Tumor intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast cancer

Full text

Original Article
Tumor-intrinsic CDC42BPB confers
resistance to anti-PD-1 immune checkpoint
blockade in breast cancer
Ravindra Pramod Deshpande,
1
Kerui Wu,
1
Shih-Ying Wu,
1
Abhishek Tyagi,
1
Eleanor C. Smith,
1
John Hunting,
1
Jimmy Ruiz,
1
Wencheng Li,
1
and Kounosuke Watabe
1
1
Department of Cancer Biology, Wake Forest University School of Medicine, Winston-Salem, NC 27157, USA
Immune checkpoint blockade has been used to treat breast can-
cer, but the clinical responses remain relatively poor. We have
used the CRISPR-Cas9 kinome knockout library consisting of
763 kinase genes to identify tumor-intrinsic kinases conferring
resistance to anti-PD-1 immune checkpoint blockade. We have
identified the CDC42BPB kinase as a potential target to over-
come the resistance to anti-PD-1 immune checkpoint blockade
immunotherapy. We found that CDC42BPB is highly ex-
pressed in breast cancer patients who are non-responsive to
immunotherapy. Furthermore, a small-molecule pharmacolog-
ical inhibitor, BDP5290, which targets CDC42BPB, synergized
with anti-PD-1 and enhanced tumor cell killing by promoting
T cell proliferation in both in vitro and in vivo assays. More-
over, anti-PD-1-resistant breast cancer cells showed higher
expression of CDC42BPB, and its inhibition rendered the resis-
tant cells more susceptible to T cell killing in the presence of
anti-PD-1. We also found that CDC42BPB phosphorylated
AURKA, which in turn upregulated PD-L1 through cMYC.
Our results have revealed a robust link between tumor-intrinsic
kinase and immunotherapy resistance and have provided a
rationale for a unique combination therapy of CDC42BPB in-
hibition and anti-PD-1 immunotherapy for breast cancer.
INTRODUCTION
Triple-negative breast cancer (TNBC) is characterized by the absence
of estrogen receptor), progesterone receptor, and epidermal growth
factor receptor 2 (HER-2) and accounts for 15%–20% of breast can-
cers.
1,2
TNBC is associated with a high risk of distant metastasis,
particularly to the lung and brain, and a high rate of recurrence,
3,4
leading to an average overall survival of only 13 months.
5
TNBC is
a heterogeneous disease with poor responsiveness to chemotherapy
6
and other targeted therapies, and therefore requires specially curated
approaches, including combinations of chemotherapy, radiation ther-
apy, and immune checkpoint inhibitors (ICIs).
7–9
ICI therapy con-
sists of blocking antibodies mainly to programmed cell death protein
1 (PD-1) expressed on T cells or its ligand PD-L1 expressed on tumor
cells. Abrogation of PD-1 or PD-L1 reinvigorates the T cells to un-
leash their cytotoxic and effector functions on the target tumor cells.
10
However, the efficacy of single-agent ICIs is in the range of 5%–24%
in TNBC, and the initial responders often develop disease relapse,
indicating the presence of intrinsic resistance.
11,12
Potential mecha-
nisms of this resistance include cancer stem cells, low antigen presen-
tation, less immune cell infiltration, and genes such as PD-L1, IFN-g,
JAK2, and STAT1.
13–15
Constitutive activation of several kinases
(RAS/RAF/mTOR/ERK/JAK) is also known to trigger ICI resistance
through the expression of immune checkpoint genes.
16
Similarly,
MAPK1 and EGFR kinases suppress the expression of antigen presen-
tation receptors such as MHC-I and hamper T cell-mediated antigen
presentation and concomitant cytotoxicity.
17
Small-molecule kinase
inhibitors have been evaluated in clinical trials to treat patients with
TNBC.
18
Thus, targeting kinase inhibitors in combination with ICIs
represents an attractive strategy to enhance the therapy’s responsive-
ness in TNBC. The development of CRISPR-Cas9 screening has
emerged as a powerful tool to investigate resistance mechanisms,
paving the way to develop novel combination therapies.
19,20
In the
present study, we have employed screening of a CRISPR-Cas9-medi-
ated kinome knockout library and have identified CDC42BPB as a
tumor-intrinsic kinase conferring resistance to ICIs in TNBC. Our re-
sults indicate that inhibition of CDC42BPB, by the small-molecule
pharmacological inhibitor BDP5290, sensitizes TNBC to anti-PD-1
immune checkpoint blockade therapy.
RESULTS
CRISPR library screening to identify tumor-intrinsic kinase
conferring resistance to anti-PD-1 treatment
To identify tumor-intrinsic kinases conferring anti-PD-1 resistance,
we first transduced the human Brunello kinome knockout library,
which consisted of over 3,000 unique single-guide RNAs (sgRNAs)
targeting 763 human kinase genes, into MCF10CA1a cells followed
by puromycin selection. The surviving cells were cultured for 3 weeks
to allow the gene editing to take place (Figure 1A). Seven single-cell
clones from the library-transduced cells were isolated, and sgRNA in-
serts were amplified with PCR to examine the quality of the library.
Received 24 January 2024; accepted 28 July 2024;
https://doi.org/10.1016/j.ymthe.2024.07.021.
Correspondence: Kounosuke Watabe, PhD, Department of Cancer Biology, Wake
Forest University School of Medicine, Winston-Salem, NC 27157, USA.
E-mail: kwatabe@wakehealth.edu
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021
Molecular Therapy Vol. 32 No 10 October 2024
Published by Elsevier Inc. on behalf of The American Society of Gene and Cell Therapy.
1

Figure 1. CRISPR library screening to identify tumor-intrinsic kinase conferring resistance to anti-PD-1 treatment
(A) Scheme of experiment. (B) In vitro co-culture of MCF10CA1a cells in the presence of educated PBMCs, anti-PD-1, and anti-CD8 at 5:1 PBMC/cancer cell ratio with
40 mg/mL anti-PD-1 and -CD8 antibody. One -way ANOVA with Tukey’s multiple comparison post hoc test was used to determine statistical significance between the
groups. (C) Diagram of data analysis. (D) MF10CA1a cells were transfected with either control or PTK2 siRNA followed by examining the expression of PTK2 RNA by real-time
PCR (n= 3). GAPDH was used as an internal control (left). The co-culture assay was performed in the presence of PBMCs, anti-PD-1 (n= 3/group) (right). Statistic al
significance was calculated using the unpaired two-tailed Student’s t test with data presented as mean ±SEM. (E) MF10CA1a cells were transfected with either control or
CDC42BPB siRNA followed by examining the expression of CDC42BPB RNA by real-time PCR (n= 3/group). GAPDH was used as an internal control (left). The co-culture
(legend continued on next page)
Molecular Therapy
2 Molecular Therapy Vol. 32 No 10 October 2024
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

All clones were found to have only one sgRNA, indicating that each
cell was transduced with only one guide (Figure S1A). To prepare
T cells for library screening, human peripheral blood mononuclear
cells (PBMCs) were educated with the cell lysate of MCF10CA1a fol-
lowed by T cell expansion using CD3/CD28 beads and recombinant
interleukin-2 (IL-2) (Figure 1A). The CD3 T cell population among
the expanded PBMCs was 80% (Figures S1B and S1C), while 40%
of the T cells were interferon-g(IFN-g) positive, indicating that the
dendritic cells cross-presented the tumor antigen to CD8 T cells
(Figure S1D).
To confirm the cancer cell killing mediated by the educated CD8
T cells, we co-cultured MCF10CA1a cells with the educated PBMCs
at a 1:5 cancer cell/PBMC ratio in the presence of anti-PD-1 immune
checkpoint blockade for 48 h with or without an anti-CD8 antibody.
We found that cell death was rescued when the CD8 T cells were
blocked (Figure 1B), indicating that the cancer cell death in the co-cul-
ture system is dependent on the educated CD8 T cells. To confirm that
the educated PBMCs contributed to cell death, uneducated PBMCs
were treated with IL-4 and granulocyte macrophage colony-stimu-
lating factor (GM-CSF) as a control and then used for co-culture
with the cancercells. We found that only the educated PBMCs induced
MCF10CA1a cell death and that the fraction of surviving cells was
significantly less when the MCF10CA1a cells were co-cultured with
educated PBMCs as compared with uneducated PBMCs, indicating
that cell death was dependent on the educated CD8 T cells (Figure S1E).
In addition, we also quantified Caspase3/7-positive cells by fluores-
cence-activated cell sorting. When the MCF10CA1a cells were co-
cultured in the presence of educated PBMCs and anti-PD-1 immune
checkpoint blockade, the number of Caspase3/7-positive cells was
significantly increased in the treatment set as compared with the con-
trol. In addition, cell death was rescued when CD8 T cells were blocked
(Figure S1F). To verify that the cell death in the co-culture system was
antigen dependent, we blocked MHC-I and observed that the reduced
cell viability was rescued (Figure S1G).
With this established screening assay system, the MCF10CA1a cells
were transduced with the kinome library with 600-fold representation
and were co-cultured with the educated PBMCs. Dead tumor cells
and PBMCs were then washed off with ice-cold PBS, and genomic
DNA was isolated to perform deep sequencing. The genes with all
four sgRNAs depleted at p< 0.05 were analyzed using the
MAGeCK algorithm (Table S1). Additionally, the fold change was
calculated by analyzing the depleted sgRNAs in the anti-PD-1-treated
set by adding pseudo count 1, and the genes which were 3-fold en-
riched in the anti-PD-1 immune checkpoint blockade-treated set
were selected (Figure 1C).
To validate the result of this screening, we silenced the four
candidate genes with a small interfering RNA (siRNA)-based
approach (Figures 1D–1G, left) and performed the co-culture assay
with the educated PBMCs in the presence of anti-PD-1 immune
checkpoint blockade. We found that the knockdown of PTK2 sup-
pressed tumor cell viability, but the intensity of cell death was more
pronounced when CDC42BPB expression was silenced (Figures 1D
and 1E, right). On a similar note, we performed a co-culture assay
by silencing NAGK (Figure 1F) and ITPKA (Figure 1G) but observed
that the cell viability was unaffected in the presence of anti-PD-1.
Therefore, we chose CDC42BPB for subsequent experiments. Of
note, PTK2 was previously found to contribute to immunotherapy ef-
ficacy,
21,22
indicating that our in vitro co-culture screening system is
translationally relevant. Interestingly, we examined data from the Hu-
man Protein Atlas portal
23
and observed that CDC42BPB expression
is correlated with poor survival in breast cancer, lung adenocarci-
noma, and lung squamous cell carcinoma (Figures S1H–S1J). We
analyzed the reverse-phase protein array data available in the
Kaplan-Meier (KM) plotter portal
24
and found that CDC42BPB
expression was associated with poor prognosis in breast cancer pa-
tients (Figure S1K). Furthermore, we examined the expression of
CDC42BPB in the treatment-naive breast cancer patient tissues of
23 responders and 16 non-responders to ICIs. We found that the
expression of CDC42BPB in non-responders was significantly higher
than that in responders (Figures 1H and 1I). To further verify this
phenomenon we again used the KM plotter portal,
25
which included
data of patients treated with anti-PD-1 immune checkpoint blockade.
The result of our analysis indicated that patients with lower expres-
sion of CDC42BPB showed significantly better overall survival after
immunotherapy treatment (Figure S1L). In addition, we extracted
the transcriptomic data of lung squamous cell carcinoma patients
treated with anti-PD-1 immunotherapy from the Tumor Immuno-
therapy Gene Expression Resource (TIGER) portal.
26
Again, we
found that patients with higher expression of CDC42BPB had poorer
prognosis compared to the patients with lower expression of
CDC42BPB (Figure S1M). CDC42BPB is expressed in a wide range
of cancer types, indicating that targeting its expression may be a
feasible approach (Figure S1N). These data further support the notion
that CDC42BPB may serve as a predictive biomarker for anti-PD-1
immune checkpoint blockade therapy.
assay was performed in the presence of PBMCs, anti-PD-1 as described in (B). Statistical significance was calculated using the unpaired two-tailed Student’s t test with data
presented as mean ±SEM. (F) MF10CA1a cells were transfected with either control or NAGK siRNA followed by examining the expression of NAGK RNA by real-time PCR
(n= 3/group). GAPDH was used as an internal control (left). The co-culture assay was performed in the presence of PBMCs, anti-PD-1 (n= 3/group) (right). Statistical
significance was calculated using the unpaired two-tailed Student’s t test with data presented as mean ±SEM. (G) MF10CA1a cells were transfected with either control or
ITPKA siRNA followed by examining the expression of ITPKA RNA by real-time PCR (n= 3/group). GAPDH was used as an internal control (left). The co-culture assay was
performed in the presence of PBMCs, anti-PD-1 (n= 3/group) (right). Statistical significance was calculated using the unpaired two-tailed Student’s t test with data presented
as mean ±SEM. (H) CDC42BPB expression was examined by immunohistochemistry in formalin-fi xed paraffin-embedded tumor tissue sections from breast cancer patients
who either responded (R) or not (NR) to immunotherapy. Scale bars, 100 mm. (I) Quantification of CDC42BPB protein expression by immunohistochemistry in (H). Statistical
significance was calculated using the unpaired two-tailed Student’s t test and presented as mean ±SEM. *p< 0.05, **p< 0.01, ***p< 0.001 ; NS, not significant.
www.moleculartherapy.org
Molecular Therapy Vol. 32 No 10 October 2024 3
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

CDC42BPB inhibition synergizes with anti-PD-1 treatment in
breast cancer in vitro
We have shown that knockout or knockdown of CDC42BPB
significantly sensitized tumor cells to anti-PD-1 treatment. Therefore,
we hypothesized that CDC42BPB inhibition with the pharmacolog-
ical small-molecule inhibitor, BDP5290,
27
should sensitize
cancer cells to anti-PD-1 treatment. We educated the PBMCs with
MCF10CA1a lysate and performed in vitro co-culture in the presence
of anti-PD-1 antibody and BDP5290 (20 and 40 nM, IC
50
150 nM).
As shown in Figures 2A, 2B, and S2A, we found that BDP5290 exerted
a significant synergistic effect when combined with anti-PD-1. The
results were verified in another triple-negative cell line, MDA-MB-
231 (Figures 2C, 2D, and S2B). To validate our results using an im-
mune-competent mouse system, two other triple-negative cell lines
of murine origin, 4T1 and E0771, were used. We first isolated CD8
T cells by MACS from a syngeneic mouse spleen inoculated with
4T1 or EO771 cells followed by activation and expansion of these cells
in the presence of murine recombinant IL-2 (Figure S2C). We then
co-cultured the 4T1 cells with CD8 T cells for 48 h. We found that
BDP5290 significantly synergized the effect of the ICI (Figure S2D).
Similar results were obtained with syngeneic mouse model E0771
cells (Figure S2E). In addition, we have examined effect of the
PTK2 inhibitor, fostamatinib,
28
on anti-PD-1 treatment. We found
that fostamatinib showed a synergistic effect only at 30 nM, while
the CDC42BPB inhibitor (BDP5290) exerted a synergistic effect
even at 10 nM. Additionally, the magnitude of inhibition was signif-
icantly higher for the CDC42BPB inhibitor than the PTK2 inhibitor
(Figure S2F). Furthermore, we performed the synergy experiment us-
ing two lung cancer cell lines (LL2, CMT167) and one melanoma cell
line (SKMEL-28). The results demonstrate that CDC42BPB plays a
pivotal role in conferring anti-PD-1 immunotherapy resistance in
lung cancer (Figures S2G and S2H) and melanoma (Figure S2I),
thus supporting the notion that CDC42BPB is an attractive target
to enhance anti-PD-1 immune checkpoint blockade in a wide range
of tumors. To confirm the role of CDC42BPB in immunotherapy
resistance, we generated anti-PD-1-resistant MCF10CA1a by
culturing with PBMCs in the presence of anti-PD1 for four rounds
(Figure 2E). We then examined the RNA and protein expression of
CDC42BPB in parental and anti-PD-1-resistant MCF10CA1a (Fig-
ure 2F). CDC42BPB was significantly overexpressed in anti-PD-1 im-
mune checkpoint blockade-resistant MCF10CA1a. Importantly, we
found that BDP5290 sensitized the resistant cells to T cell killing (Fig-
ure 2G). We examined the effect of CDC42BPB silencing, inhibition,
and overexpression in MCF10CA1a cells in the absence of T cells. We
found that genetic and pharmacological CDC42BPB inhibition
reduced cell viability by 10% (Figures S2J and S2K). A previous
study reported that BDP5290 inhibited cell viability with an IC
50
value of 5 mM,
27
thus supporting our observation that it has minimal
effect on cell viability at a lower dose. Additionally, there was no sig-
nificant impact of CDC42BPB overexpression on cell growth (Fig-
ure S2L). This confirms that CDC42BPB inhibition minimally con-
tributes to cell death. We also performed the co-culture assay by
silencing PD-L1 expression and observed that the absence of PD-L1
does not affect cell viability. There was a statistically insignificant dif-
ference between the PD-L1-silenced group and the BDP5290-treated
group in combination with PBMCs (Figure S2M), indicating that PD-
L1 is the downstream target of BDP5290. CDC42BPB is not expressed
on the immune cells (Figure S2N), and BDP5290 appears to show its
effect through inhibiting CDC42BPB in the cancer cells. Together,
these results indicate that CDC42BPB plays a critical role in anti-
PD-1 immunotherapy resistance in breast cancer, and its inhibition
with BDP5290 sensitizes cancer cells to T cell-mediated killing.
CDC42BPB inhibition synergizes with anti-PD-1 treatment in
breast cancer in vivo
To examine the efficacy of CDC42BPB inhibition with BDP5290
in vivo, we implanted E0771 cells in a syngeneic C57BL/6 host
(Figure 3A) and treated the animals with a combination of anti-
PD-1 immune checkpoint blockade and BDP5290. We found that
the combination treatment significantly suppressed tumor growth
(Figure 3B) and reduced tumor burden (Figure 3C) in mice. We
also examined the lungs of these mice (Figure 3D) and found that
the combination treatment significantly reduced micrometastasis
(Figure 3E). These results support the notion that BDP5290 could
be used to synergize with anti-PD-1 to suppress metastatic disease
of breast cancer. We examined the toxicity of BDP5290 by measuring
body weight and serum aspartate transaminase (AST) activity at the
endpoint and found no significant difference in weight (Figure S3A)
or serum AST levels in the animals from four groups (Figure S3B). We
next investigated the effects of CDC42BPB inhibition on CD8 im-
mune cells in the tumors. Our results indicated that there was a sig-
nificant degree of CD8 T cell infiltration in the group with combina-
tion treatment (Figures 3F and 3G). We further confirmed the results
by measuring the correlation between CDC42BPB expression and
CD8 T cell infiltration using TIMER. There was a significant associ-
ation between CD8 cell infiltration and CDC42BPB expression
in breast cancer (Figure 3H). Our results demonstrated that
CDC42BPB protein was highly expressed in immunotherapy non-re-
sponding patients with breast cancer (Figures 1H and 1I). To
again examine the utility of CDC42BPB as a predictive marker of
immunotherapy responsiveness, we compared the expression of
CDC42BPB in E0771 and 4T1 cells and found that the expression
of CDC42BPB was significantly higher in 4T1 than in E0771
(Figures 3I and 3J). This is in line with previous evidence that 4T1
is resistant to anti-PD-1 immune checkpoint blockade.
29
We then
silenced CDC42BPB in 4T1 cells and examined the efficacy of anti-
PD-1 immune checkpoint blockade in a syngeneic mouse model (Fig-
ure 3K). We found that silencing CDC42BPB sensitized 4T1 cells to
anti-PD-1 immune checkpoint blockade and formed less aggressive
tumors (Figures 3L and 3M). The combination of CDC42BPB
silencing and anti-PD-1 immune checkpoint blockade significantly
reduced tumor burden in animals (Figure 3N). Similar to the results
of the E0771 model (Figures 3D and 3E), we found that anti-PD-1 im-
mune checkpoint blockade in addition to CDC42BPB silencing
reduced lung metastasis (Figures 3O and 3P) and enhanced T cell
proliferation (Figures 3Q and 3R) without significantly affecting
serum AST (Figure S3C) or animal body weight (Figure S3D) at the
endpoint. siRNA-based therapeutics were previously shown to have
Molecular Therapy
4 Molecular Therapy Vol. 32 No 10 October 2024
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

Figure 2. CDC42BPB inhibition synergizes with anti-PD-1 treatment in breast cancer in vitro
(A) Co-culture assay for MCF10CA1a cells in the presence of vehicle, BDP5290 (10 nM and 20 nM), and anti-PD-1 (20 mg/mL and 40 mg/mL) alone or in combination for 48 h
at effector/tumor cell ratio of 5:1 (n= 3/group). (B) Normalized isobologram for (A) was generated with CompuSyn based on the inhibitory effect of anti-PD-1 and BDP5290
treatment. (C) Similar to (A), an in vitro synergy assay was performed for MDA-MB-231 with BDP5290 alone (20 nM and 40 nM) or in combination with anti-PD-1 (20 mg/mL
and 40 mg/mL) for 48 h (n= 3/group). (D) Normalized isobologram for (C) generated with CompuSyn based on the inhibitory effect of anti-PD-1 and BDP5290 treatment. (E)
Schematic illustration of experimental procedures in obtaining anti-PD1-resistant MCF10CA1a cells. (F) CDC42BPB RNA and protein expression were examined in parental
and anti-PD1-resistant MCF10CA1a cells by real-time PCR (n= 3/group) and western blot. GAPDH was used as an internal control. Statistical significance was calculated
using the unpaired two-tailed Student’s t test with data presented as mean ±SEM. (G) Anti-PD-1-resistant MCF10CA1a cells (n= 500) were cultured in 96-well plates and
treated with BDP5290 (20 nM) and anti-PD-1 (40 mg/mL) in the presence of educated PBMCs at effector/tumor cell ratio of 5:1 for 48 h (n= 3/group). Statistical significance
between the groups was determined by one-way ANOVA with Tukey’s multiple comparison post hoc test, with data presented as mean ±SEM. *p< 0.05, ,***p< 0.001,
****p< 0.0001.
www.moleculartherapy.org
Molecular Therapy Vol. 32 No 10 October 2024 5
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

Figure 3. CDC42BPB inhibition synergizes with anti-PD-1 treatment in breast cancer in vivo
(A) Schematic illustration of the in vivo experiment. (B) Tumor size was measured, and statistical significance was determined by two-way ANOVA. (C) Tumor weight (n= 10/
group) was measured at the endpoint. Statistical significance between the groups was determined by one-way ANOVA with Tukey’s multiple comparison post hoc test. (D)
Representative H&E images showing lung micrometastasis from each group. Scale bars, 100 mm. (E) The lung micrometastases were quantified in random fields from the
lung tissue of each mouse. Statistical significance between the groups was determined by one-way ANOVA with Tukey’s multiple comparison post hoc test. (F) Tumor tissues
isolated at the endpoint were examined for intratumoral CD8 staining by immu nohistochemistry (n= 10/group). Scale bars, 100 mm. (G) Stained intratumoral CD8 T cells were
quantified in random high-power fields from each group. Statistical significance between the groups was calculated by one-way ANOVA with Tukey’s multiple comparison
(legend continued on next page)
Molecular Therapy
6 Molecular Therapy Vol. 32 No 10 October 2024
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

great potential in overcoming docetaxel and taxol resistance in pros-
tate
30
and breast cancer.
31
Additionally, a large number of clinical and
preclinical studies have demonstrated the specificity of siRNA in can-
cer treatment.
32
Therefore, CDC42BPB silencing in combination with
anti-PD-1 immune checkpoint blockade represents an attractive
strategy to improve the clinical outcomes for breast cancer patients.
CDC42BPB upregulates PD-L1 expression and is associated
with immune evasion
To investigate how CDC42BPB affects responsiveness to anti-PD-1
immune checkpoint blockade, we hypothesized that CDC42BPB
may regulate PD-L1 expression, which is known to play an important
role in conferring resistance to immune checkpoint blockade. To test
this, we ectopically expressed CDC42BPB in MCF10CA1a and found
that PD-L1 RNA and protein expression were significantly increased
(Figures 4A, S4A, and S4B). CDC42BPB has been known to regulate
AURKA expression,
27
and AURKA was demonstrated to regulate
PD-L1 expression through cMYC in TNBC.
33
Therefore, it is possible
that CDC42BPB regulates PD-L1 expression through AURKA and
cMYC. AURKA is a serine-threonine kinase known to be
activated through Thr288 phosphorylation. We found that ectopic
CDC42BPB expression enhanced Thr288 AURKA (Figure 4B) as
well as cMYC RNA and protein expression (Figure 4C), while knock-
ing down CDC42BPB significantly reduced the expression of cMYC,
PD-L1, and Thr288 phosphorylation of AURKA (Figures 4D–4F).
Consistent with these results, the AURKA inhibitor TC-S 7010
34
significantly reduced PD-L1 expression, while it was rescued
by ectopic expression of cMYC in MCF10CA1a cells (Figure 4G).
The inhibition of AURKA did not affect CDC42BPB expression (Fig-
ure S4C), demonstrating that AURKA does not regulate CDC42BPB.
We also performed an in vitro kinase assay to examine whether
CDC42BPB can directly phosphorylate AURKA. Our results indi-
cated that AURKA is a substrate for CDC42BPB (Figure S4D).
Furthermore, the ectopic expression of cMYC rescued the inhibitory
effect on cell viability observed in the presence of PBMCs and anti-
PD-1 immune checkpoint blockade (Figure S4E), indicating that
cMYC is the downstream target of CDC42BPB. As shown in Fig-
ure 4G, our results suggest that CDC42BPB promotes PD-L1 expres-
sion on tumor cells by activating AURKA. This PD-L1 on tumor cells
engages PD-1 on T cells, blocking T cell proliferation and rendering
tumor cells more resistant to the checkpoint inhibitor. However,
PD-1 expression on T cells is independent of CDC42BPB (Fig-
ure S4F). Therefore, suppressing the expression of PD-L1 by inhibit-
ing CDC42BPB in tumor cells while blocking PD-1 in T cells with a
checkpoint inhibitor is expected to synergistically stimulate the effi-
cacy of the checkpoint inhibitor and overcome therapy resistance,
as shown in Figure 2. To further validate this hypothesis, we treated
the MCF10CA1a cells with anti-PD-L1 or anti-PD-1 in combination
with the CDC42BPB inhibitor BDP5290. We found that the degree of
tumor inhibitory effect was similar between the treatment of PD-L1
and BDP5290, but the combination of PD-1 and BDP5290 synergis-
tically suppressed tumor growth (Figure S4G). To examine the impact
of CDC42BPB on immune cells in the tumor microenvironment, we
pretreated the PBMCs with BDP5290 and then co-cultured them with
the cancer cells. Our results indicate that there is no significant differ-
ence in the cell death induced by PBMCs pretreated with or without
BDP5290. This may demonstrate that BDP5290 does not influence
the cells of the tumor microenvironment (Figure S4H). Moreover,
we examined the MHC-I expression in MCF10CA1a cells that were
treated with 10 nM or 20 nM of BDP5290 and found that
BDP5290 did not affect the mRNA and protein expression of
MHC-I, confirming that BDP5290 enhances the efficacy of anti-
PD-1 immune checkpoint blockade in a PD-L1-dependent manner
(Figures S4I and S4J). Interestingly, gene co-dependency data from
the human cancer dependency map (DepMap) project, which con-
tained the screening data of 33 cancer cell lines with the GeCKO
CRISPR-Cas9 library,
34
revealed that CDC42BPB and PD-L1 expres-
sion was co-dependent (Figure 4H). These data also support the view
that CDC42BPB plays a significant role in PD-L1 expression. More-
over, AURKA Thr288 (Figures 4I and 4J) and cMYC (Figures 4K and
4L) expressions were significantly decreased in tumor tissues from an-
imals treated with BDP5290 and anti-PD-1 (Figure 3). We examined
pAURKA and CD8 T cell expression in breast cancer patients who
were immunotherapy responders and non-responders in Figure 1.
The non-responders with high expression of CDC42BPB were found
to have high pAURKA expression (Figures 4M and 4N) and low CD8
T cell staining (Figures 4O and 4P). These results strongly suggest that
CDC42BPB regulates PD-L1 expression through phosphorylation of
AURKA and modulates immunotherapy efficacy (Figure 4Q).
DISCUSSION
In this study, we performed kinome library screening to investigate
the role of tumor-intrinsic kinases in conferring immunotherapy
resistance in TNBC. To our knowledge, this is the first study to use
the CRISPR-Cas9 kinome knockout library to identify tumor-
intrinsic kinases conferring resistance to immune checkpoint
post hoc test. (H) Scatterplot analysis showing a correlatio n between CDC42BPB with CD8 T cells in The Cancer Genome Atlas breast cancer samples. (I) RNA was isolate d
from E0771 and 4T1 cells followed by real-time PCR to examine the CDC42BPB RNA expression (n= 3/group). Statistical significance between two groups was determined
by the unpaired two-tailed Student’s t test. (J) CDC42BPB protein was examined in E0771 and 4T1 cell lysate by western blot and quantified using ImageJ. (K) Schematic
illustration of the in vivo experiment. 4T1cells (2 10
4
) were implanted in the fourth mammary fat pad of both sides (n= 12/group). After 7 days, mice were treated with either
vehicle control or anti-PD-1 as indicated. (L) Tumor size was measured by a caliper and calculated by using the formula V=LW
2
/2. Statistical significance was determined by
two-way ANOVA. (M) Representative tumor images from the four groups. (N) Tumor weight ( n= 12/group) was measured at the endpoint. Statistical significance between the
groups was determined by the two-tailed Student’s t test. (O) Representative H&E images showing lung micrometastases from each group. Scale bars, 100 mm. (P) The lung
micrometastases were quantified in random fields from the lung tissue of each mouse. Statistical significance between the groups was determined by the unpaired two-tailed
Student’s t test. (Q) Tumor tissues isolated at the endpoint were examined for intratumoral CD8 staining by immunohistochemistry (n= 5/group). Scale bars, 100 mm. (R)
Stained intratumoral CD8 T cells were quantified in random high-power fields from each group. Statistical significance was determined by one-way ANOVA with Tukey’s
multiple comparison post hoc test. Data presented as mean ±SEM. *p< 0.05, **p< 0.01, ***p< 0.001, ****p< 0.0001.
www.moleculartherapy.org
Molecular Therapy Vol. 32 No 10 October 2024 7
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

(legend on next page)
Molecular Therapy
8 Molecular Therapy Vol. 32 No 10 October 2024
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

blockade using an in vitro co-culture in a humanized system. Our re-
sults indicated the critical role of CDC42BPB in mediating anti-PD-1
immune checkpoint blockade resistance and uncovered the synthetic
lethality between CDC42BPB and the immune checkpoint inhibitor
anti-PD-1. CDC42BPB kinase is known to play an important role
in regulating the actin cytoskeleton organization, cellular migration,
35
and invasion
36
in skin cancer. CDC42BPB is localized near the plasma
membrane
37
and affects the dendrite outgrowth and spine formation
through phosphorylation of downstream MLC2. However, there is no
report indicating the role of this gene in conferring immunotherapy
responsiveness.
Our results show that CDC42BPB inhibition significantly improved
anti-PD-1 immunotherapy through PD-L1-regulated T cell prolifera-
tion. T cell dysfunction has been reported to influence immunotherapy
effectiveness through the action of various kinases. For instance, Hou
et al. showed the role of tumor-intrinsic PRMT1 and RIPK1 in regu-
lating sensitivity to T cell-mediated killing in colon cancer models.
38
Akin to this, targeting FAK2,
21
PRKCI,
39
and PDGFRB
40
has been
shown to enhance immunotherapy efficacy in ovarian, lung, and
gastric carcinoma through potentiating T cell activity. Furthermore,
the upstream kinase effector, MAPK, has been shown to induce PD-
L1 expression in lung adenocarcinoma and is viewed as an attractive
target to enhance immunotherapy efficacy.
41
Moreover, Sow et al.
have shown that transforming growth factor-breceptor-1 kinase inhi-
bition synergized with anti-PD-L1 in colon and pancreatic tumors.
42
As kinases have emerged as potential pharmacological targets to
enhance the effectiveness of immune-based treatments,
43,44
the roles
of the TNBC-intrinsic kinase CDC42BPB in the previously identified
kinase network may warrant further investigations.
We found that CDC42BPB exerts its effect through AURKA. AURKA
is a serine-threonine kinase overexpressed in breast cancer, and its
expression is associated with poor prognosis.
45
Additionally,
AURKA is known to upregulate PD-L1 through cMYC in TNBC.
33
In the present study, we showed that CDC42BPB activated AURKA
in murine as well as human TNBC cell lines, indicating the critical
role of the cMYC-AURKA-PD-L1 axis in immunotherapy resistance.
Consistent with our results, Vilgelm et al. previously demonstrated
that AURKA inhibition enhanced the immunotherapy efficacy in
melanoma by enhancing T cell activity.
46
Moreover, cMYC was
shown to upregulate PD-L1 in multiple malignancies including
breast
47
and squamous cell carcinoma.
48
These results indicated
that the role of this signaling pathway in immune evasion is widely
conserved in multiple cancers. Therefore, targeting the upstream acti-
vator of AURKA may be an effective strategy to treat breast cancer
patients in combination with an immune checkpoint inhibitor. A
recent study by Wang et al. has shown that Aurora A kinase inhibition
upregulated PD-L1 expression both in vitro and in vivo in colon can-
cer.
49
Similarly, Sun et al. have found that AURKA activates PD-L1
upregulation through MYC in TNBC.
33
These observations are in
alignment with our findings. Furthermore, Aurora A kinase is known
to play a vital role in cellular proliferation, including spindle assembly
and chromosomal segregation during mitosis,
50
centrosome matura-
tion,
51
and the integrity of astral microtubules.
52,53
Additionally,
pharmacological inhibition of AURKA has been shown to promote
mitotic arrest and induce apoptosis in pancreatic adenocarcinoma.
54
Intriguingly, our screening did not identify known kinases such as
TBK1, which is recognized for its role in tumor immunogenicity.
55
This contrasts with a recent study by Sun et al., which utilized a mouse
whole-genome knockout library in an in vivo screening platform to
identify TBK1 as an immunotherapy resistance gene in melanoma.
56
Sun et al. employed a syngeneic melanoma mouse model with intact
innate immune cells, including B cells, dendritic cells, macrophages,
and stromal cells, in the tumor microenvironment. These immune
components were absent in our system, as we utilized human-derived
PBMCs and breast cancer cells in an in vitro screening platform.
Additionally, TBK1 genetic and pharmacological inhibition has
Figure 4. CDC42BPB upregulates PD-L1 expression and is associated with immune evasion
(A) MCF10CA1a cells were ectopically expressed with CDC42BPB. Total RNA was isolated, and PD-L1 mRNA was examined by real-time PCR (n= 3/group). Protein
expression was examined by western blot. GAPDH was used as an internal control. Statistical significance was calculated using the unpaired two-tailed Student’s t test. (B)
MCF10CA1a cells were ectopically expressed with CDC42BPB. Total protein was isolated and pAURKA protein expression was examined by western blot. GAPDH was
used as an internal control. (C) MCF10CA1a cells were ectopically expressed with CDC42BPB. Total RNA was isolated and cMYC mRNA expression examined by real-time
PCR (n= 3/group). Protein expression was examined by western blot. GAPDH was used as an interna l control. Statistical significance between the groups was calculated
using the unpaired two-tailed Student’s t test. (D and E) MCF10CA1a cells transfected with CDC42BPB siRNA. Total RNA was isolated, and the expression of cMYC and PD-
L1 was examined by real-time PCR (n= 3/group). Statistical significance between the groups was calcu lated using the unpaired two-tailed Student’s t test. (F) MCF10CA1a
cells were transfected with CDC42BPB siRNA. Total protein was isolated, and a western blot was performed for pAURKA, cMYC, and PD-L1. GAPDH was used as an
internal control. (G) CDC42BPB was ectopically expressed in MCF10CA1a cells. The cells were then treated with the AURKA inhibitor (AURKAi, 1 mM) for 24 h in the presence
or absence of ectopic cMYC expression. Cell lysates were prepared, and PD-L1 protein expression was examined by western blot. GAPDH was used as an internal control.
(H) Scatterplot analysis showing CRISPR CARE score correlation between CDC42BPB and PD-L1 in 32 cell lines (Gecko). (I–L) The protein expression of pAURKA (I and J)
and cMYC (K and L) was examined in the tumor tissues from the animals treated with control, anti-PD-1, BDP5290, and anti-PD-1 + BDP5290 from the experiment
performed in Figure 3 by immunohistochemistry. One-way ANOVA with Tukey’s multiple comparison post hoc test was used to determine statistica l significance between the
groups. Scale bars, 100 mm. (M) pAURKA expression was examined by immunohistochemist ry in formalin-fixed paraffin-embedded tissues from breast cancer patients who
either responded (R) or not (NR) to immunotherapy. Scale bars, 100 mm. (N) Quantification of pAURKA protein expression by immunohistochem istry in non-responders (n=
16) and responders (n= 23). Statistical significance was calculated using the unpaired two-tailed Student’s t test. (O) CD8 T cell infiltration was examined by immunohis-
tochemistry in formalin-fixed paraffin-embedded tissues from breast cancer patients who either responded (R) or not (NR) to immunotherapy. Scale bar, 100 mm. (P) Stained
intratumoral CD8 T cells in random high-power fields were quantified in breast cancer non-responders (n= 16) and responders (n= 23). Statistical significance was calculated
using the unpaired two-tailed Student’s t test. (Q) Proposed mechanism of action. All western blots were quantified with ImageJ. Data presented as mean ±SEM. *p< 0.05,
**p< 0.01, ***p< 0.001, ****p< 0.0001.
www.moleculartherapy.org
Molecular Therapy Vol. 32 No 10 October 2024 9
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

been shown to inhibit cellular proliferation in bladder cancer.
57
The
CRISPR knockout approach used in our screening might have
severely affected the proliferation of cells with AURKA and TBK1
knockouts, potentially explaining their absence in our results. There
are advantages and disadvantages of targeting CDC42BPB in combi-
nation with anti-PD-1 immune checkpoint blockade. First,
CDC42BPB was found to synergize with anti-PD-1 immune check-
point blockade in breast, lung, and melanoma, as shown in Figure 2.
Second, CDC42BPB is expressed in many cancer types including
breast cancer, lung cancer, and melanoma (Figure S1N). Third,
CDC42BPB expression was found to be associated with poor prog-
nosis in breast cancer, lung adenocarcinoma, and lung squamous
cell carcinoma (Figures S1H–S1K). Lastly, CDC42BPB inhibition
was found to sensitize immunologically cold tumors to anti-PD-1 im-
mune checkpoint blockade without any toxic effects (Figure 3). On
the other hand, inhibiting CDC42BPB may have some limitations.
First, CDC42BPB is upstream of genes including ETS1 and MZF1
58
that are known to play a role in cell growth and division.
59
Inhibiting
CDC42BPB with BDP5290 may affect cell growth to some extent in
the normal cells. Second, BDP50290 has a molecular weight of
371.82 and may cross the blood-brain barrier. This factor should be
taken into account before clinical use, as it may have an effect on
the normal cells in the brain. Third, BDP5290 in combination with
anti-PD-1 immune checkpoint blockade may systemically activate
the T cells in patients and cause hyperactive immune reactions. Lastly,
the dose of BDP5290 to be used in patients is not yet standardized and
needs to be studied for the potential toxicity.
In summary, we used a CRISPR-based kinome library screening sys-
tem to identify tumor cell-intrinsic kinase conferring anti-PD-1
immunotherapy resistance in breast cancer. Our findings have direct
clinical implications, as the identified small-molecule inhibitor
BDP5290 showed synergistic efficacy in our preclinical experimental
system and therefore serves as an attractive therapeutic reagent in
combination with anti-PD-1 immune checkpoint blockade.
MATERIALS AND METHODS
Cell lines, plasmids, and reagents
MCF10CA1a, MDA-MB-231, E0771, LL2, SKMEL-28, and 4T1 cell
lines were purchased from ATCC. MCF10CA1a, E0771, SKMEL-
28, and 4T1 cells were cultured in RPMI medium while the MDA-
MB-231 and LL2 cells were cultured in DMEM. RPMI-1640 medium
or DMEM was supplemented with 10% fetal bovine serum (FBS),
penicillin (100 U/mL), and streptomycin (100 mg/mL). The cells
were grown at 37C in a 5% CO
2
atmosphere and routinely tested
for mycoplasma. CDC42BPB and cMYC plasmids were obtained
from Addgene. Plasmid transfections were performed with Lipofect-
amine 3000, while the siRNAs were transfected with RNAiMAX
transfection reagent as per the manufacturer’s instructions. The sour-
ces of materials are listed in Table S2.
Cell-proliferation assay
To assay cell proliferation, 1,000 cells were seeded in a 96-well plate
overnight. The cells were then treated with the indicated dose of
the inhibitor and incubated for 24–48 h. The dead cells were washed
off with ice-cold PBS, and adherent cells were fixed with methanol at
room temperature for 20 min. The cells were then stained with 0.5%
crystal violet, and the dye was dissolved in 10% acetic acid. The read-
outs were obtained at 590 nm with a plate reader.
T cell isolation, in vitro education, and expansion
PBMCs derived from healthy donors were purchased from BioIVT.
Monocytes in the PBMCs were differentiated to dendritic cells
(DCs) by culturing them in DMEM supplemented with 10% FBS,
25 ng/mL GM-CSF, and 5 ng/mL IL-4 for 6 days as described previ-
ously.
60,61,62,63,64
The culture media and cytokines were replaced
every 3 days. DCs were matured by culturing them in the presence
of 500 U/mL IFN-gand 100 ng/mL lipopolysaccharide for 48 h
and then pulsed with cell lysate (100 mg/mL). The pulsed DCs were
co-cultured with PBMCs in the presence of IL-2 (20 U/mL) for
7 days. The educated T cells in the PBMCs were then activated and
expanded with human CD3/CD28 activation beads. To educate the
murine CD8 T cells with E0771 and 4T1 cells, syngeneic (C57BL/6
for E0771, LL2 and BALB/C for 4T1) mice were implanted with 1
million cells on the fourth mammary fat pad. After 3 weeks, spleno-
cytes were isolated post red blood cell lysis as per the manufacturer’s
instructions. CD8 T cells were separated from the splenocytes
using the CD8 T cell isolation kit and expanded with CD3/28
microbeads in RPMI medium supplemented with 10% FBS and
10 U/mL IL-2.
CRISPR screening
Lentivirus for Brunello human kinome CRISPR knockout library
(3,052 unique sgRNAs targeting 763 human kinase genes for four
guides per target) was prepared by Duke University Genomics Core
(Durham, NC). MCF10CA1a cells were transduced with the viral li-
brary at a multiplicity of infection of 0.2 so that the majority of cells
were transduced with one sgRNA and then cultured with 3 mg/mL pu-
romycin. To validate the quality of the library, DNA was isolated from
seven representative single-cell clones, and PCR was performed with
gRNA-specific primers. The PCR conditions were: 98C for 30 s, fol-
lowed by 30 rounds of 98C for 5 s, 64C for 10 s, and 72C for 15 s,
with a final extension at 72C for 5 min. The cells transduced with the
MCF10CA1a-kinome library were then co-cultured with in vitro-
educated PBMCs at a 1:5 ratio (cancer cells/PBMCs) in triplicate
for 48 h in the presence or absence of 40 mg/mL human anti-PD-1.
The cells were then washed with ice-cold PBS to dislodge dead cells
and PBMCs, and DNA was isolated using the genomic DNA isolation
kit (Zymo Research). PCR amplification and sequencing were per-
formed to determine sgRNA abundance. After CRISPR screening,
the MAGeCK algorithm was used to screen all four depleted sgRNAs
with significant pvalue (<0.05). In addition, the fold change in sgRNA
enrichment in the anti-PD-1 immunotherapy-treated set was calcu-
lated after adding pseudo count 1 in both groups. We then selected
four kinase genes, PTK2, CDC42BPB, NAGK, and ITPKA, which
all met the two criteria: (1) all four sgRNAs depleted with significant
pvalue; and (2) 3-fold enrichment in the experimental set of anti-PD-
1 immunotherapy treatment. To validate our results, we knocked
Molecular Therapy
10 Molecular Therapy Vol. 32 No 10 October 2024
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

down each gene using siRNA and examined the effect on anti-PD-1-
PBMC treatment.
Generation of anti-PD-1 immune-checkpoint-resistant cells
MCF10CA cells that are resistant to anti-PD-1 ICIs were generated by
culturing them with cell-lysate-pulsed PBMCs at the indicated
effector/target ratio in the presence of 40 mg/mL anti-human PD-1
antibody for 48 h. Dead cells and PBMCs were washed out with
ice-cold PBS. Surviving cells were again co-cultured to obtain cells
non-responsive to anti-PD-1 treatment. These non-responsive and
parental cells were used for co-culture as described in the text.
In vitro synergy
To examine the synergy between the anti-PD-1 immune checkpoint
blockade and the small-molecule pharmacological inhibitor
BDP5090, 500 cancer cells were seeded in a 96-well plate for 2 h.
They were then treated with PBMCs or CD8 T cells that were isolated
from the mouse spleen in the presence or absence of 40 mg/mL anti-
PD-1 immune checkpoint blockade at the indicated dose of BDP5290.
After 48–72 h, PBMCs/T cells and the dead cells were washed out
with ice-cold PBS. The surviving cells were then fixed in methanol
for 20 min and stained with 0.5% crystal violet. The dye was dissolved
in 10% acetic acid, and readouts were measured at 590 nm. The syn-
ergy was calculated by combination index method with CompuSyn.
65
Flow cytometry
The reactivity of T cells was evaluated as described previously.
66
In
brief, PBMCs were stimulated with CD3/CD28 beads in the presence
of IL-2 (20 U/mL) for 24 h and washed with PBS. The cells were
stained with CD3, CD8, and IFN-gantibodies at 4C for 30 min
and analyzed by flow cytometry (Accuri-BD). For caspase3/7 assay,
the cells were washed with PBS to dislodge PBMCs and incubated
with 5 mM CellEvent Caspase3/7 green detection reagent for 1 h at
37C. After incubation, cells were trypsinized and analyzed by flow
cytometry (BD Canto II).
Real-time PCR
Total RNA was isolated using the Direct-zol RNA Miniprep Kit as per
the manufacturer’s instructions. The purity of isolated RNA was
examined with a NanoDrop spectrophotometer, and 200 ng of
RNA was used for cDNA synthesis using the iScript cDNA synthesis
kit. The cDNA was then used for SYBR green-based real-time PCR.
The results were analyzed with the relative quantification method.
GAPDH was used as an internal control. The list of primer sequences
is provided in Table S3.
Immunohistochemistry
For immunohistochemical analysis, the tumor tissues were fixed in
formalin followed by paraffin embedding. Sections were cut and de-
paraffinized by heating at 100C for 30 min followed by dehydration
and rehydration in xylene and grades of ethanol. Antigen retrieval
was performed in sodium citrate buffer (10 mM, pH 6). Thereafter,
slides were washed in PBS and blocked with 2% BSA for 1 h followed
by incubation with primary antibodies for CD4 (1:200), CD8 (1:200),
pAURKA (1:200), cMYC (1:200), CDC42BPB (1:200), and the
respective secondary antibodies (1:400). The sections were washed
with PBS and visualized with a horseradish peroxidase-based detec-
tion kit. The percentage of pAURKA-, cMYC-, and CDC42BPB-pos-
itive cells was estimated by analyzing the staining intensity (low,
medium, and high). The T cells were estimated by calculating the
cell number in a high-magnification field.
Western blot
Cell lysates were prepared using RIPA buffer. The proteins in the ly-
sates were quantified using the Bradford assay, and an equal amount
was loaded on 10% acrylamide gel for separation. The proteins were
transferred to the nitrocellulose membrane by semi-dry method (100
mA, 2 h). The membrane was then blocked with nonfat skim milk
powder dissolved in PBS followed by incubation with primary
antibodies for CDC42BPB (1:500), PD-L1 (1:500), cMYC (1:500),
pAURKA (1:500), and GAPDH (1:2,000) overnight at 4C. Blots
were incubated with a secondary antibody (1:5,000) and developed
with West Pico PLUS Chemiluminescent Substrate and Amersham
Imager. The information on antibody providers is listed in Table S2.
In vitro kinase assay
The in vitro kinase assay was performed as described previously.
67
In
brief, the 100 ng of recombinant CDC42BPB and AURKA proteins
were mixed in kinase buffer supplemented with 200 mM ATP at desig-
nated time points for 30C. The reaction was terminated by adding
SDS-PAGE sample buffer. The denatured proteins were separated
by 10% acrylamide gel and stained with pAURKA primary antibody
(1:1,000) overnight at 4C. The blots were washed and incubated with
secondary antibody (1:3,000) at room temperature for 2 h and devel-
oped with Amersham imager.
Animal experiments
Animal experiments were performed in line with a protocol approved
by Wake Forest Institutional Animal Care and Use Committee (Refer-
ence ID: A21-132). Animals were housed in a temperature-controlled
and pathogen-free environment with a 12-h light/12-h dark cycle and
free access to food and water. E0771 cells (50,000) were suspended
in Matrigel and injected into the mammary fat pad of C57BL/6 mice
(n= 5/group). Animals were then treated with BDP5290 (5 mg/kg,
through i.p., 3/week), anti-PD-1 (200 mg/mouse, through i.p.,
1/week), or a combination of both. For the 4T1 experiment, 20,000
cells were injected into mammary fat pads of BALB/C mice (n=
6/group). Animals were then treated with anti-PD-1 or isotype control
as indicated. Tumor volume was measured using the formula LðW2Þ
2,
where Lis the length and Wis the width of tumor. All animals were
randomized before the experiment and sacrificed when the human
endpoint was reached. Toxicity assessment was performed by
measuring serum AST and the animal body weight at the endpoint.
Human tissue samples
Human breast cancer tissues were obtained from Atrium Health Wake
Forest Baptist Comprehensive Cancer Center and BioRepository &
Precision Pathology Center, Duke University School of Medicine,
www.moleculartherapy.org
Molecular Therapy Vol. 32 No 10 October 2024 11
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

Durham, NC. The tumor tissue samples were surgically resected before
immunotherapy treatment. The patients were classified as responders
if they did not observe progression 6 months after the start of treat-
ment. Details of patients’information is provided in Table S4.
Statistical analysis
All of the statistical analysis was performed using GraphPad Prism
(version 7.0). Data were expressed as mean ±SEM. The pvalue
was calculated by an unpaired Student’s t test or ANOVA as indi-
cated. Significance between each group was represented as p< 0.05.
SUPPLEMENTAL INFORMATION
Supplemental information can be found online at https://doi.org/10.
1016/j.ymthe.2024.07.021.
ACKNOWLEDGMENTS
This work was supported by grants R01CA173499, R01CA185650,
and R01CA205067 from NIH and W81XWH2110075 from the
Department of Defense (to K.W.). This study also utilized the Cancer
Center Shared Resources including Cancer Genomics, Tumor Tissue
and Pathology, Cell Engineering, Flow Cytometry, and Biostatistics
and Bioinformatics that are supported by the Comprehensive Cancer
Center of Wake Forest University NCI, National Institutes of Health
grant (P30CA012197).
AUTHOR CONTRIBUTIONS
R.P.D. and K. Watabe designed the study and wrote the manuscript.
R.P.D. conducted all experiments and acquired, analyzed, and inter-
preted the data. J.H., J.R., and W.L. helped with acquisition of immu-
notherapy-treated patients’samples and curation of pathological
slides. R.P.D., K. Wu, S.-Y.W., A.T., E.C.S., and K. Watabe reviewed
and edited the manuscript and interpreted the data. K. Watabe super-
vised the study.
DECLARATION OF INTERESTS
The authors declare no competing interests.
REFERENCES
1. Mediratta, K., El-Sahli, S., D’Costa, V., and Wang, L. (2020). Current Progresses and
Challenges of Immunotherapy in Triple-Negative Breast Cancer. Cancers (Basel) 12,
3529. https://doi.org/10.3390/cancers12123529.
2. Waks, A.G., and Winer, E.P. (2019). Breast Cancer Treatment: A Review . JAMA 321,
288–300. https://doi.org/10.1001/jama.2018.19323.
3. Garrido-Castro, A.C., Lin, N.U., and Polyak, K. (2019). Insights into Molecular
Classifications of Triple-Negative Breast Cancer: Improving Patient Selection for
Treatment. Cancer Discov. 9, 176–198. https://doi.org/10.1158/2159-8290.Cd-
18-1177.
4. Kumar, P., and Aggarwal, R. (2016). An overview of triple-negative breast cancer.
Arch. Gynecol. Obstet. 293, 247–269. https://doi.org/10.1007/s00404-015-3859-y.
5. Li, C.H., Karantza, V., Aktan, G., and Lala, M. (2019). Current treatment landscape
for patients with locally recurrent inoperable or metastatic triple-negative breast can-
cer: a systematic literature review. Breast Cancer Res. 21, 1 43. https://doi.org/10.1186/
s13058-019-1210-4.
6. Nedeljkovi
c, M., and Damjanovi
c, A. (2019). Mechanisms of Chemotherapy
Resistance in Triple-Negative Breast Cancer-How We Can Rise to the Challenge.
Cells 8, 957. https://doi.org/10.3390/cells8090957.
7. Al-Mahmood, S., Sapiezynski, J., Garbuze nko, O.B., and Minko, T. (2018). Metastatic
and triple-negative breast cancer: challenges and treatment options. Drug Deliv.
Transl. Res. 8, 1483–1507. https://doi.org/10.1007/s13346-018-0551-3.
8. Khosravi-Shahi, P., Cabezón-Gutiérrez, L., and Custodio-Cabello, S. (2018).
Metastatic triple negative breast cancer: Optimizing treatment options, new and
emerging targeted therapies. Asia. Pac. J. Clin. Oncol. 14,32–39. https://doi.org/10.
1111/ajco.12748.
9. Mahmoud, R., Ordóñez-Morán, P., and Allegrucci, C. (2022). Challenges for Triple
Negative Breast Cancer Treatment: Defeating Heterogeneity and Cancer Stemness.
Cancers (Basel) 14, 4280. https://doi.org/10.3390/cancers14174280.
10. Deshpande, R.P., Sharma, S., and Watabe, K. (2020). The Confounders of Cancer
Immunotherapy: Roles of Lifestyle, Metabolic Disorders and Sociological Factors.
Cancers (Basel) 12, 2983. https://doi.org/10.3390/cancers12102983.
11. Young, J.A., and Tan, A.R. (2021). Targeted Treatment of Triple-Negative Breast
Cancer. Cancer J. 27,50–58. https://doi.org/10.1097/ppo.0000000000000495.
12. Hanna, A., and Balko, J.M. (2021). Breast cancer resistance mechanisms: challenges
to immunotherapy. Breast Cancer Res. Treat. 190,5–17. https://doi.org/10.1007/
s10549-021-06337-x.
13. Keenan, T.E., and Tolaney, S.M. (2020). Role of Immunotherapy in Triple-Negative
Breast Cancer. J. Natl. Compr. Canc. Netw. 18, 479–489. https://doi.org/10.6004/
jnccn.2020.7554.
14. Won, K.A., and Spruck, C. (2020). Triple-negative breast cancer therapy: Current and
future perspectives (Review). Int. J. Oncol. 57,1245–1261.https://doi.org/10.3892/ijo.
2020.5135.
15. Henriques, B., Mendes, F., and Martins, D. (2021). Immunotherapy in Breast Cancer:
When, How, and What Challenges? Biomedicines 9, 1687. https://doi.org/10.3390/
biomedicines9111687.
16. Escors, D., Gato-Cañas, M., Zuazo, M., Arasanz, H., García-Granda, M.J., Vera, R.,
and Kochan, G. (2018). The intracellular signalosome of PD-L1 in cancer cells.
Signal Transduct. Targeted Ther. 3, 26. https://doi.org/10.1038/s41392-018-0022-9.
17. Brea, E.J., Oh, C.Y., Manchado, E., Budhu, S., Gejman, R.S., Mo, G., Mondello, P.,
Han, J.E., Jarvis, C.A., Ulmert, D., et al. (2016). Kinase Regulation of Human MHC
Class I Molecule Expression on Cancer Cells. Cancer Immunol. Res. 4, 936–947.
https://doi.org/10.1158/2326-6066.Cir-16-0177.
18. Mehlich, D., and Marusiak, A.A. (2022). Kinase inhibitors for precision therapy of
triple-negative breast cancer: Progress, challenges, and new perspectives on targeting
this heterogeneous disease. Cancer Lett. 547, 215775. https://doi.org/10.1016/j.canlet.
2022.215775.
19. Belk, J.A., Yao, W., Ly, N., Freitas, K.A., Chen, Y.T., Shi, Q., Valencia, A.M., Shifrut,
E., Kale, N., Yost, K.E., et al. (2022). Genome-wide CRISPR screens of T cell exhaus-
tion identify chromatin remodeling factors that limit T cell persistence. Cancer Cell
40, 768–786.e7. https://doi.org/10.1016/j.ccell.2022.06.001.
20. Quan, M., Oh, Y., Cho, S.Y., Kim, J.H., and Moon, H.G. (2022). Polo-Like Kinase 1
Regulates Chromosomal Instability and Paclitaxel Resistance in Breast Cancer Cells.
J. Breast Cancer 25, 178–192. https://doi.org/10.4048/jbc.2022.25.e28.
21. Ozmadenci, D., Shankara Narayanan, J.S., Andrew , J., Ojalill, M., Barrie, A.M., Jiang,
S., Iyer, S., Chen, X.L., Rose, M., Estrada, V., et al. (2022). Tumor FAK orchestrates
immunosuppression in ovarian cancer via the CD155/TIGIT axis. Proc. Natl.
Acad. Sci. USA 119, e2117065119. https://doi.org/10.1073/pnas.2117065119.
22. Symeonides, S.N., Anderton, S.M., and Serrels, A. (2017). FAK-inhibition opens the
door to checkpoint immunotherapy in Pancreatic Cancer. J. Immunother. Cancer 5,
17. https://doi.org/10.1186/s40425-017-0217-6.
23. Uhlén, M., Fagerberg, L., Hallström, B.M., Lindskog, C., Oksvold, P., Mardinoglu, A.,
Sivertsson, Å., Kampf, C., Sjöstedt, E., Asplund, A., et al. (2015). Proteomics. Tissue-
based map of the human proteome. Science 347, 1260419. https://doi.org/10.1126/
science.1260419.
24. }
Osz, Á., Lánczky, A., and Gy}
orffy, B. (2021). Survival analysis in breast cancer using
proteomic data from four independent datasets. Sci. Rep. 11, 16787. https://doi.org/
10.1038/s41598-021-96340-5.
25. Kovács, S.A., Fekete, J.T., and Gy}
orffy, B. (2023). Predictive biomarkers of immuno-
therapy response with pharmacological applications in solid tumors. Acta Pharmacol.
Sin. 44, 1879–1889. https://doi.org/10.1038/s41401-023-01079-6.
Molecular Therapy
12 Molecular Therapy Vol. 32 No 10 October 2024
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

26. Chen, Z., Luo, Z., Zhang, D., Li, H., Liu, X., Zhu, K., Zhang, H., Wang, Z., Zhou, P.,
Ren, J., et al. (2023). TIGER: A Web Portal of Tumor Immunotherapy Gene
Expression Resource. Genomics Proteomics Bioinformatics 21, 337–348. https://
doi.org/10.1016/j.gpb.2022.08.004.
27. Unbekandt, M., Croft, D.R., Crighton, D., Mezna, M., McArthur, D., McConnell, P.,
Schüttelkopf, A.W., Belshaw, S., Pannifer, A., Sime, M., et al. (2014). A novel small-
molecule MRCK inhibitor blocks cancer cell invasion. Cell Commun. Signal. 12,54.
https://doi.org/10.1186/s12964-014-0054-x.
28. Rolf, M.G., Curwen, J.O., Veldman-Jones, M., Eberlein, C., Wang, J., Harmer, A.,
Hellawell, C.J., and Braddock, M. (2015). In vitro pharmacological profiling of
R406 identifies molecular targets underlying the clinical effects of fostamatinib.
Pharmacol. Res. Perspect. 3, e00175. https://doi.org/10.1002/prp2.175.
29. Bucsek, M.J., Qiao, G., MacDonald, C.R., Giridharan, T., Evans, L., Niedzwecki, B.,
Liu, H., Kokolus, K.M., Eng, J.W.L., Messmer, M.N., et al. (2017). b-Adrenergic
Signaling in Mice Housed at Standard Temperatures Suppresses an Effector
Phenotype in CD8(+) T Cells and Undermines Checkpoint Inhibitor Therapy.
Cancer Res. 77, 5639–5651. https://doi.org/10.1158/0008-5472.Can-17-0546.
30. Ye, Q.F., Zhang, Y.C., Peng, X.Q., Long, Z., Ming, Y.Z., and He, L.Y. (2012). Silencing
Notch-1 induces apoptosis and increases the chemosensitivity of prostate cancer cells
to docetaxel through Bcl-2 and Bax. Oncol. Lett. 3, 879–884. https://doi.org/10.3892/
ol.2012.572.
31. Bai, Z., Zhang, Z., Qu, X., Han, W., and Ma, X. (2012). Sensitization of breast cancer
cells to taxol by inhibition of taxol resistance gene 1. Oncol. Lett. 3, 135–140. https://
doi.org/10.3892/ol.2011.416.
32. Ngamcherdtrakul, W., and Yantasee, W. (2019). siRNA therapeutics for breast can-
cer: recent efforts in targeting metastasis, drug resistance, and immune evasion.
Transl. Res. 214, 105–120. https://doi.org/10.1016/j.trsl.2019.08.005.
33. Sun, S., Zhou, W., Li, X., Peng, F., Yan, M., Zhan, Y., An, F., Li, X., Liu, Y., Liu, Q., and
Piao, H. (2021). Nuclear Aurora kinase A triggers programmed death-ligand
1-mediated immune suppression by activating MYC transcription in triple-negative
breast cancer. Cancer Commun. 41, 851–866. https://doi.org/10.1002/cac2.12190.
34. Lee, D.H., Kim, C.G., Lim, Y., and Shin, S.Y. (2017). Aurora kinase A inhibitor
TCS7010 demonstrates pro-apoptotic effect through the unfolded protein response
pathway in HCT116 colon cancer cells. Oncol. Lett. 14, 6571–6577. https://doi.org/
10.3892/ol.2017.7023.
35. Unbekandt, M., Belshaw, S., Bower, J., Clarke, M., Cordes, J., Crighton, D., Croft,
D.R., Drysdale, M.J., Garnett, M.J., Gill, K., et al. (2018). Discovery of Potent and
Selective MRCK Inhibitors with Therapeutic Effect on Skin Cancer. Cancer Res.
78, 2096–2114. https://doi.org/10.1158/0008-5472.Can-17-2870.
36. Gaggioli, C., Hooper, S., Hidalgo-Carcedo, C., Grosse, R., Marshall, J.F., Harrington,
K., and Sahai, E. (2007). Fibroblast-led collective invasion of carcinoma cells with
differing roles for RhoGTPases in leading and following cells. Nat. Cell Biol. 9,
1392–1400. https://doi.org/10.1038/ncb1658.
37. Wang, X.X., Zhang, S., Dong, P.P., Li, Y.H., Zhang, L., Shi, S.H., Yu, Z.Q., and Chen, S.
(2021). MRCKblinks Dasm1 to actin rearrangements to promote dendrite develop-
ment. J. Biol. Chem. 296, 100730. https://doi.org/10.1016/j.jbc.2021.100730.
38. Hou, J., Wang, Y., Shi, L., Chen, Y., Xu, C., Saeedi, A., Pan, K., Bohat, R., Egan, N.A.,
McKenzie, J.A., et al. (2021). Integrating genome-wide CRISPR immune screen with
multi-omic clinical data reveals distinct classes of tumor intrinsic immune regulato rs.
J. Immunother. Cancer 9, e001819. https://doi.org/10.1136/jitc-2020-001819.
39. Yin, N., Liu, Y., Weems, C., Shreeder, B., Lou, Y., Knutson, K.L., Murray, N.R., and
Fields, A.P. (2022). Protein kinase Cimediates immunosuppression in lung adeno-
carcinoma. Sci. Transl. Med. 14, eabq5931. https://doi.org/10.1126/scitranslmed.
abq5931.
40. Akiyama, T., Yasuda, T., Uchihara, T., Yasuda-Yoshihara, N., Tan, B.J.Y., Yonemura,
A., Semba, T., Yamasaki, J., Komohara, Y., Ohnishi, K., et al. (2023). Stromal
Reprogramming through Dual PDGFRa/bBlockade Boosts the Efficacy of Anti-
PD-1 Immunotherapy in Fibrotic Tumors. Cancer Res. 83, 753–770. https://doi.
org/10.1158/0008-5472.Can-22-1890.
41. Stutvoet, T.S., Kol, A., de Vries, E.G., de Bruyn, M., Fehrmann, R.S., Terwisscha van
Scheltinga, A.G., and de Jong, S. (2019). MAPK pathway activity plays a key role in
PD-L1 expression of lung adenocarcinoma cells. J. Pathol. 249,52–64. https://doi.org/
10.1002/path.5280.
42. Sow, H.S., Ren, J., Camps, M., Ossendorp, F., and Ten Dijke, P. (2019). Combined
Inhibition of TGF-bSignaling and the PD-L1 Immune Checkpoint Is
Differentially Effective in Tumor Models. Cells 8, 320. https://doi.org/10.3390/
cells8040320.
43. Yu, H.A., Riely, G.J., and Lovly, C.M. (2014). Therapeutic strategies utilized in the
setting of acquired resistance to EGFR tyrosine kinase inhibitors. Clin. Cancer Res.
20, 5898–5907. https://doi.org/10.1158/1078-0432.Ccr-13-2437.
44. Lynce, F., Shajahan-Haq, A.N., and Swain, S.M. (2018). CDK4/6 inhibitors in breast
cancer therapy: Current practice and future opportunities. Pharmacol. Ther. 191,
65–73. https://doi.org/10.1016/j.pharmthera.2018.06.008.
45. Kahl, I., Mense, J., Finke, C., Boller, A.L., Lorber, C., Gy}
orffy, B., Greve, B., Götte, M.,
and Espinoza-Sánchez, N.A. (2022). The cell cycle-related genes RHAMM, AURKA,
TPX2, PLK1, and PLK4 are associated with the poor prognosis of breast cancer pa-
tients. J. Cell. Biochem. 123, 581–600. https://doi.org/10.1002/jcb.30205.
46. Vilgelm, A.E., Johnson, C.A., Prasad, N., Yang, J., Chen, S.C., Ayers, G.D.,
Pawlikowski, J.S., Raman, D., Sosman, J.A., Kelley, M., et al. (2016). Connecting
the Dots: Therapy-Induced Senescence and a Tumor-Suppressive Immune
Microenvironment. J. Natl. Cancer Inst. 108, djv406. https://doi.org/10.1093/jnci/
djv406.
47. Maeda, T., Hiraki, M., Jin, C., Rajabi, H., Tagde, A., Alam, M., Bouillez, A., Hu, X.,
Suzuki, Y., Miyo, M., et al. (2018). MUC1-C Induces PD-L1 and Immune Evasion
in Triple-Negative Breast Cancer. Cancer Res. 78, 205–215. https://doi.org/10.
1158/0008-5472.Can-17-1636.
48. Liang, M.Q., Yu, F.Q., and Chen, C. (2020). C-Myc regulates PD-L1 expression in
esophageal squamous cell carcinoma. Am. J. Transl. Res. 12, 379–388.
49. Wang, X., Huang, J., Liu, F., Yu, Q., Wang ,R., Wang, J., Zhu, Z., Yu, J., Hou, J., Shim,
J.S., et al. (2023). Aurora A kinase inhibition compromises its antitumor efficacy by
elevating PD-L1 expression. J. Clin. Invest. 133, e161929. https://doi.org/10.1172/
jci161929.
50. Yu, F., Jiang, Y., Lu, L., Cao, M., Qiao, Y., Liu, X., Liu, D., Van Dyke, T., Wang, F., Yao,
X., et al. (2017). Aurora-A promotes the establishment of spindle assembly check-
point by priming the Haspin-Aurora-B feedback loop in late G2 phase. Cell
Discov. 3, 16049. https://doi.org/10.1038/celldisc.2016.49.
51. Hannak, E., Kirkham, M., Hyman, A.A., and Oegema, K. (2001). Aurora-A kinase is
required for centrosome maturation in Caenorhabditis elegans. J. Cell Biol. 155,
1109–1116. https://doi.org/10.1083/jcb.200108051.
52. Giet, R., McLean, D., Descamps, S., Lee, M.J., Raff, J.W., Prigent, C., and Glover, D.M.
(2002). Drosophila Aurora A kinase is required to localize D-TACC to centrosomes
and to regulate astral microtubules. J. Cell Biol. 156, 437–451. https://doi.org/10.1083/
jcb.200108135.
53. Carmena, M., Ruchaud, S., and Earnshaw, W.C. (2009). Making the Auroras glow:
regulation of Aurora A and B kinase function by interacting proteins. Curr. Opin.
Cell Biol. 21, 796–805. https://doi.org/10.1016/j.ceb.2009.09.008.
54. Duan, Z., Chinn, D., Tu, M.J., Zhang, Q.Y., Huynh, J., Chen, J., Mack, P., Yu, A.M.,
and Kim, E.J. (2019). Novel Synergistic Combination of Mitotic Arrest and
Promotion of Apoptosis for Treatment of Pancreatic Adenocarcinoma. Transl.
Oncol. 12, 683–692. https://doi.org/10.1016/j.tranon.2019.01.009.
55. Miranda, A., Shirley, C.A., and Jenkins, R.W. (2024). Emerging roles of TBK1 in can-
cer immunobiology. Trends Cancer 10, 531–540. https://doi.org/10.1016/j.trecan.
2024.02.007.
56. Sun, Y., Revach, O.Y., Anderson, S., Kessler, E.A., Wolfe, C.H., Jenney, A., Mills, C.E.,
Robitschek, E.J., Davis, T.G.R., Kim, S., et al. (2023). Targeting TBK1 to overcome
resistance to cancer immunotherapy. Nature 615, 158–167. https://doi.org/10.1038/
s41586-023-05704-6.
57. Chen, W., Luo, K., Ke, Z., Kuai, B., He, S., Jiang, W., Huang, W., and Cai, Z. (2017).
TBK1 Promote Bladder Cancer Cell Proliferation and Migration via Akt Signaling.
J. Cancer 8, 1892–1899. https://doi.org/10.7150/jca.17638.
58. Rafn, B., Nielsen, C.F., Andersen, S.H., Szyniarowski, P., Corcelle-Termeau, E., Valo,
E., Fehrenbacher, N., Olsen, C.J., Daugaard, M., Egebjerg, C., et al. (2012). ErbB2-
driven breast cancer cell invasion depends on a complex signaling network activating
myeloid zinc finger-1-dependent cathepsin B expression. Mol. Cell 45, 764–776.
https://doi.org/10.1016/j.molcel.2012.01.029.
www.moleculartherapy.org
Molecular Therapy Vol. 32 No 10 October 2024 13
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021

59. Baselga, J., and Swain, S.M. (2009). Novel anticancer targets: revisiting ERBB2 and
discovering ERBB3. Nat. Rev. Cancer 9, 463–475. https://doi.org/10.1038/nrc2656.
60. Su, S., Liao, J., Liu, J., Huang, D., He, C., Chen, F., Yang, L., Wu, W., Chen, J., Lin, L.,
et al. (2017). Blocking the recruitment of naive CD4(+) T cells reverses immunosup-
pression in breast cancer. Cell Res. 27, 461–482. https://doi.org/10.1038/cr.2017.34.
61. Kawahara, M., and Takaku, H. (2015). A tumor lysate is an effective vaccine antigen
for the stimulation of CD4(+) T-cell function and subsequent induction of antitumor
immunity mediated by CD8(+) T cells. Cancer Biol. Ther. 16, 1616–1625. https://doi.
org/10.1080/15384047.2015.1078027.
62. Hoffmann, T.K., Meidenbauer, N., Dworacki, G., Kanaya, H., and Whiteside, T.L.
(2000). Generation of tumor-specific T-lymphocytes by cross-priming with human
dendritic cells ingesting apoptotic tumor cells. Cancer Res. 60, 3542–3549.
63. Wen, Y.J., Min, R., Tricot, G., Barlogie, B., and Yi, Q. (2002). Tumor lysate-specific
cytotoxic T lymphocytes in multiple myeloma: promising effector cells for immuno-
therapy. Blood 99, 3280–3285. https://doi.org/10.1182/blood.v99.9.3280.
64. Schnurr, M., Galambos, P., Scholz, C., Then, F., Dauer, M., Endres, S., and Eigler, A.
(2001). Tumor cell lysate-pulsed human dendritic cells induce a T-cell response
against pancreatic carcinoma cells: an in vitro model for the assessment of tumor vac-
cines. Cancer Res. 61, 6445–6450.
65. Chou, T.C. (2010). Drug combination studies and their synergy quantification using
the Chou-Talalay method. Cancer Res. 70, 440–446. https://doi.org/10.1158/0008-
5472.Can-09-1947.
66. Dijkstra, K.K., Cattaneo, C.M., Weeber, F., Chalabi, M., van de Haar, J., Fanchi, L.F.,
Slagter, M., van der Velden, D.L., Kaing, S., Kelderman, S., et al. (2018). Generation of
Tumor-Reactive T Cells by Co-culture of Peripheral Blood Lymphocytes and Tumor
Organoids. Cell 174, 1586–1598.e12. https://doi.org/10.1016/j.cell.2018.07.009.
67. Cao, H., Dzineku, F., and Blackshear, P.J. (2003). Expression and purification of re-
combinant tristetraprolin that can bind to tumor necrosis factor-alpha mRNA and
serve as a substrate for mitogen-activated protein kinases. Arch. Biochem. Biophys.
412, 106–120. https://doi.org/10.1016/s0003-9861(03)00012-2.
Molecular Therapy
14 Molecular Therapy Vol. 32 No 10 October 2024
Please cite this article in press as: Deshpande et al., Tumor-intrinsic CDC42BPB confers resistance to anti-PD-1 immune checkpoint blockade in breast
cancer, Molecular Therapy (2024), https://doi.org/10.1016/j.ymthe.2024.07.021