Phosphatidylserine expression on apoptotic lymphocytes of Xenopus laevis, the South African clawed toad, as a signal for macrophage recognition

Developmental and Comparative Immunology 24 (2000) 641±652 www.elsevier.com/locate/devcompimm

Phosphatidylserine expression on apoptotic lymphocytes of Xenopus laevis, the South African clawed toad, as a signal for macrophage recognition M. Sherleen Nera a, Gretchen Vanderbeek a, Rachel O. Johnson a, Laurens N. Ruben a,*, Richard H. Clothier b a Department of Biology, Reed College, 3203 SE Woodstock Blvd, Portland OR 97202-8199, USA School of Biological Sciences, University of Nottingham, Queens Medical Centre, Clifton Blvd, Nottingham NG7 2UH, UK

b

Received 29 October 1999; accepted 10 February 2000

Abstract In¯ammation is avoided in apoptosis by early removal of dying cells by macrophages (Més). In mammalian cells, an early aspect of apoptosis is the translocation of phosphatidylserine (PS) from the inner lea¯et of the cell membrane to the surface. PS recognition can serve as a signal for triggering removal of dying cells. PS expression on splenocytes and thymocytes of Xenopus laevis was quanti®ed using FITC-Annexin and ¯ow cytometry following exposure in vitro to several known apoptogens for this species. All apoptogens used induced PS expression. Dose dependency and the kinetics of PS expression following exposure to the calcium ionophore, A23187, were also examined. Peritoneal exudate cells (PEC's) were cultured with A23187-treated thymocytes to test Mé capacity for recognition of PS. Mé binding to apoptotic thymocytes was reduced following exposure of PEC's to a water soluble analogue of PS, phospho-L-serine. The presence of a phagocytic PS-dependent recognition system in amphibia is supportive of the evolutionary conservation of this function in mammals that is crucial in limiting in¯ammation induced by dying cells. 7 2000 Elsevier Science Ltd. All rights reserved. Keywords: Apoptosis; Phosphatidylserine; Macrophage recognition; Amphibian

1. Introduction Abbreviations: PS, phosphatidylserine; PI, propidium iodide; Mé, macrophage; PECs, Peritioneal exudate cells; DEX, dexamethasone; CER, ceramide; CHX, cycloheximide; PMA, phorbol 12-myristate 13-acetate; PHA, phytahemagglutinin. * Corresponding author. Tel.: +1-503-777-7276; fax: +1503-777-7773. E-mail address: [email protected] (L.N. Ruben).

Apoptosis, a form of programmed cell death, plays an important role in many physiological phenomena, including homeostasis and the elimination of unwanted or damaged immune cells in normal adult tissues. The elimination of

0145-305X/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 5 - 3 0 5 X ( 0 0 ) 0 0 0 2 0 - 3

642

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

unwanted cells by apoptosis has been reported in every multicellular animal having phagocytic cells that function to remove apoptotic cells by recognizing and engul®ng them [1]. Apoptosis is present in the tissues of many species [2]. Macrophages and other phagocytic cells selectively remove individual dead and dying cells from tissues and organs, thus in¯ammation in the region of the dying cells is avoided [3]. Loss of plasma membrane phospholipid asymmetry and the subsequent surface expression of phosphatidylserine (PS) occurs early in apoptosis. PS is an aminophospholipid normally found in the inner lea¯et of the plasma membrane [4]. PS translocation to the outer cell membrane lea¯et acts as a recognition signal for neighboring macrophages to engulf and degrade the dying cells [5]. The phosphoserine headgroup may serve as the binding site for macrophage recognition [3]. Membrane phospholipid asymmetry appears to be ubiquitous in mammalian cells [6]. PS-dependent recognition is found with both human and murine macrophages, while phagocytosis is inhibited in the absence of the naturally occurring L-isomer of PS [7]. The physiological importance of prompt phagocytosis is obvious. If apoptotic cells are allowed to progress into secondary necrosis, their cell membranes become permeable and they can leak lysosomal enzymes into their local environment causing in¯ammation [8]. However, translocation of PS is not unique to apoptosis, it also occurs during primary cell necrosis [9]. The di€erence between these two forms of cell death is that during the initial stages of apoptosis, the cell membrane remains intact, while with necrosis the cell swells and bursts quickly. Cells in late stage of apoptosis permit propidium iodide (PI) entry and DNA binding, as a result of having belatedly lost membrane integrity. The appearance of PS on the cell surface occurs prior to alterations in membrane permeability in all human and murine cell types examined [9]. In vitro studies using the calcium ionophore, A23187, show increased expression of PS on the external surface of plasma membranes of rat lymphocytes [10]. PS expression is assayed using FITC-Annexin V ¯uorescence with ¯ow

cytometry. Annexin V is a phospholipid-binding protein with a high anity for externalized PS [11]. Adult Xenopus splenocyte apoptosis has been previously examined in vitro following treatment with the calcium ionophore, A23187, the phorbol diester, phorbol 12-myristate 13-acetate (PMA), and the synthetic glucocorticoid, dexamethasone (DEX) [12]. Using the Tunel method, in vivo apoptosis of Xenopus has also been quanti®ed over time after T and B cell-speci®c lectin injections [13]. Additionally, apoptosis in thymocytes of adult [14] and developing [15] Xenopus has been studied following exposure in vitro to an antigen, lectin or DEX. DNA extractions were performed selectively in these earlier tests. The presence of DNA ladders on agarose gels is used to characterize apoptosis. All reagents examined here serve e€ectively as apoptogens of lymphocytes of Xenopus, although some a€ect apoptosis in thymocytes or splenocytes di€erentially. Here we ask whether expression of PS, as an early step in apoptosis, is limited to mammalian cells within the vertebrates. Moreover, if PS is expressed on apoptotic cells of Xenopus laevis, the South African clawed toad, does it serve as a recognition signal for macrophages? Here, A23187 has been utilized, because of the known relationship between calcium ¯ux and function of the translocases responsible for phospholipid asymmetry in the plasma membrane of mammalian cells [16] and its consistent ecacy as an apoptogen with Xenopus splenic lymphocytes [12]. All prior apoptotic studies in vitro with Xenopus lymphocytes have used a propidium iodide incorporation assay, which identi®es late stages in the apoptotic process. Here, we have used FITC-Annexin V (green)/PI (red). ¯uorescence to test its capacity to provide detection of early apoptosis in cells of this species. 2. Materials and methods 2.1. Animals Adult X. laevis, South African clawed toads, produced from laboratory breedings were used.

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

They were raised in ®ltered and dechlorinated tap water, kept at 23 2 28C in a light controlled room and fed twice each week with frog ``brittle'' (NASCO, Ft. Atkinson, WI). Their care was in accord with institutional and NIH standards for experimental vertebrate animals. 2.2. Cell preparation and culture The animals were anesthetized in 1:500 methyl n-ethyl-m-aminobenzoate (Fisher Scienti®c, Pittsburgh, PA) in distilled water for 3±5 min. The splenocytes or thymocytes were mechanically dissociated into cell suspensions and cultured at 5  105 cells/well in round-bottomed 96-well plates at 23 2 28C in 200 ml 60% Liebovitz (L-15) complete medium (GIBCO, Long Island, NY), 30% distilled water and 10% fetal calf serum plus penicillin and streptomycin. The cells were cultured for various times up to 20 h with the following apoptogens: the calcium ionophore, A23187 (Sigma, St. Louis, MO) in dimethyl sulfoxide (DMSO) in L-15 culture medium at concentrations of 50, 100 and 500 ng/ml. Additionally, dosages of from 1  10 ÿ4 to 1  10 ÿ8 M water soluble dexamethasone (DEX, Sigma, St. Louis, MO) diluted in L-15 media, phytahemagglutinin (PHA-P, Sigma, St. Louis, MO) used as 1 mg/ml/L-15 culture media, phorbol 12-myristate 13-acetate (PMA, Sigma, St. Louis, MO) used as 150 ng/ml/L-15 media, ceramide-1-phosphate (Sigma, St. Louis, MO) as a 3:3  103 mM stock diluted in DMSO, 10 mg/ml cycloheximide (CHX, Sigma, St. Louis, MO), and ®nally, 8E4, a monoclonal anti-Xenopus IgM developed in this laboratory from a mouse myeloma cell line, diluted 1:8 and 1:16 in L-15. Controls were cultured in L-15 complete medium or in one part DMSO/106 parts of L-15, the highest concentration of DMSO that would be in any culture. As no di€erences were observed between the L-15 and DMSO controls, the untreated L-15 controls were used in the data presentation that follows. While it was subsequently learned that 3 h was the time of maximal PS expression following exposure to A23187, this was not known when Annexin V ¯uoresence was assayed following a variety of other apoptogens. The 10-h. cul-

643

ture period was chosen for these tests, because it had been an e€ective culture time for earlier apoptotic analyses when propidium iodide incorporation had been used as the assay [14,15], (Beachdel et al., unpublished results). No attempt was made here to study the time-course of PS expression in these experiments. Thymocytes to be used in tests of macrophage recognition and removal of apoptotic cells, were prepared and exposed to 500 ng/ml A23187 for 4 h as described above. A23187-treated cells were washed thoroughly to remove any residual ionophore, before they were added to a thioglycollate-activated macrophage population and cultured together for 2 h. The water-soluble PS analogue, phospho-L-serine, used as an inhibitor of macrophage recognition, was dissolved into L15 at a concentration of 30 mM, protected from light and stored at 48C. 2.3. Isolation and culture of peritoneal macrophages from adult Xenopus laevis Adult X. laevis were injected i.p. with 1 ml of aged 4% Brewers' thioglycollate (generously provided by Dr. Valerie Fadok, National Jewish Medical Center, Denver, CO). All purchased thioglycollate (DIFCO, Detroit, MI) was aged 30 days before use. After 4 days, peritoneal exudate cells (PEC's) were collected by a lavage with L15 complete medium (GIBCO, Long Island, NY). Twenty-four well plates were seeded with 1  105 PEC's/well. Lavage without prior injection of thioglycollate produced too few PEC's to work with. Activated adherent PEC's were cultured for 4 h in L-15 complete medium at 23 2 28C. Non-adherent cells were aspirated and each well rinsed three times with culture medium. Some adherent PEC's were placed on ice to allow their release from adherence and transfered to slides to be stained for neutral esterase (Kit 91-C, Sigma, St. Louis, MO) to survey the population for macrophages. The thioglycollate-activated macrophages were also identi®ed by India ink uptake. Cultured PEC's were incubated overnight with 500 ml per well of culture medium in a 24-well plate with India ink. The cells were rinsed three times with chilled L-15 and evaluated for

644

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

Fig. 1. (Caption opposite)

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

uptake of the ink with light microscopy. A minimum of three runs with four animals in each run were used. 2.4. Annexin V assay Apoptosis was monitored using an ``Annexin V'' kit (Oncogene Res. Prod., Cambridge, MA). The splenocyte or thymocyte suspensions were centrifuged for 5 min at 2000 rpm. Five hundred ml aliquots of fresh media, 10 ml of media binding bu€er, and 0.75 ml FITC-Annexin V were added to resuspend the pelleted cells. The cells were incubated in the dark at room temperature for 30 min. After incubation, they were centrifuged at 2000 rpm for 5 min and resuspended in 500 ml of 1  Media Binding Reagent with 10 ml of 30 mg/ ml propidium iodide. They were kept on ice in the dark until analysis with ¯ow cytometry at the Oregon Health Sciences University, Portland, OR. See [17] for a protocol used with mammalian cells. 2.5. Flow cytometry Aliquots of cells were analyzed on a Becton Dickinson ``FACScan'' ¯ow cytometer (Becton Dickinson, San Jose, CA). Data analysis was done using ``CELL QUEST'' software (Becton Dickinson, San Jose, CA). Five thousand cells were analyzed in each sample examined. Four animals were used as cell donors for each of three runs. The Annexin V-FITC green signal was detected

645

at 518 nm, while the red propidium iodide signal was detected at 620 nm. The lymphocyte population was de®ned by their intermediate size in forward (FSC-H) and side (SSC-H) scatter dot plots and electronically ``gated''. Four sub-populations of lymphocytes are obtained in the dot plot (Fig. 1). Viable cells, negative for both Annexin V and PI ¯uorescence, are distributed in the lower left (LL) quadrant. Annexin V positive only cells, with PS on their cell surface, are located in the lower right (LR) quadrant, while Annexin V and PI ¯uorescent cells, are seen in the upper right (UR) quadrant of the plot. Small numbers of PI only ¯uorescent necrotic cells are included in the upper left (UL) quadrant. Positive ¯uorescence was always assayed in comparison with appropriate control, untreated samples. All experiments were conducted at room temperature. A representative sample of the apoptotic data produced is shown in comparative dot plots for untreated and ionophore-treated splenocytes over time in Fig. 1. The ®gure presents the raw, untransformed data in the upper right corner of each quadrant. They represent the percentages of the 5000 cells sampled by ¯ow cytometry that fell into each quadrant based on their ¯uorescence or lack thereof. 2.6. Statistical analysis All experimental samples were compared to the control samples from the same cell population using the natural log of the odds ratio

Fig. 1. Representative ¯ow cytometric dot plots of a splenocyte population with FITC-Annexin V (green) and propidium iodide (red) ¯uorescence 1, 2 and 3 h following exposure to 500 ng/ml of A23187. Four cellular subpopulations are created in each dot plot as a consequence of Annexin V and PI-¯uorescence. LL (the lower left quadrant of each dot plot) has the ¯uorescent negative or viable cells. LR (the lower right quadrant) shows splenocytes in early apoptosis. They are FITC-Annexin V positive, but not PI positive. They have lost membrane asymmetry and have expressed phosphatidylserine (PS) on cell surfaces. UR (the upper right quadrant) displays those cells that have lost both membrane asymmetry and integrity. They are positive for both Annexin V and PI uptake. PI binds DNA. UL (the upper left quadrant) includes the small number of necrotic cells and debris with PI uptake only. The plots are presented to illustrate what the raw data look like before transformation to Log Odds Ratios (LORs). Controls for each of the three time periods are on the left (A±C), while the e€ects of A23187 exposure appear on the right (D±F). The percent of the total population found in each quadrant is displayed in the upper right corner of each quadrant. In order to see what they represent in LORs, the LORs for the Annexin V positive population only, are to the right of (D±F). Early apoptosis (Annexin V positive, the lower right (LR) quadrant), increases quickly following treatment with the apoptogen. Late stages of apoptosis (Annexin V and PI positive, upper right (UR) quadrant) increase more gradually.

646

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

Fig. 2. The time-course of expression of phosphatidylserine (PS) with FITC-Annexin V ¯uorescence, on Xenopus splenocytes (open circles) plotted as Log Odds Ratios (LORs). The dotted line in each quadrant represents the control (zero) level for each time point. Apoptosis was induced with three concentrations of A23187 at di€erent times over 20 h. (A) shows the time-course of early apoptosis (FITC-Annexin V binding) on splenocytes treated with 50 ng/ml A23187 in vitro. (B) is of splenocytes exposed to 100 ng/ml, while (C) has the kinetics of PS expression for cells treated with 500 ng/ml of the ionophore. Increases in LORs re¯ect increases in percentages of cells with Annexin V binding. D±2F show the corresponding time-courses for Annexin V and PI positive ¯uorescence (solid squares) and for Annexin V and PI negative or viable cells (triangles) of adult splenocytes after induction of apoptosis. Double positive cells are those that have lost their membrane lipid asymmetry, as well as membrane integrity. They are in the late stages of apoptosis.  = statistical signi®cance when compared with untreated-controls …p ˆ R0:05 by paired Student ttests). N ˆ 12 (four animals in each of three runs). Five thousand cells were analyzed for each data point in each run.

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

3. Results

(LOR): LOR ln

647

P1 =1 ÿ P1 P0 =1 ÿ P0

P1 ˆ % cells of treated samples P0 ˆ % cells of untreated …control † samples The advantage provided by transforming the raw data to LORs for statistical evaluation is that it provides zeroing of all control levels of apoptosis. Individual animals have di€erent natural baseline levels of apoptosis that change over time [12]. See Fig. 1 for an illustration of this and of the e€ect of the transformation to LOR when Annexin V ¯uorescence is assayed. The LOR levels related to the percentages of green ¯uorescent cells appear to the right of the appropriate quadrants. Additionally, unlike standard normalization of percentages of increases or decreases of controls, this transformation recognize that the same percentage di€erences are not always equivalent. For instance, the di€erence between 30% and 50%, is greater than that between 3% and 5%. Paired student t-tests were employed to determine whether the di€erent treatments or doses were signi®cantly di€erent from controls and each other. 2.7. In vitro rosetting assay Peritoneal macrophages, cultured as above, were incubated at 23 2 28C. for 30 min with or without phospho-L-serine. A23187-exposed cells, after three washes in L-15 culture media, were then overlaid in 10-fold excess of the number of macrophages and incubated at 23 2 28C for 2 h. The overlaid monolayers were washed three times with L-15 medium to remove non-adherent cells. The remaining cells were ®xed with citrate± acetone±formaldehyde (CAF) solution: 25 ml citrate solution (Sigma, St. Louis, MO), 65 ml acetone, 8 ml 37% formaldehyde solution. The CAF solution was applied for 30 s and rinsed o€ three times with distilled water. Plates were evaluated for ``rosette'' formation with phase microscopy.

3.1. A23187-treated splenocytes and thymocytes elicit PS expression in a dose- and time-dependent manner Xenopus splenocytes or thymocytes were incubated for up to 20 h with 50, 100 or 500 ng/ml of A23187 and tested with FITC-Annexin V and PI. All concentrations used, produced signi®cant increases in apoptosis …p ˆ R0:05† by 3 h. One hundred …p ˆ 0:0001† and 500 ng/ml …p ˆ 0:001† induced signi®cantly more apoptosis, than did exposure to 50 ng/ml of the ionophore. Expression of PS in apoptotic mammalian cells has been found as early as 1 h after treatment with an ionophore [10]. The data here reveal (Fig. 2(A)) PS expression of the Xenopus splenocyte population over time following 50 ng/ml of A23187 stimulation is signi®cantly increased between 0 and 3 hours …p ˆ 0:0002). Following exposure of 100 ng or 500 ng/ml of A23187 (Fig. 2(B)), increases are signi®cant between 0 and 3, 6, 10, and 15 hours with p values not greater than 0.002, i.e. when 100 or 500 ng/ml of the ionophore are used, the splenocytes exhibit statistically signi®cant increases in PS expression, except at 20 hours, when no further signi®cant increases in early stage apoptosis take place. Unlike the splenocytes, the kinetics of PS expression, using 50 ng/ml of A23187, showed signi®cant increases in thymocytes expressing PS after 3, 6 and 20 hours of culture (Fig. 3(A)) with p values not greater than 0.0003. However, when higher dosages were used, higher levels of apoptosis were stimulated and signi®cant increases in PS expression were observed at more time periods, as was the case with splenocytes (Figs. 2 and 3(A±C)). While the time-course of PS expression in thymocytes was similar to that seen with splenocytes, the principal di€erence was that a higher proportion of thymocytes were still FITC-Annexin V positive after 3 h of induction with 50 ng/ml, the lowest dose tested. This should not be surprising, as thymocytes are largely immature, and therefore, are par-

648

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

Fig. 3. The time-course of expression of phosphatidylserine (PS) with FITC-Annexin V ¯uorescence, on Xenopus thymocytes (open circles) plotted as Log Odds Ratios (LORs). The dotted line in each quadrant represents the control (zero) level for each time point. Apoptosis was induced with three concentrations of A23187 at di€erent times over 20 h. (A) shows the time-course of early apoptosis (FITC-Annexin V binding) on thymocytes treated with 50 ng/ml A23187 in vitro. (B) is of thymocytes exposed to 100 ng/ml, while (C) has the kinetics of PS expression for thymocytes treated with 500 ng/ml of the ionophore. Increases in LORs re¯ect increases in percentages of cells with Annexin V binding. (D±F) show the corresponding time-courses for Annexin V and PI positive ¯uorescence (solid squares) and for Annexin V and PI negative or viable cells (triangles) of adult thymocytes after induction of apoptosis. Double positive cells are those that have lost their membrane lipid asymmetry, as well as membrane integrity. They are in the late stages of apoptosis.  = statistical signi®cance when compared with untreated-controls …p ˆ R0:05 by paired Student t-tests). N ˆ 12 (four animals in each of three runs). Five thousand cells were analyzed for each data point in each run.

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

ticularly susceptible to apoptosis. The dotted line in the center of Figs. 2(A±F) and 3(A±F) is representative of the control level of apoptosis (zeroed by transformation into LORs) to which the level at each experimental time point is to be compared. Thus, a consistent rise in the percentage of A23187-treated cells showing PS on their cell surface is observed between 0 and 3 hours, regardless of the dose or the cell type is observed.

649

Moreover, a decline in the proportion of apoptotic cells expressing PS is seen between 3 and 6 hours. In order to explore the basis for this decrease in the percentage of PS-expressing cells, the progress of splenocytes and thymocytes into the late stages of apoptosis was explored, i.e. those positive for both FITC-Annexin V (green; AN+) and PI (red; PI+) ¯uorescence were examined (the cells in the upper right quadrants seen in Fig. 1). The plotted LOR's in Figs. 2(D±F) and 3(D±F) show a temporal relationship between the proportion of splenocytes that are ANÿ, PIÿ and double positives (AN+, PI+). The group exposed to 500 ng/ml A23187 treatment show this transition most clearly as cells that were ANÿ, PIÿ become AN+, PI+ over time, i.e. there was a comparable loss in the viable cell population as increasing numbers of late apoptotic cells developed in culture. 3.2. Which apoptogens elicit cell surface expression of PS?

Fig. 4. Xenopus splenocytes were incubated with a variety of unrelated apoptogens for 10 h, e.g. dexamethasone (DEX), cycloheximide (CHX), ceramide (CER), phorbol-12-myristate 13-acetate (PMA), phytahemagglutinin (PHA), and a monoclonal rabbit-anti-Xenopus IgM. The induction of apoptosis is shown following treatment with the most e€ective dose tested in preliminary studies. Increase in mean Log Odds Ratios (LORs) represent increases in the percentages of cells with FITC-Annexin V ¯uorescence, and hence, they re¯ect increases in the ratios of cells with plasma membranes expressing phosphatidylserine.  = statistical signi®cance when compared with untreated control cultures within a cell type …p ˆ R0:05, paired Student t-tests). N ˆ 12 (four animals per run; three runs were made for each condition). Five thousand cells are analyzed for each data point.

Xenopus splenocytes were cultured for 10 h at 23 2 28C with each one of the following apoptogens independently to determine whether other reagents, known from previous studies to a€ect apoptosis in lymphocytes of this species in the time period selected (Beachdel et al., unpublished results), are able to elicit cell surface expression of PS. The apoptogens were DEX, CHX, PMA, CER, PHA, and an anti-Xenopus IgM mAb. While several doses of each has been used, only data with the maximally e€ective dose tested has been shown in Fig. 4. No attempt was made to study the time-course for PS expression with this group of apoptogen tests. A water soluble form of DEX, a synthetic glucocorticoid, when incubated at 10ÿ6 M with splenocytes for 10 h, will promote signi®cant increases in PS expression in Xenopus splenocytes (Fig. 4). Cycloheximide, a non-speci®c protein inhibitor, at 10 mg/ml, also induced a signi®cant increase in PS expression with splenocytes after 10 h of culture (Fig. 4). Phorbol-12-myristate 13acetate (PMA), a cancer promoting reagent in mammals, can function as both a mitogen and an apoptogen with Xenopus splenocytes [13,19].

650

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

Five hundred ng/mL of PMA stimulated signi®cantly enhanced FITC-Annexin V ¯uorescence after 10 h in vitro (Fig. 4). Ceramide, another apoptogen in this species, is a backbone lipid molecule that serves as a precursor for the synthesis of complex sphingolipids. Three hundred and twenty mg/ml enhanced the expression of PS on Xenopus splenocytes after 10 h in vitro (Fig. 4). PHA, is a plant lectin that has the ability to function as a weak mitogen and a strong apoptogen with adult Xenopus splenocytes [18]. One mg/ ml exposure succeeded in stimulating signi®cant PS expression on splenocytes after 10 h in culture (Fig. 4). Finally, a monoclonal rabbit-anti-Xenopus IgM antibody prepared earlier in this laboratory [19] also had been used previously to induce enhanced apoptosis with Xenopus splenocytes. The anti-Xenopus IgM mAb speci®cally binds B cells. A 1:8 dilution stimulated signi®cantly enhanced FITC-Annexin V ¯uorescence in splenocytes after 10 h in vitro (Fig. 4). Thus, Xenopus splenocytes treated with a wide range of apoptogens, produced signi®cant increases in the proportion of cells expressing PS, when compared to their own controls. 3.3. The role of PS in macrophage recognition Ionophore-treated and washed thymocytes were layered over adherent PECs for 2 h and many of the unbound, non-adherent cells were washed o€ so as to make it possible to count individual cells on the previously too crowded surface. The remaining cells were ®xed and evaluated for rosette formation using phase contrast microscopy. Rosettes were counted only when four or more thymocytes bound a central macrophage. A rosetting index, calculated from the average number of rosettes in three randomly selected microscopic ®elds, viewed at a magni®cation of 20 , divided by the total number of cells in that ®eld and multiplied by 100, was used to evaluate the ecacy of phospho-L-serine Ð a water soluble derivative of phosphatidylserine Ð to reduce rosette formation. Replicate wells were visually scored blind by two observers independently. Three separate counts of di€erent ®elds

of view were made of the cells in each well and two runs of the experiment were made with cells from di€erent donor animals. On average, 200± 300 cells were counted in each ®eld. The rosette index averaged 18 in wells with macrophages that had been pre-incubated for 30 min with 30 mM phospho-L-serine, whereas in control wells, not exposed to phospho-L-serine, the rosette index average was 30. Thus, in wells pre-incubated with rosette formation was phospho-L-serine, decreased by around 40%, when compared with controls. 4. Discussion Splenocytes and thymocytes of adult X. laevis express cell surface PS and bind FITC-Annexin V, as assayed by ¯ow cytometry, a procedure used previously with mammalian cells [9,20,21]. Thus, amphibian cells may share with mammalian cells similar mechanisms for translocating PS to the outer lea¯et of the plasma membrane following apoptotic induction [16]. The mechanism by which PS is exposed on the outer lea¯et of the plasma membrane has not yet been explored with amphibian cells. However, that Xenopus cells may also utilize aminophospholipid translocases, as mammalian cells do, is suggested by the rapid induction of PS expression following exposure to A23187 (Fig. 1). The data in Figs. 2 and 3 reveal temporal relationships in Xenopus lymphocytes with respect to early and late stage transitions of apoptosis. Splenocytes and thymocytes respectively, express PS almost immediately after exposure to the apoptogen, the calcium ionophore, A23187 (Figs. 1±3). It should be noted that rises in PS expression seen at zero hour actually represent the time required to perform the assay of slightly less than 1 h. A consistent rise in the proportion of cells expressing PS was observed between 0 and 3 hours of treatment, regardless of the concentration of ionophore used or the cell type treated. The proportion of cells with surface PS subsequently declines, as these cells enter later stages of apoptosis. The rate of splenocytes and thymocytes entering the late stages of apoptosis

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652

increases between 3 and 6 hours. The percentages of Xenopus cells expressing PS in the population appear to be similar to those found for mammalian cells [10]. Unlike the propidium iodide incorporation assay for apoptosis [14,15], which detected cells in their late stages of apoptosis, the Annexin V assay identi®es cells in the early phases of cell death. The loss of plasma membrane lipid asymmetry leading to the expression of PS appears to be a potentially generalizable phenomenon in Xenopus apoptosis, since such a varied group of unrelated apoptogens caused the cells to respond similarly. This membrane modulation is one of the major mechanisms by which some apoptotic mammalian cell types are recognized [20]. Macrophages of Xenopus recognize thymocytes expressing PS following apoptotic induction and, as a consequence, form rosettes. The involvement of phosphatidylserine in this recognition and subsequent binding is con®rmed by the inhibition of macrophage-thymocyte rosette formation by its analog, phospho-L-serine. Phospho-L-serine has served as an e€ective partial inhibitor of macrophage rosetting with mammalian cells [21±23]. Treatment with phospho-L-serine results in approximately a 30% inhibition of mammalian macrophage binding of apoptotic cells, a level similar to the 40% inhibition found for Xenopus macrophages and apoptotic thymocytes. The successful inhibition of binding also suggests that an, as yet uncharacterized, PS-dependent receptor system may be present and functional on X. laevis macrophages. It is likely that, as in mammals, other macrophage receptor systems may also be involved. This is the ®rst demonstration of PS-expression on Xenopus lymphocytes and of the involvement of a PS-dependent mechanism in macrophage recognition in an amphibian. It suggests that this mechanism has been conserved in evolution. Indeed, analogous recognition of phospholipids has been suggested for insect phagocytes [24].

[2] [3]

[4]

[5] [6] [7]

[8] [9]

[10]

[11]

[12]

[13]

[14] [15]

References [1] Sekizawa A, Fujii T, Tochinai S. Membrane receptors on

[16]

651

Xenopus macrophages for two classes of immunoglobulins (IgM and IgY) and the third complement component (C3). J Immunol 1984;133:1431±5. Wyllie AH, Kerr JF, Currie AR. Cell death: the signi®cance of apoptosis. Intern Rev Cytol 1980;68:251±306. Pigault C, Follenius-Wund A, Schmutz M, Freyssinet J, Brisson A. Formation of two-dimensional arrays of Annexin V on phosphatidylserine-containing liposomes. J Mol Biol 1994;236:199±208. Zwaal R, Comfurius P, Bevers E. Mechanism and function of changes in membrane-phospholipid asymmetry in platelets and erythrocytes. Biochem Soc Trans 1993;21:248±53. Zhang G, Gurtu V, Kain S, Yan G. Early detection of apoptosis using a ¯uorescent conjugate of Annexin V. Biotechniques 1997;23:525±31. Serra MV, Kamp D, Haest C. Pathways for ¯ip-¯op of mono- and dianionic phospholiplids in the erythrocyte membrane. Biochim Biophys Acta 1996;1282:263±73. Fadok VA, Voelker DR, Campbell PA, Cohen JJ, Bratton DL, Henson PM. Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers speci®c recognition and removal by macrophages. J Immunol 1992;148(7):2207±16. Savill J. Recognition and phagocytosis of cells undergoing apoptosis. Brit Med Bull 1997;53(3):491±508. Martin SJ, Reutelingsperger C, McGahon A, Rader J, van Schie R, LaFace D, Green D. Early distribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: inhibition by overexpression of Bcl-2 and Abl. J Exp Med 1995;182:1545±56. Dachary-Prigent J, Pasquet J, Freyssinet J, Nurden A. Calcium involvement in aminophospholipid exposure and microparticle formation during platelet activation: a study using Ca2+-ATPase inhibitors. Biochemistry 1995;34:11625±34. Tait J, Gibson D. Measurement of membrane phospholipid asymmetry in normal and sickle-cell erythrocytes by means of Annexin V binding. J Lab Clin Med 1994;123:741±8. Haberfeld M, Johnson RO, Ruben LN, Clothier RH, Shiigi S. Adrenergic modulation of apoptosis in splenocytes of Xenopus laevis in vitro. Neuroimmunomod 1999;6:175±81. Grant P, Clothier RH, Johnson RO, Schott S, Ruben LN. The time course, localization and quantitation of Tand B-cell mitogen-driven apoptosis in vivo. Immunol Letts 1995;47:227±31. Ruben LN, Buchholz DR, Ahmadi P, Johnson RO, Clothier RH, Shiigi S. Apoptosis in thymus of adult Xenopus laevis. Devel Comp Immunol 1994;18:231±8. Ruben LN, Ahmadi P, Johnson RO, Buchholz DR, Clothier RH, Shiigi S. Apoptosis in the thymus of developing Xenopus laevis. Devel Comp Immunol 1994;18:343±52. Williamson P, Bevers E, Smeets E, Comfurius P, Schlegel

652

[17]

[18]

[19]

[20]

M.S. Nera et al. / Developmental and Comparative Immunology 24 (2000) 641±652 R, Zwaal R. Continuous analysis of the mechanism of activated transbilayer liplid movement in platelets. Biochemestry 1995;34:10448±55. Vermes I, Haanen C, Richel DJ, Schaafsma M, Kaslbeek-Batenburg E, Reutelingsperger C. Apoptosis and secondary necrosis of lymphocytes in culture. Acta Haematol 1997;98:8±13. McMahan R, Johnson RO, Ruben LN, Clothier RH. Apoptosis and the cell cycle in Xenopus laevis. PHA and PMA exposure of splenocytes. Immunol Letts 1999;70:179±83. Langeberg L, Ruben LN, Clothier RH, Shiigi S. The characterization of toad splenocytes which bind mouse anti-human IL-2 receptor antibody. Immunol Letts 1987;16:43±8. Stuart M, Damoiseaux J, Frederk P, Arends J, Reutelingsperger C. Surface exposure of phosphatidylser-

[21] [22]

[23]

[24]

ine during apoptosis of rat thymocytes precedes nuclear changes. Euro J Cell Biol 1998;76:77±83. Okazaki T, Kondo T, Kitano T, Tashima M. Diversity and complexity of ceramide signalling in apoptosis. Cell Signal 1998;10:685±92. Fadok VA, Savill JS, Haslett C, Bratton DL, Doherty DE, Campbell PA, Henson PM. Di€erent populations of macrophages use either the vitronectin receptor or the phosphatidylserine receptor to recognize and remove apoptotic cells. J Immunol 1992;149:4029±35. Fadok VA, Laszlo DJ, Noble PW, Weinstein L, Riches DWL, Henson PM. Particle digestibility is required for induction of the phosphatidylserine recognition mechanism used by murine macrophages to phagocytose apoptotic cells. J Immunol 1993;151:4274±85. Ratner S, Schroit AJ, Vinson SB, Fidler IJ. Analogous recognition of insect phagoctyes and mammalian macrophages. Proc Soc Exp Biol Med 1986;182:272±6.