intra articular lentivirus mediated gene therapy targeting cracm1 for the treatment of collagen induced arthritis

Full paperIntra-articular lentivirus-mediated gene therapy targeting CRACM1 for the treatment of collagen-induced arthritis Shuang Liua,*, Takeshi Kiyoib, Erika Takemasaa, Kazutaka Maeya

Full paper

Intra-articular lentivirus-mediated gene therapy targeting CRACM1

for the treatment of collagen-induced arthritis

Shuang Liua,*, Takeshi Kiyoib, Erika Takemasaa, Kazutaka Maeyamaa

a Department of Pharmacology, Ehime University Graduate School of Medicine, Shitsugawa, Toon-shi, Ehime 791-0295, Japan

b Department of Bioscience, Integrated Center for Sciences, Ehime University, Shitsukawa, Toon-shi, Ehime, 791-0295, Japan

a r t i c l e i n f o

Article history:

Received 18 August 2016

Received in revised form

16 January 2017

Accepted 6 February 2017

Available online xxx

Keywords:

Intra-articular injection

Lentivirus

Gene silencing

Calcium release-activated calcium channel 1

Collagen induced arthritis

a b s t r a c t

Abnormal store-operated calcium uptake has been observed in peripheral T lymphocytes of rheumatoid arthritis (RA) patients, and sustained intracellular calcium signalling is known to mediate the functions of many types of immune cells Thus, it is hypothesized that regulating calcium entry through CRACM1 (the pore-forming subunit of calcium release-activated calcium (CRAC) channels; also known as ORAI1) may

be beneficial for the management of RA Localized CRACM1 knockdown in the joints and draining lymph nodes (DLNs) of mice with collagen-induced arthritis (CIA) was achieved via lentiviral-based delivery of shRNA targeting mouse CRACM1 Consistent with CRACM1 knockdown, calcium influx in synovial cells and the histopathological features of CIA were reduced These effects were also associated with reduced levels of several notable inflammatory cytokines, such as IL-6, IL-17A, and IFN-g, in the joints Addi-tionally, CRACM1-shRNA reduced the number of bone marrow-derived osteoclasts in vitro as well as osteoclasts in CIA joints, which was associated with reduced RANKL levels in the serum and joints In summary, inhibiting calcium entry by CRACM1 knockdown suppressed arthritis development and may

be therapeutically beneficial for RA patients

© 2017 The Authors Production and hosting by Elsevier B.V on behalf of Japanese Pharmacological Society This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/

licenses/by-nc-nd/4.0/)

1 Introduction

Rheumatoid arthritis (RA) is a systemic autoimmune disease

and one of the most prevalent forms of arthritis The treatment of

RA has improved dramatically over the past decade since the

introduction of disease-modifying antirheumatic drugs (DMARDs)

In addition to synthetic chemical compounds, the systemic delivery

of biological agents, especially the protein antagonists of tumour

necrosis factor (TNF), has helped patients achieve RA remission

However, less than 30% of patients show a robust response to these

drugs, which are costly and associated with many side effects

related to their systemic mode of delivery(1) Moreover, even in

responsive patients, individual joints may not respond to therapy

Therefore, novel therapies that deliver drugs locally and specifically

to a small number of target joints must be considered for the

management of RA

In recent studies, the regulation of Ca2þentry through a store-operated Ca2þrelease-activated channel (CRAC), known as ORAI, has shown benefits in the treatment of RA CRAC activation drives gene expression and promotes growth and proliferation of T cells, B cells, and osteoclasts, which are important cellular targets for the management of RA(2) The modulation of Ca2þentry through the systemic gene silencing of CRACs showed great therapeutic po-tential for the control of global immune responses in an experi-mental arthritis model(3)

Of the three CRAC homologues, CRAC modulator 1 (CRACM1) is widely expressed in human and murine tissues and is fundamental

to cell physiology In our previous cross-sectional study, the expression level and functional status of CRACM1 in peripheral CD4þT cells was increased in active RA patients compared with osteoarthritis patients and healthy donors (4) However, in the experimental arthritis model, systemic high doses or repeated low doses of inhibitors targeting CRACM1 caused severe side effects, including low body weight and splenomegaly, in addition to low survival rates(3) Because normal Ca2þhomeostasis is fundamental and its efficiency cannot be achieved via the systemic knockdown

of CRACM1, as a new attempt, local immunosuppression by the

* Corresponding author.

E-mail address: liussmzk@m.ehime-u.ac.jp (S Liu).

Peer review under responsibility of Japanese Pharmacological Society.

Contents lists available atScienceDirect Journal of Pharmacological Sciences

j o u r n a l h o m e p a g e :w w w e l s e v i e r c o m / l o c a t e / j p h s

http://dx.doi.org/10.1016/j.jphs.2017.02.001

1347-8613/© 2017 The Authors Production and hosting by Elsevier B.V on behalf of Japanese Pharmacological Society This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

Journal of Pharmacological Sciences xxx (2017) 1e9

silencing of CRACM1 may be beneficial in controlling the symptoms

of individual joints with RA

To investigate the feasibility and efficiency of the local

regula-tion of Ca2þ entry following treatment, we achieved local gene

silencing of CRACM1 using intra-articular lentivirus-delivered short

hairpin RNA (shRNA) in a collagen-induced arthritis (CIA) model

Systemic and local immunological and pathological assessments

were performed to evaluate the therapeutic effects of the

regula-tion of Ca2þentry via CRACM1 in CIA mice

2 Methods

2.1 Lentiviral constructs

The lentiviral vectors were produced by the co-transfection of

293T kidney cells with the following three plasmids: (1) a specific

(M1shRNA) (NM_175423) (pLKO.1-puro): CCGGCACAACCTCAACTCG

GTCAAACTCGAGTTTGACCGAGTTGAGGTTGTGTTTTTG and a negative

control shRNA vector (NCshRNA) (SigmaeAldrich, Tokyo, Japan); (2)

the packaging plasmid psPAX2 (Addgene, Cambridge, MA, USA); and

(3) the envelope plasmid pMD2.G (Addgene) Lentiviral particles

containing M1shRNA and NCshRNA were produced, and their titres

were determined using a Lenti-X™ P24 Rapid Titer Kit (Clontech

Laboratories, Mountain View, CA, USA) After testing the multiplicity

of infection, the lentiviral particles were used for the reported series

of experiments

2.2 Intra-articular delivery of CRACM1 shRNA lentivirus in

collagen-induced arthritis (CIA) mice and arthritis evaluation

Male DBA/1J mice (7e10 weeks of age) were used for the

in-duction of arthritis The mice were bred and maintained under

standard conditions The experimental protocols were performed

in accordance with the guidelines of the Animal Care Committee of

Ehime University and were approved by the University Committee

for Animal Research

Arthritis was induced as described previously(3) Briefly, the

mice were injected subcutaneously (s.c.) with bovine collagen type

II (CII; Chondrex, Redmond, WA, USA) emulsified in complete

Freund's adjuvant (CFA) containing heat-killed Mycobacterium

tuberculosis H37 RA Twenty-one days later, a booster injection of CII

emulsified in incomplete Freund's adjuvant (IFA) was administered

s.c at the base of the tail To synchronize the onset of arthritis,

lipopolysaccharide was injected intraperitoneally (i.p.) on day 28

after thefirst CII immunization Mice injected s.c with saline were

used as a negative control

CRACM1-shRNA and Control-shRNA were administered by a

single intra-articular injection to each ankle joint of the CIA mice

using a virus dose of 108particles in 5ml of phosphate-buffered

saline on day 21 or day 30 after thefirst CII immunization On

day 35, the number of integrated provirus copies in tissue was

determined using a Lenti-X Provirus Quantitation Kit (Clontech

Laboratories, Mountain View, CA, USA) Genomic DNA extracted

from the tissues of lentivirus-treated mice was subjected to

quan-titative real-time PCR amplification with a dilution of the calibrated

provirus control template according to the manufacturer's

instructions

Following the development of CIA, the ankle circumferences and

the articular indexes of the mice were evaluated every 5 days, from

day 20e50, under blinded conditions by two independent

exam-iners Clinical arthritis scores were evaluated from 0 to 4 as follows:

0¼ no swelling, 1 ¼ slight swelling and erythema, 2 ¼ moderate

swelling and oedema, and 4¼ joint rigidity Each limb was graded,

and a total score for all 4 limbs was summed for a maximum possible score of 16 per animal

The swelling rates of each joint were evaluated using MRI ex-aminations after scoring for signs of arthritis every 5 days, from day

20e50 MRI was performed using a MRmini SA110 scanner (DS Pharma Biomedical, Co., Ltd., Osaka, Japan) The animals were anesthetized using 1.5 vol% isoflurane vapourized in 100% medical oxygen Imaging was performed using a 3-dimensional T2-weighted flash sequence The coronal and sagittal images were collected and reconstructed to obtain the volume of the arthritic ankles and forepaws

The structural changes of the hind paws were visualized by micro-computed tomography (micro-CT) imaging at day 106 after thefirst CII immunization The animals were anesthetized using 1.5 vol% isoflurane vapourized in 100% medical oxygen Then, the mice were subjected to micro-CT imaging using a LaTheta LCT-200 CT scanner (Hitachi Aloka Medical, Ltd., Tokyo, Japan), and the pro-jection images were reconstructed into 3-dimensional images us-ing the VGStudio MAX 2.2 software (Volume Graphics, Heidelberg, Germany)

Histological examination of the joints was undertaken on day 50

of the experiment after routine fixation, decalcification, and paraffin embedding Tissue sections from the hind knee joints were stained with H&E The slides were evaluated for synovial hyper-trophy, formation and cartilage/subchondral bone destruction by two blinded observers Each joint was scored from 0 to 3, and 2 joints were analysed for each animal

2.3 Inflammation-related cytokine measurements For the preparation of murine joint protein extracts, the syno-vium of the infra-patellar fat pad and joint cartilage was harvested Total proteins were extracted from tissues using a ReadyPrep™ extraction kit in accordance with the manufacturer's instructions (BIO-RAD, Hercules, CA, USA)(5)

Cytokine levels in the serum and joint protein extracts were detected on day 50 of the experiment using a bead-based immu-noassay with the LEGENDplex™ Mouse Inflammation Panel (Bio-legend, San Diego, CA, USA) according to the manufacturer's instructions The lower limits of detection were 1.3 pg/ml for IL-1a, 2.8 pg/ml for IL-1b, 0.9 pg/ml for IL-6, 2.1 pg/ml IL-10, 0.7 pg/ml for 12p70, 1.8 pg/ml for 17A, 4.2 pg/ml for 23, 9.8 pg/ml for

IL-27, 1.7 pg/ml for MCP-1, 4.0 pg/ml for IFN-b, 0.8 pg/ml for IFN-g, 1.9 pg/ml for TNF-a, and 1.9 pg/ml for GM-CSF

2.4 Cell differentiation of splenocytes and synovial cells Splenocytes were isolated from spleens on day 50 of the experiment using a standard dissection technique and resuspended

in RPMI-1640 for staining(6) For the isolation of synovial cells, the synovia in the infra-patellar fat pad of the mice were harvested on day 50 of the experiment and digested in Dulbecco's modified Ea-gle's medium (SigmaeAldrich, At Louis, MO, USA) containing 0.1% collagenase D and 0.005% deoxyribonuclease I at 37C for 1 h The digested synovial cells werefiltered through a 70-mm cell strainer For the detection of T cell subsets and B cell subsets, splenocytes and synovial cells were stained using a PE-mouse B220 anti-body and FITC-anti-mouse CD3 antianti-body For Th17 cell staining, dispersed cell suspensions from the spleens and synovia were resuspended at 5 105cells/ml and incubated with a cell activation cocktail (Biolegend), which consists of optimized concentrations of phorbol 12-myristate-13-acetate, ionomycin, and Brefeldin A, for

6 h After extracellular staining with PE-anti-mouse CD4, the cells were permeabilized with Fix/Perm solution (Biolegend) for 20 min

at 4C The cells were stained with Alexa Fluor 488-anti-mouse

IL-S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 2

17A All the antibodies were obtained from Biolegend The data

were acquired on a BD FACSAria (BectoneDickinson, Oakville, ON,

Canada) and analysed using FlowJo software (Tree Star Inc.,

Ash-land, OR, USA)

2.5 Calcium influx in splenocytes and synovial cells

Single-cell suspensions (105cell/ml) pooled from the spleen and

the synovial membrane were obtained 35 days post-immunization

After lyse the red blood cells, the cells were loaded with fura-2-AM,

and the assay was performed using afluorometric imaging plate

reader (FLIPR) (FlexStation II, Molecular Devices Japan, Tokyo,

Japan)(7) The plates were read for a total of 700 s, including an

initial 100-s reading window to measure the baselinefluorescence

levels before the application of any compound with a 0 mM-Ca2þ

bath solution The plates were read for an additional 300 s after

0.5mM thapsigargin was applied using the FLIPR Then, 2 mM Ca2þ

was applied to the wells, and the plates were read for an additional

300 s The calcium signals were read using the 340/380 nm

exci-tation and 510 nm emission settings The results are presented as a

ratio of RFUs (340 nm/380 nm) and the initial rates of Ca2þinflux

(in thefirst 15 s after Ca2 þaddition).

2.6 Isolation of differentiated osteoclasts and bone-resorption

assay

Differentiated osteoclasts were purified using collagen films

according to a previous protocol(3) Briefly, bone marrow cells,

which include osteoclast precursors, were plated on a type I

collagen-coated culture plate on day 50 of the experiment The cells

were fed recombinant mouse macrophage colony-stimulating

fac-tor (M-CSF) (20 ng/ml) (ProSpec-Tany Technogene Ltd., East

Brunswick, NJ, USA) and soluble receptor activator of nuclear factor

kappa-B ligand (sRANKL) (200 ng/ml) (ProSpec-Tany Technogene

Ltd) every other day One week later, the collagenfilm was digested

with a solution of 0.1% collagenase A The harvested cells were

plated on chamber slides or bone slices and fed with M-CSF (20 ng/

ml) and RANKL (250 ng/ml) ina-MEM After 48 h of incubation, the

cells that were plated on the chamber slides were stained for

tartrate-resistant acid phosphatase (TRAP) using a TRAP/ALP

staining kit The bone slices were subjected to Mayer's

hematoxylin staining and scanning electron microscopy for pit

formation assay

2.7 Statistical analysis

All experiments were designed in a completely randomized

multifactorial format The results are expressed as the mean± SEM

Repeated measures were probed with analysis of variance

(ANOVA), followed by the Scheffe F-test, to evaluate the statistical

significance of the differences according to the arthritis scores and

the swelling rates of the arthritic joints Other data were analysed

using a two-sample t-test assuming unequal variance P

values< 0.05 were considered statistically significant

3 Results

3.1 The inhibition of Ca2þinflux in synovial cells by the local gene

silencing of CRACM1

Lentiviral shRNA targeting CRACM1 was intra-articularly

administered (2  108 lentiviral particles) to experimental CIA

mice The CIA model shares immunological and pathological

sim-ilarities with human RA, making it useful for RA studies The

sur-vival rates for the CRACM1-expressing lentivirus

(CRACM1-shRNA)-treated mice and the control-shRNA-expressing lentivirus (Control-shRNA)-treated CIA mice were both 100% The bio-distribution of the inoculated vector was described using the number of integrated provirus copies in individual tissue-derived cells (Fig 1a) The majority of the provirus copies were observed

in the joints and popliteal lymph nodes (LNs) Few virus copies were detected in the blood, lungs, liver, spleen, and heart Quan-titative PCR (qPCR) demonstrated the inhibition of CRACM1 in in-dividual tissues, especially in the synovium and draining LNs (Fig 1b) The lack of gene silencing of CRACM1 in the kidneys, bone marrow, and blood suggests that the lentivirus-mediated gene delivery was limited to specific sites

Cell suspensions pooled from spleen, thymus, and knee syno-vium were obtained 35 days post-immunization for Ca2þ influx analysis Intracellular Ca2þwas labelled with fura 2, and the oscil-lation of intracellular Ca2þwas reflected by the relative fluores-cence units at 340 nm/380 nm (RFU340/380) After the addition of thapsigargin to the 0 mM Ca2þ extracellular bath solution, the intracellular Ca2þstores were depleted A sustained increase in the

Ca2þinflux was observed in thapsigargin-stimulated cells upon the addition of 2 mM Ca2þto the extracellular bath solution According

to the RFU340/380values for each group, the peaks and initial rates of

Ca2þinflux were higher for the splenocytes, thymocytes, and sy-novial cells from CIA mice than those from the naive mice (peaks: 1.51-fold, 2.52-fold (P< 0.05), and 2.16-fold (P < 0.05), respectively;

influx rates: 1.61-fold, 1.40-fold, and 1.43-fold, respectively) (Fig 1c and d) The Ca2þ influx peaks in synovial cells from CRACM1-shRNA-treated CIA mice were suppressed to 37.8% (P < 0.05) of the peaks in cells from control-shRNA-treated CIA mice The initial rates of Ca2þinflux in the cells from CRACM1-shRNA-treated CIA mice decreased to 19.76% (P< 0.05) of the rates in the cells from control-shRNA-treated CIA mice No significant differences were found in the peaks and initial rates of Ca2þinflux in splenocytes and thymocytes from control-shRNA-treated CIA mice and CRACM1-shRNA-treated mice

These results indicated that Ca2þinflux via CRACs in synovial-derived cells was suppressed by local lentivirus-mediated CRACM1 silencing Moreover, the influence of the CRACM1-shRNA treatment was locally limited

3.2 Intra-articular CRACM1-shRNA reduced CIA severity

We next investigated the potential therapeutic application of CRACM1-shRNA in CIA mice Control-shRNA was given on the onset

of disease (day 21 post-immunization, day 0 being the day offirst immunization), whereas CRACM1-shRNA was given on day 21 or after the onset of arthritis (day 30) A single intra-articular injection

of CRACM1-shRNA at disease onset (21 days after thefirst sensiti-zation) abrogated early symptoms or when arthritis was already established, completely abrogated clinical symptoms (30 days after the first sensitization), significantly reducing further disease (Fig 2a) An intra-articular injection of the lentiviral vector temporarily promoted joint swelling, and the effect was transient (Fig 2b) The administration of CRACM1-shRNA at disease onset delayed the establishment of arthritis but led to a less favourable prognosis compared with injections administered after the disease was established based on the clinical scores and swelling rates (Fig 2a and b) Intra-articular injections of CRACM1-shRNA in CIA mice with full clinical symptoms (30 days after thefirst sensitiza-tion) progressively attenuated the severity of CIA and reduced late bone deformity (60 days after thefirst sensitization) (Fig 2c) The histopathological analysis of the joints on day 50 of the experiment using haematoxylin/eosin (H&E) staining showed that the delayed administration of CRACM1-shRNA reduced CIA-characteristic

S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 3

chronic inflammation, including joint inflammation, cartilage

destruction, and bone histomorphometry analysis (Fig 2d and e)

Taken together, these results indicated that local gene silencing

of CRACM1 at the acme phase decreased arthritis severity and

improved prognosis in the CIA murine model

3.3 Reduced infiltration of inflammatory cells by local gene

silencing of CRACM1

To investigate the cellular and molecular mechanisms

underly-ing the decreased severity of CIA, we evaluated the differentiation

potential of immune cells derived from mouse synovium and

spleen on day 50 of the experiment Based on theflow cytometry

analysis of mouse synovium-derived and spleen-derived primary

cells, the intra-articular injection of lenti-M1shNRA decreased the

percentage of B220þlymphocytes and CD4þIL-17þ lymphocytes

(Fig 3aef) No significant differences were observed in the

per-centages of CD3þ lymphocytes in synovial cells and splenocytes

between CRACM1-shRNA-treated mice and control-shRNA-treated

mice

We also evaluated inflammatory mediators in the joint protein

extracts and serum on day 50 of the experiment The cytokine

profile of the joint protein extracts suggested that the local gene

silencing of CRACM1 reduced the protein expression of in

flamma-tory cytokines, including IL-6, IL-12, IL-17A, IL-27, TNF-a, IFN-g, and

GM-CSF in the local joints of CIA mice (Fig 3g) In addition, lower

levels of cytokines, such as IL-6, IL-17A, IL-23, IL-27, and GM-CSF, were detected in the serum of lenti-M1hRNA-treated mice than

in the serum of control-shRNA-treated mice (Fig 3h) Release of the suppressive cytokine IL-10 was increased in the joint protein ex-tracts and serum after intra-articular CRACM1-shRNA treatment These data indicated that intra-articular CRACM1-shRNA treatment modulated the inflammatory response, including the differentia-tion of inflammatory cells and the release of inflammatory medi-ators, in the joints and periphery of CIA mice

3.4 Intra-articular gene silencing of CRACM1 inhibited the activity

of mature osteoclasts Osteoclastogenic cells may be potential therapeutic targets of intra-articular CRACM1-shRNA treatment; therefore, we evaluated the activity of mature osteoclasts in joints Bone marrow cells were isolated from CRACM1-shRNA-treated CIA mice and control-shRNA-treated CIA mice on day 50 of the experiment and cultured in the presence of M-CSF and RANKL After 6 days of cul-ture, mature osteoclasts were harvested from detached collagen films

The total number of TRAP-positive cells in CIA mice increased 3.62-fold compared with saline-sensitized naive mice (P< 0.01) (Fig 4aec, m) Gene silencing of CRACM1 around the arthritic joint resulted in a lower number of TRAP-positive cells (47.27%, P< 0.05)

in CRACM1-shRNA-treated mice than in control-shRNA-treated CIA

Fig 1 CRACM1-shRNA-mediated local inhibition of CRACM1 expression in mice (a) The biodistribution of the inoculated vector Four weeks after the intra-articular delivery of lentiviral particles, the number of integrated provirus copies in individual tissues was determined (b) Real-time PCR amplification of the reverse-transcription products predicted shRNA-mediated mRNA suppression in individual tissues 4 weeks after the intra-articular delivery of lentiviral particles (Control-shRNA vs CRACM1-shRNA; *, P < 0.05; **, P < 0.01;

n ¼ 8e10) (c) The Ca 2þ influx peaks in splenocytes, thymocytes, and synovial cells obtained from naive mice, CRACM1-shRNA-treated mice, control-shRNA-treated CIA mice and CRACM1-shRNA-treated CIA mice Single-cell suspensions (10 6 cells/ml) were obtained 35 days post-immunization The assay was performed using a fluorometric imaging plate reader (FLIPR) Calcium signals were read using the 340/380 nm excitation and 510 nm emission settings The results are presented as the peak ratio of relative fluorescence units (RFUs) (340 nm/380 nm) (d) The Ca2þinflux initial rates (in the first 15 s after Ca 2þ addition) in splenocytes, thymocytes, and synovial cells obtained from naive mice, CRACM1-shRNA-treated mice, control-CRACM1-shRNA-treated CIA mice and CRACM1-CRACM1-shRNA-treated CIA mice The results are expressed as the mean ± SEM (naive vs CRACM1-shRNA; *, P < 0.05; CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; **, P < 0.01; n ¼ 5).

S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 4

mice Additionally, TRAP staining of the knee joint sections

sug-gested that CRACM1-shRNA-treated CIA mice had significantly

reduced osteoclast activity compared with control-shRNA-treated

CIA mice (Fig 4def) The quantification analysis revealed that

local CRACM1 silencing reduced osteoclast activity to 19.21%

(P< 0.01) (Fig 4n) The pit formation area on the bone slides of

CRACM1-shRNA-treated CIA mice was significantly decreased

compared with bone slides of control-shRNA-treated CIA mice

(Fig 4gel, o) These results suggest that the bone-resorption

ca-pacity of primary mature osteoclasts from CRACM1-shRNA-treated

CIA mice was significantly reduced compared with those from

control-shRNA-treated CIA mice

Because pro-inflammatory cytokines in CIA mice may trigger

the differentiation of osteoclasts by stimulating T cells to produce

RANKL(8), we assessed the RANKL levels in the serum and protein

extracts from joints on day 50 of the experiment (Fig 4p) Local treatment with CRACM1-shRNA inhibited the production of RANKL

in the serum by 44.64% (P< 0.05) and in the knee joints to 6.56% (P< 0.001)

Taken together, these results demonstrated that intra-articular treatment with CRACM1-shRNA down-regulated osteoclast activ-ity by reducing local and systemic RANKL production in CIA mice

4 Discussion

To date, the therapeutic targets for biological DMARDs in clinical trials and those approved for use include T cells, B cells, and oste-oclasts, or their associated cytokines and activation-related intra-cellular signalling proteins(9) In the current study, by targeting T cells, B cells, and osteoclasts in arthritic joints, we demonstrated

Fig 2 Application of CRACM1-shRNA reduces the severity of collagen-induced arthritis (CIA) (a) The clinical severity of arthritis was scored every 5 days from day 20 to day 50 of the experiment CRACM1-shRNA viral particles were injected on day 20 or 30 of the experiment; Control-shRNA viral particles were injected on day 20 (naive vs other experimental groups; *, P < 0.05; CIA þ CRACM1shRNA vs CIA; ƚ, P < 0.05; ǂ, P < 0.01; n ¼ 13e22) (b) The swelling rates of the arthritic ankles and forepaws Coronal and sagittal MRI images were collected to obtain the volume of arthritic ankles and forepaws every 5 days from day 20e50 The ratio of the increased volume after antigen sensitization to the volume in the pre-sensitization period was calculated (naive vs other experimental groups; *, P < 0.05; CIA þ CRACM1shRNA vs CIA; ƚ, P < 0.05; n ¼ 13e22) (c) Upper panel: MRI images of arthritic ankles and forepaws on day 50 of the experiment Imaging was performed using a 3-dimensional T2-weighted flash sequence Lower panel: Micro-CT images of arthritic ankles and forepaws on day 81 of the experiment (d) Histological analysis of the paw joints of control-shRNA- and Lenti-M3shRNA-treated CIA mice using haematoxylin/eosin (H&E) staining

on day 50 of the experiment (e) The results of the histological analysis were quantitated by scoring for inflammation, bone histomorphometry analysis (BHA), and cartilage damage (CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; *, P < 0.05) The results are expressed as the mean ± SEM (n ¼ 5e8).

S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 5

the feasibility of inhibition of CRACM1 in the treatment of CIA.

CRACM1 protein locates on the plasma membrane of T cells, B cells,

and osteoclasts and it is essential for forming CRAC channels Since

store-operated Ca2þentry (SOCE) via CRACM1 is fundamental to

the function of most non-excitable cells and intracellular Ca2þ

homeostasis is also vital to these cells, in vivo systemic

manipula-tion of the funcmanipula-tion of CRACM1 is a challenge Systemic suppression

of SOCE via CRAC channels by perturbing the function of CRACM1

caused systemic toxicity by inducing hepatorenal syndrome and

glycaemia (unpublished data), and is not considered safe In present

study, we attempted to locally instead of systematically suppress

CRACM1 Intra-articular manipulation of CRACM1 appeared to be

well-tolerated by CIA mice and the treatment of local joints

suc-cessfully improved the systemic status

The goal of the intra-articular CRACM1-shRNA treatment was

to provide a high synovialfluid concentration that acted locally at

the joints to silence CRACM1 and to reduce local inflammation

through anti-inflammatory effects while limiting the systemic

response We demonstrated that the local immune responses in

an experimental arthritis model were suppressed by targeting

CRACM1 using gene therapy The results indicated that efficient

local gene silencing with specifically designed CRACM1 shRNA

led to the suppression of synovial inflammation by inhibiting the

activity and differentiation of immune cells Moreover, the strong functional blockage of T cells downregulated the production of RANKL The functional inhibition of CRACM1 in arthritic joints reduced subsequent bone loss and late bone deformity as a result

of decreased osteoclast activity Therefore, our results suggest that intra-articular gene therapy targeting CRACM1 is beneficial

in the treatment of individual arthritic joints

To deliver sustained, therapeutic amounts of molecules to joints via the circulation requires repeated systemic administration, which is effective for RA because the condition has systemic, extra-articular involvement However, systemic administration is a disadvantage because this type of drug delivery does not specif-ically target drugs to the joints Therefore, non-target organs are exposed to high concentrations of the drug, which increases the potential for unwanted side effects(10) The systemic delivery of CRACM1 suppressors, including YM-58483, which is a small mo-lecular blocker of store-operated Ca2þ entry via CRAC(11), and lentiviral delivery of shRNA formulations to CIA mice caused severe side effects Moreover, the risk of insertional mutagenesis based on viral vectors should always be a consideration in systemic gene therapy The direct injection of lentiviral particles into the articular cavity could overcome physiological barriers to entry while limiting systemic side effects The gene-based transduction of joint cells,

Fig 3 Administration of CRACM1-shRNA modulates the differentiation potential of immune cells and inhibits the inflammatory response in CIA mice CRACM1-shRNA viral particles were injected on day 30 of the experiment For Th17 cell staining, dispersed cell suspensions were incubated with the cell activation cocktail The cells were stained using specific antibodies, and the data were acquired on a BD FACSAria flow cytometer (a) Representative plots and combined data of the CD3 and B220 expression levels in (b) splenocytes and (c) synovial cells are shown (d) Representative plots and combined data of the CD4 and IL-17 expression levels in (e) splenocytes and (f) synovial cells are shown (CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; *, P < 0.05; **, P < 0.01; n ¼ 5e6) The cytokine profiles in (g) joint protein extracts and (h) serum were determined on day 35 post-immunization using the LEGENDplex™ Mouse Inflammation Panel The results are expressed as the mean ± SEM (CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; *, P < 0.05; **,

P < 0.01; n ¼ 8e10).

S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 6

whereby CRACM1-targeting shRNA is maintained locally for an

extended period, may provide sustained, therapeutic

concentra-tions that lack the peaks and troughs of intermittent application By

local application to peripheral joint tissues, the systemic toxic risks

of viral vectors may be much more moderate than those in systemic

application The cells carrying random integration sites, which are

probably mutagenic, are locational limited In this study, a single intra-articular injection using relatively low copies of virus particles silenced CRACM1 around the joint and ameliorated the symptoms

of CIA The local delivery of CRACM1-shRNA at the acme phase successfully suppressed CRACM1 activity and showed anti-CIA benefits

Fig 4 CRACM1-shRNA-mediated CRACM1 silencing inhibits the activity of mature osteoclasts in collagen-induced arthritis (CIA) mice CRACM1-shRNA viral particles were injected

on day 30 of the experiment Bone marrow cells were isolated on day 50 of the experiment and differentiated into osteoclasts (aec) TRAP staining of the cells, which were plated on chamber slides (def) Joint sections were detected using TRAP staining on day 50 of the experiment Paraffin-embedded joint sections were stained using a TRAP/ALP-staining kit and counterstained with haematoxylin The sections were using a by light microscope (600) The representative red field shows the TRAP-positive area (gei) Bone slide were stained with Mayer's haematoxylin and observed using a light microscopy (100) The arrows represent resorption pits (jel) Surfaces of bone slides were assessed using scanning electron microscopy (500) (m) The total number of TRAP-positive cells isolated from each mouse was counted (CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; *, P < 0.05; naive

vs CIA þ Control-shRNA; **, P < 0.01; n ¼ 5) The results are expressed as the mean ± SEM (n) The quantification of the TRAP-positive area on the joint slides A representative red-field image showing the TRAP-positive area was segmented from the background objects and scored The total area of the tissue slide was segmented using the greyscaled image (1 ¼ 0.237 mm 2 ) The TRAP-positive area was divided by the area of the total tissue sections to determine the TRAP-positive area per mm 2 (CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; *, P < 0.05; naive vs CIA þ Control-shRNA; **, P < 0.01; n ¼ 5) The results are expressed as the mean ± SEM (o) The quantification of the pit area on bone slides The area of resorption pits was segmented from background objects and scored The pit area is represented by the resorption pit area/bone slides (CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; *, P < 0.05; naive vs CIA þ Control-shRNA; *, P < 0.05; n ¼ 5) The results are expressed as the mean ± SEM (p) RANKL levels in the serum and joint protein extracts The levels of RANKL in the serum and protein extracts, which were obtained on day 35 post-immunization from the joints of normal naive mice, control-shRNA-treated CIA mice, and CRACM1-shRNA-treated CIA mice, were determined using a sandwich ELISA (CIA þ Control-shRNA vs CIA þ CRACM1-shRNA; *, P < 0.05; ***, P < 0.001, n ¼ 7e15).

S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 7

The intra-articular administration of CRACM1-shRNA to CIA

mice resulted in decreased pro-inflammatory cytokine release in

the joints and the periphery Among the pro-inflammatory

cyto-kines, the IL-23-IL-17 axis was regulated by local

CRACM1-shRNA-treatment High expression of CD4þ IL-17þcells, known as Th17

cells, along with their major effector molecule, IL-17A, was detected

among the splenocytes and synovial cells of CIA mice The

increased levels of IL-23, a cytokine mainly secreted from

macro-phages and dendritic cells, in synovial extracts and serum indicated

that the IL-23eIL-17 axis was activated at the onset of CIA Evidence

also suggests that Th17 cells and the effector cytokine IL-17 play

pivotal roles in the pathogenesis of CIA, which is related to T cell

activation and bone resorption, as well as inflammatory cell

infil-tration and joint destruction(12)

Intra-articular injections of CRACM1-shRNA had no effect on the

IL-23 levels in the synovium but down-regulated the number of

Th17 cells and the amount of IL-17A in the synovium Moreover, the

local regulation of the IL-23-IL-17 axis caused systemic functional

suppression of Th17 cells, which was reflected by the decreased

Th17 population among splenocytes and the reduction of IL-17A

and IL-23 in the serum from CRACM1-shRNA-treated CIA mice A

lower circulating concentration of cytokines, as shown in the

cytokine profile of serum, induced by local immunosuppression in

joint could systemically affect the deviation of Th17 cells Also,

osteoclastogenic Th17 cells play an important role in bone

destruction Therefore, inhibition of the IL-23-IL-17 axis may have a

beneficial effect on RA(13) The decreased activation of osteoclasts

and the progressive attenuation of late bone deformity in the

cur-rent study may have partially resulted from suppression of the

IL-23-IL-17 axis in CRACM1-shRNA-treated mice

In many cells, CRAC channels function as an essential route for

SOCE The CRAC channels of mast cells and lymphocytes were the

first to be characterized and exhibit high Ca2þ selectivity (14)

Increased Ca2þsignalling via CRAC channels plays a fundamental

role in mast cells and lymphocytes in processes ranging from gene

expression to the regulation of proliferation CRAC channel families

regulate mast cell functions not only through the modulation of

intracellular Ca2þsignalling but also by directly affecting the

sta-bility of vesicle fusion during exocytosis, which is a critical process

in mast cell degranulation(15,16) Mast cells are a strongly

sug-gested therapeutic target in RA because activated

IL-17A-expressing mast cells existing in the synovium from RA patients

contribute to the IL-17-rich environment and therefore may

pro-mote the amplification of the innate immune responses in the

joints(17) Although the present study was mainly focused on the

functional assessment of lymphocytes, CRAC inhibitors may also

lead to beneficial effects in RA through the in situ inhibition of mast

cell activation in the joints

The timing of the local CRACM1-shRNA injection is important

for improving prognosis A single injection during the onset phase

delayed the progression of arthritis However, this approach did not

affect prognosis, whereas treatment with CRACM1-shRNA in the

acme phase with clinical symptoms suppressed ongoing disease

and attenuated late bone deformity, which fulfils an essential

prerequisite for anti-arthritic therapy The local blockade of Ca2þ

entry via CRAC channels, which are the cellular targets in mainly T,

B, and osteoclastogenic cells but not macrophages, may not disturb

antigen presentation and delay the onset of autoimmune arthritis

In the acme phase, the local inflammatory microenvironment is

already established, and inflammatory cells are overwhelmingly

dominant The therapeutic effects of the CRAC inhibitor are evident

in this phase of CIA, which is mainly characterized by the T

cell-mediated activation of osteoclastogenesis Therefore, regarding

CRACM1-shRNA treatment, it is important to consider the timing of

administration

In this study, we found that the intra-articular injection of len-tiviral vectors temporarily promoted joint inflammation, as shown

by the joint swelling rates, which were quantified using 3-dimensional magnetic resonance imaging (MRI) The safety of the intra-articular CRACM1-shRNA injection is an important prerequi-site for the local application of gene therapy The increased swelling that was observed in mice may have been because of a site infection

or the lentiviral product Infection of the joint could have occurred

as a response to local immunosuppression or contamination during the infection process However, because the increased swelling was mild and transient, it is difficult to attribute this swelling to deep infection Other requirements for the intra-articular application of this novel treatment include sufficient viral counts and appropriate composition of the viral solutions Non-specific host immune re-sponses may lead to unexpected swelling(18) The amount of some pro-inflammatory cytokines increased after lentiviral treatment in CIA mice, such as IL-1b, IL-12, and TNF-a According to a report by Guo et al., infection with HIV-1, from which the lentiviral vector was derived, induces the expression of pro-IL-1bthrough Toll-like re-ceptor (TLR) 8-dependent mechanisms and IL-1bmaturation via the leucine-rich repeat-containing protein 3 in flammasome-dependent pathway in human monocytic cells (19) Neutrophils also recognize HIV-1 through TLR7/8 and produce reactive oxygen species, which are associated with cell activation and secretion of pro-inflammatory cytokines, including TNF-a, which favours HIV-1 replication(20) A similar effect was previously reported in human microglial cells and in HIV-1-latently infected promonocytic cells as well In the present study, an increased level of IL-12 protein in the joints and systemically increased levels of TNF-a in lentiviral-treated mice may be partially due to the immunological re-sponses of the application of lentiviral particles Future studies using different quantities of viral particles or different composi-tions of viral solucomposi-tions should be undertaken to investigate the feasibility and safety of intra-articular CRACM1-shRNA injection for the treatment of CIA

In summary, our approach involved the delivery of the CRAC inhibitor to joints in CIA mice with the goal of achieving local immunosuppression by regulating SOCE via CRAC channels The intra-articular injection of CRACM1-shRNA at the acme phase of CIA resulted in the amelioration of symptoms, the prevention of joint damage and bone resorption, and an improvement in late bone deformity Further modifications of lentivirus-mediated local gene therapies are needed to translate this treatment approach into the clinical setting; however, we suggest that thefindings of this study provide the basis and will generate enthusiasm for the use of CRACM1 inhibitors, which are already widely studied as a novel target of drug development to treat autoimmune diseases such as RA

Conflicts of interest The authors have nofinancial conflicts of interest to disclose Acknowledgement

This work was supported by the Japan Society for the Promotion

of Science (JSPS) (KAKENHI Grant Number 15K19575)

References (1) Mandema JW, Salinger DH, Baumgartner SW, Gibbs MA A dose-response meta-analysis for quantifying relative efficacy of biologics in rheumatoid arthritis Clin Pharmacol Ther 2011;90:828e835

(2) Parekh AB Store-operated CRAC channels: function in health and disease Nat Rev Drug Discov 2010;9:399e410

S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 8

(3) Liu S, Kiyoi T, Takemasa E, Maeyama K Systemic lentivirus-mediated delivery

of short hairpin RNA targeting calcium release-activated calcium channel 3 as

gene therapy for collagen-induced arthritis J Immunol 2015;194:76e83

(4) Liu S, Watanabe S, Shudou M, Kuno M, Miura H, Maeyama K Upregulation of

store-operated Ca entry in the naive CD4 T cells with aberrant cytokine

releasing in active rheumatoid arthritis Immunol Cell Biol 2014;92:752e760

(5) Liu X, Cai F, Zhang Y, Yang A, Liu L Celastrol, an NF-kappaB inhibitor,

ame-liorates hypercalciuria and articular cartilage lesions in a mouse model of

secondary osteoporosis J Pharmacol Sci 2016;130:204e211

(6) Cato MH, Yau IW, Rickert RC Magnetic-based purification of untouched

mouse germinal center B cells for ex vivo manipulation and biochemical

analysis Nat Protoc 2011;6:953e960

(7) Tagishi K, Shimizu A, Endo K, Kito H, Niwa S, Fujii M, et al Defective splicing of

the background Kþ channel K2P5.1 by the pre-mRNA splicing inhibitor,

pla-dienolide B in lectin-activated mouse splenic CD4þ T cells J Pharmacol Sci.

2016;132:205e209

(8) Kay J, Calabrese L The role of interleukin-1 in the pathogenesis of rheumatoid

arthritis Rheumatology (Oxford) 2004;43(Suppl 3):iii2eiii9

(9) Venkatesha SH, Dudics S, Acharya B, Moudgil KD Cytokine-modulating strategies

and newer cytokine targets for arthritis therapy Int J Mol Sci 2015;16:887e906

(10) Evans CH, Ghivizzani SC, Robbins PD Arthritis gene therapy and its tortuous

path into the clinic Transl Res 2013;161:205e216

(11) Ohga K, Takezawa R, Arakida Y, Shimizu Y, Ishikawa J Characterization of

YM-58483/BTP2, a novel store-operated Ca2þ entry blocker, on T cell-mediated

immune responses in vivo Int Immunopharmacol 2008;8:1787e1792

(12) Korn T, Bettelli E, Oukka M, Kuchroo VK IL-17 and Th17 cells Annu Rev Immunol 2009;27:485e517

(13) Sato K, Suematsu A, Okamoto K, Yamaguchi A, Morishita Y, Kadono Y, et al Th17 functions as an osteoclastogenic helper T cell subset that links T cell activation and bone destruction J Exp Med 2006;203:2673e2682 (14) Prakriya M Store-operated Orai channels: structure and function Curr Top Membr 2013;71:1e32

(15) Liu S, Sahid MNA, Takemasa E, Kiyoi T, Kuno M, Oshima Y, et al CRACM3 regulates the stability of non-excitable exocytotic vesicle fusion pores in a Ca2þ-independent manner via molecular interaction with syntaxin4 Sci Rep 2016;6:28133

(16) Ma HT, Beaven MA Regulation of Ca2þ signaling with particular focus on mast cells Crit Rev Immunol 2009;29:155e186

(17) Hueber AJ, Asquith DL, Miller AM, Reilly J, Kerr S, Leipe J, et al Mast cells express IL-17A in rheumatoid arthritis synovium J Immunol 2010;184: 3336e3340

(18) Rothe M, Schambach A, Biasco L Safety of gene therapy: new insights to a puzzling case Curr Gene Ther 2014;14:429e436

(19) Guo H, Gao J, Taxman DJ, Ting JP, Su L HIV-1 infection induces interleukin-1beta production via TLR8 protein-dependent and NLRP3 inflammasome mechanisms in human monocytes J Biol Chem 2014;289:21716e21726 (20) Saitoh T, Komano J, Saitoh Y, Misawa T, Takahama M, Kozaki T, et al Neutrophil extracellular traps mediate a host defense response to human immunodeficiency virus-1 Cell Host Microbe 2012;12:109e116

S Liu et al / Journal of Pharmacological Sciences xxx (2017) 1e9 9