Reduction of Sporadic and Neurofibromatosis Type 2-Associated Vestibular Schwannoma Growth in Vitro and in Vivo After Treatment With The C-Jun N-Terminal Kinase Inhibitor AS602801

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

LABORATORY INVESTIGATION

Reduction of sporadic and neurofibromatosis


type 2–associated vestibular schwannoma growth in vitro
and in vivo after treatment with the c-Jun N-terminal
kinase inhibitor AS602801
Mark C. Dougherty, MD,1 Seiji B. Shibata, MD, PhD,2 J. Jason Clark, MS,2 Franklin J. Canady, MD,2
Charles W. Yates, MD,3 and Marlan R. Hansen, MD2
1
Department of Neurosurgery, University of Iowa, Iowa City; 2Department of Otolaryngology–Head and Neck Surgery, University
of Iowa, Iowa City, Iowa; and 3Department of Otolaryngology, Indiana University School of Medicine, Indianapolis, Indiana

OBJECTIVE  Vestibular schwannomas (VSs) are benign nerve sheath tumors that result from mutation in the tumor sup-
pressor gene NF2, with functional loss of the protein merlin. The authors have previously shown that c-Jun N-terminal
kinase (JNK) is constitutively active in human VS cells and plays a central role in their survival by suppressing accumula-
tion of mitochondrial superoxides, implicating JNK inhibitors as a potential systemic treatment for VS. Thus, the authors
hypothesized that the adenosine 5ʹ-triphosphate–competitive JNK inhibitor AS602801 would demonstrate antitumor
activity in multiple VS models.
METHODS  Treatment with AS602801 was tested in primary human VS cultures, human VS xenografts, and a genetic
mouse model of schwannoma (Postn-Cre;Nf2flox/flox). Primary human VS cell cultures were established from freshly
obtained surgical tumor specimens; treatment group media was enriched with AS602801. VS xenograft tumors were
established in male athymic nude mice from freshly collected human tumor. Four weeks postimplantation, a pretreatment
MRI scan was obtained, followed by 65 days of AS602801 (n = 18) or vehicle control (n = 19) treatment. Posttreatment
MRI scans were used to measure final tumor volume. Tumors were then harvested. Finally, Postn-Cre;Nf2flox/flox mice
were treated with AS602801 (n = 10) or a vehicle (n = 13) for 65 days. Posttreatment auditory brainstem responses were
obtained. Dorsal root ganglia from Postn-Cre;Nf2flox/flox mice were then harvested. In all models, schwannoma identity
was confirmed with anti-S100 staining, cell proliferation was measured with the EdU assay, and cell death was mea-
sured with terminal deoxynucleotidyl transferase–mediated dUTP nick-end labeling staining. All protocols were approved
by the local institutional review board and Institutional Animal Care and Use Committees.
RESULTS  Treatment with AS602801 decreased cell proliferation and increased apoptosis in primary human VS
cultures. The systemic administration of AS602801 in mice with human VS xenografts reduced tumor volume and cell
proliferation. Last, the AS602801-treated Postn-Cre;Nf2flox/flox mice demonstrated decreased cell proliferation in glial cells
in the dorsal root ganglia. However, AS602801 did not significantly delay hearing loss in Postn-Cre;Nf2flox/flox mice up to 3
months posttreatment.
CONCLUSIONS  The data suggest that JNK inhibition with AS602801 suppresses growth of sporadic and neurofibroma-
tosis type 2–associated VSs. As such, AS602801 is a potential systemic therapy for VS and warrants further investigation.
https://thejns.org/doi/abs/10.3171/2022.7.JNS22934
KEYWORDS  vestibular schwannoma; acoustic neuroma; JNK inhibitor; bentamapimod

V
schwannomas (VSs) are benign nerve
estibular bromatosis type 2 (NF2).1 Currently, the treatment for most
sheath tumors that originate from Schwann cells VSs is limited to surgical resection, stereotactic radiation
(SCs) lining cranial nerve VIII. Approximately 95% therapy, or observation with serial imaging.2 There are no
of VSs occur sporadically, whereas the remainder occur FDA-approved drugs for the treatment of VS. Compounds
secondary to the autosomal dominant syndrome neurofi- with promising preclinical findings, such as lapatinib and

ABBREVIATIONS  ABR = auditory brainstem response; DRG = dorsal root ganglia; JNK = c-Jun N-terminal kinase; NF2 = neurofibromatosis type 2; PBS = phosphate-
buffered saline; SC = Schwann cell; TUNEL = terminal deoxynucleotidyl transferase–mediated dUTP nick-end labeling; VS = vestibular schwannoma.
SUBMITTED  April 26, 2022.  ACCEPTED  July 12, 2022.
INCLUDE WHEN CITING  Published online September 9, 2022; DOI: 10.3171/2022.7.JNS22934.

©AANS 2022, except where prohibited by US copyright law J Neurosurg  September 9, 2022 1
Dougherty et al.

everolimus, have shown limited efficacy in clinical trials.3,4 Given previous findings, we hypothesized that
Bevacizumab—an anti–vascular endothelial growth factor AS602801 would increase VS apoptosis, decrease prolif-
(VEGF) antibody used for other neoplasms—has demon- eration, decrease tumor size, and improve auditory out-
strated clinical efficacy in patients with NF2 who have an comes. We therefore explored the impact of AS602801
advanced tumor burden; however, its effect is temporary on VS tumor growth, cell proliferation, and cell death in
and long-term use is typically precluded due to systemic primary human VS cultures, patient-derived VS mouse
side effects.5,6 Thus, there is an urgent need, especially in xenografts, and a genetic schwannoma mouse model with
patients with NF2, for the development of novel therapies conditional Nf2 deletion using Postn-Cre;Nf2flox/flox mice.
that limit VS growth and/or reduce tumor volume.
Inactivation of the NF2 gene is essential for tumori- Methods
genesis in both sporadic and NF2-associated VSs.7–9 NF2
resides on chromosome 22q12 and encodes the protein All animal procedures followed NIH guidelines and
merlin. Merlin shares homology with the ezrin–radixin– were approved by the University of Iowa’s and the Uni-
moesin (ERM) family of membrane–cytoskeleton-linking versity of Indiana’s Institutional Animal Care and Use
proteins. Merlin regulates interactions of transmembrane Committees. All protocols involving human tissues were
and signaling molecules with cytoskeletal actin, thereby reviewed and approved by the local institutional review
affecting cell motility, intercellular attachments, and sub- board.
cellular localization in response to cellular contact inhibi-
tion.10–12 Therefore, merlin plays a central role in contact Primary VS Cultures
inhibition. Depletion of merlin disrupts contact inhibition, Primary VS cultures were prepared as previously de-
thus promoting cell proliferation and tumorigenesis. Sev- scribed.27–29 In brief, following tumor harvest, the tissue
eral lines of evidence suggest that merlin interacts with was minced into 1-mm3 fragments, digested in 0.25% tryp-
several progrowth kinase signaling cascades, leading sin and 0.1% collagenase at 37°C for 30–40 minutes, and
to the suppression of mitogenic activity at both the cell then dissociated by trituration. Cells were resuspended in
membrane and nucleus.13,14 Merlin blocks signaling at the DMEM supplemented with N2 supplement (Sigma), bo-
cell membrane by integrins and receptor tyrosine kinases vine insulin (Sigma; 10 µg/mL), and 10% fetal calf serum.
(RTKs) such as ErB2 and platelet-derived growth factor The suspensions were then plated on cell culture slides
(PDGF) receptor; this in turn regulates multiple down- that had been pretreated with poly-l-ornithine followed by
stream pathways including the Rac/PAK/JNK, mTORC1, laminin (20 µg/mL). Cultures were grown in a humidi-
PI3K/AKT, Wnt/β-catenin, Ras/Raf/MEK/ERK, and fied incubator at 37°C with 6.0% CO2. After 1–2 days the
FAK/Src pathways.15 Furthermore, upstream suppression medium was exchanged for serum-free medium. The cells
of the Hippo pathway by merlin attenuates Yes-associated were subsequently kept in serum-free conditions with reg-
protein 1 (YAP1) function.16 Within the nucleus, merlin ular media exchanges for 7–10 days without passage. The
inhibits proliferation via downregulation of the E3 ubiqui- JNK inhibitor AS602801 (Selleck Chemicals) was added
tin ligase CRL4 (DCAF1).17 to the treatment group cultures and kept at a constant con-
The c-Jun N-terminal kinase (JNK) is an evolutionarily centration for the duration of the experiment. Doses of 2,
conserved member of the mitogen-activated protein kinase 20, and 50 µg/mL AS602801 were tested to establish a
(MAPK) family. The canonical JNK signaling cascade dose-response profile. Control cultures were treated in an
regulates numerous cellular processes, particularly in- identical manner but were not given AS602801.
flammatory responses and cell death.18,19 Previous reports
indicate that merlin suppresses JNK activity. In the case of VS Xenograft Mice
nonneoplastic SCs with intact merlin, nerve injury results VS xenograft models were developed from tumor spec-
in JNK activation via the p75NTR death domain, ultimately imens harvested from 4 separate patients without NF2 as
leading to SC death.20 In contrast, reduced merlin expres- previously described.2 In short, acutely resected VS tissue
sion in human VS cells leads to persistent phosphorylation was placed in Hanks balanced salt solution (Invitrogen)
and activation of JNK. This persistent JNK activity para- on ice and transported to the laboratory. The specimens
doxically enhances VS cell survival in part by suppress- were then separated into 10-mm3 fragments and implant-
ing oxidative stress in the mitochondria, which implicates ed in the interscapular fat pad of male athymic Ncr Nu/
JNK inhibitors as potential therapeutic agents.21–23 Indeed, Nu mice (National Cancer Institute, NIH; n = 37). Mice
it has been demonstrated that JNK inhibitors SP6000125 were anesthetized with a cocktail of ketamine (100 mg/
and I-JIP lead to VS cell mitochondrial superoxide accu- kg; Hospira) and xylazine (10 mg/kg; Phoenix Scientific)
mulation and subsequent cell death in vitro, and increase during the procedure. Grafts were developed for 4 weeks
the sensitivity of human VS to radiation treatment.24 prior to initiating treatment. Animals had unlimited food
The adenosine 5ʹ-triphosphate–competitive JNK inhib- and water access and were housed in constant temperature
itor AS602801 (also called bentamapimod or PGL5001) and humidity conditions with a 12-hour light/dark cycle.
has demonstrated cytotoxicity against cancer stem cells VS xenograft mice were randomly assigned to receive
derived from pancreatic cancer, lung cancer, ovarian can- 60 mg/kg/day of AS602801 (n = 18; treatment group) or
cer, prostate cancer, and glioblastoma.25 Importantly, the vehicle (H2O, n = 19; control group) twice daily for 65
safety of systemic AS602801 administration has already days. Treatments were delivered by oral gavage. Pretreat-
been demonstrated in phase 2 clinical trials for endome- ment MRI scans of the VS xenograft mice were obtained
triosis (NCT01630252, www.ClinicalTrials.gov).26 4 weeks after human VS implantation to determine initial

2 J Neurosurg  September 9, 2022


Dougherty et al.

tumor volume. Posttreatment MRI scans were obtained Sigma-Aldrich) and diluted in blocking buffer for 1 hour at
to determine final tumor volume. Xenograft tumors were 37°C. Following several washes in PBS, a secondary anti-
subsequently harvested and fixed in 4% paraformaldehyde. body (goat anti–rabbit Alexa Fluor 488, 1:800; Invitrogen)
was applied for 1 hour at room temperature. Sections were
Periostin Mouse Strain (Postn-Cre;Nf2flox/flox) mounted with Prolong Gold + DAPI (Life Technologies).
Transgenic mice carrying Cre expression upstream of Images were obtained using Metamorph software (Mo-
Periostin gene (Postn-Cre) were crossed with mice car- lecular Devices, LLC) on a Leica DMIRE 2 microscope
rying loxp sequences flanking exon 2 of the Nf2  gene​ equipped with epifluorescent filters (Leica Microsystems).
(Nf2flox/flox), generating Postn-Cre;Nf2flox/flox mice and their Nuclei were counted using ImageJ. Analogous methods
Cre-negative control littermates. This Postn-Cre;Nf2flox/flox were used to stain for S100 and DAPI in VS cultures.
mouse model has a 100% penetrance of developing schwan-
nomas in their sensory ganglia, including both cranial Cell Proliferation
and dorsal root ganglia (DRG), by the age of 10 months. VS cultures were labeled with EdU (10 µM; Life Tech-
Nf2flox2 and Nf2∆2 bands were detected by polymerase nologies) for 48 hours prior to fixation as previously de-
chain reaction analysis as described in Giovannini et al.30 scribed.27 Briefly, following fixation, VS cultures were
The following primers were used to detect the Postn-Cre treated with 2 N HCl for 15 minutes, immunostained
transgene via polymerase chain reaction: P1 (CAT TTG for DAPI, S100, and EdU. EdU was detected using the
GGC CAG CTA AAC AT) and P2 (CCC GGC AAA Click-iT EdU detection kit (Life Technologies). Cell pro-
ACA GGT AGT TA), as previously described.31 liferation was estimated by dividing the total number
Postn-Cre;Nf2flox/flox mice were randomized to treat- of EdU-positive/S100-positive nuclei by the number of
ment (n = 10) or control (n = 13) groups. Mice then re- DAPI-positive/S100-positive nuclei. Nuclei were counted
ceived oral treatment of 60 mg/kg/day of AS602801 (n = from 4 randomly selected fields at ×20 magnification per
10) or H2O vehicle (n = 13) divided into 2 daily doses for condition.
65 days. Serum levels of AS602801 were measured at 24 Starting 24 hours prior to euthanasia, mice bearing
hours. At the end of the experiment, auditory brainstem VS xenografts or Postn-Cre;Nf2flox/flox mice were injected
responses (ABRs) were obtained as described below. The with 50 mg/kg EdU intraperitoneally 4 times at regular
animals were then euthanized and DRG were harvested intervals (10 µM, Life Technologies) in order to label ac-
and fixed in 4% paraformaldehyde. Because all sensory tively replicating cells. Once the mice were euthanized,
ganglia in this model demonstrate similar SC character- the tumor and 2 pieces of small intestine (EdU-positive
istics and tumorlet formation, we evaluated DRG rather control) were collected. The xenograft tumors and DRG
than vestibular ganglia, because DRG are more numerous were dissected and prepared as described above. EdU in-
and allow for more consistent evaluation of cell activity. corporation into tumor cell nuclei was detected using the
This practice is consistent with previous literature.31,32 Click-iT EdU kit per the manufacturer’s instructions (Life
Technologies).32 Sections were mounted on coverslips
Tumor Volume with Prolong Gold + DAPI (Life Technologies). Stained
Tumor volume was assessed in VS xenografts as de- sections were imaged using a Leica confocal microscope.
scribed previously.2 In brief, T2-weighted images were Cell proliferation was estimated by dividing the total num-
acquired in the axial and sagittal planes using a small- ber of EdU-positive/S100-positive nuclei by the number of
animal MRI unit (Varian Unity/INOVA 4.7-T scanner; DAPI-positive/S100-positive nuclei. A total of 4 randomly
Varian, Inc.). Tumor volume was measured on ImageJ selected microscopic fields per section and 2 sections per
software (NIH) using the ABC/2 method by two indepen- slide were counted. During cell counting, the investigator
dent reviewers who were blinded to the experimental con- was blinded to treatment conditions.
ditions.33 Final dimensions were the average of each in-
dependent measurement. Tumor growth was calculated as Cell Apoptosis
final tumor volume (Vf ) minus initial tumor volume (Vi) Apoptotic cells were detected in VS cultures and in
divided by initial tumor volume (i.e., [Vf − Vi]/Vi), thus VS xenografts by terminal deoxynucleotidyl transferase–
accounting for variability in the initial xenograft volumes. mediated dUTP nick-end labeling (TUNEL) using the In
Situ Cell Death Detection Kit (Roche Diagnostics).2,20,27,34
Immunohistochemistry Tumor cell apoptosis was estimated as the percentage of
Following fixation, the VS xenografts or Postn- TUNEL-positive/S100-positive nuclei with typical con-
Cre;Nf2flox/flox lumbar DRG were harvested, washed in densed morphology identified in 4 randomly selected
phosphate-buffered saline (PBS), cryoprotected in 30% microscopic fields at ×20 magnification, as previously
sucrose, embedded in optimum cutting temperature com- described.23,27 The investigator was blinded to treatment
pound (OCT; Thermo Fischer Scientific), and finally cryo- conditions.
sectioned at 10–15 µm. Frozen sections were treated with
2 N HCl for 30 minutes, permeabilized with 0.8% Triton ABR Measurement
X-100 in PBS for 20 minutes, and then blocked with block- ABR measurements were recorded as described previ-
ing buffer (5% goat serum, 2% bovine serum albumin, and ously in Gehlhausen et al.31 ABRs were first performed on
0.8% Triton X-100) for 30 minutes. Next, the samples were Postn-Cre;Nf2flox/flox mice at 4 months of age as a baseline.
incubated in polyclonal rabbit anti-S100 antibody (1:400; Mice were then randomized to receive either AS602801

J Neurosurg  September 9, 2022 3


Dougherty et al.

FIG. 1. Treatment with AS602801 reduces proliferation of primary hu- FIG. 2. Treatment with AS602801 increases apoptosis in primary human
man VS cultures. Primary VS cultures were grown in the presence or VS cultures. VS cultures were treated with JNK inhibitor AS602801 for
absence of JNK inhibitor AS602801 for 24 hours. A and B: Representa- 24 hours. A and B: Representative images of control culture staining
tive cultures immunolabeled with anti-EdU (red) and anti-S100 (green) with S100 antibody (green). Apoptotic cells were determined by TUNEL
antibodies. Nuclei were identified with DAPI (blue). Arrows indicate EdU- (red) and nuclei were labeled with DAPI (blue). Arrows indicate TUNEL-
positive nuclei determined by overlap of red and blue channels. C: The positive nuclei. C: The percentages of TUNEL-positive cells, S100-posi-
percentages of EdU-positive, S100-positive nuclei to total S100-positive tive nuclei to total S100-positive nuclei were counted by ×20 microscopic
nuclei were counted by ×20 microscopic fields and are expressed in the fields and are expressed here as numbers of TUNEL-positive nuclei per
graph as numbers of EdU-positive nuclei per DAPI-stained nuclei. Treat- DAPI-stained nuclei. Treatment with AS602801 significantly increased
ment with AS602801 significantly reduced proliferation of primary hu- the percentage of TUNEL-positive cells in dose-dependent fashion (p <
man VS cells in a dose-dependent fashion (p < 0.0001, 1-way ANOVA). 0.0001, 1-way ANOVA). Error bars represent SEMs. Bar = 50 µm.
Error bars represent SEMs. Bar = 50 µm.

or vehicle control for 65 days, after which posttreatment 50-µM group) compared to vehicle control cultures (3.29
ABR measurements were obtained. ± 0.43; p < 0.0001, 1-way ANOVA), indicating decreased
cellular proliferation following AS602801 treatment; this
Statistical Analysis reduction occurred in a dose-dependent manner (Fig. 1).
Differences in tumor growth, cell proliferation, and Likewise, there was a significant increase (p < 0.0001,
cell death were analyzed by a Student 2-tailed t-test using 1-way ANOVA) in TUNEL-positive VS cells in cultures
GraphPad Prism software (GraphPad Software). Differ- treated with 20 or 50 µM AS602801 (2.97 ± 0.39, 50-µM
ences in relative tumor growth were analyzed by 1-way group) compared to nontreated cells (1.03 ± 0.33). Thus,
ANOVA followed by a post hoc Kruskal-Wallis test using there is a dose-dependent increase in apoptotic schwan-
SigmaStat (Systat Inc.). noma cells following AS602801 treatment (Fig. 2).

Systemic Administration of AS602801 Reduces Tumor


Results Cell Proliferation In Vivo but Does Not Slow Progressive
JNK Inhibitor Reduces Cell Proliferation and Increases Hearing Loss in Postn-Cre;Nf2flox/flox Mice
Apoptosis in Human VS Culture Building on these in vitro results, we sought to inves-
S100 staining confirmed successful culture of primary tigate whether AS602801 reduces glial cell prolifera-
VS. Schwannoma cells were immunolabeled with S100 tion in the Postn-Cre;Nf2flox/flox genetic mouse model of
and counterstained for EdU (Fig. 1) or TUNEL (Fig. 2). schwannoma.31,32,35 In order to establish standard experi-
There were significantly fewer EdU-positive VS cells in mental dosing protocols, we measured the serum levels
AS602801-treated cultures (1.01 ± 0.15 [mean ± SEM], of AS602801 in Postn-Cre;Nf2flox/flox mice at multiple

4 J Neurosurg  September 9, 2022


Dougherty et al.

FIG. 3. Pharmacokinetics of AS602801 in Postn-Cre;Nf2flox/flox mice.


Serum levels of AS602801 were measured for 24 hours after a single
60-mg/kg treatment. The AS602801 serum levels peak within the first
2 hours, followed by a sharp decline within the next 8 hours. Error bars
represent SDs.

time points after oral administration. This demonstrated


a peak drug level of 1500 ng/mL within the first 3 hours, FIG. 4. Auditory function in Postn-Cre;Nf2flox/flox mice treated with
AS602801 (n = 14) or vehicle control (n = 15). ABR thresholds were
followed by a decrease to 100 ng/mL at 8 hours (Fig. 3). measured at 6 months of age. The mice were administered AS602801
Animals tolerated the diet without difficulty, body weight 30 mg/kg twice daily (60 mg/kg/day) throughout the treatment period.
was tracked daily, and no systemic side effects were noted. ABR threshold levels were widely variable in both groups; although there
We next assessed the functional impact of AS602801 is tendency of reduced average threshold shift in AS602801, this did not
on hearing in Postn-Cre;Nf2flox/flox mice. Mice were ran- reach statistical significance (p = 0.61, Student 2-tailed t-test). Error bars
domized into 2 groups and treated twice a day for 65 represent SEMs. SPL = sound pressure level.
days with 60 mg/kg/day of AS602801 (n = 10) or vehicle
control (H2O; n = 13) by gavage. After the completion of volume of 29.2 ± 7.7 mm3 (mean ± SEM) (Fig. 6C).36,37
treatment, mice were followed for an additional 65 days, This difference was statistically significant (p = 0.025,
at which time ABRs were measured to determine hearing 1-way ANOVA followed by Kruskal-Wallis). Eleven of 19
thresholds (Fig. 4). Control mice had an average hearing VSs in control mice demonstrated tumor growth, whereas
threshold of 51.0 ± 9.7 dB (mean ± SEM), and AS602801- tumor shrinkage was seen in 14 of 18 AS602801-treated
treated mice had an average threshold of 49.2 ± 8.1 dB. mice (Fig. 6D). Thus, although on average the tumors de-
Although an overall trend toward delayed elevation of creased in size with AS602801 treatment, there was some
ABR thresholds was noted, no significant difference was variability in the treatment response. No systemic toxici-
reached (p = 0.61, Student 2-tailed t-test). Thus, AS602801 ties were noted throughout the treatment, including weight
did not affect hearing thresholds in Postn-Cre;Nf2flox/flox loss, decreased oral intake, or signs of distress such as
mice at the time point assessed. No systemic toxicities hunched posture or decreased activity.
were noted throughout the treatment, including weight To determine whether these differences in tumor size
loss, decreased oral intake, loss of fur, or signs of distress were related to altered cell proliferation and/or apoptosis,
such as hunched posture or decreased activity. VS xenografts were also stained with EdU (DNA repli-
Finally, we used the EdU assay to measure glial cell cation marker) and TUNEL (apoptotic nucleus marker).
proliferation in Postn-Cre;Nf2flox/flox mice DRG. We found There was a significant reduction in EdU uptake in VS
that AS602801 significantly decreased EdU uptake in xenografts that were treated with AS602801 (0.60% ±
Postn-Cre;Nf2flox/flox mice DRG, suggesting that AS602801 1.04%, mean ± SEM) relative to those treated with H2O
decreases cellular proliferation (Fig. 5). vehicle control (1.24% ± 1.71%, mean ± SEM; p = 0.02,
Student 2-tailed t-test) (Fig. 7A–E). However, there was
AS602801 Reduces Human VS Xenograft Tumor Size and no significant difference in the percentage of TUNEL-
Cell Proliferation but Does Not Induce Apoptosis positive cells between treatment (0.85% ± 1.07%, mean
Pretreatment and posttreatment MRI scans were used ± SEM) and control (0.94% ± 1.15%, mean ± SEM; p =
to calculate relative tumor growth in patient-derived mouse 0.686) xenograft groups (Fig. 7F). Hence, AS602801 does
xenograft VSs. Relative tumor growth was determined by not appear to significantly increase apoptosis in VS cells.
comparing the percent change in tumor volume from the
initial to final MRI scan (Fig. 6A and B). Control tumors (n
= 19) demonstrated a mean increase of 19.0 ± 10.3 mm3 in
Discussion
tumor volume during the 65-day treatment period, where- JNK Inhibitors as Potential Therapies for VSs
as AS602801-treated tumors had a mean decrease in tumor Three genes (jnk 1–3) encode the JNK proteins, and

J Neurosurg  September 9, 2022 5


Dougherty et al.

FIG. 5. Treatment with AS602801 reduced proliferation of merlin-deficient SCs in Postn-Cre;Nf2flox/flox mice DRG. A–D: EdU uptake
in adult (4-month-old) mice was analyzed with confocal microscopy. Mice were injected with EdU (50 mg/kg intraperitoneally 4
times in a 24-hour period: at 0, 4, 8, and 24 hours), and DRG were dissected within 72 hours. Arrows depict EdU-positive cells.
E: DRG glial cells of Postn-Cre;Nf2flox/flox mice treated with vehicle control (n = 13) demonstrated significantly more EdU-positive
cells than AS602801-treated mice (n = 10; p < 0.001, Student t-test). Bar = 50 µm.

alternative splicing leads to at least 10 isoforms of JNK. We further identified that JNK is persistently activated
Whereas proteins derived from the JNK1 and JNK2 genes (phosphorylated) in primary VS cultures and promotes VS
are ubiquitously expressed in all cells and tissue types, survival, at least in part, by suppressing mitochondrial su-
JNK3 protein is thought to be neuron-specific.38 The JNK peroxides. Furthermore, JNK inhibition led to increased
proteins are activated by multiple stimuli in a wide range oxidative stress and apoptosis in VS cells following irra-
of tissues and regulate diverse extracellular activities, play- diation.23,24 Together, such studies suggested that further
ing a critical role in cellular proliferation, apoptosis, and investigation of JNK inhibition in VS was warranted.
differentiation.39 As such, the JNK proteins are often im- In this study we used in vitro and in vivo models to
plicated in contradictory cellular responses.40 In human VS demonstrate that the JNK inhibitor AS602801 reduces
cells, we previously identified that JNK1 and JNK2 are ex- VS cell proliferation. In primary human VS culture,
pressed in higher levels in normal nerve tissue than in VS, AS602801 significantly reduced cell proliferation and in-
although the significance of this finding remains unclear.23 creased apoptosis (Figs. 1 and 2). We also explored the
6 J Neurosurg  September 9, 2022
Dougherty et al.

FIG. 6. Treatment with AS602801 reduces growth of patient-derived VS xenografts. A and B: Representative MRI scans of
xenografts in axial planes from the beginning and end of experiment, with increased tumor size in a control xenograft (A) and a de-
creased size in an AS602801-treated tumor (B). C: Treatment with AS602801 reduced the growth of the VS xenografts compared
with controls (p = 0.025; error bars represent mean ± SEM). D: Most AS602801-treated tumors (14/18) shrank over the treatment
period, whereas most control tumors (11/19) grew. Each bar on the x-axis represents 1 mouse.

effects of AS602801 in Postn-Cre;Nf2flox/flox mice. These In these mice, SC proliferation forms tumorlets centered
mice develop proliferation of SCs—predominantly in on the cell bodies of the sensory neurons. Systemic admin-
the sensory ganglia including the DRG, geniculate gan- istration of AS602801 reduced proliferation of schwanno-
glia, and vestibular ganglia—mirroring the propensity of ma cells in Postn-Cre;Nf2flox/flox mice. However, AS602801
schwannomas to occur in sensory ganglia in humans.31,32,41 did not preserve auditory function (Figs. 4 and 5). In the
J Neurosurg  September 9, 2022 7
Dougherty et al.

FIG. 7. Treatment with AS602801 reduces proliferation of patient-derived VS xenograft cells. EdU and TUNEL uptake in adult
(3-month-old) xenograft mice was analyzed (6 AS602801-treated mice, 9 control). A–D: EdU was detected in tumor cells (red)
and counterstained with DAPI (blue); arrows indicate EdU-positive nuclei. E: Quantification of EdU demonstrates clear reduction
of proliferation in AS602801-treated mice (p = 0.02, Student 2-tailed t-test). F: In contrast, TUNEL staining did not show increased
apoptotic cells in AS602801 compared to control (p = 0.686). Bar = 50 µm.

DRG of Postn-Cre;Nf2flox/flox mice, AS602801 treatment data are particularly relevant to patients with NF2, who are
reduced cell proliferation. often afflicted with too many tumors to treat with radiation
VS tumor–bearing xenograft mice treated with and/or surgery alone.
AS602801 demonstrated radiographic reduction in tumor
size (Fig. 6). Likewise, in VS xenograft mice, AS602801 Limitations
reduced cell proliferation but did not increase apoptosis
(Figs. 6 and 7). Taken together, these data indicate that JNK Although these data are promising, discrepancies in find-
inhibition may be effective in slowing the growth of VSs. ings between our different model systems and variability in
When combined with the fact that systemic administration tumor growth responses underscore the difficulty of con-
of AS602801 has already been demonstrated in phase 2 trolling schwannoma growth by targeting a single aberrant
clinical trials, these findings implicate JNK inhibition via pathway. This is consistent with previous reports from our
AS602801 as a potential systemic therapy for VS. These group and others.32,42,43 Optimal VS treatment will be likely
8 J Neurosurg  September 9, 2022
Dougherty et al.

to require simultaneously targeting multiple pathways; thus, ing. So far, bevacizumab is the only pharmacotherapy
AS602801 may be most effective as one of several drugs to show efficacy in phase 3 clinical trials, but this use
used together. For example, VS cells can survive in the is off-label, the effect is temporary, and long-term use
presence of the high-affinity p75NTR ligand proNGF, and can have severe side effects.5,6 Here we have shown us-
proNGF rescues VS cells from JNK inhibitor–mediated ing in vitro and in vivo models of VS that the adenosine
cell death via activation of NF-κB. This suggests a paradox- 5ʹ-triphosphate–competitive JNK inhibitor AS602801 re-
ical antiapoptotic role of p75NTR leading to VS growth.20,44 duces cell proliferation and tumor growth. This suggests
Therefore, targeting the p75NTR and JNK pathways together that AS602801 may be developed as a systemic therapy
may simultaneously impair VS growth and spare normal to limit schwannoma growth in patients in whom surgery
SCs. Thus, although our results are encouraging, they also and radiation treatment fail. Such pharmacotherapies are
highlight the limitation of a monotherapy for VSs due to the especially relevant for patients with NF2.
complexity and multifactorial nature of VS tumor growth.
Further research is necessary to investigate multiple molec-
ular targets to optimize therapy against VS growth. Acknowledgments
The development of animal models that accurately We acknowledge the assistance of Linjing Xu and Brian
recapitulate human NF2-related schwannomas has also Mostaert. Financial support was provided to Dr. Hansen by DOD
been challenging. Previous reports have shown that deaf- CDMRP/NFRP NF130072, NIDCD-R01 DC009801, and NIDCD
ness in Postn-Cre;Nf2flox/flox mice first becomes apparent 5T32DC000040.
at 6 months and progresses further at 8 and 10 months.31
Mice in our study were 6–7 months of age at the time of References
ABR testing; hence, it is possible that hearing loss at this   1. Fong B, Barkhoudarian G, Pezeshkian P, Parsa AT, Gopen
age is too mild to detect significant differences between Q, Yang I. The molecular biology and novel treatments of
treatment groups, even when differences are present. The vestibular schwannomas. J Neurosurg. 2011;​115(5):​906-914.
mechanism of hearing loss in this model is incompletely   2. Clark JJ, Provenzano M, Diggelmann HR, Xu N, Hansen SS,
understood, further complicating the interpretation of our Hansen MR. The ErbB inhibitors trastuzumab and erlotinib
finding. Further research is necessary to study the un- inhibit growth of vestibular schwannoma xenografts in nude
derlying mechanism of hearing loss phenotype in Postn- mice:​a preliminary study. Otol Neurotol. 2008;​29(6):​846-853.
  3. Karajannis MA, Legault G, Hagiwara M, et al. Phase II trial
Cre;Nf2flox/flox mice. of lapatinib in adult and pediatric patients with neurofibroma-
In addition, the proliferative and apoptotic biomarkers tosis type 2 and progressive vestibular schwannomas. Neuro
we use to study tumor response provide only indirect evi- Oncol. 2012;​14(9):​1163-1170.
dence of JNK pathway inhibition. Due to the small number   4. Goutagny S, Raymond E, Esposito-Farese M, et al. Phase II
of animals and the size of tissue samples, we were unable study of mTORC1 inhibition by everolimus in neurofibroma-
to directly quantify JNK pathway inhibition in this study. tosis type 2 patients with growing vestibular schwannomas. J
Despite this, we believe that our approach is supported Neurooncol. 2015;​122(2):​313-320.
  5. Alanin MC, Klausen C, Caye-Thomasen P, et al. The effect
by previous publications from our laboratory and others, of bevacizumab on vestibular schwannoma tumour size and
which have shown that JNK is persistently phosphorylated hearing in patients with neurofibromatosis type 2. Eur Arch
(active) in VS primary human tumors and cell cultures; Otorhinolaryngol. 2015;​272(12):​3627-3633.
this activation is driven by lack of merlin and can be sup-   6. Slusarz KM, Merker VL, Muzikansky A, Francis SA, Plotkin
pressed by reexpression of wild-type NF2/merlin.21–24 SR. Long-term toxicity of bevacizumab therapy in neurofi-
Another limitation of our study is our use of a single bromatosis 2 patients. Cancer Chemother Pharmacol. 2014;​
73(6):​1197-1204.
dose and concentration of AS602801 in the mouse models.   7. Seizinger BR, Martuza RL, Gusella JF. Loss of genes on
We did not perform an assay testing different dosages of chromosome 22 in tumorigenesis of human acoustic neu-
oral AS602801 in our in vivo models; rather, the 60 mg/ roma. Nature. 1986;​322(6080):​644-647.
kg/day dosage of AS602801 was determined based on our   8. Trofatter JA, MacCollin MM, Rutter JL, et al. A novel moe-
in vitro data (Fig. 1). Twice daily dosing (30 mg/kg/dose) sin-, ezrin-, radixin-like gene is a candidate for the neurofi-
was selected based on the observed pharmacokinetics of bromatosis 2 tumor suppressor. Cell. 1993;​72(5):​791-800.
AS602801 in Postn-Cre;Nf2flox/flox mice; however, it is pos-   9. Rouleau GA, Merel P, Lutchman M, et al. Alteration in a new
gene encoding a putative membrane-organizing protein causes
sible that 3 doses per day would be more effective than 2. neuro-fibromatosis type 2. Nature. 1993;​363(6429):​515-521.
Thus, further quantitative screening assays for dose titra- 10. Welling DB, Packer MD, Chang LS. Molecular studies of
tion will be necessary. Nonetheless, the administered dose vestibular schwannomas:​a review. Curr Opin Otolaryngol
of AS602801 was well tolerated in our mice, and the safety Head Neck Surg. 2007;​15(5):​341-346.
profile of AS602801 in humans has been validated in a 11. McClatchey AI, Giovannini M. Membrane organization and
phase 2 clinical trial for endometriosis (NCT01630252).26 tumorigenesis—the NF2 tumor suppressor, Merlin. Genes
This suggests that AS602801 may be suitable for long- Dev. 2005;​19(19):​2265-2277.
12. Xiao GH, Chernoff J, Testa JR. NF2:​the wizardry of merlin.
term therapy, which is especially important for patients Genes Chromosomes Cancer. 2003;​38(4):​389-399.
with NF2. 13. Li W, Cooper J, Karajannis MA, Giancotti FG. Merlin:​a
tumour suppressor with functions at the cell cortex and in the
Conclusions nucleus. EMBO Rep. 2012;​13(3):​204-215.
14. Zhou L, Hanemann CO. Merlin, a multi-suppressor from cell
Current treatment for VS consists of resection, stereo- membrane to the nucleus. FEBS Lett. 2012;​586(10):​1403-1408.
tactic radiation therapy, or observation with serial imag- 15. Dougherty MC, Shibata SB, Hansen MR. The biological un-

J Neurosurg  September 9, 2022 9


Dougherty et al.

derpinnings of radiation therapy for vestibular schwannomas:​ microenvironment for schwannoma development. Acta Neu-
review of the literature. Laryngoscope Investig Otolaryngol. ropathol. 2016;​132(2):​289-307.
2021;​6(3):​458-468. 36. Reznitsky M, Petersen MMBS, West N, Stangerup SE, Cayé-
16. Baia GS, Caballero OL, Orr BA, et al. Yes-associated protein Thomasen P. The natural history of vestibular schwannoma
1 is activated and functions as an oncogene in meningiomas. growth-prospective 40-year data from an unselected national
Mol Cancer Res. 2012;​10(7):​904-913. cohort. Neuro Oncol. 2021;​23(5):​827-836.
17. Li W, You L, Cooper J, et al. Merlin/NF2 suppresses tumori- 37. Lees KA, Tombers NM, Link MJ, et al. Natural history of
genesis by inhibiting the E3 ubiquitin ligase CRL4(DCAF1) sporadic vestibular schwannoma:​a volumetric study of tumor
in the nucleus. Cell. 2010;​140(4):​477-490. growth. Otolaryngol Head Neck Surg. 2018;​159(3):​535-542.
18. Shen HM, Liu ZG. JNK signaling pathway is a key modula- 38. Gupta S, Barrett T, Whitmarsh AJ, et al. Selective interaction
tor in cell death mediated by reactive oxygen and nitrogen of JNK protein kinase isoforms with transcription factors.
species. Free Radic Biol Med. 2006;​40(6):​928-939. EMBO J. 1996;​15(11):​2760-2770.
19. Weston CR, Davis RJ. The JNK signal transduction pathway. 39. Davis RJ. Signal transduction by the JNK group of MAP
Curr Opin Cell Biol. 2007;​19(2):​142-149. kinases. Cell. 2000;​103(2):​239-252.
20. Ahmad I, Fernando A, Gurgel R, Jason Clark J, Xu L, Han- 40. Bode AM, Dong Z. The functional contrariety of JNK. Mol
sen MR. Merlin status regulates p75(NTR) expression and Carcinog. 2007;​46(8):​591-598.
apoptotic signaling in Schwann cells following nerve injury. 41. Tryggvason G, Barnett A, Kim J, Soken H, Maley J, Hansen
Neurobiol Dis. 2015;​82:​114-122. MR. Radiographic association of schwannomas with sensory
21. Kaempchen K, Mielke K, Utermark T, Langmesser S, Hane- ganglia. Otol Neurotol. 2012;​33(7):​1276-1282.
mann CO. Upregulation of the Rac1/JNK signaling pathway 42. Sagers JE, Beauchamp RL, Zhang Y, et al. Combination
in primary human schwannoma cells. Hum Mol Genet. 2003;​ therapy with mTOR kinase inhibitor and dasatinib as a novel
12(11):​1211-1221. therapeutic strategy for vestibular schwannoma. Sci Rep.
22. Hilton DA, Ristic N, Hanemann CO. Activation of ERK, 2020;​10(1):​4211.
AKT and JNK signalling pathways in human schwannomas 43. Fuse MA, Plati SK, Burns SS, et al. Combination therapy with
in situ. Histopathology. 2009;​55(6):​744-749. c-Met and Src inhibitors induces caspase-dependent apoptosis
23. Yue WY, Clark JJ, Fernando A, Domann F, Hansen MR. of Merlin-deficient Schwann cells and suppresses growth of
Contribution of persistent C-Jun N-terminal kinase activity schwannoma cells. Mol Cancer Ther. 2017;​16(11):​2387-2398.
to the survival of human vestibular schwannoma cells by 44. Ahmad I, Yue WY, Fernando A, Clark JJ, Woodson EA,
suppression of accumulation of mitochondrial superoxides. Hansen MR. p75NTR is highly expressed in vestibular
Neuro Oncol. 2011;​13(9):​961-973. schwannomas and promotes cell survival by activating
24. Yue WY, Clark JJ, Telisak M, Hansen MR. Inhibition of c- nuclear transcription factor κB. Glia. 2014;​62(10):​1699-1712.
Jun N-terminal kinase activity enhances vestibular schwan-
noma cell sensitivity to gamma irradiation. Neurosurgery.
2013;​73(3):​506-516. Disclosures
25. Li Z, Sun C, Tao S, et al. The JNK inhibitor AS602801 Dr. Hansen has ownership in IotaMotion, Inc.
synergizes with enzalutamide to kill prostate cancer cells in
vitro and in vivo and inhibit androgen receptor expression. Author Contributions
Transl Oncol. 2020;​13(4):​100751.
26. Palmer SS, Altan M, Denis D, et al. Bentamapimod (JNK Conception and design: Hansen, Clark, Canady, Yates. Acquisition
inhibitor AS602801) induces regression of endometriotic le- of data: Clark, Canady, Yates. Analysis and interpretation of data:
sions in animal models. Reprod Sci. 2016;​23(1):​11-23. all authors. Drafting the article: Dougherty, Shibata, Canady.
27. Hansen MR, Clark JJ, Gantz BJ, Goswami PC. Effects of ErbB2 Critically revising the article: Dougherty. Reviewed submit-
signaling on the response of vestibular schwannoma cells to ted version of manuscript: Hansen, Dougherty, Shibata, Yates.
gamma-irradiation. Laryngoscope. 2008;​118(6):​1023-1030. Approved the final version of the manuscript on behalf of all
28. Hansen MR, Roehm PC, Chatterjee P, Green SH. Constitu- authors: Hansen. Administrative/technical/material support:
tive neuregulin-1/ErbB signaling contributes to human ves- Clark, Yates. Study supervision: Hansen, Yates.
tibular schwannoma proliferation. Glia. 2006;​53(6):​593-600.
29. Schularick NM, Clark JJ, Hansen MR. Primary culture of Supplemental Information
human vestibular schwannomas. J Vis Exp. 2014;​(89):​51093. Previous Presentations
30. Giovannini M, Robanus-Maandag E, van der Valk M, et al. A portion of this work has been previously presented at the
Conditional biallelic Nf2 mutation in the mouse promotes Association for Research in Otolaryngology’s annual meeting,
manifestations of human neurofibromatosis type 2. Genes held in Baltimore, MD, January 29, 2017. Presented as “JNK
Dev. 2000;​14(13):​1617-1630. Inhibitor AS602801 (Bentamapimod) Decreases Schwannoma
31. Gehlhausen JR, Park SJ, Hickox AE, et al. A murine model of Tumor Volume and Cell Proliferation.”
neurofibromatosis type 2 that accurately phenocopies human
schwannoma formation. Hum Mol Genet. 2015;​24(1):​1-8. Current Affiliations
32. Kersigo J, Gu L, Xu L, et al. Effects of Neurod1 expression
Dr. Shibata: Rick and Tina Caruso Department of Otolaryn­
on mouse and human schwannoma cells. Laryngoscope.
gology–Head and Neck Surgery, Keck School of Medicine of the
2021;​131(1):​E259-E270.
University of Southern California, Los Angeles, CA.
33. Bathla G, Policeni B, Hansen MR, Berbaum K. Calculat-
Mr. Clark: CI-CG LLC, Louisville, KY.
ing the tumor volumes in vestibular schwannomas:​are the
Dr. Canady: Department of Anesthesiology, Stanford University
ABC/2 and volumetric methods comparable? Otol Neurotol.
School of Medicine, Palo Alto, CA.
2017;​38(6):​889-894.
34. Provenzano MJ, Minner SA, Zander K, et al. p75(NTR) expres-
sion and nuclear localization of p75(NTR) intracellular domain Correspondence
in spiral ganglion Schwann cells following deafness correlate Marlan R. Hansen: University of Iowa, Iowa City, IA.
with cell proliferation. Mol Cell Neurosci. 2011;​47(4):​306-315. marlan-hansen@uiowa.edu.
35. Schulz A, Büttner R, Hagel C, et al. The importance of nerve

10 J Neurosurg  September 9, 2022

You might also like