See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/19635599

Tumor dormancy. I. Regression of BCL1 tumor and induction of a dormant

tumor state in mice chimeric at the major histocompatibility complex

Article  in  The Journal of Immunology · September 1986

Source: PubMed

CITATIONS

47READS

93

4 authors, including:

Some of the authors of this publication are also working on these related projects:

Peptoid View project

T and B cells View project

Henry Siu

Miami Dade College

2 PUBLICATIONS   111 CITATIONS   

SEE PROFILE

Ellen S Vitetta

University of Texas Southwestern Medical Center

546 PUBLICATIONS   23,429 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Ellen S Vitetta on 25 August 2017.

The user has requested enhancement of the downloaded file.

of August 25, 2017.This information is current as

histocompatibility complex.in mice chimeric at the majortumor and induction of a dormant tumor state Tumor dormancy. I. Regression of BCL1

H Siu, E S Vitetta, R D May and J W Uhr

http://www.jimmunol.org/content/137/4/13761986; 137:1376-1382; ;J Immunol 

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. All rights reserved.Copyright © 1986 by American Association of Immunologists1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

VOi. 137. 1376-1382. NO. 4, August 15, 1986 Prfnted in U.S.A.

0022-1767/86/1374-l376$02.00/0 THE JOURNAL OF IMMUNOLOGY Copyright 0 1986 by The American Association of Immunologists

TUMOR DORMANCY

I. Regression of BCLl Tumor and Induction of a Dormant Tumor State in Mice Chimeric at the Major Histocompatibility Complex'

HENRY SIU, ELLEN S. VITETTA, RICHARD D. MAY, AND JONATHAN W. UHR2 From the Department of Microbiology, Southwestern Medical School, University of Texas Health Science Center,

Dallas, T X 75235

The growth of the BCLl tumor in murine H-2 chi- meras was studied. Lethally X-irradiated BALB/c mice were reconstituted with C57BL/6 bone marrow that had been depleted of T cells. When chimerism was established 90 to 120 days later, large doses of BCLl cells were injected. The tumor grew progres- sively, reaching a peak level of as many as lo9 tumor cells per animal by 40 days after inoculation. After that time, the tumor regressed in all the chimeric animals, and by 100 days after inoculation, vir- tually all the animals appeared disease free as judged by an absence of BCLl-idiotype-positive cells in the spleen and peripheral blood, a normal spleen size, and absence of an elevated white blood cell count. Such animals were followed for as long as 8 mo after tumor inoculation and remained disease free. However, transfer of graded numbers of splen- ocytes from these animals into normal BALB/c re- cipients resulted in development of tumor in recip- ients receiving 100 or more spleen cells. These re- sults indicate a large tumor burden in the spleen of each donor, namely, 10'to lo7 BCLl cells. The pres- ent model should facilitate characterization of the mechanisms underlying tumor dormancy.

BCL, is the first spontaneous murine B cell lymphoma/ leukemia to be described (1). It arose in an elderly BALB/ c mouse and has been subsequently maintained in vivo for 9 yr by i.v. passage of frozen cells (from the second passage] into syngeneic animals. A single viable tumor cell transferred into a virgin BALB/c mouse causes pro- gressive disease in approximately 50% of recipients de- tectable by 12 wk after injection (2). One month after i.v. injection of lo6 BCL, cells into BALB/c mice, the spleen enlarges, followed several weeks later by leukemia and hepatomegaly. At the peak of disease, there is profound splenomegaly (2 to 10 x lo9 cells/spleen) and severe leukemia (1 0' cells/ml blood], resulting in a tumor burden of approximately 10" BCL, cells (3). Such mice die 2 to 5 mo after the initial injection of 1 O6 BCLl cells. The disease

Received for publication March 10, 1986. Accepted for publication April 22. 1986.

payment of page charges. This article must therefore be hereby marked The costs of publication of this article were defrayed in part by the

aduertisement in accordance with 18 U.S.C. Section 1734 solely to indi- cate this fact.

' This work was supported by National Institutes of Health Grants CA- 28149 and AI-19640.

Address all correspondence: Department of Microbiology. University of Texas Health Science Center at Dallas, 5323 Harry Hines Boulevard. Dallas. TX 75235.

in the mouse closely resembles human prolymphocytic leukemia, a subset of chronic lymphocytic leukemia (4).

We previously described several experimental situa- tions in which BCLl tumor cells become dormant in the host: i.e., progressive disease is absent, although viable tumor cells can be demonstrated in host tissues: 1) In one of 20 lethally irradiated recipients of BCL1-containing bone marrow (BM)3 that had been treated with anti- immunoglobulin (anti-Ig) A chain-immunotoxin before transplantation, the recipient appeared tumor-free 25 wk after transplantation. However, its splenocytes trans- ferred tumor to normal secondary recipients (5. 6). 2) Animals bearing large burdens of BCL, were treated with total lymphoid irradiation and splenectomy, followed by injection of anti4 or anti-BCL, idiotype (Id)-A chain im- munotoxin. These mice remained disease free for the 4 to 7 mo of observation in three of four experiments. In one experiment such animals were followed for 7 mo, after which time their tissues were transferred to normal mice. All recipients of such tissues showed evidence of tumor (6, 7). Because a single BCLl cell can kill a virgin animal in 12 wk and the disease-free tissue donors had no evidence of leukemia or Id+ cells in their blood, the tumor cells appeared to be in a dormant state in these animals. 3) Fifty percent of BCL,-bearing mice receiving nonspecific cytoreduction (total lymphoid irradiation and cyclophosphamide), followed by the injection of T cell- depleted allogeneic BM, developed a dormant tumor state: the other 50% demonstrated total eradication of BCLl tumor (8, 9).

The present experiments describe a model that exploits the antitumor effect of the chimeric state induced by T cell-depleted allogeneic BM transplantation. The studies indicate that in H-2-chimeric mice, BCL, cells can grow progressively for 4 to 5 wk and reach a peak tumor burden of about lo9 BCL, cells per mouse before the onset of tumor regression. By 12 wk after inoculation with BCL, cells, such animals appear tumor free as judged by normal spleen size, absence of Id+ cells in the spleen, and a normal white blood cell count. Cell transfer studies, however, reveal that 100% of these animals carry a large burden of tumor cells, i.e., 1 to 10% of their splenocytes ( lo6 to lo7) are dormant BCL, cells. This

Abbreviations used in this paper: BM, bone marrow: GARIg. goat anti- rabbit immunoglobulin; Id. idiotype; MoAb. monoclonal antibody; NRIg.

immunoglobulin; RAMy. rabbit anti-mouse 7-chain; Rald, rabbit anti- normal rabbit IgG; NRtIg. normal rat 1gG; RAMIg. rabbit anti-mouse

BCL, idiotype.

1376

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

TUMOR DORMANCY IN H-2 CHIMERIC MICE 1377

model should facilitate characterization of the mecha- nisms underlying tumor dormancy.

MATERIALS AND METHODS

Mice. BALB/c and C57BL/6 mice were obtained from The Jackson Laboratory, Bar Harbor, ME.

Cells. The BCL, tumor was maintained in vivo by i.v. passage in BALB/c mice. Five weeks after tumor passage of 5 X 1 O5 spleen cells from a tumor-bearing animal, mice were sacrificed and their splen- ocytes were used as a source of tumor cells.

Spleen cell suspensions were prepared in balanced salt solution (BSSI, and the stromal tissue and debris were removed by allowing the cell suspensions to stand for 5 min. The supernatants were then removed, diluted to 40 ml with BSS a t 2 0 T , underlayed with 10 ml Ficoll-Hypaque (P = 1.09). and centrifuged at 400 X G for 20 min at 20°C. The fluid above the pellet was removed and washed with cold BSS. This cell pellet was resuspended in hemolytic Gey’s solution to lyse any remaining erythrocytes in the cell suspension. The cells were maintained at 4°C or 5 min, then underlayed with 5 ml of 100% fetal calf serum (FCS; HyClone. Logan, UT) and centrifuged. Cells were washed twice with BSS before use.

Ficoll-Hypaque, and treated with hemolytic Gey’s solution to remove Peripheral blood was collected in heparinized tubes, separated on

dead cells and erythrocytes as described above. BM cells were prepared by flushing cold BSS through the cavities

of the hind limbs of healthy mice 10 to 12 wk of age. Single cell suspensions of BM cells were obtained by gently pipetting cells with Pasteur pipettes. Bone, stromal tissue, and debris were removed by allowing the cells to stand at 4°C for 5 min. The supernatant was removed and the cells were washed twice with BSS.

Antisera. Rabbit anti-BCL, Id (Rdd) was prepared and absorbed as described (3). Briefly, the BCL, IgMX from the ascites fluid of BALB/c mice bearing BCL, X P3/X63-Ag8 hybridoma cells was purified by chromatography on S-200 and affinity purification on Sepharose-anti-g. A hyperimmune rabbit antiserum directed against this IgMX was purified by affinity chromatography on Sepharose bound to the BCLI-IgMh. Antibody was eluted in 3.5 M MgCI,, di- alyzed. concentrated. and absorbed sequentially with Sepharose bound to mouse euglobulin, MOPC-104E (p. k), MOPC-21 (y, K ) , TEPC

spleen cells. The resulting anti-Id antibody was judged to be specific 183 (p . K ) , MOPC-315 (a. X), and paraformaldehyde-fixed BALB/c

by the immunochemical and immunofluorescence criteria described previously (3).

chromatography on DEAE-Sephadex A-50. Normal rabbit IgG (NRIg) Normal rat IgG (NRtlg) was prepared from pooled rat serum by

and rabbit anti-mouse Ig (RAMIg) were prepared as described (10). Goat anti-rabbit Ig(GARlg] was generated, and the IgG fraction was coupled to fluorescein isothiocyanate (FITC: Sigma, St. Louis, MO)

directed against H-2Kb and H-2Db (28-8-6S), H-2Dd (34-5-883, and H- Culture supernatants containing monoclonal antibodies (MoAb)

2K‘ and H-2Dk (15-3-1s) were purchased from Litton Bionetics. Charleston, SC. All three MoAb are of the IgG2a isotype, and there IS no haplotypic cross-reactivity (1 2- 14). Rabbit anti-mouse y serum (RAM71 was prepared (15) and its F(ab’), fragments were coupled to FITC (1 1, 16). This antibody reacts with rat and mouse IgG. MoAb anti-Thy-1.2 (murine IgM) was isolated from culture supernatants of the HO-13.4 hybridoma by ammonium sulfate precipitation (17). The IgG fraction from MoAb anti-Thy-1.2 (rat IgG2b) was purified from the supernatants of 30-H12 hybridoma cells (18).

Irradiation of animals. Five-month-old female BALB/c mice were used, since younger mice were unable to tolerate whole body x- irradiation as well as 5-mc-old mice. Two weeks before irradiation, mice were housed in sterile cages in racks fitted with HEPA filters and laminar air flow systems (Contamination Control Inc., Lamsdale, PA]. Cages contained autoclaved wood shavings, food. and drinking water. Seven to 10 days before irradiation, an antibiotic mixture consisting of 1 mg cephalothin, 1 mg ampicillin, and 2 mg carbeni- cillin was added to each milliliter of drinking water. Lethal whole body irradiation (850 to 950 rad) was administered with a 137Cs source (Gammacell 40; Atomic Energy of Canada. Ottawa, Canada). All irradiated animals were maintained as described above.

Treatment of BM cells and reconstitution of irradiated animals. Washed BM cells were resuspended at 2 x 107/ml in BSS and incubated at 4°C for 40 min with an optimal concentration of MoAb anti-Thy-1.2 (HO-13.4). Cells were then centrifuged, resuspended at 5 X 107/mI in a 1/8 dilution of baby rabbit complement, and incu- bated at 37°C for 30 min. Cells were then washed three times in BSS and filtered through nylon mesh to remove any aggregated cells or debris. T-depleted BM cells (1 0’) from syngeneic (BALB/c) or alloge- neic (C57BL/6) donors were resuspended immediately before use in

(11).

0.2 ml of BSS containing 1 mg of cephalothin, 1 mg ampicillin, and 1 mg of carbenicillin/ml and injected into the tail vein of each of the mice that had been lethally irradiated 1 day before BM reconstitu- tion.

Indirect immunofluorescence and analysis of cells by thefluo- rescence-activated cell sorter IFACS) 119). Cells (lo6) were sus- pended in BSS containing 10 mM sodium azide (BSS-Azide) and were incubated with optimal concentrations of the primary antibodies

2Dd, or MoAb ~yH-2~) at 4°C for 30 min. Cells were then washed, resuspended in BSS-Azide. and incubated with optimal concentra- tions of the appropriate secondary antibodies (FITC-GARlg or FITC- F(ab’)Z-RAMy] a t 4OC for 15 min. Cells were subsequently washed and analyzed on the FACS-Ill (B-D FACS Systems, Becton-Dickenson and Co.. Sunnyvale, CA). Control staining values were subtracted.

Growth of BCL, cells in chimeric mice. Three to 4 mo after irradiation and BM reconstitution. individual mice were assayed for chimerism by indirect immunofluorescence and FACS analysis (see above) of erythrocyte-free peripheral blood cells. Viable BCLl cells (5 x IO5. 5 x IO6 or 5 x IO‘) were then inoculated i.v. into groups of IO chimeric animals. Control animals that had received syngeneic BM cells were inoculated with 5 X IO5 viable BCLl cells only. At various intervals after inoculation with BCLI cells. animals were sacrificed: their spleens were removed, teased into single cell suspensions, counted, and analyzed on the FACS for the percentage of Id+ cells.

Detectlon ofdormant tumor cells. At various times after injection with BCL, cells, spleen cells were obtained from the chimeric recip- ients. Graded numbers of viable cells from these spleens were inoc- ulated i.v. into groups of normal BALB/c animals. Normal BALBlc spleen cells were added as “filler cells” in the inoculum such that a total of lo6 cells was injected into each recipient.

Dounce homogenization of BCL, cells and preparation of a cellfree extract. Spleen cells (2 X 10‘) from BCL,-bearing mice in 10 ml BSS were disrupted by approximately 100 strokes in a tight- fitting Dounce homogenizer. The homogenate was centrifuged at 5200 X G for 30 min at 4°C. The supernatant was removed and filtered through a 0.22-pm Millipore filter. Groups of five to 10 normal BALB/c mice received homogenates containing 2 X IO’ or 1 x lo6 cell equivalents. Mice were followed for 8 mo and examined for the presence of tumor.

(Rald, RAM& NRIg, MoAbolThyl.2. NRtlgG. MoAb aH-Zb. MoAb aH-

RESULTS

Establishment of BM chimeras. Ninety to 120 days after lethal irradiation and transplantation of T cell- depleted allogeneic BM cells, the mice were tested for chimerism. Peripheral blood cells, free of erythrocytes, were analyzed by indirect immunofluorescence on the FACS to determine the percentages of H-2b- and H-2d- positive cells. As shown in Table I, a chimeric state was established in the six representative mice shown. Thus, 85 to 92% of the blood cells were of donor type in these mice. These results are representative of those obtained in over 200 animals that were prepared and evaluated as described above.

Kinetics of growth of BCLt in chimeras. The growth

TABLE I H - 2 typing of BM chimeras”

Percentage of Cells Stained’ with MoAb

Anti-H-Zb Anti-H-2d

Source of Cells

C57BL/6 (H-2b)

BCL, cells (H-zdl BALB/c (H-zd)

Chimera 2 Chimera 1 (H-2b

Chimera 3 Chimera 4 Chimera 5 Chimera 6

97 6 2

91 91

88 85 90 92

+ H-2d)

4 94 96

4 5 9

13 6 7

8s (antl-H-2Dd]. 15-3-1S(anti-H-2Kk and -H-2Dk), and FITC-F(ab’12-RAMr a Cells (lo6) were stained with 28-8-6s (anti-H-2Kb and -H-2Db), 34-5-

and analyzed on the FACS. Cells stained with FITC-F[ab’),-RAM? alone were less than 1 %. The

percentage of cells stained with MoAb ar1ti-H-2~ served as a negative control. These values [less than 2%) were subtracted from the values reported here.

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

1378 TUMOR DORMANCY IN H-2 CHIMERIC MICE

of BCLl tumor cells in established chimeras was moni- tored by FACS analysis of spleen cells for the presence of Id+ cells (Fig. 1). In normal BALB/c animals, 5 X lo5 BCLl cells injected i.v. grew rapidly and reached a growth plateau after 40 days. These animals had as many as 5 X 10' cells/spleen, 70 to 80% of which expressed the BCL, Id. These animals all died within 90 days. In con- trast, the group of chimeric animals injected with 5 x lo5 BCL, cells demonstrated a different growth pattern: 1) The initial rate of growth of tumor was similar, as shown by the slope of the growth curves during the first 2 to 3 wk. 2) The maximum tumor burden was 10 times smaller by 6 wk after inoculation (5 X 1 07/spleen). 3) The growth curve did not reach a plateau. After 6 wk, there was a rapid decrease in the number of Id+ cells in the spleens of these animals. The size of the spleens of these mice paralleled the kinetics of BCLl growth. By 12 wk after tumor inoculation, spleen sizes of virtually all chimeric animals were indistinguishable from those of uninocu- lated controls, and Id+ cells were no longer detectable (see below).

To determine if larger tumor inocula could alter the growth pattern in chimeric mice, the doses injected were increased 10-fold (5 X lo6) and 100-fold (5 X lo7). The numbers of BCL, cells per spleen at the peak of the curves were 2 X lo8 and 6 X 10'. respectively. However, the same general growth pattern described above for 5 X lo5 BCL, cells (Fig. 1) was seen, i.e., an apparent complete regression of BCLl tumor by 100 days after inoculation.

These results indicate that in chimeric mice, the initial growth of the tumor is similar to normal controls, but tumor cells are subsequently eliminated from the spleens of the chimeric mice.

lmmunophenotyping of splenocytes from chimeric animals inoculated with BCL, tumor. We next deter- mined whether the spleen cells from mice in which BCL, tumor had regressed have an abnormal distribution of T and B cells. Spleen cells from chimeric mice that had received BCLl 78 to 104 days previously were phenotyped for surface Ig and Thy- 1.2 as well as Id. A representative experiment (Table 11) shows that there were no detectable Id+ cells in chimeric animals 88 to 104 days after the

SI-

I d Doys After Injection

ents. Viable BCLl cells-5 X lo5 (O...O), 5 X 10' (U--0). or 5 X lo' FLgure 1 . Growth of BCL, tumor in spleens of chimeric recipi-

(A-.-.-A)-wereinjectedi.v.intoestablishedchimericrecipients.Controls ( 0 - 0 ) were injected with 5 x lo5 viable BCL, cells. Assays were performed by anti-Id staining and FACS analysis. Each point represents the average of values obtained from two experiments in which one spleen was assayed for each time point in each experlment: the 5 X lo6 dose was only

least squares fitting through experimental data points. employed in one experiment. Computer generated lines were obtained by

TABLE 11 Immunophenotyping ofsplenocytesfrom normal. tumor-bearing. and

chimeric mice

Days after Percentage of Cells Posi- Animals Tumor Injec- tive f o f

tlona Id sIg T h ~ - l . 2 ~

Normal BALBlc 1 2 3

BALB/c

None given None given None given

recipient 1 50 2 3

60 70

1 2

78 88

Chimeric recipient

0 41 0 48

52 35

0 33 28

58 74 1 0 63 72 61 72

35 30

2 67 27 0 32 23

3 104 0 30 25

recipients. -BCLI cells (5 X lo") were inoculated i.v. into normal or chimeric

RAMIg, or MoAb aThy-1.2) and the appropriate secondary antibodies bSpleen cells (lo') were stained with the primary antibodies (Rald,

(FITC-GARIg or FITC-F(ab')n-RAMy) and analyzed on the FACS. Values obtained by staining with NRIg or NRtlg, and FITC-GARIg or FITC-F(ab')2- RAMy, respectively. were subtracted as background from the values reported here. sIg = surface lg.

'Rat MoAb aThy-1.2. an IgG2b. is detected with FITC-F(ab'),-RAMy.

TABLE 111 Quantijication of dormant tumor cells tn the spleens of chimeric

animals injected 4 mo earller with BCL. cells"

No. of Splenocytes from Chimeric

Mice Transferred into Normal

BALB/c Mkeb

1 06 105 104

1 o2

10'

Days after Transfer When BALB/c

Adoptive Recipients Were Sacrificed

39 42 42 84

120 84

84 98

120

Analysis of BALB/c Adoptlve Reciplents

No. of cells Percent of BCLl per spleen Id+ cells per

x 10- spleend

3.0 f 0.2 66 & 3 2.3 f 0.3 1.3 & 0.2

1 9 f 8 1 + 1

3.3 f 0.3 2.7 f 0.3

3 9 f 6 1 2 f 3

3.6 f 0.4 1.1 f 0.1

27 & 3 1 f 1

1.2 f 0.2 1.2 f 0.1 <1

1 + 0

graded numbers of spleen cells from these animals were adoptively trans- - Proven chimeric mice were given 5 X lo5 BCLl i.v. Four months later,

ferred to normal BALB/c mice. All inocula were initially made up to a concentration of 5 x lo6 total

cells/ml. Tenfold dilutions were performed in the presence of normal BALB/c spleen cells. and 0.2 ml (10' total cells) were injected i.v.

values shown are the averages obtained from two to four mice followed 'These data are a composite of results from two experiments. The

by the range. Cells (10') were stained with Rald or NRIg (negative control subtracted

as background) and FITC-GARlg and were analyzed on the FACS to determine the percentage of BCL Id+ cells.

inoculation of BCLl cells, confirming earlier results. The levels of surface Ig+ and Thy-l.2+ cells were within nor- mal range.

Presence and quantitation of dormant tumor cells in chimeric animals inoculated with BCL, tumor. Indirect immunofluorescence studies showed no detectable Id+ cells 12 wk or later after chimeric mice had received BCL, tumor cells. To determine if BCL, tumor cells were pres- ent in the spleens of such chimeric animals, we injected graded numbers of spleen cells from these mice into normal BALB/c animals. Earlier studies had shown that one viable BCL, cell is capable of causing tumor in BALB/ c animals by 12 wk after inoculation (2).

A s shown in Table 111, these mice harbored dormant BCLl tumor cells in their spleens. All seven recipients of lo2 splenocytes obtained from chimeric animals (that had been inoculated 84 to 120 days previously with BCL, cells) developed tumor, whereas none of the eight recipi-

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

TUMOR DORMANCY IN H-2 CHIMERIC MICE 1379

ents of 10’ splenocytes from similar donors developed BCL, tumor. Hence, we estimate that 1 to 10% of splen- ocytes in these chimeric mice are BCL, tumor cells. The growth rate and/or malignant potential of dormant BCLl cells appear to be similar to those of actively growing BCL, cells in a syngeneic recipient.

To determine the sensitivity of detection by immuno- fluorescence of Id+ cells in the spleen cell population, a mixing experiment was performed. Spleen cells from tu- mor-bearing mice containing 70% Id+ cells were mixed with normal splenocytes and analyzed for Id+ cells by FACS analysis. Approximately 5% of Id+ cells could be detected, but not 2% (data not shown). Taken together with the cell transfer studies, these findings suggest that 2 to 5% of the splenocytes of the tumor-dormant mice are BCL, cells.

The BCL, tumor is not caused by a leukemogenic virus. Prior studies of other investigators (20, 2 1) indicate that the Id of a B cell tumor is a clonotypic marker. Hence, our studies of BCLl transfer into syngeneic recipients were interpreted as caused by transfer of viable BCL, cells. To formally exclude the possibility that transfer of a leukemogenic virus was responsible for the develop- ment of B cell leukemia in recipients of spleen cells from chimeric mice, we performed the following experiment. Groups of five to 10 normal BALB/c mice were inoculated with 2 X lo7 and 1 X lo6 cell equivalents of cellfree extracts of spleen suspensions containing BCL, cells. Concomitantly, three groups-each consisting of 10 nor- mal BALB/c mice-received lo6, lo3, or 10 viable BCLl cells from spleen cell suspensions. After 34 wk of obser- vation, none of the mice that received cell extracts showed any signs of tumor. The mice from all the control groups, however, developed BCL, disease with character- istic splenomegaly and Id positivity as analyzed by indi- rect immunofluorescence on the FACS (data not shown). Because one BCL, cell has been shown to transfer tumor within 12 wk (2). these results indicate that the extracts were cell free and suggest that they did not contain a leukemogenic virus that caused BCLl disease.

A state of tumor dormancy does not develop in chi- meras established after tumor inoculation. The immu- nity to BCL, displayed by chimeric mice might be useful therapeutically if the approach could be applied to treat an established tumor. To investigate this possibility, BCL, cells were injected before chimerism was estab- lished. In all cases, tumors grew normally and eventually killed their hosts. Figure 2 illustrates the results of a representative experiment. BCL, cells were injected 17 days before lethal irradiation and allogeneic BM trans- plantation. The growth curve of BCL, in the spleen showed a decrease in numbers of tumor cells due to the x-irradiation. However, subsequently, the number of tu- mor cells increased rapidly and the BCLl growth curve became indistinguishable from that of the control by the end of 8 wk.

DISCUSSION

This report describes the killing of a large proportion of highly malignant B leukemic cells (BCL,) and the even- tual development of a state of dormancy of the remaining tumor cells in mice chimeric at the H-2 locus. Such chimeric mice were generated by lethal whole body x- irradiation followed by allogeneic BM transplantation (H-

9 I - 1

‘e’

I I I I I I

10 20 30 40 50 60

Days After Injection

BCLI tumor. BALB/c mice were injected with 5 X lo6 BCLl cells i.v. 17 Figure2 Effect of induction of chimerism in BALB/c mice bearing

days before receiving 900 rad of whole body x-irradiation. The arrow in the figure indicates the day of irradiation. Bone marrow cells from C57BL/ 6 donors were depleted of T cells and transferred 1 day after irradiation. ( 0 - 0 ) . Irradiation followed by transplantation of allogeneic bone marrow: [O-.-*-O), untreated controls.

2b + H-2d). The chimeric state was established without graft-vs-host disease by eliminating T cells from the do- nor bone marrow with the use of a monoclonal anti-Thy- 1.2 and complement. After chimerism had been estab- lished, the animals received a large dose of tumor cells (5 x lo5 to 5 x lo7 per animal] that grew rapidly. Seven weeks after an inoculum of 5 X lo7 BCL, cells. the tumor cells reached a level of approximately 1 O9 cells per mouse. The tumor burden was calculated by assaying increases in the white blood cell count, spleen size, and BCLl-Id+ cells in the spleen. The tumor burden then declined gradually over an additional period of 8 wk until it reached undetectable levels and the mice appeared nor- mal. However, because one tumor cell (BCLI) can transfer leukemia to a normal recipient, it was possible to estimate the minimum tumor load in the dormant animals by cell transfer experiments to virgin recipients. Unexpectedly, the transfer studies revealed a relatively large tumor load of lo6 to lo7 BCL, cells per spleen. Similarly treated animals have been observed for as long as 8 mo after inoculation of BCL, cells, and they appear healthy and tumor free by conventional criteria, including normal white blood cell count, spleen size, and lack of detectable Id+ cells. Thus, there is no evidence that these animals cannot live out a normal life span despite the presence of a relatively large tumor burden.

Perhaps the most impressive aspect of the antitumor response is its effectiveness. Thus, although the chimeric mice are initially permissive hosts for tumor growth, they eventually reduce the very large tumor burden to below the level at which progressive growth takes place. It is provocative that although the host response can elimi- nate a large tumor mass, tumor cells are not completely eliminated. Rather, a small proportion, but a large num- ber, of viable tumor cells remain in the animal in a dormant state. It is unclear why this very effective host antitumor response is unable to eliminate all the tumor cells. One possibility is that the dormant cells represent an outgrowth of a subset that escapes from the host response and is less malignant. However, cell transfer experiments indicate that the malignant potential of the dormant BCL, cells has not been significantly reduced. Thus, 100 spleen cells routinely transferred tumor to

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

1380 TUMOR DORMANCY IN H-2 CHIMERIC MICE

normal recipients, whereas 10 cells did not. Therefore, by this criterion, the spleen of the donor mouse contained at least lo6 BCL, cells, i.e., lo6 potentially lethal doses of dormant BCL, cells. There are two additional reasons for believing that the malignant potential of the dormant cells is not significantly decreased and that the determi- nation of tumor burden by cell transfer is not an under- estimate: 1) The kinetics of tumor growth in the recipient animals is similar to that in normal BALB/c mice. 2) A higher tumor burden should have been detectable by the other assays used. These observations argue that the host response can kill a large proportion of the tumor cells and can prevent progressive growth of the remaining cells despite their high degree of malignancy.

The mechanisms responsible for immunity to the tu- mor have not yet been characterized. Because of the effectiveness of the resistance, the experimental model employed (H-2 chimeras), and the latent period of several weeks between initiation of tumor growth and the begin- ning of tumor regression, we presume that the major resistance mechanism is an immune response. Earlier studies in a similar model indicate that tumor immunity can be transferred to syngeneic recipients with spleen cells (8). The most likely response, therefore, is an allo- geneic T cell-mediated immune response against class I major histocompatibility complex antigens of the BCL, cells. Such allogeneic responses are very strong ones and are capable of inducing tumor regression (22). If the response is directed to H-2 antigens, then the antitumor effect is a more sensitive assay than graft-vs-host dis- ease. The response could be directed, however, to other leukemia-associated antigens, e.g., the BCL, Id. An anti- Id response can provide immunity to myelomas (23, 24) and to lymphocytic leukemia (25), and can contribute to immunity to BCL, in IgH-allotypic congenic strains (26). I t is also possible that an allotypic response to a B cell differentiation antigen or a response to a minor histocom- patibility antigen (27) contributes to host resistance to the BCL, tumor. Indeed, we have no evidence to exclude a contribution from nonimmunologic mechanisms such as natural killer cells in the maintenance of the dormant state.

The major question raised by the above results is why the viable leukemic cells are dormant. Presumably, they are either in Go or are replicating at the same rate a s they are dying. In either event. it is likely that there is a dynamic equilibrium between the host defense mecha- nism(s) and the growth of the tumor cells. Changes in this equilibrium are critical to the issue of survival. It is well known from studies of clinical cancer that dormancy is a major feature of this disease. Thus, there are many examples in which primary neoplasms are removed and metastatic disease does not become overt until many years or decades later (28-34). The natural history of such tumors, together with the capacity of malignant tumors to destroy basement membranes and invade blood vessels (35-38). strongly suggests that dissemination of tumor cells occurs early in the development of a tumor. Indeed, studies of experimental tumors in which the en- trance of tumor cells into the circulation has been quan- tified indicate that lo5 to lo7 tumor cells are present in the venous effluent/day (36,38). The fate of such dissem- inated cells in human cancer determines the outcome of the disease. In some cases, metastatic disease appears

early in the course of the disease: in other instances, metastatic disease may not develop for many years. It is possible, therefore, that in virtually all "cured cancer patients, there are metastatic foci that remain dormant. These dormant cells may eventually die or may persist for the lifetime of the patient. Considering the critical nature of the dormant state, it is surprising that there is relatively little experimental evidence directed to eluci- dating this facet of malignancy.

Fisher and Fisher (39) and Sugarbaker et al. (40) in- jected a Walker 256 sarcoma into Sprague-Dawley rats and observed that liver metastasis did not occur unless the rats were subjected to repeated laparotomy and liver massage. Presumably microscopic foci in the liver re- mained quiescent until additional trauma was imposed. Stein-Werblowsky (41) made similar observations with pulmonary metastases from a prostatic tumor carried intramuscularly in rats. He observed that treatment of such animals with trypsin or antibody to a-2-macroglob- ulin resulted in the development of pulmonary metas- tases. He suggested that the pulmonary vessels of tumor- bearing animals are lined with a layer of a-2-macroglob- ulin that blocks the transmigration of tumor cells. If this barrier is removed by proteases or specific antiserum, metastases can then develop. Eccles and Alexander (42) demonstrated an increased incidence of distant metas- tases in rats whose primary sarcoma was removed if whole body x-irradiation or thoracic duct drainage was employed. Gimbrone et al. (43) showed that implantation of epithelioma cells into the aqueous humor of rabbits in which no vascularization takes place resulted in the growth of very small tumors that could not enlarge fur- ther. They suggested that accumulation of growth-in- hibiting metabolites in the center of the tumor mass caused necrosis, so that only the tumor cells on the outside of the mass were capable of growth. If the tumor invaded the iris, vascularization occurred and the tumor progressively enlarged. They emphasized, therefore, the importance of vascularization of solid tumors as a re- quirement for progressive tumor growth. They describe the static nature of nonvascularized tumors as a case of population dormancy. The most extensive series of stud- ies on tumor dormancy have been carried out by Whee- lock, Robinson, et al. (44-50). They have studied the fate of live L5 178Y lymphoma cells injected into DBA/2 mice that had previously received viable tumor cells but had the tumor nodule excised 10 days later. Such animals displayed relative immunity to the second challenge of tumor cells. Rather than succumbing in approximately 1 mo, some mice lived for over 1 yr, and many harbored tumor cells, as determined by outgrowth of tumor in vitro or by cell transfer. In such animals, re-emergence of tumor appeared at a time when numbers of cytolytic T lymphocytes (CTL) had declined. By labeling tumor cells in vitro with '251-IUDR and inoculating them into the peritoneal cavity of immune mice, they found that cell lysis occurred very rapidly, was immunologically specific, and required host-tumor cell contact. They postulated that CTL and cytolytic macrophages, both individually and in synergy, maintain the dormant tumor state. Ad- ditionally, before termination of tumor dormancy, mac- rophages appear in the peritoneal cavity that can sup- press the generation of CTL in in vitro assays.

I t should be stressed that the present model of the

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

TUMOR DORMANCY IN H-2 CHIMERIC MICE 1381

dormant tumor state is not a physiologic one, because the recipient is an H-2 chimera. T cell-mediated alloreac- tivity is a much stronger antitumor mechanism than T cell-mediated syngeneic reactivity. The latter, however, can play a significant role in tumor immunity (22), in- cluding immunity to BCLl (26, 27. 51). Hence, the chi- meric model may represent an exaggeration of the phys- iologic state of tumor dormancy. In any case, the present model should facilitate the characterization of the under- lying cellular and humoral mechanisms in this tumor- host “standoff“ as a step to understanding cancer dor- mancy.

Acknowledgments. We thank Ms. L. Trahan, Ms. F. LaMontagne, Ms. R. Nisi , Ms. S. Gorman, Mr. Y. Chinn, Ms. A. Buser, and Mr. S. Korchak for technical assistance and Ms. C. Baselski for secretarial assistance.

REFERENCES

1, Slavin. S., and S. Strober. 1978. Spontaneous murine B cell leuke-

2. Krolick, K. A., P. C. Isakson, J. W. Uhr, and E. S. Vitetta. 1979. mia. Nature 272:624.

BCL,. a murine model for chronic lymphocytic leukemia: use of the surface immunoglobulin idiotype for the detection and treatment of

3. Vitetta. E. S., D. Yuan, K. Krolick, P. Isakson. M. Knapp. S. Slavin, tumor. Immunol. Reu. 48:81.

and S. Strober. 1979. Characterization of the spontaneous murine B cell leukemia (BCL,). Ill. Evidence for monoclonality using an anti-

4. Muirhead. M. J., J. M. Holberts, Jr., J. W. Uhr, and E. S. Vitetta. idiotype antibody. J. Immunol. 122: 1649.

1981. BCL,. a murine model of prolymphocytic leukemia. 11. Mor- phology and ultrastructure. Am. J. Pathol. 105306.

5. Krolick, K. A., J. W. Uhr, and E. S. Vitetta. 1982. Selective killing of leukaemia cells by antibody-toxin conjugates: implications for

6. Vitetta. E. S.. K. A. Krolick, and J. W. Uhr. 1982. Neoplastic B cells autologous bone marrow transplantation. Nature 295604.

a s targets for antibody-ricin A chain immunotoxins. Immunol. Rev. 62: 159.

7. Krolick, K. A.. J. W. Uhr. S. Slavin. and E. S. Vitetta. 1982. In vivo therapy of a murine B cell tumor (BCL,) using antibody-ricin A chain immunotoxins. J. Exp. Med. 155: I 797.

8. Weiss. L., S. Morecki, E. S. Vitetta. and S. Slavin. 1983. Suppression and elimination of BCL, leukemia by allogeneic bone marrow trans- plantation. J. Immunot. 130~2452.

9. Slavin, S., L. Weiss. S. Morecki. and M. Weigensberg. 1981. Eradi- cation of murine leukemia with histoincompatible marrow grafts-in mice conditioned with total lymphoid irradiation (TLI). Cancer Im- munol. Immunother. 11:155.

10. Vitetta, E., and K. Krolick. 1980. Allelic exclusion of 1gD allotypes on murine B cells. J. Immunol. 124:2988.

11. Ligler. F. S.. E. S. Vitetta, and J. W. Uhr. 1977. Cell surface immu- noglobulin XXII. Reappearance of surface IgM and IgD on murine

12. Ozato, K., and D. M. Sachs. 1981. Monoclonal antibodies to mouse splenocytes after removal by capping. J. Immunot. 119:1545.

MHC antigens. 111. Hybridoma antibodies reacting to antigens of the H-Zb haplotype reveal genetic control of isotype expression. J. Im-

13. Sachs. D. M., N. Mayer. and K. Ozato. 198 1. Hybridoma antibodies munol. I26:317.

directed toward murine H-2 and la antlgens. In Monoclonal Anti- bodies and T-cell Hybridomas. G. J. Hammerling. U. Hammerling, and J. F. Kearney. eds. Elsevier/North Holland Biomedical Press, Amsterdam. P. 95.

14. Ozato. K.. N. Mayer, and D. M. Sachs. 1980. Hybridoma cell lines

Immunol. 124~533. secreting monoclonal antibodles to mouse H-2 and la antigens. J.

15. Yuan. D., E. S. Vitetta, and J. R. Kettman. 1977. Cell surface immunoglobulin. XX. Antibody responsiveness of subpopulations of B lymphocytes bearing different isotypes. J. Exp. Med. 145; 1421.

16. Zan-Bar. I., S. Strober. and E. S. Vitetta. 1977. The relationship between surface immunoglobulin isotype and immune function of murine B lymphocytes. 1. Surface immunoglobulin isotypes on

17. Muirhead, M. J.. P. C. Isakson, K. A. Krolick, J. W. Uhr, and E. S. primed B cells in the spleen. J. Exp. Med. 145: 1 188.

Vitetta. 1981. W L I . a murine model of prolymphocytic leukemia. I.

J. Pathol. 105295. Effect of splenectomy on growth kinetics and organ distribution. Am.

18. Ledbetter. J. A., and L. A. Herzenberg. 1979. Xenogeneic mono- clonal antibodies to mouse differentiation antigens. Immunol. Rev. 4 7: 63.

19. Krolick, K. A., P. C. Isakson. J. W. Uhr, and E. S. Vitetta. 1979. Murine B cell leukemia (BCLI): organ distribution and kinetics of growth as determined by fluorescence analysis with an anti-idiotypic antibody. J. Immunol. 123: 1928.

20. Fu, S. M., R. J. Winchester, andH. G. Kunkel. 1975. Similar ldiotypic

J. Immunol. 114:250. specificity for the membrane IgD and IgM of human B lymphocytes.

21. Thielemans, K., D. G. Maloney, T. Meeker. J. Fujimoto. C. Doss, R. A. Warnke, J. Bindl, J. Gralow, R. A. Miller, and R. Levy. 1984. Strategies for production of monoclonal anti-idiotype antibodies

22. Smith. R. T., and M. Landy (eds.). 1970. Immune Surveillance. against human B cell lymphomas. J. Immunol. 133~495.

23. Lynch, R. G., R. Graff. S. Sirisinha. E. S. Simms, and H. N. Eisen. Academic Press. New York.

Proc. Natl. Acad. Sci. USA 69: 1540. 1972. Myeloma proteins as tumor-specific transplantation antigens.

24. Freedman, P.. J. R. Autry, S. Tokida. and R. C. Williams. 1976. Tumor immunity induced by pre-immunization with BALB/c mouse

25. Stevenson, F. K.. and J. Gordon. 1983. Immunization with idiotypic myeloma protein. JNCI 45~735.

immunoglobulin protects against development of B lymphocytic leu- kemia, but emerging tumor cells can evade antibody attack by mod-

26. Ciavarra, R. P., E. S. Vitetta. and J. Forman. 1986. Growth inhibi- ulation. J. Immunol. 130:970.

tion of a B cell leukemia: Evidence implicating an anti-idiotype immune response for protective tumor immunity. J. Immunol. 137:1371.

27. Forman. J.. R. Riblet. K. Brooks. E. S. Vitetta. and L. A. Henderson. 1984. H-40. an antigen controlled by an Igh linked gene and recog-

distribution of its product on B cell tumors. J. Exp. Med. 159:1724. nized by cytotoxic T lymphocytes. I. Genetic analysis of H-40 and

28. Callender. G. R.. H. C. Wilder, and J. E. Ash. 1942. Five hundred malignant melanomas of the choroid and ciliary body followed five years or longer. Am. J. Ophthalmol. 25:962.

29. Adair, F.. J. Berg, L. Joubert. and G. B. Robbius. 1974. Long-term follow-up of breast cancer patients: the 30 year report. Cancer 33: 1 145.

30. Einhorn. L. H., M. A. Burgess, and J. A. Gottlieb. 1974. Metastatic

31. Holland, J. F., and E. Frei. 1973. Cancer Medicfne. Lea and Febiger. patterns of chroidal melanoma. Cancer 34: 1001.

Philadelphia. P. 1841. 32. Jensen. 0. A. 1963. Malignant melanomas of the uvea in Denmark.

Acta Ophthalmol. ISuppl.l IC0penh.J 75132. 33. Hadfield, G. 1954. The dormant cancer cell. Br. Med. J. 2507. 34. Willis, R. A. 1952. The Spread of Tumors In the Human Body. 2nd

ed. Butterworth, London. 35. Liotta, L. A.. Kleinerman, J.. and Saidel, G. M. 1974. Quantitative

relationships of intravascular tumor cells, tumor vessels, and pul- monary metastases following tumor implantation. Cancer Res. 34:997.

36. Liotta, L. A., Kleinerman, J., and Saidel. G. M. 1976. The signifi- cance of hematogenous tumor cell clumps in the metastatic process.

37. Cole, W. H.. S. Roberts. A. Watne. G. McDonald, and E. McGrew. Cancer Res. 36:889.

1958. The dissemination of cancer cells. Bull. N. Y. Acad. Med. 34: 163.

38. Roberts. S., A. Watne. R. McGrath, E. McGrew, and W. H. Cole. 1958. Technique and results of isolation of cancer cells from the circulating blood. Arch. Surg. 76:334.

39. Fisher, B., and E. R. Fisher. 1959. Experimental evidence in support

40. Sugarbaker, E. V.. A. S. Ketcham, and A. M. Cohen. 1971. Studies of the dormant tumor cell. Scfence 130:918.

41. Stein-Werblowsky, R. 1978. On the latency of tumour cells. Br. J. of dormant tumor cells. Cancer 28:545.

Exp. Pathol. 59:386. 42. Eccles, S. A.. and P. Alexander. 1975. Immunologically-mediated

43. Gimbrone, M. A.. Jr., S. B. Leapman, R. S. Cotran. and J. Folkman. restraint of latent tumour metastases. Nature 257:52.

1972. Tumor dormancy in utuo by prevention of neovascularization. J. Exp. Med. 135261.

44. Weinhold, K. J.. L. T. Goldstein, and E. F. Wheelock. 1977. Tumour- dormant states established with L5 178Y lymphoma cells in immu-

45. Weinhold, K. J., L. T. Goldstein, and E. F. Wheelock. 1979. The nised syngeneic murine hosts. Nature 270:59.

tumor-dormant state: quantitation of L5178Y cells and host immune responses during the establishment and course of dormancy in syn-

46. Robinson. M. K., and E. F. Wheelock. 1981. Identification of mac- geneic DBA/2 mice. J. Exp. Med. 149:732.

rophage-mediated cytolytic activity a s a tumor suppressive mecha- nism during maintenance of the L5178Y-tumor dormant state in

47. Wheelock, E. F., M. K. Robinson, and G. A. Truitt. 1982. Establish- DBA/2 mice. J. Immunol. I26:673.

ment and control of the L5178Y-cell tumor dormant state in DBA/2 mice. Cancer Metastasfs Reu. 1:29.

48. Robinson. M. K.. and E. F. Wheelock. 1982. Synergistic cytolytic activity by combined populations of peritoneal T-lymphocytes and

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

1382 TUMOR DORMANCY IN H-2 CHIMERIC MICE

macrophages during the L5178Y cell tumor-dormant state in DBA/2 1983. Enhanced suppressor macrophage activity associated with mice. Cell. Imrnunol. 73:230. termination of the L5178Y cell tumor-dormant state in DBA/2 mice.

cells in a tumor-dormant state. cancer ~ ~ ~ ~ ~ ~ l . Immunother. 51. Forman. J., R. Ciavarra. and E. S. Vitetta. 1981. Cytotoxic T cells 1983. Cytotoxic T lymphocytes in DBA/2 mice harboring L5178Y CancerRes' 43:5831.

16:59. specific for antigens expressed on surface immunoglobulln positive

49. Marsili, M. A., M. K. Robinson, G. A. Truitt, and E. F. Wheelock.

50. Robinson, M. K., G. A. Truitt. T. Okayasu, and E. F. Wheelock. cells. J. Exp. Med. 154: 1357.

by guest on August 25, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

View publication statsView publication stats