B~c.~t~t~ e[lB~ophysica A cta, 604 (1980) I !,1 -246
© El~'evieffN¢ttth-Holland Bio~edical P~e~s
BBA 8 5 2 0 8
I N T R A C E L L U L A R PHOSPHOLIPASES A
H, VAN DEH BOSCH
State Univer,~ty o f Utrecht, L~boratory o f Biochemistry, P~du~l~'n 8, 3508 TB ,'Jtre..~t
./The Nethev~'ands) .
(Received M~rch 26th, 19~0)
Contents
1.
introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
II.
Rat flyer phospholipase~ A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
193
IlL
Purified ph~spholipases A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A. Phospholip~.s Ai . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B. Phospholipases A2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I~
1~
IV.
How is ,'he activity of membtane-bo~md phooholipases A re~,ulated? . . . . . . . . . . . .
203
A. D o in,?racellularphospholipases occur in z y m o p n
204
B.
C.
D.
E.
F.
191
fon~.~? . . . . . . . . . . . . . . . . .
Are p h ~ o H p a s e s regulated by q~ecific association w~th n~,z-enzym~tic l~Ot©U~,$? . .
F~egulation ofphospholipase A2 l~y availability of Ca~' . . . . . . . . . . . . . . . . . .
Influence of cyclic, adenosine monophosphate . . . . . . . . . . . . . . . . . . . . . . .
Effects o f h o t m o a ~ on phosphafipase activi*~t . . . . . . . . . . . . . . . . . . . . . .
Regulation of phasphol~pases b:" changes in membrane st~ctuze . . . . . . . . . . . .
V.
Sp~rific functions of inttac~llulat phospholipases . . . . . . . . . . . . . . . . . . . . . .
A. Release of pro~;tafi~nd~ pgecur~3rs . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B. I~ospholipid tu~nover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C. l~ynthesis of m~lecular species . . . . . . . . . . . . .
" ..................
D. 'Bacterial phospholipases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
VI.
Cor~cluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
207
2~ I
2t3
21~
219
22i
22i
22~
233
" 2~
23~
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References
...
~ . . . . . . . . . . . . . . . . . . . . . . . . . .
235
". . . . . . . . . . . . . . . . . . .
239
!. I n t r o d u c t i o n
T h e e x i s t e n c e o f etu~,yn~es cataly:fing t h e release o f fittty a c ~ f ~ ) m ~
p r o p o s e d over a h u n d ~ d y e a ~ ago, w h e n it w~s observed t h a t a ~ a ~ i ¢ ~ o f ~ t i d y l c h o l t a e w i t h p a n c ~ U ¢ juice resalted in f o r m a t i o n o f fr~e f a t t y ac~d~ (see R e t : I f ~
Abbrevia'tions: SDS, sodium dodecyl st.'ate; EGTA, othy~te gJyc~ ~ e ~ . V |
tetraaceti¢ acid.
e~Vj~-
192
historical review). The subsequent finding in 1903, that cobra venom converted phosphatidylcholine into a product which possessed the property of being able to hemolyze red
blood cells, d/rected much of the early research in this area towards the phospholil~idhydrolyzing enzymes in snake venoms and towards the elucidation of the structure of the
lyric product. Around i1915 it was established that this was a phosphatidylcholine from
which one fatty acid had been removed. For obvious reasons, the term lysophosphatidylcholine was introduced for this product. It was not until about 1960, however, that the
problem as to which fatty acid was actually removed by the snake venom phospholipase A was definitely settled as being the one esterified to the sn.2.position [2,3]. Such
enzymes are found in ill snake venoms and in the poisons of bees, wasps ~nd scorpions
[41.
lndlrect evidence t~)r the occu~rrence of phospholipase A in animal tissues other thar~
pancreas was obtained by Francioli [5] as earl?/as 1934, when appreciable quantities of
lysophosphatidylcholine were found in dried, but not in fresh or autoclaved, preparafion:~
of h~art, liver, spleen, brain, muscle, thymus and prostate. With the avaih~bllit:¢ of iso.
topically labelled substrates and the known positional specificity of the snake venon:L
phospholipases A it became clear in the mid-sixties that several mammalian tissues cor~tain~:d lipolytic activities attacking the fatty acyl ester bonds at either po,,~tion of the
glyc~'rol moiety [6,7]. Thus, intracellular phospholipases A were designated as At ar, al
A2 ~ctivities, produciag 2-acyl and 1-acyl lysophospholipids, respectively. The extracellular phospholipases A from snake venoms as well as the enzymes secreted by pancreas e~ll
helc,ng to the A? type [8]. The latter group of enzymes, .although studied in much mo~re
detail from both a rcechanistic and structural point of view [9-12], is not reviewed here.
Thi~t ",eview deals with the vast field of intraceHular phospholipases A and their £nvoh'emerit in membrane processes. Many proposals for the function of intracellular pi~ospholip~aes A have been put forward in the literature and I shall try to evaluate the ~.vallat~le
evidence. Of course, the question as to the possiblv structural relationship between filtra.
cel2~ular phospholipases and the well documented venom and pancreas phospholipase ,~2
will be considered. Also, some comparative aspects of intracelluiar lipases and lysophospholipe.ses will be included. While some examples of the widespread occurrence ofintracellular phospholipases A will be given, R is not the purpose of this review to give an
extensive treatise on this subject. Reasonably up-to.date enumerations can be foundl in
various reviews [8,!3-15]. Assay methods, with special reference to determine the relatively low activities generally associated with crude preparations ofintracelluiar phospholipases, were recently discussed [16].
Phospholipases A activities have been found in almost every cell where they have been
sought. Thus, such activities have been described in bacteria [17], amoeba [18], insf,'cts
[:t9], plants [20] and many mammalian tissues including liver [21], spleen [22], I!ung
[:.~3], brain [24], heart [25], retina [26], adrenal medulla [27], intestinal [28] and
s;:omach mucosa [ 29] as well as in erythrocytes [30], piatelets [31], polymorphonudear
h,ukocytes [32] a~ld alveolar macrophages [33]. Not oaly do many of these cells corrtaln
both phospholipa~e A~ and A2 activities, but fit addition their intracelluiar location i,,~not
restricted to a sin:~le subcelluiar site. Although there may be consklerable variation, ,~,spe.
dally between eukaryotic cells, in the distribution of phospholip~es A, the ubiqufl:y of
lhese enzy~les can be adequately illustrated by considering rat liver tissue.
193
li. Rat liver phespholipases A
Without going into details or discuss/ng in fuU length the controserzial res~its ~ t ?~ave
som,:iimes been obtained, it seems justified to suramarize the p h ~ h o ~ l ¢ ~ f i b u t ~
in the rat hepatoeyte as indicated in Table I.
Plasma membranes contain I~oth phospholipas¢ At [34,35] and p h o ~ h o ~
A2 [36,
38]. Both enzymes are optimally active at an alka~e pH of about 9 and req'~e Caz~ fo~
optimal activity. The ratio in which the two enzymes expr¢~ t h e ~ I v ~ m vRro a ~ _ r s
to vary considerably depending on the substrate used .'rod the method of/solafion of~he
plasma membranes [38,39]. Since pancreatic lipa.~¢has been shown to exhibit pho~q~holipas¢ Al activity [40], Newkfrk and WaRe [34] investigated wh~:thcr t h ¢ ' ~
1~¢mbrahe phosphofipase A~ activity could be ascrib~t to a lipase. They co~,]uded that this
was not the case' as no hydro!iysi3 of radioactively labeled tripalmitoy~yce~ol w'ds obsenred. However, Colbean et al. [41] reinvestigated the problem using ttiolcin e m p t i e d
in Igum arabic as substrate and found both a triglyceride and a monoglycc~d¢ I~[ms~ hq
pla..ana membranes. Both enz~,:meswere optimally active at pH 9. The d~,ercnfial eff~Ls
ob.~erved during heat treatment and solub'dization by salts and de~tergca~ tended to
gest that the triglycerLie fipau~, the monoglyc~ride Hpas¢ and the ] ~ s p h o ~
At w~e
dff['erent protein entities. In g,meral, such studies are subject to nxmy pitfalls. For exampie, it is difficult to rule out that the activity o:fa single, m c m b r . ~ t c d
ea~.-q~m~
wi~$ broad substrate specificRy is influenced to different extents.; by haat trtam~:nt,
pn,,sence of detergents or (partial) solubilization depending on the substrate used (see sub.
section IliA). The phospholipas¢ A~ activity could easily be g~h:.bilizcd fl~om ~ k ~
plasma membranes by hepa~t treatment [42]. Evidence was p~vidcd to sho~ that th~
enzyme is most li~ely identical to the phospholipas¢ At a c u i t y purifg'd f~om postheparin plasma [43]. This enzyme catalyzes bo~h hydrolytic a ~ l acybtransfer r e a c ~
with a variety of substrates and it has been suggested that the e ~
shoed -he m~re
appropriately named monoglyceride acyltran=fe~as¢, since monoglyct~'id¢ is the p~f~zied
substrate and transacylat/on is the predominant reaction ~44]. These fmd/a~ make it
very likely that the enzyme is idea-'.ical, after a~l, to the monoglyccrid¢ ~ a s ¢ in ~ s m ~
membranes, despRe the dit'feranccs observed by Colbaau et al. [41] and i~tus~-atet ~ 9~tfe,lls discussed above. The following discrepancy is probably a :Part.herexampk~ thcccof.
According to WaRe and Sisson [44], the heparia-solubilized em:ym¢ does not h y d e -
TABLE !
PHOSPHOLIPASEA DISTRIBUTIONIN RAT LIVER
+~.stimulated; - , inhibJcd; o, no effect; n.J., not investgal:ed.
Sul~-cllulatsite
Mainactivity
Plasma membrane
Micaosornes
Golgimemlnan<
Mitochondt~a
l,ysosomes
Cytosoi
At *
A2
A!
At +
A2
A2
Al + A2
At + A2
AI
pH opti~m~
Ca~e-effe:t
Regc~'g.nc~
8.0-9.5
8 0-9.5
8.0
8.0-9.0
+
+
~4-39
21,35,37,47
+
3?
+
4.0-5.0
-
7.4
3,6
n.i,
35,5~55
35,58,59
21
60,61
o
194
triacylglycerol. Inasmuch as this enzyme is thought to be identical to the post-hell~arin
plasma phosphoHpase At [43], it is relevant to note that Ehnholm et al. [45] haw~ isolated a hepariu-released phospholipase A~ from human pla:~na with almost equal hydrolytic activity against triolein and phosphatidylethanolamine. Some evidence has been obtained which suggests that a large part of the phospholipase A2 of rat liver plasma ~,llembranes is present in ma inactive form which can be converted into active enzyme under
certain conditions [39] (see subsection IVA). Ol~ly when present in its active form could
the enzyme be released from the membral~e b y h.~ghsalt concentrations. The memf)ranebound enzyme showed normal Micha~lis-Menten kL'~etics with sonicated phospinLtidylethazlolamine ~s substrate. In contrast, thc~solubilized phospholipase A2 showed irregular
kinetics with a marked increase in activity above 35 I~Msubstrate concentration [46]. A
s~milar V/S cun'e was noticed for the phospholipase Ai in plasma membranes. Critical
micelle concentration (CMC) determination in the assay medium by ~,~orescenc,~ measurements yielded CMC values in the rangr~ of 15-30 ~M. The sharp increase in activity
above 35/.tM substrate was attributed to formation of |ipid aggregates, ~nd resembles the
striking rate enhancement found for pancreatic phospholipases A2 when assayed with
substrates above the CMC [10]. It should be noted, however, tha~ the measured CI~,iCvalues are rather h i ~ and evel~ several-fold greater than others have reported for lys~phospholipids [280,313]. The kinetic behavior of the enzymes and oth~r data such as pH
dependency and preference for endogenous or exogenous substrate led the authors to hypothesize a different localization of phospholipase At and A~ in the plasma membrane
[46]. Thus, phospholipase At activity would be caused by a peripheral membrane protein
with exposure of its active site to the external medium and would be most actiw.' with
exogenous substrates. In contrast, phospholipase A~ would be much more embed,led in
the membrane, would require the penetration of exogenous substrate prior to hydrolysis
and therefore show a preference for endogenous substrate. As pointed out by the authors,
these suggestions of a different localization of active sites in the membrane should be considered ~entatively until more direct evidence becomes available (cf. Fig. 1).
Tile microsomal fraction isolated from rat liver contains both phospholipase ,~,~ L~nd
A2 activities, but invariably the A~ activity was reported to be much more prov,finent
|21,35,37,47]. Attempts to differentiate between the phospholipase A~ and a microsom~l triglyceride lipase have yielded some indications that distinct enzymes may be
involved. No phospholipase At inhibition was observed with p-chloromercuribet:tzoate
Fig. 1. Schem~ticmodel for the activation of membrane-boundphosphollpaseA2. Left, inacti'~econformation; r~tt, active conformation.As discussedin the text, two possiblemechanismsfor this activation can be put forward, i.e., proteolytic action convert.ng a prophosphoUpase into ee, active
enzyme or complex formation betweeh an inactiveenzyme ~:tdan as yet undefined platelet favorerto
yield active phospholipaseA2. The modeldepicts schematicall!~the observationsmade during tile activation of live: plasmamembranephospholipaseA2, namely a decreasein hydrophobic binding ~ccompanyingth~ conversionof inactiv.~enzymeinto activeenzym~ 139].
195
under conditions causing 30% inhibition o1" the m~cro~omal Iipase [2I i- A ~ e q ~ t
investigation [41] demonstrated equal se:as~tivity to heat tr~.atment and s~nilar extractability by increasing salt concentrations for both enzy:natic activities. ~
these ~ r rations are in iine with the existence of a single enzyme, partial d e ~ t i o n of the
microsomes wiq~ either acetone or phospholipase C had different effects on trig[yce~de
and phosphati0, !~qhanol,'u~i:l~,:-hydrolysis. Whether these difference~ are related to varying physico.chemical properties of the two substrates or point to the presence of d~ tinct prot,~in entities cannot be established with certainty at pres;~.nt.
The few studies that l~ave been performed to measure the phcspholipase A acfivifi~/~f
highly pu~.'ifiedGolgi mewbr;mes have indic:Red the presence of AI and At activic~es~~ t h
phospholipase AI predondnating [37]. The purified Golgi membranes lacked sever-~llipid
biosynthetic activities, Le., chol~lepho~photransferase, acyl-CoA : 1,2.diacyl-m-giy~zro[
acyltransfera~ and acyl-CoA : l..acyl.sn-g[ycero-3-pho~,,phochoIhleacyltr~fferase [37,
48]. Inasmuch as the latter enzvntes are l¢,cated either exc~_sively or mainly in the emloplasmic reticulum [14], these resu:!tsdemoJ1strated that the phospholipa.¢~ Al ami A2 are
true constituents of Golgi membranes and are not the result of mierosomal c o n t a m i w ~
as the similar positiona.~ specificSJies observed for the microsomal phospholipasvs m i ~ t
su[Igest. It should be noted, however, thal the exclusive localization of c h o ~ , ~ transferase in endoplasmic reticulum merabrane has recently been questi~'~d..lel~,m~
and Morr6 [49] concluded that, although nearly 90% of the total activity of c h o ~ i ~ photransferase was accounted for by endoplasmic reticulum, the specific activity of ~he
enzyme in the Golgi app.lratus was 40% of the endopl~rnic reticulum. This high v.~e
could not be accounted for on the basis of the approx. 5-10% c ~ n ~ t i o n
of the
Golgi fraction by endoplamuc reticulum. Inasmuch as the sp~zifie activities of the p h ~
pholipases A~ and A2 in the Golgi fractic,n were 50 and 130%, respectively, of the values
found for endopla~mic reticulum [37], the conclusion that these enzymes are true constituents of the Golgi apparatus seems to be still valid.
Mitochondria contain ~ Ca2+-dependent phospholipase A~ with an ~ e
pH ¢~p~imum. The enzymic activity appears to b,.• present in both inner and outer mi:~)cbondriaI
membranes [35,54], although it is not yet kne.va~whether the same protein e~fity is present in both membranes. The apparent Km value for phosphatidytethanoL~.~ae was
reported [35] to be several-fold smaller for the enzyme in th~ outer membrane (90 as
against 400/.tM), but this ,:an ~so be exFlained by a different environment for the ?rerein. The phospholipase h ~ been partially purified from tt~[ rat liver mitochoa~-i~ by
Waite and Sisson [55]. The !.60-fold ~ufified pseparation hy~o]yzed phosphatidylethanolamhle 6ptimaUy at pH 9.5, but phosphatidylserine hydrolysis was rao~ extet~ve
at pH 7A, ~hus, clearly indicating the effi;ct of environmental factors. So far, the mitochondrial enzyme constitutes the only n~a.~mbrane-boundp ~ o l i p a s e A from rat ~ver
which has been purified to a reasonable extent. Unfortunately, further purif'~atioa was
hampered by extreme inst.ability resulting in complete loss of a~tivity.
In conh'ast to the membrane-bound phospholipases referred to above, ra¢ ~iver
seines were found to contain soluble phospholipase A activities. InitiaLty, it was a ~ e d
from the lack of accumulal~ion of lysopho:,-phoglycefides that one enzyme catalyzed the
complete deacylation of p.hospboglycerides [56,57]. Stoffel and Trabe~ [58] were the
first to suggest the presence of two phosphdipa~ with different poeitior~ specW~cityon
the bash of selective inhibition studies. These u~ults were co~Lfumedby Na~L,b a ~ eta].
[35] and F~anson et aL [59]. The latter au~t~or~partially separated ~
A~ and
A~ by gel filtration of the lysoso~fl soluble fraction on Se~tdex G-200. R e m a k e ' y ,
196
both enzymes were found to be inhibited rather than stimalated by Ca2÷, although to ,:lifferent extents, with phospholipase A2 activity being more sensitive. It i~ not unlikely tllat
the pbospbolipase AI activity may be associated with acidic lysosomai lipases or ester,~:~es.
Acidic triglyceride lipase was mainly recovered in the ~3]uble fraction ol~lysosomes [41 ].
Incubation of thi.,; fraction at 60°C resulteci in a more rapid loss of lipase activity whe:~
compared with F~ospholipase acti'.'ity, but this should not be used ~:~an indicatio~ of
separate enzymes since the phospholipase A in these studies was not defined as A~ or
A2 activity.
Soluble phospholipase activities have also been detected in the cytosol fraction obtained from rat i.;ver homogenates [21]. Due to the presence of active lysophosphoEpase(s), the initial site of attack in the hydrolysis of phosphatidylethanolamine cc,uld
not be ascertained Inhibition of the lysophospholipase(s) with deoxycholate led to the
accumulation of both l-acyl and 2-acyl lysophosphatidylethanolamine, indicating the
presence of both phospholipase A2 and Al. These activ,:,*.ieswere measured at neutral pH
v,tlues. It is not known at present whether these enzymes occur in the cytoplasm of iw:act
cells or if their recovery in the 100000 Xg snpematant is due to solubilization of these
enzymes from other subcellular structures during the homogenization and fractionaLion
procedure. This question applies particularly to a recently described acidic phospholipase Az, which was found in the cytosul of many mammalian cells, including rat Iliver
[160,61]. This enzyme could not be detected in the absence of detergents, but was
markedly activated in the presence of acidic phospholipids, iysophosphatidylcholine and
•,~me non-ionic detergents such as Triton X-100 and Tween 20. Under these conditions,
the enzyme exhibited highest activity at pH 3.6, thus resembling the lysosomal sol'able
phosphc,lipase A~. Based on several criteria, the authors believe, however, that the cyto:;olic enzyme is different from the lysosomal, one.
It is clear from the above discussion that many basic questions about the intracel:inia:
;phospholipases h;~veremained unresolved. Is there any structural relationship between the
phospholipases Az and the pancreatic and venom enzymes? Do the various subce]llular
znembrene stroct.Jres contain identical protein entities with either phospholipase A2 or
Az activity or do~s each sobcellular structure possess different phospholipases? Likewise,
the relationship, ff any, between intracellalar phusphollpase At and lipase activities
remains largely unclear. Obviously, the answer to such questions must await purification
of the enzymes and a careful comparison of the properties of these enzymes in a purified
and reconstituted membrane-bound state. Considering that purification is a prerequisite
to fulI characterization of enzymes, the vas~ field of ~atracellular phospholipases bec:omes
suddenly very snuzlLNevertheless, several successful attempts to purify intracellular phospbolipases have been undertaken in the last decade. The results thereof will be di~:ussed
next.
i!i. Purified phozpholipues A
IliA.
Pho~izool~t~esA !
b Most of ~ e i n t r a c e l l ~ phozpholipa~m A that have been purified in earlier iRudies
emnged to me clam of Ai~type enzym~ (Tab!~ II), Seandella and Komberg [17], taking
advantage of ~ nuu'ked ~bflity of a meml~rane-bound phmpholipase A of Esche~'hiu
~ l i B in SDS solutions s a l t e d with butanol, purified this enzyme approx. 5000..fold to
PAl
PAl
PA i
PAt
PAl
PA 2
PAt
E. coli B
~. c o i i K . i 2
~i. phlel
~ "=~.~*':'-~;.~m
Pa~cJ-~,o
P, note, turn
Br~a
MB
MR
MB
S
S
S
S
Form
29 000
28000
45 000 *
26 000
60000
116 000
75 000
Mr
8.4
8.0
8.0 **
5.5-6.5
7.5
4.0
4.2
pit optimum
" A n o t h e r enzyme with M r 27 000 ~'~owcd ~L'nfl~r :ub*',ote s,,,-;ft,,Itv
** For neuUal phospholJpids; acidic phospholipids were optimally hydro!yzed at about pit 4.
+ I,Pase
+ LPase
+ t Pa~
+ ."A~- "~
. . . .o°¢~.
..,
+ L[,z~
Mare activity
Source
-
4.
+
Ca 24' effect.
2.9
4.7
0.'7, 7.2
1560
0.8
27
0.04
Specific activity
(units/rag)
P A l , phosphol/pase A I ; PA2, phospholipase A2 ; LPase, lysophospholipase; MB, membrane-bound; S, soluble. +, requi~ed; --, not requ~ed.
P t J R i F i E D PHGSPHOLIPAo°ES At AHD PHOSPHOL!PASF~g a
TAPLE ll
67
68
69
70,71
72
17
6~
Reference
198
near homogeneity. The purified enzyme was identified as a phospholipase Al by its product fomlation from synthetic mixed-acid substrates. The lack of triolein hydrolysis distinguished the enzyme from pancreatic and fungal lipases. Phosphatidyl¢thanolamiv,e and
phosphatidylglycerol were hydrolyzed at the same rate, either in the presence or allsence
of Triton X-100. In contrast, diphosphatidyiglycerol hydrolysis was stimulated about
100-fold by the addition of Triton and then proceeded at comparable rates as observed
for the other major phospholipids of E. coll. 1-Acyl lysophosphatidylethanolamine was
hydrolyz~.d +,wice as fast as phosphatidylethanolamine under the standard assay conditions, i.e., hi the presence of Triton X-IO0. Apparently, 2-acyl' ly~ophospholipids ~Lrenot
hydrolyzed as they represent one of the products obtained by actiou of the enzyme on
diacyl phospholipids. On the basis of its substrate specificity, the enzyme can be d~noted
as a phospholipase Al with lysophospholipase activity.
Nishijima and coworkers [62] have isolated an enzyme from E colt K-12 which most
likely represents the K-i 2 analog of the E. colt B enzyme isolated by Scandella and Kern.
berg [17]. In botl~ strains the enzyme appears to be located in the outer membran~ [6365] and is solubilized and partially purified by identical procedures [62]. However, there
appears to be some difference in the positional specificity and sensitivity to SDS of the
enzyme from the two sources° The E. colt B enzyme produc~;d only 2-acyl lysophospho.
lipids, while the K-12 enzyme formed both l-acyl and 2-acyl ]ysophospholipids. This indicates that the initial attack in the deacylation of a diacyl ph~.,~pholipid can take place at
either the 2-position or the l-position. The relative activity of the inP.ial deacylation reactions could not be ascertained because the purified enzyme also h~-drolyzed both po.'i.
tional isomers of lysophospholipid. The rate of deacylation of l-acyl lysophosphatidyl.
ethanolamine was 5-times that of the 2-acyl isomer. As for the E. colt B enzyme, these
reactions were carried out in the presence of 0.05% Triton X.IO0. The latter stimulated
the hydrolysis of l,acyl lysophosphatidyletl~anolamine somewhat, but inhihited by at
least 80% the breakdown of 2-acyl lysophospbatidylethanolamine. Under the proper conditions, the K-12 enzyme, and probably also the E. col~ B enzymu, should bf~ able to
catalyze a corLlplete dealcylation of diacyl phosphollpids and thus c~,1also he classified as
phospholiipases B according to the nomenclature rephrased by McMulriy and Mal]ee [66].
It follows from the above description of the substrate specificity tha t much of the apparent specificity depends on the detergent concentration. The action of this enzy~cnein the
in situ situation will be discussed in subsection VD.
An em:yme with similar, but not identical, substrate specificity w~s purified from the
membranes of' Mycobacterium phlei [67]. Hydrolytic rates for 1-ac:~l and 2-1tcyl lysophosphatidylethanolamine were equal and about 2-fold higher than with pho:~phatidyl.
ethanolamine when tested at the same substrate concentrations, Yet, during iJ~cubation
with phosphatidylethanolamine, small amounts of lysoph¢,sphatidylethanolamine accumulated which were shown to consist almost exch,sively of the 2-acyl isomer. Thus, this
enzyme can also be classified as a phospholipase B, but differs from the E. colt K-12
enzyme in the initial site of attack of diacyl phospholipids, which in this case :is predominantly the ester linkage at the l-position. It was stated that the rate of trioleh~ hydrolysis
was less than 1% of that for phosphatidylethanolamino, but it ~mu!d be noted J:hat values
of 0.7 an,:!7.2/.unol • rain "1~• mg-1 for phosphatidylethanolamine hydrolysis are reported
[67].
The most active and probably also the most specific phospholipase A~ il~Jsbeen obtained from Bacillus megaterium spores [68]. In its action on phosphatidylg]ycero| the
purified ¢,m:yrae was stimldated about 2S-fold by either the non-ionic deteri~ent 'l~riton
~99
X-100 or ani,3nic detergents such as taurc,chola'.~e or dcoxycholate. In the p r e ~
of
Triton X..IC~, anionic phosphatidylglycerol was by far the best substrate ~'ld ~ t a tidylethanolamine was virtually inert, tlowever, phosphatidylethanolam~e hydro|y~ was
stimulated ~bout 100-fold in the presence cf taurochohte. Under these conditD~s, ~ c ~
phatidyletharLolamine hydrolysis amount~:d to 50% of l ~ t of ph~phatkiy~gtycc~d.
These data support the v.ew that the enzyme Frefers nega~tivelycharged sab~rate-detergent complexes. Tributyfin, even ir~ the presence of taurocholate, was hydro]yzed at a
rate l~ss them 0.2% of th~Lt of phosphatid~lcl~c.line and p h o s p h a t i d y l e t h ~ .
S/~ce
the lysophospholipids ~,roduced were the 2-acyl isomers, the enzyme seems to be f ~ l y
specific for the 1-acyl ester h*nkage of pho:;pholipids, i.e., it acts as a ph~phc~rpase A~. It
should be mentioned, howeser, that some activity was also noted with l - ~ y | [ysopho~phatidylglycerol and that its substrate wa~ only tested at low concentrat~ns !~ ~he presence of detergants. The latter are know~ o inhibit most lysophospho~pase sc~t/e~z. It
cannot be ruled out, therefore, that the enzyme ,:an cataly:~e the comptete deacylatg.q~ of
diacyl phospholipids in the absence of detergents.
It is well known that detergents, by m3difying the physico-chemk:a] state of the s~bstrate, exert a great influence on the rates of lipo~yt,~ reacti.ons. The discu~on L~ this section has provided several examples of thi~; phenomenon. It sht~ld be real/zeal, however,
that when carded to it~ extreme this can completely char~ge the apperent su~s~rate specificity of art enzyme. We have demonstrated this for a protein isolated to homogeneity
from bovh~e pancreas. In the absence of deoxyeholate the en:~me was fully active on
l.acyl and 2-acyl lysophosphoglycerides with a 5- to lO-t~aldhigher rate with ~he |-acy[
isomer [69]. Intact nat'ural ]?hospholipids were hydrolyzed at rates less than 1% of those
observed for l-acyl ly~pho~pholipids. These rates were ',timulated 25-fold by add~c2oaof
optimal amounts of 1 mg l~¢r ml of deoxycholate. Under these conditions, e q ~
amounts of fatty acid and 2-acyl lysophospholipid were produced, indicating tha~ Lhe
enzyr~le acted as a phc,spholipase A~. It turned out ha subsequent studies tha~ the deoxychelate concentrations used to obtain m~dmal stimulation of pkospholipase A~ acti~fity
caused complete inhibition ,~f the lysophospholipase acti~'ity of this enzyme, in ~he presence of intermediate levels of detergent, both activities were expressed an~ the enzyme
acted as a phosphofipase B catalyzing the complete deacylation of pho~hatidylchog,,~e
and phosphatidyletha~olamine.
Similar observations were made for a phospholipase li~ purified to h~aogeneity, from
Penicillium notatum [70,71]. The rati,~ of lysophosi~atidylcholL'~e to p h ~ k ~ y b
choline hyd~:olysis changed from 100 : 1 in the absence of Triton X-100 to about ~ : ~ in
the presence of this detergent. Thus, o~dy ha the p~sence of Triton X-100 w ~ some
accumulation of lysophosphatidylcho]ine seen. This enabled the authors to d e t e m ~ e the
sequence of release of acyl groups d u ~ g incubation with phosphatidyh:ho[in,, lr~ contrast ~o the pancreatic enzyme, the P. notatum phosphc/iipase B pvedommmtly a~t~cked
the diacyl phospholipi~d initially at the 2.position.
An 80-f01d purified phospholipase At, presumably of lygksonmi origin, was ~ b ~
from acetone powders of human brain, l'he e~zyme specific~y release~ fatty acids f~om
the l.position of phosphatidylcholine and phosphatidylethaadamia¢ in ~a ~,~ay m e ~ m
containing Triton X-100 and taurocl'o~ate. Hydrolysis of l y s o ~ o ~ , ~ p / ~ s m ~ e
absence o f detergents was not studied [72].
In su~mnary, the enzy~es described in t l ~ sectio~ ba~e withy v~yiag m o i e ~
weights and specific activities, Some require Ca~ for their a~tivi~, wh~e ~ g ~ s do ~ .
Some occur intracellulzdy in membrane-bound form and others in sotub~ fom~. In n ~
200
cases, the iiysophospholipase activity is at least comparable to ~a~ usually r~ch greate:~
than the p!~ospholipase A activity, but this varies greatly with ertviro~tmental I~actorssuch
as the nalure and concentration of detergent. Probably, most of the enzyr~tes listed h~
Table II @~owphospholipese B ~.-ti~ity under ~.ppropriate conditio.us. In this regard most
enzymes attack the diacyl phospholipid initially at the sn.l.p,3siticn, while: others firft
attack the sn.2-position (/~. notatum) or show less preference fi)r either posit~on (E. coli
K-I 2). Virtually nothing is known about the factors determining the seque~tial release of
acyl groups from the different positions in the diacyl phosphblipid molecule. Little is also
known about which substrates are attacked in vivo by these enzy.~es and t!~e nat:~re ~,r
fate of the produ,:ts fcrmed. Such information is difficult to deduce frorr~ the in vitro
substrate specificity of these enzymes in view of the influence,~ detergents e:~er~ thereo~.
In this re:;pect, it will be interesting to see what action the enzymes have o~ membraneembedd-.d phospholipids and lysophospholipids and which in tracellular fa ~tors, if any,
mimic the in vitro function of detergents,
The lack of specificity for the sn-2-position and in most c ~ s the lack ~3f a clear-cut
Ca~÷ requirement as well as the molecular weights in SDS gels clearly dis'lingnish these
enzymes from the venom and pancreas phospholipases A2.
lllB. Phospholipases A 2
The purl.fled pho.~pholipases A2 are listed in Table I!I~ "i~,.* rat spleen er~yme was obtained in ~olubie form fto.m rat spleen bomogenates by son',:ation, indicatin 8 that this is
probably not a membrane-bouna enzyme. The purified eeL yme showed an isoelectric
point of pH 7.4, a requirement for Ca2~ and specificity for t~,e sn-2-positi~:mof phosphatidylath~molamine. Hydrolysis of phosphatidylcholine wa~ .~hibited by deoxycholate,
whereas no significant inhibition was found with phosphatid yiethanolan~te as substrate.
These properties are similar to those shown by pancreatic phospholipase A2, except that
phosphatidylcholine hydrolysis by the latter enzyme was stimulated by adding deoxycholate [74].
The presence of an active phospholipase A2 in the insoluble pulmonal:y secretions of
patients with alveolar proteinosis was demonstrated by Sahu :rod Lynn [7'i ]. This enzyme
is included in Table Ill, ~nce it is one of the very few manm!~alianphosphc)llpeses A2 that
have b~n purified, to ,date, from sources other than pancreatic tissue. It is uncertain,
howew:r, whether this ~ an intraceUular phospholipase and from which ~:elltype it originates. Since alnvay cells or inflammatory cells are potentia~ sources of t~is enzyme, the
authors also studied pho~,hollpase A2 in isolated human polymorphonu,:]ear leukc~'yt.es,
rabbit alveolar macropb~ges and pig tracheal mucosa. The l~Lttercell types contained any.
where from S. to 10-t;mes less phospholipese activity, on a mg proi:ein basis, when
assayed Imder optimal conditions far the pulmona.ry secreti~m enzyme. S,~arting from lyophili~;d powder, obtained after therapeutic bronchoalveoktr lavage, the enzyme was sol.
ubiliz,~d by dellpidation and purified to homogeneity [76]. The molecular weight as estimate(~ by gel filtration and SDS-polyacrylamide gel electrophoresis amc,unted to 75 000
in both cases, strongly suggesting tha~:the enzyme consiste(~ of a single p~)lypeptide c!:~n.
This was corroborated by the flndinl~ that only a single N-terminal resid~le, alanine; could
be detected. It is interesting *.onote that the same N-terminus has been found in thr. pan.
creat,2c phospholipases A2 [77]. No sugar residue could be detected in the phospholipa~; Az from pulmonary secretion. This property as well as the Caz~ ~equirement and
stimulation of phosphatidylcholine hydrolysis by deoxycholate is shared with the pancreatic enzyn~i.
S
U
U
MB
MB
MB
MB
Rat spleen
Human pulmonary secretions *
Rabbit exudate *
P.~b~t PMN lankocytes
$ h ~ p erythrc,,:ytes
Rabbit platflf4s
Human platelets
" (lntra) cellular ~urc~ unknown.
Foh.'~
Source
i 5 000
75 000
14 800
14 000
12 000-18 500
12 000
44 000
+..+u-
7.0
8.0
8.0
7.5
80
9.0
9.5
pH optimum
+, requiP~d; S, soluble; MB, membrane-bound; U, unknown; PMN, polymorphonuclear.
PURIF,~ED PHOSPHOLIPASE A2
TABLE 111
Ca2+ effect
5.0
1140
0.5
5.0
4.7
0.8
0.5
t , . + , , , + w ,,.~p
Specific activity
78
317
79
o3
314
73
....
I~, In
Reference
202
Peritoneal exudates, produced in rabbits by injection of ~ycogen h:l saline, contained
a soluble phospholipase A2. A 300,fold purified preparation sh~wed ,:~.~leband in normal
and SDS-polyacrylamide gel electrophoresis. Gel filtration and SD$-I[[el electrophoresis
both gave an estimated molecular weight of 14800. The en~yme required Ca2+ and
showed a strict specificity for the sn.2-position [78]. The striking! similarity in the
enzymic and l~hysica~ properties of the peritoneal fluid phespholipa~e A2 and a membrane-associated phospholipase A2 from polymorphonuclear leukoc~,tes suggested that
the soluble phospholipase in the exudate is probably of leukocyte ¢ "igi~l.
A membrane-associated phospholipase A2 from rabbi~i polymorpilonuclear leukocytes
wa~ recently extracted from intact cells by treatment with 0.16 N H2S 0~. The so[ubilized
enzyme was purified over 8000-fold to yield a preparation with a specific activity of
5 ~anol. rain "i •mg -! when assayed with autoclaved F coil as substJ:ate [317]. According to the authors, this specific activity is comparable lo that of pul:e phospholipase A2
from other sources in this assay system. The molecniar weight of' this phospholipase A2 is
about 14000.
Another clearly membrane-b0und phospholipase A2 which, has ~een ~urified to near
homogeneity was obtained from sheep erythrocytes by Kramer et al. [79]. Ghost membranes were extracted at low ionic strength to effect the release ,>f weakly bound memb)~ne proteins. Optimal solubilization of membrane prot.eins, including the phospholipase A2, was achieved with buffer containing 0.5% SDS. This detergent inhibited the
phospholipase A~ almost completely and was replaced by chelate, wh;ch itself was unable
to solul~ilize the enzyme from the membranes, in a suhseq~ent gel-exclusion chromatography step using Sephadex G-7S. The active fractioa was further purified on an affinity
adsorbent to yield a 2800-fold purified e~zyme, which gave a single band on SDS-gel electrophoresis. The affinity adsorbent, initially developed by Rock and Snyder [80], was an
alkyl ether analog of phosphatidylcholine coupled as li~gand to AH-Sepharose 4B. The
procedme is based on the fact that some, but not all, phnspholipases A2 only bind to
their substrate in the presence of Ca2~. Taking advantag(~ of this property, the enzyme can
be ehited from the affinity column with buffer containing EDTA. As demonstrated by
Kramer et al. [79], this principle can still be applied in buffers containing 0.5% (w/v)
chelate. A molecular weight of approx. 12 000 was estimated for the sheep erythr:~cyte
phospholipase A2 from its behavior on Sephadex G-75 ,:ohmms. Co:.~parison with marker
proteins in SDS gels yielded an apI~irent molecular weight of 18 500. Studies on the substrate specificity of this enzyme demonstrated [81] a preferential re~leaseof polyunsaturated fatty acids, especially C22 fatty acids, compared to mono- and diunsaturated C18
acyl chains. This preference was observed equally w~ll with phosphatidylcholine and
phosphatidylethanolamine. Such a selective cleavage of acyl ester bonds in response to
variation in chain length and unsaturation was not found with L~ancreatic phospho.
lipase A~. The sheep erythrccyte enzyme was about 10-thnes more active on phospha.
tidylcholine than on phosphatidylethanolamine and phosphatidylglycerol, while phosphatidylserine was not attacked at all. This preferen¢~ for phosphc~pid polar groups is
thought to be related to the lipid composition of ruminant eryt~ocyte membranes,
which are remarkably low in phosphatidylcholine referent. It is interesting to note in this
respect that the phospbatidylcholine content of sheep erythrocytes increased up to 5-fold
upon incubation with sheep [82] or human [83] serum provided i ls phospholipase was
inhibited by EGTA or inactivated by pronase treatra,~nt. Inactivation of the phospholipa~e in intact erythrocytes by pronase or in resealed and leaky ghosts by chymotrypsin
[84] led to the conclusion that the phospholipase is ol~iant~,.dtowar~ls the exterior of the
203
cell. Even after extensive degradation of memb~ne p r o t e ~ , approx. |0% of the in/~/ai
phospholipase activity remained. This may indk:ate that a part of ~h¢ enzyme, ~[u<~fl~g
the active .~ite, is embedded in the hydrophobb: core of the lipid b~ayer. [n ~
raembranes the phospholipase was found ~o be inactive at Ca:" concentrations bek~w !O-s M.
A rapid increase in activity w~s obs~,:ved between 5 - 10 -s ~'d 5 - | 0 "4 M.~f~.er w[~:h a
plateau was reached. Considering that the plasma conc,:r,:~ ,tion of Ca z+ "~ 1.5 m2~
that the en~,yme is localized c~n the outside of the cell, these data seem to Lad/care that
enzyme aclivity is not regulated by the availability of Ca2" [84].
A conskierable purification G020-fc4d) has recently been described for a p h ~ . ~
lipase A2 from rabbit platelets [85] Heat treatment of platdet s/o~,ates at pH 5, fo~lowed by e:~traction of particulate material with 1 M KCI, so~bflized ab,,~t 50% of t~e
alkal/ne phcspholipase activity. Furt,~er purification was achieved through get f'fl~ration
and CM-celhlose chromatog~ai~,hy. No detergents were required to keep the e~.yme ~a
soluble forr~ throughout these step~. The purified enzyme preparat~a w ~ not y ~
homogeneot;s. The authors stated ~:ha:tthe final preparation contained oniy a a e ~
amount of [~hospholipid, but the actual data, i.e., 1.4 nmol phosphoh'p~ per #g protek~,
are about rwice the phospholipid-to.proteia ratio found in most biomerabh~an~.
The mol~,~cularweight of the rabbit platelet phospbotip~, A~ is at variance with
molecular weight of 44 000 reported recently for a homogeneoas p h ~
A:~ from
human platelets [314]. The latter enzyme w~s purified by a two-step procedure ¢zap~ying solubil~,,ation from the membranes with HzSO~ tbllowed by a f f ~ t y c l ~ a - ~ a t ~ p ~ y .
An overall enriclunent of about 1300-fold was obtained by these p~oeed~e~ Oa ~ e
other hand, Franson et al. [315,316] have obtained preparations of the m e m s ~
associated phosphofipase A: from human platelets which were 3500-fo1~ par~ted compared with platelet homogenates. The molecular weight of their isolated p ~ o ~ o lipase A2 was estimated to be about 15000 (Fran:~on, R.C., pem~na~ c o m m ~ i o a ) .
The reason for this discrepaJ~,cyis currently unknown.
In summary, the purified phospholipa.~es listed in Table 111 clea~y s~ow Az spec~city
and province equimohr a~aounts of l-acyl lysophosphthipid and free fa~y ~c~L AR
enzymes require Ca2÷ for their activity, although the latter ~
~y
fron~ c ~
enzyme to the other. Most of the proteins have n~olecuiar weights corapa~ab~e ~o ~.~ ~ creatic phospholipase Az.
IV. How is the activity of membrane-bonad phespl~lipases A regahte~.
With the availability of purified p~-ospholipases A derived from m~ace~u~r membranes it is possible to examine several obvious questions, Is the p h o s p ~
~a i a t e ~
or a peripheral membrane l,,rctein? On which side of the membrane is the e ~ ,
o~ at
least i.ts active site, located? How much of the protein is buffed in the ~ip~d~ilaye~'? Can
the membrane-associatet e~yme ~nly hydrol~ze phospholipids ha c~te ngmohye~' ~" a
biornemb:ane or does it hav,~ access to ~ e phospho~pids ha both moaolaye~s? If so, ~s t ~
achk.,ved through protein movemen~ :n the tra~.'e~se plane of the b ~ y ~ or ~
transmembrane movement of the ph~xsphv$1ycerides? The asTmmeL~¢ ~ b ¢ ~ m
of
phospholipids and the transmemb~ane movement of these ¢0~astituents ha b~c~e~-a~ra~es
have re~.~enflybeen reviewed [86,87]. ~etha~ the most important q~.~est~ is ~ a s ~ - t
of ~egulation of the phospholipase activity, t h e raerab,raae-asso~ted ~ p a ~ , s
¢~a
be viewed as enzymes floa~.hag ha a sea of substrate. What factors la~evea~and ~ i ~ t e
digestion of cellular phospholipids in vivo? It is not very l~keIy that ~ a c t ~ a a s ~ to
204
these questions, ~hich apply to all membrane-associated phospholtpases, can be given.
One is theretbre fac,~d with studies of special cases and one has to attempt to abstract
general regulatory mechanis:,as front these studies. Helpful in these respects are a large
number of observations, whtich show an activation of phospholipase A activity ;as the consequence of a wide range of stimuli or treatments (see subsections IVB, C and E for specific examples). In these circumstances, the experimentally obserw~d increase in phospholipase activity leads to the question: what ehang¢~ at the mokcular level augment
enzyme activity? Unfortunately, our present knowled[,,e about such activation phenomena is scanty a:,~- the description thereof has .targely been given irl terms of models by
analogy with other enzyme systems, the regulation of which is better understood in terms
of molecular events. Such models have to consider cenversion of inactive zymogens to
active enzymes, other covalent modification of the enzyme molecule, inter~ction with
regulatory ligands or proteins and changes in membrane structure res~alting in variations in
the catalytic capacity of membrane-associated enzymes. The merits~ of such models an~
their experimental support will be discussed in the following sections.
! VA. Do intracelhdar phospholipases occur in zymogen form ?
The actual occurrence of zymogens of phospholipase A2 was demonstrated in a most
straightforward munner by isolation of the precursor molecule from porcine pancreas by
de Haas et at. [88]. Conversion of the zymogen into the active enzyme occurred by
trypsin-catalyzed removal of a hept~.peptide from the N-terminus of the molecule. This
conversion could ~dso be brought ab,mt by prolonged autolysis of a pancreatic homoge.
nate at room temperature. This procedure resulted in a 10-fold increase in the phospholipase activity in the crude homogenate. The enzyme purified from autolyzed homogenates [74] was indistinguishable from the one obtained by trypsin activation of the
zymogen [88].
Moderate activation (2-4-fold) of phospholipase A activity has been described in vari.
ous rat tissues [89|, but the positional specificity c,f the enzymes) involved was not
studied. The activation proces.~was found in cytosol fractions prepared from spleen, lung,
thymus, liver and bone marrow, it was claimed that wh,m homog~mates of these tissues
were submitted to centrifugal fraetionation, 80% or mor~ of the 9hospholipase activity
with or without trypsin treatment was found in ~he snpematant fraction after centrifuga.
tion for 4 h at 120 000 ×g. For liver, these resu!ts with non-treated homogenates are certainly at variance with the data reviewed in Sectiov~ II, ~hieh indicated that the major
part of the neutral or slightly alkaline phospholipase A acti 1ity is membrane.bound.
In early studies, the search for a phospholipase activity n erythrocytes failed to detect
such an enzyme ill the red blood cells of various animals. On the other hand, lysophos.
phoglycerides were the main accepters during fatty acid incorpc,ration in ~.'rythrocytes.
This led to the belief that acylation of lysophosphog~ycerides played an important role in
the renewal of fatty acids in the phospholipids of fllese cells. Originally, the red cell was
thou~t to be dependent on the plasma as ~ source of lysophosphc,glycerides, an idea supported by experiments showing rapid exchange of lysophospholipids between plasma and
erythrocytes. The ability of red cells to produce lysophospholipiids them~lves was first
demonstrated by Paysant et al. [90] in rat erythrocytes, albeit with a pi',osphatidylgly..
cerol, i,e., ~ ~on.erythrocyte phost~holipid, as substrate. In contrast to rat erythrocytes,
those of humans showed barely detectable phosph,~lipase activity, which could only be,'
demonstrated in lysates of human cells. Interestingly, treatment of the lysates with tryp.
2~5
sin resulted in an at least 10-fold increase ha phospho~ipag A acti~ty ['~il. ~ ~eve~-zl
criteria, this increased activity was ,;hewn not to be du:e to a pebble phosph~.pa~e in the
trypsin preparation used, :rod 'thus suggested a trypsin.catalyzed zc~/v~tion of the erythrccyte phospholipase. The question as to whether this activatio~ is caused by a z-yr,,'lo~tactive ~'nzyme conversion or by removal of an inhibitory peptide in r e ' s p ~ to tryp~n
treatment was not further invest~ga!ed. At present, it is not even c[ear whefi~er ~he observed activation is to I~.-a~a:ribed, to the pr_ot,.,olyticactivi D" of trypsLn. Zwazl et a[. [82]
have been unable to detect t?hospholipase act:ivity in human red cell ghogs, either before
or after trypsin treatment. Increased phosp!hol~pa~e~, ~ti~ty tow~zds ph~phatidvtglycerol after trypsin treatment has also beeu observed in rzt plasma [92]. 6 sL~ar, 2-3fold, in,=rease was observed durir~g storage of diluted plasma at 0°C for sever~ days. After
this activation, no further increase was obtained by trypsin treatment. Upor, et~n~t fr~etionatio~ of whole rat plasma, the active enzyme appeared in fraction I, whereas the
albumin-rich fraction V showed only phospholipas¢ A activity after treatment with trypsin. The authors have con,:luded from these data that rat plasma c o n t a ~ ~ ~mctive
prect, rsor of phospholipase A which became transformed into an active enzyme with ~fifferent solubility properties undzer the influence of either a pl~ma factor o~ t~rps~ [92].
Although a considerable purifieati.o', of the at:tire enzyme was reported [93], no attemp~
was ma ',~ to isolate the precm~,o~ :t~olecule and to show th." activation in a sy~em containing pure components. Other mechanisms to account for the activation proc.,~ c-~
therefore still be envisaged, iProgress -, this area ~hasbeen hampered by a tater observ~km
[94] that the activation of phsma phospholipas~ A by the crude commercial prepara~km
of pancreatic trypsin was due to the presence i:a this preparation of an aefiwat/n[ f a c ~
which is different from trylgsin itself. This may explain why Zw~l et ",d. ~82|, p r e p ably using pure trypsin preparations, were not able to conf'u'm the actUator/onof h ~
erythrocyte phospholipase A..
In addition to crude trypsin, intact or lysed rat ptatelcts also activated the p / ~
lipase A in platelet-deficient plasma [94]. Addition of rat ptat~lets, likewise, caused a tremendous increa~e, varying l'r<m 6- to 35-foild, in the p h ~ p a s e
acti~qD" o f h ~
serum and plasma, rabbit l~tam~a and lysed human erythroeytes. The a c ~ m , ~ f~c~er
could be solubil/zed from rat plate|ets by a variety of te~b~niqu~s including soa~c~-~,
repeated freeze-thawing and treatment with sodium d~oxyc~c~.ate 0¢ 0.I I~! C~b. a~
pH 4. The platelet factor appeared to be sensitive to heat and comp~tdy lost ~ts ~cavab
ing property after 10 min ~t 60°C. The platelet lysate has since been ~ by Pol~vs~d
and colleagues [95,97] to activate a number of other phospholipases. Tocqueb/a~,Ct.q~d
etal. [95] have shown tha~ rat liver plasma raembranes have a low phosphol~pa~ acfi,,5~y
with phosphatidylethanola~alne as substrate which could be stimulated 5 - t ~ by ~ g
platelet lysate. Perhsion of th,~, liver with saline prior to fracticnmtion of the |~ss~¢
yielded a plasma membram,' fraction devoid of phospholipase activity, bu~ after ~k]iti,: a
of platelet lysate the plamm membrane pho,.~holipase activRy was restored to the
present in plasma raembranes of non.perfuu:d livers in the ~
of phtdet l y ~ e . b-'y
using liver plasma membranes of heparinh:ed rats from which most of ~ p ~
lipase A t has been detachad (Sact~on ll), it has been s u b t l y
s h o ~ that m~a of t ~
increase in total phospholipase activity can be ascribed to a c t i v a ~ of a [ ~
iipase Aa [391. The inactive phospholipas¢ A: remained bound ~o the memb~,~es whea
these were washed with ~ i~ NaCI, In contntst, activation of the ~ . p ~ s ¢
A~ by t~e
platelet factor prior to the washing procedure led to a ~
sole~t~
of tee
active etmyme. These data suggest that aclivation is a c c ~
bY a ~vee,4o~ of v,
206
more hydrophobically bound integral memb~'ane protein into a more electrostatically
bound peripheral membrane protein [39]. As discussed by the author~; two possibl;e
mechanisms mag account for these observati~',ns. Srhc~aatical[y, these are depicted in
Fig. 1. First, proteolyti¢ action by tile platelet fi,.ctor could convert a prophospholipa~;e
into an active enzyme. Secondly, complex forma~'ion between tile inactive phospholipase
and the platelet factor could yield an active phosl)holipase. Unfortunately, no defmiti'~e
disti,"~ction between these two possibilities car b~; made at present. It may be relevavJt,
however, in this respect, to m~mdon some recent experiments cm the release of phospholipase A and triglyeeride lipase from perfused rat liver or from isolated hepatocytes [96].
Although in most experiment'= totai phospholipase activity was not differentiated into At
~nd A2 activity, it was claimed that hepatocytes re!leased a mixture of both enzymes, with
phospholipase A2 accounting tbr about one-third of the total phospholipid-hydrolyzing
~ctivity. The total activity released into the mediam was 20-times greater than that present in the cells and this was not altered by puromycin concentrations which completely
• ut off protein biosynthesis. These data suggest the conversion of inactive to active
forms of both phospholipases prior to or during re,lease.
An inactive form of phospholip~se A that could be converted into an about 2-fold
more active form by addition ,of lysed platelets ~lso h:ls been described in fetal and adult
rat lung [97].
Phospholipase A2 has been postulated to play a regulatory role in the bio~nthesis of
prostaglandins [98] (but compare subsection VA) by releasin[l free, amchidonate as a :;ubstrate for the cyclo-oxygenase system. Treatment of intact cells with proteolytic
enzymes, in several cases, gave rise to increased release of arachidonate. Bo~h thrombin
and trypsin induced phosphotipase activity in human platelets [99], suggesting a proteoiiytic modification of a phospholipase. However, compared with intact platelets, disrupted
platelets were much less responsive to thrombin and trypsin.. Thus, it ,woitld appear that
the structural integrity of the platelet is indispensable for proteolytic enzymes to activate
the phospholipase. Remarkably, EGTA was abl~ to augment the effect of thrombin but
not that of trypsin, suggesting that each .eazyme influenced different points in the pathway leading to phospholipas¢ activation, the above-mentioned fact that EGTA enhances
the thrombin effect indicates that EGTA has ~o access to the intracellular Ca2÷ required
in the active site of the phospholipase A2. Thit; would seem to indicate that it least the
active site of the phospholipase is located in ~the inner monolayer of the platelet membrane. It is still conceivable that 9art ofa prophospholipase could be reached by th~ proteolytic enzymes from the outside to yield an active enzyme. If this were the sob: event
responsible for the activation, one would expect that acti',ration czuld have been simulated by treatment of platelet membranes with proteolytic enzymes. Since this was not
observed, an alternative ex#ana~ion was postulated in that EGTA treatment of intact
cells would chelate surface Ca2÷, thereby facilitating thrombin proteolysis. A.,;with trypsin, this proteolytic attack of thb membrane ~ould perhapt; have activated the phospho.
lipase in a non-specific manner by altering membrane structure. Similar explanations can
be put forward for the observed stimulation ,of arachidotmte rele~se, measured as prostaglandin production, in a tramfc;med mouse BALB/3T3 cell line by thTombin and trypsin.
Several other proteases (chymotrypsin, collagenase, elast:ase and thermolysiu) did not
stimulate prostaglaodin production [100].
In conclusion, except for pancreatic phospholipase A2, the occurrence of inactive
zymogens of intracellular phospholipeses has not been demonstrated convincingl,/at the
molecular level. This would require the isolation of a, at least partially, purified pro-
enzyme, which can be converted by proteolysis into an active enzyme urtd¢: c ~ d / t / o ~
which unequivocally show a modification of the covalent structure of ff~epcoenzyme, tt
is obvious that studies on the activation of membrane-bound phc~pho~pas~, alth-~zgh
interesting and of utmo:;t importance in themselves, are not likely to provide an answer to
this question. In the proteolytic activation of membrane-bound pho~pho~pases it ~d2Jbe
very difficult to unravel specific zymogen-active enzyme convers/ons from other ~ b l e
activation mechanisms such as those due to non-specific alterations in memh ar~e structure and proteolytic re,loyal of inhibitow proteins. Before discussing the evidence for the
presence of proteins or peptides which specifically regulate pho~hoIipase A z~ctivities, the
difficulty of inttrpreting prc.teolyfic activation of membranebound enzymes is further
illustrated in the following examples. Adenyhte cyclase levels in a rat embryo fibrob~t
cell line were found to be much higher in cells detached by tryp~Jzation as compared t~
cells obtained by mechanical scrap~ag [ 101 ]. The latter cells showed an 8-fold increase/~
cyclase activity upon trypsin treatment. A 2-fold activation of rat liver plas~la n-tembr~z~e
adenylate cyclase by several p:.~.'ep:roteolytic enzymes was reported and exl:lained by unmasking of active sites [102]. In contrast, proteoiyt~c activation of adenylate cycla~e in
several cultured fibroblasts seemed to be more in line with an effect on t ~ regulato~"
functions of the enzyme rather than on the catalytic unit [103]. In these cells, s e ~
other membrane-associated enzyme activities such as 5'-nucleotidase aJ~d b~th Mg2"- and
(Na÷ + K*)-ATPase were not appreciably modified by trypsin treatment. "l~is wag used as
an argument to indicate that proteolytic activation of the ¢yclase system is specific and
not the result of general alterations in membrane structure, thereby defying the ~ [ l
known fact that not :ill membrane-ass0ciated enzymes are equally sensitive to memb~z~e
perturbations. These ,~esults are only vsentioned to indicate that ~ A l ~ u e
acfiva~
by proteolytic enzymes has its analogies in other areas of membrane research.
Finally, it should be noted that activation of phospholipases by c o n v e l s ~ of zymogen into active erlzymes represents an irreversible modulation of ea..~matic activity. For
physiological control processes, reversible regulation mechanisms would se~m to be much
more appropriate.
IVB. Are phospholil~ses tegulc~teclby specific assoclotion with non-enzyng~tic ~ t ¢ f ~ s ?
As discussed in the previous section, the activation of phospholipase activities ob~e~ed
in a number of cases upon treatment with proteolytic enzyng~ has been ~
by at
least two general mechanisms, i.e., zymogen-active enzyn~ ,.-onvenk~s or p t o ~ e o ~ :
inactivation of a no:a-enzymatic inlubitor of the phospholipas~. Since no data on the ~.~
lecular events regtlting in activation are available it has beea~ impossible to
between these possibilities. This section deals with the evidence for the l~'ese~ce ofpcoteins or peptide~ specifically modifying phospholipase activities, lnhibl~m of eaz-yma~c
activities by protein h~hibit.ors is bes,t known for proteolytic enzymes, w~d'e the ~ g b
ing protein acts by binding stoichiometrically to the eazym~ [104]. Does th~ type of
regulation necm with phospholipases as well?
In their search for mut'.eats of i~cillus subtilis with fragile membraae~, Kent and Le~natz [105] isolated a mutant which, in coatrast to wild-type ceils, s h ~ - d up to 70%
phosO~olipid degradation durh~g pmtoplast formation. This appeared ~:o be ~ o to the
sequential action o:["/, membrane-~:iated p h ~
Az and a c~'toplasm~ lysophospholipase, in subsequent expe:riments to explain why no c o m ~
degradation toc~k phce in wild-type protoptasts, it was fotmd that will-type ~
con-
208
tained a potent inhibitor of the phospholipase A~.in mutant cells. This inhibitor, apparently
absent in the mutant, was present in both sol~Jble and membrane.associated form in wildtype cells. The soluble form was heat-stable, non-dialvzablie and sensitive to trypsin. Fur.
ther evidence for the protein nature of the inhibitor was 'l)rovided by Krag and Lennarz
[106]. The soluble form of the inhibitor, lo,:ated in the periplasmic space, was purified
over 2000.fold to homogeneity by classical protein.purification techniques, Its estimated
molecular weight amounted to about 30000 in gel filtral:ion and about 36000 in SDSurea gel electrophoresis experinlents. Inactivation of the membr;me.bound phospholipase A~ appeared to occur via an enzymatic inactivatif:m process rather than via stoichiometric binding of the iahihitor ~o the enzyme. Addiiion of purified inhibitor to mutant cells during protoplast fmmation not only prevent~l hydrolysis of membrane lipids
bu so restored the osmotic stability of the mutant pro toplasts, ind,icating a correlation
between reduced me:nbra~e l:ipid content and increased fragility. On the bases of these
results, it was postulated that the wild-type cell would contain a pho~pholipase AI as
well, but that the expre~on of its activity would be raasked by the inhibitor protein.
This idea was corroborate~l by the finding :hat not only mutant cells but also wild.type
cells secreted the lipolytic enzyme into th~,~growth medium. In addition, several properties of the partially purified extracellulal ee~,yme from wild-type cells were fo~and to be
identical with those of the partially purLqed lipolytic enzyme obtained from mutant
membranes [107]. Both enzymes were not specific fez phospholdpids, but hydrolyzed
~iglycerides and diglucosyldiglycerides at :;lightly lower rates. The partially purified membrane-bound and extraceilul.u (phospho)t,ipases were i;naetivated in an identical fashion
by the inhibitor protein. Th,~ strong similarity betw.¢¢~ the membrane-bound and extra.
cellular form of the (phospho)lipase suggested that the membrane-bound foma might be
an intermediate in the ~ecretion of the (pl'tospho)lipase. This hypothesis was supported by
the finding that under appr,~priate conditions, about 60% of the (phospho)lipase activity
wa~ lost from the membral)es of intact cells and an equis~ent m o u n t appeared in the
medium. If the membrane-bound (phospho)lipase is in fact a secretion intermediate, then
the role of the protein inhibiior in the ~ild.type cells could be to protect the membranes
of the cell against the action of the (pho~;pho)lipase driving or after secretion. Iit should be
rec~dled, however, that thence is no evidence that the inlUbiror protein exerts its irreversible inactivation of the (phospho)lipase I:hrough re:motion of a stoichiometric (phospho)lipase-inhibRor protein complex.
Other evidence for the involvement ~f proteins in the regulation of intra¢¢llular phospholipase A2 activi~, has recently emerged from stTadies on prostagland~, release. It is
generally accepted that prost~glandins :are not stored wthin cells but are biosynthe-.,ized
in response to various stimuli. The rale4imiting step in the biotranffonnation of arach.
idonic acid into endoperoxide and further metabolites is believed to be a J:elease of the
precursor acid from complex ceIlular Iipids by pho~gholipases [108-.110| or other acyl
hydrolases (see subsection VA). Prostaglandins have been implicated in infl~anmation and
non-steroidal aspirin-like an.~i.inflammatory drugs like aspirin and indomethacin inhibit
prostaglandin generation by inhtbitiort of the cyclo.oxygenase enzyme [! 11]. Also, antiinflammatory corticosteroids have been shown to ir,terfere with prostaglandin production
in, for example, perfu~..d guinea.plg lungs [112] and transformed mouse flbrobla.~ts
[110]. Since corticoster;~ had.no ¢~ppreciable effect on the cyclo,oxygelaase in vitro, it
was suggested that thesv u:ti-inflan~,natory agent~ acted by reducing the availability of
intracellular subst~te. ]nd0od. additk~n of hyd~c~::ortisone to flbroblasts did not inhibit
prostaglandin formation, from exogenously suppli,!d ~-rachidonic acid but instead reduced
209
the release of arachidonic acid flora the cellular pheapho)~p~d~ [l !OJ. S~JIar~y, p c t f ~ : a
of phosphatidyl,:holine, labeled with oleate ~n lhe ~ - 2 - p ~ f i o n , throngh ~a)~t¢-~ ~
allowed the det¢clion of" a phospholipase A2 activity which was mh~b~.-~lby ar~fi-~af~mmatory steroids [! 13]. [t was the~ discow;red that Lqh~bifion of p r o s t ~ a n ~ f~rm,~t~o~
in renal papillae by corticosteroMs was completely abolished in the p r ~
of ~ t o ~
of RNA and protein biosynthesis, suggesting that $ y n t h ~ of a proteEa factor was
involved in the hd~ibition of prostaglandin production [! 14]. However, no ~rc~-t measurement of phospholipase A2 activities was made in these studies. Flower and D~ck~e~
[115] were the first to provide evidence that protein synthesis was re~g~re~ for ~ , ~ t ~ o ~
of the phosphc,lipase A2 in perfused lungs by c~)rtico~tero~..~tf~:,~ of ~
~th
dexamethasone markedly reduce@ phospholipase A2 activity as e s t . , t e d by the re~a~e
of (hremboxan(; A2 or by hydrolysis o¢ radioactive phosphatidyleho~Jne. D~rL~ ~ b ~ quent perfus~on without dexarnethasor.,e, contto| levels of pho~ph~pas¢ a c t ~ t y were
reached witM~t 1 h, indict.ring a reversible inactivation. In the presence of ~ . ~ b ~ t ~ of
com,coj___~~_ !PTOR
S~ERO,~S---F-----T----~
(cs,
/ ~ / [R.C~,o.O.:CS]
--~
RCSI~RF .
-~- '- ---FtCS-RF-LiKtZ material
!
\
Fig. 2. Tentative scheme fog the regulation of phosphoUpa.~A2 activity by rc~ .~ymi¢ inh~c~ aa~
activator pe.pl[idcs. The f,~l;owing ~luence of events is sch(m~tica~" ~,.'I~'t~. C o ~ s
induce the synthesis in lung of a pzotein which somehowis in~)tvcd in the macfivat~a of a
lipa~ [ 115]. "/'hedepicted compl(,'xformation of active enzymeand ~ t ~ ) ~ " ~
to y ~ a~ ~ . ~
tire complex i~ only one of sev¢£~lpossible, as yet unresolved,m e c h .
~
~z~ ~u-~c~
mbttance-zele~ factoz (RCS~F) has t~en d e s c n ~ as a ~
of ~
zw~. ~
of
stim~datingpk,ospholipaseA2 activity [1~$] when petfused thtoegh t~$~. At ~ t ~
~e ~ t e ~
action of RC$-RF, i.e., dissocia~m of the L~lihitozprotein ~ tLw~ t i ~ • ~
A2 ~ h ~
tot ~oluplex t.3 yield active p h o ~
A2, shoekl only be ~-"~';~ ~ ~ ~ s(t%,~ h ~ "~c~
modet of action, Stimulation of p~hospho~J~seactivity in tz~fextaed a ~ s e ~
,,%Vea~~a
protein synthesi~ [ 129,130], pedmps zesult~ in the p~t~.~,= ~-.¢~.CS-l~lF-~kem ~ ~
f~.~
indde the adls,
210
RNA and/or protein bio.,,ynthesis, de~:amethasone perfusion failed to inh!ibit phospho.
lipase activity. The perfit~ate of dexamethasone-tre~.ted lungs, but not that of untreated
lungs~ was shown to contain a factor which inhibited the phospholipase A2 of a second
lung. In view of |he above-mentioned effect of pro rein synthesis ;nhibiton;, this factor is
likely to be a peptide or proterm [115]. This experimental approach will eventually allow
isolation of the factor so that its mode of action can be studied. At present, the results
can most easily be explained by assuming ,hat the factor reversibly modifies the phospho.
lipase A2, possibly by combining with the enzytr~e to yield an inactive enzyme-factor
complex (Fig. 2). However: an alternative explanation in which the phospholipase is kept
in an active form throul,h association with a second factor, X, is also compatible with the
experin~entai results, provided it is the factor X which is subject to modification by the
protein synthesized in n~sponse to the anti-inflamn~atory steroids.
The previous paragr:~phs have dealt with the possibility of phospholipase A inhibition
through interaction with proteins. In the followirig paragraphs the evidence for stimulation of phospholipases by non-enzymic proteinz will be reviewed. The possibility of this
occuning has clearly been demonstrated for ~'~eral highly purified phospholipases or
lipolytic enzymes in vitro, although even in the~; cases the exact molecular mechanisms
underlying the activation process remain as yet unresolved. Lipoprotein lipases not only
hydrolyze primary acyl ester bonds in triglycerides, but also those in phospholipids, thus
expressing phospholipase A~ activity. The ~nzym,,~can be released from the vascular endothelial surface by intravenous injec'tion of beparia or during perfusion w~th heparin. Characteristic of the purified enzyme is that both its lipase and phospho'lipase At activity
[116] are greatly stimulated by ~@olipopro~eh:~C-I[, a peptide consisting of 78 amino
acids. Similar activations have been desc ibed ~br th~ action of bovine milk lipoprotein
lipase towards phosphatidylcholine vesic.es [117]. To explain this activation it may be of
importance to note that high affraity of apolipoprotein C-II for both phospholipid vesicles [ 117] and lipoprotein lipase [ 1 ! 8] has been demonstrated. The a'polipoprotein may
thus function to link the enzyme to its substr:;~te. While the exact mechanism is unclear,
the fmding that maximal activation was obtained when a stoichiometr~.ic 1 : 1 complex of
lipoprotein lipase and apolipopr¢tein C-II hacl been formed strongly supports the hypothesis that this complex fonnation is the basis for the activation phenomenon [118].
Lecithin-cholesterol acyltransferase catalyzes the transfer of long-chain acyl groups
primarily from the sn-2-position of phosphatidylcholine to cholesterol. The activity of
the purified enzyme depends on the presence of an apolipoprotein from high.density lipoproteins [119]. Recently, the two partial reactions of lecithin-choleslierol acyltransferase,
i.e., hydrolysis of phosphatidylcholine and esterification of choleslterol, have been unco~tpled by demonstrating that the enzyme has major phospholipase activity when cholesterol is not present [120]. Not only the transferase activity, but also the phospholipase
activity appeared to be completely d.~pendent on the presence of apolipoprotein A-l. In
this case, it was shown that the apolJ~oprotein A-I was not required for the binding of the
enzyme to the lil~omes, sugj~esti'.lga dire~;;t activation of the pho~pholipase (and transferase) function of the enzy~; 0 aough complex formation with thc~apo A-I.
Activation by n o n - c h i c pj oteins has been demomtrated also for hydrolysis of complex glyeolipil ~.~by l y ~
hy4rolases. In some cases, e.g,, cerebroside-sulfatase [121,
122], the activating protein Ibnm a 1 : 1 Complex with the lip!d mbstrate and this complex is then recognized as a subst~te by the enzyme. In other cases, e.g., glucocerebrosidase [123], the activatc r forms ,a complex with the enzyme, when the latter is present in soluble JOrm. The association bet-,,,ean activator and enz~nne reaches a l : 1 stoi-
211
chiometry when the activator concentration is limiting, lncorpcration ~f the activator
into a membrane preparation containi, o, the ca'ta!ytic protein was t i m e - d ~ d e n t and
paralleled the appearance of enzyme activity in the membrane [124]. This indicates tb~t
activity of the membrane-be and enzyme is also affected by the ~ctivator. The lysosorr~.
actiwtors are relatively heat-stable, low molecular weight (about 25 000) ac/d/c #ycoproteins.
Some evidence for the existence of a peptide which stimulate., pho~holipam A2 activity has beea obtained by Nijkamp et al. [125]. It has long been known that ~ a D h y l a ~
in isolated lungs from sensitized guinea-pigs leads to relea.~einto the perfu~te of both an
unstable rabbit aorta-contractin8 substance (RC$) and a much more stable R C S - r e l ~ g
factor (RCS-RF). The latter releasing factt~l c,as able to promote a further release of RCS
during peffusion of unsem,itLzed lungs [126]. The main active component of RCS was
later iden~:ified as thromboxane Az, although some cyclo-endoperoxides were present as
well [127,128]. RCS-releasing factor has been partially purifie~i by Nijkamp et al. [125].
The almost 6000-fold pm~fied material exhibited properties ,:ompatible with RCS-RF
being a neptide of less than 10 arn~ao acids. Hydrolysis of cir.'ate-labeled phosphatidyL
choline d,urmg perfusion of lungs was stimulated 2-fold by simultaneous injection of
RCS-RF and this hydrolysis was blocked when steroids were present. These results suggested that RCS-RF activated a phospholipase A2. Combination of these resutts with the
steroid-:.,duced synthesis .of a phospholipase A2 inhibitor leads t ) a tentative scheme for
phospho?ipase A~ regulation as indicated schematically in Fig. 2. The phosphoL/pase can
associate with an inhibitor protein to yield an inactive enzyme-inhibitor c o m ~ x . Antiinflammatory steroids enhance the synthesis of inhibitor molecules with ~he net effect of
decreased phospholipase activity. Other peptides, such as RCS-RF, release the inhtbito~
from the.'inactive complex, perhaps through a direct interaction ~-ith the inhibitor, to #re
augmen :ed phospholipase activity. Similar peptides may be synthesized in other o~b~n~
and cells in addition to lung. Some evidence for this can ,~-erlmpshe extracted from observations made by Levine et al. [129] and Pong et al. [1"-~0] or, tramformed mouse fibreblasts. Prostaglandin production b-~ these cells is stimulated by serum, thrombm and
bradykinha. This stimulation is blocked, however, by inhibitors of protein and RNA ~ynthesis, albeit after a certain lag time. The requirement of protein synthesis for the expression of the stimuli suggests the synthesi.~ of a peptide or protein factor which e~m~.-es
prostaglandin productio~ ~ince the stimuli had no effect on the cycM.oxygetmse it.serf, it
could be that the peptia~ :actor, perhaps RCS-RF-like m~terial, stimulated prostaglamd/n
formation by providing, more arachidonic acid mbstrate tluough stimulation of the ~ o v
pholifase A2. AIt~,matively, the observation could be explained by de nero synthes/s o f
new phospholipa-~ molecules. It should be remembered, therefore, that the scheme ~
Fig. 2 is highly tentative at present. Neither the inhibitor prc~tein nor the activato~ prote~r~
has been isolated at present in sufficient purity to allow studies ca the/r mode of action.
Mu~, work will have to be done to establish f'Lrmlythe ~ x t y h)vothetical inter~ct2~oe,~
put forward in this scheme.
IVC Regulation of phospholipase Az by ava~tabih'tyo.f C~'~
Membrane-bound phospholipases A~ show an absolute requirement for Caz*- Kunze et
al, !i131] have found that prostaglandm biosynthesis in ~ t e s
of bovine sem~ml
vesioles is markedly stimulated by Ca~'. Sinea Ca~ s]~tJy inhibit~ p i x x s t ~ d m syr~
the:gs from free arachidonate but stimulated # ~ t s e
Az, these t ~
s~r~ed
212
file assumption that phospbolipase A2 phe,ys an important role :in stimulating prostaglandin biosynthesis by providing free precu'rsor a~:ids. In studying the patterns of fatty
acid release from endogenous substrates by human platelet merabranes, Derksen and
Cohen [132] discovered a lipolytic enzyme releasing arachidonate. An equivalent amount
of arachidonate 'was lost from the phospholipid fraction, thus suggesting that the lipolyt*ic
enzyme was a phospholipase A:. Additior~ of Ca2~ augmented arachidonate release 20fold. If these and o~.ber calcium-dependent: phospholipases A2 ind,.'ed function as regulators in intracellular archidonate release fox ~ndoperoxide synthesis one migbt expect that
events which inc:rease the avai~labilityof Ca ~+to the phospholipase A2 in intact cells might
stimulate the fi)rmation of endoperoxide.*; and related products. ,q;everal research groups
have provided q.,vidence to support this hypothesis, by using ionophores to elevate cytoplasmic Ca=÷ at the expense of intracellul;~Lrstorag~Jcompartments, perhaps mitochondria
[133] or the ~.ense tubular system [134].. Pickett et al. [135] demonserated that when
washed human platelets were incubated i~a the presence of 5 mM EGTA, addition of the
calcium ionophor¢, A 2318~r, led to the sadden rt~lease of arachidonat¢ in much the same
way as that produced with thrombin. In the presence of 10 gM ionophore, 46% of the
total phosphoilipid arachidonate conten't was n~leased within 1 rain. The stimulatory
effect of the ionophore wa~ maximal in the pres~:nce of 5 mM EGTA and declined progressively as the concentration of free C~.~=÷ in the medium was raised. These data imply
that in intact ,:ells he EGTA did not penetrate ,~o the locus of phospholipase A2, which
therefore is likely ~o be not at the outside surface. Independently, Rittenhouse-Simmons
and Deykin [ 136] confirmed and extend~:d the.,~ observations by showing that the abrupt
release in prelabelled cells was from phosphatidylcholine and phosphatidylinositol. Interestingly, in cellls deprived of metabolic A'I~ by pr,.~incubation with deoxyglucose and antimycin A, the release of a~achidonate a~; promoted by thrombin, but not that by ionophore, was inhibited coml?'~ed to control cells. Others have |curtal that the thrombininduced burst in oxygen ut,lization by pgatelets, an ind.;cator of free arachidonate oxygenation, is markedly diminished by antimycin A. This suggested also tlmt thtombin.induced
release of arachidonate was largely dependent on electron tnmsport and, presumably,
ATP synthesis [137]. This impalrmemr .of the thrombin-induc~:d arachidonate release in
ATP-depleted cells was found even in 'the presence of extracellular Ca=÷, indicating that
thromhin alone does not render the m,.'mbrane sufficiently permeable to Ca2÷ to account
for the phos.pholipase activation [136]o The combined data were interpreted to indicate
that a Ca2÷ :flux, resulting in a rise in cytoplasmic free Ca2÷ concentration, activated phospholipase A,. This flux c,f intracellnla~r Ca2÷ could be brought about either by the iono.
.~hore or by thrombin, h was proposed only in the latter cas~,,that a c~ntractile or alternative ATP.dependent p~ocess is required to liberate internally stored Ca2+. In this case,
no covalent modification, such*as, e.g., zyraogen-active enz~qne conversion (subsection
IVA), would be necex~a'y for thrombin-induced phospholipase activation [136]. It
should be mentioned, however, that ,this explanation is comg,letely based cn the assumption that i~mophore- and thrombin-induced activation sha~'ean identical last step, namely
saturation of pre.existing enzyme wi~h Ca~+. The availabifity of free cytoplasmic Ca~÷ for
the phos~holipase supposedly is cm~trolled by cyclic AMP (compare next section and
|oeophore stimulation of platek, t thromboxane B2 fonna~ion and cyclo-oxygenase
~,xy~e~s consumption was not restricted to the divalent cat~onophore A 23187. The
moao- and divalent ¢ationophore, X 537A, produced comparable increases, while the
mox~ovaleat cationophores, nigericirt, monensi~ A and valinomycin, had no effect [ 137|.
213An interesting new dimension to the concept that plateh.'t phospholipase Az/s r e . t e d
by the availability of intracellular Caz* to the enzyme lo,:us was recently ~Idcd by Woag
and Cheung [138]. These investigators pre~nted some p~Jiminary evidence suggestiag
that the stinmlatio~ of human platelet phospholipase A~ by Ca2÷ could be raed~ted
through calmodulin, a ubiquitous Ca2+-binding protein.
Data suggesting that A 23187-induced stimulati6n o:["thxomboxane or prosta#~adm
b:,osynthesis may t~.= a more generalized phenomenon were obtained by Znapp et al.
[137]. The ionophore produced this effect also in rat renal medulla, rat stomach and
trachea, human lymphoma cells and guinea-pig polymorphonnclear leuk~/tes Contrary
to what has been observed with platelets, A 23187 only stimulated prom~andm synt:aesis in renal med~lla when Ca2* was added to the external medium [137]. The sfimulatory effects of Ca:°, Mn2. and Sr~÷ on prostag;landin fo:anation in rabbit I d ~ " medulla
slice,~ in the absence of ionophores were found to correlate with their stL,z:,zlato~" effects
on tire release of arachidonate and linoleate from tissue lipids [ 139].
IVD. Influence of c.vclic adenosine mop.ophosphate
As mentioned in the previous section, Pickett et al. [! 35 ] noticed that dibutyry. ] cyclic
AMP inhibited thrombin-, but not ionophore-induced, activation of #ateiet phospho]ipase A2. Using different techniques, other investigators arrived at similar condus/op.~
:oncerning the effect of dibutyryl cyclic AMP on thrombin-indnced arachidon,~te reIegse
from platelet phospholipids. Thus, Minkes et al. [140] found that incubation of phtelets
with dibutyryl cyclic AMP before thromt in addition blocked the subsequent foEmation
of oxygenated derivatives of arachidona~e. In contra.'~, when arachid~.~ate was added
directly to platelets preincubated with dibutyryl cyclic AMP no inhibition of thromboxane A2 formatiov, was obserced.
This demonstrated that dibutyryl cyclic AMP had no direct inhibitory effect on cyc[ooxygenase or thromboxan ~.sy~thetase, a conclusiori in conflict wit h findings of Mahnsten
et al. [141] but independently confirmed by LapetiJla et al. [142]. T~¢ ~tter authors
used horse platelets and a mot(.' direct measurement of phospholipa~c A2 activity, i.e., the
loss of radioacth~e arachidonate from phospholipids of prelaheled ~telets. Not o ~
dibutyryl cyclic AMP but also agents which elevated platelet cyclic AMP levels, such as
cyclic AMP phosphodiesterase inhibitors [142] or ad(mylase cy~Ls¢ stimulate, such ~
prostaglandin Et [140,158] or prostacyclin [142], w,,~re found tc prevent the thromb~induced deaeylation of arackldonate4abeled phospholipids. Contra~ to the initial observation by Pickett et al. [135], it has later been fourLd that also ,~123187-induced [~hospholipase Az activation is inhibited by prior incubation with dibutyq,'l cyclic AMP [143,
144] or cyclic tLMP phosphodiesterase inhibitors [[43]. Many of these effects depend
apparently on the relative concentrations of stimuli used, as Fehtstein et al. [158| have
described conditions under which arachidouate release in the pre~nc¢ ofionophore ~ds
normal despite high cellular cyclic AMP levels. It rca~as to be ducidatcd by what mechahism cyclic AMP inhibits the release of arachidonic acid. Exclud:~.gme~,eg ¢ ~ effects
of cyclic AMP on cellular metabolism and assuming that the :aachidonatc~-clcasiag phespholipase is directly regulated by cyclic AMP, several possibifiti¢~ still eaa be ¢ ~ ,
Among these are direct binding of dibutyryl cyclic AMP oI cyclic A.MF to the ~ ,
cyclic AMP-detendent phosphorylation of the phc~holipase ot ~ m t e ~ c ~ protein
resulting in inhibition of the phcx~pholipase. Altemativ~y, cy'~c AMP could ~ t e
Ca~'
fluxes and stimulate the storage of Ca~, thereby ~ , ~ g
Caz" a ~
to
214
phospholipase. E,4dence to support the notion that phospholipase activity in platelets is
solely dependent upon the cytoplasmic levels of Ca a+ was provided by RittenhouseSimmons and Deykin [ 1441. They showed that thrombin-induced release of arachidonate
from human plalelet phosphaIlidylcholine was not only impaired by dibutyryl cyclic AMP
but also by t2~¢.~Jntre#~.:iular calcium antagonist, 8.(N,N.diethylamino)octyl-3,4,5-trimethoxybenzoate. Ad<~don of external Ca 2+ did noTt abolish these inhibitions. However,
when the platek)t phospholip~lse was activated by the divalent cationophore A 23187, the
inhibition produced by eith(~r dibutyryl cyclic AMP or the calcium antagonist could be
overcome by addition of..~,xt(~mal Ca 2+. In platelet lysates the phospholipase A2 was stimulated by addition of Ca ~'~, but neither A 23187 nor dibutyryl cyclic AMP exerted under
these conditions the res)~ective stimulatory or inhibitory effect observed in whole cells.
This shows that dibutyryl cyclic AMP does not inl~.bit the phospholipase by direct binding to enzyme or substrate, a conclusion confirmed by others [203]. It was suggested
[ 144,158] that in intact cel~s cyclic AMP may promote a compartmentalization of intracellular Ca =÷, thereby reducing cytoplasmic free Ca 2+ and inhibiting phospholipase A~
activity. It is worth noting tn this respect that K/is~r.Glanzmann et aL [145] h~.ve shown
that platelet vesicles, possibly derived from the derise tubular syster~, could concentrate
Ca 2* when incubated in the presence of cyclic AMP, ATP and prot,,~in kinase. Furthermore, the accumulated Ca 2<' could be released frc,ca the vesicles by ~. 23187. These findings allow fo~ the model elf plate[et phospholipase A2 regulation by the availability of
c*v
jJJj//
IONOPHOI~E
A 23187
Fig. 3. Model for the regulation of platelet phospholipase A2 by availability of Ca2+. The conversion of
an inactive phosphofipase A2 into an active phosphotipase A2 : Ca2+ complex is depicted. A rise in
cytoplasmic Ca2÷ level can be induced by ionophore A 23187 either in the presence or absence of
external Ca2+. In the latt~ case, rite Caa÷ is thought to be released from intracellular Ca2+ stores.
Likewise, addition of thronLbin is thought to lead to xelease of Ca2÷ from this or another store in an
ATP-depend,mt process, in the depicted model for the regulation of phosphoUpase A= activity, the
inhibition ol:~ervedby cyclic AMP (cAMP) is ascribed to its lowering of cytoplazmle Ca2÷ levels.
215
Ca2÷ as d,.,picted schematically in Fig. 3. "l'hlombin or ionophore induce the release of
Ca 2÷ from an intracellular store which acti~rates the p h o ~ o l i p ~ , e Az. A~ents vcb.ich
enhance the intrac,ellular cyclic AMP level or Ca~ ~mtagonists counteract the ~.-~ease L,~
cytoplasmic Ca 2÷ by sequestering the Ca2÷,, lhus resulting in an h~bition of t~omb~aor ionop.hore.activated phospholipase Ae. Tht: inhibition of the ionophore-, but ~o~ that
of thrombin-induced phospholipase activation, by cyclic AMP can be overcome by external Ca2+. Only in the former case does the e:~tra;ellular Ca2+ gala access to the phospholipase, thus overruling the decrease in cytoplasmic free Caz÷ as produced by eyc~c AMP.
If indeed Ca2* sequestering is an ATP-requiring process [145], the observat~n that ATP
deprivation inhibits thrombin4nduced phospholipase A2 a c t i v a t ~ [136,137] is diff'~u]¢
to reconcile with this model, unless Ca2. release is also energy-requiring and more ser~.
tire to lowered ATP levels than sequestering of Ca2÷. On the other hand, the free cytoplasmic Ca2÷ concentration as modulated by cyclic AMP may not be the only factor regulating phospholipa~e A2 activity. In addition, there is no co~pelling evidence to s h ~
that phospholipase Az activation by ionophore follows the same mechanism as ~.hromb~aor collagenqnduced activation. ~n fact, Lindgren et al. [146] have recently repoae~
almost complete inhibition of collagen-induced arachidonate release from p|ate|et-rich
plasma under conditions where intracellulal cyclic AMP levels were not yet ele-cated. Dffferentiatl inhibition of thrombin- and colla~en- as compared to ionophore4nduced phospholip~,se activation was reported by Fcinstein et al. [i58]. Using malond~dehyde formation as a measure for arachidonate relea.'e, it was shown that prostaglandm E~, cau~ag
increased cyclic AMP levels, almost comp!t~tely inhibited thrombin- and collagea&aduced
malondialdehyde formation, but had no el!['ect on ionophore A 23187-stiraulated
dialdehyde production. Although this would still be explainable by compensation fo~ zhe
cyclic AMP decrease in cytoplasmic Caz+ with ionophore, other experiments deraoastrated that phospholipase activation by thrombin or collagen, but not that by ionop~ore,
was blocked by the protease inhibitor, phenylmetham,,sulfoayl fluoride. In the ~ase of
thrombin, kinetic evidence was provided to show that this effect was not due to ~aact~vation of thrombin itself. On the basis of these results, it was postulated that two d~tTerent
mechanisms underlie the phospholipase activati,m in platelets depending on the stimulus
[158]. lonophore-induced stimulation woald be due to Caz+ activation of ar, existing
active form of the phospholipase. If stimulation by collagen or thrombin f a ~ to n~ob~e
sufficient Ca ~÷ to stimulate maximally tills acti'¢e form, an additional m e c ~
is p~oposed to operate. The latter stimuli would acti,/ate a serine-protease, which m turn catalyze:; the conversion of an inactive form of the ?~,ospholipase to an active form (compare
subsection IVA).
Inhibition of prostaglandin biosynthesis by cyclic AMP appears to be a unique feature
of platelets. In many other systems, inchltding cultured ~
eeU$, Graaf'~n follicles, thyroid calls, adrenal cortex and adipocytes, a sthnulation has been observed (see
Ref, 147 for review and references therein). Therefore, the inhibition of
lipase A2 by cyclic AMP, either direct c,r indirect, as observed in plate~ts may not ~., 9f
gen,eral occurrence, in thyroid, dibutyu~l cyclic AMP, although stimulating p r o s t a g l a ~
release, was shown to have no effect on a phospholipase Aa releasing a r ~ t e
f~n
phosphatidylinositol [148]. In contrast, indications of stimulation of phos~oli0cses by
cyclic AMP were obtained in lung [149] a~adadipoeyte [150] h~ao~aates. In the ~atter
case, the use of doubly labelled phosphatidylethandamine, ~o~Lag for a d c t ~ t ~ a
of the positional ~,'pecificityof the phcspholipases, ~ t the complete depe~'u~mey o ftke
hydrolysis process on a~ded Ca~ , gave strong indication of th~ stimu~t~a o f a phosph~
216
iipase Ai by cyclic A~IP [151 ]. A similar com:lusion was reached by Lindgren et al. (Ref.
153 and unpublished data) for 3T3 fibroblasts. Prostaglandin release from these cells was
stimulated markedly t~ the presence of either dibutyryl cyclic AMP or N~-butyryl.cyclic
AMP, provided that theoifllylline was simultaneously added. Elevated 'levels ofintracellular cyclic AMP, indu~l by adenosine and a plAosphodiesterase inhibitor, preceded prostaglandin production, F,iom prelabelled cells, dibutyryl cyclic AMP plus theophylline stimulated radioactivity rel,~ase only in those cells labelled with homo-3,-linolenic acid or arachidonic acid. No such. increase~l release ove, control cells was measured when the pre.
labellin~ was carried out with palmitate, stesratc, oleate, linoleate or linolenate. These
results would be eorn~iible with the stimulation by cyclic AMP of a rather specific phospholipase Al, which preferentially releases homo-7-1inoleate and arachidonate either
through acyl.chain selectivity or substrate avl~ability.
IVI~: Effects of hormone~ on p~ospholipase ~tivities
Although various hormones are known to stimulate prostaglandin synthesis or release
in their respective target tissu(:s (e.g., Table IV), only a few studies have appeared so far
which have conceatt:ated on the specific irtfluence of homlones on the tissue phospholipe,,es (Table V). H~.ye--~:al. [ 154], when litudying the increased biosynthesis of prostaglandins in porcine thy:o.~d slices diseovered that thyreostimulin stimulated an endogenous phospholipase A2. This c,nzyme hydrolyzed unsaturated fatty acids from exogenous
phosphatidylcholine and ph(~phatidyliito~itol. Whereas the fatty acids released front
phnsphatidylcholine were m.ainly mono.unsaturated, phosphatidylinositol hydrolysis
yielded mainly arachidonate. Since in thyroid ~lices both thyreostimulin and dibutyr./l
cyclic AMP stimuhJ.ted prostaglaadin biosynthesis, the possibility existed that thyreo.
stimulin stimulatiorl of the phospholipase A2 was mediated by cyclic AMP. This possibility was made unlikely by the subsequent fatdings [148] that cyclic AMP had no effec~ on
the pho~holipese A2 activity in thyroid Immogenates under conditions where thyreo-
TABL~EIV
HORMONALSTIMULATIONOF PROSTAGLANDINPRODUCTION
ACIH, adrenocortict,tropin hormone; TSH, thyreostimulin; FSH, foUicle~timulatinghormone; LH,
luteinizing hormone.
Hormone
Target tissue
Reference (examples)
Angic,tensin41
kidney
cultdre¢~nl~malian ce~is
heart
lung
kidney
165.166
161,167
168
169
162,171
cultured mammalian ci~lls
161,167,170
lpleen
adrenal
thyroid
ovt~y
ovary
172
173
154.174
175
175
Bradykinin
Epiw~phrine
AC-'I'tl
TSH
FSH
LH
ad|po~e
adrenal
adrenal
t'nytuld
mammary gland
heart
kidney
RMIC
lung
skin
kidney
RM|C
Target
homojenate or
cytosol
cytosol
cytosol
homo~e~tate
membranes
perfasion
peffusion
cells
perfu~don
membranes
perful~n
cells
Assay
* This ~pa~e may be identical to the cholcstcro| cstera~ [! 82-184 J.
Epin~
ACTH
T~I
ACTH
(B)
Angioten~4!
(A)
TSH
ProL~/n
Bradykinin
Hormoae
lipase *
chol. est. *
lipase *
PA2
PA2
PA2 or lipase
PA2 or tipase
PA2 or lipase
PA2
PA2
PA2 or lipase
PA2 c,r lipase
Enzyme
1.8-12
1.4
1.4
4
2
2.8
3.8
36
2.3
3.0
3.1
72
Stimulation
(-fold)
cAMP
cAMP
cAMP
?
?
?
cAMP-independent
?
?
?
Implied mechanism
H ~ M O N A L STIMULATION OF LIPOLYTIC ENZYMES
Abbreviations as i~ "iabl~ IV; PA2, phospholipase A2 ; choLest., cholesterol esterase; RMIC, renomedullary interstitial cells; cAMP, cyclic AMP.
TABLE V
180--184
177-179
177
148,154
157
13")
160
!61
i ~~,,i 76
185
160
161
Reference
211]
stimulin stimulated this activity 4-fold.. Dibutyryl cyclic AMP stimulated, however, arachidonate release from triglycerides. It i.,; therefore believed that thyreostimulin activates a
phospholipase A2 in thyroid by an as yet unknown, but cyclic AMP.J~tdependent, mechanism. Thyreosl:imulin also raises the cyclic AMP level [ 154,155 ], which activates a iLipase.
Both enzymes potentially can release arachidonate and thus contribute to the thyreo.
stimulin-stimuJated prostaglandin bio~ynthesis. Indeed, in the presence of albumin and
indomethacin, thyreostimulin-stimulat:ed thyroid slices were found to contain free arachidonate levels ':wice those of control slices [ 156].
Prolactin, when added to membranes prepared from mammary gland, effected a 2-fold
increase in thl~ release of arachidonate from exogenous, arachidonate-labeUed, phosphatidylcholine [157]. The concentratiort of prolactin necessary to produce this eff~;ct was
rather high, i.e., at least two orders of magnitude greater than those required to stimulate
'biochemical events in mammary glan,:l explants. On the other hand, specificity of the prolactin effect 'teas suggested by the observation that similar concentrations of growth hormone did not stimulate arachidonate release.
When hearts, prelabelled with arachidonate, were pert'used, injection of bradykinir~
resulted in a 5-fold stimulated relea',~e of radioactive prostaglandin, without a significant
increase in tke amount of free arachidonate in the effluent. However, when fatty acid-fret;
bovine serum albumin was added to the perfu~te, the fre:e arachidonate was apparently
trapped before reincorporation into tissue lipids masked completely the release., of free
acid. Under these co,lditions, bra~iykinin administration caused an elevation in arachidonate in the cardiac effluent. The increase in fatty acid release upon hormone stimul:l.
tion appeared to be a specific phenomenon fi>r arachidonate-labelled hearts anti was not
observed with hearts prelabelled with linoleate, oleate or palmitate [i59]. This ~lective
bradykinin ,fffect on fatty acid release suggests that an enzyme is activated that either distinguishes different fatty acids or is selectively compartmentalized with arac'aidona~e.
containing ~issue lipids. Ischemia, on the other hand, exhibited a non-specific st'~zaulation
resulting in release of oleic as well as arachidonic acid. Since most of the labelle,(~ unsatu.
rated fatty acids in the prelabelled heart were present at the 2-position of phospholipids,
the activated enzyme(s) most likely belongs to the phospholipase A2 type, but this was
not provec[ by identification of lysophospholipids as reaction products, or by a correlation of released radioactivity with loss of label from th~ phospholipid pool. At the begin.
ning of the experiments, the neutral lipid pool contaMed a substantial amount, about
15-20%, of the total radioa:tivity, so that activation of lipases cannot be exclu(~ed completely. It was notice,] that during ischemia the ratio of released arachidonate to prostaglandins (lO : 1) was much high~r than that found after bradykinin stimulation (3 : I).
This was !interpreted to indicate that the hormone-sti~aulated deacylation of tissue lipids
is more tightly coupled to the cyclo-oxygenase than that occurring during ischemia [ 159,
162]. Similar conclutions were r~;ached by ~hwartzmann and Raz [ 160] for hradykinin.
or an#ot,~:nsin-ll-~timulated release of arachidonate and prostaglandin E2 from rabbit kidney. The:;e authors avoided the u:~eof prelabelling procedures to exclu~.e the possibilJ.tyof
selective :release of radioactive acids from preferentially labelled pools, Appro~:. :;% of the
aracludo,:late released under ba.~ conditions was converted to prostaglandin E~. Hormone stimulation not only resulted in a 2-3-fold selective increase in arachidonate
release not seen for other fatty acids, but also led to a more efficient coneersion into
prostaghndin E2, which now amounted to about 3~Yo.The activated enzyme, apparently
tightly coupled to the pros~agl;mdin-generating system, was not further ideatified. This
holds aJso for tl~e acylhydrolase responsible for the large stimulation of arachidonate
219
release fr,sm prelabelled rabbi: renomeduliary intecstitial ceils in t i ~ e c~fiture as cau~d
by bradykini~q and ~agiontews[a-ll [ 161].
in view of the existing ua<ertainty as to which enzyme is ac~u~ly activated, it ~ ~bvious that no rigor.us data a~out the exact mechanism underlying these activatio~ ¢~n
be given. It is only for thyrec stimulin-activated thyroid phosphoI/pase Az (Tabge V) that
some indications have been obtained which point to a cycE~cAMP-independent activatg~.n
mechanism. In all other sy.,~tems, cyclic AMP could either directly or ind/rectly be
involved. "[his holds especial]/for those cases where honnone-sth-nulated ]/pases, rather
than phospholipases Az, migL~ be responsible for elevated arachidonate re~ase. For comparison, some well known ex.~mples of cyclic AMP-mediated, hermone-st/mulated, ~ b
hydrolases are included in T~.ble V. It should be noted, however, that some recea~ data
appear to indicate that epinzphrine-induced lipolysis in fat cells cannot be explained
solely on the basis of cyclic: AMP-dependent phosphorylation of the h ~ - s c a ~ t [ v e
lipase. It has been suggested that hormone treatment leads to mod/ficafion in the nature
of the lipase-substrate interac:ion, possibly through interaction of epinephrine with phospholipids resulting in facilita'Iedhydrolysis by the hormone-sensitive lipase [163,164].
IVF. Regulation of phospholb~ases by changes in membrane structure
It has recently been pro p,,sed [ 186] that activation of phospholipases might occ~¢rby
changes in tile structural arrangement of phospholipids in (plashy) membranes. That ~ach
a proposal is not merely sFeculative can be deduced ~,om the large influences •h/ch
membrane phospholipids ,'an exert on the activity of membrane-bound ¢~Tmes (see
Refs. 187 and 188 for recen: reviews). For example, Sinensky et al. [189] have recent"
correlated adenylate eyclase activatio:, with increased ordering of ~ acy[ c h a ~ of
membrane lipids. In contrast, reduction of the fluidity of phospholipid fatty, acyl c h a ~
gave rise to a decrease in (~a* + K÷)-s~imu~atedATPase activity [190]. Much o f ' ~ e information of this type has bcen obtained using reconstituted systems in whk:h the
lipid acyl-chain and polar-h~adgr-~ap composition can be varied at will Such recoa~ituted systems for membrane-bound phospholipases are not yet available. It c~.zmct be
excluded that reconstitution in fact takes place during the assay of purh'k,d membranebound phospbolipases A ('gables 11 and Ill), but this has not been proved rig~oa~..'l'here
is some indirect evidence w~ch might suggest that membrane-associated ~hosph~ipase
activity is hlfluenced by the fluidity of the membrane. Local anesthetics ca,-~cause both
membrane fluidization [191] and facilitated phospholipid cleavage [192,194]. The htte~
effect is strongly concentration-dependant and a direct correlaf~n between ~ a ~
and enzyme activity was '~ot establi'~.ed.
The effect of detergents on hhe activity of membrane.associated p h o s p h o ~ i ~ s h~s
been documented in nualerous reports. In most ca.s0s, ¢xogcmms . s ~ t e s w¢~,~ used
and the detergent effect is most likely due to facilitation e f the i a t e r a c ~ of the mem~
.brahe-behind enzyme with ~ e dispersed lipids. In general terms, this can p K ~ , y be
accomplished by events such as loosening of She membrane str~ture, pe~aps accom~
panied by a partial solubilization of enzyme or enlmnced ~acoq~ratioa of the
lipid s,:~strate in the residual membrane structure. On the other hand, the dete~fats may
only exert their effect at the substrate level by providing the ~
with a sabstrate
aggregate of mixed lipid.detergent micelles te ,shieSt the eazyme has a greater a f f ~ t y c~
penetration power rest/~ting in enhanced catal~lic capacity. An ohvio~ d i f ' ~
is t ~ t
220
various p,ar~:leters, e.g., surface charge, spacing of lipid molecules, fluidity of acyl chains
and absorpticm of enzyme to th,.• surface, as modified by the detergent may not be mutually independent. No systematic study unravelling these various possibilities has been
carried c,ut .~o far for membrar~e.associated phospholipases. Some analogies with much
better &.'fined systems using purified soluble phospholipases may be drawn here. The
latter sy:!;temshave been extensiYely reviewed by Verger and de Haas [9]. E~pecially from
monolayer studies, it has become clear that three factors are particularly important in
governing the activity of phospholipases added to lipid films, i.e., zeta potential on the
substratt~ s~rfac,v [195], packinl~ of the lipid molecules and absorption of enzyme molecules to the surface followed b.v penetration into the lipid monolayer [196,197]. Thus,
densely !packed phosphatidylcholine monolayers were not attacked by phcspholipase A2,
unless dicetyl phosphoric acid was also preselt [198]. Although this r~sult could be
explained by the increased distance between adj~zent lecithin molecules, similar enhanced
hydrdly.~;is was not observed when stearylamine was used as spacer. Apparently, a negative
surfac,.~ charge is of utmost importance in the initiation of this phospholivase (Na/a ha~a)
,~c:ivity. E~lEanced activity, also of membrane-boui,d phospholipases, has often been
reported by a~dition of anionic detergents, e.g., bil~ acids. In some cas~s, th,~ negative
charge provided by bile acids to the mixed-micelle surface appears to be an important factor in itself as similar effects can be obtained by ~ncluding negatively charged phospholipids h'~ the substrate aggregate. It is possible then, that membrane p'aosphorylation
resulting in ml increased negative charge might influence the regulatory membrane
environment of a phospholipase. It should be realized, however, that it is highly question.
able whether effects observed for the attack of lipid monolayers or aggregates by soluble
phospholipases can be extrapolated to situations where both enzyme and substrate are
part of the same membrane. A similar situation exists for the influence of the packing of
the phospholipid molecules in the membranes. Monolayer studies have clearly shown
[ 196,197] that a surface pressure can be created above which the lipolytic enzymes phospholipa:~e Az and lipase can no Longerpenetrate the lipid film. Below a surihce pressure of
about 2;0 dyn~:/cm for pancreatic phospholipase A2, there appears to be an optimum in
the velocity of enzyme action as a function of surface pressure. In elegant experiments
this wa:~shown to be the result of at least two counteracting factors [ 197]. Wi.th ine~ee~ing surface pressure the amount of enzyme in the interface decreased, wltile on the other
hand the specific activity of the interracial enzyme increased with the closer packing of
the substrate molecules at hi~er surface pressure.,;. That the packing of lil~id molecules in
the s,tt, strate aggregate plays an important role in the initial attack by pho~holipases was
also demonstrated in bulk experiments. Hydrolysis of artificial lipid bihye~, prepared
from well defined molecular sr~ecies of phosphatidyleholine, was strongly enhanced at the
phase ~:ran~tion iemperature. It was concluded that the coexistence of ~;olid and liquid
phases in the bilayer created irregularities in lipid packing at the border o["these domains
facilitating the penetration of the phospholipa.~¢A2 [199]. Similar conductions were
reached for a 1020-fold purified preparation of rabbit platelet phosphelipase A= [85].
Extrapolating from such obs.~rvations it is tempting to speculate that membrane-as.soelated phospholipase activities could be enhanced by any event that crea~!esirregularities
in the bilayer arrangement of phospholipids. Again, however, the experimental conditions
employing soluble phospholipases added to lipid substrates are essentially diflbrent from
those occurrh~g in natural membranes. In this case, enzyme and substra',~ are present in
the same membrane to begh with, thus precluding effects of facilitated penetration.
These differences stress the n~ed mentioned earliier in this section for a reconstituted sys-
211
tem of a membrane-associated phospholipase Az to enable a systema~/c ~tudy of the influence of the structural arrangement of the membrane lipids on phosphol~pzse acridly.
V. Specifi,- functious of intracelhdar phosphol[pases
VA. Release o f prostagl, mdin precursors
The r,.le of acylhydrolases in the release of precursor-free fatty acids for prostaggand/n
formation was firmly establishe~! during the last decade. Prustaglandin.s do not ~ : c ~ m
stored form in tissue phospho~pids [98]. Conversion ofarachidonate from ph,:~pho~p£d
into prostaglandins does not take place unless the arachidonate is fast releas~ by add/tion of a phospholipas~ A2 [200-202]. Once formed, the prostaglandins cannot be re/acorporated into phospi~olipids under conditions that allow efficient incorpo~'ation of
arachidonate [151]. T.Flese and other experiments convincingly demonstrated that the
cyclo-oxy~enase is limited in it,,; production of prostaglandins by the amount of precursor-free f~:ty acid. It was Iogica~to assume then, that endogenous acylhy&oIases fulfdJed
an impcr,.mt role in l:,roviding free fatty acid from endogenous lipids. Much wo~k h ~
since been carried out to answer ~he question as to which lipid class serves as the exclusive
or most important store of the arachidonate released prior to prostaglandin product~n.
On a quantitative basis, the sn-2-position of phuspholipids is, by far, richcs~ L~ gra~hidonate, suggesting that the enzyme involved in arachidoaat¢ release was a p h o ~ c ~
lipase A2. Much of th~ evidence for this involvement of phospholipase A2 has been discussed in. the previous subsections of Section IV. Especially in more intact b/o~oggcalsystems, such as perfused organs, often only fatty acid release was measured. Thus, R
not always clear whether arachidonate release was exclusively from phospho]/p/ds and, ;,t"
so, from which phosphofipid class(es) and by which mechanisms. This proble:la w/~ not
be discussed here in g~.meraltexans, but will be considered in detail in the sy,s',em which
has been investigated most intensively, i.e., platelets. A recent review [204] on the role of
lipids in platelet functi,m has dealt with data published until 1977.
The possible pathways for arachidonate release in plat61ets are summa~ec:, in Fig. 4.
One of the earliest studies on lipolytic activities in platelets was can'ied out by Smith and
Silver [205]. Using specifically labelled phosphatidylchoiines they detected ~,~phosphc~
lipase Am activity, which was optimal at pH 4.8 and required taurocholate. Treatmen~ of
human platelets with thrombin released the enzyme into the medium. These cha~'acteristics supported.the belief that the enzyme was of lysosomal origin. The same enzyrae has
also
•
been detected in rabbit platelet soaicates [203]; Ca2+ did not affect the acridly. The
substrate specificity of this enzyme has not been incesti~ted. As ~ t c d
in Fig. 4, this
enzyme can only con tlSbute to arachidonate release in conjunction with a lysc~:ospholipas¢. The latter enzyzne, optimally active at pH 8.5, was reported to be falsest ix~rabbit
[203] and human [I:1] platel,Rs. At pH 4.5 the lysophospholipas¢ activR~, ~ s almost
completely abolished. This an¢. the finding that the optimal p h o s p h o l i [ ~ A~ ac~v/ty at
pH 4.5 amounted to o:aly abou125% of that of a Ca'-requiring phosphc~pase A~ activity.
at pH 9.5 have direcled mucl:~ attention to the latter enzyme as a gkdy c a ~ d ~ e fo~
arachidonate release in platelet,,;.
Studies on the sub:;t~te sp(:chqcity of this phospholipase A~ have not yet i ~ d a
clear picture. Depon¢ing on whether endogenous or exogenous p h o s p h ~
~
u~d,
different results for the prefer, race for phuspholipid class and acyt cha~'~at rig, s~-2.posi-
222
Ph
"3e'Ph°"
i.iDa,~
A1 I.....
~/~'IL._p~(
.............. -] PhosphoLipas~C
~. X= Inosltot
~
,-- OH
i
PhosphoLipas¢ A2
~~"~
- - ~/N~
s ~
ATP Phosphatldlc
" Ki'na,s~~ Acid
ON
Lysophospho!OH~
I,ipasQ
HO--Lp X
---
/
AA~ r rl
/ i
-~-HO OH
.....
' LYS°Al~cl~'atidi¢
Fig. 4.. Possible pathways for atachidonate release in platelets.
tion were reported for this phospholipase A= (compare Table VI). When human platelets
were labelled with arachidonate in platelet-rich plasma, subsequent treatment o f washed
p'atelets caused a major decrease in the radioactivity o f phosphatidylcholine and phospnatidylinositol [206,207], Similar patterns for arachidonate release were noticed when
washed platelets were prelabelled [208] or when the phospholipase A2 was stimulated
TABLE VI
SUBSTRATE SPECIFICITY OF PLATELET PHOSPHOLIPASE A2
PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, phospbatidylinodtol; PS, phosphatidylsedne.
S~.'cies
Preparation
Substtate
pH
Human
pselabelled cells
pselabelled ceils
prelabelled cells
prelabelled cells
endogenous
endogenous
efidogenous
endogenous
7A
7.4
7.4
7.4
Horse
Rabbit
prelabelled cells
prelabelled cells
endogenous
endogenous
7A
7.4
EIuman
l~,abbit
membranes
membranes
endogenous
exogenous
9.5
9.5
* Release of.radioactive arachidonate:
** Release of total arachidonate.
ConhAbufion to fatty acid
release (%)
PC
PE
70
78
68
0
7
0
PI
Reference
PS
25
5
11
4
32
0
arachidonoylPCspecies only,
all speck~ of Pl + PS, moPE .
species
31
34
35
42
22
29
?*
14
58
25
3 **
28
69
<3
<3
preference for PE, no preference
for fatty acid
206
207
208
211
142
209,210
209,210
132,217
203
223
with ionophore A 22;187 [144]. In contrast, phosphatidylcholine, p h o s p h a t i d y ~ t o l
and phosphatidylethanolsmine contributed about equally to the thromb'm-h~uecd arachidonate release from prelabelled horse platehts [142,143]. Remarkably, when the c¢'~
were treated with ionophore A 23187 rather than with thrombin to induce ptmspholipase A=, radioactivdy labelled arachidonate was only lost from phosphatidyidmlinz and
phosphatidylinositol~ bul'~no longer from phosphatidylethanolamine [ 143]. It is presently
unknown whether this can be used as an argument for the presence of different actiratable phospholipa~s A2. Some of the contrasting findings with regard to the question
as to which phospholipid(s) donate free ara.'hidonate after phospholipase A2 stimulation
can, at least, partially be explained by the different extents to which the v a ~ phospholipid classes incorporate arachidonate during prelabelling [142,143,206,207]. Hardly any
labelling of phosphatidyl[serine was observed in horse platelets [142], thus excluding the
possibility to measure arachidonate release from this phospholipid. Only minor anmunts,
if any, of arachido~Late were released from phosphatidylserine and p h o s p h a t i d y ~
amine in intact hmrmn platelets, despite the fact that about 6 and 12% of the labelled
arachidonate had been incorporated in these phospholipids, respectively [206,208].
Blackwell et al. [209,210] emphasized the disadvantages of using prelahelled plat¢lcts in
which the specific radioactivity of the arachidonate may vary with the phospholipid class.
When the collagen~,;timulated release of radioactive arachidonate from prelahelled rabbit
platelets was measured, 41.6% originated from phosphatidylcholine, 28.7% from phosplmtidylinositol, 22.5% from phosphatidylethanolamine and 7.2% from phosphatidylserine. When the Ice01 sizes of arachidonate in the phospholipids as deduced from gas
chromatographic analysi:s were taken into consideration, it could be calculated that phosplmtidylcholine ~ntributed 14-~%, phosphatidylinositol 24.6%, phosphatidylothandamine 57.7% and phosphatidylserine 3.4% to the total arachidonate release. Two assumptions were made irL these calculations. First, the phospholipase A2 does not differentiate
between [1-*2C]a:rachidonate and [1-~4C]arachidonate. This assumption is generally
made in radioactive-tracer studies and in all likelihood is valid. The second asmmption is
that complete mixing of phospholipid species containing uulahelled arachidonat¢ with the
corresponding species containing [l-z4C]arachidonate has taken place with respect to
availability for phospholipase A= hydroly'~. The validity of this assumption is much less
likely and at least should he verif'~d experimentally. In other words, the possibility exists
that a small pool of, e.g., phosphatidylethanolamine becomes heavily labelled during the
prelahelling procedure, perhaps because of its localization in the plasma nmnbran¢ or certain areas~thereof. Then, the percentage hydrolysis measured for this small pool does not
necessarily apply t o the total platelet phosphatidylethanolamine pool and calculations
based on this amlmption may very well yield erroneous results. The validity of such an
as0umption shou/d he checked independently by showing that the specific rad/oactivity
o f the phosphol/I~id class does not change during phosphofipase A= action. In this light,
the conclusion that most of the arachidonate, i¢lcased during collagen-induced
tion :of rabbit platclcts, originates from phosphatidylethanolamh.ae should still b¢ contentative, Appalantly, species differences contribute to the variabilRy as to which
phospholipi~;oc~tributc~most ~to arachidonate re.lease.Thus, p h ~ k t i m m C ~ u i a ¢
is n o t h y ~ z e d in human platelets [206-208,21 ]], while it is in rabb~ [209,210] and
horse~[142;l~] plmle~s.
~
detai~:studies o n ~ specificity of the phospholipase A, in hmma platelcts
w e m ~ m ~ d bl~lills ot al. [21,]. Using cells prclabelled with various u m m ~
fatty
acRIs,~~_hoiLq¢
hydrolysis upon thrombin trcatm~.t was only found m araclb
224
idonate-labeUed platelets. This suggested that the phospholipase A2 which utilizes pho.,~.
phatidylchol~ne is specific for the arachidortoyl.cont~ining species. In support of this
notion, a 25.6% decrease in radioactivity from phosphatidylcholine wa~ accompanied by
a 7.6% loss ~f phosphatidylcholine phosphorus, in agreement with an~dytical data indicating that human platelet phc~sphatidylcholine consists of about 24-30% of arachidonoyl '.;pecies [213]. Arachidonate release was also observed from the phosphatidylinositol + phosphatidylserine fraction, but not from phosphatidylethanolamine. In contrast
~o the specific release of arachidonate from phosphatidylcholine, hydrolysis of phosphatidylinositol and/or phosphatidylserine was also noticed in cell:; radioactively labelled
with oleate or linoleate, Due to much less efficient incorporation of these acids in comparison to arachidonate, the total release of radioactivity amounted to only a few perce'at
of that observed with ~rachidonate. Corroboratively, when the amounts of non-radicactive fatty acids were measurec[ only arachidonate was seen to ac :umulate in ~espons~: t,~
thrombin [211]. Such data suggest a selective mechanism for arachidonate release either
through acyl chain spe,-ificity of the phospholipase A2 involved or through compartmentalization of the substratc's in relation to the enzyme. Acy| ~.hain specificity cannot
be excluded, but it should be noted that this has generally not been found with other
phospholipases A2. In adtiition, such specificity is not found w~th platelet membrane fractions in combination with exogenous substrates (see below). C')mpartmentalizatic.m of
enzyme and substrale appears to be, therefore, a mote likely explanation for the specific
release of arachidonate upon thrombin stimulation of human platelets. It beco~.s interesting then to review the scarce data on ph6spholipid and a~c$idonate localization in
platelets.
Ch~p et al. [214J estimated that 63% of the phospholipids ht pig platelets are locat,:d
in the plasma membrane. The composition of those plasma membrane phospholipids is
26% sphingomyelin, 30% phosphatidylcholine, 27% pkosphatidylethanolamine and 16~o
phosphatidylinositol + phosphatidylserine. From a comparison ,)f the degradative action
of N. na/a phosphclipase A2 and Staphylococcus aureus sphinliomyelinase on lysed axd
intact platelets, it was concluded that the outer mov.olayer contained 46% of the total
plasma membrane phospholipids. This tayer containe#, 91% of the sphingomyelin, 40~ of
the phosphatidylckoUne, 34% of the phosphatidylethanolamino and less than 6% of the
phosplhatidylinositol + phosphatidylserine in the plasma membzane. Similar distributions
were found in hur[mn plateh.,ts, except that somewhat less of the phosphatidylethancd.
amine was present in the out~.~rplasma membrane leaflet. This appears to be compcnsat0d
by higher percentages of phosphatidylcholine and phosphalidylinositol ha ti~-. outer
monolayer [215]. These data extend observations made by Schick et al. [216] who u~.~d
trinitrobenzanesullbnete to hlbel aminophospholipids. In intact human platelets, no phosphati~ylserine reacted withl ~:beprobe, while phosphatidylethanolamine labelling leveI,.~d
off at 12-17% in reasonable agreement with the 12% of the total platelet phosphatidyb
ethanq~unine which could be hydrolyzed by exo~mous phospllolipase A2 umler non4ytic
cond/tions [214,215]. What bearing do these ph~holipid looalizations have on the <~bserved selective hydrolysis of'phoslphatidylcholin¢~and phosphatidylinositol by the thrombin.stimulated phospholipase A2 of human platelets? One sl~raighfforward explanation
could be given by asrauning that the phospholipase A2 is located in the outer monolayer
of the plamna membrane where it hydrolyses phosphatidylchcline and phosplmtidylinosi.
tel. This necemtates the further assumption that the mall ~mount of phosphatidylethanolamine in the outer monolayer either is not ivailable for the er,~zyme or does not
contain arae.hidonate. Two arguments at leas'~ can be provided agah,.st this reasoning.
225
Firstly, evidence is available which seems to argue against a Iocat~.at~on of the ~ > z f ~ lipase on the outer surface of platelets (discussed in subsection IVA). : % c ~ t y , atth~gh
about 60% of the platelet phospholipids is thought to be present in the [~ta~r~amembrane
[214,215] whh about 30% than being in the outer monolayer, studies on the d~v:but[~a
of total arachidonate have led to the conclusion that only 10% thereof is located kn the
outer surface monolayor. Thus, the outer surface seems to be rather ~ni~over=~hed in
arachidopate. On the other hand, if the platelet phospholip~_~
~ Az is located in the L,~r.er
l,~aflet of the plasma membrar~e or in an intracdluiar membrane it is not o b ~ s , at ~esent, why specifically phes~l:atidylcholine and phosphatidylethanola~e are attacked,
considering the fact that most of the arachidonate-containlng other phospholipid ~ . - ' c ~
ate present at these cellular sites as well. Compartmentalization of the enzyme and substrate wcuId then have to be proposed as the basis for the observed selective de~adatkm
of phospholip~d classes or species. In all likelihood, such compartmentalizatkms occur in
the intact platflet prior to anti during stimulation of the endogenous ph~sph~pase AzAt the same time, it should be realized that the ex~rimental evider~ce to s u p p ~ sach
comp~xtmenta'lizations is very poor indeed.
Other experiraents to elucidate the substrate specificity of the platdet ph<~holipase A2 have employed isolated total membrane fractions rather than intact c¢~. Derksen arLd Cohen [132] incubated isolated membranes in the presence of Ca~ and determined 8as-cl~romatographically the amount of fatty acid released from e n d c ~ e ~ s sabstrates. Optimal release of amehidonate was observed at pH 9.5, while pahmtate, stearzte
and oleate :release peaked at pH 8.5. This led the authors to postulate the p r ~
of
phospholipase Aa and A2 activities with optima at pH 8.5 and 9.5, respectively, h c ~ m ~
he concluded with certainty, at the moment, whether the optimal release of :~tarated
fatty ~ids and oleate at pH 8.5 is due to the presence of a phosph~l~pase A~ oe to the
presence of phospholipase A~ plus lysophospholipase [203]. At pH 9.5, the increase
free arachidonate was accompa~ied by an equivalent loss ofarach~d~,ate fr~a the ~ pholipid fraction. The putative phospholipase A~ responsible for this release ~ c ~ d he
expected to release linoleate wRh equal facility as this acid shows a propens/D" ~
~
arachidonate for este~cation at the sn-2-position of phosphog~ycer~des. L~ f ~ . i~ w ~
observed that arachidonate and linoleat~- wet~ released in a molar ratio o~ ~2-~5 :
whexeas the phosphofipids of freshly isolated platelet membranes c ~ t ~ e d these f~Lv
acids in a molar ratio of about 6 : 1 [132]. In a subsequcm, more , ~ t ~ d s~qy h v~gs
shown that these molar ratios in phosphatidyl~holiae and ~ h a t k / y l e ~ , ~ o ~ ¢
amounted to 1.7 : 1 and 17.8 : 1, respectively. This sugg~tcd flint, in t ~ a b s ~ o f ~ y
acyl chain specificity of the phospholipase Aa, p h o s p h a f i d y l o t h a n o ~ w<~u~dbe the
preferred substrate. Direct measurement of the arachidonate coat~nts of th:.~ p ~ h ~ lipids and free fatty acid chains before and after incubation corroborated th/s s u ~ .
Of ,:he total arachidonate re!eased, 28% originatod from idaosl~t/dyk~',~'~, 6 ~ from
phosphatidylethanolamine and only about 3% was released from sc~ces older t ~ these
two phosphog|yccrides [217]. Using ~, particulate fr~tioa is~ated from rabb~ ~ t e b ~
soaicates ~ad exogonous labelled subst~ates, Kannagi and K~zun~ [203] s ~ m ~ " arrived
at th¢ con :usion that phosphatidylethanolamiae was the paffea~l sa~tra~e~ Und¢~
conditiors, the enzyme did not exlfibit strict acyl daaia spgcit~ity. In fact. hy&~ys/s of
l.acyt.2.l.l:~C]oleoyl phosphatidylcholine was slightly ~ ' - r thaa that of :l-~cyb2[l:~C]a~chidonoyl phosphatidylcholine. It r e m a ~ to be ¢ s ~
why ~ i~:~:~
plateiet membranes phosph;~tidylethanolamine is hh~ ~referred substrat~ [217~, w b ~
intact ht,man platolets no ]:,reakdown of this p h ~ p i d
is obser~'ed [206~07~ ~t ].
22~
Two po:~ble explanations for this discrepancy ,~ome to mind. The phospholipid degrada.
tion in intact platelets is always measured ~,fler stimulation, ~Jsually by thrombin. Does
the basal phospholipase As, as assayed in isolated membranes, show a different substrate
specificity from that of the thrombin-stimulated phospholipa,,e A=? The other difference
which should not be overlooked is that the membranes were incubated at pH 9.5, while
studies with intact plat.elets were performed at pH 7-7.4. It raay be r,~.levant to mention,
in this respect, that a part,ally purified phospholipase As from rat liver mitochondria
hydrolyzed phospha~:idylethanolamine more rapidly at pH 9.5 than at pH 7.4, whereas
Me reverse was true for pho:;phatidylserine [55].
In most studies with either platelets or men~branes thereof', attention was focussed on
fatty acid release At best, this was correlated with a.decrease in the fatty acid under considerati,3n in the phcspholipids or in a given phospholipid chss[132,211,217]. This cot'relatioll in itself doe~ not prove the action of a phospholip~.se~A2, a~.~similar results would
be expected for the conjunctive action of a phospholipa.ce C and a lipase (compare
Fig. 4). It is therefore important to note that the presence of phospholipase A2 in platelet
membranes has been demonstrated unequivocally through isolation of both fatty acid and
lysopho.~holipids ~203,217]. Jesse and Cohen [217] analyzed the lysophosphatidylethanoLamine that was produced from endogenous phosphaiidylethanolamine along with
the fatty acid relea~ and 'showed that it contained virtually no fatty aldehydes. This h.,d
the authors to conclude that exclusively diacyl phospaatid/lethanolamine, but no pl~smalogen phosph,,~,~-ylethanolamine, was attacked by the phospholipase A2. it was not
stated, however, ~!Lether accumulation of lysophosphatidylethanolamine was stoichiometric with pllosphatidylethanolamine breakdown. It should be considered that rapid
incorporation o~" a.rachidonate from phosphatidylcholine ~uld phosphatidylethanolamine
into plasmaloge:a phosphatidylethanolamine has been repotted in human platelets [207].
Thus, th~ pauclity of fatty aldehydes in the isolated lysophosphatidylethanalamine could
also be due to a ~elective reacylation of the plasmalogen lysuphosphatidylethanolamine.
That the above mentioned possibility of the concerted action of phospholipase C and
lipase could explain both fatty acid release and phospholipid breakdown in intact platelets is not purely hypothetical and has been ~'eported by several investigators. Rittenhouse-Simmons [212] demonstrated that human platelets generate diglyceride:; within 5 s
of exposure to tbrombin to yield up to 30-fold incceased levels. This diglyceride accumulation was transito .ry and within 2 rain the cells had essen~:iallyachieved the low levels of
diglyceride characteristic of the resting state. In cells prelahelled with arachidonate within
the time cour~ of diglyce,ide accumulation, only phospkatidylinositol show~.d a sizable
loss of radioactivity, which could easily account for the gain in diglyceride. Other potential diglyceride precursors, either did not lose ~dioactis'ity (triglycerides) or showed an
increased amount of label (phosphatidic' acid). Phosphatidylethanolamine and phos.
phatidyl~rine did not lose radioactivity in the examined period, while phosphatidyl.
choline started to show signific~lt loss of labelled arachidonate only after 30 s and con.
tinued to do so well after the time that diglycerlde had returned to normal levels. These
data strongly suggested that phosphatidylinoaitul was the source cf the diglycefide produced. In line w~th this conclusion, a phog~hatidylinositol.specific phmphodieaterase was
discovered itt the soluble fraction of platelet sonicates. T i ~ phospholipase C-type enzyme
hydrolyzed phosphatidylinositol to form diglyceride and ino$itol phmphate, but did not
produce diglyceride from phosphatidylcholine, phc~phatidylcthanolamine or phosphatidylserine. The presence of this enzyme in platelets was independently discovered
also by Mauco et al. [21:S] and Boil et al. [219]. The enzyme required Ca~* [212,218]
227
and, when assayed with exogenous phosphatidylinos~tol, apparendy a n~tively ¢har~'d
lipid surface. The enzyme was active only in the presence of deoxychohte [2 t 2] or ~
phosphatidylino-itol concentrations exceeded 50 gM [218]. Unfb~':lr~tebl, no da~ are
yet available to explain the sudden activation of this phosphafidy~o~toi-g~ec~ ~ y ~
phodiesterase upon thrombin treatment of platelets. Interegingly, P d t t e ~ c ~ e ~
[212] d,~monstrated that pretreatment of platelets with dibutyryl cyclic AMP ¢ o ~ e t ¢ l y
inhibited both diglyceride production and serotonin release. Dibuty~,t cyclic AMP, v ~ h
is thought to act, in part, to restrict Caz÷ mobilization, could thus p~event the a c ~ / o n
of the pho~hodiesterase, which in the absence of dibutyryl cyclic AMP ~ g be c d ~
by making stoJred Ca2÷ available to the enzyme in much the ~ n e way as ~ bee~ proposed f~r the platelet phospholipase Az (compare subsection WC). A ~
co~ct~ion
was rea,-hed by Billah et al. [318].
Bell et ai. ['219] have examined the fatty acid1 compo~tion ofthe accum~ted ~ ceride, whi~ch consisted of aimost equimolar amounts of stearate and a ~ c h i d ~ t e , h~
with the fatty acid composition of platelet phosphatidylinositol [2|3]. Fuzthem~re,
confirmation o:' preliminary data obtained by Manco et al. [2120], they ~ demor~r~ed
the presence o¢ diglyceride lipase in the particular fraction of platele~s. Tb.~s enzyme
could generate the arachidonate for the burst of prostag~ar~din and t~ombox,~e ~y~thesis that follows thrombin stimulation. Thrombin stimulation releases about ~-lO
nmol of arachidonate per 10~ platelets. Since the activity of the cytosoh~ phosp~odiesterase amounted to 15-20 nmol phosphatidylinositol hy~ro]yzed/min pe~ 10~ ce~
[218] and that of the particnlate diglyceride lipase to 3i mnol fatty acid r e l ~ / m L ~ per
10~ cells [219], it follows that the activity of these enzyme,s is sufHcicat to account for
the arachidonate release. I) should be pointed out, how~,ver, that the aboYe-ment~oae~
activities were measured wit.~ saturating substrate levels, ~ hic.h may not rg~-es~* ~pp~'
to the in vivo situation. This holds especially for the low levels of tr~x~¢o_,y ~g~yce~¢
substrate for the lipase. Oq the other hand, the involvement of a ph~tidy]L,~e~o~specific phospholipase C and a diglycetide lipase in arachidormto release from s~Lm~ed
plateI.ets offers an explanation for several findings observed shortly after ~ b / n
activation. Firstly, that only arachidonate is released upon thr~mbin stiroula~o~ n~y be
explained by the fact that the diglyceride substrate originates from 9 h o s ~ y L ~ w g
and ,.'ontains virtually only arachidonate in the sn-2-position. The release of ~ c h ~ d o ~ e
and other unsaturated fatty acids from other phospholipids by an ac~va~ed p ~
lipase A~ would in this view be a secondary event following the burst of ~ t b
dytinositol degradation after thrombin addition. Secoud~ty, transient dig|yceri~e, a c c ~ laticn followed by phosphorylation of the diglyceride could e x ~ ' ~ the ~rea~cd k~e~s
of phosphatidlc acid repeatedly observed in platdets after thrombL~ s ~ m ~
[~43,
212,220]. This indicates that the diglyceride lipase is not the only e . ~ e in p ~ t e ~ s
that acts on the diglyceride. The diglyceride lipase in pl~tetet membr~es s~owe~ ~t~e, ff
any, activity towards triolein [21~a]. Also, in another aspect, th'~ lipase is unique ~a that it
apparently removes fatty atids from the sn.2.position. Most other ~pascs stucL~ so f~
sho¢~ an overwhelming preference for the acyl ester bo~d at the p ~ y
hyd~x~
and fatgy acids from the sn-2-position are only hydrot,/z~ a f ~ ~ ! m/~atio~. Such a
non;enzynmtic step is difficult to accept in a ~
which ~ e s
~ ~ e
rel0ase from platolet lipids. When the pasticu|ate plate]et fract~o~ was mcubate~ ~
diglycerid¢ labelled with ~H in the glycerol backbo~ and ~C in the fatty acid at t ~
sn-2-position, n o radioactive, moaoglyceride was detected [219]. This m~'ates ~ the
en~,yme hydrolyzes stearate from the sq-l-position as fast as arachidoaate f ~ the ~-2-
228
position, ldterrmtivety, other enzymes in the particulate fraction might prevent the
accumulation o f monoglycerides. Two possibilities come to mind. A monoa,.-ylglycerol
hydrolase laas recently been des,~ribed for human platelets [221]. This enzyme has not
been a,~sayed with diacylglycerol as substratJ.'. Several prope,~ties seem to be rathe~ sirMlar
to the diglyceride lipase described by Bell e~ ai'. [219]. The possibility of both enzymatic
actMties residing +m a single protein seems worthy of inve.,~tigation. Whether or not the
diglycerid:: lipase and the monoglyceride lipase at,.+ the same or enzyme~; a~ting in
~equence, this pathway does not explain why upon thrombin stimulation only ",~achidonate arm no stearate is released. This could, perhaps, be explained by the presence of a
second enzyme preventing the accumulation of monoglyceride, i.e., monoglyce+ride
kinase. The latter enzyme, perhaps identic.'d to diglyceride kinase, as well as the acct~mu.
lation of iysophosphatidic acid during thrambin-induced platelet aggregation have been
described by Mauco etal. [220]. These authors also pr6vided evidence to show that the
lysophosphatidic acid was not formed from phosphfLtidic acid through phospholipase A
action. Obviously, many of the enzyme specificities in thi.,~recently discovered phospholipase C plus diglyceride lipase pathway for arachidormte release remain to be established.
VB. Phospholipid turnover
Phospholipids have long been viewed as static building blocks of subcellular structures.
As in many areas of biochemistry, an extr'~ dimension was added to this structural role of
lipids when radioactive isotopes became available. The early exF.+riments employing 32p
sh~wed that the phospholipids of all tissues became :~apidlylabelled. This rate of labelling
of tissue phosphokipids was hardly influenced by h~.'patectomy, except that under there
experim,mtal conditions no labelling of plasma pho:~pholipids occurred [222]. These and
many other experiments using isolated organs, tissue slices or homogenates have established the view that virtually all t.:ssues are more or less au~onomons in their phospholipid
synthesis. Not only was a rapid labelling qf the phospholipids observed after a pulse with
radioactively labelled phosphate, but the incorponted radioactivity disappeared rapidly
again from the phospholipid fraction. These observations have led to the concept of phospholipid turnover, in which a constant supply of newly synthesized phospholipids is balanced by a removal of existing ~olecules. The half-lives of the total phospholipid pool or
of specific phospholipid classes in only a few tissues or cell-types are summarized as
examples in Table VII. A consideration of' the data in Table VII clearly shows the complexity of the turnover phenomenon. Relatively short half-lives o f disappearance of
labelle~l #tospholipids are notic~,d in all tissues, except for brain, notably the myelin fraction (ExpL 5). Considerable variation is found in dopend,.nce of the precursor t ~ d (compare Eztpt. 3 with 6 and the re:mRs for kuag in Expt. 8). This has to be ascribed mos;
likely to extensive re-utilization of the labeUed atoms, thus giving rise to falsely high values for the troll.life of the component under invvstigation. The early experiments have
often t+sed the disappearance of label from the total phospholipid pool of whole tissues~
Although .~uchexperiments were adequate to damonstratc the dymunic character of tissue
phospholipids, it ha~ since been tealized and ,demonstrated that the individual phospholipid
classes turnover indepen~tly arld at qu/te different rat~s (Expts, 5 ' 7 , 16). Eve~ within
a givett phospholipid c h u i the different mol,~ular species which compriN that class tun,over a~I different rates (Expt. 9). Turnover studies in whole organscan obviously be influ.
enced strongly by secretion of intact phospholipid molecnles. This certainly contributes
229
a problem in studies with livel, which is known to s:¢nth~ze plasma and bile phospholipids. This raises the question as to whether the phospholipids in the liver itsetf ~um over
a( all This has been studied by isolation of liver subcq.'llularmembranes at vvrious ~ r ~ d ~
after pulse-labelling. The data in Expts. 10--14 of Table VII clearly clemormrate that
pi~ospholipid turnover occurs in all suoce!~u',!armembranes. In accerdaz~ with expectation, the observed half-lives are considerably longer, however, than found in whole liver
(compare Expts. 1-4 and 10-14). Although considerable variation in the data can be
noticed, it seems justified to conclude that the glycerol-pho,,;phate-base (hydropt~=)
backbone in the merabrane phospholipids turns ow:r faster than the h y d r o ~
acyl
chain part of ~he molecules (Expts. 10 and 14 as oplmSed to 12 an,:[ 13). Thh poL~ts to a
~nore extensive re-utilization of acyl chains when compared to the glycerol-phosphatebase backbone.
It is an interesting and, as yet, unresolved question as to why turnover of p h ~ o lipids, constituting metabolic endproducts, should be necessary for the celt's physiolo~'.
Provocative ideas on this subject have been formulated by Dawson [242], who corn/tiered
natural c2eath and autolysis of ceils, movement of ceil membranes in exo- and endocyto~
processes and maintenance of structural integrity of celluhr membranes to support meml.~rane functions as processes which might have some bearing en the phenomenon of phospholipid turnover. Arguments have since been put foraard which seem to suggest that
replacement of dead cells in tissues through cell division and cell growth is v.ot the orJy
mechanism underlying phosphoEpid turnover. Studies with cells in tissu~e culture have
shown that phospholipid turnover was similar for growing and non-growing cells umder
conditions where no secretion of phospholipids into the medium took place [239,24t lWith mouse fibroblasts [239], metabolic inhibitors which suppressed the synthesis of
phospholipids caused an equivalent decline in the rate of degradation. Thus, the ar~bolbc
and catabolic events of turnover appear to be tightly coupled in these cells, l n c r e ~ d
turnover of acidic phospholipids, notably phospl~atidic acid and phosphatidylLno~tol,
during phagocytosis or horn~one stimulation of ~arious glan~ (~¢e Ref. 243 for review) is
usually accompanied by metabolic stability of the main phospholipids phosFhatidylchcAine, phosphatidylethanolamine and sphingomyelin, during the period of investigatk>n.
The most likely reason for phospholipid turnover than appears to be that it sust',~s membrane integrity and membrane function. Such a vague conclusion is not vet/" sat~factory,
however, since it in no way explains the necessity for continuous synthesis and degra~lation of phospholipids. A plausible ~xplanation could be, as suggested by Dawson [244~.
that turnover is requ~ed to replace damaged phospholipid molecules due to auto-oxk~tion of the polyunsaturated fatty acid chains in ph(x~'pholipids.It has recently been s h o ~
that lipid peroxidation of membrane suspensions in vitro leads to incensed phospho~pic[
bilayer rigidity, which in turn can cause impah-n~nt of membrane functions [245]. If
auto-oxidation were the main reason for turnover, one would expect shorter half-Hves for
",he more susceptible (poly)unsaturated phospholipid Slmci~ and appreciable m e t a ~
stability for the di~turatcd phospholipids.
Clearly, this is not the case Gable VII, Expt. 9). Trewhelh and Cdlh-~s [246| have
likewise concluded that of all the phosphatidylcholine molecules pt-esent in r~t live~, the
l-stearoyl-2-arachidonoyl species had the slowest rate of tumoset. We a-e, therefore,
faced with the problem of being unable to ptovi~le a compelling reason as to why ph~pholipid ~urnovor is required.
In this review I shall not deal with the bioffnthetic part of this tumove~. A fe~
remarks on the catabolic sequence of turnover seem a p p ~ t e
in a ~e~ew on
[ 14Cl~rine
["*Clm
Rat
Rat inte*~tne
Rabbit lung
(7)
(8)
.
to)
s2;,
tc
[Sl-llcholine
[~iijiOycerol
114Ch~dndtate
[3~Plphosphate
[_~=Plphmphate
[= "Plphmpitate
[~4Clcholine
.[32P]pbosphate
h i liver
Fm~_liver
Pat liver
Pat liver
Rat brain
(1)
(2)
~3.
(4)
(5)
Precursor
Tissue
Expt.
total
total
total
PC
PC
PC
PC
total
PC
_PC_
PC
PC
PE
P!
PC
PE
P!
PC
PE
PS
PE
total
total
tot~l
total
mito
myelin
Phosp]i~-~pid class
Tissue fraction
6 - 10
1.8- 3.0
10.9
10
340
680
5O
1350
4750
850
1.2
2.1
7.a
0.4
1.0
88
13
8
10
Half-life (h)
231.232
230
230
225-227
223
224
229
228
Reference
TABLE Vii
TURNOVER O F PltOSPHOLIPIDS
Abbreviatiom m in Table VI. ~.0, &l,/,2 and/*4, disathrated, monoenoic, dienoic and tetraenoie species, respectively; mito, mitochondria; OM, outer membrane;
IM, imm= memlmme;microL: micr.o~
_mm_..~;..m~_,p!~.--,~am~mbrane; BHK, baby hamster kidney.
Pat aveJ
Pat I~er
Pat liver
Rat~ver
Moow flbrobla~
BiiK ceib
Ne,oplutic m u t cells
1~
(12)
(13)
(14)
(15)
(16)
(i 7)
• ~
pat ~
,
kat l i r a
(1o)
(9)
all orl;anelles
[ 14CJcholine
I SH lslycerol
total
total
['4C]~).co~
[ ! 4Cjacetat e
Phi
mito
mlto
micros.
mito
OM
IM
micros.
micros.
[32PJphosphate
[3Hlglycero[
[ !4Clcholine
micros.
[ 14Clglycerol
[=Hlethanol
PC
PE
total
PC
PE
Pl
PC
total
PC
PC
PC
PC
total
PC
PE
total
backbones
acylchalns
te~d
,~4
PC ~0
AI
a2
12
0.7-1.8
2.5-4.0
7.8-18
44
29
140
38
[6
16
15
16
3
8
48
4~
48
48-72
48
48
24
4
4-6
15
15-25
241
239
24O
238
236
235
237
234
233
232
lipases A. It is generally thought that the citabolism of diacyi phosphoglycerides in most,
if not all, cells proceeds through a coasecutive deacylation catalyzed by phospholipases AI or A2 znd lysopl~.ospholipases. Phospholipase C activity is mainly found in
some bacteria, while phospholipase D is mostly confined to the plant kingdom. A few
exceptions to these general rules are known. A phospholipase C-type enzyme with absolute specificity for phosphatidylinositoi is present in many mammalian cells, but its activity tow~trds other phosphoglycerides has so far not been reported unequivocally. Phospholipase D has recently been partially purified from rat brain [247], but the role c,f this
enzyme in turnover has yet to be established.
As early as 1955, Dawson [248] concluded from specific radioactivity versus time
relation~dlips that glycerophosphocholine and glycerophosphoethanolamine were intermediates in the catabolism rather than in the biosynthesis of phosphatidylcholine and
phosphatidylethanolanline in rat liver. "[he intracellular deacylating enzymea catalyzing
these c~mversions have only later.been detected (Section II). As might be expected for a
consectttive deacylation pathway, the specific radioactivity of lysophosphatidyicholine
was f,~und to be intermediate be',ween those of phosphatidylcholine a.~ld,glycerophosphocholine [249]. The deacylat~on proces,; apparently proceeds without any appreciable
accunmlation of the intermediary lysophosphatidylcholine, the amount of which appears
~,o be kept rather constant at a level of about 0.5-1% of :he total lipid phosphorus This
implies that the intracellular enzymes utilizing lysophospholipids, i.e., acyl~oA : lysophospholipid acyltransferases and lysol:hospholipases should be able to cope with the
production of lysophospholi[,ids throu[~ phospholipase action. Over 8(F/b of t~e lysophospholipase activity in rat liver is pr~sent in the 100 000 × g supernatant of homogenates [250]. The subcellular distribution of the total lysophospholipase activity varies
considerably from tissue to tissue (des,:ribe'd in Ref. 250). For bovine liver, it has later
been demonstrated that the tissue contained at least two enzymes with lysophospholipase
activity. Both enzymes were purified to homogeneity [251 ] and their individual subceliular localization was determined [252]. One enzyme appeared to be in the micxosomal
fraction, where,as the other lysophospholipase was mainly recovered in the cytosol. The
presence of soluble ly~phospholipases in both rat and bovine liver raised the question as
to whether these enzymes can act on ly.,~ophosphatidylcholine embedded in membrane
structures. Most likely, this is the form :in which the lysophospholipid is present when
produced during phospholipid degradation by intracellular membrane-bound phospholipases. Studies employing the purified b o ~ e liver enzymes have ~own that this enzyme
is active against iysophosphatidylcholine embedded in both model membranes [253] and
liver microsomal membranes [254]. Thus, the cytosolic enzyme is likely to contribute to
the further catabolism of phosphoglycerides in those subceilular membranes that can be
reached by the enzyme and in this way. to function in preserving membranc ~'Itegrity. A
rigorous proof that these hypotheses airily to the in vivo situation could only be given .~f
specific inhil:itors were ai'a~able that could be used in vivo to establish whether inhibition
of the lysophospholipase would result in accumulation of lysophospholipids, and to what
extent, in the individual subeellular i~ractiions. The availability of such inhibitors would
also allow an investigation of wheth~,'r the anabolie and catabolic parts of turnover are
coupled. In the absence of such inhi!bitors, results of in vitro experiments have to be
extrapolated with caution to the in ~¢i~'osituation. It is interesting to note that in vitro
!253254], the velocity of lysophosphatio:ylcholine hydrolysis ~ncrea~d almost linearly
with the mole percentage of lysopho~;phatidylcholme in the membranes. This would !end
~ suggest that the cytosolic lysophospholipase acts under intracellular conditions, i.e.,
233
with lysophosphati~rlylcholine concentrauons in the subceUular m e m b ~ e s of about I%
of the total phospholipids [255 ], far below its r~,aximal ~tivity. In thB ~ , an ~ r e ~ e d
concentration of ly:~ophosphatidylcholine in su(~ membranes may lead to e ~
activity of the lysophospholipase in an attemp': to restore low lysophosphatklyIcholine ~eve[~.
The role of lysosomal phospholipases (Section II) in phosphoilpid catabo~m has
recently been reviewed extensively [256] It is interesting to note that when the phosphoIlpids in isolated microsomal membranes were subjected to hydroly~ by h ~
extracts, the above-mentioned pathways for phosphofipid degradatkm were confirmed.
Catabolism of phosphatidylcholine and phosphatidylethanolamine procee~d v~ a
dea~:ylation pathway with some accumulation of the respective lysophospho~p~, while
phosphatidylinositol was mainly degraded by the phospholipase C pathway [257]. It
remains to be determined then whether the activity of intraceIlular p h o ~ p ~ e s A~
and A2 is sufficien~t to account for the half-lives of d i s ~ p ~
given in Table VII. The
enzymatic activities of the membrane-bound phospholipases discussed in Section II v~W
roughly from 1 to I0 nmol substrate hydrolyzed/min per mg protein. Mo~t ~ver membranes contain about 600 nmol lipid phosphorus per mg protein, so that ,half of ~ e membrane's phosphoglycerides can be degraded in a period of 30-300 n~a. Since the bulk of
liver phospholipids occurs in membrane structures, even the most conservative estimate
shows that the activity of the intracellular phospholipases is sufficient to account for the
half-lives in Tabh~ VII. A valuable criti,:~tm against such calculations is that ~he specific
activity of the membrane-bound enzymes was determined, in most case~, with saturat~
sub~trate concentrations. It seems more appropriat~ to use hydrolysis rates d e t ~
with endogenous substrates. Using this approach, Bj~rnstad [258] observed in rat
microsom~s a dBe.ppearance of 10 and 50%, respectively, for pho~obatidy~otme ~ d
phosphatidylethacolamine per h. The corresponding hydrolysis rates in m i t ~
were about 20 and 50%, respectively ~52]. Also, these values are not in con~'ct with the
conclusion that l~e intraccUular phosph(dipases A possess sufficient hydrolytic aclivity
towards membranc-bound substrates to allow the turnover rates obse~ed in vivo.
VC Synthesis of molecular species
The phosphoglycerides of any cell comprise a heterogeneous n~xture of ~
molecular species due to the specific combination of saturated and um~turated f~t~~
acids. There is now a large body of evidence (reviewed extensively in Ref. 259) to in,~cate that not all molecular species are synthesized de novo. In this respect, the stater~m
in Section VB that phosphoilpids are metabolic and-products to suppo~ membrane structure and function appears to be an oversimplification. The de novo synthesis ofphosp~tidylcholine in rat liver produces prirrtarily monoenoic and dienoic species (e~., see Refs.
260 and 261). Sim~afly, the specificky of the enzymes involved in the de novo s y ~
of phosphatidyl,~thanolamine is such that mainly dienoi¢ and hexaen~ species are
duced [260-26.2]. Both phog~hatid~lcholine and p h m p h a t i d y t e ' ~
are kJtown
to contain considerable ~mounts of tetraenoic species, with a saturated fatty acid at the
sn-l- and arachidonate at the m-2-p*~sition. Cunent opinion takes the view that ~r~chidonate is mainly introduced by remodelling of de n o v o - ~ n ~
species via a ~ - y b t tion-reacylation me~anism, l~efly, this view is based on the foSowing lines of ev~mce.
Arachidonate tmds to be excluded in the synthe~ of ~ t e
[263,269]; the
phospho0yce~Jes coar,~n much higher percentages of arad,Adonate species t h ~ the~ ira.
mediate precursor, I ~2.d/glycerides [264], and a selective utili~tion of ~ ¢ e ~
spe~es
234
containing arachidonate by either choline phosphotransferase or ethanolamine phospho~ransferase has been exluded [262,265]. The replacement of oleate and linoleate in de
hove-synthesized monoenoic and dienoi¢ phosphoglyceride species, respectively, by
,'lrackidona~e through a deacylation-reacylation mechanism requires that at least part of
the monoenoic and dienoic species cannot be considered as metabolic end-prnducts. In
this view, one would expect these species to show a higher turnover than the accumulating species with arachidonate. This has indeed been o~served for both phosphoglycerides
in in vivo experiments [233,261,266]. The half-lives of phosphatidylche.line monoenoic
and dienoic species with palmitate as saturated acid were in the order of 1 - 2 h. The
arachidonate species, on file contrary, showed half-lives of 4-21 h, depending on the
saturated fatty acid present, it was noticed in all these studies that molecular species with
palmitate turned over much more rapidly t h ~ species containing stearate. These and
other observations [272] support the notion that also stearate is to a large extent incorporated via a deacylation-reacylation cycle.
Again, such remodelling mechanisms imply the action of phospholipases At and A2 on
de nero-synthesized species to produce the 2-~cyl and l-acyl :lysophospholipids as substrates for specific acyl-CoA : lysophospholipid acyltransferases. The presence of these
phospholipases in liver has been amply discussed (Section II) and, indeed, the lysophosphaudylcholine fraction from rat liver has been demonstrated to consist of an almost
equimolar mixture of the l-acyl and 2-acyl isomers [7]. Very suggestive evidence for the
operation of deacylation-reacylation cycles in vivo was obtained in dietary studies [267].
When essential fatty acid-deficiem rats were vhifted to a corn-oil diet, the rapid fall in
eicosatrienoyl species of phosphatidylcholine was balanced by an increase of arachidonoyl
species in an almost perfect mirror-image manner. This suggested the replacement of only
the eicosatfienoic acid at the sn-2-position. The acyl-CoA : l-acyl-sn-glycero-3-phosphocholine acyRnmsferase in total rat liver has hter been shown to acylate intravenously
injected lysophosphatidylcholine preferentially with arachidonate [268]. The enzyme
could be separated from other acyltransferases of ~rat liver microsomes [269] and has
recently been obtained in a solubilized form with a 30-fold increased specific activity
[270]. Both preparations showed a preferential incorporation of arachidonate into the
accepter, l-acyl-sn-glycero-3-phosphozholine.
Additional arguments for the operation of deacylation-reacylation cycles in the biosynthesis of particular phosph:,dpid species were provided by Kanoh and Akesson [271].
These authors prepared rat hepatocytes enriched in dilinoleoyl, dipalmitoyl or dimyristoyl phosphatidylcholine by pulse incubation of isolated cells with [3H]glycerol and 14Clabelled fatty acids. The cells were chased in tracer-free medium to follow the disappearance of the labelled species. The indicated species do not occur to any appreciable extent
in the absence of the added fatty acids, so that their degradation could be studied without slguificant resynthesis" of ~ese species. Al:out 40% of the dilinoleoyl, 70% of the
dipalmitoyl and 30% of the dimyristoyl phosph.atidylcholine were degraded during a 2 h
chase. Data on the positional distribution of the labelled fatty acids during the chase
demonstrated that the degradation of dilinoleoyl phosphatidylcholine and dilinoleoyl
phosphatidylethanolamine proc~g~ded through phospholipase At action. In contrast, the
degradation of dipalmitoyl .phosphatidylcholin,,- could be accounted for by the action
of phospholipase A2. A consi~lerable part of the 2-1inoleoyl and l-palmitoyl glycerophosphocholine produced during the chase was reutiliz¢~ for synthesis of the normal phosphatidylcholine specks with saturated fatty acids at the sn-l-position and unsaturated fatty
acids at the sn-2-posltion. The data clearly show the involvement of phospholipases At
23~
and A2 in mechanisms modulating lhe acy! chain composition of pho~hogtycerides in
:intact cells, alLbeit so far only for rather unnatural species. It is worth noting, however,
that the differential attack of diunsaturated species by phosphoh'pase At and that of
disaturated species by phospholipase Az is in keeping with the product~eu of l y ~ h ~ pholipid isomers that, through their specific acylation with :aturated or un~turated fat~.y
,acids by acyitransferas0s, will yield lhe naturally occurring phosphog]ycefide spec/e~.
The involvement of phospholipases At and Az in the ,:em~lellingof diuns~turated
disaturated phosphatidylcholines ~as suggested further by the f'mding that the degra~tion of t]-ese species was inhibited up to 95% when the chase was performed in the p~esonce of i0 mM tetracaine [271]. However, ~he effect of local anesthetics on membr~ebound phospholipases is complex, with stimulation at low and inhibition at high concentrations (compare subsection IVF). Since the intracellular localization and concentration
,af tetracainc is unknown, its mode of action is difficult to assess. A more specific ~ b i tar for intraceUular phospholipa~;es would be highly desirable. Potentially, p - b r o r ~
phenacyl bromide could serve thi~ function for phospholipase A2, although its effect in
:intact cells laas not been studied systematically. In vitro experiments, however, demonstrated the usefulness of this inhibitor in studies on fatty acid elongation in brain refitoehondria and microsomes [273]. Both fractions produced a C22 : 4 fatty zcid as an dongation prod~act of ¢ndogeno,Js asachidonate. The first step in this elongation process is
thought to be the release of arachidonate from the sn-2-position ofmembr,~ne p h o ~ o glycerides and subsequent utilization cf the released arachidonate by ¢ n . z y ~ of the
elongation system. The elongated fatty acid may be reincorpomted into p h o ~ .
thus providing another example af modulating the fatty acid c o m p o s i t ~ of memb~ne
phospholipids. In line with the above reasoning, it was t'on~ad that p-bromopher~cy[
bromide inhibited strongly the elongation of endogenous fatty acids, wh/~ rio ©ff~ct w ~
seen on the elongation of added fatty acids. These results demonstrate that p-bK~ophenacyl bromide inhibited the phospholipase A2 without having an effect on Lhe
enzymes involved in the elongation process. The hd~ibitor is known to react with zn
active-site histidine in pancreatic [274] and snake venom [275] phosphotipase A~ and its
inhibiting a,.'tion on some other enzymes, therefor©, cannot be excluded.
The incorporation of arachidonate via arachidonoyl-CoA : lysophosphet/pid acyhransferase is not limited to phosphatidylcholine and p h O s p h a t i d y l e t h a n ~ . The [ a b e ~ g
patterns of liver phosphatidylin3sitol species as a function of time after rejection Gf either
phosphate or glycerol were consistent with an active de nero synthesis of the
and dienoi,: phosphatidylinosi~:ols, followed by a deacylation-reacylation cycle for the
introduction of arachidonate [276]. Marked l~eferences for ~ I - C o A
in the
acylation of 1-acyl.sn-glycero-3-phosphoinositolhave been reposed for ~cylt~nsfemses
from rat liver [277] and brain [1278].
An important role is asctib~:d also to deacylation-~eacylationprocesses in the m o d ' ~ tion of de novo-synthesized raonoer:oic and dienoic species of I ~ ' o ~ t i d y ~ c h o l / ~ in
lung to provide the fully saturated, m~sdy dipalmitoyl, phosphatidylch~a¢ species. The
.~att¢~ constitute the major component supporting pulmon~-T suffactant function. F~c~pholipase A~ activities, necess~.ry for the removal of moao~.aoi¢ and ~
acyl c ~
have 0een reported in lung tissue [23,279] ~ a p~f©~e,nc~ for uasat~-ated over sackrated species was recently de,,~ribed [281]. The synthesis of disatm~tted phc~hat~ylcholh~e in lung tissue ,,viii no"~.be discussed h©r~ further since some ~ ' e a f ~ v ~ w s ~re
available [282-284].
236
VD. B,Tcterial phospholipases
The function of bacterial ~hospholipases, with emphasis on b2 coil phospholipases, is
less clear at'the moment [301]. E. coil contains a well characterized phospholipase At
[ 17], located in the outer memblane [ 17,64]. This enzyme hydrolyzes both diacyl phosphoglycerides and l.acyl lysophosphoglycerides [17,285] (compare subsection IliA).
Some phosphol[pase A2 activity has been detected as well in the cell wall. The properties
of the A~ and A~ activities in membranes [285,305] and those of a purified phospholipase from E. coli K-12 [62] strongly suggest that a single enzyme is responsible for the
cleavage of acyl ester bonds at both positions. Genetic evidence is in line with this interpretation [303]. A less well characterized phosph'olipase A ~ the cytoplasm exhibits a
preference for phosphatidylglyce~:ol [293]. In addition to these enzymes, several lysophospholipases have been detected in E. coll. Notably, the inner membrane contained an
enzyme which was twice as active towards 2-acyl lysophosphatidylethanolamine when
compared to the l-acyl isomer 1285]. Using classical protein-purification techniques, a
soluble lysophospholipase was purified about 1500-fold to near homogeneity by Doi and
Nojima [286]. This enzyme hydrolyzed the l.acyl isomer of lysophospbatidylethanolamine at a rate about 3-times fa~ter than that measured for the 2-aeyl isomer. It would
thus seem that the inner membrane and cytosol fraction is equipped with enzymes to
catabolize further lysophospholipids formed initially by either the phospholipase At in
the outer membrane or the cytoplasmic phospholipase. As noted above, the function of
these lipolytic activities is still rather obse~4re. It has been repeatedly reported that no
detectable ~urnover of the major E. coil Fhosphoglyceride, phosphatidylethanolamine,
takes place in exponentially grouting cells [287-290]. Turnover of phosphatidylglycerol
and cardiolipin is easily measurable, but has heen explained by conversion of phosphatid3,1glycerol into cardiolipin anti of the polar headgroups of these phospholipids into a
family of water.soluble oligosaccharides contain.lng glycerophosphate moieties [291,292].
Phospholipases A have not been implicated, in these conversions. These data would seem
to exclude a role for the phospholipases A in normally growing cells.
In their original article on the, purification of the membrane-bound phospholipase At,
Scandella and Kornberg [17] have considered a number of other possible functions for
this enzyme. Among these is that the enzyme may be responsible for phospholipid break.
down and subsequent changes ill membrane integrity following phage infection. Indeed,
when cells were infected with bacteriophage ]'4, radioactive lysophosphatidylethanolamine accumulated in parallel with loss of radioactivity from phosphatidylethanolan~ne.
However, lysophospholipid production could equally well be elicited by osmotic lysis of
cells in the absence of phage. ~hi!e the~ experiments clearly show that the latent phos.
pholipase can be activated to degrade the phospholipids in its own membrane, they also
demonstrate that phage infection is not the only membrane.perturbing mechanism that
leads to activation of the phospholipase. A variety of other treatments, including additior~
of colicins [294] and chemicals such as chloramphenicol, penicillin, toluene and formaldehyde [294], heat treatment [17,295] and spheroplut formation [295] have been
shown to induce phospholipid hydroly.sis in E. coll. Nevertheless, seversl authors have
attempted to inves~'.gate whether phage.induced lysis is preceded by phospholipid hydrolysis. It is well known that lysozym¢ is produced after infection of E. coli with bac.
teriophage T4 and that mutants of the phage which lack the lysozyme gone fail to produce lysis of the ho:~t cell. It appeared an attractive hypothesis that mombrane perturbation was required in ,~rder to permit lysozyme to reach the host peptidoglycan substrate
237
~nd that formation of free f~tty acids and lysophosphotipids by p h o ~ o l i ~ u e s constitute,~ such a perturbation mechanism. Working along these lines, Cron~a ~'ld Wutff
[296,2!~7] and Bradley and Astrachan [298] demonstrate1 that phosphoti~i~ hydrolys~
can substantiaUy precede lysis ir~duced by infection with 1"4 phage or ?hug~ m u l L % On
the other hand, Josslin [299] arrived at the conclusion that phage infect~',n ~/d not necessarily enhance the rate of ph,~sphoiipid hydrolysis de~.ected in uninfec~¢~ cdL~. ~ :
findings were reported by Bradl,.'y and Astrachan [29F.| when wild-tyT~e~,~lge was used.
It was concliuded that phospholipid hydrolysis played no essential rob:/~ ~e phage ~fecycle and that phospholipid hy~rolysi~ was the consequence rather t ~ n th" cause of the
lysis proces,.; [298,299]. Such a conclusion has also been drawn with ;egat;i to ~ earb"
effects of colicins. Increases in phospholipase A activity were found to occur weU after
coldcin had caused changes in erLvelope structure and had exerted its complete effect on
intracollulal ATP and potassium level~ [300].
A :more direct approach appli,.~dby Doi et al. [293] to elucidate the fuactk~n ofF.. coli
phosphnlipases has largely confirmed these conclusions. On the b a ~ of t~'Leiri n i t ~
studies, these authors [293] distinguished two phospholipase A activitie~ in E. co//. A
detergent-resistant phospholipas~ A (dr-phosphofipase A), stable or even activated in
presence of detergents and organic solvents, was located in the part:iculate fraction and is
now known to be present in th,'. outer membrane. The enzyme is stable or activaled by
heat treatment at 60°C, requires Ca2÷ and hydrolyzes both p h o s i ~ h a t i d y l e t ~ n ~ e
and phosphatidylglycerol, prefi.'rentiaUy at the sn-l-position. In ~:onuast, a detergentsensitive phospholipase A (ds.phosphofipase A) hydrolyzing only phosphatidy|gb,.'cerol
was detected in the soluble fraction. This enzyme is inactivated by detergents, o ~ g ~ c sob
vents and heat treatment at 60')C and its Caz÷ requirement is much less pronounced. ~ve
preferential position of attack could not be determined since i:o accumulation of lysophosphatidylglycerol accompanied the hydrolysis of phnsphatidylglycerol [293]. By
using an ingeneous selection technique, in which the product f~e fatty ~c~ of the
dr-phospholipase A was used to support growth of a fatty acid anxotroph orE. co//. mutants l a c i n g either dr-phospholipase or both dr- and ds-phospholipase were ~so~ated [302,
30~t]. A mutant deficient in only the ds-phospholipase has been coastrucr.ed from the
dr-ds'-mutant by conjugation [306]. Yhe availability of dr--, ds-- and d~-ds--mutangs has
greatly facilitated studies on the possible functions orE. co/~ phospholipa~. Thu% Lusk
and Park [307] were able to demonstrate that the dr-ds--mutant was as se~tive to
colicin g as the dr÷ds*-parent, despite the fact that colicin did not promote hyd~ob~s~ of
phosphatidylethanolamine in the mutant. Both phage lambda [308] and phage 1"4 [308311 ] induced normal cell lysis in the mutants, while no phospho~/~ hyd~ob~s was
detectable. Although the use of these mutants has clearly establis~hed that ~hospl~pid
hydrolysis is not a prerequisite for colicin action or p~ge4ndu~ed b ' ~ , studies ~
the
mutants have thus far provide:! no hint as to the function ofE. co/i idaog~.o1~p~ses.~
and Nojirna [306] have compared the growth characteristics, lipid composition and pl~spholip~d 'turnover of the three mutants and the parent strain. No marked d/ffe~ence w~s
detected in these parameters. It is relevant to note here that homogenates of the dr-~s'mutant, when assayed with aqueous dispersions of phosphatidylethanolan~,~e, c o n r ~ g ~
less than 1% of the parent-homogenate phospholipase A activity. However, autolysis of
endogenous phospholipids in mutant homogenates amounted to abou~ 5% of the ~ s
obtained with the parent str~n. This could, perhaps, Ladicate that E. co~ contains additional lipolytic enzymes, yet to be ide~tifiod. This possibility was also ~
by
studies of Nelson and Buller [311]. The authors found a phospholipase A assoc~ted w~fi~
238
T4 phage or 1"4 particles. This phage-associated phospholipase A was also found when the
phage was grown in the dr-ds--mutv.nt. Although this could mean that this phospholipa~e
is phage gene specific, the alternative explanation could be that a hitherto undetected
phospholirase in the dr-ds--mutant specifically adheres to the phage particle. The phageassociated phosphollpase showed specificity for phosphatidylglycerol and in this respect
resembled the ds.phospholipase of E. coli. It differed from this enzyme, however, in that
it was activated rather than intaibited by methanol and Triton X-100. When extracts ,of
the dr-ds--mutant were incubated with phosphatidylglycerol in the presence of 2C~
methanol, the residual phospholipase activity amounted to n;ady 50% of the level found
in dr÷ds'-parent extracts. Thea~ data could be in line with the presence of other phosph~olipases in E. coil which were not previously detected due to the assay conditions.
Also, the possibility that the d,r-phospholipase A, which is located in the outer membrane, would play a role in the digestion of phospholipids in the medium was not supported by experimental findirgs. No bacterial growth was observed in the presence of
phospholipid:; as the sole carbon source [306]. What, if any, function the phospho~ipase
A-catalyzed degradation of phospholipids has in the physiology of/~: coil remains to be
established.
~/~. Concluding remarks
It is clear from the preced~ag diseussien that phospholipid degradation by intracellular
phospholipases A and lysophospholipases is an exceedingly complex phenomenon. Despite numerous studies on ph~.sphollpid catabolism, much remains to be learned in the
future. Progress in science ~s l~receded by asking questions. Many of the most elemen~tary
questions to be answered in future studies were put forward in the discussion of the
remits obtained so far and will not be repeated here. It is evident that some basic questions concerning subeellu;ar distribution and functiordng of particular phospholipases
were treated superP,cially, if at all. For example, the effect of lysophosphollpids on membrane structure and activ~.ty of membrane-bound enzymes was rarely discussed. In this
regard, the reader is referred to a recent review in this series [312]. Nevertheless, it is
hoped that the available ~:vidence compiled in the foregoing pages will be of help in the
design of experiments to un.ntvel further the function and regulation of intracellular phospholipases. This undoubtedly will be accomplished by two lines of research. On the one
hand, to understand the molecular basis of phospholipase action, it will be necessaly to
achieve solubilization of thes: enzymes from membranes followed by the~ complete purification with maintenance of enzymatic activity. Currently, in a limited number of cases,
this goal has been reached (Section !!1). On the; other hand, continued research is necessary on the factors which govern enzymatic activity of phospholipases in their natural
environment, especially in membrane-bound form. Hopefully, in the near future, th~ gap
between these two approaches can be bridged by the ,;[ev.elepment of reconstituted systems of purified enzymes in model membranes. The availability of such systems will allow
a systematic approach to ~e study of the physico.chemic~ factors which govev~ the
action of membrane-bound, phospholipases.
Acknowledgements
The author is much indebted to Dr. R.C. Franso~, Richmond, U.S.A., for critically
reading and correcting the manuscript. I like to thank Dr. J.A. IJndgren, Stockholm,
Sweden, for making unpublished results available to me (subsection WD).
239
References
1
2
3
4
Wittcoff, H. (1951) The Pho~;phatides, pp. 99-109, Reinhold, New York
Van Deepen, L.L.M. and de Haas, G.H. (1963) Biochim. Biophys. Acta 70, 538-553
Van Deenen, L.L.M. and de !~aas, G.H. (1966) Annu. Rev. Biochem. 35,157-194
Kate,s, M. (1960) in Lipid Metabolism (Bloch, K., ed.), pp. 1[~1- 196, John W~'~y ~rut Sons, Ne~
York
5 Franciol~, M. (1934) Fermer~tforschung 14,241-249
6 LIoveras, J., Douste-Blazy, L. and Valdigui~, P. (1963) C.R. A ~ d . ScL 256. 1861-1862
7 Van den Bosch, H. and van Deenen, LL.M. (1965) Biochim. I::iophys. Ac~ 106, 326-337
8 Bro~kerhoff, tl. and Jensen, R.G. (1974) Lipolytic Enzymes, pp. 197-207. Ac~clemic Pte~. Ne~~
York
9 Verger, R. and de tlaas, G.tt. (1976) Annu. Rev. Biophys. Bioeng. 5, 77-117
10 Slo~tboom, A.J., van Dam-Mieras, M.C.E., Jansen, E.H.J.M., Pattus, F., Verheij. H.M. ~nd ¢~e ~ .
G.H. (1978) in Enzymes c.f Lipid Metabolism (Gatt, S., F:reysz, L. and M~mie[, P., ed~.), pp.
137-152, Plenum Press, New York
! 1 Dij~cst,a, B,W,, Drenth, J., l~,alk, K.H. and Vandermaelen, P.J. (1978) J. Mo]. Biol. 124, 5 3 - 5 9
12 Heinrikson, R.L, Krueger, E.T. and Keim, P.S. (1977) J. Biol. Chem. 252, 4913-492!
13 Va:a den Bosch, H., van Golde, L.M.G. and van Deenen. L.L.,~L (1972) Rev. P h y S . 6 6 , 1 3 - 1 4 5
14 Van den Bosch, H. (1974) Annu. Rev. Biochem. 4 3 , 2 4 3 - 2 7 T
15 Sh,~n, B.W. and Law, J.H. (1979) in The Biochemistry of A:herosclerosis (Scanu. A.M., eeL), ~p.
275-291, M. Dekker, Inc., New York
16 Van den Bosch, H. and A, :sman, A.J. (1979) Agents Action~ 9, 382-?89
17 Scandella, C.J. and Kornberg, A. (1971) Biochemistry 10,4447-4456
1~ Ferber, E., Munder, P.G., Fischer, ,'t. and Gerisch, G. (1970) Eur. ]. Biochem. 14,253-257
19 Kumar, S.S., Millay, R.H. and Bieber, L.L. (1970) Biochem~.try 9, 754-759
20 Bligny, R. and Douce, R. (1978) Biochim. Biophys. Acta 529,419-428
21 Wai~.e,M. and van Deenen, LL.M. (1967) Biochim. Biophys. Acta 137,498-517
22 Paysant, M., Wald, R. and Polonovski, J. (1968) Bull. Soe. Chim. Biol. 50, 1445-1453
23 Garcia, A., Newkitk, J.D. and Mavis, R.D. (1975) Biochenl. Biophys. Res. C~'nmun. 64, 12813.~,
24 Woelk, H. and PorceUati, G. (1973) Hopp~.-Seylers's Z. Physiol. Chem. 354, '30- i~O
25 Weglicki, W.B., Waite, M., Sisson. P. and Shohet, S.B. (1971) Biochim. Biophys. Ac~ 23!. 5t2-519
26 Swartz, J~G. and Mitchell, J.E. (1973) Biochemistry 12, 5273-5278
27 Stair|h, A.D. and Winkler, H. (1968) Biochem. J. 108, 867-874
28 Gallai-Hatchard, J.J. and Thompson, R.H.S. (1965) Bioeh~. Biophys. Acta 98, !28-136
29 Wassef, M.K., Lin, Y.N. and Horowitz, M I. (1978) Biochin:,. Biophys. A c ~ 528. 318-330
30 Faysant, M., Delbanffe, D., Ward, R. and Polonovski, J. (1967) Bull. Soc. Ch.~. Bk~L4 9 , 1 6 9 ~76
31 Trugnan, G., Bereziat, G , Manier, M.C. attd Poionovski, J. (1979) Biochh-~. Bk~hys. A c ~ 573.
61-72
32 Eisbach, P., van den Berg, J.W.O., van den Bosch, H. and van Deenen, E L M . ~l~h~5) Bk'~ch~'~.
Biophys. Acta 106, 338-347
33 Franson, R., Beckerdite, S., Wang, P., Waite, M. and Elsbach, P. (1973) B ~ I ~ . B~phys. A c ~
296,365-373
34 Newkitk, J.D. and Waite, M. (1971) Biochim. Biophys. Acta 225,224-233
35 Nachbaur, J., Colbeau, A. and Vignais, l~.t'~l.(1972) Biocl'~im.Biophys. Ac~a 274, 426--44~
36 'Victoria, EJ., van Golde, L.M.G., Hos~:etler, K.Y., Scherphof, G . L ami '~n Deemm, L.L.M.
(1971) Biochim. Biophys. Acta 239,443-457
37 Van Golde, L.M.G., Fle'tschet, B. and Fleischer, S. (1971) Biochim. Biophy.- Act~ 249, 318-330
38 Newki~k, J.D. and Waite, M. (1973) Biochim. Biophys. Acta 298, 562-576
39 Colatd-Torquebiau, O., Beteziat, G. and l'oionovski, ~. (1975) Biochhrfie 57,1221-1227
40 De Haas, G.H., Sa~da, L. and Roger, J. (1965) Bioctfira. B ~ y s . Ac~a 106,638-640
41 Colbeau, A., Cuanlt, F. and Vignais, P.M. (1974) Biochimie 56, 275-288
42 Waite, M. and Sisson, P. (1973) J. Biol. Chem. 248, 7201.-7206
43 WaRe, M., Sisson, P., Fteude, K,A, and Zieve, F J . (197.';) Bkw.hem. B ~ y s . Res. C ~ m u n . 67,
471-477
24ll
44
45
46
47
48
WaRe, M. and Sissoa,P. (1971~)J. Biol.Chem. 249. 6401-6405
Ehnholm, C., Shaw, W., Greten, H. and Brown, W.V. (1975) J. Biol.Chem. 250, 6756-6761
Colard-Torqueblau,O., Wolf, C. alxdB:rezi~t,G. (1976) Biochlmie 58, 587-592
Lumb, R.H. and Allen, K.F. (1976) Biochim. Biophys. Acta 450,175-184
Van Golde, L.M,G., Raben, J., Batenburg, JJ., Flelscher, B,, Zambrano, F, and Fleischer, S.
(I 974) Biochim. Biophys. Acta 360,179-192
49 Jelsema, C,L. and MorrO,DJ. (1978) J. Biol. Chem. 253, 7960-7971
50 Rossi, C,R., Sartorelli, L,, Tato, L., Baretta, L. and Siliprandi, N. (1965) Biochim. Biophys. Acta
98, 207-209
51 Scherphof, G.L. and van Deeaen, L.L.M. (1965) Blochim. Biophys. Acta 98,204-206
52 Bj~rnstad, P. (1960) J. Lipid Res, 7,612-620
53 Scherphof, G.L., Waits:, M. and van Deenen, L.L,M. (1966) Biochim. Bi,~phys. Acta 125, 4.06409
54 Waite, M. (1969) Biochemistry8, 2536-2542
55 V;aite, M. and Sisson, P. (1971) Biochemistry 10, 2377-2383
56 M¢llors,A. and Tappel, A.L. (1967) J. Lipid Res. 8,479-485
57 Fowler, S. and de Duve,C. (1969) J. Bic,LChem. 244, 471-481
58 Stoffel, W. and Trabe~t, U. (1969) Hoppe-Seyler'sZ. Physiol. Chem. 350, 836-844
•~9 Franson, R., WaRe, M. and [a Via, M. (1971) Biochemistry 10, 1942-1946
60 Suzuki, Y. and Matsumoto, M. (1978) J. Biochem. 84,1411-1422
61 Suzuki, Y. and Matsumoto, ?tL (1978) FEBS Left. 92,173-176
62 Nishijima,M,, Nakaike, S., Tamori, Y. and Nojima, S. (1977) Eur. J. Biochem. 73,115-124
63 White, D.A., Alhright, F.R.. Lennarz, WJ. and Schnaitman, C.A. (1971) Blochim. Biophys. Acta
.249,636-642
64 Bell, R.M., Ma~s, R,D., O~bom, M.J. and Vagelos, P.R. (1971) Biochim. Biophys. Act;t 249,
628-635
'
65 Yamato, I.,Anr~ku, Y. and Hicosawa, K. (I.q75~1. Biochem. 77,705-718
66 McMurray, W.C. and Magee. W.U (1972) Annu. Rev. Biochem. 41,129--160
67 Nishijima,M., Akama~su,J. and Nojima, S. (1974) J. Biol. Chem. 249, 5658-5667
68 .~.aybin,D.M., Bertsch, L.L and Kornbe:g, A. (1972) Biochemistry 11,1754-1760
69 Van den Bosch, H., Aarsm:m, AJ. and van Deenen, L.L.M. (1974) Biochim. Biophys. Acl:a 348,
197-209
70 Kawasaki,N., Sugatani, J. ~nd Saito, K. (1975) J. Biochem. 77,1233-1244
71 Sugatani, J., Kawasaki,N. ~nd Saito, K. (1978) Biochim. Biophys. Acta 529, 29-37
72 Woelk, H., FOrniss,H. and Debuch, H. (1972) Hoppe-Seyler's Z. Physiol. Chem. 353, 1111-1119
73 Rahman. Y.E., Cemy, E.A. and Peraino. C. (1973) Biochim. Biophys. Aata 321,526-535
74 De Haas, G.H., Postema, lq.M., Nieuwenhuizan,W. and van Deenen, L.L.M. (1968) Biochim. Biophys. Acta 159,103-117
75 $ahu, S. and Lynn, W.S. (1977) Biochim, Biophys. Acta 487,354-360
76 Sahu, S. and Lynn, W.S. (] 977) Biochim. Biophys. Acta 489, 307-317
77 Fleer, E.A.M.,Verheij, H.M. and de Haas, G.H. (1978) Eur. J. Biochem. 82, 261-269
78 Franson, R., Dobrov, R., Weiss,J., Elsbaeh, P. and Weglicki,W.B. (1978) J. Lipid Res. 19, 18-23
7q Kramer, R.M., WOtluich, C., Bollier. C., Allegrini, P.R. and Zahler, P. (1978) Biochim. F;iophys.
Acta 507, 381-394
80 Rock, C.O. and Snyder, F. (1975) J. Biol. Chem. 250, 6564.-6566
81 Jimeno-Abenda~o J. and gahlcr, P. (1979) Biochim. Biophys. Acta 573,266-275
82 Zwaal, R.F.A., Fl~ckigtr, R., Moscr,S. and Zahlet, P. (197,1) Biochim. Biophy.s.Acta 373,416424
83 Bomchov, H., Zahle~, P.r,Wilbrandt, W. and ShinRzky, M. (I~77) Blochlm. Biophys. Acta 470,
382-388
84 Frei, E. and Zah~z, P. (1'~79) Biochim. l)iophys. Acta 550, 450-463
85 KarmaS, R. and Koizun~, K. (1979) Biochim, Biophys. Acta 556, 42~-433
86 Ro~nmn, J.E. al~ Lemud, J. (1977) Science 195,743-753
87 Op den IOunp,J.A.F. (1979) Annu. Rev. Biochem.48, 47-71
88 De Haas, G.H., Postcma, N.M., Nieuwonhuizon,W. and van Deenen, L.L.M. (1968) Biocldm. Biophys. Acta 159, 118-129
89 Ottolenghi, A. (1967) Lipids 2, 303-307
7AI
90 ]?aysant, M., Delbanf~, D., Wald, R. and Polonovsld, J. (1967) Bull. Soc. CF./m. BIOL49, Z69176
91 iPaysant, M, Bitran, M,, Wald, R. and Polonovski, J. (1~70) B ~ . S~c. Ch/m. Biol. 52, 12571269
92 Paysant, M., Bitran, M., Etienne, J. and PolonoVski,J. (1969) Bu||. Soc. Chim. Biol. 51,863873
93 Paysant, M., Bitran, M. and Polonov~i, J. (1969) C.R. Acad. ScL 269. 93-95
94 Ducheme, MJ., Etienne, J., Grtiber, A. ~ad Polonovski, J. (1972) Biocldmi¢ 54, 257-260
95 Torquebiau-Colard,O., Etienne, J. and Polonovski,J. (1972) C.R. Acad. Sd. 275, 2775-2777
96 Sundaram, G.S., Shakir, K.M.M., Barnes, G. and Margolih S. (1978) L Bi~. Chem. 253, 77037710
97 M~:lin,B., Maximilian, It., Friedlander, G., Etiemle, J. and Ale|rider, ~,.G. (1977) Biockim. ~ phys. Acta 486, 590-594
98 Vogt, W., Suzuki, T. and BabiBi,S. (1966) in Memoirsof the Soci~w for E~oc~ao~l~y (l[r~k-~%
V.R. and Fitzpatrick, RJ., eds.), pp. 13~-142, Universit~fPress. C a ~ r ~
99 Pickett, W.C, Jesse, R.L. and Cohen, P. ,11976) Biochem.J. 160,405-408
100 Hong, S.L., Polsky-Cynkin,R. and Levine, L. (1976) J. BioLChem. 251,776-780
101 Ryan, W.L., Short, N.A. and Curtis, G.L (1975) ttoc. Sec. Exp. Bio! Med. 150, 699-702
102 Lacombe, M.L., Stengel, D. and Hanoun¢,J. (1977) FEBS LetL 77, !59-163
103 Anderson, W.B., Jaworski, CJ. and Vlahakis,f;. (1978) J. ~i<,~.Chem. 253, 2921-2926
104 Laskowski,M., Jr. and Sealock, R.W. (1971) ~ The Ena]~nte~(Boyer, P.D., ed.), VoL 3~ pp. 375473, Academic Press, Hew York
105 Kent, C. and Lannarz, WJ. (1972) Proc. Haft. Aced. Sci. U.S.A. 69, 2793-2797
106 Kra~, S.S. and Lennarz, WJ. (1975) J. Biol. Cb-.,n. 250, 2813-2822
107 Kennedy, M,B. and Lennarz, W.J. (t979) $. Biol. Chem. 254,1080-1089
108 Kunze, H. and Vogt, W. (1971) Ara-,.N.Y. Acad. Sci. 180, |23-125
109 Samuelsson, B. (1972) Fed. Proc. 3~, 1442-1450
110 Hong, S.L. and Levine, L. (1976) lhoc. Natl..A:ad. ScL U.S.A. 73,1730--1734
111 Vane, J.R. (1971) Natuie 231,232--235
112 Gryglewski, R.J., Panc~enko, B., Korbnt, R., Grodzin~, L. and Ocetkiewicz, A. (1975)P~Lqa,$andins 10, 343-355
113 Blackweil,G.J., Flow,:~:,RJ., Nijkamp, F.P. and Vane, I.R. (1978) B~. J. PharmacY. 62, 79-89
11,1 Danon, A. and Assoultae, G. (1978) Nature 273,552-554
115 Flower, R.J. and Blackwell,GJ. (1979) Nature 278, 456---~.59
116 Groot, P.H.E., Oerlem;ms,M.C.and Schule, L.M. (1978) Biec.l-dm.Biophys. Acta 530. 9 1 - ~
117 Muntz, H.G., Matsuoka, N. and Jackson, R.L. (1979) l~iochem.B ~ y s . Res. Coa~at~a. ~ . x~21
118 (?hung, J. and Scanu, A.M. (1977) J. Biol. Che~.. 252, 4204--4209
119 Fielding, C.J., Shore, V.G. and Fiel~fing, P.E. (1972) Biochem. Biot~ys. Res. C o ~ .
46.
1493-1498
120 ArCh,L., Jones, S. an,! Fielding,CJ. (1978) J. Biol.Chem. 253, 7220-7226
121 Fischer, G. and Jatzkcwitz, H. (1978) Biochim. Biophys. Acta 528, 69-76
122 Fischer, G. and Jatzk~:witz, H. (1977) Biochim. Biophys. Acta 481,561-572
• 123 Ho, M.W. (1975) FE~S Lett. 53, 243-247
1"!4 Ho, M.W. and O'Brien, J.S. (1971) Prec. Haft. Acad. ScL U.S.A. 68, 2810-2813
125 Hijkamp, F.P., Flower, RJ., Moncada, S. and Vane, J.R. (1976) Natt~re263, 479-482
126 Piper, PJ, and Vane, J.R. (1969) Nature 223, 29-35
127 Samuebson, B. (197'7) in BiochemicalAspects of Pmsta~landm~aa~t T~tx~mb~,xaaes( ~ , g s ¢ ~ ,
N. and Fried, J., ¢ds.), pp. 113-154, Academic Pres~ New Ym'k.
128 Moncada, S. and Vane, J.R. (1977) in BiochemicalAspects o f ~ ~
a ~ "I'~mbo~x~es
(Kh~,
N. and Fried, J., eds.), pp. 155-177, AcademicPt'ea, New yo~rk
129. Levine, L., Pong, S.S., Hon~, S.L. and Tam, S. (1977) in ~
Aspects o t ' ~
and Thmmboxanes (Kharasch, N. and Fried, J., eds.), pp. 15-38, AcademicPre~ New Y~k
130 Pon8,S.S,, Hong, S.l,, and Levine, L. (1977) J. Biol,Chem, 252,1408-1413
131 Kunze, H,, Bohn, E, and Vogt, W, (1~74) Biochim, Biophys. Acta 360, 260-269
132 De~ksen,A. and Cohen, P. (1975) J. Biol. Chem. 250, 9342-9347
133 Babcock, D.F., Chen, JJ., Yip, B.P. and Lardy, H.A. (1979) J. BioLChem. 254, 8117--8120
242
134
135
136
137
White, J.G., Rao, G.H.R. and Gerrard, J,M. (1974) Am. J. Pathol. 77,135-150
Pickett, W.C., Jesse, R.|.. and Cohen, P. (1977) Biochlm. Biophys. Acta 486,209-213
Rittenhou~-Simmons, S. and Deykin, D. (1977) J. Clin. Invest. 60, 495-498
Knapp, HR., Oelz, O., Roberts, L.J., Sweetman, B.J., Oates, J.A. and Reed, P.W. (1977) Proc.
Natl. AcacLSci. U.S.A. 74,4251-4255
138 Wong, P.Y.K. ~nd Cheung, W.Y. ti979) Biochem. Biopllys. Re~. Commnn. 90, 473-48~)
139 Erman, A. and Raz, A. (i979) Biochem. J. 182,821-825
140 Minkes, M., Stanford. N., Chi, M.M., Roth, GJ., Raz, A, Needleman, P. and Majerus, P.W.
(1977) J. Clin. Invest. 59, 449-454
141 Malmsten, C., Gransts6m, E. and Samuelsson, B. (1976) Biochem. Bic,phys. Res. Co~nmun. 68,
569-5715
142 Lapetinzt,E.G., SchmJtges,C.J., Chandrabose, K. and Cuatreca~as, P. (1977) Biocher~. Biophys.
Res. Comnlun. 76, 82~-835
143 Lapetina, E.G., Chandrabose, K.A. and CuaUecasas, P. (1978) Proc. Natl. Acad. Sci. U.S.A. 75.
818 -822
144 Rittenhouse-Simmon.,;,S. and Deykin, D. (1978) Biochim. Biophys. Acta 543,409-422
145 K~ser-Glanzmann,R., Jak~bov~, M., George, J.N. and L(Ischer, E.F. q1977) Biochim. Biophys.
Acta 466,429-440
146 Lind~en, J.A., Claes..;on, H.E., gindahl, H. and HammarstrOm, S. (19~r9) FEB$ Lett. 98, 247-25O
147 Samuelsson, B., Goldyne, M., Granstr6m, E., Hamberg, M., Hammarstc6m, S. and lVlalmsten,C.
(1978) Annu. Rev. B~:ochem.47,997-1029
148 Haye, B., Champion, S. and Jacquemin, C. (I 974) FEBS Lett. 41, 89-93
149 hnre, S. (1972) Acta Biochim. Biophys. Acad. Sci. flung. 7, 24"/-249
150 Chiappe de Cingolani,G.E, van den Bosch, H. and van Deenen, L.L.M. (1972) Biochim. Biophys.
Acta 260, 387-392
151 V~,nden Bosch, H. and van den Besselaar, A.M.H.P. (1978) in Advances in Prost;lglandin and
Thromboxane Rese~xch (GaUi, C., GaBi, G. and Porcellati, G., eds.), Vol. 3, pp. 69-75, Raven
P.:ess,New York
152 l~eferencedeleted
153 Lindgren, J.A., Cla~.'sson,H.E. and Hammarstr~m, S. (1978) in Advances in Prostaglandin and
Thromboxane Rese~trch(GalU, C., Galli, G. and Porceliati, G., eds.), Vol. 3, pp. 167 -174, Raven
Press, New York
154 Haye,B., Champiot~,S. and Jacquemin, C. (1973) FEBS Lett. 30, 253-260
155 Spaulding, S.W. and Burrow, G.N. (1975) Endocrinology 96,1018-1021
156 Haye, B. and Jacqw~min,C. (1977) Biochim. Biophys. Acta 487,231-242
15"I Rillema, J.A. and ~'ild~ E.A. (1977) Endocrinology lf~, 1219-1222
158 Feinstein, M.B., Be,:ker, E.L. and Fraser, C. (1977) i~:ostaglandins14,1075-1093
159 Hsueh, W., lsakson. P.C. and Needl~man,P. (1977) Prostaglandins 13,1073-1091
160 Schwartzn'.an,M. and Raz, A. (1979) Biochim. Biophys. Act~ 472, 363-369
161 Zusman, R.M. and Keiser, H.R. (1977) J. Biol. Chef,1.252, 21)69-2071
162 Isakson, P.C., Raz, A., Hsueh, W. and Needlemar,, P. (1978) in Advances in l~os,taglandinand
Thromboxane Research (GalE, C., GaHi, G. and l~rcellati, G., eds.), Vol. 3, pp. 133 -120, Raven
Press, New York
163 Wise,L.S. and Jungas, R.L. (1978) L Biol. Chem. 253, 2624-2627
164 Okuda, H., Saito, Y., Matsuoka, N., Takeda, E. and KumN~al, A. (1978) J. Bioch~:m. 83, 887-892
165 McCiff, J.C., Terragno, N;A., Malik, K.U. and Lon~ro, AJ. (1972) Circ. Rus. 31, 36-43
166 Danun, A., CI~[~, L.C.T., Sweetman, B.J,, Nle.s,A.S. and Oates, J.A. (1975) ~io~in~. Biophys.
Acta 388, 71-83
167 Alexander, W. and Gimbrone, M.A. (1976) Proc. Natl. Acad. ~t. U.S.A. 73,1617-i' ??~n
168 Needleman, P., Key, $.L., Denny, S.E., It~kmn, P.C. and Marshall,G.R. (1975) l~oc. Natl. Acad.
Sci. U.S.A. 72, 2060:-2063
169 Varpftig, B.B. and Dan Hal, N. (197~) J, Ph,am. Pluilmaeei. 24,159~!61
170 Hot,g, S.L. and Levine, L. (1976) J. Biol.Chem. 251, $a14-$816
171 lsakmn, P.C., Raz, A. and Needleman, P. (1976) l ~ o a t a l ~ s 12, 739-748
172 Gilmore, N., Vain;, J.R. and Wylllc,J.H. (1968) Nature 218,1135-1140
243
IL73 Laychock, S.G., Warner, W. and Rut, in, P.R. (1977) E ~ o g y
100, 74-81
'[74 Burke, G., Chang, L.L. and Szabo, M. (1973) Sciee~e 180, 872.-874
[ 75 Bauminger, S. and L~dner, H.R. (1975) Prostag[a~lins 9, 737~-751
|76 Flower, R. (1978) in Advances in prostaglandin and Tl~,~,nb~xat~ Reu~arch ( G ~ i , C . , G ~ . G .
and PorceHati, G., eds.), Vol. 3, pp. 105-112, Raven Press, New York
177 Gorban, A.M.S. aml Boyd, G.S. (19'77) FEBS Lett. 79, 54-58
178 Beckett, G.J. and Boyd, G.S. (1977) Eur. J. Biochem. 72, 223-233
179 Wailat, S. and Kunan, W.H. (1976) Hoppe Seyler's Z. Physi~L Chem. 357,949-960
180 Vaughan, M., Berges, J.E. and Steinberg. D. (1964) J. Biol. Chem. 239,401-409
181 Huttunen, J.K. and Steinberg, D. (1971) Biochim. Biophytk A¢ta 239, 411-427
182 Pittman, R.C., Khoo, J.C. and Steinberg, D. (1975) $. ffK~l.Chem. 250, 4505-45II
! 83 Khoo, J.C., Steinberg, D., Huang, J.J. and Vagelos, P.R. (1976j J. Biol. Chem. 251. 2882-2890
184 Severson, D.L., Kl:,oo, J.C. and Steinberg, D. (1977) J. BioL Chem. 252,1484-1489
185 Ziboh, V.A. and Lord, J.T. (1979) Biochem. J. 184, 283-290
186 Vogt, W. (1978) in Advances in Prostaglandin and Thrombox._~¢ Research (Ga~, C.. C~ili, G.
and Porcellati, G., eds.), Vol. 3, pp. 89-95, Raven press, New Yoxk
187 Chapman, D., G6mez-Fexnfmdez, J.C. and G6ni, F.M. (1979) FEB$ Lett. 98, 211-223
188 Sandermann, H. (1978) Biochim. Biophys. Acta 515,209-238
189 Sinensky, M., Minneman, K.P. and Molinoff, P.B. (1979) J. Biol. Chem. 254, 9135-9!41
190 Kimelberg, H.K. and Papa.hadjopoulos, D. (1974) J. Bh>l.Chem. 249,1071-1080
191 Papahadjopoulos, D., Jakobson, K., Poste, G. and Shepherd, G. (1975) Biochim. B~phyg Ac|g
394,504-519
192 Kunze, H., Nahas, N., Tmynor, J.R. and Wuri, M. (]976) Biod~n. Biophys. Acta 4 4 1 , 9 3 - | 0 2
193 Scherphof, G.L. ~nd Westenberg, H. (t975) Biochhn. Biophys. Acta 398, 442-451
194 WaRe, M. and Stsson, P. (1972) Biochemistry 11, 3098-3105
195 Dawson, R.M.C ~1969) Methods Enzymol. 14, 633-648
196 Verger, R., Mieras, M.C.E. and de Haas, G.H. (1973) J. Biol. Cl~em. 248, 4023-4034
197 Verger, R., Rietg~h, J., van Dam-Mieras, M.C.E. and de Haas, G.H. (1976) .1. Biol. Chem. 251.
3128-3t33
198 Dawson, R.M.C. (1966) Bio,:hem. J. 98.35c-37c
199 Op den Kamp, I,A.F., Kauerz, M.T. and van Deenen, LL.M. (1975) Biochtm. B ~ y s . Ac~
406,169-177
200 Lands, W.E.M. and Samuelsson, B. (1968) Biochim. Biophys. Acta 164,426-429
201 Vonkeman, H. and van Dorp, D.A. {1968) Biochim. Bgophys. Acta 164. 430-432
202 Kunze, H. (1970) Binchhn. Biophys. Acta 202,180-183
203 Kannagi, R. and Koizum~, K. (1979) Arch. Biochem. Biophys. 196,534-.542
204 Mazcus, AJ. (1978) J. Lipid Res. i9, 793-826
205 Smith, J.B. and Silver, M.J. (1973) Biochem. J. 131,615-618
206 BRIs, K.T., Smith, J.B. and Silvex, M.J. 0976) Biochira. Biophys. Acta 424,303-314
207 Rittenhouse-Simmons, S., Russell, F.A. and Deykin, D. (1977) Bk~chi~a. B/ophys. Ac~ 488,
370-380
208 Schoene, N.W. (1978) in Advances in Prostaglanc~inand Tht-omboxane Research (Ga~, C.,
G. and Porcellati, G., eds.), Vol. 3, pp. 121-126, Raven Press, New York
209 Blackwell, G.J., Duncombe, W.G., Flower, R.J., P,~,'sons, M.F. and Vane, .LR. (1977) Br..L
•
Phatmacol. 59.. 353-366
210 Blackwell, G.J. (1978) in Advances in Prostaglandin and T ~ b o x a n e I~ese~n~ (Ga~, C., ~ .
G. and Porcellati, G., eds.), Vol. 3, pp. 137-142, Raven Press, New York
211 Bills, T.K., Smith, J.B. and Silver, M.J. (1977) J. C]ia. Invest. 6 0 , 1 - 6
212 Rittenhouse-Simmons, S. (1979) J. Clin. Invest. 63,580-587
213 Maxcus, A.J., IJUman, H.L. argl SaBer, L.B. (1969) J. Lipid Res. 10,108-114
214 Chap, HJ., Zwul, R.F.A. and van Deenen, L.L.M, (1977) Biochim. Biophys. A¢ta 467,146-164
215 Pertet, B., Ch~.p, HJ. and Doaste-Blazy, L. (1979) Biochim. Biophys. Acta 556, 434-44~
216 Schick, P.K., Kutica, K.B, and Chacko, G.K, (1976) J. Oin. Invest. 57,1221-1226
217 Jesse, R J , and Cohen, P. (1976) Biochem, J. 158, 283-287
218 Menco, G., Chap, H. and Doust~Blazy, L. (1979) FESS Left. 100, 367-370
219 Boll, R.I.., Kcnnotly, D.A., SUmfotd. N. and Ma~ontg P.W. (1979) Pro¢. Nat1. Acad. ~.~ U.S.A.
76, 3238-32~1
244
220 Mauco, G., Cl',~p,H., Simon, M.F. and Douste-Blazy, L, (1978) Biochimie 60, 653-661
:221 Bry, K., And~;rsson, L.C., Kuusl, T. and Kinnuner~, P,K.J. (1979) Biochlm. Biophys Acta 575,
121-127
222 Fischer, M.C., Enter,man, C., Montgomery, M.L. and Chaikoff, I.L. (1943) J. Biol.CITtem. 150,
47-55
223 Zilve~smi~,D.B., Entmman, C, and Chaikoff, I,L.(1948) J. Biol.Chem, 176,193.-201;
224 Tolbert, M.E. and Okey, R. (1952) J. Biol. Chem. 194, 755-768
225 Bollmnn, J.L., Flock, E.V. and Berkson, J. (1948) Proc. Soc. Exp. Biol. Med. 67, 308--:~17
226 Popj~k, (]. and Muir, H. (1950) Biochem. J. 46,103-113
227 Campbell, ~,~.M.and Kosteriitz, H,W. (1952) Biochim. Biophys. Acta 8,664-679
228 Smith, M.E. and Eng, L.F. (1965) J. An,. Oil Chem. So¢, 42,1013-1018
229 Arvidson. G,A.E. (I¢.~58)Eur. J. Biocbem. 5,415-421
230 Wise, E.Iv[.,Jr. and Elwyn, D. (1965) J. Biol. Chem. 240,1537-1548
231 Jobe, A. (1977) Biochim. Biophys. Acta 489,440-453
232 Jobe, A., Kkkpatricl:, E. and Gluck, L. (1978) J. Biol. Chem. 253, 3810-3816
233 Curstedt, T. (1974) 3iochim. Biophys. Acta 369,196-208
234 Omura, T., Siekevitz, P. and Palade, G,E, (1967) J, Biol. Chore, 242, 2389-2396
235 Bailey, E., Taylor, C,B, and Bartley, W. (1967) Biochem. J. 10q, 1026-1032
236 Bygrave, F.L. (1969) J. Biol. Che,~, 244, 4768-4772
237 Eriksson, L.C. and DaHner,G, (1973) FEBS Lett. 29, 351-354
238 Lee. T.C., Stephens, N., Moehl, A. a~d Snyder, F. (1973) Biochim. Biophys, Acta 291.86-92
239 Warren, L. and Gllci:, M.C. (1968" J. Cell Biol. 37, 729.-746
240 Gallaher, W.R. and ]|lough, H,A. (197S) Arch. Biochem. Biophys. 168, 104-I 14
241 Pastemak, C.A. and Be~e~n, J,J.M. (1970) Biochem. J, 119, 473-480
242 Dawson, R.M.C. (1966) in Essay's in Biochemistry (Campbell, LN. and Greville, G,D,, eds.),
Vol. 2, pp. 69-I 15, Academic Prt:ss, London
243 Michell, R.H. (1975) Biochim. Biophys. Acta 415, 81-147
244 Dawson, R.M.C. (1978) Adv. Exp. Med. Biol. I01,1-13
245 Dobretsov, G.E., Borschevrkaya, T.A., Petrov, V.A. and Vladimirov, Y.A. (1977) FE!~S Lett. 84,
125-128
246 TrewhcUa, M.A. and Collins, F.D. (1973) Biochim. Biophys. Acta 296, 34-50
247 Taki, T. and Kanfer, J.N. (1979) J. Biol. Chem. 25/~,976! -9775
248 Dawson, R.M.C. (I!~55) Biocbem. J. 59, 5 - 8
249 Kewitz, H. and PIeui, O. (1976) Proc. Nail. Acad. Sci. U.S.A. 73, 2181-2185
250 Van den Bosch, H., Amp.man,A.J., Siotboom, A.J. and van Deenen, L,L.M, (1968) Biochim. Binphys, Acta 154, 215-225
251 De Jong, J.G.N., van den Bosch, H., R~-jken,D. and van Deanen, L.L.M. (1974) Biochim. Biophys. Acta 369, 50-63
252 Van den Bosch, H. and de Jo~g, J.G.N. (1975) Biochim. Biophys. Acta 398, 244-~157
253 Van den Besscla~, ~.M.H.P., Verheijen, J.H. and van den Bosch, H. {1976) Bioch!,m. Biophys.
Acta 431,75-85
254 Moonen, H., Trienekcns, P. and ~m den Bosch, H. (1977) Biochim. Biophys. Acta 489, 423-430
255 Poonhuis, B.J.H.M., Yazaki, pJ. ~md Hostetler, K.Y. (1976) J. Lipid. Res. 17,433-,i37
256 Waite, M., Griffin, H.D. and Franson, R. (1976) in Lysosomes in Biology and Patltolosy (Dingle,
J.T. and Dean, R.T., eds.), Vol. 4, pp. 257-305. North-Holland, Amsterdam
257 Rkha~s, D.E., Irvine, R.F. and Dawson, R.~,I.C. (1979) Biochem. J. 182, $99-606
2.~8 Bj~stad, P. (1966) Biochim. BIophys. Acta 116, $00-510
259 Van Golde, L.M.G. and van den I~rgh, S,G. (1977) in Lipid Metabolism in Mammals (Snyder, F.,
cal.), Vol. I, pp. 35-149, Plenum Press, New York
260 Akesson, B., Elovson, J. and Arvidson, G. (1970) Biochim. Biophys. Acta 210,15-27
261 Tsewhella, M.A. and Collins, F.D. (1973) Biochim. Biophys. Acta 296, $1-61
262 Kanoh, H. and Ohno, K. (1975) Biochim. Biophys. Acta 380; 199-207
263 Hill, E.E. and Lairds, W.E.M.(1961l) Biochim. Biophys. Acta 152; 645-;-648
264 Akesson, B., Elovson, J. and Anddson, G. (1970) Biochim. Biophys. Acta 218, 44-:5~
265 De KruUff, B., van Goide, L.M.G. and van l)eenen, L.L.M, (1970) Biochim, Bk,phys, Ac~a 210,
~25-435
266 Sundler, R. and Akesson, B. (1975) Biochem. J. 146, 309-315
245
267 Van Golde, L.M.G., Pieterson, WA. and van Deenen~LL.M. (|~-~8) Biochhn. ~ y ~
Acta 152,
84-95
268 Akino, T., Yamazaki, I. and Abe, M. (1972) Tohoku J. Exp. Meal. !08,133-139
269 YamasMta,S., Nakaya, ~|., Miki, Y. and Numa, S. (19751 Prec. Natl. Acad. Sci. U.S.A. 72, 6 ~ 603
270 Hasagawa-Sasakt,H. and Ohno, K. (19801 Biochi,-a~Biophys. Acta 617, 205-217
271 Kanoh, H. and Akesson, B. (1977) Biochtm. l~iophyr,.Acta 486, 511-523
272 Akes~,.on,B., Arner, A. and Sundler, R. (19761 BioeJltm.Bio~ys. Acta 441,453-464
273 Gan-Elepano, M. and Meacl,J.F. (19781 Biochem. |~iophys.Res. Commun. 83. 247-251
274 Volwerk, J.J., l~eterson, W.A. and de Haas, G.H. (!i974) Biochem~try 13, .~446-1454
275 Roberts, M.F., Deemer, R.A., Mincey, T.C. and Dennis, E.A. (19771 L B~L ~ . 252. 24052410
276 Holub, B3. and Kuksis, A. ~,i 97! ) J. Lipid Res. 12,699-705
277 Holub, B,J. (1976) Lipi~s ~ !, 1-5
278 Baker, R.R. and Thompson, W. (!973) J. Biol. Chem. 248, 7060-7065
279 Ohta, M, Hasagawa, H. and Ohno, K. (19721 Bioc~im. Biophys. Acta 280, 552-558
280 Bonsen, P.P.M., de Haas, G.H., Pietetson, W.A. and van Deenen, LL.M. (1972) ~ c h i m . ~ phys, Acla 270, 364-382
281 Longmore, W.J., Oldenborg, V. and van Golde. L.M.G. (19791 Biochim. Bio~y~. AcUt 572.
452-460
282 Van Golc~e,L.I~.G. (19761 Am. Rev. Respir. Dis. 114,977-1000
283 Frosolono, M.F. (19771 Jn Lipid Metabolism in Mammals (Snydet, F., cd.), VoL 2, pp. | - 3 8 ,
Plenum l~ess, New York
284 Ohno, K., Akh~o, T. and Fujiwara, T. (19781 ir Reviews in Perinatal Medicine ( S c ~ ,
E.M.
and Comfi, E.V., eds.), Vol. 2, pp. 227-318, Ravin Pre~, New York
285 A!bright, F.R., White, D.A. and Lennarz, W.J. (IS~,73)J. Biol. Chem. 248, 3968-3977
286 Doi, O. and N~jima, S. (19751 J. Biol. Chem. 250, 5208-5214
287 Kanfer, 1. and Kennedy, E.P. (1963) J. Biol. Chela. 238, 2919-2922
288 Ames, G.F. (19681 J. Bacteriol. 95,833-843
289 Cronan, LE. (19681 J. Bacterioi. 95, 2054-2{161
290 Kanemasa, Y., Akamatsu, Y. and Nojima, S. (19~,7) Biochim. Biophys. Acl~t 144, 3~2-390
291 Van Go!~e, L.M.G., Schulman, H. and Ke~,'tedy,E.P. (1973) Pm¢'. Nail Acad. Sci. U~S.A. 70,
1368-1372
292 Kennedy, E.P.. Rumley, M.K., Sehulman, H. and van CoMe, L.M.G. (I976) .L BioL~ .
251,
4208-4213
293 Doi, O., Ohki, M. and Nojima, S. (19721 Biochint. Biophys. Acta 260, 244-258
294 Cava~d, D., R~mpini, C., Barbt,, E. ad ?oi~t:u~-~ki,J. (19~8) Bull. Soc. Chi~. Bill 50. 14551471
295 ~atri~ca, P., Beekerdite, S. and Elsbach, P. (19721 Biocl~.~. Biophy~.Act~ 260, 593-600
296 Cronan, I.E. ~ d Wulff, D.L. (19691 Virology 38,241-2,6
297 Cmnan, J.E. and Wulff, D.L. (1971) Vbology 46,977-978
298 Bradley, W.E.C. and Astrachan, L. (19711J. Vtrol. 8,437-445
299 Joutlin, R. (19711 Vkology 44, 94-100
300 Cramer, W.A. and Keenan, T.W. (19741 Biochem.Bioph~~ Res. Co~mtm. 56, 60-67
301 Raetz, C.R.H. (19781 Micmbicl. Rev. 42, 614-659
302 Ohki, M., Doi, O, and Nojima, S. (19721 J. Bactefiol. llf,~ 864-869
303 Abe, M,, Akamoto. N.. DoioO. and Nojima, S. (19741 J. ~tet'ioL I I9, 543-546
304 Nojima, S., Doi, O.~ Akaraoto, N. and Abe, M. (1972) ia Membrane Research (Fc~x,C.F.. ed.),
pp. 135-144, Academic Press. New York
305 Doi, O. and l~;ojima,S. (19731 J. Biochem. 74,667-674
306 Doi, O. and l~ojinuz,S. (19761 J. Biochem. 80,1247-12:~8
307 Lusk, J,E. and Park, M.H. (19751 Biochim. Biophys. Act~,394,129-~34
308 ~ , h a m , ~'., Boi~O. and Nojima, S. (19721 Biochem. ~ l i o l ~ Re~ C~mmua. 46,1434-144~
309 Van de~ Matin, M., Nelson, E.T, and Bul~e~,C.S. (1974) 1. Vii'eL 14,1617-1619
310 Bullet, C.S., van dot Maten, M., Faurot, D. and Nehoa, E 1". (1975) 1. Vkol. 15.1141-I147
311 Nehon, E.T. and Bullet, C.S. (19741, J. Vi~ol. 14, 479-484
312 Weltziea, H.U. (19791 Biochim. Biophys. Acta 559, 259-287
246
313 Aarsman, A.J. end van den Bosch, It. (1979)Biochim. Biophys. Acta 572, 519-530
314 Apitz-Castro, R,J., Mas, M.A,, Cruz, M.R. and Jain, M.K. (1979) Biochem, Biophys. Res, Commun, 91, 63-71
315 Franson, I~.C,, Ei.o,en, D., Jesse, R, and Lanni, C, (1980) Blochem, J. 186,633-638
316 Jesse, R.L. and Franson, R.C. (1979) Biochim. Biophys. Acta 575,467-470
317 Elsbach, P., Weiss, J., Franson, R.C., Beckerdite-Quagliata, S., Schneider, A. and Flalris, L. (1!}79)
J. Biol. Chem. 254,11000-1100~
318 Biilah, M.M., Lapetina, E.G. and Cuatrecasas, P. (1979) Biochem, Biophys. Re~).Commun. 90,
c)~---98