Embed Size (px)
Citation preview
R E S EARCH ART I C L E
CANCER
http://stm.sci
Dow
nloaded from
Cytokine-induced memory-like natural killer cellsexhibit enhanced responses against myeloid leukemiaRizwan Romee,1* Maximillian Rosario,1,2* Melissa M. Berrien-Elliott,1* Julia A. Wagner,1
Brea A. Jewell,1 Timothy Schappe,1 Jeffrey W. Leong,1 Sara Abdel-Latif,1 Stephanie E. Schneider,1
Sarah Willey,1 Carly C. Neal,1 Liyang Yu,3 Stephen T. Oh,3 Yi-Shan Lee,2 Arend Mulder,4
Frans Claas,4 Megan A. Cooper,5 Todd A. Fehniger1†
Natural killer (NK) cells are an emerging cellular immunotherapy for patients with acute myeloid leukemia (AML);however, the best approach to maximize NK cell antileukemia potential is unclear. Cytokine-induced memory-likeNK cells differentiate after a brief preactivation with interleukin-12 (IL-12), IL-15, and IL-18 and exhibit enhanced re-sponses to cytokine or activating receptor restimulation for weeks to months after preactivation. We hypothesizedthat memory-like NK cells exhibit enhanced antileukemia functionality. We demonstrated that human memory-likeNK cells have enhanced interferon-g production and cytotoxicity against leukemia cell lines or primary human AMLblasts in vitro. Usingmass cytometry, we found that memory-like NK cell functional responses were triggered againstprimary AML blasts, regardless of killer cell immunoglobulin-like receptor (KIR) to KIR-ligand interactions. In addition,multidimensional analyses identified distinct phenotypes of control and memory-like NK cells from the same indi-viduals. Humanmemory-like NK cells xenografted into mice substantially reduced AML burden in vivo and improvedoverall survival. In the context of a first-in-human phase 1 clinical trial, adoptively transferred memory-like NK cellsproliferated and expanded in AML patients and demonstrated robust responses against leukemia targets. Clinicalresponses were observed in five of nine evaluable patients, including four complete remissions. Thus, harnessingcytokine-induced memory-like NK cell responses represents a promising translational immunotherapy approachfor patients with AML.
ence
by guest on September 4, 2020
mag.org/
INTRODUCTION
Acute myeloid leukemia (AML) is a hematologic malignancy primarilyof older individuals that remains a substantial clinical challenge (1).Currently, less than 30%ofAMLpatients are curedwith standard thera-pies, and relapsed/refractory (rel/ref) AML patients who are not can-didates for hematopoietic cell transplantation (HCT) have a poorprognosis and no curative treatment options (2, 3). Cellular immuno-therapy mediated by alloreactive T and NK cells administered in thecontext of an allogeneic HCT is an effective treatment for AML; how-ever, most AMLpatients are not candidates for this procedure becauseit is associated with substantial treatment-related morbidity and mor-tality (4,5).Analternative approach thatprovides the immunotherapeuticbenefits of allogeneic HCTwithout severe toxicity is the adoptive transferof allogeneic lymphocytes that mediate the “graft versus leukemia” effect.This strategy may expand the option of cellular immunotherapy to mostAML patients.
Natural killer (NK) cells are innate lymphoid cells that are importantfor host defense against pathogens and mediate antitumor immune re-sponses (6,7).Major histocompatibility complex (MHC)–haploidenticalNK cells exhibit antileukemia responses without causing “graft versushost disease” (GVHD) afterHCT (8), providing evidence of their utilityas a cellular effector for leukemia patients. Allogeneic NK cell adoptive
1DivisionofOncology, Department ofMedicine,WashingtonUniversity School ofMedicine,St. Louis,MO63110, USA. 2Department of Pathology,WashingtonUniversity School ofMed-icine, St. Louis, MO 63110, USA. 3Division of Hematology, Department of Medicine, Wash-ington University School of Medicine, St. Louis, MO 63110, USA. 4Department ofImmunohematology and Blood Transfusion, Leiden University Medical Center, 2333 ZCLeiden, Netherlands. 5Division of Rheumatology, Department of Pediatrics, WashingtonUniversity School of Medicine, St. Louis, MO 63110, USA.*These authors contributed equally to this work.†Corresponding author. Email: [email protected]
www.ScienceTra
transfer is safe and can induce remissions in patients with leukemia(9–12); however, these studies have been limited by inadequate per-sistence, expansion, and in vivo antileukemia activity of the adop-tively transferred NK cells. Thus, one key barrier in the field is theneed for biology-driven approaches to enhance NK cell antitumorfunctionality before adoptive transfer.
Although NK cells have traditionally been considered members ofthe innate immune system, paradigm-shifting studies in mice haveidentified memory-like properties after hapten exposure, virus infec-tion, or combined interleukin-12 (IL-12), IL-15, and IL-18 cytokine pre-activation (13, 14). Cytokine-induced memory-like NK cells weredefined by briefly preactivating murine NK cells with IL-12, IL-15, andIL-18, followed by adoptive transfer into syngeneic mice. Weeks tomonths later, memory-like NK cells had proliferated and exhibitedenhanced restimulation responses to cytokines or triggering via activat-ing receptors (15, 16). This preactivation approach also resulted in anti-tumor responses to murine NK cell–sensitive cell lines after adoptivetransfer inmice (17). The potential translation of these findings as im-munotherapy was established by the identification of human IL-12,IL-15, and IL-18–induced memory-like NK cells (18). Key propertiesof human memory-like NK cells include enhanced proliferation, ex-pression of the high-affinity IL-2 receptorabg (IL-2Rabg), and increasedinterferon-g (IFN-g) production after restimulation with cytokines or viaactivating receptors (19, 20). However, the ability of humanmemory-likeNK cells to respond to cancer target cells has not been extensively re-ported. We hypothesized that human cytokine-induced memory-likeNK cells exhibit enhanced antileukemia properties. This was testedin vitro against primary AML blasts and in vivo in nonobese diabetic(NOD)/severe combined immunodeficient (SCID)/common gammachain−/− (gc−/−) (NSG) mouse xenograft models and in AML patients
nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 1
R E S EARCH ART I C L E
who were administered memory-like NK cells as part of a first-in-humanclinical trial. Our results demonstrate that combined preactivation withIL-12, IL-15, and IL-18 differentiates cytokine-induced memory-likeNK cells that have potent antileukemia functionality in vitro and in vivoand thus represent a promising immunotherapy strategy forAMLpatients.
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
RESULTS
Memory-like NK cells exhibitenhanced functional responsesagainst leukemia target cellsAlthough the enhanced recall response ofhuman IL-12, IL-15, and IL-18–inducedmemory-likeNKcellswaswell establishedafter cytokine receptor restimulation (18),their response to leukemia target cells, in-cluding primary AML blasts, was not ex-tensively studied. To address this issue, wepreactivated human NK cells with IL-12,IL-15, and IL-18 or control conditions(low-dose IL-15 only) overnight and thendifferentiated memory-like or control NKcells from the same individual for 7 days(Fig. 1A).Memory-like NK cells exhibitedenhanced IFN-g production in responseto restimulation with both K562 leukemiacells (Fig. 1, A and B) and primary AMLblasts (Fig. 1, A and C). In addition,memory-like NK cells demonstrated sig-nificantly higher cytotoxicity comparedto control NK cells from the same donors,which was dependent on NKG2D andDNAM-1 (P = 0.02; Fig. 1D and fig. S1).Consistent with augmented cytotoxicity,we observed increased expression of thecytotoxic effector protein granzyme B(Fig. 1, E and F). To further define the dif-ferences between memory-like and con-trol NK cells, we established a masscytometry panel that included functional-ly relevant proteins to deeply immuno-phenotype NK cells.
Mass cytometry defines thedifferences between memory-likeand control NK cellsMass cytometry allows for high-throughputanalysis of a large number of parameterson single cells, and it has been used todeeply immunophenotype and track thediversity of human NK cells (21, 22). Wedeveloped custom mass cytometry panelsthat include NK cell lineage, maturation,receptor repertoire, and functional capac-ity (tables S1 and S2), and using the viSNEclustering and visualization strategy (23),we identified the differences between con-
www.ScienceTra
trol andmemory-like NK cells (Fig. 2, A and B, and figs. S2 and S3). Con-trol and memory-like NK cells were localized within separate areas ofthe viSNEmap, indicative of distinct cellular islands (Fig. 2A).We con-firmed significant differences in the expression of a number of markersthat were previously reported (18, 19), including CD94 (P = 0.002),
0
5
10
15
20
252550
0
5
10102540
0
5
10
15
20
252550
CD56+ CD56bright CD56dim
P = 0.0003 P = 0.0002P < 0.0001
0
10
20
304060
0
20
40
60
Cntl Mem Cntl Mem Cntl Mem Cntl Mem Cntl Mem Cntl Mem0
20
40
60
CD56+ CD56bright CD56dim
P = 0.0005 P = 0.0005P = 0.0003
A
B C
D
0.0 2.5 5.0 7.5 10.00
20
40
60
80
CntlMem
P = 0.02
Effector /target ratio
Per
cent
spe
cific
kill
ing
Pre activated
Control
Wash 16 hours 1 week
IL-15
NK
IL-12+ IL-15+ IL-18
CD
56
IFN-γ
Leukemia targets
Memory-like
6 hours Control
26.4
Memory-like
11.1
Control
38.2
Memory-like
9.17
Control
K562 Primary AML
Cntl Mem Cntl Mem0
10
20
30P = 0.005 P = 0.007
CD56bright CD56dimG
zmB
MF
IF E
GzmB
Eve
nts
CD56bright CD56dim
Cntl Mem Cntl Mem
Per
cent
IFN
-γ–p
ositi
ve
6 hours
Per
cent
IFN
-γ–p
ositi
ve
Fig. 1. Memory-like NK cells exhibit enhanced functional responses against leukemia targets. Puri-fied NK cells were preactivated with IL-12, IL-15, and IL-18 or control (cntl; IL-15) for 16 hours, washed, and
then rested in low concentrations of IL-15 to allow for differentiation. (A) Schema of memory-like (mem) NKcell in vitro experiments and representative flow plots showing enhanced IFN-g production by NK cellsafter K562 (left) and primary AML (right) triggering. Inset numbers are the percentages of IFN-g–positiveNK cells within the indicated regions. (B and C) Summary of data showing enhanced IFN-g production byall memory-like NK cells (CD56+) as well as each of the two major human NK cell subsets (CD56dim andCD56bright) restimulated with K562 (B) or primary allogeneic AML blasts (C). (D) Increased killing of K562leukemia target cells by purified memory-like NK cells as compared to control NK cells from the sameindividuals. (E) Representative flow cytometry data showing increased granzyme B (GzmB) protein inmemory-like compared to in control NK cells. (F) Summary of data from (E) showing granzyme B medianfluorescence intensity (MFI). Data represent two to six independent experiments and were comparedusing Wilcoxon signed-rank test with means ± SEM displayed in all graphs.nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 2
R E S EARCH ART I C L E
NKG2A(P=0.004),NKp46 (P=0.04), andCD25 (P=0.03) on the viSNEgated control and memory-like populations. Leveraging mass cytome-try’s expanded view of NK cell proteins, we observed increases in NKp30,NKp44, CD62L, CD27, and TRAIL and in the cytotoxic molecules perforin
www.ScienceTra
andgranzymeBandalsodecreases inNKp80, compared to controls (Fig. 2Band fig. S3). Thus, memory-like and control NK cells were identifiable asdistinct populations when tracked with high NK cell marker dimensionalityby mass cytometry. In addition, mass cytometry confirmed flow cytometry
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
A C
B
D
0
2
4
6
Nve Cntl Mem
Div
ersi
tyin
dex
0
10
20
30
2DL1 2DL2/2DL3
3DL1 2DS4 2DL5
P = 0.02
Per
cent
KIR
-po s
itiveComposite
tSN
E2
tSNE1
Cntl
Mem
Cntl
Mem
Density ControlMemory-like
CD56* CD16* CD94*
tSN
E2
tSNE1
NKG2D* NKG2C*NKG2A*
Cntl
Mem
Cntl
Mem
Cntl
Mem
Cntl
Mem
Cntl
Mem
Cntl
Mem
CD62L* CD25* NKp30*
tSN
E2
tSNE1
NKp46* NKp80*NKp44*
Cntl
Mem
Cntl
Mem
Cntl
Mem
Cntl
Mem
Cntl
Mem
Cntl
Mem
CD56 CD16 CD94 NKG2D NKG2CNKG2A
CD62L CD25 NKp30 NKp46 NKp80NKp44
78%
70%
Med
ian
exp
ress
ion
(arc
s inh
)M
edi a
nex
pres
sion
(arc
sinh
)
Ki-67 TRAIL Perforin
tSN
E2
tSNE1
Granzyme B
Cntl
Mem
Cntl
Mem
Cntl
Mem
Cntl
Mem
Med
ian
expr
essi
on( a
r csi
nh)
ControlMemory-like
0
100
200
300
400
500P = 0.008
0
10
20
30
40P = 0.008
0
50
100
150
200
250 P = 0.002
0
100
200
300
400P = 0.004
0
5
10
15
20
P = 0.012
0.00
0.05
0.10
0.15
0.20P = 1.0
0
20
40
60
80 P = 0.004
0
10
20
30
40 P = 0.03
0
20
40
60
80
100 P = 0.004
0
5
10
15P = 0.008
0
10
20
30
40 P = 0.04
0
10
20
30 P = 0.01
0
500
1000
1500 P = 0.004
0.0
0.1
0.2
0.3 P = 0.06
0
100
200
P = 0.03
0
50
100 P = 0.004
Ki-67 TRAIL Perforin Granzyme B
MaxMin
Expression
Fig. 2. Multidimensionalanalyses define the differ-
ences between memory-likeand control NK cells. NK cellswere assessed at baseline or af-ter in vitro control and memory-like differentiation at day 7 (Fig.1A) for the expressionof 36mark-ers using mass cytometry. (A andB) Comparison of control andmemory-likeNKcells fromarepre-sentative healthy individual usingviSNE, clusteredon21NKcell phe-notypic markers. (A) Memory-like(orange) and control NK cell (blue)events overlaid in the tSNE1/tSNE2 fields (left) show their dif-ferential localization within theviSNEmap. Density plots of con-trol and memory-like NK cells inthe tSNE1/tSNE2 fields (right).Inset values indicate the fre-quency of cells that fall withinthe control or memory-like gate.(B) The composite of the con-trol and memory-like popula-tions from (A), displaying themedian expression of the indi-catedmarkers. Dataare represen-tative of nine individuals. Colorscale indicates the intensity ofexpression of each marker sig-nal. Minimum (min) and maxi-mum (max) correspond to the2nd and the 98th percentile val-ues for each indicated marker,respectively. For additional sum-marydata and statistical compar-isons, see fig. S3. (C) InhibitoryKIR receptor inverse Simpson di-versity index of baseline [naïve(nve)], control, andmemory-likeNK cells from nine individuals;box with whiskers displayingminimumtomaximum. (D) Sum-mary of data showing the per-cent positive of each indicatedKIR for control and memory-likeNKcells. All summarygraphsdis-play means ± SEM, unless other-wise indicated.Comparisonsweremade using Wilcoxon signed-rank test.nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 3
R E S EARCH ART I C L E
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
findings of increased cytotoxic effector proteins, which likely contribute tothe enhanced cytotoxicity ofmemory-likeNKcells against leukemia targets.
Memory-like NK cells respond to AML regardless ofinhibitory KIR to KIR-ligand interactionsBecause allogeneic NK cell responses to AML blasts are influencedby the interaction of the inhibitory killer cell immunoglobulin-likereceptors (KIRs) with their cognate human leukocyte antigen (HLA)ligands (24–26), we next examined whether memory-like differentiationaltered KIR diversity. We used mass cytometry to examine NK cells atbaseline (naïve) and after control or memory-like differentiation invitro at day 7 (table S1 and Fig. 1A). The inverse Simpson diversity indexof KIR expressed by naïve, control (day 7), and memory-like (day 7) NKcells was unaltered (Fig. 2C) (21, 27), demonstrating that inhibitoryKIR phenotypes present at baseline were stable in vitro. Consistentwith this, the percent of KIR-positive NK cells was similar at day 7between control and memory-like NK cell populations (Fig. 2D).
In the setting of allogeneicHCT, donorKIR repertoire affects patientprognosis, highlighting the importance of the KIR to KIR-ligand mis-match for conventional NK cell anti-AML activity (24–26, 28). Toexperimentally identify NK cell subsets that preferentially respondto primary AML blasts, we restimulated control and memory-likeNK cells for 6 hours with primary AML blasts and assessed the ex-pression of 36 parameters via mass cytometry (tables S2 to S4).Memory-like NK cells again produced more IFN-g when triggeredwith primary AML blasts compared to controls (Fig. 3, A and B).Weused spanning-tree progression analysis of density-normalized events(SPADE) to cluster the NK cells according to the expression of coreNK cell markers and assessed each node (representing a distinct NK cellspecificity) for IFN-g expression (Fig. 3C) (29, 30). The KIR-ligand in-teraction status of each node was defined to investigate their impact onmemory-like and control NK cell responses. For a representative indi-vidual, we assigned the SPADE nodes as matched if they expressedKIR2DL2/KIR2DL3, which recognizes HLA-C1 present on the AMLblasts. KIR2DL1 and KIR3DL1 did not have ligands present, and nodesthat lackedKIR2DL2/KIR2DL3were categorized asmismatched. Basedon this approach, we observed the expected increased response in KIR-mismatched versus KIR-matched node in IFN-g production in thecontrol cells when triggered with AML blasts (Fig. 3, D to F). However,memory-like NK cells produced increased IFN-g compared to controlcells in both the KIR-ligand–matched and KIR-ligand–mismatchedsituations. Moreover, the IFN-g production by memory-like NK cellswas similar in the KIR-matched and KIR-mismatched setting (Fig. 3,D and E), which was reproducible across the seven individuals tested(Fig. 3E). Furthermore, when we expanded our analysis to four dif-ferent primary AML blasts, we observed a consistent increase inIFN-g production in KIR-matched memory-like NK cells comparedto KIR-mismatched control-treated cells (Fig. 3F). Because mostmemory-like NK cells express NKG2A, we assessed IFN-g productionby NKG2A− or NKG2A+ KIR-ligand–matched or KIR-ligand–mismatched NK cells and observed increased IFN-g production bymemory-like NK cells compared to controls in both subpopulations(figs. S4, A to C). Cell surface CD107a, tumor necrosis factor, and mac-rophage inflammatory protein-1a were also increased on memory-like NK cells in these assays, compared to controls (fig. S4D). Thus,memory-like differentiation enhances IFN-g production in response toprimary AML, regardless of whether memory-like NK cells expressedinhibitory KIR with a ligand present on the target AML blast.
www.ScienceTra
Memory-like NK cells exhibit potent antileukemiafunctionality after transfer into NSG miceTo assess whether humanmemory-likeNK cellsmaintainmemory-likefunction in vivo, we used an adoptive transfer model with NSG mice(Fig. 4A). After 7 days, memory-like NK cells were found to localizein key hematopoietic tissues, including bone marrow (BM), spleen,and blood (Fig. 4B), and were found in similar numbers compared tocontrolNKcells (Fig. 4C). The expressionof selected chemokine receptorswasminimally changed after IL-12, IL-15, and IL-18 pre-activation, withthe exception of reduced expression of CX3CR1 (fig. S7). Memory-likeNK cells also maintained their enhanced functionality as evidenced byincreased IFN-g production in response to ex vivo restimulation withK562 leukemia cells after 7 days in NSG mice (Fig. 4, D and E).
To test the ability of human memory-like NK cells to control leuke-mia in vivo, we engrafted K562-luc (luciferase-expressing) leukemiacells in NSG mice and adoptively transferred human IL-12, IL-15, andIL-18–preactivated or control NK cells into groups of NSG mice onday 4 (Fig. 4F).Memory-likeNK cells were significantlymore effectivein controlling K562 tumor cell growth in vivo using whole-body bio-luminescence imaging (BLI) (P < 0.001; Fig. 4, G andH). The enhancedleukemia control afforded by a single injection of human memory-likeNK cells also resulted in improved survival of these mice (Fig. 4I).
Donor memory-like NK cells proliferate and expand afteradoptive transfer into patients with AMLOn the basis of these preclinical findings with humanmemory-like NKcells and those frommouse models evaluating NK cell–sensitive tumorcell line challenge (17), we initiated a first-in-human phase 1 clinical trialof allogeneic, HLA-haploidentical, IL-12, IL-15, and IL-18–preactivatedNK cells in patients with rel/ref AML (fig. S5). Here, we report memory-likeNKcell biology and antileukemia activitywithin the first nine evaluablepatients treated at three different dose levels (Table 1).DonorNKcellswerepurified by CD3 depletion followed by CD56-positive selection (10), pre-activated for 12 to 16 hours with rhIL-12, rhIL-15, and rhIL-18 in a goodmanufacturing practice (GMP) laboratory, washed, and infused into AMLpatients who were preconditioned with fludarabine/cyclophosphamide (9)on day 0. After adoptive transfer, low-dose rhIL-2 was administered to sup-portmemory-likeNKcells throughtheir inducedhigh-affinity IL-2Rabg (19).
Three patients at dose level 1 (0.5 × 106/kg), three patients at doselevel 2 (1.0 × 106/kg), and three patients at dose level 3 (all NK cellsgenerated, capped at 10 × 106/kg) were evaluable. Donor memory-likeNK cells were tracked in the blood of all patients with informative HLA[using donor- or patient-specific anti-HLA monoclonal antibodies(mAbs)], peaked in frequency at 7 to 14 days after infusion and, asexpected (9), decreased in number after recipient T cell recovery (Fig.5, A to C). Memory-like NK cells comprised >90% of blood NK cellsat day 7 (Fig. 5, A and B), with an average of 419 ± 166–fold increase(range, 39 to 1270) comparing day 1 and day 7 counts (Fig. 5C). Donormemory-likeNKcells had increasedproliferation (Ki-67+) at days 3 and7(Fig. 5, D and E). Similarly, assessments of the BM at day 8 after infusionrevealed large percentages and absolute numbers of donor memory-likeNK cells (Fig. 5, F toH), although therewas heterogeneity between donorand recipient pairs, especially between NK cell dose levels (fig. S6).
Donor memory-like NK cells exhibit enhanced functionalityafter adoptive transfer into patients with AMLWe next assessed donor and recipient blood NK cell IFN-g productionfrom these patients after a short-term ex vivo restimulation triggered by
nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 4
R E S EARCH ART I C L E
K562 leukemia cells (Fig. 6, A to D). These analyses revealed an increasedfrequency of IFN-g–positive donor, compared to recipient, NK cells (Fig.6A). More strikingly, these experiments revealed that the absolute numberof donor IFN-g–producingNK cells wasmarkedly increased, compared tothe recipient NK cells in these samples (Fig. 6, B toD). Similarly, the num-ber of IFN-g–positive donor memory-like NK cells was greater than reci-pient NK cells in BM, although fewer patients had BM samples withadequate cell numbers for functional analyses (Fig. 6, E to G).
www.ScienceTra
Complete remissions were observed after IL-12, IL-15, andIL-18–preactivated donor NK cell infusion in patients withrel/ref AMLPerforming a phase 1 study of memory-like NK cell adoptive immuno-therapy in older patients with active rel/ref AML is a clinical challenge,but it does permit investigation of anti-AML responses. Thirteen pa-tients who had progressed after multiple previous treatments wereadministered with memory-like NK cells at dose levels 1, 2, and 3. Four
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
0
1
2
3
4
Matched Mismatched
P < 0.001 P < 0.001P < 0.001P = 0.001
P = 0.53
Cntl Mem0
10
20
30
40 P = 0.016BA
25.3
52.4
Cntl
CD
56
IFN-γ
Mem
19.0 12.5
50.9 40.3
5 61Donor ID:
0
1
2
3
4
Matched Mismatched
P < 0.0001
P < 0.0001
P < 0.0001
P < 0.0001P < 0.0001
C Control Memory-like
>5 cells>710 cells
Median IFN-γexpression (arcsinh)
D
E F
4737
34
20
49
22
50
2118
17
3846
36
4737
34
20
49
22
50
21
18
17
3846
3644 4333
42
7
19
1523
10
4443 33
42
7
19
15
23
10
16
113
14
142848
31
24
32
2
641
4539
12
2526
29
83540
1327
305
16
113
1
1428
48
31
24
32
641
45
39
12
2526
29
83540
1327
305
24
AML: 614 AML: 614, 895, 734, 925
ControlMemory-like
KIR /KIR-ligand: Mismatched Matched
0 4Median expression (arcsinh)
9 9
Per
cent
IFN
-γ–p
ositi
ve
Med
ian
IFN
-ex
pres
sion
Med
ian
IFN
expr
essi
on-
Fig. 3. Response of memory-like NK cells to primary AML
blasts is enhanced regardlessof KIR-ligand interactions. Con-trol and memory-like NK cellswere stimulated with primaryAML blasts for 6 hours and as-sessed for the expression of36 markers using mass cytome-try. (A) Representative bivariatemass cytometryplotsof IFN-g pro-duction by control and memory-like NK cells stimulated at thebulk population level. Numbersdepict percentages of cellswithinthe indicated regions. (B) Sum-mary of data (means ± SEM) fromseven individuals showing per-centages of IFN-g–positive NKcells. (C to E) NK cellswere furtheranalyzed on 21 clustering param-eters using SPADE. (C) Repre-sentative SPADE diagram of invitro–differentiated control andmemory-like NK cells from ahealthy individual. Node size de-picts relative number of cells pernode, and color indicatesmedianIFN-g expression for each node.Numbers next to each node rep-resent the node ID. (D) Heat mapof the nodes from (C). KIR to KIR-ligandmatched andmismatchedstatus was assigned on the basisof the presence or absence ofKIR2DL2/KIR2DL3, which recog-nizes HLA-C1 expressed by theprimaryAML. (E) Summary of datafrom seven individuals analyzedas in (C). (F) Summary of data fromfour to sevendifferent individualsstimulated with four differentAMLblasts showing reproducibil-ity of these findings. Control andmemory-likedatawere comparedusing the Wilcoxon signed-ranktest. Matched and mismatcheddata were compared using theMann-Whitney test.nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 5
R E S EARCH ART I C L E
patients were not evaluable because of inadequate donor cell collectioncaused by apheresis technical failure (n= 1) or death resulting frombac-teremia with septic shock before day 14 (n = 1) or before day 35 (n = 2).
www.ScienceTra
Of the nine evaluable patients (Table 1), there were four CR/CRi andone MLFS by the IWG response criteria (31), yielding an overall re-sponse rate of 55% and a CR/CRi rate of 45%. The changes in leukemia
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
0
5
10
1515
35 Spleen
0.0
0.3
0.50.5
1.5 BM
02468
1010
30 BloodCntlMem
hCD
45+C
D56
+ /m
CD
45+
P = 0.052 P = 0.27
P = 0.25
hCD
45+C
D56
+/
mC
D45
+
0
20
40
60
80
Per
cent
IFN
-γ+
NK
c ells
0
10
20
30
40
50CD56bright CD56dim
Cntl Mem Cntl Mem
P = 0.002 P = 0.003
B C
D
A
mC
D45
hCD45 hCD3 hCD16
hCD
56
hCD
56
Spleen
89.5
BM
39.1
90.8
80.7
87.1
97.5
1.3
95.4
3.4
83.7
3.0
89.0
Blood
E
21 28134
Cntl
Mem
Day:G
0 10 20 300.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
MemCntl
P < 0.001
Time (days)
Leuke
mia
burd
en
(10
10
photo
n fl
ux)
H
0 20 40 60 80 1000
20
40
60
80
100
Cntl NKMem NK
PBS
*
Time (days)
Per
cent
surv
ival
I
NSG
Wash16 hours
IL-15
IL-12+
IL-15+
IL-18
Control
NK
Day 7
EngraftmentFunctionality
Preactivated
NSG
1 106
K562-lucPreactivated or cntl
NK cells
0Day 43
Wash16 hours
IL-15
IL-12+
IL-15+
IL-18
NK
Tumor burden (BLI)SurvivalrhIL-2 QOD
F
hCD
56
hIFN-γ
21.7 21.527.2
1.8 1.52.8
Cntl
Mem
CD56 CD56dimCD56bright
rhIL-2 QOD
*P = 0.027
treated as in (F), monitored for survival, and analyzed using the log-rank test. PBS, phosphate-buffered saline. Summary of
Fig. 4. Human memory-likeNK cells control human leuke-
mia in an NSG xenograft mod-el. (A) Experimental design for(B) to (E). rhIL-2, recombinanthuman IL-2; QOD, every otherday. (B to E) NSG mice receivedhumanNKcell adoptive transfersas indicated in (A). Representa-tive flow cytometry at day 7 aftertransfer shows engraftment ofhuman memory-like NK cellsin the indicated tissues froma representative donor. BothCD56bright and CD56dim subsetsare detectable. (C) Summary ofdata from (B) demonstratingthe engraftment of control andmemory-likeNKcells,withabun-dance identified as the ratioto murine CD45+ mononuclearcells. (D) Control or memory-like NK cells were administeredtoNSGmiceas in (A).After7days,splenocytes were isolated andrestimulated with K562 for6 hours, followed by assess-ment of human NK cells forIFN-g production. Numbers de-pict the percentages of cellswithin the indicated regions.(E) Summary of data from (D)showing the means ± SD ofpercent IFN-g–positive NK cellsfrom the indicated NK cell sub-sets. Statistical analysis was per-formed with Mann-Whitneytest. (F) Experimental designfor (G) to (I). (G to I) K562-lucwas injected intravenously intoNSG mice. After 4 days, BLI wasperformed to ensure leukemiaengraftment, and control ormemory-like NK cells wereadministered to the mice. Themice were treated with rhIL-2every other day andmonitoredfor tumor burden (BLI) and sur-vival. (G) Representative BLI ofrecipient mice engrafted withK562-luc on the indicated dayafter tumor administration. (H)Summary of serial BLI measure-ments that show reduced tu-mor burden in mice receivingmemory-likeNKcells comparedto control NK cells. Differenceswere determined using analysis of variance (ANOVA). (I) Mice weredata are from two to three experiments with n = 12 to 24 mice per group represented as means ± SEM.nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 6
R E S EARCH ART I C L E
blast percentages in the BM of responding and nonresponding patientsare shown in fig. S6. There was no detectable correlation between pre-therapy BMblast percentage, donorKIR haplotype, the presence of pre-dictedKIR-ligandmismatch, and the frequency of donorNKcells in theblood or BM and clinical response, but the number of patients wassmall. A complete set of clinical vignettes describing each patient andhis or her clinical course is included in the SupplementaryMaterials andMethods. Additional parameters of relevant NK cell receptor biologyare shown in table S5. Thus, allogeneic human IL-12, IL-15, and IL-18–induced memory-like NK cells proliferate, expand, and exhibit anti-leukemia function after adoptive transfer into rel/ref AML patientswith active disease.
by guest on Septe
http://stm.sciencem
ag.org/D
ownloaded from
DISCUSSION
Motivated by the need to develop innovative treatment options forAML patients (4, 5), we investigated the ability of a recently definedfunctional class of NK cells, cytokine-induced memory-like NK cells,to mediate antileukemia responses (20). Here, we demonstrated thatprimary human NK cells, differentiated in vitro into memory-like NKcells via brief preactivation with IL-12, IL-15, and IL-18, exhibit potentantileukemia responses in the form of IFN-g production and cyto-toxicity. Usingmass cytometry andmultidimensional analyses to assessa large number of NK cell relevant proteins simultaneously, we foundthatmemory-likeNKcells were clearly distinguishable fromcontrolNKcells from the same individual. Despite stable inhibitory KIR receptorexpression, memory-like NK cells responded more robustly to primaryAML blasts, regardless of KIR-ligand interactions. Moreover, humanmemory-like NK cells were superior, in terms of both controlling leu-kemia burden and prolonging survival, in an in vivo NSG xenograftmodel. Translating this to the clinic, we examined the biology of theMHC-haploidentical donor IL-12, IL-15, and IL-18–preactivated NKcells administered to patients with active rel/ref AML in the contextof a first-in-human NK cell trial. After transfer, memory-like NK cellswere detectable in the blood and BM of patients for weeks, proliferated
www.ScienceTra
extensively, and expanded in vivo. Further, memory-like NK cells thatdifferentiated in patients in vivo exhibited enhanced functionalityagainst leukemia target cells. Notably, five of the nine evaluable AMLpatients receiving IL-12, IL-15, and IL-18–preactivatedNK cells had clin-ical responses, suggesting preliminary evidence of in vivo antileukemiaactivity. Thus, preactivation of NK cells with IL-12, IL-15, and IL-18 thatresults in memory-like NK cell differentiation represents a promisingapproach to enhancing adoptive allogeneic NK cell therapy.
Published studies have identified comparable IL-12, IL-15, and IL-18–induced memory-like NK cell biology in mice and humans (15–19). Inmouse models, preactivation with IL-12, IL-15, and IL-18 has beenshown to improve NK cell responses to lymphoma and melanoma celllines in vivo, which required T cell–derived IL-2 (17). Although somephenotypic differences have been identified (18), a clear approach todistinguishing memory-like from conventional NK cells from the samedonor was lacking. Through mass cytometry and multidimensionaldata reduction algorithms, control and memory-like NK cells wereidentified in separate areas of viSNE maps, indicating a distinct pheno-type and providing an approach to tracking memory-like NK cellsin vivo. Currently, memory-like NK cells that emerge after IL-12, IL-15,and IL-18 preactivation are thought to result from a differentiation pro-cess that yields a long-term alteration of functional capacity; however,the molecular mechanisms that control this process are yet to bedefined. This concept of differentiation is consistent with the distinctphenotype tracked by multidimensional analyses, and the observationsthat enhanced memory-like NK cell responses to restimulation are re-tained after extensive cell division and persisted for weeks to monthsafter the initial preactivation (15, 18). The phenotypic changes observedafter IL-12, IL-15, and IL-18 preactivation include increased expressionof inhibitory, activating, and cytokine receptors (CD94/NKG2A,NKp30, NKp44, NKp46, NKG2D, CD62L, and CD25), whereas otherreceptors appear unchanged (KIR, CD57, NKG2C, DNAM-1, CD137,and CD11b) or decreased (NKp80). Such dynamic changes in activat-ing, inhibitory, cytokine, and adhesion receptors are also consistentwithdifferentiation. Further, prolonged IL-12, IL-15, and IL-18 activation(for 5 days) resulted in reduced methylation of the IFN-g conserved
mber 4, 2020
Table 1. Summary of clinical characteristics of evaluable patientstreated with cytokine-induced memory-like NK cells. All patientswere Caucasian. KIR-ligand mismatch is reported in the NK cell donorversus patient vector. Patient 012 was reenrolled at a higher dose levelas patient 019. UPN, universal patient number; WHO, World Health Or-
ganization; DLT, dose-limiting toxicity; M, male; F, female; IWG, Interna-tional Working Group; TF-PD, treatment failure due to progressivedisease; CR, complete remission; MLFS, morphologic leukemia-freestate; CRi, CR with incomplete blood count recovery; MDS, myelodys-plastic syndrome.
UPN
Dose level Gender Age(years)WHOdiagnosis
Number of previoustherapies
nsla
PretreatmentBM blast (%)
tionalMedicine.or
KIR-ligandmismatch
g 21 September 2016 V
IWGresponse
ol 8 Issue 357
DLT
357ra1
GVHD
001
1 M 73 M2 2 16 Yes TF-PD No No006
1 M 70 M0 3 28 Yes TF-PD No No007
1 M 77 M0 1 47 Yes CR No No008
2 M 76 t-AML 3 17 Yes TF-PD No No009
2 F 73 M1 3 80 No MLFS No No012
2 F 71 M5 3 15 Yes CR No No017
3 M 64 t-AML 3 69 Yes TF-PD No No019
3 F 71 M5 4 15 Yes CR No No020
3 M 60 MDS-AML 1 13 Yes CRi No No23 7
R E S EARCH ART I C L E
noncoding sequence 1 locus in human NK cells, indicating that cyto-kines can affect epigenetic control mechanisms (32). One alternativehypothesis to the differentiation theory is that IL-12, IL-15, and IL-18–
www.ScienceTra
preactivated NK cells exist at an enhanced activation state without afundamental change in their molecular programs. Although most pub-lished data are consistent with the idea of differentiation rather than
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
A
Num
ber
of P
B
dono
r NK
cel
ls p
erm
l
1 3 7 14 210
20
40
60
80
100
DonorRecipientP < 0.0001
Time after infusion (days)
Per
cent
age
of P
B N
K c
ell s
Donor 7.0
Donor 1.0
Rcpt 99.0
Rcpt 93.0
Donor 91.0
Rcpt 9.0
Donor 0.1
Rcpt 99.9
Day 1 Day 3 Day 7 Day 14
CD
56
HLA marker
Time after infusion (days)
C
3 70
20
40
60
80
100P < 0.001 P = 0.02
Recipient Donor
Time after infusion (days)
Per
cent
age
Ki -6
7–po
sitiv
e N
K c
ells
Cou
nts
Day 7
Ki-67
Day 3
Rcpt Donor
Iso
0
20
40
60
80
100
8 14
RecipientDonor
Time after infusion (days)
Per
cent
age
ofB
M N
K c
ells
P = 0.012
D G
E
B
020
1 3 7 14 210
2000
40006000
8000
10,000019
1 3 7 14 210
20,000
40,000
60,000
80,000017
1 3 7 14 210
10,000
20,000
30,000
40,000012
1 3 7 14 210
2000
4000
6000
8000001
1 3 7 14 210
100
200 007
1 3 7 14 210
20
40
60008
1 3 7 14 210
10002000300040005000 009
1 3 7 14 210
100
200
300
Donor 88.0
Rcpt 12.0
Day 8
CD
56
HLA marker
F
007 009 012 017 019 0200
5000
10,000
15,000
20,000
25,000
100,000200,000300,000400,000500,000
Dose 2 Dose 3Dose 1
Num
ber
of B
M d
onor
NK
cel
ls p
er m
l
Donor 56.0
Rcpt 44.0
Day 14
H Day 8
Fig. 5. Donor memory-like NK cells expand and proliferate in vivo inAMLpatients. (A) Representative flow cytometry data of donor (HLA+) versus
blood on the indicated day after infusion. Isotype control (iso) staining forKi-67 is indicated by the gray histogram. (E) Summary of data of the percen-
recipient (rcpt) (HLA−) NK cells on the indicated day after infusion. (B) Percen-tages of the peripheral blood (PB) NK cell compartment composed of donorversus recipient NK cells at the indicated time points after infusion, summariz-ing the eight patients with informative HLAmAbs available for all time points.(C) Absolute numbers of donor NK cells permilliliter of peripheral blood at theindicated times after infusion for each patient (UPN above individual graphs).Absolute numbers were calculated by multiplying the number of mono-nuclear cells obtained per milliliter of blood by the fraction of CD45+ cellsthat consisted of donor NK cells. (D) Representative flow cytometry datashowing Ki-67 staining in donor and recipient NK cells from the peripheral
tagesof Ki-67–positivedonor versus recipientperipheral bloodNKcells at days3 and 7 after infusion. (F) Representative flow cytometry data of donor (HLA−)versus recipient (HLA+) NK cells on the indicatedday after infusion. (G) Percen-tagesof theBMNKcell compartment thatweredonor versus recipientNKcellsat the indicated timepoint after infusion, summarizing the sevenpatientswithinformative HLA mAbs available for all time points. (H) Absolute numbers ofdonor and recipient NK cells per milliliter of BM at day 8 after infusion. Ab-solutenumberswereobtainedas in (C). Statistical comparisonwasperformedby two-way ANOVA. In representative bivariate flow plots (A and F), numbersindicate the frequency of cells within the indicated gate.
nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 8
R E S EARCH ART I C L E
http://stm.sciencem
aD
ownloaded from
activation, elucidation of molecular mechanisms controlling cytokine-induced memory would provide additional clarification of this issue.
Earlier work has set the stage for the use of allogeneic NK cellsas adoptive immunotherapy for AML patients (9–12), harnessing theleukemia-targeting potential of NK cell allorecognition of MHC-mismatched AML first identified after HCT (8). Initial pioneeringadoptive NK cell therapy clinical trials identified that an availablelymphoid niche was essential for donor NK cell engraftment and expan-sion; however, clinical responses occurred in a minority of patients withlimited duration (9). This modest effect is consistent with the conceptthat combined modulatory approaches will be required to fully optimizeNK cell adoptive immunotherapy for cancer patients, including (i) en-hancement of NK cell targeting of cancer cells, (ii) augmentation of NKcell functional capacity, and (iii) elimination of inhibitory checkpointsor cellular negative regulators (33S). In the context of AML, NK cellallorecognition of MHC-haploidentical blasts has provided anestablished mode of targeting NK cells to myeloid leukemia via inhib-itory receptor mismatch and activating receptor ligation. We discoveredthat IL-12, IL-15, and IL-18–induced memory-like NK cells exhibitenhanced triggering against AML regardless of KIR to KIR-ligand inter-actions, resulting in an expanded NK cell pool of AML-reactive effectorcells. Because NKG2A is expressed on most memory-like NK cells, theinteraction with HLA-E on AML blasts or bystander cells in the micro-environment remains a potentially important inhibitory pathway formemory-like NK cells. The functionality of memory-like NK cells is alsoenhanced in terms of cytotoxicity and IFN-g production. Thus, throughbrief combined cytokine stimulation, two aspects of the NK cell anti-tumor response have been improved.
www.ScienceTra
Most adoptive NK cell therapy studies to date use IL-2 or IL-15 toactivate NK cells overnight before transfer. This strategy results in ashort-term priming signal that increases NK cell functional capacity(33), but the effect is rapidly lost after removal from the in vitro cytokinemilieu and transfer into the patient. Here, we reasoned that the longer-lasting increase in functional capacity afforded by memory-like NK celldifferentiation, combined with improved AML recognition, would en-hance in vivo expansion and antileukemia responses, resulting in a sev-eral week “window of opportunity” to attack AML blasts. Donormemory-like NK cells consistently expanded to become >90% of bloodNKcells, as well asmost BMNKcells. Because of concerns for cytokine-release syndrome, our phase 1 study was intentionally initiated with astarting NK cell dose of 1/10 to
1/20 of the typical adoptive NK cell dose(9–12). Considering the relatively low doses of highly purified IL-12,IL-15, and IL-18–preactivated NK cells administered to patients, thedonor NK cell frequencies and numbers observed here in the bloodand BM are remarkable. Two published studies have transferred dosesof CD56+CD3−-purified NK cells similar to our dose level 3 as adoptiveimmunotherapy for patients with active AML (11, 12). The overall re-sponse rate of 55% (5 of 9) and CR/CRi rate of 45% (4 of 9) in our studycompare favorably to 7%(1of 15) observedwithin cohort 2 of Bachanovaet al. (0 of 10) (12) and patients with active AML on the work of Curtiet al. (1 of 5) (11). Because our sample size was limited, this finding ishypothesis-generating andwill need to be studied inmore patients trea-tedwith IL-12, IL-15, and IL-18–preactivatedNKcells. Inour small sampleset treated to date, these responses did not segregate with any KIR-based rules for NK cell allorecognition, althoughmost patients studiedhad aKIR toKIR-ligandmismatch in thedonor versus recipient direction.
by guest on Septem
ber 4, 2020g.org/
007 009 019 0200
5001000150020002500
10,00030,00050,00070,000
BM Dose 1
Dose 2
Dose 3
0
20
40
60
80
100
P < 0.0001
Nor
mal
ized
nu m
ber o
f PB
I FN
-γ–
p osi
tive
NK
cel
ls/m
l
0
20
40
60
80
100
Nor
mal
ized
num
ber o
f BM
I FN
- γ –
po
sitiv
e N
K c
ells
/ml
0
5
10
15
20
25
RecipientDonor
P = 0.006
Unstimulated K562
P = 0.6
Per
cent
age
of I F
N-γ
–pos
itive
IFN-γ
CD
56
3
Recipient 22
Donor Blood
13 Donor
CD
56
IFN-γ
Recipient 2
BM
B C D
E F G
A Blood
001 007 008 009 012 017 019 0200
500
1000
3000500070009000
Blood Dose 1 Dose 2 Dose 3
Recipient Donor
Num
ber o
f IFN
-γ–p
ositi
ve
NK
cel
ls/ m
l
Num
ber o
f IFN
-γ–p
ositi
ve
NK
cel
ls/m
l
P < 0.0001
Fig. 6. Donor memory-like NK cells display enhanced antileukemia re-sponses at 1 week after adoptive transfer. Freshly isolated peripheral
ing relative IFN-g production by recipient and donor NK cells from the blood;the datawere normalized to donor IFN-g–positive NK cells, whichwere set to
bloodmononuclear cells (PBMCs) frompatient blood or BMwere stimulatedwith K562 leukemia cells at an effector/target ratio of 10:1 for 6 hours andassessed for IFN-g production by flow cytometry. (A) Percentage of IFN-g–positive donor versus recipient NK cells in the peripheral blood for all pa-tients. Data were compared using paired t test. (B) Representative flowcytometric data showing donor and recipient NK cell IFN-g responses fromtheperipheralblood. (C)Absolutenumbersof recipient (blue)anddonor (green)NK cells producing IFN-g in the blood. (D) Summary of data from (C) depict-
100%. (E) Representative flow cytometry data showing donor and recipientNK cell IFN-g responses from the BM. (F) Absolute numbers of recipient anddonor NK cells producing IFN-g in the BM. (G) Summary of data from (F) de-picting relative IFN-g production by recipient and donor NK cells from theBM; the data were normalized to donor IFN-g–positive NK cells, which wereset to 100%. Normalized control NK cell responses were tested against 100%(donor) using a one-sample t test. Numbers represent percentage of cellswithin the indicated quadrant. All summary data depict means ± SEM.
nslationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 9
R E S EARCH ART I C L E
Similarly, there were no associations with NK cell number or IFN-g pro-duction measurements, which had substantial variability within patientswho received different doses of NK cells. Thus, this report provides proofof concept that human IL-12, IL-15, and IL-18–preactivated NK cellsmay be translated into the cancer immunotherapy clinic, with consistentmemory-like NK cell biology in vitro, in xenograft models in vivo and inAML patients in vivo.
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
MATERIALS AND METHODS
Study designPatients treated on an open-label, nonrandomized, first-in-humanphase 1 dose escalation trial (NCT01898793) are included in this study.The primary objective of the clinical trial was to identify the maximumtolerated or tested dose of memory-like NK cells administered to pa-tients with rel/ref AML. Sample size was based on a standard 3 + 3 doseescalation design. Patients with rel/ref AML who were not candidatesfor immediate HCT were eligible to participate. Patients were treatedwith fludarabine/cyclophosphamide between days −7 and −2 for immu-nosuppression, followed on day 0 by allogeneic donor IL-12, IL-15, andIL-18–preactivated NK cells in escalating doses: 0.5 × 106/kg (doselevel 1), 1.0 × 106/kg (dose level 2), and all NK cells that could begenerated from a single leukapheresis capped at 10 × 106/kg (doselevel 3). After donor NK cell adoptive transfer, patients received low-dose rhIL-2 (1 × 106 IU/m2) subcutaneously every other day for atotal of six doses. Donor NK cell products were generated by purifyingNK cells from a nonmobilized leukapheresis product using CD3 deple-tion followed byCD56-positive selection on aCliniMACSdevice (>90%CD56+CD3−). Purified NK cells were preactivated with IL-12 (10 ng/ml),IL-15 (50 ng/ml), and IL-18 (50 ng/ml) for 12 to 16 hours under cur-rent GMP conditions. Samples were obtained from the peripheralblood (before treatment and at days 1, 3, 7, 8, 14, 21, 30, 60, and100) and BM (before treatment and at days 8, 14, 30, 60, and 100) afterthe NK cell infusions. Clinical responses were defined by the revisedIWG criteria for AML (31). If cell numbers were limiting, analyses ofNK cell number and phenotype were prioritized. All patients providedinformedconsentbeforeparticipating andwere treatedat theWashingtonUniversity Institutional Review Board (IRB)–approved clinical trial(Human Research Protection Office #201401085).
Reagents, mice, and cell linesAntihumanmAbswere used for flow andmass cytometry (tables S1, S2,andS6), includingpreviously reported anti-HLAmAbs (34). Endotoxin-free, recombinant human cytokines are described in the Supplemen-tary Materials and Methods. NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ mice(4 to 8weeks old)were obtained fromThe JacksonLaboratory,maintainedunder specific pathogen–free conditions, and used in accordance with ouranimal protocol approved by the Washington University Animal StudiesCommittee. K562 cells [American Type Culture Collection (ATCC),CCL-243] were obtained in 2008, viably cryopreserved and stored inliquid nitrogen, thawed for use in these studies, and maintained for<2 months at a time in continuous culture according to ATCC instruc-tions. K562 cells were authenticated in 2015 using single-nucleotide poly-morphism analysis andwere found to be exactlymatched to theK562 cellsfrom the Japanese Collection of Research Bioresources, GermanCollectionof Microorganisms and Cell Cultures (DSMZ), and ATCC databases(2015, Genetic Resources Core Facility at Johns Hopkins University).
www.ScienceTran
NK cell purification and cell cultureNormal donor PBMCswere obtained from anonymous healthy plateletdonors and isolated by Ficoll centrifugation (18, 19, 35). NK cells werepurified using RosetteSep (STEMCELL Technologies; routinely ≥95%CD56+CD3−). KIR genotyping of donor PBMC was performed byKashi Clinical Laboratories. Memory-like and control NK cells weregenerated as previously described (see the Supplementary Materialsand Methods for more details) (18).
Patient samplesPatients with newly diagnosed AML provided informed consent undertheWashingtonUniversity IRB-approvedprotocol (2010-11766) andwerethe source of primary AML blasts for in vitro resimulation experiments(table S4).After informedconsent, patient sampleswere alsoobtained fromthe first-in-human phase 1 clinical study of allogeneic donor memory-likeNK cells in rel/ref AML (fig. S5). Patient peripheral blood or BM aspiratewasobtainedat the indicated timepoint afterNKcell infusion, andPBMCswere isolatedbyFicoll centrifugationand immediately used in experiments.
Functional assays to assess cytokine productionControl and memory-like NK cells were harvested after a rest period of7 days to mimic transfer into a patient and to allow memory-like NKcell differentiation to occur. Cells were then restimulated in a standardfunctional assay (18, 19, 35). Cells were stimulated with K562 leukemiatargets or freshly thawed primary AML blasts (effector/target ratio of5:1, unless otherwise indicated). Further details are provided in the Sup-plementary Materials and Methods.
Flow-based killing assayFlow-based killing assays were performed by co-incubating memory-like or control NK cells with carboxyfluorescein diacetate succinimidylester–labeled K562 cells for 4 hours and assaying 7-aminoactinomycinD uptake as described (18, 19, 35).
Adoptive transfer of humanmemory-like NK cells into NSGmicePurified control (5 × 106;≥95% CD56+CD3−) or IL-12, IL-15, and IL-18–preactivated NK cells were injected retro-orbitally into NSG mice.Mice received rhIL-2 (50,000 IU) through intraperitoneal injection ev-ery other day to support the adoptively transferredNKcells. After 7 days,the mice were sacrificed, and organs (spleen, blood, and BM) wereassessed for the presence of transferred cells. NSG splenocytes werealso examined for intracellular IFN-g production after restimulationwith K562 leukemia targets.
In vivo BLIForBLI experiments,micewere treated as described in the SupplementaryMaterials andMethods. For imaging,mice were injected intraperitoneallywith D-luciferin (150 mg/g) (Biosynth) in PBS and imaged using the IVIS100 (Caliper). BLI was performed at the indicated time points.
Flow cytometric analysisCell staining was performed as described (18, 19, 35), and data wereacquired on a Gallios flow cytometer (Beckman Coulter) and analyzedusing Kaluza (Beckman Coulter) or FlowJo (Tree Star) software.
Mass cytometryAll mass cytometry data were collected on a CyTOF2 mass cytometer(Fluidigm) and analyzed using Cytobank (36). Mass cytometry data
slationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 10
R E S EARCH ART I C L E
Dow
were analyzed using previously described methods (37). For detailedinformation, refer to the SupplementaryMaterials andMethods. Diver-sity was assessed on eight individual donors at baseline (freshly thawedPBMCs or thawed pure NK cells), after control treatment or memory-like differentiation (described above) using the Boolean gating strategydescribed in fig. S2 (21, 22). Event counts were used to generate inverseSimpson index for KIR2DL1, KIR2DL1/KIR2DL2, KIR3DL1,KIR2DS4, and KIR2DL5 expression.
Statistical analysisBefore statistical analyses, all data were tested for normal distribution(D’Agostino-Pearson omnibus normality test). If datawere not normal-ly distributed, the appropriate nonparametric tests were used (GraphPadPrism v5.0), with all statistical comparisons indicated in the figurelegends. Uncertainty is represented in the figures as SEM, except as in-dicated in the figure legends. All comparisons used a two-sideda of 0.05for significance testing.
by guest on Septem
ber 4, 2http://stm
.sciencemag.org/
nloaded from
SUPPLEMENTARY MATERIALS
www.sciencetranslationalmedicine.org/cgi/content/full/8/357/357ra123/DC1Materials and MethodsClinical vignettesFig. S1. Reduction of memory-like NK cell cytotoxicity against K562 leukemia cells by blockadeof NKG2D or DNAM-1.Fig. S2. Mass cytometry gating strategies.Fig. S3. Phenotypic marker expression on viSNE gated control and memory-like NK cells.Fig. S4. Enhanced effector responses of memory-like NK cells compared to control-treated NKcells.Fig. S5. Schema of allogeneic memory-like NK cell phase 1 clinical trial (NCT01898793).Fig. S6. Distribution of BM blast percentages and donor NK cell numbers sorted by clinicaloutcomes.Fig. S7. Chemokine receptor expression on control versus IL-12, IL-15, and IL-18–preactivatedhuman NK cells.Table S1. NK cell phenotypic mass cytometry panel design, reagents, and clustering usage.Table S2. NK cell functional mass cytometry panel design, reagents, and clustering usage.Table S3. Characteristics of normal donors used in mass cytometry functional experiments.Table S4. Characteristics of AML samples used for in vitro NK cell functional assays.Table S5. Patient HLA and donor KIR characteristics for evaluable donor-patient pairs treated inthe phase 1 clinical trial.Table S6. Flow cytometry mAbs.Reference (38)
020
REFERENCES AND NOTES
1. G. Juliusson, P. Antunovic, Å. Derolf, S. Lehmann, L. Möllgård, D. Stockelberg, U. Tidefelt,A. Wahlin, M. Höglund, Age and acute myeloid leukemia: Real world data on decision totreat and outcomes from the Swedish Acute Leukemia Registry. Blood 113, 4179–4187(2009).
2. E. Estey, H. Döhner, Acute myeloid leukaemia. Lancet 368, 1894–1907 (2006).3. J. M. Rowe, M. S. Tallman, How I treat acute myeloid leukemia. Blood 116, 3147–3156
(2010).4. H. Döhner, E. Estey, S. Amadori, F. R. Appelbaum, T. Büchner, A. K. Burnett, H. Dombret,
P. Fenaux, D. Grimwade, R. A. Larson, F. Lo-Coco, T. Naoe, D. Niederwieser, G. J. Ossenkoppele,M. A. Sanz, J. Sierra, M. S. Tallman, B. Löwenberg, C. D. Bloomfield, Diagnosis and managementof acute myeloid leukemia in adults: Recommendations from an international expert panel,on behalf of the European LeukemiaNet. Blood 115, 453–474 (2010).
5. K. Mrózek, G. Marcucci, D. Nicolet, K. S. Maharry, H. Becker, S. P. Whitman, K. H. Metzeler,S. Schwind, Y.-Z. Wu, J. Kohlschmidt, M. J. Pettenati, N. A. Heerema, A. W. Block, S. R. Patil,M. R. Baer, J. E. Kolitz, J. O. Moore, A. J. Carroll, R. M. Stone, R. A. Larson, C. D. Bloomfield,Prognostic significance of the European LeukemiaNet standardized system for reportingcytogenetic and molecular alterations in adults with acute myeloid leukemia. J. Clin.Oncol. 30, 4515–4523 (2012).
www.ScienceTran
6. M. A. Caligiuri, Human natural killer cells. Blood 112, 461–469 (2008).7. J. Pahl, A. Cerwenka, Tricking the balance: NK cells in anti-cancer immunity. Immunobiology
S0171-2985(15)30031-0 (2015).8. L. Ruggeri, M. Capanni, E. Urbani, K. Perruccio, W. D. Shlomchik, A. Tosti, S. Posati, D. Rogaia,
F. Frassoni, F. Aversa, M. F. Martelli, A. Velardi, Effectiveness of donor natural killer cell allo-reactivity in mismatched hematopoietic transplants. Science 295, 2097–2100 (2002).
9. J. S. Miller, Y. Soignier, A. Panoskaltsis-Mortari, S. A. McNearney, G. H. Yun, S. K. Fautsch, D. McKenna,C. Le, T. E. Defor, L. J. Burns, P. J. Orchard, B. R. Blazar, J. E. Wagner, A. Slungaard, D. J. Weisdorf,I. J. Okazaki, P. B. McGlave, Successful adoptive transfer and in vivo expansion of humanhaploidentical NK cells in patients with cancer. Blood 105, 3051–3057 (2005).
10. J. E. Rubnitz, H. Inaba, R. C. Ribeiro, S. Pounds, B. Rooney, T. Bell, C.-H. Pui, W. Leung,NKAML: A pilot study to determine the safety and feasibility of haploidentical natural killercell transplantation in childhood acute myeloid leukemia. J. Clin. Oncol. 28, 955–959(2010).
11. A. Curti, L. Ruggeri, A. D’Addio, A. Bontadini, E. Dan, M. R. Motta, S. Trabanelli, V. Giudice,E. Urbani, G. Martinelli, S. Paolini, F. Fruet, A. Isidori, S. Parisi, G. Bandini, M. Baccarani, A. Velardi,R. M. Lemoli, Successful transfer of alloreactive haploidentical KIR ligand-mismatched naturalkiller cells after infusion in elderly high-risk acute myeloid leukemia patients. Blood 118,3273–3279 (2011).
12. V. Bachanova, S. Cooley, T. E. Defor, M. R. Verneris, B. Zhang, D. H. McKenna, J. Curtsinger,A. Panoskaltsis-Mortari, D. Lewis, K. Hippen, P. McGlave, D. J. Weisdorf, B. R. Blazar, J. S. Miller,Clearance of acute myeloid leukemia by haploidentical natural killer cells is improvedusing IL-2 diphtheria toxin fusion protein. Blood 123, 3855–3863 (2014).
13. G. Min-Oo, Y. Kamimura, D. W. Hendricks, T. Nabekura, L. L. Lanier, Natural killer cells:Walking three paths down memory lane. Trends Immunol. 34, 251–258 (2013).
14. A. Rölle, J. Pollmann, A. Cerwenka, Memory of infections: An emerging role for naturalkiller cells. PLOS Pathog. 9, e1003548 (2013).
15. M. A. Cooper, J. M. Elliott, P. A. Keyel, L. Yang, J. A. Carrero, W. M. Yokoyama, Cytokine-induced memory-like natural killer cells. Proc. Natl. Acad. Sci. U.S.A. 106, 1915–1919 (2009).
16. M. P. Keppel, L. Yang, M. A. Cooper, Murine NK cell intrinsic cytokine-induced memory-likeresponses are maintained following homeostatic proliferation. J. Immunol. 190, 4754–4762(2013).
17. J. Ni, M. Miller, A. Stojanovic, N. Garbi, A. Cerwenka, Sustained effector function of IL-12/15/18–preactivated NK cells against established tumors. J. Exp. Med. 209, 2351–2365 (2012).
18. R. Romee, S. E. Schneider, J. W. Leong, J. M. Chase, C. R. Keppel, R. P. Sullivan, M. A. Cooper,T. A. Fehniger, Cytokine activation induces human memory-like NK cells. Blood 120, 4751–4760(2012).
19. J. W. Leong, J. M. Chase, R. Romee, S. E. Schneider, R. P. Sullivan, M. A. Cooper, T. A. Fehniger,Preactivation with IL-12, IL-15, and IL-18 induces CD25 and a functional high-affinity IL-2 recep-tor on human cytokine-induced memory-like natural killer cells. Biol. Blood Marrow Transplant.20, 463–473 (2014).
20. M. M. Berrien-Elliott, J. A. Wagner, T. A. Fehniger, Human cytokine-induced memory-likenatural killer cells. J. Innate Immun. 7, 563–571 (2015).
21. A. Horowitz, D. M. Strauss-Albee, M. Leipold, J. Kubo, N. Nemat-Gorgani, O. C. Dogan, C. L. Dekker,S. Mackey, H. Maecker, G. E. Swan, M. M. Davis, P. J. Norman, L. A. Guethlein, M. Desai,P. Parham, C. A. Blish, Genetic and environmental determinants of human NK cell diversityrevealed by mass cytometry. Sci. Transl. Med. 5, 208ra145 (2013).
22. D. M. Strauss-Albee, J. Fukuyama, E. C. Liang, Y. Yao, J. A. Jarrell, A. L. Drake, J. Kinuthia,R. R. Montgomery, G. John-Stewart, S. Holmes, C. A. Blish, Human NK cell repertoire diversityreflects immune experience and correlates with viral susceptibility. Sci. Transl. Med. 7, 297ra115(2015).
23. E.-a. D. Amir, K. L. Davis, M. D. Tadmor, E. F. Simonds, J. H. Levine, S. C. Bendall, D. K. Shenfeld,S. Krishnaswamy, G. P. Nolan, D. Pe’er, viSNE enables visualization of high dimensionalsingle-cell data and reveals phenotypic heterogeneity of leukemia. Nat. Biotechnol. 31,545–552 (2013).
24. S. Cooley, D. J. Weisdorf, L. A. Guethlein, J. P. Klein, T. Wang, C. T. Le, S. G. E. Marsh, D. Geraghty,S. Spellman, M. D. Haagenson, M. Ladner, E. Trachtenberg, P. Parham, J. S. Miller, Donor se-lection for natural killer cell receptor genes leads to superior survival after unrelated transplan-tation for acute myelogenous leukemia. Blood 116, 2411–2419 (2010).
25. S. Cooley, E. Trachtenberg, T. L. Bergemann, K. Saeteurn, J. Klein, C. T. Le, S. G. E. Marsh,L. A. Guethlein, P. Parham, J. S. Miller, D. J. Weisdorf, Donors with group B KIR haplotypesimprove relapse–free survival after unrelated hematopoietic cell transplantation for acute my-elogenous leukemia. Blood 113, 726–732 (2009).
26. J. M. Venstrom, G. Pittari, T. A. Gooley, J. H. Chewning, S. Spellman, M. Haagenson, M. M. Gallagher,M. Malkki, E. Petersdorf, B. Dupont, K. C. Hsu, HLA-C–dependent prevention of leukemia re-lapse by donor activating KIR2DS1. N. Engl. J. Med. 367, 805–816 (2012).
27. D. M. Strauss-Albee, A. Horowitz, P. Parham, C. A. Blish, Coordinated regulation of NK re-ceptor expression in the maturing human immune system. J. Immunol. 193, 4871–4879 (2014).
28. L. L. Lanier, NK cell recognition. Annu. Rev. Immunol. 23, 225–274 (2005).29. S. C. Bendall, E. F. Simonds, P. Qiu, E.-a. D. Amir, P. O. Krutzik, R. Finck, R. V. Bruggner,
R. Melamed, A. Trejo, O. I. Ornatsky, R. S. Balderas, S. K. Plevritis, K. Sachs, D. Pe’er, S. D. Tanner,
slationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 11
R E S EARCH ART I C L E
http://stm.scienc
Dow
nloaded from
G. P. Nolan, Single-cell mass cytometry of differential immune and drug responses across ahuman hematopoietic continuum. Science 332, 687–696 (2011).
30. P. Qiu, E. F. Simonds, S. C. Bendall, K. D. Gibbs Jr., R. V. Bruggner, M. D. Linderman, K. Sachs,G. P. Nolan, S. K. Plevritis, Extracting a cellular hierarchy from high-dimensional cytometrydata with SPADE. Nat. Biotechnol. 29, 886–891 (2011).
31. B. D. Cheson, J. M. Bennett, K. J. Kopecky, T. Büchner, C. L. Willman, E. H. Estey, C. A. Schiffer,H. Doehner, M. S. Tallman, T. A. Lister, F. Lo-Coco, R. Willemze, A. Biondi, W. Hiddemann,R. A. Larson, B. Löwenberg, M. A. Sanz, D. R. Head, R. Ohno, C. D. Bloomfield, Revised recom-mendations of the international working group for diagnosis, standardization of responsecriteria, treatment outcomes, and reporting standards for therapeutic trials in acute myeloidleukemia. J. Clin. Oncol. 21, 4642–4649 (2003).
32. M. Luetke-Eversloh, Q. Hammer, P. Durek, K. Nordström, G. Gasparoni, M. Pink, A. Hamann,J. Walter, H.-D. Chang, J. Dong, C. Romagnani, Human cytomegalovirus drives epigeneticimprinting of the IFNG locus in NKG2Chi natural killer cells. PLOS Pathog. 10, e1004441 (2014).
33. R. Romee, J. W. Leong, T. A. Fehniger, Utilizing cytokines to function-enable human NKcells for the immunotherapy of cancer. Scientifica 2014, 205796 (2014).
34. A. Mulder, M. J. Kardol, J. S. Arn, C. Eijsink, M. E. I. Franke, G. M. T. Schreuder, G. W. Haasnoot,I. I. N. Doxiadis, D. H. Sachs, D. M. Smith, F. H. J. Claas, Human monoclonal HLA antibodiesreveal interspecies crossreactive swine MHC class I epitopes relevant for xenotransplanta-tion. Mol. Immunol. 47, 809–815 (2010).
35. M. Rosario, B. Liu, L. Kong, L. I. Collins, S. E. Schneider, X. Chen, K. Han, E. K. Jeng, P. R. Rhode,J. W. Leong, T. Schappe, B. A. Jewell, C. R. Keppel, K. Shah, B. Hess, R. Romee, D. R. Piwnica-Worms,A. F. Cashen, N. L. Bartlett, H. C. Wong, T. A. Fehniger, The IL-15-based ALT-803 complexenhances FcgRIIIa-triggered NK cell responses and in vivo clearance of B cell lymphomas.Clin. Cancer Res. 22, 596–608 (2016).
36. N. Kotecha, P. O. Krutzik, J. M. Irish, Web-based analysis and publication of flow cytometryexperiments. Curr. Protoc. Cytom. Chapter 10, Unit10.17 (2010).
37. K. E. Diggins, P. B. Ferrell Jr., J. M. Irish, Methods for discovery and characterization of cellsubsets in high dimensional mass cytometry data. Methods 82, 55–63 (2015).
38. E. R. Zunder, R. Finck, G. K. Behbehani, E.-a. D. Amir, S. Krishnaswamy, V. D. Gonzalez,C. G. Lorang, Z. Bjornson, M. H. Spitzer, B. Bodenmiller, W. J. Fantl, D. Pe’er, G. P. Nolan,Palladium-based mass tag cell barcoding with a doublet-filtering scheme and single-cell de-convolution algorithm. Nat. Protoc. 10, 316–333 (2015).
www.ScienceTran
Acknowledgments: We would like to thank W. Yokoyama, A. French, T. Ley, P. Westervelt, andJ. DiPersio for insightful discussion. We thank C. Keppel, K. Shah, and A. Ireland for technicalassistance. We also thank our patient volunteers and the BM transplant/leukemia physicianand nurse coordinator teams at the Washington University School of Medicine (WUSM).Funding: This work was supported by the American Society of Hematology Foundation,the Conquer Cancer Foundation of the American Society of Clinical Oncology, the WUSM SitemanCancer Center Developmental Research Award and Team Science Award, the WUSM Institute ofClinical and Translational Research Award, the Leukemia Specialized Program of Research Excel-lence (P50 CA171963) Development Research Award, the Howard Hughes Medical Institute Med-ical Fellow Award, the Translational TL1 program, NIH/National Cancer Institute (NCI) grant F32CA200253, the V Foundation for Cancer Research, and the Gabrielle’s Angel Foundation for CancerResearch. Technical support was provided by the Immunomonitoring Laboratory (also supportedby the Center for Human Immunology and Immunotherapy Programs), the Biological TherapyCore, and the Small Animal Cancer Imaging Core (also supported by P50 CA94056), which aresupported by the NCI Cancer Center Support grant P30CA91842. We acknowledge the use ofthe Protein Production and Purification Facility for CyTOF mAb conjugation (P30 AR048335). Au-thor contributions: R.R., M.R., M.M.B.-E., S.T.O., M.A.C., and T.A.F. conceived and designed thestudy; R.R., M.R., M.M.B.-E., J.A.W., B.A.J., T.S., S.A.-L., S.E.S., S.W., L.Y., Y.-S.L., and C.C.N. collected,analyzed, and assembled the data; A.M. and F.C. provided critical reagents; R.R., M.R., M.M.B.-E.,J.W.L., M.A.C., and T.A.F. wrote the manuscript; and all authors reviewed the data and edited andapproved the final version of the manuscript. Competing interests: The authors declare that theyhave no competing interests.
Submitted 22 February 2016Accepted 4 August 2016Published 21 September 201610.1126/scitranslmed.aaf2341
Citation: R. Romee, M. Rosario, M. M. Berrien-Elliott, J. A. Wagner, B. A. Jewell, T. Schappe,J. W. Leong, S. Abdel-Latif, S. E. Schneider, S. Willey, C. C. Neal, L. Yu, S. T. Oh, Y.-S. Lee,A. Mulder, F. Claas, M. A. Cooper, T. A. Fehniger, Cytokine-induced memory-like naturalkiller cells exhibit enhanced responses against myeloid leukemia. Sci. Transl. Med. 8,357ra123 (2016).
em
slationalMedicine.org 21 September 2016 Vol 8 Issue 357 357ra123 12
by guest on Septem
ber 4, 2020ag.org/
myeloid leukemiaCytokine-induced memory-like natural killer cells exhibit enhanced responses against
Yi-Shan Lee, Arend Mulder, Frans Claas, Megan A. Cooper and Todd A. FehnigerJeffrey W. Leong, Sara Abdel-Latif, Stephanie E. Schneider, Sarah Willey, Carly C. Neal, Liyang Yu, Stephen T. Oh, Rizwan Romee, Maximillian Rosario, Melissa M. Berrien-Elliott, Julia A. Wagner, Brea A. Jewell, Timothy Schappe,
DOI: 10.1126/scitranslmed.aaf2341, 357ra123357ra123.8Sci Transl Med
of which produced clinical remissions.acute myeloid leukemia, where the natural killer memory-like cells again demonstrated antileukemia effects, some against myeloid leukemia models in vitro and in mice. They also performed a clinical trial in human patients withcompared human natural killer memory cells to non-memory control cells, then demonstrated their effectiveness
. investigated these memory cells' potential as a cancer therapy. The authorset alnatural killer cells, and Romee pathogens as well as cancer. Recent studies have identified the existence of memory-like characteristics in some
Natural killer cells, part of the innate immune system, play a role in immune responses against exogenousNatural killers of leukemia
ARTICLE TOOLS http://stm.sciencemag.org/content/8/357/357ra123
MATERIALSSUPPLEMENTARY http://stm.sciencemag.org/content/suppl/2016/09/19/8.357.357ra123.DC1
CONTENTRELATED
http://science.sciencemag.org/content/sci/367/6478/643.fullhttp://science.sciencemag.org/content/sci/362/6421/1416.fullhttp://science.sciencemag.org/content/sci/362/6421/1355.fullhttp://science.sciencemag.org/content/sci/361/6407/1056.fullhttp://science.sciencemag.org/content/sci/359/6383/1537.fullhttp://stke.sciencemag.org/content/sigtrans/10/485/eaai9062.fullhttp://stm.sciencemag.org/content/scitransmed/9/413/eaao1214.fullhttp://stm.sciencemag.org/content/scitransmed/9/374/eaaj2025.fullhttp://science.sciencemag.org/content/sci/355/6322/eaaf8399.fullhttp://stm.sciencemag.org/content/scitransmed/8/368/368ra171.fullhttp://stm.sciencemag.org/content/scitransmed/5/208/208ra145.fullhttp://stm.sciencemag.org/content/scitransmed/6/231/231ra49.fullhttp://stm.sciencemag.org/content/scitransmed/6/243/243ra87.fullhttp://stm.sciencemag.org/content/scitransmed/7/283/283ra55.full
REFERENCES
http://stm.sciencemag.org/content/8/357/357ra123#BIBLThis article cites 37 articles, 22 of which you can access for free
PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions
Terms of ServiceUse of this article is subject to the
registered trademark of AAAS. is aScience Translational MedicineScience, 1200 New York Avenue NW, Washington, DC 20005. The title
(ISSN 1946-6242) is published by the American Association for the Advancement ofScience Translational Medicine
Copyright © 2016, American Association for the Advancement of Science
by guest on Septem
ber 4, 2020http://stm
.sciencemag.org/
Dow
nloaded from
