Glycobiology vol. 11 no. 10 pp. 831–841, 2001
The major gangliosides of human peripheral blood monocytes/macrophages: absence of
ganglio series structures
Herbert C. Yohe1,2, Paul K. Wallace3, Charles S. Berenson4,
Song Ye5, Bruce B. Reinhold5, and Vernon N. Reinhold5
2Research
Service, Veterans Administration Medical and Regional Office
Center, 215 North Main Street, White River Junction, VT 05009, USA and
Department of Pharmacology and Toxicology, Dartmouth Medical School,
Hanover, NH 03755, USA; 3Department of Microbiology, Dartmouth Medical
School, Hanover, NH 03755, USA; 4Infectious Disease Section, Department
of Veterans Affairs, Western New York Healthcare System, State University
of New York at Buffalo, School of Medicine, Buffalo, NY 14215, USA; and
5Department of Chemistry, University of New Hampshire, Durham,
NH 03824, USA.
Received on May 31, 2001; accepted on June 27, 2001
Sialoglycosphingolipids (gangliosides) are membrane
components of eukaryotic cells that modulate cell signal
transduction events. Discrepancies exist in the published
descriptions of the gangliosides present in the human
peripheral monocyte/macrophage. Macrophages were
isolated from healthy human volunteers by two different
methods. Their ganglioside fractions were isolated and
examined by 2D thin-layer mobility, enzymatic susceptibility, and mass spectral-collision induced dissociationmass spectral analyses. Thin-layer ganglioside chromatographic patterns displayed four major doublets and were
similar for monocytes/macrophages isolated by either
apheresis/elutriation or density gradient centrifugation. All
gangliosides were resistant to β-galactosidase but sensitive
to Clostridium perfringens sialidase, indicating the absence
of terminal galactose residues and sialidase-resistant sialic
acid moieties. Mass spectra indicated only three major sets
of glycolipid components with mass heterogeneity in the
ceramide portion of each set. In all the gangliosides,
the ceramide moiety contained only C18 sphingosine with
the heterogeneity produced by the presence of C16 or C24
fatty acid. One doublet was resistant to Newcastle disease
virus sialidase, indicating the presence of an α(2-6)-linked
sialic acid residue with the same mass as another doublet.
All data was consistent with the following structures as the
major gangliosides of human peripheral monocyte/macrophages: II3NeuAcLacCer (sialolactosyl ceramide, GM3),
IV3- and IV6NeuAcnLcOse4Cer (sialoparagloboside, nLM1),
and IV3NeuAcnLcOse6Cer (a sialohexosylceramide).
Key words: glycolipids/gangliosides/human/monocytes/
tandem mass spectrometry
1To
whom correspondence should be addressed
© 2001 Oxford University Press
Introduction
Gangliosides, sialic acid–containing glycosphingolipids, are
components of the cell plasma membrane, where they appear
mainly in lipid rich areas of the plasmalemma termed rafts
(Simons and Ikonen, 1997). These glycolipids are now
emerging as important modulators of cell signal transduction
(Yates and Rampersaud, 1998). Though early investigations
centered on glycolipids of brain, work in murine models has
shown that substantial glycolipid alterations can occur in
immune cells on cell differentiation and in cells with genetic
deficiencies not directly related to glycolipid biosynthesis
(Schwarting and Gajewski, 1981; Yohe and Ryan, 1986; Yohe
et al., 1991). Like brain, glycolipid patterns of immune cells
are conserved within a species (Nakamura et al., 1988; Yohe
et al., 1991), unlike glycolipid patterns from other peripheral
tissues (Nakamura et al., 1988).
Macrophage gangliosides possess a unique array of
structural and immunoregulatory attributes, distinct from
gangliosides of other tissues. Furthermore, the ability of
glycolipids to affect cell function appears dependent on not
only the carbohydrate portion of these compounds but the
ceramide as well (Kannagi et al., 1982). This is supported by
several lines of evidence. Murine macrophage gangliosides
inhibit T cell proliferation with far greater potency than do
brain gangliosides and act at the level of the cell membrane,
while the effect of neural gangliosides is primarily extracellular (Ryan et al., 1985; Berenson and Ryan, 1991).
Furthermore, the immunologic effect of macrophage gangliosides is reversible, but brain ganglioside-mediated inhibition of
T cell function is not (Berenson and Ryan, 1991). Murine
macrophage gangliosides are also far more effective at downregulating human monocytic CD4 expression than are brain
gangliosides (Berenson et al., 1991).
Monoclonal antibodies raised specifically against human
macrophage gangliosides activate human macrophages and
functionally inhibit macrophage migration, much as antibody
binding to GD3 initiates T cell activation (Norihisa et al., 1994;
Ortaldo et al., 1996). Thus gangliosides of human macrophages
and lymphocytes are likely receptors for activating ligands. In
particular, minor endogenous human macrophage gangliosides
bind nontypeable Haemophilus influenzae with high specificity,
while gangliosides of brain origin do not (Fakih et al., 1997).
Most notably, all gangliosides of normal human macrophages
are represented in macrophages of patients with AIDS, but
become progressively inaccessible to surface molecules with
progression of HIV disease (Berenson et al., 1998). This is
accompanied by diminished responsiveness to gangliosidestimulated activation, further emphasizing the clinical importance of macrophage gangliosides (Berenson et al., 1998). The
831
H.C. Yohe et al.
conspicuous immunologic properties of macrophage gangliosides may be due to distinct molecular structures, and correct
identification of such structures is critical to determining their
function in immune cells. Unfortunately, the glycolipid structures of many peripheral cell systems are not well defined.
Despite increasing evidence for their importance in cellular
interactions, conflicting reports exist on the ganglioside
structures present in human monocytes/macrophages. Investigations have demonstrated effects on macrophage function by
Vibrio cholerae toxin, a ligand known to bind the sialidaseresistant monosialoganglioside, GMla (Corcoran et al., 1994;
Krakauer, 1996) and previous investigations have claimed the
presence of GMla in human peripheral blood monocytes.
Progressive expression of GM1a on human mononuclear
phagocytes was reported with advancing HIV infection (Auci
et al., 1992; Fantini et al., 1998).
In contrast, Berenson et al. (1998) were able to detect neither
any ganglioside peaks of human macrophages that comigrated
with GM1a standards, nor any significant differences in
relative expression of any ganglioside peaks on chromatograms between HIV-seronegative and -seropositive donors.
Furthermore, an investigation of the glycolipids of human
peripheral blood immune cells by Kiguchi et al. (1990) also
did not report the presence of GMla or asialoGM1 structures in
human monocytes. Kiguchi et al. (1990) indicated all human
monocytic gangliosides were sialidase-sensitive, terminally
sialylated, monosialo structures. This further contrasts with a
report indicating the presence of the disialoganglioside, GD3,
in human monocytes (Fantini et al., 1998). Nonetheless, all
investigations agree that sialyllactosyl ceramide, GM3, is the
major ganglioside component present. The immunologic functions of human monocyte/macrophage gangliosides are very
likely modified by the sialidases released by many infectious
bacteria. Such sialidase sensitivity would be dependent not
only on the accessibility of these molecules but also on specific
carbohydrate structures. However, the conflicting published
information on ganglioside structure made it imperative to first
examine definitively the core ganglioside structures of human
monocytes and macrophages. The work presented here
confirms the carbohydrate structures previously suggested by
Kiguchi et al. (1990). Our investigation further expands these
findings by demonstrating that ceramide structures of murine
immune cell gangliosides (Muthing et al., 1987; Yohe et al.,
1991, 1997), which differ from those of mammalian brain
glycolipids, are also present in human monocyte/macrophage
glycolipids.
Results
Characterization of isolated monocyte fractions
Isolated monocyte preparations were examined for purity prior
to adherence by using a series of defined antibodies, selected to
detect not only monocytes but also the presence of other cells.
The results are given in Table I. Results indicate the minimum
monocyte purity before adherence was at least 80%, with most
preparations contained 5–10% T and B cells. Staining with a
pan-leukocyte marker was consistently very high, indicating
that contamination from nonlymphoid cells was very low.
Adherent cells examined microscopically for esterase staining
were 97–99% nonspecific esterase positive.
832
Table I. Antibody analyses of the isolated monocyte preparations
Percentage of total (N)a
Antigen
Major cell type(s) assessed
CD3
T cells
0.5–9.5 (6)
CD19
B cells
4.2–6.5 (6)
CD14
monocytes
82.2–98.1 (6)
CD64
monocytes
78.8–98.2 (6)
CD32
monocytes
88.4–99.6 (6)
CD16
NK cells, macrophages
18.3–29.7 (6)
CD15
myeloid marker
47.0–55.2 (2)
CD40
antigen presenting cells
63.5–91.8 (4)
CD56
NK cells
4.1–8.6 (4)
DRA 12 (HLA Class II) class II
87.6–91.2 (6)
CD45
85.6–100 (4)
Pan leucocyte marker
aPercentages are shown as the range of values obtained for the number of
samples analyzed. As indicated, not all preparations were tested with all antibodies
HPTLC of the monocyte gangliosides
Two-dimensional thin-layer analyses of the monocyte gangliosides displayed minor differences between the samples from
the two different isolation procedures employed (Figure 1). In
all samples, the ganglioside pattern consisted of four major
doublets with both visual (Figure 1) and distribution analyses
(Table II) showing the majority of the sialic acid density
residing in the fastest moving doublet. A slight reduction in the
doublet designated 3,4 in density gradient-isolated monocytes
was the only difference noted for the two methods of isolation.
When apheresed cells were allowed to differentiate in culture,
patterns were similar to the ganglioside patterns of freshly
isolated monocytes.
Enzymatic and lectin analyses of monocyte gangliosides
Clostridium perfringens sialidase treatment. Experiments were
performed to determine if sialic acid residues of any human
macrophage gangliosides are susceptible to C. perfringens
sialidase. Control samples demonstrated that the sialidasesusceptible standard GM3 treated with C. perfringens sialidase
was no longer resorcinol-positive, whereas sialidase-resistant
standards GM1a and GM2 were unaffected. Equimolar
volumes of sialidase-treated and untreated macrophage
ganglioside samples were run on thin-layer chromatograms
(TLCs) and sprayed with resorcinol. Greater than 98% of the
human monocyte ganglioside sialic acid was removed by
sialidase treatment, compared with untreated controls. The
presence of desialylated substrates was confirmed by exposure
of the chromatograms to iodine vapors prior to resorcinol
spraying. Incubation with buffer alone did not cause any
degradation. Results indicate the absence of any C. perfringens
sialidase-resistant residues.
β-galactosidase treatment. Human macrophage ganglioside
susceptibility to β-galactosidase was measured to determine if
any possessed external galactose residues. Equimolar volumes
of β-galactosidase-treated and untreated samples were
prepared on TLC and sprayed with resorcinol. To confirm
activity of the enzyme lot and conditions used, control samples
were simultaneously treated. A shift in the chromatographic
mobilities of ganglioside standards which possess an external
Major ganglosides of human monocytes
galactose was observed; GD1b converted to GD2 and GM1a
converted to GM2. No change in the chromatographic mobility
of standards GD1a or GD3, which both lack an external
galactose, was seen. No changes in either the relative chromatographic mobility or content were seen in any ganglioside
from human monocyte/macrophage preparations in either the
buffer or β-galactosidase-treated samples. The results indicate
the absence of terminal galactose residues.
Fig. 1. Two-dimensional thin-layer chromatograms of gangliosides from human
monocytes isolated by two different procedures. Monocytes were isolated by
leukapheresis/elutriation (a) and by gradient centrifugation of whole blood (b).
Gangliosides were isolated and chromatographed as described in Materials and
methods. Chromatographic origin is in the lower right with solvents for the first
and second dimensions developed as indicated by the numbered arrows.
Gangliosides are visualized by the reaction of their sialic acid moieties with
resorcinol.
Newcastle disease virus (NCDV) sialidase treatment. Experiments were performed to independently verify the anomeric
sialic acid linkage of human monocyte gangliosides by
treatment with NCDV sialidase, which has relative specificity
for removal of α2,3-linked sialic acids (Corfield et al., 1983;
Berenson et al., 1995). Equimolar volumes of NCDV
sialidase-treated and untreated samples were prepared, placed
on TLCs, and sprayed with resorcinol. Sialic acid residues
were lost from nearly all ganglioside peaks, except for the
minor doublet of peaks designated 5 and 6 in Figure 2. The
most abundant peak, peak 1 (GM3), which comprised approximately 50% of the total volume (Table II), also retained a faint
amount of resorcinol positivity, likely due to its greater relative
volume. The presence of resorcinol-negative glycolipid, in
sialidase-treated samples, was confirmed in all TLCs by
reversibly staining with iodine vapor prior to resorcinol
spraying (Berenson et al., 1989). The results support previous
data indicating a ganglioside sialylated by an α2,6 anomeric
linkage (Kiguchi et al., 1990).
To confirm the sialidase activity, control samples were
treated and demonstrated that externally α2,3 sialylated
gangliosides, GM3 and GD3, treated with NCDV sialidase
were no longer resorcinol-positive, and internally sialylated
gangliosides, GM1a and GM2, remained resorcinol-positive.
Sambucus nigra lectin binding. In a single experiment, an
in situ thin-layer binding analysis was done using S. nigra
lectin, which binds α(2-6)-linked sialic acid (Shibuya et al.,
Table II. Relative percentages of human monocyte/macrophage gangliosides
on thin layer chromatograms isolated by two different methods
Relative percentage
Peak no.
Apheresis/elutriation (N = 5)
Gradient centrifugation (N = 4)
1
50.4 ± 1.5
49.8 ± 3.7
2
28.2 ± 0.5
26.6 ± 4.8
3
7.3 ± 0.5
3.8 ± 0.3
4
4.2 ± 0.3
2.5 ± 0.4
5
3.0 ± 0.4
3.3 ± 0.4
6
2.7 ± 0.4
4.5 ± 0.4
7
0.8 ± 0.3
2.5 ± 0.9
8
0.7 ± 0.2
2.0 ± 0.6
9
2.6 ± 0.4
2.0 ± 0.5
Gangliosides from 6 × 107 to 1.2 × 108 cells were separated by 2D thin-layer
chromatography. Thin-layer chromatograms of macrophage gangliosides were
scanned by densitometer. Based on density compared to brain ganglioside
standards, human monocytes contain 3.5–4 µg ganglioside sialic acid /108
cells. Each value was calculated as a percentage of the total measured ganglioside
content of the chromatogram to compensate for differences in total quantity of
ganglioside applied to each TLC plate. Peak number designations (left column)
correspond to those on the schematic diagram (Figure 2).
Fig. 2. Schematic of human monocyte 2D thin-layer ganglioside chromatogram
pattern. Pattern schematic corresponds to images in Figure 1 and to the sialic acid
density distributions for cells from the two isolation methods given in Table II.
833
H.C. Yohe et al.
1987). S. nigra lectin linked to horseradish peroxidase was
purchased from E Y Laboratories (Costa Mesa, CA). Thinlayer plates were prepared and analyzed as detailed previously
for horseradish peroxidase–linked antibodies (Yohe et al.,
1991). The doublet numbered 5,6 (Figure 2) displayed a weak
but positive binding, also indicating this set contained an
α(2-6)-linked sialic acid.
Radioisotope labeling and detection of macrophage
glycolipids
Autoradiographic analyses of labeled monocyte gangliosides
failed to reveal the presence of additional gangliosides
following extended exposure of TLCs to film as described in
Materials and methods. A sample autoradiogram following a
48-h exposure to Kodak BioMax™ MS film is shown in
Figure 3. Extension of the exposure to over 2 weeks revealed a
few trace components migrating near the identified gangliosides, suggesting the components may be present as triplets
rather than doublets (data not shown). One salient observation
was that the components containing the α(2-6) linked sialic acid
(designated 5, 6 in Figure 2) appeared to be synthesized at a
different rate, as they were barely evident on the films exposed for
less than 48 h (marked by central arrows in Figure 3).
Mass spectral analyses of monocyte gangliosides
Initial mass spectral analyses of the total methylated ganglioside fraction gave two intense peaks with parent ions of m/z
1372 and 1484; fragmentation analyses indicated the ganglioside GM3 with ceramides containing C16 and C24 fatty acids
and a single sphingosine base of C18 (data not shown).
Subsequent preparations of monocyte/macrophage gangliosides were prepared by altering the final elution of the
ganglioside fraction from the Iatrobead columns (2 ml bed
volume) by using 5 ml of chloroform:methanol (4:1, by vol.)
followed by 15 ml of chloroform:methanol (1:2, by vol.).
These solvent mixtures and volumes caused a large portion of
the major peak, now identified as GM3, to elute in the first
fraction, whereas the rest of the ganglioside fraction and a
portion of the GM3 eluted in the second fraction. Methylation
of this second ganglioside fraction followed by MS provided
the ion pattern shown in Figure 4. Under the ionization
conditions employed, the glycolipid acquires charge by
adducting sodium cations. This results in a molecular ion peak
at m/z = [M + n23]/n where M is the molecular mass of the
permethylated glycolipid, n is the number of adducted sodium
cations, and 23 is the mass of the sodium cation. Mass relationships suggest that the parent ion profile consists of three sets of
ions containing two components each.
The six major ion peaks of the parent fraction were
selectively transmitted into the collision cell followed by mass
analyses of the collision products of mass spectral–collisioninduced dissociation–mass spectral analysis (MS-CID-MS).
Fragmentation analyses of the parent ion m/z 922.2 and 978.2
indicated that these ions were doubly charged components of
monosialylated tetraosylceramides with the heterogeneity
residing only in the alkane (ceramide) portion of the molecules.
The resulting spectra are shown in Figure 5 with the derived
structures in Figure 6. Fragmentation analyses gave a common
carbohydrate structure with their ceramides having common
C18-sphingosine bases, but differing fatty acids of C16 (m/z 922.2)
and C24 (m/z 978.2). Carbohydrate fragmentation indicated
terminal sialylation, consistent with its sialidase sensitivity.
The assignment of the structures as sialoneolactosylceramides
(sialoparaglobosides) was also based on the absence of any
N-acetylgalactosamine when the monosaccharide composition
of a single total monocyte ganglioside preparation was
examined by gas-liquid chromatography as done previously
for murine monocytes (Yohe et al., 1991; Griffin and Yohe,
unpublished observations). Fragmentation analyses of the
parent ions m/z 1147 and 1202 also indicated terminally
sialylated structures, consistent with the sialidase analyses, but
with a hexosyl carbohydrate backbone containing the same
ceramide heterogeneity. An example of the fragmentation of
the m/z 1147 ion is given in Figure 7. The last set of doublets
of m/z 1372 and 1484 was again identified as singly charged
ions from GM3 having the same ceramide pattern as seen for
the other ganglioside components.
Discussion
Fig. 3. Autoradiogram of human monocyte 2D thin-layer gangliosides
following intracellular labeling. Autoradiograms were obtained as described in
Materials and methods. Image shown was obtained from a film after a 48-h
exposure to a chromatogram containing 400 DPM. Chromatographic origin is
in the lower right with solvents for the first and second dimensions developed
as indicated by the numbered arrows. Arrows with the chromatogram indicate
the weakly labeled moieties corresponding with the gangliosides designated 5,
6 in the schematic.
834
Samples from human monocytes/macrophages isolated by two
different methods show 2D thin-layer human monocyte/
macrophage ganglioside patterns to consist of four doublets,
with the fastest migrating doublet as the major component.
Consistent with all prior investigations, this major component
was determined to be the sialylated lactosylceramide molecule,
GM3. Published discrepancies exist between investigators
(Kiguchi et al., 1990; Auci et al., 1992; Fantini et al., 1998) as
to the identity of the other ganglioside components of the
human monocyte/macrophage. In a detailed report of the
glycolipids of peripheral blood lymphocytes, Kiguchi et al.
(1990) described the gangliosides of monocytes as terminally
sialylated monosialo species consisting of GM3, sialoparagloboside, and a monosialohexosyl structure. In our investigations,
Major ganglosides of human monocytes
Fig. 4. Parent ion mass spectrum derived from the isolated methylated monocyte ganglioside preparation. Conditions for preparation of the sample and the method
for obtaining the mass spectrum are given in the text.
combined enzymatic data show only sialidase susceptible
species with no susceptible galactose residues present. The
mass spectra show that all structures analyzed were terminally
monosialylated. All data is consistent with the absence of
GM1a as a component of the human monocyte/macrophage.
Our structural data, including the mass spectral collision
analyses of gangliosides, are in complete agreement with that
described by Kiguchi et al. (1990).
Previous investigations on murine B-cells indicated that the
method of isolation had a profound impact on the glycolipid
profile (Berenson et al., 2001). However, the method of
isolation or culture of human macrophage/monocytes had little
effect on their sialylated glycolipid patterns or distribution.
Therefore, it seems unlikely that the discrepancies in the literature
are due to cell isolation methods. Auci et al. (1992) described
an increase in the surface expression of the GM1a component
of human monocytes with AIDS. This report is also in direct
contrast with that of Berenson et al. (1998) who found a
decrease in surface accessibility but no decrease in ganglioside
content in macrophages of adults with AIDS. An examination
of the report reveals that Auci et al. (1992) used an antiasialoGM1a antibody as the tool for detection. Although the
cross-reactivity of the antibody used is not described, they may
have detected an HIV-induced alteration in neutral glycolipid
component(s) content or accessibility of the neutral sphingolipid constituents of the monocyte membrane. Berenson et al.
(1998) did not examine the neutral glycolipid fraction for
content or accessibility. Regardless, based on the antibody
used, the data of Auci et al. (1992) cannot definitively show
GM1a to be a component of the human monocyte/macrophage
membrane. A detailed report by Fantini et al. (1998) also
describes alterations in glycolipid metabolism in the human
monocyte on HIV infection. These investigators described a
slow moving component as the disialolactosylceramide
structure, GD3. However, our mass spectra examination found
no evidence for a disialo moiety. The data for the presence of
GD3 as reported by Fantini et al. (1998) is mainly based on
thin-layer mobility and the failure to detect a neutral tetraosyl
structure following sialidase treatment of the ganglioside
fraction. These investigators may have had the same problem
that we observed in our early attempts to obtain mass spectra:
the large percentage of GM3 overwhelmed the ability to obtain
structural information on the other components. The analyses
are further compounded by similar thin-layer chromatographic
mobilities of some of the monocyte monosialo gangliosides
and of disialo entities with a lactosyl backbone.
Kiguchi et al. (1990) indicated the presence of a ganglioside
in human monocytes that was susceptible to Clostridium
perfringens sialidase but not to sialidase derived from NCDV
and concluded that this component contained an α(2-6) linked
residue. The residue yielded a paragloboside neutral glycolipid
and was thus identified as α (2-6) linked sialoparagloboside
(Kiguchi et al., 1990). We also found a single ganglioside pair
(5, 6 in Figure 2) that was resistant to NCDV sialidase. In
addition, we noted the same ganglioside pair weakly bound an
α(2-6) linked-sialic acid-binding lectin. The results based on
all the data indicate the presence of α(2-6)-sialoparagloboside
and explain the observation of four major ganglioside pairs on
thin-layer chromatography (Figure 1), but only three pairs seen
in the mass spectra parent ion profile (Figure 4).
Select ceramide moieties of macrophage gangliosides may
also comprise important conserved immunoregulatory
components. The diverse effect of specific ceramide structure
on glycolipid activity is well established (Kannagi et al., 1982,
1983; Ladisch et al., 1994). The ceramide structures of human
monocytes/macrophages seen in this work were originally
observed in murine macrophages. Brain gangliosides have
predominantly C18 as the ceramide fatty acid, with the major
heterogeneity in the sphingosine as C18 and C20 entities
(Ando and Yu, 1984). The ceramide component of gangliosides
835
H.C. Yohe et al.
Fig. 5. Mass spectra derived from fragmentation of parent ions m/z 922.2 (a)
and m/z 978.2 (b). Fragmentation analyses with derived structures are given in
Figure 5. Fragmentation also shows the parent ions are doubly charged ions.
The spectra show a large number of common fragments for the two parent ions
with heterogeneity only in the ceramide containing fractions.
of macrophages, from either murine and now human sources,
is structurally distinct from ceramide of gangliosides of brain
origin in having a single C18-sphingosine entity with heterogeneity in the fatty acid portion as either C16 or C24 structures.
The structural distinctions of murine and human macrophage
gangliosides may well explain many of the unique immunoregulatory properties that distinguish them from brain
gangliosides. These properties include far more effective downregulation of human macrophage CD4 expression and more
potent, reversible down-regulation of lectin-induced T cell
proliferation, compared with brain gangliosides (Berenson and
Ryan, 1991; Berenson et al., 1991).
The minor gangliosides of human macrophages, designated
as 3 and 4 in this investigation, bind nontypeable H. influenzae
with high specificity and high affinity, while gangliosides of
brain origin do not (Fakih et al., 1997). The selective affinity of
gangliosides of human macrophages for activating ligands has
836
only recently been explored and is likely to be dependent on
specific core carbohydrate structures of ganglioside receptors.
Binding of nontypeable H. influenzae to gangliosides is
dependent on the presence of sialic acid, indicating the
importance of the core carbohydrate structure to this interaction (Fakih et al., 1997). The findings of this current study
now permit us to identify this putative receptor as α(2-3)
sialoparagloboside.
Effects of the GM1a-binding ligand, cholera toxin, on
human monocyte/macrophages have been described. Cholera
toxin has been shown to induce production of cytokines in
human monocytes via protein kinase c activity (Krakauer,
1996). This toxin has also been shown to alter other human
monocyte activities via adenylyl cyclase (Corcoran et al.,
1994). Our data as given here suggest the toxin could be
binding to a membrane component other than GM1a. Pessina
et al. (1989) demonstrated a profound effect of cholera toxin
on the murine myelomonocytic cell line WEHI-3D, which we
found also contained no GM1a (Yohe et al., 1992). In WEHI
cells, an extended GM1b structure containing a galactose-Nacetylgalactosamine terminus was the probable target for the
toxin. Cholera toxin has also been shown to bind fucosyl-GM1
(Masserini et al., 1992) and, to a reduced extent, the disialo
ganglioside GD1b (Cumar et al., 1982). However, even with
extended exposure of film to radiolabeled samples, no such
alternate glycolipid target appeared in the human monocyte
ganglioside fraction. Attempts to analyze fluorescein isothiocyanate (FITC)-tagged cholera toxin b subunit binding by
fluorescence-assisted cell sorting (FACS) or by fluorescent
microscopy indicated low nonspecific binding or were
completely negative (Bergeron et al. unpublished data).
Conceivably, a low level of surface protein with the requisite
glycan structure could bind this toxin. An alternate possibility
is a small, but still effective, level of contaminating endotoxins
in the cholera toxin preparations as has been recently noted for
lipoteichoic acid preparations (Gao et al., 2001). This possibility
is now under investigation.
It has long been known that many infectious bacteria release
sialidase (Corfield et al., 1981). Such exogenous bacterial
enzymes may serve as important virulence factors by
modifying accessible sialic acid residues of host cell sialoglycoconjugates. The data presented here, and that of
Kiguchi et al. (1990), indicate all major sialylated glycolipids
in the human monocyte are sialidase-susceptible and, along
with sialylated proteins, may be targets for exogenous bacterial
sialidases. The enzymatic alteration of a sialoconjugate in a
cell is dependent not only on its core carbohydrate structure
and the specific sialidase, but also on its accessibility. In situ
accessibility of these compounds and the effect of sialidases on
monocyte function are now under investigation.
Materials and methods
Isolation of human blood monocytes
Human blood monocytes were isolated by two different
procedures.
In one series of experiments, peripheral mononuclearenriched cell preparations were obtained by leukapheresis
using a Cobe Spectra Apheresis cell separator (Lakewood, CO)
and citrate-acid dextrose as anticoagulant. For each procedure,
Major ganglosides of human monocytes
Fig. 6. Fragmentation analyses of ions m/z 922.2 (a) and m/z 978.2 (b). Fragmentation analyses indicate the only difference between the two structures resides in
the ceramide moiety as indicated in the diagrams and specifically in the fatty acid entity.
3.1–5.3 L of whole blood was processed at a flow rate of 4.9 to
7.0 ml/min to collect a total of 68 ml of product. The
leukapheresis preparation was counted and checked for
sterility.
Monocytes were further purified from the leukapheresis
preparation by countercurrent flow elutriation using a
Beckman J2-MI centrifuge equipped with a JE-6B elutriation
rotor, standard (4 ml) elutriation chamber (Beckman Instruments,
Palo Alto, CA) and Masterflex L/S variable flow peristaltic
pump (Cole-Palmer, Niles, IL). Elutriation was performed
under sterile conditions using disposable, single-use tubing
and collection bags (Baxter Fenwal, Deerfield, IL) designed
for blood collection and processing. Before each procedure the
elutriator chamber, chamber assembly, rotating seal and
transfer tube were cleaned and rinsed with sterile distilled
water for injection. The assembled rotor with attached inflow
and outflow tubing was then sterilized by running 6%
hydrogen peroxide through the system for 30 min and then
rinsed with sterile elutriation buffer (physiological saline,
0.9%; D-glucose, 1 mg/ml; sodium heparin, 10 U/ml; human
serum albumin, 0.5%; American Red Cross, Washington DC).
During the elutriation procedure a centrifuge temperature of
22°C and a fixed rotor speed of 600 × g was used with a gradually increasing counterflow rate. Cells from the leukapheresis
preparation were pumped into the chamber at 12 ml/min.
Platelets, red blood cells, and most lymphocytes were eluted
under these conditions, but the monocytes were retained in the
chamber.
After loading was completed, elutriation buffer was pumped
through at an initial flow rate of 12 ml/min. After a total of
300 ml had passed through, the counterflow rate was
sequentially increased to 15, 18, and 25 ml/min and 150 ml of
elutant was collected at each speed. Each fraction was
centrifuged at 250 × g, resuspended in AIM V (Life Technologies,
Rockville, MD), analyzed by Coulter Counter with
channelyzer for white blood cell differential based on cell size,
and immunophenotyped by flow cytometry using an antiCD14 PE/anti-CD45 FITC cocktail (Becton Dickinson, San Jose,
CA). The fractions containing greater than 85% monocytes
(typically the fractions collected at 18 and 25 ml/min flow
rates) were pooled for additional analysis and culture. Monocyte purity was verified by FACS analyses using a selected
837
H.C. Yohe et al.
Fig. 7. Fragmentation mass spectrum (a) and analysis (b) of the parent ion m/z 1147.2. Fragmentation showed many mass units in common with m/z 922.2. Analysis
shows the compound to be a terminally sialylated monosialostructure with an extended saccharide backbone and the same ceramide moiety as m/z 922.2.
battery of fluorescent antibodies. Pooled monocyte fractions
were allowed to adhere overnight to sterile glass 100 mm petri
dishes in RPMI 1640 supplemented with 5% heat-inactivated
fetal calf serum at 5% CO2, 95% humidity and 37°C. Nonadherent cells were then removed with serial rinses of warm
phosphate-buffered saline and the cells extracted for glycolipids as detailed below. With two separate preparations,
freshly elutriated monocytes were resuspended in hydrophobic
culture bags at a density of 2.5–3 × 106 cells/ml in monocyte
culture medium (Iscove’s Modified Dulbecco’s Medium
[provided by IDM]) or AIM V, supplemented with 2.5%
autologous serum; 2 mM fresh L-glutamine; 25 mM HEPES
(BioWhittaker); 5 × 10–5 M 2-mercaptoethanol (Sigma, St.
Louis, MO) containing 10 ng/ml GM-CSF, generously supplied
by Immunex (Seattle, WA) and 1 × 10–8 M 1,25-dihydroxyvitamin D3 (Abbott Pharmaceuticals, Abbott Park, IL).
838
Cells, in a volume of 100–200 ml per bag, were cultured for
5–7 days at 37°C in a 5% CO2 humidified incubator. During
the final 18 h of culture, the cells were further activated by the
addition of 750 IU/ml of recombinant human interferon
gamma (IFNγ; Genetech, San Francisco, CA) and then centrifuged to obtain a cell pellet at 400 × g for 10 min. The pellet
was then resuspended in a 0.32 M isotonic nonionic pentaerythritol solution, repelleted, and finally extracted to obtain
lipids as described below.
In a second series of experiments, human mononuclear
phagocytes were purified from buffy coat suspensions
obtained from HIV-seronegative volunteers from the Red
Cross of Western New York as previously described in detail
by Berenson et al. (1998). All donors were also seronegative
for hepatitis B and C. Briefly, buffy coat suspensions were
prepared from whole blood centrifuged at 3800 rpm for 4 min
at 4°C. Mononuclear cells were further purified by Ficoll-Hypaque
Major ganglosides of human monocytes
density centrifugation and seeded onto 100-mm glass petri
dishes (5 × 106 cells/ml) in RPMI 1640 supplemented with
10% heat-inactivated human AB serum. After incubation at
5% CO2, 95% humidity, 37°C for 7 days, nonadherent cells
were removed with serial rinses of warm phosphate-buffered
saline. Remaining monocyte-derived macrophages were
incubated in RPMI 1640 with 10% fetal calf serum until
extracted for lipids as detailed below.
Isolation of glycolipids
The total human monocytic glycolipid fraction was isolated,
separated by ion-exchange chromatography into neutral and
ganglioside fractions and the fractions purified using reverse
phase and spherical silica gel chromatographies as previously
reported (Yohe et al., 1991; Macala and Yohe, 1995). Total
lipid extracts are prepared by overnight extraction of the cell
pellet using 10 ml chloroform:methanol (1:1, by vol)/108 cells.
Cells adhered to glass petri dishes were extracted as previously
detailed (Macala and Yohe, 1995). To remove membrane
fragments, all resulting lipid extracts were filtered through
sintered glass funnels (15 ml, medium porosity) overlaid with
a glass fiber mat. The filtered residue was rinsed with 3 × 5 ml
of chloroform:methanol (1:1, by vol). The combined extracts
and rinses were taken to dryness by rotary evaporation. Using
a slight modification of our earlier procedure (Yohe et al.,
1991), the total lipid extract was redissolved, with sonication,
in 20 ml of chloroform:methanol:water (30:60:8, by vol) and
applied to a 4-ml bed volume column of DEAE-Sephadex A-25,
acetate form (Pharmacia, Piscataway, NJ). The neutral lipid
fraction was eluted with an additional 30 ml of the same
solvent. The acidic lipid fraction, containing the gangliosides,
was then eluted with 35 ml of chloroform:methanol:aqueous
0.8 M sodium acetate (30:60:8, by vol). The acidic lipid fraction
was taken to dryness by rotary evaporation, redissolved in 5 ml
of 0.1 N aqueous NaOH, and heated at 37°C for 90 min in a
water bath to saponify any acidic phospholipids. The sample
was chilled in an ice-water bath and the sample pH reduced to
5–6 by adding 0.1 N HCl. The sample was then diluted to
20 ml with chilled water and immediately desalted on a 2 ml
bed volume reversed phase silica gel column (SepPak, Waters
Associates, Waltham, MA). Each sample was applied to the
column three times by passing the first and second eluates back
through the same column. Following the third application of
sample, flask rinses of 5 ml of 0.1 N NaCl and 2 × 5 ml of
water were applied to the column. Remaining salts were eluted
with 30 ml of water. Lipids were then eluted with 7 ml of
methanol followed by 35 ml of chloroform:methanol (1:2 by
vol). The desalting column was regenerated by rinsing with
7 ml of methanol followed by 40–50 ml of water.
The desalted lipid sample was taken to dryness by rotary
evaporation, frozen, and then lyophilized to remove all traces
of water. The lyophilized sample was dissolved, with sonication, in 1.5 ml of chloroform:methanol (1:1, by vol) followed
by 3.5 ml of chloroform, making the final proportion of
chloroform:methanol, 85:15. The sample and two 2.5 ml
flask rinses were loaded sequentially onto a 2-ml Iatrobead
6RS-8060 column (Iatron Laboratories, Tokyo). Low polarity
contaminants were eluted with an additional 10 ml of
chloroform:methanol (85:15, by vol). The total ganglioside
fraction was eluted with 20 ml of chloroform:methanol (1:2, by
vol). After removal of the solvent by rotary evaporation, the
purified ganglioside sample was transferred to a 12 × 75
conical screw cap tube with 4 × 1.5 ml rinses of
chloroform:methanol (1:2, by vol), and dried under nitrogen.
Samples were stored at –20°C until analyzed.
HPTLC of the ganglioside fractions
The macrophage ganglioside patterns were examined by
2D high-performance thin-layer chromatography (HPTLC)
(Yohe et al., 1988, 1991). Briefly, the ganglioside sample (2–5
µg sialic acid) was spotted 15 mm in and up from the lower left
corner. A human brain ganglioside standard (0.4 µg sialic) was
spotted 5 mm in and 15 mm up from the lower right corner.
The plate was chromatographed for 45 min in chloroform:methanol:0.25% aqueous KCl (50:45:10, by vol) and
dried with forced air for 5 min and then over P2O5 for 90 min
in a vacuum desiccator. The dried plate was rotated 90° counterclockwise, and a second brain ganglioside standard was
spotted 5 mm in and 15 mm up from the new lower left corner
of the plate. The plate was chromatographed in the second
dimension in chloroform:methanol:2.5 N aqueous NH4OH
containing 0.25% KCl (50:40:10, by vol) for 30 min. The
plates were dried with forced air until the NH3 odor was gone.
Ganglioside sialic acid was visualized by resorcinolhydrochloric acid spray (Svennerholm, 1957) and the distribution determined by densitometry (Yohe et al., 1988; Berenson
et al., 1998).
Enzymatic degradation of human macrophage gangliosides
C. perfringens sialidase treatment. Human macrophage gangliosides containing 5–10 µg sialic acid were incubated with
C. perfringens sialidase (2 U/ml) (Sigma) in 0.5 ml of 50 mM
sodium citrate-phosphate buffer, pH 5.5, at 37°C, for 2 h, as
previously described (Yohe et al., 1991). A duplicate sample
of equal quantity of gangliosides was incubated in buffer
alone. Reactions were terminated with addition of 0.1 M
NaOH and neutralized with 0.1 M HCl. Solutions were
desalted on SepPak columns and tested for hydrolytic products
on TLCs. TLCs were run in chloroform:methanol:0.25% KCl
(50:45:10, by vol), sprayed with resorcinol, and heated (92–
94°C). Resorcinol-positive intensity was quantitated by scanning densitometry and desialylation was determined by loss of
resorcinol-positivity, compared with untreated samples. The
percent desialylation was expressed as: [volume of (resorcinolpositive) sialidase-treated sample ÷ volume of (resorcinol-positive)
untreated sample] × 100. To verify efficacy of enzymatic
activity, positive and negative control gangliosides were
treated with sialidase, including gangliosides with external
sialic residues (GM3) and gangliosides lacking external sialic
acid residues (GM1a, GM2), respectively (Matreya, Pleasant
Gap, PA).
NCDV sialidase treatment. To further investigate the anomeric
sialic acid linkage of minor monocyte gangliosides, human
monocyte gangliosides were incubated with sialidase from
NCDV (Glyko, Novato, CA), which is specific for removal of
α2,3-linked sialic acid (Corfield et al., 1983). Human monocyte
gangliosides containing 5–10 µg sialic acid were incubated
with 0.2 U of NCDV sialidase in 100 µl of 50 mM sodium
acetate buffer containing 2 µg/µl sodium cholate (Sigma), pH 5.5,
at 37°C, for 18 h as previously described (Berenson et al., 1995).
839
H.C. Yohe et al.
A duplicate sample of equal quantity of gangliosides was
incubated in buffer alone. Reactions were terminated by
placement of samples on ice in 5 ml of 0.1 M NaCl. After
adjustment of pH to 5.0, samples were desalted on SepPak columns,
and re-eluted over Iatrobead columns, as described earlier. The
entire content of each sample was loaded onto a TLC plate, which
was run in two dimensions (chloroform:methanol:0.25% KCl,
50:45:10, by vol, and chloroform:methanol:2.5N NH3
containing 0.25% KCl, 50:40:10, by vol), also as described
earlier. TLC plates were sprayed with resorcinol and heated
(92–94°C). Resorcinol-positive intensity was quantitated by
scanning densitometry, and desialylation was determined by
loss of resorcinol-positivity, compared with untreated samples.
The presence of resorcinol-negative spots on NCDV sialidase–
treated samples was confirmed on TLCs by reversible staining
with iodine vapor prior to resorcinol spraying (Berenson et al.,
1989).
Conditions for effective desialylation of α2,3-linked
gangliosides were established, using known ganglioside
(GM1a, GM2, GM3, GD3) standards (Matreya). Under the
established conditions, gangliosides with external α2,3-linked
sialic acids (GM3, GD3) were successfully desialylated, while
those with internal α2,3-linked sialic acids (GM1a, GM2) were
not.
β-galactosidase treatment. Human macrophage gangliosides
containing 5–10 µg sialic acid were incubated with bovine
testes β-galactosidase (0.3 U/ml) (Sigma) at 37°C in 100 µl of
50 mM citrate phosphate buffer (pH 4.3), containing Triton
X-100 (0.5 µg/ml), for 18 h, as previously described (Yohe
et al., 1991). A duplicate sample of an equal quantity of
gangliosides was incubated in buffer alone. Reactions were
terminated with 0.1 M NaOH, and neutralized to pH 4–5 with
0.1 M HCl. As with sialidase treatment, solutions were
desalted on SepPak columns and tested for hydrolysis on
TLCs, run in chloroform:methanol:0.25% KCl (50:45:10, by
vol) and sprayed with resorcinol. Enzymatic degradation was
determined by a shift in chromatographic mobility of bands
compared with untreated samples and with controls of known
structures. To verify efficacy of enzymatic activity, positive
and negative control gangliosides were treated with β-galactosidase and included gangliosides with external galactose
residues (GM1a, GD1b) and gangliosides lacking external
galactose residues (GD3, GD1a), respectively (Matreya).
Radioisotope labeling and detection of macrophage
glycolipids
Monocyte/macrophages were isolated by apheresis/elutriation
and allowed to adhere as described above. The isolated cells
were exposed to 14C-galactose (ARC, St. Louis, MO) (5µCi/
20 × 106 cells) for 48–96 h and then extracted for glycolipids as
detailed above. The isolated ganglioside fractions containing
400–1000 DPM were separated by 2D thin-layer chromatography as above and the resulting chromatograms exposed to
Kodak BioMax™ MS film enhanced with a BioMax TranScreen™ LE intensifying screen (Eastman Kodak, Rochester,
NY). Films were exposed from 24 h to 4 weeks at –70°C and
developed in a Kodak X-omat™ processor using the standard
conditions. Extensive testing with 14C-labeled brain gangliosides indicated discrete spots containing as little as 30 DPM
840
could be detected with a 24 h exposure and less than 5 DPM at
4 weeks (Degregorio et al., unpublished data).
Mass spectrometric analyses of macrophage gangliosides
The total ganglioside fraction was permethylated and subjected
to electrospray ionization mass spectrometry performed on a
triple quadrapole mass spectrometer as described previously
for murine glycolipids (Reinhold et al., 1994; Yohe et al., 1997).
Acknowledgments
The authors wish to acknowledge the assistance of Mary E.
Griffin, Tamar J. Kitzmiller, and Robin H. Rasp in different
phases of this investigation. This work was supported by
Veterans Affairs Merit Reviews (H.C.Y. and C.S.B.) and
supported in part by grants from The National Institute for
Heath, AI-34478 (P.K.W.) and GM05445 (S.Y., B.B.R., and
V.N.R.)
Abbreviations
FACS, fluorescence-activated cell sorting; FITC, fluorescein
isothiocyanate; HPTLC, high-performance thin-layer chromatography; MS-CID-MS, mass spectral–collision-induced
dissociation–mass spectral analysis; NCDV, Newcastle
disease virus; TLC, thin-layer chromatograms.
References
Ando, S. and Yu, R.K. (1984) Fatty acid and long chain base composition of
gangliosides isolated from adult human brain. J. Neurosci. Res., 23, 205–211.
Auci, D.L., Chice, S.M., Durkin, H.G., and Murali, M.R. (1992) Constitutive
production of PAI-II and increased surface expression of GM1 ganglioside
by peripheral blood monocytes from patients with AIDS: evidence of
monocyte activation in vivo. J. Leukoc. Biol., 52, 282–286.
Berenson, C.S. and Ryan, J.L. (1991) Macrophage gangliosides inhibit
mitogen-induced lymphocyte proliferation. J. Leukoc. Biol., 50, 393–401.
Berenson, C.S., Gallery, M.A., Katari, M.S., Foster, E.W., and Pattoli, M.A.
(1998). Gangliosides of monocyte-derived macrophages of adults with
advanced HIV infection show reduced surface accessibility. J. Leukoc.
Biol., 64, 311–321.
Berenson, C.S., Patterson, M.A., Miqdadi, J., and Lance, P. (1995) n-Butyrate
mediation of ganglioside expression of human and murine cancer cells
demonstrates relative cell specificity. Clin. Sci., 88: 491–499.
Berenson, C.S., Rasp, R.H., Gau, J.-T., Ryan, J.L., and Yohe, H.C. (2001)
Differences in splenic B lymphocyte expression and accessibility in
normal and endotoxin-hyporesponsive mice. J. Leukoc. Biol., 69, in press.
Berenson, C.S., Yohe, H.C., and Ryan, J.L. (1989) Factors mediating
lipopolysaccharide-induced ganglioside expression in murine peritoneal
macrophages. J. Leukoc. Biol., 45, 221–230.
Berenson, C.S., Yohe, H.C., Ryan, J.L., and Buckley, P.J. (1991)
Immunomodulation of human macrophage CD4 by macrophage gangliosides. FASEB J., 5, A1349.
Corcoran, M.L., Stetler-Stevenson, W.G., DeWitt, D.L., and Wahl, L.M. (1994)
Effect of cholera toxin and pertussis toxin on prostaglandin H synthase-2,
prostaglandin E2, and matrix metalloproteinase production by human
monocytes. Arch. Biochem. Biophys., 310, 481–488.
Corfield, A.P., Higa, H., Paulson, J.C., and Schauer, R. (1983) The specificity
of viral and bacterial sialidases for alpha(2-3)- and alpha(2-6)-linked sialic
acids in glycoproteins. Biochim. Biophys. Acta, 744, 121–126.
Corfield, A.P., Michalski, J.-C., and Schauer, R. (1981) The substrate
specificity of sialidases from microorganisms and mammals. Perspect.
Inher. Metab. Dis., 4, 3–70.
Cumar, F.A., Maggio, B., and Caputto, R. (1982) Ganglioside-cholera toxin
interactions: a binding and lipid monolayer study. Mol. Cell. Biochem., 46,
155–160.
Major ganglosides of human monocytes
Fakih, M.F., Pattoli, M.A., Murphy, T.F., and Berenson, C.S. (1997) Specific
binding of Haemophilus influenzae to minor gangliosides of human
respiratory epithelial cells. Infect. Immun., 65, 1695–1700.
Fantini, J., Tamalet, C., Hammache, D., Tourres, C., Duclos, N., and Yahi, N.
(1998) HIV-1-induced perturbations of glycosphingolipid metabolism are
cell-specific and can be detected at early stages of HIV-1 infection.
J. Acquir. Immune Defic. Syndr. Hum. Retrovirol., 19, 221–229.
Gao, J.J., Xue, Q., Zuvanich, E.G., Haghi, K.R., and Morrison, D.C. (2001)
Commercial preparations of lipoteichoic acid contain endotoxin that
contributes to activation of mouse macrophages in vitro. Infect. Immun.,
69, 751–757.
Kannagi, R., Nudelman, E., and Hakomori, S.-I. (1982) Possible role of
ceramide in defining structure and function of membrane glycolipids.
Proc. Natl Acad. Sci. USA, 79, 3470–3474.
Kannagi, R., Stroup, R., Cochran, N.A., Urdal, D.L., Young, W.W.J., and
Hakomori, S.-I. (1983) Factors affecting expression of glycolipid tumor
antigens: influence of ceramide composition and coexisting glycolipid on
the antigenicity of gangliotriosylceramide in murine lymphoma cells.
Cancer Res., 43, 4997–5005.
Kiguchi, K., Henning-Chubb, C.B., and Huberman, E. (1990) Glycosphingolipid patterns of peripheral blood lymphocytes, monocytes, and
granulocytes are cell specific. J. Biochem., 107, 8–14.
Krakauer, T. (1996) Evidence for protein kinase C pathway in the response of
human peripheral blood mononuclear cells to cholera toxin. Cell. Immunol.,
172, 224–228.
Ladisch, S., Li, R., and Olson, E. (1994) Ceramide structure predicts tumor
ganglioside immunosuppressive activity. Proc. Natl Acad. Sci. USA, 91,
1974–1978.
Macala, L.J. and Yohe, H.C. (1995) In situ accessibility of murine macrophage
gangliosides. Glycobiology, 5, 67–75.
Masserini, M., Freire, E., Palestini, P., Calappi, E., and Tettamanti, G. (1992)
Fuc-GM1 ganglioside mimics the receptor function of GM1 for cholera
toxin. Biochemistry, 31, 2422–2426.
Muthing, J., Egge, H., Kneip, B., and Muhlradt, P.F. (1987) Structural
characterization of gangliosides from murine T lymphocytes. Eur. J.
Biochem., 163, 407–416.
Nakamura, K., Hashimoto, Y., Yamakawa, T., and Suzuki, A. (1988) Genetic
polymorphism of ganglioside expression in mouse organs. J. Biochem.,
103, 201–208.
Norihisa, Y., McVicar, D.W., Ghosh, P., Houghton, A.N., Longo, D.L.,
Creekmore, S.P., Blake, T., Ortaldo, J.R., and Young, H.A. (1994)
Increased proliferation, cytotoxicity, and gene expression after stimulation
of human peripheral blood T lymphocytes through a surface ganglioside
(GD3) [published erratum appears in J. Immunol., 1994 Jul
15;153(2):910]. J. Immunol., 152, 485–495.
Ortaldo, J.R., Mason, A.T., Longo, D.L., Beckwith, M., Creekmore, S., and
McVicar, D.W. (1996) T cell activation via the disialoganglioside GD3:
analysis of signal transduction. J. Leuko. Biol., 60, 533–539.
Pessina, A., Mineo, E., Masserini, M., Neri, M.G., and Cocuzza, C.E. (1989)
Inhibition of murine leukemia (WEHI-3B and L1210) proliferation by
cholera toxin B subunit. Biochim. Biophys. Acta, 1013, 206–211.
Reinhold, B.B., Chan, S.-Y., Chan, S., and Reinhold, V.N. (1994) Profiling
glycosphingolipid structural detail: periodate oxidation, electrospray,
collision-induced dissociation and tandem mass spectrometry. Org. Mass
Spectrom., 29, 736–746.
Ryan, J.L., Inouye, L.N., Gobran, L., Yohe, W.B., and Yohe, H.C. (1985) Lack
of specificity of brain gangliosides in the modulation of lymphocyte
activation. Yale J. Biol. Med., 58, 459–467.
Schwarting, G.A. and Gajewski, A. (1981) Glycolipids of murine lymphocyte
subpopulations: a defect in the levels of sialidase sensitive sialosylated
asialoGM1 in beige mouse lymphocytes. J. Immunol., 126, 2403–2407.
Shibuya, N., Goldstein, I.J., Broekaert, W.F., Nsimba-Lubaki, M., Peeters, B.,
and Peumans, W.J. (1987) The elderberry (Sambucus nigra L.) bark lectin
recognizes the Neu5Ac(alpha 2–6)Gal/GalNAc sequence. J. Biol. Chem.,
262, 1596–1601.
Simons, K. and Ikonen, E. (1997) Functional rafts in cell membranes. Nature,
387, 569–572.
Svennerholm, L. (1957) Quantitative estimation of sialic acids. II. A colorimetric
resorcinol-hydrochloric acid method. Biochim. Biophys. Acta, 24, 604–611.
Yates, A.J. and Rampersaud, A. (1998) Sphingolipids as receptor modulators:
an overview. In Ledeen, R.W., Hakomori, S.-I., Yates, A.J., Schneider, J.S.,
and Yu, R.K. (eds), Sphingolipids as signaling modulators in the nervous
system. The New York Academy of Sciences, New York. 845, 57–71.
Yohe, H.C. and Ryan, J.L. (1986) Ganglioside expression in macrophages
from endotoxin responder and nonresponder mice. J. Immunol., 137,
3921–3927.
Yohe, H.C., Cuny, C.L., Berenson, C.S., and Ryan, J.L. (1988) Comparison of
thymocyte and T lymphocyte gangliosides from C3H/HeN and C3H/HeJ
mice. J. Leuko. Biol., 44, 521–528.
Yohe, H.C., Cuny, C.L., Macala, L.J., Saito, M., McMurray, W.J., and
Ryan, J.L. (1991) The presence of sialidase-sensitive sialosylgangliotetraosyl ceramide (GM1b) in stimulated murine macrophages. Deficiency of
GM1b in Escherichia coli-activated macrophages from the C3H/HeJ
mouse. J. Immunol., 146, 1900–1908.
Yohe, H.C., Macala, L.J., Giordano, G., and McMurray, W.J. (1992) GM1b
and GM1b-GalNAc: major gangliosides of murine-derived macrophagelike WEHI-3 cells. Biochim. Biophys. Acta, 1109, 210–217.
Yohe, H.C., Ye, S., Reinhold, B.B., and Reinhold, V.N. (1997) Structural
characterization of the disialogangliosides of murine peritoneal
macrophages. Glycobiology, 7, 1215–1227.
841