-
Cancer Therapy: Preclinical
Effective Activity of Cytokine-Induced Killer Cells
againstAutologous Metastatic Melanoma Including Cells withStemness
Features
Loretta Gammaitoni1, Lidia Giraudo1,5, Valeria Leuci4,5, Maja
Todorovic1,5, Giulia Mesiano1,5,Franco Picciotto2, Alberto
Pisacane3, Alessandro Zaccagna2, Maria Giuseppa Volpe1, Susanna
Gallo1,5,Daniela Caravelli1, Elena Giacone2, Tiziana Venesio3,
Antonella Balsamo3, Ymera Pignochino4,5,Giovanni Grignani4,
Fabrizio Carnevale-Schianca1, Massimo Aglietta1,5, and Dario
Sangiolo1,5
AbstractPurpose:We investigate the unknown tumor-killing
activity of cytokine-induced killer (CIK) cells against
autologous metastatic melanoma and the elusive subset of
putative cancer stem cells (mCSC).
Experimental Design:We developed a preclinical autologous model
using same patient-generated CIK
cells and tumor targets to consider the unique biology of each
patient/tumor pairing. In primary tumor cell
cultures, we visualized and immunophenotypically defined a
putative mCSC subset using a novel gene
transfer strategy that exploited their exclusive ability to
activate the promoter of stemness gene Oct4.
Results: The CIK cells from 10 patients with metastatic melanoma
were successfully expanded (median,
23-fold; range, 11–117). Primary tumor cell cultures established
and characterized from the same patients
were used as autologous targets. Patient-derived CIK cells
efficiently killed autologousmetastaticmelanoma
[up to 71% specific killing (n¼ 26)]. CIK cells were active in
vivo against autologousmelanoma, resulting indelayed tumor growth,
increased necrotic areas, and lymphocyte infiltration at tumor
sites. The metastatic
melanoma cultures presented an average of 11.5% � 2.5% putative
mCSCs, which was assessed by Oct4promoter activity and
stemnessmarker expression (Oct4, ABCG2, ALDH,MITF). Expression was
confirmed
on mCSC target molecules recognized by CIK cells (MIC A/B;
ULBPs). CIK tumor killing activity against
mCSCs was intense (up to 71%, n ¼ 4) and comparable with results
reported against differentiatedmetastatic melanoma cells (P ¼
0.8).Conclusions: For the first time, the intense killing activity
of CIK cells against autologous metastatic
melanoma, including mCSCs, has been shown. These findings move
clinical investigation of a new
immunotherapy for metastatic melanoma, including mCSCs, closer.
Clin Cancer Res; 19(16); 4347–58.
�2013 AACR.
IntroductionThe incidence of malignant melanoma in western
popu-
lations has increased in recent decades. Although
surgicalresection of primary lesions has high cure rates,
metastaticmelanoma remains largely refractory to conventional
ther-apies with a dismal prognosis (1, 2). Targeted
strategiesagainst key molecules such as B-RAF and MEK (3–5),
andimmune system antitumor activity restoration to block the
CTLA4 (6, 7) or PD-1 molecules (8) are two recent break-throughs
that have positively impacted metastatic melano-ma
treatment.Despite some survival advantage, results havealmost
always been short-lived; it seems that as hypothe-sized for other
tumors, metastatic melanoma includes asmall cell subpopulation
endowed with the stemness fea-tures that sustain drug resistance
and disease relapse (9).This challenging clinical scenario demands
new therapeuticapproaches, ideally the ones that are able to target
mela-noma cancer stem cells (mCSC).
Much promise lies in adoptive immunotherapy for thetreatment of
metastatic melanoma as supported by resultsreported with ex vivo
expanded tumor infiltrating lympho-cytes (TIL; ref. 10) or T cells
engineered with melanomaantigen-specific T-cell receptors (TCR;
refs. 11, 12). Theextended clinical application of TIL is, however,
limited byavailable, suitably sized resectable tumor lesions,
whereasspecific human leukocyte antigen (HLA) restriction, com-plex
expansion conditions, and stringent regulatory require-ments
confound the TCR transfer approach.
Authors' Affiliations: Units of 1Stem Cell Transplantation and
Cell Ther-apy, 2Surgical Dermatology, 3Pathology, and 4Sarcoma,
Fondazione delPiemonte per l'Oncologia, I.R.C.C.S.; 5Department
ofOncology, Universityof Torino Medical School, Candiolo, Torino,
Italy
Note: Supplementary data for this article are available at
Clinical CancerResearch Online
(http://clincancerres.aacrjournals.org/).
Corresponding Author: Dario Sangiolo, Cancer Cell Therapy
Laboratory,I.R.C.C.S., Provinciale 142Candiolo 10060, Torino,
Italy. Phone: 390-1199-33503; Fax: 390-1199-33522; E-mail:
[email protected]
doi: 10.1158/1078-0432.CCR-13-0061
�2013 American Association for Cancer Research.
ClinicalCancer
Research
www.aacrjournals.org 4347
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
Cytokine-induced killer (CIK) cells are ex vivo
expandedT-natural killer (NK) lymphocytes potentially able
toaddress some of the issues currently limiting clinicalapplication
of other immunotherapies (13, 14). CIK cellscan be massively
expanded from peripheral blood mono-nuclear cells (PBMC) cultured
with the timed addition ofIFN-g , Ab-anti-CD3, and interleukin
(IL)-2 through sim-ple standardized culture conditions (15–18). CIK
cellactivity is not dependent on HLA restriction; it is
mediatedmostly by the interaction of its NKG2D receptor
withstress-inducible molecules [MHC class I-related chain Aand B
(MIC A/B) and UL-16–binding proteins (ULBPs)]on tumor targets (19,
20). Initial clinical trial resulting invarious tumor settings are
encouraging, but data are gen-erally absent on CIK cell potential
activity against autol-ogous metastatic melanoma (13, 14, 21, 22),
let alonework that might take into account the unique biologicand
immunogenic features of a specific tumor in a specificpatient.
Furthermore, whether or not the tumor-killing ability ofCIK
cells affects the subpopulation of mCSCs is completelyunexplored.
Currently, clear identification of mCSCs isintensely debated.
Several membrane molecules or geneshave been proposed as putative
markers, but agreementremains elusive. Quintana and colleagues
recently failed toidentify any marker that robustly distinguished
mCSCsdespite examining 85 markers, including ATP-binding cas-sette
B5 (ABCB5), CD271 (also known as nerve growthfactor receptor;
NGFR), and CD133 (23, 24).
Several facts germane to this topic are commonly accept-ed.
First, expression of human embryonic stem cell plur-ipotency
markers SOX2, Klf4, and Oct4 indicate high plas-ticity (25, 26).
Second, inhibition of microphthalmia-associated transcription
factor (Mitf), the master regulatorof melanocyte differentiation,
increases the tumorigenicpotential of melanoma cells and
upregulates stem cellmarkers Oct4 and Nanog as shown by Cheli and
colleagues(27). Third and recently reposted is that forced
expressionofthe Oct4 gene promoted dedifferentiation of
melanomacells toward mCSCs with decreased expression of melano-
cytic markers, acquisition of multipotent
differentiationcapacity, membrane expression of ABCB5 and
CD271,resistance to chemotherapy, hypoxia, and increased
tumor-igenic capacity (28).
Here, we first report preclinical activity of patient-derivedCIK
cells against autologous metastatic melanoma includ-ing putative
mCSCs. To highlight the mCSCs, we intro-duced a gene transfer
strategy whereby bulk melanoma cellcultures were transduced with a
lentiviral vector encodingthe enhanced eGFP under expression
control of the Oct4promoter. We visualized mCSCs exploiting their
exclusiveability to activate theOct4 promoter and sorted themon
thebasis of eGFP expression. CIK cells efficiently killed
autol-ogous melanoma targets regardless of their stemness
ordifferentiated phenotype.
Materials and MethodsCIK culture and expansion
Human peripheral blood samples were obtained fromsubjectswith
histologically confirmed stage IVmelanoma atthe Fondazione del
Piemonte per l’Oncologia-–Institute forCancer Research and
Treatment (FPO-IRCC; Candiolo,Torino, Italy). All individuals
provided their informedconsent.
Cultures were started with PBMCs collected from samepatients
with metastatic melanoma and isolated by densitygradient
centrifugation using cell separation media Lym-phoprep
(SentinelDiagnostic). For some experiments, freshPBMCs collected
from volunteer donors were used. PBMCswere cultured overnight in
cell culture flasks at a cell densityof 1.5 � 106/mL RPMI (Gibco
BRL) supplemented with10% FBS (Sigma) in 1,000 U/mL IFN-g
(PeproTech). After18 to 24 hours in culture at 37�C and 5% CO2, 50
ng/mLanti-CD3 antibody (OKT3, PharMingen) and 300 U/mLrecombinant
human IL-2 (Proleukin, Aldesleukin, ChironCorporation) were added.
Fresh medium with IL-2 wasadded as needed.
Establishment of primary melanoma cell culturesHuman melanoma
tissues were obtained from surgical
specimens; patients provided consent under institutionalreview
board–approved protocols. Human melanoma tis-sues were cut into 3
mm3 pieces and processed for cellisolation. Tumor tissue was
processed by mechanical andenzymatic dissociation (Collagenase Type
I, Invitrogen) for3 hours and then subsequently for an additional
12 hours.Cells were then resuspended in KnockOut
Dulbecco’smod-ified Eagle medium: nutrient mixture F-12 medium
(KODMEM:F12 medium Gibco BRL) with the addition ofpenicillin (50
U/mL), streptomycin (50 mg/mL), Glutamax100X (all from Gibco BRL);
cells could be seeded either inserum-free conditions with the
addition of N2 supplement(Gibco BRL) or in 10% heat-inactivated FBS
(Euroclone).Cells were plated at clonal density (104–105
cells/cm2)either in ultra-low attachment multi-well plates for
suspen-sion cell culture applications or in multi-well plates
treatedfor anchorage-dependent cultures (Corning/Costar).
Translational RelevanceThis work is the first report of
effective antitumor
activity of patient-derived cytokine-induced killer (CIK)cells
against autologousmetastaticmelanoma, includingputative cancer stem
cells (CSC). The translational rel-evance of this work is borne out
in two ways. First, theautologous model (same patient CIK cells and
tumortargets) uniquely considers the intrinsic biology ofeach
patient—tumor, and second, chemoresistant- andrelapse-associated
putative melanoma CSCs are identi-fied and targeted. The ability of
CIK cells to kill bothdifferentiated melanoma cells and CSCs may
promotethem into clinical trials, either alone or in
associationwith emerging molecular targeted therapies.
Gammaitoni et al.
Clin Cancer Res; 19(16) August 15, 2013 Clinical Cancer
Research4348
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
In vivo tumorigenesisFour-week-old nonobese
diabetic/LtSz-scid/scid [nonob-
ese diabetic/severe combined immunodeficiency (NOD/SCID)]
(Charles River) female mice were injected subcuta-neously with 106
cells from melanoma primary culturesresuspended in sterile PBS1X
and BD Matrigel BasementMembrane Matrix (Becton Dickinson), 1:1.
Tumor growthwasmonitored weekly with calipers; volumewas
calculatedusing formula, V ¼ 4/3 � p � (l/2)2 � (L/2), where L
isthe length and l the width diameter of the tumor. When thetumor
volume reached 2 cm in diameter, the animal waseuthanized, the
tumor was recovered and fixed overnight in4% paraformaldehyde, then
dehydrated, paraffin-embed-ded, sectioned (5 mm), and stained with
hematoxylin andeosin (H&E; Bio.Optica).
hOct4.eGFP lentiviral vector generationVSV-G pseudotyped
third-generation lentiviral vectors
were produced by transient four-plasmid cotransfectioninto 293T
cells as described by Follenzi and colleagues(29). The transfer
vector pRRL.sin.PPT.hPGK.EGFP.Wpre(LV-PGK.EGFP) was kindly provided
by Dr. Elisa Vigna(Gene Transfer and Therapy, IRCC Candiolo,
Torino, Italy)and has been described elsewhere (29).The
phOCT4.EGFP1 vector (30)was provided byWei Cui
(IRDB, Imperial College London, London, United King-dom). The
pRRL.sin.PPT.hOct4.eGFP.Wpre (LV-Oct4.eGFP) was obtained by
replacing the expression cassettehPGK.eGFP into LV-PGK.eGFP with
the hOct4-eGFP1cleaved from the phOct4-eGFP1 vector by insertion
intothe SalI and XhoI restriction enzyme sites. Physical titers
forlentiviral vector stocks were determined on the basis of
p24antigen content (HIV-1 p24 ELISA kit; PerkinElmer).
Melanoma primary cell transductionFor each lentiviral vector
transduction, melanoma pri-
mary cells were resuspended in fresh KODMEM-F12 with10%FBS.
Virus-conditionedmediumwas added at a dose of400 ng P24/100,000
cells. After 16 hours, cells were washedtwice and grown for a
minimum of 10 days to reach steady-state eGFP expression and to
rule out pseudotransductionbefore flow cytometry analysis. As a
transduction efficiencycontrol, the same melanoma primary cells
were transducedwith LV-PGK.eGFP. Murine embryonic cells and
PBMCswere transduced with LV-Oct4.eGFP as positive and nega-tive
expression controls.
Analysis of LV-Oct4.eGFP presence in eGFP-positiveand -negative
cell fractionsDetection of the presence of LV-Oct4.eGFP in both
frac-
tions of freshly sorted cells was verified by
PCR-basedamplification of the expression cassette Oct4.eGFP.
Geno-mic DNA was extracted separately from eGFPþ and eGFP�
cells using a commercial kit (Qiagen). PCR was conductedusing
100 ng of gDNA per sample and Phusion High-Fidelity DNA Polymerase
(Thermo Scientific) according tothe manufacturer’s protocol.
Annealed on the lentiviral vector backbone sequencesboth
upstream and downstream the expression cassettewere primers LV
forward primer 50-AGG CCCGAAGGAA-TAGAAGA-30 and LV reverse primer
50-CCACATAGCG-TAAAAGGAGCA -30.
The PCR products were separated by electrophoresis on1% agarose
gel.
In vitro proliferation assay and PKH26 stainingTo evaluate the
proliferation rate of eGFPþ versus eGFP�
cell sorted fractions, cells had been labeled with lipophylicdye
PKH26, for which fluorescence intensity decreased byhalf at each
cell division per kit protocol (PKH26GL kit,Sigma). Briefly, an
adequate quantity of Diluent C labelingvehicle was added to the
previously washed cell pellet (es.1� 106 cells/0.5 mL) to obtain a
2� single-cell suspension.A 2� (4 mmol/L) PKH26dye solution was
prepared andadded to 2� single-cell suspension. Cell membrane
dyeuptake was stopped by adding an equal volume of heat-inactivated
serum. PKH26-stained cells were then washedtwice with culture
medium added with 10% heat-inacti-vated serum (5 minutes at 400 �
g). An aliquot of labeledand counted cells was read on a Flow
Cytometry Cyan(Cyan ADP, Dako) and analyzed using Summit Softwareto
set the baseline fluorescence level. The remaining cellswere seeded
in culture under optimal conditions as previ-ously described. After
7, 14, and 21 days, the reduction influorescence was quantified by
flow cytometry.
Cytotoxicity assayCIK tumor-killing ability was assessed against
an alloge-
neic cell line (DettMel) and melanoma primary tumor cells.The
allogeneic melanoma cell line DettMel was a kind giftfrom Dr V.
Russo (Cancer Gene Therapy Unit, ScientificInstitute S. Raffaele,
Milan, Italy), derived from metastasesof malignant melanomas as
described elsewhere (31). Theeffector cells were assayed against
both autologous andallogeneic tumor targets (32). A nonradioactive
stain ofmelanoma target cells was used from PKH26 kit
(Sigma-Aldrich) to conduct in vitro assays. Their
immune-mediatedkillingwas analyzedbyflowcytometry
(CyanADP,Dako)bypropidiumiodidepermeability of target cells (PKH26þ
gate).CIK cells were coculturedwith either autologous or
allogenicmelanoma primary cells with a 40:1, 20:1, 10:1, and
5:1effector:target ratio for 2 to6hours in200mLofmediumwithIL-2at a
concentrationof300U/mLat37�C/5%CO2. Tumorcells, in the absence of
CIK cells, were used as a control toassess spontaneous mortality.
The percentage of tumor-spe-cific lysis for each effector/target
ratio was calculated accord-ing to the following formula:
(experimental�spontaneousmortality/100�spontaneous mortality) �
100.
In vivo activity assayNOD/SCID (Charles River) female mice were
subcuta-
neously injected with an 8 mm3 tumor fragment from
apatient-derived melanoma biopsy (mMel2). Starting 1week after
tumor implantation, mice received 8 weeklyintravenous infusions
with 1� 107 mature autologous CIK
CIK Cells Kill Autologous Melanoma and mCSCs
www.aacrjournals.org Clin Cancer Res; 19(16) August 15, 2013
4349
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
cells resuspended in 1� PBS (200 mL total volume injected).Mice
injected with PBS alone were used as the untreatedcontrol. Tumor
growth was monitored weekly with calipersand volume calculated
according to the formula: V¼ 4/3�p � (l/2)2 � (L/2), where L is the
length and l the widthdiameter of the tumor. Animals were
euthanized when thetumor size reached 2 cm in itsmain diameter. The
recoveredtumor was fixed overnight in 4% paraformaldehyde,
dehy-drated, paraffin-embedded, sectioned (5 mm), and
finallystained with H&E (Bio.Optica). Immunohistochemicalassay
was conducted with human anti-CD5 and anti-CD56(NCMA) antibodies
(Novocastra, Leica Biosystem).A certified pathologist evaluated the
necrotic areas.
Statistical analysisStatistical analysis was conducted using
software Graph-
Pad Prism 5. A descriptive statistical analysis of CIK
andmelanoma primary cell culture median values and ranges,ormean�
SEM, was used as required. Subgroup phenotypeand necrotic area
extension were compared with theunpaired, two-tailed t test. The
mixed model ANOVA wasused to assess CIK cytotoxic activity curves
in vitro and tocompare tumor volumes in vivo. Statistical
significance hasbeen expressed as true P value and all less than
0.05 wereconsidered statistically significant.
ResultsEstablishment and characterization of autologousmelanoma
primary cell cultures
Autologous melanoma cell lines were successfully estab-lished
from 10 patients with stage IV melanoma. Thecharacteristics of the
10 patients are shown in Table 1.
Primary cell cultures were generated from tumor tissuebiopsies
onmetastatic sites in 4 to 12weeks. All cell culturesdisplayed
morphologic features consistent with the pathol-ogy evaluation of
the corresponding tumor; a representativepicture of the primary
tumor cell cultures is shown inSupplementary Fig. S1.
We analyzed each cell culture for expression of previous-ly
described main melanoma surface antigens: CD271,melanoma-associated
chondroitin sulfate proteoglycan(MCSP), CD34, and VEGF receptor 1
(VEGFR1). All testedtumors (Table 2) showed heterogeneous
expression ofCD271 and MCSP at medians of 55% (range: 15%–93%)and
78% (range: 2%–99%), respectively (SupplementaryFig. S2A and S2B).
CD34þwas expressed at a median of 7%(range: 2%–25%) and
VEGFR1bright at 10% (range: 2%–19%), whereas CD34 and VEGFR1 were
coexpressed in 4%of the cells (range: 2%–21%; Supplementary Fig.
S2E).mMel6 and mMel10 were the only two melanoma cellcultures found
to express CD117 (also known as c-kit) at89% and 70%, respectively
(Supplementary Fig. S2K).
Some of the principal molecules reported as
associatedwithmCSCphenotypewere also evaluated.Oct4,ATP-bind-ing
cassette G2 (ABCG2), and aldehyde dehydrogenase(ALDH) were detected
in all samples that averaged expres-sion of 10.9% � 0.9%, 11.1% �
1.4%, and 9.7% � 2.3%,respectively (Supplementary Fig.
S2F–S2H).Melanomacellsnegative for Mitf expression were 10.8% �
1.5% (Supple-mentary Fig. S2I). All data are expressed as mean �
SEM.
Each melanoma cell culture was analyzed for expressionof the
principal known ligands recognized by the NKG2Dreceptor on CIK
cells (MIC A/B, ULBP1, ULBP2, andULBP3). MIC A/B and ULBP2 were
found expressed in
Table 1. Main characteristics of melanoma patients and
corresponding samples
Subject number Age/sex Status Lesion siteaPrimary
cellcultureb
CIK cellexpansionc
Autologouscytotoxicity assayd
mMel1 67/M PD SC Ye 115 YmMel2 64/F PD SC Ye,f 133 YmMel3 69/F
PD SC Ye,f 419 YmMel4 54/M PD LN Ye 314 YmMel5 67/M PD LN Ye,f 1870
YmMel6 78/M PD LN Ye,f 109 YmMel7 79/F PD LN Y 125 YmMel8 87/F PD
SC Y 1387 YmMel9 74/F PD LN Y 49 YmMel10 67/M PD LN Y 195 Y
Abbreviations: Y, yes; mMel, metastatic melanoma.aBiopsy was
conducted at different metastatic sites including subcutaneous (SC)
or lymph node (LN) sites.bIn vitro primary cell cultures derived
from tumor biopsies.cCD3þ/CD56þ cell number-fold increase after 3
weeks of expansion.dCytotoxicity assay with CIK cells versus
autologous tumor target cells.emMel samples assessed for
tumorigenic capacity in vivo in NOD/SCIDmice. The samemMel
sampleswere transducedwith LV-Oct4.eGFP to visualize mCSC and
subsequently were sorted on the basis of eGFP expression.fmMel
samples used for cytotoxicity assay with CIK cells versus
LV-Oct4-eGFP–transduced target cells sorted on the basis of
eGFPexpression.
Gammaitoni et al.
Clin Cancer Res; 19(16) August 15, 2013 Clinical Cancer
Research4350
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
every tumor tested to varying levels with medians of 24%(range:
7%–90%) and 61% (range: 40%–98%), respective-ly. ULBP 1 and 3
expression was negligible except inmMel10 (16.57% ULBP3þ cells;
Supplementary Fig. S2Cand S2D).All melanoma cell cultures, except
two (mMel9 and
mMel10), retained membrane expression of HLA class Imolecules
(>99%HLA-ABCþ).Melanoma cell cultures gen-erated from selected
representative patients (n ¼ 6) weresubcutaneously inoculated in
NOD/SCID mice to showtheir tumorigenicity. All mice inoculated with
6 differentmelanoma primary cell cultures produced tumor
growthstarting between 1 and 8 weeks after injection (Fig. 1).
Expansion and phenotype of CIK cellsThe ex vivo expansion of CIK
cells was evaluated in the
same 10 patients from whom we had generated melanomaprimary
cultures. CIK cells were classically expanded fromfresh or
cryopreserved PBMCs cultured with the timedaddition of IFN-g ,
Ab-anti-CD3, and IL-2. CIK cells fromall patients were successfully
expanded within 3 to 4 weeksof culture.Median expansion of bulk CIK
cells was 23-fold (range
11–117-fold) after 3 weeks of culture, whereas
252-foldexpansionwasobtained for theCD3þCD56þ
fraction(range49–1870-fold). The presence of pure NK
(CD3�CD56þ)cells was negligible at less then 5% in all cases at the
end ofexpansion. The subset of mature CIK cells coexpressingCD3 and
CD56molecules (CD3þCD56þ) was present witha median of 49% (range
23%–80%) while 78% (59–91%)of CD3þ cells were also CD8þ
(Supplementary Fig. S3).The median membrane expression of the NKG2D
recep-
tor, which is the main molecule responsible for
tumorrecognition, on expanded CD3þ CIK cells was 84%
(range:57%–93%).A summary of patient characteristics and relative
CIK
expansion data is reported in Table 3.
Killing activity of CIK cells against melanoma cell lineTo test
the antitumor activity of CIK cells expanded from
the 10 patients, we evaluated their ability to kill in vitro
amelanoma cell line (DettMel). The cytotoxicity test was
conducted at the end of ex vivo expansion and showedefficient
killing that varied among patients. The averagespecific tumor
killing was 63% � 4%, 52% � 5%, 39%� 5%, and 28% � 5% (mean � SEM)
at 40:1, 20:1, 10:1,and 5:1 effector/target ratio, respectively (n
¼ 18, Fig. 2Aand B).
In vitro and in vivo killing activity of CIK cells
againstautologous metastatic melanoma cells
Patient-derived CIK cells efficiently killed in vitro
autol-ogous metastatic melanoma targets with an average
specifickilling of 71%�2%, 61%� 3%, 49%� 3%, and37%� 3%(mean � SEM)
at a 40:1, 20:1, 10:1, and 5:1 effector/targetratio, respectively
(n ¼ 26). The intensity of killing againstautologous targets was
comparable (P ¼ 0.9991) with thatobservedwith allogeneicCIK cells
assessed inparallel versusthe same tumor cells with an average
specific killing of 70%� 4%, 61� 4%, 49%� 5%, and 35%� 4% (mean�
SEM)at a 40:1, 20:1, 10:1, and 5:1 effector/target ratio,
respec-tively (n ¼ 20). A summary of cytotoxicity in vitro
againstautologous or allogeneic tumor targets is reported in Fig.
2A.
We evaluated also the activity of patient-derived CIK cellsin
vivo against autologous metastatic melanoma. NOD/SCID mice (n ¼ 12)
were subcutaneously implanted withan 8 mm3 tumor fragment from a
patient-derived melano-ma biopsy (mMel2). One week after tumor
implantation, agroup of implanted mice (n ¼ 8) were infused
weeklyby tail vein injection withmature autologous CIK cells
(1�107/week for 6 weeks). When tumor growth in untreatedmice (n¼ 4)
wasmore than 2 cm in at least one dimension,all animals were
euthanized. Tumors were excised andanalyzed for the presence of
lymphocytic infiltration andthe extension of necrotic tissue areas.
At the end of theexperiment, tumors fromanimals treatedwith CIK
cells hadsignificantly larger necrotic areas compared with
untreatedcontrols (Fig. 2B; P ¼ 0.0255), and we could confirmthe
presence of CIK cells infiltrating the autologous tumor(Fig.
2C).Moreover, a significant delay in the tumor growthcurve was
observed in treated mice compared with untreat-ed controls (P ¼
0.0305) after two-way ANOVA analysis(Fig. 2D).
Table 2. Phenotype characterization of primary melanoma cell
cultures
Subjectnumber
% MICA/B
%ULBP1
%ULBP2
%ULBP3
%NGFR
%MCSP
%Oct4
% ALDHbright
%ABCG2
%Mitf-
% VEGFR1bright
%CD34
%CD117
%HLA-ABC
mMel1 33.7 0 62.8 0 74.2 64.7 15.5 8.3 11.8 16.8 19.3 7.6 neg
99.9mMel2 15.2 1.6 40.7 1.9 58.3 71.4 11.5 8.7 12.4 16.9 10.0 18.0
neg 99.5mMel3 90.1 1.6 97.9 0 93.0 81.5 8.3 28.9 12.2 14 12.5 25.1
neg 99.6mMel4 64.9 0 60.6 7.4 19.9 95.0 10.1 8.2 1.3 10.7 10.6 6.3
neg 99.2mMel5 7.3 1.1 40.3 1.3 82.4 88.9 10.7 11.6 7.3 6.8 12.3
12.3 neg 99.7mMel6 32.4 0 55.0 0 50.8 95.1 7.3 2.8 11.2 6.5 10.7
6.4 88.8 99.1mMel7 20.5 0 64.5 0 39.1 78 10.7 8.9 14.1 9.8 5.0 3.1
30.1 99.4mMel8 53.5 0 52.7 0 28.9 74.2 11.4 3.3 17.5 15 4.4 3.1 neg
99.6mMel9 7.4 0 65.5 0 82.3 99.7 8.3 6.0 10.2 3.4 2.7 2.1 neg
negmMel10 14.2 1.17 62.1 16.6 33.6 73.6 15.2 9.9 13 8.3 5.6 7.0
70.3 neg
CIK Cells Kill Autologous Melanoma and mCSCs
www.aacrjournals.org Clin Cancer Res; 19(16) August 15, 2013
4351
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
Activity ofCIK cells against
autologousputativemCSCsVisualization of putative mCSCs was
accomplished by
stably transducing the primary melanoma cell culturesusing a
lentiviral vector that carried eGFP controlled by thepromoter
regulatory element of the Oct4 gene (LV-Oct4.eGFP; Fig. 3A and B).
The average eGFP expression, 7 daysafter transduction, was 11.5%�
2.5%. As a positive control,a murine embryonic cell line expressing
Oct4 (mES) wassuccessfully transduced with LV-Oct4.eGFP up to 90.5%
ofeGFP expression (Fig. 3B), whereas no eGFP expression wasdetected
on differentiated PBMCs from healthy donorstransduced with the same
vector (LV-Oct4.eGFP). As anadditional control, we confirmed that
both primary mela-noma cell cultures (Fig. 3B) and mES could be
transducedefficiently (>90% of eGFP expression; data not
shown)when the strong ubiquitous promoter (Phospho GlyceratoKinase,
PGK, regulatory element) was used in place of theOct4 promoter to
control eGFP expression (Fig. 3B).
On the basis of eGFP expression, transduced melanomacells were
sorted into two fractions (eGFPþ and eGFP�) thatserved as targets
to assess separately the antitumor activity ofpatient-derived CIK
cells against their own putative mCSCs
(eGFPþ) and bulk eGFP� melanoma cells (SupplementaryFig.
S4A–S4D). As further evidence of stem cell enrichmentwithin the
eGFPþ fraction, we transduced bulk primarymelanoma cells with
LV-Oct4.eGFP vector and culturedthem in anchorage-independent and
serum-free conditionsat low cell concentration (104 cells/cm2) to
observe theformation of mainly fluorescent (eGFPþ) spheroids
(Sup-plementary Fig. S4E).
The integration of LV-Oct4.eGFP was confirmed by PCRin both
eGFPþ and eGFP�melanoma cell subsets (Fig. 3A).Additional evidence
that the eGFPþmelanoma cell fractionwas enriched in putative mCSCs,
LV-Oct4.eGFP–trans-duced cells were evaluated on the basis of ABCG2
expres-sion. The average cell percentage expressing ABCG2 was10.6%
� 2.3%, 72.4% of which coexpressed eGFP (Sup-plementary Figs. S5
and S6A).
We then measured the distribution of the main NKG2Dligands (MIC
A/B, ULBP2) expressed in the target cells foreGFP expression. The
percentage of MIC A/Bþ or ULBP2þ
cells were equally represented in eGFPþ and eGFP� frac-tions
without a statistically significant difference (P ¼0.5181 and
1.000, respectively; Supplementary Fig. S6).
A mMel3mMel2mMel1
mMel6mMel5mMel4
Weeks after tumor cell inoculation (1 ×× 106)
Tu
mo
r vo
lum
e (c
m3 )
1W2 W3 W4 W5 W6 W7 W8 W9 W10 W
11 W
12 W
13 W
0
1
2
3
4
5mMel1mMel2
mMel3
mMel4
mMel5
mMel6
Melanoma primary cell growth curves in NOD/SCID mice B
Figure 1. Primary melanomacultures successfully generatetumor
xenografts in vivo. A,primary metastatic melanoma cellcultures were
successfully provedfor their tumorigenic activity in
vivo.Representative H&E-stainedparaffin-embedded sections
fromtumor generated aftersubcutaneous injection in NOD/SCID mice (1
� 106 metastaticmelanoma cells): mMel1, mMel2,mMel3, mMel4, mMel5,
mMel6.Tumor xenograft morphology wasconsistent with the original
humantumor as confirmed by a pathologyreview.Scalebars, 12mm.B,
tumorgrowth curves of the melanomacell lines transplanted into
theNOD/SCID mice.
Gammaitoni et al.
Clin Cancer Res; 19(16) August 15, 2013 Clinical Cancer
Research4352
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
Through a functional assay, we evaluated the prolifera-tion
ability of putative mCSCs. Melanoma cells transducedwith
LV-Oct4.eGFP were stained with PKH26 dye to distin-guish the rare
quiescent/slowly dividing cells of putativeCSCs from themore
differentiated fast-growing population.Cell fluorescence was
acquired initially and weekly for 3weeks. Putative eGFPþ mCSCs
displayed a proliferativepotential in vitro that was, on average,
three times less thantheir eGFP� counterparts after 21 days of
culture (n ¼ 5),showing a slow-growing phenotype typical of CSCs;
arepresentative histogram is reported in Fig. 3C.Patient-derived
CIK cells efficiently killed in vitro autol-
ogous eGFPþ melanoma cells with an average specifickilling of
71% � 5%, 56% � 7%, 44% � 7%, and 40%� 6% (mean� SEM) at a 40:1,
20:1, 10:1, and 5:1 effector/target ratio, respectively (n¼ 4).
Comparable (P¼ 0.8224)tumor killing intensity was reported against
autologouseGFP� targets with an average specific killing of 66%
�6%, 54%� 8%, 44%� 10%, and 36%� 8% at a 40:1, 20:1,10:1, and 5:1
effector/target ratio, respectively (n ¼ 4). Thekilling activity of
CIK cells remained equally effectiveagainst both eGFPþ and eGFP�
autologous melanoma cells(n¼ 5; P¼ 0.6286) even when the assay was
conducted ontotal tumor cells without preemptive sorting of
putativeeGFPþ mCSCs. A summary of cytotoxicity against autolo-gous
tumor targets is reported in Fig. 3D.
DiscussionThe present work addresses two main issues
regarding
the activity of immunotherapy in preclinical models:first, to
show tumor-killing activity towards autologoussolid tumor cells;
second, to show effective killing ofautologous putative cancer stem
cells hitting one of thereservoirs responsible for tumor resistance
to standardtreatments.
For the first time, we report the strong preclinical activityof
patient-derived CIK cells against autologous metastaticmelanoma,
with insight on their potential to target putativemCSCs. CIK cells
represent promise for cancer adoptiveimmunotherapy and carry
biologic features that comparepositively with other immunotherapies
for reliable andeffective clinical translation.
The intense expansibility is the first of such features (13).Our
data confirmed that CIK cells from patients withmetastatic melanoma
were expanded at clinically relevantlevels. The simplicity and
relative low expense of the expan-sion protocol, already validated
in good manufacturingpractice–controlled conditions (33, 34), may
positivelyimpact the clinical perspective. Recent clinical trials
withCIK cells have reported encouraging observations in
chal-lenging settings like metastatic lung, renal, and
gastrointes-tinal tumors (13, 21, 35–37). However, such trials
lackformal demonstration of CIK cell antitumor activity
againstautologous metastatic targets; indeed, no data are
currentlyavailable for metastatic melanoma.
For 10 different patients, CIK cells efficiently killed
autol-ogous metastatic melanoma cells. Such findings are bothnovel
and potentially clinically relevant. They overcomeimportant
limitations linked to the use of commerciallyavailable allogeneic
tumor cell lines, avoiding confoundingresults based on alloreactive
HLA mismatches and allowedfull appreciation of the unique biology
of each tumor.Furthermore, within each patient, it is possible to
hypoth-esize the existence of additional and important
biologicdifferences between metastases and the primitive
tumor,supporting the importance of the observed killing of CIKcells
against autologous metastatic melanoma.
Our in vitro cytotoxicity assays were evaluated conserva-tively,
within a 6-hour experimental timeframe, to favorkilling specificity
of the extremely delicate autologoustumor targets. Results are
indicative of CIK killing capacity
Table 3. Expansion rate and phenotype characterization of CIK
cells
Subjectnumber
FIcellsa
%iCD3b
%fCD3c FId CD3
%iCD3/CD56b
%fCD3/CD56c FId CD3/CD56
%iCD3/CD8b
%fCD3/CD8c
%iCD3/NKG2Db
%fCD3/NKG2Dc
mMel1 11 69 99 16 6 60 115 25 91 10 93mMel2 21 74 99 28 4 25 133
20 69 10 74mMel3 41 60 99 67 4 41 419 15 73 17 85mMel4 55 93 99 58
4 23 314 20 82 20 57mMel5 117 73 99 159 5 80 1870 11 76 14 91mMel6
11 50 94 20 5 50 109 17 80 11 91mMel7 22 68 100 32 10 58 125 37 69
37 79mMel8 27 49 95 53 1 51 1387 4 59 16 83mMel9 14 76 99 18 14 48
49 59 80 50 93mMel10 13 57 98 22 3 45 195 15 79 14 79
aFold increase [FI ¼ (cell number T ¼ week 3)/(cell number T ¼
0)] of total cell number after 3 weeks of expansion.bPercentage of
cells expressing different surface antigens at the basal time (T ¼
0).c Percentage of cells expressing different surface antigens
after 3-week expansion.dFold increase [FI = (cell number T ¼ week
3)/(cell number T ¼ 0)] of absolute cell number after 3 weeks of
expansion calculated for everysubpopulation of cells expressing
different surface antigens.
CIK Cells Kill Autologous Melanoma and mCSCs
www.aacrjournals.org Clin Cancer Res; 19(16) August 15, 2013
4353
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
but a linear projection and quantification of
prospectiveclinical efficacy is difficult to be predicted. In
vivopersistenceof patient-infused CIK cells is expected to be about
2 weeksand multiple infusions will be possible based on
theirintense ex vivo expansion and production simplicity.
A selected experiment to assess the in vivo activity
ofpatient-derived CIK cells against autologous melanomatargets
engrafted into NOD/SCID mice has shown delayedtumor growth, along
with increased extension of necroticareas and infiltration of CIK
cells at tumor sites. Thisadditional work was intended to provide
proof of in vivoactivity of CIK cells; however, a deeper and more
definitivein vivo analysis will require a dedicated study.
The second important feature of CIK cells is their
HLA-unrestricted tumor killing, which extends to virtually
allpatients the possibility to benefit from this
approach,regardless of the expression of specific
tumor-associatedantigen (TAA) restricted by precise HLA haplotypes.
Themechanistic investigation of CIK cell tumor killing was notthe
aim of our study; however, we showed the expression of
target molecules, recognized by NKG2D receptor, on allmetastatic
melanoma primary cell cultures from ourpatients. The ULPB2 molecule
was most consistently repre-sented while more variability among
patients was observedfor MIC A/B.
Direct expression of MIC A/B molecules has been des-cribed on
both primary andmetastaticmelanomas (38, 39).The possible
upregulation of NKG2D ligands in varioustypes of treatment (e.g.,
chemotherapy, statins, doxycycline),with a consequent increased
susceptibility toMHC-indepen-dent immune-mediated lysis, has been
described in variousexperimental models (40–42) and may provide
intriguingprospective synergies with immunotherapy.
Downregulation of MHC expression is one of the mainimmune-escape
mechanisms developed by tumor cells. Inour study, CIK cells were
effective against two MHC class Inegative melanoma samples, which
confirmed their poten-tial in melanomas with immunogenicity
alterations.
Melanoma primary cell cultures were derived frompatient tumor
biopsies. These cultures retained original
Effector:Target ratio
% o
f T
um
or-
spec
ific
lysi
s
40:1
20:1
10:1 5:
10
20
40
60
80CIKs vs. autologous melanoma
CIKs vs. DettMel
Effector:Target ratio
% o
f T
um
or-
spec
ific
lysi
s
40:1
20:1
10:1 5:
10
20
40
60
80CIKs vs. allogenic melanoma
CIKs vs. DettMel
CIK in vitro tumor killing activityA
C
Anti-CD56Anti-CD5
Are
a o
f ti
ssu
e n
ecro
sis
(%)
Trea
ted
(aut
olog
ous C
IK)
Untre
ated
0
10
20
30
*
B
D
Weeks of treatment
Tu
mo
r vo
lum
e (c
m3 )
1 2 3 4 5 6 70
1
2
3
Untreated
CIK treated (1×107 CIK/mouse)
Figure 2. In vitro and in vivo activity of CIK cells against
autologous melanoma. A, patient-derived CIK cells efficiently
killed in vitro all bulk autologousmelanoma targets (n ¼ 35; left);
results were comparable with those obtained with allogeneic CIK
cells assessed in parallel versus the same tumor cells(n ¼ 20;
right). Tumor killing was evaluated by flow cytometry assay
conducted after coculturing mature CIK cells with PKH-26–stained
targets for 4 hours.B,NOD/SCIDmice (n¼12)were subcutaneously
implantedwith an8mm3 tumor fragment of
patient-derivedmelanomabiopsy (mMel2).Oneweekafter
tumorimplantation, 107 CIK cells were infused weekly by tail vein
injection (n ¼ 8). Percentage of tissue necrosis on tumor growth
was calculated at the endof the experiment on paraffin-embedded
histologic sections. The results were expressed by mean� SEM and
the extension of necrotic areas was analyzedby an unpaired,
two-tailed t test (�, P ¼ 0.0255). C, at the end of the infusions,
infiltration of CIK cells at tumor sites were shown by
immunohistochemistryusing antibodies against CD5 and CD56. Scale
bars, 12 mm. D, a significant delay in tumor growth was observed in
NOD/SCID–treated mice comparedwith the controls (n ¼ 4). Tumor
volume increments were expressed as mean � SEM and CIK activity was
analyzed by two-way ANOVA (P ¼ 0.0308).
Gammaitoni et al.
Clin Cancer Res; 19(16) August 15, 2013 Clinical Cancer
Research4354
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
tumor characteristics and displayed great immunopheno-typic
heterogeneity among samples. Most differentiationantigens detected
on metastatic melanoma cells showedvariable levels of expression,
as others have described (23,24). In contrast, the expression and
average levels of puta-tive stemness markers, Oct4 and ABCG2, as
well as the lackof Mitf expression, were quite comparable among
differentmelanoma samples (27, 43–45). Together, these data
sug-gest that the cell fraction endowed with stemness features
isstably detectable and retained in different samples. This
isconsistent with the decade-old CSC theory that tumorscontain a
subset of cells that both self-renew and generatedifferentiated
progeny (9, 46, 47). CSCs are, therefore, thedriving force of the
tumor.Truthfully, the identification of molecular and pheno-
typic markers for CSCs still remains partially unsolved. Infact,
CSCs seem to have a dynamic phenotype, more likelyas expression of
a functional state rather than a precisecellular entity (48, 49).
The expression of the Oct4 geneseems able to be reliably associated
with cancer cells ofvarious histotypes endowed with stemness
features (50–55). Recently, Oct4 expression was correlated to
dediffer-entiation of melanoma cells, reacquisition of stem
pheno-type, increased tumorigenic capacity, and resistance
tochemotherapy (28). We exploited these observations by
designing a gene transfer strategy to detect mCSCs andassess
their susceptibility to CIK-mediated killing. Bulk,patient-derived
metastatic melanoma cells were transducedwith a lentiviral vector
that encodes eGFP under control ofthe humanOct4 regulatory element,
with the idea that onlymCSCs are able to activate the Oct4 promoter
to expresseGFP, which allows their specific killing by CIK cells to
betracked and evaluated.
This approach uncovered a small subpopulation ofeGFPþ putative
mCSCs, consistent with the expected rateof mCSCs given detected
Oct4 protein. This small fractionappeared to preferentially
coexpress the stemness markerABCG2 relative to its eGFP�
counterpart. To identify func-tionally rare quiescent/slowly
dividing CSCs, a lipophilicfluorescent dye, PKH26, was used to
visualize relativelyquiescent cells within a proliferating
population (56).Indeed, eGFP� cells encountered up to 5 cell
divisionsduring a 3-week culture period, whereas eGFPþ
cellsencountered a maximum of 2 cell divisions in the sameelapsed
time. Moreover, as CSCs can withstand anoikis,they
proliferate/differentiate in anchorage-independentconditions and
give rise to clonal spheroids. Melanomaprimary cells after
LV-Oct4.eGFP transduction were thencultured in
anchorage-independent and serum-free condi-tions. Only a small
fraction of cells retained the ability to
T=0eGFP– T=+21 days
eGFP+ T=+21 days
PKH26 Dye fluorescence
C Proliferation assay
mM
el10
O
ct4.
eGF
P+
mM
el10
Oct
4.eG
FP
-
mM
el3
Oct
4.eG
FP
+
mM
el3
Oct
4.eG
FP
-
mM
el2
Oct
4.eG
FP
+
mM
el2
Oct
4.eG
FP
-
LV-Oct4.eGFPExpression cassette
A
LV-Oct4.eGFP–Transduced
melanoma cells
LV-Oct4.eGFP–Transduced
mES cells
LV-PGK.eGFP–Transduced
melanoma cells
B
Effector:Target ratio%
of
Tu
mo
r-sp
ecif
ic ly
sis
40:1
20:1
10:1 5:
10
20
40
60
80CIKs vs. eGFP+ sorted cellsCIKs vs. eGFP– sorted cells
CIK in vitro tumor killing activityD
Figure 3. CIK cells efficiently killed autologous melanoma
putative mCSC. A, schematic representation of lentiviral vector
LV-Oct4.eGFP used to visualizeputative mCSC. The presence of
integrated LV-Oct4.eGFP was confirmed by PCR in both eGFPþ and
eGFP� fractions of freshly sorted cells. Pictureshows
representative PCR electrophoresis gel (100 ng gDNA/each sample)
with primers annealing on the lentiviral vector backbone upstream
anddownstream the Oct4.eGFP expression cassette. B, representative
eGFP expression in melanoma primary cell cultures transduced with
LV-Oct4.eGFPor LV-PGK.eGFP; as positive transduction control, mES
cells transduced with LV-Oct4-eGFP (scale bars, 15 mm). C,
proliferation assay in vitro wasevaluated by staining
LV-Oct4.eGFP-transduced tumor cells with the vital dye PKH26 and
assessing the fluorescence intensity decrement over time.
Arepresentative experiment is reported in figure. D, the antitumor
activity of patient-derived CIK cells was equally intense against
autologous Oct4.eGFPþ
mCSC (n ¼ 4); results were comparable with those observed
against Oct4.eGFP� bulk melanoma cells.
CIK Cells Kill Autologous Melanoma and mCSCs
www.aacrjournals.org Clin Cancer Res; 19(16) August 15, 2013
4355
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
grow, and spheroidswere generated exclusively fromeGFPþ
cells that maintained the fluorescence even when
hetero-geneously distributedwithin the spheres. CIK cells
intenselykilled the autologous eGFPþ-sorted fraction at a lysis
ratecomparable with that observed against eGFP� tumortargets.
Killing the truemelanoma stem cells currently remains anideal
concept. Yet, our data suggest that CIK cells kill asubset of
autologous metastatic melanoma cells able toactivate Oct4 that,
based on current knowledge, reliablydefines a subpopulation of
tumor cells with stemness fea-tures (28). Dedicated studies are
required and currentlyongoing to investigate deeply the functional
and tumori-genic characteristics of eGFPþ mCSCs. Nevertheless,
ourfindings provide new and additional weight to the potentialof
cancer immunotherapy with CIK cells. Data from Kumarand colleagues
confirmed that the Oct4 expression corre-lates with putative CSC
features and that Oct4-expressingcells display a significantly
higher chemotherapy agentresistance (28). Indeed, the observed
killing ability of CIKcells against putative mCSCs may reveal other
valuableperspectives on the potential of this immunotherapy
strat-egy. Moreover, the elevated safety profile of CIK cells
doesnot preclude their use in associationwith other approaches.One
appealing possibility would be to explore their poten-tial
synergism with conventional chemotherapies or
evenmolecular-targeted treatments. This strategy could
reduceresistance occurrence by improving the odds of targeting
thecrucial CSC subset from which tumor regrowth is specu-lated to
start.
The MHC-unrestricted tumor killing mechanism of CIKcells showed
in this study may advantage it over otherimmunotherapy approaches
because it addresses the diffi-cult quest of targeting mCSCs and
also HLA-negativetumors. We showed that the membrane expression
ofNKG2D ligands is maintained on putative mCSCs. Viceversa, it is
still unknown whether or not such a peculiartumor subpopulation
retains the same antigenic features(specific TAA or MHC molecule
expression), as do othertumor cells.
Overall, shown here, for the first time, is the intensetumor
killing activity of CIK cells against autologous met-astatic
melanoma, including putative mCSCs. These data
point toCIK cells as favorable candidates for clinical trials
inpatients withmelanoma. The biologic basis is set for
furtherpreclinical and clinical investigations on the
prospectivepotential of targeting mCSCs with CIK cells, either
inde-pendently or in synergismwith other therapeutic
strategies.
Disclosure of Potential Conflicts of InterestNo potential
conflicts of interest were disclosed.
Authors' ContributionsConception and design: L. Gammaitoni, F.
Picciotto, G. Grignani, M.Aglietta, D. SangioloDevelopment of
methodology: L. Gammaitoni, V. Leuci, F. Picciotto, A.Pisacane, T.
Venesio, A. Balsamo, Y. Pignochino, D. SangioloAcquisitionofdata
(provided animals, acquired andmanagedpatients,provided facilities,
etc.): L. Gammaitoni, L. Giraudo, V. Leuci, M. Todoro-vic, G.
Mesiano, A. Pisacane, A. Zaccagna, M.G. Volpe, S. Gallo, D.
Caravelli,E. Giacone, T. Venesio, A. Balsamo, G. GrignaniAnalysis
and interpretation of data (e.g., statistical analysis,
biosta-tistics, computational analysis): L. Gammaitoni, L. Giraudo,
V. Leuci, M.Todorovic, G. Grignani, D. SangioloWriting, review,
and/or revision of the manuscript: L. Gammaitoni, L.Giraudo, V.
Leuci, M. Todorovic, G. Mesiano, F. Picciotto, A. Zaccagna,
E.Giacone, G. Grignani, F. Carnevale-Schianca, M. Aglietta, D.
SangioloAdministrative, technical, or material support (i.e.,
reporting or orga-nizing data, constructing databases): L.
GammaitoniStudy supervision: L. Gammaitoni, F. Picciotto, G.
Grignani,M. Aglietta, D.Sangiolo
AcknowledgmentsThe authors thank Dr. W. Cui (IRDB, Imperial
College London, London,
United Kingdom) who provided the phOCT4.EGFP1 vector and Dr.
ElisaVignawhoprovided the transfer vector
pRRL.sin.PPT.hPGK.EGFP.Wpre (LV-PGK.EGFP). The authors also thank
Joan Leonard (Leonard Editorial Ser-vices, LLC) for the linguistic
revision and editorial assistance and E. Lantelmefor sorting
services.
Grant SupportThis work was supported by grants from Progetti di
Ricerca Rete Onco-
logica Piemonte-Valle d’Aosta, Associazione Italiana Ricerca sul
Cancro–AIRC I.G. grant no. 11515, Associazione Italiana Ricerca sul
Cancro–AIRC 5� 1000, and University of Torino-Progetti di Ateneo
2011 grant RETHE-ORTO 11RKTW. The fellowships of L. Giraudo, M.
Todorovic, and Y.Pignochino are sponsored byMIUR (University of
Turin) and the fellowshipof G. Mesiano is sponsored by an
Associazione Italiana Ricerca sul Cancro–AIRC I.G. grant no.
11515.
The costs of publication of this article were defrayed in part
by thepayment of page charges. This article must therefore be
hereby markedadvertisement in accordance with 18 U.S.C. Section
1734 solely to indicatethis fact.
Received January 8, 2013; revised May 15, 2013; accepted June
10, 2013;published OnlineFirst June 21, 2013.
References1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics,
2010. CA Cancer
J Clin 2010;60:277–300.2. Middleton MR, Grob JJ, Aaronson N,
Fierlbeck G, Tilgen W, Seiter S,
et al. Randomized phase III study of temozolomide versus
dacarbazinein the treatment of patients with advanced metastatic
malignantmelanoma. J Clin Oncol 2000;18:158–66.
3. Chapman PB, Hauschild A, Robert C, Haanen JB, Ascierto P,
Larkin J,et al. Improved survival with vemurafenib in melanoma with
BRAFV600E mutation. N Engl J Med 2011;364:2507–16.
4. Hauschild A, Grob JJ, Demidov LV, Jouary T, Gutzmer R,
Millward M,et al. Dabrafenib in BRAF-mutated metastatic melanoma: a
multi-centre, open-label, phase 3 randomised controlled trial.
Lancet 2012;380:358–65.
5. Flaherty KT, Robert C, Hersey P, Nathan P, Garbe C, Milhem M,
et al.Improved survival with MEK inhibition in BRAF-mutated
melanoma.N Engl J Med 2012;367:107–14.
6. Hodi FS, O'Day SJ, McDermott DF, Weber RW, Sosman JA,
HaanenJB, et al. Improved survival with ipilimumab in patients with
metastaticmelanoma. N Engl J Med 2010;363:711–23.
7. Wolchok JD, Neyns B, Linette G, Negrier S, Lutzky J, Thomas
L, et al.Ipilimumab monotherapy in patients with pretreated
advanced mela-noma: a randomised, double-blind, multicentre, phase
2, dose-rang-ing study. Lancet Oncol 2010;11:155–64.
8. Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC,
McDer-mott DF, et al. Safety, activity, and immune correlates of
anti-PD-1antibody in cancer. N Engl J Med 2012;366:2443–54.
Gammaitoni et al.
Clin Cancer Res; 19(16) August 15, 2013 Clinical Cancer
Research4356
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
9. Reya T, Morrison SJ, ClarkeMF,Weissman IL. Stem cells,
cancer, andcancer stem cells. Nature 2001;414:105–11.
10. Rosenberg SA, Yang JC, Sherry RM, Kammula US, Hughes MS,
PhanGQ, et al. Durable complete responses in heavily pretreated
patientswith metastatic melanoma using T-cell transfer
immunotherapy. ClinCancer Res 2011;17:4550–7.
11. Robbins PF, Morgan RA, Feldman SA, Yang JC, Sherry RM,
DudleyME, et al. Tumor regression in patients with metastatic
synovial cellsarcoma and melanoma using genetically engineered
lymphocytesreactive with NY-ESO-1. J Clin Oncol 2011;29:917–24.
12. Rosenberg SA. Cell transfer immunotherapy for metastatic
solid can-cer–what clinicians need to know. Nat Rev Clin Oncol
2011;8:577–85.
13. Mesiano G, Todorovic M, Gammaitoni L, Leuci V, Giraudo Diego
L,Carnevale-Schianca F, et al. Cytokine-induced killer (CIK) cells
asfeasible and effective adoptive immunotherapy for the treatment
ofsolid tumors. Expert Opin Biol Ther 2012;12:673–84.
14. Hontscha C, Borck Y, Zhou H, Messmer D, Schmidt-Wolf IG.
Clinicaltrials on CIK cells: first report of the international
registry on CIK cells(IRCC). J Cancer Res Clin Oncol
2011;137:305–10.
15. Lu PH, Negrin RS. A novel population of expanded
humanCD3þCD56þ cells derived from T cells with potent in vivo
antitumoractivity in mice with severe combined immunodeficiency. J
Immunol1994;153:1687–96.
16. Schmidt-Wolf IG, Negrin RS,KiemHP,BlumeKG,Weissman IL.Use
ofa SCID mouse/human lymphomamodel to evaluate
cytokine-inducedkiller cells with potent antitumor cell activity. J
Exp Med 1991;174:139–49.
17. Schmidt-Wolf IG, Lefterova P, Johnston V, Huhn D, Blume KG,
NegrinRS. Propagation of large numbers of T cells with natural
killer cellmarkers. Br J Haematol 1994;87:453–8.
18. Baker J, Verneris MR, Ito M, Shizuru JA, Negrin RS.
Expansion ofcytolytic CD8(þ) natural killer T cells with limited
capacity for graft-versus-host disease induction due to interferon
gamma production.Blood 2001;97:2923–31.
19. Karimi M, Cao TM, Baker JA, Verneris MR, Soares L, Negrin
RS.Silencing human NKG2D, DAP10, and DAP12 reduces cytotoxicityof
activated CD8þ T cells and NK cells. J Immunol
2005;175:7819–28.
20. Verneris MR, Karami M, Baker J, Jayaswal A, Negrin RS. Role
ofNKG2D signaling in the cytotoxicity of activated and expanded
CD8þT cells. Blood 2004;103:3065–72.
21. Olioso P, Giancola R, Di Riti M, Contento A, Accorsi P,
Iacone A.Immunotherapy with cytokine induced killer cells in solid
and hema-topoietic tumours: a pilot clinical trial. Hematol Oncol
2009;27:130–9.
22. Schmidt-Wolf IG, Finke S, Trojaneck B, Denkena A, Lefterova
P,Schwella N, et al. Phase I clinical study applying autologous
immu-nological effector cells transfected with the interleukin-2
gene inpatients with metastatic renal cancer, colorectal cancer and
lympho-ma. Br J Cancer 1999;81:1009–16.
23. Quintana E, Shackleton M, Foster HR, Fullen DR, Sabel MS,
JohnsonTM, et al. Phenotypic heterogeneity among tumorigenic
melanomacells from patients that is reversible and not
hierarchically organized.Cancer Cell 2010;18:510–23.
24. Quintana E, Shackleton M, Sabel MS, Fullen DR, Johnson TM,
Morri-son SJ. Efficient tumour formation by single human melanoma
cells.Nature 2008;456:593–8.
25. Nagata S, Toyoda M, Yamaguchi S, Hirano K, Makino H, Nishino
K,et al. Efficient reprogramming of human and mouse primary
extra-embryonic cells to pluripotent stem cells. Genes Cells
2009;14:1395–404.
26. Nichols J, Zevnik B, Anastassiadis K, Niwa H, Klewe-Nebenius
D,Chambers I, et al. Formationof pluripotent stemcells in
themammalianembryo depends on the POU transcription factor Oct4.
Cell 1998;95:379–91.
27. Cheli Y, Giuliano S, Fenouille N, Allegra M, Hofman V,
Hofman P, et al.Hypoxia and MITF control metastatic behaviour in
mouse and humanmelanoma cells. Oncogene 2012;31:2461–70.
28. Kumar SM, Liu S, LuH, ZhangH, ZhangPJ,Gimotty PA, et al.
Acquiredcancer stem cell phenotypes through Oct4-mediated
dedifferentia-tion. Oncogene 2012;31:4898–911.
29. Follenzi A, Ailles LE, Bakovic S, Geuna M, Naldini L. Gene
transfer bylentiviral vectors is limited by nuclear translocation
and rescued byHIV-1 pol sequences. Nat Genet 2000;25:217–22.
30. Gerrard L, Zhao D, Clark AJ, Cui W. Stably transfected human
embry-onic stem cell clones express OCT4-specific green fluorescent
proteinand maintain self-renewal and pluripotency. Stem Cells
2005;23:124–33.
31. Orfanelli U, Wenke AK, Doglioni C, Russo V, Bosserhoff AK,
LavorgnaG. Identification of novel sense and antisense
transcription at theTRPM2 locus in cancer. Cell Res
2008;18:1128–40.
32. Fischer K, Andreesen R, Mackensen A. An improved flow
cytometricassay for the determination of cytotoxic T lymphocyte
activity.J Immunol Methods 2002;259:159–69.
33. Introna M, Borleri G, Conti E, Franceschetti M, Barbui AM,
Broady R,et al. Repeated infusions of donor-derived
cytokine-induced killer cellsin patients relapsing after allogeneic
stem cell transplantation: a phaseI study. Haematologica
2007;92:952–9.
34. Laport GG, Sheehan K, Baker J, Armstrong R, Wong RM, Lowsky
R,et al. Adoptive immunotherapy with cytokine-induced killer cells
forpatients with relapsed hematologic malignancies after
allogeneichematopoietic cell transplantation. Biol Blood Marrow
Transplant2011;17:1679–87.
35. Wu C, Jiang J, Shi L, Xu N. Prospective study of
chemotherapy incombination with cytokine-induced killer cells in
patients sufferingfrom advanced non-small cell lung cancer.
Anticancer Res2008;28:3997–4002.
36. Jiang JT, Shen YP,Wu CP, Zhu YB,Wei WX, Chen LJ, et al.
Increasingthe frequency of CIK cells adoptive immunotherapymay
decrease riskof death in gastric cancer patients. World J
Gastroenterol 2010;16:6155–62.
37. Liu L, Zhang W, Qi X, Li H, Yu J, Wei S, et al. Randomized
study ofautologous cytokine-induced killer cell immunotherapy in
metastaticrenal carcinoma. Clin Cancer Res 2012;18:1751–9.
38. Vetter CS, Groh V, thor Straten P, Spies T, Br€ocker EB,
Becker JC.Expression of stress-induced MHC class I related chain
molecules onhuman melanoma. J Invest Dermatol 2002;118:600–5.
39. Vetter CS, Lieb W, Br€ocker EB, Becker JC. Loss of
nonclassical MHCmolecules MIC-A/B expression during progression of
uveal melano-ma. Br J Cancer 2004;91:1495–9.
40. PichC, Teiti I, Rochaix P,Mariam�eB, Couderc B, Favre G, et
al. Statinsreduce melanoma development and metastasis through MICA
over-expression. Front Immunol 2013;4:62.
41. Tang H, Sampath P, Yan X, Thorne SH. Potential for
enhancedtherapeutic activity of biological cancer therapies with
doxycyclinecombination. Gene Ther 2013;20:770–8.
42. Hervieu A, R�eb�e C, V�egran F, Chalmin F, Bruchard M,
Vabres P, et al.Dacarbazine-mediated upregulation of NKG2D ligands
on tumor cellsactivates NK and CD8 T cells and restrains melanoma
growth. J InvestDermatol 2013;133:499–508.
43. Monzani E, Facchetti F, Galmozzi E, Corsini E, Benetti A,
Cavazzin C,et al. Melanoma contains CD133 and ABCG2 positive cells
withenhanced tumourigenic potential. Eur J Cancer
2007;43:935–46.
44. La Porta C. Cancer stem cells: lessons frommelanoma.
StemCell Rev2009;5:61–5.
45. La Porta CA, Zapperi S. Human breast and melanoma cancer
stemcells biomarkers. Cancer Lett. 2012 Mar 20. [Epub ahead of
print].
46. Pardal R, Clarke MF, Morrison SJ. Applying the principles of
stem-cellbiology to cancer. Nat Rev Cancer 2003;3:895–902.
47. Dick JE. Acute myeloid leukemia stem cells. Ann N Y Acad
Sci2005;1044:1–5.
48. La Porta CA, Zapperi S, Sethna JP. Senescent cells in
growing tumors:population dynamics and cancer stem cells. PLoS
Comput Biol2012;8:e1002316.
49. Visvader JE, Lindeman GJ. Cancer stem cells in solid
tumours: accu-mulating evidence and unresolved questions. Nat Rev
Cancer2008;8:755–68.
50. Asadi MH, Mowla SJ, Fathi F, Aleyasin A, Asadzadeh J, Atlasi
Y.OCT4B1, a novel spliced variant ofOCT4, is highly expressed in
gastriccancer and acts as an antiapoptotic factor. Int J Cancer
2011;128:2645–52.
CIK Cells Kill Autologous Melanoma and mCSCs
www.aacrjournals.org Clin Cancer Res; 19(16) August 15, 2013
4357
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
51. Chang CJ, Chien Y, Lu KH, Chang SC, Chou YC, Huang CS, et
al.Oct4-related cytokine effects regulate tumorigenic properties
ofcolorectal cancer cells. Biochem Biophys Res Commun
2011;415:245–51.
52. Chiou SH, Wang ML, Chou YT, Chen CJ, Hong CF, Hsieh WJ, et
al.Coexpression of Oct4 and Nanog enhances malignancy in
lungadenocarcinoma by inducing cancer stem cell-like properties
andepithelial-mesenchymal transdifferentiation. Cancer Res
2010;70:10433–44.
53. Dong Z, Zeng Q, Luo H, Zou J, Cao C, Liang J, et al.
Increasedexpression of OCT4 is associated with low differentiation
and tumor
recurrence in human hepatocellular carcinoma. Pathol Res
Pract2012;208:527–33.
54. Gu G, Yuan J, Wills M, Kasper S. Prostate cancer cells with
stem cellcharacteristics reconstitute the original human tumor in
vivo. CancerRes 2007;67:4807–15.
55. Wen J, Park JY, Park KH, ChungHW, Bang S, Park SW, et al.
Oct4 andNanog expression is associated with early stages of
pancreatic car-cinogenesis. Pancreas 2010;39:622–6.
56. Pece S, Tosoni D, Confalonieri S, Mazzarol G, Vecchi M,
Ronzoni S,et al. Biological and molecular heterogeneity of breast
cancers corre-lates with their cancer stem cell content. Cell
2010;140:62–73.
Gammaitoni et al.
Clin Cancer Res; 19(16) August 15, 2013 Clinical Cancer
Research4358
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/
-
2013;19:4347-4358. Published OnlineFirst June 21, 2013.Clin
Cancer Res Loretta Gammaitoni, Lidia Giraudo, Valeria Leuci, et al.
FeaturesAutologous Metastatic Melanoma Including Cells with
Stemness Effective Activity of Cytokine-Induced Killer Cells
against
Updated version
10.1158/1078-0432.CCR-13-0061doi:
Access the most recent version of this article at:
Material
Supplementary
http://clincancerres.aacrjournals.org/content/suppl/2013/06/21/1078-0432.CCR-13-0061.DC1
Access the most recent supplemental material at:
Cited articles
http://clincancerres.aacrjournals.org/content/19/16/4347.full#ref-list-1
This article cites 55 articles, 13 of which you can access for
free at:
Citing articles
http://clincancerres.aacrjournals.org/content/19/16/4347.full#related-urls
This article has been cited by 5 HighWire-hosted articles.
Access the articles at:
E-mail alerts related to this article or journal.Sign up to
receive free email-alerts
Subscriptions
Reprints and
[email protected]
To order reprints of this article or to subscribe to the
journal, contact the AACR Publications Department at
Permissions
Rightslink site. Click on "Request Permissions" which will take
you to the Copyright Clearance Center's (CCC)
.http://clincancerres.aacrjournals.org/content/19/16/4347To
request permission to re-use all or part of this article, use this
link
on June 3, 2021. © 2013 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst June 21, 2013; DOI:
10.1158/1078-0432.CCR-13-0061
http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-13-0061http://clincancerres.aacrjournals.org/content/suppl/2013/06/21/1078-0432.CCR-13-0061.DC1http://clincancerres.aacrjournals.org/content/19/16/4347.full#ref-list-1http://clincancerres.aacrjournals.org/content/19/16/4347.full#related-urlshttp://clincancerres.aacrjournals.org/cgi/alertsmailto:[email protected]://clincancerres.aacrjournals.org/content/19/16/4347http://clincancerres.aacrjournals.org/