γ‐Secretase inhibitor reduces immunosuppressive cells and...

c-Secretase inhibitor reduces immunosuppressive cells andenhances tumour immunity in head and neck squamous cellcarcinoma

Liang Mao1, Zhi-Li Zhao1, Guang-Tao Yu1, Lei Wu1, Wei-Wei Deng1, Yi-Cun Li1, Jian-Feng Liu1, Lin-Lin Bu1, Bing Liu1,2,

Ashok B. Kulkarni3, Wen-Feng Zhang1,2, Lu Zhang1 and Zhi-Jun Sun 1,2,3

1 The State Key Laboratory Breeding Base of Basic Science of Stomatology & Key Laboratory of Oral Biomedicine, Ministry of Education, School and

Hospital of Stomatology, Wuhan University, Wuhan, 430079, China2 Department of Oral Maxillofacial-Head Neck Oncology, School and Hospital of Stomatology, Wuhan University, Wuhan, 430079, China3 Functional Genomics Section, Laboratory of Cell and Developmental Biology, National Institute of Dental and Craniofacial Research, National Institutes of

Health, Bethesda, MD

Immune evasion is a hallmark feature of cancer, and it plays an important role in tumour initiation and progression. In addi-

tion, tumour immune evasion severely hampers the desired antitumour effect in multiple cancers. In this study, we aimed to

investigate the role of the Notch pathway in immune evasion in the head and neck squamous cell carcinoma (HNSCC) microen-

vironment. We first demonstrated that Notch1 signaling was activated in a Tgfbr1/Pten-knockout HNSCC mouse model. Notch

signaling inhibition using a c-secretase inhibitor (GSI-IX, DAPT) decreased tumour burden in the mouse model after prophylac-

tic treatment. In addition, flow cytometry analysis indicated that Notch signaling inhibition reduced the sub-population of

myeloid-derived suppressor cells (MDSCs), tumour-associated macrophages (TAMs) and regulatory T cells (Tregs), as well as

immune checkpoint molecules (PD1, CTLA4, TIM3 and LAG3), in the circulation and in the tumour. Immunohistochemistry (IHC)

of human HNSCC tissues demonstrated that elevation of the Notch1 downstream target HES1 was correlated with MDSC, TAM

and Treg markers and with immune checkpoint molecules. These results suggest that modulating the Notch signaling pathway

may decrease MDSCs, TAMs, Tregs and immune checkpoint molecules in HNSCC.

Immune evasion is a hallmark feature of tumours as theyemploy various strategies to suppress the immune system’sability to recognize and destroy cancer cells.1 Accumulatingevidence of the antitumour activity of the immune systemsupports recent success in cancer immunotherapy.2 Manyimmune cells play a central role in immune evasion in thehead and neck squamous cell carcinoma (HNSCC) tumourmicroenvironment; for example, the presence of myeloid-

derived suppressor cells (MDSCs), immunosuppressive regu-latory T cells (Tregs) and tumour-associated macrophages(TAMs) is imbalanced and enhanced in HNSCC.3 MDSCsrepresent heterogeneous immunosuppressive cells in multiplecancer types, dampening T cell-mediated immune surveil-lance, attenuating the cytotoxicity of NK cells and supportingstem cell-like cancer cells.4,5 MDSCs have been recognized inmice by the expression of CD11b and Gr-1 (Ly6C1/Ly6G1),

Key words: notch1, head and neck squamous cell carcinoma, knockout mouse, immune checkpoint, myeloid-derived suppressor cell,

tumour-associated macrophage, regulatory T cell

Abbreviations: CTLA4: cytotoxic T lymphocyte-associated antigen 4; DLN: draining lymph node; EDTA: thylene diamine tetraacetic

acid; EMT: epithelial mesenchymal transition; HNSCC: head and neck squamous cell carcinoma; IHC: immunohistochemistry; LAG3:

lymphocyte activation gene 3; MDSCs: myeloid-derived suppressor cells; PD1: programmed cell death protein 1; PTEN: phosphatase

and tensin homolog; STAT3: signal transducer and activator of transcription 3; TAMs: tumour-associated macrophages; TGFBR1: trans-

forming growth factor receptor 1; TIM3: T cell membrane protein 3; TMA: tissue microarray; Tregs: regulatory T cells; TSCC: tongue

squamous cell carcinoma; WT: wild type

Additional Supporting Information may be found in the online version of this article.

Grant sponsor: National Natural Science Foundation of China; Grant numbers: 81672668, 81472529, 81472528, 81672667; Grant sponsor:

NIH (Division of Intramural Research, NIDCR; A.B.K); Grant sponsor: Ministry of Education of China (program for new century excellent

talents in university; Z.J.S.); Grant number: NCET-13–0439

DOI: 10.1002/ijc.31115

History: Received 14 Feb 2017; Accepted 9 Oct 2017; Online 19 Oct 2017

Correspondence to: Prof. Zhi-Jun Sun, School and Hospital of Stomatology, Wuhan University, Wuhan, 430079, China, Tel.: 186-27-8768-

6108, Fax: 186-27-8787-3260, E-mail: [email protected] and Lu Zhang, School and Hospital of Stomatology, Wuhan University, Wuhan,

430079, China, Tel.: 186-27-8768-6108, Fax: 186-27-8787-3260, E-mail: [email protected]

Tum

orIm

mun

olog

yan

dMicroenvironment

Int. J. Cancer: 00, 00–00 (2017) VC 2017 UICC

International Journal of Cancer

IJC

http://orcid.org/0000-0003-0932-8013

and subpopulations of MDSCs have been shown to exist:polymorphonuclear (PMN)- and monocytic (M)-MDSCs.6

The PMN-MDSCs were defined as CD11b1Ly6G1Ly6Clo,and M-MDSCs were defined as CD11b1Ly6G-Ly6Chi.6

CD41CD251Foxp31 regulatory T cells (Tregs) play animportant role in regulating the immune system by suppress-ing self-reactive T cells during excessive immune response inperipheral lymphoid tissues.7 TAMs secrete an array of cyto-kines and deplete L-arginine in the tumour microenviron-ment to suppress T cell activity.8,9 Given theseimmunosuppressive effects, it has been proposed that elimi-nation of these immature myeloid cells and Tregs may signif-icantly improve antitumour responses and enhance thebeneficial effects of cancer immunotherapy.10,11

Immune checkpoint blockade antibodies enhance T cellanti-tumour activity and have produced exciting and consis-tently encouraging results in the treatment of various cancers,including head and neck cancer.7 Tumours co-opt certainimmune-checkpoint pathways such as programmed cell deathprotein 1 (PD1), cytotoxic T-lymphocyte-associated antigen 4(CTLA4), T cell membrane protein 3 (TIM3) and lympho-cyte activation gene 3 (LAG3) as a major mechanism ofimmune resistance, particularly against T cells that are spe-cific for tumour antigens.12 Increased expression of immunecheckpoint proteins on T cells may reflect an exhausted oranergic state.

Notch1 signaling plays a fundamental role in the imple-mentation of differentiation and proliferation in the regula-tion of self-renewal programs in development anddisease.13–15 Accumulating evidence suggests that excessiveactivation of the Notch1 signaling is implicated in a varietyof tumours, particularly in leukemia and HNSCC.16 Activat-ing mutations of Notch1 signaling are found in approxi-mately 50% of Chinese HNSCC patients16,17 and only 10% ofCaucasian patients.18,19 Our previous study indicated thatactivation of the Notch1 pathway plays a pivotal role in can-cer stem-like cell maintenance in HNSCC, participating inepithelial-mesenchymal transition (EMT) and increasing che-moresistance.20 Several reports have suggested that theNotch1 signaling pathway plays vital roles in T cell expansionand Treg induction.21 However, the exact role of Notch1 sig-naling in myeloid cell expansion and immune tolerance israther controversial.22,23 In addition, the comprehensive effectof the Notch1 signaling pathway in cancer cells and immunecells in the solid tumour microenvironment remains unclear.

This study was designed to investigate the precise role ofNotch1 signaling in the modulation of tumour-infiltratingMDSCs, TAMs, Tregs and immune checkpoint molecules inHNSCC.

Material and MethodsImmunocompetent transgenic HNSCC mouse models

All experiments were conducted in accordance with guidelinesof the Institutional Animal Care and Use Committee of theWuhan University, Wuhan, China. The time inducible tissue-specific Tgfbr1/Pten double knockout mice (K14-CreERtam6;Tgfbr1flox/flox; Ptenflox/flox, Tgfbr1/Pten 2cKO) were maintainedand genotyped according to published protocols.24,25 All ani-mal studies were carried out in compliance with the NIHguidelines for the use of laboratory animals in pathogen freeASBL3 animal center of the Wuhan University and approvedby the Animal Care and Use Committee of the Wuhan Univer-sity (2014LUNSHENZI06, 2016LUNSHENZI62). All the micewere in FVBN/CD1/129/C57 mixed background.

Drugs and DAPT treatment

The g-secretase inhibitor DAPT (N-[N-(3,5-Difluorophenace-tyl-L-alanyl)]-S- phenylglycinet-Butyl Ester, GSI-IX), pur-chased from Selleck Chemicals (Houston, TX), was initiallydissolved in 100% ethanol at a concentration of 20 mg/ml,and the aliquots were stored at–208C. The working solutionwas further diluted in PBS at a final concentration of 2 mg/ml immediately before use. The vehicle was used as a nega-tive control. The 10- to 12-week-old Tgfbr1/Pten 2cKO mice(Tgfbr1flox/flox; Ptenflox/flox; CK14-CreERtam6) were orally gav-aged with tamoxifen (200 ll of a 10 lg/ll solution in cornoil) for 5 consecutive days as previously described.24 For thechemopreventive experiment, the Tgfbr1/Pten 2cKO micewere randomly divided into a control group (vehicle, i.p.daily, n5 7 mice) and a DAPT treatment Group (10 mg/kg,i.p. daily, n5 9 mice) on Day 21. Mice were treated witheither DAPT or vehicle for 3 weeks. Tumour size was mea-sured with a micrometre calliper, and tumours were photo-graphed every other day. The mice were euthanized at theend of the studies, and tissues were harvested for flow cytom-etry, Western blot and immunohistochemical analysis.

Flow cytometry

A single cell suspension was prepared from the spleen, drain-ing lymph nodes (DLNs), blood and tumours of the HNSCC

What’s new?

Tumor survival depends on sneaking past the body’s own immune defenses. Here, the authors probed how cancer cells exploit

the Notch signaling pathway to evade immune destruction. Looking at HNSCC cells, they first showed that Notch1 signaling is

activated in the tumor cells. Then, they showed that inhibiting the Notch signaling pathway decreased the tumor burden, as

well as markedly reducing the production of immunosuppressive cells, such as tumor associated macrophages and immuno-

suppressive regulatory T cells. Thus, treatments targeting Notch1 could be useful against HNSCC.

Tum

orIm

mun

olog

yan

dMicroenvironment

2 g-Secretase inhibitor reduces immunosuppressive cells


mouse models as previously described.26 Specifically, thetumour tissues and immune organs were harvested and proc-essed with a gentle Macs dissociator and a mouse tumourdissociation kit (Miltenyi Biotec, Germany). Flow cytometrywas performed with a FACS Caliber flow cytometer equippedwith CellQuest software (Becton Dickinson, Mountain View,CA). The following anti-mouse antibodies were used forstaining: FITC-conjugated anti-CD3, anti-CD4, anti-CD11b,PE-conjugated anti-Ly6G, anti-PD1, anti-TIM3, anti-LAG3,anti-CTLA4, anti-IFN-g, PE-Cy7-conjugated anti-Ly6C,Percp-Cy5.5-conjugated anti-F4/80, anti-CD8 and mouse reg-ulatory T cell staining kit #3 (all from eBioscience, SanDiego, CA), as well as isotype-matched IgG controls (eBio-science, San Diego, CA). All the antibodies were used follow-ing the manufacturer’s recommended protocols. The cellswere analyzed using FlowJo (Tree Star, Ashland, OR) andgated by the side scatter and forward scatter filters. Nonviablecells were excluded by staining with 7AAD (Invitrogen).

Proliferation assay

The MDSCs were sorted as PMN-MDSCs(CD11b1Ly6G1Ly6Clo) or M-MDSCs (CD11b1Ly6G-Ly6Chi)subtype from spleens of Tgfbr1/Pten 2cKO mice by mouseMyeloid-Derived Suppressor Cell Isolation Kit (Miltenyi Bio-tec, Germany). The na€ıve CD81 T cells were sorted fromlymph nodes of WT mice by Na€ıve CD8a1 T Cell IsolationKit (Miltenyi Biotec, Germany). CD41CD25hi cells fromspleens of Tgfbr1/Pten 2cKO mice were sorted as Tregs cellsby flow cytometry, and CD11b1F4/801 cells were sorted asTAMs. The na€ıve CD81 T cells were stained with CFSE, andthen were stimulated with 25 lg/ml aCD3, 2.5 lg/ml aCD28and 30 IU/ml rIL-2 (BD PharmingenTM). The activated Tcells were, respectively, cocultured with MDSCs, TAMs, andTregs at ration of 1:1, 1:0.5, and 1:0.25 in 96-well round-bot-tom plate. 72 hrs later, cells were harvested for CFSE analysisby flow cytometry.

Western blot

The Western blot analysis was performed according to ourpreviously described procedures.27 Briefly, tumours werecarefully dissected from Tgfbr1/Pten 2cKO mice and lysed ina T-PER buffer containing 1% phosphatase inhibitors andcomplete mini cocktail (Roche). About 40 lg of protein fromeach sample was denatured and then loaded in each lane ofNuPAGE 4–12% Bis-Tris precast gel. Subsequently, proteinswere transferred onto a PVDF membrane and blocked withmilk for 1 hrs, then incubated with primary antibodies over-night. On Day 2, the membrane was incubated with horse-radish peroxidase-conjugated secondary antibody (Pierce,Rockford, IL). The blotting was visualized by an enhancedchemiluminescence kit (West Pico, Thermo). The followingprimary antibody dilutions were used: 1:1,000 for Notch1,HES1, CCL2 (Cell signaling Technology) and CXCL1(Abcam) were detected on the same membrane. GAPDH(Cell Signaling Technology) was used as a loading control.

Human HNSCC samples

HNSCC specimens were obtained from the patients in the Hos-pital of Stomatology of Wuhan University. The pathologicalanalysis was performed by two independent pathologists of theDepartment of Oral Pathology, Wuhan University. All theHNSCC specimens were made into human HNSCC tissue array,including 97 primary HNSCC cases (Outdo Biotech, Shanghai).The custom made human HNSCC tissue microarrays were usedin this study as described previously.28 The study was performedwith the approval of the Institutional Medical Ethics Committeeof School and Hospital of Stomatology, Wuhan University(2014LUNSHENZI06, 2016LUNSHENZI62).

Immunochemistry, images analyses and scoring system

Tumours and control tissues such as tongue mucosa were care-fully dissected from the mice and fixed in 10% buffered forma-lin overnight. The tissues were embedded in paraffin. Thesections of tissues were deparaffinized, rehydrated and sub-jected to antigen retrieval by sodium citrate (pH 6.0) or ethyl-ene diamine tetraacetic acid (EDTA) for 20 min. Endogenousperoxidase activity was blocked with 3% H2O2 followed byserum block. The sections were incubated with primary anti-bodies overnight. On the second day, the sections were reactedwith appropriated secondary antibodies. The staining was per-formed by ABC kits (Vector). The following antibodies wereused in this study: Notch1, HES1, PD1, TIM3 (Cell SignalingTechnology), LAG3 (Abcam), CTLA4 (Santa Cruz). All immu-nohistochemically stained slides were scanned using an AperioScanScope CS scanner (Vista, CA). Percent positivity (totalnumber of positive pixels/total number of pixel) of whole sec-tion were quantified using Aperio ScanScope quantificationsoftware (Version 9.1). Area of interest was selected either inthe epithelial or the stromal area for quantification. The histo-score of calculated value as a percentage of different percentpositive cells were performed using the formula (31) 3

31 (21) 3 21 (11) 3 1 as previously described.24 Thethreshold for scanning of different positive cells was set accord-ing to the standard controls slices provided by Aperio.

Statistical analysis

Statistical data analysis was performed with GraphPad Prism 5.03(GraphPad Software, La Jolla, CA) statistical packages. The result-ing data among each group were analyzed through One-wayANOVA, followed by the post-Tukey multiple comparison testsor unpaired t test. Two-tailed Pearson statistical analyses wereused for correlated expression of these markers. Mean values6

SEM with a difference of p< 0.05 were considered statistically sig-nificant (ns, p> 0.05; *p< 0.05; **p< 0.01; ***p< 0.001).

ResultsActivation of Notch1 signaling in a Tgfbr1/Pten 2cKO

mouse model

Our previous study showed that Notch1/HES1 signaling wasactivated in human HNSCC tissues and cell lines.20 However,

Tum

orIm

mun

olog

yan

dMicroenvironment

Mao et al. 3


there is still a lack of preclinical research on the role ofNotch1 signaling in the HNSCC tumour environment. Con-ditional knockout of Pten and Tgfbr1 in head and neck epi-thelia of immunocompetent mice using a CK-14Cre-Loxpsystem gives rise to the initiation and progression of HNSCCthrough tamoxifen induction.29 This Tgfbr1/Pten 2cKOmouse model shows full penetration and fast tumour forma-tion and is suitable for therapeutic intervention, especiallyusing immunotherapy. To determine whether Notch1 signal-ing is activated or down-regulated in the Tgfbr1/Pten 2cKOmouse model, we examined the protein levels of Notch1 sig-naling in tongue mucosa from wild-type (WT) mice and2cKO mice and HNSCC tissue from 2cKO mice. The immu-nohistochemical (IHC) staining and Western blot analysisshowed that Notch1 and HES1, which is a putative down-stream target of Notch1, were strongly expressed in Tgfbr1/Pten 2cKO mouse tongue squamous cell carcinoma (TSCC)compared to WT tongue and Tgfbr1/Pten 2cKO mousetongue (Figs. 1a and 1b). These results indicate that Notch1signaling is potentially involved in the development ofHNSCC, and pharmacological intervention of Notch1 signal-ing in vivo in this transgenic mouse model will reveal theaction of Notch1/HES1 in HNSCC.

Inhibition of Notch1 signaling reduces tumourigenesis in a

Tgfbr1/Pten 2cKO mouse model

To functionally assess the effect of Notch1 signaling onHNSCC progression, we performed a chemopreventive experi-mental therapeutic investigation on tumour-bearing Tgfbr1/Pten 2cKO mice. DAPT, a novel g-secretase inhibitor, whichprevents proteolytic processing of Notch receptors,30 was usedin this preclinical study. As shown in Figure 2a, conditionalknock-out of Tgfbr1 and Pten in epithelia was initiated by oralgavage with tamoxifen (2 mg/kg) for 5 consecutive days. Micewere treated with DAPT (10 mg/kg) or vehicle (100 ml) admin-istered intraperitoneally every day starting at Day 21 aftertamoxifen induction. The results of this treatment indicate thattumour volume was remarkably reduced after 3-weeks ofDAPT treatment compared with the control group (Figs. 2band 2c; p < 0.01 on Day 36, p< 0.001 on Days 39 and 42).

Considering the adverse effects of DAPT on the digestive sys-tem,31 we assessed the toxicity of DAPT by examining changesin body weight. No statistical significance was observedbetween the DAPT-treated group and the control group (Fig.2d), which indicates that 10 mg/kg DAPT treatment was welltolerated by the transgenic mice. Likewise, as shown in Figure2e, the on-target inhibition of Notch1 signaling was confirmedby Western blot, which indicated that Notch1 and HES1 weresignificantly decreased by the DAPT treatment. Takentogether, these results demonstrate that inhibition of Notch1signaling by DAPT suppressed HNSCC progression.

Effect of DAPT on immunosuppression in an HNSCC mouse

model

Notch signaling reportedly has diverse functions in the cross-talk between cancer-associated cells in the tumour microenvi-ronment,3,5,32 participating in multiple immune responses.33–35

Our previous findings indicated that tumour-bearing Tgfbr1/Pten 2cKO mice were in an immunosuppressive state, withincreased MDSCs, TAMs and Tregs.26 Therefore, we exploredthe potential effect of Notch1 signaling inhibition on theseimmunosuppressive cells using flow cytometry. First, we foundthat CD11b1Ly6G1Ly6Clo PMN-MDSCs represented a sizablefraction of the spleen and tumour MDSCs population. In con-trary, CD11b1Ly6G- Ly6Chi M-MDSCs were a dominant sub-set of the MDSCs in DLNs. In addition, after DAPT treatment,M-MDSCs in DLNs (p5 0.0168) and the spleen (p5 0.0290)as well as PMN-MDSCs in the spleen (p5 0.0018) and tumour(p5 0.0588) were significantly decreased compared with thecontrol group (Figs. 3a and 3b). Interestingly, we also observedthat the populations of two other types of immunosuppressivecells, CD11b1F4/801 TAMs and CD41CD251Foxp31 Tregs,were remarkably decreased in the spleen and peripheral bloodafter DAPT treatment (Figs. 3c–f). However, DAPT treatmentfailed to significantly reduce the population of TAMs and Tregsin DLNs.

To confirm the immunosuppressive activity ofCD11b1Ly6G1Ly6Clo, CD11b1Ly6G– Ly6Chi, CD11b1F4/801

and CD41CD25hi cells, we, respectively, sorted these subsetsfrom the spleens of tumour-bearing transgenic mouse and

Figure 1. Increased Notch1 pathway activity in a Tgfbr1/Pten 2cKO mouse model of HNSCC. (a) Representative immunohistochemical stain-

ing of the Notch1 and HES1 isotype, Notch 1 and HES1 in wild-type (WT) tongue, Tgfbr1/Pten 2cKO tongue and 2cKO tongue squamous

cell carcinoma (TSCC). Scale bar, 50 lm. (b) Notch1 and HES1 exhibit relative overexpression in a 2cKO mouse model. The Western blot

probed for Notch1 and HES1, with GAPDH as a loading control. [Color figure can be viewed at wileyonlinelibrary.com]

Tum

orIm

mun

olog

yan

dMicroenvironment



http://wileyonlinelibrary.com

evaluated their capacity to suppress proliferation of effector Tcells. Results indicated that CD11b1Ly6G1Ly6Clo MDSCsinhibited T cell proliferation by nearly 90% at T/MDSCs rationof 1:1 and remained about 50% suppression at T/MDSCsration of 1:0.25 (Figs. 4a and 4b). And CD11b1Ly6G- Ly6Chi

MDSCs also showed potent immunosuppressive effect, butthe suppressive capacity of this subsets were less thanCD11b1Ly6G1Ly6Clo MDSCs (Fig. 4c). And we also foundthat CD11b1F4/801 TAMs and CD41CD25hi Tregs showedsuppressive function on effector T cells (Figs. 4d and 4e).

Taken together, these results suggested that DAPT iseffective in reducing the number of immunosuppressive cellsin the transgenic mouse model and that it showed differentlevels of efficacy in different immune organs.

Effect of DAPT on immune checkpoints in an HNSCC

mouse model

Immune-checkpoint pathways are a major mechanism ofimmune resistance, and blockade of these immune checkpoints

can enhance the activation of antitumour immune responses.12

Based on the findings reported above, we investigated whetherDAPT has an effect on anti-tumour immune modulation. Todetermine the effect of DAPT on immune checkpoints in theTgfbr1/Pten 2cKO mouse, we measured the expression of fourmajor checkpoint molecules (PD1, TIM3, CTLA4 and LAG3)on T cells after DAPT treatment in vivo. Of interest, the resultsindicated that PD1, TIM3, CTLA4 and LAG3 expression on cir-culating T cells was decreased by DAPT treatment (Figs. 5a–d).Consistent with the flow cytometry results, these findings werefurther substantiated by IHC analysis, which indicated that thecheckpoint molecules in the tumour microenvironment wereremarkably reduced or diminished in HNSCC tissues in themouse model after DAPT treatment (Fig. 5e). And we also ana-lyzed the populations of CD81IFN-g1 T cells in spleen andtumour microenvironment to investigate the immune activa-tion. The results demonstrated that DAPT treatment can signif-icantly increase CD81IFN-g1 T cells in the spleen and tumour(Fig. 5f). Collectively, these observations suggest that DAPT

Figure 2. Notch1 inhibition attenuates tumour growth in Tgfbr1/Pten 2cKO mice. (a) Schematic shows the chemopreventive experimental

therapeutic treatment of Tgfbr1/Pten 2cKO mice with DAPT (n 5 9; 10 mg/kg, i.p) or PBS (n 5 7; 10 mg/kg, i.p) administered intraperitone-

ally every day for 3 weeks. (b) Representative photograph of head and neck tumours of the DAPT groups and PBS groups on Day 21 and

Day 42 after tamoxifen gavage. (c) Average volumes of HNSCC in control and DAPT treated mice. (Mean 6 SEM. Unpaired t test, **p<0.01;

***p<0.001). (d) Box and whiskers plot reveals the weight changes of mice in the DAPT- and PBS-treated groups (unpaired t-test). (e)

DAPT treatment decreased Notch1 and HES1 in Tgfbr1/Pten 2cKO mice compared with the control group. The Western blot was probed for

Notch1 and HES1, with GAPDH as a loading control. [Color figure can be viewed at wileyonlinelibrary.com]

Tum

orIm

mun

olog

yan

dMicroenvironment

Mao et al. 5



treatment might reverse the immunosuppressive state oftumour-bearing mice and decrease the expression of immunecheckpoints molecules in effector T cells.

Notch1 signaling is associated with immunosuppressive cells

markers and immune checkpoints molecules in human HNSCC

Given the effect of DAPT on immunosuppression in themouse model, we determined the relationship between

Notch1 signaling and immunosuppressive cells in human tis-sues. We quantified the expression of Notch1, Foxp3, PD1,TIM3, CTLA4, LAG3 and of MDSCs markers (CD11b andCD33) and TAMs markers (CD163 and CD68) in 97 primaryHNSCC patients using IHC on serial tissue microarrays (Fig.6a). The results indicate that the protein expression ofNotch1 was correlated with TAMs markers (CD68:r5 0.2839, p5 0.0048; CD163: r5 0.4203, p< 0.0001) and

Figure 3. Notch1 inhibition reduces the number of immune suppressor cells in Tgfbr1/Pten 2cKO mice. Representative FACS plots (a) and

quantification (b) of MDSCs from the spleen, DLNs and tumours on Day 42 after tamoxifen gavage (Mean 6 SEM. Unpaired t test, NS, No

Significance). Isotype staining is shown in the Supporting Information Figure S1. (c) Representative FACS plots of CD11b1F4/801TAMs in

the spleen. (d) Quantification of CD11b1F4/801TAMs in the spleen, DLNs and blood (Mean 6 SEM. Unpaired t test; NS, No Significance).

(e) Representative FACS plots of CD41CD251Foxp31 Tregs in the spleen. (f) Quantification of CD41CD251Foxp31 Tregs in the spleen, DLNs

and blood (Mean 6 SEM. Unpaired t test; NS, No Significance).

Tum

orIm

mun

olog

yan

dMicroenvironment



immune checkpoints (PD1: r5 0.5310, p< 0.0001; CTLA4:r5 0.7309, p< 0.0001; LAG3: r5 0.2598, p5 0.0102; TIM3:r5 0.3421, p5 0.0008) in human HNSCC (Supporting Infor-mation Fig. S3). However, Notch1 expression was not signifi-cantly correlated with Foxp3 (r5 0.1535, p5 0.1334) orMDSCs markers (CD11b: r5 0.1040, p5 0.3106; CD33:r5 0.08176, p5 0.4259; Supporting Information Fig. S4).Additionally, we performed IHC of HES1, which is a down-stream target of Notch1 signaling, to study whether the activa-tion of Notch1 signaling is correlated with immune resistance

in HNSCC. Surprisingly, not only TAMs markers (CD68:r5 0.2825, p5 0.0051; CD163: r5 0.3619, p5 0.0003) andimmune checkpoints (PD1: r5 0.5994, p< 0.0001; CTLA4:r5 0.5607, p< 0.0001; LAG3: r5 0.4335, p< 0.0001; TIM3:r5 0.4345, p< 0.0001) but also Foxp3 (r5 0.2774, p5 0.0059)and MDSCs markers (CD11b: r5 0.4983, p< 0.0001; CD33:r5 0.2443, p5 0.0159) showed a significant positive correla-tion with HES1 expression (Figs. 6be). Together, these resultssuggest that activation of Notch1/HES1 signaling represent apotential immunotherapy target in human HNSCC.

Figure 4. Immunosuppressive activity of MDSCs, TAMs, and Treg in Tgfbr1/Pten 2cKO mice. (a) Representative FACS histogram of CFSE

staining of anti-CD3/CD28 stimulated T cells after cocultured with CD11b1Ly6G1Ly6Clo MDSCs in 72 hrs (NEG, negative control). Quantifica-

tion of the suppressive activity of CD11b1Ly6G1Ly6Clo MDSCs (b), CD11b1Ly6G–Ly6Chi MDSCs (c), CD11b1F4/801 TAM (d) and

CD41CD25hi Tregs (e) (Mean 6 SEM. One-way ANOVA, followed by the post-Tukey multiple comparison tests; NS, No Significance).Tum

orIm

mun

olog

yan

dMicroenvironment

Mao et al. 7


DiscussionNumerous recent studies have suggested that additional ther-apeutic agents should be developed to broaden the utility ofactive drugs and increase the therapeutic efficacy of immuno-oncology treatment modalities because of the complex immu-nosuppressive milieu.36 In this study, our investigations pro-vide translational evidence supporting the hypothesis thatblocking Notch1 with a g-secretase inhibitor can decrease thenumber of immunosuppressive cells to retune the anti-

tumour immune response. Most importantly, we demonstratethat inhibition of Notch1 signaling can reduce the expressionof the negative immune checkpoint molecules PD1, CTLA4,TIM3 and LAG3 on effector T cells.

Aberrant biological process and genetic change in HNSCCmay correlate with clinical and epidemiologic variables.37

Through next-generation sequencing and high-throughputmolecular profiling, mutation of Notch1 was discovered to beextensive in HNSCC, ranging from approximately 10% in

Figure 5. Notch1 inhibition blockades immune checkpoints in Tgfbr1/Pten 2cKO mice. PD1 (a), TIM3 (b), CTLA4 (c) and LAG3 (d) expression

on CD4 and CD8 T cells in the spleen, DLNs and blood (Mean 6 SEM. Unpaired t test; NS, No Significance). The gating strategy is shown in

the Supporting Information Figure S2. (e) Representative immunohistochemical staining of the protein isotypes, PD1, TIM3, CTLA4 and LAG3

in control group and DAPT treatment group. (f) Quantification of CD81IFN-g1 T cells in the spleen and tumour of control group and DAPT

treatment group (Mean 6 SEM. Unpaired t-test). The gating strategy is shown in the Supporting Information Figure S3. [Color figure can be

viewed at wileyonlinelibrary.com]

Tum

orIm

mun

olog

yan

dMicroenvironment




Caucasian to above 50% in the Chinese population.16–19,38,39

Discrepancies in the potential role of Notch1 mutations mayrely on the different mutation spectrum in different cohortstudies or population.16–19,38,39 Recent reports and our previ-ous studies revealed a critical fact to be considered, namelythat Notch1 signaling status perhaps is mostly gain-of-function and highly activated in the Asian population,16,17

which is profoundly different from the Caucasian popula-tion.18,19,38,39 Activation of the Notch 1 signaling pathway

was observed in the transgenic mouse model, which is suit-able for studying the role of Notch 1 in HNSCC. Specifically,the transgenic mouse is an immunocompetent model andfacilitates research exploring the change in immune statusafter inhibition of Notch 1 signaling.

Although accumulated studies have confirmed that theNotch1 signaling pathway plays vital roles in T cell21 andmyeloid cell expansion,22,23 the exact role of Notch in thesolid tumour environment is still unclear. Our remarkable

Figure 6. Notch1 signaling is associated with immune escape in human HNSCC. (a) Representative images of immunohistochemical staining

of Notch1, HES1, Foxp3, CD11b, CD33, CD68, CD163, PD1, CTLA4, TIM3 and LAG3 in human HNSCC. Correlation and linear regression of

HES1 expression with Foxp3 (b), CD11b/CD33 (c), CD68/CD163 (d) and PD1/CTLA4/TIM3/LAG3 (e) in stroma. Each dot presents an inde-

pendent core, two-tailed Pearson’s statistic including 97 primary HNSCC patients. [Color figure can be viewed at wileyonlinelibrary.com]

Tum

orIm

mun

olog

yan

dMicroenvironment

Mao et al. 9



results in this study indicate that Notch1 inhibition canreduce the population of MDSCs in the tumour microenvi-ronment. Moreover, our previous studies showed that DAPTdecrease the phosphorylation of signal transducer and activa-tor of transcription 3 (STAT3), which plays an importantrole in myeloid cell expansion and inhibition of matura-tion.27,40 Nonetheless, due to certain limitations of our stud-ies, more investigations are required to confirm themechanisms of how Notch1 signaling regulates MDSCs inHNSCC. Because haematopoietic cytokines such as IL-6 andVEGF can modulate the expansion and recruitment ofMDSCs by regulating Notch signaling,41 and previous workhas revealed that Notch1 pathways are tightly linked withVEGF, which plays a vital roles in angiogenic expansion ofthe local vasculature in the tumour microenvironment,23,42

VEGF has been recognized as a robust chemokine in therecruitment of MDSCs to the tumour environment.43 Withthe present data alone, we cannot completely rule out thepossibility that Notch1 signaling blockade reduces the recruit-ment and expansion of MDSCs through other signaling path-ways or combinatorial signaling activity as discussed above.Therefore, how Notch1 signaling regulates MDSCs in HNSCCremains an open question and further studies are required toelucidate the detailed regulation mechanisms. In addition, thisstudy also demonstrated that DAPT treatment could reducethe TAMs population in vivo. However, this phenomenon con-tradicts previous reports that TAMs with forced Notch activa-tion exhibited an M1 phenotype and antitumour activity.22,44 Itshould be noted that Notch signaling could be activated to per-form biological functions in cancer cells and tumour microen-vironments. The complex crosstalk between tumours and the

microenvironment has been recognized as a major contributorto the progression of cancer,45,46 which cannot be ignored inthis study. The high correlation between the activation ofNotch signaling and the infiltration of TAMs in humanHNSCC (Fig. 6d), may indicate the possibility that Notch sig-naling is involved in the tumour-microenvironment interac-tions. However, further mechanistic investigations are requiredto explain this observation.

The rapidly emerging clinical successes of blocking immunecheckpoint molecules, such as CTLA4 and PD1, have openedprospects and extended the potential of cancer immunother-apy.47–49 Several immune checkpoint molecules such as PD1,CTLA4, LAG3 and TIM3 have been confirmed to be over-expressed in HNSCC patients. However, the modulation andactivation of these immune checkpoint molecules are still uncer-tain. Of particular interest, the present study revealed that Notch1inhibition by DAPT can reduce PD1, CTLA4, LAG3 and TIM3in the tumour microenvironment and macroenvironment, whichindicated the reinvigoration of effector T cells. Keeping in mindthat these results are rather descriptive, a detailed molecularmechanism needs to be worked out in the future.

In summary, our results clearly indicate that targetingNotch signaling in a Tgfbr1/Pten 2cKO mouse modeldecreased the number of immunosuppressive cells, andreduced the expression of negative immune checkpoint mole-cules on T cells in the HNSCC tumour microenvironmentand macro-environment. Over-expression of Notch1/HES1 inhuman HNSCC may be correlated with immunosuppressivecells markers and immune checkpoint molecules. Mostimportantly, our study points to Notch1 as a potential immu-notherapeutic target for treatment of HNSCC patients.

References

1. Hanahan D, Weinberg RA. Hallmarks of cancer:the next generation. Cell 2011;144:646–74.

2. Hoos A. Development of immuno-oncologydrugs - from CTLA4 to PD1 to the next genera-tions. Nat Rev Drug Discov 2016;15:235–47.

3. Azoury SC, Gilmore RC, Shukla V. Molecularlytargeted agents and immunotherapy for the treat-ment of head and neck squamous cell cancer(HNSCC). Discov Med 2016;21:507–16.

4. Pyzer AR, Cole L, Rosenblatt J, etet al. Myeloid-derived suppressor cells as effectors of immune sup-pression in cancer. Int J Cancer 2016;139:1915–26.

5. Peng D, Tanikawa T, Li W, et al. Myeloid-derived suppressor cells endow stem-like qualitiesto breast cancer cells through IL6/STAT3 andNO/NOTCH cross-talk signaling. Cancer Res2016;76:3156–65.

6. Bronte V, Brandau S, Chen SH, et al. Recommen-dations for myeloid-derived suppressor cellnomenclature and characterization standards. NatComms 2016;7:12150.

7. Gubin MM, Zhang X, Schuster H, et al. Check-point blockade cancer immunotherapy targetstumour-specific mutant antigens. Nature 2014;515:577–81.

8. Noy R, Pollard JW. Tumor-associated macro-phages: from mechanisms to therapy. Immunity2014;41:49–61.

9. Tsujikawa T, Yaguchi T, Ohmura G, et al. Auto-crine and paracrine loops between cancer cellsand macrophages promote lymph node metastasisvia CCR4/CCL22 in head and neck squamouscell carcinoma. Int J Cancer 2013;132:2755–66.

10. Predina JD, Kapoor V, Judy BF, et al. Cytoreductionsurgery reduces systemic myeloid suppressor cellpopulations and restores intratumoral immuno-therapy effectiveness. J Hematol Oncol 2012;5:34.

11. Arce Vargas F, Furness AJS, Solomon I, et al. Fc-optimized anti-CD25 depletes tumor-infiltratingregulatory T cells and synergizes with PD-1blockade to eradicate established tumors. Immu-nity 2017;46:577–86.

12. Pardoll DM. The blockade of immune check-points in cancer immunotherapy. Nat Rev Cancer2012;12:252–64.

13. Artavanis-Tsakonas S, Rand MD, Lake RJ. Notchsignaling: cell fate control and signal integrationin development. Science 1999;284:770–6.

14. Artavanis-Tsakonas S, Matsuno K, Fortini ME.Notch signaling. Science 1995;268:225–32.

15. Vinson KE, George DC, Fender AW, et al. TheNotch pathway in colorectal cancer. Int J Cancer2016;138:1835–42.

16. Izumchenko E, Sun K, Jones S, et al. Notch1mutations are drivers of oral tumorigenesis. Can-cer Prev Res (Phila) 2015;8:277–86.

17. Song X, Xia R, Li J, et al. Common and complexNotch1 mutations in Chinese oral squamous cellcarcinoma. Clin Cancer Res 2014;20:701–10.

18. Agrawal N, Frederick MJ, Pickering CR, et al.Exome sequencing of head and neck squamouscell carcinoma reveals inactivating mutations inNOTCH1. Science 2011;333:1154–7.

19. Stransky N, Egloff AM, Tward AD, et al. Themutational landscape of head and neck squamouscell carcinoma. Science 2011;333:1157–60.

20. Zhao ZL, Zhang L, Huang CF, et al. NOTCH1inhibition enhances the efficacy of conven-tional chemotherapeutic agents by targetinghead neck cancer stem cell. Sci Rep 2016;6:24704.

21. Franklin RA, Liao W, Sarkar A, et al. The cellularand molecular origin of tumor-associated macro-phages. Science 2014;344:921–5.

22. Wang YC, He F, Feng F, et al. Notch signalingdetermines the M1 versus M2 polarization ofmacrophages in antitumor immune responses.Cancer Res 2010;70:4840–9.

23. Cheng P, Kumar V, Liu H, et al. Effects of notchsignaling on regulation of myeloid cell differentia-tion in cancer. Cancer Res 2014;74:141–52.

24. Sun ZJ, Zhang L, Hall B, et al. Chemopreventiveand chemotherapeutic actions of mTOR inhibitorin genetically defined head and neck squamous

Tum

orIm

mun

olog

yan

dMicroenvironment



cell carcinoma mouse model. Clin Cancer Res2012;18:5304–13.

25. Zhang L, Sun ZJ, Bian Y, et al. MicroRNA-135bacts as a tumor promoter by targeting thehypoxia-inducible factor pathway in geneticallydefined mouse model of head and neck squamouscell carcinoma. Cancer Lett 2013;331:230–8.

26. Yu GT, Bu LL, Zhao YY, et al. CTLA4 blockadereduces immature myeloid cells in head and necksquamous cell carcinoma. Oncoimmunology 2016;5:e1151594.

27. Bu LL, Yu GT, Deng WW, et al. TargetingSTAT3 signaling reduces immunosuppressivemyeloid cells in head and neck squamous cellcarcinoma. Oncoimmunology 2016;5:e1130206.

28. Wu L, Deng WW, Yu GT, et al. B7-H4 expres-sion indicates poor prognosis of oral squamouscell carcinoma. Cancer Immunol Immunother2016;65:1035–45.

29. Bian Y, Hall B, Sun ZJ, et al. Loss of TGF-betasignaling and PTEN promotes head and necksquamous cell carcinoma through cellular senes-cence evasion and cancer-related inflammation.Oncogene 2012;31:3322–32.

30. Takebe N, Miele L, Harris PJ, et al. Targeting Notch,Hedgehog, andWnt pathways in cancer stem cells:clinical update. Nat Rev Clin Oncol 2015;12:445–64.

31. Yuan X, Wu H, Xu H, et al. Notch signaling: anemerging therapeutic target for cancer treatment.Cancer Lett 2015;369:20–7.

32. Ortiz-Martinez F, Gutierrez-Avino FJ, SanmartinE, et al. Association of Notch pathway down-regulation with triple negative/basal-like breastcarcinomas and high tumor-infiltrating FOXP31

Tregs. Exp Mol Pathol 2016;100:460–8.

33. Ghorpade DS, Kaveri SV, Bayry J, et al. Coopera-tive regulation of NOTCH1 protein-phosphatidylinositol 3-kinase (PI3K) signaling byNOD1, NOD2, and TLR2 receptors rendersenhanced refractoriness to transforming growthfactor-beta (TGF-beta)- or cytotoxic T-lymphocyte antigen 4 (CTLA-4)-mediatedimpairment of human dendritic cell maturation.J Biol Chem 2011;286:31347–60.

34. Mathern DR, Laitman LE, Hovhannisyan Z, et al.Mouse and human Notch-1 regulate mucosal immuneresponses.Mucosal Immunol 2014;7:995–1005.

35. Fasnacht N, Huang HY, Koch U, et al. Specificfibroblastic niches in secondary lymphoid organsorchestrate distinct Notch-regulated immuneresponses. J Exp Med 2014;211:2265–79.

36. Topalian SL, Drake CG, Pardoll DM. Immunecheckpoint blockade: a common denominatorapproach to cancer therapy. Cancer Cell 2015;27:450–61.

37. Bergamo NA, da Silva Veiga LC, dos Reis PP,et al. Classic and molecular cytogenetic analysesreveal chromosomal gains and losses correlatedwith survival in head and neck cancer patients.Clin Cancer Res 2005;11:621–31.

38. Pickering CR, Zhang J, Yoo SY, et al. Integrativegenomic characterization of oral squamous cellcarcinoma identifies frequent somatic drivers.Cancer Discov 2013;3:770–81.

39. Lawrence MS, Sougnez C, Lichtenstein L, et al.Comprehensive genomic characterization of headand neck squamous cell carcinomas. Nature 2015;517:576–82.

40. Kumar V, Cheng P, Condamine T, et al. CD45phosphatase inhibits STAT3 transcription factor

activity in myeloid cells and promotes tumor-associated macrophage differentiation. Immunity2016;44:303–15.

41. Saleem SJ, Conrad DH. Hematopoietic cytokine-induced transcriptional regulation and Notch sig-naling as modulators of MDSC expansion. IntImmunopharmacol 2011;11:808–15.

42. Zhang Z, Dong Z, Lauxen IS, et al. Endothelialcell-secreted EGF induces epithelial to mesenchy-mal transition and endows head and neck cancercells with stem-like phenotype. Cancer Res 2014;74:2869–81.

43. Gabrilovich DI, Ostrand-Rosenberg S, Bronte V.Coordinated regulation of myeloid cells bytumours. Nat Rev Immunol 2012;12:253–68.

44. Zhao JL, Huang F, He F, et al. Forced activation ofnotch in macrophages represses tumor growth byupregulating miR-125a and disabling tumor-associated macrophages. Cancer Res 2016;76:1403–15.

45. Gabrilovich D, Nefedova Y. ROR1C regulates dif-ferentiation of myeloid-derived suppressor cells.Cancer Cell 2015;28:147–9.

46. Correia AL, Bissell MJ. The tumor microenviron-ment is a dominant force in multidrug resistance.Drug Resist Updat 2012;15:39–49.

47. Larkin J, Chiarion-Sileni V, Gonzalez R, et al.Combined nivolumab and ipilimumab or mono-therapy in untreated melanoma. N Engl J Med2015;373:23–34.

48. Woods K, Cebon J. Tumor-specific T-cell help isassociated with improved survival in melanoma.Clin Cancer Res 2013;19:4021–3.

49. Villaruz LC, Kalyan A, Zarour H, et al. Immuno-therapy in lung cancer. Transl Lung Cancer Res2014;3:2–14.

Tum

orIm

mun

olog

yan

dMicroenvironment

Mao et al. 11


本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具

http://www.xuebalib.com/cloud/

http://www.xuebalib.com/

http://www.xuebalib.com/cloud/


http://www.xuebalib.com/vip.html

http://www.xuebalib.com/db.php

http://www.xuebalib.com/zixun/2014-08-15/44.html


γ‐Secretase inhibitor reduces immunosuppressive cells and...

Documents

Transcript of γ‐Secretase inhibitor reduces immunosuppressive cells and...