| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institut d’Hématologie et d’Immunologie, Faculté de Médecine, Université Louis Pasteur, Strasbourg, France;
Institut National de la Santé et de la Recherche Médicale, Unité 143, Hôpital de Bicêtre, Le Kremlin-Bicêtre; and
Laboratoire de Biochimie, Hôpital Lariboisière, Paris, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Owing to a strong affinity for phosphatidylserine (14), annexin V is now widely used for probing cell stimulation or death (4, 10, 15, 16, 17). Depending on local Ca2+ conditions and on the proportion of available phosphatidylserine, annexin V has been shown to form two-dimensional arrays on membranes (18, 19, 20) in an essentially interfacial interaction (21, 22). This probably explains its ability to counteract plasma membrane vesiculation when present during the whole course of platelet activation, but it does not prevent phosphatidylserine transmembrane migration (23). Although it is abundantly present at the surface of cells fulfilling a barrier function such as trophoblasts or placental endothelial cells, where it could exert an anticoagulant potential (24, 25), its genuine physiologic role remains to be unequivocally established (26).
The above considerations prompted us to assess the effect of annexin V on the execution of induced cell death programs in human CEM T cells. This cell line was selected because of its ability to express CD4 and cellular prion protein (PrPc)3 (27), two constitutive membrane Ags of particular significance with respective participation in the onset of cell degeneration in AIDS (2, 28) and prion diseases (29). In addition, CD4 is an integral membrane protein while PrPc is linked to the lipid bilayer by a glycosylphosphatidylinositol (GPI) anchor (29, 30). Annexin V was indeed observed to interfere in apoptosis with a direct impact on membrane features, and an indirect modulating effect on the cytoplasmic caspase activation cascade, with possible repercussions in vivo.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell culture reagents were obtained from Bioproducts (Gagny, France). FCS was obtained from Life Technologies (Paisley, U.K.). Actinomycin D, etoposide (VP-16), propidium iodide, type I-A RNase A, FITC, and human serum albumin were obtained from Sigma (St. Louis, MO). HN-2 was obtained from Laboratoires Delagranges (Paris, France). Natural annexin V was purified from human placenta, characterized according to a previous report (21), and conjugated with FITC (annexin VFITC) as already described (16). Recombinant annexin V from Bender MedSystems (Vienna, Austria) was also used in preliminary experiments and yielded identical results as the natural counterpart. mAb to CPP32 (caspase-3) was purchased from Transduction Laboratories (Lexington, KY). Biotinylated mAb to CD4, goat anti-mouse IgG, HRP-conjugated secondary Ab, and irrelevant biotinylated IgG1 were obtained from Leinco Technologies (Ballwin, MO). Biotinylated mAb to human PrPc (3F4) was obtained from Senetek (Maryland Heights, MO). z-Val-Ala-Asp.fluoromethyl ketone (z-VAD.fmk) was obtained from Calbiochem (La Jolla, CA). High binding capacity streptavidin-coated microtitration plates and Chromozym TH were obtained from Boehringer Mannheim (Mannheim, Germany). Human blood coagulation factors were the same as those used in a recent study reported by our group (10). The caspase-3 cellular activity assay kit was obtained from Biomol Research Laboratories (Plymouth Meeting, PA).
Cell culture and induction of apoptosis
Human lymphoid CEM T cells were cultured in X-vivo 15 medium (Bioproducts, Gagny, France) under standard conditions. Human HL-60 promyelocytic leukemia cells (CCL-240; American Type Culture Collection, Manassas, VA) were grown in RPMI 1640 supplemented with 10% heat-inactivated FCS. Cell counts were determined using a hemocytometer. The cell viability was checked by trypan blue exclusion. Cells were seeded at 5 x 105 cells/ml in the presence or absence of actinomycin D (0.3 µg/ml) or VP-16 (5 µM) for 18 h. In some experiments, cells and released microparticles were examined separately after centrifugation at 12,000 x g for 30 s.
Animals and treatments by HN-2 and annexin V
Male BALB/c (Institut Français de la Fièvre Afteuse Centre de Recherche et d’êlevage des oncins, Les Arbresles, France), 6 wk old, were used in these experiments. They were housed under conventional conditions and provided with standard feed and water. HN-2 was diluted in sterile 0.15 M NaCl shortly before use and injected i.p. at 6 mg/kg. Annexin V (50 µg) was also diluted in sterile 0.15 M NaCl and injected i.v. immediately after HN-2 injection. At 24 h following the injection, mice were sacrificed and their spleen were aseptically excised, weighed, and put in cold RPMI 1640. Single-cell suspensions were prepared by teasing the spleens with the plunger of a 2-ml syringe. Red cells were eliminated by incubating the cells for 1 min at 4°C in a solution referred to as ACK (NH4Cl 0.83%, 10 mM Tris buffer, pH 7.4). After two washing steps, the pellets were resuspended in RPMI 1640. Cell number and viability were determined by counting trypan blue excluding cells using a Neubauer hemocytometer.
Flow cytometry analysis
The dilution or suspension buffer for flow cytometry experiments was HBSS. Phosphatidylserine exposure was evidenced using annexin VFITC added at a final concentration of 140 nM. Incubation at room temperature was allowed to proceed for 10 min in the dark before data acquisition. Samples were analyzed using a FACScan flow cytometer (Becton Dickinson, San Jose, CA). The sheath fluid was Isoton II balanced electrolyte solution (Coulter, Krefeld, Germany). Suspensions to be analyzed contained 5 x 105 cells/ml. Data acquisition, 10,000 events in each case, and analysis were conducted using CELLQuest software (Becton Dickinson). The forward light scatter setting was E-01 when cells and microparticles were analyzed and E00 when cells only were analyzed.
Determination of hypodiploid DNA
After the different treatments, cells were harvested and numbered. Concentration was adjusted to 5 x 105 cells/ml of 70% ethanol solution in H2O, and fixation was allowed to proceed during at least 1 h at 4°C. Cells were washed once in HBSS before resuspension in a solution containing type I-A RNase A (0.5 mg/ml) in HBSS and were incubated for 10 min at 37°C. Propidium iodide was then added at a final concentration of 0.1 mg/ml. Samples were allowed to stand another 15 min in the dark at room temperature before flow cytometry analysis.
Antigenic capture, characterization of released membrane particles, and prothrombinase assay
Particles were captured from the supernatant, after centrifugation of the cell suspension at 600 x g during 10 min at room temperature, using appropriate biotinylated Abs insolubilized onto 96-well streptavidin-coated microtitration plates, and were estimated through their phosphatidylserine content determined by prothrombinase assay as previously described (10). The only difference was that to remove annexin V from bound microparticles, three washing steps were performed in the presence of EDTA (0.5 mM in TBS, consisting of 50 mM Tris buffer, pH 7.5, containing 120 mM NaCl and 2.7 mM KCl), followed by three washing steps in the presence of CaCl2 (1 mM in TBS) before prothrombinase assay. Phosphatidylserine is the rate-limiting parameter of the reaction. The assay is sensitive enough to allow the detection of minute amounts of generated thrombin corresponding to a minimum value of 125 pM catalytic phosphatidylserine. Background values obtained with insolubilized irrelevant IgGs were subtracted from those measured with mAbs. Linear absorbance changes were recorded at 405 nm using a microtitration plate reader equipped with a kinetics software. Results were expressed as nanomolar phosphatidylserine equivalent by reference to a standard curve constructed by using liposomes of defined composition and of known concentration. The liposomes containing 33% phosphatidylserine and 67% phosphatidylcholine (mol/mol) were prepared and observed by electron microscopy according to Pigault et al. (20). Background values obtained in the absence of biotinylated mAbs never exceeded 0.5 nM phosphatidylserine equivalent, even in samples with the highest particle content, and were subtracted from all the data presented in this study.
Western blotting
After incubation with apoptosis-inducing agent, cells were sedimented, washed once with HBSS and solubilized in lysis buffer consisting of 50 mM Tris buffer containing 8 mM MgCl2, 5 mM EGTA, 0.5 mM EDTA, 10 µg/ml leupeptin, 10 µg/ml pepstatin, 10 µg/ml aprotinin, 1 mM PMSF, 250 mM NaCl, and 1% (v/v) Triton X-100, adjusted to pH 7.5. Samples containing 15–20 µg protein (Peterson’s protein assay kit; Sigma) were separated on 10% SDS-PAGE (31). Separated proteins were then blotted onto Protran nitrocellulose membrane (Schleicher & Schuell, Dassel, Germany). Blots were probed with mAb to CPP32 and developed with the appropriate HRP-conjugated secondary Ab. Bound Abs were detected by chemiluminescence (Pierce, Rockford, IL).
Caspase-3 activity assays
A caspase-3 cellular activity enzyme assay kit was used according to the manufacturer’s instructions adapted to the use of CEM T cells. Briefly, after incubation with an apoptosis-inducing agent, cells were sedimented, washed once with HBSS, and resuspended to 20 x 106 cells/ml in ice-cold lysis buffer containing 0.03% 3-((3-cholamidopropyl)dimethylammonio)-1-propanesulfonic acid. After lysis, 20 µl of each cell extract was added to the assay buffer, incubated for 10 min at 37°C before the addition of 10 µl DEVD-p-nitroanilide (the final volume of each reaction was 100 µl). Linear absorbance changes were recorded at 405 nm using a microtitration plate reader equipped with a kinetics software. No endogenous inhibitor could be detected in cell extracts incubated with known amounts of purified caspase-3. Furthermore, in each sample, it was verified that the measured caspase-3 activity was totally inhibited by Asp-Glu-Val-Asp-aldehyde (DEVD-CHO), a potent caspase-3 reversible inhibitor.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Two agents, VP-16 and actinomycin D, inhibitors of topoisomerase II and DNA-primed RNA polymerase, respectively, were used to induce apoptosis in CEM T and HL-60 cells. The CEM T cells were treated with 5 µM VP-16 or with 0.3 µg/ml actinomycin D during 18 h. Under these conditions, the content of hypodiploid DNA in CEM T cells was 31 ± 2% (mean ± SEM) after treatment with either VP-16 (n = 17) or actinomycin D (n = 13). Similar results were obtained with HL-60 cells treated during 6 h with 8 µM VP-16 or with 0.3 µg/ml actinomycin D (not shown).
Phosphatidylserine externalization by apoptotic cells was revealed by
flow cytometry using annexin VFITC as a probe. A large
majority of untreated cells (
95%) showed a low affinity for annexin
VFITC (mean fluorescence intensity = 5 ± 1
arbitrary units, n = 5); the remaining 5%,
corresponding to basal cell death, had a mean fluorescence intensity in
the range indicated below for treated cells. After treatment with VP-16
or actinomycin D, two cell populations were clearly evidenced, one with
a low labeling similar to that of untreated cells and corresponding to
the normal growing population, the other one highly labeled (mean
fluorescence intensity = 516 ± 55 arbitrary units for
VP-16-treated CEM T cells, n = 5; and 421 ± 50
arbitrary units for actinomycin D-treated CEM T cells,
n = 5) and mainly containing apoptotic cells (4). Other
characteristic features of apoptosis, including cleavage of DNA into
nucleosomal multiples of 200 bp visualized as a ladder (not shown),
dramatic plasma membrane blebbing (as assessed by optical microscopy;
not shown), and collapse of cells into numerous vesicles (as observed
by flow cytometry; Table I
) were observed
in treated cells. When examined by flow cytometry, released particles
had light scatter parameters comparable to those of liposomes of mean
diameter of 0.15 µm with extremes at 0.03 and 0.3 µm (20). All
these observations confirmed the drastic changes of the plasma membrane
organization that are hallmarks of apoptotic cells. The results
obtained with actinomycin D as well as with HL-60 cells (not shown)
were comparable with those shown for VP-16-treated CEM T cells.
|
To examine the possible effect of annexin V on the apoptotic
process, VP-16 or actinomycin D-treated cells were maintained in the
presence or absence of annexin V during the whole course of the 18-h
apoptosis-inducing treatment. Annexin V counteracted apoptosis induced
by 5 µM VP-16 in a dose-dependent manner with a plateau,
corresponding to 40–45% inhibition at 18 h, at about 30 µg/ml
annexin V (850 nM) (Fig. 1). The
binding of annexin V to phosphatidylserine is a relatively rapid
reaction—about 20 min is sufficient to reach a plateau (15). Hence, it
is not a rate-limiting step with respect to the apoptotic process,
which requires several hours. However, the addition of annexin V after
the first 2 h of treatment resulted in a progressive loss of
effect. Comparable results were obtained with actinomycin D and with
HL-60 cells, but at 6 h in this latter case. To verify the binding
of annexin V, the apoptosis-delaying effect of annexin
VFITC was also tested at 20 and 30 µg/ml under the same
conditions; identical results were obtained with a binding of annexin
VFITC to apoptotic cells reflected by a mean fluorescence
intensity in the range of 500 arbitrary units. In the presence of
annexin V, the release of particles was highly to totally reduced in
the population of VP-16-treated cells as observed by photonic
microscopy and by flow cytometry (Table I
). The events recorded in the
microparticle gate were considerably reduced in the presence of annexin
V, and those recorded in the cell gate increased accordingly.
|
. At
20 µg/ml (560 nM), annexin V totally inhibited the release of
apoptotic particles bearing either of CD4 or PrPc Ag, and
at half this concentration (10 µg/ml) the inhibition reached 90%
(not shown).
|
The permeable peptidic agent z-VAD.fmk is an irreversible
inhibitor of several members of the caspase family of proteases, which
become activated during the apoptotic cascade. It prevents DNA
digestion, intracellular acidification, increased membrane
permeability, and cell shrinkage (33, 34). When present during the
18 h proapoptotic treatment, z-VAD.fmk inhibited the generation of
hypodiploid DNA normally induced by 5 µM VP-16 (Fig. 2) or 0.3 µg/ml actinomycin D (not
shown) in CEM T cells. As shown in Fig. 2, the inhibitory effect of
z-VAD.fmk measured by propidium iodide assay was dose-dependent and
reached 100% at 100 µM. As indicated in Table II, at 10 µM this
inhibitor had a limited effect on the shedding of CD4+
particles and no effect at all on that of PrPc+ ones.
DEVD-CHO, another caspase inhibitor more specifically directed to
caspases 8 and 3, elicited a reduction of 50% in apoptotic cell
population at a concentration of 50 nM, but had no effect on the
release of membrane particles. Annexin V enhanced the inhibitory effect
of z-VAD.fmk on VP-16-induced apoptosis as shown in Fig. 2. Thus
the external effect of annexin V is additive to that of an
intracellular inhibitor of apoptosis. This finding was verified by DNA
ladder analysis (not shown).
|
Caspase-3 is a member of the caspase family of proteases (35) for
which the processing to the active form is inhibited by z-VAD.fmk.
Because this peptide is a strong inhibitor of apoptosis-associated
phosphatidylserine externalization (5), it was assessed whether annexin
V also interfered in the caspase cascade. Caspase-3 underwent
proteolytic processing during VP-16 treatment of CEM T cells, which was
clearly reduced in the presence of 30 µg/ml (850 nM) annexin V, as
qualitatively shown in Fig. 3. As
expected, z-VAD.fmk was efficiently protective against caspase-3
proteolysis. This processing was concomitant with an increased
caspase-3 activity in VP-16-treated CEM T cells, and with the reduction
of this activity of about 50% and 80% in the presence of annexin V
and z-VAD.fmk, respectively (Table III).
|
|
HN-2 or nitrogen mustard, a bifunctional alkylating agent still
used in cancer chemotherapy, induced apoptosis in cultured lymphocytes
at submicromolar concentrations (36) and a reduction of splenic weight
and cellularity in mice (37). To examine the possible effect of annexin
V on the apoptotic process in vivo, mice were treated by HN-2 in the
presence or absence of annexin V. HN-2 induced a significant reduction
of spleen weight, which was counteracted in annexin V-injected animals
(Fig. 4). Comparable results were
observed for cellularity (data not shown). It has to be noted that
annexin V alone had no noticeable effect on spleen weight (Fig. 4a) or cellularity. These results suggest that annexin V
exerts a protective effect in vivo.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
When present during the execution of induced death programs, annexin V
elicited a dose-response apparent inhibition of apoptosis. A maximum of
about 45% reduction of events with hypodiploid DNA was achieved, at
18 h, at 30 µg/ml (850 nM) annexin V. Such a concentration
appears compatible with physiologic values in certain organ
compartments as the intracellular concentration of annexin V is
certainly much higher than 1 µM in the endothelium (24, 41) and
abundant annexin V has been evidenced at the surface of several types
of cells (24, 25). Under our experimental conditions, CEM T cells did
not contribute for >0.2% to the concentration of total annexin V,
owing to the low amount in leukocytes (24, 41). The external action of
annexin V was additive to the intracellular antagonist effect of
z-VAD.fmk, a potent irreversible inhibitor of caspase-1-like proteases,
showing at the same time that annexin V did not significantly interfere
in the cell permeability of relatively small molecules. The
interference of annexin V in apoptosis was further demonstrated by a
decreased proteolytic processing of CPP32 to its active form. None of
the cells exposed to VP-16 or actinomycin D could be rescued by annexin
V, z-VAD.fmk, or a combination of both agents after a 24-h culture
period following the 18-h proapoptotic treatment. This suggests that
annexin V effect takes place between the irreversible mitochondrial
collapse (42) and the caspase checkpoint.
The obvious question arising from the above observation is: how can
annexin V counteract apoptosis, resulting in reduced exposure of
phosphatidylserine to which it precisely binds? The same holds true in
respect of experiments performed in the presence of z-VAD.fmk. The
answer takes into account the complex Ca2+-mediated
interaction between annexin V and phosphatidylserine; in this respect,
the X-vivo 15 synthetic culture medium used in our experiments contains
1.8 mM Ca2+. Annexin V has been shown to bind to
unstimulated cells, but the number of binding sites is considerably
lower when compared with stimulated cells (15). Several targets can be
considered, but a proportion of phosphatidylethanolamine is accessible
at the surface of a variety of unstimulated cells (7), which makes it a
likely candidate allowing the binding of annexin V, with a half-maximal
Ca2+ concentration requisite of 0.9 mM (14). This
suggests that bound annexin V molecules can be viewed as latent
"nucleation" sites for the formation of two-dimensional arrays as
soon as phosphatidylserine becomes exposed, allowing the binding of
more annexin V molecules, even when the proportion of
phosphatidylserine is low (20) in the early stages of the death
program. The process must reach an asymptotic limit because the
interaction of annexin V with phosphatidylserine has a delaying
consequence on the progress of the cell death program, precisely
resulting in a reduction of phosphatidylserine exposure. Such a mutual
neutralization explains why the inhibitory effect of annexin V cannot
be >50% (Fig. 1), and this is also true for the generation of
caspase-3 activity (Table III).
In addition to an apoptosis-delaying effect, extracellular annexin V considerably reduced the degree of membrane particle shedding regardless of their Ag content. But cell permeable z-VAD.fmk had a limited action on the release of CD4+ membrane fragments and no effect at all on PrPc+ ones, at least below 10–15 µM because at higher concentrations it might destabilize the membrane, leading to artifactual vesiculation. Interestingly, other investigators have reported the absence of morphological change and surface blebbing in apoptotic cells treated with peptidic caspase inhibitors (5, 6, 33, 34, 43). The differential exportation of membrane proteins is suggestive of cluster formation before shedding, which may have consequences on the dissemination of particular Ags. Because CD4 and PrPc do not have the same attachment to the membrane, it remains to be established whether this may account for discriminating shedding, especially for PrPc, which could be preferentially associated with sphingolipid-cholesterol rafts as being GPI-anchored (44). CD4+ particles, believed to reflect ongoing apoptosis of CD4+ T cells, have been detected in peripheral blood samples from certain HIV-1-infected subjects (10). An intriguing issue is the part of apoptotic membrane particles in the possible convoying and addressing of PrPc and its pathological misfolded form, scrapie prion protein, in the development of prion diseases. The converse points to the significance of the inhibition of the release of such particles by annexin V.
Hence, if an intracellular proapoptotic cascade of events leads to the external remodeling of the plasma membrane, symmetrically, an external constraint can delay the activation in the internal cascade. Although upstream coupling element(s), possibly of mitochondrial origin (45), remain(s) to be identified, these observations raise the question of a relevant role of annexin V in the process of apoptosis. In this respect, its discussed ability to generate Ca2+ channels (46, 47) may be considered in the light of a possible impaired Ca2+-mediated phospholipid scrambling (48). Despite the lack of a signal peptide, this canonical member of the annexin family (26, 47) may well be translocated to the exoplasmic leaflet attached to phosphatidylserine during the loss of asymmetry of the plasma membrane. Once externalized from cells that contain a substantial proportion, annexin V has been proposed to fulfill an anticoagulant function by neutralizing the procoagulant potential of phosphatidylserine (24, 25). More recently, an apoptosis-inducing activity was localized in the fraction of annexin V-binding Abs from 10 patients with lupus anticoagulant. Furthermore, annexin V neutralized the apoptotic potential of such Abs, but these authors did not consider that annexin V could interfere in the apoptotic process itself (49). It is also of interest to point to the modulation of cell growth or death by other members of the annexin family when exogeneously added (50, 51, 52, 53).
When injected in vivo, annexin V concentrates preferentially in kidney and spleen (54), which led us to examine its effect on spleen weight loss in mice treated by HN-2, a potent alkylating agent previously shown to induce lymphocyte apoptosis (36). Annexin V appeared indeed protective against spleen cell loss, in agreement with its ability to delay apoptosis in vitro when present in the extracellular medium as suggested from the above results.
These in vitro and in vivo observations are suggestive of a role of annexin V in the control of cell death, which deserves further attention.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Jean-Marie Freyssinet, Institut d’Hématologie et d’Immunologie, Faculté de Médecine, 4, rue Kirschleger, 67085 Strasbourg, France. E-mail address:
3 Abbreviations used in this paper: PrPc, cellular prion protein; GPI, glycosylphosphatidylinositol; VP-16, etoposide; z-VAD.fmk, z-Val-Ala-Asp.fluoromethyl ketone; DEVD-CHO, Asp-Glu-Val-Asp-aldehyde.
Received for publication July 9, 1998. Accepted for publication February 17, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(VAC) to planar phospholipid bilayers. J. Biol. Chem. 265:4923.
-irradiated mice. Anticancer Res. 18:3289.[Medline]
This article has been cited by other articles:
|
|
 |
|
  E. Camors, V. Monceau, and D. Charlemagne Annexins and Ca2+ handling in the heart Cardiovasc Res, March 1, 2005; 65(4): 793 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Hugel, M. C. Martinez, C. Kunzelmann, and J.-M. Freyssinet Membrane Microparticles: Two Sides of the Coin Physiology, February 1, 2005; 20(1): 22 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Kenis, H. van Genderen, A. Bennaghmouch, H. A. Rinia, P. Frederik, J. Narula, L. Hofstra, and C. P. M. Reutelingsperger Cell Surface-expressed Phosphatidylserine and Annexin A5 Open a Novel Portal of Cell Entry J. Biol. Chem., December 10, 2004; 279(50): 52623 - 52629. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Martin, A. Tesse, B. Hugel, M. C. Martinez, O. Morel, J.-M. Freyssinet, and R. Andriantsitohaina Shed Membrane Particles From T Lymphocytes Impair Endothelial Function and Regulate Endothelial Protein Expression Circulation, April 6, 2004; 109(13): 1653 - 1659. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Gerke and S. E. Moss Annexins: From Structure to Function Physiol Rev, April 1, 2002; 82(2): 331 - 371. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
