1. This article was downloaded by: [University of Aberdeen]
On: 21 July 2009
Access details: Access Details: [subscription number 908669290]
Publisher Informa Healthcare
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Cytotherapy
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713656803
Clinical-grade ex vivo-expanded human natural killer cells up-regulate
activating receptors and death receptor ligands and have enhanced cytolytic
activity against tumor cells
Maria Berg a; Andreas Lundqvist a; Philip McCoy Jr b; Leigh Samsel b; Yong Fan c; Abdul Tawab c; Richard
Childs a
a
Hematology Branch, b Flow Cytometry Core Facility, National Heart, Lung and Blood Institute, c Department
of Transfusion Medicine, Cell Processing Section, National Institutes of Health, Bethesda, Maryland, USA
First Published:May2009
To cite this Article Berg, Maria, Lundqvist, Andreas, McCoy Jr, Philip, Samsel, Leigh, Fan, Yong, Tawab, Abdul and Childs,
Richard(2009)'Clinical-grade ex vivo-expanded human natural killer cells up-regulate activating receptors and death receptor ligands
and have enhanced cytolytic activity against tumor cells',Cytotherapy,11:3,341 — 355
To link to this Article: DOI: 10.1080/14653240902807034
URL: http://dx.doi.org/10.1080/14653240902807034
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.

2. Cytotherapy (2009) Vol. 11, No. 3, 341Á355
Clinical-grade ex vivo-expanded human natural
killer cells up-regulate activating receptors and
death receptor ligands and have enhanced
cytolytic activity against tumor cells
Maria Berg1, Andreas Lundqvist1, Philip McCoy JR2, Leigh Samsel2, Yong Fan3,
Abdul Tawab3 and Richard Childs1
1
Hematology Branch, 2Flow Cytometry Core Facility, National Heart, Lung and Blood Institute, and 3Department of Transfusion Medicine,
Cell Processing Section, National Institutes of Health, Bethesda, Maryland, USA
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
Background aims NK cells, co-cultured with K562 and renal cell carcinoma tumor
Cancer immunotherapy involving natural killer (NK) cell infusions targets, secreted significantly higher levels of soluble Fas ligand 6;
and administration of therapeutic agents modulating the susceptibility fgjhd, IFN-g, GM-CSF, TNF-a, MIP-1a and MIP-1b compared
of tumors to NK-cell lysis has been proposed recently. We provide a with resting NK cells. Secretion of the above cytokines and NK-cell
method for expanding highly cytotoxic clinical-grade NK cells in vitro cytolytic function were IL-2 dose dependent. Cryopreservation of
for adoptive transfer following bortezomib treatment in patients with expanded NK cells reduced expression of NKG2D and TRAIL and
advanced malignancies. NK-cell cytotoxicity, although this effect could be reversed by exposure
of NK cells to IL-2.
Methods
NK cells were expanded with irradiated EpsteinÁBarr virus- Conclusions
transformed lymphoblastoid cells. Expanded cells were evaluated for We describe a method for large-scale expansion of NK cells with
their phenotype, cytotoxicity, cytokine secretion, dependence on inter- increased expression of activating receptors and death receptor ligands
leukin (IL)-2 and ability to retain function after cryopre- resulting in superior cytotoxicity against tumor cells. This ex vivo
servation. NK-cell expansion technique is currently being utilized in a clinical
trial evaluating the anti-tumor activity of adoptively infused NK cells
Results in combination with bortezomib.
A pure population of clinical-grade NK cells expanded 4909 260-fold
over 21 days. Expanded NK cells had increased TRAIL, FasL and Keywords
NKG2D expression and significantly higher cytotoxicity against Bortezomib, expansion, immunotherapy, natural killer cells.
bortezomib-treated tumors compared with resting NK cells. Expanded
Introduction receptors expressed on target cells. They can also mediate
Natural killer (NK) cells are innate immune lymphocytes antibody (Ab)-dependent cellular cytotoxicity (ADCC) via
that are identified by the expression of CD56 surface antigen the membrane receptors FcgRIII (CD16) [3]. Unlike Tcells,
(Ag) and lack of CD3 [1,2]. NK cells have the ability to kill NK cells do not require the presence of a specific tumor Ag
target cells directly through the release of granules contain- to kill cancer cells, rather their recognition of targets is
ing perforin and serine proteases (granzymes) and/or by regulated through a balance of activating and inhibitory
surface-expressed ligands that engage and activate death signals. Even in the presence of an activating ligand,
Correspondence to: Richard Childs, Hematology Branch, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD
20892Á1652, USA. Room 3Á5140, Building 10-CRC, 10 Center Drive MSC 1202, Bethesda, MD 20892Á1202, USA. E-mail: childsr@nih.gov
– 2009 ISCT DOI: 10.1080/14653240902807034

3. 342 M. Berg et al.
inhibitory ligands can initiate overriding signals that of these changes, NK-cell cytolytic activity against
culminate in a net suppression of NK-cell function. The bortezomib-treated tumors is significantly higher with
inactivation of NK cells by self-HLA molecules is a expanded compared with fresh NK cells.
potential mechanism by which malignant cells evade host
NK-cell mediated immunity [4,5]. Methods
Recently, we and others observed that the proteasome Cell isolation, culture and cryopreservation
inhibitor bortezomib (Velcade, PS-341) sensitized malig- Human NK cells were isolated from PBMC obtained from
nant cells to tumor necrosis factor (TNF)-related multiple different healthy volunteers and one patient with
apoptosis-inducing ligand (TRAIL)-dependent NK-cell metastatic sarcoma. Depletion of CD3 ' T cells and a
lysis [6Á8]. This effect appeared to overcome Killer cell subsequent positive selection of CD56 ' cells were per-
immunogloblin Á like receptors (KIR)-mediated suppres- formed on a CliniMACS system (Miltenyi Biotec Inc.,
sion of NK-cell function, enhancing autologous NK-cell Auburn, CA, USA). The cells were analyzed immediately
cytotoxicity against patient tumor cells in vitro. Based on after purification for phenotypic markers and cytotoxicity
this finding, we pursued a method for large-scale and were then either expanded or cryopreserved for future
expansion of clinical-grade NK cells to evaluate the
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
analysis. For NK expansions the following parameters were
anti-cancer effects of autologous adoptively infused NK tested: autologous/allogeneic PBMC versus EBV-LCL as
cells following bortezomib treatment in patients with feeder cells; culture vessels (flasks versus bags); feeder cell
cancer. irradiation doses (25, 50 and 75 Gy); feeder to NK-cell
Only a few trials investigating adoptive NK-cell infu- ratios (90:1, 50:1, 20:1, 10:1, 5:1 and 1:1) and plasma
sions in humans with cancer have been conducted to date (obtained from NK-cell donors or from PBMC donors)
(reviewed in 9,10). Because NK cells represent only a versus serum (2%, 5% and 10% pooled AB plasma, AB
minor fraction of human lymphocytes, the small number of serum and six different lots of commercial AB serum).
NK cells isolated following a typical leukapheresis
procedure has precluded phase I trials evaluating NK- NK-cell expansion in flasks
cell dose-dependent tumor cytotoxicity in humans with (small-scale expansions)
cancer. Twenty million 100 Gy-irradiated and washed EBV-LCL
Several methods for expansion and activation of NK cells were co-cultured with 106 magnetic bead-purified
cells in vitro have been investigated, including overnight NK cells in upright 75-cm2 tissue culture flasks in 15 mL
and long-term culture with cytokines [11,12] and the use X-VIVO 20 (Lonza, Walkersville, MD, USA) supplemen-
of peripheral blood mononuclear cells (PBMC) [13], K562 ted with 10% heat-inactivated human AB serum (Gemini
cells [14] and EpsteinÁBarr virus-transformed lympho- Bio-Products, West Sacramento, CA, USA) or 10% heat-
blastoid cell lines (EBV-LCL) as feeder cells [15,16]. We inactivated AB single donor or pooled plasma or serum
have previously developed [17] and now optimized an [obtained from The Department of Transfusion Medicine
improved method for large-scale expansion of human NK (DTM), National Institutes of Health (NIH), Bethesda,
cells in bags using irradiated EBV-LCL feeder cells and MD, USA], 500 IU/mL recombinant human (rh)IL-2
interleukin (IL)-2. The EBV-LCL cell line, used in our (50 ng/mL; TecinTM, Hoffmann-La Roche Inc., Nutley,
studies, has been proven previously [18] to be safe for use NJ, USA) and 2 mM GlutaMAX-1 (Invitrogen, Carlsbad,
in clinical trials; cells have met release test criteria for the CA, USA) at 378C and 6.5% CO2. The effect on NK-cell
presence of viral contaminants and infectious EBV. We proliferation of varying the percentage of CO2 from 5%
explored the phenotype, cytotoxic potential against tumor to 8% was investigated systematically; proliferation was
cells and cytokine secretion of these expanded NK cells greatest at 6.5% CO2 (data not shown). Therefore, all
compared to freshly isolated cells. We also investigated the NK-cell expansions, both small- and large-scale, were
effects of IL-2 withdrawal on phenotype and function of performed in incubators using 6.5% CO2. After 5 days
expanded cells and, finally, the effects of cryopreservation half of the culture medium was replaced. Starting on day
and thawing. 7, NK cells were diluted to 0.6 )106 cells/mL with
We show that NK-cell phenotype and function are growth medium containing IL-2 every 24Á72 h for up
modulated following in vitro expansion. As a consequence to 28 days. In some experiments, following 14 days of

4. Expansion of NK cells for cancer immunotherapy 343
culture, 1.0 )106 expanded NK cells were co-cultured and 10 U/mL heparin. Cells were thawed at 378C, slowly
with 20 )106 irradiated feeder cells and the culture was diluted with 10 mL thawing medium, and left at room
expanded for an additional 14 days. temperature for 1Á2 h before being centrifuged to avoid
cell breakage. Thawed cells were used for expansion
experiments, in cytotoxicity assays and for flow cytometry
NK-cell expansion in bags
1.5Á2 h following thawing.
(large-scale expansions)
At DTM, under good manufacturing practice (GMP) con-
ditions, 12Á24)106 magnetic bead-purified NK cells were Flow cytometry analysis of resting and expanded
combined with 120Á240 )106 irradiated EBV-TM-LCL NK cells
cells in 100Á140 mL medium containing rhIL-2 obtained The phenotype of freshly isolated or expanded NK cells was
from the NIH Pharmacy Development Service (NIH PDS, assessed by flow cytometry on a FACSCaliburTM (BD
Bethesda, MD, USA) in Baxter 180-cm2 300-mL bags Biosciences, San Jose, CA, USA) with the following anti-
(Fenwal Lifecell, Baxter Healthcare Corporation, Deer- human monoclonal antibodies (MAb): anti-CD56Áallophy-
field, IL, USA). Four to 5 days after initiation of culture, cocyanin (APC) (clone B159), anti-CD16Áfluorescein iso-
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
half of the medium was replaced. Two days later, the thiocyanate (FITC) (clone 3G8), anti-CD3Áphycoerythrin
concentration of NK cells was adjusted to 106 cells/mL (PE) (clone UCHT1), anti-CD25ÁPE (clone M-A251),
using growth medium containing IL-2. Expanding cells anti-NKG2DÁAPC (clone 1D11), anti-CD244ÁPE (2B4,
were counted and diluted every 24Á72 h until day 28. ¨
clone 269), antiÁCD48ÁFITC (clone TU145), anti-CD11a/
The GMP-certified human EBV-transformed B-cell LFA-1ÁPE (clone G43-25B), anti-FasLÁbiotin (clone
line EBV-TM-LCL was obtained from the Fred Hutch- NOK-1), anti-perforinÁFITC (clone dG9) and anti-
inson Cancer Research Center (FHCRC, Seattle, WA, CD158bÁPE (KIR2DL2/3, clone CH-L); cell viability was
USA); it had been supplied originally to FHCRC by the determined by staining with Via-ProbeTM (7AAD). Intra-
Beckman Research Institute of the City of Hope (Duarte, cellular staining was performed on cells that were permea-
CA, USA) and maintained in our laboratory in RPMI- bilized and fixed using BD Cytofix/CytopermTM. The above
1640, 10% heat-inactivated human AB serum (Gemini Ab and reagents were purchased from BD Biosciences
Bio-Products), 2 mM GlutaMAX-1 and 15 mM HEPES Pharmingen (San Diego, CA, USA) and used according to
(Invitrogen). For large-scale expansions of NK cells, EBV- the manufacturer’s specifications. Anti-granzyme AÁFITC
TM-LCL cells were maintained at DTM in 162-cm2 flasks (clone CB9), anti-granzyme BÁPE (clone GB11) and anti-
or 300-mL cell culture bags at 0.2Á1.0)106/mL in the TRAILÁPE (clone RIK-2) were purchased from Abcam Inc.
above medium supplemented with 10% heat-inactivated (Cambridge, MA, USA). Anti-NKG2AÁAPC (CD94/
human AB-type plasma. The human erythroid leukemia CD159a, clone 131411) and anti-NKG2CÁPE (CD94/
cell line K562 (ATCC, Manassas, VA, USA) and human CD159c, clone 134591) were purchased from R&D Systems
renal cell carcinoma cell lines (RCC) established in our (Minneapolis, MN, USA). Anti-KIR3DL1ÁPE (clone DX9)
laboratory were cultured in DMEM (Lonza), 10% fetal was obtained from BioLegend Inc. (San Diego, CA, USA).
bovine serum (FBS; Quality Biological Inc., Gaithersburg, Cells were also stained with their corresponding isotype-
MD, USA) and 2 mM GlutaMAX-1. matched control MAb.
Freshly isolated or expanded NK cells were cryopre-
served in PlasmaLyte A medium (Baxter) supplemented
with 4% human serum albumin (HAS; Talecris Biother- Cytotoxicity assays
apeutics Inc., Research Triangle Park, NC, USA), 6% Standard 51Cr-release assays were performed as described
pentastarch (hypoxyethylstarch; NIH PDS), 10 mg/mL previously [17] with the following modifications: after a
DNase I (pulmozyme; Genentech Inc., South San Fran- 5-h incubation of NK cells with target cells at various
cisco, CA, USA), 15 U/mL heparin (Abraxis Pharmaceu- effector to target ratios, 25 mL culture supernatants were
tical Products, East Schaumburg, IL, USA) and 5% DMSO transferred onto Luma plates (Perkin Elmer, Wellesley,
at 20Á50 )106 cells/mL/vial. Thawing medium contained MA, USA) and analyzed using a MicroBeta scintillation
X-VIVO 20, 10% human AB serum or plasma, 4% HSA counter (Perkin Elmer).

5. 344 M. Berg et al.
Assay for cytokine production by NK cells Results
co-cultured with K562 and RCC target cells Expansion kinetics of NK cells
One-hundred thousand NK cells expanded for 14 days in Previously, small-scale laboratory-based experiments have
Baxter bags under GMP conditions or NK cells expanded shown that NK-cell lines can be expanded in vitro using a
in tissue culture flasks were washed twice in X-VIVO 20 variety of different methods [16,17]. We sought to optimize
medium and plated into triplicate wells in 96-well tissue the conditions for large-scale NK-cell expansions using
culture plates with 104 K562 or RCC target cells in 200 GMP conditions for NK-cell-based clinical trials in
mL X-VIVO 20 containing 10% human AB serum and 2 humans with cancer.
mM GlutaMAX-1. RCC cells were left untreated or When allogeneic PBMC were used as feeder cells, NK
treated with 10 nM bortezomib (Millennium Pharmaceu- cells were most efficiently expanded by 25 Gy-irradiated
ticals, Cambridge, MA, USA) for 16 h prior to co-culture feeder cells added to cultures at a 20:1 ratio in culture
with expanded NK cells. After 5-h incubation at 378C, medium containing 500 IU/mL IL-2 and 10% single
supernatants were collected and centrifuged, and cleared donor or pooled plasma in upright culture flasks or Baxter
bags at a starting density of 1.0)106 NK cells/mL in 6.5%
supernatants were stored at (208C. Beadlyte† Human
CO2. Under these conditions, up to a 100-fold increase in
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
Multi-Cytokine BeadmasterTM Kit and BeadmatesTM were
cell number was achieved in 15 days, and after a second
obtained from Millipore Corporation (Billerica, MA,
round of expansion for an additional 14 days increases of
USA) and used according to the manufacturer’s specifica-
up to 200Á400-fold could be achieved, although results
tions. Data were acquired on a Luminex IS100 (Luminex
varied depending on the NK-cell donor (Figure 1A).
Corp, Austin, TX, USA) and analyzed using MasterPlex
Cryopreservation and subsequent thawing of purified NK
QT 3.0 (MiraiBio Group, Hitachi Software Engineering
cells before the start of expansion did not affect the
America, South San Francisco, CA, USA). The same
expansion kinetics of NK cells compared with NK cells
culture supernatants were also analyzed by Quantikine†
that were isolated and expanded fresh from the blood.
ELISA (R&D Systems) according to the manufacturer’s We next evaluated whether EBV-LCL (EBV-TM-LCL)
instructions. that had been previously manufactured under GMP con-
ditions would achieve more efficient and consistent NK-cell
yields. Freshly isolated or cryopreserved and thawed non-
IL-2 withdrawal from expanded NK cells
expanded NK cells were cultured in upright 75-cm2 flasks in
NK cells that were expanded for 13Á19 days with EBV-
the presence of irradiated EBV-TM-LCL cells at a 20:1
LCL feeder cells were washed twice in X-VIVO 20 and
feeder to NK-cell ratio. NK cells from five normal donors
cultured at 106 cells/mL in medium without IL-2 or in
cultured for 16 days expanded 815Á3267-fold (Figure 1B).
media containing 5, 50 or 500 IU/mL IL-2 for 24 h. Cells
To facilitate conditions for expanding NK cells at a
were assessed for viability with 7AAD and CD56, CD16,
larger scale under GMP, we then optimized NK-cell
TRAIL and NKG2D expression by flow cytometry. The
expansions in bags rather than flasks. NK cells isolated
lytic capability of NK cells incubated with 5, 50 or 500 IU/ from four normal donors and a sarcoma patient who had
mL IL-2 or without IL-2 against K562 and RCC target previously undergone an autologous transplant were co-
cells was determined by 51Cr-release assays. Cytokine cultured with EBV-TM-LCL feeder cells. The total yield
secretion was measured in culture supernatants with a of NK cells in bags was comparable to yields obtained
Millipore kit or Quantikine† ELISA as above. when NK cells were grown in flasks (Figure 1C).
Treatment of tumor cells with bortezomib Phenotype of resting versus expanded NK cells
RCC cells were seeded into 10-cm2 tissue culture dishes in We next evaluated phenotypic changes associated with
12 mL culture medium; 24 h later 10 nM bortezomib was expanding NK cells in vitro. Resting and expanded NK cells
added. After 16Á18 h, RCC cells were trypsinized, washed were analyzed by flow cytometry at baseline and ]10 days
in DMEM and used in cytotoxicity assays. following in vitro expansion.

6. Expansion of NK cells for cancer immunotherapy 345
cells did not express TRAIL, FasL or NKG2C, while
NKG2D, LFA-1, CD244, CD48, perforin and granzymes A
and B were constitutively expressed. CD25 expression
varied amongst donors but was typically low or absent on
resting NK cells.
NK cells obtained from nine different donors and
expanded over 10Á22 days had a mean expression of
CD56'CD16' and CD56'CD16Á of 84.397.8% (range
66.5Á97.5%) and 14.797.7% (range 2.1Á31.9%), respec-
tively, and did not contain CD56 ( CD16 ' populations.
After expansion, there was a substantial increase in NK-
cell surface expression of CD56, TRAIL, NKG2D, CD48
and CD25; on expanded versus resting NK cells from three
different donors, CD56 expression increased from a
median 85.393.4% to 99.390.3% [mean CD56 fluores-
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
cence intensity (MFI) increased from 70.4939.9 to
470.6966.6], TRAIL expression increased from a median
0.690.4% to 80.8915.4% (mean TRAIL MFI increased
from 6.095.1 to 37.993.2), NKG2D surface expression
increased from a median MFI of 48.3916.3 to 432.0970.9,
CD48 surface expression increased from a median MFI of
36.999.1 to 121.0938.8, and CD25 expression increased
from a median 2.391.6% to 48.6919.7% (mean CD25
MFI increased from 4.891.8 to 20.796.5). The expression
of perforin did not change, although there was a small but
consistent increase in surface expression of LFA-1, FasL,
NKG2C, CD244 and intracellular expression of granzymes
A and B, respectively (Figure 2A).
Surface expression of the NK-cell inhibitory receptor
CD158b increased in expanded NK cells; compared with
fresh NK cells, the MFI of CD158b increased 1.790.4 and
Figure 1. Expansion kinetics of NK cells grown ex vivo under 3.790.0 fold in NK cells expanded for 10 and 22 days,
various conditions. NK cells were isolated from PBMC by immuno- respectively. The MFI of NKG2A and KIR3DL1 remained
magnetic bead selection of CD56 ' CD3( cells. Irradiated PBMC unchanged, although the percentage of expanded NK cells
were used as feeder cells for expansion of NK cells from four healthy expressing NKG2A increased 3.791.8 fold (Figure 2B).
donors in flasks (NK1772 and NK1257) and Baxter bags (NK0772
and NK0155) (A). NK cells from five healthy donors were co-cultured
with irradiated EBV-TM-LCL cells in flasks at a 20:1 feeder to NK Cytotoxic function of expanded NK cells
cell ratio (B). NK cells grown in Baxter bags in the presence of EBV- We next evaluated the lytic effects of expanded versus
TM-LCL feeder cells at a 10:1 feeder to NK cell ratio at DTM under resting non-expanded NK cells against K562 and RCC cell
GMP conditions (C). lines. NK cells expanded in culture from 10 to 21 days
consistently demonstrated increased cytotoxicity against
K562 and RCC cells compared with resting NK cells
NK cells enriched from PBMC by immunomagnetic (Figure 3A). At a 1:1 effector to target ratio, lysis of RCC
bead selection contained 1Á30% monocytes, B1% CD3 ' cells was significantly higher with expanded NK cells
T cells, no CD56'/CD3' cells, no CD19' B cells and (27.699.3%) compared with resting NK cells (3.492.1%)
70Á92% CD56 '/CD3Á NK cells. Resting CD56 ' NK (P 0 0.005).

7. Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
346
M. Berg et al.
Figure 2 (Continued)

8. Expansion of NK cells for cancer immunotherapy 347
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
Figure 3. Specific lysis of K562 and RCC cell lines by resting non-expanded versus expanded NK cells and the effect of bortezomib on NK-cell
cytolytic function. 51Cr-release assays were performed using freshly isolated, cryopreserved and thawed NK cells in parallel with 12-day expanded
cells from the same donor at the indicated effector-to-target ratios (E:T). Experimental results for two of three donors are shown as mean9 SD (A).
RCC tumor cells were treated with 10 nM bortezomib for 16 h or left untreated. Percentage specific lysis of tumor cells by freshly isolated or
expanded NK cells at a 1:1 NK to target cell ratio from three donors was determined in a 5-h 51Cr-release assay (B).
Treatment of RCC cells with proteasome inhibitor resting and expanded NK cells was augmented by pre-
bortezomib has previously been shown to up-regulate treating tumor cells with bortezomib for 16 h. However, in
surface expression of the TRAIL death receptor DR5 contrast to resting NK cells, there was a dramatic increase
(TRAIL-R2), which sensitizes tumors to NK-cell cyto- in bortezomib-treated tumor killing by expanded NK cells;
toxicity [6Á8]. Therefore, we compared lysis by resting at a 1:1 effector to target ratio, resting NK cells lysed 3.49
versus expanded NK cells against bortezomib-treated 2.1% and 5.092.7% (P 00.44, unpaired t-test) of un-
versus untreated RCC cells. Lysis of RCC cells by both treated and bortezomib-treated RCC tumor cells, respec-
Figure 2. Flow cytometry analysis of freshly isolated resting and expanded NK cells. NK cells were stained with the indicated MAb immediately
after isolation and after 12 days of expansion. Cell-surface expression on viable cells is shown. Data from a representative experiment from one of
three donors are shown. The shaded areas represent negative isotype controls.

9. Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
348
M. Berg et al.
Figure 4 (Continued)

10. Expansion of NK cells for cancer immunotherapy 349
tively, compared with NK cells expanded for 12Á18 days, (68.3923.5% to 26.3913.1%) and in TRAIL MFI
which killed 27.699.3% and 55.898.3% (P00.001, (52.2924.0 to 18.492.9) within 16Á24 h of IL-2 removal
unpaired t-test) of untreated versus bortezomib-treated from the medium. TRAIL expression was restored by
RCC tumor cells, respectively (Figure 3B). subsequent addition of IL-2 back into the medium, and was
IL-2 dose dependent (data not shown). Similar to TRAIL,
Cytokine secretion profiles of NK cells the MFI of NKG2D expression in expanded NK cells
We then compared cytokine secretion profiles of freshly declined significantly (2.190.2 fold) 24 h following IL-2
isolated versus expanded NK cells. Resting cells sponta- removal from the medium. After the addition of IL-2,
neously produced very low levels of TRAIL, macrophage NKG2D expression was restored in a dose-dependent
inflammatory protein (MIP)-1a and MIP-1b, and high manner.
amounts of Il-1 receptor antagonist (IL-1ra). Co-culture Reductions and subsequent increases in TRAIL and
with K562 target cells for 5 h in the absence of IL-2 induced NKG2D surface expression that occurred with the removal
NK-cell secretion of TNF-a, interferon (IFN)-g, gran- and addition of IL-2 directly correlated with NK-cell
ulocyteÁmacrophage colony-stimulating factor (GM-CSF), cytotoxicity against K562 and RCC target cells (Figure 5A).
FasL, MIP-1a, MIP-1b and IL-1ra but not TRAIL. NK cells Culturing previously expanded NK cells in media contain-
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
expanded for 14 days spontaneously secreted IL-2, IFN-g, ing no or low doses of IL-2 (0Á5 IU/mL IL-2) for 24 h
GM-CSF, FasL, TRAIL, MIP-1a and MIP-1b but not IL- resulted in a substantial decline in NK-cell cytotoxicity
1ra (the only cytokine secreted by resting but not expanded against K562 and RCC target cells compared with cultures
NK cells). With the exception of IL-2 and TRAIL, the containing 50Á500 IU/mL IL-2 where cytotoxicity was
secretion of the above cytokines was augmented by co- maintained. Likewise, spontaneous secretion of FasL and
culturing expanded NK cells with K562 and RCC target TRAIL and multiple cytokines, including GM-CSF, TNF-
cells (Figure 4A, B). RCC cells pretreated with bortezomib a and IFN-g, was also IL-2 dose dependent, declining
stimulated NK cells to produce higher levels of TNF-a, rapidly in cultures in which the concentration of IL-2 was
whereas the secretion of other cytokines remained un- decreased or where IL-2 was removed (Figure 5B).
changed (Figure 4B). There was no spontaneous TNF-a Expanded NK cells did not secrete IL-1a, IL-1b, IL-10,
release from K562 and RCC cells. In these experimental G-CSF and TNF-b regardless of IL-2 content in culture
conditions, neither resting nor expanded NK cells produced medium. In one of four donors, IL-13 was detected (260Á
IL-1a, IL-1b, TNF-b, IL-10, G-CSF and IL-13. 280 pg/mL) in cultures of expanded NK cells when 50 and
500 U/mL IL-2 were added for 24 h (data not shown).
The effect of IL-2 deprivation on expanded
NK cells The effect of cryopreservation on phenotype and
Whether exogenous IL-2 would be required to support function of expanded NK cells
NK-cell cytotoxicity and proliferation following adoptive In order to assess the impact of cryopreservation, the
NK-cell infusions in humans is unclear. Thus, we phenotype and cytolytic function against K562 and RCC
evaluated the effects of IL-2 withdrawal and add-back on cells of expanded versus cryopreserved NK cells were
the phenotype and function of expanded NK cells. TRAIL compared. Lysis of untreated and bortezomib-treated RCC
expression on expanded NK cells declined rapidly in cells by expanded thawed NK cells was significantly lower
association with IL-2 deprivation; there was a decline in compared with lysis by non-frozen expanded NK cells
both the percentage of NK cells expressing TRAIL (Figure 6A). Lysis of K562 cells by thawed NK cells was
Figure 4. Cytokine secretion profile of NK cells after co-culture with K562 and RCC target cells. Resting NK cells or NK cells from two normal
donors, expanded for 14 days, were cultured for 5 h in 96-well plates in 200 mL NK-cell growth medium (no IL-2) in triplicate either without
target cells or with 104 K562 at a 10:1 NK to target cell ratio (A). Expanded NK cells were cultured as above with K562 or RCC cells, which were
untreated or treated with 10 nM bortezomib for 16 h. Representative data for one of three donors are shown (B). Cytokine content in cell-free
supernatants was measured with both the Beadlyte† Human Multi-Cytokine BeadmasterTM Kit and R&D Quantikine ELISA. ELISA data are
shown for FasL, MIP-1a and MIP-1b; Luminex data are shown for other cytokines.

11. 350 M. Berg et al.
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
Figure 5. Cytotoxicity of expanded NK cells and their cytokine production after IL-2 deprivation. IL-2 dose-dependent lysis of K562 and RCC
cells with and without bortezomib treatment. The percentage specific lysis of target cells was determined in a 5-h 51Cr-release assay. Data for NK
cells from two normal donors are presented as mean9 SD (A). Cytokine secretion and TRAIL and sFasL release by expanded NK cells from one of
three donors after culture in medium without IL-2 or with varying doses of IL-2. NK cells were expanded for 12 days, washed twice in X-VIVO 20
medium, and incubated at 106 cells/mL for 24 h. Cell-free culture supernatants were harvested and assayed for cytokine secretion as described for
Figure 4 (B).

12. Expansion of NK cells for cancer immunotherapy 351
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
Figure 6. Cytolytic function of expanded NK cells after cryopreservation and thawing; correlation with CD56/CD16, TRAIL and NKG2D
expression. Expanded NK cells were cryopreserved then subsequently thawed and analyzed in parallel with cells maintained in culture or thawed
cells incubated for 16 h in medium with 500 U/mL IL-2. Chromium release assay data for a 1:1 ratio of NK cells from two normal donors to K562
cells and untreated or bortezomib-treated RCC target cells are presented as mean 9 SD (A). Flow cytometry analysis of expanded NK cells. Cell
viability was assessed by trypan blue stain exclusion and 7AAD staining. Representative flow cytometry data for NK cells from one of the above
donors are shown. Dot-plots show the percentage of viable cells and the gates, and numbers in histograms represent the MFI (B).

13. 352 M. Berg et al.
also diminished, although this effect was only evident at 1:1 achieves substantial NK expansions, in the range of
and 0.5:1 effector to target ratios (data not shown). 250Á850-fold, over a 2Á3-week interval. With a starting
Decreased cytotoxicity of thawed NK cells against tumor population of 200 million immunomagnetic bead-purified
targets correlated with their reduced surface expression of CD3Á CD56 ' NK cells isolated after a typical 15-L
both TRAIL and NKG2D, together with an increase in the apheresis, this expansion technique would achieve a final
percentage of cells that were either negative or had NK-cell product in the range of 3 )1010 cells, a number
dim expression of CD16 (Figure 6B). Expanded NK that would seem sufficient for phase I studies. To address
cells maintained in culture contained 89.292.7% the safety of using EBV-LCL cells for NK-cell expansion,
CD56 ' CD16 ' (88.099.6% co-expressed TRAIL) and three expanded NK-cell products were tested by in situ
7.490.3% CD56 ' CD16Á (62.4910.9% co-expressed hybridization for EBV-encoded early small RNAs (EBER)
TRAIL), while thawed cells were 57.9924.6% and were all found to be negative. The TM-LCL feeder
CD56 ' CD16 ' double-positive and 35.4920.6% cell line used here to expand NK cells has previously been
CD56 ' CD16Á, with only 27.794.9% double-positive used by others to expand T-cell lines in vitro utilizing
cells co-expressing TRAIL. Incubation of thawed cells in GMP-compliant components [18,19]. To avoid expanding
medium containing 500 IU/mL IL-2 for 6 h increased NK T cells that proliferate rapidly under these culture
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
cytolytic function and surface expression of NKG2D and conditions (data not shown), a two-step CD3 ' T-cell
TRAIL to about 50% of baseline (data not shown), while a depletion followed by a CD56' selection was used to
16-h treatment with IL-2 restored NKG2D and TRAIL enrich for an NK-cell population that typically had
and cytotoxicity to levels seen with non-frozen cells B0.5% T-cell contamination.
(Figure 6A, B). Although the addition of IL-2 to medium The most efficient large-scale NK-cell expansions were
was able to restore NK-cell cytotoxicity, the viability of achieved when cells were cultured in Baxter Lifecell bags.
thawed NK cells (assessed by 7AAD staining) declined In contrast, cultures generated in Teflon-coated bags
from 93Á97% immediately after thawing to 38Á50% at resulted in relatively limited NK-cell expansions (data
16 h. This decline in thawed NK-cell viability did not not shown). The viability and expansion rates of NK cells
correlate with the time NK cells were maintained in were at their greatest 9Á15 days following the initiation of
culture prior to cryopreservation. These results suggest cell cultures and declined after 21 days. Several attempts to
that expanded NK cells that have been cryopreserved may re-expand NK cells with EBV-LCL feeder cells after cells
require culturing in IL-2-containing medium following had been cultured for ]14 days were mostly unsuccessful.
thawing to restore function prior to infusion in patients. Regardless of culture vessels used for expansions of NK
However, the substantial decline in viability of thawed NK cells, the phenotype and lytic activity of the expanded cells
cells rescued with IL-2-containing medium highlights the were similar.
limitation of this approach. Although NK cells can be activated by IL-2, IL-2 alone
fails to expand NK cells in vitro. In contrast, NK cells
Discussion stimulated with EBV-LCL feeders expanded dramatically,
Although there has been increased interest in exploring had an activated phenotype and, as a consequence of up-
the anti-tumor effects of adoptively infused NK cells in regulated expression of NKG2C, NKG2D, FasL, TRAIL
cancer patients, the small number of cells isolated follow- and granzymes A and B, were significantly more cytotoxic
ing a typical apheresis procedure has precluded trials against tumor cells compared with fresh NK cells. We also
assessing a relationship between NK-cell dose and tumor observed that expression of CD244 (2B4) and CD48 was
response. We present a functionally closed in vitro system augmented in expanded compared with resting NK cells.
using irradiated EBV-LCL feeder cells resulting in large- The function of CD48 and CD244 on expanded human
scale expansion of highly cytotoxic clinical-grade NK cells. NK cells is not entirely understood. Although one study
In contrast to NK-cell expansion protocols that require reported increased expression of CD244 could have an
culturing in plastic flasks and multiple rounds of stimula- inhibitory effect on the function of NK cells [20], murine
tion with feeder cells, the expansion technique presented data have shown that homotypic interactions between
here utilizes sterile bags, requires only a single round of these molecules prevent fratricide and enhance NK-cell
stimulation with irradiated EBV-LCL feeder cells and expansion and cytolytic activity [21].

14. Expansion of NK cells for cancer immunotherapy 353
Compared with non-expanded NK cells, expanded NK NKG2A expression remained unchanged or increased
cells secreted, either spontaneously or following co-culture slightly following NK-cell expansion.
with tumor targets (K562 and RCC cells), higher levels of The changes in phenotype and maintenance of cyto-
IFN-g, IL-2, FasL and TRAIL. In contrast, non-expanded toxicity against tumor cells by expanded NK cells were
NK cells secreted higher levels of IL-1ra, which was not dependent on IL-2. Withdrawal of IL-2 from expanded
produced by expanded cells. An unexpected and pre- NK-cell populations rapidly resulted in substantial reduc-
viously unobserved finding was that TNF-a secretion tions in NK-cell killing of tumor cells. Whether the
increased when NK cells were co-cultured with bortezo- exogenous administration of IL-2 would be required to
mib-treated RCC cells. The biologic significance of this maintain high levels of NKG2D, TRAIL and tumor
finding is unknown, although TNF-a can be directly cytotoxicity in vivo of adoptively infused expanded NK
cytotoxic to tumor cells and can have a positive immunor- cells is currently being investigated in an animal model.
egulatory function, inducing dendritic cell (DC) matura- The ability to cryopreserve and subsequently thaw NK
tion, activation and Ag cross-presentation, resulting in cells while maintaining their cytolytic activity could
augmented T-cell cytokine secretion [22,23]. In contrast to logistically facilitate clinical trials evaluating multiple
previous reports, only very low levels of IL-10 were
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
rounds of adoptive NK-cell infusions. Although expanded
detected in expanded NK cells cultured in IL-2. In NK cells that were frozen then subsequently thawed
contrast to when IL-12 is combined with IL-2, IL-2 alone maintained high viability, their cytolytic capacity was
is a weak stimulator of IL-10 secretion. IL-10 has been substantially lower than that of expanded NK cells that
shown to have anti-inflammatory effects, inhibiting macro-
had never undergone cryopreservation. Thawed NK cells
phage and DC activation and maturation and secretion of
had lower surface expression of TRAIL and NKG2D and
multiple pro-inflammatory cytokines [24,25]. Therefore,
were more likely to contain populations that were dim or
lack of IL-10 secretion would seem desirable when
negative for CD16. These findings suggest thawed adop-
expanded NK cells are used in the context of tumor
tively infused NK cells might have reduced cytotoxic
immunotherapy.
potential compared with expanded NK cells that are
The net effect of changes in NK-cell phenotype and
maintained fresh in culture. Importantly, the cytotoxicity
cytokine secretion resulted in expanded NK cells having
of expanded NK cells that were frozen then thawed could
markedly higher levels of cytotoxicity against tumor cells
be rescued by culturing in IL-2-containing medium for 16
compared with non-expanded cells. Although it is likely
h, although the overall viability of these populations was
that an increase in expression of activating receptors and
lower than that of non-thawed cells.
molecules that induce tumor apoptosis (TRAIL, FasL,
In conclusion, we describe a method for the large-scale
granzyme B, etc.) in expanded NK cells contributed to
production of in vitro-expanded NK cells using irradiated
their enhanced cytotoxicity, blocking experiments
to define the exact contribution of individual pathways EBV-LCL feeder cells and a functionally closed ‘bag-
to augmented NK-cell cytolytic function were not per- based’ culture system. In vitro-expanded NK cells had
formed in this analysis. altered cytokine secretion profiles, were phenotypically
Previously, we and others have shown that the proteo- distinct from non-expanded NK cells and were signifi-
some inhibitor bortezomib enhances TRAIL-mediated cantly more cytotoxic to tumor cells. Expanded cells had
cytotoxicity against tumor cells in vitro [26,27] and in vivo increased surface expression of the NKG2D and TRAIL
[6Á8]. In experiments conducted in this study, we observed and greatly enhanced TRAIL-mediated cytotoxicity
that lysis of bortezomib-treated RCC tumors was dramati- against bortezomib-treated tumors compared with non-
cally higher with expanded compared with resting NK cells, expanded NK cells. Based on these findings, a phase I trial
providing strong evidence that increased surface expression has recently been initiated in patients with advanced
of TRAIL on expanded NK cells substantially augmented metastatic tumors and hematologic malignancies to in-
their tumor lysis at least in part via TRAIL apoptotic vestigate the safety and anti-tumor effects of escalating
pathways. In contrast, it is unlikely that changes in NK-cell doses of adoptively infused ex vivo-expanded autologous
inhibitory receptors played any role in augmenting NK-cell NK cells. NK-cell doses in this trial will range from 5 )
cytotoxicity, as CD158b/KIR2DL2/3, KIR3DL1 and 106 to 108 NK cells/kg and will be given every 3 weeks

15. 354 M. Berg et al.
following treatment with bortezomib and concomitant 8 Sayers TJ, Brooks AD, Koh CY, Ma W, Seki N, Raziuddin A et al.
with IL-2 administration. The proteasome inhibitor PS-341 sensitizes neoplastic cells to
TRAIL-mediated apoptosis by reducing levels of c-FLIP. Blood
2003;102:303Á10.
Acknowledgements 9 Ljunggren HG, Malmberg KJ. Prospects for the use of NK cells
This research was supported by the intramural research in immunotherapy of human cancer. Nat Rev Immunol
program of NIH, National Heart, Lung, and Blood 2007;7:329Á39.
Institute, Hematology Branch. We wish to acknowledge 10 Passweg JR, Koehl U, Uharek L, Meyer-Monard S, Tichelli A.
ACKC (Action to Cure Kidney Cancer) and The Dean R. Natural-killer-cell-based treatment in haematopoietic stem-cell
transplantation. Best Pract Res Clin Haematol 2006;19:811Á24.
O’Neill Memorial Fellowship for generous contributions
11 Carlens S, Gilljam M, Chambers BJ, Aschan J, Guven H,
supporting this research. The authors would also like to Ljunggren H-G et al. A new method for in vitro expansion of
thank Dr E. J. Read, Dr David Stroncek, Dr Hanh Khuu, cytotoxic human CD3 ( CD56 ' natural killer cells. Hum
Vicki Fellows and Virginia David-Ocampo from the Immunol 2001;62:1092Á8.
Department of Transfusion Medicine in NIH for their 12 McKenna DH Jr, Sumstad D, Bostrom N, Kadidlo DM, Fautsch
valuable contribution to the development of clinical-grade S, McNearney S et al. Good manufacturing practices production
of natural killer cells for immunotherapy: a six-year single-
NK-cell expansion protocols, Dr Stefania Pittaluga (NIH/
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009
institution experience. Transfusion 2007;47:520Á8.
NCI) for performing EBER testing of NK cells, and Drs 13 Luhm J, Brand JM, Koritke P, Hoppner M, Kirchner H, Frohn C.
Shelly Heimfeld, Brenda Sandmaier and Kimberly Boyt Large-scale generation of natural killer lymphocytes for clinical
from Fred Hutchinson Cancer Research Center. The application. J Hematother Stem Cell Res 2002;11:651Á7.
authors have no conflicting financial interests. 14 Imai C, Iwamoto S, Campana D. Genetic modification of
primary natural killer cells overcomes inhibitory signals and
induces specific killing of leukemic cells. Blood 2005;106:376Á83.
Declaration of interest: The authors report no conflicts of
15 Perussia B, Ramoni C, Anegon I, Cuturi MC, Faust J, Trinchieri
interest. The authors alone are responsible for the content G. Preferential proliferation of natural killer cells among
and writing of the paper. peripheral blood mononuclear cells cocultured with B lympho-
blastoid cell lines. Nat Immun Cell Growth Regul 1987;6:171Á88.
16 Rabinowich H, Sedlmayr P, Herberman RB, Whiteside TL.
Increased proliferation, lytic activity, and purity of human
References natural killer cells cocultured with mitogen-activated feeder
1 Robertson MJ, Ritz J. Biology and clinical relevance of human cells. Cell Immunol 1991;135:454Á70.
natural killer cells. Blood 1990;76:2421Á38. 17 Igarashi T, Wynberg J, Srinivasan R, Becknell B, McCoy P Jr,
2 Trinchieri G. Biology of natural killer cells. Adv Immunol Takahashi Y et al. Enhanced cytotoxicity of allogeneic NK cells
1989;47:187Á376. with killer immunoglobulin-like receptor ligand incompatibility
3 Perussia B, Trinchieri G, Jackson A, Warner NL, Faust J, against melanoma and renal cell carcinoma cells. Blood
Rumpold H et al. The Fc receptor for IgG on human natural 2004;104:170Á7.
killer cells: phenotypic, functional, and comparative studies with 18 Cooper LJ, Ausubel L, Gutierrez M, Stephan S, Shakeley R,
monoclonal antibodies. J Immunol 1984;133:180Á9. Olivares S et al. Manufacturing of gene-modified cytotoxic T
4 Farag SS, Fehniger TA, Ruggeri L, Velardi A, Caligiuri MA. lymphocytes for autologous cellular therapy for lymphoma.
Natural killer cell receptors: new biology and insights into the Cytotherapy 2006;8:105Á17.
graft-versus-leukemia effect. Blood 2002;100:1935Á47. 19 Wang J, Jensen M, Lin Y, Sui X, Chen E, Lindgren CG et al.
5 Karre K, Ljunggren HG, Piontek G, Kiessling R. Selective Optimizing adoptive polyclonal T cell immunotherapy of
rejection of H-2-deficient lymphoma variants suggests alter- lymphomas, using a chimeric T cell receptor possessing CD28
native immune defence strategy. Nature 1986;319:675Á8. and CD137 costimulatory domains. Hum Gene Ther 2007;18:712Á
6 Lundqvist A, Abrams SI, Schrump DS, Alvarez G, Suffredini D, 25.
Berg M et al. Bortezomib and depsipeptide sensitize tumors to 20 Chlewicki LK, Velikovsky CA, Balakrishnan V, Mariuzza RA,
tumor necrosis factor-related apoptosis-inducing ligand: a novel Kumar V. Molecular basis of the dual functions of 2B4 (CD244).
method to potentiate natural killer cell tumor cytotoxicity. J Immunol 2008;180:8159Á67.
Cancer Res 2006;66:7317Á25. 21 Lee KM, Forman JP, McNerney ME, Stepp S, Kuppireddi S,
7 Lundqvist A, Rao S, Smith A, Berg M, Su S, Yokoyama H et al. Guzior D et al. Requirement of homotypic NK-cell interactions
Bortezomib enhances the antitumor activity of adoptively- through 2B4 (CD244)/CD48 in the generation of NK effector
infused natural killer cells in vivo: a novel approach to override functions. Blood 2006;107:3181Á8.
KIR-mediated inhibition of NK cell cytotoxicity. ASH Annual 22 Cooper MA, Fehniger TA, Fuchs A, Colonna M, Caligiuri MA.
Meeting Abstracts. Blood 2007;110:530a. NK cell and DC interactions. Trends Immunol 2004;25:47Á52.

16. Expansion of NK cells for cancer immunotherapy 355
23 Sun K, Li M, Sayers TJ, Welniak LA, Murphy WJ. Differential 26 An J, Sun Y, Fisher M, Rettig MB. Antitumor effects of
effects of donor T-cell cytokines on outcome with continuous bortezomib (PS-341) on primary effusion lymphomas. Leukemia
bortezomib administration after allogeneic bone marrow 2004;18:1699Á704.
transplantation. Blood 2008;112:1522Á9. 27 Lashinger LM, Zhu K, Williams SA, Shrader M, Dinney CP,
24 Loza MJ, Perussia B. The IL-12 signature: NK cell terminal McConkey DJ. Bortezomib abolishes tumor necrosis factor-
CD56 ' high stage and effector functions. J Immunol 2004; related apoptosis-inducing ligand resistance via a p21-depen-
172:88Á96. dent mechanism in human bladder and prostate cancer cells.
25 Wolk K, Kunz S, Asadullah K, Sabat R. Cutting edge: immune Cancer Res 2005;65:4902Á8.
cells as sources and targets of the IL-10 family members?
J Immunol 2002;168:5397Á402.
Downloaded By: [University of Aberdeen] At: 11:47 21 July 2009