Hierarchically tumor-activated nanoCRISPR-Cas13a facilitates efficient microRNA disruption for multi-pathway-mediated tumor suppression

Rationale: CRISPR-Cas13a is an efficient tool for robust RNA knockdown with lower off-target effect, which may be a potentially powerful and safe tool for cancer gene therapy. However, therapeutic effect of current cancer gene therapy that targeting monogene was compromised by the multi-mutational signal pathway alterations of tumorigenesis. Methods: Here, hierarchically tumor-activated nanoCRISPR-Cas13a (CHAIN) is fabricated for multi-pathway-mediated tumor suppression by efficient microRNA disruption in vivo. A fluorinated polyetherimide (PEI; Mw=1.8KD) with graft rate of 33% (PF33) was utilized to compact the CRISPR-Cas13a megaplasmid targeting microRNA-21 (miR-21) (pCas13a-crRNA) via self-assemble to constitute a nanoscale 'core' (PF33/pCas13a-crRNA), which was further wrapped by modified hyaluronan (HA) derivatives (galactopyranoside-PEG2000-HA, GPH) to form CHAIN. Results: The dual-tumor-targeting and tumor-activated CHAIN not only manifested long-term circulation, but augmented tumor cellular uptake and endo/lysosomal escape, thus achieving efficient transfection of CRISPR-Cas13a megaplasmid (~ 13 kb) in tumor cells with minimal toxity. Efficient knockdown of miR-21 by CHAIN restored programmed cell death protein 4 (PDCD4) and reversion‐inducing‐cysteine‐rich protein with Kazal motifs (RECK) and further crippled downstream matrix metalloproteinases-2 (MMP-2), which undermined cancer proliferation, migration and invasion. Meanwhile, the miR-21-PDCD4-AP-1 positive feedback loop further functioned as an enhanced force for anti-tumor activity. Conclusion: Treatment with CHAIN in hepatocellular carcinoma mouse model achieved significant inhibition of miR-21 expression and rescued multi-pathway, which triggered substantial tumor growth suppression. By efficient CRISPR-Cas13a induced interference of one oncogenic microRNA, the CHAIN platform exerted promising capabilities in cancer treatment.


Introduction
Along with the mounting knowledge of tumorigenesis molecular mechanisms, gene therapy offers a potential choice for cancer treatment [1,2]. Current cancer gene therapy mainly focuses on delivering RNA/DNA system targeting one single gene [3][4][5][6][7]. However, neoplastic cells are driven by a combination of gene abnormalities or multiple signal pathway alterations on the genetic level, which led to Ivyspring International Publisher tumorigenesis [8][9][10]. Due to the sophisticated mechanism of cancer, targeting only one signaling molecules in cancer has shown only short-lived or modest clinical benefit [11]. Thus, it is becoming increasingly clear that cancer gene therapy should regulate several targets in the multi-pathway signal network.
MicroRNAs (miRNAs), one class of small noncoding RNAs, were found to interfere the gene expression and subsequently promote or inhibit cell differentiation, proliferation and apoptosis via regulating multiple downstream genes [12,13]. The abnormal expression of miRNAs contributes to various cancers [14]. Many solid cancers contain high levels of the oncogenic microRNA miR-21 (miR-21), whose overexpression encourages cancer cell invasion, metastasis, proliferation, and tumor formation [15][16][17]. Therefore, inhibiting miR-21 expression and modulating the downstream multi-signal pathways may be an ideal strategy for cancer therapy.
The clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated protein (Cas) adaptive immunity systems evolved to protect bacteria and archaea against foreign nucleic acids of viruses and other genetic elements [18][19][20][21]. Among them, CRISPR-Cas13a (originating from Leptotrichia wadei), has been reported as an efficient RNA knockdown tool with competitive efficiency to RNA interference (RNAi) technology [22][23][24]. Cas13a recognizes and cleaves single-stranded RNA target with a protospacer flanking sequence (PFS: A, U, or C) by the direction of a single CRISPR RNA (crRNA) containing a 28-nt spacer sequence, which largely minimized the off-target effects [22]. Furthermore, due to its direct RNA targeting, CRISPR-Cas13a system can avoid potential risks induced by manipulating genomic DNA, such as chromosomal instability and oncogene activation [25,26], which showed great potential for cancer treatment by gene edting. However, the large size of CRISPR-Cas13a plasmid (~ 13 kb) still challenges its safe and effective delivery in vivo.
The exertion of genetic therapy system calls for sufficient editing system molecules inside target cells [6,27,28]. Therefore, an efficient and non-cytotoxic delivery system will be the basic prerequisite for effective therapeutic application. Currently, due to their long-term expression, high infection efficiency and wide hosts variety, viral vectors are powerful weapons for gene therapy and oncolytic viral therapy [29][30][31]. Most viruses are nanoscale particles with a 'core-shell' structure: a nucleic acid packaged core and a surrounding envelope protein shell [32,33]. The excellent gene delivery ability of viral vectors is due to their sophisticated infection mechanisms. Usually, viruses enter host cells via receptor-mediated endocytosis, sequentially unpacking the envelope and capsid to release the genetic cargo [34]. However, undesirable immune responses, limited gene packaging capability or aberrant gene expression limited their clinical applications [35][36][37]. In solving these problems, several reports had revealed that 'core-shell' virus-mimicking gene delivery systems displayed excellently in cancer gene therapy [38][39][40][41][42].
Herein, inspired by the structure and infection pathway of viruses, we designed hierarchically tumor-activated nanoCRISPR-Cas13a (CHAIN) for multi-pathway-mediated tumor suppression by efficient microRNA disruption in vivo (Figure 1). A fluorinated polyetherimide (PEI; Mw=1.8KD) with graft rate of 33% (PF 33 ) was utilized to compact the pCas13a-crRNA megaplasmid via self-assemble to constitute a nanoscale 'core' (PF 33 /pCas13a-crRNA), which was further wrapped by modified hyaluronan (HA) derivatives (galactopyranoside-PEG 2000 -HA, GPH) to form CHAIN. The CHAIN could long circulate in the blood due to its PEGylated anionic shell, and actively targeting HCC cells via overexpressed asiaglycoprotein receptor (ASPGR) and CD44 receptor with the help of galactopyranoside (Gal) and HA [43,44], respectively. After internalization, HA layer could be decomposed by hyaluronidase (HAase) in the endo/lysosomal system, and the re-exposed cationic inner 'core' which could efficiently promote lysosomal escape and released the pCas13a-crRNA megaplasmid into the cytoplasm. Following the miR-21 downregulation by CRISPR/Cas13a system, reversion-inducing-cysteine-rich protein with Kazal motifs (RECK) and programmed cell death protein 4 (PDCD4) were restored and matrix metalloproteinases-2 (MMP-2) was inhibited. Furthermore, the miR-21-PDCD4-AP-1 positive feedback loop further serves as a reinforcement for cancer therapy. Therefore, CHAIN collectively induced cancer cell apoptosis and minimized cell proliferation, metastasis and invasion. Prospectively, the CHAIN platform offers a promising strategy for multi-signal-pathways regulation in cancer therapy by efficient CRISPR-Cas13a-mediated interference of one oncogenic microRNA.

Results and discussion
Preparation and characterization of CHAIN PF 33 (substitution degree of fluorine on PEI is about 33%) and GPH (substitution degree of Gal-PEG on HA is about 33%) were synthesized and then characterized via 19 F-NMR and 1 H-NMR, respectively ( Figure S1 and Figure S2). Herein, we first utilized PF 33 to bind the CRISPR-Cas13 megaplasmid (~ 13 kb). As shown in Figure 2A-B, PF 33 /pCas13a showed a narrow size distribution with an average diameter of 117 ± 4 nm and a positive zeta potential of 23.1 ± 0.3 mV. When PF 33 /pCas13a was coated with the multifunctional polymer (GPH) to form CHAIN/ pCas13a, the hydrodynamic diameter shifted to 153 ± 3 nm, while the zeta potential was reduced to -20.8 ± 0.4 mV. Next, the morphology of PF 33 /pCas13a and CHAIN/pCas13a was confirmed by TEM ( Figure 2C). Then, we investigated the enzymatic sensitivity of CHAIN/pCas13a. After incubation with HAase, the 'shell' GPH was degraded and dissociated from the PF33/pCas13a core ( Figure 2C) and the particle size potential of CHAIN/pCas13a decreased from 153 ± 3 nm to 129 ± 5 nm, while the zeta potential reversed to 5.4 ± 0.4 mV ( Figure 2B). These results suggested that CHAIN was a well-organized 'core-shell' structure and had agile enzyme sensitiveness.
A gel shift assay further confirmed the pCas13a binding ability of PF33. When PF 33 and pCas13a was at mass ratios 1:1, the pCas13a was completely bound by PF 33 , and coating with the negatively charged GPH had no influence on pCas13a encapsulation ( Figure  2D). Furthermore, CCK-8 assays were performed to test the cytotoxicity of PF 33 and GPH on HepG2 and LO2 cells. Compared with PEI 25K, GPH and PF 33 showed little cytotoxicity on both cell lines ( Figure  2E-F).

Cellular internalization, endo/lysosomal escape, and transfection in vitro
The CD44 expression of HepG2 cells was evaluated by flow cytometry before analyzing the cellular internalization efficiency of CHAIN [45]. The results showed CD44 was highly expressed (~ 100%) on HepG2 cells ( Figure S3). Since ASGPR is an endocytotic receptor expressed primarily on the surface of hepatocytes [46], we did not analyze its expression levels in HepG2 cells.
Then, the cellular uptake efficiency of the YOYO-1-labeled CHAIN/pCas13a was measured in HepG2 cells. As shown in Figure 3A and Figure S4, PF 33 /pCas13a and CHAIN/pCas13a exhibited comparable cellular uptake efficiency (> 95%), which performed more excellent than that of PEI 25K/pCas13a (~ 74%, p < 0.001, p < 0.01, respectively). Moreover, HAC/pCas13a performed slightly less inferior than CHAIN/pCas13a ( Figure S5). In order to verify whether the excellent cellular uptake efficiency of CHAIN/pCas13a was associated with the CD44 receptor and ASGPR, we performed competitive binding experiments. When HepG2 cells were incubated with CHAIN/pCas13a under free galactopyranoside and/or HA, a relatively low mean fluorescence intensity (MFI) was observed, implying that cellular uptake of CHAIN/pCas13a was  Figure S6). These results indicated that the CHAIN/pCas13a were taken up by HepG2 cells through both ASGPRand CD44-mediated endocytosis.
In addition to internalization, CHAIN/pCas13a need to escape from the endo/lysosomal system and transfer into the nucleus to initiate the transcription of pCas13a. To evaluate the endosomal escape ability of CHAIN/pCas13a, we performed CLSM to observe the intracellular distribution at different time points. After incubation for 0.5 h, most of the pCas13a (green) was located in the endo/lysosomes (red). As time increased, the pCas13a was gradually transferred from the endo/lysosomes into the nuclei (blue). After 8 h of incubation, most of the pCas13a entered the nuclei of HepG2 cells ( Figure 3D), suggesting that CHAIN/pCas13a exhibited excellent endosomal/ lysosomal escape ability. Furthermore, the colocalization of red, green and blue fluorescence were analyzed in Figure S7.

Construction of the pCas13a-crRNA expression vectors
For miR-21 knockdown with the CRISPR-Cas13a system, three individual crRNAs were designed to target primary miR-21. Mature miR-21 sequences and all three crRNA sequences are shown in the precursor miR-21 (pre-miR-21) hairpin structure ( Figure 4A). As depicted in Figure 4B, we adopted an 'all-in-one' system, which expresses both the crRNA and Cas13a nuclease driven by the U6 and EF1α promoters, respectively.
After introducing CHAIN/ pCas13a-crRNA into HepG2 cells by separate transfections, we performed quantitative PCR (qPCR) assays and found that crRNA1 and crRNA2 could induce ~ 37% and ~ 27% reductions in mature miR-21 expression, respectively ( Figure 4C), suggesting that this strategy was effective. Moreover, we found that PDCD4, one known miR-21 target genes 14 , was significantly upregulated, as detected by western blot analysis ( Figure 4D). Due to its high efficiency, pCas13a-crRNA1 was chosen for further studies.

Antitumor analysis in vitro
Since PDCD4 was previously identified as one of the direct target genes of miR-21 in cancer cells, we hypothesized that knockdown of miR-21 may upregulate the expression of PDCD4, thereby inhibiting cell proliferation and inducing apoptosis.
A number of previous studies have revealed miR-21 was able to promote HCC cells invasion, migration and proliferation. Therefore, we aimed to examine whether RNA knockdown mediated by the CHAIN/pCas13a-crRNA1 abrogated the oncogenic activity of miR-21. To evaluate migration and invasion, we transfected HepG2 cells with the CHAIN/pCas13a-crRNA1 or CHAIN/pCas13a. As shown in Figure 5E-H, compared to the CHAIN/pCas13a-treated or blank group, the CHAIN/pCas13a-crRNA1-treated group showed significantly decreased migration and invasion of HepG2 cells. These results were further confirmed by analysis of MMP2 and RECK protein expression ( Figure 5I-J), which was related to cell migration and invasion. The above results indicated CHAIN/ pCas13a-crRNA1 may be a promising choice to treat HCC.

Targeting efficacy, biodistribution and antitumor activity in vivo
To investigate the tumor targeting capability and biodistribution of CHAIN/pCas13a-crRNA1 in subcutaneous xenograft tumor mice, we tracked the fluorescence of CHAIN/pCas13a-crRNA1 (labeled with TOTO-3) at different time points. As described in Figure 6A-B, while both groups rapidly gathered in the tumors at 2 h, the fluorescence intensity was substantially stronger in mice treated with CHAIN/pCas13a-crRNA1 than in mice treated with HAC/pCas13a-crRNA1.
The CHAIN/pCas13a-crRNA1 gradually accumulated in the tumor over time. Twenty-four hours after injection, strong fluorescence intensity was still observed at the tumor area of mice injected with CHAIN/pCas13a-crRNA1, while the fluorescence intensity of mice injected with HAC/pCas13a-crRNA1 was weakened. Similarly, stronger fluorescence was observed after CHAIN/ pCas13a-crRNA1 treatment than HAC/pCas13a-crRNA1 treatment in ex vivo photos ( Figure 6C). These results demonstrated that the versatile polymer GPH played an important role in long-term retention and active targeting in HCC tumor tissue in vivo. Encouraged by the excellent inhibitory effect of the CHAIN/pCas13a-crRNA1 on cancer cells in vitro, we further investigated the therapeutic effect in vivo using HCC xenograft subcutaneous mice model. Compared to those of the groups treated with PBS, GPH, pCas13a-crRNA1 (free plasmid) and CHAIN/ pCas13a, the tumor growth of the CHAIN/pCas13a-crRNA1-or HAC/pCas13a-crRNA1-treated group was strongly inhibited ( Figure 6D-E). Notably, CHAIN/pCas13a-crRNA1 induced a substantially stronger antitumor effect than HAC/pCas13a-crRNA1 (p < 0.05). The antitumor activity of CHAIN/pCas13a-crRNA1 was further validated by the average weight of tumors harvested at the end of the experiment ( Figure 6F).
Next, mir-21 expression in the tumors was quantified by qRT-PCR, about 29% decreased in CHAIN/pCas13a-crRNA1-treated group ( Figure  S11). Moreover, to evaluate the mechanisms that underlie the anti-miR-21 therapy, we performed an immunohistochemical assay. As depicted in Figure  6G, we found that the PDCD4 level was significantly increased after CHAIN/pCas13a-crRNA1 treatment, indicating that miR-21 was dramatically knocked down by the CRISPR-Cas13a system. Then, Ki-67 staining assays and transferase-mediated dUTP nick end labeling (TUNEL) staining assays were used to analyze cell proliferation and apoptosis in the tumors, respectively. The CHAIN/pCas13a-crRNA1 treatment induced a notable reduction in Ki-67-positive cells and a marked augment in TUNEL-positive cells in vivo. The above results revealed that anti-miR-21 therapy via the CRISPR-Cas13a system was an effective therapeutic choice.

Toxicity evaluation in vivo
The blood from each mouse was collected for analysis of the complete blood count (CBC) and blood chemistry profile to explore the toxicity of our CHAIN in vivo. As shown in Figure S12, the data of all groups demonstrate no significant differences and were comparable to those of normal mice. Meanwhile, histological analysis of the major organs revealed no significant pathological changes ( Figure S13). Therefore, the CHAIN/pCas13a-crRNA1 may be a safe therapeutic candidate for applications in vivo.

Synthesis of PF 33
PF 33 was synthesized as described previously [50]. Briefly, 400 mg PEI 1.8 K and 220 μL heptafluorobutyric anhydride were dissolved separately and mixed in anhydrous methanol. Then, trimethylamine was added to the mixture above and further stirred for 48 h. Eventually, the solution was purified by dialysis with distilled water for 72 h and lyophilized. The final product was verified by 19 F nuclear magnetic resonance ( 19 F NMR).

Synthesis of GPH
First, 67.75 mg 4-aminophenyl β-D-galacto-pyranoside and 106 mg NHS-PEG 2000 -OH were dissolved separately and mixed in N, N-dimethylformamide (DMF). The reaction proceeded under gentle stirring for 6 h at 25 ℃. The product (Gal-PEG 2000 -OH) was purified by dialysis and subsequently lyophilized. 84.6 mg HA, 4.94 mg EDCI and 3.65 mg DMAP were dispersed in 15 mL formamide and stirred for 2 h for activating the carboxyl groups of HA. Then, the mixtures were added to 80 mg Gal-PEG2000-OH and stirred for 1 d. Eventually, the solution was dialyzed, lyophilized and the final product was analyzed by 1 H NMR.

Preparation and characterization of the CHAIN/pCas13a
PF 33 and pCas13a were incubated with a series of mass ratios for 20 min, subsequently run by electrophoresis (1% agarose gel) for 15 min with 150 V. Afterward, the bands were imaged and recorded by a Gel Doc system (Bio-rad, USA). PF 33 (20 μg) was mixed with pCas13a (2 μg) by pipetting gently and incubated for 20 min at room temperature. Then, 60 μg pre-dispersed GPH was added and incubated for an additional 25 min to obtain the CHAIN/pCas13a. The nanoparticle sizes and their zeta potentials were measured using a dynamic light scattering (DLS) detector (Zetasizer, Nano-ZS, Malvern, UK). Transmission electron microscopy (TEM, H-600, Hitachi, Japan) was used to observe the morphology of the nanoparticles.
The enzymatic sensitivity of the CHAIN/ pCas13a was investigated by incubated with hyaluronidase (HAase) according to the manufacturer's instruction. Then, the nanoparticle characteristics of CHAIN/pCas13a were analyzed as described above.

Cell lines and cell culture
HepG2 (human HCC cell line) and LO2 (human nontumor hepatic cell line) were provided by the American Type Culture Collection (ATCC, MD). The cells were cultured in Dulbecco's modified Eagle's medium (Gibco, USA), which contained 10% fetal bovine serum (FBS) and 100 units/mL penicillin/ streptomycin antibiotics. All cells were cultured in humidified condition at 37 °C with 5% CO2.

Cell viability assay
CCK-8 assay was used to analyze the cytotoxicity of PF 33 and GPH. LO2 and HepG2 cells were seeded into 96-well plates with incubation for 16 h. Subsequently, PF 33 , GPH, PEI 1.8K, and PEI 25K were administered at concentrations ranging from 0 μg/mL to 40 μg/mL. Two days later, CCK-8 was added and further incubated for 2 h. Finally, the absorbance at 450 nm of each well was recorded via a microplate reader (Bio-Rad 680, USA).

Cellular uptake
The CD44 expression of HepG2 cells were verified with APC anti-mouse CD44 (Biolegend, USA) antibody staining and analyzed by a flow cytometer (ACEA NovoCyte, USA).
For cellular uptake analysis, HepG2 cells (1 × 10 5 cells per well) were seeded into 12-well plates and incubated overnight. Next, pCas13a was stained with the nucleic acids probe YOYO-1. CHAIN loaded with 1 μg of pCas13a were incubated with the cells for 2 h. For the competitive assay, HepG2 cells were preincubated with free HA (10 mg/mL) and/or with galactopyranoside (1 mM) for 2 h to block the CD44 and/or ASGPR receptors. Then, cells were either analyzed by flow cytometry or washed, fixed and observed by a fluorescence microscope (Olympus, Japan).

Intracellular trafficking
The intracellular colocalization of CHAIN/ pCas13a in the HepG2 cells were performed using confocal laser scanning microscopy (Zeiss, LSM 880, Germany). Cells were seeded into a 12-well plate pre-covered with glass cover slips (1 × 10 5 cells/well) for 24 h incubation. Next, the CHAIN/pCas13a containing 1 μg YOYO-1-labeled plasmid was added to each well. The cells were incubated for 0.5, 1, 4 or 8 h, rinsed with PBS, stained with LysoTracker Red probe, fixed with 4% paraformaldehyde, and stained with Hoechst in sequence. At last, the cells were subjected to confocal laser scanning microscopy for observation.

Transfection efficiency
HepG2 cells (1 × 10 5 cells/well) were seeded into 12-well plates and incubated overnight. Next, the medium was replenished with 500 µL per well fresh medium supplemented with 0 ~ 30% serum. CHAIN encapsulating 1 μg pCas13a-msfGFP were incubated with cells for 6 ~ 8 h, then the medium was discarded and replenished with complete medium and further incubated for 48 h. Meanwhile, PEI 1.8K and PEI 25K were exploited as controls. Finally, GFP expression was imaged and analyzed by flow cytometry.

Western blotting
Western blotting analysis was performed according to the previously described method [53]. Briefly, cells were lysed using RIPA reagent (Thermo, USA) on ice for 0.5 -1 h, and centrifuged at 12,000 rpm for 10 min. Lastly, supernatants were acquired and quantified using a BCA protein assay kit (Thermo, USA), and then, 35 µg protein sample was loaded into each well and subsequently run by SDS-PAGE gels for protein separation and then transferred to 0.45 µm PVDF membranes (Millipore, Germany) for immunoblotting. The PVDF membranes were incubated with 5% nonfat milk and then stained with primary antibodies against β-actin, GAPDH, PDCD4, MMP2 and RECK (Santa Cruz, USA) at room temperature for 2 h. Additionally, the membranes were washed and hybridized with a HRP-conjugated secondary antibodies (Santa Cruz, USA). A chemiluminescence detection system (Clinx, Shanghai) was used to detect the targeted bands.

In vitro migration and invasion assays
HepG2 cells migration and invasion were analyzed using transwell chambers (Millipore, Germany). For the detection of tumor cell migration, after transfection for 48 h, the CHAIN/pCas13a-crRNA1 and CHAIN/pCas13a-transfected cells were trypsinized and suspended, and then, HepG2 cells (50000 cells/well) in serum-free medium were seeded into the upper chamber. The lower chamber of each well was loaded with complete medium. For the detection of tumor cell invasion, matrigel (BD, USA) was used to precoat transwell chambers. Cells were added in serum-free medium, then the complete medium in the lower chamber was served as a chemoattractant. One day or two days later, cells were treated with cooled ethanol, followed by 0.1% crystal violet staining and counted via a microscope.

Proliferation and apoptosis analysis
For analysis of proliferation and apoptosis, HepG2 cells were incubated in 12-well plates for 16 h, and then transfected by CHAIN/pCas13a-crRNA1 or CHAIN/pCas13a in serum-free medium. Six hours later, the medium was discarded and replenished with complete medium and maintained for another 2 d. Proliferation analysis and apoptosis assays were performed with an EdU detection kit and Annexin V-APC/PI Apoptosis Detection Kit.

In vivo biodistribution
6-week-old female Balb/c nude mice were kept in a specific pathogen-free condition. The animal care and experiment conductions were in accordance with the relevant protocol, which was approved by the Institutional Animal Care and Treatment Committee of Sichuan University (Chengdu, China).
For construction of the HepG2-bearing tumor model, the right flank of each mouse was subcutaneously injected with 1 × 10 7 cells. Then, mice were randomly divided into 3 groups until the tumor size reached ~ 200 mm 3 , and then intravenously injected with PBS, CHAIN/pCas13a-crRNA1 or HAC/pCas13a-crRNA1 (pCas13a-crRNA1 labeled with TOTO-3). At 2 h, 12 h and 24 h, imaging data were collected utilizing an IVIS Lumina imaging system (Caliper, USA). 24 h later, mice were sacrificed, and tumors and major organs were harvested and imaged.

In vivo antitumor effect
The HepG2 xenograft tumor model was established in accordance with the method described above. 10 days later, mice were divided randomly into 6 groups (5 mice per group). PBS, GPH, pCas13a-crRNA1, CHAIN/pCas13a, HAC/pCas13a-crRNA1, and CHAIN/pCas13a-crRNA1 were prepared freshly and intravenously injected at an interval of 2 days. The individual tumor volumes were measured using a digital caliper every 3 days. 31 days later, the mice were sacrificed, and then the blood was collected for routine blood test and blood chemistry profile. The tumors were collected, weighed and then fixed with 4% paraformaldehyde. Then hematoxylin and eosin (H&E) and immunohistochemical analysis were conducted. Additionally, the major organs were fixed for H&E analysis.

Statistical analysis
Quantitative data were expressed as standard error of the mean or mean ± standard deviation. P-values analysis between groups was calculated using one-way ANOVA method. Significant differences are suggested by NS (not significantly different), * (p < 0.05), ** (p < 0.01) and *** (p < 0.001).

Conclusion
In summary, inspired by the structure and infection pathway of viruses, we constructed a versatile 'core-shell'-shaped CHAIN for multipathway-mediated tumor suppression by efficient delivery CRISPR-Cas13a megaplasmid system in vivo. The CHAIN enhanced the cellular uptake by the dual-targeting effect, promoted endo/lysosomal escape, and thus achieving high transfection efficiency of CRISPR-Cas13a megaplasmid in HCC cells. Furthermore, the versatile polymer GPH endowed the CHAIN with stabilization in physiological conditions, long-lasting circulation in the blood and a tumor active targeting capability in subcutaneous HCC-bearing mouse models. Finally, when the CRISPR-Cas13a system was delivered by the CHAIN in vivo, it knocked down the targeted oncogene miR-21, restored PDCD4 and RECK, crippled downstream MMP-2, and eventually suppressed tumor growth. Therefore, our CHAIN provides a high-efficient system for CRISPR-Cas13a megaplasmid transfection in vivo and achieves multi-signal-pathways regulation by efficient interference of one oncogenic microRNA.
While our work successfully demonstrated the use of CRISPR-Cas13a to target miRNA-21 and regulate RECK, PDCD4, and MMP2, it is important to note that miRNA-21 plays critical roles in other signaling pathways that warrant further research. In addition, it is crucial to investigate the potential impact of unwanted nanoparticle accumulation in other organs on these pathways. Finally, we aim to