[CANCERRESEARCH57, 3511—3516,August 15, 1997]

synergistic effect with other murine cytokines (7). Flt3-L alone has aweak stimubatory effect on in vitro progenitor cell colony growth buthas additive or synergistic effects when combined with GM-CSF,granubocyte-CSF, IL-3, IL-i 1, IL- 12, erythropoietin, stem cell factor(steel factor), and other cytokines (7—9).Fbt3-L also appears to havea robe in bymphopoiesis by supporting the proliferation of B-cellprogenitors (8).

Recently, F1t3-L has been noted to possess a growth-stimulatoryeffect on dendritic cells (10). Because dendritic cells are importantantigen-presenting cells, we hypothesized that tumor cells expressingFlt3-L may be useful in inhibiting tumor growth and in promotingtumor immunization. To test this hypothesis, we chose the C3L5munne breast cancer cell model (1 1, 12). Animals were immunizedwith non-irradiated, vector-expressing tumor cell lines, and tumorfree animals were challenged 4 weeks after inoculation with parentC3L5 cells. The antitumor and immunizing activity of recombinantFbt3-L was compared with that of other cytokines with known antitumor activity. Because F1t3-L exists in a variety of isoforms, weconstructed retroviral vectors that encode soluble and membranebound forms of Fbt3-L and compared the antitumor activity of theseconstructs in our model. In this study, Flt3-L was found to possessantitumor properties, whether given systematically or in the context ofa tumor vaccine. Interestingly, sustained antitumor immunity wasobserved only when F1t3-L was administered as a tumor vaccine.

MATERIALS AND METHODS

Retroviral Vector. Vectors used in this study are shown in Fig. 1. Fragments of cDNA sequence encoding human Fbt3-L and specific isoforms ofmarine Flt3-L (provided by Immunex Corp., Seattle, WA) were inserted into

the retroviralvector pLXSN originally developed by A. Dusty Miller. Thevectors L(hFlt3-L)SN and L(mFbt3-L)SN were shuttle packaged through theecotropic packaging cell line GP + E86 (13) into the amphotropic PA317packaging cell line (14). Three additional vectors were obtained through acollaboration with Genetic Therapy, Inc. (Gaithersburg, MD) and used in thisstudy (Fig. 1). GiNa and LXSN retroviral vectors are control vectors containing the neomycin resistance gene alone. GbNaSvi2G contains the NeoR geneand the human IL-2 gene (titer, 4.2 X 10'; ELISA activity, >2000 pg/mI IL-2).The Gbil3SvNa vector contains the NeoR gene and the human IL-1(3 gene(titer, 2.4 X 10'; ELISA activity, 482 ng/ml).

Cell Lines and Transduction. Retrovirabvectors were introducedby supernatant transduction into the syngeneic murine breast cancer cell line C3L5(kindly provided by Dr. P. K. Lala, University of Western Ontario, London,Ontario, Canada; Refs. 11 and 12). One million C3L5 cells were incubatedwith vector at a multiplicity of infection of I :2 for 2 h in the presence ofPolybrene (8 @.tg/ml).C3L5 cells containing vector were obtained by selection

in G418 (400 @.tg/ml)for 10—14days. C3L5 cells were maintainedin DMEMsupplemented with 2 mmol/liter glutamine, 10%FCS, and I mmol/liter sodiumpyruvate and incubated at 37°C,5% CO2. Cell lines used in this study testednegative for the presence of Mycoplasma and replication-competent amphotropic retrovirus.

In Vivo Studies. To assess the effect of i.p. administered human Chinesehamster ovary-derived Flt3-L (Ref. 15; 500 pg/kg daily for 10 days) on thegrowth of cytokine-transduced C3L5, mice were challenged with 1 X b0'@C3L5 cells injected s.c. into the anterior chest wall of C3HIHeN female mice

3511

Antitumor Activity and Immunotherapeutic Properties of F1t3-Ligand in a Murine

Breast Cancer Model'

Keyue Chen, Stephen Braun, Stewart Lyman, Yi Fan, Christie M. Traycoff, Eric A. Wiebke, Jay Gaddy,George Sledge, Hal E. Broxmeyer, and Kenneth Cornetta2

Departments of Medicine (K. C., Y. F., C. M. T., G. S., H. E. B., K. C.], Medical and Molecular Generics (K. CI, Microbiology/immunology (S. B., H. E. B., K. C.], SurgeryIE. A. W.J, and the Walther Oncology Center (S. B., I. G., H. E. B.], Indiana University School of Medicine, indianapolis, indiana 46202, and the Immunex Corporation, Seattle.Washington (S. LI

ABSTRACT

Flt3-Ligand (Flt3-L) is a stimulatory cytokine for a variety of hematopoletic lineages, including dendritic cells and B cells. The antitumorproperties ofFlt3-L were evaluated in C3H/HeN mice challenged with thesyngeneic C3L5 murine breast cancer cell llne@Eighty % of animalsreceiving 500 gLg/kg/day of Chinese hamster ovaryderived human F1t3-L

for 10 days were protected from tumor growth, whether the tumorchallenge was administered on the first or fourth days of Flt3-L adminIstration. The protection provided by soluble F1t3-L was transient. Alltumor-free animals rechallenged 4 weeks after the primary challengedeveloped tumor. Transduction of C3L5 with retroviral vectors expressmg human or marine Flt3-L did not influence in vitro growth or MHCexpression but decreased in vivotumor development to 0 and 10% of mice,respectively. This compares with tumor growth of52% with interleukin-2transduced C3LS and over 85% with untrausduced and control vectortransduced C3L5. Unlike animals treated with soluble F1t3-L, administratlon of F1t3-L as a tumor vaccine protected mice from a subsequentchallenge with untransduced C3L5 in 60—78% of mice, compared to 0%

ofcontrols. Our Initial work used the most common Flt3-L isoform, whichis membrane bound but can undergo proteolytic cleavage to generate asoluble form. To evaluate the role of the various F1t3-L isoforms inpreventing tumor formation, retroviral vectors encoding only the membrane-bound form or only the soluble isoform were evaluated In the C3L5model. Tumor formation was similar with either isoform, preventingtumor formation in 80-90% of mice after the primary challenge and

88-89% after the secondary challenge. Splenocytes obtained 4 weeksafter the secondary challenge conferred adoptive immunity to naive mice

in 60% of animals. This initial report of antitumor activity by F1t3-Lisconsistent with its known stimulatory effect on antigen-presenting cellsand suggests It may enhance the development of tumor vaccines.

INTRODUCTION

F1t3, also known as flU and Stk-l, is a type ifi tyrosine kinasemembrane receptor first identified in hematopoietic stem and progenitor cells, including CD34+ bone marrow and cord blood cells, and isbelieved to play a role in the regulation of hematopoiesis. (1—3)TheF1t3-L3 is a cytokine with potent costimubatory activity. The murineand human homobogues of Flt3-L have been cloned and share 72%amino acid homology (4—6).Human Flt3-L stimulates proliferationof cells expressing the murine Flt3-L receptor, has a similar biologicalactivity on murine bone marrow cells as munne Fbt3-L, and has a

Received 3/7/97; accepted 6/11/97.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 fact.

I This work was supported by a Breast Cancer Research Grant from the Catherine

PeachyFoundationandPublicHealthServiceGrantP01 CA59348(toK.C.).Support(toH. E. B.) includes USPHS Grants ROl HL54037, ROl HL56416, and a project on P01 HL53586. During the course ofthis work, S. B. was supported by NIH training program T32DK07519 (to H. E. B.) and is currently a Fellow ofthe Leukemia Society ofAmerica, Inc.

2 To whom requests for reprints should be addressed, at the Division of Hematology/

Oncology, Department of Medicine, Indiana University School of Medicine, IB442, 975WestWalnutStreet,Indianapolis,IN 46202.Phone:(317)274-0843;Fax: (317)274-0396.

3 The abbreviations used are: Flt3-L, Flt3-ligand; hFlt3-L human F1t3-L; GM-CSF,

granulocyte/macrophage-colony-stimulating factor; IL, interleukin.

on March 6, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACrIvITY OF Flt3-LIGANDIN BREASTCANCER

A.@ TAt@@@ GiNaLTR SEC LIP

B. E@ TA@1@'///A-@ 1@ L(hFIt3-L)SNLTR hFIt3@L SV4O N@ LIP

C. TA@[@'@' I@ L(muFIt3-L)SNLiP mFtt3@L SV4O 1@EO LiP

D. E@F TA@@ L(5H)SNLTP mFIt3-L SV4O F@EO LTP

K TAc1@@@@ — L(Ex6)SNLTP mFlt3-L SV4O ,@ LTP

F.@ Giii3SvNaLTP h-ILl beta SV4O ,@ LTP

G.@ I I I@ G1NaSvI2GLIP TAG t€O SV4O hIL•2 LIP

I.I KB

(primary challenge). Animals were assessed for evidence of local tumorgrowth. Tumors were measured weekly in two dimensions, and tumor size wascalculated using the formula: volume (width2 X length)/2. If animals weretumor-free at 4 weeks, they were challenged with 1 X 10―untransduced C3L5cells (secondary challenge). In a subset of animals tumor-free 4 weeks aftersecondary challenge, splenocytes were harvested and injected into naive C3HIHeN mice that simultaneously received 5 X l0@untransduced C3L5 cells.Splenocytes were enriched for T cells and CD4-positive lymphocytes using theIsoCell Murine I Cell Isolation and IsoCell Murine CD4 Cell Isolation kits(Pierce, Inc., Rockford, IL).

Flow Cytometry. Class I and class II MHC antigens were evaluatedusing the phycoerythrin-conjugated anti-mouse H-2Kk and FITC-conju

gated anti-mouse I-Ak (PharMingen, San Diego, CA). Simultest Control‘Yt@'Y2aFITC/phycoerythrin antibodies (Becton Dickinson, San Jose, CA)

were used as the isotype control. For cytometric analysis, 5 X l0@cellswere washed in PBS, pelleted, and incubated with the appropriate antibodyfor 20 mm on ice. Antibody-exposed cells were washed once in PBS with

I% human serum albumin and then resuspended in a volume of 0.5 ml andanalyzed on a FACScan (Becton Dickinson Immunocytometry Systems).Membrane-bound Flt3-L was evaluated using a human Flt3 receptor-humanFc fusion protein (4). The receptor-Fc protein was incubated with cells for30 mm at room temperature. Cells were washed twice with PBS containing2% FBS + 0.1% sodium azide and incubated with biotin-conjugated

anti-human Fc (Caltag, Burlingame, CA) for 30 mm. Tertiary staining wasperformed with streptavidin-conjugated with TRICOLOR (Caltag) and

analyzed on a FACScan.

Fig. 1. Retroviral vectors. A. GINa vector containing the neomycin resistance gene(NEO). B. L(hFlt3-L)SN vector containing the hFlz3.L gene and the NEO gene 3' to theSV4O early promoter. C. L(muFlt3-L)SN vector containing the murine F1t3-L gene(@nFlt3-L) with proteolytic cleavage site (black box) and the transmembrane region(speckled box). D, L(E6)SN vector containing the soluble isoform of the mFlt3-L, whichlacks the transmembrane portion. E, L(5H)SN vector containing the membrane-boundisoformofthemFlt3-Lgene,whichretainsthetransmembraneportion(speckledbox)butlacks the proteolytic cleavage site. F, G Ii l2SvNa vector containing the human 1L-l@(h-IL! beta) gene. G. GlNaSvi2G vector containing the human IL-2 gene 3' to the SV4Oearly promoter. LTR. long terminal repeat; TAG. stop codon replacing the gag start codon.5. LTR are derived from the Moloney sarcoma virus, whereas 3' LTR are from theMoloney leukemia virus.

.. ..

.. ..

00 • 00 I

00 00

NoTumor

4_

g

I 2'j

0 —

. FIt3L . FIt3L

DATO DAY4

Fig. 2. Tumor growth in mice challenged with C3L5 breast cancer cells. Mice werechallenged with I X 1ff' C3L5 cells during administration of Flt3-L. Mice received dailyi.p. injection of 500 @zg/kg human Flt3-L daily for 10 days. The diluent, PBS, was

administered to the control group. Tumor challenge was given on the first (day 0) or fourth(day 4) day of Flt3-L administration. 0, PBS-treated mice; @,Flt3-L group. Mice wereexamined weekly for tumor growth.

RESULTS

Soluble Flt3-L Administration. The antitumor activity of solubleFlt3-L was evaluated using the murine breast cancer cell line C3L5.Mice received daily i.p. administration of human Flt3-L (500 @g/kg/day) for 10 days. This dose was selected because it has no apparenttoxicity in mice (15), is known to mobilize murine progenitor cells(15),andstimulatesmurinedendriticcellproliferation(10).Primarytumor challenge consisted of 1 X lO@C3L5 cells. Animals receivedtumor challenge on the first (day 0) or fourth (day 4) day of F1t3-Ladministration. As shown in Fig. 2, all mice injected with controlvehicle (PBS) developed tumors by week 2, whereas four of five micein the Flt3-L group were tumor free at 4 weeks. Tumor-free animals

were then injected with 5 X b0@C3L5 cells in the contralateral chestwall and monitored for tumor growth. Each of the four animals in theday 0 and day 4 groups developed tumors, indicating a lack ofimmunity (data not shown).

Cytokine Expressing Retroviral Vector in C3LS. To assess theeffectiveness of Fbt3-L in a tumor vaccine, populations of cytokinesecreting C3L5 cells were generated using the retroviral vectorsshown in Fig. I . Prior to in vivo evaluation, the in vitro growth ofvector-transduced cells was assessed to determine potential effects on

growth from cytokine expression. As shown in Fig. 3, the growth ratesof all vector-transduced C3L5 populations were comparable to un

transduced C3L5, suggesting the cytokines studied did not influencecell growth. The one exception noted was C3L5 expressing IL-b j3,which demonstrated reduced proliferation after transduction.

Vector-transduced cells were also evaluated for production of human Flt3-L and human IL-2 using assays described previously (16,17). The parent C3L5 cell line does not produce detectable levels ofsoluble hFbt3-L or human IL-2 when evaluated by ELISA. Cells

3512

on March 6, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF F1t3-LIGAND IN BREAST CANCER

were used as a challenge control. Our findings, shown in Fig. 5B,

demonstrate that the majority of animals injected with C3L5 cells

expressing murine Flt3-L, hFlt3-L, and IL-2 are immunized to asubsequent challenge with untransduced C3L5 cells (P < 0.01 for allgroups when compared to control). Interestingly, although C3L5 cells

transduced with IL-lf3 did not form tumors in vivo (Fig. 5A), this didnot result in immunization against a subsequent challenge with theuntransduced C3L5 (Fig. 5B).

Membrane-Bound versus Soluble Flt3-L Isoforms. The murineF1t3-L gene gives rise to three functional isoforms, which can bemembrane bound, soluble, or both (19). To evaluate the relative role

of Flt3-L isoforms in preventing tumor growth, we constructed retroviral vectors producing either a strictly membrane-associated form or

a purely soluble F1t3-L [L(5H)SN and L(Ex6)SN, respectively; Fig.

1]. As shown in Table 1, both isoforms significantly decreased primary tumor formation (P < 0.001) and immunized mice against a

second challenge of the parent C3L5 cell line (P < 0.01) when

compared to control C3L5. As shown in Table 2, inhibition of tumor

growth by F1t3-L vectors can be overcome by increasing the dose ofF1t3-L-transduced C3L5 cells. There was no clear difference betweeneither isot'orm with regard to preventing tumor growth or slowing

tumor development.

Protection from Tumor Challenge by Adoptive Transfer ofImmunity. The mechanisms by which tumor rejection was mediatedwere initially evaluated using cytotoxicity assays described previously(20). Coculture of chromium-labeled C3L5 cells and splenocytes from

animals immunized with untransduced and vector-transduced C3L5

cells failed to demonstrate significant levels of cytotoxicity. These

results were seen using freshly explanted splenocytes or splenocytes

restimulated in vitro for 1 week with irradiated C3L5 cells prior to

cytotoxicity testing (data not shown). Because this in vitro assay may

not accurately reflect the in vivo environment, we chose to evaluate

adoptive transfer of splenocytes from immunized animals. These

transduced with GINaSvi2G vector produced 551 pg of IL-2/l06cebbs/24 h. C3L5 cells transduced with the L(hFlt3-L)SN vector,which encodes a transmembrane isoform of human F1t3-L that isexpressed on the cell surface and can be proteolytically cleaved togenerate a soluble form (18), produces 6058 pg of hFlt3-U106cells/24 h. Using a human Flt3 receptor-human Fc hybrid protein,membrane-bound human F1t3-L was also detected on transduced

C3L5 cells (Fig. 4). Murine F1t3-L bigand production was not assessedin this study, due to the lack of an anti-murine Flt3-L antibody.

Primary Tumor Challenge. To determine whether tumor formation would be affected by cytokine gene expression, C3HIHeN femalemice received s.c. injections of 1 X i0@ cytokine-transduced C3L5cells. Animals were followed weekly for 4 weeks for evidence of

tumor growth. Palpable tumors were measured weekly in two dimensions, and the results of this analysis are shown in Fig. 5A. C3L5 cellsexpressing IL-2, murine Fbt3-L, hFbt3-L, and human IL- I f3 preventedtumor growth in 40, 90, 100, and 100% of animals, respectively. Atwo-tailed Fisher' 5 Exact test was used to compare the proportion oftumor-bearing mice in each group to control. No significant differ- _________ences were noted in the number of unimuls developing tumors in theuntrttnsduced and GiNa control groups (P = 1) Significantly lesstumor growth was noted in cytokine-expressing C3L5 groups(P < 0.001 for all groups).

Tumor Immunization, To determine if tumor-free animals injected with eytokine@expressing C3L5 were immunized against theparent tumor, we performed a secondary challenge with untransducedC3L5 cells 4 weeks after the primary tumor challenge Naive mice

3513

--@0-- C3LS

.....o@.. GiN

,—*-.-- IL-2

- IL.i

----a... F1t3-L

7

106

10@

2 4 6 8

0.@

Ez0C.)

Days

Fig. 3. Growth characteristics of gene-transduced C3L5. Vector-containing C3L5 cellswere growth for 8 days in triplicate cultures. Cell expansion was determined for cellscontainingthe neomycinresistancegene(GIW),the GlNaSvi2Gvector(IL-2).theGlil2SvNa vector (IL.1), and the L(hFlt3-L)SN vector (Flt3-L). Bars, SE.

: Control

¶ -@@ — C3L5-L(huFlt3L)SN

@::@ i@1t@1L@@ J.@

:@ @11•@@ .@ .,, Ht1@

@ 2i@@

@ @r@j n@r-@ ,@l

@@ fl@

@i@ p@_J@i/@ ________

FluorescenceFig.4. Surfaceexpressionof humanF1t3-L.C3LScellswereanalyzedt@rexpressit)fl

of membrane-bound F1t3-L using a human Flt3 receptor extracellular domain-human Fehybrid protein. Afier incubation with the hybrid protein, cells were stained with hiotinconjugated human anti-Fe followed by staining with strepuvidin-laheled with TRICOI.A)Rreagent. Untransduced C3L5 cells (C3L5) sod vector-containing C3LS cells IC.1L5-L(huFlliL.)SNI were analyzed. Control cells represents vector-containing cells stainedwith all antibodies except the hybrid protein.

on March 6, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

C3L5 Neo IL-i 1L.2 hFtt3'L mFIt3.L

Table 1 Tumor growth in mice challenged with C3LS cells containingretroviralvectorsexpressing two isoforms ofFlt3-LData

indicated the number of mice developing tumor/total number of micechallenged.TheL(5H)SN vector encodes a membrane-bound isoform, whereas L(Ex6)SN encodes a

soluble isoform of murine Flt3-L. Initially, mice received a challenge of I Xl0@vector-transducedor untransduced C3L5. After 4 weeks, tumor-bearing mice were enu

merated (Primary challenge). Tumor-free mice at 4 weeks were challenged with 5 Xl0@untransducedC3L5. After an additional 4 weeks, the number of tumor-free micewasdetermined

(Secondary challenge).Primary

challenge SecondarychallengeC3L5

9/9C3LSINeovector10/10C3L5/L(5H)SN

2/101/8C3L5/L(Ex6)SN1/101/9Naive

mice 5/5

IL-i IL'2 hFlt3'L mFlt3-L

Tumor volumes were measured weekly for 4 weeks in mice injected with untransducedcells (C3L5) or C3L5 transduced with the retroviral vectors L(Ex6)SN or L(5H)SNexpressing the soluble and membrane-bound isoforms of murine Flt3-L, respectively. Fivedoses were used in the tumor challen@eranging from I X l0@to 5 X 106tumor cells. Datarepresent the estimated volume (mm ) ±SE from threeanimals.Tumor

doselXlO4

lXlO@ 5Xl05 lXlO65X106Week

1C3L5 2 ±I NT 33 ±7 NT 697 ±106Ex6 0±0 0±0 0.8±0.2 12±4 54±11SH 0±0 0±0 2±1 8±3 25±4

Week 2C3L5 18±8 NT 109±17 NT 1580±110Ex6 0±0 2±2 13±3 79±20 204±85H 0±0 2±2 23±5 91±26 355±30

Week 3C3L5 155 ±21 NT 410 ±45 NT 3048 ±333Ex6 0±0 13±13 66±9 216±25 453±48SH 0 ±0 13±13 86±10 223±50 562±64

Week 4C3L5 599 ± 133 NT 1480 ± 163 NT 4595 ±511

Ex6 0 ±0 42 ±42 196±0 472 ±55 1073±73SH 0±0 36±36 216±20 761±2311374±68a

@‘not tested.

ANTITUMOR ACTIVITY OF Flt3-LIGAND IN BREAST CANCER

addition, tumor-free animals treated previously with Flt3-L-expressing C3L5 cells are protected from a second challenge with the parentC3L5 cells. This is in contrast to IL-l@-transduced C3L5 cells, whichprevented tumor growth in vivo but did not immunize animals againsta secondary challenge. IL-1f3-transduced C3L5 cells demonstrateddecreased in vitro growth that may have contributed, at beast in part,to the back of tumor growth in vivo. Tumor formation after a secondary challenge in IL-l(3-C3L5-treated mice and in the occasionalcontrol mouse that did not develop tumor after the primary challenge(Fig. 5) suggests that inoculation of tumor cells alone is insufficient toelicit immunity to C3L5.

A number of investigators have demonstrated tumor immunizationusing cytokine-expressing tumor cells. Previously tested cytokineshave a variety of stimulatory and inhibitory activities and includeagents such as GM-CSF, IL-2, IL-4, IL-6, IL-7, tumor necrosis factor,and the interferons (23—27).In the present model, human and murineFbt3-L prevented primary tumor growth in the majority of animals,whereas IL-2 protected approximately 50% of animals. Flt3-L is thefirst early-acting cytokine with known stimulatory activities of dendritic and pre-B cells to exhibit tumor-immunizing properties.

The primary translation products of the human and murine F1t3-Lgenes are transmembrane proteins that can undergo proteolytic cbeavage, generating a soluble protein (4, 6). These cDNAs were used togenerate the L(hFlt3-L)SN and L(mFlt3-L)SN vectors. In addition,alternative splicing of Fbt3-L mRNA can generate isoforms that arepurely soluble or only membrane bound (18, 19). Inclusion of intron

A.

B.

8,,.@i

0@0

4,U

4,,

1oo@@_@

40 -

@1

Fig. 5. Tumor growth after primary and secondary challenge with vector-transducedC3L5 cells. In A. the primary challenge consisted of I X I0'@vector-transduced C3L5 cellss.c. into the chest of C3H/HeN mice. Animals were followed for 4 weeks for evidence oftumor growth. In B. a portion of tumor-free mice received a secondary challenge with5 x lO@untransduced C3L5 cells into the contralateral chest. Treatment groups consistedof untransduced cells (C3LS) or C3L5 cells transduced with the GlNa vector (Neo), theGlil2SvNa vector (IL-i), the GlNaSvi2G vector (IL-2), the L(hFlt3-L)SN vector (hFlt3-L), or the L(muFlt3-L)SN vector (mF!t3-L). Control for the secondary challenge consistedof untransduced C3L5 cells in naive mice. The numbers presented along the X-axisindicate the number of animals in each group.

studies used mice immunized with either the soluble or membrane

isoforms of murine F1t3-L. Spleens were obtained from tumor-free

animals 4 weeks after the secondary challenge with untransduced

C3L5 cells. As shown in Table 3, splenocytes and enriched T-cellpopulations provide protection in 60% of animals (P = 0.01). Asshown in Table 3, CD4-enriched splenocytes did not appear to confersignificant immunity (P = 1.0).

DISCUSSION

F1t3-L is a recently identified cytokine that demonstrates stimulatory activity for myeloid and lymphoid lineages (7—9).The recentlyidentified dendritic cell and B-cell growth-promoting activity ofFbt3-L prompted us to evaluate the activity of Flt3-L as an antitumoragent. Based on our interest in breast cancer (21, 22), we used themurine breast cancer model C3L5 and found that i.p.-administeredFlt3-L prevented tumor formation in the majority of animals. Despitethe antitumor activity observed after the primary tumor challenge,systemic administration of F1t3-L did not provide long-basting immu

nity at the dose schedule evaluated.Expression of Flt3-L by murine breast cancer cells did not affect

their in vitro growth but did prevent tumor formation in vivo. In

80 -

60-

20 -

@—.----,-@- - -@-@-

Control Neo

Table 2 Tumor growth in C3LS cells expressing the soluble (Er6) and membranebound (5H)forins of murine Flt3-L

3514

on March 6, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table 3 Adoptive transfer ofimmunity in mice challenged with C3L5 cells

Data indicated the number of mice developing tumor/total number of mice challenged.Mice were infused with spleen cells from previously immunized mice to assess adoptivetransferof immunity.TheimmunizingagentswereC3L5cellstransducedwitheithertheL(5H)SN vector encoding a membrane-bound isoform or the L(Ex6)SN encoding asoluble isoform of murine F1t3-L.Spleen cells were obtained from tumor-free animals 4weeks after a secondary challenge of C3L5 cells (see Table 1). Control splenocytes wereharvested from animals who were previously untreated. Naive mice did not receivesplenocytes. In the initial experiment, 5 X 106 splenocytes were iv. infused into naivemice. In a second experiment, 2 X l0@spleen T cells or 2 X l0@spleen CD4 cells wereinfused. At the time of splenocyte infusion, mice were challenged with 1 X l0@ untrans

duced C3L5.

ANTITUMOR @crivrry OF F1t3-LIGANDIN BREAST CANCER

Spleen sourceSplenocyte protectionTotalT-cell

protectionCD4cell

protectionC3L5IL(5H)SN2/52/54/5C3L5/L(Ex6)SN2/52/54/5Control

splenocytes5/55/55/5Naivemice3/33/3

5 producesa hydrophobicdomainthatdoesnotundergoproteolyticcleavage and was used to generate the L(5H)SN vector. Alternatesplicing of exon 6 beads to a premature stop codon before the transmembrane region and was used to generate the L(Ex6)SN vector.Both the soluble and membrane-bound forms of F1t3-L are biobogically active, but the relative biological significance of either isoformis unknown (4). A wide variety of hematopoietic cell lines express the

membrane-bound form of F1t3-L (28, 29), but other cells, includingC3L5 cells, do not express detectable amounts of membrane-boundFlt3-L. We observed that either F1t3-L isoform introduced into C3L5cells prevents tumor growth and confers long-lasting immunity thatcan be adoptively transferred to naive mice. The C3L5 dose requiredfor tumor formation and the rate of in vivo tumor growth are alsosimilar. These findings suggest that both isoforms possess equivalentantitumor properties in this model.

The precise mechanisms by which Flt3-L-expressing tumor cellsconfer tumor immunization are unknown. Immunizing properties havebeen described for a wide variety of cytokines (23—27),and thebreadth in structure and function of these cytokines suggests that avariety of pathways may lead to tumor immunization. Although the

immunizing properties of Flt3-L may be a direct effect on T bymphocytes, its superiority to IL-2 in our study suggests that other mechanisms, such as an in vivo enhancement of antigen presentation bydendritic cells, may be responsible for the development of T cellmediated immunity. The bow bevel of class I and class II MHCexpression on C3L5 cells4 may require the recruitment of accessorycells for tumor recognition, a requirement that could explain the lack

of in vitro cytotoxicity noted in our work but successful tumorrejection by adoptive transfer of splenocytes. Our findings are consistent with previous reports of tumor-immunizing activity in cytokines, the physiological activity of which has been ascribed to nonT-cell hematopoietic lineages (e.g., IL-6 and GM-CSF; Refs. 23 and26). In particular, GM-CSF has been found to be superior to a widevariety of cytokines in the B16 melanoma model, including IL-2 (23).

Because increasing concentrations of C3L5/Flt3-L cells lead to tumorformation, evaluation of the infiltrating cells at the sight of the tumorchallenge may also provide an insight into the local effect of Flt3-L in

this model. The B-cell growth properties of F1t3-L warrant evaluation

of humoral immunity. Beyond its potential as a tumor vaccine, wehope that C3L5IF1t3-L may prove useful in determining the immunological properties of Flt3-L.

4 K. Corneua, unpublished data.

3515

ACKNOWLEDGMENTS

We thank Dr. Yawen Chaing at Genetic Therapy, Inc. for supplying anumber of retroviral vectors used in this work, Mark Hanna for statistical

support, and Dr. Michael Robertson for critical review of the manuscript.

REFERENCES

I. Rosnet, 0., Marchetto, S., de Lapeyriere, 0., and Bimbaum, D. Murine Flt3. a geneencoding a novel tyrosine kinase receptor of the PDGFRJCSFI R family. Oncogene,6: 1641—1650,1991.

2. Matthews, W., Jordan, C. T., Wiegand, G. W., Pardoll, 0., and Lemischka, I. R. Areceptor tyrosine kinase specific to hematopoietic stem and progenitor cell-enrichedpopulations. Cell, 65: 1143—1152, 1991.

3. Small, D., Levenstein, M., Kim, E., Carow, C., Amin, S., Rockwell, P., Witte, L.,Burrow, C., Ratajczak, M. Z., Gewirtz, A. M., and Civin, C. I. STK-l, the humanhomolog of Flk-2/Flt-3, is selectively expressed in CD34+ human bone marrow cellsand is involved in the proliferation of early progenitor/stem cells. Proc. NatI. Acad.Sci. USA, 91: 459—463, 1994.

4. Lyman, S. D., James, L., Vanden Bos, T., de Vries, P., Brasel, K., Gliniak, B.,Hollingsworth, L. T., Picha, K. S., McKenna, H. J., Splett, R. R., Fletcher, F. A.,Maraskovsky, E., Farrah, T., Foxworthe, D., Williams, D. E., and Beckmann,M. P. Molecular cloning of a ligand for the flt3/flk-2 tyrosine kinase receptor: aproliferative factor for primitive hematopoietic cells. Cell, 75: 1157—1167,1993.

5. Hannum, C., Culpepper, J., Campbell, D., McClanahan, T., Zurawski, S., Bazan, J.,Kastelein, R., Hudak, S., Wagner, J., Mattson, J., Luh, J., Duda, G., Martina. N.,Peterson, D., Menson, S., Shanafelt, A., Muench, M., Kelner, G., Namikawa, R.,Rennick, D., Roncarolo, M-G., Ziotnik, A., Rosnet, 0., Dubreuil, P., Birnbaum, D.,and Lee, F. Ligand for FLT3IFLK2 receptor tyrosine kinase regulates growth ofhaematopoietic stem cells and is encoded by variant RNAs. Nature (Lond.), 368:643—648,1994.

6. Lyman, S. D., James, L., Johnson, L., Brasel, K., de Vries, P., Escobar, S. S.,Downey, H., Splett, R. R., Beckmann, M. P., and McKenna, H. J. Cloning of thehuman homologue of the murine flt3 ligand: a growth factor for early hematopoieticprogenitor cells. Blood, 83: 2795—2801, 1994.

7. Broxmeyer, H. E., Lu, L., Cooper, S., Ruggieri, L., Li, Z. H., and Lyman, S. D. Flt3ligand stimulates/costimulates the growth of myeloid stem/progenitor cells. Exp.Hematol., 23: 1121—1129,1995.

8. Hirayama, F., Lyman. S., Clark, S. C., and Ogawa. M. The flt3 ligand supportsproliferation of lymphohematopoietic progenitors and early B-lymphoid progenitors.Blood,85: 1762—1768,1995.

9. Jacobsen, S. E. W., Okkenhaug, C., Myklebust, J., Veiby, 0. P., and Lyman, S. D.The FLT3 ligand potently and directly stimulates the growth and expansion ofprimitive murine bone marrow progenitor cells in vitro: synergistic interactions withinterleukin (IL) 11, IL-12, and other hematopoietic growth factors. J. Exp. Med., 181:1357—1363,1995.

10. Maraskovsky, E., Brasel, K.. Teepe, M., Roux, E. R., Lyman, S. D., Shortman, K.,and McKenna, H. J. Dramatic increase in the numbers of functionally maturedendritic cells in flt3 ligand-treated mice: multiple dendritic cell subpopulationsidentified. J. Exp. Med., 184: 1953—1962,1996.

11. Khoo, N. K. S., Chan, F. P. H., Saarloos, M. N., and I..ala. P. K. Immunotherapy ofmammary adenocarcinoma metastases in C3HIHeN mice with chronic administrationof cyclo-oxygenase inhibitors alone or in combination with IL-2. Clin. & Exp.Metastasis, 10: 239-252, 1992.

12. Saarloos, M., Khoo, N. S., and Lala, P. Effects of cancer immunotherapy withindomethacin and interleukin-2 on murine hemopoietic stem cells. Cancer Res., 52:6452—6462, 1992.

13. Markowitz, D., Goff, S., and Bank, A. A safe packaging line for gene transfer:separating viral genes on two different plasmids. J. Virol., 62: 1120—1124,1988.

14. Miller, A. D., and Buttimore, C. Redesign of retrovirus packaging cell lines to avoidrecombination leading to helper virus production. Mol. Cell. Biol., 6: 2895—2902,1986.

15. Brasel, K., McKenna, H. J., Momssey, P. J., Charrier, K., Morris, A. E., Le, C. C.,Williams, D. E., and Lyman, S. D. Hematologic effects of flt3 ligand in viva in mice.Blood, 88: 2004—2012,1996.

16. Lyman, S. D., Seaberg, M., Hanna, R., Zappone, J.. Brasel. K.. Abkowitz, J. L.,Prchal, J. T., Schultz, J. C., and Shahidi, N. T. Plasma/serum levels of flt3 ligand arelow in normal individuals and highly elevated in patients with Fanconi anemia andacquired aplastic anemia. Blood, 86: 4091—4096, 1995.

17. Traycoff, C. M., Srour, E. F., Dun, P., and Cometta, K. The 30/35 kDa chymotrypticfragment of fibronectin enhances retroviral-mediated gene transfer in purified chronicmyelogenous leukemia bone marrow progenitors. Leukemia (Baltimore), 11: 159—167, 1997.

18. Lyman. S. D., Stocking, K., Davison, B., Fletcher, F.. Johnson, L., and Escobar, S.Structural analysis of human and murine flt3 ligand genomic loci. Oncogene. 11:1165—1172,1995.

19. Lyman, S. D., James, L., Escobar, S., Downey, H., de Vries, P., Brasel, K., Stocking,K., Beckmann, M. P., Copeland. N. G., Cleveland, L. S., Jenkins, N. A., Belmont,J. W., and Davison, B. L. Identification of soluble and membrane-bound isoforms ofthe murine flt3 ligand generated by altemative splicing of mRNAs. Leukemia (Baltimore),9: 149—157,1995.

20. Pross, H. F., Baines, M. G., Rubin, P., and Shragge, P. Spontaneous human lympho

on March 6, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF Flt3-LIGAND IN BREAST CANCER

cyte mediated cytotoxicity against tumor target cells. IX. The quantitation of naturalkiller cell activity. J. Clin. Immunol., 1: 51—63,1981.

21. Cometta, K., Berebitsky, D., Behnia, M., Traycoff, C., Srour, E., and Sledge, G. W.A retroviral vector expressing human interferon y upregulates MHC antigen expression in human breast cancer and leukemia cell lines. Cancer Gene Ther., 1: 91—98,I994.

22. Cometta, K.. Moore, A., Johannessohn, M., and Sledge, G. W. Clonal dominancedetected in metastases but not primary tumors of retrovirally marked human breastcarcinoma injected into nude mice. Clin. Exp. Metastasis, 12: 3—12, 1994.

23. Dranoff, G., Jaffee, E., Lazenby. A.. Golumbek, P., Levitsky. H., Brose, K., Jackson,V.. Hamada, H., Pardoll, D., and Mulligan, R. C. Vaccination with irradiated tumorcells engineered to secrete murine granulocyte-macrophage colony-stimulating factorstimulates potent. specific, and long-lasting anti-tumor immunity. Proc. NatI. Acad.Sci. USA, 90: 3539—3543, 1993.

24. Gansbacher, B., Zier, K., Daniels, B., Cronin, K., Bannerji, R., and Gilboa, E.Interleukin-2 gene transfer into tumor cells abrogates tumorigenicity and inducedprotective immunity. J. Exp. Med., 172: 1217—1224,1990.

25. Gansbacher, B., Bannerji, R., Daniels, B., Zier, K., Cronin, K., and Gilboa, E.Retroviral vector-mediated y-interferon gene transfer into tumor cells generatespotent and long-lasting antitumor immunity. Cancer Res., 50: 7820—7825, 1990.

26. Porgador, A., Tzehoval, E., Katz, A., Vadai, E., Revel, M., Feldman, M., andEisenbach, L. Interleukin 6 gene transfection into Lewis lung carcinoma tumor cellssuppresses the malignant phenotype and confers immunotherapeutic competenceagainst parental metastatic cells. Cancer Res., 52: 3679—3686, 1992.

27. Golumbek, P. T., Lazenby, A. J., Levitsky, H. I., Jaffee, L. M., Karasuyama, H.,Baker, M., and Pardoll, D. M. Treatment of established renal cancer by tumor cellsengineered to secret interleukin 4. Science (Washington DC), 254: 713—716,1991.

28. Brasel, K., Escobar, S., Anderberg, R., de Vries, P., Gruss, H-J., and Lyman, S. D.Expression of the flt3 receptor and its ligand on hematopoietic cells. Leukemia(Baltimore),9: 1212—1218,1995.

29. Meierhoff, G., Dehmel, U., Gruss, H-J., Rosnet, 0., Bimbaum, D., Quentmeier, H.,Dirks, W., and Drexler, H. G. Expression of FLT3 receptor and FLT3-ligand inhuman leukemia-lymphoma cell lines. Leukemia (Baltimore), 9: 1368—1372, 1995.

3516

on March 6, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1997;57:3511-3516. Cancer Res   Keyue Chen, Stephen Braun, Stewart Lyman, et al.   Flt3-Ligand in a Murine Breast Cancer ModelAntitumor Activity and Immunotherapeutic Properties of

  Updated version

  http://cancerres.aacrjournals.org/content/57/16/3511

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/57/16/3511To request permission to re-use all or part of this article, use this link

on March 6, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from