Self-Assembled Nanomicelles of Affibody-Drug Conjugate with Excellent Therapeutic Property to Cure Ovary and Breast Cancers

Highlights The nanoagent of affibody-drug conjugate made of ZHER2:342 and MMAE was successfully fabricated through molecular self-assembly of the conjugate in aqueous solution for targeted cancer therapy, which is recorded as Z-M ADCN for short. Z-M ADCN shows extraordinary anticancer ability in HER2-positive ovary and breast tumor models with good biosecurity, that nearly eradicated both small solid tumors (about 100  mm3) and large tumors (exceed 500  mm3). Supplementary Information The online version contains supplementary material available at 10.1007/s40820-021-00762-9.


Introduction
Breast and ovary cancers have been the leading cause of cancer-related death for women worldwide [1]. Meanwhile, human epidermal growth factor receptor 2 (HER2), a member of tyrosine kinase cell membrane receptors, has been proved amplified and overexpressed in lots of breast and ovary cancers [2]. Over the last decades, the technology of monoclonal antibodies which target the HER2 receptor developed rapidly and corresponding antibody-drug conjugates (ADCs) have been successfully explored for HER2targeted cancer therapy by using antibodies as vehicles to deliver cytotoxic agents into tumor cells efficiently and selectively [3][4][5][6]. However, there are still some ineluctable deficiencies such as large size, complicated fabrication, unspecific conjugate site, and poor tissue penetration which may influence somewhat the therapeutic efficiency of ADC drugs [7][8][9]. To circumvent the limitations, various smaller protein fragments such as monobodies [10], anticalins [11], DARPins (designed ankyrin repeat proteins) [12], and nanobodies [13] have been developed as alternative drug carriers. Beside these candidates, affibody, a small (6-7 kDa) affinity protein of 58 amino acids organized into a three-helix bundle, has got intensive attention due to its high affinity to a large number of target proteins or peptides [14][15][16]. Compared with antibodies, affibody molecules possess several potential advantages, such as rapid tissue penetration due to the small size, high selectivity with picomolar affinities, and easily obtained by microbial fermentation [17,18]. More importantly, the absence of cysteine in the original affibody sequence provides us an opportunity to introduce cysteine into the sequence for the site-specific conjugation with payloads via thiol chemistry [19,20]. The small size of affibody benefits the tissue penetration, but simultaneously resulting in rapid clearance by kidney. The performances of rapid tumor penetration and fast blood clearance make the affibody molecules suitable for a variety of medical imaging applications, such as positron emission tomography (PET) imaging [21,22], optical and magnetic resonance imaging (MRI) [23,24], and fluorescence-guided surgery [25,26], but evidently are inappropriate for cancer therapy [27].
Recently, a few researchers tried to conjugate affibody molecules with cytotoxic drugs forming affibody mediated targeting anticancer medicine. For instance, Jacek Otlewski and co-workers constructed two affibody-drug conjugates by binding monomethyl-auristatin E with monomeric Z HER2:2891 and dimeric Z HER2:4 [28,29]. However, these works did not involve in molecular self-assembly, and only provided the experiment data in vitro without in vivo. In order to prolong the half-life of affibody-based drugs, Torbjörn Gräslund and co-workers expressed fusion proteins containing albumin binding domain (ABD) and dimeric Z HER2:2891 , and then combined them with maytansine derivates [30]. The resulting prodrug could lengthen survival time of the mice bearing HER2 overexpressing tumors, while the data concerning tumor volumes had not been discussed. Overall, it is a feasible way that combining affibody molecules with anticancer drugs for targeted cancer therapy, but how to improve pharmacokinetic property is still an imperious demand for clinic application. It is known that self-assembled nanoagents have a longer retention time in the bloodstream compared with their precursor molecules [31,32]. It can be imagined that if an affibody-drug conjugate could self-assemble into nanoagents, that would substantially improve its pharmacokinetics profile. Such a kind of precisely targeted delivery and highly efficient therapeutic system is promising.
Inspired by this concept, we designed and prepared a nanoscale agent consisting of affibody-drug conjugate for targeted cancer therapy ( Fig. 1). At first, a unique cysteine codon was attached to the 5' end of the specific DNA sequence of Z HER2:342 , and then the Z HER2:342 -Cys was expressed by the E. coli system [15,33]. Subsequently, the affibody Z HER2:342 -Cys was conjugated with anticancer drug monomethyl-auristatin E (MMAE) through a linker containing a maleimide attachment group, a cathepsin B-responsive valine-citrulline (Val-Cit) dipeptide, and a para-amino-benzyloxycarbonyl (PABC) self-immolative spacer [34,35]. Finally, the conjugate as a precursor was self-assemble into nanomicelles in water owing to its inherent amphiphilic nature (Fig. 1a). We coined such nanomicelles as affibody-drug conjugate nanoagent (ADCN). In this work, the affibody-drug conjugate is made of Z HER2:342 and MMAE, so we designate these nanomicelles as Z HER2:342 -MMAE affibody-drug conjugate nanoagent (Z-M ADCN). The nanoscale characteristics of Z-M ADCN results in elevating both pharmacokinetics and in vivo targeting performance than those of free Z HER2:342 because there are a number of Z HER2:342 on the nanoagent surface, Z-M ADCN can more effectively accumulate in tumor and be internalized by cancer cells through HER2-specific receptor-mediated endocytosis. Because of the abundant protease in lysosomes of tumor cells [36,37], MMAE can be rapidly released in chemically 1 3 unmodified form through facile proteolysis, resulting in effective inhibition on cellular proliferation (Fig. 1b).

Biosynthesis and Purification of Recombinant Affibody
Plasmid pET19b-Cys-Affibody was constructed for recombinant expression of Z HER2:342 -Cys under the IPTG-inducible T7 promoter. The plasmid was transformed into E. coli BL21 (DE3) and the expression was induced with isopropyl-β-D-thiogalactopyranoside (IPTG) at 1 mM overnight at 16 °C. The obtained cells were sonicated with an Ultrasonic Homogenizer. The resulting supernatant was purified by Ni-nitrilotriacetyl (Ni-NTA) agarose column. The protein was collected and the purity was analyzed by 15% SDS-PAGE. The obtained affibody was purified by dialysis against water with a dialysis bag (MWCO, 3.5 kDa) and lyophilized for storage. Finally, the molecular weight was verified by MALDI-TOF-MS.

Cleavage of His-tag from Recombinant Affibody
The experiment of removing His-tag from recombinant affibody were executed as the manufacturer's instructions. Briefly, 50 mg His-EK-Cys-Z HER2:342 was dissolved in 50 mL of enterokinase reaction buffer, followed by adding with enterokinase of 100 U, then the solution was stirred for 24 h at 25 °C and the His-tag residue was removed subsequently by dialysis against water with a dialysis bag (MWCO, 3.5 kDa) for 12 h. Finally, the obtained Z HER2:342 -Cys was lyophilized for storage. The molecular weight was verified by MALDI-TOF-MS.  10 h at 25 °C. After that, a slightly blue solution was obtained and followed by dialysis against PBS with a dialysis bag (MWCO, 14 kDa) for 6 h to remove the uncombined raw materials, during which the PBS was exchanged every 2 h. Finally, a stable and bluish Z-M ADCN solution was obtained. The molecular weight was verified by MALDI-TOF-MS.

Labeling the Conjugates
The Cy5.5-labeled Z HER2:342 -Cys were prepared as follow: 2 µmol of water soluble Sulfo-Cyanine5.5 NHS ester (Cy5.5) and 1 µmol Z HER2:342 -Cys were dissolved in 5 mL PBS (pH 7.4), and the mixture was incubated at 25 °C for 4 h. Then Cy5.5-labeled Z HER2:342 -Cys solution was purified by using the PD MiniTrap G-25 column. The content of Cy5.5 was determined by a NanoDrop 2000/2000C spectrophotometer. The Cy5.5-labeled Z-M ADCN were prepared as follow: 2 µmol of water soluble Sulfo-Cyanine5.5 NHS ester (Cy5.5) and 1 µmol Z-M ADCN was added into 4 mL PBS at 25 °C and the solution was stirred for 4 h. After that, a bluish-green solution was obtained and followed by dialysis against PBS with a dialysis bag (MWCO, 3 kDa) for 4 h to remove the uncombined Cy5.5, during which the PBS was exchanged every 2 h. Finally, a stable and bluish-green solution was obtained. The content of Cy5.5 was measured by a NanoDrop 2000/2000C spectrophotometer.

Biospecific Interaction Analysis
The affinity of the interactions between Z HER2:342 -Cys or Z-M ADCN and extracellular domain (ECD) of HER2 were analyzed by a Biacore 8 K instrument. HER2 ECD was immobilized on a CM5 chip by amine coupling firstly and the immobilization level was about 1000 RU for experiments. Then the affinity constants were detected by injecting a series dilution concentration as required. antibiotic-antimycotic. The cells were cultured at 37 °C under an atmosphere containing 5% CO 2 .

Cellular Uptake Analysis
Flow cytometry (FCM) and confocal laser scanning microscopy (CLSM) were employed to study the cellular uptake behaviors. For the CLSM analysis, 2 × 10 5 SKOV-3 cells or MDA-MB-231 cells were planted into each well and incubated overnight, followed by the addition of Cy5.5-labeled Z-M ADCN (10 μg mL −1 ), the cells were incubated at 37 °C for timed intervals. Then the cell nuclei were stained with Hoechst 33,342 for 15 min. Images were taken by using a laser scanning confocal microscope. For the FCM analysis, the SKOV-3 cells were incubated with Cy5.5-labeled Z-M ADCN (10 μg mL −1 ) at 37 °C for timed intervals. Then the cells were collected and analyzed by flow cytometry. For the binding specificity assay, SKOV-3 cells were preincubated with Z HER2:342 -Cys (10 μg mL −1 ) for 1 h and then treated with Cy5.5-labeled Z-M ADCN (10 μg mL −1 ) for an additional 4 h. Then the cells were collected and analyzed by flow cytometry.

Cytotoxicity Evaluation
CCK-8 cell proliferation assay was used to investigate the in vitro cytotoxicity of free MMAE and Z-M ADCN. Briefly, 5 × 10 3 cells were planted into each well and incubated overnight, followed by the addition of Z-M ADCN or free MMAE at the required concentrations. With a subsequent 48 h incubation, CCK-8 solution of 10 μL was added to each well and incubated for an additional 1 h. Finally, the absorbance of each well was detected at 450 nm by a BioTek Synergy H4 spectrophotometer.

Apoptosis Study
SKOV-3 cells were planted at 5 × 10 5 cells per well and incubated overnight, followed by the treatment of PBS, MMAE (3 nM), Z-M ADCN (equivalent dose of MMAE at 3 nM) for 24 h. After that, the cells were harvested and managed as the manufacturer's protocol. The treated samples were finally measured by flow cytometry.

Animals and Tumor Models
The Animal Ethics Committee of Shanghai Jiao Tong University approved all the animal experiments. All procedures were in compliance with the relevant guidelines and regulations.
As the establishment of tumor models, 200 µL of cell suspension containing 3 × 10 6 SKOV-3 cells or a mixture of 100 µL of cell suspension containing 3 × 10 6 BT474 cells and 100 µL of Matrigel Membrane Matrix were injected subcutaneously in the right flank region of nude mice. Antitumor experiments were performed when the tumors volumes were around 100 or 500 mm 3 .

In vivo Optical Imaging and ex vivo Biodistribution Analysis
Two groups (n = 3) of SKOV-3 xenografts-bearing mice were prepared for the in vivo optical imaging study, followed by treated with 200 μL of Cy5.5-labeled Z HER2:342 -Cys and Cy5.5-labeled Z-M ADCN which possessed similar intensity of Cy5.5, respectively. The fluorescence distribution was measured by a Kodak multimode imaging system after injection of 1, 2, 4, and 8 h. Two groups (n = 9) of SKOV-3 xenografts-bearing mice were prepared for the ex vivo biodistribution analysis, followed by treated with 200 μL of Cy5.5-labeled Z HER2:342 -Cys and Cy5.5-labeled Z-M ADCN which possessed similar intensity of Cy5.5, respectively. At the timepoint of 1, 4, and 8 h after injection, three of the mice in each group were sacrificed, the tumor tissues and major organs were collected and subsequently analyzed by a Kodak multimode imaging system.

In vivo Antitumor Activity
For small SKOV-3 tumor models: when the volumes of tumors were about 100 mm 3 , mice were divided into five groups (n = 5) randomly and administered with 200 μL of PBS, MMAE (0.6 mg kg −1 ), Z-M ADCN (MMAE-equiv. dose at 0.6, 0.8, and 1 mg kg −1 , respectively) once every three days for five times, respectively. The tumor volumes as well as body weights were monitored every three days for 21 days in total. The formula: V (mm 3 ) = 1/2 × length (mm) × width (mm) 2 was used to calculate tumor volume (V). On the day 21 after the initial treatment, all the mice were euthanized and the tumor were collected for further study.
For large SKOV-3 tumor models: when the volumes of tumors were about 500 mm 3 , mice were divided into three groups (n = 5) randomly and administered with 200 μL of PBS, MMAE (1 mg kg −1 ), Z-M ADCN (MMAE-equiv. dose at 1 mg kg −1 ) once every five days for five times, respectively. The tumor volumes and body weights were monitored every five days for 35 days in total. After the period of 35 days, the cured mice were observed for further 20 days, the unhealed mice were euthanized and the tumors and major organs were collected for further study.
For small BT474 tumor models: when the volumes of tumors were about 100 mm 3 , mice were divided into three groups (n = 5) randomly and administered with 200 μL of PBS, MMAE (1 mg kg −1 ), Z-M ADCN (MMAE-equiv. dose at 1 mg kg −1 ) once every five days for five times, respectively. The tumor volumes and body weights were monitored before each injection. After the total treatment, all the mice were euthanized, the tumors were collected for further study.
For large BT474 tumor models: when the volumes of tumors were about 500 mm 3 , mice were divided into two groups (n = 5) randomly and administered with 200 μL of PBS, Z-M ADCN (MMAE-equiv. dose at 1 mg kg −1 ) once every five days for five times, respectively. The tumor volumes and body weights were monitored before each injection. After the treatment of 25 days, the cured mice were observed for further 30 days, the unhealed mice were euthanized and the tumors were collected for further study.

Preparation and Characterization of Z HER2:342 -MMAE Conjugate Nanoagent
Firstly, the initial recombinant affibody molecule His-EK-Cys-Z HER2:342 was produced in Escherichia coli BL21(DE3) (Fig. S1) . 2a). Besides, the high binding affinity performance of affibody molecules intensively related to the α-helical structure [33,38]. So circular dichroism (CD) spectroscopy was adopted to characterize the structure of Z HER2:342 -Cys and the results verified α-helical secondary structure dominating in Z HER2:342 -Cys. This is consistent with the previous report ( Fig. S3) [33], and confirms its HER2-binding ability. All these data demonstrate that Z HER2:342 -Cys were successfully prepared. Afterward, Z HER2:342 -Cys was dissolved in PBS and then a dimethyl sulfoxide (DMSO) solution of Mc-VC-PABC-MMAE was added dropwise into it under the stirring (Fig.  S4). As the reaction progress, a light blue solution with opalescence was obtained and then purified by dialysis against PBS to remove the residues of raw materials. The MALDI-TOF-MS analysis of the resulting product showed a single peak at 8438.14 ( Fig. 2b), which is consistent with the theoretical mass of Z HER2:342 -MMAE conjugate (8394.48). Furthermore, some parallel experiments were carried out to study the self-assembly behavior of the resulting Z HER2:342 -MMAE conjugates (  (Fig. 2c).
The zeta potential of Z-M ADCN in PBS was − 2.3 mV (Fig.  S6), which inidcated the high density of affibody located in the corona of Z-M ADCN to form negative-charged surface. The transmission electron microscopy (TEM) images of Z-M ADCN clearly exhibited that they were spherical nanomicelles with an average size of 121.9 nm (Fig. 2d).
On the contrary, the TEM images of Z HER2:342 -Cys and Mc-VC-PABC-MMAE only displayed some irregular fragments, obviously different from the morphology of Z-M ADCN (Fig. S7). In addition, the critical micellar concentration (CMC) of the Z HER2:342 -MMAE conjugate was determined as 8.2 μg mL −1 (Fig. S8), indicating the relatively high stability of Z-M ADCN in aqueous solution. Furthermore, Z-M ADCN demonstrated the good storage stability because their average size and distribution were substantially unchanged during 15 days (Fig. S9). Meanwhile, the time-dependent changes in the diameter of the Z-M ADCN in water containing 5 or 10% FBS were determined by DLS, respectively, revealing the high serum stability of Z-M ADCN (Fig. S10).
All above data confirmed that Z-M ADCN were successfully prepared.

Binding Specificity and Affinity Analysis
The high binding affinity of affibody molecules can be retained still even after they are conjugated with some attachments [29,30,39]. Here, the biospecific interaction analysis between Z HER2:342 -Cys (or Z-M ADCN) and ECD of HER2 was performed by a Biacore 8 K instrument (Figs. 2e, f and Table S1). The association rate constant (k a ) of Z-M ADCN was 1.49 × 10 5 M −1 s −1 , obviously higher than that of Z HER2:342 -Cys (9.07 × 10 4 M −1 s −1 ). This can be attributed to the synergistic effect of Z HER2:342 segments on the surface of Z-M ADCN binding with ECD of HER2. Meanwhile, the dissociation rate constant (k d ) of Z-M ADCN was 9.59 × 10 -4 s −1 , also higher than that of Z HER2:342 -Cys (4.82 × 10 -4 s −1 ). But the equilibrium dissociation constant (K D ) of Z-M ADCN was 6.44 × 10 -9 M, similar to that of Z HER2:342 -Cys (5.31 × 10 -9 M). All above results indicated that conjugating with MMAE ramification did not affect the affinity of Z HER2:342 -Cys to the tumor cells with HER2 receptors. Actually, the affinity of both Z HER2:342 -Cys and Z-M ADCN to bind ECD of HER2 was in the nanomolar range, which was sensitive enough for targeting therapy [40].

In Vitro Studies of Z-M ADCN
The fluorescence dye of Cy5.5 was adopted to label Z-M ADCN to study the cellular uptake of them. The morphology of Cy5.5-labeled Z-M ADCN was consistent with that of Z-M ADCN by DLS and TEM measurements ( In addition, the fluorescence signal in MDA-MB-231 cells (human breast cancer cell lines with low HER2 expression) was also obviously weaker than that in SKOV-3 cells after incubated with Cy5.5-labeled Z-M ADCN for 4 h (Fig. 3c).
All the data demonstrated that the interaction of Z-M ADCN with the cancer cells was HER2-specific receptor-mediated indeed [39]. It has been proved that the toxicity of affibody-drug conjugates is receptor dependent [28][29][30]. Therefore, we speculated Z-M ADCN would exhibit evidently selective cytotoxicity toward the cells with various HER2 expression levels. Herein, the in vitro cytotoxicity of Z-M ADCN was evaluated against SKOV-3 cells, BT474 cells (human breast cancer cell lines with high HER2 expression) and  (Fig. 3g). The ratio of apoptosis cells induced by MMAE and Z-M ADCN was 95.5% and 91.2%, respectively, indicating Z-M ADCN has a substantially similar ability to induce cell apoptosis compared with that of free MMAE.

Pharmacokinetic and Biodistribution Analysis
It is known that the half-life of free affibody is short due to its low molecular weight, alternatively, self-assembled aggregates always have enhanced blood circulation time compared with that of their precursors [43][44][45]. Herein, pharmacokinetics evaluation of Z-M ADCN was performed by i.v. injection of Cy5.5-labeled Z-M ADCN to Sprague-Dawley (SD) rats (∼ 200 g), and the rats injected with Cy5.5-labeled Z HER2:342 -Cys were used as control (Fig. 4a). Obviously, the metabolic rate of Z-M ADCN was slower than that of free Z HER2:342 -Cys. The fluorescence intensity in the bloodstream for Z-M ADCN group is obviously stronger than that of Z HER2:342 -Cys group. Specifically, the fluorescence intensity in the bloodstream for Z-M ADCN group keeps at a high level of 0.86 × 10 4 a.u. after injection for 8 h, whereas that for Z HER2:342 -Cys group is only 0.4 × 10 4 a.u. after injection for the same time. These results clearly indicate that the pharmacokinetics performance of Z-M ADCN is considerably improved than that of free Z HER2:342 -Cys.
The prolonged blood retention time of Z-M ADCN facilitates the HER2-specific drug accumulation in tumor. The in vivo fluorescence imaging of nude mice bearing SKOV-3 tumor was adopted to investigate the accumulation of Z-M ADCN in tumor (Fig. 4b). From 1 to 8 h after injection, the fluorescence intensity of Cy5.5-labeled Z HER2:342 -Cys group decreased quickly due to the small molecular weight of Z HER2:342 -Cys. On the contrary, the fluorescence intensity of Cy5.5-labeled Z-M ADCN group was kept high in the same period after injection, especially at the tumor site. These results confirm that Z-M ADCN have the longer retention time in bloodstream and efficiently accumulate in tumor site. To further investigate the biodistribution of Z-M ADCN, SKOV-3 xenograft-bearing mice were euthanized at 1, 4, and 8 h after injection and the tumors and major organs were collected subsequently for the ex vivo imaging to quantitatively assess the accumulation of Z-M ADCN in tissues. Strikingly, Cy5.5-labeled Z-M ADCN mainly accumulated in tumor site as well as the fluorescence intensity in tumor tissues remains similar within 8 h after injection (Fig. 4c), which is consistent with the results of in vivo imaging. For Cy5.5-labeled Z HER2:342 -Cys group, the strong fluorescence was observed mainly in kidney at 1 h after injection and it became rather weaker in main organs as well as tumor at 8 h after injection (Fig. S13). Taken together, Z-M ADCN has significantly enhanced pharmacokinetics property, extraordinary tumor-homing ability, and great potential for targeting cancer therapy.

In Vivo Antitumor Activity Studies
Encouraged by the excellent targeting performance, the antitumor activities of Z-M ADCN were further assessed using the SKOV-3 xenograft-bearing mice (Fig. 5a). When the tumors grew up to about 100 mm 3 , the mice were divided into five groups randomly and then intravenously injected with PBS, MMAE (0.6 mg kg −1 ), Z-M ADCN (MMAEequiv. dose, 0.6, 0.8, and 1 mg kg −1 , respectively) once every three days for five times, respectively. Herein, MMAE-equiv. dose means the content of free MMAE in Z-M ADCN. It was exhibited that MMAE-equiv. dose at 0.6 or 0.8 mg kg −1 could significantly inhibit the tumor growth and several tumors were eradicated after five treatments (Fig. 5b). Moreover, with increasing MMAE-equiv. dose to 1 mg kg −1 , the growth of the tumors was almost completely suppressed, and four out of five mice were cured and the tumor of the last mice shrank into a small scab and kept unchanged during a nine-day treatment-free period (Fig. 5d-f). After the total treatment, the mice were euthanized and tumors were collected. The tumors weight was recorded to calculate the tumor inhibitory rate (TIR) (Fig. S14). Compared with that of the PBS group, the TIRs of Z-M ADCN groups (dose of MMAE at 0.6, 0.8, and 1 mg kg −1 ) are 98.7%, 99.6%, and 99.8%, respectively, confirming the excellent tumor inhibition capability of the Z-M ADCN. Meanwhile, owing to serious side effects of free MMAE, the mice treated with MMAE at 0.6 mg kg −1 exhibited weight loss strikingly and all the mice in this group died after four injections (Fig. 5c). However, for Z-M ADCN groups, the body weights of mice only decreased slightly even at the equivalent dose of MMAE being 1 mg kg −1 , but recovered to normal levels once the treatment was finished, which verified the controllable side effects and the acceptable security of Z-M ADCN.
In clinic, robust therapeutic effects usually depend on very early treatment of tumors, and large tumors are still highly challenging to cure and always respond poorly to therapy [46,47]. Even in mouse cancer models, large tumors also lead to limited therapeutic effect. Nonetheless, we challenged to treat the tumors around 500 mm 3 by using Z-M ADCN (MMAE-equiv. dose, 1 mg kg −1 ) and free MMAE used as a control (Fig. 5g). Herein, we extended the time interval between two injections from three days to five days to avoid the slight side effect of Z-M ADCN at MMAEequiv. dose of 1 mg kg −1 (Fig. 5c). As expected, free MMAE at 1 mg kg −1 exhibited limited inhibition ability to such large tumors and led to severe weight loss of the mice, while one injection of Z-M ADCN initiated significant tumor shrinking, and almost all the tumors shrank to inappreciable ones after five injections (Fig. 5h, i), even though the initial volume of two tumors in this group were about 1000 mm 3 (Figs. 5k and S15). After five injections, three out of the five treated mice were tumor-free and the cured mice kept disease-free and never relapse in the next 30 days (Fig. S16). The average weight of the residual tumors of Z-M ADCN group was about 0.004 g, which means a high tumor inhibition rate of 99.8% (Fig. 5j). To further confirm the antitumor and biosecurity performance of Z-M ADCN (MMAE-equiv. dose, 1 mg kg −1 ), H&E and TUNEL staining and serum biochemistry assays after different treatments were analyzed. For H&E and TUNEL staining, the results showed that Z-M ADCN group induced obvious apoptosis and large area of necrosis within the residue tumors, while the control group showed regular proliferation of tumor cells (Figs. 5l and S17). The outcomes of H&E and TUNEL assay combining with the fact that the thin tissue of the two residue tumors did not relapse during a fifteen days treatment-free period after the last injection (Fig. 5h), and the tiny mass as shown in the bottom line kept unchanged for nine days till the mice was euthanized (Fig. 5f), so we suppose that the tiny tissue would be a scab of fibroblasts without living tumor cells. In addition, serum biochemistry assays of liver function parameters (alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (ALP)) and kidney function parameters (blood urea nitrogen (BUN), creatinine (CRE), and uric acid (UA)) were also measured (Fig. S18). The results indicated that the treatment with free MMAE in mice led to greatly increased hepatotoxicity and nephrotoxicity, whereas the mice treated by Z-M ADCN showed no signs of related toxicity. Among the hepatotoxicity related parameters of ALT, AST, and ALP, and the nephrotoxicity related parameters of BUN, CRE, and UA, the average levels increased in the free MMAE group by 1.9-and 1.3fold, respectively, compared with that of Z-M ADCN group. Meanwhile, regional cell atrophy was also observed from the liver and kidney histology analysis for free MMAE group, while the morphology of tissues treated with Z-M ADCN was similar to those treated with PBS (Fig. S19), proving the extraordinary biosafety performance of Z-M ADCN.
Inspired by the extraordinary anticancer performance in SKOV-3 tumor model, we further evaluated the anticancer activity of Z-M ADCN against BT474 tumor model. When tumors grew up to around 100 mm 3 (Fig. 6a), the mice were divided into three groups randomly and then intravenously injected with PBS, MMAE (1 mg kg −1 ) and Z-M ADCN (MMAE-equiv. dose at 1 mg kg −1 ) once every 5 days for five times, respectively (Fig. 6b-d). The treatment with free MMAE at 1 mg kg −1 showed negligible inhibition ability, but led to unacceptable weight loss and death of the mice, while Z-M ADCN at 1 mg kg −1 MMAE-equiv. dose could significantly reduce the volumes of tumors accompanied with no weight loss of the mice during the treatment period.
Finally, all the mice of Z-M ADCN group were cured after the treatment (Fig. 6e), confirming the extraordinary tumor inhibition capability of Z-M ADCN.
Furthermore, the antitumor activity of Z-M ADCN against large BT474 tumors with volumes around 500 mm 3 were also evaluated (Fig. 6f). Following the injection of Z-M ADCN (MMAE-equiv. dose, 1 mg kg −1 ), obviously tumor shrinking were observed, four out of the five tumors shrank to undetectable dimension after five treatments (Figs. 6g, j and S20) and the cured mice did not relapse in the next 30 days (Fig. S21), meanwhile, the body weights of mice kept unchanged during the treatment period (Fig. 6h). After 25 days treatment, the only one residual tumor was dissected, and the average weight of the tumor in Z-M ADCN group was about 0.003 g, which means a 99.8% tumor inhibition rate compared with that of PBS group (Fig. 6i). Moreover, H&E and TUNEL staining of the residual tumor in Z-M ADCN group exhibited more extensive cell apoptosis, comparing with that of PBS group (Figs. 6k and S22). All above results further verify the effective biosecurity and superior anticancer ability of Z-M ADCN.

Conclusions
In summary, a precisely targeting nanodelivery system (Z-M ADCN) was developed by the molecular self-assembly of Z HER2:342 -MMAE conjugate. Z-M ADCN can specifically bind to HER2 which overexpressed in various cancer cells and enter tumor cells by endocytosis, then MMAE can be released through the degradation of valinecitrulline (Val-Cit) dipeptide group due to the cathepsin B enzyme in cancer cells. Meanwhile, the nanoscale characteristic of Z-M ADCN leads to longer retention time in blood and higher drug accumulation in tumor. Besides, the core shell structure of Z-M ADCN with the cytotoxic MMAE inside the Z HER2:342 corona results in lower side effect to normal organs. All of these result in superior antitumor efficacy and high biosecurity of Z-M ADCN in vitro and in vivo. The relative tumor inhibition rate reaches 99.8% for both ovary (SKOV-3) and breast (BT474) cancer models and the all mice including the cured ones and that with a scab did not relapse until sacrificed. Looking forward, different categories of affibody molecules can be integrated with diverse anticancer drugs for various cancer therapy, and we expect that our strategy will be expanded and developed into a big family of Affibody-Drug Conjugate Nanoagent.