Immunology Letters, 26 (1990) 227-232 Elsevier

1MLET 01491

A monoclonal antibody against the human IL-2 receptor binds to paraformaldehyde-fixed but not viable frog (Xenopus) splenocytes

Laura Haynes t, J an A. M o y n i h a n 2 and Nicholas C o h e n ~ Departments of 1Microbiology and Immunology and 2psychiatry, University of Rochester School of Medicine and Dentistry,

Rochester, NY, US.A.

(Received 30 April 1990; revision received and accepted 19 July 1990)

1. Summary

Others have reported that a monoclonal anti- human IL-2 receptor antibody (anti-CD25) specifi- cally binds a membrane receptor on Xenopus laevis PHA-induced and paraformaldehyde-fixed splenic blasts [1, 2]. In this paper, we present evidence sug- gesting that this binding is an artifact of membrane damage. Specifically, significant binding of anti- CD25 could only be achieved if the lymphoblasts were acid-washed and/or paraformaldehyde-fixed prior to being incubated with the fluoresceinated an- tibody. For example, in a representative experiment 95°7o of paraformaldehyde-fixed blasts, about 19°70 of acid-washed but not fixed blasts, but fewer than 2°7o of viable (untreated) blasts were positive for the CD25 epitope. Paraformaldehyde is known to alter membrane permeability. The DNA dye propidium iodide (PI) was used to demonstrate that the acid washing procedure also causes membranes to be- come permeable. Flow cytometric analyses of acid- washed PHA-induced splenic blasts doubly stained with the anti-CD25 antibody and PI showed that only 1.5°70 of the cells that were positive for CD25 did not stain with PI. Additionally, the anti-CD25 antibody, which immunoprecipitated a molecule from human lymphoblasts of between 50 and 60 kDa, did not immunoprecipitate any surface

Key words: IL-2 receptor; Xenopus; splenocyte

Correspondence to: Laura Haynes, Department of Microbiolo- gy and Immunology, Box 672, University of Rochester Medical Center, Rochester, NY 14642, U.S.A.

molecules from ~25I-labeled Xenopus splenic blasts. Since binding of anti-CD25 to Xenopus splenic blasts appears to occur only after membrane dam- age, the antibody may be recognizing a cross- reactive internal epitope that is not involved in ligand binding on the cell surface.

2. Introduction

The South African frog Xenopus laevis exhibits T lymphocyte responses that are comparable to those of mammals [3]. In vitro, Xenopus T cells proliferate in response to alloantigens [4] and T cell mitogens such as phytohemagglutinin (PHA) and concanava- lin A (Con A) [4, 5]. Additionally, Xenopus T cells are capable of killing allogeneic targets in a haplotype-specific manner [6, 7]. In vivo, T cells play a helper role in antibody production [8] and participate in acute graft rejection [9]. Xenopus T lymphocytes also produce a T cell growth factor (TCGF) which is similar in its in vitro activities to mammalian interleukin 2 (IL-2) [10]. TCGF is produced by T cells upon stimulation with P H A or alloantigens and has a glycosylated M r of 14000 (unpublished observation). Both Xenopus splenic and thymic blasts proliferate in response to TCGF, whereas resting lymphoid cells do not respond. TCGF can also support the long-term growth of Xenopus T cell lines [7, 10]. While Xenopus TCGF and mammalian IL-2 molecules appear to be physi- cally similar and have at least some identical func- tional properties, they are not cross-reactive in vitro I101.

The human IL-2 receptor is a heterodimer of two

0165-2478 / 90 / $ 3.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division) 227

polypeptide chains of 55 and 75 kDa [11-13]. The 55-kDa molecule is commonly referred to as the Tac antigen, p55, or the CD25 molecule. A report by Langeberg et al. [1] describes the specific binding of an anti-human CD25 monoclonal antibody (mAb) to Xenopus splenocytes. These investigators demon- strated that 43.5% of resting Xenopus splenocytes and 92.5% of PHA-activated splenic blasts specifi- cally bind this antibody after the cells are acid washed and paraformaldehyde fixed. These data seemed inconsistent with the lack of in vitro func- tional cross-reactivity of human recombinant IL-2 and Xenopus lymphocytes and the ability of PHA- activated but not resting Xenopus lymphocytes to absorb TCGF activity from TCGF-rich superna- tants [10]. Evidence supporting the proposition that this positive staining of Xenopus tymphoblasts is an artifact of damage to the membrane will be present- ed in this report.

3. Materials and Methods

3.1. Preparation of Xenopus splenic blasts

Spleens were removed aseptically from adult fe- male Xenopus frogs (Scientific Animal Import, Glen Ridge N J) that had been anesthetized in 0.5% (w/v) tricaine methane sulfonate (Argent Chemi- cals, Redmond, WA). Spleens were dissociated be- tween two frosted microscope slides and the cells were cultured in complete medium (Leibovitz' s L-15 medium (Gibco, Grand Island, NY) adjusted to am- phibian osmolarity (220 mOsm) and supplemented with 1.25 × 10 -5 M Hepes buffer (Gibco, Grand Is- land, NY), 100 U/ml penicillin, 100/xg/ml strep- tomycin (Gibco, Grand Island, NY), 1x10 -2 M NaHCO3, 5x10 -5 2-mercaptoethanol (Sigma, St. Louis, MO), and with 10% fetal bovine serum (FBS) (Hyclone, I~gan, UT)). Splenocytes were then cen- trifuged (350 x g) over Histopaque 6 = 1.119 (Sigma, St. Louis, MO) washed twice and then put into cul- ture with 1/zg/ml PHA-P (Difco, Detroit, MI) at 5 × 10 6 cells/ml in 24-well culture plates (Costar, Cambridge, MA) for 3 days at 26 °C.

3.2. Preparation of human blasts

Human peripheral blood was collected in heparinized vacutainer tubes (Becton-Dickinson,

228

Mountain View, CA), diluted 1:2 with phosphate- buffered saline (PBS) and centrifuged (350xg) on Isolymph 6=1.077 (Gallard-Schlesinger Chemical, Carle Place, NY). Lymphocytes were recovered, washed twice and cultured in RPMI-1640 (Gibco, Grand Island, NY) containing 50 mM Hepes buffer, 10 U/ml penicillin, 10/zg/ml streptomycin, 10°70 FBS and 1 /zg/ml PHA at 5 x 106 cells/ml in 24-well culture plates for 3 days at 37 °C.

3.3. Fixation, staining and flow cytometry

The fixation and staining protocol of Langeberg et al. was used [1]. All cells and solutions were kept at 4 °C. Xenopus PHA-induced splenic blasts were washed twice in complete medium. Some aliquots were acid-washed to remove endogenous IL-2 from its receptor by treatment with complete medium plus containing 25 mM sodium acetate, pH 4.0 for 60 sec, washed once in L-15, and then washed twice in amphibian PBS (APBS) + 1% bovine serum al- bumin (BSA) + 0.1% NaN 3 (ABN). Blasts were fixed by treatment with 1% paraformaldehyde (Sig- ma) for 30 min and then washed twice with ABN. Live cells were washed twice in ABN prior to incuba- tion with mAbs.

Either the fluoresceinated anti-CD25 (anti-TAC) or fluoresceinated anti-KLH mAbs (Becton- Dickinson), both IgG~K, were then added to 1 x 106 blasts at 0.2 txg per aliquot; the fluoresceinated anti- Xenopus IgM mAb 6.16 (IgGlx) [14] was added at 100/~g per sample. Samples were incubated over- night at 4°C and then washed twice with ABN. Propidium iodide (PI) was added (to detect the proportion of live versus damaged cells) at the indi- cated concentrations 15 min before the cells were analyzed on an EPICS V multiparameter flow cytometer/cell sorter (Coulter Electronics, Hialeah, FL) with a 5 W UV-enhanced argon-ion laser (Spec- tra Physics, Mountain View, CA, Model 202S) tuned to 488 nm at 500 mW. Cells were analyzed for fluorescein (515 nm to 560 nm) and/or propidium iodide (>610 nm) fluorescence and for low angle light scatter. Twenty thousand cells were analyzed per sample.

3.4. Immunoprecipitation and SDS-PAGE

Xenopus P H A blast surface proteins were ira-

munoprecipitate~l with the anti-CD25 or anti-IgM mAbs, and human P HA blasts were immunoprecipi- tated with the anti-CD25 mAb according to the method of Schwager and Hadji-Azimi [15]. Briefly, 1 × 107 blasts were surface-iodinated with 125I using lactoperoxidase (Sigma) and glucose oxidase (Sig- ma). Cells were lysed in 100 tzl of buffer containing 20 mM Tris/1 mM MgCl2/150 mM NaC1/2% NP 40/0.1 mM phenylmethylsulfonyl fluoride pH 8.0. Lysates were precleared twice with 50 #1 of protein A-Sepharose (Pharmacia, Piscataway, N J) conjugated to a subclass control (IgGl~) myeloma (ICN Biochemicals, Costa Mesa, CA). Samples were incubated with anti-CD25 (1 /~g) or 6.16 (50 tzg) for 4 h, 50 #1 of a 50% protein A-Sepharose solution was added and the incubation was continued over- night. The protein A complexes were spun out in a microfuge, washed and boiled for 3 min in sample buffer containing 0.125 M Tris-C1, pH 6.8/4°?0 SDS/20% glycerol/10% 2-mercaptoethanol. Each sample was run on an SDS-PAGE reducing gel [16]. Samples from Xenopus blasts were run on at a 7.5-15°70 gradient gel and the sample from human blasts was run on a 7.5°7o gel.

4. Results

Live and fixed PHA-induced splenic blasts were stained according to the protocol of Langeberg et al. [1]. In 4 of 4 experiments, a significant higher per- centage of fixed cells than live cells stained positive with the anti-CD25 mAb. In the representative ex- periment, 95°?o of fixed blasts were positive for CD25 (Fig. 1A)whereas less than 1°70 of live (i.e. not fixed) blasts stained positive for CD25 (Fig. 1B). Fig. 2 shows a representative experiment (N= 2) in-

LOG RELATIVE FLUORESCENCE INTENSITY

Fig. 1. Comparison of anti-CD25 (shaded) and subclass control anti-KLH antibody (unshaded) binding to paraformaldehyde- fixed (A) or live (B) Xenopus splenic blasts. Histograms A and

B are on different scales.

100-

8 0 -

60".

4 0

2 0 -

0 FIXED, ACID WASHED FIXED

[ ] ANTI-IgM [ ] ANT~-CD25

UNFIXED, ACID WASHED UNFIXED

Fig. 2. Representative experiment (N= 2) comparing the staining of live and paraformaldehyde-fixed Xenopus splenic blasts with either anti-IgM (striped bars) or anti-CD25 (open bars)• Live,

acid-washed blasts were not assayed with anti-IgM.

dicating that the acid-washing procedure has little effect on the staining of fixed cells. Conversely, the acid-washing treatment of unfixed blasts was as- sociated with an increase (from 3.1°70 to 18.5°70) in the percentage of cells staining positive for CD25. Fig. 2 also shows that essentially all fixed blasts ap- peared to express surface IgM (slgM) as detected by 6.16 staining, whereas 30°7o of the live blasts were slgM-positive (approximately 30% of 3-day P H A splenic blasts are slgM-positive B cells which appear to have been stimulated by cytokines produced by ac- tivated T cells [17]. In all experiments, fixation and acid washing did not increase the binding of the con- trol anti-KLH mAb.

The increase in the percentage of cells staining positive for CD25 after fixation can be correlated with the cell membrane alterations that occur during fixation [19, 20]. Since acid washing also causes an increase in the percentage of cells staining positive for CD25, it seems likely that this procedure also causes cell membrane alterations. Propidium iodide is an intercalating dye which reacts with DNA and double stranded RNA in cells only if the membrane is made permeable [21-23]. Therefore, PI red fluorescence was used to gate out damaged cells. Ta- ble 1 demonstrates that PI is not toxic for non-acid washed (i.e., undamaged) cells at concentrations up to 10 tzg/ml. Acid-washed cells appear to take up all of the PI that is available since the higher the PI con- centration, the greater the percentage of cells ex- hibiting red fluorescence.

In two experiments, splenic PHA blasts (acid washed but not fixed) were stained with both anti- CD25 and PI and gated on light scatter alone such that all cells were counted by the flow cytometer. In

229

TABLE 1

Compar ison of acid-washed and not acid-washed Xenopus splenocytes and PHA-induced splenic blasts.

P.I. (t~g/ml) % Positive for P.I. a

Not acid-washed Acid-washed

Splenocytes Blasts Splenocytes Blasts

0 0 0 0 0 1.25 1.5 1.0 18 33 2.5 1.6 1.5 47 67 5.0 1.5 2.0 65 82

10.0 2.0 2.7 72 88

a Cells were acid washed as described in Materials and Methods and the indicated concentrations o f P I were added 15 min before

the cells were assayed on the flow cytometer.

the experiment shown in Fig. 3A, 22°70 of these cells were positive for CD25. Fig. 3B shows the same population of ceUs gated on light scatter and red fluorescence such that cells displaying red fluores- cence (membrane permeable or damaged) were omitted from the count by the flow cytometer. In the resulting population, only 1.5% of the cells were positive for CD25. This indicates that most, if not all, of the acid-washed blasts staining positive for CD25 had suffered membrane damage caused by acid washing.

To determine if reactivity of the anti-CD25 mAb to fixed and acid-washed cells was artifactual, anti- CD25 and 6.16 mAbs were used to immunoprecipi- tate 125I-labeled surface molecules from both live and fixed splenic blasts. As seen in Fig. 4, no bands were evident after either fixed (lane A) or live (lane

B. <1% CD25+

LOG RELATIVE FLUORESCENCE INTENSITY

Fig. 3. Compar ison of anti-CD25 (shaded) and subclass control an t i -KLH antibody (unshaded) binding to propidium iodide (PI)-treated (1/~g/ml), unfixed acid-washed Xenopus splenic blasts gated on light scatter (A) or light scatter and red fluores-

cence (B). Histograms A and B are on different scales.

230

A B C D E

66

45

31

21 14

6 6

4 5

O 21

Fig. 4. lmmunoprecipi tat ion of 125I-labeled P H A lymphoblast surface proteins. Lanes A - D are Xenopus splenic blast proteins immunoprecipitated with: A and B, anti-CD25; C and D, 6.16 (anti-IgM). Lanes A and C represent fixed cells and lanes B and D represent live cells. All were separated on a 7.5-15°70 reducing gradient gel. Lane E shows human P H A blast surface proteins im- munoprecipitated with anti-CD25 and separated on a 7.5°7o

reducing gel. M r is shown in kDa.

B) surface-labeled Xenopus splenic blast proteins were immunoprecipitated with anti-CD25. As a positive control, both fixed (lane C) and live (lane D) splenic blast surface proteins were immunoprecipi- tated with anti-Xenopus IgM (6.16). The bands in lane D show immunoglobulin/z-heavy chains (H#) at approximately 70 kDa and light chains from 25 to 30 kDa. Lane C indicates that fixation causes the surface molecules to become extensively cross- linked. As a positive control for the anti-CD25 mAb, human peripheral blood PHA blasts were im- munoprecipitated with anti-CD25 (lane E) and the Tac antigen is seen as a broad band between 50 and 60 kDa.

5. Discussion

Alloantigen- or mitogen-stimulated Xenopus T lymphocytes produce TCGF which is similar in its in vitro activity and physical characteristics to mam- malian IL-2 [10]. Xenopus TCGF stimulates proliferation of splenic and thymic blasts, but not resting cells, and it can also support the long term growth of T cell lines. Additionally, lymphoblasts, but not resting cells, can absorb TCGF activity from

a T cell supernatant . Despite these similarities, Xenopus T C G F and mammal i an IL-2 are not func- t ionally cross-reactive in vitro [10]. Furthermore, at least 4 mAbs against h u m a n IL-2 fail to recognize Xenopus TCGF by Western blot analysis or by im- munoprecipi ta t ion (unpublished data).

In contrast to published reports on fixed cells [1, 2], our data demonstra te that an ant i -CD25 m Ab does not recognize an epitope on the surface o f via- ble Xenopus splenic blasts. In a representative ex- periment, a significant percentage o f paraformalde- hyde-fixed or acid-washed splenic blasts stained positive for CD25 (78.7°7o and 18.5070, respectively)

' compared with 3.1070 o f live untreated blasts. Recent- ly, Ruben et al. have reported comparable values for ant i-CD25 staining o f unfixed Xenopus P H A - induced splenic blasts [18]. Fixation also increased the number o f cells staining positive for surface im- munoglobul in from about 3007o to almost 10007o. Ad- ditionally, significant numbers o f fixed Xenopus thymocytes stained positive for CD25 (28070) and surface immunoglobul in (71%), whereas live Xeno- pus thymocytes were not stained ( < 1070 positive) by either fluoresceinated reagent (data not shown). The absence o f s lgM ÷ cells in preparat ions o f thymo- cytes is consistent with other published data [3, 14]. data [3, 14].

Both fixation and acid washing treatments alter membrane integrity. Fixation causes cross-linking of surface proteins which alters membrane permeabili- ty [19, 20] and the acid washing procedure, which is per formed to remove endogenous T C G F from its receptor, also causes membrane damage as assayed by PI staining [21-23]. I f the Xenopus T C G F recep- tor is similar to the mammal i an IL-2 receptor, acid washing should not be necessary in order to see bind- ing o f ant i -CD25 to the CD25 epitope. The CD25 molecule (p55 peptide o f the IL-2 receptor) is the low-affinity IL-2 receptor and is present in ten-fold excess o f the high-aff ini ty receptor, which is com- posed o f bo th 55 and 75-kDa polypeptides [11-13]. The affinity o f p55 for IL-2 is sufficiently low (K d 10 -8 M) that acid washing is not necessary to re- move IL-2 f rom these molecules. The only instance in which acid washing is required is when at tempting to quant i fy high-aff ini ty IL-2 receptors [24]. Ad- ditonal evidence that ant i -CD25 is not specifically binding surface T C G F receptors on Xenopus blasts is the failure o f the m Ab to immunoprecipi ta te any

surface proteins f rom ~25I-labeled Xenopus cells. Since there seems to be antigen-specific binding o f the ant i -CD25 MAb to fixed Xenopus splenic blasts (the subclass control MAb did not bind), some inter- nal protein may be cross-reactive with this antibody. I f this is the case, it is unlikely that this internal pro- tein is involved in binding TCGF on the cell mem- brane. Perhaps a more appropria te way to study the Xenopus TCGF receptor is to use its physiological ligand, TCGF, in experiments similar to those in which the human IL-2 receptor was characterized [11-131.

Acknowledgements

This research was supported by U S P H S grant HD-07901. Review o f this manuscr ipt by Fiona Harding is gratefully appreciated.

References

[1] Langeberg, L., Ruben, L. N., Malley, A., Shiigi, S. and Bea- dling, C. (1986) Immunol. Lett. 14, 103.

[2] Langeberg, L., Ruben, L. N., Clothier, R. H. and Shiigi, S. (1987) Immunol. Lett. 16, 43.

[3] Du Pasquier, L., Schwager, J. and Flajnik, M. F. (1989) Annu. Rev. Immunol. 7, 251.

[4] Nagata, S. (1986) Dev. Comp. Immunol. 10, 259. [5] Green-Donnelly, N. and Cohen, N. (1979) Cell. Immunol.

46, 59. [6] Bernard, C. C. A., Bordmann, G., Blomberg, B. and Du

Pasquier, L. (1979) Immunogenetics 9, 442. [7] Watkins, D., Harding, F. and Cohen, N. (1988) Transplanta-

tion 45,499. [8] Bernard, C. C. A., Bordmann, G., Blomberg, B. and Du

Pasquier, L. (1981) Eur. J. Immunol. 11, 151. [9] Horton, J. D. and Manning, M. J. (1972) Transplantation

14, 141. [10] Watkins, D. and Cohen, N. (1987) Immunology 62, 119. [11] Sharon, M., Klausner, R. D., Cullen, B. R., Chizzonite, R.

and Leonard, W. J. (1986) Science 234, 859. [12] Tsudo, M., Kozak, R. W., Goldman, C. K. and Waldman,

T. A. (1986) Proc. Natl. Acad. Sci. USA 83, 9694. [13] Teshigawara, K., Wang, H. M., Kato, K. and Smith, K. A.

(1987) J. Exp. Med. 165, 223. [14] Bleicher, P. A. and Cohen, N. (1981) J. Immunol. 127, 1549. [151 Schwager, J. and Hadji-Azimi, I. (1984) Differentiation 27,

182. [16] Laemmli, W. K. (1970) Nature 227, 680. [17] Cohen, N. and Haynes, L. (In Press) in: Phylogenesis of Im-

mune Functions (G. W. Warr and N. Cohen, Eds.) CRC Press, Boca Raton, FL.

[181 Ruben, L. N., Clothier, R. H., Murphy, G. L., Marshall, J. D., Lee, R., Pham, T., Nobis, C. and Shiigi, S. (1989) Gen. Comp. Endocrinol. 76, 128.

231

[19] Fox, C. H., Johnson, F. B., Whiting, J. and Roller, E P. (1985) J. Histochem. Cytochem. 33, 845.

[20] Hopwood, D. (1985) Histochem. J. 17, 389. [21] Krishan, A. (1975) J. Cell Biology 66, 188. [22] Loken, M. R. and Stall, A. M. (1982) J. Immunol. Methods

50, 485. [23] Horan, P.K. and Kappler, J.W. (1977) J. Immunol.

Methods 18, 309. [24] Robb, R. J., Greene, W. C. and Rusk, C. M. (1984) J. Exp.

Med. 160, 1126.

232