genetic analysis of in vitro leukemogenesis induced by thymus epithelial reticulum cells...

Znt. J . Cancer: 21, 115-120 (1978)

GENETIC ANALYSIS OF IN VITRO LEUKEMOGENESIS INDUCED BY THYMUS EPITHELIAL RETICULUM CELLS TRANSMITTING MURINE LEUKEMIA VIRUSES

Martin HAAS Department of Cell Biology, The Weizmann Institute of Science, Rehovot, Israel

Lymphatic leukemia developed in C57BL/6 mice following inoculation of normal thymocytes that had been co-cultured on leukemic thymus epithelial reticulum monolayer cells. Using thymocytes genetically marked in Ly membrane antigens, we showed that the thymomas which developed were produced by the co-cultured thymocytes rather than by leukemic cells derived from the monolayer. Thus, leukemogenic conversion of normal thymocytes took place in vitro. Inoculation of cultured leukemic thymus epithelial reticulum monolayer cells (LTER) gave rise mainly to reticulum cell sarcomas and myeloid leukemias, rather than to lymphatic leukemias (which developed following inoculation of thymocytes that had been cultured on the LTER monolayers). Thus LTER cells may themselves be tumor cells capable of producing RCNA (reticulum cell neoplasm type A) or myeloid tumors in addition to their ability to convert normal thymocytes into leukemic cells.

In vitro lymphomagenesis has been shown to take place by incubation of normal thymocytes on leukemic thymus epithelial reticulum (LTER) monolayers. Waksal er al. (1976) used the high- leukemia-incidence mouse strain AKR and incubated young adult thymocytes on LTER monolayers derived from spontaneous leukemic AKR thymuses. Haas (1976) and Haas et al. (1977) used young adult normal C57BL/6 (B6) thymocytes incubated on LTER monolayers derived from radiation leukemia virus (RadLV)-inoculated B6 thymic leukemias (lymphosarcomas). The incubated thymocytes produced lymphatic leukemia when inoculated in vivo into normal or thymectomized recipient mice.

Spontaneous induction of viral leukemia in AKR mice and of X-ray-induced or RadLV-induced leukemia in B6 mice is dependent on the thymus. Thymectomy was shown to effectively prevent development of leukemia (McEndy et al., 1944; Kaplan, 1950), whereas implantation of a thymus graft into thymectomized animals restored their susceptibility to leukemia development (Law, 1952; Kaplan el al., 1953). Law postulated that thymic reticular elements, rather than thymus lymphoid cells, are responsible for restoration of sensitivity to leukemogenesis. Moreover, Hays (1968) has shown, by grafting thymus reticulum epithelial cells into thymectomized mice, that these are the cells which restore susceptibility to leukemia development. Thus, whatever the specific role of TER cells in the induction of viral lymphatic leukemia, in vitro incubation of thymocytes on LTER monolayers endows them with functions which are normally provided by the thymus in an intact animal.

Co-cultures of LTER monolayers and normal mouse thymocytes contain many different cell types. Apart from the various normal cortical lymphocytes which are added in suspension, the leukemic monolayer contains medullar epithelial cells, cortical epithelial reticular cells and periodic acid-Schiff (PAS)-positive cells that also originate in the thymic cortex. Fibroblasts are essentially absent from our cultures; when present, these can be quantitatively removed by mild trypsin-versene treatment of the monolayers (Haas ef al., 1977). One cannot exclude the possibility that some viable lymphocytes remain within this group of diverse cells of the monolayer. Such lymphocytes would induce leukemia when inoculated into recipient mice. Microscopically, the LTER monolayers are free of lymphocytes. More- over, no leukemia-inducing cells could be concen- trated from supernatants of LTER monolayers (Haas et al., 1977). Nevertheless, the possibility that some leukemic cells from the LTER monolayer sneak through and are the cause of leukemias in the inoculated mice, rather than the in vitro transformation of the suspended thymocytes, needs to be rigorously investigated. Data supporting the in vitro transformation thesis are the subject of this communicat ion.

We used the Ly thymocyte genetic marker system (Cantor and Boyse, 1975) to differentially label cells of the LTER monolayer as well as the suspended thymocytes. The use of three congeneic cell lines on a C57BL/6 (B6) background (B6; B6-Ly-1.1 and B6-Ly-2.1) enabled this approach to be used (Shiku et al., 1975). In this communication we also report that most of the tumors produced by inoculation of LTER monolayers into recipient animals were not lymphosarcomas, unless thymocytes were co-inoculated with the epithelial cells.

MATERIAL A N D METHODS

Mice C57BL/6 (B6), B6-Ly-1 .I and B6-Ly-2.1 female

mice were used. Mice were bred at the Weizmann Institute Animal Breeding Center. Thymomas were induced in 6-week-old mice by intrathymic injection of the radiation leukemia virus (Haran-Ghera et al., 1966, 1977). When used, mice were thymectomized at the age of 5-6 weeks and received the injection of in vitro-induced leukemic cells 3-4 weeks

Received: August 15, 1977 and in revised form October 26, 1977.

116 HAAS

after the operation. Thymectomized mice were always checked when moribund for total absence of the thymus. For Ly marker analysis only intact mice were used, since indirect immunofluorescence was possible only on thymic tumor cells. Normal thymocytes were removed from 6-week-old mice and were dispersed under sterile conditions through a stainless steel grid, and washed in PBS. Injection of thymocytes after in vitro cultivation on LTER monolayers was performed intraperitoneally.

Thynurs epithelium and other culture methods Preparation of TER monolayers from normal

and leukemic mice, as well as the trypsinization of the monolayers to remove any fibroblasts, have been described previously (Haas, 1976; Haas et al., 1977). Thymocytes were cultivated on LTER cells in Eagle's medium. Normally, 20 x lo6 thymocytes were added to a LTER culture containing 1-2x 10n reticulum monolayer cells in a 60-mm tissue culture dish. In some experiments as many as 50 x lo6 thymocytes were cultivated per LTER culture with no apparent difference in experimental results. Removal of thymocytes at the end of the cultivation period was done by gentle shaking of the medium and one or two additional washes. LTER cells were quantitatively retained on the dish by this method. During incubation of thymocytes on LTER monolayers, thymocytes underwent marked blast transformation.

TABLE I

TUMOR DEVELOPMENT BY INOCULATION OF Ly-1.1 OR Ly-2.1 THYMOCYTES CULTURED ON B6 LEUKEMIC

THYMUS EPITHELIAL RETICULUM MONOLAYERS '

Exp, Source of Tumor incidence and No' Of No, thymocytes average latent period l y m p ~ ~ ~ ~ m a s '

tumors obtained (days)

1 B6-Ly-1.1 14/17 (37) 12/14 2 B6-Ly-2.1 10/12 (41) 9/10

'loa live thymocytes cultured for 72 h on 8 6 LTER monolayers were inoculated IP into B6 recipient mice. Mice were checked for palpable tumors at weekly intervals. Mice with tumors were killed, internal organs were prepared for histology and one half of each thymoma was analyzed by indirect immunofluorescence for Ly surface antigens.

It~iniunoJluurescence techniques Ly antigens on the surface of thymocytes were

determined by indirect immunofluorescence on live cells. All operations were done at 20" C. Normal thymuses or thymomas were passed through a stainless steel grid and the cells were washed in PBS. Twenty million cells in 0.05 ml PBS were mixed with 0.05 ml antiserum at a dilution of 1:40 in PBS. We used Ly 1.1 antiserum (lot ESlp.49) and anti-Ly 2.1 serum (lot ESlp.45) (obtained from Drs. F.W. Shen and E.A. Boyse, Memorial Sloan- Kettering Cancer Center), as well as anti-Ly 1.1 serum (lot Z 625) received from Dr. Marc Feld- mann, University College, London. After incubation

for 1 h at 37" C the cells were washed three times in PBS then incubated at 37" C for 1 h with fluorescent FITC-conjugated anti-mouse IgG serum, made in rabbit, at a dilution of 1 :20 in PBS (Miles- Yeda, Rehovot, Israel). Cells were washed three times in PBS and immediately observed under an epi-illumination UV microscope. Only intact, live

TABLE I1

INDIRECT IMMUNOFLUORESCENCE OF Ly MEMBRANE ANTIGENS ON THYMOCYTES OF B6-Ly-1.1, B6-Ly-2.1

AND B6 CONGENEIC MICE, AND ON RadLV-INDUCED 8 6 THYMOMA CELLS '

Reacted with antiserum to

Thymus cells Ly-1.1 Ly-2.1

No. % No. % ~

B6-Ly-1.1 280/340 82 3/402 < 1 B6-Ly-2.1 0/122 0 158/186 85 B6 4/185 2 OjlOS 0 B6 thymoma2 12/214 6 18/278 7

I Live thymus or thymoma cells were incubated for 1 h at 37" C with the indicated antisera, washed extensively and incubated asdin with fluorescent anti-mouse 1gG. Cells were washed and immediately observed under UV microscope with alternate visible and UV light. Numbers represent number of cells showing membrane fluorescence out of total counted in the same fields. - Fluorescence of B6 thymomas was due to the presence of immunoglobulin producing cells that stained also when reacted only with fluorescent anti-mouse IgC.

cells were included when determining the percentage of cells showing thin peripheral membrane fluorescence by counting the cells alternately with UV and visible light exposure. A typical preparation contained 1 % or less dead cells. Normal thymocyte controls contained less than 1 % immunoglobulin- producing cells (cells fluorescing after staining with anti-mouse IgG serum), whereas thymoma cells typically contained 2-5 % such cells. Ly membrane antigen determinations were, of course, done only on thymic tumors and not on any of the other tumors obtained.

R ESU LTS

Tumor induction and Ly immunojuorescence controls Thymocytes of B6-Ly-1.1 or of B6-Ly-2.1 young

female mice were cultivated for 72 h on leukemic thymus epithelial reticulum monolayers grown from RadLV-inoculated B6(Ly-2.2) mice. One million live cultivated thymocytes were then inoculated I P into 86 recipient mice. Table I shows two typical experiments. Some 80% of the inoculated mice developed tumors within 5-6 weeks. All thymic tumors were suspected of being lymphosarcomas and were immediately assayed for Ly membrane antigens by indirect immunofluorescence (IF). Tumors and internal organs were processed for histology. Most, but not all, of the tumors obtained were lymphosarcomas. The efficacy of using indirect IF for the determination of Ly antigens of thymocytes is demonstrated in Table 11. Immunofluorescence

GENETIC ANALYSIS OF IN

of Ly antigens was consistent in many experiments and enabled clear-cut determination of Ly-1.1, Ly-2.1 and Ly-2.2 antigens. The background was low, as compared to complement-mediated cytotoxicity. Table 11 also contains results of IF-staining of a 136-virus-induced thymoma. In order to identify Ly antigens on the surface of thymoma cells it was important to establish that the antisera at our disposal did not recognize leukemia-related antigens. As shown in Table 11, the antisera did not stain B6-thymoma cells; the 6-7 % staining thymoma cells were immunoglobulin-bearing cells which also stained IF-positive when Ly antisera were omitted from the reaction.

VITRO LEUKEMOGENESIS 117

Cellular origin of in vitro-induced thymomas

Thymomas obtained following inoculation into B6 mice of Ly-1.1 or Ly-2.1 marked thymocytes following cultivation on B6 LTER monolayers, could originate from any one of three different sources: they could be leukemic lymphocytes trans- ferred from the LTER monolayer; they could be lymphocytes of the recipient rendered leukemic after induction by inoculated LTER cells; or, alternatively, they could originate from in vitro- transformed Ly-marked thymocytes. In the first two contingencies the thymoma cells would carry the B6 marker Ly-2.2, whereas in the latter contin- gency the thymoma cells would carry the Ly-1.1 or Ly-2.1 markers. Table 111 shows that the thymomas obtained originated from in vitro-transformed thymocytes. B6-Ly-I. 1-incubated thymocytes induced thymomas that carry Ly-1.1 antigen but no Ly-2.1 antigen; B6-Ly-2.1 -incubated thymocytes induced

thymomas that only stained with Ly-2.1 antiserum. The percentage of Ly-positive cells in the thymomas is low (27 to 34%), compared with that found by immunofluorescent staining for Ly antigens of normal thymocytes. The reason for this may be that thymocytes cultivated on LTER monolayers first ‘‘ home ” to the spleen, and only later cause thymic lymphosarcomas, probably through dissemination.

TABLE 111

Ly SURFACE MARKER ASSAY OF I N VITRO-INDUCED THYMOMAS ’

Genotype of No. of Thymoma cells reacting with antiserumx thymocytes cultured thymomas

on LTER cells assayed Ly-1.1 (X) Ly-2.1 (%)

B6-Ly-1.1 9 34 (1 8-54) 4 (0-7) B6-Ly-2.1 9 2 (0-5) 27 (11-46)

In vivo B6 thymomas that were obtained by inoculation of B-6- Ly-1.1 or B6-Ly-2.1 thymocytes cultivated on 86 LTER monolayers were stained for Ly markers by means of indirect immunofluorescence. - The percentage of thymoma cells reacting with fluorescent anti-mouse IgC serum (5-8 %) was subtracted from the percentage of cells staining with the Ly antisera. The numbers refer to the average percentage of stained cells in the group of nine thymomas tested. The range is given in parenthesis.

Thus, the thymomas are relatively small and only part of the thymus cells may have been replaced by tumor cells.

These results show that some 80% of the tumors induced in the recipient mice were lymphosarcomas

FIGURE 1 - Mouse ‘60-1-1. Lymphatic leukemia infiltration in the pan- creas. The mouse was inoculated IP with log live Ly-1.1 thymocytes that had been cultured for 72 h on a B6 LTER monolayer. The mouse was found to have a solid tumor (lymphosarcoma) at the site of inoculation, an enlarged leukemic thymus, an infiltrated liver and almost normal spleen and peripheral lymph nodes. Other mice in the same group (not shown) had disseminated lymphatic leukemia with grossly enlarged spleens and lymph nodes. x 800.

118 H A A S

FIGURE 2 - Mouse 50-2-11. Reticulum cell neoplasm type A in mouse thymus. The mouse was inoculated with lo6 cultured LTER cells. Twenty- eight days later, when moribund, the mouse had a solid tumor at the site of injection, an enlarged spleen and enlarged lymph nodes. x 800.

that carried the marker of the cultivated thymocytes. Figure 1 shows a thymic lymphosarcoma induced by B6-Ly-I .I-incubated thymocytes. In these experiments we observed that some other, non-lymphosarcoma tumors were also induced. The question arose as to what types of tumors these were and whether they arose from leukemic thymocytes or from cells carried over from the LTER monolayer.

LTER mor1olayer cells cause tumors that are not lymphosarconias

Five out of a total of 29 mice inoculated with thymocytes that were cultivated on LTER monolayers developed tumors that were not lyniphosar- comas. We surmised that these tumors were caused by concurrent inoculation of epithelial reticulum cells. Therefore we inoculated several B6 mice

FIGURE 3 - Mouse 50-2-10. Myeloid leukemia in the liver. The mouse was inoculated with lo8 cultured LTER cells. Spleen and liver tumors were myeloid leukemias. The thymus was not involved, but peripheral lymph nodes were enlarged. x 800.

GENETIC ANALYSIS OF I? ‘I VITRO LEUKEMOGENESIS 119

with lo6 cells of every LTER monolayer prepared. Out of a total of 22 mice inoculated IP with epithelial cells, two mice developed disseminated lymphosarcoma; the other tumors were equally divided between reticulum cell neoplasms (RCNA) (Dunn, 1954) and myeloid leukemias. Figures 2 and 3 show the histology of some of the tumors which we obtained. A complete study of the tumors induced by leukemic thymus epithelial cells is beyond the scope of this communication, and will be presented elsewhere. It is of relevance here that the LTER monolayer cells on which no thymocytes were cultivated induced few lymphosarcomas. This point strengthens the premise that thymocytes underwent leukemogenic transformation after incubation on LTER monolayers, whereas the monolayer cells themselves caused few lymphosarcomas. The majority of the monolayers (20/22) harbored no leukemic cells.

DISCUSSION

The experiments reported here show that normal thymocytes incubated on monolayers of leukemic thymus epithelial reticulum cells underwent leukemogenic transformation. There was some doubt whether the leukemia-inducing cells originated from the transformed thymocytes or from cells of tae monolayer. The use of differentially Ly-antigen- labelled thymocytes and monolayers showed that in vitro leukemogenic conversion of thymocytes took place efficiently (Table 111). During co-cultivation of thymocytes on LTER monolayers, thymocytes underwent blast transformation in addition to infection with virus (Haas et a[., 1977). Both events may be necessary for leukemogenic conversion of the thymocytes, since either their infection with LTER supernatant virus (Haas, unpublished), or blast formation of thymocytes on normal TER cells were not sufficient to induce in vitro leukemogenic conversion. It must also be emphasized at this point that, although viability of thymocytes incubated on LTER cells was four times longer than their viability when incubated on B6 fibroblast feeder layers, or without any feeder layer, the leukemic cells did not grow to become in vitro immortal lines.

Additional evidence for in virro leukemogenic transformation of the thymocytes comes from control experiments in which the LTER monolayer

cells were inoculated into mice. Had the monolayers harbored leukemic lymphocytes, mice inoculated with the LTER cells should have developed lymphatic leukemia. Moreover, on the basis of experiments done by Law (1952), Hays (1968) and others, we expected that inoculated leukemic TER cells might recruit recipient lymphocytes and induce lymphatic leukemia. The above authors showed that this is what happened upon grafting of normal TER cells and subsequent injection of Gross virus. Our in vitro-grown LTER cells may have multiplied and caused mostly myeloid leukemia or RCNA, but little lymphatic leukemia. The question whether the inoculated LTER cells caused RCNA or myeloid leukemia in the recipient through induction rather than through independent cell multiplication cannot be settled from the present experiments. It will be interesting to see whether similarly inoculated mice will develop lymphatic leukemia at a later time if they do not succumb to early RCNA or myeloid leukemia, e.g. by inoculation of a small number of LTER cells.

We have shown previously (Haas et al., 1977) that LTER cells grown from RadLV-injected mice were virus-infected. Data presented here suggest that the LTER cells themselves are capable of inducing non-lymphoid tumors in mice inoculated with such cells. Why do RadLV-injected mice develop mostly lymphatic leukemia, whereas epithelial tumors are infrequently observed, even though virus-infected epithelial cells are present in their thymus? Our results prompt us to speculate that thymus epithelial cells may play a mediating role in the development of RadLV lymphomas. Support for this view is given by the pathology of murine leukemias (Siegler, 1968).

ACKNOWLEDGEMENTS

This work was supported by grants from the US-Israel Binational Science Foundation and by the Leukemia Research Foundation, Inc., USA. I thank Dr. Asher Meshorer for assistance with histological determinations and Profs. Michael Feldrnan and Irun Cohen for support. Mrs. Rivka Karakash, Mrs. Tamara Reshef and Mr. Shalom Gohali provided excellent technical assistance.

I thank Drs. Edward A. Boyse, Fung-Win Shen and Marc Feldmann for generously supplying the Ly antisera used in this study.

ANALYSE GENETIQUE DE LA LEUCEMOGENESE I N VfTRO INDUITE PAR DES CELLULES EPITHELIALES DU THYMUS TRANSMETTANT LES VIRUS

DE LA LEUCEMIE MURINE

Une leucemie lympholde est apparue chez des souris C57BL/6 apres inoculation de thymocytes normaux en co-culture sur des monocouches de cellules leucemiques du reticulum epithelial du thymus (LTER). En utilisant des thymocytes gene- tiquement marques par les antigenes memlxanaires Ly, on a demontre que les cellules qui produisaient les thymomes ttaient les thymocytes co-cultives plutbt que les cellules leucemiques en monocouches. La conversion leucemogene des thymocytes normaux s’est donc produite in virro. L’inoculation de cellules LTER cultivks en monocouches a produit surtout des reti- culosarcomes et des leucemies myeloldes plutBt que lympholdes (ces dernieres sont apparues apres inoculation de thymocytes cultives sur les monocouches de LTER). Les cellules LTER peuvent donc &tre des cellules tumorales capables de produire des neoplasmes a cellules reticulaires ou des tumeurs myeloldes, outre qu’elles peuvent transformer les thymocytes normaw en cellules leucemiques.

1 20 HAAS

REFERENCES

CANTOR, H., and BOYSE. E. A., Functional subclasses of T lymphocytes bearing different Ly antigens. J . exp. Med., 141, 1376-1389 (1975).

DUNN, T. B., Normal and pathologic anatomy of the reticular tissue in laboratory mice, with a classification and discussion of neoplasms. J. nat. Cancer Inst., 14, 1281-1433 (1954).

HAAS, M., Radiation leukemia virus infection of thymocytes during in vitro leukemogenesis on thymus epithelium monolayer cultures. In: J. F. Duplan (ed.), International symposium on radiation-induced leukemogenesis and related viruses. INSERM Symposium No . 4 , pp. 149-165, North-Holland Biomedical Press, Amsterdam (1976).

HAAS, M., SHER, T., and SMOLINSKY, S., Leukemogenesis in vilro induced by thymus epithelial reticulum cells transmitting murine leukemia viruses. Cancer Res., 37, 1800-1807 (1 977).

HARAN GHERA, N., LIEBERMAN, M., and KAPLAN, H. S., Direct action of a leukemogenic virus on the thymus. Cancer Res., 26, 438-442 (1966).

HARAN GHERA, N., BEN YAAKOV, M., and PELED, L., Immunologic characteristics in relation to high and low leukemogenic activity of radiation leukemia virus variants. J. Immunol., 118, 600-606 (1977).

HAYS, E. F., The role of thymus epithelial reticular cells in viral leukemogenesis. Cancer Res., 28, 21-26 (1968).

KAPLAN, H. S., Influence of thymectomy, splenectomy and gonadectomy on the incidence of radiation-induced lymphoid tumors in strain C57B1 mice. J. naf. Cancer Inst., 11, 83-90 (1950).

KAPLAN, H. S., BROWN, M. B., and PAULL, J., Influence of post-irradiation thymectomy and of thymic implants on lymphoid tumor incidence in C57BI mice. Cancer Res., 13. 677-680 (1953).

LAW, L. W., Increase in incidence of leukemia in hybrid mice bearing thymic transplants from a high leukemic strain. J. nut. Cancer Insf . , 12, 789-805 (1952).

MCENDY, D. P., BOON, M. D., and FURTH, J., On the role of the thymus, spleen and gonads in the development of leukemia in a high leukemia stock of mice. Cancer Res., 4, 377-383 (1944).

of the thymus, spleen and gonads in the development of leukemia in a high leukemia stock of mice. Cancer Res., 4, 377-383 (1944).

SHIKU, H., KISIELOW, P., BEAN, M. A., TAKAHASHI, T., BOYSE, E. A,, OETTGEN, H. F., and OLD, L. J., Expression of T-cell differentiation antigens on effector cells in cell-mediated cytotoxicity in virro. J. exp. Med., 141, 227-241 (1975).

SIEGLER, R., Pathology of murine leukemias. In: M. A. Rich (ed.), Experimental leukemia, pp. 51-95 (1968).

WAKSAL, S. D., SMOLINSKY, S., COHEN, I. R., and FELDMAN, M., Transformation of thymocytes by thymus epithelium derived from AKR mice. Nature (Lond.), 263, 512-514 (1976).

genetic analysis of in vitro leukemogenesis induced by thymus epithelial reticulum cells...

Documents