This article was published in an Elsevier journal. The attached copy
is furnished to the author for non-commercial research and
education use, including for instruction at the author’s institution,
sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
Author's personal copy
Evidence for a relationship between bovine erythrocyte lipid
membrane peculiarities and immune pressure from ruminal ciliates
Guadalupe Gimeneza,*, Mo ´nica Florin-Christensenb, Marı ´a L. Belaunzara ´na,
Elvira L.D. Isolaa, Carlos E. Sua ´rezc, Jorge Florin-Christensena,?
aDepartment of Microbiology, Parasitology and Immunology, School of Medicine,
University of Buenos Aires, Paraguay 2155, piso 13, C1121ABG Buenos Aires, Argentina
bCICVyA, INTA-Castelar, Buenos Aires, Argentina
cAnimal Disease Research Unit, US Department of Agriculture, Washington State University, Pullman, WA 99164-6630, USA
Received 23 January 2007; received in revised form 11 April 2007; accepted 3 May 2007
Erythrocytes of bovines and other ruminants have a strikingly anomalous phospholipid composition, with low or absent
phosphatidylcholine (PC) together with high sphingomyelin (SM) content. Here, we report the presence in normal bovine serum of
high levels of anti-phospholipid antibodies of IgM isotype against, PC and the phosphono analogue of phosphatidylethanolamine,
acyl-phosphatidylethanolamine (NAPE), the major components of bovine erythrocytes. In addition, we found that exposure of the
ciliate Tetrahymena thermophila to bovine serum results in rapid lysis, an effect that was inhibited by adsorption of the serum with
SM/AEPL liposomes. Furthermore, incubation with bovine serum had a similar effect on freshly obtained ruminal ciliates, and the
lytic activity was eliminated by pre-adsorption of the serum with SM/PE liposomes.
The ruminant mode of lifewith its concomitant ciliate fauna is hereby linked to the peculiar conformation of bovine erythrocyte
membranes. We propose that the unique phospholipid composition of bovine erythrocytes appears as an evolutionary adaptation to
tolerate the lytic effects of anti-phospholipid antibodies generated against AEPL, a membrane component of the huge mass of
ruminal ciliates, necessary commensals of this group of mammals.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Phospholipids; Antibodies; Erythrocytes; Bovines; Ciliates
In mammals, phosphatidylcholine (PC) is generally
the most abundant erythrocyte membrane phospholipid
ruminants are a particular zoological group in that their
erythrocytes lack PC and possess abundant sphingo-
myelin (SM) (Christie, 1981; Matsumoto and Miwa,
1973). Moreover, ruminant erythrocyte membranes
present a particular derivative of phosphatidylethano-
(NAPE) (Fig. 1) (Christie, 1981; Matsumoto and Miwa,
Veterinary Immunology and Immunopathology 119 (2007) 171–179
Abbreviations: PC, phosphatidylcholine; SM, sphingomyelin;
CL, cardiolipin; PE, phosphatidylethanolamine; AEPL, aminoethyl-
phosphonolipid; NAPE, N-acyl-phosphatidylethanolamine; DOcPC,
* Corresponding author. Tel.: +54 11 59509500x2191;
fax: +54 11 59509577.
E-mail address: email@example.com (G. Gimenez).
?The author Dr. Jorge Florin-Christensen passed away on 10
0165-2427/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
Author's personal copy
1973). The biological explanation of this phenomenon
has remained elusive.
We have recently found that, in addition to the global
peculiarities in phospholipid composition of bovine
erythrocytes, the distribution throughout the lipid
bilayers is also unique. Indeed, the amounts of PE
exposed on their outer surface are 10-fold lower than in
human erythrocytes. Therefore, we have proposed a
model that involves the combined activities of an
endogenous phospholipase A2and an acyl-transferase
located in the inner surface of the erythrocyte
membrane (Florin-Christensen et al., 2001). So far,
the biological significance of these highly unusual
features has not been understood. Also unexplained has
remained the reason why these abnormalities occur
exclusively in ruminant artiodactylans.
The rumen is the largest of the four chambers that
constitute the ruminant stomach. The rumen and the
reticulum, the second chamber, are needed for digestion
of cellulose, and act as fermentation vats. Notably, this
digestive process requires a rich microbial community,
including a dense and complex ciliate fauna (Ogimoto
and Imai, 1981). It has been estimated that an adult cow
digests about 4 kg of ciliates daily (Ogimoto and Imai,
1981). Characteristic of ciliates is their abundance of
phosphonolipids, particularly aminoethylphosphonoli-
pid (AEPL) on their cell surface (Harfoot and Hazle-
wood, 1988). AEPL is the phosphonic analogue of PE
and its distinctive trait is the presence of a direct C–P
bond (Horiguchi and Kandatsu, 1959), which confers
high resistance to enzymatic hydrolysis (Kittredge and
Roberts, 1968; Rosenthal and Pousada, 1968). Because
of its resistance, AEPL synthesized by ciliates can be
absorbed into the blood of bovinesas an intact molecule
and, indeed, this lipid has been detected in bovine blood
(Tamari and Kametaka, 1980). Thus, exposure of ciliate
phosphonolipids to the bovine immune system is a
necessary consequence of ruminant digestion. As these
lipids are structurally different from their mammalian
analogues (Fig. 1), they could be targets for an immune
response. At the same time, their similarities with
common phospholipids could easily result in antibodies
to ciliate lipids that cross-react with their mammalian
host counterparts. We here characterize bovine anti-
phospholipid antibodies and analyze their relation to
bovine erythrocyte membrane architecture.
2. Materials and methods
Egg L-a-phosphatidylcholine (PC), egg sphingo-
myelin (SM), egg L-a-phosphatidylethanolamine (PE),
cardiolipin (CL) and dioctanoylphosphatidylcholine
(DOcPC) were purchased from Avanti Polar Lipids.
Aminoethylphosphonolipid (AEPL) was purified from
Tetrahymena thermophila as described below. N-Acyl-
phosphatidylethanolamine (NAPE) was synthesized as
described by Caramelo et al. (2003).
2.2. Preparation of erythrocyte suspensions
Erythrocyte suspensions were prepared from freshly
drawn blood samples obtained with heparin from: fetal
(n = 3) and adult bovines (Bos taurus) (n = 5) and adult
pigs (n = 3) obtained from INTA-Castelar, Argentina,
healthy human donors (n = 3) and adult mice (Balb-c)
(n = 5) obtained from the Department of Microbiology,
Parasitology and Immunology, School of Medicine,
UBA, Argentina. Erythrocytes were washed four times
in 20 mM Tris–HCl/150 mM NaCl/5.5 mM glucose/
1 mM CaCl2/0.1 mM MgCl2, pH 7.4, discarding the
buffy coat and the top 2-mm layer in each wash, to
minimize leukocyte contamination. They were then
resuspended in Tris-buffered saline (TBS).
2.3. Serum samples
Fetal (n = 3) and adult bovine (n = 5) sera were
obtained from INTA-Castelar, Argentina. Murine sera
(n = 5) were obtained from the Department of Micro-
biology, Parasitology and Immunology, School of
Medicine, UBA, Argentina.
T. thermophila, strain CU399, was obtained from Dr.
Arno Tiedtke, University of Mu ¨nster, FRG. It was
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179172
Fig. 1. Structure of phosphatidylethanolamine (PE) and N-acyl-phosphatidylethanolamine (NAPE) from bovine erythrocytes and aminoethylpho-
sphonolipid (AEPL) from rumen ciliates. Arrows indicate the chemical differences between PE and its two related compounds.
Author's personal copy
axenically grown in 1% tryptone, 0.1% yeast extract,
1% glucose, 0.003% iron citrate (TGYF medium) for
48 h at 30 8C (Florin-Christensen et al., 1986a),
collected by centrifugation and washed twice in
10 mM Tris–HCl,pH7.4.Thepellets were immediately
used for AEPL extraction.
2.5. AEPL purification from T. thermophila
Lipids from cell pellets were extracted according to
Bligh and Dyer (1959). Briefly, the pellets were
solubilized with water/chloroform/methanol (0.9:1:1,
v/v), and the organic phase (lower phase), containing
the total cell lipids, was collected. In order to obtain T.
thermophila AEPL, the lipid fraction was run by
preparative thin layer chromatography (TLC) as
described by Kapoulas (1969) using chloroform/acetic
acid/water (60:36:4, v/v). Lipids were identified by
comparison with authentic standards. AEPL was eluted
from the plates with chloroform/methanol (1:1, v/v).
Quantification was performed by phosphorous deter-
mination according to Ames (1966).
2.6. Hemolytic activity assay
Suspensions (0.5%) of the different erythrocytes in
TBS were incubated for 1 h at 37 8C under various
conditions. Samples were centrifuged and absorbance
at 540 nm was measured in the supernatants. Percen-
tage hemolysis was calculated as the absorbance value
of each sample divided by the absorbance of an
identical erythrocyte sample resuspended in distilled
2.7. Liposome preparation
Liposomes were prepared by mixing 5 mg of each
phospholipid in the case of SM/PE, SM/PC, SM/AEPL
and SM/NAPE liposomes or 10 mg in SM liposomes in
chloroform solution, evaporated under N2and finally,
subjected to high vacuum. SM was added to the
different phospholipids to allow liposome formation in
all cases. SM alonewas used as control. The lipids were
then resuspended in TBS by vigorous vortexing in the
presence of glass beads and then sonicated for 5 min in
Torbeo tip sonifier (Cole Palmer).
2.8. Liposome adsorption of bovine sera
Two volumes of the liposome suspensions (SM, SM/
PE, SM/PC, SM/AEPL and SM/NAPE) were added to
8 volumes of bovine serum. The mixture was then
incubatedfor1 hat37 8Cwithgentlerocking.Afterthis
period, samples were centrifuged at 15,000 ? g for
15 min. Supernates were removed to assay hemolytic
activity. Liposome pellets werewashed three times with
TBS and prepared with sample buffer and dithiothreitol
(DTT) for electrophoretic analysis (SDS-PAGE and
stained with Coomasie blue R250).
2.9. Detection of anti-PL antibodies
(a) ELISA assay: This was performed as described in
the literature (Bordmann et al., 1998). A 100 ml aliquot
of PC, PE, SM, NAPE, AEPL and CL in ethanol at
50 mg/ml was loaded into test wells of Maxisorp
microtiter plates (Nunc). The ethanol was evaporated at
37 8C and blocking was performed with a commercial
blocking system (Roche blocking reagent1) for 2 h at
37 8C.Theplates were then washed fivetimeswithTBS
and incubated with 1:100 serum dilutions for 2 h at
37 8C. The plates were washed again five times with
TBS. Antibodies bound to the wells were detected with
anti-bovine IgM–horse radish peroxidase (HRP) con-
jugate (KPL). After a 2-h incubation at 37 8C, plates
were washed five times with TBS, and incubated with
O-phenylendiamine (OPD, DakoCytomation). After
4 min the absorbance was measured in a Microplate
reader Model 450 (Biorad) at 490 nm. (b) Dot blot
assays were conducted by depositing 5 mg of lipid from
1 mg/ml chloroform solutions on nitrocellulose mem-
branes, which were then blocked and incubated with
ELISA except that HRP-labeled conjugates were
detected using a chemiluminescent substrate (ECL
plus Western Blotting Detection System1, Amersham
Biosciences). Membranes were scanned using Storm1
Gel and Blot Imaging System. (c) Immunoblot analysis
of liposome pellets: An alternative antibody capture
method was used, which involved the incubation of
bovine sera with liposomes made of different mixtures
of phospholipids, as described above. Liposome
pellets were dissolved in Laemmli’s sample buffer
and DTT. After SDS-PAGE, proteins were transferred
to nitrocellulose membranes using a Trans-blot1semi
dry transfer cell (Biorad, California, USA). Mem-
branes were blocked with Phosphate-buffered sal-
ine + 3% skimmed milk for 2 h at room temperature.
IgM was detected, by over night incubation at 4 8C,
with anti-IgM–HRP conjugate 1:500 (v/v) in Phos-
phate-bufferedsaline + 0.1%
skimmed milk. Chemiluminiscence was used to
visualize the reactions and membranes were scanned
as described above.
Tween20 + 1%
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179173
Author's personal copy
2.10. IgM purification
A sample of freshly obtained bovine serum (10 ml)
was dialyzed overnight against water to precipitate the
IgM fraction. The dialyzed serum was centrifuged
(10,000 ? g, 30 min, 4 8C) and the pellet resuspended
in TBS, pH 7.5. IgM was purified using Sepharose 6B
(Sigma). The protein composition of the different
fractions was verified by SDS-PAGE in 10% gels
stained with Coomasie Blue. The fractions were also
analyzed by Immunoblot following the protocol
2.11. Protease treatment
Bromelain (6 U/ml,final conc.,SigmaChemicalCo.)
human erythrocytes in TBS, and along with control
60 min. Erythrocytes were washed three times with
cold TBS and resuspended in either bovine or murine
sera at a 0.5% concentration (v/v). Hemolysis was
measured after 1h of incubation at 37 8C by determina-
tion of absorbance at 540 nm in the supernatants, after
bovine and human erythrocyte ghosts was confirmed by
SDS-PAGE. Erythrocyte ghosts were prepared as
described by Florin-Christensen et al. (2001).
2.12. Incubations with
DOcPC (100 mg/ml final conc.) was incubated with
washed erythrocytes (5%, v/v) in TBS/1 mM disodium
ethylenediaminetetraacetic acid for 30 min at 37 8C
with stirring. Erythrocytes were washed once with the
same buffer and resuspended in: bovine serum, heat-
inactivated and SM/AEPL pre-adsorbed bovine serum
and then assayed for hemolysis as described above.
2.13. Phospholipid analysis of erythrocytes
Lipids from samples containing the same number of
erythrocytes from fetal and adult bovines and humans
were extracted according to Bligh and Dyer (1959). The
solvents were evaporated under nitrogen to a constant
using chloroform/methanol/water (65:35:2.5, v/v) to
separate polar lipids. Lipids were identified by
comparison with authentic standards. Plates were then
dried, sprayed with 10% CuSO4in 8% H3PO4and
charred by exposure to 150 8C for 13 min in order to
visualize the lipids (Baron et al., 1984). The TLC was
scanned in an hp scanjet 2400 apparatus. Each result is
representative of three similar experiments.
2.14. Incubations of ciliates with bovine serum
(a) T. thermophila CU 399 cells grown in TGYF
medium were washed three times in 10 mM Tris–HCl,
pH 7.4, and resuspended in this buffer in a cell density
of 5 ? 105cells/ml. One volume of this suspension was
added to onevolume ofa 1:8 dilution ofheat inactivated
(56 8C, 30 min) bovine serum, or sera adsorbed with
either SM or SM/AEPL liposomes. Aliquots were
removed for microscopic observation. (b) A fresh
sample of bovine ruminal fluid, maintained under
anaerobic conditions throughout the procedure, was
used as source of rumen ciliates which were exposed to
1:4 dilutions of either SM or SM/PE adsorbed bovine
sera in TBS. Incubations were conducted for different
time periods (5, 10, 15, 20, 25, 30 min) and ciliate
motility and cell lysis were microscopically examined.
2.15. Statistical analysis
The statistical significance of the results was
analyzed using one-way ANOVA, followed by Bon-
ferroni’s multiple comparison test (GraphPad Prism 4
Software Inc., San Diego, CA, USA).
3.1. Bovine serum contains anti-phospholipid
Taking into consideration the phospholipid compo-
sition of non-ruminant mammals’ erythrocytes (Zwaal
et al., 1975), we initially tested whether bovine serum
could be naturally hemolytic to these erythrocytes in a
phospholipid-dependent fashion. The results shown in
Fig. 2 demonstrate that all adult bovine sera tested
(n = 5) hemolysed human erythrocytes from different
ABO/Rh blood groups, as well as erythrocytes from
mouse and pig (88 ? 2% and 91 ? 2% percentage
hemolysis, respectively). Hemolysis was dependent on
serum complement (C0), since it was fully suppressed
by heat-inactivation of bovine serum at 56 8C for
30 min, suggesting that it was mediated by complement
fixing antibodies. To determine whether hemolysis was
mediated by anti-phospholipid antibodies, aliquots of
bovine serum were then adsorbed with liposomes,
formed by mixtures of SM with different phospholipids,
and the effect of these treatments on the hemolysis of
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179174
Author's personal copy
human erythrocytes was studied. Adsorption of sera
with SM/PC liposomes largely blocked hemolysis,
while complete inhibition was attained with SM/PE or
SM/AEPL liposomes. On the other hand, identical
treatment with liposomes made of SM only or SM/
NAPE, the distinctive lipids of bovine erythrocytes did
not inhibit hemolysis (Fig. 2A). Incubation with
liposomes did not affect the levels of complement as
shown by the following experiment. If heat-inactivated
bovine serum (as a source of antibodies) was mixed
with SM/PE-adsorbed serum (as a source of C0),
hemolysis was restored and the same result was
obtained supplementing SM/PE-adsorbed serum with
IgM to restore hemolytic activity was, in turn, lost by
pre-adsorption with SM/PE liposomes (Fig. 2B). These
experiments demonstrate the involvement of bovine
IgM reacting with PE, AEPL and PC in the observed
Direct demonstration of the presence of anti-
phospholipid IgM antibodies in bovine serum was
attained using three different approaches: enzyme-
linked immunoassay (ELISA) (Fig. 3A), dot blot
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179175
Fig. 2. Hemolysisofhumanerythrocytesbybovineserumisdependent
on complement, blocked by adsorption with phospholipids and IgM-
serum adsorbed with liposomes composed of PC, SM, SM/PE,
SM/NAPE or SM/AEPL. Hemolysis was recorded spectrophotometri-
1h to normal bovine serum (control), heat inactivated bovine serum,
Values represent the mean ? SD of 5 samples.
Fig. 3. Characterization of anti-phospholipid antibodies in normal
bovine serum by ELISA, dot-blot and immunoblot. (A) Antibodies in
normal bovine serum against the different lipids (SM, NAPE, PC, PE,
AEPL, CL) were detected by ELISA, using anti-bovine IgM-HRP
conjugate. The results show the average absorbance ? S.D. obtained
from five animals. Assays were performed in triplicate. The absor-
bancevalues obtained when using PC, PE and AEPL as antigens were
and CL (p < 0.001). (B) Antibodies against the different lipids (PE,
NAPE, PS, SM, PC, AEPL) were detected by dot blots using anti-
bovine IgM-HRP and detection by chemiluminiscence. The results
shown are representative of three different serum samples. (C) Detec-
tion of IgM adsorbed to the different liposomes was performed by
immunoblotting using anti-bovine IgM-HRP conjugate and chemilu-
miniscent detection. The results shown are representative of three
different serum samples. PS, phosphatidylserine; CL, cardiolipin.
Author's personal copy
(Fig. 3B) and immunoblot (Fig. 3C). Collectively, the
results concur in demonstrating that normal bovine
serum contains high amounts of IgM antibodies against
PE and its phosphono analogue AEPL, and lower levels
against PC. By contrast, no detectable antibody activity
was found against SM, NAPE or cardiolipin using these
methods. Cardiolipin was included in the repertoire of
phospholipid antibodies in some human diseases
(Schultz, 1997). No reactivity was found using anti-
bovine IgG conjugates in ELISAs, dot blots and
immunoblots, thus IgM is largely the class of antibody
responsible for the anti-phospholipid activity. Further-
more, we found that IgM purified from fresh bovine
serum in Sepharose 6B, bound mainly to PE as
determined by ELISA (Fig. 4). These observations
are in agreement with the effects of the adsorption of
serum aliquots with different lipids on the hemolytic
activity (Fig. 2A).
3.2. Sensitivity to bovine anti-phospholipid
antibody-mediated hemolysis depends on the
phospholipid composition of the target erythrocytes
We initially investigated whether the treatment with
bromelain, a protease that exposes membrane phos-
pholipids, could sensitize bovine erythrocytes to
hemolysis by bovine serum, as observed in mice
(Cox and Hardy, 1985). Incubation with bromelain had
no lytic effect on bovine erythrocytes exposed either to
bovine serum, which contains mainly anti-PE anti-
bodies,or tomurine serum whichcontains mainlyanti-
PC antibodies (Mercolino et al., 1988). Bromelain
cleavage of surface proteins in bovine erythrocyte
ghosts was confirmed by SDS-PAGE. Significant
differences were observed between control and
bromelain-treated erythrocyte ghosts, where protein
(data not shown). In addition, incubation of washed
bovine erythrocytes with dioctanoyl-PC (DOcPC)
resulted in its rapid uptake and consequent modifica-
tion of their phospholipid composition (unpublished
data). After 1h of incubation in the different sera,
the DOcPC modified erythrocytes showed 80 ? 4%
hemolysis for normal serum, 6 ? 1% for heat-
inactivated serum and 7 ? 1% for SM/AEPL lipo-
some-adsorbed serum. These results show that DOcPC
modified erythrocytes became sensitive to lysis by
bovine serum obtained from the same blood samples
but not by heat-inactivated bovine serum or SM/AEPL
liposome-adsorbed serum. Furthermore, when brome-
lain-treated human erythrocytes were incubated with
bovine serum, hemolysis of these cells persisted
(84 ? 2%). Therefore, phospholipid composition and
not surface proteins, appears to be the key determinant
inthe sensitivitytobovine anti-phospholipidantibody-
Regarding fetal calf erythrocytes, we found that they
already presented the phospholipid composition of
adults, with low or absent PC (Fig. 5). However, fetal
bovineserum didnot show detectableanti-phospholipid
antibody activity and lacked lytic effects on human
erythrocytes, in accordance with the fact that fetal
bovine immune system is not mature.
3.3. A possible biological role of anti-phospholipid
We examined whether bovine anti-phospholipid
antibodies could be needed to prevent invasion of the
bloodstream and internal body cavities by ruminal
ciliates. This was analyzed by testing the effects of
bovine sera on axenic cultures of the ciliate T.
thermophila that shares with rumen ciliates the
occurrence of phosphonolipids exposed on the plasma
membrane (Holz and Conner, 1973; Kennedy and
Thompson, 1970). We observed that when T. thermo-
phila cells were exposed to bovine serum, they were
rapidly lysed and this effect was completely prevented
by adsorption of the serum with SM/AEPL liposomes
(Fig. 6A), but not with SM liposomes, similar to
observations with human erythrocytes (Fig. 2A). More-
over, bovine serum had similar effects on freshly
obtained ruminal ciliates. After 5 min of exposure to
SM-adsorbed bovine serum, in a 1:4 dilution, all ciliate
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179 176
bovine IgM was purified by precipitation after dialysis against dis-
tilled water and gel filtration on Sepharose 6B. The presence of IgM
antibodies against PE, PC and SM was analyzed by ELISA using
average ? S.D. of triplicate ELISA determinations.
Author's personal copy
motility in ruminal preparations ceased and lysis was
almost complete after 30 min. On the other hand;
serum pre-adsorbed with SM/PE liposomes had
no effect (Fig. 6B). These results support a role for
anti-phospholipid antibodies in host defense against
Bovine erythrocytes possess a peculiar phospholipid
composition in their plasma membranes that consists of
low or absent PC, PE confined to the inner leaflet and
high SM and NAPE contents (Christie, 1981; Florin-
Christensen et al., 2001; Matsumoto and Miwa, 1973).
We here report the presence of hemolytic activity in
normal bovine serum against human, pig and mouse
erythrocytes that present high levels of PC and PE
(Zwaal et al., 1975).
Anti-PC antibodies are known to occur in normal
murine serum (Mercolino et al., 1988). However,
treatment of mouse erythrocytes with bromelain, a
protease that exposes membrane phospholipids, renders
them sensitive to lysis by isologous serum (Cox and
Hardy, 1985). In consequence, masking of phospholi-
pids by membrane proteins appears to be the mechan-
ism whereby murine erythrocytes tolerate anti-PC
antibodies (Cox and Hardy, 1985). In contrast, we here
observed the persistence of non reactivity of bovine
anti-phospholipid antibodies towards bovine erythro-
cytes after bromelain treatment. This fact could be due
phospholipids exposed on the membranes, but not to a
membrane protein masking effect, as demonstrated in
mice (Cox and Hardy, 1985).
We here undertook the characterization of anti-
phospholipid antibodies in normal bovine serum. We
suggesting a T-independent immune response. This is
consistent with the fact that phospholipids are T-
independent antigens (Abbas et al., 1994). We observed
that bovine erythrocytes lack or do not expose at all,
respectively. In addition, we also detected IgM
antibodies against AEPL, the phosphono analogue of
PE normally produced by ciliates present in the rumen
(Harfoot and Hazlewood, 1988).
Anti-phospholipid antibodies are known to be
heterogeneous in their specificity towards either: pure
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179177
Fig. 5. Phospholipid analysis of erythrocytes from humans and fetal
and adult bovines. Total erythrocyte lipids were extracted from the
different samples according to Bligh & Dyer and loaded on silica gel
thin layer chromatography (TLC) plates. Polar lipids were separated
using chloroform/methanol/water (65:35:2.5, v/v) as solvent mixture.
The TLC is representative of three independent experiments. O,
origin; H: human erythrocytes; AB, adult bovine erythrocytes; FB,
fetal bovine erythrocytes.
Author's personal copy
phospholipids,asdetectedinhuman bacterial infections
(Wicher and Wicher, 1991), or neo-antigens generated
by serum proteins conformational changes induced by
phospholipids, as observed in human autoimmune
diseases (Sugi and McIntyre, 1995, 1996). Herein, we
report that bovine anti-phospholipid antibodies appear
to be of the type that directly recognizes phospholipids
since purified bovine IgM bound mainly to PE and in a
lesser extent to PC, but not to SM.
Our results show that the anti-phospholipid anti-
bodies present in bovine serum are non-hemolytic to
self-erythrocytes due to the peculiar membrane
architecture (Florin-Christensen et al., 2001). In
addition, we also report that this membrane composi-
tion is already present in erythrocytes from fetal
bovines. Thus, the striking paucity of PE and PC
from the surface of bovine erythrocytes could be the
reason why these lipids become omitted from the
repertoire of auto-antigens, permitting the survival of
these B-cell clones throughout the ontogenic develop-
ment of bovine immune system. Then, the constant
stimulation of these B-cell clones by AEPL, from
rumen ciliates, present in bovine bloodstream (Tamari
antibodies that could cross-react with glyceropho-
spholipids. Nevertheless, bovine erythrocytes exclude
those phospholipids that could be recognized by the
circulating antibodies, thus preventing autoimmune
The ciliate T. thermophila presents phosphonolipids
exposed on the plasma membrane (Holz and Conner,
1973; Kennedy and Thompson, 1970). T. thermophila is
known to secrete diverse hydrolases (Florin-Christen-
sen et al., 1989), including cytolytic phospholipases
(Florin-Christensen et al., 1985; Florin-Christensen
et al., 1986a,b). Therefore, phosphonolipids appear to
play an important role in protecting these ciliates from
their own secretions (Florin-Christensen et al., 1986c).
As regards the biological role of anti-phospholipid
antibodies, a distinct possibility is that they could
prevent invasion of the bloodstream and internal body
cavities by ruminal ciliates. Our results regarding the
incubation of T. thermophila and ruminal ciliates with
normal bovine serum strongly support the fact that anti-
phospholipid antibodies could be useful in order to
prevent the systemic invasion by these protozoa.
In conclusion, these results are the first to link the
ruminal ciliates with the striking lipid peculiarities that
characterize bovine erythrocytes, through the presence
of anti-phospholipid antibodies. The unique bovine
erythrocyte phospholipid composition may be an
evolutionary adaptation to prevent erythrocyte hemo-
lysis while allowing the generation of cytolytic anti-
phospholipid antibodies against ruminal ciliates.
Finally, as regards a possible application of our
findings, we suggest that the enhancement of the natural
anti-phospholipid antibodies in bovines, could improve
other immunoprophylactic strategies against pathogens
where phospholipids are relevant antigens.
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179178
Fig. 6. Lysis of Tetrahymena thermophila and ruminal ciliates by bovine serum is blocked by adsorption with SM/AEPL or SM/PE liposomes. (A)
Microscopic observation (400?) of T. thermophila cells exposed for 30 min to heat-inactivated bovine serum diluted 1:8 (control), bovine serum
adsorbed with SM liposomes (SM-ads S) or with SM/AEPL liposomes (SM/AEPL-ads S) at 30 8C, followed by fixation with 2% formalin in PBS.
Note lytic effects only in the SM-ads treated suspensions. (B) Arrows indicate moving rumen ciliates incubated with SM/PE-adsorbed bovine sera
diluted 1:4. No such movement was observed in rumen ciliate suspensions treated with SM-ads sera after 5 min of exposure. The experiments were
performed thrice with similar results.
Author's personal copy Download full-text
We thank Dr. Estela Lammel, Dr. Terry McElwain,
Dr. Guy Palmer, and Dr. Wendy C. Brown for helpful
discussions and manuscript editing. This work was
supported by Agencia Nacional de Promocio ´n Cientı ´-
fica y Tecnolo ´gica (FONCYT), Consejo Nacional de
Investigaciones Cientı ´ficas y Te ´cnicas (CONICET) and
Immunology. W.B. SaundersCompany, Philadelphia, pp. 188–203.
Ames, B.N., 1966. Methods in Enzymology. Academic Press, New
York, pp. 115–118.
Baron, C.B., Cunningham, M., Strauss III, J.F., Coburn, R.F., 1984.
Pharmacomechanical coupling in smooth muscle may involve
phosphatidylinositol metabolism. Proc. Natl. Acad. Sci. USA
Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction
and purification. Can. J. Med. Sci. 37, 911–917.
phosphatidylcholine drastically reduces the parasitaemia of sub-
sequent Plasmodium chabaudi chabaudi blood-stage infections.
Immunology 94, 35–40.
Caramelo, J.J., Florin-Christensen, J., Delfino, J.M., 2003. Phospho-
lipase activity on N-acyl phosphatidylethanolamines is critically
dependent on the N-acyl chain length. Biochem. J. 374, 109–115.
Christie, W.W., 1981. Lipid Metabolism in Ruminant Animals. Per-
gamon Press, Oxford, pp. 95–192.
modified RBC are specifically inhibited by a common membrane
phospholipid, phosphatidylcholine. Immunology 55 (2), 263–269.
Florin-Christensen, J., Florin-Christensen, M., Knudsen, J., Rasmus-
sen, L., 1985. Cytolytic activity released from Tetrahymena. J.
Protozool. 32, 657–660.
Florin-Christensen, J., Florin-Christensen, M., Rasmussen, L., Knud-
sen, J., 1986a. An acid phospholipase C from Tetrahymena culture
medium. Comp. Biochem. Physiol. 85B, 143–148.
Florin-Christensen, J., Florin-Christensen, M., Rasmussen, L., Knud-
sen, J., Hansen, H.O., 1986b. Phospholipase A1 and triacylgly-
cerol lipase: two novel enzymes from Tetrahymena culture
medium. Comp. Biochem. Physiol. 85B, 149–155.
Florin-Christensen, J., Florin-Christensen, M., Knudsen, J., Rasmus-
attack and defence system? TIBS 11, 354–355.
Florin-Christensen, J., Sua ´rez, C.E., Florin-Christensen, M., Wains-
zelbaum, M., Brown, W.C., McElwain, T.F., Palmer, G.H., 2001.
A unique phospholipid organization in bovine erythrocyte mem-
branes. Proc. Natl. Acad. Sci. USA 98, 7736–7741.
Florin-Christensen, M., Florin-Christensen, J., Tiedtke, A., Rasmus-
sen, L., 1989. New aspects of extracellular hydrolytic enzymes in
lower eukaryotes. Eur. J. Cell Biol. 48, 1–4.
Harfoot, C.G., Hazlewood, G.P., 1988. The Rumen Microbial Eco-
system. Elsevier Applied Science, pp. 285–322.
Holz, G.G., Conner, R.L., 1973. Biology of Tetrahymena. Hutchinson
& Ross Inc., Dowden, pp. 99–122.
Horiguchi, M., Kandatsu, M., 1959. Isolation of 2-aminoethane
phosphonic acid from rumen protozoa. Nature 184, 901–902.
Kapoulas, V.M., 1969. The chromatographic separation of phospho-
nolipids from their phospholipid analogs. Biochim. Biophys. Acta
Kennedy, K.E., Thompson Jr., G.A., 1970. Phosphonolipids: localiza-
Kittredge, J.S., Roberts, E.A., 1968. A carbon-phosphorus bond in
nature. Science 164, 37–42.
Matsumoto, M., Miwa, W., 1973. Study on the new phospholipid, N-
acyl-I-alkyl glycerophosphorylethanolamine, from bovine ery-
throcytes. Biochim. Biophys. Acta 296, 350–364.
Mercolino, T.J., Arnold, L.W., Hawkins, L.A., Haughton, G.J., 1988.
Normal mouse peritoneum contains a large population of Ly-1+
(CD5) B cells that recognize phosphatidylcholine. J. Exp. Med.
Ogimoto, K., Imai, S., 1981. Atlas of Rumen Microbiology. Japan
Scientific Societies Press, Tokyo.
Rosenthal, A.F., Pousada, M., 1968. Inhibition of phospholipase C by
phosphonate analogs of glycerophosphatides. Biochim. Biophys.
Acta 164, 226–237.
Schultz, D.R., 1997. Antiphospholipid antibodies: basic immunology
and assays. Semin. Arthritis Rheum. 26 (5), 724–739.
Sugi, T., McIntyre, J.A., 1995. Autoantibodies to phosphatidyletha-
nolamine (PE) recognize a kininogen-PE complex. Blood 86,
Sugi, T., McIntyre, J.A., 1996. Phosphatidylethanolamine induces
specific conformational changes in the kininogens recognizable
by antiphosphatidylethanolamine antibodies. Thromb. Haemosta-
sis 76, 354–360.
Tamari, M., Kametaka, M., 1980. Agric. Biol. Chem. 44, 1957–1959.
Wicher, K., Wicher, V., 1991. In: Harris, E.N., Exner, T., Hughes,
G.R.V., Asherson, R.A. (Eds.), Phospholipid-Binding Antibodies.
CRC Press, Boston, pp. 97–105.
Zwaal, R.F.A., Roelofsen, B., Comfurius, P., Van Deenen, L.L.M.,
as detected by the action of various purified phospholipases.
Biochim. Biophys. Acta 406, 83–96.
G. Gimenez et al./Veterinary Immunology and Immunopathology 119 (2007) 171–179 179