Oncogene (2013) 32, 42314242

& 2013 Macmillan Publishers Limited All rights reserved 0950-9232/13

www.nature.com/onc

ORIGINAL ARTICLE

Poly(ADP-ribose) polymerase family member 14 (PARP14)

is a novel effector of the JNK2-dependent pro-survival
signal in multiple myeloma
A Barbarulo1, V Iansante2, A Chaidos3, K Naresh4, A Rahemtulla3, G Franzoso1, A Karadimitris3, DO Haskard5, S Papa2,6
and C Bubici1,6
Regulation of cell survival is a key part of the pathogenesis of multiple myeloma (MM). Jun N-terminal kinase (JNK) signaling has
been implicated in MM pathogenesis, but its function is unclear. To elucidate the role of JNK in MM, we evaluated the specic
functions of the two major JNK proteins, JNK1 and JNK2. We show here that JNK2 is constitutively activated in a panel of MM cell
lines and primary tumors. Using loss-of-function studies, we demonstrate that JNK2 is required for the survival of myeloma cells and
constitutively suppresses JNK1-mediated apoptosis by affecting expression of poly(ADP-ribose) polymerase (PARP)14, a key
regulator of B-cell survival. Strikingly, we found that PARP14 is highly expressed in myeloma plasma cells and associated with
disease progression and poor survival. Overexpression of PARP14 completely rescued myeloma cells from apoptosis induced by
JNK2 knockdown, indicating that PARP14 is critically involved in JNK2-dependent survival. Mechanistically, PARP14 was found to
promote the survival of myeloma cells by binding and inhibiting JNK1. Moreover, inhibition of PARP14 enhances the sensitization of
MM cells to anti-myeloma agents. Our ndings reveal a novel regulatory pathway in myeloma cells through which JNK2 signals cell
survival via PARP14, and identify PARP14 as a potential therapeutic target in myeloma.
Oncogene (2013) 32, 42314242; doi:10.1038/onc.2012.448; published online 8 October 2012
Keywords: PARP14; c-Jun N-terminal kinase; survival; multiple myeloma; apoptosis

INTRODUCTION
Multiple myeloma (MM) is a neoplasm of terminally differentiated
B cells, characterized by accumulation of clonal, long-lived plasma
cells in the bone marrow (BM) as well as extramedullary sites.1
Despite the signicant progress that has been made in the
identication of deregulated pathways, MM remains incurable.
Further understanding of the biology of this disease is therefore
needed to develop more effective therapies.2,3
Notable characteristics of myeloma cells include a low
proliferative capacity and a weak apoptotic index in vivo.1,2
Thus, pro-survival mechanisms have central roles in the
accumulation of the malignant clone in MM.2,4,5 Among the
signaling pathways involved in the regulation of cell survival are
members of the mitogen-activated protein kinase (MAPK) family,
which includes three members: extracellular-regulated kinase
(ERK1/2), Jun N-terminal kinase (JNK) and p38-MAPK signaling.68
Although the functions of ERK1/2 and p38-MAPK in MM have
been dened in recent studies,9,10 the role of JNK in the
pathogenesis of the disease is unclear.
Signaling via JNK has key roles in induction of apoptosis, cell
proliferation, transformation and survival through the phosphorylation and regulation of its substrates.11 As these cellular
responses are important in tumorigenesis, JNK signaling has

been implicated in human cancers.11,12 Two major JNK proteins,

JNK1 and JNK2, are expressed ubiquitously and each has two
different splicing isoforms (p46 and p54).11 Although JNK1 and
JNK2 have been considered redundant kinases, there is evidence
for distinct or even opposing functions of JNK1 and JNK2 in
tumorigenesis.1215 Thus, it has been reported that JNK2, but not
JNK1, is essential for growth of various human cancer cells
including lung,16,17 glioblastoma,18,19 prostate and skin.2022 In
contrast, a recent study reported that JNK1 is essential for liver
carcinogenesis, whereas JNK2 appears to be dispensable.12,23
Taken together, these data infer a context- and tissue-specic
functions for JNK1 and JNK2 in cancer development.12 Although it
has been suggested that differential regulation of specic cellular
targets is the molecular basis for the distinct functions of JNK1 and
JNK2 in different tumors,12 there is a paucity of understanding
regarding the identication of targets regulated by the individual
JNK members.
A role for JNK signaling in MM has also been reported,
although the proposed functions are controversial.2427 Whereas
JNK activation was shown to promote cell-cycle progression
in MM,24 JNK activation mediates apoptosis induced by treatment with anti-myeloma agents, often through activation of
caspases.2527 Furthermore, inhibition of JNK with the chemical

1
Section of Inammation and Signal Transduction, Department of Medicine, Imperial College, London, UK; 2Cell signalling and Cancer Laboratory, Institute of Hepatology,
Foundation for Liver Research, London, UK; 3Centre for Haematology, Imperial College Healthcare NHS Trust and Imperial College, London, UK; 4Department of Histopathology,
Imperial College Healthcare NHS Trust and Imperial College, London, UK and 5British Heart Foundation Vascular Sciences Unit, National Heart and Lung Institute, Imperial College
Healthcare NHS Trust and Imperial College, London, UK. Correspondence: Dr S Papa, Cell signalling and Cancer Laboratory, Institute of Hepatology, Foundation for Liver Research,
69-75 Chenies Mews, London WC1E 6HX, UK or Dr C Bubici, Section of Inammation and Signal Transduction, Deparment of Medicine, Imperial College London, Commonwealth
Building R10.N6, Du Cane Road, London W12 0NN, UK.
E-mail: s.papa@researchinliver.org.uk or c.bubici@imperial.ac.uk
6
These authors contributed equally to this work.
Received 8 May 2012; revised 31 July 2012; accepted 9 August 2012; published online 8 October 2012

PARP14 in multiple myeloma

A Barbarulo et al

4232
inhibitor, SP600125, induces growth arrest in MM cell lines.24
However, the individual contribution of JNK1 and JNK2 to the
pathogenesis of MM is not known due to the lack of chemical
inhibitors specic for each JNK protein.28
Here we show that JNK2 is constitutively active in myeloma
cells, and inhibits the pro-apoptotic activity of JNK1 through
effects on poly(ADP-ribose) polymerase (PARP)14, thus promoting
survival in myeloma cells. PARP14, a member of the macro-PARP
subfamily of PARP proteins, is an ADP ribosyltransferase that
transduces survival signals in murine primary B cells by regulating
expression of B-cell survival factors as well as by repressing an
apoptotic program involving caspases.29,30 In the mouse, PARP14
has been shown to facilitate B-lymphoid oncogenesis driven by
oncogenic c-Myc.31 However, the mechanisms underlying the
survival function of PARP14 have not been previously investigated
in MM or other types of human cancer.
We demonstrate that PARP14 acts as a physiological downstream effector of the JNK2-dependent pro-survival signal by
binding to JNK1 and inhibiting its activity. We also nd that
expression of PARP14 correlates with disease progression and
poor prognosis in MM. Therefore, this study describes a novel
regulatory pathway in myeloma cells through which JNK2
promotes survival and suggests that selective inhibition of PARP14
might be of therapeutic value.

RESULTS
JNK2 is constitutively active in myeloma cells
Activation of JNK in BM trephine biopsies from four patients with
newly diagnosed MM was assessed by immunohistochemistry
using an antibody that recognizes both phospho-active JNK1 and
JNK2 (p-JNK). All myeloma cells showed strong nuclear staining for
p-JNK in all the samples examined. In contrast, no p-JNK staining
was detectable in the normal plasma cells in tonsillar samples
(n 3) (Figure 1a). Consistently, analysis of JNK activation by
immunoblotting in six additional MM cases showed considerable
increase of p-JNK in BM plasma cells (selected as CD138
CD38hiCD45low/  )32 in ve out of six cases compared with normal
B cells (Figure 1b, left). Interestingly, we also observed that
intensity of p-JNK2 (p-p54, which is the prominent spliced form of
JNK2) was signicantly higher than p-JNK1 (p-p46, the prominent
spliced form of JNK1)11 (Figure 1b, right), indicating that JNK2
activity was increased in myeloma plasma cells. Increased levels of
p-JNK2 were also found in a panel of MM cell lines compared with
p-JNK1 (Figure 1c). The differences in the levels of p-JNK2 versus
p-JNK1 in MM cell lines and primary tumors were not due to
differential protein expression (Figures 1b and c). Further
validation was obtained using a JNK2 kinase assay. A dramatically
elevated JNK2 activity was detected in MM cell lines, whereas
normal B cells lacked a comparable JNK2 activity (Figure 1c). In
contrast, JNK1 kinase assay showed that JNK1 activity was nearly
undetectable in all MM cells tested, and comparable to that found
in normal B cells (Figure 1c). These data indicate that JNK2 is
constitutively activated in myeloma plasma cells.
To establish the function of constitutive JNK2 activity in
myeloma cells, expression of JNK2 was knocked down in the
human MM cell line, RPMI-8226, using JNK2-specic shRNA
lentiviruses. We used the pLL3.7 lentiviral vector to expresses
JNK2 shRNA (shJNK2) together with GFP in a bicistronic
fashion.33,34 Lentiviruses expressing JNK1 shRNA (shJNK1) or
non-specic shRNA (shNS) were used as controls. Knockdown
was specic, as shRNAs did not affect expression of other related
MAPK proteins (ERK1/2 and p38), as veried by western blots (WB)
(Figure 1d). As shown in Figure 1e, shJNK2 expression caused a
progressive reduction of the GFP-positive cells over time
compared with shJNK1 or shNS, despite the three lentiviruses
each yielding GFP-positive cells in RPMI-8226 cells with
Oncogene (2013) 4231 4242

comparable efciency 2 days after infection (Supplementary

Figure S1A). This demonstrates that cells expressing shJNK2 and
GFP were negatively selected from the population due to cell
death and indicates a role for JNK2 but not JNK1 in RPMI-8226 cell
viability. In contrast, BJAB cells, which are derived from Burkitts
lymphoma (BL) tissue, were unaffected by infection with shJNK1,
shJNK2 or shNS lentiviruses (Figure 1f) despite JNK1 or JNK2
protein being effectively knocked down (Figure 1g). These results
are consistent with other studies demonstrating redundant roles
for JNK1 and JNK2 proteins in BLs.14,35
Knockdown of JNK2 induces apoptosis
To provide further evidence for cell death in shJNK2-infected MM
cells, we quantied apoptosis measuring the sub-G1 population
using propidium iodide (PI) staining as well as AnnexinV/7AAD
staining and activation of caspases. Flow cytometric (FCM) analysis
of PI-stained RPMI-8226 cells as early as 4 days after infection
revealed that JNK2 silencing resulted in an approximate sixfold
increase in apoptosis, compared with JNK1 or control silencing
(Figure 2a; Supplementary Figure S1B). The percentage of
apoptotic cells in shJNK2-infected RPMI-8226 cells further
increased with time, and by 8 days post transduction had reached
89.8% (Figure 2a; Supplementary Figure S1B). We extended these
analyses to two additional MM cell lines, MM1.S and JJN3. Efcient
knockdown of JNK2 correlated with a signicant induction of
apoptosis (Figures 2a and b; Supplementary Figures S1C,D).
Depletion of JNK2 also resulted in a robust cleavage of
procaspase-9, -3 as well as of downstream PARP1, and the
appearance of active cleavage products of the expected size in
each case (shJNK2-infected cells). As expected, in BJAB cells, JNK2
depletion did not induce apoptosis (Figure 2a; Supplementary
Figure S1F), conrming that JNK2 is not essential for basal survival
of BJAB cells. In contrast to JNK2 silencing, the silencing of JNK1
did not affect survival in any of the cell lines tested (Figures 2a and
b). Together, these data indicate that JNK2 is required for
myeloma cell survival.
JNK2 constitutively inhibits JNK1-dependent apoptosis in MM cells
As previous studies have shown that JNK2 negatively regulates
JNK1 activity and promotes survival in colorectal cancer cell lines
by constitutively suppressing JNK1-mediated apoptosis,36,37 we
investigated whether this was the case for MM. In-vitro kinase
assays revealed that JNK2 knockdown led to a robust increase in
JNK1 activity in MM cells compared with control shNS-infected
cells (Figure 2c, JNK1 K.A.). This effect of JNK2 knockdown was
specic, as p38 and ERK1/2 phosphorylation was comparable in
shNS-, shJNK1- and shJNK2-infected cells (Figure 2c). Notably,
JNK2 depletion did not result in a signicant increase of JNK1
activity in BJAB cells (Supplementary Figure S1F, compare lanes NS
and JNK2). As predicted, in all cell lines tested, JNK2 activity was
unaffected by JNK1 depletion (Figure 2c; Supplementary Figure
S1F, JNK2 K.A., compare lanes NS and JNK1). The data indicate that
JNK2 specically inhibits basal JNK1 activity in MM cells.
Next, we examined whether JNK1 silencing can rescue JNK2depleted cells from apoptosis. As shown in Figures 2d and e,
apoptosis induced by JNK2 depletion was completely reversed by
co-silencing JNK1 with JNK2. Remarkably, cleavage of casapse-3
was abolished when JNK1 was co-silenced with JNK2 (Figure 2f),
indicating that JNK1 is an essential mediator of apoptosis induced
by JNK2 depletion. Our data therefore indicate that JNK2 regulates
myeloma cell survival by suppressing basal JNK1-mediated
apoptosis.
JNK2 regulates expression of PARP14 in MM cells
To dissect the mechanism of JNK1 inhibition by JNK2 in myeloma
cells, we screened for known B-cell survival factors responsive to
& 2013 Macmillan Publishers Limited

PARP14 in multiple myeloma

A Barbarulo et al

4233

Figure 1. JNK2 is constitutively activated in MM primary tumors and cell lines and targeting JNK2 results in MM cell death. (a) Immunohistochemical
analysis of phospho-JNK (p-JNK) expression in BM biopsies of newly diagnosed MM patients (Pt 4) and normal tonsillar samples (n 3).
Representative sections are shown. Positive staining for p-JNK is demonstrated in myeloma plasma cells (white arrowheads) by brown nuclear signal.
Positivity for p-JNK is completely absent in normal plasma cells (black arrowheads) in the tonsillar sample. Images are showing  600 magnification.
(b) WBs (left) with antibodies against phosphorylated or total JNK1 and JNK2 showing increased expression of p-JNK in CD138-purified myeloma
from six additional newly diagnosed patients with MM (Pt. 510) compared with normal B cells isolated from human peripheral blood mononuclear
cells (PBMCs) of three healthy donors. Right, densitometry analysis of p-p54 and p-p46 bands versus total JNK2p54 and JNK1p46, respectively.
Intensity of p-p54 (p-JNK2) is significantly higher than p-p46 (p-JNK1). The two-tailed Wilcoxon signed-rank test was used to compare expression of
p-p54 and p-p46 in CD138-purified myeloma cells from MM patients. Each symbol represents an individual patient; horizontal lines indicate the
median. (c) WBs showing increased expression of p-JNK in a collection of human MM cell lines compared with normal B cells, and increased levels of
p-p54 (p-JNK2) compared with p-p46 (p-JNK1). In-vitro kinase assays (K.A.) with GST-c-Jun as substrate showing JNK1 and JNK2 activities in
indicated MM cell lines and normal B cells. Total JNK1 and JNK2, and b-actin were used as loading control. (d) WBs with antibodies against both
spliced variants of JNK1 (JNK1p46, JNK1p54) and JNK2 (JNK2p46 and JNK2p54) showing the knockdown of either JNK1 or JNK2 proteins in RPMI8226 cells infected with lentiviruses expressing JNK1, JNK2, but not control (lane NS), shRNA. Total p38, ERK1/2 and b-actin were used as loading
control. (e, f) Analysis of GFP expression in RPMI-8226 MM and BJAB BL cells infected as in d showing a reduction of GFP-positive cells over time
in RPMI-8226 shJNK2 cells, but not in shJNK1, control or BJAB cells. Proportion of GFP-positive cells were assessed at indicated times and
expressed relative to the proportion on day 2 after infection. Values are the means.d. of three independent cultures. (g) JNK1 and JNK2 levels
were assessed by WB in BJAB cells infected as in f. A full colour version of this figure is available at the Oncogene journal online.
& 2013 Macmillan Publishers Limited

Oncogene (2013) 4231 4242

PARP14 in multiple myeloma

A Barbarulo et al

4234

Figure 2. JNK2 promotes MM cell survival by suppressing JNK1-mediated apoptosis. (a) RPMI-8226 and MM1.S MM cell lines and BJAB cells
were infected with lentiviruses expressing JNK1, JNK2 or control non-specific (NS) shRNA, and after the times indicated, stained with PI and
analyzed by FCM. Shown is the percentage of apoptotic cells in the population (sub-G1 DNA content). Values are the meanss.d. of three
independent experiments. (b) Human myeloma JJN3 cells were infected as described in (a), and 6 days after infection the percentage of
AnnexinV cells showing either AnnexinV /7AAD  (early apoptosis) and AnnexinV /7AAD (late apoptosis) was assayed by FCM using
AnnexinV/7AAD staining. Data are expressed as means.d. of three separate experiments. (c) WBs, in-vitro JNK1 and JNK2 kinase assay (K.A.)
of lysates from RPMI-8226 and JJN3 cells infected as in (a) for 8 or 10 days, respectively, showing cleavage of caspases and PARP1 as well as
activation of JNK1 and JNK2 using the indicated antibodies and GST-c-Jun as substrate, respectively. Closed and open arrowheads indicate
the pro-cleaved and cleaved (active) products of the indicated proteins, respectively. Activation of p38 and ERK1/2 were analyzed by
immunoblotting with the indicated phospho-specific (p) antibodies. b-actin was used as loading control. (d, e) Co-silencing of JNK1 and JNK2
rescues MM cells from apoptosis induced by JNK2 depletion. RPMI-8226, MM1.S and JJN3 cells were infected with lentiviruses expressing JNK2
shRNA or JNK1/JNK2 shRNA (that is, silencing simultaneously both proteins), and the percentage of sub-G1 population and AnnexinV cells
were determined by FCM at day 6, 10 or 8 of infection. (f ) WBs of extracts from MM cells used in d and e showing levels of cleaved-active
caspase-3 (Cl. caspase-3) and total JNK1/JNK2. b-actin was used as loading control. Representative results from at least three experiments are
shown.

JNK2 silencing. PARP14 is a well-established pro-survival protein

that has been implicated in protecting B lymphocytes against
apoptosis.30 PARP14 expression has been also shown to promote
B-cell lymphomagenesis through supporting cell survival signals
in mice.31 These observations led us to investigate whether levels
of PARP14 were affected by JNK2 depletion. Indeed, PARP14
expression levels were markedly reduced in MM cells by
knockdown of JNK2 but not JNK1 or control (Figure 3a). The
reduction in PARP14 levels was specic, as there was no difference
between shJNK1-, shJNK2-, and shNS-infected cells in terms of
PARP9, another macro-PARP family member.38 This indicates that
JNK2 modulates PARP14 expression in myeloma cells (Figure 3a).
Oncogene (2013) 4231 4242

Intriguingly, reduced PARP14 protein levels were not observed in

JNK2-silenced BJAB cells, suggesting that PARP14 expression is
controlled by cancer cell-specic mechanisms (Figure 3a).
To examine the expression of PARP14 further, BM biopsies from
four MM patients were examined by immunohistochemistry. Over
80% of the neoplastic cells in all myeloma samples tested had
high expression of PARP14. Normal control sections (n 3)
showed very low PARP14 expression in a proportion of plasma
cells within the tonsillar tissue. As with p-JNK, PARP14 expression
was stronger in the nucleus as compared with the cytoplasm
(Figure 3b). The expression of PARP14 was also conrmed by
immunoblotting using MM cell lines and plasma cells obtained
& 2013 Macmillan Publishers Limited

PARP14 in multiple myeloma

A Barbarulo et al

4235

Figure 3. PARP14 is overexpressed in MM cells, and its expression correlates with disease progression and depends on JNK2. (a) WBs showing
a reduction of PARP14 levels in MM cells (RPMI-8226, MM1.S and JJN3) but not in BJAB cells infected with lentiviruses expressing JNK2 shRNA
compared with JNK1 shRNA or control NS shRNA. PARP9 and b-actin levels were analyzed with appropriate antibodies as controls.
(b) Immunohistochemical analysis of PARP14 expression in BM biopsies of newly diagnosed MM patients (Pt 4) and normal tonsillar samples
(n 3). Representative sections are shown. Positive staining for PARP14 is demonstrated in myeloma plasma cells (white arrowheads) by
brown nuclear signal. Positivity for PARP14 is completely absent in plasma cells (black arrowheads) in the tonsillar sample. Images are showing
 600 magnification. (c) WBs showing PARP14 expression in CD138 -selected cells from the BM aspirates of six additional MM patients
(Pt. 510) and B cells from peripheral blood mononuclear cells (PBMCs) of three healthy donors as well as indicated MM and BL cell lines.
b-actin was used as loading control. (d, left) Gene-expression analysis of PARP14 from oligonucleotide microarray data GSE2658 and GSE5900
(Zhan et al.39,40) showing significantly higher expression of PARP14 transcripts (log2 transformed, median-centered values) in MM plasma cells
(n 414) compared with normal BM (n 22) or monoclonal gammopathy of unknown significance (MGUS) (n 44) plasma cells (Po0.0001,
nonparametric KruskalWallis one-way analysis of variance; *Po0.05 and ***Po0.001, Dunns multiple comparison post-hoc tests). For each
box plot, the whiskers represent the 2.597.5th percentile range of values, the lower and up boundaries denote the 25th and the 75th
percentile of each data set, respectively, and the horizontal line represents the median value for each group. (d, right) Gene-expression
profiling of PARP14 from oligonucleotide microarray data GSE12453 (Brune et al.41) showing no significant differences of PARP14 transcripts
levels between normal purified tonsillar germinal center B cells (centrocytes and centroblasts) (n 5) and Burkitts lymphoma (BL) cases (n 5)
(P 0.1013, nonparametric KruskalWallis one-way analysis of variance). The horizontal line indicates the mean. A full colour version of this
figure is available at the Oncogene journal online.

from the BM aspirates of six additional MM patients. PARP14 levels

were increased in MM lines (sensitive to JNK2 depletion) and
primary cells compared with normal B cells (Figure 3c). In contrast,
& 2013 Macmillan Publishers Limited

two of three JNK2-independent BL cell lines (data not shown),

Ca46 and Raji, did not have detectable PARP14 protein (Figure 3b).
The third cell line, BJAB, had high levels of PARP14 protein
Oncogene (2013) 4231 4242

PARP14 in multiple myeloma

A Barbarulo et al

4236
(Figure 3b), conrming that JNK2-independent mechanisms of
regulation exist. To gain additional clinical insights, we examined
the expression of PARP14 transcripts in MM and BL using publicly
available gene-expression data sets.3941 We observed that
PARP14 expression signicantly increases in MM plasma cells
compared with monoclonal gammopathy of undetermined
signicance (MGUS) or normal BM plasma cells, indicating a
positive correlation between PARP14 expression and disease
progression (Figure 3d, left). In contrast, no signicant differences
were observed between normal germinal center B cells and BL
cases (Figure 3d, right).
Knockdown of PARP14 induces apoptosis
We next asked if PARP14, like JNK2, is required for MM cell
survival. As with JNK2, depletion of PARP14 by shRNA (shPARP14)
induced apoptosis (Figures 4ac). Silencing was specic, as
shPARP14 did not affect expression of PARP9 (Figure 4d). The
impact of PARP14 knockdown in two additional MM cell lines,
OPM-2 and KMS12-BM, was monitored by the progressive
disappearance of GFP-positive cells in shPARP14-infected cells
(Figure 4e). Interestingly, PARP14 knockdown markedly increased
JNK1 activity with magnitude similar to JNK2 knockdown,
indicating that PARP14 is required for JNK1 inhibition (Figure 4c,
Supplementary Figure S2A). This effect of PARP14 knockdown was
specic, as p38 and ERK1/2 phosphorylation/activation was
comparable in shNS- and shPARP14-infected cells (Supplementary Figure S2B). Similarly, JNK1 activation and a dose-dependent
apoptosis were also induced by the pan-PARP inhibitor
PJ-34, which has been previously used to examine PARP14
function (Supplementary Figures S2C,D).29,30,42 Notably, PARP14
knockdown did not affect survival of BJAB cells (Figures 4a and c),
despite these cells exhibit high PARP14 protein levels (Figure 3c)
that were effectively knocked down (Figure 4d), in accordance
with their insensitivity to JNK2 depletion (see Figures 1f and 2a).
Together, these data indicate that knockdown of PARP14
resembles JNK2 knockdown, and suggest that JNK2 may mediate
MM cell survival by regulating the levels of PARP14.
PARP14 mediates survival signaling by JNK2
To test directly the role of PARP14 in JNK2-mediated survival in
MM, we examined the effects of ectopic expression of PARP14 in
JNK2-depleted cells using the lentiviral expression vector, pWPI.34
Because the large size of the PARP14 ORF precluded efcient
packaging of lentivirus particles, we generated PARP14(DN553), a
truncated form of PARP14 that has previously been found to
mimic its physiological functions (Figure 5a).29 We tested whether
apoptosis caused by JNK2 knockdown could be reversed by
ectopic expression of PARP14(DN553). RPMI-8226 cells infected
with either PARP14(DN553) or empty pWPI lentiviruses were
double infected with shJNK2 lentivirus. PARP14(DN553) overexpression was conrmed by WB (Figure 5b). Of note in
PARP14(DN553)-overexpressing cells, levels of total PARP14
(endogenous plus exogenous) were comparable to those in
control cells (Figure 5b; compare lane 1 with lane 3). As shown in
Figure 5c (left panel), apoptosis induced by JNK2 knockdown was
completely reversed by PARP14(DN553) overexpression in RPMI8226 cells. Notably, the protective efcacy of PARP14(DN553)
overexpression against this killing was comparable to that of
control shNS-infected cells (see Figure 2a), whereas pWPI
expression afforded no protection to apoptotic shJNK2-infected
cells. Similar results were observed in additional MM cells (Figures
5c and d and data not shown). Remarkably, ectopic expression of
PARP14(DN553) suppressed JNK1 activity in JNK2-depleted cells
(Figure 5b; JNK1 K.A.). These results show that restoration of
PARP14 expression rescues the apoptotic phenotype in myeloma
cells caused by JNK2 depletion.
Oncogene (2013) 4231 4242

PARP14 binds and inhibits JNK1 to promote survival

As a tool to further dissect the mechanism of JNK1 inhibition by
JNK2 and PARP14, we co-expressed FLAG-tagged PARP14 in
HEK293T cells with HA-JNK2p54, HA-JNK1p46 or HA-empty vector,
and investigated protein associations by combined immunoprecipitations and WB analyses. PARP14 specically bound to
JNK1p46, but not JNK2p54, and this binding was conrmed with
PARP14(DN553) and PARP14(DN1470) (a mutant lacking the three
macro domains) (Figure 6a; see also Figure 5a), indicating that the
C-terminal of PARP14 is involved in this interaction. Importantly,
an association between endogenous PARP14 and JNK1 was
conrmed in both RPMI-8226 and BJAB cells. Anti-JNK1 antibodies
co-immunoprecipitated PARP14 from parental cells, but not from
cells depleted of PARP14 (Figure 6b). JNK1 was immunoprecipitated at comparable levels in both parental and shPARP14infected cells (Figure 6b; WB: JNK1), and was not co-precipitated
by isotype-matched control antibodies (data not shown). We
conclude that endogenous PARP14 and JNK1 associate under
physiological conditions.
To test whether the binding of PARP14 to JNK1 is sufcient to inhibit JNK1, we generated recombinant GST-tagged
PARP14(DN1470) protein for use in JNK1 kinase assay. Neither
recombinant PARP14(DN1470) nor control protein inhibited active
JNK1 in vitro (Supplementary Figures S3A,B), indicating that
PARP14/JNK1 binding is not alone sufcient to explain the
inhibitory effect of PARP14 on JNK1. Moreover, JNK1 still bound
to MKK7, an upstream JNK activator, in HEK293T cells overexpressing PARP14 (Supplementary Figure S3C).43 Together, these
results suggest that an additional inhibitory mechanism of JNK1
by PARP14 exits, in vivo.
We next investigated whether the protective function of
PARP14 also depended on inhibition of JNK1. As shown in
Figure 6c, co-silencing JNK1 with PARP14 completely rescued MM
cells depleted of PARP14 from apoptosis, indicating that PARP14
suppresses JNK1-mediated apoptosis. The protective effect of
JNK1 silencing was reminiscent to that seen in JNK2-depleted cells
(Figures 2df), indicating that PARP14 and JNK2 have similar
effects on JNK1-mediated apoptosis. Together, these ndings
identify PARP14 as the mechanistic link by which JNK2 suppresses
JNK1, thus promoting myeloma cell survival.
PARP14 inhibition can potentiate cytotoxicity of anti-myeloma agents
As the JNK2-PARP14-JNK1 regulatory axis sustains the survival of
MM cells, we investigated whether inhibition of PARP14 could be
used as a therapeutic strategy in MM. First, we evaluated whether
PARP14 expression had any impact on anti-MM therapy.
Interrogation of the data set GSE2658 revealed that high
expression of PARP14 transcripts correlated with a signicantly
shorter survival in patients treated with the Total Therapy 3
protocol,39,40 suggesting that PARP14 may be implicated in
treatment outcomes (Figure 7a). Next, we assessed the effects of
PJ-34 in combination with the anti-myeloma agent dexamethasone on MM cells. PJ-34 signicantly enhanced cytotoxicity
induced by dexamethasone (Figure 7b), suggesting an additive
effect. To evaluate whether this effect of PJ-34 was attributable to
PARP14 inhibition, we either silenced or overexpressed PARP14 in
MM1.S cells and assayed their viability after treatment with
dexamethasone. Depletion of PARP14 signicantly enhanced the
sensitization of MM1.S cells to dexamethasone compared with the
control cells (Figure 7c). Conversely, MM1.S cells overexpressing
truncated PARP14 were resistant to dexamethasone-induced
apoptosis compared with control vector. Moreover, addition of
PJ-34 overcame this resistance and resulted in cell death
(Figure 7d). Finally, PARP14 knockdown also sensitized MM1.S
cells to bortezomib, lenalidomide and, to a lesser extent,
thalidomide and other clinically relevant compounds used for
the treatment of MM (Figures 7e and f).
& 2013 Macmillan Publishers Limited

PARP14 in multiple myeloma

A Barbarulo et al

4237

Figure 4. PARP14 expression is required for survival of MM cells. (a) RPMI-8226, MM1.S and BJAB cells were infected with lentiviruses
expressing PARP14 or control NS shRNA, and after the times indicated, stained with PI and analyzed by FCM. Shown is the percentage
of apoptotic cells in the population (sub-G1 DNA content). Values are the meanss.d. of three independent experiments. (b) JJN3 cells were
infected as described in (a), and the percentage of AnnexinV cells showing either AnnexinV /7AAD  (early apoptosis) and
AnnexinV /7AAD (late apoptosis) was assayed by FCM using AnnexinV/7AAD staining. Data are expressed as means.d. of three separate
experiments. (c) WBs and in-vitro JNK1 kinase assay (K.A) of lysates from RPMI-8226, JJN3 and BJAB cells infected as in a and b for 10 days,
showing cleavage of indicated caspases and PARP1 as well as activation of JNK1 using the indicated antibodies and GST-c-Jun as substrate,
respectively. Closed and open arrowheads indicate the pro-cleaved and cleaved (active) products of the indicated proteins, respectively.
b-actin was used as loading control. (d) WBs showing the knockdown of PARP14 levels in indicated MM cell lines and BJAB cells infected with
lentiviruses expressing PARP14, but not control (lane NS), shRNA. PARP9 and b-actin were used as loading control. (e) Analysis of GFP
expression in OPM2 and KMS12-BM infected as in d showing a depletion of GFP-positive cells over time in shPARP14 cells, but not in control
cells. Proportion of GFP-positive cells were assessed at indicated times and expressed relative to the proportion on day 4 after infection.
A representative experiment out of three is shown.

& 2013 Macmillan Publishers Limited

Oncogene (2013) 4231 4242

PARP14 in multiple myeloma

A Barbarulo et al

4238

Figure 5. PARP14 suppresses JNK1-mediated apoptosis in JNK2-depleted cells. (a) Schematic representation of full-length (FL) (amino-acid
residues 1 to 1801) FLAG-tagged PARP14 (FLAG-PARP14 (FL)) and the C-terminal fragments of PARP14, FLAG-PARP14(DN553) and FLAGPARP14(DN1470), used in complementary and co-immunoprecipitation experiments. The three macro domains (Macro), the PARP-domain and
the proteinprotein interaction domain WWE are indicated. (b) Representative WBs showing ectopic expression of FLAG-PARP14(DN553) in
RPMI-8226 cells infected with lentiviruses expressing FLAG-PARP14(DN553) or empty virus (pWPI), and then double infected with lentiviruses
expressing JNK2 shRNA. Endogenous (FL) and exogenous PARP14 (FLAG-PARP14(DN553)) are indicated. Kinase assays (K.A.) showing inhibition
of JNK1 activity in the same RPMI-8226-infected cells. GST-c-Jun was used as substrate. Total JNK1/2 and b-actin was used as loading control.
(c, d) Ectopically expressed FLAG-PARP14(DN553) rescues JNK2-depleted cells from apoptosis. Myeloma cell lines were infected as in b, and
apoptosis was assessed by FCM using PI (c) or AnnexinV/7AAD staining (d). Shown are bar graphs representing the mean (s.d.) percentage
of apoptotic cells in the population (sub-G1 DNA content or AnnexinV cells) in three independent experiments.

DISCUSSION
The identication of pathways that govern survival of myeloma
plasma cells is of great interest for possible therapeutic applications, as deregulated survival is a key feature of myelomagenesis.1,2,4
In this study, we show that JNK2 is a master regulator of survival
of myeloma cells, and dene a new regulatory pathway linking
JNK2 to JNK1 through regulation of a novel survival factor of
B-lineage cells, PARP14.30 Moreover, we report for the rst time a
correlation of PARP14 expression with tumor survival and
progression in MM.
Elevated activity of p-JNK was found in myeloma plasma cells in
all the BM trephine biopsies of newly diagnosed MM patients
tested, but was nearly undetectable in normal plasma cells in
the tonsillar samples. This basal JNK activity in myeloma plasma
cells exhibits exclusive nuclear localization, which is consistent
with published reports demonstrating the accumulation of p-JNK
in the nucleus of cancer cells.18,23 Unlike JNK1, JNK2 has autophosphorylation activity that leads to its constitutive activation
in several tumor cell types.16,18,44 Consistent with this, we found
that the major JNK protein that is phosphorylated in MM cell
Oncogene (2013) 4231 4242

lines and primary patient-derived cells is JNK2 (predominantly

present as p-p54 isoform), but not JNK1 (predominantly present as
p-p46 isoform), despite p46 and p54 JNK proteins being expressed
at similar levels. Moreover, p-JNK2 levels and JNK2 activities in
myeloma cells were higher than in normal B cells, whereas JNK1
activities were weak and remained unchanged, suggesting a
pathogenic role for JNK2 in this disease.
We systematically evaluated the functional signicance of
constitutive JNK2 activity using shRNA interference, and demonstrated that JNK2 is essential for survival of myeloma cells. In this
system, silencing of JNK2 or JNK1 expression was found to be
highly specic, as levels of ERK- and p38-MAPK were not affected.
The survival function of JNK2 is not shared by JNK1. Indeed, while
depletion of endogenous JNK2 protein enhanced apoptosis in MM
cell lines, reduction in the protein levels of JNK1 did not affect the
growth rates of these cancerous cells. Moreover, constitutive
activation of JNK2 negatively regulated activity of JNK1, as
increased JNK1 activity was observed in JNK2-depleted cells.
Oncogenic function of JNK2 on survival was depending on JNK1mediated apoptosis. Our ndings demonstrate that co-silencing of
JNK1 with JNK2 completely rescued MM cells from apoptosis
& 2013 Macmillan Publishers Limited

PARP14 in multiple myeloma

A Barbarulo et al

4239

Figure 6. PARP14 specifically interacts with JNK1. (a) HEK293T cells

were transfected with the indicated FLAG-tagged PARP14 and
HA-tagged empty vector (HA-vector), HA-JNK1 or HA-JNK2. After
24 h, cells were harvested and lysed. Lysates were immunoprecipitated (IP) with anti-FLAG and subject to WBs as indicated. Results are
representative of three experiments. (b) WBs with anti-JNK1
immunoprecipitates (top) or lysates (bottom) from RPMI-8226 (left)
and BJAB (right) shPARP14 or uninfected (parental) cells, showing
association of endogenous PARP14 with JNK1. (c) Co-silencing of
JNK1 with PARP14 rescues MM cells from apoptosis induced by
shPARP14. RPMI-8226, MM1.S and JJN3 cells were infected with
lentiviruses expressing JNK1 or control NS shRNA, double infected
with shPARP14, and then the percentage of AnnexinV cells
showing either AnnexinV /7AAD- (early apoptosis) and AnnexinV /
7AAD (late apoptosis) was assayed by FCM using AnnexinV/
7AAD staining. Values are the means.d. of three independent
experiments.

induced by JNK2 depletion. This was further supported by our

observations that JNK1 co-silencing inhibited activation of effector
caspase-3 triggered by silencing of JNK2 in myeloma cell lines. The
fact that no alteration of phosphorylation of ERK and p38-MAPK
was observed in JNK2-depleted cells further suggested that JNK1
is the specic MAPK involved in shJNK2-inducing cell apoptosis.
& 2013 Macmillan Publishers Limited

Our results indicate that JNK2 is not a universal oncogene, as

knockdown of JNK2 expression in the BL cell line BJAB did not
alter the basal cell viability. Recent data has implied that JNK2mediated cell survival involves the regulation of selective prosurvival proteins or tumor suppressors in specic cellular contexts.
JNK2 cooperates with Ras to promote epidermal neoplasia,45
phosphorylates oncoproteins such as STAT3 in non-small cell lung
carcinoma and Akt in glioblastoma,16,18 and contributes to the
inhibition of tumor suppressors in breast and colon cancer
cells.46,47 Interestingly, we found that JNK2 depletion leads to a
reduction of PARP14 levels in MM cells, but not in BJAB,
suggesting that the JNK2-dependent regulation of PARP14 is
limited to terminally differentiated B cells. Elevated expression
levels of PARP14 were found in MM cell lines as well as primary
CD138 MM cells and, intriguingly, correlated with the nuclear
activity of JNK2 in MM primary tumors. PARP14 levels also appear
to have important clinical implications for patients with MM, as
PARP14 gene expression increases with disease progression from
normal BM plasma cells to early (monoclonal gammopathy of
undetermined signicance) and then to active stages of MM.
Furthermore, knockdown of PARP14 recapitulated knockdown of
JNK2 in MM cells by mediating apoptosis, consistent with the
hypothesis that JNK2 mediates survival by regulating PARP14
expression. Complementary experiments further supported this
conclusion by showing that the increased apoptosis in JNK2depleted cells was completely suppressed by expression of
PARP14. Thus, the function of JNK2 to inhibit apoptosis can be
replaced by PARP14, consistent with the established role of this
protein as a protective factor in primary murine B lymphocytes.30
The obvious question that remains to be addressed relates to the
mechanisms of reduced PARP14 expression caused by JNK2
depletion. JNK proteins regulate the activity of several
transcription factors, such as c-Jun and ATF-2, and the stability
of non-transcription factors such as Bcl2.11 Therefore, it is possible
that JNK2 may indirectly regulate PARP14 expression at either
transcriptional or posttranscriptional levels. Further studies are
required to address this question.
What is the molecular mechanism by which JNK2 and PARP14
proteins cooperate in regulating MM cell survival? The JNK2dependent suppression of JNK1 is apparently mediated through
PARP14, because JNK1 activity is similarly increased in PARP14depleted cells and the suppression of JNK1 was restored by
overexpression of PARP14(DN553) in cells depleted of JNK2.
Consistent with this, apoptosis caused by PARP14 knockdown in
MM cells was completely reversed by JNK1 co-silencing, indicating
JNK1 as a direct mediator of apoptosis triggered by PARP14
knockdown.
Along with the three tandem macro domains, PARP14 contains
a WWE domain and a catalytic PARP domain at the C-terminal.38,48
The WWE domain is predicted to mediate proteinprotein
interactions in ubiquitination and ADP-ribose conjugation systems.49 Indeed, it has been shown that PARP14 associates with
Stat6 and phosphoglucose isomerase through its C-terminus.29,50
Furthermore, PARP14 activity was shown to be essential in
mediating IL-4-induced protection of murine B lymphocytes
against apoptosis.30 Although we have not addressed whether
the enzymatic activity of PARP14 is strictly required for the
inactivation of JNK1 kinase, we found that PARP14 specically
interacts with JNK1 through its C-terminal portion and that
overexpression of the latter completely suppressed JNK1 activity.
We also found that blocking PARP activity using PJ-34 resulted in
JNK1 activation and apoptosis. Accordingly, neither PARP14/JNK1
interaction was alone sufcient for JNK1 suppression nor
prevented JNK1 from binding to MKK7, an upstream JNK1
activator. Therefore, it is possible that the PARP14 catalytic
activity might be involved in JNK1 inactivation. Studies to dene
the role of the PARP14 activity in inhibiting JNK1 kinase activity in
more detail are underway.
Oncogene (2013) 4231 4242

PARP14 in multiple myeloma

A Barbarulo et al

4240

Figure 7. PARP14 inhibition enhances cytotoxicity of anti-myeloma agents. (a) High PARP14 expression significantly correlates with
poor survival outcomes post therapy. Shown are the KaplanMeier overall survival (OS) curves of MM patients treated with total therapy 3
(TT3) according to PARP14 expression above or below the median value of 1082 based on a data set GSE2658 (Zhan et al.39,40) (P 0.0202).
(b) RPMI-8226 and MM1.S viability, as assessed by MTT assay, 48 h after treatment with dexamethasone (Dex) [5 mM or 10 mM], PJ-34 [50 mM],
or both in combination, is shown. Values are the means.d. of triplicate cultures. *Po0.05, **Po0.01 and ***Po0.001, two-tailed t-tests.
(c) MM1.S were infected with the indicated lentiviruses and treated with dexamethasone [10 mM]. Cell viability was measured by MTT assay at
48 h after treatment. Values are the means.d. of triplicate cultures. **Po0.01 and ***Po0.001, two-way ANOVA with Bonferronis post-hoc
tests. (d) MM1.S infected with lentiviruses expressing FLAG-PARP14(DN553) or empty virus (pWPI) were cultured for 48 h with dexamethasone
[5 mM or 10 mM] alone or in combination with PJ-34 [50 mM]. The bar graphs show the percentage of apoptotic cells in the population (sub-G1
DNA content) analyzed by FCM. Values are the means.d. of triplicate cultures. ***Po0.001, two-way ANOVA with Bonferronis post-hoc tests.
(e) MM1.S were infected as in c and then treated with bortezomib (Bortez) [5 nM]. Cell viability was measured by MTT assay at 24 h after
treatment. Values are the means.d. of triplicate cultures. ***Po0.001, two-way ANOVA with Bonferronis post-hoc tests. (f ) MM1.S were
infected as in c and then treated with thalidomide (Thal) (25 mg/ml), lenalidomide (Len) [1 mM] or left untreated for 72 h. Cell viability
was measured by MTT assay. Values are the means.d. of triplicate cultures. ***Po0.001, two-way ANOVA with Bonferronis post-hoc tests.
(g) Schematic model for JNK2-dependent myeloma cell survival. JNK2 suppresses basal apoptosis in myeloma cells via PARP14. Constitutive
JNK2 activity regulates levels of PARP14, which in turn inhibits JNK1, thereby suppressing the activation of endogenous apoptotic pathways.

Oncogene (2013) 4231 4242

& 2013 Macmillan Publishers Limited

PARP14 in multiple myeloma

A Barbarulo et al

4241
Finally, we observed that high PARP14 mRNA expression
correlated with poor survival outcomes in newly diagnosed MM
patients treated with total therapy 3 protocol.39,40 In line with
these clinical observations, we showed that inhibition of PARP14
can cooperate with anti-myeloma agents in inducing more
effective cell death, suggesting that PARP14 could be targeted
to improve anti-MM therapies. Therefore, it will be of interest to
examine the role of PARP14 in development of drug resistance.
In summary, we identify a novel pathway whereby constitutively active JNK2 promotes survival via PARP14, which in turn
associates with JNK1 and inhibits its kinase activity; this inhibition
is crucial for the suppression of basal apoptosis in myeloma cells
(Figure 7g). Our ndings dene for the rst time the individual role
of JNK1 and JNK2 in MM and demonstrate that PARP14 is a novel
physiological downstream effector of JNK2, thus expanding the
number of specic targets downstream of JNK2. These results may
open new avenues for therapeutic intervention in MM.
MATERIAL AND METHODS
Patient samples
After approval from the Hammersmith Hospitals NHS Trust Research Ethics
Committee and informed consent according to the Declaration of Helsinki,
BM samples were obtained from patients with MM undergoing routine
diagnostic procedure.
Mononuclear cells were isolated from BM aspirates by Ficoll gradient
centrifugation, and primary CD138 MM cells were positively sorted using
CD138-conjugated magnetic beads (Miltenyi-Biotec, Surrey, UK) as
previously reported.32

Cells and reagents

Human MM cell lines (RPMI-8226, JJN3, MM1.S, OPM-2, KMS12-BM, KMS11
and U266) and BL lines (BJAB, Ca46 and Raji) were cultured in RPMI1640
(Invitrogen, Paisley, UK) supplemented with 10% fetal bovine serum
(Sigma, Gillingham-Dorset, UK), 100 U/ml penicillin, 100 mg/ml streptomycin and 2mM L-glutamine (Invitrogen) in a 5% CO2 incubator at 37 1C.32,51
D-MEM (Invitrogen) was used as culture medium to grow human HEK293T
cells. Normal B cells were isolated from peripheral blood of healthy donors
using the B-cell isolation kit II (Miltenyi-Biotec), according to the
manufacturers instructions. Dexamethasone and PJ-34 were obtained
from Sigma. Bortezomib was obtained from the pharmacy of
Hammersmith Hospital. Thalidomide and Lenalidomide were obtained
from Tocris Bioscience (Bristol, UK) and Selleck Chemicals (Huston, TX,
USA), respectively.

Plasmids, co-transfection, protein purication, lentiviral

preparations and infections
pcDNA-FLAG, pSRa-HA-JNK1, pcDNA-HA-JNK2 were used previously.52
pcDNA-FLAG-PARP14 was obtained by sub-cloning the full-length human
complementary DNA (cDNA) of PARP14 (provided by Dr Michael Johns)
between the KpnIXhoI restriction site of pcDNA-FLAG. pcDNA-FLAGPARP14(DN553) and pcDNA-FLAG-PARP14(DN1470), encoding amino acids
5531801 and 14701801 of human PARP14, respectively, were obtained
by linearizing pcDNA-PARP14(FL) with appropriate restriction enzymes and
replacing original fragments with oligonucleotide linkers. For co-immunoprecipitations, HEK293T cells were co-transfected using the calcium
phosphate method with HA-tagged and FLAG-tagged plasmids, as
described before.52
To generate pGEX-PARP14(DN1470), the cDNA PmeIPstI fragment
of pcDNA-FLAG-PARP14(DN1470) was inserted into the BamHISmaI
sites of pGEX1 along with an appropriate oligonucleotide linker. GSTPARP14(DN1470) were expressed in E. coli and puried as previously
described.52
Oligonucleotides to obtain shRNA constructs were cloned into the
lentiviral vector, pLL3.7, using the HpaIXhoI restriction sites.33,34
Oligonucleotides used are listed in Supplementary Table S1. Non-specic
shRNA (shNS) was previously described.33 We also used a lentiviral vector
pWPI for complementary experiments.34 The expression constructs pWPIFLAG-PARP14, pWPI-FLAG-PARP14(DN553) and pWPI-HA-JNK2 were
generated by sub-cloning the full-length PARP14, PARP14(DN553) or
JNK2 cDNA sequences into the PmeI restriction site of pWPI. All constructs
& 2013 Macmillan Publishers Limited

were conrmed by nucleotide sequencing. Lentiviral preparations were

obtained, as previously described.33 Supernatants were harvested,
concentrated and then used to infect MM cell lines in presence of
8 mg/ml polybrene (Sigma).

Apoptosis and cell viability assays

Apoptosis was assessed by FCM analysis using PI nuclear staining,
as previously described,33,51 and Annexin-V-eFluor450 and 7-aminoactinomycin D (7AAD) (eBioscience, Hateld, UK) staining, according to
the manufacturers instructions. The percentage of apoptotic cells was
calculated on the total population, gating out the cellular fragments using
the FSC/SSC parameter. Cell viability was assessed using MTT reagent
according to the manufacturers protocol (Sigma). Proportion of live GFP
was evaluated over time by FCM analysis, and expressed relative to the
proportion at 2 or 4 days after lentiviral infection.

Western blot, kinase assay, co-immunoprecipitation and

antibodies.
WB, co-immunoprecipitation, JNK1 and JNK2 kinase assays (KA) were
performed as described previously.33,51,53 Antibodies used in these
protocols included JNK1 and JNK1/2 (BD Biosciences, Heidelberg,
Germany); caspase-3, caspase-9, phospho-JNK, phospho-p38, phosphoERK1/2, JNK2, p38, ERK1/2 (Cell Signaling, Danvers, MA, USA); anti-HAprobe, PARP9, b-actin (Santa Cruz, Heidelberg, Germany); PARP1
(Calbiochem, Lutterworth, UK); PARP14 and anti-FLAG-probe (Sigma).

Immunohistochemistry
Sample preparation was performed as previously described.54 Rabbit antihuman PARP14 and phospho-JNK were used in a dilution of 1:50. The
staining in the myeloma samples was reviewed and scored by a qualied
hematopathologist (K.N.) in a blinded fashion. Stained sections were
analyzed with a light microscope (BX51, Olympus, Hamburg, Germany)
carrying a DP70 digital camera.

Gene-expression proling
Publicly available gene-expression data for a series of MM (GSE2658 and
GSE5900) and BL (GSE12453) primary samples proled on Affymetrix U133
(Affymetrix, Santa Clara, CA, USA) platforms were downloaded and
analyzed against human PARP14 (probe 224701_at).3941

Statistical analysis
Data are shown as means.d. of three experiments. Statistical signicances of differences were determined using appropriate tests as
described in gure legends. The minimal level of signicance was Po0.05.

CONFLICT OF INTEREST
The authors declare no conict of interest.

ACKNOWLEDGEMENTS
We are grateful to S Chokshi, R Williams and J Dyson for critical comments on the
manuscript. We also thank D Trono (Ecole Polytechnique Federale de Lausanne,
Switzerland) for the pWPI lentiviral vector and M Johns (Imperial College) for the
human PARP14 cDNA; H Cooksley, A Riva (Institute of Hepatology), and PM Choy, MH
Lam and V Patel (MSc students) for technical support. GF receives funding from NIH
grants R01 CA084040 and R01 CA098583 and Cancer Research UK grant C26587/
A8839. DOH receives professorial support from the British Heart Foundation. This
study was primarily supported by a research grant from Kay Kendall Leukemia Fund
(KKL443) to CB, and in part by grants from Italian Association for Cancer Research
(Start-up 8557) and the Foundation for Liver Research to S.P.

REFERENCES
1 Anderson KC, Carrasco RD. Pathogenesis of myeloma. Annu Rev Pathol 2011;
6: 249274.
2 Raab MS, Podar K, Breitkreutz I, Richardson PG, Anderson KC. Multiple myeloma.
Lancet 2009; 374: 324339.
3 Rajkumar SV. Treatment of multiple myeloma. Nat Rev Clin Oncol 2011;
8: 479491.

Oncogene (2013) 4231 4242

PARP14 in multiple myeloma

A Barbarulo et al

4242
4 Hallek M, Bergsagel PL, Anderson KC. Multiple myeloma: increasing evidence for a
multistep transformation process. Blood 1998; 91: 321.
5 Witzig TE, Timm M, Larson D, Therneau T, Greipp PR. Measurement of apoptosis
and proliferation of bone marrow plasma cells in patients with plasma cell
proliferative disorders. Br J Haematol 1999; 104: 131137.
6 Platanias LC. Map kinase signaling pathways and hematologic malignancies.
Blood 2003; 101: 46674679.
7 Geest CR, Coffer PJ. MAPK signaling pathways in the regulation of hematopoiesis.
J Leukoc Biol 2009; 86: 237250.
8 Raman M, Chen W, Cobb MH. Differential regulation and properties of MAPKs.
Oncogene 2007; 26: 31003112.
9 Kim K, Kong SY, Fulciniti M, Li X, Song W, Nahar S et al. Blockade of the MEK/ERK
signalling cascade by AS703026, a novel selective MEK1/2 inhibitor, induces pleiotropic
anti-myeloma activity in vitro and in vivo. Br J Haematol 2010; 149: 537549.
10 Hideshima T, Akiyama M, Hayashi T, Richardson P, Schlossman R, Chauhan D et al.
Targeting p38 MAPK inhibits multiple myeloma cell growth in the bone marrow
milieu. Blood 2003; 101: 703705.
11 Bogoyevitch MA, Kobe B. Uses for JNK: the many and varied substrates of the
c-Jun N-terminal kinases. Microbiol Mol Biol Rev 2006; 70: 10611095.
12 Wagner EF, Nebreda AR. Signal integration by JNK and p38 MAPK pathways in
cancer development. Nat Rev Cancer 2009; 9: 537549.
13 Sabapathy K, Jochum W, Hochedlinger K, Chang L, Karin M, Wagner EF. Defective
neural tube morphogenesis and altered apoptosis in the absence of both JNK1
and JNK2. Mech Dev 1999; 89: 115124.
14 Anbalagan M, Sabapathy K. JNK1 and JNK2 play redundant functions in
Myc-induced B cell lymphoma formation. Int J Cancer 2012; 130: 19671969.
15 Kennedy NJ, Davis RJ. Role of JNK in tumor development. Cell Cycle 2003; 2: 199201.
16 Nitta RT, Del Vecchio CA, Chu AH, Mitra SS, Godwin AK, Wong AJ. The role of the
c-Jun N-terminal kinase 2-a-isoform in non-small cell lung carcinoma tumorigenesis. Oncogene 2011; 30: 234244.
17 Bost F, McKay R, Bost M, Potapova O, Dean NM, Mercola D. The Jun kinase 2
isoform is preferentially required for epidermal growth factor-induced
transformation of human A549 lung carcinoma cells. Mol Cell Biol 1999; 19:
19381949.
18 Cui J, Han SY, Wang C, Su W, Harshyne L, Holgado-Madruga M et al. c-Jun NH(2)terminal kinase 2alpha2 promotes the tumorigenicity of human glioblastoma
cells. Cancer Res 2006; 66: 1002410031.
19 Yoon C-H, Kim M-J, Kim R-K, Lim E-J, Choi K-S, An S et al. c-Jun N-terminal
kinase has a pivotal role in the maintenance of self-renewal and tumorigenicity
in glioma stem-like cells. Oncogene 2012; 31: 46554666.
20 Yang YM, Bost F, Charbono W, Dean N, McKay R, Rhim JS et al. c-Jun NH(2)terminal kinase mediates proliferation and tumor growth of human prostate
carcinoma. Clin Cancer Res 2003; 9: 391401.
21 Ke H, Harris R, Coloff JL, Jin JY, Leshin B, Miliani de Marval P et al. The c-Jun NH2terminal kinase 2 plays a dominant role in human epidermal neoplasia. Cancer Res
2010; 70: 30803088.
22 Chen N, Nomura M, She QB, Ma WY, Bode AM, Wang L et al. Suppression of skin
tumorigenesis in c-Jun NH(2)-terminal kinase-2-decient mice. Cancer Res 2001;
61: 39083912.
23 Hui L, Zatloukal K, Scheuch H, Stepniak E, Wagner EF. Proliferation of human HCC
cells and chemically induced mouse liver cancers requires JNK1-dependent p21
downregulation. J Clin Invest 2008; 118: 39433953.
24 Hideshima T, Hayashi T, Chauhan D, Akiyama M, Richardson P, Anderson K.
Biologic sequelae of c-Jun NH(2)-terminal kinase (JNK) activation in multiple
myeloma cell lines. Oncogene 2003; 22: 87978801.
25 Hideshima T, Mitsiades C, Akiyama M, Hayashi T, Chauhan D, Richardson P et al.
Molecular mechanisms mediating antimyeloma activity of proteasome inhibitor
PS-341. Blood 2003; 101: 15301534.
26 Kajiguchi T, Yamamoto K, Iida S, Ueda R, Emi N, Naoe T. Sustained activation of
c-jun-N-terminal kinase plays a critical role in arsenic trioxide-induced cell
apoptosis in multiple myeloma cell lines. Cancer Sci 2006; 97: 540545.
27 Chauhan D, Li G, Hideshima T, Podar K, Mitsiades C, Mitsiades N et al.
JNK-dependent release of mitochondrial protein, Smac, during apoptosis in
multiple myeloma (MM) cells. J Biol Chem 2003; 278: 1759317596.
28 Bogoyevitch MA, Ngoei KR, Zhao TT, Yeap YY, Ng DC. c-Jun N-terminal kinase
(JNK) signaling: recent advances and challenges. Biochim Biophys Acta 2010; 1804:
463475.
29 Goenka S, Boothby M. Selective potentiation of Stat-dependent gene expression
by collaborator of Stat6 (CoaSt6), a transcriptional cofactor. Proc Natl Acad Sci USA
2006; 103: 42104215.

30 Cho SH, Goenka S, Henttinen T, Gudapati P, Reinikainen A, Eischen CM et al.

PARP-14, a member of the B aggressive lymphoma family, transduces survival
signals in primary B cells. Blood 2009; 113: 24162425.
31 Cho SH, Ahn AK, Bhargava P, Lee CH, Eischen CM, McGuinness O et al. Glycolytic
rate and lymphomagenesis depend on PARP14, an ADP ribosyltransferase
of the B aggressive lymphoma (BAL) family. Proc Natl Acad Sci USA 2011; 108:
1597215977.
32 Spanoudakis E, Hu M, Naresh K, Terpos E, Melo V, Reid A et al. Regulation
of multiple myeloma survival and progression by CD1d. Blood 2009; 113:
24982507.
33 Pham CG, Bubici C, Zazzeroni F, Knabb JR, Papa S, Kuntzen C et al. Upregulation of
Twist-1 by NF-kappaB blocks cytotoxicity induced by chemotherapeutic drugs.
Mol Cell Biol 2007; 27: 39203935.
34 Wiznerowicz M, Trono D. Conditional suppression of cellular genes:
lentivirus vector-mediated drug-inducible RNA interference. J Virol 2003; 77:
89578961.
35 Gururajan M, Chui R, Karuppannan AK, Ke J, Jennings CD, Bondada S. c-Jun
N-terminal kinase (JNK) is required for survival and proliferation of B-lymphoma
cells. Blood 2005; 106: 13821391.
36 Liu J, Minemoto Y, Lin A. c-Jun N-terminal protein kinase 1 (JNK1), but not JNK2, is
essential for tumor necrosis factor alpha-induced c-Jun kinase activation and
apoptosis. Mol Cell Biol 2004; 24: 1084410856.
37 Ahmed SU, Milner J. Basal cancer cell survival involves JNK2 suppression of a
novel JNK1/c-Jun/Bcl-3 apoptotic network. PLoS One 2009; 4: e7305.
38 Schreiber V, Dantzer F, Ame JC, de Murcia G. Poly(ADP-ribose): novel functions for
an old molecule. Nat Rev Mol Cell Biol 2006; 7: 517528.
39 Zhan F, Huang Y, Colla S, Stewart JP, Hanamura I, Gupta S et al. The molecular
classication of multiple myeloma. Blood 2006; 108: 20202028.
40 Zhan F, Barlogie B, Arzoumanian V, Huang Y, Williams DR, Hollmig K et al.
Gene-expression signature of benign monoclonal gammopathy evident in multiple myeloma is linked to good prognosis. Blood 2007; 109: 16921700.
41 Brune V, Tiacci E, Pfeil I, Doring C, Eckerle S, van Noesel CJ et al. Origin and
pathogenesis of nodular lymphocyte-predominant Hodgkin lymphoma as
revealed by global gene expression analysis. J Exp Med 2008; 205: 22512268.
42 Gangopadhyay NN, Luketich JD, Opest A, Visus C, Meyer EM, Landreneau R et al.
Inhibition of poly(ADP-ribose) polymerase (PARP) induces apoptosis in lung
cancer cell lines. Cancer Invest 2011; 29: 608616.
43 Bardwell AJ, Frankson E, Bardwell L. Selectivity of docking sites in MAPK kinases.
J Biol Chem 2009; 284: 1316513173.
44 Cui J, Holgado-Madruga M, Su W, Tsuiki H, Wedegaertner P, Wong AJ. Identication of a specic domain responsible for JNK2alpha2 autophosphorylation.
J Biol Chem 2005; 280: 99139920.
45 Ke H, Harris R, Coloff JL, Jin JY, Leshin B, Miliani de Marval P et al. The c-Jun NH2terminal kinase 2 plays a dominant role in human epidermal neoplasia. Cancer Res
2010; 70: 30803088.
46 Ford J, Ahmed S, Allison S, Jiang M, Milner J. JNK2-dependent regulation of SIRT1
protein stability. Cell Cycle 2008; 7: 30913097.
47 Pruitt K, Zinn RL, Ohm JE, McGarvey KM, Kang SH, Watkins DN et al. Inhibition
of SIRT1 reactivates silenced cancer genes without loss of promoter DNA
hypermethylation. PLoS Genet 2006; 2: e40.
48 Goenka S, Cho SH, Boothby M. Collaborator of Stat6 (CoaSt6)-associated poly
(ADP-ribose) polymerase activity modulates Stat6-dependent gene transcription.
J Biol Chem 2007; 282: 1873218739.
49 Aravind L. The WWE domain: a common interaction module in protein ubiquitination and ADP ribosylation. Trends Biochem Sci 2001; 26: 273275.
50 Yanagawa T, Funasaka T, Tsutsumi S, Hu H, Watanabe H, Raz A. Regulation of
phosphoglucose isomerase/autocrine motility factor activities by the poly(ADPribose) polymerase family-14. Cancer Res 2007; 67: 86828689.
51 Zazzeroni F, Papa S, Algeciras-Schimnich A, Alvarez K, Melis T, Bubici C et al.
Gadd45beta mediates the protective effects of CD40 costimulation against Fasinduced apoptosis. Blood 2003; 102: 32703279.
52 Papa S, Zazzeroni F, Bubici C, Jayawardena S, Alvarez K, Matsuda S et al.
Gadd45beta mediates the NF-kappaB suppression of JNK signalling by targeting
MKK7/JNKK2. Nat Cell Biol 2004; 6: 146153.
53 Blonska M, Pappu BP, Matsumoto R, Li H, Su B, Wang D et al. The CARMA1-Bcl10
signaling complex selectively regulates JNK2 kinase in the T cell receptorsignaling pathway. Immunity 2007; 26: 5566.
54 Naresh KN, Lampert I, Hasserjian R, Lykidis D, Eldereld K, Horncastle D et al.
Optimal processing of bone marrow trephine biopsy: the Hammersmith Protocol.
J Clin Pathol 2006; 59: 903911.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene (2013) 4231 4242

& 2013 Macmillan Publishers Limited