAG generation was first examined by use of the total

DAG kinase assay described 15 years ago in studies of rat

mast cells �i1]. Figure 1 shows the kinetics of DAG gen-

eration following stimulation with 1 �tM fMLP. Figure 1,

left, shows the kinetics over a 20-mm period following stim-

ulation whereas Figure i, right, expands the early time

frame, placing the results in the context of the kinetics of

histamine release. The data show that elevations in total

DAG occurred with this stimulus, beginning with an aver-

age resting level of 15 pmol/i06 basophils; net peak eleva-

tions of 30 pmol/i06 basophils occurred between 3 and

5 mm after the addition of fMLP. However, as seen in Fig-

ure 1, right, no elevations were detectable before or during

histamine release. Therefore, although this stimulus could

cause elevations in DAG, no elevations were observable in

basophils. Comparing these results with the data in Figure 2,

it can be appreciated that if DAG were generated at similar

(instantaneous) steady state levels early in the reaction, the

elevations would not be measurable (a minimum of 3- to 5-

pmol elevation would be discernible given the intra-replicate

variation).

Previous studies in other cell types have noted that phos-

phatidylinositols have a prevalence of arachidonic acid in

the 2-acyl position [17]. Indeed, a predominant species

has stearic acid in the snl position and arachidonic acid

in the sn2 position, leading to the generation of 1-stearoyl,

2-arachidonoyl, sn-3-glycerol (SA-DAG) following hydro-

lysis with phospholipase C. We have recently developed an

extremely sensitive method of measuring DAG molecular

species but have thus far only applied it to the measure-

the time frame of histamine release.

Previous studies of PKC activation indicated that in- 2.5

creased membrane activity occurred within 5 s of stimula-

tion 121. Furthermore, previous studies of cytosolic calcium

demonstrated an increase that occurred within 1-2 s after 2

the addition of fMLP. It is highly likely that this initial ele-

vation of Ca2� results from the generation of 1P3, although

this hasn’t been formally proven for human basophils. fMLP-

mediated stimulation of basophils is pertussis toxin sensitive

and it is expected that this GTP-binding protein-mediated

reaction results in the activation of PLC�3, which should

.Ea.0� � .5Cl)

.�

0�a)� -iEa.

subsequently result in the generation of both IP3 and DAG

from phosphatidylinositols. Therefore, we examined IP3

cv�

0.5

generation following stimulation of basophils with 1 �iM

fMLP. The results are shown in Figure 2. As expected,

1P3 levels increased at the earliest measured point, 5 s,

and decayed toward resting levels by 5 mm. There was sig-

nificant variation in the response of different basophil prep-

0� ....i....t....j.... I

0 50 100 150 200

Time After fMLP Stimulation, S

arations, which is reflected in the wide range of the data,

with a rapid decay occurring in some preparations and a

somewhat sustained response occurring in others. On aver-

age, the peak elevation was approximately 2.5 pmol/i06basophils (ranging from 1 to 5 pmol/i06 basophils). As

Fig. 2. 1P3 elevations: basophils were stimulated with 1 �tM IMLP,

cells extracted at the times shown, and 1P3 levels measured with a

receptor-based binding assay (n = 4). Basophil purity averaged 85 ±

2% and histamine release averaged 78 ± 28%. Basal levels of lP3 aver-

aged < 0.25 pmol/106 basophils (below the S!N of 2 in the 1P3 assay).

range = 1 -5 pmoIes/10� basophils n=4

35

Oriente et al. Diacyiglycerol changes in human basophils 639

Fig. 1. Total DAG elevations: basophils

were stimulated with 1 �tM fMLP, cells

extracted at the times shown, and total

DAG measured with the DAG kinase as-

say (n = 5). Basophil purity averaged 81

± 9% and histamine release averaged

63 ± 34%. (A) The entire time period ex-

amined; (B) the first � 2 mm. Basal levels

of DAG averaged 15 pmol!106 basophils.

Data is not adjusted for DAG recoveries

(see text). The sketched-in curve labeled

HR is the normal IMLP histamine release

kinetics as determined from numerous

previous studies, with a half-maximum at

15 s.

0 200 400 600 800 1000 1200

Time After fMLP, S

640 Journal of Leukocyte Biology Volume 61, May 1997

U)

0.

0

a0

Ea

a

c,J

‘9

az

Fig. 3. Elevations in 1-stearoyl 2-arachidonoyl-sn-3-glycerol after stim-

ulation of basophils with 1 �iM fMLP (n = 5). Basophil punities aver-

aged 87 ± 2% and histamine release averaged 79 ± 18%. The gray

tracing represents the 1P3 data. derived from Figure 2, to demonstrate

the temporal relationship of the two kinetic curves.

ment of SA-DAG and two related species. Figure 3 shows

the changes in SA-DAG after stimulation with i �tM fMLP.

In these experiments, resting levels ofSA-DAG averaged 1.1

pmol/106 basophils. fMLP induced a biphasic response,

with a net average increase in SA-DAG, in the 5- to 60-s

time frame, of 1.5 pmol/i06 basophils. The range of this

increase varied among preparations from 0.9 to 4 pmol/

106 cells. A second much stronger increase occurred with

a peak at 3 mm. Indeed, approximately 98% of the SA-

DAG mass was generated in the time frame previously as-

sociated with the late increase in total DAG (measured with

the kinase assay). Like the total DAG, SA-DAG generation

subsided by 15-20 mm. The stimulation index (stimulated!

resting levels) for this species in the early time frame was

2.2 ± 0.3 and averaged 6.0 ± 1.1 for the 3-mm measure-

ment. Histamine release averaged 79% but calculated as a

stimulation index (stimulated release/spontaneous release)

was 6.4 ± 0.7. A close molecular relative, i-oleoyl, 2-

arachidonoyl-sn-3-glycerol is also generated, starting with

a resting level of 0.4 pmol/106 basophils and increasing to

an average net of 0.5 pmol/106, whereas 1-linoleoyl-2-

arachidonoyl-sn-3-glycerol was not detected in resting or

stimulated cells. It is of interest to note that the summation

of the two measured DAG species, � 2 pmol/i06 cells, was

close to the average IP3 increase of 2.5 pmol!i06 baso-

phils (see Discussion).

The above data indicate that SA-DAG generation occurs

within the proper time frame for histamine release, mirror-

ing the increase in IP3 and elevations in Ca. However, fol-

lowing stimulation with fMLP, the greatest increases in

both SA-DAG and total DAG levels occurred after the com-

pletion of histamine release. As noted in the Introduction,

our goal was to determine whether the generation of DAG

followed a time course that could explain the transient na-

ture of PKC activation. Membrane PKC activity following

stimulation with fMLP returns to resting levels within 5

mm, whereas DAG is actually peaking at this time point.

Clearly, the DAG data doesn’t support the transient nature

of the PKC activity curve. We have previously noted that

basophils undergo remarkable morphological changes dur-

ing stimulation with fMLP. These changes can be observed

and measured by fluorescent light microscopy. By fluo-

rescence microscopy, the cells develop cytoplasmic holes,

which by electron microscopy represent a stage in the res-

olution of the degranulation process. In particular, these

holes probably represent extremely large degranulation

sacs observed by electron microscopy. When these hole for-

mations are measured (see Methods) after stimulation with

fMLP, their appearance rises and falls on the same time

scale as the later phase of DAG generation. Figure 4 plots

the hole formation as determined by a combination of im-

age processing algorithms. The data for DAG generation

is sketched in for comparison. Like total DAG generation,

there are no discernible holes in the first 60 s of the re-

action while the holes peak by 5 mm and gradually disap-

pear by 20 mm.

DISCUSSION

These studies were undertaken to determine the nature of

DAG generation in stimulated human basophils and to de-

termine whether the kinetics of DAG generation were tran-

sient in a manner consistent with the transient nature of the

putative PKC activity previously observed. At the level of

both the total DAG and SA-DAG, the kinetic profile was tran-

sient, but on a time scale not clearly appropriate to the ac-

tivity of PKC previously observed. It is interesting to note

that the total DAG kinetic profile is qualitatively similar to

the total DAG profile observed in cytochalasin B-treated

fMLP-stimulated neutrophils, i.e., a lag in generation fol-

lowed by a stronger later elevation 1181. However, in con-

trast, in the absence of cytochalasin B, there was an ob-

servable early peak in fMLP-stimulated neutrophils. In

basophils, these changes occur in the absence of cyto-

chalasin B and the initial lag was longer in duration than

that observed in neutrophils. It became apparent that the

standard DAG kinase assay was missing important early

data, the period between 0 and 60 s. It was necessary to

develop a more sensitive technology for measuring DAG

mass. For the purposes of this study, we did not anticipate

that it was essential to measure specific molecular species

of DAG but the nature of GCMS technology, by default, al-

lowed us to examine three arachidonic acid-containing spe-

cies thought to be dominant in phosphatidylinositol lipids.

For the purposes of this study, it has turned out to be suf-

ficient to draw one important conclusion on the basis of

only one of the species, SA-DAG. As expected, changes in

DAG levels do occur on the time frame of histamine re-

lease. The instantaneous steady-state levels of SA-DAG/OA-

DAG and IP3 were similar. There was no a priori reason

to believe that the levels would closely match because both

1P3 and DAG are presumably present as a consequence of

competing processes that generate or metabolize the two

types of molecules. Therefore, even if the two molecules

C

aU)a

E

E

0

C0

aU.

200 400 600 800 1000 1200 0 50 100 150

Time After fMLP Addition, s

Oriente et al. Diacylglycerol changes in human basophils 641

Fig. 4. Kinetics of the appearance of

cytoplasmic free regions in basophils la-

beled with Fura-2!AM, stimulated with

1 �tM fMLP and observed by fluores-

cence microscopy (n = 2). (A) Results

for the entire period of observation and!

or measurement, � 20 mm. (B) Expands

the early time points. Black squares, re-

plot of the data for SA-DAG changes;

open circles, replot of the data for total

DAG measurements. Small black circles

represent the holes measurement which,

like the DAG measurements, is plotted

as fraction of the maximum response for

the purposes of comparing the three

measures of the basophil response.

were generated by the same mechanism, their metabolism

might have differed. Until a larger number of species is ex-

amined, the similarity in the steady state levels for DAG

and IP3 in the first 15 s does not have any signifi-

cance except to demonstrate that metabolism is not grossly

dissimilar (if other species were present at significant 1ev-

els, the total DAG measurements would have revealed an

early peak; the specific rise in arachidonic acid-containing

species of DAG is consistent with studies in fibroblasts 1191).Although there is plenty of precedence in the literature for

DAG generation being sustained for periods much longer

than the functional end point (in this case, histamine re-

lease), the markedly larger generation of DAG late in this

particular reaction, and its subsequent return to restinglevels by 20 mm, suggested an alternate role for DAG.

There are a multiplicity of reasons that DAG could remain

elevated and several other functional endpoints that might

be regulated by a prolonged generation of DAG. However,

in human basophils, other known functional endpoints re-

quire either shorter (LTC4 secretion 1201) or longer (IL-4

secretion [7J) periods of time. Our studies indicate that, al-

though the DAG generation is prolonged compared with his-

tamine release or the PKC-like activity we have previously

described, it is still relatively transient. One line of spec-

ulation has placed DAG central to granule fusion reactions

[21-251. It is not unreasonable to speculate that it may

have a role in the resolution of degranulation. By electron

microscopy, the morphological changes observed follow-

ing fMLP stimulation are remarkable. The amount of mem-

brane restructuring is extreme so that, if DAGs are involved,

the period of granule resolution would dictate the period of

DAG generation. Thus, the similarity between DAG kinetics

and the resolution of the degranulation structures could be

coincidental or suggest a role for DAG in this process.

The fluorescent microscopic studies of hole formation re-

semble the process of degranulation resolution described

in previous electron microscopic studies. It is useful to note

that, under resting conditions, basophils are more spheri-

cal, even though they rest on the surface of the microscope

chamber. Upon stimulation the cells flatten considerably

and spread. There are no holes measured in unstimulated

cells for the duration of the period of observation [see the

figure in ref. 15J. In addition, basophils may move around

without there being any apparent holes. The appearance

of holes is strictly related to stimulated secretion and, in

a series of experiments comparing this parameter to hista-

mine release (from the supernatants of the cells under ob-

servation in the microscope chamber), we found a close cor-

relation. For example, a decrease in histamine release by 60

± 5% using suboptimal stimulation, also led to a 59 ± 6%

decrease in the average hole parameter (pixel counts along

the periphery of the holes). In contrast, and as previously

noted, the perimeter of the cell has a different relation-

ship to histamine release [15, 26J. Direct proof that these

fluorescent-free regions represent the same degranulation

sacs observed by electron microscopy is not yet available.

If our previous studies of a PKC-like activity that trans-

locates to the cell membrane(s) following stimulation with

IMLP actually reflects the activity of PKC isozymes then the

prolonged generation of DAG appears disconnected from

its regulation of PKC. We are currently re-evaluating the role

and kinetics of PKC changes during these reactions and are

finding that translocation ofseveral PKC isozymes is not quite

as transient as the activity studies would suggest. These

newer studies will be presented in a future publication.

There appears to be selective generation of specific spe-

cies of DAG during the early period after fMLP stimulation.

We have yet to develop a complete profile of the species

generated but these studies clearly demonstrate that not all

species (with respect to those phospholipid species pres-

ent) are generated at all time points. Adjusting the total

DAG levels for imperfect recovery (see Methods), the late

reaction is characterized by peak changes of around 45

pmol/106 basophils. This is eight to nine times the peak

changes in SA-DAG observed. If the same profile of DAG

species were generated in the early reaction, this would

translate to 10- to 15-pmol (/106 basophils) elevations in

total DAG, which would have been detectable with kinase

642 Journal of Leukocyte Biology Volume 61, May 1997

assay. Therefore, the absence of observable changes in total

DAG but elevations in SA-DAG and OA-DAG suggest a se-

lective generation. As noted in Results, the early changes

in SA-DAG and OA-DAG were similar to the changes in

IP3 mass. This could be considered fortuitous because

there is no a priori reason to believe that the metabolic

rates of the two second messengers would be similar even

if they were derived from the same parent phospholipids.

This result also suggests that the predominant species of

phosphatidylinositol contain arachidonic acid in the sn2

position. We should be in a position to verify this predic-

tion in future studies of basophil phospholipid species with

the use of newly developed tandem mass spectroscopy tech-

niques [unpublished resultsl. These differences in the ki-

netics of SA-DAG and IP3 generation also suggest that SA-

DAG is derived from another source of phospholipids later

in the reaction. This appears to be a common theme be-

cause many other studies have noted that the source of

DAG appears to change during the course of a reaction.

It has been noted in rat peritoneal mast cells where it ap-

pears that the early phase of the reaction is dominated by

DAG derived from phosphatidylinositol [27J. In the later

phase of the reaction, a larger component of the DAG gen-

erated appears to be derived from phosphatidylcholine. As

noted, similar conclusions have been drawn from studies

in other cell types Ifor example, see refs. 19 and 281. In

addition to changes in the sources of DAG, the enzymes

responsible for its generation may also change 129-311.

To summarize, DAG generation in basophils is pro-

longed in a manner not consistent with an exclusive role

in histamine release or PKC (-like) activity measurements

but temporally more consistent with the phenomenon of de-

granulation resolution. Selective generation of DAG spe-

cies appears to occur early in signal transduction.

REFERENCES

1. Warner. J. A., MacGlashan, D. W., Jr. (1989) Protein kinase C (PKC)

changes in human basophils. IgE-mediated activation is accompanied by

an increase in total PKC activity. J. Immunol. 142, 1669-1677.2. Warner. J. A.. MacGlashan, D. W.. Jr. (1990) Signal transduction events in

human basophils-A comparative study of the role of protein kinase-C

in ba.sophils activated by anti-IgE antibody and formyl.methionyl-leucyl-

phenylalanine. J Immunol. 145, 1897-1905.3. Bishop. W. R., Bell, R. M. (1988) Functions ofdiacylglycerol in glycerolipid

metabolism. sjgnal transduction and cellular transformation. Oncogene Res.

2, 205-218.4. Bell, R. M., Hannun, Y. A., Loomis, C. R. (1986) Mechanism of regulation

of protein kinase C by lipid second messengers. Symp. Fundamental Cancer

Res. 39, 145-156.5. Kennerly. D. A., Sullivan, T. J., Sylvester, P., Parker, C. W. (1979) Diacyl-

glycerol metabolism in mast cells: a potential role in membrane fusion andarachidonic acid release. j Exp� Med. 150. 1039-1044.

6. Kennerly, D. A. (1987) Diacylglycerol metabolism in mast cells. Analysisof lipid metabolic pathways using molecular species analysis of intermedi-

ates. J. Biol. C/tern. 262. 16305-16313.7. MacGlashan, D. W., Jr., White, J. M., Huang, S. K., Ono, S. J., Schroeder,

J.. Lichtenstein, L. M. (1994) Secretion of interleukin-4 from human baso-phils: The relationship between IL-4 mRNA and protein in resting and stim-ulated basophils. J. Immunol. 152. 3006-3016.

8. Gilbert, H. S., Ornstein, L. (1975) Basophil counting with a new stainingmethod using alcian blue. Blood 46, 279-286.

9. Siraganian. R. P. (1974) An automated continuous-flow system for the cx-traction and fluorometric analysis of histamine. Anal. Bloc/tam. 57,383-394.

10. Bligh, E. G.. Dyer. W. J. (1959) A rapid method oftotal lipid extraction andpurification. Can. J. Biochem. Physiol. 37, 911-917.

1 1 . Kennerly. D. A., Parker, C. W., Sullivan, T. J. (1979) Use of diacylglycerolkinase to quantitate picomole levels of 1,2.diacylglycerol. Anal. Biochem.

98. 123-131.12. Preiss, J. E., Loomis, C. R., Bell, R. M., Niedel, J. E. (1987) Quantitative

measurement of sn-1,2,diacylglycerols. Met/i. Enzymol. 141, 294-301.13. Yamada, K., Jelsema, C. L.. Beaven. M. A. (1992) Certain inhibitors of pro-

tein serine/threonine kinases also inhibit tyrosine.phosphorylation of phos-pholipase Cgl and other proteins and reveal distinct roles for tyrosinekinase(s) and protein kinase C in stimulated rat basophilic RBL-2H3 cells.

J. Immunol. 149, 1031-1037.14. Hubbard, W. C., Hundley. T. H.. Oniente. A., MacGlashan. D. W.. Jr.

(1996) Quantitation of 1-stearoyl-2-arachidonyl.sn-3-glycerol in humanbasophils via gas chromatography-negative ion chemical ionization massspectrometry. Anal. Biochem. 236. 309-321.

15. MacGlashan, D. W., Jr., Bochner, B., Warner. J. A. (1994) Graded changesin the response of individual human basophils to stimulation: Distributionalbehavior of early activation events. J. Leukoc. Biol. 55, 13-23.

16. Dvorak, A. M., Warner, J. A., Kissell, S., Lichtenstein, L. M.. MacGlashan,D. W.. Jr. (1991) F-met peptide-induced degranulation of human basophils.

Lab. Invest. 64, 234-253.17. Lee, C.. Hajra. A. K. (1991) Molecular species of diacylglycerols and phos-

phoglycerides and the postmortem changes in the molecular species of di-

acylglycerols in rat brains. J Neurochem. 56. 370-379.18. Honeycutt, P. J., Niedel, J. E. (1986) Cytochalasin B enhancement of the

diacylglycerol response in formyl peptide-stimulated neutrophils. J. Blot.

Chem. 261. 15900-15905.19. Pessin, M. S., Raben, D. M. (1989) Molecular species analysis of

1.2-diglycerides stimulated by thrombin in cultured fibroblasts. J. Biol.Chem. 264, 8729-8738.

20. MacGlashan. D. W.. Jr., Peters. S. P.. Warner. J.. Lichtenstein, L. M. (1986)

Characteristics of human basophil sulfidopeptide leukotriene release: re-leasability defined as the ability ofthe basophil to respond to dimeric cross-links. J. Immunol. 136. 2231-2239.

21. Das, S., Rand, R. P. (1984) Diacylglycerol causes major structural transi-tions in phospholipid bilayer membranes. Biochem. Biophys. Res. Commun.124. 491-496.

22. Coorssen, J. R. , Rand, R. P. (1990) Effects of cholesterol on the structuraltransitions induced by diacylglycerol in phosphatidylcholine and phospha-tidylethanolamine bilayer systems. Biochem. Cell Biol. 68. 65-69.

23. Burger. K. N., Nieva. J. L.. Alonso, A.. Verkleij, A. J. (1991) Phospholipase

C activity-induced fusion of pure lipid model membranes. A freeze fracturestudy. Biochim. Bioph)s. Acta 1068. 249-253.

24. Francis, J. W., Balazovich. K. J. . Smolen, J. E., Margolis. D. I.. Boxer. L. A.(1992) Human neutrophil annexin I promotes granule aggregation and mod-ulates Ca(2+)-dependent membrane fusion. .1 Clin. Invest. 90. 537-544.

25. Agwu, D. E.. McPhail. L. C.. Wykle. R. L.. McCall. C. E. (1989) Mass de-termination of receptor-mediated accumulation of phosphatidate and diglyc-erides in human neutrophils measured by coomassie blue staining and den-sitometry. Bwchem. Bioplrps. Res. Commun. 159. 79-86.

26. MacGlashan, D. W.. Jr. (1995) Graded changes in the response of individ-

ual human basophils to stimulation: Distributional behavior of events tem-porally coincident with degranulation. .1 Leukoc. Biol. 58. 177-188.

27. Kennerly, D. A. (1990) Phosphatidylcholine is a quantitatively more impor-

tant source of increased 1,2-diacylglycerol than is phosphatidylinositol in

mast cells. J. Immunol. 144, 3912-3919.28. Lee, C.. Fisher. S. K., Agranoff. B. W., Hajra. A. K. (1991) Quantitative

analysis of molecular species of diacylglycerol and phosphatidate formed

upon muscaninic receptor activation of human SK-N-SH neuroblastomacells. J. Bail. Chem. 266, 22837-22846.

29. Mullmann, T. J., Cheewatrakoopong, B., Anthes, J. C.. Siegel, M. I., Egan.R. W., Billah, M. M. (1993) Phospholipase C and phospholipase D are ac-tivated independently of each other in chemotactic peptide-stimulated hu-man neutrophils. J. Leukoc. BiOL 53, 630-635.

30. Gruchalla, R. S., Dinh, T. T., Kennerly. D. A. (1990) An indirect pathway

of receptor-mediated 1.2-diacylglycerol formation in mast cells. I. IgEreceptor-mediated activation of phospholipase D. J. Immunol. 144. 2334-.

31. Dinh, T. T., Kennerly, D. A. (1991) Assessment of receptor-dependent ac-tivation of phosphatidylcholine hydrolysis by both phospholipase D andphospholipase C. Cell Reg. 2, 299-309.