![Generation of NK92MI/HN3 and NK92MI/HS20 cells
(A) Schematic design of HN3-CAR and HS20-CAR in SFG retroviral vector. The SFG retroviral vector contains the HN3 or HS20 single chain antibody fragment, a human IgG1 CH2CH3 hinge region and CD28 transmembrane region, followed by the intracellular domains of co-stimulatory CD28, 4-1BB, and the intracellular domain of CD3ζ. (B) Generation of NK92MI/HN3 and NK92MI/HS20 cells. 293T cells were transfected with Retro-HN3 or Retro-HS20 for 72 h for CAR retrovirus packaging and transduced into NK92MI cells. (C) Determination of CAR expression by flow cytometry. NK92MI/HN3 and NK92MI/HS20 cells were harvested after 4–5 days and then stained with anti-CD56 and CAR F(ab)2 domain [IgG (H+L)] for flow cytometry.](https://www.researchgate.net/publication/388277243/figure/fig1/AS:11431281304871412@1737574042788/Generation-of-NK92MI-HN3-and-NK92MI-HS20-cells-A-Schematic-design-of-HN3-CAR-and_Q320.jpg)
Access to this full-text is provided by PLOS.
Content available from PLOS One
This content is subject to copyright.
RESEARCH ARTICLE
Exploring Glypican-3 targeted CAR-NK
treatment and potential therapy resistance in
hepatocellular carcinoma
Lei Yang
1
, Kien Pham
1
, Yibo Xi
1
, Qunfeng Wu
2
, Dongfang LiuID
2
, Keith D. Robertson
3
,
Chen LiuID
1
*
1Department of Pathology, Yale School of Medicine, Yale University, New Haven, Connecticut, United
States of America, 2Department of Pathology and Laboratory Medicine, New Jersey Medical School,
Rutgers University, Newark, New Jersey, United States of America, 3Department of Molecular
Pharmacology and Experimental Therapeutics, Mayo Clinic, Rochester, Minnesota, United States of America
*chen.liu@yale.edu
Abstract
Hepatocellular carcinoma (HCC) is the most prevalent form of primary liver cancer and the
second leading cause of cancer-related mortality globally. Despite advancements in current
HCC treatment, it remains a malignancy with poor prognosis. Therefore, developing novel
treatment options for patients with HCC is urgently needed. Chimeric antigen receptor
(CAR)-modified natural killer (NK) cells have demonstrated potent anti-tumor effects, mak-
ing them as a promising immunotherapy strategy for cancer treatment. Glypican-3 (GPC3),
a cell surface oncofetal glycoprotein, is highly expressed in most HCC tissues, but not in nor-
mal tissues, and functions as a key driver of carcinogenesis. Given its high expression level
on the cell surface, GPC3 is considered as an attractive immunotherapy target for HCC. In
this study, two GPC3-specific CAR-NK cells, NK92MI/HN3 and NK92MI/HS20, were estab-
lished using NK92MI cells, a modified IL-2-independent NK cell line. These cell lines were
engineered with third generation GPC3-specific CARs, and their activities were subse-
quently evaluated in the treatment of HCC. We found that NK92MI/HN3 cells, rather than
NK92MI/HS20 cells, exhibited a significant cytotoxicity effect against GPC3
+
HepG2 cells in
vitro and efficiently suppressed tumor growth in a xenograft model using NSG mice. In addi-
tion, irradiated NK92MI/HN3 cells displayed similar anti-tumor efficacy to unirradiated
NK92MI/HN3 cells. Furthermore, we observed that NK92MI/HN3 cells showed higher killing
activity against the GPC3 isoform 2 overexpression cell line (Sk-Hep1-v2) than those with
GPC3 isoform 1 overexpression cell line (Sk-Hep1-v1). This suggest that the presence of
different GPC3 isoforms in HCC may impact the cytotoxicity activity of NK92MI/HN3 cells
and potentially influence therapeutic outcomes. These findings highlight the effective anti-
HCC effects of NK92MI/HN3 cells and reveal the role of GPC3 isoforms in influencing ther-
apy outcomes, suggesting that isoform analysis should be considered to optimize CAR-NK
therapies to improve patient outcomes.
PLOS ONE
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 1 / 19
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Yang L, Pham K, Xi Y, Wu Q, Liu D,
Robertson KD, et al. (2025) Exploring Glypican-3
targeted CAR-NK treatment and potential therapy
resistance in hepatocellular carcinoma. PLoS ONE
20(1): e0317401. https://doi.org/10.1371/journal.
pone.0317401
Editor: Trung Quang Nguyen, Center for Research
and Technology Transfer, VIET NAM
Received: June 11, 2024
Accepted: December 27, 2024
Published: January 22, 2025
Copyright: ©2025 Yang et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.
Funding: The author(s) received no specific
funding for this work.
Competing interests: The authors have declared
that no competing interests exist.
1. Introduction
Liver cancer ranks as the sixth most commonly diagnosed cancer and the fourth leading cause
of cancer-related death worldwide, with continuously increasing incidence and mortality
rates, in recent years [1]. By 2030, the burden of liver cancer is projected to exceed 1 million
cases [2]. Hepatocellular carcinoma (HCC), accounting for 75%-85% of all diagnosed cases, is
the most common type of primary liver cancer [1]. Despite extensive exploration of therapeu-
tic options, HCC remains poor prognosis due to postoperative recurrence and metastasis.
Therefore, developing novel strategies and long-life therapies for the patients with HCC are
still urgently needed.
Natural killer (NK) cells mediate potent cytotoxicity against tumor cells, making them
attractive candidates for effective immunotherapies in the treatment of HCC [3,4]. NK cells
possess the unique ability to recognize target cells via a major histocompatibility complex
(MHC)-independent mechanism, allowing them to directly eliminate tumor cells, without
being sensitized. Like T cells, NK cells can be genetically modified with chimeric antigen
receptors (CARs) that can recognize antigens expressed by tumors. These CAR-engineered
NK cells are also equipped with signaling components that enhance NK cell activity, thereby
increasing their cytotoxicity against tumor cells. In comparison to CAR-T cell-based immuno-
therapy, which is known to have toxic side effects, the use of CAR-NK cells offers a promising
approach to enhance efficacy while mitigating adverse effects such as acute cytokine release
syndrome (CRS), neurotoxicity and graft-versus-host disease (GvHD) [5,6]. Several studies
have demonstrated that CAR engineering of NK cells significantly enhances their cytotoxicity
against various types of cancers [7–11]. For instance, NK92 cells engineered with CARs target-
ing CD19 displayed an increased cytotoxicity activity against B-cell malignancies [12]. Simi-
larly, CAR-NK cells designed to recognize CD20 and Flt3 have demonstrated effective anti-
tumor effects against B-cell tumors [13]. Thus, CAR-NK cell therapy has emerged as a promis-
ing immunotherapy strategy for the treatment of cancers.
Glypican-3 (GPC3), a member of the heparan sulfate proteoglycan family, functions as an
oncofetal glycoprotein that is attached to the cell membrane via glycosylphosphatidylinositol
(GPI) [14]. Several studies have consistently shown that both mRNA and protein levels of
GPC3 are significantly elevated in HCC tissue, but not in healthy adult liver tissues [15]. This
overexpression of GPC3 has been found to be strongly correlated with a poorer prognosis in
individuals diagnosed with HCC [16,17]. Survival analysis of HCC patients with high GPC3
expression has revealed significantly reduced overall survival compared to those with low
GPC3 expression. Moreover, the risk of recurrence after liver resection was found to be
increased up to three-fold in HCC patients with high GPC3 expression, as compared to those
with low GPC3 expression [18]. Given the oncogene functions and high expression level of
GPC3 on the cell surface in HCC, it is considered an attractive target for HCC therapy [19–
27].
The human GPC3 gene is transcribed and alternatively spliced into four distinct mRNA
isoforms [28], with isoform 2 being the most commonly expressed [29]. All GPC3 isoforms
share the same C-terminal subunit, while the N-terminal subunits exhibit slight differences.
Indeed, the alternative splicing of variants leading to the generation of different isoforms has
been implicated in resistance to immunotherapy, as observed in B-cell acute lymphoblastic
leukemia (B-ALL) patients with different CD19 isoforms contributing to CAR-T cell escape
and resistance to CAR-T immunotherapy [30]. Isoforms arise from the combination of differ-
ent exons through RNA alternative splicing, resulting in proteins with diverse biological prop-
erties [31]. The differential isoform expression can alter cellular activities, including cell
proliferation, drug responsiveness and therapy outcomes [32,33]. The existence of various
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 2 / 19
GPC3 isoforms and their potential impact on CAR-NK cytotoxicity may raise questions about
the correlation between alternative mechanisms of GPC3 isoforms and their effects on
CAR-NK immunotherapy. Further investigation of the expression and functional properties
of GPC3 isoforms in the context of CAR-NK immunotherapy may provide valuable insights
into their role in determining the outcomes of immunotherapeutic approaches targeting
GPC3 and other antigens with alternative splicing isoforms, and potentially guide the develop-
ment of more effective CAR-NK therapies for cancer treatment.
In the study, we developed two types of CAR-NK cells, NK92MI/HN3 and NK92MI/HS20,
by genetically engineering NK92MI cells, a highly cytotoxic NK cell lines. We then evaluated
their function and effectiveness against different HCC cells to identify a promising cell-based
therapeutic strategy for HCC and investigate the potential impact of GPC3 isoforms on ther-
apy efficacy.
2. Materials and methods
2.1 Ethics statement
All experimental manipulation of mice was undertaken following the National Institute of
Health Guide for the Care and Use of Laboratory Animals. The protocol was approved by the
Committee on the Ethics of Yale University (Protocol Number: 2023–20315).
2.2 Cell culture
Human HCC cell lines, including HepG2, Huh7, Huh7.5 and Sk-Hep1, as well as the 293T cell
line, were purchased from ATCC (Manassas, VA). HCO2 and LH86 hepatoma cell lines were
established in our laboratory [34,35]. 293T cells and HCC cell line HCO2, HepG2, Huh7,
Huh7.5, Sk-Hep1, LH86, Sk-Hep1-v1, and Sk-Hep1-v2 cells were maintained in Dulbecco’s
minimal essential medium (DMEM) supplemented with 10% (v/v) FBS and antibiotics (100
IU/ml of penicillin and streptomycin) at 37˚C in a 5% CO2 air-humidified incubator. Cells
were routinely passaged with 0.25% trypsin. Human NK92MI, NK92MI/HN3 and NK92MI/
HS20 cells were cultured in α-minimum essential medium supplemented with 0.2 mM inosi-
tol, 0.1mM 2-mercaptoethanol, 0.02 mM folic acid, 12.5% horse serum and 12.5% FBS.
2.3 Construction of retroviral vector and transduction of NK92MI cells
Both HN3 and HS20 antibodies, which target GPC3, have demonstrated significant tumor
growth inhibition in HCC in vitro and in vivo [36–38]. The CAR construct used in this study
contains the HN3 single-domain antibody or HS20 single-chain antibody fragment, human
IgG1 CH2CH3 hinge region, and CD28 transmembrane region, followed by the intracellular
domains of co-stimulatory CD28, 4-1BB, and CD3z(Fig 1A) [39]. For the production of the
GPC3-CAR retrovirus, 293T cells were transfected with a combination of plasmids containing
the GPC3-CAR in SFG backbone, RDF, and PegPam3, as previously described [39]. After 72
hours, the supernatant containing retroviral particles, Retro-HN3 or Retro-HS20, was har-
vested. NK92MI cells were then transduced with Retro-HN3 or Retro-HS20 retrovirus using
RetroNectin (Clontech) coated plates. Two days post-transduction, the cells were transferred
to flask for further expansion. To confirm the surface expression of the CARs, the transduced
NK92MI/HN3 and NK92MI/HS20 cells were harvested 4–5 days after transduction. The cells
were stained with anti-CD56 (clone HCD56) and CAR F(ab)
2
domain [IgG (H+L)] (Jackson
ImmunoResearch Laboratories INC, PA) for flow cytometry analysis. After validation, the
transduced NK92MI/HN3 and NK92MI/HS20 cells were collected, sorted for CAR-positive
populations, and expanded for subsequent experiments.
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 3 / 19
2.4 Analysis of NK92MI and CAR-NK92MI cell phenotype
The NK92MI, NK92MI/HN3 and NK92MI/HS20 cells were collected and stained with the
following antibodies for flow cytometry: anti-F(ab)
2
domain [IgG (H+L)] (Jackson
ImmunoResearch Laboratories INC, PA), PE-conjugated anti-NKG2C Ab (clone S19005E),
AF647-conjugated anti-CD56 Ab (clone HCD56), PECy7-conjugated anti-NKG2A (clone
S19004C), FITC-conjugated NKp46 (clone 9E2), PE-conjugated CD94 Ab (clone DX22),
APC-conjugated CD69 Ab (clone H1.2F3), APC-Cy7 conjugated NKp44 Ab (clone P44-8),
FITC-conjugated CD25 Ab (clone BC96), PE-conjugated NKG2D Ab (clone 1D11), APC-con-
jugated NKp30 Ab (clone P30-15) and APC-Cy7 conjugated PD-1 Ab (29F.1A12).
2.5 Polymerase Chain Reaction (PCR)
The cultured cells were collected for RNA extraction and reverse transcription. cDNA was syn-
thesized from the extracted RNA, and this cDNA was then used as a template for PCR
Fig 1. Generation of NK92MI/HN3 and NK92MI/HS20 cells. (A) Schematic design of HN3-CAR and HS20-CAR in SFG retroviral vector. The
SFG retroviral vector contains the HN3 or HS20 single chain antibody fragment, a human IgG1 CH2CH3 hinge region and CD28 transmembrane
region, followed by the intracellular domains of co-stimulatory CD28, 4-1BB, and the intracellular domain of CD3z. (B) Generation of NK92MI/
HN3 and NK92MI/HS20 cells. 293T cells were transfected with Retro-HN3 or Retro-HS20 for 72 h for CAR retrovirus packaging and transduced
into NK92MI cells. (C) Determination of CAR expression by flow cytometry. NK92MI/HN3 and NK92MI/HS20 cells were harvested after 4–5 days
and then stained with anti-CD56 and CAR F(ab)2 domain [IgG (H+L)] for flow cytometry.
https://doi.org/10.1371/journal.pone.0317401.g001
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 4 / 19
amplification of different GPC3 using the following primers: GPC3F1: 5’-CCGTATAGGGCTA
GACTTACAG-3’; GPC3R1: 5’-CAGCTCATGGAGATTGAACTGG-3’; GPC3F2: 5’-TAGAAACTCC
CGTGCCAG-3’; GPC3R2: 5’-GCCGTAGAGAGACACATCTGTG-3’. Primers were synthesized by
Yale Keck oligo synthesis. The protocol for PCR cycle was as follows: denaturation at 94˚C for
30 seconds, annealing at 56˚C for 30 seconds, and primer extension at 72˚C for 1 minute, for a
total of 30 cycles. A final extension step of 10 minutes at 72˚C was performed. PCR products
were analyzed on a 2% agarose gel stained with SYBR safe and visualized under UV light.
2.6 Construction of GPC3 isoform overexpression lentiviral vector
RNA was extracted from HepG2 cells using RNeasy Mini Kit (QIAGEN, MD) according to
the manufacture’s instructions, followed by reverse transcription with a cDNA synthesis kit
(Thermo Scientific, IL). HepG2 cDNA was used as the template for PCR amplification. The
GPC3 isoform 1 and 2 fragments were amplified by PCR with primers as follows: GPC3-V1-
F1: 5’-TATATTATATTATATTTATGCTAGCGCCACCATGGCCGGGACCGTGCGCACCGCGT-3’;
GPC3-V1-R1: 5’-CTGTAAGTCTAGCCCT-3’; GPC3-V1-F2: 5’-CTTCCTGTGTATAGGGCTAGA
CTTACAG-3’; GPC3-V1-R2: 5’-ACTAGCGCATATGGATCCTCAGTGCACCAGGAAGAAGAAGC
A-3’ and with primers GPC3-V2-F: 5’-TATATTATATTATATTTATGCTAGCGCCACCATGGCC
GGGACCGTGCGCACCGCGT-3’ and GPC3-V2-R: 5’-ACTAGCGCATATGGATCCTCAGTGCACC
AGGAAGAAGAAGCA-3’, respectively. The pLenti-GIII lentiviral vector, containing puromycin
antibiotic selection and green fluorescence protein (GFP) reporter, was used for construction.
The obtained GPC3 variant fragments were cloned into the pLenti-GIII plasmid, which con-
tains the NheI and BamHI restriction enzyme sites. The recombinant pLenti-GIII-hGPC3-v1
and pLenti-GIII-hGPC3-v2 plasmids were used for further recombinant lentivirus particle
generation.
2.7 Lentivirus production and transduction of Sk-Hep1 cells
The 3
rd
generation packaging system from abm company (Applied Biological Materials Inc,
Canada) was utilized for recombinant lentivirus generation. A total of 2 x 10
6
293T cells were
seeded into a 100 mm dish and incubated overnight. Plasmid transfection was performed using
TransIT Lenti reagent (Mirus, WI) according to the manufacturer’s instructions. The 293T cells
were transfected with the vector plasmid pLenti-GIII-hGPC3-v1 or pLenti-GIII-hGPC3-v2,
along with a helper plasmid mixture, and maintained at 37˚C in a 5% CO2 for 48 h. The recom-
binant lentiviruses were then collected from the supernatant, concentrated, and stored at -80˚C
for future use. Subsequently, Sk-Hep1 cells were incubated with medium containing GPC3-v1
or GPC3-v2 lentivirus to establish the Sk-Hep-1 GPC3 isoform 1 (Sk-Hep1-v1) or Sk-Hep-1
GPC3 isoform 2 (Sk-Hep1-v2) cell models, respectively. After 48 hours of incubation, the
medium was aspirated and replaced with fresh medium containing puromycin at 4 μg/ml for
cell selection of Sk-Hep1-v1 and Sk-Hep1-v2. The selective medium containing puromycin was
replaced every 2–3 days for up to 2 weeks. Expression validation was subsequently performed.
2.8 RNA isolation and qRT-PCR analysis
Total RNA isolation was performed on cultured cells using RNeasy Mini Kit (QIAGEN, MD),
followed by reverse transcription with a cDNA synthesis kit (Thermo Scientific, IL). Quantita-
tive real-time polymerase chain reaction (qRT-PCR) was then performed using a two-step
SYBR green qPCR assay and the target genes were amplified using following primers. Human
GPC3, forward: 5’-CATTGGAGGCTCTGGTGATGGA-3’; reverse: 5’-TTGTCCTTCGGAGTTGCC
TGCT-3’; Human GAPDH, forward: 5’-GTCTCCTCTGACTTCAACAGCG-3’; reverse: 5’-ACCA
CCCTGTTGCTGTAGCCAA-3’. The qRT-PCR data were acquired using the Step One real-time
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 5 / 19
PCR system (Applied Biosystem, CA). The cycling procedure was as follows: one cycle at 95˚C
for 30 seconds, followed by 40 cycles at 95˚C for 5 seconds and 64˚C for 31 seconds. Each
assay plate included negative control with no template. The mRNA levels of the gene of inter-
est were normalized to the mRNA levels of GAPDH and analyzed using the 2
−ΔΔCt
method.
2.9 Western blot analysis
Samples from the cultured cells were lysed in ice-cold RIPA buffer (Sigma, MA) containing
phosphatase/protease inhibitors (Thermo Scientific, IL). The lysates were then centrifuged at
11,000 g for 10 min at 4˚C, and the supernatant was collected and quantified using a BCA pro-
tein assay kit (Thermo Scientific, IL). After quantification, cell lysates were loaded and sepa-
rated by electrophoresis with a 12% sodium dodecyl sulfate-polyacrylamide gel. The protein
was then transferred onto a polyvinylidene fluoride membranes (Millipore, MA), which were
blocked for 1 h at room temperature in PBS and probed with anti-human GPC3 and GAPDH
monoclonal antibodies at 4˚C overnight. After washing with PBS containing 0.05% Tween 20
three times at 10-min intervals, the membrane was incubated with goat anti-rabbit IgG or goat
anti-mouse IgG secondary antibodies for 1 h at room temperature. After further wash steps,
the membranes were treated with the substrate (Thermo Scientific, IL) to visualize the protein
bands using the ChemiDoc Imaging System (BioRad, CA).
2.10 Cytotoxicity effects of CAR-NK92MI cells to HCC cells
HepG2 cells at a density of 2 x 10
5
were co-cultured with NK92MI, NK92MI/HN3 and
NK92MI/HS20 at E:T ratios of 2.5:1, 5:1 and 10:1 for 24 h and 48 h in vitro. GFP florescence
signals were observed under a microscope, which served as an indicator to evaluate the killing
efficacy of CAR-NK92MI cells. Similarly, Sk-Hep1, Sk-Hep1-v1 and Sk-Hep1-v2 cells at a den-
sity of 2 x 10
5
were cultured in a 24-well plate overnight. NK92MI/HN3 cells were added and
incubated with the target cells with an E:T ratio of 5:1 for 24 h and 48 h in vitro. Florescence
signals were observed under a microscope.
2.11 Functional assay of CAR-NK92MI cells
CAR-NK92MI cells were co-cultured with target cells at an E:T ratio of 1:1 for 24h. After incu-
bation, the supernatant from the co-culture system was collected and the production of IFN-γ
was detected using an enzyme-linked immunosorbent assay kit (R&D Systems, USA) accord-
ing to the manufacturer’s instructions. Similarly, the CAR-NK92MI cells were co-cultured
with target cells at an E:T ratio of 1:1 for 24h in 96-well V bottom plates. The cells were col-
lected and washed with PBS, and then stained with anti-CD107 for flow cytometry analysis.
2.12 Xenogeneic tumor-grafted mouse models
Six to eight-week-old NOD-scid IL2Rgnull (NSG) mice were purchased from the Jackson Lab-
oratory. The NSG mice were randomly divided into four groups. A total of HepG2 cells (1 x
10
6
) were suspended in 50% Matrigel and subcutaneously injected into the right flack of each
mouse. When the tumor size reached approximately 100mm
3
, the mice were intravenously (i.
v.) injected with NK92MI, NK92MI/HN3, NK92MI/HS20 cells or PBS twice at seven-day
intervals. Tumor sizes were monitored and calculated every two days thereafter using the for-
mula V = length x width
2
/2.
The animals will be euthanized using CO
2
asphyxiation if the tumor size reaches 1.5 cm in
diameter or if tumor ulceration occurs. The tumor volume will not exceed 1 cm
3
. In addition
to standard endpoint parameters, animals will also be monitored for other clinical signs,
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 6 / 19
including, but not limited to, constipation, weight loss and hunching. Animals showing signs
of significant suffering were euthanized before the study’s endpoint to prevent unnecessary
pain.
2.13 Effects of irradiation on cytotoxicity of NK92MI/HN3 cells in vitro
NK92MI/HN3 were irradiated with 5 Gy using an X-ray irradiator. HepG2 cells were co-cul-
tured with irradiated and unirradiated NK cells at various ratios of E:T at 1:1, 2.5:1, 5:1 and
10:1. Florescence signals were observed under the microscope to assess the cytotoxicity of
CAR-NK92MI cells towards HepG2 cells.
2.14 CFSE cell proliferation assay
1 x 10
6
irradiated and unirradiated NK92MI/HN3 cells were labeled with CFSE at a final work-
ing concentration of 10 μM, according to manufacturer’s instructions (CellTrace
TM
CFSE Cell
Proliferation Kit, Thermo Fisher), and incubated for 20 min at 37˚C. The staining was washed
using 5 volumes of culture medium. The stained cells were maintained at 37˚C in a 5% CO2
for 120 h to analyze the proliferation.
2.15 Effects of irradiation on cytotoxicity of NK92MI/HN3 cells in vivo
Six-week-old mice were randomly divided into three groups. A total of 1 x 10
5
HepG2 cells
were suspended in 50% Matrigel and subcutaneously injected into the right flack of each
mouse. When tumor size reached approximately 100mm
3
, the mice were intravenously (i.v.)
injected with irradiated or unirradiated NK92MI/HN3 cells, or PBS, at the indicated time
points. Tumor sizes were then monitored and calculated every two days.
2.16 Statistics analysis
Data are shown as mean ±SD. All calculations and statistical analyses were performed using
Graphpad PRISM 5.0 (GraphPad Software, San Diego, CA) for Mac. Comparisons between
groups were conducted using analyses of unpaired t-tests, and P<0.05 was considered as statis-
tically significant.
3. Results
3.1 Generation of NK92MI/HN3 and NK92MI/HS20 cells
Given its oncogenic function, GPC3 has already been suggested as a potential target for cancer
immunotherapy in the treatment of HCC, including CAR-T and CAR-NK strategies [40].
These immunotherapeutic approaches aim to specifically target and eliminate GPC3-expres-
sing tumor cells while minimizing off-tumor effects. Considering the splendid advantages of
easy expansion, cultivation and activation, NK92MI cell line with indefinite expansion capac-
ity have been used in clinical practice [41]. To develop an NK cell-based immunotherapy for
HCC patients, we genetically engineered NK92MI cells with the third-generation CAR mole-
cules, HN3 or HS20, which specifically target GPC3, a highly expressed antigen in HCC. The
human single-domain antibody, HN3, recognized a conformational epitope that requires both
the amino and carboxy terminal domains of GPC3 [37], and the HS20 recognizes the heparan
sulfate chains on GPC3 [36]. As shown in Fig 1A, we cloned the HN3 single-domain antibody
or HS20 single-chain antibody fragment into an SFG retroviral vector. The CAR construct
contains human IgG1 hinge CH2-CH3 domain, CD28 transmembrane (TM) domain and
intracellular domain, 4-1BB ligand intracellular domain, and CD3zeta intracellular domain.
After construction, the 293T cells were transfected with a combination of plasmids containing
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 7 / 19
HN3-CAR or HS20-CAR in the SFG backbone, RDF, and PegPam3, as previously described
[42]. The SFG retrovirus particles in the supernatant were utilized to transduce NK92MI cells
with the GPC3-specific CAR construct to generate NK92MI/HN3 or NK92MI/HS20 cells (Fig
1B). After 4–5 days, the cells were collected and labeled with CD56 and human IgG (H+L) for
sorting to enrich the CAR-NK92MI cell population. Flow cytometry analysis demonstrated
that more than 96% of CD56+CAR+ NK92MI cells were observed (Fig 1C).
3.2 Phenotypic characterization of NK92MI/HN3 and NK92MI/HS20 cells
To assess the immunophenotyping of NK92MI, NK92MI/HN3 and NK92MI/HS20 cells, flow
cytometry was performed to characterize the expression of several important activating and
inhibitory receptors and markers on the surface of the cells, including CD56, CD25, CD69,
PD-1, NKp30, NKp44, NKp46, NKG2A, NKG2C, NKG2D and CD94 (Fig 2). The results
revealed that the expression profile of these receptors and markers did not show significant dif-
ferences between NK92MI, NK92MI/HN3 and NK92MI/HS20 cells.
3.3 Cytotoxicity activity of NK92MI/HN3 and NK92MI/HS20 cells against
HepG2 cells in vitro and in vivo
To determine the recognition of CAR-NK cells to the GPC3+ HCC cells, we first evaluated the
expression of GPC3 in several HCC cell lines, including HCO2, HepG2, Huh7, Huh7.5, Sk-
Fig 2. Phenotypic characterization of NK92MI/HN3 and NK92MI/HS20 cells. Representative histograms of the
expression of CD56, CD25, CD69, PD-1, NKp30, NKp44, NKp46, NKG2A, NKG2C, NKG2D and CD94 on the cell
surface of NK92MI, NK92MI/HN3 and NK92MI/HS20 cells. The expression of thesereceptors was determined by
flow cytometry.
https://doi.org/10.1371/journal.pone.0317401.g002
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 8 / 19
Hep1, LH86 and Hep3B cell lines. As shown in Fig 3A, we observed higher mRNA levels of
GPC3 in HepG2, Huh7, Huh7.5 and Hep3B cells, and lower levels in HCO2 and LH86 cells.
On the other hand, the protein expression of GPC3 was significantly higher in HepG2 cells
(Fig 3B), while Huh7, Huh7.5, and Hep3B cells showed relatively lower GPC3 expression lev-
els. Consistent with previous data from other groups, Sk-Hep1 cells did not exhibit detectable
GPC3 protein expression [43]. Thus, HepG2 cells were chosen as the cell model for subsequent
studies.
To assess the anti-tumor efficacy of different CAR-NK92MI cells, we co-cultured the
HepG2 cells with NK92MI, NK92MI/HN3 and NK92MI/HS20 cells and observed the
Fig 3. Cytotoxicity activity of NK92MI/HN3 and NK92MI/HS20 cells against HepG2 cells in vitro and in vivo.(A) mRNA level and (B) protein expression
levels of GPC3 in different HCC cell lines. (C) Cytotoxicity activity of NK92MI, NK92MI/HN3 or NK92MI/HS20 against HepG2 cells in vitro. GFP signal was
observed under the florescence microscope. (D) The IFN-γproduction of the co-cultured NK cells with HepG2 cells in vitro. IFN-γconcentration was
measured in the supernatant. (E) Cytotoxicity activity of NK92MI/HN3 against HepG2 cells at an E:T ratio of 1:1, 2.5:1, 5:1 and 10:1 for different time points.
(F) Procedure and growth curve of HepG2 Xenografts treated with NK92MI, or NK92MI/HN3 or NK92MI/HS20 cells. The mRNA levels were normalized
with the mRNA levels of GAPDH. n = 5 mice per group. Data are represented as mean ±SD. *, P <0.05; **, P<0.01; ***, P<0.001.
https://doi.org/10.1371/journal.pone.0317401.g003
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 9 / 19
florescence signal under a fluorescence microscope. Compared to the NK92MI and NK92MI/
HS20 cells, co-culturing with NK92MI/HN3 cells displayed less GFP fluorescence signals, indi-
cating a high anti-tumor effect against HepG2 cells (Fig 3C). In addition, we found that CAR-
mediated killing activity was associated with the elevated levels of IFN-γproduction, as mea-
sured via ELISA, in the supernatant of HepG2 cells co-cultured with NK92MI/HN3 cells (Fig
3D). To further validate the cytotoxicity of NK92MI/HN3 cells, we co-cultured the NK92MI/
HN3 cells with HepG2 cells by utilizing different effector-to-target (E:T) cell ratios ranging
from 1:1 to 10:1. As shown in Fig 3E, the NK92MI/HN3 cells effectively eliminated the tumor
cells at different ratios in vitro. To explore the in vivo anti-tumor activity of NK92MI/HN3 and
NK92MI/HS20 cells, we established a xenograft model using immunodeficient (NOD/SCID)
mice bearing subcutaneous HepG2 cells. Approximately 2 weeks after tumor cell inoculation,
mice were grouped and treated with PBS, NK92MI, NK92MI/HN3 or NK92MI/HS20 cells.
The results showed that administration of NK92MI/HN3 effectively inhibited the growth of
the HepG2 xenografts (Fig 3F), indicating the effective anti-tumor activity of NK92MI/HN3
cells.
3.4 Impact of irradiation on anti-tumor efficacy of NK92MI/HN3 In Vitro
and in vivo
To ensure the safety and minimize the risk of NK lymphoma development in patients, NK92
cells are commonly irradiated prior to clinical application. Studies have shown that a radiation
dose of 5Gy is sufficient to limit the lifespan of CAR-NK92 cells while maintaining their effec-
tiveness [44]. Accordingly, we assessed the specific cytotoxicity of both irradiated and unirra-
diated CAR-NK against HepG2 cells in vitro and in vivo. It was observed that both irradiated
and non-irradiated NK92MI/HN3 cells displayed cytotoxicity effects against HepG2 cells at
varying effector-to-target ratios in vitro (Fig 4A). To evaluate the impact of 5Gy irradiation on
the longevity and proliferation of NK92MI/HN3, we performed a CFSE-based proliferation
assay. Flow cytometry analysis showed that while irradiated NK92MI/HN3 cells remained via-
ble for 5 days post-irradiation, they were unable to proliferate (Fig 4B). Furthermore, both
irradiated and non-irradiated NK92MI/HN3 cells effectively inhibited tumor growth in
HepG2 xenograft models (Fig 4C). These results suggest that irradiation does not compromise
the anti-tumor efficacy of NK92MI/HN3 cells, as the irradiated NK92MI/HN3 cells displayed
similar anti-GPC3 malignancy activity to unirradiated NK92MI/HN3 cells. To further opti-
mize the treatment procedure, we administered irradiated NK92MI/HN3 cells to HepG2
xenografts every 4 days, for a total five-time injections. It was observed that this treatment regi-
men significantly inhibited the tumor growth (Fig 4D). Thus, the use of irradiated NK92MI/
HN3 cells as a safe and effective option for CAR-NK immunotherapy in the treatment of
HCC, preserving both safety and therapeutic efficacy.
3.5 Identification and analysis of GPC3 isoforms in HCC
To analyze the expression of GPC3 and its variants in malignancies, we performed expression
analysis with ISOexpresso, a web-based platform for isoform-level expression analysis in can-
cer [45]. The results showed that the mRNA level of GPC3 was markedly increased in HCC
and is also detectable in several other cancers, suggesting a broad-spectrum activity of GPC3
in tumor development. In addition, the subtype expression profile data showed that GPC3 var-
iants 1 and 2 are present in a variety of human malignancies, while GPC3 variants 3 and 4 are
not (Fig 5A). This discrepancy presence may contribute to the differential regulation of GPC3
variants during tumor growth. Moreover, unlike GPC3 variant 2, variant 1 was only found in
HCC, implying that GPC3 variant 1 presence may exert distinct biological activities in the
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 10 / 19
development of HCC. In contrast, the expression level of GPC3 variant 2 showed much higher
than that of GPC3 variant 1 in HCC, which is considered to be the dominant form of existence
[29].
It was reported that there exist different kinds of GPC3 variants in HCC (Fig 5A), whereas
no obvious evidence to show the impact of different GPC3 isoforms on cytotoxicity of
CAR-NK cells. To explore the potential impact of GPC3 isoforms on anti-tumor efficacies, we
first compared the sequence information of four GPC3 variants (NM_001164617.2,
NM_004484.4, NM_001164618.2, and NM_001164619.2), which have been identified in Gen-
Bank and encode alternatively spliced forms. Through the sequence comparison, it was found
that the main difference among the GPC3 isoforms was the deletion of part or entire sequence
of exon 2 and exon 4 (Fig 5B). This exon loss phenomena could potentially result in different
Fig 4. Impact of irradiation on anti-tumor efficacy of NK92MI/HN3 in vitro and in vivo.(A) Cytotoxicity activity of irradiated and unirradiated NK92MI/
HN3 cells against HepG2 cells at an E:T ratio of 1:1, 2.5:1, 5:1 and 10:1. (B) CFSE-based proliferation assay of irradiated and unirradiated NK92MI/HN3 cells.
(C) Procedure and growth curve of HepG2 Xenografts treated with PBS, irradiated NK92MI/HN3 or unirradiated NJ92MI/HN3 cells. (D) Treatment
procedure optimizing of irradiated NK92MI/HN3 cells. n = 5 mice per group. Data are represented as mean ±SD. *, P <0.05; **, P<0.01; ***, P<0.001.
https://doi.org/10.1371/journal.pone.0317401.g004
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 11 / 19
biological functions of GPC3 in the development of HCC. Based on sequence information, we
designed two primer sets, GPC3F1/R1 and GPC3F2/R2, to examine the existence and distribu-
tion of GPC3 variants in HCC cell lines. The GPC3 variant 1 was detected with the GPC3F1
and GPC3R1 primer set at the size of 348 bp. The total GPC3 variants 1 and 2, variant 3, and
variant 4 were detected at the size of 311bp, 286bp and 145bp, respectively, using the GPC3F2
and GPC3R2 primer set (Fig 5C). The results showed that GPC3 variants 1 and 2 were
detected in HepG2, Huh7, Huh7.5, and Hep3B cells, only GPC3 variant 2, but not variant 1,
was present in HCO2 and LH86 cells (Fig 5D). There were no GPC3 variant expression
detected in Sk-Hep1 cells. Interestingly, we found that NK92MI/HN3 cell showed different
killing activities to different HCC cells (Fig 5E), indicating that GPC3 isoform existence may
have potential impact on anti-tumor efficiencies.
3.6 Impact of GPC3 isoforms on the cytotoxicity activity of NK92MI/HN3
cells
Since GPC3 isoforms are not detectable in Sk-Hep1 cells, it would be a good model to generate
two stable cell line Sk-Hep1-v1 and Sk-Hep1-v2 that were overexpressing with GPC3 variants
1 and GPC3 variant 2, respectively, to elucidate different biological activities on regulating
HCC tumorigenesis and potential anti-tumor efficacies of CAR-NK92MI cells. To assess the
expression levels of GPC3, we collected both protein and mRNA from Sk-Hep1, Sk-Hep1-v1,
and Sk-Hep1-v2 cells for western blot and qPCR, respectively. Notably, we found that the
Fig 5. Identification and analysis of GPC3 isoforms in HCC. (A) Analysis of GPC3 and its variants in cancer. A web-based platform, ISOexpresso, for
isoform-level expression analysis in human malignancies. (B) Sequence comparison between different GPC3 variants. (C) Designing specific primers for
recognizing specific GPC3 variants. Band sizes of the PCR products representing different GPC3 variants are shown in the graph, using GPC3 primer sets 1
and 2. (D) Distribution and composition of GPC3 variants in different HCC cell lines. (E) Different killing activities of NK92MI/HN3 cells to various HCC
cells. The mRNA levels were normalized with the mRNA levels of GAPDH. Data are represented as mean ±SD. *, P <0.05; **, P<0.01; ***, P<0.001.
https://doi.org/10.1371/journal.pone.0317401.g005
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 12 / 19
GPC3 protein expression level of Sk-Hep1-v1 cells was significantly higher than in Sk-
Hep1-v2 cells, despite mRNA levels were comparable in both cell lines (Fig 6A). To further
investigate the cellular localization of GPC3 expression, its distribution on the cell membrane
and in the cytosol was assessed. The results revealed that GPC3 variant 1 exhibited a higher
level of expression on the cell membrane compared to GPC3 variant 2 (Fig 6B). These findings
suggest that the existence of different GPC3 isoforms may impact the properties of the protein
and cellular activities and could potentially influence the CAR-NK therapy efficacy.
To evaluate the killing efficacy of NK92MI/HN3 cells in relation to different GPC3 iso-
forms, we co-cultured NK92MI/HN3 with Sk-Hep1, Sk-Hep1-v1 and Sk-Hep1-v2 cells in vitro
at an E/T ratio of 5:1. Fluorescence signal of the HCC cells was observed under the microscope,
which was served as an indicator to evaluate the killing efficacy of NK92MI/HN3 cells. It was
observed that NK92MI/HN3 cells exhibit high cytotoxic effects against Sk-Hep1-v2 than Sk-
Hep1 and SkHep1-v1 cells (Fig 6C), followed by an increased IFN-γproduction in the super-
natant (Fig 6D), indicating that the presence of different isoforms may have an impact on ther-
apy outcomes. To further validate the activity of NK92MI/HN3 cells, the expression of
CD107a, a sensitive marker for NK cell functional activity, was analyzed by flow cytometry.
Co-culturing NK92MI/HN3 cells with Sk-Hep1-v2 cells led to a significant upregulation of
Fig 6. Cytotoxicity activity of NK92MI/HN3 cells to Sk-Hep1-v1 and Sk-Hep1-v2 cells. (A) Establishment of GPC3 variant overexpression cell model.
Construction of the lentiviral vector containing GPC3 variant 1 and variant 2, respectively. The cells were checked for fluorescencesignal under a microscope.
The expression validation of GPC3 expression levels. (B) The assessment of GPC3 expression on the membrane and in the cytosolic of Sk-Hep1, Sk-Hep1-v1
and Sk-Hep1-v2 cells. (C) NK92MI/HN3 cells were co-cultured with Sk-Hep1, Sk-Hep1-v1 or Sk-Hep1-v2 cells at a E:T ratio of 5:1 for 12h, 24h and 48h.
Cytotoxicity efficacies of NK92MI/HN3 cells were observed under the florescence microscope. (D) IFN-γproduction was measured via ELISA in the
supernatant. (E) CD107a expression levels were assessed by flow cytometry. The mRNA levels were normalized with the mRNA levels of GAPDH. Data are
represented as mean ±SD. *, P <0.05; **, P<0.01; ***, P<0.001.
https://doi.org/10.1371/journal.pone.0317401.g006
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 13 / 19
CD107a expression in NK92MI/HN3 cells (Fig 6E), indicating the enhanced NK cell cytotox-
icity against Sk-Hep1-v2 cells, but not against Sk-Hep1-v1 cells. Thus, further analysis and
investigation may be required to better understand the implications of GPC3 isoform 1 in the
clinical outcome of HCC treatment with NK92MI/HN3 cells.
4. Discussion
HCC is a highly aggressive form of cancer, and curative treatment options are typically limited
to patients. Therefore, novel strategies for the patients with HCC are urgently needed. One
promising approach involves the use of chimeric antigen receptor (CAR) technology to har-
ness cell-based immunity. In the study, we constructed two CAR-NK cell lines, NK92MI/HN3
and NK92MI/HS20 cells, using GPC3-targeting HN3 single-domain antibody or HS20 single-
chain antibody fragment, respectively. Compared to NK92MI/HS20 cells, NK92MI/HN3 cells
exhibited a higher cytotoxicity against HepG2 cells (Fig 3). This discrepancy in cytotoxicity
may be due to differences in antibody recognition. The HN3 antibody targets a conformational
epitope on the GPC3 core protein, which requires both the amino and carboxy terminal
domains [37]. In contrast, the HS20 antibody recognizes the heparan sulfate chains on GPC3
[36]. As a result, HN3 is more potent because it binds directly to the GPC3 core protein, con-
tributing to its enhanced cytotoxicity. However, despite the higher cytotoxicity observed in
vitro, the in vivo tumor growth inhibition by NK92MI/HN3 cells was less significant. This dis-
crepancy may be attributed to challenges in persistence and trafficking of NK92MI/HN3 cells
to tumor sites, both of which are crucial for the in vivo efficacy of CAR-NK cell therapies.
Additionally, the complex tumor microenvironment (TME) in vivo, which is absent in in vitro
conditions, could play a role in modulating NK cell activity. Immunosuppressive factors and
cell-cell interactions in the TME may diminish the efficacy of NK92MI/HN3 cells in vivo,
reducing their ability to effectively target and kill tumor cells. To overcome these challenges,
further optimization of the dosing schedule and treatment regimen may be necessary to
enhance the persistence and trafficking of CAR-NK cells. Potential strategies include repeated
dosing, combining CAR-NK therapy with TME-modulating agents, or implementing genetic
modifications to enhance CAR-NK cell survival and activity, which could improve the thera-
peutic outcomes in HCC.
Chimeric antigen receptor (CAR) T cell therapy is recognized as a promising immunother-
apeutic strategy, especially in the treatment of relapsed and refractory B-cell malignancies [46].
However, CAR-T cell therapy faces several challenges. One of the major obstacle is the need to
collect and utilize autologous cells, which requires labor-intensive processes, including isola-
tion and in vitro expansion [47,48]. This concern may limit the broader clinical application of
CAR-T-cell therapy, leading to increased interest in exploring alternative CAR platforms [49].
The modification of NK cells with CARs is being investigated as an alternative to T cells in var-
ious therapeutic areas. One advantage of NK cells over T cells is their non-major histocompati-
bility complex (MHC) restriction, which enables them to activate or inhibit target cells
through germline-encoded activating or inhibitory receptors that interact with specific ligands
on the target cells [50]. This characteristic makes NK cells suitable as an “off-the-shelf” product
for patients in a cost-effective manner, as allogenic NK cells have shown a reduced risk of allor-
eactivity. Allogeneic NK cells can contribute to a graft-versus-tumor effect without causing
GvHD, as shown in mouse models [51] and clinical studies [52]. Various sources of NK cells
are being used to generate CAR-NK cells, including primary NK cells from cord blood, iPSC-
derived NK cells, and NK cell lines [44]. In this study, we utilized the NK92MI/HN3 cell line
to develop NK92MI/HN3 cells. NK92MI cells demonstrated sustained activity and prolifera-
tion rates after flow cytometry sorting or recovery from liquid nitrogen cryopreservation. As a
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 14 / 19
result, the NK92MI cells are more amenable to stable in vitro expansion and exhibit higher
CAR transduction efficiency, making them a convenient source of NK cells for the production
of CAR-NK cells. However, it is important to note that NK92MI cells, a modified IL-2-inde-
pendent NK92 cell line derived from an NK lymphoma patient, is indeed a human cancer cell
line. This presents potential risks related to uncontrolled growth and the possible formation of
secondary malignancies. Although NK92MI cells have shown promise in clinical studies, their
safety in clinical settings requires careful evaluation and further assessments. To mitigate the
risk of secondary NK lymphoma development in patients, large doses of radiation are applied
to NK92MI cells prior to clinical use. While this approach ensures safety, it also reduces the
cytotoxicity of NK92 cells. Studies have identified a radiation dose of five Gy as optimal for bal-
ancing the safety and effectiveness of the CAR-NK92MI cells [44].
GPC3 plays a crucial role in HCC carcinogenesis, and human GPC3 gene can be alterna-
tively spliced into four variants, however, the biological significance of these variants in HCC
and therapies remains largely unknown. In the study, various HCC cell lines were analyzed for
the distribution of GPC3 variants. Consistent with the database, GPC3 variant 2 was found in
a broader range of cells, suggesting it is more commonly expressed [29]. Studies have demon-
strated that alternative splicing generates multiple protein isoforms with distinct biological
properties, such as protein interaction, subcellular localization, and catalytic ability [31]. These
differences in isoform expression can lead to alterations in cellular behavior, influencing pro-
cesses like cell proliferation and therapy responsiveness [32]. Several studies have demon-
strated that the abnormal regulation of alternative splicing contributes to tumor occurrence
and development in human malignancies [53–56]. In addition to the role of alternative splic-
ing in tumorigenesis, that alternative splicing variants are implicated in resistance to immuno-
therapy [30,55,57]. In the study, we observed that NK92MI/HN3 exhibited varying cytotoxic
efficacies against Sk-Hep1-v1 and Sk-Hep1-v2 cells (Fig 6). This was accompanied by
increased IFN-γproduction and CD107a expression. These discrepancies suggest that the exis-
tence of GPC3 alternative splicing variants may increase isoform diversity, potentially impact-
ing therapy outcomes. Understanding the diversity of GPC3 isoforms and their influence on
treatment efficacy is critical. The differential cytotoxicity observed between HCC cells express-
ing GPC3 isoform 1 and isoform 2 underscores the importance of considering GPC3 isoform
expression patterns in patient populations. Validating GPC3 isoform profiles could help
understand the potential mechanism of escape and improve the precision of CAR-NK therapy,
ensuring both efficacy and safety in clinical trials. This approach could optimize therapeutic
outcomes and enhance the overall effectiveness of CAR-NK treatments for HCC. Additionally,
combining CAR-NK therapy with other treatments, such as immune checkpoint therapy, may
improve NK cell antitumor activity and enhance long-term complete remission in these
patients.
In this study, we demonstrated the anti-tumor effects of GPC3-specific NK92MI/HN3 cells
both in vitro and in vivo. Our findings emphasize the therapeutic potential of targeting GPC3
in HCC using CAR-NK cells, particularly in light of the observed variation in cytotoxic efficacy
between different GPC3 isoforms. These results suggest that isoform profiling may be crucial
in optimizing CAR-NK therapies for better patient outcomes. Further exploration of GPC3
isoform diversity in clinical settings and its impact on therapy resistance could guide the devel-
opment of more personalized and effective immunotherapies for HCC.
Supporting information
S1 Raw images.
(PDF)
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 15 / 19
Author Contributions
Data curation: Lei Yang.
Formal analysis: Lei Yang.
Funding acquisition: Chen Liu.
Investigation: Chen Liu.
Methodology: Qunfeng Wu, Dongfang Liu, Keith D. Robertson.
Writing – original draft: Lei Yang.
Writing – review & editing: Kien Pham, Yibo Xi, Chen Liu.
References
1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics
2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA:
a cancer journal for clinicians. 2021; 71(3):209–49. https://doi.org/10.3322/caac.21660 PMID:
33538338
2. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA: a
cancer journal for clinicians. 2015; 65(2):87–108. https://doi.org/10.3322/caac.21262 PMID: 25651787
3. Daher M, Basar R, Gokdemir E, Baran N, Uprety N, Nunez Cortes AK, et al. Targeting a cytokine check-
point enhances the fitness of armored cord blood CAR-NK cells. Blood, The Journal of the American
Society of Hematology. 2021; 137(5):624–36. https://doi.org/10.1182/blood.2020007748 PMID:
32902645
4. Tian Z, Chen Y, Gao B. Natural killer cells in liver disease. Hepatology. 2013; 57(4):1654–62. https://
doi.org/10.1002/hep.26115 PMID: 23111952
5. Sonpavde G. PD-1 and PD-L1 inhibitors as salvage therapy for urothelial carcinoma. Mass Medical
Soc; 2017. p. 1073–4.
6. Xie G, Dong H, Liang Y, Ham JD, Rizwan R, Chen J. CAR-NK cells: A promising cellular immunother-
apy for cancer. EBioMedicine. 2020;59.
7. Zhang C, Burger MC, Jennewein L, Genßler S, Scho
¨nfeld K, Zeiner P, et al. ErbB2/HER2-specific NK
cells for targeted therapy of glioblastoma. Journal of the National Cancer Institute. 2016; 108(5):djv375.
https://doi.org/10.1093/jnci/djv375 PMID: 26640245
8. Chen KH, Wada M, Pinz KG, Liu H, Lin K-W, Jares A, et al. Preclinical targeting of aggressive T-cell
malignancies using anti-CD5 chimeric antigen receptor. Leukemia. 2017; 31(10):2151–60. https://doi.
org/10.1038/leu.2017.8 PMID: 28074066
9. Chu J, Deng Y, Benson DM, He S, Hughes T, Zhang J, et al. CS1-specific chimeric antigen receptor
(CAR)-engineered natural killer cells enhance in vitro and in vivo antitumor activity against human multi-
ple myeloma. Leukemia. 2014; 28(4):917–27. https://doi.org/10.1038/leu.2013.279 PMID: 24067492
10. Liu E, Marin D, Banerjee P, Macapinlac HA, Thompson P, Basar R, et al. Use of CAR-transduced natu-
ral killer cells in CD19-positive lymphoid tumors. New England Journal of Medicine. 2020; 382(6):545–
53. https://doi.org/10.1056/NEJMoa1910607 PMID: 32023374
11. Pfefferle A, Huntington ND. You have got a fast CAR: chimeric antigen receptor NK cells in cancer ther-
apy. Cancers. 2020; 12(3):706. https://doi.org/10.3390/cancers12030706 PMID: 32192067
12. Romanski A, Uherek C, Bug G, Seifried E, Klingemann H, Wels WS, et al. CD 19-CAR engineered NK-
92 cells are sufficient to overcome NK cell resistance in B-cell malignancies. Journal of cellular and
molecular medicine. 2016; 20(7):1287–94. https://doi.org/10.1111/jcmm.12810 PMID: 27008316
13. Oelsner S, Waldmann A, Billmeier A, Ro
¨der J, Lindner A, Ullrich E, et al. Genetically engineered CAR
NK cells display selective cytotoxicity against FLT3-positive B-ALL and inhibit in vivo leukemia growth.
International journal of cancer. 2019; 145(7):1935–45. https://doi.org/10.1002/ijc.32269 PMID:
30860598
14. Li L, Jin R, Zhang X, Lv F, Liu L, Liu D, et al. Oncogenic activation of glypican-3 by c-Myc in human
hepatocellular carcinoma. Hepatology. 2012; 56(4):1380–90. https://doi.org/10.1002/hep.25891 PMID:
22706665
15. Shih T-C, Wang L, Wang H-C, Wan Y-JY. Glypican-3: a molecular marker for the detection and treat-
ment of hepatocellular carcinoma. Liver research. 2020; 4(4):168–72.
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 16 / 19
16. Chen I-P, Ariizumi S-i, Nakano M, Yamamoto M. Positive glypican-3 expression in early hepatocellular
carcinoma predicts recurrence after hepatectomy. Journal of gastroenterology. 2014; 49:117–25.
https://doi.org/10.1007/s00535-013-0793-2 PMID: 23532638
17. Wang Y-L, Zhu Z-J, Teng D-H, Yao Z, Gao W, Shen Z-Y. Glypican-3 expression and its relationship
with recurrence of HCC after liver transplantation. World Journal of Gastroenterology: WJG. 2012; 18
(19):2408. https://doi.org/10.3748/wjg.v18.i19.2408 PMID: 22654434
18. Ning S, Bin C, Na H, Peng S, Yi D, Xiang-hua Y, et al. Glypican-3, a novel prognostic marker of hepato-
cellular cancer, is related with postoperative metastasis and recurrence in hepatocellular cancer
patients. Molecular biology reports. 2012; 39:351–7. https://doi.org/10.1007/s11033-011-0745-y PMID:
21655958
19. Sawada Y, Yoshikawa T, Ofuji K, Yoshimura M, Tsuchiya N, Takahashi M, et al. Phase II study of the
GPC3-derived peptide vaccine as an adjuvant therapy for hepatocellular carcinoma patients. Oncoim-
munology. 2016; 5(5):e1129483. https://doi.org/10.1080/2162402X.2015.1129483 PMID: 27467945
20. Abou-Alfa GK, Puig O, Daniele B, Kudo M, Merle P, Park J-W, et al. Randomized phase II placebo con-
trolled study of codrituzumab in previously treated patients with advanced hepatocellular carcinoma.
Journal of hepatology. 2016; 65(2):289–95. https://doi.org/10.1016/j.jhep.2016.04.004 PMID:
27085251
21. Ishiguro T, Sano Y, Komatsu S-i, Kamata-Sakurai M, Kaneko A, Kinoshita Y, et al. An anti–glypican 3/
CD3 bispecific T cell–redirecting antibody for treatment of solid tumors. Science translational medicine.
2017; 9(410):eaal4291. https://doi.org/10.1126/scitranslmed.aal4291 PMID: 28978751
22. Li D, Li N, Zhang Y-F, Fu H, Feng M, Schneider D, et al. Persistent polyfunctional chimeric antigen
receptor T cells that target glypican 3 eliminate orthotopic hepatocellular carcinomas in mice. Gastroen-
terology. 2020; 158(8):2250–65. e20. https://doi.org/10.1053/j.gastro.2020.02.011 PMID: 32060001
23. Xie C, Monge B MC, Mabry-Hrones D, Coffman KL, Hicks S, Redd B, et al. A phase I study of GPC3 tar-
geted CAR-T cell therapy in advanced GPC3-expressing hepatocellular carcinoma (HCC). American
Society of Clinical Oncology; 2023.
24. Fleming BD, Urban DJ, Hall MD, Longerich T, Greten TF, Pastan I, et al. Engineered Anti-GPC3 Immu-
notoxin, HN3-ABD-T20, Produces Regression in Mouse Liver Cancer Xenografts Through Prolonged
Serum Retention. Hepatology. 2020; 71(5):1696–711. https://doi.org/10.1002/hep.30949 PMID:
31520528
25. Bell MM, Gutsche NT, King AP, Baidoo KE, Kelada OJ, Choyke PL, et al. Glypican-3-targeted alpha
particle therapy for hepatocellular carcinoma. Molecules. 2020; 26(1):4. https://doi.org/10.3390/
molecules26010004 PMID: 33374953
26. Labadie KP, Hamlin DK, Kenoyer A, Daniel SK, Utria AF, Ludwig AD, et al. Glypican-3–Targeted 227Th
α-Therapy Reduces Tumor Burden in an Orthotopic Xenograft Murine Model of Hepatocellular Carci-
noma. Journal of Nuclear Medicine. 2022; 63(7):1033–8.
27. Lin F, Clift R, Ehara T, Yanagida H, Horton S, Noncovich A, et al. Peptide Binder to Glypican-3 as a
Theranostic Agent for Hepatocellular Carcinoma. Journal of Nuclear Medicine. 2024; 65(4):586–92.
https://doi.org/10.2967/jnumed.123.266766 PMID: 38423788
28. Ho M, Kim H. Glypican-3: a new target for cancer immunotherapy. European journal of cancer. 2011;
47(3):333–8. https://doi.org/10.1016/j.ejca.2010.10.024 PMID: 21112773
29. Kolluri A, Ho M. The role of glypican-3 in regulating Wnt, YAP, and hedgehog in liver cancer. Frontiers
in oncology. 2019; 9:708. https://doi.org/10.3389/fonc.2019.00708 PMID: 31428581
30. Fischer J, Paret C, El Malki K, Alt F, Wingerter A, Neu MA, et al. CD19 isoforms enabling resistance to
CART-19 immunotherapy are expressed in B-ALL patients at initial diagnosis. Journal of Immunother-
apy (Hagerstown, Md: 1997). 2017; 40(5):187. https://doi.org/10.1097/CJI.0000000000000169 PMID:
28441264
31. Stamm S, Ben-Ari S, Rafalska I, Tang Y, Zhang Z, Toiber D, et al. Function of alternative splicing.
Gene. 2005; 344:1–20. https://doi.org/10.1016/j.gene.2004.10.022 PMID: 15656968
32. Skotheim RI, Nees M. Alternative splicing in cancer: noise, functional, or systematic? The international
journal of biochemistry & cell biology. 2007; 39(7–8):1432–49. https://doi.org/10.1016/j.biocel.2007.02.
016 PMID: 17416541
33. Venables JP. Unbalanced alternative splicing and its significance in cancer. Bioessays. 2006; 28
(4):378–86. https://doi.org/10.1002/bies.20390 PMID: 16547952
34. Zhu H, Dong H, Eksioglu E, Hemming A, Cao M, Crawford JM, et al. Hepatitis C virus triggers apoptosis
of a newly developed hepatoma cell line through antiviral defense system. Gastroenterology. 2007; 133
(5):1649–59. https://doi.org/10.1053/j.gastro.2007.09.017 PMID: 17983809
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 17 / 19
35. Trinh TL, Zhu G, Xiao X, Puszyk W, Sefah K, Wu Q, et al. A synthetic aptamer-drug adduct for targeted
liver cancer therapy. PLoS One. 2015; 10(11):e0136673. https://doi.org/10.1371/journal.pone.0136673
PMID: 26523833
36. Gao W, Kim H, Feng M, Phung Y, Xavier CP, Rubin JS, et al. Inactivation of Wnt signaling by a human
antibody that recognizes the heparan sulfate chains of glypican-3 for liver cancer therapy. Hepatology.
2014; 60(2):576–87. https://doi.org/10.1002/hep.26996 PMID: 24492943
37. Feng M, Gao W, Wang R, Chen W, Man Y-G, Figg WD, et al. Therapeutically targeting glypican-3 via a
conformation-specific single-domain antibody in hepatocellular carcinoma. Proceedings of the National
Academy of Sciences. 2013; 110(12):E1083–E91.
38. Gao W, Tang Z, Zhang Y-F, Feng M, Qian M, Dimitrov DS, et al. Immunotoxin targeting glypican-3
regresses liver cancer via dual inhibition of Wnt signalling and protein synthesis. Nature communica-
tions. 2015; 6(1):6536. https://doi.org/10.1038/ncomms7536 PMID: 25758784
39. Yang Y, Badeti S, Tseng H-c, Ma MT, Liu T, Jiang J-G, et al. Superior expansion and cytotoxicity of
human primary NK and CAR-NK cells from various sources via enriched metabolic pathways. Molecular
Therapy-Methods & Clinical Development. 2020; 18:428–45. https://doi.org/10.1016/j.omtm.2020.06.
014 PMID: 32695845
40. Baumhoer D, Tornillo L, Stadlmann S, Roncalli M, Diamantis EK, Terracciano LM. Glypican 3 expres-
sion in human nonneoplastic, preneoplastic, and neoplastic tissues: a tissue microarray analysis of
4,387 tissue samples. American journal of clinical pathology. 2008; 129(6):899–906. https://doi.org/10.
1309/HCQWPWD50XHD2DW6 PMID: 18480006
41. Zeng J, Tang SY, Toh LL, Wang S. Generation of “off-the-shelf” natural killer cells from peripheral blood
cell-derived induced pluripotent stem cells. Stem cell reports. 2017; 9(6):1796–812. https://doi.org/10.
1016/j.stemcr.2017.10.020 PMID: 29173894
42. Xiong W, Chen Y, Kang X, Chen Z, Zheng P, Hsu Y-H, et al. Immunological synapse predicts effective-
ness of chimeric antigen receptor cells. Molecular Therapy. 2018; 26(4):963–75. https://doi.org/10.
1016/j.ymthe.2018.01.020 PMID: 29503199
43. Takai H, Kato A, Kato C, Watanabe T, Matsubara K, Suzuki M, et al. The expression profile of glypican-
3 and its relation to macrophage population in human hepatocellular carcinoma. Liver International.
2009; 29(7):1056–64. https://doi.org/10.1111/j.1478-3231.2008.01968.x PMID: 19141032
44. Liu Q, Xu Y, Mou J, Tang K, Fu X, Li Y, et al. Irradiated chimeric antigen receptor engineered NK-92MI
cells show effective cytotoxicity against CD19+ malignancy in a mouse model. Cytotherapy. 2020; 22
(10):552–62. https://doi.org/10.1016/j.jcyt.2020.06.003 PMID: 32747298
45. Yang IS, Son H, Kim S, Kim S. ISOexpresso: a web-based platform for isoform-level expression analy-
sis in human cancer. BMC genomics. 2016; 17:1–14.
46. Kochenderfer JN, Rosenberg SA. Treating B-cell cancer with T cells expressing anti-CD19 chimeric
antigen receptors. Nature reviews Clinical oncology. 2013; 10(5):267–76. https://doi.org/10.1038/
nrclinonc.2013.46 PMID: 23546520
47. Boyiadzis MM, Dhodapkar MV, Brentjens RJ, Kochenderfer JN, Neelapu SS, Maus MV, et al. Chimeric
antigen receptor (CAR) T therapies for the treatment of hematologic malignancies: clinical perspective
and significance. Journal for immunotherapy of cancer. 2018; 6(1):1–12.
48. Schmidts A, Maus MV. Making CAR T cells a solid option for solid tumors. Frontiers in immunology.
2018; 9:2593. https://doi.org/10.3389/fimmu.2018.02593 PMID: 30467505
49. Klingemann H. Are natural killer cells superior CAR drivers? Oncoimmunology. 2014; 3(4):e28147.
https://doi.org/10.4161/onci.28147 PMID: 25340009
50. Lanier LL. NK cell receptors. Annual review of immunology. 1998; 16(1):359–93.
51. Olson JA, Leveson-Gower DB, Gill S, Baker J, Beilhack A, Negrin RS. NK cells mediate reduction of
GVHD by inhibiting activated, alloreactive T cells while retaining GVT effects. Blood, The Journal of the
American Society of Hematology. 2010; 115(21):4293–301. https://doi.org/10.1182/blood-2009-05-
222190 PMID: 20233969
52. Miller JS, Soignier Y, Panoskaltsis-Mortari A, McNearney SA, Yun GH, Fautsch SK, et al. Successful
adoptive transfer and in vivo expansion of human haploidentical NK cells in patients with cancer. Blood.
2005; 105(8):3051–7. https://doi.org/10.1182/blood-2004-07-2974 PMID: 15632206
53. Tazi J, Bakkour N, Stamm S. Alternative splicing and disease. Biochimica et Biophysica Acta (BBA)-
Molecular Basis of Disease. 2009; 1792(1):14–26. https://doi.org/10.1016/j.bbadis.2008.09.017 PMID:
18992329
54. Ghigna C, Giordano S, Shen H, Benvenuto F, Castiglioni F, Comoglio PM, et al. Cell motilityis con-
trolled by SF2/ASF through alternative splicing of the Ron protooncogene. Molecular cell. 2005; 20
(6):881–90. https://doi.org/10.1016/j.molcel.2005.10.026 PMID: 16364913
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 18 / 19
55. Poulikakos PI, Persaud Y, Janakiraman M, Kong X, Ng C, Moriceau G, et al. RAF inhibitor resistance is
mediated by dimerization of aberrantly spliced BRAF (V600E). Nature. 2011; 480(7377):387–90.
https://doi.org/10.1038/nature10662 PMID: 22113612
56. Ma X, Zhao Y, Li Y, Lu H, He Y. Relevance of Bcl-x expression in different types of endometrial tissues.
Journal of Experimental & Clinical Cancer Research. 2010; 29:1–7. https://doi.org/10.1186/1756-9966-
29-14 PMID: 20178603
57. Castiglioni F, Tagliabue E, Campiglio M, Pupa S, Balsari A, Menard S. Role of exon-16-deleted HER2
in breast carcinomas. Endocrine-related cancer. 2006; 13(1):221–32. https://doi.org/10.1677/erc.1.
01047 PMID: 16601290
PLOS ONE
Glypican-3 targeted CAR-NK therapy in hepatocellular carcinoma
PLOS ONE | https://doi.org/10.1371/journal.pone.0317401 January 22, 2025 19 / 19
