Nanoenabled Modulation of Acidic Tumor Microenvironment Reverses

Apr 3, 2019 - Nanoenabled Modulation of Acidic Tumor Microenvironment Reverses Anergy of Infiltrating T Cells and Potentiates Anti-PD-1 Therapy...
3 downloads 0 Views 2MB Size
Subscriber access provided by Queen Mary, University of London

Communication

Nano-Enabled Modulation of Acidic Tumor Microenvironment Reverses Anergy of Infiltrating T Cells and Potentiates Anti-PD-1 Therapy Yuxue Zhang, Yangyang Zhao, Jizhou Shen, Xun Sun, Yi Liu, Hang Liu, Yucai Wang, and Jun Wang Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b04296 • Publication Date (Web): 03 Apr 2019 Downloaded from http://pubs.acs.org on April 3, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

Nano-Enabled Modulation of Acidic Tumor

2

Microenvironment Reverses Anergy of Infiltrating T

3

Cells and Potentiates Anti-PD-1 Therapy

4

Yu-Xue Zhang,†,# Yang-Yang Zhao,†,# Jizhou Shen,†,# Xun Sun,† Yi Liu,† Hang Liu,†

5

Yucai Wang†,* and Jun Wang‡

6



7

Microscale, the CAS Key Laboratory of Innate Immunity and Chronic Disease, School

8

of Life Sciences, University of Science and Technology of China, Hefei 230027, China

9



Division of Molecular Medicine, Hefei National Laboratory for Physical Sciences at

Institutes for Life Sciences, School of Medicine and National Engineering Research

10

Center for Tissue Restoration and Reconstruction, South China University of

11

Technology, Guangzhou 510006, China

12

1

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

ABSTRACT

2

While tumor-infiltrating cytotoxic T lymphocytes play a critical role in controlling

3

tumor development, they are generally impotent in an acidic tumor microenvironment.

4

Systemic treatment to neutralize tumor acidity thus holds promise for the reversal of

5

the anergic state of T cells and the improvement of T cell-associated immunotherapy.

6

Herein, we report a proof-of-concept of RNAi nanoparticle-mediated therapeutic

7

reversion of tumor acidity to restore the anti-tumor functions of T cells and potentiate

8

the checkpoint blockade therapy. Our strategy utilized an in vivo optimized vesicular

9

cationic lipid-assisted nanoparticle, as opposed to its micellar counterpart, to mediate

10

systematic knockdown of lactate dehydrogenase A (LDHA) in tumor cells. The

11

treatment resulted in the reprogramming of pyruvate metabolism, a reduction of the

12

production of lactate, and the neutralization of the tumor pH. In immunocompetent

13

syngeneic melanoma and breast tumor models, neutralization of tumor acidity

14

increased infiltration with CD8+ T and NK cells, decreased the number of

15

immunosuppressive T cells, and thus significantly inhibited the growth of tumors.

16

Furthermore, the restoration of tumoral pH potentiated checkpoint inhibition therapy

17

using the antibody of programmed cell death protein 1 (PD-1). However, in

18

immunodeficient B6/Rag1-/- and NOG mice, the same treatment failed to control tumor

19

growth, further proving that the attenuation of tumor growth by tumor acidity

20

modulation was attributable to the activation of tumor-infiltrating immune cells.

21

2

ACS Paragon Plus Environment

Page 2 of 33

Page 3 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

KEYWORDS

2

Cancer immunotherapy, checkpoint blockade, lactate dehydrogenase A, nanomedicine,

3

T cell anergy, tumor acidity

4

3

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 33

1

CD8+ T lymphocytes play a central role in anti-cancer immunotherapy due to their

2

capacity to kill malignant cells upon T-cell receptor (TCR)-mediated recognition of

3

specific antigenic peptides presented by cancer cells.1 Enhanced intratumoral

4

infiltration of CD8+ T cells correlates with good clinical outcomes in various tumor

5

types.2 However, the infiltrated T cells can be subject to various immunosuppressive

6

mechanisms, acquire functional defects in the tumor, and enter a state of anergy.3-4

7

Systemic administration of antibodies that target co-inhibitory receptors on T cells such

8

as the programmed cell death protein 1 (PD-1) or cytotoxic T lymphocyte antigen 4

9

(CTLA-4), known as immune checkpoint blockade therapy, can reactivate the anergic

10

tumor-specific T cells and reinstate cancer immune-surveillance.5-10 Nevertheless,

11

despite some recent clinical success, T cell checkpoint blockade therapy is only

12

effective in a small fraction of patients with certain cancers, suggesting that additional

13

immunosuppressive pathways are still active.11-12 Thus, strategies to reverse the anergy

14

of infiltrated T cells and broaden the population benefitting from T cell-based treatment,

15

especially checkpoint blockade therapy, are urgently needed.

16

Solid tumors are characterized by a highly acidic microenvironment that results from

17

the enhanced production of lactate by tumor cells due to their high rates of aerobic

18

glycolysis.13-14 The acidity of tumor contributes to the chemotherapeutic resistance,

19

proliferation

20

immunosuppressive role of tumor acidity in impeding effective antitumor T-cell

21

immune responses. Specifically, CD8+ T cells tend to become anergic and apoptotic

22

when exposed to a low pH environment.16-20 On the other hand, excessive lactate

and

metastasis.15

Emerging

evidence

4

ACS Paragon Plus Environment

is

elucidating

the

Page 5 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

enhances the function of immunosuppressive cells such as myeloid subsets3,

2

regulatory T cells (Tregs),22 polarizes macrophages to the immunosuppressive

3

phenotype, and thereby blunts antitumor immune responses.23-24 Hence, antagonizing

4

the tumor acidity may reverse the detrimental effects of lactate and lead to the recovery

5

of functions of antitumor T cells.16, 25 Neutralization of tumor acidity with orally fed

6

bicarbonate20 or with intraperitoneally injected esomeprazole16 have been reported to

7

increase tumor response to checkpoint inhibitors and lead to cures in combination with

8

adoptive T-cell therapy. Despite that, to the best of our knowledge, systemic treatment

9

strategies to potentiate the efficacy of T cell immunotherapy by interfering with tumor

10

21

and

acidity have rarely been reported.

11

Herein, we propose a concept of nano-mediated modulation of the tumor acidic

12

microenvironment, which can therapeutically reverse the anergic state of infiltrating T

13

cells and further potentiate anti-PD-1 antibody (-PD-1) therapy. Nanomedicine-based

14

approaches have been demonstrated to function effectively in directly improving

15

systemic immune response or by modulating certain tumor microenvironment,26-27

16

including hypoxia to normoxia conversion,28 dense to sparse extracellular matrix,29-32

17

non-immunogenic to immunogenic conditions,33-37 suppressive to stimulatory milieu,38

18

or their combinations.39 Tumor acidity has been widely exploited as an external

19

stimulus for the design of environment responsive nanomedicines for cancer

20

treatment.40-41 Moreover, recent studies have demonstrated the potential of using

21

calcium carbonate or calcium bicarbonate nanoparticles to elevate tumor pH and

22

improve anticancer drug activity.42-43 Nevertheless, to the best of our knowledge, using 5

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

nanomedinces as an effective neutralizer of tumor acidity to reverse T cell anergy has

2

not been demonstrated yet. Our proof-of-principle strategy is based on the RNA

3

interference-mediated silencing of lactate dehydrogenase A (LDHA), which converts

4

pyruvate to lactic acid, accelerates glycolysis, and causes tumor acidity. LDHA has

5

been demonstrated to be associated with T cell anergy and poor prognosis in cancer

6

patients. Moreover, reduced expression of LDHA in melanoma cells exhibited slower

7

tumor growth through immune-mediated mechanisms.25 We employed cationic lipid-

8

assisted nanoparticles (CLAN) based on FDA-approved polymers for systemic siRNA

9

delivery, as they hold great promises for clinical translation and have been extensively

10

validated both pre-clinically and clinically for the delivery of siRNA, miRNA, and

11

CRISPR-Cas9 system.44-48 We demonstrated that a vesicular CLAN,44 as opposed to its

12

micellar counterpart, could prevent the premature release of encapsulated siRNA

13

during blood circulation and improve the tumor accumulation of siRNA (Scheme 1).

14

The vesicular CLAN-mediated gene silence efficiently downregulated LDHA

15

expression, decreased lactate secretion, and raised the pH of the tumor. Normalization

16

of the tumor acidity in turn decreased the number of immunosuppressive cells,

17

increased infiltration of CD8+ T cells, and restored the function of T cells. Tumor

18

acidity neutralization in combination with -PD-1 checkpoint blockade therapy

19

improved the anti-tumor response and produced synergistic therapeutic effects (Scheme

20

1).

6

ACS Paragon Plus Environment

Page 6 of 33

Page 7 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12

Scheme 1. Nano-mediated reversion of tumor immunosuppressive microenvironment through tumor acidity modulation. The pH of tumor microenvironment modulates the activation and proliferation of infiltrating immune cells and thereby regulates the balance between the immune surveillance and escape (also known as immune response and tolerance). Tumor cells overexpressing LDHA converts glucose into lactate, which blunts tumor surveillance by T cells. Nanomediated knockdown of LDHA therapeutically reversed the tumor acidic immunosuppressive microenvironment, which decreased the number of immunosuppressive cells, increased infiltration of CD8+ T cells, and restored their antitumor functions. Tumor acidity neutralization prior to -PD-1 checkpoint blockade therapy improved the anti-tumor response and produced synergistic effects.

13

7

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Results and Discussion

2

Tumor acidity affects proliferation, apoptosis, activation, and function of CD8+ T

3

cells. We first investigated the influence of tumor acidity on the proliferative and

4

functional properties of CD8+ T cells in vitro. Naïve mouse CD8+ T cells were isolated

5

from the spleen of C57BL/6 mice by flow cytometry (FACS) followed by CD3/CD28

6

stimulation. Culture media of pH 6.5 and 7.4 were used to mimic the tumor acidic

7

microenvironment and normal physiological conditions, respectively. After 72 h of

8

culture, the cells were labeled with fluorescent carboxyfluorescein succinimidyl ester

9

(CFSE) to monitor their proliferation. A significant inhibition of the proliferation of

10

CD8+ T cells was observed under acidic conditions as compared to neutral pH

11

conditions (proliferating cells: 63.7% at pH 6.5 versus 82.4% at pH 7.4, Figure S1A

12

and S1B). In addition, approximately ~2-fold more T cells were induced to apoptosis

13

under acidic, compared with neutral conditions (Figure S2A). The expression of CD25,

14

a T cell activation marker, was dramatically suppressed in cells cultured at pH 6.5

15

(Figure S2B). In line with this, we also observed a decreased secretion of IFN-γ under

16

acidic conditions, indicating the inhibition of T cell functions (Figure S2C). The above

17

data was consistent with previous reports which showed adverse effects of low pH on

18

T cell activation and function.16, 49

19

We next established mouse 4T1 mammary tumor models and mapped intratumoral

20

pH via multispectral fluorescence imaging (MSFI) using a pH-sensitive fluorescent dye

21

(SNARF-4F) intravenously injected prior to animal sacrifice. Dual channel

22

fluorescence images of the tumor were then acquired ex vivo and correlated with local 8

ACS Paragon Plus Environment

Page 8 of 33

Page 9 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

pH from the calibration curves obtained earlier with biological tissue-like phantoms

2

(Figure 1A and Figure S3). The results revealed low intra-tissue pH at ~6.6 for the 4T1

3

tumors, as compared to the neutral pH in the liver (Figure 1B). In B16-F10 mouse

4

melanoma tumor models, the pH values in the tumors were determined to be 6.5~6.6

5

as assessed with an MI-407 microelectrode (data not shown), which were in line with

6

previous reports on the tumor acidity of the same models.50-52 As anticipated for tumors

7

with an acidic microenvironment, -PD-1 treatment failed to delay their growth,

8

although the failure of -PD-1 therapy can be subject to other co-existing intrinsic or

9

environmental suppressive mechanisms (Figure 1C and 1D).53-57

10 11 12 13 14 15 16 17 18 19 20

Figure 1. Tumor acidity may be associated with the poor therapeutic outcome of -PD-1 therapy in B16-F10 and 4T1 tumor models. (A) Multispectral fluorescence imaging and (B) the corresponding quantification of pH of 4T1 tumor and liver of BALB/c mice. The mice were intravenously injected with a pH-sensitive fluorescent dye (SNARF-4F) at 20 min before sacrifice. The tissues were cut along the central line and intra-tissue pH was determined by fluorescence imaging (em = 580 and 640 nm) using an external calibration curve. Data are shown as mean  SD (n = 3), ****p < 0.0001. (C and D) Growth curves of (C) B16-F10 and (D) 4T1 tumors of mice intravenously administered with -PD-1 (each injection: 75 g per mouse) at times as indicated by red arrows (n = 5).

21 22

Knockdown of LDHA reduced the secretion of lactate and attenuated the acidity

23

in vitro. LDHA is generally upregulated in human cancers due to the accelerated

24

glycolysis of cancer cells, which require the lytic enzyme LDHA to convert 9

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

accumulated pyruvate to lactic acid.58 Using datasets from R2 (http://r2.amc.nl), we

2

confirmed the negative correlation between LDHA expression and the survival of

3

patients with breast cancer and melanoma (Figure S4-S6). Furthermore, a number of

4

studies have also reported that elevated level of serum LDHA could be a poor

5

prognostic factors of patients with various solid tumors including melanoma and breast

6

cancers.59-62 The elevated expression of LDHA was observed in B16-F10 and 4T1 cells,

7

as well as in other cancer cells including EMT-6 (breast cancer cells), CT26 (colorectal

8

cancer cells), and Hepa 1-6 (hepatoma carcinoma) cells (Figure S7). We next screened

9

for appropriate siRNAs against Ldha from three sequences complementary to Ldha (i.e.,

10

siLdha-1, siLdha-2, and siLdha-3, Table S1) in B16-F10 cells. Robust downregulation

11

of Ldha mRNA and the LDHA protein were observed using the siLdha-1 and siLdha-

12

3 sequences, compared to the negative control (siNc, scrambled siRNA), and thus

13

siLdha-1 was used in the rest of the study (Figure S8). The B16-F10 and 4T1 cells were

14

next transfected with selected siLdha at different concentrations (2.5, 5, 10, and 20 nM)

15

and siNc (20 nM) using Lipofectamine® RNAiMAX. Decreased levels of both Ldha

16

mRNA and the LDHA protein were observed in B16-F10 (Figure 2A and 2B) and 4T1

17

(Figure S9) cells with increasing concentrations of siLdha. Upon the knockdown of

18

Ldha, the concentrations of lactate in the cell culture media decreased from 0.32 to 0.20

19

mg/mL (Figure 2C and Figure S10), consistent with the change of pH values from 6.6

20

to 7.4 (Figure 2D and Figure S11). The effect of Ldha knockdown on cell metabolism

21

and proliferation varied a lot among different cells and different reports.25, 63 In our

22

study, the knockdown of Ldha by RNAi did not obviously affect the proliferation of 10

ACS Paragon Plus Environment

Page 10 of 33

Page 11 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

the cells as revealed by a cell metabolic activity assay (Figure S12).

2 3 4 5 6 7 8 9 10 11 12

Figure 2. RNAi-mediated knockdown of LDHA reduced the secretion of lactate by cancer cells and attenuated the acidity of the culture media in vitro. (A) qRT-PCR analyses of Ldha mRNA in B16-F10 cells transfected with siLdha at different concentrations (siLdha, 2.5, 5, 10, and 20 nM) and a scrambled siRNA (siNc, 20 nM) for 24 h. GAPDH (glyceraldehyde 3-phosphate dehydrogenase) was used as the loading control. (B) Western blot analyses of LDHA proteins in whole-cell extracts from B16F10 cells that were untransfected (Ctrl) and transfected with different concentrations of siLdha for 72 h. Tubulin was used as the loading control. (C) The concentrations of lactate and (D) pH values in the supernatants of culture media of B16-F10 cells. Data are shown as mean  SD (n = 3), *p < 0.05, **p < 0.005, ****p < 0.0001.

13 14

Vesicular and micellar nanoparticles for in vivo siRNA delivery. Encouraged by the

15

efficient regulation of lactate and extracellular pH in vitro, we moved forward to in vivo

16

studies in mouse tumor models. We, along with other groups, have reported the use of

17

lipid-polymer hybrid NPs for efficient in vivo RNAi.46, 64-71 We further optimized the

18

formulations by comparing in vivo performances of the two most common types of

19

NPs, i.e., vesicular CLAN nanoparticles (VNPs) and micellar CLAN nanoparticles

20

(MNPs), which were prepared by the double emulsion and single emulsion methods, 11

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 33

1

respectively (Figure 3A). Both NPs were formulated by a cationic lipid N, N-bis(2-

2

hydroxyethyl)-N-methyl-N-(2-cholesterloxycarbonyl

3

bromide (DOTAP) and a FDA-approved polymer poly(ethylene glycol)-block-

4

poly(lactide-co-glycolide) (PEG-PLGA, Mw, PEG = 5000, Mw, PLGA=11000, LG : GA =

5

3 : 1, mol: mol). DOTAP can bind with the negatively charged siRNA, induce fusion

6

with the cell membrane, and enhance the cell uptake. PEG modifications can reduce the

7

clearance of reticuloendothelial system (RES) or phagocyte and increase the circulation

8

time. VNPsiLdha and MNPsiLdha displayed spherical morphology (Figure S13), similar

9

size (average diameter ~90-100 nm) and zeta potential (Figure S14). However, the

10

release of siRNA from VNPs was much slower than that from MNPs, presumably due

11

to the delayed release of siRNA encapsulated within the inner core of the VNP (Figure

12

3A, Figure S15, and Scheme 1). Flow cytometric analysis showed that VNPCy5-siRNA

13

and MNPCy5-siRNA could be efficiently internalized by B16-F10 cells (Figure S16).

14

Consequently, both VNPsiLdha and MNPsiLdha efficiently downregulated Ldha mRNA

15

(Figure S17A) and the LDHA protein (Figure S17B) in B16-F10 cells without causing

16

obvious cytotoxicity (Figure S18). Moreover, transfections with VNPsiLdha and

17

MNPsiLdha did not affect the viabilities of the normal cells of NIH/3T3 and L929

18

fibroblast cells (Figure S19).

amino

ethyl)

ammonium

19

To further compare their in vivo siRNA delivery capabilities, we used time-lapse

20

intravital microscopy (IVM) dual imaging to real-time and simultaneously monitor the

21

time course blood concentrations of NPs and siRNA following systemic administration

22

(Figure 3B). We used spectrally complementary fluorescent dyes rhodamine B (RhoB) 12

ACS Paragon Plus Environment

Page 13 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

and Cy5 to label the VNPs and MNPs (hereafter referred to as RhoB-VNP and RhoB-

2

MNP, respectively) and siRNA (hereafter referred to as Cy5-siRNA), respectively. The

3

Cy5-siRNA exhibited high colocalization with both RhoB-VNP and RhoB-MNP in the

4

blood vessels at 2 min post-injection (Figure 3C). For the MNPCy5-siRNA group, a fast

5

clearance of the Cy5-siRNA was observed at the first 30 min (Ct/Cmax = 36.1% at 30

6

min) followed by a later phase of slower release (Figure 3D). However, the

7

corresponding vehicle RhoB-MNP exhibited relatively slower pharmacokinetics with

8

Ct/Cmax of 70.6% and 39.4% at 30 and 120 min, respectively, indicating the quick

9

separation of Cy5-siRNA and MNP in blood, which were consistent with its burst

10

release of siRNA in vitro. In line with the more gradual in vitro siRNA release from

11

VNP, VNPsiRNA displayed much slower vascular pharmacokinetics for both vesicle and

12

siRNA, as compared to its micellar counterpart. We quantified the colocalization

13

changes of RhoB-NP and Cy5-siRNA by plotting the intensities of their pixels on a

14

scatter plot and calculated the Pearson’s r and slope of siRNA intensity/NP intensity at

15

different time points (Figure 3E). The pixels in the scatter plots of well-colocalized NP

16

and siRNA would distribute along the diagonal, whereas pixels of decreased

17

colocalization would be located off the diagonal. Thus, the decrease of r and slope

18

values indicated the separation of siRNA with NP. At 2 min post-injection, both VNP

19

and MNP were well-colocalized with siRNA (r > 0.7). However, as time progressed,

20

the r and slope value of MNP-siRNA decreased much faster than VNP-siRNA (Figure

21

3F), indicating that siRNA is easier to detach from MNP than VNP, presumably due to

22

the release of surface bound siRNA replaced by serum proteins. On the contrary, the 13

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

vesicle shell of VNP protected the siRNA encapsulated in the core from premature

2

exposure to serum proteins, prolonged its circulation time, and enhanced the tumor

3

accumulation of siRNA. Agarose gel electrophoresis analyses supported the better

4

siRNA protection capability of VNP compared to MNP (Figure S20).

5

We further assessed the tumoral accumulation behaviors of Cy5-siRNA loaded in

6

VNP and MNP using an IVIS in vivo imaging system. The time-dependent fluorescence

7

imaging of 4T1 tumor-bearing mice showed that VNPCy5-siRNA delivered significantly

8

more Cy5-siRNA into the tumor as compared to MNP, which was supported by the

9

superior blood circulation of VNPCy5-siRNA (Figure 3G and Figure S21). In addition,

10

direct imaging of excised organs and tumor tissues at 24 h post-injection confirmed the

11

above observation (Figure S22). Together, these data showed that VNP enabled a

12

longer circulation half-life and superior tumor accumulation of siRNA.

14

ACS Paragon Plus Environment

Page 14 of 33

Page 15 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Figure 3. Vesicular and micellar cationic lipid-assisted polymeric nanoparticles for in vivo siRNA delivery. (A) Schematic showing the preparation of vesicular (VNPs) and micellar (MNPs) nanoparticles via the double and single emulsion methods, respectively. (B) Schematic of intravital microscopy imaging to simultaneously monitor pharmacokinetics of both nanoparticles and siRNA in the ear vessels of mice. C) Time-lapse confocal fluorescence microscopy observation of mice revealing the pharmacokinetics of rhodamine B-labelled VNP (RhoB-VNP) and MNP (RhoB-MNP) encapsulating Cy5 labelled siRNA (Cy5-siRNA). (D) Time-dependent changes in the concentrations of vehicles (i.e., VNP and MNP) and payloads (i.e., Cy5-siRNA) in blood were simultaneously monitored and quantified. (E) Scatter plots of siRNA and NPs distributed in vessels at the indicated time points. The selected non-colocalization pixels from the scatter plot were indicated by the blue boxes. Slopes of siRNA intensity/NP intensity are indicated by red dashed lines. (F) Time-dependent changes of correlation coefficients (referred to as Pearson’s r) of RhoB-VNPs or RhoB-MNPs with Cy5-siRNA. (G) Quantitative analyses of accumulation and retention of Cy5siRNA in tumors of mice received intravenous injections of VNPCy5-siRNA and MNPCy5siRNA. Data are shown as mean  SD (n = 3 per group). 15

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2

VNPsiLdha treatment attenuated tumor acidity, inhibited tumor growth, and

3

potentiated -PD-1 response. Based on the above results, we next used VNP as the

4

vehicle for in vivo gene silencing and investigated the tumor growth after Ldha

5

knockdown. When the tumor size reached 80-100 mm3, the mice bearing subcutaneous

6

B16-F10 tumors received the following intravenous treatments: PBS; free siLdha; -

7

PD-1, VNPsiNc, VNPsiLdha, and VNPsiLdha+-PD-1 at the siLdha dose of 2 mg/kg and -

8

PD-1 dose 75 ug/mouse at times as indicated in Figure 4A. The tumor volume and

9

weight of mice were measured every day after the first treatment. The B16-F10 tumors

10

grew at a fast rate without any treatments (Figure 4B). Knockdown of Ldha via

11

VNPsiLdha for 7 times (group II) resulted in significant inhibition of tumor growth (by

12

35.1%) as compared to the control group (group I, Figure 4B). We further assessed

13

whether VNPsiLdha treatment would augment antitumor responses in combination with

14

-PD-1 therapy. Administration of -PD-1 alone had no obvious therapeutic effect due

15

to the immunosuppressive tumor microenvironment (group VI, Figure 4C). However,

16

we observed that combinations of VNPsiLdha with -PD-1 significantly slowed the

17

tumor growth by 68.2% (Group III, Figure 4B). The results also confirmed that free

18

siLdha treatment had no influence on the tumor growth due to the fast clearance and

19

degradation of siRNA during circulation and the lack of nanoparticle protection (group

20

IV, Figure 4C). The tumor volume (Figure 4D) and weight (Figure 4E) measurements

21

of the tumor tissue excised on the 19th day post-inoculation confirmed the superior

22

therapeutic effect of VNPsiLdha combined with -PD-1. The collected B16-F10 tumors 16

ACS Paragon Plus Environment

Page 16 of 33

Page 17 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

were fixed and sectioned for immunohistochemical staining. Representative images and

2

quantitative analyses confirmed the remarkable downregulation of the LDHA protein

3

(Figure 4F and 4G). Next, the pH values in the tumors were assessed with an MI-407

4

microneedle, which showed that the tumoral pH increased from ~6.5-6.6 for the control

5

mice to ~7.0 of the mice received VNPsiLdha treatment (Figure 4H). The treatment did

6

not cause significant body weight change (Figure S23), blood biochemical indicators

7

change (Figure S24), and morphological changes of major organs of mice (Figure S25),

8

indicating its safety potential.

9

Similar trends in the growth inhibition by VNPsiLdha and VNPsiLdha+-PD-1 were

10

observed in mice with 4T1 tumors (Figure 4I-4L and Figure S26). Measurement of

11

tumor pH using MSFI imaging confirmed the normalization of the tumoral acidic

12

microenvironment (Figure 4M and 4N). Therefore, the above results demonstrated that

13

the delayed tumor progression was related to the reversion of the tumor acidity. In

14

addition, T cells might participate in an enhanced anti-cancer process based on Ldha

15

knockdown.

17

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Figure 4. VNPsiLdha treatment normalized the tumor acidic microenvironment, inhibited tumor growth, and potentiated checkpoint blockade therapy in B16-F10 and 4T1 tumor models. (A) Treatment schedule for immunocompetent mice bearing B16-F10 or 4T1 tumors using VNPsiLdha combined with -PD-1-based checkpoint blockade therapy. The mice were intravenously administered VNPsiLdha (each injection: 2 mg siLdha per kg of mouse weight) and α-PD-1 (each injection: 75 g α-PD-1 per mouse) at times indicated by the arrows. (B and C) Growth curves of B16-F10 tumors during treatment. (D) Volume and (E) weight of the tumor tissues excised on the 19th day. (F) Representative immunohistochemical staining of LDHA in the tumor tissues and (G) the corresponding quantitative expression analyses. (H) The pH value of the B16-F10 tumor tissues measured on the 19th day after treatment. The pH values were measured in vivo with a pH microneedle probe. (I and J) Growth curves of 4T1 tumors during treatment. (K) Volume and (L) weight of the tumor tissues excised on the 19th day. (M) Typical fluorescence images of all of the tumor tissues. The mice were intravenously injected with a pH-sensitive fluorescent dye (SNARF-4F) at 20 min before sacrifice. The tumors were cut along the central line and intratumoral pH was 18

ACS Paragon Plus Environment

Page 18 of 33

Page 19 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1 2 3 4

determined with multispectral fluorescence imaging (MSFI) using an external calibration curve. (N) The pH value of the 4T1 tumor tissues measured on the 19th day after tumor inoculation. Data are shown as mean  SD (n = 5 per group), **p< 0.005, ***p