Med. Oncol. & Tumor Pharmacother. Vol. 3, No. 3/4, pp. 269-273. 1986 I)736-4)118/86 $3.00 + .(XI Printed in Great Britain Pergamon Journals Ltd.

A U T O C R I N E G R O W T H O F H U M A N B L Y M P H O C Y T E S : M A I N T A I N E D R E S P O N S E T O A U T O S T I M U L A T O R Y F A C T O R S IS T H E S P E C I A L F E A T U R E O F I M M O R T A L I Z A T I O N BY E P S T E I N - B A R R VIRUS m A H Y P O T H E S I S

JOHN GORDON,* GRAEME GUY,t LEONIE WALKER,* PAUL NATHAN,* RUTH EXLEY~ and MIKE CLEMENS~t

Department of *Immunology and tBiochemistry, University of Birmingham; .~Department of Biochemistry, St. George's Hospital Medical School, London, U.K.

(Receired 28 May 1986)

The autocrine growth profile of human B lymphocytes transformed with Epstein-Barr virus (EBV) was found to comprise three distinct components: a B-cell growth factor (BCGF); an interleukin-I (IL-1)-Iike activity; an activity requiring ceil-to-cell contact for its action. Observations on the inhibition of the EBV-carrying Daudi lymphoma line by a-interferon indicated that loss of response to these autostimulatory factors was underlying growth cessation. Furthermore, a putative receptor for BCGF was found to be down-regulated on B cells stimulated with non-transforming mitogens but constitutively expressed following EBV-transformation. Taken together with recent evidence that normal B ceils produce autostimulatory factors, these findings suggest that the special feature of autocrine growth by EBV-immortalized cells is a maintenance of what should normally be a transient phenotype, possibly through deregulation of receptor expression. This hypothesis is discussed.

Key words: B lymphocytes, Growth factors, Epstein-Barr virus, Receptors, Lymphoma.

INTRODUCTION

Deregulation of normal cell growth can occur at a variety of leveLs with lesions impinging on growth factors, their receptors and the intracellular mes- sages which they convey. A useful model for studying some of the disturbances which accompany deregulation of growth is afforded by the interaction of the transforming Epstein-Barr virus (EBV) with its host cell, the human B lymphocyte. Resting B cells form the major targets for this potentially oncogenic virus t and during the first round of activation the sequels are indistinguishable from those observed with other T-independent, but non- transforming, B-cell mitogens (e.g. Staphylococcus aureus Cowan Strain IMSAC). 2'3 However, unlike other mitogens, EBV does not require exogenous influences to sustain its effects and, as a result, autonomous lines can be generated possessing an immortal phenotype.

It has been shown recently that the continued growth of B cells transformed by EBV was depen- dent on the production and utilization of auto-

Correspondence to: Dr J. Gordon, Department of Immunology, University of Birmingham Medical School, Birmingham B15 2TJ, U.K.

269

stimulatory factors. 4"s Since these reports, evidence has emerged to indicate that the autocrine growth of B cells is not exclusive to the transformed state in that normal activated cells also produce factors which may be autostimulatory. 6"7 In this paper we look further at the autocrine profile of transformed B cells and investigate the effects of the anti- proliferative drug a-interferon on this phenotype. Furthermore, studies on a putative receptor for a B-cell growth factor (BCGF) allows us to present a hypothesis where the special feature of EBV- immortalization, in terms of autocrine growth, is the maintenance of a response to autostimulatory fac- tors which, under physiological conditions, would be transient only.

EVIDENCE FOR A MINIMUM THREE-FACTOR- MODEL OF B-CELL AUTOSTIMULATION

Initial reports on the autostimulatory component released by EBV-transformed cells pointed to a BCGF. s Supernatants from these cells synergised with anti-immunoglobulin (Ig) in activating resting B cells and enhanced DNA synthesis in pro-activated B cells. However, observations that the same cell types released molecules with IL-1 activities and that

270 John Gordon et al.

IL-1 was growth-promoting for EBV-transformed cells compromised the notion of a unique B-cell derived BCGF being involved. '~'l°

Data presented in Table 1 reconcile these initally conflicting observations by suggesting the existence of both IL-l-like and BCGF activities in super- natants conditioned by transformed B cells (LCL- CM). First, it should be noted that purified monocyte-derived IL-1, while supporting the growth of transformed B cells, did not augment DNA synthesis in assays designed to demonstrate BCGF activities. By contrast, LCL-CM was active in all three BCGF assays. Furthermore, this activity was destroyed on heating to 56°C for 1 h, conditions under which both IL-1 and IL-2, but not BCGF, are stable. Like Scala and colleagues 9 we found it necessary to fractionate the conditioned medium for IL-1 activity to be revealed due to the simultaneous presence of a high molecular weight inhibitor. The major peaks of IL-1 activity were evident on gel-filtration with molecular weights corresponding to 15-kDa and 35-kDa although numerous peaks above 70-kDa and activity below 10-kDa were also seen. All these peaks were autostimulatory. The BCGF activity was characterized by poor yield and was not associated with the major IL-1 peaks. From some lines, material eluting at 25-kDa exhibited anti-Ig co-stimulating activity while in others BCGF of a higher molecular weight was predomi- nant (45-55-kDa). In addition to the two soluble activities described above, it ws recently demon-

strated that the pronounced "clumping" accompany- ing EBV-transformation was a requisite component of the autocrine response and could not be replaced by soluble factors contained in the conditioned medium. H The accumulated data prompt us to present a minimum three-factor-model for B-cell autostimulation based on the following components: (i) Cell contact factor; (ii) Interleukin-l-like factor; (iii) B-cell growth factor.

EVIDENCE THAT A MAINTAINED RESPONSE TO AUTOSTIMULATORY FACTORS IS THE

KEY TO IMMORTALIZATION

Recent evidence suggests that the generation of autostimulatory factors is not the prerogative of B cells with a transformed phenotype. IL-l-like activi- ties, BCGF and growth-promoting cell contacts have all been described as products of normal B cells when activated in an appropriate fashion. 6"7"~1"12 What then is the special feature which sustains autocrine growth in EBV-transformed cells? The first clue came from a study on a transformed B-cell clone which displayed programmed senescence. It was shown that terminal differentiation of the cells was accompanied by a failure to respond to the autostimulatory factors that they were still pro- ducing, t3 Further evidence is now presented in support of the notion that a maintained response is the key feature of sustained autocrine growth in

Table 1. Evidence for IL-1 and non-IL-1 soluble factors in autocrine growth

[3H]Tdr incorporation in following assays:

anti-Ig a LD B cells 3d SAC LCL Thy+PHA

Control 216 1317 652 1074 473 IL- lb 249 1659 583 4529 42,620 LCL-CM ¢ 6302 12,073 5968 7370 29,497 LCL-CM

(56 °, 1 h) 855 2143 nd nd nd

aGrowth factor assays were as follows: anti-Ig co-stimulator assay with purified, resting tonsillar B cells; s enhanced proliferation of 'pre-activated', low density (LD) B cells; s maintenance of stimulation at day 6 following activation of resting B cells for 3 d with SAC; 19 enhanced growth of lymphoblastoid cells seeded at 2000 per microweii; IL-1 assay using sub-optimal stimulation of mouse thymocytes with PHA. l° In all cases [3H]thymidine (Tdr) incorporation was assessed between 48 and 64 h after addition of factor. blnterleukin-1 (IL-1) was purchased from Genzyme and represented an affinity- purified 16k monocyte-derived product. It was used at 1 unit per ml which was near optimal in all the assays. eLCL-CM represent medium conditioned by a lymphoblastoid line as described

8 o previously and used at 10 Yo final concentration. When assayed in the thymocyte assay for IL-1 such supernatants were negative. The positive activity indicated represents that associated with material fractionating on G-100 at 15--40-kDa.

transformed B ceils. The EBV-harbouring Daudi Burkitt lymphoma line is exquisitely sensitive to the anti-proliferative effects of a-interferon. Fol- lowing treatment with a-interferon we noted that the cells developed a plasmacytoid morphology, ac- cumulated cytosplasmic Ig and expressed antigens associated with plasma cells (details to be published elsewhere). The results presented in Table 2 reveal that accompanying these processes was a diminution in the ability of cells to respond to growth factors whether autogenous or exogenously supplied. In contrast, factor activity actually rose although this could reflect a diminished utilization rather than any real increase in production.

Table 2. Growth cessation of Daudi by a-interferon corresponds to a diminished response to auto-stimulatory

factors

Treatment for 2 d with interferon ~

No Yes

GF Production b 100% 190% GF Response 100% 23%

aDaudi lymphoma cells were taken from their log phase of growth and cultured in the presence or absence of Namalwa derived (Wellcome) a-interferon (50 units ml -I) for 2 days. ~Following culture cell supernatants were collected, fil- tered and the presence of factors capable of supporting the growth of a B-lymphoblastoid line was assayed. The results are represented as change over untreated cells. Similarly, the ability of untreated compared to treated cells to respond to an exogenous source of B-cell derived growth factors was determined.

So, as with the senescent lymphoblastoid line, these studies demonstrate that when a breakdown in the ' immortalized' state occurs, this is accompanied by a failure to respond to autostimulatory factors rather than an inability to continue producing them. This study also confirms in a clonal system the previously suggested notion that a-interferon is capable of functioning as a B-cell differentiation factor (BCDF). 14

CHARACTERIZATION OF A RECEPTOR INVOLVED IN THE GROWTH-REGULATION OF

ACTIVATED B CELLS

The CD23,p45 antigen (also known as BLAST-2; EBVCS) appears within 2-4 h of activating resting B cells with minimal, optimal and transforming stimuli. 2'3 Its constitutively high expression on EBV- transformed cells led early workers to believe that

Autostimulation of B cells 271

this B-lineage restricted antigen was confined to the viraily-transformed state, hence its designation of EBVCS.'5 A recent report demonstrated that liga- tion of the CD23 antigen on B cells prompted into early G~ with the phorbol ester TPA delivered a signal for cell-cycle progression to the activated cell. 16 Results outlined in Table 3 reveal that the effects of perturbing CD23 with the monoclonal antibody MHM6 could not be distinguished from the action of purified, low molecular weight, T-cell- derived BCGF whether measuring augmentation of DNA synthesis, the induction of IL-2 receptors or indeed the induction of IL-2 responsiveness in TPA-activated B cells. These results indicate that, at the very least, the signals conveyed to the B cell through CD23 and via low molecular weight BCGF are quantitatively and qualitatively similar.

Table 3. MHM6 antibody to CD23 mimics low molecular weight BCGF

Control MHM6 BCGF

TPA ~' stimulations 2074 37,532 39,962 IL-2 b response 1788 14,593 17,438 Tac ~ expression 17% 38% 36%

~'Purified resting tonsillar B cells were cultured in micro- wells at 105 per 200 Ixl in the presence of 2 ng ml -I of the phorbol ester TPA as described previously/ Results are expressed as cpm [3H]Tdr incorporation between 48 and 64 h in the presence or absence of IgG from MHM6 ascites at ~6 50 [xg ml -~ or 10% semi-purified T-cell- derived, 12-kDa BCGF (Cellular Products Inc.). bCells were cultured as for (a) except that at 48 h recombinant IL-2 (Genzyme) was added at 200 units per mi and DNA synthesis assessed between 120 and 136 h. In the absence of rlL-2, [3H]Tdr incorporation was below 2000 cpm in all cases. ¢Cells were cultured as for (~) except that at 64 h cells were taken, washed and stained with FITC-labelled antibody to IL-2-receptor (Tac). Results are given as % cells positive above background as assessed on a FACS IV (Becton- Dickinson).

We next explored the detailed kinetics of CD23 expression following the stimulation of highly puri- fied resting B cells both with SAC and with EBV. It should be noted (Table 4) that over the first 72 h the kinetics of appearance and, later, of disappearance of the antigen were virtually indistinguishable be- tween SAC and EBV stimulations. However, whereas with SAC, the receptor remained down- regulated, cells infected with EBV revealed a re-expression of the antigen by day 6 which was constitutively maintained from thereon.

272 John Gordon et al.

Table 4. Kinetics of CD23 expression

Percentage cells rosette positive"

Hours post-activation SAC EBV

contribution of other EBV-genes to the autocrine phenotype, is the subject of current and future investigations in our laboratories.

0 2 2 4 31 23

24 49 57 48 46 50 72 9 18

144 2 35

Acknowledgement--This work was funded by grants from the Medical Research Council (U.K.) and the Cancer Research Campaign.

REFERENCES

"Purified resting B cells were prepared 2 and cultured at 5 x 105 ml -I in the presence of SAC (1:100,000) or an optimal transforming titre of B95-8-derived EBV. The percentage of cells forming rosettes in an indirect assay with MHM6 as first layer 2 was assessed at the times indicated.

HYPOTHESIS

The special feature of E B V-immortalization is the maintenance of an otherwise transient response to autostimulatory factors

This conclusion stems both from the studies on transformed lines where the autocrine loop has been broken and from a detailed comparison of the T-independent activation of resting B cells by EBV and non-transforming mitogens. The preliminary studies on the CD23 antigen suggest that its de- regulated expression on EBV-transformation may be pivotal to the whole process.

If we accept that CD23 provides a growth- promoting signal by serving as a receptor for BCGF and consequently regulates the autocrine loop, then we need to address the mechanism underlying its deregulation by EBV. One potential explanation might lie in the observation that the gene encoding the latent membrane protein (LMP) is, in fact, transforming. 17 LMP is the most abundantly tran- scribed gene in latent EBV-infection and from the predicted amino acid sequence its product repre- sents a six-looped transmembrane protein exposing small sequences only of each domain to the outside of the cell. It has been suggested that its function might be to anchor and stabilize in the membrane receptors which bind autostimulatory factors. 17 It therefore becomes attractive to speculate that the constitutive expression of CD23 on EBV- transformation relates to the presence of LMP, particularly in view of the observation that CD23 is secreted from an EBV-negative B-cell line (Ramos) but is integrated into the membrane on EBV- conversion. ~8 This possibility, together with the

1. Aman P, Gordon J, Lewin N, Nordst6m M, Ehlin- Henriksson B, Klein G, Carstensson A: Surface marker characterization of EBV target cells in normal blood and tonsil B lymphocyte populations. J lmmun 135, 2362 (1985).

2. Walker L, Guy G, Brown G, Milner A, Gordon J: Control of human B-lymphocyte replication. I. Characterization of novel activation stages which precede the entry of G~j B cells into cycle. Immu- nology 58, 583 (1986).

3. Gordon J, Walker L, Guy G, Brown G, Rowe M, Rickinson A: Control of human B-lymphocyte replication. II. Transforming Epstein-Barr virus ex- ploits three distinct viral signals to undermine three separate control points in B-cell growth. Immunology 58, 591 (1986).

4. Blazar B A, Sutton L M, Strome M: Self-stimulating growth factor production by B-cell lines derived from Burkitt's lymphoma and other lines transformed in vitro by Epstein-Barr virus. Cancer Res 43, 4562 (1983).

5. Gordon J, Ley S C, Melamed M, Aman P, Hughes- Jones N C: Soluble factor requirements for the autostimulatory growth of B lymphoblasts immortal- ized by Epstein-Barr virus. J Exp Med 159, 1554 (1984).

6. Matsushima K, Procopio A, Abe H, Scala G, Ortaldo J R, Oppenheim J J: Production of interleukin-1 activity by normal human peripheral blood B cells. J Immun 135, 1132 (1985).

7. Witzel N, Ambrus J L, Jurgensen C H, Mostowski H, Fauci A S: Production of a monoclonal antibody recognizing a sub-population of human B lympho- cytes producing B cell growth factor (BCGF). Fed Proc Abs 4905 (1986).

8. Gordon J, Melamed M D, Ley S C, English L S, Hughes-Jones N C: Immortalized B lymphocytes produce B-cell growth factor. Nature 310, 145 (1984).

9. Scala G, Kuang Y D, Hall R E, Muchmore A V, Oppenheim J J: Accessory cell function of human B cells. I. Production of both IL-l-like activity and an IL-1 inhibitory factor by an EBV-transformed human B-cell line. J Exp Med 159, 1637 (1984).

10. Gordon J, Guy G, Walker L: Autocrine models of B-lymphocyte growth. II. Interleukin-1 supports the proliferation of transformed lymphoblasts but not the stimulation of resting B cells triggered through their receptors for antigen. Immunology 57, 419 (1986).

11. Gordon J, Guy G, Walker L: Autocrine models of B-lymphocyte growth. I. Role of cell contact and

soluble factors in T-independent B-cell responses. Immunology 56, 329 (1985).

12. Spicker-Polet H, Hagen K, Teodorescu M: The role of inter-cellular contacts in the activation of B lymphocytes by anti-immunoglobulin antibodies. J Immun 134, 2827 (1985).

13. Melamed M D, Gordon J, Ley S C, Edgar D, Hughes-Jones N C: Senescence of a human lympho- blastoid clone producing anti-Rhesus (D). Eur J Immun 15, 742 (1985).

14. Choi J S, Lim K H, Sanders F K: Effect of Interferon on pokeweed mitogen-induced differentiation of human peripheral blood B lymphoctyes. Celhtlar bnmun 64, 20 (1981).

15. Slovin S F, Frisman D M, Tsoukas C D, Royston I, Baird S M, Wormsley S B, Carson D A, Vaughn J H: Membrane antigen on EBV infected human B cells

Autostimulation o f B cells 273

recognised by a monoclonal antibody. Proc Nat/Acad Sci USA 79, 2649 (1982).

16. Gordon J, Rowe M, Walker L, Guy G: Ligation of the CD23 (BLAST-2, EBVCS) antigen triggers the cell-cycle progression of activated B lymphocytes. Eur J Immunol (in press).

17. Wang D, Liebawitz D, Kieff K: An EBV membrane protein expressed in immortalized lymphocytes trans- forms established rodent cells. Cell 43, 831 (1985).

18. Thorley-Lawson D A, Mann K P: Early events in EBV infection provide a model for B-cell activation. J Exp Med 162. 45 (1985).

19. Minegari M C. Geroson F, Moretton A, Zubler R H, Moretta L: B cell growth factor activity of immuno- affinity-purified and recombinant human interleukin 2. Eur J hnmunol 15, 193 (1985).