Biosensors and Bioelectronics 166 (2020) 112452

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

journal homepage: [Link]

Identification of programmed death ligand-1 positive exosomes in breast

cancer based on DNA amplification-responsive metal-organic frameworks
Ya Cao a, Ying Wang a, b, Xiaomeng Yu a, Xihui Jiang a, c, Gang Li b, **, Jing Zhao a, *
a
Center for Molecular Recognition and Biosensing, School of Life Sciences, Shanghai University, 200444, PR China
b
Minhang Branch, Fudan University Shanghai Cancer Center, Shanghai, 200240, PR China
c
Department of Medical Science and Technology, Suzhou Chien-shiung Institute of Technology, Taicang, 215411, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: Cancer-derived exosomes have recently emerged as potent candidates for diagnosis and prognosis of breast
Metal-organic framework cancer. As an example, programmed death ligand-1 positive (PD-L1þ) exosomes are found to be correlated with
Hyperbranched rolling circle amplification the progression and immunotherapy response of breast cancer, and therefore show great potential in liquid
Breast cancer
biopsy. Herein, we propose an electrochemical biosensing method for accurate identification of PD-L1þ exosomes
Programmed death ligand-1
Electrochemical technique
by using DNA amplification-responsive metal-organic frameworks, PVP@HRP@ZIF-8. Specially, PD-L1þ exo­
somes are captured by anti-CD63 functionalized magnetic beads and bound with anti-PD-L1-linked capture
probe. Then, in situ hyperbranched rolling circle amplification, a typical DNA amplification reaction, is con­
ducted using the surface-attached capture probes as primers, which lows environmental pH. As a result, disas­
sembly of PVP@HRP@ZIF-8 takes place, leading to the release of enzymes, which can arouse amplified
electrochemical responses for the identification of target exosomes. Experimental results reveal that the bio­
sensing method displays a linear range for PD-L1þ exosomes identification from 1 � 103 to 1 � 1010 particles/mL
and the detection limit reaches 334 particles/mL. What is more, by using the method, elevated level of circu­
lating PD-L1þ exosomes is found in the undiluted serum samples from patients with breast cancer, particularly
for metastatic breast cancer, revealing a positive correlation of the PD-L1þ exosome level with the tumor staging
and disease progression of breast cancer. Therefore, the biosensing method may be valuable for not only exosome
identification but also providing reference information for diagnosis and real-time monitoring of breast cancer in
the future.

1. Introduction et al., 2015). Cancer-derived exosomes are also gaining increasing

attention in liquid biopsy, and represent the superiority in absolute
Cell-secreted exosomes, a type of extracellular vesicles, are produced quantity over circulating tumor cells that are extremely rare even at the
through the intracellular trafficking routine and widely distributed in late stage of cancers and provide more information than cell-free nucleic
body fluids after releasing from cell membrane, especially blood (Kalluri acids (Fitts et al., 2019; Poudineh et al., 2018). Moreover,
et al., 2020). The nano-sized vesicles contain typical phospholipid surface-expressed proteins on exosomes have close relationship with the
bilayer structures and carry abundant biological molecules derived from original tumors and may reflect abundant biological information for
the original cells (Jeppesen et al., 2019). Normally, cell-secreted exo­ precious management of the tumors. For instance, programmed death
somes engage in multiple physiological processes relating to cellular ligand-1 (PD-L1), an important immune checkpoint molecule, is found
communications, and differentiation and development of organisms to be expressed on exosomes derived from breast cancer, the most
(Mathieu et al., 2019; Thery et al., 2020). Besides, exosomes are found in common malignant tumor among women worldwide (Chen et al.,
several pathological processes. For example, emerging evidences reveal 2018a; Yang et al., 2018). The level of PD-L1 positive (PD-L1þ) exo­
that cancer-derived exosomes are associated with tumorigenesis, inva­ somes is potentially a prognostic prediction marker of anti-PD-1
sion, metastasis and drug resistance in tumors (Wortzel et al., 2019; Yu immunotherapy, which may be critical for cancer management

* Corresponding author.
** Corresponding author.
E-mail addresses: 13391071510@[Link] (G. Li), jingzhao@[Link] (J. Zhao).

[Link]
Received 3 June 2020; Accepted 13 July 2020
Available online 26 July 2020
0956-5663/© 2020 Elsevier B.V. All rights reserved.
Y. Cao et al. Biosensors and Bioelectronics 166 (2020) 112452

Scheme 1. Schematic illustration of identification of PD-L1þ exosomes based on HRCA-responsive PVP@HRP@ZIF-8.

(Daassi et al., 2020; Huang et al., 2020; Tang et al., 2020). Whereas, dynamically changing environmental pH under a low-buffered condi­
determination of PD-L1þ exosome level is still challenging in breast tion (Hua et al., 2019; Mao et al., 2019; Toumazou et al., 2013).
cancer, mostly due to the limited biosensing method for accurate iden­ Combining DNA amplification-promoted pH changes and pH-responsive
tification of PD-L1þ exosomes (Boriachek et al., 2018; Cheng et al., protein-encapsulating ZIF-8, we herein reported a biosensing method for
2019). accurate identification of PD-L1þ exosomes in breast cancer by virtue of
Metal-organic frameworks (MOFs) are currently the most attractive electrochemical technique. To best of our knowledge, it is the first time
porous nanomaterials for biosensing application (Fang et al., 2018). The to quantify PD-L1þ exosomes in the blood samples of breast cancer pa­
nanoscaled inorganic-organic hybrids exhibit several distinct advan­ tients, which reveals a potential correlation between the PD-L1þ exo­
tages over other existing nanomaterials, including adjustable structure some level and the advanced stage and poor prognosis of breast cancer.
and surface functionalization for diverse functions, large pore volumes Scheme 1 illustrates the principle of the method. pH-responsive
for high loading capabilities, and biodegradable and biocompatible protein-encapsulating ZIF-8 (PVP@HRP@ZIF-8) were prepared by
nanoformula for in vivo transportation. Encouraged by biomineraliza­ encapsulation of horseradish peroxidase (HRP) into ZIF-8 and subse­
tion, multiple biological molecules (e.g. biomacromolecules and cells) quent surface coating by polyvinylpyrrolidone (PVP), which were stable
are encapsulated into MOFs, thus MOFs are popular for intracellular at weak alkaline pH and disassembled at acidic pH. PD-L1þ exosomes
delivery of nucleic acids and proteins (Cheng et al., 2018; Wang et al., derived from breast cancer were first accumulated on the surface of anti-
2019). More interestingly, the disassembly of MOFs can be modulated CD63 functionalized magnetic beads (anti-CD63@MBs) due to the se­
by different endogenous or exogenous stimulus, such as pH, ions, tem­ lective recognition of the established exosomal biomarker, CD63, and
perature, pressure, redox and magnetic field (Chen et al., 2018b; Yang also specially bound with anti-PD-L1-linked DNA strand, namely cap­
et al., 2019; Zhong et al., 2019; Zhou et al., 2018). For instance, zeolitic ture probe, utilizing the antibody part of capture probe. Subsequently,
imidazolate framework-8 (ZIF-8), a typical kind of MOFs, is known for DNA part within capture probe worked as a primer to initiate a typical
inherent pH-responsive ability, which displays improved release effi­ DNA amplification reaction, hyperbranched rolling circle amplification
ciency at an acidic environment rather than the neutral or alkaline (HRCA). Environmental pH was gradually decreased during the HRCA
condition. Protein-encapsulating ZIF-8 prepared by process because hydrogen ions was sustainably released along with the
biomineralization-facilitated method is able to realize pH-responsive continuous addition of nucleotides, which then mediated the disas­
release of native active proteins in the live cells via endo-lysosomes sembly of PVP@HRP@ZIF-8 at a mildly acidic environment. As a result,
(Chen et al., 2018b). At the meanwhile, DNA polymerase that cata­ HRP was released and used to arouse amplified electrochemical re­
lyzes the incorporation of nucleotides into DNA chain is also demon­ sponses, and thus realized identification of breast cancer-derived PD-
strated to promote steady release of protons from 30 -hydroxyl group of L1þ exosomes.
extensive DNA chain during nucleic acid amplification, thereby

2
Y. Cao et al. Biosensors and Bioelectronics 166 (2020) 112452

2. Material and methods supplemented with 10% FBS and MCF-7 cells were grown in DMEM with
10% FBS. All cells were cultured at 37 � C with 5% CO2 and 80% relative
2.1. Materials humidity. Cell-derived exosomes were isolated from the supernatant of
culture medium of the three cell lines according to previous work with
HRP, 2-methylimidazole (2-MIM), zinc nitrate hexahydrate, sodium some modifications (Tian et al., 2020). Specially, cells were separated
dodecyl sulfate (SDS) and hydrogen peroxide (H2O2) were purchased from the culture medium when grown to 70–80% confluence and
from Sangon Biotech Co., Ltd (Shanghai, China). Phi29 DNA polymerase washed twice with PBS. After incubation for 24–48 h in serum-free
and dNTPs were obtained from New England Biolabs (Beijing, China). medium, the culture medium was collected and centrifuged at 200�g
PVP, o-phenylenediamine (OPD), sulfosuccinimidyl-4-(N- for 20 min and 3000�g for 15 min at 4 � C in sequence to remove intact
maleimidomethyl)cyclohexane-1-carboxylate (sulfo-SMCC), tris(2- cells and cellular debris. Thereafter, the supernatant was transferred to
carboxyethyl)phosphine (TCEP), N-hydroxysuccinimide (NHS), and N- 12000�g for 30 min at 4 � C. After being filtered with a 0.22 μm filter, the
(3-Dimethylaminopropyl)-N’-ethlcarbodiimide hydrochloride (EDC) supernatant was further purified with a 100 KD ultrafiltration centrifuge
were ordered from Sigma-Aldrich (Shanghai, China). MDA-MB-231, tube at 4000�g for 10 min. Finally, the upper filtrate containing exo­
MCF-7, and L02 cell lines were purchased from the Institute of somes was dispersed in 1.5 mL centrifuge tube, and stored at 80 � C
Biochemistry and Cell Biology of Chinese Academy of Science before further use. For expression analysis of CD63 and PD-L1 on exo­
(Shanghai, China). Dulbecco’s modified Eagle medium (DMEM), RPMI- some surface, immune-MBs (anti-CD63@MB and anti-PD-L1@MB,
1640 medium, and fetal bovine serum (FBS) were obtained from Gibco respectively) were used to capture exosomes. After magnetic separa­
Co., Ltd. (Beijing, China). Carboxylated MBs and 3,30 -dioctadecylox­ tion, the captured exosomes were stained by Dio (30 μM) for 30 min, and
acarbocyanine perchlorate (Dio) were purchased from Invitrogen then brought to analysis on a MoFlo XDP flow cytometer (Beckman
(Shanghai, China). Anti-CD63 antibody and anti-PD-L1 antibody were Coulter).
purchased from Abcam (Shanghai, China). All oligonucleotides used in
this work were synthesized by Sangon Biotech Co., Ltd (Shanghai, 2.5. Characterization of HRCA
China), and their sequences were listed in Table S1.
HRCA was conducted in a 60 μL mixture solution, including P1 (500
2.2. Synthesis of PVP@HRP@ZIF-8 nM), P2 (1 μМ), Template (500 nM), dNTPs (10 mM) and phi29 DNA
polymerase (12 U). The initial pH was adjusted with KOH (20 mM) to
PVP@HRP@ZIF-8 was synthesized according to previous report with 8.0, and the reaction was performed at 37 � C for 2 h, which was then
some modifications (Chen et al., 2018b). First, HRP (0.7 mg) was diluted terminated at 85 � C for 20 min. Agarose gel electrophoresis (AGE) and
in an aqueous solution of 2-MIM (3.15 mM, 0.9 mL) and stirred at 30 � C pH test were used for the characterization of HRCA. For AGE, DNA
for 10 min. Then, a solution of zinc nitrate hexahydrate (0.045 mM, 0.1 samples (10 μL) were mixed with 6 � loading buffer (2 μL) containing
mL) was added into the above mixture and stirred at 30 � C for another SYBR Green I and separated in 1 � TAE at 80 V for 10 min. After being
10 min to prepare HRP@ZIF-8. The prepared HRP@ZIF-8 was centri­ turned into 120 V for 30 min, gel imaging was performed using a Gel Doc
fuged at 3500 rpm for 20 min, and washed twice with ethanol. Then, XR Imaging system. For pH test, DNA samples (10 μL) were dropped on
HRP@ZIF-8 was dissolved in SDS (5%, w/w) aqueous solution at 50 � C the pH test papers, and the colors were photographed 90 s later.
to remove free proteins, and centrifuged at 14000 rpm for 10 min and
washed three times with deionized water to remove excessive SDS. 2.6. Electrochemical identification of PD-L1þ exosomes
Finally, HRP@ZIF-8 was mixed with PVP (1 mL, 3%, w/w) and stirred
for 30 min to obtain PVP@HRP@ZIF-8, which was stored at 4 � C for For identification of PD-L1þ exosomes, anti-CD63@MB (100 μL) and
future use. Transmission electron microscopy (TEM) characterization of capture probes (80 μL) were first mixed with different concentrations of
PVP@HRP@ZIF-8 was performed with a HT7700 electron microscope PD-L1þ exosomes at room temperature for 60 min. Then, the MBs were
(Hitachi, Japan) at 200 kV. Fourier transform infrared spectroscopy washed twice by PBS, and further mixed with HRCA reaction solution
(FTIR) was performed on a VERTEX70 FTIR spectrometer (Braker, (60 μL) containing P2 (1 μМ), Template (500 nM), dNTPs (10 mM) and
Germany). Dynamic light scattering (DLS) characterization was per­ phi29 DNA polymerase (12 U). HRCA was conducted at 37 � C for 2 h and
formed on a Zetasizer Nano (Malvern, UK). then terminated at 85 � C for 20 min. Thereafter, PVP@HRP@ZIF-8 (6.5
μL), H2O2 (7.5 μL, 40 mM) and OPD (7.5 μL, 40 mM) were added to the
2.3. Preparation of immune-MBs and capture probes HRCA reaction mixture and incubated for 45 min. Finally, the mixture
was diluted to 2 mL with PBS and used for electrochemical measure­
For the preparation of immune-MBs (anti-CD63@MB and anti-PD- ments. Electrochemical measurements were performed on CHI 660C
L1@MB), carboxyl MBs (100 μL, 10 mg/mL) were washed three times workstation (Shanghai, China) with the use of a three-electrode system,
by PBS, and then activated using EDC/NHS (500 μL, 0.22 M) at room including a platinum wire as the auxiliary electrode, a saturated calomel
temperature for 30 min. After washing by PBS, the activated MBs were electrode as the reference electrode, and a glassy carbon electrode as the
further reacted with antibodies (anti-CD63 or anti-PD-L1) (10 μL, 20 working electrode. Electrochemical responses were recorded with
μM) at room temperature for 2 h, and the obtained immune-MBs were square wave voltammetry (SWV) which was performed from 1.1 V to
magnetically separated from the solution and re-dispersed in PBS for 0.2 V.
further use. For the preparation of capture probes, anti-PD-L1 (10 μL,
2.4 μM) and sulfo-SMCC (10 μL, 10 mM) were mixed in PBS (80 μL) at 2.7. Preparation of clinical samples
30 � C for 1 h. At the meantime, P1 (10 μL, 20 μМ) and TCEP (10 μL, 200
μМ) were also mixed in PBS (80 μL) at 30 � C for 1 h. Then, the activated Clinical experiments were approved by the Scientific Ethical Com­
antibodies were mixed with the activated P1 at room temperature and mittee of Fudan University Shanghai Cancer Center and performed in
incubated for 2 h. Afterward, the mixture was transferred to 30 KD ul­ accordance with the ethical standards. Blood samples were collected
trafiltration centrifuge tube and centrifuged at 14000 rpm for 10 min, from healthy volunteers and breast cancer patients at Minhang Branch
and repeated for 3 times to collect the capture probes. of Fudan University Shanghai Cancer Center (Shanghai, China) after
obtaining informed consent. Each blood sample was left at room tem­
2.4. Cell culture and exosome isolation perature for 1 h, and then the supernatant was taken. Afterward, the
sample was centrifuged at 1550�g for 20 min to pellet all cells, and then
MDA-MB-231 and L02 cells were grown in RPMI-1640 medium centrifuged at 1200�g for 20 min to remove residual cellular debris.

3
Y. Cao et al. Biosensors and Bioelectronics 166 (2020) 112452

Fig. 1. Preparation and characterization of PVP@HRP@ZIF-8: (A) Preparation process of PVP@HRP@ZIF-8; B) SDS-PAGE analysis of HRP released from HRP@ZIF-8
and PVP@HRP@ZIF-8 after incubation in PBS at pH 6.5 or 8.0 for 45 min, respectively; (C) Absorbance at 425 nm for HRP-catalyzed oxidation of OPD using
HRP@ZIF-8 and PVP@HRP@ZIF-8 at pH 6.5 and 8.0, respectively.

Fig. 2. Characterization of exosomes derived from triple-negative breast cancer cell line MDA-MB-231: (A) TEM image and (B) size distribution analysis using NTA.
(C) to (F) represent flow cytometry analysis of surface expression of CD63 using anti-CD63@MB and PD-L1 using anti-PD-L1@MB..

4
Y. Cao et al. Biosensors and Bioelectronics 166 (2020) 112452

Fig. 3. Fluorescence images after staining anti-CD63@MB by Dio with (A) and without (B) MDA-MB-231-derived exosomes. (C) and (D) represent fluorescence
images after binding anti-CD63@MB-captured exosomes with Cy3-labeled capture probe and Cy3-labeled P1.

Finally, the supernatant was gently collected and stored at 80 � C PVP@HRP@ZIF-8 than HRP@ZIF-8 at pH 8.0, suggesting the improved
before use. stability for PVP coating under the alkaline condition. Overall, SDS-
PAGE and UV–vis spectra demonstrated the successful preparation and
3. Results and discussion pH-responsive ability of PVP@HRP@ZIF-8, and the encapsulated HRP
well maintained catalytic ability toward OPD oxidation.
3.1. Preparation and characterization of PVP@HRP@ZIF-8
3.2. Isolation and characterization of breast cancer-derived PD-L1þ
Fig. 1A shows schematic illustration of the preparation of exosomes
PVP@HRP@ZIF-8. To this end, HRP@ZIF-8 was first synthesized
through a biomimetic mineralization method, and PVP coating was Exosomes derived from triple-negative breast cancer line MDA-MB-
further applied to control pore penetration and increase long-term sta­ 231 were used as model breast cancer-derived PD-L1þ exosomes in
bility and dispersibility of the nanocomposites (Chen et al., 2018b). this work. TEM was first performed to characterize the exosomes, con­
Fig. S1 displays the characterization of PVP@HRP@ZIF-8. TEM was forming a typical structure of lipid bilayer membrane (Fig. 2A). The
used to demonstrate their nanoscale monodisperse morphology nanoparticle tracking analysis (NTA) revealed the size of 30–170 nm
(Fig. S1A) and DLS analysis showed an average size of ~330 nm (Fig. 2B). Expression of CD63 and PD-L1 on the surface of MDA-MB-231-
(Fig. S1B), which were in line with the previous report (Chen et al., derived exosomes were investigated by using flow cytometry (Fig. 2C to
2018b). FTIR was also performed. As shown in Fig. S1C, characteristic E). For this purpose, anti-CD63 and anti-PD-L1 were used to prepare
absorption bands at 1660 cm 1 and 1560 cm 1 were observed for immune-MBs (anti-CD63@MB and anti-PD-L1@MB) respectively. After
HRP@ZIF-8 and PVP@HRP@ZIF-8, respectively, attributing to amide I staining MB-captured MDA-MB-231-derived exosomes by Dio, fluores­
and II bands from HRP, which demonstrated the encapsulation of the cence shifts from the background confirmed the existence of surface-
enzymes. expressed CD63 and PD-L1. Furthermore, simultaneous recognition of
As one of the core elements of this work, pH-responsive ability of CD63 and PD-L1 on the PD-L1þ exosomes was revealed by using fluo­
PVP@HRP@ZIF-8 was fully investigated. Fig. 1B shows the results of rescence microscopy, in which anti-CD63@MB and Cy3-labeled capture
SDS-polyacrylamide gel electrophoresis (SDS-PAGE) analysis. Emerged probe that consisted of anti-PD-L1 and Cy3-labeled P1 were used for
protein bands indicated that encapsulated HRP could be preferably target binding. The fluorescence image first demonstrated the capture of
released from PVP@HRP@ZIF-8 at pH 6.5 but rarely released at pH 8.0. exosomes on anti-CD63@MB by Dio staining. A bright green fluores­
Fig. S2A shows the UV–vis spectrum for OPD oxidation by HRP released cence was observed with the addition of exosomes (Fig. 3A) but nearly
from PVP@HRP@ZIF-8. The maximum absorption peak at 425 nm no fluorescence was observed without exosomes (Fig. 3B), suggesting
demonstrated the well maintenance of catalytic ability of released HRP. the capture of exosomes through CD63-based immune-affinity interac­
It can also be seen from Fig. S2B that the absorption peak increased with tion. Moreover, intense yellow fluorescence of Cy3 was observed on MB-
the decrease of pH values from 8.0 to 6.5 and the solution color corre­ captured exosomes after adding Cy3-labeled capture probe (Fig. 3C), but
spondingly changed from colorless to yellow, which suggested gradually no fluorescence signaling was observed when using Cy3-labeled P1
enhanced release of HRP, confirming the pH-responsive ability of instead of Cy3-labeled capture probe (Fig. 3D). The results suggested the
PVP@HRP@ZIF-8. Fig. 1C further compares the pH-responsive abilities formation of sandwich-like structure at MB surface, which strongly
of PVP@HRP@ZIF-8 and HRP@ZIF-8 by using UV–vis spectra. relied on the recognition of exosomal PD-L1 and anti-PD-L1 within
Comparably high absorption at the wavelength of 425 nm was observed capture probe.
at pH 6.5 for both PVP@HRP@ZIF-8 and HRP@ZIF-8, attributing to
HRP-catalyzed oxidation of OPD with the addition of H2O2 after the
MOFs disassembly. In contrast, much lower absorption was observed for

5
Y. Cao et al. Biosensors and Bioelectronics 166 (2020) 112452

Fig. 4. (A) Electrochemical responses

(a) with or (b) without PD-L1þ exo­
somes, and (c) that with PD-L1þ exo­
somes but without HRCA. (B) Peak
currents of electrochemical responses
obtained upon analyzing 1010 particles/
mL exosomes isolated from MDA-MB-
231, MCF-7 and L02 cells, respectively.
(C) Electrochemical responses with
different concentrations of PD-L1þ exo­
somes (From a to i: 0, 103, 104, 105, 106,
107, 108, 109, and 1010 particles/mL).
(D) The relationship between peak cur­
rent and exosome concentration and the
linear relationship between peak cur­
rent and logarithmic value of exosome
concentration in the range from 103 to
1010 particles/mL (inset).

3.3. Electrochemical identification of PD-L1þ exosomes oxidation of OPD by the released HRP for amplified electrochemical
response. Conversely, the lack of exosomes inhibited HRCA for the
HRCA was essential to induce pH changes of the reaction microen­ missing surface-attached capture probe, identical to that with PD-L1þ
vironment under a low-buffered condition, so the process of HRCA was exosomes but without DNA polymerases. Fig. 4B shows the electro­
examined by AGE and pH test papers. Fig. S3A shows the schematic chemical responses with the addition of same amounts of exosomes
illustration of the reaction process and accompanying release of protons derived from different cell lines respectively. High electrochemical
involved in HRCA. Specially, P1 triggered polymerase-catalyzed DNA response was observed for exosomes derived from MDA-MB-231, but
amplification with the use of a circular template (Template); a reverse lower electrochemical response was observed for MCF-7-derived exo­
primer, P2, was then recruited to favor primer extension and strand somes and quite low electrochemical response approaching background
displacement simultaneously, together constituting HRCA. Fig. S3B signaling was observed for L02-derived exosomes. The results revealed a
shows the AGE image with the input of different DNA strands with or distinct PD-L1þ exosome level in the three cell lines: PD-L1 was highly
without DNA polymerase. Lane 1 displayed the amplification band of expressed in MDA-MB-231-derived exosomes, but expressed in MCF-7-
HRCA, but no amplification band was observed in the absence of P1 in derived exosomes at a low level and negatively expressed in L02-
Lane 2, suggesting the critical role of P1 in initiation of HRCA. Lane 3 derived exosomes. This was in agreement with the previous report
and 4 showed no bands for the amplification reaction without the input (Yang et al., 2018), demonstrating the feasibility of the method for ac­
of DNA polymerase even after the addition of P1, proving the necessity curate identification of PD-L1þ exosomes.
of DNA polymerase in HRCA. Fig. S3C shows the corresponding pH Afterward, several optimization experiments were performed.
changes using pH test paper. The solution pH remained stable at ~8.0 in Fig. S4 shows the peak currents with different reaction time for the
the absence of P1 or/and DNA polymerase (Lane 1–3) while decreased to capture of exosomes and binding with capture probe, suggesting the
~6.5 after HRCA (Lane 4). The results were in agreement of DNA optimal time of 60 min. Fig. S5 shows the peak currents with different
amplification-promoted proton release under low-buffered condition. reaction time for HRCA, suggesting the optimal time of 120 min. Fig. 4C
Fig. S3D further shows a high absorbance peak of OPD oxidation after shows the electrochemical responses with different concentrations of
HRCA but quite low absorbance without HRCA, which indicated HRCA- PD-L1þ exosomes under the optimal conditions. Obviously, peak cur­
mediated release of HRP from PVP@HRP@ZIF-8, confirming the HRCA- rents increased with the enhancement of exosome concentrations. This
responsive property of the nanocomposites. is reasonable, because elevated amounts of surface-enriched PD-L1þ
Electrochemical techniques were then applied to validate the feasi­ exosomes provided increased binding sites for the loading of capture
bility of the method. Fig. 4A shows the electrochemical responses with probes, and thus enhanced the reaction efficiency of HRCA to promote
or without the addition of PD-L1þ exosomes. High electrochemical the disassembly of PVP@HRP@ZIF-8 and the subsequent oxidation of
response was observed with the addition of PD-L1þ exosomes after OPD. Fig. 4D shows the correlation of peak currents and concentrations
HRCA-mediated release of HRP and subsequent OPD oxidation (curve of exosomes, and demonstrated a linear relationship between the peak
a), but quite low electrochemical responses were observed without PD- current and the logarithmic value of PD-L1þ exosome concentration in
L1þ exosomes (curve b) or that with PD-L1þ exosomes but without the the range from 103 to 1010 particles/mL. The linear equation was I (μA)
occurrence of HRCA (curve c). The results were as expected. Exosomes ¼ 0.181 � LogCexosomes (particles/mL) þ 0.037, R2 ¼ 0.992, and the limit
were enriched on anti-CD63@MB, and also bound with the capture of detection (LOD) was estimated to be 334 particles/mL at the signal-to-
probe through anti-PD-L1 to trigger HRCA. Sustainably released protons noise ratio of 3. Compared to prior studies, the method demonstrated an
promoted the disassembly of PVP@HRP@ZIF-8, and then catalyzed the improved sensitivity with wider linear range and lower LOD (Table S2).

6
Y. Cao et al. Biosensors and Bioelectronics 166 (2020) 112452

Fig. 5. (A) Peak currents of electrochemical responses for clinical serum samples. (B) Scatter plots of the peak currents of serum samples from healthy controls, non-
metastatic breast cancer patients and metastatic breast cancer patients measured by using the method.

Electrochemical measurements for each concentration were repeated at and HRP-catalyzed oxidation of OPD was regarded as the second signal
least three times with the average of RSDs as 3.85% for satisfactory amplification. The multiple signal amplification contributed to high
reproducibility. In addition, the anti-disturbance capability of the sensitivity of the method with a wider linear range and lower LOD. At
method was evaluated by using diluted FBS as complex sample. As the meanwhile, the selective and high-affinity binding of surface bio­
shown in Fig. S6, the peak currents obtained upon analyzing PD-L1þ markers and the corresponding antibodies ensured the high specificity of
exosomes in PBS buffer and diluted FBS were comparable, confirming the method. Moreover, the method could be used for identification of
high selectivity and recovery of the method in complex environment. PD-L1þ exosomes in clinical samples and found that circulating PD-L1þ
exosomes might be a valuable marker for diagnosis and progression
3.4. Identification of PD-L1þ exosomes in breast cancer clinical samples monitoring of breast cancer. In view of these features, we anticipate that
the method may have a cheerful application prospect in exosome
We also examined the application of the method in monitoring the identification and breast cancer research.
personalized level of PD-L1þ exosomes in breast cancer clinical samples.
To this end, undiluted serum from healthy persons (n ¼ 6), breast cancer Declaration of competing interest
patients with non-metastatic diseases (n ¼ 8) and metastatic diseases (n
¼ 7) were collected as clinical samples. The information of all the sub­ The authors declare that they have no known competing financial
jects was listed in Table S3. Fig. 5A shows the obtained peak currents for interests or personal relationships that could have appeared to influence
serum samples. Low peak currents were observed for the serum of the the work reported in this paper.
healthy persons due to the lack of PD-L1þ exosomes, and increased peak
currents were observed for the serum samples from the breast cancer CRediT authorship contribution statement
patients, suggesting the presence of PD-L1þ exosomes. In the meantime,
peak currents for breast cancer patients with metastatic diseases were Ya Cao: Methodology, Investigation, Writing - original draft. Ying
higher than those with non-metastatic diseases, which was consistence Wang: Methodology, Validation, Investigation. Xiaomeng Yu: Investi­
with the advanced disease progression and worse prognosis in breast gation, Writing - original draft. Xihui Jiang: Investigation. Gang Li:
tumors. Fig. 5B further shows the scatter plots of peak currents for serum Methodology, Writing - review & editing. Jing Zhao: Conceptualization,
samples. The low average value of 0.5040 μA was calculated for con­ Writing - review & editing, Supervision.
trols, and the average of 0.9011 μA and 1.525 μA were accounted for
patients with non-metastatic and metastatic diseases, respectively. The Acknowledgements
average peak currents for breast cancer patients were significantly
higher than that for healthy controls, and the average peak currents for This work was supported by the National Natural Science Foundation
metastatic patients was also significantly higher than that for non- of China (Grant Nos. 81972799, 81671781, and 81871449), and spon­
metastatic patients. It was worth noting that normal cell-derived exo­ sored by Shanghai Pujiang Program (2019PJD018).
somes in the serum samples that were also CD63-positive had little
impact on the electrochemical measurements, reconfirming the accurate Appendix A. Supplementary data
identification of PD-L1þ exosomes by using the method. The electro­
chemical identification of PD-L1þ exosomes in clinical samples sug­ Supplementary data to this article can be found online at [Link]
gested that PD-L1þ exosomes might be a potential biomarker to predict org/10.1016/[Link].2020.112452.
the occurrence of breast cancer at an early stage, and help to monitor the
disease progression during the treatment. References

Boriachek, K., Islam, M.N., M€ oller, A., Salomon, C., Nguyen, N.T., Hossain, M.S.A.,
4. Conclusion Yamauchi, Y., Shiddiky, M.J.A., 2018. Small 14, 1702153.
Chen, G., Huang, A.C., Zhang, W., Zhang, G., Wu, M., Xu, W., Yu, Z., Yang, J., Wang, B.,
In summary, we proposed an electrochemical biosensing method for Sun, H., Xia, H., Man, Q., Zhong, W., Antelo, L.F., Wu, B., Xiong, X., Liu, X., Guan, L.,
Li, T., Liu, S., Yang, R., Lu, Y., Dong, L., McGettigan, S., Somasundaram, R.,
identification of PD-L1þ exosomes using HRCA-mediated disassembly of Radhakrishnan, R., Mills, G., Lu, Y., Kim, J., Chen, Y.H., Dong, H., Zhao, Y.,
PVP@HRP@ZIF-8. The prepared PVP@HRP@ZIF-8 were stable at Karakousis, G.C., Mitchell, T.C., Schuchter, L.M., Herlyn, M., Wherry, E.J., Xu, X.,
alkaline environment to reduce background signaling. The HRCA- Guo, W., 2018a. Nature 560, 382–386.
Chen, T., Yi, J., Zhao, Y., Chu, X., 2018b. J. Am. Chem. Soc. 140, 9912–9920.
induced changes of environmental pH facilitated the disassembly of
PVP@HRP@ZIF-8, which was known as the first signal amplification,

7
Y. Cao et al. Biosensors and Bioelectronics 166 (2020) 112452

Cheng, G., Li, W., Ha, L., Han, X., Hao, S., Wan, Y., Wang, Z., Dong, F., Zou, X., Mao, S., Thery, C., Zitvogel, L., Amigorena, S., 2002. Nat. Rev. Immunol. 2, 569–579.
Zheng, S.Y., 2018. J. Am. Chem. Soc. 140, 7282–7291. Tian, Y., Gong, M., Hu, Y., Liu, H., Zhang, W., Zhang, M., Hu, X., Aubert, D., Zhu, S.,
Cheng, N., Du, D., Wang, X., Liu, D., Xu, W., Luo, Y., Lin, Y., 2019. Trends Biotechnol. 37, Wu, L., Yan, X., 2020. J. Extracell. Vesicles 9, 1697028.
1236–1254. Toumazou, C., Shepherd, L.M., Reed, S.C., Chen, G.I., Patel, A., Garner, D.M., Wang, C.
Daassi, D., Mahoney, K.M., Freeman, G.J., 2020. Nat. Rev. Immunol. 20, 209–215. A., Ou, C., Amin-Desai, K., Athanasiou, P., Bai, H., Brizido, I.M.Q., Caldwell, B.,
Fang, X., Zong, B., Mao, S., 2018. Nano-Micro Lett. 10, 64. Coomber-Alford, D., Georgiou, P., Jordan, K.S., Joyce, J.C., La Mura, M., Morley, D.,
Fitts, C.A., Ji, N., Li, Y., Tan, C., 2019. Adv. Healthcare Mater. 8, 1801268. Sathyavruthan, S., Temelso, S., Thomas, R.E., Zhang, L., 2013. Nat. Methods 10,
Hua, X., Yang, E., Yang, W., Yuan, R., Xu, W.J., 2019. Chem. Commun. 55, 641–646.
12463–12466. Wang, S., Chen, Y., Wang, S., Li, P., Mirkin, C.A., Farha, O.L., 2019. J. Am. Chem. Soc.
Huang, M., Yang, J., Wang, T., Song, J., Xia, J., Wu, L., Wang, W., Wu, Q., Zhu, Z., 141, 2215–2219.
Song, Y., Yang, C., 2020. Angew. Chem. Int. Ed. 59, 4800. Wortzel, I., Dror, S., Kenific, C.M., Lyden, D., 2019. Dev. Cell 49, 347–360.
Jeppesen, D.K., Fenix, A.M., Franklin, J.L., Higginbotham, J.N., Zhang, Q., Yang, X., Tang, Q., Jiang, Y., Zhang, M., Wang, M., Mao, L.Q., 2019. J. Am. Chem. Soc.
Zimmerman, L.J., Liebler, D.C., Ping, J., Liu, Q., Evans, R., Fissell, W.H., Patton, J.G., 141, 3782–3786.
Rome, L.H., Burnette, D.T., Coffey, R.J., 2019. Cell 177, 428–445. Yang, Y., Li, C.W., Chan, L.C., Wei, Y., Hsu, J.M., Xia, W., Cha, J.H., Hou, J., Hsu, J.L.,
Kalluri, R., LeBleu, V.S., 2020. Science 367, eaau6977. Sun, L., Hung, M.C., 2018. Cell Res. 28, 862–864.
Mao, D.S., Chen, T.S., Chen, H.N., Zhou, M.R., Zhai, X.W., Chen, G.F., Zhu, X.L., 2019. Yu, D.D., Wu, Y., Shen, H.Y., Lv, M.M., Chen, W.X., Zhong, S.L., Tang, J.H., Zhao, J.H.,
Analyst 144, 4060–4065. 2015. Canc. Sci. 106, 959–964.
Mathieu, M., Martin-Jaular, L., Lavieu, G., Thery, C., 2019. Nat. Cell Biol. 21, 9–17. Zhou, Y., Zhang, D., Zeng, J., Gan, N., Cuan, J., 2018. Talanta 181, 410–415.
Poudineh, M., Sargent, E.H., Pantel, K., Kelley, S.O., 2018. Nat. Biomed. Eng. 2, 72–84. Zhong, M., Yang, L., Yang, H., Cheng, C., Deng, W., Tan, Y., Xie, Q., Yao, S.Z., 2019.
Tang, Y., Zhang, P., Wang, Y., Wang, J., Su, M., Wang, Y., Zhou, J., Xiong, W., Zeng, Z., Biosens. Bioelectron. 126, 493–500.
Zhou, Y., Nie, S., Liao, Q., 2020. Front. Immunol. 11, 604.