Hypoxia-Induced Pro-Protein Therapy Assisted by a Self-Catalyzed Nanozymogen

Article abstract of DOI:10.1002/anie.202004008

The success of intracellular protein therapy demands efficient delivery and selective protein activity in diseased cells. Therefore, a cascaded nanozymogen consisting of a hypoxia-activatable pro-protein, a hypoxia-inducing protein, and a hypoxia-strengthened intracellular protein delivery nanovehicle was developed. RPAB, an enzymatically inactive pro-protein of RNase, reversibly caged with hypoxia-cleavable azobenzene, was delivered with glucose oxidase (GOx) using hypoxia-responsive nanocomplexes (NCs) consisting of azobenzene-cross-linked oligoethylenimine (AOEI) and hyaluronic acid (HA). Upon NC-mediated delivery into cancer cells, GOx catalyzed glucose decomposition and aggravated tumoral hypoxia, which drove the recovery of RPAB back to the hydrolytically active RNase and expedited the degradation of AOEI to release more protein cargoes. Thus, the catalytic reaction of the nanozymogen was self-accelerated and self-cycled, ultimately leading to a cooperative anti-cancer effect between GOx-mediated starvation therapy and RNase-mediated pro-apoptotic therapy.

Full text of DOI:10.1002/anie.202004008

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Hypoxia-instructed pro-protein therapy assisted with self-
catalyzed nanozymogen
Authors: Xudong Li, Yuansong Wei, Yuchen Wu, and Lichen Yin
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202004008
Link to VoR: https://doi.org/10.1002/anie.202004008

10.1002/anie.202004008
Angewandte Chemie International Edition
1 Hypoxia-instructed
pro-protein
therapy
assisted
with
2 self-catalyzed nanozymogen
3
4
Xudong Li#,Yuansong Wei#, Yuchen Wu, and Lichen Yin*
5
X. Li, Y. Wei, Y. Wu, Prof. Dr. L. Yin
6
Institute of Functional Nano & Soft Materials (FUNSOM),
Jiangsu Key Laboratory for Carbon-Based Functional Materials and Devices,
Collaborative Innovation Center of Suzhou Nano Science and Technology,
Soochow University,
7
8
9
10
11
Suzhou 215123, China
E-mail: lcyin@suda.edu.cn
12
13
14
15
16
17
X. Li and Y. Wei contributed equally.
Keywords
pro-protein engineering; intracellular protein delivery; hypoxia responsiveness; self-catalysis;
starvation therapy
18
19
Abstract
20
21
22
23
24
25
26
27
28
29
The success of intracellular protein therapy demands potent intracellular delivery
efficiency and selective protein activity in diseased cells. To fulfil these requirements, a
cascaded nanozymogen incorporating hypoxia-instructed pro-protein engineering and
hypoxia-strengthened intracellular protein delivery was developed. RPAB, a pro-protein of
RNase A reversibly caged with hypoxia-cleavable azobenzene yet deprived of the enzymatic
activity, was designed and co-delivered with glucose oxidase (GOx) using
hypoxia-dissociable nanocomplexes (NCs) consisting of azobenzene-cross-linked
oligoethylenimine (AOEI) and hyaluronic acid (HA). Upon NCs-mediated tumor targeting,
cancer cell internalization, and endolysosomal escape, GOx catalyzed intracellular glucose
decomposition and aggravated tumoral hypoxia, which drove the recovery of RPAB back to
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
4
5
6
7
the hydrolytically active RNase and concurrently expedited the degradation of AOEI to
release more protein cargoes. Thus, the catalytic reaction of the nanozymogen was
self-accelerated and self-cycled, ultimately leading to cooperative anti-cancer efficacies
between GOx-mediated starvation therapy and RNase-mediated pro-apoptotic therapy. This
study represents the first example of hypoxia-instructed pro-protein engineering, which
expands the available chemical tools for the conditional control over protein functions. The
self-catalyzed nanozymogen design also provides an effective strategy to mediate intracellular
protein delivery and highly sensitive and selective intracellular pro-protein activation.
8
9
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
Protein therapeutics hold great potentials in the treatment of human diseases, mainly due
to their higher specificity and less side effects over small-molecule drugs.[1] To date, clinically
applied protein drugs are largely limited to those functioning extracellularly.[2] Although
advancement in molecular biology has suggested that proteins targeting the intracellular
machinery could be even more potent toward disease treatment,[3] protein therapeutics that
exert bioactivity in the cytosol have achieved limited success,[4] because of the lack of
efficient intracellular delivery vehicles that can transport the protein cargo to the diseased
tissues, mediate trans-membrane delivery into target cells, avoid endolysosomal entrapment,
and release the biologically active protein in the cytosol.[5] In the past years, various systems
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
have been developed to enable intracellular protein delivery with exciting successes, as
[7]
exemplified by dendrimer,[6] polymeric micells,
polymerized capsules,[8] cationic
liposomes,[9] nanogels,[10] inorganic nanoparticles,[11] and etc.[12] However, effective vehicles
that can harmonize all the above-mentioned delivery processes to mediate protein therapy in
vivo are still of great demand.[13]
Another critical issue related to intracellular protein therapy is how to realize selective
pharmacological activity in target cells, thus minimizing non-specific adverse effects to
normal cells.[14] While the use of stimuli-responsive carriers can enable selective protein
release in target cells, the inevitable pre-leakage or non-specific protein release would
increase the risk of side toxicity.[15] Recently, the strategy of reversible caging of proteins
with trigger-cleavable motifs has been explored to enable spatiotemporal control of protein
function in response to chemical, biological, or physical signals, such as acid,[16] reactive
oxygen species (ROS),[17] enzyme,[18] or light.[19] However, the potential of these protein
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
prodrugs in mediating selective protein therapy may still remain doubtful, mainly because of
the insufficient sensitivity to exogeneous/endogenous triggers or the low specificity to
discriminate between diseased and normal cells.[20] It is thus highly imperative to design
caged protein prodrugs that are capable of highly efficient and selective de-caging in diseased
cells, thus achieving maximized therapeutic efficacy and minimized systemic toxicity.
To fulfill the demands for the intracellular pro-protein therapy toward cancer treatment,
we herein designed a hypoxia-activatable pro-protein, assisted with the construction of
self-catalyzed cascaded nanozymogen to realize hypoxia-strengthened intracellular
pro-protein delivery as well as sensitive and selective control over the hypoxia-instructed
pro-protein activation in cancer cells. While various protein prodrugs capable of
trigger-responsive activation have been reported,[16-19] hypoxia-instructed pro-protein
engineering is still lacking, which makes it the first example of hypoxia-responsive protein
prodrug. Particularly, RNase A was reversibly caged with hypoxia-cleavable azobenzene
domains on its lysine residues, generating the pro-protein (RPAB) deprived of the enzymatic
activity along with elevated net anionic charge density (Scheme 1). Thus, RPAB was
deprived of the non-specific toxicity to normal cells. The azobenzene group could be reduced
under hypoxic condition by the various reductases that are over-expressed and mainly
localized inside tumor cells, thus restoring the hydrolytic activity of RNase intracellularly
upon hypoxia-triggered uncaging of the azobenzene protection group. While the tumoral
microenvironment is often hypoxic due to fast metabolism and proliferation of cancer cells,
the hypoxia level in many solid tumors is heterogeneous and insufficient, which hurdles the
effective re-activation of RPAB. As such, RBAP was co-delivered with glucose oxidase
4
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
(GOx), an enzyme that catalyzes the decomposition of glucose accompanied with concurrent
exhaustion of O2 and generation of H2O2.[21] Thus, GOx served to aggravate tumoral hypoxia
and accordingly drive the recovery of RPAB back to RNase. To enable the targeted
accumulation of protein drugs into tumor tissues and trans-membrane internalization into
cancer cells, a cationic, hypoxia-degradable polymer based on azobenzene-cross-linked
oligoethylenimine (AOEI) was further designed, which was allowed to form the
nanozymogen upon electrostatic complexation with negatively charged hyaluronic acid (HA)
and concurrent encapsulation of RPAB and GOx. The nanozymogen facilitated tumor
accumulation via the permeability and retention (EPR) effect and the binding affinity between
HA and over-expressed CD44 on cancer cell membranes. The cationic AOEI then promoted
internalization of protein cargoes (RPAB and GOx) into cancer cells, and enabled
endolysosomal escape via the “proton sponge” effect. The hypoxia condition along with the
intracellular reductases initiated the NCs dissociation and release of GOx. GOx catalyzed the
decomposition of glucose in cancer cells and simultaneously aggravated tumoral hypoxia,
which not only propelled the recovery of RNase, but also expedited the degradation of AOEI
to release more protein cargoes. The released GOx could further amplify the hypoxia signals
and thus self-accelerate the catalytic reaction of the nanozymogen. Finally, GOx-mediated
starvation therapy and RNase-mediated pro-apoptotic therapy cooperated to enable effective
anti-cancer outcomes.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
RPAB, the hypoxia-activatable prodrug of RNase, was synthesized via caging of the
primary amines on lysine residues of RNase with azobenzene motifs. PEG-azobenzene (PAB,
compound 4) was first synthesized via the diazotization reaction between
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
4-aminobenzylalcohol and phenol followed by conjugation with tosylated PEG (compound 1)
to improve the aqueous solubility (Scheme S1, Figure S1-S4). PAB was then conjugated onto
the lysine residues of RNase via a carbamate linker to yield RPAB. MALDI-TOF analysis
confirmed successful conjugation of PAB onto RNase (MW = 13600 Da), and the peaks at
14998, 16194, 17537, and 18733 Da after PAB conjugation indicated that each RNase
molecule was modified with 1-4 azobenzene groups (Figure 1A). After 6-h treatment with
Na2S2O4 (155 mM), the MW of RPAB recovered to that of RNase, indicating liberation of the
native RNase due to hypoxia-mediated cleavage of the azobenzene linkage followed by
self-immolation reaction (Scheme 1). Sodium dodecyl sulfate polyacrylamide gel
electrophoresis (SDS-PAGE) analysis further demonstrated successful conjugation of PAB
onto RNase, as revealed by the disappearance of the RNase band while appearance of a new
band (RPAB) with increased MW (Figure 1B). When RPAB was treated with Na2S2O4, the
band of the native RNase reappeared, suggesting hypoxia-triggered liberation of the native
RNase. The circular dichroism (CD) spectra of RPAB and Na2S2O4-treated RPAB were found
to be identical to that of the native RNase, indicating that the PAB caging and
hypoxia-triggered PAB uncaging did not alter the secondary structure of the protein drug
(Figure 1C). In support of the above findings, the hydrolytic activities of RNase and RPAB
against RNA were measured using the ethidium bromide (EB) fluorescence assay (Figure
1D).[22] After addition of RNase, the fluorescence intensity was decreased due to
RNase-mediated RNA degradation. In comparison, RPAB revealed minimal enzymatic
activity, which however, could be completely restored after pre-treatment with Na2S2O4. It
therefore substantiated that caging of RNase with the azobenzene moiety could temporally
6
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
deactivate RNase, and hypoxia-triggered PAB cleavage could completely restore the protein
function in a selective manner.
3
AOEI was then synthesized via chemical cross-linking of OEI (600 Da). Briefly, an
azobenzene-containing cross-linker (compound 5) was firstly synthesized via reductive
coupling of 4-nitrobenzoic acid (Scheme S2 and Figure S5), and AOEI was thereafter
synthesized via the amidation reaction between compound 5 and OEI (600 Da) (Figure S6).
The MW of AOEI was determined to be 4094 Da using MALDI-TOF, and it decreased to 799
4
5
6
7
8
Da after 12-h treatment with Na2S2O4 (155 mM, a reducing agent that mimics the hypoxia
[23]
9
condition),
indicating the successful cross-linking of OEI and hypoxia-triggered
10
11
12
13
14
15
16
17
18
19
20
21
22
degradation of AOEI (Figure S7). BOEI, a non-responsive analogue of AOEI with, was
similarly synthesized from OEI and (1,1’-biphenyl)-4,4’-dicarboxylicacid (Scheme S2).
The inactivated RBAP and GOx were then used to construct the nanozymogen with AOEI
and HA. After reaction with PAB, the negative charge density of RNase was increased due to
partial caging of the cationic lysine residues (Table S1). GOx also possessed negative net
charges at physiological pH. Thus, RBAP and GOx, along with the negatively charged HA,
were able to form NCs with the positively charged AOEI via electrostatic interactions.
Because positively charged NCs are usually toxic and unstable during circulation, it is crucial
to adjust the surface charge of NCs to a negative value to minimize non-specific binding of
serum proteins and enhance their biocompatibility. When the HA/AOEI weight ratio
increased from 0.5 to 4, NCs with diameters of 150-200 nm could be obtained, while the zeta
potential continuously decreased from +28.9 mV to -23.1 mV (Figure S8). Further
enhancement of the HA/AOEI ratio did not significantly increase the negative zeta potential.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
Thus, at the optimal HA/AOEI weight ratio of 4, RPAB and GOx were allowed to form the
HA/AOEI/RPAB+GOx NCs (HARPG NCs, HA/AOEI/RPAB/GOx = 160/40/10/1, w/w/w/w)
with the particle size of ~140 nm and negative zeta potential of -18.1 mV as determined by
dynamic laser scanning (DLS) measurement (Figure 2A and Table S2). Transmission electron
microscopy (TEM) observation also revealed that the HARPG NCs possessed diameter of
~120 nm with a relatively uniform size distribution (Figure 2B). The loading capacities of
RPAB and GOx in the HARPG NCs were determined to be 4.4% and 0.3%, respectively. As
controls, HA/AOEI/RNase NCs (HAR NCs), HA/AOEI/RPAB NCs (HARP NCs),
HA/AOEI/GOx NCs (HAG NCs), and HA/AOEI/RNase+GOx NCs (HARG NCs) were
similarly prepared and characterized (Table S2). After incubation with serum-containing
medium for 24 h, size of the HARPG NCs remained almost constant (Figure S9), which
indicated their desired serum stability due to the presence of HA that contributed to the
negative surface charges. Such property would provide implications for prolonged blood
circulation in vivo.[24]
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
The hypoxia-responsive dissociation of NCs was explored by TEM and DLS. As shown
in Figure 2A and 2B, the NCs remarkably swelled after treatment with Na2S2O4 for 15 min,
and they partly disassembled after 4-h treatment. It thus suggested that the azobenzene groups
in the cross-linked AOEI were effectively reduced under hypoxia condition, which resulted in
the dissociation of NCs as a consequence of decreased molecular entanglement with protein
cargoes or HA. To evaluate the hypoxia-responsive drug release profile, fluorescein
isothiocyanate (FITC)-labeled RPAB and rhodamine B (RhB)-labeled GOx were synthesized
and allowed to form NCs with AOEI and HA. As shown in Figure 2C, in the absence of
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
Na2S2O4, only 12.2% of the total GOx was released from HARPG NCs within 24 h,
suggesting the excellent GOx retention capability of HARPG NCs. In comparison, the amount
of GOx released from HARPG NCs increased to 73.8% within 24 h in the
Na2S2O4-containing medium. Such a hypoxia-responsive release profile was similarly noted
for RPAB, which further confirmed that hypoxia-triggered de-crosslinking of AOEI could
allow “on-demand” protein release under hypoxia condition in tumor cells.
3
4
5
6
7
GOx could catalyze the conversion of glucose to gluconic acid, accompanied with O2
consumption, pH decrease, and H2O2 production. Based on this reaction, the catalytic
capability of HARPG NCs was evaluated in terms of gluconic acid and H2O2 production. As
illustrated in Figure 2D and S10, the pH value of the glucose solution (1 mg/mL) slightly
decreased from 7.0 to 6.1 after treatment with HARPG NCs for 1 h. In contrast, a drastic pH
drop (from 7.0 to 3.9) occurred after incubation with Na2S2O4-treated HARPG NCs or free
GOx for 1 h, indicating the hypoxia-triggered GOx release from NCs that subsequently
catalyzed glucose decomposition. In accordance, H2O2 concentration in the glucose solution
(1 mg/mL) increased rapidly and extensively to ~2300 μmol/L after 1-h incubation with free
GOx or Na2S2O4-treated HARPG NCs (Figure 2E and S11), while the H2O2 concentration was
only slightly increased upon incubation with HARPG NCs in the absence of Na2S2O4 (Figure
2E). These findings demonstrated that hypoxia triggered GOx release from HARPG NCs and
the released GOx preserved its enzymatic activity to decompose glucose and concomitantly
consume oxygen that would potentially starve cancer cells and aggravate the hypoxia
condition.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
The capability of NCs to deliver protein cargoes into cancer cells was then evaluated in
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
vitro. As shown in Figure S12, after HeLa cells were incubated with HARPG NCs containing
FITC-RPAB and RhB-GOx for 4 h, extensive cytoplasmic distribution of green and red
fluorescence was noted, indicating the strong capability of the NCs to mediate intracellular
delivery of protein cargoes. Such result was further supported by flow cytometry analysis,
wherein NCs showed remarkable cellular uptake level compared to the minimal uptake level
of free RPAB (Figure 3A and S13). Moreover, pre-treatment of cells with free HA
significantly reduced the uptake level of HARPG NCs in HeLa cells (tumor cell) but not in
3T3 cells (fibroblast), indicating that HA facilitated cancer cell uptake of HARPG NCs by
binding to the CD44 overexpressing on cancer cell surfaces. While HA mediates effective
targeting of CD44 on cancer cells, it has to be admitted that CD44 is also expressed in many
other cell types, which would compromise the cancer targeting selectivity of HA-based
nanosystem. Anionic polymers, such as chondroitin sulfate (CS), polyglutamic acid (PGA),
poly(propylacrylic acid) (PPAA), heparin, and fucosylated polysaccharide could be utilized as
HA alternatives to neutralize the positive charges of NCs, and covalent conjugation with
tumor-specific antibodies or aptamers may potentially enable highly effective and selective
tumor targeting.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
After endocytosis, NCs will experience endosomal entrapment, which prevents them
from entering the cytosol. Therefore, we further evaluated the intracellular distribution of
HARPG NCs containing FITC-RPAB. CLSM observation revealed extensive internalization
of FITC-RPAB in HeLa cells post 4-h incubation, and the green fluorescence of FITC-RPAB
largely separated from the red fluorescence of Lysotracker Red-stained endosomes (Figure
S14), suggesting that the HARPG NCs were capable of mediating effective endosomal escape,
10
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
presumably due to the AOEI-assisted proton sponge mechanism.[25]
Then, hypoxia-triggered intracellular protein release, GOx-mediated H2O2 generation, and
HARPG NCs-aggravated hypoxia in tumor cells were investigated. A hypoxia/oxidative stress
(ROS) detection kit was first used to detect the intracellular H2O2 generation and the
concomitant aggravation of local hypoxia after incubation of HeLa cells with HARPG NCs.
As shown in Figure 3B, free GOx induced a slightly higher level of intracellular hypoxia than
deferoxamine (DFO), a hypoxia inducer, as indicated by the intensified magenta fluorescence
generated by the hypoxia probe. Cells treated with HARPG NCs showed the strongest
magenta fluorescence intensity, indicating the capabilities of NCs to facilitate intracellular
GOx delivery that catalyzed the decomposition of glucose accompanied with oxygen
consumption. Cells treated with HARPG NCs also showed the strongest green fluorescence of
the ROS probe, which consistently demonstrated the generation of intracellular H2O2 as a
product of glucose decomposition (Figure S15). Such finding was further supported by the
quantitative analysis using flow cytometry, wherein cells treated with HARPG NCs showed
remarkably higher levels of cellular hypoxia (48.6%) and ROS (46.1%) levels than
GOx-treated and un-treated cells (Figure 3C).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
To further investigate the hypoxia-triggered intracellular protein release, HARP NCs
consisting of rhodamine-labeled AOEI (RhB-AOEI) and FITC-labeled RPAB (FITC-RPAB)
were incubated with HeLa cells for 4 h and the colocalization between the fluorescence of
RhB-AOEI and FITC-RPAB was calculated. As shown in Figure 3D and 3E, notable
fluorescence separation was noted under hypoxia concentrations (5% and 1% O2) compared
to the fluorescence overlap under normoxia (20% O2), conferring decreased colocalization
11
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
ratios of 81.1%, 64.0%, and 41.1% at the oxygen levels of 20%, 5%, and 1%, respectively.
Similar fluorescence separation was also noted for cells treated with HARPG NCs containing
RhB-AOEI and FITC-RPAB under norxomia condition (Figure S16). In contrast,
unappreciable fluorescence separation was noted for HA/BOEI/RPAB (HBRP) NCs
containing the non-responsive RhB-BOEI and FITC-RPAB either under hypoxia or normoxia
conditions (Figure S17). These results thus collectively substantiated that sufficient hypoxia
could lead to protein release as a consequence of AOEI degradation, and HARPG NCs could
enhance oxygen consumption upon initial GOx release induced by mild hypoxia, which
enabled cyclic amplification of the hypoxia trigger signals and thus self-catalyzed the protein
release.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Polycatioic materials tend to cause cell damage, because of their strong interactions with
cell membranes. Thus, prior to the in vitro anti-cancer efficacy study, the cytotoxicity of AOEI
was first monitored in HeLa cells under both normoxia (20% O2) and hypoxia (1% O2)
environmenets. AOEI displayed concentration-dependent cytotoxicity under normoxia
condition, while it showed negligible cytotoxicity under hypoxia at the concentration up to 35
μg/mL (Figure 4A), indicating that hypoxia-triggered degradation of AOEI into low-MW
segments to diminish the materials toxicity. As a control, the non-responsive BOEI revealed
similar cytotoxicity under both normoxia and hypoxia conditions (Figure S18), which further
confirmed that hypoxia-triggered polymer degradation could greatly reduced the materials
toxicity.
Then, the in vitro anti-cancer efficacy of various NCs in HeLa cells was evaluated using
the MTT assay. As shown in Figure 4B and Figure 4C, HARP NCs showed minimal
12
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
cytotoxicity under normoxia at the RPAB concentration up to 2 μg/mL, indicating deprivation
of the catalytic activity of the protein drug. However, HARP NCs showed
concentration-dependent cytotoxicity under hypoxia (1% O2), which was similar to that of
HAR NCs under either hypoxia or normoxia conditions (Figure 4C and 4D), substantiating
hypoxia-mediated recovery of the prodrug RPAB to the catalytically active RNase. HAG NCs
also exhibited a concentration-dependent anticancer activity in presence of O2 (Figure 4B),
which was ascribed to the consumption of glucose that starved the tumor cells and generation
of intracellular H2O2 that caused apoptosis of tumor cells. Because O2 was continuously
consumed during GOx-mediated starvation therapy, it created a more hypoxic environment
that would facilitate the dissociation of NCs to release more GOx, thus realizing the cyclic
amplication of hypoxia signals. In accompany with such enzyme-linked reaction, more RPAB
could also be released from HARPG NCs and be recovered to RNase upon hypoxia. In
support of such mechanism, the anti-cancer efficacy of HARPG NCs was significantly
stronger than HAG NCs and similar to that of HARG NCs (Figure 4B), indicating
GOx-assisted recovery of the prodrug RPAB back to RNase which subsequently provoked
cooperative anti-cancer effect with GOx. Similar trend was noted when the cancer cell
apoptosis level was determined using the Annexin V-fluoresceine isothiocyanate (Annexin
V-FITC)/propidine iodide (PI) apoptosis detection assay. As shown in Figure 4E, HARP NCs
induced minimal cancer cell apoptosis, while HARPG NCs and HARG NCs provoked similar
and the highest apoptosis levels of 50.7% and 56.2%, notably outperforming the HAR NCs
(19.4%) and HAG NCs (31.6%). When AOEI was replaced by the non-degradable BOEI, the
anti-cancer efficacies of NCs were significantly decreased (Figure S19), which further
13
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
evidenced that the hypoxia-responsive AOEI degradation was critical to the cytosolic release
and bioactivity of protein cargoes.
3
The pharmacokinetics and biodistribution of HARPG NCs after i.v. injection were then
explored. As shown in Figure 5A, the blood circulation half-life (t1/2) of HARPG NCs
containing Cy5-RPAB (~4.16 h) was about 8-fold longer than that of free Cy5-RPAB (~0.51
h), mainly because HA shielded the positive surface charges of NCs and sterically hindered
the binding of blood components. The biodistribution of Cy5-RPAB-containing HARPG NCs
and free Cy5-RPAB was then evaluated in HeLa xenograft tumor-bearing mice using the live
animal imaging system. As shown in Figure 5B, the fluorescence of HARPG NCs at the
tumor site steadily increased and peak at 24 h post i.v. injection, while in comparison,
negligible fluorescence was observed in tumors administered with Cy5-RPAB. Thus, tumors
and major organs were harvested at 24 h post injection and subjected to ex vivo imaging. The
fluorescence intensity in NCs-treated tumors was 4.1-fold of that in free Cy5-RPAB-treated
tumors, which could be attributed to the prolonged circulation and HA-mediated tumor
targeting capability of NCs (Figure 5C and 5D).
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
To investigate whether the HARPG NCs could exacerbate hypoxia in the solid tumors, in
vivo photoacoustic (PA) imaging was adopted to monitor the blood oxygenation status (sO2)
in tumors via detection of oxyhemoglobin saturation. As shown in Figure 5E and 5F, the sO2
level in tumors dramatically decreased from 46.6% at 0 h to 14.1% at 24 h post i.v. injection
of HARPG NCs. In comparison, no obvious change of tumoral sO2 level was observed in
mice injected with HARP NCs deprived of GOx. Free GOx, after i.v. injection, led to slight
decrease of the tumoral sO2 level, mainly due to the low tumor accumulation level and fast
14
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
clearance. These findings therefore evidenced that HARPG NCs could lead to aggravation of
hypoxia in the tumor tissues in accompany with GOx-mediated glucose decomposition, which
could subsequently accelerate the hypoxia-instructed NCs dissociation, protein cargo release,
and activation of RPAB.
3
4
5
The in vivo anti-tumor efficacy of the nanozymogen was evaluated in HeLa xenograft
tumor-bearing mice following i.v. injection. As shown in Figure 6A and 6B, no significant
tumor inhibition was observed for HARP NCs, because RPAB was deprived of the enzymatic
activity after caging of RNase with azobenzene motifs. HAR NCs capable of RNase-mediated
tumor inhibition and HAG NCs capable of GOx-mediated tumor starvation showed moderate
anti-cancer efficacies, conferring the tumor growth inhibition (TGI) rates of 75.5% and 77.2%,
respectively. Comparatively, HAPRG NCs led to notably higher tumor remission,
representing the TGI rate of 97.6% that was similar to that of HARG NCs. In accordance,
HARPG NCs led to the highest animal survival rate within the 45-d observation period
(Figure 5C), and they induced the highest levels of tumor cell apoptosis and remissioin
(Figure 5D and S21). These findings thus substantiated the re-activation of RPAB in tumors in
response to GOx-aggravated hypoxia, which concomitantly provoked cooperative anti-cancer
outcomes. The body weights of mice were not significantly affected within the 19-d
observation period upon administrations of various NCs (Figure S20), implying lack of
appreciable systemic toxicity of NCs.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
The in vivo biocompatibility of NCs was finally evaluated in terms of hematological
analysis, blood biochemical measurement, and histological observation. On day 1, 3, and 5,
PBS or various NCs were i.v. injected using the same protocol as in the efficacy study, and
15
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
biocompatibility evaluation was performed at 24 h post the third injection. Lack of
abnormality was noted for all hematological and biochemical parameters after administration
of various NCs (Figure S22). In the homogenate of liver, spleen, and kidney, alanine
aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), lipid
peroxidation (LPO), and creatinine (CR) levels showed no significant change after treatment
with various NCs (Figure S23). Additionally, the levels of pro-inflammatory cytokines
(TNF-α, IL-6, and IL-1β) in these organs did not significantly change after administration of
NCs (Figure S23). HE staining of major organs also revealed lack of pathological indications
(Figure S24). These results therefore collectively indicated that the NCs did not cause
appreciable systemic toxicity or inflammatory response.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
In summary, were herein reported a cascaded anti-cancer nanozymogen that incorporated
hypoxia-instructed pro-protein engineering and hypoxia-strengthened intracellular protein
delivery to enable potent anti-cancer protein therapy. RPAB, a hypoxia-activatable
pro-protein of RNase, was designed and co-delivered with GOx using a hypoxia-dissociable
nanosystem to enable efficient intracellular protein therapy in cancer cells. To the best of our
knowledge, this study represents the first example of hypoxia-responsive pro-protein
engineering, and when coupled with the GOx-mediated, self-catalyzed hypoxia aggravation
mechanism, it provides highly sensitive and selective control over the protein activities in
tumor cells. We thus believe that the current approach expands the available chemical tools
for the conditional and precision control over the structure and function of proteins. In
addition, such a hypoxia-dissociable nanosystem that is capable of GOx-strengthened,
hypoxia-assisted intracellular protein release also provides innovative insights into the design
16
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
4
5
6
7
8
9
of intracellular protein delivery vehicles.
Acknowledgment
This work was supported by the Ministry of Science and Technology of China
(2016YFA0201200), the National Natural Science Foundation of China (51722305 and
51873142), 111 project, and Priority Academic Program Development of Jiangsu Higher
Education Institutions (PAPD).
References
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
[1]
[2]
B. Leader, Q. J. Baca, D. E. Golan, Nat. Rev. Durg Discov. 2008, 7, 21-39.
a) S. Marchal, A. El Hor, M. Millard, V. Gillon, L. Bezdetnaya, Drugs 2015, 75, 1601-1611; b) J. Shi, P. W.
Kantoff, R. Wooster, O. C. Farokhzad, Nat. Rev. Cancer 2017, 17, 20-37.
[3]
[4]
[5]
L. Urquhart, Nat. Rev. Durg Discov. 2018, 17, 309-309.
A. N. Zelikin, C. Ehrhardt, A. M. Healy, Nat. Chem. 2016, 8, 997-1007.
a) P. J. Carter, Nat. Rev. Immunol. 2006, 6, 343-357; b) J. Lv, Q. Fan, H. Wang, Y. Cheng, Biomaterials
2019, 218, 119358; c) S. Mitragotri, P. A. Burke, R. Langer, Nat. Rev. Durg Discov. 2014, 13, 655-672.
a) C. Liu, T. Wan, H. Wang, S. Zhang, Y. Ping, Y. Cheng, Sci. Adv. 2019, 5, eaaw8922; b) Z. Zhang, W. Shen,
J. Ling, Y. Yan, J. Hu, Y. Cheng, Nat. Commun. 2018, 9, 1-8; c) L. Ren, J. Lv, H. Wang, Y. Cheng, Angew.
Chem. Int. Ed. 2019, 132, 4741-4749.
[6]
[7]
[8]
a) K. Dutta, D. Hu, B. Zhao, A. E. Ribbe, J. Zhuang, S. Thayumanavan, J. Am. Chem. Soc. 2017, 139,
5676-5679; b) J. Ren, Y. Zhang, J. Zhang, H. Gao, G. Liu, R. Ma, Y. An, D. Kong, L. Shi, Biomacromolecules
2013, 14, 3434-3443.
a) M. Yan, J. Du, Z. Gu, M. Liang, Y. Hu, W. Zhang, S. Priceman, L. Wu, Z. H. Zhou, Z. Liu, Nat.
Nanotechnol. 2010, 5, 48-53; b) M. Zhao, Y. Liu, R. S. Hsieh, N. Wang, W. Tai, K.-I. Joo, P. Wang, Z. Gu, Y.
Tang, J. Am. Chem. Soc. 2014, 136, 15319-15325; c) D. Wu, M. Qin, D. Xu, L. Wang, C. Liu, J. Ren, G.
Zhou, C. Chen, F. Yang, Y. Li, Adv. Mater. 2019, 31, 1807557.
[9]
M. Wang, J. A. Zuris, F. Meng, H. Rees, S. Sun, P. Deng, Y. Han, X. Gao, D. Pouli, Q. Wu, Proc. Natl. Acad.
Sci. USA 2016, 113, 2868-2873.
[10]
a) M. Liu, S. Shen, D. Wen, M. Li, T. Li, X. Chen, Z. Gu, R. Mo, Nano Lett. 2018, 18, 2294-2303; b) J. R.
Clegg, A. S. Irani, E. W. Ander, C. M. Ludolph, A. K. Venkataraman, J. X. Zhong, N. A. Peppas, Sci. Adv.
2019, 5, eaax7946; c) S. Su, Y. Y. Wang, F. S. Du, H. Lu, Z. C. Li, Adv. Funct. Mater. 2018, 28, 1805287.
a) T.-T. Chen, J.-T. Yi, Y.-Y. Zhao, X. Chu, J. Am. Chem. Soc. 2018, 140, 9912-9920; b) P. Yuan, H. Zhang, L.
Qian, X. Mao, S. Du, C. Yu, B. Peng, S. Q. Yao, Angew. Chem. Int. Ed. 2017, 56, 12481-12485; c) P. Yuan,
X. Mao, X. Wu, S. S. Liew, L. Li, S. Q. Yao, Angew. Chem. Int. Ed. 2019, 131, 7739-7743; d) D. Shao, M. Li,
Z. Wang, X. Zheng, Y. H. Lao, Z. Chang, F. Zhang, M. Lu, J. Yue, H. Hu, H. Yan, L. Chen, W. F. Dong, K. W.
[11]
17
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
Leong, Adv. Mater. 2018, 30, 1801198.
[12]
a) D. Y. Ng, M. Arzt, Y. Wu, S. L. Kuan, M. Lamla, T. Weil, Angew. Chem. Int. Ed. 2014, 53, 324-328; b) Y.
Ma, Y. Zhao, N. K. Bejjanki, X. Tang, W. Jiang, J. Dou, M. I. Khan, Q. Wang, J. Xia, H. Liu, ACS Nano 2019,
13, 8890-8902; c) L. Zhang, Z. Wang, Y. Zhang, F. Cao, K. Dong, J. Ren, X. Qu, ACS Nano 2018, 12,
10201-10211; d) D. Yuan, Y. Zhao, W. A. Banks, K. M. Bullock, M. Haney, E. Batrakova, A. V. Kabanov,
Biomaterials 2017, 142, 1-12.
3
4
5
6
7
[13]
[14]
F. Wu, T. Jin, Aaps Pharmscitech 2008, 9, 1218-1229.
8
a) Q.-Y. Hu, F. Berti, R. Adamo, Chem. Soc. Rev. 2016, 45, 1691-1719; b) M. R. Battig, B. Soontornworajit,
Y. Wang, J. Am. Chem. Soc. 2012, 134, 12410-12413.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
[15]
[16]
P. Mahlumba, Y. E. Choonara, P. Kumar, L. C. Du Toit, V. Pillay, Molecules 2016, 21, 1002-1002.
a) X. Liu, P. Zhang, D. He, W. Rꢀdl, T. Preiß, J. O. Rꢁdler, E. Wagner, U. Lꢁchelt, Biomacromolecules 2016,
17, 173-182; b) A. Kim, Y. Miura, T. Ishii, O. F. Mutaf, N. Nishiyama, H. Cabral, K. Kataoka,
Biomacromolecules 2016, 17, 446-453; c) M. Wang, K. Alberti, S. Sun, C. L. Arellano, Q. Xu, Angew.
Chem. Int. Ed. 2014, 53, 2893-2898.
[17]
a) M. Wang, S. Sun, C. I. Neufeld, B. Perez‐Ramirez, Q. Xu, Angew. Chem. Int. Ed. 2014, 53,
13444-13448; b) B. Liu, M. Ianosi-Irimie, S. Thayumanavan, ACS Nano 2019, 13, 9408-9420; c) T. T.
Hoang, T. P. Smith, R. T. Raines, Angew. Chem. Int. Ed. 2017, 56, 2619-2622.
J. Chang, W. Cai, C. Liang, Q. Tang, X. Chen, Y. Jiang, L. Mao, M. Wang, J. Am. Chem. Soc. 2019, 141,
18136-18141.
[18]
[19]
a) H. He, Y. Chen, Y. Li, Z. Song, Y. Zhong, R. Zhu, J. Cheng, L. Yin, Adv. Funct. Mater. 2018, 28, 1706710;
b) J. Li, C. Xie, J. Huang, Y. Jiang, Q. Miao, K. Pu, Angew. Chem. Int. Ed. 2018, 57, 3995-3998.
D. Huo, X. Jiang, Y. Hu, Adv. Mater. 2019, 1904337, doi:1904310.1901002/adma.201904337.
a) L. H. Fu, C. Qi, Y. R. Hu, J. Lin, P. Huang, Adv. Mater. 2019, 31, 1808325; b) S. B. Bankar, M. V. Bule, R.
S. Singhal, L. Ananthanarayan, Biotechnol. Adv. 2009, 27, 489-501.
[20]
[21]
[22]
X. Liu, P. Zhang, D. He, W. Rꢀdl, T. Preiß, J. O. Rꢁdler, E. Wagner, U. Lꢁchelt, Biomacromolecules 2015, 17,
173-182.
[23]
[24]
S. H. Lee, E. Moroz, B. Castagner, J.-C. Leroux, J. Am. Chem. Soc. 2014, 136, 12868-12871.
K. Y. Choi, H. S. Han, E. S. Lee, J. M. Shin, B. D. Almquist, D. S. Lee, J. H. Park, Adv. Mater. 2019, 31,
1803549.
[25]
R. V. Benjaminsen, M. A. Mattebjerg, J. R. Henriksen, S. M. Moghimi, T. L. Andresen, Mol. Ther. 2013,
21, 149-157.
18
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
Scheme 1. Schematic illustration of cascaded, self-catalyzed nanozymogen incorporating
hypoxia-instructed pro-protein engineering and hypoxia-strengthened intracellular protein
delivery. RPAB, a pro-protein of RNase A reversibly caged with hypoxia-cleavable
azobenzene, was designed and co-delivered with GOx using hypoxia-dissociable NCs
consisting of AOEI and HA. Upon systemic injection, the nanozymogen mediated efficient
tumor accumulation, targeted internation in cancer cells, and effective endolysosomal escape.
GOx catalyzed intracellular glucose decomposition and aggravated tumoral hypoxia, which
drove the recovery of RPAB back to the hydrolytically active RNase and concurrently
expedited the degradation of AOEI to release more protein cargoes. Thus, the catalytic
reaction of the nanozymogen was self-accelerated and self-cycled, ultimately leading to
4
5
6
7
8
9
10
11
19
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
cooperative anti-cancer efficacies between GOx-mediated starvation therapy and
RNase-mediated pro-apoptotic therapy.
20
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
4
5
6
7
Figure 1. Characterization and hypoxia-triggered re-activation of RPAB. A) MALDI-TOF
spectra of (a) RPAB, (b) RNase, and (c) RPAB treated with Na2S2O4 (155 mM). B)
SDS-PAGE of RNase and RPAB with or without Na2S2O4 treatment (155 mM). C) Enzymatic
activity of RNase, RPAB, and Na2S2O4 (155 mM)-treated RPAB as measured by the EB assay
(n = 3). D) CD spectra of RNase, RPAB, and Na2S2O4 (155 mM)-treated RPAB.
21
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
Figure 2. Hypoxia-triggered protein release from NCs and GOx-mediated glucose
decomposition. Size A) and TEM images B) of HARPG NCs before and after Na2S2O4
treatment (155 mM) for different time. Scale bar = 100 nm. C) RPAB and GOx release from
HARPG in PBS (pH 7.4) with or without Na2S2O4 (155 mM) (n = 3). D) pH values of glucose
solutions at different concentrations after 1-h incubation with free GOx, HARPG NCs, and
Na2S2O4 (155 mM)-treated HARPG NCs (n = 3). E) H2O2 concentrations of glucose solutions
at different concentrations after 1-h incubation with free GOx, HARPG NCs, and Na2S2O4
(155 mM)-treated HARPG NCs (n = 3).
4
5
6
7
8
9
10
22
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
Figure 3. Cancer cell-targeted intracellular protein delivery, hypoxia-triggered intracellular
protein release, and aggravation of intracellular hypoxia mediated by HARPG NCs in vitro. A)
Uptake of HARPG NCs containing FITC-RPAB in HeLa cells following 4-h incubation as
evaluated by flow cytometry. HeLa cells were pre-treated with HA for 4 h to block CD44 and
the HA-mediated targeting effect. B) CLSM images of HeLa cells treated with free GOx,
HARP NCs, and HARPG NCs for 4 h followed by staining with hypoxia/ROS detection
probes. DFO and Pyo served as positive controls for ROS and hypoxia induction, respectively.
Scale bar = 50 μm. C) Flow cytometry analysis of intracellular ROS and hypoxia levels
following the same treatment in B) (n = 3). D) CLSM images of HeLa cells following
incubation with HARP NCs containing FITB-RPAB and RhB-AOEI under hypoxia or
4
5
6
7
8
9
10
11
12
13
14
o
normoxia at 37 C for 4 h. Cell nuclei were stained with DAPI. Scale Bar = 15 μm. E)
Colocalization ratios between the fluorescence of FITC-RPAB and RhB-AOEI as calculated
from CLSM images in D) (n = 20).
23
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
4
5
6
7
8
9
Figure 4. Anti-cancer efficacy of HARPG NCs in HeLa cells in vitro. A) Cytotoxicity of
HA/AOEI NCs (w/w = 4/1) following 48-h incubation under normoxia or hypoxia at various
AOEI concentrations (n = 3). B) Cytotoxicity of different NCs after 48-h incubation under
normoxia (n = 3). Cytotoxicity of HARP NCs and HAR NCs after 48-h incubation under
normoxia C) or hypoxia D) at various RNase/RPAB concentrations (n = 3). (E) Flow
cytometric analysis of HeLa cells treated with different NCs for 48 h and stained with
Annexin V-FITC/PI.
24
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
Figure 5. Pharmacokinetics and biodistribution HARPG NCs and their capabilities to
aggravate hypoxia in tumors in vivo. A) Pharmacokinetics of Cy5-RPAB-containing HARPG
NCs and free Cy5-RPAB in mice following i.v. injection at 2.5 mg/kg RPAB (n = 3). B)
Fluorescence imaging of HeLa tumor-bearing mice at different time points post i.v. injection
of Cy5-RPAB-containing HARPG NCs and free Cy5-RPAB (2.5 mg/kg RPAB). The red
circle indicates the tumor site. C) Ex vivo fluorescence imaging of excised tumors and major
organs at 24 h post i.v. injection (1: heart; 2: liver; 3: spleen; 4: lung; 5: kidney; 6: tumor). D)
Calculated fluorescence intensity of each tissue in (C) (n = 3). E) PA imaging of tumors by
measuring deoxygenated hemoglobin and oxygenated hemoglobin at different time points
post i.v. injection of PBS, HARP NCs, HARPB NCs, or free GOx (1 mg/kg GOx). (F)
Calculated oxyhemoglobin saturation levels in tumors from PA imaging in (D) (n = 3).
4
5
6
7
8
9
10
11
12
25
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
2
3
Figure 6. In vivo cooperative anti-tumor efficacy of HARPG NCs against HeLa xenograft
tumors. PBS, HARP NCs, HAR NPs, HAG NCs, HARG NCs, and HARPG NCs were i.v.
injected on day 1, 3, and 5 (RPAB, 2.5 mg/kg; RNase, 2.5 mg/kg; GOx, 1 mg/kg). (A) Tumor
volume progression within the 19-d observation period (n = 11). Arrows represent the
injection time. (B) Representative images of tumors harvested on day 19. (C) Survival rate of
mice within the 45-d observation period (n = 8). (D) H&E-stained tumor tissues harvested on
day 6. Scale bar = 100 μm. (E) Cell apoptosis in tumor tissues harvested on day 6 as
determined by TUNEL staining. Scale bar = 100 μm.
4
5
6
7
8
9
10
26
This article is protected by copyright. All rights reserved.

10.1002/anie.202004008
Angewandte Chemie International Edition
1
TOC image
GOx catalysis
RPAB
RNase
CD44 receptor
2
3
4
5
Cascaded nanozymogen was designed to incorporate hypoxia-instructed pro-protein
engineering and hypoxia-strengthened intracellular protein delivery. Hypoxia-activatable
pro-protein and GOx were encapsulated within cell-ingestable and hypoxia-dissociable NCs.
Upon GOx-catalyzed hypoxia aggravation, cascaded intracellular reactions of pro-protein
reactivation, NCs dissociation, and protein release were self-catalyzed and self-cycled, thus
provoking efficient and selective anti-cancer efficacies.
6
7
8
9
10
11
12
27
This article is protected by copyright. All rights reserved.