(CANCER RESEARCH 58. 5439-5446, December 1. 1998]

EblacZ Tumor Dormancy in Bone Marrow and Lymph Nodes: Active Control ofProliferating Tumor Cells by CD8+ Immune T Cells1

Markus Müller,Fotini Gounari, Sophia Prifti, Hans JörgHacker, Volker Schirrmacher, and Khashayarsha Khazaie2

Tumor Immunology Program /M. M., F. G., S. P., V. 5., K. KJ and Department of Pathology ¡H.J. H.¡,German Cancer Research Center (Deutsches Krebsforschungzentrum),Heidelberg 69120. Germany

ABSTRACT

A well-defined /«<•/gene tagged DBA/2 lymphoma (EblacZ) was used to

examine the role of host immune responses in controlling tumor dissemination and persistence, as well as metastasis. In s.c. and intra-ear pinna-

inoculated mice, low numbers of EblacZ cells homed to the bone marrowand lymph nodes. The frequency of bone marrow-residing tumor cells did

not change with the growth of primary tumor or with multiple inoculations of tumor cells. The bone marrow-residing tumor cells expressed theproliferation-associated KÕ67antigen and expanded upon CDS ' deple

tion. In contrast, inoculation of nu/nu or severe combined immunodeficiency mice or of immune-suppressed DBA/2 mice led to the rapid out

growth of EblacZ cells in the bone marrow and their metastasis to otherorgans.

Transfer of bone marrow from EblacZ immunized Ml K congenie orsyngeneic DBA/2 donors, but not from naive donors, protected s.c.-inoc-

ulated DBA/2 mice. Protection was abrogated by in vitro depletion ofCD8+ T cells prior to transfer of bone marrow. These experiments show

that bone marrow and lymph nodes are privileged sites where potentiallylethal tumor cells are controlled in a dormant state by the immune system.Metastasis may be a consequence of the breakdown of this immunecontrol.

INTRODUCTION

A well-characterized animal model for tumor metastasis is represented by the methylcholanthrene-induced lymphoma LSI78Y (Eb),induced originally in a female DBA-2 (H-2d) mouse. The Eb tumor

line expresses characteristic MHC I-restricted CTL epitopes and is

nonmetastatic in syngeneic hosts; however, several metastatic variants, such as ESb and ESb-MP, have been further established from

this original tumor line (1,2). The transition from the nonmetastatic tothe metastatic phenotype required successive i.p. transplantations andin situ selection (3, 4). The Eb and derived cell lines express characteristic class I major histocompatibility Kd-associated tumor antigens

and elicit specific CTL responses in immunized syngeneic hosts (5).L5178 (Eb) cells were also among the first tumor cell lines used in

tumor dormancy studies. To establish dormancy, DBA/2 mice werefirst inoculated s.c. with a lethal dose of Eb cells, and the resultingtumor mass was then surgically excised. Such mice were challengedi.p. with Eb cells at a dose that produced ascitic tumors and death in100% of naive mice. The vaccinated mice were initially protected buteventually succumbed to rapid tumor cell outgrowth and ascites afterseveral weeks, some as late as 340 days after the challenge (6-8).

This delay in tumor outbreak was shown to be due to cytotoxic andcytostatic effects mediated by interleukin 2, IFN-y and tumor necrosisfactor-a secreted by lymphocytes and macrophages (9, 10).

We have shown that immunization through s.c. inoculation ofDBA/2 mice with Eb tumor cells genetically modified to express thebacterial lacT. gene (EblacZ) and rendered nontumorigenic, either

Received 3/26/98: accepted 9/25/98.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this faci.

1Supported by the Mildred Scheel Stiftung and the Deutsche Forschung Gemeinschaft.2 To whom requests for reprints should be addressed, al U. 373 INSERM, Facultéde

MédecineNecker-Enfanls Malades. 156 rue de Vaugirard. 75730 Paris cedex 15. France.Phone: 33 1 4061 5590: Fax: 33 1 4061 5590; E-mail: khazaie@necker.fr.

through irradiation or genetic modification, resulted in the persistenceof these cells in the host bone marrow. This long-term persistence

correlated with the duration of antitumor immunity (11). Similarly,potentially lethal live Eb-lacZ cells did not grow as a tumor wheninjected iep3 but rather disseminated to the bone marrow and estab

lished a state of tumor dormancy. Because EblacZ cells were lethalwhen injected s.c., the persistence of these cells after iep inoculationin part satisfied the definition for tumor dormancy (12) as a state inwhich potentially lethal tumor cells persist for a prolonged period oftime in a clinically normal host with little or no increase in the tumorcell population.

Here, we have exploited our earlier observations on the dissemination and persistence of EblacZ cells to the bone marrow of vaccinatedmice to define the factors responsible for the dormancy of tumorigenicEblacZ cells. We report that the rapid dissemination and persistence oflow numbers of Eb-lacZ tumor cells in the bone marrow is found inmice with or without lethal primary Eb-lacZ tumors. Our observationssuggest that in tumor-resistant hosts, the bone marrow and lymph

nodes can be privileged sites for the persistence of tumor cells in anapparent dormant state. The maintenance of low numbers of Eb cellsin these organs was not caused by the tumor cells resting in the G()state of the cell cycle but by active T-cell immunity, resulting in low

numbers of proliferating tumor cells. Transplantation of bone marrowfrom Eb dormant mice to naive mice yielded both tumor cells andprotective immune effector cells. The bone marrow thus appears to bea privileged site where protective immune T cells keep tumor cells ina dormant state.

MATERIALS AND METHODS

Recombinant Retroviruses, Transduction, Characterization, and Detection of Modified Tumor Cells. Tumor cells were genetically modified byretroviral gene transduction using the M48 retroviral vectors (13) encoding asingle bacterial lacZ gene fused to the nuclear localization signal of the SV40large T, driven by the phosphoglucokinase gene promoter. The vector waspackaged using the ecotropic CRE packaging cell line (14). Several independent modified clones were isolated and tested. The most aggressive of theseclones was used for additional experiments. To visualize tumor cells, wholeorgans or dispersed cells were fixed in 2% formaldehyde/0.2% glutaraldehydeand stained overnight with 5-bromo-4-chloro-3-indolyl ß-i>-galactosidase.Dispersed cells from bone marrow, lymph node, or spleen were cytospun (IO6

cells) onto poly-L-lysine-treated glass slides before staining for ß-galactosidase(11). This method allowed confident scoring of single tumor cells among IO6

nontumor cells, as determined by mixing of/«rZ-positive and -negative cells.The Ki67 (anti-mouse rabbit polyclonal) was a kind gift from Prof. J. Gerdes

(Forschungsinstitut Borstel, Kiel, Germany). For double staining, cells werefixed with 3.7% formaldehyde and stained with KÕ67rabbit anti-mouse serum( 1:100 dilution) and secondary Cy3 conjugated a-rabbit, followed with X-gal

staining overnight and incubation with DAPI.Vaccination and Irradiation. Mice were between 8 and 12 weeks of age

at the beginning of the experiments. Tumor cells (2 x IO6 live EblacZ cells)

were washed with PBS and resuspended in 50 ¿dof PBS for injection iep or

1The abbreviations used are: iep. intra-ear pinna; X-gal. 5-bromo-4-chloro-3-indolyl-ß-D-galaclopyranoside; DAPI. 4'.6-diamidino-2-phenylindole: mAb. monoclonal anti

body: SCID. severe combined immunodeficiency; BMT. bone marrow transplantation.

5439

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TUMOR DORMANCY IN BONE MARROW

1(K) ¡j.\s.c. Mice were irradiated at 5 Gy (500 rads) with a "7Cs source

(Gannitaceli I(XX):Atomic Energy, Ottawa. Canada).Immunostaining. Staining of bone marrow cells was done with biotin/

streptavidin directly coupled «-CD8-FITC and a-CD4-PE (Life Technologies,Inc.), and pan-0-TCR-Red670 (15).

T-Cell Depletion m Vivo and in Vitro. For in vivo depletion of CDS*

cells, the monoclonal antibody YTS-169.4.2 (purified IgG; Ref. 16) was

injected once or several times (as indicated) i.p. in a pretested concentration of0.5 mg in 2(X) p.\ of PBS. Three-color fluorescence-activated cell sorteranalysis ( IO5events) with «-CD8.and a-CD4 and pan-ß-TCRrevealed 0.26%

mature bone marrow T-cells before depletion and 0.01% after depletion. Invitro depletion was performed using or-CD8 IgM mAb and treatment with

complement and DNase.Bone Marrow Transplantation. In bone marrow transplantation experi

ments, donor mice were injected iep or s.c. with PBS or 5 X IO5 to 2 X IO6EblacZ cells in a total volume of 50 fil. At indicated time points. 2 x IO7 bone

marrow cells were prepared and injected i.v. into recipient mice. One day later,the recipient mice were challenged s.c. with 5 X 10s or 2 X IO6 EblacZ cells,

and their survival was followed.Quantitation of CTL Activity. Quantitations of in vitro and in situ acti

vated CTLs were performed as described earlier (5. 11). Briefly, for in vitrorestimulation. 9 days after inoculation with EblacZ cells (2 X IO6), spleen cells(2 X IO6) were removed and restimulated in vitro for 5 days with EblacZ cells(2 X IO5 cells, 1(X)Gy irradiated). For in situ restimulation, after 7 days themice were restimulated i.p. with EblacZ cells (2 X IO6 cells. 100 Gy irradiated), and 3 days later, peritoneal exúdate cells (PEC) were isolated. "Cr

release assays were performed after 4 h of incubation of the CTLs withMCr-labeled target.

Klispot. ELISPOT assays were performed by culturing 2 X IO7 immunespleen cells (9 days after bone marrow transfer) with 2 x IO6 irradiated EblacZ

cells for 5 days. The cells were then plated in 96-well cellulose-based micro-tiler plates (MAHA N45; Millipore) coated with rat a-mouse INF-y mAb; after20 h. the wells were washed with PBS/Tween/l'/r BSA and incubated with

biotin-conjugated rat a-mouse mAb for 2 h. After additional washes, treatmentwith avidin peroxidase (Pharmigen; 1:I(XX)) for 3-10 min relieved positive

spots, which were scored visually using a light microscope.ELISA. For ELISA, 0-galactosidase-coated (Sigma Chemical Co.; 2.5

fig/ml in 0.05 M NaCO,/0.05 M NaHCO,, pH 9.6) Falcon-ELISA plates were

washed three times with PBS/Tween. blocked with 0.2% gelatin/PBS ( 1 h.37°C). incubated with test sera for 1 h at room temperature, washed, and

incubated with goat «-mouse IgG-POX Ab-9()D (Dianova, Hamburg, Ger

many: dilution, 1:5000). After wash with PBS/Tween. OPD/H,O, was addedfor 5-30 s in the dark, the reaction was stopped with IN H2SO4, and meas

urements were performed at 492 nm.

RESULTS

Vaccination and Tumor Development. When 2 x IO6 EblacZ

cells were s.c. inoculated into syngeneic DBA/2 mice, growth ofprimary tumors and death was observed within 1 month of injection(Fig. 1A, a and b). When the same number of live tumor cells wasinjected iep at 14 or 56 days prior to the s.c. inoculation, the mice wereprotected against Eb tumor cell challenge (11). In agreement withearlier results, iep injection of the mice did not result in any tumors,and mice vaccinated by this protocol exhibited long-term protection

against further challenge with tumor cells (Fig. \A, a and b).When s.c.-injected mice received 2 weeks after the first inoculation

an equal number of EblacZ cells in the opposite flank, they did notexhibit any significant change in survival (Fig. Iß«),but at the secondsite, only regressive small tumors were found (Fig. IBb). This observation indicated existence of concomitant immunity in the EblacZ-

¡noculatedmice.Active and Specific Host T- and B-Cell Responses Correlate

with Immunity against EblacZ Cells. CTL assays were performedto analyze the status of tumor-specific T-cell responses in tumor-bearing mice, iep-vaccinated or s.c.-inoculated tumor-bearing mice

were used as source of immune spleen cells, which were restimulated

100li;

75-i50-S?

2S

0laœQe

^r"!¿o

o—oo.i

75\

50*

25

n .UTT;

...To1

01234567 01234567

Wecks after inoculation

B 40E£ 30

= 20

E.2 10

456 0 I

Wecks after ¡noeulation

Fig. 1. Induction of concomitant immunity after s.c. tumor inoculation and of protective systemic immunity after iep tumor inoculation. In A. prophylactic ear pinna vaccination leads to protective systemic immunity. Groups of at least H) DBA/2 mice werevaccinated first before they were challenged at day 0 with 2 X IO6 EblacZ tumor cells.Vaccination was performed with 2 x IO* live EhlacZ cells inoculated into the ear pinna,

which is a subtumorigenic site. In experiment «.vaccination was performed at day 14 (A),and in experiment h, at day 56 (A). O. survival in nonvaccinated control groups. B, s.c.tumor growth and concomitant immunity. Groups of 10 DBA/2 mice were inoculated inone flank s.c. with 2 X IO* EblacZ cells or first in the leti tlank and after 2 weeks in theright flank with 2 X IO6 EblacZ cells, a, survival of the immunocompetent animalschallenged only once (O) or twice ( H >w'as followed and is plotted as a function of weeksafter challenge, h, average diameters of primary1 tumors in groups of 10 DBA/2 miceinoculated s.c. with 2 X 106 EblacZ cells. Tumor growth in the right flank ( y ) or of the

same mice injected again after an interval of 2 weeks in the left flank (B), or of miceinjected only once (O), injected once but 5 Gy preirradialed 1 day before inoculation ( 0 )are shown.

with EblacZ in vitro. The activated spleen cells were used as effectorcells in CTL assays against EblacZ cells. Alternatively, some inoculated mice were restimulated i.p. with irradiated tumor cells to examine the ability of the mice to elicit a specific T-cell response in vivo.

Peritoneal exúdate cells were used as source of effector T cells.Significant tumor-specific CTL activities were detected both in thevaccinated and the tumor-bearing DBA/2 mice, using either method(Fig. 2A). Anti-ß-galactosidase ELISA tests of sera from EblacZ-

transplanted mice revealed comparable humoral immune responses inboth instances (Fig. 2B).

Fig. 2C shows the production of IFN-y by spleen cells fromindividual, immune bone marrow-reconstituted DBA/2 mice upon in

vitro restimulation with either the same (EblacZ) or different (ESb)tumor-associated antigen(s). In these experiments, donor B10.D2mice were inoculated i.v. with EblacZ cells. After 2 weeks, 2 X IO7

bone marrow cells from these mice were transplanted to naive recipient DBA/2 mice. Nine days later, spleen cells were obtained from therecipient mice, stimulated in vitro with EblacZ or with the closelyrelated ESb cells, and analyzed for production of INF-y using anELISPOT assay. An antigen-dependent increase in the number ofcytokine-releasing cells was observed in all recipient mice but not in

naive mice. Thus, specific cellular (Fig. 2, A and C) and humoral (Fig.2B) immune responses against EblacZ cells were detected in tumor-bearing mice as well as immune cell-reconstituted mice.

T-Cell Immunity Is Required for the Control of EblacZ TumorMetastasis. Staining for lacZ readily allowed detection of singletumor cells in organs, frozen tissue sections, or in cytospins of

5440

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TUMOR DORMANCY IN BONE MARROW

CytotoricHyB

ELISA

100

-

O

80 -

60-

40-

20.

OH

^ 19 V >3

target ratio serum dilution

ELISPOT

100

er

50

.fi

9C

V\ci

</} O "}IH 1-5 o

number of spleen cellsxlO5

Fig. 2. Active and specific host T- and B-cell responses correlate with concomitantimmunity against EblacZ cells. A, EblacZ-specific CTL activities measured in groups oftour DBA/2 mice injected with 2 X IO6 EblacZ cells iep (i.e.; A, A) or s.c. (O, •)after

in vitro restimulation of immune spleen cells (A. O. D) or in siln restimulalion of day 7peritoneal CTLs (A. •.* ). ESb large! cells were used as control after I'Mvitro (D) or in

situ ( * ) restimulation with EblacZ cells. B, ELISA for antibodies against 0-galactosidase

in sera of mice injected 14 days earlier with EblacZ cells iep (A), s.c. (O). or left untreated(D). O.D., ahsorbance. C. ELISPOT assay for production of INF- y was performed usingspleen cells from three individual, bone marrow-grafted DBA/2 mice. The DBA/2 micehad been grafted 9 days earlier each with 2 X IO7 bone marrow cells derived from fiveB10.D2 donor mice, which in turn had been inoculated 14 days earlier with 2 X IO6

EblacZ cells i.V.. Spleen cells were restimulated in vitro with EblacZ (D) or ESb (A) for5 days before being assayed as described in "Materials and Methods."

1 X 10'' nonstained cells. Using whole organ staining, it was established that iep inoculation of 2 x IO6 EblacZ cells into syngeneic

DBA/2 mice did not result in any detectable tumor growth or métastases in peripheral organs (Fig. 3/4«).In contrast, in immune-compro

mised mice, metastatic nodules were clearly visible in liver sections(Fig. 3, Ah and Ba), and the lungs showed a scattered distribution ofX-gal-positive cells (Fig. 3ßr).Only nonspecific staining, due toendogenous ß-galactosidase activity, was observed in the kidneys

(Fig. 3, A and Be). This pattern of metastasis was observed in nu/nu.SCID, sublethally irradiated or CD8+ T-cell-depleted DBA/2 mice,

whereas no tumor cells were detected in similar organ sections fromtumor cell-inoculated, immunocompetent DBA/2 mice (Fig. 35, b, d,

and /). Thus, the most critical factor limiting metastasis of EblacZcells was a T-cell-dependent immune response.

In immune-compromised animals, liver and the lymphatic organs

were targets of métastases.Thus, nu/nu mice exhibited massivespleen, lymph node, and bone marrow involvement (Fig. 4A, a. b, andc). In late-stage disease, the bone marrow and spleen were heavily

infiltrated by tumor cells (Fig. 4/4, a and c), and the draining lymphnodes were overgrown by tumors (Fig. 4/46). A similar metastatic

5441

spread was seen in SCID mice and in sublethally irradiated DBA/2mice (data not shown). These observations underline the importanceof T-cell immunocompetence for the control of EblacZ tumor

metastasis.Active Immune Responses of Tumor-bearing Mice Prevent Mé

tastases but Fail to Control the Dissemination and Persistence ofTumor Cells in Privileged Sites. In DBA/2 mice the iep-inoculated

EblacZ cells disseminated rapidly into the lymphatic system, appearing in the draining lymph nodes within 30 min and in nondraininglymph nodes, bone marrow, and spleen by 24 h. Over several monthsof observation, only residual numbers (1-100/10'1 normal cells) of

tumor cells were detected in the bone marrow and lymph nodes (Fig.4/4, a and b), whereas the tumor cells were rapidly cleared from thespleen (Fig. 4/4r). The bone marrow of tumor-bearing DBA/2 miceresembled closely that of iep-vaccinated mice. Thus, in contrast toimmune-compromised, tumor-bearing mice, immune-competent mice

presented only small numbers of EblacZ cells in the bone marrow thatdid not increase with time (Fig. 4ß).A second inoculation with tumorcells did not significantly affect the number of EblacZ cells in thebone marrow (Fig. 40). These observations show a close correlationof immunity and control of EblacZ tumor metastasis. In this particularanimal model, the host immune response effectively cleared all examined tissues from tumor cells, with the exception of two privilegedsites: the bone marrow and the lymph nodes.

A Significant Fraction of the Residual Tumor Cells Detected inthe Marrow Express an S-Phase Proliferation Marker. To analyzethe tumorigenic potential of bone marrow-residing EblacZ cells, bonemarrow from iep-inoculated DBA/2 mice was placed in growth me

dium with 5% PCS, conditions under which only tumor cells wereexpected to survive. Staining for lacZ indicated that the expandednonadherent cells indeed represented EblacZ cells, s.c. inoculation ofvarying numbers of these cells into DBA/2 mice clearly showed thatthe bone marrow-derived EblacZ cells had maintained their tumori

genic potential (Fig. 5/4).To investigate the proliferation potential of the tumor cells in the

bone marrow of DBA/2 mice directly ex vivo, cytospin preparationsfrom bone marrow of mice inoculated with EblacZ were triple stainedfor lacZ (to visualize tumor cells; Fig. 50«),DAPI (to visualize DNA;Fig. 5Bb), and KÌ67(to identify potentially dividing cells; Fig. 5Bc).In contrast to the broad staining for lacZ that marked the shape ofintact nuclei in cytospins and histológica! sections (see Fig. 3), nuclearlacZ appeared as a collapsed aggregate when cells were fixed forstaining with the Ki67 antibody (Fig. 5Ba). In the iep-inoculated.disease-free, immune-competent mice, 21.3 ±4.0% of the bonemarrow-derived tumor cells were positive for the S-phase-specificKÕ67proliferation marker (Fig. 5Bc). The fraction of Ki67-positivetumor cells in the diseased bone marrow of immune-compromisedDBA/2 mice was 40 ±13.2%, whereas tissue culture-derived EblacZcells contained 87 ±3.5% Ki67-positive cells. Thus, a fraction ofbone marrow-residing EblacZ cells were proliferating in both s.c-inoculated. tumor-bearing mice and iep-inoculated. tumor-free mice,

but the absolute number of cells did not increase, indicating that thecell number was controlled by the on-going immune response.

CDS Immune T Cells Control Tumor Cell Numbers and Transfer Protective Immunity. To address the potential role of CD8^

immune T-cells in the control of tumor dissemination and metastasis,

DBA/2 mice were either injected with depleting monoclonal CDSantibodies or irradiated with 5 Gy to prevent primary antitumorresponses. In both instances, iep inoculation of immune-suppressed

mice with EblacZ cells led to rapid dissemination and continuousgrowth of the tumor cells in the host bone marrow, resulting in thedeaths of the animals (Fig. 6, a and h). In contrast, similarly inoculated, immune-competent DBA/2 mice did not develop any métastases

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TUMOR DORMANCY IN BONE MARROW

A a

f

Fig. -VT-cell immunocompctence is a critical requirement tor the eontrol of bblacZ tumor metastasis. X-gal and periodic acid-Schiff staining of sections of DBA/2 and mi/wi mice,inoculateti with HhlacZ cells. .\. ^MIIL-MCII:DBA/2 (u) and nii/nu (hi mice were injected with 2 X IO6 EblacZ cells iep and sacrificed 24 days later. Organs were removed, fixed in2'i formaldehyde and 0.2'i glulardialdehyde. X-gal stained for 24 h. and cut on a cryostat. A, intact organs (lop, liver: middle, lungs; hoiiom, kidneys) of immune DBA/2 or mi/numice, stained for Im-Z. B, histológica! analysis of the same organs, a. liver of nu/nu mouse with fixai metastasis; h, unstained liver of DBA/2 mice; i, lung of nu/nu mouse with sparse

distributed, single tumor cells; il. unstained lung/heart of DBA/2 mouse; e, unspecific (not cellular) staining of nu/nu mouse kidney;/ unspecific (not cellular) staining of DBA/2 mousekidney.

hut rather established tumor dormancy in the bone marow and lymph quency of bone marrow-residing dormant tumor cells, DBA/2 mice

nodes as described before. were vaccinated iep with EblacZ tumor cells. After 3 weeks, boneTo investigate the influence of immune suppression on the fre- marrow of three mice was checked for establishment of tumor dor-

5442

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TUMOR DORMANCY IN BONE MARROW

Fig. 4. Active immune response of tumor-hearing mice prevents métastasesbut tails to control thedissemination and persistence of tumor cells inprivileged sites. A. groups of DBA/2 or nit/nit micewere injected iep with 2 x IO6 EblacZ cells. At the

indicated lime points, organs were removed from20 animals per group, and single-cell suspensionswere prepared from hone marrow («I.lymph nodes{hi. or spleen (<•).Two cytospins with IO6cells each

for each individual mouse were fixed and stainedovernight for UicZ. Stained cells were scored visually under a light microscope. Results were averaged: huís.SD. DBA/2 (•).DBA/2 irradiated 5 GyI day before (A). SC1D ( 0 ). nii/nu O mice. fl.groups of 10 immunocompetent (O. H) or suhle-thally irradiated DBA/2 (A) mice were injected s.c.with 2 X IO6 EblacZ cells once in the right Hank

(O, A) or in both flanks (H). The frequency ofEblacZ cells in the bone marrow was scored asbefore as a function of days after the first challenge.

ng cel

ed

B

« .>v) '£— O)

s IN nSJ S£n ui,:

10s-104-jIO3-

i/i2io-!{a

AÖafP,'•'•«

«•fT??

ffbone

marrow•

>•- •*

P Ivmph node

01234 25 01234 25

spieen

01234

-*-8 25

Weeks after injection of 2x IO6 Ehi acZ cells i.e.

tf¡oi

lrauJCu-u

—KJ<U«s1*w

2io6-io5-104IO3.io2.IO1.1.bone

marrowA//g^i-H—

-Hg.oA-;ó--'°O'"'

—•-DBA/2

--C-- nu/nu

--0-- SCID

A 5 Gy irnul.DBA/2

-©—DBA/2, one flank inoculated

-B- DBA/2, both flanks inoculated

Weeks after injection of 2x IO6 Ebl acZ cells s.c.

mancy by staining bone marrow cytospins for /at'Z-positive tumor

cells. In parallel, other mice were subjected to a course of treatmentwith depleting monoclonal CDS antibodies or irradiation at 5 Gy andexamined at regular intervals for presence of tumor cells in the bonemarrow. Tumor-dormant mice responded to CDS depletion (Fig. 6c),but not to irradiation (Fig. 6d) with expansion of bone marrow-

residing EblacZ cells. Thus, after four courses of antibody treatment,the frequency of bone marrow-residing tumor cells increased by two

orders of magnitude. These observations are in accord with ourprevious findings that 5 Gy of irradiation suppresses the generation ofbut not the function of effector cells (17). The data also indicate thatCD8+ T cells play a critical role in containing bone marrow-residing

tumor cells.To investigate whether effector CD8+ T cells reside in the bone

marrow and provide protection against a s.c. challenge with EblacZcells, BMT experiments were performed. MHC congenie B10.D2 andsyngeneic DBA/2 mice were inoculated i.v. or iep, respectively, withEblacZ cells. Four weeks later, 2 X IO7 bone marrow cells from the

inoculated mice were injected into naive recipient DBA/2 mice thatwere challenged 1 day later s.c. with EblacZ cells. DBA/2 micereceiving EblacZ-primed, B10.D2-derived immune bone marrow[B10.D2 mice are resistant against i.v.-inoculated tumor cells (18)]showed significant (P = 0.028) protection against challenge with live

EblacZ (Fig. 7, group I), whereas control DBA/2 mice receivingnonimmune B10.D2 bone marrow showed no protection (Fig. 7,group III). Preirradiation (5 Gy) of recipient mice did not significantlyinfluence the protective effect of BMT (data not shown). The protective effect was completely abolished by in vitro depletion of CDS + T

cells prior to injection of the bone marrow cells into the hosts (Fig. 7,group II). The protective effect is consistent with the observed elevated levels of EblacZ-specific INF-y production by spleen cells fromthe immune bone marrow-reconstituted animals (Fig. 3C). The bonemarrow-mediated effector response was sufficiently strong to allow

therapy of established tumors: the combination of sublethal irradiation

of tumor-bearing DBA/2 mice with transplant of bone marrow from

immune B10.D2 mice resulted in regression and significantly(P = 0.028) prolonged survival (Fig. 7, groups IV and V).

Transplantation of DBA/2 immune bone marrow from vaccinated(Fig. 7, group VI) or tumor-bearing (Figure. 7, groups VII and VIII)hosts into normal DBA/2 mice also transferred CDS-dependent pro-

tectitve immunity against challenge with live EblacZ cells, but theeffects were not as significant and long-lasting as those seen with

BMT from B10.D2 donors.These observations provide evidence for the coexistence of tumor cells

and immune CD8+ T-cells with the potential for protective immunity in

the bone marrow of mice inoculated with EblacZ tumor cells.

DISCUSSION

Our data show that potentially metastatic EblacZ cells can reside asdormant tumor cells in lymph nodes and bone marrow as long asCDS + T cells can mediate protective immunity that prevents the atleast partially cycling cell population from expanding and disseminating. Although the persistence of dormant tumor cells has beendescribed previously in tumor cell-vaccinated mice (11, 19), ourresults here show that they also exist in tumor-bearing, s.c.-injected

mice, and more importantly, that in both cases they are kept in adormant state by an active immune response with the critical involvement of CDS + lymphocytes. The local tumor growth in s.c.-inocu

lated mice may be related to a local overriding of immune control.Local versus systemic defeat of the immune control by tumor cells hasbeen reported and discussed before (20). Local rate-limiting factors

have been reported, even in the case of cancers caused by viralinfections. The local growth of Rous sarcoma virus-induced tumors in

chicks was related to local tissue damage and production of transforming growth factor ßat the site of virus inoculation (21-24).

The expression of nuclear lacZ by Eb cells provides a sensitivemethod for detecting tumor cells at the single-cell level. Both mac-

5443

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Il MOK IX)KMANCY IN BONE MARROW

Fig. 5. Tumorigenic potential and proliferation of residualtumor cells delected in ihe marrow. A, tumorigenic potential ofbone niarrow-deríved EblacZ cells. To lest the tumorigenic po-tenlial of bone marrow-derived tumor cells, 2 X IO5 ( 4 ) or2 X IO6 (0) EblacZ cells expanded in vitro from the bone

marrow of ¡ep-inoculated DBA/2 mice were injected s.c. intoeach of at least 10 naive DBA/2 mice. As control. 2 X 10s •

normal, in nfro-passag ed EblacZ cells were inoculated. B. proliferation potential of bone marrow-residing EblacZ cells. Bone

marrow cytospins from three independent experiments each usinggroups of three mice were stained with anti-Ki67, anti-0-gatac-losidase, and DAPI. Each panel shows three representative fieldsfrom independent cytospins al X I(MX).At least 100 bone marrow-residing, ß-galactosidase-stained cells were scored to quantitatethe frequency of Ki67-positive tumor cells, ti, ß-galactosidase-siained tumor cells, showing precipitation of (he blue staining innuclei of tumor cells under the fixing conditions used. h. DAPIstaining for DNA showing general outline of nuclei in the >amefield; note the uppermost EblacZ cell with condensed chromosomes, undergoing cell division, c. Ki-67 staining in the samefield. Arrows, tumor cells. See main text for quantitations.

aca

0 7 14 21 28 35

days after inoculation

Ba

\

\

\\

roscopic and histológica! analyses showed that from the organs examined, only the hone marrow and lymph nodes harbored residualiunior cells. Because iep or s.c. inoculation of immune-suppressedanimals reproducing led to the outgrowth of bone marrow-residing

EblacZ cells and their metastasis and because ¡mmunosuppressivemanipulations could break tumor dormancy in immunocompetentanimals, one can conclude that in immunocompetenl mice, it was thehost immune response that actively controlled and determined the

number and distribution of residual tumor cells. This balance betweentumor cell number and immune response could eventually break downfor a number of reasons, including immune escape or tolerance.

It is not clear why in the presented animal model the bone marrowand lymph nodes but not other organs should be the site of tumor cellpersistence and immune control. The development of secondary tumors at distant sites many years after successful therapy of theprimary tumor is a common observation in cancer patients and

Fig. 6. Control of EblacZ Iunior cell number in thebone marrow requires CDS immune T cells. DBA/2mice were injected iep (d) I day after depletion withanti-CDS monoclonal antibody (~ebni arrow), 1 day

after irradiation (closedarrow, bì.Ìweeks before thestart of a course of depletion (:t-hru tirntws; c), or 3

weeks before irradiation (closed arrow: d). At eachindicated lime point, three mice <a-CD8-trealed) orsix mice (irradiated! were sacrificed, and the frequency of /(/('/-positive cells in the bone marrow was

determined in duplicate cytospins for untreated (•.A). CD8 ' -depleted (D). or irradiated (A) mice.

naive animals tumor dormant animals

_L0.5meT o-CDSi.p.

|5Gyirradiation i 0.5m

T 0-CI15 Gy

irradiationJ2 10°

ga105-i

1io4-'•-,?1103~8

SMS I«2"

¡|„.W^

1 1a

1•

•fb

AAA

11U"

105-104-103-102

-

10 -

iC§DD•

••yleftdi £u

01 234501 2345 012 3 456789 10 12 3 456789 10

Weeks after injection of 2x 10^ EblacZ i.e.

5444

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TUMOR DORMANCY IN BONüMARROW

day:-281

2xl05i.v.

EblacZ11

2xl05i.v.EblacZ"I

lOOuli.v.PBSIV

2xl05i.v.

EblacZV

lOOuli.v.PBS-145xl05s.c.

EblacZ5xl05s.c.

EblacZ0BMTbm

CD8-

BMTBMT5

Gy irradiation &BMT5

Gy irradiation &BMT12xl06s.c.

EblacZ2xl06s.c.

EblacZ2xl06s.c.

EblacZl2xIO¿s.c.

EblacZ2xl06s.c.

|EblacZ

DBA/2(donor)

day: -28

VI 2xl06i.eEblacZ

v" 2xl06i.e

EblacZ

Vili 2xl06s.c.EblacZ

IX 100 ul s.c.PBS

DBA/2 (recipient)

0

BMT

bmCD8-

BMT

BMT

BMT

2xl05s.c.EblacZ

surviving mice, % (at 10 weeks)

10 20 30 40 50 60 70 80 90 100

2xl05s.c.EblacZ

2xl05s.c.EblacZ

2xl05s.c.

EblacZ

0 10

surviving mice, % (at 3 D and 5 •weeks)! ï I T^^^^ î I _^ ^ ^

20 30 40 50 60 70 80 90 100

Fig. 7. EblacZ-specific CDS immune T cells coexist in the bone marrow and can transfer protective immunity and INF-y response. Groups of at least IO B10.D/2 (groups I-V) orDBA/2 mice (groups VI-IX) were inoculated with 2 X IO5 or 2 X 10" EblacZ cells, respectively. The mode of inoculation was i.v. (groups I-V). iep (groups VI and VII), or s.c. (groupsVili and IX). Four weeks later, 2 X IO7 bone marrow cells from these mice were transferred (BMT) i.v. into each of at least 10 naive DBA/2 mice. The bone marrow was (CDS")

or was not depleted in vitro from CDS cells prior to transfer, as indicated. Survival of recipient DBA/2 mice challenged s.c. with 2 X 10h or 2 X 10s EblacZ cells 1 day after honemarrow transfer was followed. Group IV. a therapy experiment where DBA/2 mice were inoculated s.c. with 5 X IO5 EblacZ 2 weeks before irradiation at 5 Gy and BMT. Group V,

control donors injected i.v. with PBS. The Table is designed in the form of a flow diagram with the chronology of (he manipulations being presented at the top of the table. The threedifferent shades of ho.\es (Hack, gray, and while) signify different lime points at which survival was checked (10. 5. or 3 weeks after challenge). *. experiments with BIO.D2 donorswere statistically significant as determined by the log-rank test (P - 0.028 exact comparing group I with group III; P = 0.0635 exact comparing group I with group II; P = 0.444

exact comparing group II with group III). Exact tests were performed using the Slat Xact software package (Cytel Software Corp.). bm. bone marrow; i.e. iep.

strongly supports the notion of postoperative persistence of tumorcells in a dormancy state. Several clinical observations point to thebone marrow as a major site of persistence of tumor cells in curativelyoperated or treated cancer patients. A significant fraction of patientswith non-small-cell lung carcinomas (25), prostate cancer (26, 27),

colon cancer (28, 29), squamous cell carcinomas of the head and neckregion (30), gastric cancer (31, 32), and breast cancer (33, 34) harbored low numbers of tumor cells in the bone marrow at the time ofand after operation of the primary tumor. Similarly, tumor cells werefound in regional lymph nodes in metastasis-free cancer patients (35).The detection of bone marrow-residing tumor cells has been consid

ered to be a significant and independent predictor for a later clinicalrelapse in distant organs (25, 28, 31).

The microenvironment of the tumor cells is likely to play a determining role in the maintenance of tumor dormancy state as well as thetransition to tumor metastasis (12, 20, 36, 37). Although neovascu-

larization may be a critical factor for the eventual outbreak of metastasis (38), it is highly unlikely to be limiting the number of tumor cellsin the bone marrow or lymphatic system. Bone marrow and lymphnodes are two prominent organs involved in antigen presentation andimmune response (39,40). The bone marrow of immunized animals isa rich source of protective immune cells (41,42). Our observations areconsistent with these organs offering favorable conditions for homingof tumor cells and their survival, in spite of the presence of protectiveimmune cells at the same sites.

5445

ACKNOWLEDGMENTS

We are grateful to H. von Böhmer for critical comments and interest. S.Eccles, T. Blankenstein. U. Haberkorn, U. Moebius, and S. Psarras are thankedfor support and encouragement. We thank Dr. J. Gerdes (ForschungsinstitutBorstel. Kiel, Germany) for the generous gift of anti-mouse KÌ67antibody.

Anja Fre/.za and Anja Sierrnann are thanked tor excellent technical assistance.

REFERENCES

1. Schirrmacher. V.. Bosslet. K., Shantz. G.. Clauer, K., and Hübsch.D. Tumormétastasesand cell-mediated immunity in a model system in DBA/2 mice. IV.Antigenic differences between a metastasizing variant and the parental tumor linerevealed by cytotoxic T lymphocytes. Int. J. Cancer. 23: 245-252, 1979.

2. Schirrmacher. V.. Shantz, G.. Clauer. K.. Komitowski. D.. and Zimmermann. H. P.Tumor métastasesand cell-mediated immunity in a model system in DBA/2 mice. I.Tumor invasiveness in vitro and métastasesformation in vivo. Int. J. Cancer, 23:233-244, 1979.

3. Schirrmacher. V.. and Bosslet. K. Tumor métastasesand cell mediated immunity ina model system in DBA/2 mice. X. Immunoselection of tumor variants differing intumor antigen expression and metastatic capacity. Int. J. Cancer, 25: 78I-78S. 1980.

4. Schirrmacher, V., Fogel. M., Russmann. E.. Bosslet, K.. Altevogt. P.. and Beck, L.Antigenic variation in cancer metastasis: immune escape versus immune control.Cancer Metastasis Rev.. 1: 241-274. 1982.

5. Schirrmacher, V., Leidig. S., and Griesbach. A. In siln activation of syngeneictumour-specific cytotoxic T lymphocytes: intra-pinna immunization followed byrestimulation in the peritoneal cavity. Cancer Immunol. Immunother., JJ: 299-306.

1991.6. Weinhold. K. J. L. T.. Goldstein, L. T., and Wheelock, E. F. Tumour-dormant state

established with LSI78 lymphoma cells in immunized syngeneic murine hosts.Nature (Lond.). 270: 59-60. 1977.

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TUMOR DORMANCY IN BONK MARROW

7. Weinhold, K. J., Goldslein. L. T.. and Wheelock. E. F. The Iunior dormant siale:quaniiiation ot L5178 Y cells and hos! immune responses during the establishment andcourse of dormancy in syngeneic DBA/2 mice. J. Exp. Med.. 149: 732-744. 1979.

K. Weinhold. K. J.. Miller. D. A., and Wheelock, E. F. The tumor dormant stale:comparison of L5178Y cells used to establish dormancy with those that emerge afterits termination. J. Exp. Med.. 14t: 745-757. 1979.

9. Chen. L. P.. Suzuki. Y., Liu, C. M., and Wheelock. E. F. Maintenance and cure of theL5178Y murine lumor-domiant state by interleukin 2: dependence of intcrleukin 2 oninduced Interferon-7 and on tumor necrosis factor for its anlilumor effects. CancerRes., 50: 1368-1374. 1990.

K). Whcelock, E. F.. Yang. G.. and Chen. L. Immune regulation of a murine T-cell

lymphoma dormant slate. In: T. H. M. Stewart and E. F. Wheelock (eds.). CellularImmune Mechanisms and Tumor Dormancy, pp. 53-65. Boca Raton. FL: CRC Press,

1992.11. Khazaie. K.. Prifli. S.. Beckhove. P.. Griesbach. A.. Russell. S.. Collins. M.. and

Schirrmacher, V. Persistence of dormant tumor cells in the bone marrow of tumorcell-vaccinated mice correlates with long-term immunological protection. Proc. Nail.Acad. Sci. USA. 91: 7430-7434, 1994.

12. Wheclock. E. F. An overview of mechanisms responsible for tumor dormancy. In:T. H. M. Stewart and E. F. Wheelock (eds.). Cellular Immune Mechanisms andTumor Dormancy, pp. 1-17. Boca Raton, FL: CRC Press. 1992.

13. Danos, O., and Heard, J. M. Retroviral mediated gene transfer into hepatocytes invivo. Proc. Nati. Acad. Sci. USA, 88: 8377-8381. 1991.

14. Danos. O., ami Mulligan. R. Safe and efficient generation of recomninant retroviruseswith amphotropic and ecotropic host range. Proc. Nati. Acad. Sci. USA. 85: 6460-

6464. 1988.15. Kubo. R. T.. Born. W.. Kappler. J. W.. Marrack. P.. and Pigeon, M. Characterization

of a monoclonal antibody which delects all murine aßT cell receptors. J. Immunol..142: 2736-2742, 1989.

16. Schild, H-J., Kyewski, B.. Von Hoegen. P.. and Schirrmacher. V. CD4+ helper T

cells are required for resistance to a highly metastalic murine tumor. Eur. J. Immunol..17: 1863-1866. 1987.

17. Schirrmacher, V., and Zangemeister-Wiuke, LI.Gamma-irradiation suppresses T cell

mediated protective immunity against a metaslalic tumor in the afferent phase of theimmune response but enhances it in the efferent phase if given before immune celltransfer. Int. J. Oncol., 4: 335-346. 1994.

18. Schirrmacher. V.. Beckhove. P.. Krüger.A., Rocha. M.. Umansky. V.. Fichtner, K. P..Hull. W. E., Zangemeister-Wittke, U., Griesbach. A.. Jurianz. K.. and von-Hogen. P.

Effective immune rejection of advanced metastasized cancer. Int. J. Oncol., 6:505-521. 1995.

19. Morecki, S.. Pugalsch. T.. Levi, S., Moshel. Y., and Slavin. S. Tumor-cell vaccinationinduces tumor dormancy in a murine model of B-cell leukemia/lymphoma (BCL-1 ).Int. J. Cancer. 65: 204-208. 19%.

20. Eccles, S. Dormane)' in experimental solid tumor systems, in: T. Stewart and F.Wheelock (eds.), Cellular Immune Mechanisms and Tumor Dormancy, pp. 27-53.

Boca Raton, FL: CRC Press. 1993.21. Dolberg. D. S.. Hollingsworth. R.. Hertle. M.. and Bissell, M. J. Wounding and its

role in RSV-mediated tumor formation. Science (Washington DC). 230: 676-678,

1985.22. Sieweke. M. H.. Thompson. N. L.. Sporn. M. B., and Bissell. M. J. Mediation of

wound-related Rous sarcoma virus tumorigenesis by TGF-ß.Science (WashingtonDC). 248: 1656-1660. 1990.

23. Martins. G. M., Boudreau. N.. and Bissell. M. J. Inflammation is responsible for thedevelopment of wound-induced tumors in chickens infected with Rous sarcoma virus.Cancer Res.. 54: 4334-4341. 1994.

24. Wisner, T. W.. England. J. M., Pan. D. Y.. Stoker, A. W.. and Halpern. M. S. Tumorimmunity generated in the course of regression of v-src-induced sarcomas. J. Viro].,65: 7020-7024, 1991.

25. Panlel. K.. Izbicki. J.. Passlick, B.. Angstwurm. M.. Haussinger. K.. Thelter. O.. andRiethmuller, G. Frequency and prognostic significance of isolated tumour cells in

bone marrow of patients with non-small-cell lung cancer without overt métastases.Lancet, 347: 649-653, 1996.

26. Pantel. K.. Aignherr. C.. Kollermann. J.. C'aprano. J.. Riethmuller. G.. and Koller-

iii.um. M. W. Immunocytochemical detection of isolated tumour cells in bone marrowof patients with untreated stage C prostatic cancer. Eur. J. Cancer, 31A: 1627-1632.1995.

27. Riesenberg, R.. Oberneder. R.. Kriegmair, M., Epp, M.. Bitzer. U., Hofstetter. A.,Braun. S., Riethmuller. G.. and Pantel. K. Immunocytochemical double staining ofcytokeratin and prostate specific antigen in individual prostatic tumour cells. Histo-chemistry, 99: 61-66, 1993.

28. Braun. S., and Pantel. K. Immunodiagnosis and immunotherapy of isolated tumorcells disseminated to bone marrow of patients with colorectal cancer. Tumori. 81:78-83. 1995.

29. Schlimok. G., Funke. I.. Bock. B.. Schweiherer. B.. Witte, J.. and Riethmuller, G.Epithelial tumor cells in bone marrow of patients with colorectal cancer: immuno-

cytochcmical detection, phenotypic characterization, and prognostic significance.J. Clin. Oncol.. 8. 831-837. 1990.

30. Gath. H. J.. Heissler. E., Hell. B.. Bier. J.. Riethmuller. G.. and Pantel. K. Immuno-cytologic detection of isolated tumor cells in bone marrow of patients with squamouscell carcinomas of the head and neck region. Int. J. Oral Maxillofac. Surg., 24:351-355, 1995.

31. Jauch. K. W„Heiss, M. M.. Gruetzner. U., Funke. I.. Pantel. K.. Babic, R., Eissner,H. J., Riethmuller, G.. and Schildbcrg. F. W. Prognostic significance of bone marrowmicrometastases in patients with gastric cancer. J. Clin. Oncol.. 14: 1810-1817, 1996.

32. Schlimok, G., Funke, I.. Pantel. K., Strobel. F.. Lindemann. F.. Witte. J.. andRielhmuller. G. Micrometastatic tumour cells in bone marrow of patients with gastriccancer: methodological aspects of detection and prognostic significance. Eur. J.Cancer, 27: 1461-1465. 1991.

33. Panlel. K.. Schlimok. G., Braun. S.. Kutter. D.. Lindemann. F.. Schaller, G.. Funke.I.. Izhicki, J. R.. and Riethmuller. G. Differential expression of proliferation-associ

ated molecules in individual micrometastalic carcinoma cells. J. Nati. Cancer Inst.,85: 1419-1424, 1993.

34. Funke. I.. Fries, S., Rolle, M.. Heiss, M. M., Unteti, M.. Böhmen. H.. Schildberg.F. W., and Jauch. K. W. Comparative analyses of bone marrow micrometastases inbreast and gastric cancer. Int. J. Cancer. 65: 755-761. 1996.

35. Passlick. B.. Izbicki, J. R.. Kubuschok. B., Nalhrath, W., Thetter, O.. Pichlmeier. U..Schweiberer. L.. Rielhmuller. G.. and Pantel, K. Immunohistochemical assessmenl ofindividual tumor cells in lymph nodes of patients with non-small-cell lung cancer.J. Clin. Oncol.. 12: 1827-1832. 1994.

36. Schirrmacher. V. Cancer metastasis: experimental approaches, theoretical concepls.and impacts for trcatmenl strategies. Adv. Cancer Res., 43: 1-73. 1985.

37. Kohn, E. C.. and Liotta. L. A. Molecular insights into cancer invasion: strategies forprevention and intervention. Cancer Res.. 55.- 1856-1862. 1995.

38. Holmgren. L., O'Reilly, M.. and Folkman. J. Dormancy of micrometastases: balanced

proliferation and apoptosis in the presence of angiogcnesis suppression. Nat. Med., I:149-154. 1995.

39. Huang. A. Y., Bruce, A. T., Pardoll, D. M., and Levitsky, H. 1. Does B7-I expressionconfer antigen-presenting cell capacity lo lumors in vivo? J. Exp. Med.. 183: 769-

776. 1996.40. Doe, B.. Selby. M.. Barnett, S., Baenziger, J., and Walker. C. M. Induction of

cylotoxic T lymphocytes by intramuscular immunization with plasmid DNA isfacilitated by bone marrow-derived cells. Proc. Nati. Acad. Sci. USA. 93: 8578-8583. 1996.

41. Hornung. R. L.. Longo. D. L.. Bowersox. O. C.. and Kwak. L. W. Tumor antigen-specific immunization of bone marrow transplantation donors as adoptive therapyagainst established tumor. J. Nati. Cancer Inst., 87: 1289-1296, 1995.

42. Kwak, L. W., Pennington. R.. and Longo, D. L. Active immunization of murineallogeneic bone marrow transplant donors with B-cell tumor-derived idiotype: astrategy for enhancing the specific anlilumor effect of marrow grafts. Blood. 87:3053-3060. 1996.

5446

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1998;58:5439-5446. Cancer Res   Markus Müller, Fotini Gounari, Sophia Prifti, et al.   Cells

Immune T+Active Control of Proliferating Tumor Cells by CD8EblacZ Tumor Dormancy in Bone Marrow and Lymph Nodes:

  Updated version

  http://cancerres.aacrjournals.org/content/58/23/5439

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/23/5439To request permission to re-use all or part of this article, use this link

on May 16, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from