USP15 Regulates Neuromuscular Functionsthrough the Control of RNA Splicing

著者 KIM Jaehyun内容記述 この博士論文は内容の要約のみの公開（または一部

非公開）になっていますyear 2018その他のタイトル USP15によるRNAスプライシングを介在した神経筋の

機能制御メカニズムの解析学位授与大学 筑波大学 (University of Tsukuba)学位授与年度 2017報告番号 12102甲第8557号URL http://hdl.handle.net/2241/00152444

USP15 Regulates Neuromuscular Functions

through the Control of RNA Splicing

A Dissertation Submitted to

the Graduate School of Life and Environmental Sciences,

the University of Tsukuba

in Partial Fulfillment of the Requirements

for the Degree of Doctor of Philosophy in Science

(Doctoral Program in Biological Sciences)

Jeahyun KIM

i

Table of Contents

Abstract 1

Abbreviations 3

Introduction 5

Materials and Methods 7

Results

USP15 deficient mice exhibit ataxia-like symptoms 13

USP15 interacts with TUT1 and promotes its deubiquitination 14

USP15 deubiquitinates K63-linked polyubiquitin chins on TUT1 15

SART3 enhances interaction between USP15 and TUT1 16

USP15 promotes TUT1 subnuclear translocation 18

Loss of USP15 reduces U6-snRNA level in cerebellum 19

Loss of USP15 results in altered RNA splicing 19

USP4 regulates deubiquitination and translocations of TUT1 20

KLHL7 leads to relocation of TUT1 in nucleus 22

Discussion 24

Acknowledgements 26

References 27

Tables 33

Figures 39

1

Abstract

Ubiquitin system controls various physiological functions in cells by modulating

cellular processes such as protein degradation and signaling pathway. Recent studies

have implied that deubiquitinating enzyme (DUB) responsible for the removal of

ubiquitin from substrate is important as a regulator of neurological function. For

example, dysfunction or deficiency of certain DUBs results in neurodegenerative and

psychiatric disorders. Furthermore, it is also revealed that ubiquitin specific protease

15 (USP15), a member of large family of DUBs, is involved in several neurological

disorders including autism, ataxia, Parkinson's disease and glioblastoma, providing a

clue about the close relationship between USP15 and nervous system. However, the

detailed molecular mechanism of how USP15 works on nervous system is yet to be

elucidated.

In this study, I found that USP15 deficient mice show ataxia-like behavior

resulting from the functional defects of cerebellum and skeletal muscle. In

biochemical analysis, I revealed that USP15 deubiquitinates terminal uridylyl

transferase 1 (TUT1), which carries out polyuridylation on 3'-end of U6 snRNA, and

regulates TUT1 subnuclear localizations. SART3, which is a recycling factor of

U4/U6 di-snRNP, facilitates TUT1 deubiquitination as a substrate targeting factor of

USP15. Furthermore, USP4, highly homologous with USP15, also promotes

deubiquitination and translocation of TUT1 indicating that USP4 may play a same

role with USP15 in regulating TUT1. As to regulation of subnuclear translocation of

2

TUT1, KLHL7, an ubiquitin ligase, which interacts with USP15, also regulates

TUT1 translocation implying that KLHL7 is involved in USP15-TUT1 pathway.

In gene expression analysis, loss of USP15 reduces U6-snRNA level and

widely affects splicing pattern of multiple genes potentially related to neuromuscular

disorders, suggesting that USP15 is essential for proper neuromuscular functions

mediated by RNA splicing. Thus, this study may give a novel understanding between

RNA metabolism controlled by ubiquitin system and neuromuscular functions.

3

Abbreviations

ADP Adenosine diphosphate

ALS Amyotrophic lateral sclerosis

Cul Cullin

DUB Deubiquitinating enzyme

E1 Ubiquitin activating enzyme

E2 Ubiquitin conjugating enzyme

E3 Ubiquitin ligase

EDTA Ethylenediaminetetraacetic acid

FL Flag

IB Immunoblot

IP Immunoprecipitation

JAMMs JAB1/MPT/Mpv43 metalloenzymes

KLHL7 Kelch like protein 7

LSm Like-Sm

MEF Mouse embryonic fibroblast

NPM Nucleophosmin

OUT Ovarian tumor protease

PCR Polymerase chain reaction

PD Pull down

SART3 Squamous cell carcinoma antigen recognized by T cells 3

SDS-PAGE Sodium dodecyl sulphate-polyacrylamide gel electrophoresis

4

SMA Spinal muscular atrophy

snRNA Small nuclear RNA

snRNP Small nuclear ribonucleic particles

TUT1 Terminal uridylyl transferase 1 (U6 snRNA-Specific)

Ub Ubiquitin

UCH Ubiquitin C-terminal hydrolases

USP Ubiquitin-specific proteases

WCL Whole cell lysates

WT Wild type

5

Introduction

Ubiquitin (Ub) system is responsible for regulating various cellular processes such as

protein degradation, DNA transcription, signal transduction and protein quality

control [1]. Ub is covalently attached to a target protein through a sequential action

of Ub activating enzyme (E1), Ub conjugating enzyme (E2), Ub ligase (E3), and is

removed from the target by deubiquitinating enzymes (DUBs). This reversible

reaction, which governs a balance of ubiquitination status of target proteins, is also

important for the control of nervous system functions including neurite growth [2],

synaptic transmission [3-5], receptor turnover [6, 7] and synaptic plasticity [8, 9] and

receive attention as a key regulator of nervous system functions.

DUBs have also attracted attention as therapeutic targets for

neurodegeneration. The importance of DUB function at nervous system was first

highlighted by specific mutation of DUB genes that link to several neurological

disorders [10, 11]. Furthermore, it has been established that dysfunction or

deficiency of DUBs results in disruption of synapse development and function,

neurodegenerative disorders and psychiatric disorders [12], indicating that DUBs

perform crucial functions in both the central and peripheral nervous system. DUBs

are composed of five distinct subfamilies: ubiquitin-specific proteases (USPs),

ubiquitin C-terminal hydrolases (UCHs), ovarian tumor protease (OTUs), Josephin

proteases, and JAB1/MPT/Mpv43 metalloenzymes (JAMMs) [12]. One of this

subfamily, USPs, which form a large family of deubiquitinating enzymes, is also

6

involved in several neurological disorders such as spinocerebellar ataxia type 1 [13],

Down’s syndrome [14, 15] and Parkinson’s disease [16, 17].

USP15 is a member of the USP family. Recent studies have revealed that

USP15 promotes oncogenesis in glioblastoma by activation of the TGF-β signaling

pathway [18] and the expression levels of both USP15 mRNA and protein decreases

in mouse models of spinocerebellar ataxia type 3 and Huntington’s disease [19].

These suggest that USP15 may play important roles in neuromuscular functions.

However, the detailed molecular mechanism of how USP15 functions in nervous and

muscle systems were yet to be elucidated.

In this study, I have focused on both physiological and molecular role of

USP15 in both neurons and muscle cell. Through the analyses of USP15 deficient

mice, I found that USP15 deficient mice exhibit motor dysfunctions, and lack of

USP15 induces global misregulation in RNA splicing. In molecular analyses, I

identified novel target of USP15. USP15 deubiquitinated TUT1, responsible for

polyuridylation of U6-snRNA, and regulated its subnuclear translocation. This

process was facilitated by SART3, substrate targeting factor of USP15. In these

analyses, I also found that USP4 and KLHL7 regulate TUT1 translocation implying

that these two proteins are involved in TUT1 regulation. These data, therefore,

suggest that USP15 is involved in the regulation of RNA splicing through the control

of TUT1, which is a key process that controls nervous and muscle system.

7

Materials and Methods

Rotarod performance test

2-month old wild-type and USP15 deficient mice were placed on rod. Rotation speed

was 5 r.p.m. at the beginning and then gradually accelerated to 40 r.p.m. The time to

fall from the rod was measured for 240 seconds.

Plasmids

LSms (1, 2, 4, 6 and 7), SART3 and TUT1 were amplified from HCT116 cells

cDNA library by using KOD plus DNA polymerase (TOYOBO) and inserted into

pcDNA3-Flag, pCS4-Myc, pCS4-HA, pCS4-EGFP vectors. SART3 deletion mutants

were amplified from SART3 wild-type plasmids and constructed likewise. The

pCAGEN-His-Ub (wild-type and mutants) was described previously [20]. pcDNA3-

Flag-USP15, pcDNA3-Myc-USP15, pcDNA3-Myc-USP15 D879A, pcDNA3-Flag-

USP4 were constructed previously in Tomoki Chiba laboratory (University of

Tsukuba). pcDNA-Myc-KLHL7 WT, pcDNA3-Myc-KLHL7 S150N, pcDNA3-

Myc-KLHL7 A153T and pcDNA3-Myc-KLHL7 A153V were describe previouly

[21]. The primers used in this study are described in Table 1.

Antibodies

For immunoblot analyses, anti-USP15 (Bethyl, A300-923A), anti-SART3

(Proteintech, 180251AP), anti-Tubulin (SIGMA, DM1A), anti-FLAG (SIGMA, M2),

anti-cMyc (Santa Cruz, 9E10), anti-GFP (MBL, 598) and anti-His (GE Helthcare,

8

27-4710-01) were used as primary antibodies. The peroxidase-conjugated anti-mouse

and anti-rabbit antibodies (SeraCare) were used as secondary antibodies.

Anti-Calbindin (SIGMA, CB-955) was used for immunohistochemistry as

primary antibodies. For immunofluorescence analyses, anti-Flag (SIGMA, M2), anti-

Flag (SIGMA, F7425), anti-HA (Roche, Roche) were used as primary antibodies.

Anti-NPM antibody was provided by Dr. Mitsuru Okuwaki [22]. Alexa Fluor 488

and 594 conjugated anti-rabbit, anti-rat and anti-mouse antibodies were used as

secondary antibodies.

Cell culture

Wild type and USP15 deficient MEFs, HEK293, HEK293T and HeLa cells were

cultured in Dulbecco's modified Eagle's medium (high glucose) (WAKO)

supplemented with 10% fetal bovine serum, 1% penicillin streptomycin (Gibco) in a

37°C incubator with 5% CO2.

Histology

Brains and skeletal muscles of 3-month-old and brains of 9-month-old mice were

perfused with 4% paraformaldehyde in phosphate buffered saline (PBS). Tissues

were fixed in the same fixative for 48 hours and then embedded in paraffin. Sections

were stained with Meyer's Hematoxylin and Eosin or subjected to

immunohistochemical analyses.

Immunohistochemistry

9

Deparaffinized tissue specimens were immersed in 0.01 M citrate buffer (10mM

Citric Acid, 0.05% Tween 20, pH 6.0) and boiled in microwave oven for 10 minutes.

After antigen retrieval, tissue sections were blocked for an hour in 3% bovine serum

albumin (BSA) in TBST (25 mM Tris-HCl (pH 7.5), 0.14 M NaCl, 0.1% TritonX-

100) and incubated with a primary antibody, mouse anti-Calbindin antibody diluted

1/300 in TBST for overnight at 4°C. Following wash with TBST, tissue sections

were incubated with a secondary antibody (Alexa Fluor 594 anti-mouse IgG) diluted

1/500 in TBST for an hour at room temperature. Tissue specimens were observed

using a fluorescence microscope (Keyence, BIOREVO BZ-9000).

Immunoprecipitation

Cells were lysed with ice-cold lysis buffer (20 mM Tris-HCl (pH 7.5), 150 mM NaCl,

1 mM EDTA, 0.5% NP-40, 1 mM DTT) and centrifuged at 14,000 r.p.m. for 10

minutes. The supernatant was subjected to immunoprecipitation with anti-USP15 or

anti-cMyc antibodies and protein G agarose beads (Thermo Scientific), or anti-Flag

M2-agarose beads (SIGMA). Samples were incubated at 4°C overnight. The beads

were then washed with lysis buffer. The protein samples were added to 4 X SDS

sample buffer (250 mM Tris-HCl (pH 6.8), 40% Glycerol, 10% SDS, 0.04%

bromophenol blue, 20% β-mercaptoethanol), boiled for 3 minutes and subjected to

immunoblot analysis.

His-Ub Pull down assay

10

Cells were washed with PBS and lysed in extraction buffer (6M guanidinium-HCl,

50 mM sodium phosphate buffer (pH 8.0), 300 mM NaCl and 5 mM imidazole). Cell

lysates were sonicated for 30 seconds on ice, centrifuged and then incubated with

Talon metal affinity resin (Clonetech) at 4°C overnight. The precipitants were

washed with buffer (50 mM sodium phosphate buffer (pH 8.0), 300 mM NaCl and 5

mM imidazole) and subjected to immunoblot analysis.

Immunoblot

The protein samples after immunoprecipitation and his-tagged pull down assay were

separated by SDS-PAGE, transferred to PVDF membranes (Millipore). Membranes

were incubated overnight at 4°C with primary antibodies. The proteins on membrane

were detected with HRP-conjugated secondary antibodies and chemiluminescence

reagent (Amersham ECL Prime Western Blotting Detection Reagents, GE

Healthcare).

Immunocytochemistry

HeLa cells were cultured on cover slips and after 18 hours, transfected with indicated

plasmids. The cells were fixed with 4% paraformaldehyde in PBS for 10 min at room

temperature. The cells on coverslips were blocked in 0.4% Triton X-100 in blocking

solution (3% BSA in PBS) for 30 min at room temperature, and then incubated with

primary antibodies diluted in blocking solution for an hour or overnight at 4 °C.

After washing with PBS, the cells were incubated with secondary antibodies diluted

in blocking solution for 30 minutes at room temperature. Nuclei were stained with

11

Hoechst 33342 (Life Technologies). The coverslips were then mounted onto slides

using the Fluoromount Plus mounting solution (Diagnostic BioSystems). Images

were obtained using a fluorescence microscope (Keyence model BZ-9000).

Quantitative real-time PCR (qPCR)

Total RNAs from wild-type and USP15 deficient cerebellum, cortex, skeletal muscle,

liver, spleen, kidney, heart and MEFs were prepared by ISOGEN II (NIPPON

GENE). The cDNA were synthesized by SuperScript III CellsDirect cDNA

Synthesis Kit (Life Technologies) using random hexamer primer. qPCR was

performed with THUNDERBIRD SYBR qPCR Mix (TOYOBO). The data were

analyzed using Thermal Cycler Dice Real Time System Software (TAKARA).

Following Primers were used.

5.8 S rRNA qPCR Forward, 5'-CGGCTCGTGCGTCGAT-3’;

5.8 S rRNA qPCR Reverse, 5'-CCGCAAGTGCGTTCGAA -3’;

U6-snRNA qPCR Forward, 5'-GCTTCGGCAGCACATATACTAAAAT-3';

U6-snRNA qPCR Reverse, 5'-ACGAATTTGCGTGTCATCCTT-3'.

Exon array

One-month-old mice of wild type and USP15 deficient mice (n=3) were euthanized

using carbon dioxide. Their cerebellum and skeletal muscles were then dissected and

snap frozen in liquid nitrogen. Total RNAs were extracted from each tissue using

ISOGEN II (NIPPON GENE) according to the manufacture’s instructions.

Fragmented and labelled total RNA of each sample were hybridized on Affymetrix

12

GeneChip mouse exon 1.0 ST arrays. After hybridization, each probe array was

washed and stained with Affymetrix GeneChip Fluidics Station 450 and scanned by

Affymetrix GeneChip Scanner 3000. Data were analyzed with GeneSpring 12.6

Software and filtered by more than Splicing Index 0.5 and with P-value<0.05.

Splicing Index was calculated with following calculation:

Splicing index = log2 (NIi1/NIi2)

NIij = Eij/Gj

NIij means normalized intensity (NI) for exon i in experiment j. Eij is the estimated

intensity level for exon i in experiment j. Gj is the estimated gene intensity.

13

Results

USP15 deficient mice exhibit ataxia-like symptoms.

Although USP15 appears to be related with neuromuscular disorders [19], little is

known about the detailed function of USP15 in the nervous systems. To understand

the physiological roles of USP15 in the neuromuscular system, I analyzed USP15

deficient mice. USP15 deficient mice were much smaller than wild-type mice in

body size (Figure 1). It also showed hind limb clasping reflexes in tail suspension

test, which is a typical phenotype observed in mice with ataxia (Figure 2). To further

investigate whether lack of USP15 impairs motor ability, rotarod performance test

was performed. USP15 deficient mice easily lost balances and tended to grip the rod

instead of walking on the rotating rod, and eventually fell to the ground earlier than

wild-type ones (Figure 3). These results indicate that lack of USP15 leads to motor

disorder.

It is widely known that dysfunctions of cerebellum or skeletal muscle may

cause ataxia-like symptoms. Therefore, I next examined the tissue specimens derived

from cerebellum and skeletal muscle of USP15 deficient mice. In USP15 deficient

cerebellum, abnormal branching and position of lobules were observed (Figure 4). In

addition, immunohistochemical analyses using Purkinje cell marker, anti-Calbindin

antibody, revealed that loss of Purkinje cells, which contribute to motor coordination,

was observed in USP15 deficient cerebellum (Figure 5). In skeletal muscle, 70%

reduction in muscle fiber thickness was observed in USP15 deficient mice (Figure 6).

14

Theses data shows that USP15 is important for the regulation of cerebellar and

muscular morphogenesis and lack of USP15 causes motor dysfunction.

USP15 interacts with TUT1 and promotes its deubiquitination.

In order to figure out the mechanism of how the loss of USP15 leads to motor

dysfunction, I explored deubiquitination substrates of USP15. Mass spectrometry

analysis of USP15 interacting proteins identified several proteins related to U6-

snRNP, which is a core component of spliceosome and carries out mRNA splicing

(Table 2) implying that USP15 may be involved in spliceosomal functions. LSm

proteins assists the association of U4/U6 di-snRNP by assembly of LSm ring [23].

SART3 is known as a recycling factor of U4/U6-snRNP. It facilitates disassembly of

U4/U6 di-snRNP by recruiting USP4 to its deubiquitination substrates Prp3, a

component of U4-snRNP [24-26]. TUT1 is responsible for polyuridylation on 3'- end

of U6-snRNA [27]. To further investigate the interaction between USP15 and these

U6-snRNP related proteins, co-IPs in HEK293 cells were performed. I first

examined whether LSm proteins bind to USP15. Although several LSm proteins

were identified as the binding targets by mass spectrometry, co-IP with USP15 could

not be seen (Figure 7). I next tested whether SART3 bind to USP15. Protein lysates

from wild-type and USP15 deficient MEFs as a negative control, were

immunoprecipitated with anti-USP15 antibodies and immunoblotted with both anti-

USP15 and anti-SART3 antibodies. Endogenous SART3 was co-immunoprecipitated

with endogenous USP15, indicating that USP15 binds to SART3 (Figure 8). Lastly, I

investigated if TUT bind to USP15 by expressing Myc-USP15 and GFP-TUT1 in

15

HEK293 cells. Myc-USP15 was immunoprecipitated and immunoblotted with anti-

Myc and anti-GFP antibodies. GFP-TUT1 was co-immunoprecipitated with Myc-

USP15, indicating that USP15 binds to TUT1 (Figure 9).

Since USP15 binds to TUT1, I then investigated whether TUT1 is a substrate

of USP15 by His-Ub pull down assay. HEK293 cells were transfected with Flag-

TUT1, His-Ub and either Myc-USP15 or mock. Ubiquitinated proteins were pulled-

down with Talon metal affinity resin and immunoblotted with anti-Flag, anti-Myc

and anti-His antibodies. Ubiquitinated TUT1 was detected in the absence of USP15

overexpression. On the other hand, overexpression of Myc-USP15 decreased the

amount of ubiquitinated TUT1, while that of Myc-USP15 D879A mutant that lacks

its enzyme activity did not, suggesting that USP15 deubiquitinates TUT1. Together,

theses data suggest that TUT1 is the target substrate of USP15 (Figure 10).

USP15 deubiquitinates K63-linked polyubiquitin chains on TUT1.

To figure out which polyUb chain on TUT1 is targeted by USP15, I used his- wild-

type and mutant (K48R, K63R, 48K and 63K) Ub expression plasmids for his-tagged

pull down assay. As polyUb chains could be formed via lysine residues, mutant Ub

of lysine 48 to arginine (K48R) or lysine 63 to arginine (K63R) are unable to form

polyUb chains via lysine 48 or 63 linkages respectively. Ub of every lysine mutated

to arginine except lysine 48 (K48) or lysine 63 (K63) could form only polyUb chains

via lysine 48 or 63 linkages. HEK293 cells expressing Flag-TUT1, Myc-USP15 and

either WT or mutant His-Ub were pulled-down with Talon metal affinity resin and

immunoblotted with anti-Flag, anti-Myc and anti-His antibodies. TUT1

16

deubiquitination by USP15 was observed in overexpression of WT, K48R and 63K

Ub, but not that of K63R and 48K (Figure 11). These results indicate that USP15

deubiquitinates K63-linked polyUb chains on TUT1. Unlike K48-linked polyUb that

is related to degradation by proteasome, K63-linked polyUb is involved in signaling

pathway such as inflammation and translocation. Thus, these data suggest that

USP15 controls the function of TUT1 through its deubiquitination.

SART3 enhances interaction between USP15 and TUT1.

Previous studies revealed that SART3 binds to USP4 and guide it to Prp3, and

thereby promotes deubiquitination of Prp3 as a substrate targeting factor of USP4

[24]. I hypothesized that SART3 may play an equal role in regulating USP15. To test

this, I first investigate whether SART3 binds to TUT1 by expressing Flag-TUT1 and

Myc-SART3 in HEK293 cells. Myc-SART3 was immunoprecipitated and

immunoblotted with anti-Myc and anti-GFP antibodies. As expected, TUT1 was co-

immunoprecipitated with SART3 indicating that SART3 also binds to TUT1 (Figure

12). Next, I examined interaction between USP15 and TUT1 with or without

overexpression of SART3. HEK293T cells expressing Flag-TUT1, GFP-SART3 and

Myc-USP15 were immunoprecipitated with anti-Myc antibody and immunoblotted

with anti-Flag, anti-GFP and anti-Myc antibodies. Though Co-IP of Flag-TUT1 with

Myc-USP15 could be seen without GFP-SART3 overexpression (Figure 9 and

Figure 13, lane 6), it increased in GFP-SART3 overexpression (Figure 13, lanes 7

and 8). This suggests that SART3 is a substrate targeting factor of USP15

17

To characterize the region of SART3 where TUT1 binds, SART3 deletion

mutants (SART31~649 and SART3660~963) were used for co-IP experiment. Myc-

SART3 WT or deletion mutants were co-expressed with Flag-TUT1 in HEK293T

cells. Cell lysates were immunoprecipitated with anti-Flag antibody and

immunoblotted with anti-Myc and anti-Flag antibodies. Flag-TUT1 co-

immunoprecipitated Myc-SART3 and Myc-SART3660~963, but did not Myc-

SART1~649 (Figure 14). It indicates that TUT1 binds to SART3 C-terminal region

(660~963 amino acids region). Previous study reported that USP15 binds to SART3

278~649 amino acids region [28]. Therefore, USP15 and TUT1 bind to different

regions of SART3 and I thought SART3278~649 and SART3660~963 deletion mutants

could be used as a dominant negative mutant.

Next, I investigated the interaction between TUT1 and USP15 with or

without SART3 WT and deletion mutants by overexpressing GFP-TUT1, Flag-

USP15 and either Myc-SART3 or deletion mutants in HEK293T cells. Flag-USP15

was immunoprecipitated and immunoblotted with anti-GFP, anti-Myc and anti-Flag

antibodies. Co-IP of GFP-TUT1 with Flag-USP15 did not increase in Myc-SART3

deletion mutants (either SART3 227~649 or SART3 660~963) overexpression, like in

Myc-SART3 WT overexpression (Figure 15). This indicates that SART3 deletion

mutants could not increase the interaction between USP15 and TUT1.

To further investigate the regulation of USP15 by SART3, his-Ub pull down

assay was performed by expressing Flag-TUT1, Myc-USP15, His-Ub and either

Myc-SART3 WT or Myc-SART3660~963 in HEK293 cells. Ubiquitinated proteins

were pulled-down with Talon metal affinity resin and immunoblotted with anti-Flag,

18

anti-Myc and anti-His antibodies. Myc-SART3 WT overexpression increased TUT1

deubiquitination by USP15, but Myc-SART3660~936 overexpression did not (Figure

16). These results show that SART3 enhances interaction between USP15 and TUT1

by binding to both proteins, and lead to TUT1 deubiquitination as a substrate

targeting factor of USP15.

USP15 promotes TUT1 subnuclear translocation.

Next, to figure out the physiological role of deubiquitination of TUT1, I investigated

subcellular localization of TUT1. HeLa cells were transfected with Flag-TUT1 and

co-stained with anti-NPM and anti-Flag antibodies. Flag-TUT1 was co-localized

with NPM, showing that TUT1 is mainly localized in nucleolus (Figure 17). As

USP15 deubiquitinates TUT1 and SART3 promotes this process, I then examined if

overexpression of WT or mutants USP15 and SART3 affects TUT1 localization.

Flag-TUT1 was co-expressed with Myc-USP15 or Myc-USP15 D879A or both Myc-

USP15 and HA-SART3 or Myc-SART3660~963 in HeLa cells, and stained with anti-

Flag and anti-Myc antibodies. Flag-TUT1 was relocated to nucleoplasm in Myc-

USP15 only or both Myc-USP15 and HA-SART3 overexpression, but was still

localized within the nucleolus in Myc-USP15 D879A and Myc-SART3660~963

overexpression (Figure 18A). Furthermore, Myc-USP15 was localized in nucleus in

HA-SART3 overexpression indicating that SART3 recruits USP15 to nucleus where

TUT1 is localized. The ratio of TUT1 localized in nucleolus (Type1) increased in

both Myc-USP15 and HA-SART3 overexpression more than in Myc-USP15

overexpression indicating that SART3 regulates TUT1 translocation (Figure 18B). It

19

is interesting that TUT1 localized in nucleolus ratio (Type1) pattern shown in

quantification data is similar to the ubiquitinated TUT1 ratio pattern seen in the his-

tagged pull down assay (Figure 16 and 18B). Together, these results indicate that

USP15 and SART3 promote TUT1 translocation from nucleolus to nucleoplasm

through its deubiquitination.

Loss of USP15 reduces U6-snRNA level in cerebellum.

As TUT1 polyuridylates 3'- end of U6-snRNA, I hypothesized that lack of USP15

affects U6-snRNAs stability. Therefore, I analyzed the U6-snRNA level, a core

spliceosomal component, via quantitative polymerase chain reaction (qPCR). Total

RNAs were purified from WT and USP15 deficient cerebellum, cortex, skeletal

muscle, liver, spleen, kidney, heart and MEFs. The RNAs were reverse transcribed,

and the U6-snRNA levels were measured by qPCR. The U6-snRNA level decreased

only in USP15 deficient cerebellum. It was 0.8 fold lower compared to WT (Figure

19A). On the contrary, The U6-snRNA levels in other tissues did not decrease

(Figure 19 B, C, D, E, F G and H). Therefore, it is likely that the mechanisms that

underlie the regulation of spliceosomal components mediated by USP15 may be

tissue-specific.

Loss of USP15 results in altered RNA splicing.

Since the loss of USP15 induces lower U6-snRNA level in cerebellum, USP15 could

be involved in regulation of RNA splicing machinery. It has been known that

misregulation of RNA splicing causes neuromuscular disease such as amyotrophic

20

lateral sclerosis and spinal muscular atrophy (Gubitz et al., 2004; Kwiatkowski et al.,

2009; Vance et al., 2009). Therefore, ataxia-like symptoms observed in USP15

deficient mice may be caused by abnormal RNA splicing.

To determine whether the loss of USP15 affects RNA splicing, I performed

exon array analysis, which can measure the relative levels of individual exons. Total

RNAs extracted from the cerebellums and skeletal muscles of three pairs of one-

month-old female littermates were used for exon arrays. From the datasets, I

identified a large number of genes with exon transcripts that were altered

significantly in USP15 deficient mice compared to wild-type mice. This analysis

identified 246 genes in cerebellum and 206 genes in skeletal muscle (Table 3, 4, 5

and 6: data only shows the 10 genes with the highest splicing index score). These

results demonstrated that RNA splicing in USP15 deficient cerebellum and skeletal

muscle was widely changed in multiple genes which may include the candidate

genes related to neuromuscular diseases. Thus, loss of USP15 appears to trigger a

huge change in mRNA splicing critical for nervous and muscle systems and may

result in failure of neuromuscular functions. Interestingly, genes altered in RNA

splicing due to the lack of USP15 are quite different between cerebellum and skeletal

muscle. These data imply that RNA splicing modulated by USP15 may be tissue-

specific, and this specificity enables to maintain proper neuromuscular functions via

regulation of RNA splicing.

USP4 regulates deubiquitination and translocations of TUT1.

21

Given that USP4 is highly homologous with USP15 and recruited to nucleoplasm by

SART3 like USP15 [24, 28], it may be possible that USP4 also targets TUT1. To test

this assumption, I investigated interaction between USP4 and TUT1 by expressing

GFP-TUT1, Flag-USP4 and either Myc-SART3 WT, Myc-SART3278~649 or Myc-

SART3660~963 in HEK293T cells. Flag-USP4 was immunoprecipitated with anti-Flag

and immunoblotted with anti-GFP, anti-Myc and anti-Flag antibodies. Co-IP of GFP-

TUT1 with Flag-USP4 could be seen indicating that TUT1 binds to USP4 (Figure 20,

lane 9). Interestingly, Myc-SART3 overexpression decreased interaction of GFP-

TUT1 with Flag-USP4. However, either Myc-SART3278~649 or Myc-SART3660~963

overexpression did not change it (Figure lanes 20, 10, 11 and 12) suggesting that

SART3 prefers to recruit USP15 than USP4 in regulating TUT1.

I then examined if TUT1 is a deubiquitination target of USP4 by His-Ub pull

down assay. HEK293 cells were transfected with Myc-TUT1, Flag-USP4, His-Ub

and either GFP-SART3 WT or GFP-SART3660-963. Ubiquitinated proteins were

pulled-down with Talon metal affinity resin and immunoblotted with anti-Flag, anti-

Myc, anti-GFP and anti-His antibodies. TUT1 was deubiquitinated in Flag-USP4

overexpression, which was stronger with GFP-SART3 overexpression. On the other

hands, GFP-SART3660~963 overexpression did not lead to TUT1 deubiquitination

(Figure 21). Taken together, these data show that USP4 is also responsible for TUT1

deubiquitination and SART3 functions as a substrate targeting factor of USP4 for

TUT1.

Since both USP15 and USP4 are in charge of TUT1 deubiquitination, I

assumed that both USP15 and USP4 overexpression could increase TUT1

22

deubiquitination much more. HEK294 cells were transfected Myc-TUT1, Flag-

USP15, Flag-USP4 and His-Ub. Ubiquitinated proteins were pulled-down with

Talon metal affinity resin and immunoblotted with anti-Flag, anti-Myc and anti-His

antibodies. TUT1 deubiquitination increased slightly in both Flag-USP15 and Flag-

USP4 overexpression, compared to each Flag-USP15 and Flag-USP4 overexpression

individually (Figure 22). It indicates that both USP15 and USP4 can deubiquitinate

TUT1. I also investigated the subnuclear localization of TUT1 with USP4. HA-

TUT1 was co-expressed with Flag-USP4 in the presence or absence of Myc-SART3

or Myc-USP15 in HeLa cells, and stained with anti-HA and anti-Flag antibodies.

Flag-USP4 expression translocated HA-TUT1 to nucleoplasm like USP15 (Figure

23A). Co-overexpression with Myc-SART3 or Myc-USP15 also induced TUT1

translocation (Figure 23B). Thus, these results suggest that USP15 and USP4 carry

out the same role for the regulation of TUT1 at least in cells.

KLHL7 leads to relocation of TUT1 in nucleus.

In our preliminary study, KLHL7 is shown to interact with USP15 and several

proteins involved in mRNA splicing machinery. KLHL7 is a component of Cul3-

based ubiquitin ligase [21] and its mutations cause retinitis pigmentosa [29, 30]. As

KLHL7 interacts with USP15 and seems to be related to RNA splicing machinery, I

assumed that KLHL7 might be involved in USP15-TUT1 regulatory pathway. To

test this assumption, I analyzed TUT1 subnuclear localization. Flag-TUT1 was co-

expressed with Myc-KLHL7 WT or disease-causative mutants (A153T, A153V and

S150N) in HeLa cells, and stained with anti-Myc and anti-Flag antibodies. TUT1

23

was localized in nucleoplasm with expression of KLHL7 WT and S150N mutant. On

the contrary, TUT1 still stayed in nucleolus with expression of KLHL7 A153T and

A153V mutants (Figure 24). Given that KLHL7 A153 mutations inhibit Cul3

interaction but A150N does not [21], these data indicate that KLHL7 activity, as an

ubiquitin ligase seems to be important for TUT1 translocation to nucleoplasm.

24

Discussion

Alternative mRNA splicing plays an important role in generating enormous

proteomic diversity. It allows eukaryotic cell to produce a huge number of proteins

from the limited number of genes (20,000~25,000 in the human genome) through

selective elimination of introns and exon joining and may thereby contribute to

tissue-specific functions. In brain, alternative splicing is involved in neuronal

functions through the regulation of gene expression, which acts in the synapse such

as neurotransmitter receptors, cation channels, adhesion and scaffold proteins [31].

Given the importance of alternative mRNA splicing in regulating neuronal functions,

it is not surprising that disruption of RNA splicing leads to neuronal dysfunction, and

thereby neurodegenerative disorder. Indeed, recent studies have indicated that

disruption and misregulation of RNA splicing results in neuromuscular disorders

such as amyotrophic lateral sclerosis (ALS) and spinal muscular atrophy (SMA) [32-

34]. However, the biological mechanisms of how the failure in RNA splicing leads to

neuromuscular diseases remains unclear.

In this study, I report that TUT1 is the target substrate of USP15. USP15

deubiquitinated K63-linked polyUb chains on TUT1. Given that K63-linked polyUb

is mainly involved in signaling pathway, not degradation by proteasome, USP15

could regulate TUT1 functions through its deubiquitination. Indeed, USP15

promotes TUT1 subnuclear translocation, suggesting that deubiquitination of K63-

linked polyUb chains. TUT1 is important for translocation to nucleoplasm from

nucleolus. Since TUT1 is responsible for uridylation of 3'-region of U6-snRNA, it is

25

possible that USP15 may be associated with spliceosomal functions through

regulation of TUT1 and, furthermore, affect on RNA splicing. Indeed, lack of USP15

caused reduction in U6-snRNA level in cerebellum and a global alteration in RNA

splicng of various genes indicating that lack of USP15 could lead to misregulation on

RNA splicing machinery. Since dysfunction of alternative mRNA splicing results in

neuromuscular disorder, it is likely that disruption of USP15-TUT1 pathway, which

arouses the misregulation on RNA splicing machinery, induces accumulation of mis-

spliced forms of various genes that could affect neuromuscular functions and

therefore, eventually results in motor disorder shown in USP15 deficient mice.

It is interesting that RNA splicing enacts on every cells but disruption of

RNA splicing mainly results in neuromuscular disorder. In USP15 deficient mice,

USP15 is defective on every tissue but the symptom is quite tissue-specific. In this

study, I found that not only USP15 but also USP4 targeted TUT1 implying that

USP15 and USP4 may assist each other's deubiquitination functions. Furthermore,

SART3 and KLHL7 also regulate TUT1. It is possible that activity and function of

USP15, USP4, SART4 and KLHL7 may differ in each tissue, and thereby TUT1

could be regulated in a tissue-dependent manner. This notion may provide a clue

about tissue-specificity of neuromuscular disorder resulted from misregulation of

RNA splicing.

26

Acknowledgements

I express my deep thanks to Prof. Tomoki Chiba and Assistant Prof. Fuminori

Tsuruta for their persistent mentoring and guidance throughout this research. I also

thank Dr. Tohru Natsume (AIST) for mass spectrometry data and all the members of

Chiba laboratory for helpful discussions and technical supports.

27

References

1. Wagner SA, Beli P, Weinert BT, Nielsen ML, Cox J, Mann M and

Choudhary C (2011) A proteome-wide, quantitative survey of in vivo ubiquitylation

sites reveals widespread regulatory roles. Mol Cell Proteomics 10:M111.013284. doi:

10.1074/mcp.M111.013284

2. Hamilton AM and Zito K (2013) Breaking it down: the ubiquitin proteasome

system in neuronal morphogenesis. Neural Plast 2013:196848. doi:

10.1155/2013/196848

3. Willeumier K, Pulst SM and Schweizer FE (2006) Proteasome inhibition

triggers activity-dependent increase in the size of the recycling vesicle pool in

cultured hippocampal neurons. J Neurosci 26:11333-41. doi:

10.1523/JNEUROSCI.1684-06.2006

4. Yao I, Takagi H, Ageta H, Kahyo T, Sato S, Hatanaka K, Fukuda Y, Chiba T,

Morone N, Yuasa S, Inokuchi K, Ohtsuka T, Macgregor GR, Tanaka K and Setou M

(2007) SCRAPPER-dependent ubiquitination of active zone protein RIM1 regulates

synaptic vesicle release. Cell 130:943-57. doi: 10.1016/j.cell.2007.06.052

5. Jiang X, Litkowski PE, Taylor AA, Lin Y, Snider BJ and Moulder KL (2010)

A role for the ubiquitin-proteasome system in activity-dependent presynaptic

silencing. J Neurosci 30:1798-809. doi: 10.1523/JNEUROSCI.4965-09.2010

6. Lin A, Hou Q, Jarzylo L, Amato S, Gilbert J, Shang F and Man HY (2011)

Nedd4-mediated AMPA receptor ubiquitination regulates receptor turnover and

trafficking. J Neurochem 119:27-39. doi: 10.1111/j.1471-4159.2011.07221.x

28

7. Lussier MP, Herring BE, Nasu-Nishimura Y, Neutzner A, Karbowski M,

Youle RJ, Nicoll RA and Roche KW (2012) Ubiquitin ligase RNF167 regulates

AMPA receptor-mediated synaptic transmission. Proc Natl Acad Sci U S A

109:19426-31. doi: 10.1073/pnas.1217477109

8. Dong C, Upadhya SC, Ding L, Smith TK and Hegde AN (2008) Proteasome

inhibition enhances the induction and impairs the maintenance of late-phase long-

term potentiation. Learn Mem 15:335-47. doi: 10.1101/lm.984508

9. Cajigas IJ, Will T and Schuman EM (2010) Protein homeostasis and synaptic

plasticity. EMBO J 29:2746-52. doi: 10.1038/emboj.2010.173

10. Wilson SM, Bhattacharyya B, Rachel RA, Coppola V, Tessarollo L,

Householder DB, Fletcher CF, Miller RJ, Copeland NG and Jenkins NA (2002)

Synaptic defects in ataxia mice result from a mutation in Usp14, encoding a

ubiquitin-specific protease. Nat Genet 32:420-5. doi: 10.1038/ng1006

11. Saigoh K, Wang YL, Suh JG, Yamanishi T, Sakai Y, Kiyosawa H, Harada T,

Ichihara N, Wakana S, Kikuchi T and Wada K (1999) Intragenic deletion in the gene

encoding ubiquitin carboxy-terminal hydrolase in gad mice. Nat Genet 23:47-51. doi:

10.1038/12647

12. Todi SV and Paulson HL (2011) Balancing act: deubiquitinating enzymes in

the nervous system. Trends Neurosci 34:370-82. doi: 10.1016/j.tins.2011.05.004

13. Hong S, Kim SJ, Ka S, Choi I and Kang S (2002) USP7, a ubiquitin-specific

protease, interacts with ataxin-1, the SCA1 gene product. Mol Cell Neurosci 20:298-

306.

29

14. Valero R, Bayés M, Francisca Sánchez-Font M, González-Angulo O,

Gonzàlez-Duarte R and Marfany G (2001) Characterization of alternatively spliced

products and tissue-specific isoforms of USP28 and USP25. Genome Biol

2:RESEARCH0043.

15. Adorno M, Sikandar S, Mitra SS, Kuo A, Nicolis Di Robilant B, Haro-Acosta

V, Ouadah Y, Quarta M, Rodriguez J, Qian D, Reddy VM, Cheshier S, Garner CC

and Clarke MF (2013) Usp16 contributes to somatic stem-cell defects in Down's

syndrome. Nature 501:380-4. doi: 10.1038/nature12530

16. Li Y, Schrodi S, Rowland C, Tacey K, Catanese J and Grupe A (2006)

Genetic evidence for ubiquitin-specific proteases USP24 and USP40 as candidate

genes for late-onset Parkinson disease. Hum Mutat 27:1017-23. doi:

10.1002/humu.20382

17. Zhang X, Zhou JY, Chin MH, Schepmoes AA, Petyuk VA, Weitz KK,

Petritis BO, Monroe ME, Camp DG, Wood SA, Melega WP, Bigelow DJ, Smith DJ,

Qian WJ and Smith RD (2010) Region-specific protein abundance changes in the

brain of MPTP-induced Parkinson's disease mouse model. J Proteome Res 9:1496-

509. doi: 10.1021/pr901024z

18. Inui M, Manfrin A, Mamidi A, Martello G, Morsut L, Soligo S, Enzo E,

Moro S, Polo S, Dupont S, Cordenonsi M and Piccolo S (2011) USP15 is a

deubiquitylating enzyme for receptor-activated SMADs. Nat Cell Biol 13:1368-75.

doi: 10.1038/ncb2346

30

19. Menzies FM, Huebener J, Renna M, Bonin M, Riess O and Rubinsztein DC

(2010) Autophagy induction reduces mutant ataxin-3 levels and toxicity in a mouse

model of spinocerebellar ataxia type 3. Brain 133:93-104. doi: 10.1093/brain/awp292

20. Tsuruta F, Takebe A, Haratake K, Kanemori Y, Kim J, Endo T, Kigoshi Y,

Fukuda T, Miyahara H, Ebina M, Baba T and Chiba T (2016) SCFFbl12 Increases

p21Waf1/Cip1 Expression Level through Atypical Ubiquitin Chain Synthesis. Mol

Cell Biol 36:2182-94. doi: 10.1128/MCB.00174-16

21. Kigoshi Y, Tsuruta F and Chiba T (2011) Ubiquitin ligase activity of Cul3-

KLHL7 protein is attenuated by autosomal dominant retinitis pigmentosa causative

mutation. J Biol Chem 286:33613-21. doi: 10.1074/jbc.M111.245126

22. Hisaoka M, Nagata K and Okuwaki M (2014) Intrinsically disordered regions

of nucleophosmin/B23 regulate its RNA binding activity through their inter- and

intra-molecular association. Nucleic Acids Res 42:1180-95. doi: 10.1093/nar/gkt897

23. Beggs JD (2005) Lsm proteins and RNA processing. Biochem Soc Trans

33:433-8. doi: 10.1042/BST0330433

24. Song EJ, Werner SL, Neubauer J, Stegmeier F, Aspden J, Rio D, Harper JW,

Elledge SJ, Kirschner MW and Rape M (2010) The Prp19 complex and the

Usp4Sart3 deubiquitinating enzyme control reversible ubiquitination at the

spliceosome. Genes Dev 24:1434-47. doi: 10.1101/gad.1925010

25. Bell M, Schreiner S, Damianov A, Reddy R and Bindereif A (2002) p110, a

novel human U6 snRNP protein and U4/U6 snRNP recycling factor. EMBO J

21:2724-35. doi: 10.1093/emboj/21.11.2724

31

26. Trede NS, Medenbach J, Damianov A, Hung LH, Weber GJ, Paw BH, Zhou

Y, Hersey C, Zapata A, Keefe M, Barut BA, Stuart AB, Katz T, Amemiya CT, Zon

LI and Bindereif A (2007) Network of coregulated spliceosome components revealed

by zebrafish mutant in recycling factor p110. Proc Natl Acad Sci U S A 104:6608-13.

doi: 10.1073/pnas.0701919104

27. Trippe R, Guschina E, Hossbach M, Urlaub H, Luhrmann R and Benecke BJ

(2006) Identification, cloning, and functional analysis of the human U6 snRNA-

specific terminal uridylyl transferase. RNA 12:1494-504. doi: 10.1261/rna.87706

28. Park JK, Das T, Song EJ and Kim EE (2016) Structural basis for recruiting

and shuttling of the spliceosomal deubiquitinase USP4 by SART3. Nucleic Acids

Res 44:5424-37. doi: 10.1093/nar/gkw218

29. Friedman JS, Ray JW, Waseem N, Johnson K, Brooks MJ, Hugosson T,

Breuer D, Branham KE, Krauth DS, Bowne SJ, Sullivan LS, Ponjavic V, Granse L,

Khanna R, Trager EH, Gieser LM, Hughbanks-Wheaton D, Cojocaru RI, Ghiasvand

NM, Chakarova CF, Abrahamson M, Goring HH, Webster AR, Birch DG, Abecasis

GR, Fann Y, Bhattacharya SS, Daiger SP, Heckenlively JR, Andreasson S and

Swaroop A (2009) Mutations in a BTB-Kelch protein, KLHL7, cause autosomal-

dominant retinitis pigmentosa. Am J Hum Genet 84:792-800. doi:

10.1016/j.ajhg.2009.05.007

30. Angius A, Uva P, Buers I, Oppo M, Puddu A, Onano S, Persico I, Loi A,

Marcia L, Hohne W, Cuccuru G, Fotia G, Deiana M, Marongiu M, Atalay HT, Inan

S, El Assy O, Smit LM, Okur I, Boduroglu K, Utine GE, Kilic E, Zampino G,

Crisponi G, Crisponi L and Rutsch F (2016) Bi-allelic Mutations in KLHL7 Cause a

32

Crisponi/CISS1-like Phenotype Associated with Early-Onset Retinitis Pigmentosa.

Am J Hum Genet 99:236-45. doi: 10.1016/j.ajhg.2016.05.026

31. Ule J, Ule A, Spencer J, Williams A, Hu JS, Cline M, Wang H, Clark T,

Fraser C, Ruggiu M, Zeeberg BR, Kane D, Weinstein JN, Blume J and Darnell RB

(2005) Nova regulates brain-specific splicing to shape the synapse. Nat Genet

37:844-52. doi: 10.1038/ng1610

32. Kwiatkowski TJ, Bosco DA, Leclerc AL, Tamrazian E, Vanderburg CR,

Russ C, Davis A, Gilchrist J, Kasarskis EJ, Munsat T, Valdmanis P, Rouleau GA,

Hosler BA, Cortelli P, de Jong PJ, Yoshinaga Y, Haines JL, Pericak-Vance MA, Yan

J, Ticozzi N, Siddique T, McKenna-Yasek D, Sapp PC, Horvitz HR, Landers JE and

Brown RH (2009) Mutations in the FUS/TLS gene on chromosome 16 cause familial

amyotrophic lateral sclerosis. Science 323:1205-8. doi: 10.1126/science.1166066

33. Vance C, Rogelj B, Hortobágyi T, De Vos KJ, Nishimura AL, Sreedharan J,

Hu X, Smith B, Ruddy D, Wright P, Ganesalingam J, Williams KL, Tripathi V, Al-

Saraj S, Al-Chalabi A, Leigh PN, Blair IP, Nicholson G, de Belleroche J, Gallo JM,

Miller CC and Shaw CE (2009) Mutations in FUS, an RNA processing protein,

cause familial amyotrophic lateral sclerosis type 6. Science 323:1208-11. doi:

10.1126/science.1165942

34. Gubitz AK, Feng W and Dreyfuss G (2004) The SMN complex. Exp Cell

Res 296:51-6. doi: 10.1016/j.yexcr.2004.03.022

33

34

Table 2. USP15 interacting proteins identified by mass spectrometry.

* Proteins involved in RNA spliceosomal functions

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

Figure 17. TUT1 is localized at nucleolus.

HeLa cells transfected with Flag-TUT1 were stained with anti-NPM (nucleophosmin)

and anti-Flag antibodies and Hoechest 33342. Scale bar, 20 µm.

56

57

58

59

60

61

62

A B

Figure 24. Overexpression of Wild-type KLHL7 and KLHL7 SN mutant

induces relocation of TUT1 from nucleolus to nucleoplasm.

A, Fluorescent images of HeLa cells expressing Myc-TUT1 together with each

KLHL7 wild-type and dieses causative mutant plasmids. Scale bar: 20 µm. B,

Quantification of the data in (A). Subnuclear localization of Myc-TUT1 was

examined. Cells that express Flag-TUT1 in nucleolus (Type 1) and in diffusely

distribution (Type 2) were counted. n=30 ~ 34.